The Septins

The Septins

The Septins Edited by Peter A. Hall, S.E. Hilary Russell and John R. Pringle © 2008 John Wiley & Sons, Ltd. ISBN: 978-0-470-51969-1

The Septins Edited by

Peter A. Hall Queen’s University Belfast, Belfast, Northern Ireland, UK

S.E. Hilary Russell Queen’s University Belfast, Belfast, Northern Ireland, UK

and

John R. Pringle Stanford University Medical Centre, Stanford, CA, USA

A John Wiley & Sons, Ltd., Publication

This edition first published 2008  2008 by John Wiley & Sons, Ltd. Wiley-Blackwell is an imprint of John Wiley & Sons, formed by the merger of Wiley’s global Scientific, Technical and Medical business with Blackwell Publishing. Registered office: John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK Other Editorial Offices: 9600 Garsington Road, Oxford, OX4 2DQ, UK 111 River Street, Hoboken, NJ 07030-5774, USA For details of our global editorial offices, for customer services and for information about how to apply for permission to reuse the copyright material in this book please see our website at www.wiley.com/wiley-blackwell The right of the author to be identified as the author of this work has been asserted in accordance with the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books. Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. The publisher is not associated with any product or vendor mentioned in this book. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold on the understanding that the publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought. Library of Congress Cataloging-in-Publication Data The septins / edited by Peter A. Hall, S.E. Hilary Russell, and John R. Pringle. p. cm. Includes index. ISBN 978-0-470-51969-1 1. Septins. I. Hall, Peter A. II. Russell, S.E. Hilary. III. Pringle, John R., 1943QP552.S37S47 2009 572
.6 – dc22 2008022154 ISBN: 978-0-470-51969-1 A catalogue record for this book is available from the British Library Typeset in 10.5/12.5pt Times by Laserwords Private Limited, Chennai, India Printed and bound in Great Britain by Antony Rowe Ltd, Chippenham, Wiltshire First impression

2008

Cover images (from left to right): • Localization of the septin Sep1 (green) to the cleavage furrows in dividing cells of a post-gastrulation Drosophila embryo. DNA (stained with propidium iodide) is shown in red. Image kindly provided by Johnny Fares and John Pringle and used with their permission. • Septin (green) in Candida albicans pseudohyphae (wildtype) • Septin (green) in Candida albicans pseudohyphae (rga2
/
bem3
/
mutant lacking Cdc42 GAPs). C. albicans images kindly provided by Helen Court and Peter Sudbery and used with their permission.

Contents Authors and Affiliations

vii

An introduction to the septins Peter A. Hall, S. E. Hilary Russell and John R. Pringle

1

Section I

Setting the scene

5

Chapter 1

Origins and development of the septin field John R. Pringle

7

Chapter 2

Evolution and conserved domains of the septins Michelle Momany, Fangfang Pan and Russell L. Malmberg

Section II Septins in model systems Chapter 3

Biochemical properties and supramolecular architecture of septin hetero-oligomers and septin filaments Michael A. McMurray and Jeremy Thorner

Chapter 4 Yeast septins: a cortical organizer Yves Barral Chapter 5

Chapter 6

Septins in four model fungal systems: diversity in form and function Amy S. Gladfelter and Peter Sudbery Septins in the metazoan model systems Drosophila melanogaster and Caenorhabditis elegans Christine M. Field, Amy Shaub Maddox, John R. Pringle and Karen Oegema

Section III Septins in mammals Chapter 7

The genomics and regulation of the human septin genes S.E. Hilary Russell

35

47

49 101

125

147

169 171

CONTENTS

vi

Chapter 8

The functions of septins in mammals Carol D. Froese and William S. Trimble

187

Chapter 9

Septin-interacting proteins in mammals Brandon E. Kremer and Ian G. Macara

211

Chapter 10 Septin functions in the mammalian cytoskeleton Elias T. Spiliotis and W. James Nelson

229

Chapter 11 Septins and the synapse Jing Xue, Victor Anggono and Phillip J. Robinson

247

Chapter 12 Septins and platelets Jerry Ware, Constantino Mart´ınez and Barbara Zieger

269

Chapter 13 Septins and apoptosis Marie-Jeanne Carp and Sarit Larisch

281

Chapter 14 Septins and human disease Peter A. Hall and Fern P. Finger

295

Chapter 15 Insight into septin functions from mouse models Makoto Kinoshita

319

Section IV Envoi

337

Chapter 16 Septins: 2008 and beyond Peter A. Hall, S.E. Hilary Russell and John R. Pringle

339

Appendix A

Septin and septin-like sequences 343 Michelle Momany, Fangfang Pan and Russell L. Malmberg

Appendix B

Mammalian septin nomenclature Peter A. Hall, Elspeth Bruford, Hilary Russell, Ian G. Macara and John R. Pringle

351

Appendix C

Septin meetings and workshops Peter A. Hall and John R. Pringle

355

Index

361

Authors and Affiliations Victor Anggono

Cell Signalling Unit, Children’s Medical Research Institute, Locked Bag 23, Wentworthville, NSW 2145, Australia

Yves Barral

Institute of Biochemistry, ETH Zurich, Schafmattstrasse 18, 8093 Zurich, Switzerland

Elspeth Bruford

Human Genome Nomenclature Committee, European Bioinformatics Institute, Wellcome Trust Genome Campus, Hinxton, Cambridge CB10 1SD, UK

Marie-Jeanne Carp

Apoptosis and Cancer Research Laboratory. Department of Pathology, Rambam Medical Center, Haifa 31096, Israel

Fern P. Finger

Department of Biology and Center for Biotechnology and Interdisciplinary Studies, Rensselaer Polytechnic Institute, 110 Eighth Street, Troy, NY 12180-3590, USA

Carol D. Froese

Program in Cell Biology, Hospital for Sick Children, 555 University, Avenue, Toronto, Ontario M5G 1X8, Canada

Christine Field

Department of Systems Biology, Harvard Medical School, 200 Longwood Avenue, WA 536, Boston, MA 02115, USA

Amy Gladfelter

Department of Biology, Dartmouth College, 409 Gilman Hall, Hanover, NH 03755, USA

Peter A. Hall

Institute of Pathology, School of Medicine, Queen’s University Belfast, Belfast BT12 6BL, Northern Ireland, UK

viii

AUTHORS AND AFFILIATIONS

Makoto Kinoshita

Cell Biology and Biochemistry Unit, Kyoto University Graduate School of Medicine and CREST, Japan Science & Technology Agency, Yoshida Kanoe, Sakyo, Kyoto 606-8501, Japan

Brandon E. Kremer

Department of Microbiology, University of Virginia School of Medicine, Charlottesville, VA 22908-0577, USA

Sarit Larisch

Apoptosis and Cancer Research Laboratory, Department of Pathology, Rambam Medical Center, Haifa 31096, Israel

Ian G. Macara

Department of Microbiology, University of Virginia School of Medicine, Charlottesville, VA 22908-0577, USA

Amy Shaub Maddox

Institute for Research in Immunology and Cancer, Department of Pathology and Cell Biology, University of Montr´eal, Montr´eal H3C 3J7, Canada

Russell L. Malmberg

Plant Biology Department, University of Georgia, Athens, GA 30602, USA

Constantino Mart´ınez

Centro Regional de Hemodonaci´on, University of Murcia, Calle de Ronda de Garay S/N, Murcia 30003, Spain

Michael A. McMurray

Department of Molecular and Cell Biology, Division of Biochemistry and Molecular Biology, University of California, Berkeley, CA 94720, USA

Michelle Momany

Plant Biology Department, University of Georgia, Athens, GA 30602, USA

W. James Nelson

Department of Biological Sciences, The James H. Clark Center, The Bio-X Program, 318 Campus Drive, E-200, Stanford University, Stanford, CA 94305-5430, USA

Karen Oegema

Ludwig Institute for Cancer Research, Department of Cellular & Molecular Medicine, University of California San Diego, CMM East, Room 3053, 9500 Gilman Drive, La Jolla, CA 92093-0653, USA

AUTHORS AND AFFILIATIONS

ix

Fangfang Pan

Plant Biology Department, University of Georgia, Athens, GA 30602, USA

John R. Pringle

Department of Genetics, MC 5120, 300 Pasteur Drive, M-322 Alway Building, Stanford University Medical Center, Stanford, CA 94305-5120, USA

Phillip J. Robinson

Cell Signalling Unit, Children’s Medical Research Institute, Locked Bag 23, Wentworthville, NSW 2145, Australia

S.E. Hilary Russell

Ovarian Cancer Research Laboratory, Centre for Cancer Research and Cell Biology, Queen’s University Belfast, A floor, Belfast City Hospital, Lisburn Road, Belfast BT9 7AB, Northern Ireland, UK

Elias T. Spiliotis

Department of Biological Sciences, The James H. Clark Center, The Bio-X Program, 318 Campus Drive, E-200, Stanford University, Stanford, CA 94305-5430, USA

Peter Sudbery

Department of Molecular Biology and Biotechnology, Sheffield University, Western Bank, Sheffield S10 2TN, UK

Jeremy Thorner

Department of Molecular and Cell Biology, Division of Biochemistry and Molecular Biology, University of California, Berkeley, CA 94720, USA

William S. Trimble

Program in Cell Biology, Hospital for Sick Children, 555 University Avenue, Toronto, Ontario M5G 1X8, Canada

Jerry Ware

Department of Physiology and Biophysics, University of Arkansas for Medical Sciences, 4301 West Markham Street, Little Rock, AR 72205, USA

Jing Xue

Vascular Biology Centre, Medical College of Georgia, 1459 Laney Walker Boulevard, Room CB 3330, Augusta, GA 30912, USA

Barbara Zieger

Department of Paediatrics and Adolescent Medicine, University Hospital Freiburg, Mathildenstrasse 1, D-79106 Freiburg, Germany

(a)

(b)

(d)

(c)

(e)

Plate 5.1 Septin organisation in C. albicans. (a) A septin ring forms at the bud neck in yeast cells. (b) Septin rings form at the bud necks of pseudohyphal cells, and nuclei divide across the mother-bud necks. (c) Basal septin bars and septin cap in a newly evaginated hyphal germ tube. (d) Septin collars form along the length of hyphal germ tubes. Nuclei migrate out of mother cells and divide across the septin collars. Inset, enlarged view of the septin collar indicated by the arrow. (e) After mitosis, the septin collar separates into two sharply defined rings between which the chitinous primary septum forms. Septin is Cdc10-YFP in all panels. Panels a, c and e are counterstained with Calcofluor white (blue), which stains cell walls and the primary septum in panel c. Panels b and d are counterstained with Concanavaline A – Texas red and the nuclei are stained with 4
,6-diamidino-2-phenylindole (DAPI) (blue). Scale bars, 5 µm in panels a, b, and d; 1 µm in panels c and e, and in inset d. Images were generated with a DeltaVision RT wide-field epifluorescence microscope and deconvolved with Softworx company software.

The Septins Edited by Peter A. Hall, S.E. Hilary Russell and John R. Pringle © 2008 John Wiley & Sons, Ltd. ISBN: 978-0-470-51969-1

S51 W260 G241 G261 H270

c.c. N55

c.c. S31K30

(Y17)

c.c.

(a) c.c.

c.c.

c.c.

c.c.

c.c.

SEPT7-GDP SEPT2-GDP SEPT2-GDP SEPT7-GDP SEPT6-GTP SEPT6-GTP

T228

D185

c.c. c.c.

(b)

Cdc11-GDP Cdc12-GDP Cdc3-GTP Cdc10-GDPCdc10-GDP Cdc3-GTP Cdc12-GDP Cdc11-GDP

c.c.

F156

c.c.

() α0 ) | β1 > () α1 ) Cdc3 70 MGITSSQSEKGQVLPDQPEIKFIRRQINGYVGFANLPKQW-HRRSIKNGFSFNLLCVGPDGIGKTTLMKTLFNNDDIEANLVKDYEEELANDQEEEEGQG 168 Cdc10 1 MDPLSSVQPAS------YVGFDTITNQI-EHRLLKKGFQFNIMVVGQSGLGKSTLINTLFASH--------LIDSATGDD-------- 65 Cdc11 1 MSGIIDASSALRKRKHLKRGITFTVMIVGQSGSGRSTFINTLCGQQ--------VVDTSTTILLP------ 57 Cdc12 1 MSAATATAAPVP-----PPVGISNLPNQR-YKIVNEEGGTFTVMLCGESGLGKTTFINTLFQTV---------LKRADGQQ-------- 66 hsSEPT2 1 MSKQQPTQFIN-PETPGYVGFANLPNQV-HRKSVKKGFEFTLMVVGESGLGKSTLINSLFLTD--------LYPERVIP--------- 69 hsSEPT6 1 MAATDIARQVGEGCRTVPLA-----GHVGFDSLPDQL-VNKSVSQGFCFNILCVGETGLGKSTLMDTLFNTK----------FEGEP---------- 71 hsSEPT7 1 MSVSARSAAAEERSVNSSTMVAQQKNLEGYVGFANLPNQV-YRKSVKRGFEFTLMVVGESGLGKSTLINSLFLTD--------LYSPEYP---------- 81 ident/simil * * *... * * *..* . .* modification P P P

Cdc3 169 Cdc10 66 Cdc11 58 Cdc12 67 hsSEPT2 70 hsSEPT6 72 hsSEPT7 82 ident/simil modification

| β2 > | β3 > () α2 ) | β4 > EGHENQSQEQRHKVKIKSYESVIEEN-GVKLNLNVIDTEGFGDFLNNDQKSWDPIIKEIDSRFDQYLDAENKINR--HSINDKRIHACLYFIEPTGHYLK ----ISALPVTKTTEMKISTHTLVED-RVRLNINVIDTPGFGDFIDN-SKAWEPIVKYIKEQHSQYLRKELTAQRE-RFITDTRVHAILYFLQPNGKELS -----TDTSTEIDLQLREETVELEDDEGVKIQLNIIDTPGFGDSLDN-SPSFEIISDYIRHQYDEILLEESRVRRN-PRFKDGRVHCCLYLINPTGHGLK ----HRQEPIRKTVEIDITRALLEEK-HFELRVNVIDTPGFGDNVNN-NKAWQPLVDFIDDQHDSYMRQEQQPYR--TKKFDLRVHAVLYFIRPTGHGLK ----GAAEKIERTVQIEASTVEIEER-GVKLRLTVVDTPGYGDAINC-RDCFKTIISYIDEQFERYLHDESGLNR--RHIIDNRVHCCFYFISPFGHGLK ------ATHTQPGVQLQSNTYDLQES-NVRLKLTIVSTVGFGDQINK-EDSYKPIVEFIDAQFEAYLQEELKIRRVLHTYHDSRIHVCLYFIAPTGHSLK ----GPSHRIKKTVQVEQSKVLIKEG-GVQLLLTIVDTPGFGDAVDN-SNCWQPVIDYIDSKFEDYLNAESRVNR--RQMPDNRVQCCLYFIAPSGHGLK . . . . .... * *.** . . . * . . * * * *.. * . * *. * P

265 158 150 158 161 163 173

Cdc3 266 Cdc10 159 Cdc11 151 Cdc12 159 hsSEPT2 162 hsSEPT6 164 hsSEPT7 174 ident/simil modification

() α3 ) |β5> () α4 ) () α5' ) |β6> PLDLKFMQSVYEKCNLIPVIAKSDILTDEEILSFKKTIMNQLIQSNIELFKPPIYSNDDAEN----------SHLSERLFSSLPYAVIGSNDIVENYS-G RLDVEALKRLTEIANVIPVIGKSDTLTLDERTEFRELIQNEFEKYNFKIYPYDSEELTDEE-----------LELNRSVRSIIPFAVVGSENEIEIN--G EIDVEFIRQLGSLVNIIPVISKSDSLTRDELKLNKKLIMEDIDRWNLPIYNFPFDEDEISDED---------YETNMYLRTLLPFAIIGSNEVYEMGGDV PIDIETMKRLSTRANLIPVIAKADTLTAQELQQFKSRIRQVIEAQEIRIFTPPLDADSKEDAKSGSNPDSAAVEHARQLIEAMPFAIVGSEKKFDNGQ-G PLDVAFMKAIHNKVNIVPVIAKADTLTLKERERLKKRILDEIEEHNIKIYHLPDAESDEDEDF---------KEQTRLLKASIPFSVVGSNQLIEAK--G SLDLVTMKKLDSKVNIIPIIAKADAISKSELTKFKIKITSELVSNGVQIYQFPTDD----ESV---------AEINGTMNAHLPFAVIGSTEELKIG--N PLDIEFMKRLHEKVNIIPLIAKADTLTPEECQQFKKQIMKEIQEHKIKIYEFPETD-DEEE-----------NKLVKKIKDRLPLAVVGSNTIIEVN--G .*. .. . *..*.* *.* .. * . * .. . . .* ...** P P P P

354 245 241 257 250 248 259

Cdc3 355 Cdc10 246 Cdc11 242 Cdc12 258 hsSEPT2 251 hsSEPT6 249 hsSEPT7 260 ident/simil modification

| β7> | β8> () α5 ) () α6 ) () α7? ... NQVRGRSYPWGVIEVDNDNHSDFNLLKNLLIKQFMEELKERTSKILYENYRSSKLAKLGIK-QDNSVFKEFDP---ISKQLEEKTLHEAKLAKLE 445 ETFRGRKTRWSAINVEDINQCDFVYLREFLIRTHLQDLIETTSYIHYEGFRARQLIALKENANSRSSAHMSSNAIQR.----------------- 298 GTIRGRKYPWGILDVEDSSISDFVILRNALLISHLHDLKNYTHEILYERYRTEALSGESVAAESIRPNLTKLNGSSSSSTTTRRNTNPFK(8)VL 341 TQVVARKYPWGLVEIENDSHCDFRKLRALLLRTYLLDLISTTQEMHYETYRRLRLEGHENTGEGNE--DFTLPAIAPAR----KLSHNPRYKEEE 346 KKVRGRLYPWGVVEVENPEHNDFLKLRTMLI-THMQDLQEVTQDLHYENFRSERLKRGGRKVENEDMNKD------------------------- 319 KMMRARQYPWGTVQVENEAHCDFVKLREMLIRVNMEDLREQTHTRHYELYRRCKLEEMGFKDTDPDSKPFSLQETYEAK----RNEFLGELQKKE 339 KRVRGRQYPWGVAEVENGEHCDFTILRNMLIRTHMQDLKDVTNNVHYENYRSRKLAAVTYNGVDNNKNKGQLTKSPLAQMEEERREHVAKMKKME 354 * * .. ** *. *. . .* * ** . * P

(c) Plate 3.3 Structural models and structural alignment of the human SEPT2-SEPT6-SEPT7 hexamer and the budding yeast Cdc3-Cdc10-Cdc11-Cdc12 octamer. (a) A ribbon representation of a space-filling depiction of the structure of the human SEPT7-SEPT6-SEPT2-SEPT2-SEPT6SEPT7 heterohexamer (adapted from Figure 4 in Sirajuddin et al., 2007), based on coordinates (PDB accession no. 2QAG) for the asymmetric unit (a SEPT2-SEPT6-SEPT7 trimer) and on the relative order of subunits as determined by EM analysis of a MBP-SEPT2-SEPT6-SEPT7 complex (Sirajuddin et al., 2007). The positions and directions in which the CTEs are predicted to extend from the globular G domains are indicated by the arrows and labeled ‘c.c.’ (for ‘coiled coil’). SEPT2, purple; SEPT6, blue; SEPT7, green; bound guanine nucleotide, red; positions of selected residues from Table 3.1, gold and labeled; disordered residues not represented in the final model, absent or shown in their most likely positions by dashed lines. (b) A ribbon representation of the budding yeast Cdc11-Cdc12-Cdc3-Cdc10-Cdc10-Cdc3-Cdc12Cdc11 hetero-octamer, based on the relative order of subunits determined by Bertin et al. (2008) [see also Figure 3.1c and modelled on (a)]. Cdc10, purple; Cdc3, blue; Cdc12, green; Cdc11, magenta; bound guanine nucleotide, red. (c) Alignment using the Clustal W algorithm of the G domains and surrounding sequences of the septins shown in (a) and (b), with the elements of secondary structure labeled as for human SEPT2 (Sirajuddin et al., 2007), plus the predicted ‘α7’ helix (based on Versele et al., 2004). Highly conserved positions are indicated below the sequences: ‘∗’, residues identical in all; ‘.’, highly similar residues in all. Residues known to be phosphorylated are in bold and indicated by a ‘P’. ‘(8)’ represents a stretch of eight residues present in Cdc11, but not in the other septins shown.

(a)

(b)

(c)

(d)

(e)

(f)

3.0 µm (g)

(h)

Plate 5.2 Septin organization in A. gossypii. Fluorescence micrographs of cells expressing Sep7-GFP are shown. (a–d) Initial appearance and deposition of septin rings at growing tips. (e– h) Assembly of septin rings at emerging lateral branches. Images prepared by Bradley DeMay.

Sept 3, 5 and 7 Localization in Neurons Septins

Synaptophysin

Merge

Sept3

Sept5

Sept7

Plate 11.3 Localization of septins in primary cultured hippocampal neurons. Hippocampal neurons cultured for 14 days were stained with antibodies against SEPT3 (top), SEPT5 (middle), and SEPT7 (bottom); and they were all co-stained with antibodies against synaptophysin. They show punctate distribution and partial co-localization with the presynaptic nerve terminal marker, synaptophysin. [Figure in the top panel was reproduced with permission from Xue, J., Tsang, C.W., Gai, W.P. et al. (2004b) Septin 3 (G-septin) is a developmentally regulated phosphoprotein enriched in presynaptic nerve terminals. Journal of Neurochemistry, 91, 579–90; figures in the middle and bottom panels were reproduced with permission from Fujishima, K., Kiyonari, H., Kurisu, J. et al. (2007) Targeted disruption of SEPT3, a heteromeric assembly partner of SEPT5 and SEPT7 in axons, has no effect on developing CNS neurons. Journal of Neurochemistry, 102, 77–92].

Dapi

SEPT4_i1

SEPT4_i2

Lymphocytes from healthy donor

Lymphocytes from ALL patient

(c) Plate 13.1 SEPT4 i2 is the only human Septin4 isoform shown to directly induce and promote apoptosis. SEPT4 i2 functions as a tumour suppressor protein. Immunofluorescence staining presents lymphocytes isolated from healthy donor (upper panel) and from acute lymphoblastic leukemia (ALL) patient (lower panel). Staining with Dapi, showing nuclei of cells (blue), the non-apoptotic isoform of SEPT4; SEPT4 i1 (red), and pro-apoptotic protein SEPT4 i2 (green). Whereas all lymphocytes in healthy control contain similar levels of SEPT4 i2 and SEPT4 i1 (upper panel), the sample from leukemia patient exhibits loss of SEPT4 i2 staining in all tumour lymphoblasts (only remaining normal cells are stained), indicating selective loss of the pro-apoptotic protein SEPT4 i2 in leukemia patients (Adopted from Elhasid et al., 2004).

Plate 15.1 Embryonic lethality of Sept9−/− mice. Hematoxylin-eosin (HE) stain indicated that the disorganized histogenesis and/or degeneration involves the neural tube, limb buds, and mesenchyme, where pyknotic nuclei predominate. Unpublished data provided by courtesy of Dr Ernst-Martin F¨uchtbauer.

(a) Plate 15.2 Sept4 function in dopaminergic nerve terminals in the striatum. A scanning laser confocal microscopic image of a mouse striatal section, immunostained for DNA (blue), the dopamine transporter (green), and tyrosine hydroxylase, a rate-limiting enzyme for dopamine synthesis (red). Note fine network of highly branched axon terminals that synthesize, release, and reuptake dopamine. The nucleated cell bodies belong to the post-synaptic GABAergic neurons.

(a) Plate 15.3 Spermiogenesis defect observed in Sept4 knockout mice. Immunofluorescence signals superimposed on a DIC image of wildtype spermatozoa. Nuclear DNA (blue), Sept4 (green), and mitochondria (red), respectively, are labelled by DAPI, an anti-Sept4 antibody, and MitoTracker Red (Molecular Probes). Note that a septin-based ring (the annulus) is localized to the caudal end of the middle piece, a segment of flagellum covered with a mitochondrial sheath.

An introduction to the septins Peter A. Hall, S. E. Hilary Russell and John R. Pringle

Few modern scientists like specialist monographs. Typically, they are out of date before they are published, and, in fast-moving fields, they show their age quickly! It was therefore with trepidation that we took on the task of planning, commissioning chapters for, and editing a monograph on septins. Why were we persuaded that this was a worthwhile project? The field is still young, yet it has grown quickly in the past few years (see Figure 0.1). Moreover, although it was once largely the preserve of a small band of yeast biologists, this field now encompasses the whole of animal and fungal biology. (Septins do not appear to be present in plants or prokaryotes.) A further stimulus has been the recognition that the septins are involved in a variety of disease processes. It was thus with the anticipation that the septin field would continue to grow rapidly, and that many newcomers would benefit from a comprehensive overview of the state of the field ca. early 2008 (probably the last time that such a thing would be possible), that we embarked on this task. The germ of the idea emerged after the First Septin Workshop in May 2005 (see Appendix C). In the summer of 2006, Andrea Baier of Wileys (now Wiley Blackwell Publishers) approached one of us (Peter Hall) about the idea. A conference call between the three editors followed, and the germ began to grow. Conversations took place with others, including potential authors, and a general view of cautious enthusiasm developed. The original intention was that contributions would be finalized immediately following the Second Septin Workshop (May 2007), but (as in many publishing ventures) reality was different. In this case, a delay was produced not only by the difficulties of some authors in finishing their chapters according to the original schedule, but also by the desirability of including the important structural analyses that were just beginning to emerge in the Summer and Fall of 2007 – so much is thrown into a new light by these important developments! Despite the delay, the contributors have produced a set of chapters that well defines the field as it stands at the end of 2007, not far from our original intention. The Septins Edited by Peter A. Hall, S.E. Hilary Russell and John R. Pringle © 2008 John Wiley & Sons, Ltd. ISBN: 978-0-470-51969-1

AN INTRODUCTION TO THE SEPTINS

2

Septin publications 1994 to 2007 500 Cumulative publication 450 400 350 300 250 200 150 Annual publications

100 50 0

1994 1995 1997 1996

1998 1999 2000 2001

2002 2003 2004 2005

2006 2007

Figure 0.1 The increase in publications on septins from October 1994, when the word was first used in the published literature (Sanders and Field, 1994). PubMed was searched using ‘septin*’ as the search string. [Note that the word septin had been employed previously to denote a factor in plasma that opsonizes lipopolysaccharide-bearing particles.]

The septin field now has a solid basis in data from a wide range of systems, providing a foundation on which its future can be built. Furthermore, there are links to a surprising array of physiological processes, not just the role in cytokinesis that for a long time dominated the field. As the contributions here show, we now have a good grip on both the biology and the biochemistry of this fascinating family of proteins. The scene is set by historical perspectives in both recent and geological time. The book begins with a detailed history of the beginnings of the field (including much previously unpublished material) by John Pringle, in whose laboratory most of the early work on the septins took place. The second chapter, by Michelle Momany, Fangfang Pan and Russell L. Malmberg, takes advantage of the many septin sequences now available, from diverse organisms, to describe the evolution of the septin family and the intriguing gain and loss of septin genes during the evolution of different phylogenetic groups. One key point is the evolution of a great diversity of septin genes, which in vertebrates can number as many as 14. A major challenge for the field will be the dissection of the roles of these multiple genes and the evolutionary pressures for their duplication and reduplication. The second section of the book focuses on septins in model systems. Michael McMurray and Jeremy Thorner provide a detailed account of the biochemical properties and supramolecular architecture of septin oligomers and filaments; the

AN INTRODUCTION TO THE SEPTINS

3

work in budding and fission yeast is placed in the context of the important recent work on the structures of the mammalian and C. elegans septin complexes. This contribution dovetails with that of Yves Barral on the function and cell biology of septins in Saccharomyces cerevisiae, where these proteins were first discovered and have since been studied in considerable detail. Amy Gladfelter and Peter Sudbery further develop this theme with a consideration of the septins in other fungi, including Schizosaccharomyces pombe, Candida albicans, Ashbya gossypii and Aspergillus nidulans. Christine Field, Amy Maddox, Karen Oegema and John Pringle complete the section with a discussion of the septins in Drosophila melanogaster and Caenorhabditis elegans. The worm is of particular interest given that it only has two septin genes, unlike the fly with five, yeasts with seven and humans with 14! The third section of the book turns to the mammalian systems. Growth in this area has been particularly dramatic since the recognition by Kinoshita in 1997 that septins are not only essential for cytokinesis, but also seemingly involved in other processes, in mammalian cells. The septin genes in mammals (and particularly in humans) present particular challenges as outlined by Hilary Russell. Not only are there at least 14 distinct human septin genes, but many of these genes show remarkably complex alternative splicing events and regulation by a plethora of strategies. Carol Froese and Bill Trimble provide an overview of the possible functions of mammalian septins, which is complemented by a detailed consideration of the known septin-interacting proteins by Brandon Kremer and Ian Macara. There follows a series of chapters that discuss some specific areas of mammalian septin research including septins and the cytoskeleton (Elias Spiliotis and James Nelson) and the burgeoning area of septins and the nervous system (Jing Xue, Victor Anggono and Philip J. Robinson). Septins have been implicated in the physiology and pathophysiology of platelets, remarkable anucleate ‘cells’ derived from megakaryoctes that have a key role in hemostasis, and Jerry Ware, Constantino Mart´ınez and Barbara Zieger review this area. Marie-Jeanne Carp and Sarit Larisch then cover the controversial area of the role of septins in apoptosis. This section culminates with two chapters that look at septins in disease (Peter Hall and Fern Finger) and at the many insights that we have gained from murine models of septin biology, and, in particular, from knockout mutants (Makoto Kinoshita). The monograph then ends with a speculative overview from the editors of the key questions in the field and of where it may be headed. In addition, several Appendices summarize some important information in a form that may be convenient for those in, or just entering, the field. This monograph could not have been produced without the help and enthusiastic contributions of those in the field. The editors wish to express their sincere thanks to all the authors for their hard work and to the many others in the field whose contributions and comments (particularly at the two Septin Workshops) have helped to push the field, and thus this monograph, forward. We also wish to

4

AN INTRODUCTION TO THE SEPTINS

thank Andrea Baier and Fiona Woods from Wiley Blackwell, who have cajoled, persuaded and supported the Editors in this journey! Finally, we wish to note that all royalties from this book have been waived by the Editors and authors and will be used to help support future Septin Workshops.

REFERENCE Sanders, S.L. and Field, C.M. (1994) Cell division. Septins in common? Current Biology, 4, 907–10.

Section I Setting the scene

The Septins Edited by Peter A. Hall, S.E. Hilary Russell and John R. Pringle © 2008 John Wiley & Sons, Ltd. ISBN: 978-0-470-51969-1

1 Origins and development of the septin field John R. Pringle Department of Genetics, Stanford University School of Medicine, Stanford, CA 94305, USA

INTRODUCTION In this chapter, I have told the story of the septin field from its conception in Lee Hartwell’s and Breck Byers’s laboratories (1967–1976), through its long gestation, awkward birth, and short childhood, mostly in my own laboratory (1977–1993), and finally through its adolescence in a growing number of laboratories until its coming of age was marked (in 2001) by a session’s being devoted to it at a major meeting (see Appendix C). Although there are many parts of the story that I find embarrassing and guilt-provoking (even many years after the events), I have told the unvarnished truth as best I can reconstruct it. I think that there is value in having an accurate historical record of a scientific field; if nothing else, this account should clarify aspects of the septins’ early history that would otherwise be quite mysterious to a newcomer. I think that this story also illustrates a great and reassuring truth about science: it moves forward even though the people involved, and their behaviour, are often imperfect.

CONCEPTION, GESTATION, AND BIRTH In 1966, Lee Hartwell, then a beginning Assistant Professor at the University of California, Irvine, began isolating a large collection of temperature-sensitive-lethal (ts) mutants of the budding yeast Saccharomyces cerevisiae. His inspiration was the work done on ts mutants of bacteriophage T4 in Bob Edgar’s lab at Caltech, where Lee had done undergraduate research. A publication summarizing The Septins Edited by Peter A. Hall, S.E. Hilary Russell and John R. Pringle © 2008 John Wiley & Sons, Ltd. ISBN: 978-0-470-51969-1

8

CH01 ORIGINS AND DEVELOPMENT OF THE SEPTIN FIELD

the properties of 400 such mutants appeared soon (Hartwell, 1967); it included brief descriptions of two of the mutants that would later define the septins, namely ts310 and ts471 (later cdc10 and cdc12 ), described at this time as defective ‘in cell-wall formation’. In 1967, another new faculty member at Irvine, Cal McLaughlin, teamed up with Lee, and the two spent the next several years characterizing mutants with defects in RNA and protein synthesis (e.g. Hartwell et al., 1970b); this collaboration continued after Genetics Department Chairman Herschel Roman lured Lee to the University of Washington in 1968. Meanwhile, I had been diverted from my intended graduate study of population ecology by fabulous courses in genetics (Matt Meselson) and cell biology (Keith Porter) that I was required to take as a first-year student. I had also fallen in love with yeast as an experimental organism on the basis of a lab course taught by Nick Gillham, a Chlamydomonas geneticist who had learned about yeast through some exposure as a postdoc. Given my somewhat contrary personality, the fact that no one else in the Boston area was working on yeast at that time only added to its appeal. I solved my immediate problem by working with Guido Guidotti, a protein chemist who was willing to sponsor my study of some yeast proteins (‘so long as they’re interesting’), but I soon also began planning to do a postdoc in Seattle, which at the time had the only concentration of yeast geneticists (three labs) in the country. In November 1968, I approached Herschel at a meeting and soon had arranged to join his Department as a postdoc when I finished my PhD; he kindly agreed to sponsor me although my proposed project (a genetic analysis of the enzymes that I was studying as a graduate student) was unrelated to his own research. In subsequent correspondence, Herschel soon began pushing me toward the laboratory of his new recruit, but I resisted because I wasn’t excited by the papers on RNA and protein synthesis. However, when I visited Seattle in September 1969, Lee told me how he, undergraduate Brian Reid, and first-year graduate student Joe Culotti had recently realized that because of yeast’s budding mode of reproduction, microscopic inspection of the ts mutants after temperature shift would allow recognition of mutants with a variety of specific cell-division-cycle (cdc) defects. Luckily, I had the wit required (it didn’t take much) to see that this project was more exciting than my original plan, and Lee and I soon agreed that I would join his lab. Thus, in July 1970, I happily became Lee’s first postdoc, just a month after the first report on the cdc mutants was published (Hartwell, Culotti and Reid, 1970a). This paper included a description of a third mutant that would later help to define the septins; it was named cdc3 and described as defective in ‘cell separation’. During the next three exhilarating years, I worked mostly on the nutritional and growth control of cell-cycle initiation, but I used the cdc mutants in many of my experiments, and we talked endlessly about them in the lab (often to the short-term detriment of our experiments!). I was particularly captivated by the mutants with grossly abnormal cell morphologies that indicated defects in cytoplasmic rather than nuclear processes (Figure 1.1). These mutants included one that could continue growth and the nuclear cycle but not make buds (cdc24 : Hartwell

CONCEPTION, GESTATION, AND BIRTH

9

Figure 1.1 The first attempt to organize the events of the yeast cell cycle into dependent and independent pathways (Figure 3 of Hartwell et al., 1974, a now-famous paper that was originally rejected without review by Nature). Events defined by particular cdc mutants are indicated by the CDC gene number. iDS and DS, initiation and continuation of DNA synthesis; mND and lND, medial and late nuclear division; BE and NM, bud emergence and nuclear migration; CK and CS, cytokinesis and cell separation; HU and TR, the DNA-synthesis inhibitors hydroxyurea and trenimon; MF, the mating pheromone α factor. Note that cdc12 was not included among the cytokinesis mutants because the alleles available at this time were too leaky to allow full characterization. Reproduced with permission from Hartwell et al ., (1974) Science, 183, 46–51, Copyright  1974 Elsevier

et al., 1974) and four that made abnormally elongated buds with multiple nuclei but could not complete cytokinesis (cdc3, cdc10, cdc11 and cdc12 : Hartwell, 1971; Figure 1.2). When I began my own lab at The University of Michigan in 1975 (after a second postdoc in Z¨urich that focused further on nutritional control), I soon decided to concentrate on these mutants, a decision that was the more

Figure 1.2 Images from the first systematic description of the ts mutants that would later define the septins (Hartwell, 1971); the cells had been incubated for several hours at restrictive temperature and then stained with Giemsa to reveal their nuclei. The defect in cytokinesis was clearly recognized at this time, using a newly developed assay to discriminate cytokinesis from the somewhat later process of separation of the daughter cell walls. (b) cdc3 ; (c) cdc11 . Similar images were also presented for cdc10 and cdc12 mutants. Reproduced with permission from Hartwell, L.H. (1971) Exp. Cell Res., 69, 265–76, Copyright  1971 Elsevier

10

CH01 ORIGINS AND DEVELOPMENT OF THE SEPTIN FIELD

appealing to me because it seemed that everyone else (including Lee) was more interested in the mutants with defects in DNA replication and/or nuclear division. My early years at Michigan were dominated by the struggle to get a lab up and running on a miniscule budget, cope with a heavy teaching load, and finish the projects related to nutritional and growth control of the cell cycle. The time and resources available for the morphogenesis mutants were put into further characterizing cdc24 (Sloat and Pringle, 1978; Sloat, Adams and Pringle, 1981) and screening for more mutants with defects in bud emergence. These efforts were quite rewarding (as reviewed elsewhere: Pringle et al., 1995; Pringle, 2006), but they delayed work on cdc3, 10, 11, and 12, and thus the birth of the septins. However, the mutant screen did yield, as a by-product, 28 new mutants with the same phenotype as cdc3, 10, 11, and 12. We were disappointed when all 28 proved to have mutations in the four known genes, but this did help to convince us that the phenotype was highly specific, so that our intended studies would be of manageable scope. Also, we recovered cdc11 and cdc12 alleles (including the now widely used cdc12-6 : Adams and Pringle, 1984) that were tighter, and thus more useful in temperature-shift experiments, than those available previously. During this period, there were also other developments important for what would become the septin field. First, Breck Byers and his outstanding technician Loretta Goetsch, working a few doors down from the Hartwell lab, undertook systematic electron-microscopic (EM) analyses of both wild-type yeast and the various cdc mutants. In wild-type cells, they observed seemingly filamentous structures in close apposition to the plasma membrane in the mother-bud neck (Byers and Goetsch, 1976a; Byers, 1981; Figure 1.3a and b; see also Figure 3.4 of Chapter 3 by McMurray and Thorner). In addition, in the cdc3, 10, 11, and 12 (a)

(c)

(b)

(d)

(e)

Figure 1.3 (a and b) Conventional EM images of the neck filaments (Byers, 1981); these images have better contrast than those published earlier (Byers and Goetsch, 1976a). (c–e) ImmunoEM images obtained using antibodies against (c) Cdc3p and (d and e) Cdc10p (J. Mulholland, B. Haarer, S. Ketcham, D. Preuss, J. Pringle and D. Botstein, unpublished results). (a) and (b) reproduced from Byers, B. (1981) with permission from Cold Spring Harbor Laboratory Press (c–e) ImmunoEM images obtained using antibodies against (c) Cdc3p and (d and e) Cdc10p (Reproduced with permission from Mulholland J., Haarer B., Ketcham S., Preuss D., Pringle J. and Botstein D., unpublished results)

CONCEPTION, GESTATION, AND BIRTH

11

Figure 1.4 The abstract that reported that certain cdc mutants lacked the ‘neck filaments’ (Byers and Goetsch, 1976b); note that the names of these mutants are not actually mentioned in the abstract, although this information was provided when the poster was presented at the International Congress on Cell Biology (B. Byers, personal communication). Reproduced from Byers, B. and Goetsch, L. (1976a) J. Cell Biol., 69, 717–21, with permission from Rockefeller University Press

mutants, but not in 20 other cdc mutants, they observed that these filament-like structures were lost after temperature shift and that the rates of filament loss in the various mutants correlated with the rates at which cytokinesis ability was lost in the same mutants (Figure 1.4; this is an interesting example of a highly influential study that was never published in detail). Second, my own group’s interest in cell polarization led us to study the actin and microtubule cytoskeletons; this in turn led to the adaptation of immunofluorescence methods to yeast by Alison Adams in collaboration with John Kilmartin in England (Adams and Pringle, 1984; Kilmartin and Adams, 1984). In these studies, we also exploited the elongated-bud phenotype of the cdc3, 10, 11, and 12 mutants to demonstrate the highly polarized distribution of actin structures that first suggested their role in the polarized delivery of secretory vesicles to the tip of the bud. Finally, other yeast geneticists discovered that genes could be cloned by plasmid rescue of the corresponding mutants, and this approach was used to clone CDC10 as a step toward the analysis of the nearby centromere (Clarke and Carbon, 1980). The availability of this CDC10 clone allowed Kaback and Feldberg (1985) to observe that this gene was among those strongly induced during sporulation, a harbinger of later studies showing a role for the septins in this process (see below). In 1983, Sue Lillie and Brian Haarer in my lab finally began systematic molecular analysis of CDC3, 10, 11, and 12, and they were soon joined by several others. Such work was still fairly challenging in those days, and we encountered a number of confusing complications in the course of cloning the genes and establishing that the clones were correctly identified and faithfully represented what was present in the genome. For example, because the common phenomenon of dosage suppression had not been recognized previously, it took us months to

12

CH01 ORIGINS AND DEVELOPMENT OF THE SEPTIN FIELD

realize that the several plasmids isolated by rescue of a cdc11 mutant all actually contained CDC12 (Pringle et al., 1986)! However, we had the clones sorted out by 1986, and Brian had been able to fuse portions of CDC12 in frame to E. coli lacZ and trpE (because we didn’t yet have sequence information, he cloned random fragments of CDC12 into the fusion vectors and screened for clones that produced a fusion protein), use the fusion proteins to make Cdc12p-specific antibodies, and use these antibodies to demonstrate that Cdc12p localized to the mother-bud neck (Haarer and Pringle, 1987). This demonstration required Brian to overcome another confusing artefact that resulted from the presence in most rabbit sera of antibodies that recognized a yeast cell-wall component (probably chitin) that is also localized to the neck and often remains there during the preparation of cells for immunofluorescence. Thus, we had two moments of high excitement in this project: first, when the unpurified antiserum (but not a control serum) gave us staining at the neck, the exact place where we expected to find Cdc12p, and second, when affinity-purified antibodies also stained the neck. Of course, these highs were separated by the profound low of realizing that the first results were an artefact! Before long, we had also obtained good antibodies to Cdc3p, 10p, and 11p, and shown that these proteins also localized to the neck (Figure 1.5a and b; Kim, Haarer and Pringle, 1991; Ford and Pringle, 1991). Importantly, we demonstrated that all four proteins were lost simultaneously from the neck when any one of the four ts mutants was shifted to restrictive temperature; taken together with the results of Byers and Goetsch (see above), this suggested that Cdc3p, 10p, 11p, and 12p formed a complex that contributed to the filament-like structures seen by EM. Further support for this conclusion was obtained in 1991, when Jon Mulholland and Daphne Preuss in David Botstein’s lab obtained some beautiful immunogold images using the antibodies that we provided (Figure 1.3c–e). Although we unfortunately did not find a vehicle for publication of these images, they were shown widely and mentioned in a review (Longtine et al., 1996), and they helped establish two points that were not clear from fluorescence images, namely that the proteins of interest, like the filaments described by Byers and Goetsch, are closely associated with the plasma membrane and are present throughout the neck region (so that the structure in which they are found during most of the cell cycle is an hourglass-shaped band and not a pair of discrete rings, despite the common appearance of fluorescence images resulting from the geometry of the neck). Meanwhile, in the summer of 1986, Brian and others in the lab had begun sequencing the four genes, another process that was rather laborious in those days. In Brian’s report to his thesis committee in December 1986 (which I still have), all four genes were described as partially sequenced, but there was no hint that any sequence similarities had been found. But in early February 1987, I wrote letters of recommendation for Brian that described the ‘very interesting result’ that Cdc3p, 10p, 11p, and 12p formed a family of related proteins. So although I have no record of the precise date, the ‘Eureka moment’ must have been in January 1987. It took quite a bit longer actually to finish the sequencing of both strands and to feel confident that we had all the bugs out of both this and the original

CONCEPTION, GESTATION, AND BIRTH

13

(a)

(b)

(c)

Figure 1.5 Excerpts from abstracts for talks at the biannual Cold Spring Harbor Yeast Cell Biology Meetings of 1987 (a), 1989 (b) and 1991 (c). In (c), cdc103+ was the gene later named spn1 (Longtine et al., 1996); the Drosophila gene was sep1 (Fares, Peifer and Pringle, 1995); and the mammalian (mouse) genes were Sept1 (Nottenburg, Gallatin and St. John, 1990) and Sept4 (Kato, 1990), originally called DIFF6 and H5 . For a Cold Spring Harbor cytoskeleton meeting in April 1991, we submitted an abstract nearly identical to (c), except that we did not yet know about the mouse genes. Although we were very slow to publish our results, we were not secretive about them! Reproduced with permission from the authors

14

CH01 ORIGINS AND DEVELOPMENT OF THE SEPTIN FIELD

characterization of the clones, but we were comfortable with presenting the central results at a major meeting in August 1987 (Figure 1.5a). (After this presentation, Jeremy Thorner referred, prophetically, to the project as a ‘gold mine’, but most other people did not seem very interested in this out-of-the-mainstream work.) It also took many months of hard work to write it all up: the work had been done by seven people (most of whom had already left the lab) over a five-year period; the story had some intrinsic complications; and methods had evolved rapidly during the years that the work was being done (so that the ways we had actually done things were not always how we would have done them at the time of writing). However, by July 1990, we thought that we had satisfactorily addressed every complication and produced a pair of tightly written manuscripts, which we submitted with great pride to Genetics (Figure 1.6). In a telephone conversation on about August 25, the editor indicated that the reviews were generally positive (as indeed they proved to be), and on this basis we described the papers as ‘in press’ when we submitted the Cdc3p-localization manuscript (Kim, Haarer and Pringle, 1991) on August 27. Thus, it was with shock bordering on disbelief that I received an editorial decision letter on August 29 that said, among other unpleasant things, that the manuscripts were ‘intensely irritating to read’ and that I should have known that the two manuscripts would need to be combined into a single manuscript. (Actually, this thought had never even crossed my mind.) The letter was the more shocking in that it came from an editor with whom I had worked on other projects and thought of as someone who shared my deeply held beliefs about the importance of addressing, rather than glossing over, the technical details and complications of a study. I couldn’t really see how to re-write the material as

Figure 1.6 The two manuscripts in which we attempted to report the cloning and sequencing of the original four septin genes were submitted to Genetics in July 1990. The postcards acknowledging receipt are shown. Authors of the first paper were Sue Lillie, Brian Haarer, Laird Bloom, Kevin Coleman, and myself; authors of the second paper were Brian Haarer, Stuart Ketcham, Susan Ford, David Ashcroft, and myself

CHILDHOOD

15

a single paper of reasonable length, and although it was years before I gave up on publishing these papers, both life (I moved from Michigan to North Carolina in 1991, and none of the people involved in this phase of the project moved with me) and science (see below) moved on without its ever getting done. We did talk widely about the results, refer to them in the 1991 protein-localization papers, and provide details and reagents to the few who were interested. We also entered the sequences into the public databases, although, embarrassingly, this didn’t happen until May 1993 (as a result of a miscommunication and a lack of follow-up by me, rather than any attempt to be secretive), with some consequences as described below. At some point in 1989 or 1990, we began in the lab to refer to the protein family as ‘septins’ (for their role in septation), but we were too diffident to apply this term in our 1991 publications, a 1992 paper in which we proposed to introduce the term was aborted (see below), and we published nothing more on septins until 1995. Thus, when publication of the Neufeld and Rubin (1994) Pnut paper (see below) elicited the minireview by Sanders and Field (1994), Sylvia and Chris had to call to ask if we had yet given the proteins a name. I told them, and they kindly used it (with attribution) in their review, thus introducing this convenient term into the literature.

CHILDHOOD As the S. cerevisiae septin picture was coming into focus in the late 1980s, we naturally began to wonder whether such proteins were also found in other organisms. EM observations by Soll and Mitchell (1983) had shown that Candida albicans also had ‘neck filaments’, but even if these proved to contain septins (as indeed found later: see Chapter 5 by Gladfelter and Sudbery), this didn’t seem to expand our horizons very much. Thus, we turned to the fission yeast Schizosaccharomyces pombe, which is more distant phylogenetically from S. cerevisiae and also distinct morphologically in that it divides by medial fission rather than by budding. I thought that we would be competent to handle another yeast in the lab, and this proved more-or-less true, although we also got a huge assist when S. pombe expert Peter Fantes came from Edinburgh on sabbatical early in the project. I also thought that the comparison of septin function in these two yeasts might be highly informative, just as the comparative analysis of cell-cycle control had been (Hartwell, 2002; Nurse, 2002); this has arguably been the case, albeit in the rather perverse way of revealing to us mostly the complexity of septin biology and the major gaps in our understanding of cytokinesis (see below and Chapter 5)! Our initial approach, in 1988, was to screen our antibodies against the S. cerevisiae proteins for cross-reaction with S. pombe proteins; we found a plausible potential homologue only with an anti-Cdc3p antibody (Figure 1.7a). A long struggle by Annette Healy and Hyong Kim with the λgt11 expression-vector system then led eventually to the cloning of the corresponding S. pombe gene (spn1 ),

16

CH01 ORIGINS AND DEVELOPMENT OF THE SEPTIN FIELD

(a)

(b)

(c)

Figure 1.7 Initial identification and localization of an S. pombe septin (H. Kim, A. Healy, T. Pugh, P. Fantes, and J. Pringle, unpublished results). (a and b) Recognition of Spn1p in immunoblots. (a) Antibodies raised against Cdc3p (Kim, Haarer and Pringle, 1991) recognized Cdc3p in a blot of S. cerevisiae proteins (lane 1) and a protein of similar molecular weight in a blot of S. pombe proteins (lane 2). As Cdc3p and Spn1p are only ∼45% identical in amino-acid sequence, this cross-reaction sounds a cautionary note for investigators working in organisms with multiple septins. (b) After cloning spn1 , Spn1p-LacZ and Spn1p-TrpE fusion proteins were used to raise and affinity purify antibodies specific for Spn1p, as shown by staining blots of proteins from wild-type cells (lane 1), an spn1 deletion strain (lane 2), and a strain carrying a high-copy spn1 plasmid (lane 3). (c) The antibodies used in (b) were also used for immunofluorescence staining of wild-type S. pombe cells; the cells shown were fixed at the time of cytokinesis. Reproduced with permission from Kim, H.B., Haarer, B.K. and Pringle, J.R. (1991) J. Cell Biol., 112, 535–44, Copyright  1991 Elsevier

and we had enough sequence information to report the result at 1989 summer meetings (Figure 1.5b). Tom Pugh joined the project at this point, and he, Hyong, and Peter (and later Omayma Al-Awar) proceeded both to investigate Spn1p function and to isolate four additional genes (spn2-spn5 ) by a combination of PCR (new to the lab at this time) with degenerate primers and rescue of the S. cerevisiae septin mutants using a library of S. pombe cDNAs in an appropriate expression vector. They found that Spn1p localized to the division site (Figure 1.7b and c) but that knockout of spn1 did not block cytokinesis or septum formation, although it did produce a delay in the separation of the daughter cells. We reported our progress at 1991 summer meetings (Figure 1.5c) and began drafting a manuscript, but our zeal to finish it was reduced by the supposition that the mild phenotype of the spn1 mutant probably resulted from redundancy in function with one or more of the other septins. At this critical juncture came the move to North Carolina and the dispersal of the S. pombe group, of whom only Omayma moved with me. She was joined in a few months by Maria Valencik, and the two of them began to look for additional genes (finding spn6 ) and to construct the required single, double, and multiple mutants (which was challenging in those days because of a paucity of selectable markers for S. pombe gene knockouts and the lack, at the time, of a PCR method for generating the knockout constructs). Discouragingly, even the multiple mutants still displayed only the mild phenotype of cell-separation delay, leaving the puzzle, which endures to this day (see below and Chapter 5), as to why the seemingly conserved septin array at the division site is essential for cytokinesis in some cell types (such as S. cerevisiae) and non-essential in others.

CHILDHOOD

17

Maria understandably became discouraged and eventually abandoned the project for a look at the mouse septins (see below); Omayma bravely soldiered on and completed what became half of her thesis, but it was not until 2001 that work by J¨urg B¨ahler and Jian-Qiu Wu had filled in all the gaps (including the identification of spn7 based on a partial sequence released by the S. pombe genome project). In the meantime, we had described many of the results at the 1993 Cold Spring Harbor Yeast Meeting, deposited the sequences in the public databases (although not until August 1995), and described some of the results in our 1996 review (Longtine et al., 1996). Although other groups have now begun to work on the S. pombe septins (see Chapter 5), and we have provided them with information and reagents, we still have considerable material that really ought to be published, and I still have hopes, but with a trail so long and complicated it is not a simple matter. Meanwhile, armed with degenerate PCR and the knowledge that the septins were evolutionarily ancient at least in the fungi, we had been emboldened to look farther afield, and by mid-1990 Johnny Fares had identified a Drosophila gene (sep1 ), which we reported (eliciting no great interest) at meetings in April and August 1991 (Figure 1.5c). Functional analysis of the Drosophila septin was stymied for some time because we had no experience with flies and no appropriate collaborator at Michigan, Johnny’s first attempts to generate antibodies failed, and Johnny himself also had several yeast projects that took much of his time. Then, in May 1991, there was an exciting development: after four years of checking the sequence databases intermittently without success, we noticed a mouse sequence that was clearly homologous to the yeast and fly septins. Nottenburg, Gallatin and St. John (1990) had been trying to clone the gene for a glycoprotein thought to be involved in the binding of lymphocytes to endothelial layers. They recovered a cDNA (‘DIFF6 ’) whose expression levels in different cell lines matched those of the glycoprotein of interest, but, to their disappointment, the clone did not encode that protein, and they found no DIFF6 homologues when they searched the databases (for the reasons described above). Thus, DIFF6 (now SEPT1) has the honour of being the first septin whose sequence was published and entered in the public databases. [Kato (1990) described H5 (now Sept4 ) as one of a set of mouse cDNAs isolated on the basis of their interesting expression patterns in the brain, but the sequence was not entered into the databases until August 1991; Steensma and van der Aart (1991) sequenced a chromosome region that included CDC10 as an early step in the S. cerevisiae genome project, but their sequence (checked against ours, which they knew about from our meeting presentations) wasn’t deposited until May 1992.] Immediately after seeing the DIFF6 sequence, I called Tom St. John (whom I knew slightly from his earlier work on yeast) and told him the septin story, which he was pleased to hear. Remarkably, he had also recently been contacted by George Miklos, a Drosophila geneticist in Australia who had also sequenced a gene encoding a DIFF6 homologue. George’s group was interested in several genes defined by mutations that produced behavioural and/or neurological phenotypes

18

CH01 ORIGINS AND DEVELOPMENT OF THE SEPTIN FIELD

and mapped near the centromere of the X chromosome. To identify the genes, they had cloned this chromosome region and were characterizing all the transcription units. Between the genes that proved to correspond to the mutations small optic lobes (Delaney et al., 1991) and sluggish (Hayward et al., 1993) were two other transcribed regions that did not appear to correspond to any of the mutations of interest; one of these had shown the homology to DIFF6 that prompted the contact with the St. John group. On May 14, I sent large packets of information on the septins (including various sequence alignments) to both the St. John and Miklos groups; arriving at the latter without prior warning, I suppose that it produced a stir, especially given that our sep1 proved to be identical to their gene (whose sequence we had not yet seen when I sent our packet). In any case, a cordial and informative letter arrived by fax on May 22 and was followed by a phone call, and we soon agreed to publish a joint paper reporting that proteins homologous to the ‘neck-filament-associated proteins’ in yeast were present in both flies and mammals. I was excited, because I thought that this little paper would markedly raise the level of interest in this family of proteins. I proposed that because St. John and co-workers had not been able to say anything interesting about the DIFF6 sequence when they published it, but had put out the sequence that allowed us all to make contact, they should now be included as co-authors, and this proposal appeared to be accepted (albeit with some reluctance). Work began on the paper but proceeded rather slowly (too slowly for George’s satisfaction) because of the disruption caused by my move in August to North Carolina and the other pressures (notably the editing of the Cold Spring Harbor yeast monograph) that I was facing at that time. George also objected to several aspects of my proposed organization of the paper, as well as to my rather hesitant proposal to call the proteins ‘septins’ (arguing that this was premature without knowledge of their function in animal cells). Nonetheless, by April 1992, after lots of hard work and patient negotiation, I thought that we had a manuscript that all could agree on, and I shipped it off to Australia. Several weeks later, I received a fax announcing George’s intention to withdraw his group from the paper; the main sticking point was the inclusion of St. John and co-workers as co-authors, which was now described (to my astonishment) as ‘outrageous’. This was the last straw for me, and I made no further attempt to rescue the joint publication. The Miklos group entered their sequence into the public databases in July 1992 but never published any of their other information on sep1 , which is a shame because they apparently had genetic data (based on deletion analysis of the region) that would have shed valuable light on Sep1 function. George apparently also considered himself and his group to be the injured parties (which I could never understand), and even after we had finally obtained a good antibody and an appropriate collaborator and published our paper on Sep1 localization and possible functions (Fares, Peifer and Pringle 1995), they declined to acknowledge the septin name or the progress in the field and used their own terminology (‘innocent bystander’ for sep1 and the ‘innocent bystander family’ for the septins) on through their last publication in the area (Maleszka, De Couet and Miklos, 1998).

EARLY ADOLESCENCE

19

During the early 1990s, several other groups encountered mammalian septin genes during studies of other topics (Kumar, Tomooka and Noda, 1992; Nakatsuru, Sudo and Nakamura, 1994; H. Zogbhi, personal communication), but except for an unsuccessful attempt by Makoto Noda’s group (in collaboration with yeast expert Kuni Matsumoto) to rescue the yeast septin mutants using their mouse Nedd5 (now Sept2 ) clone (we sent strains in October 1993, along with lots of information about the yeast and fly septins, and learned of the negative results in December), there was little indication of interest in actually studying septin function. (Although the December letter mentioned that ‘we are raising antibody’, we didn’t learn until much later that Makoto Kinoshita in the Noda lab was leading a serious effort to investigate NEDD5 function, as described below.) Thus, when Maria Valencik became frustrated with the S. pombe project (see above), I encouraged her proposal to begin looking at mammalian septin localization even though our lab was not well equipped for studies of either mice or cultured mammalian cells. We obtained the Diff6 cDNA from Tom St. John and an H5 cDNA from Huda Zoghbi in November 1993 and began trying to make antibodies using fusion proteins, eventually succeeding with both the amino- and carboxyl-terminal halves of DIFF6. (The eventual success was with rabbit antibodies, but this did not happen until after we had first followed the enthusiastic advice of a colleague and tried to raise antibodies in chickens, yielding about half a coldroom full of eggs but no useful antibodies!) Maria used the cDNAs and antibodies to get a variety of interesting results (Figure 1.8), but we still had only fragments of a story when she left the lab in August 1995 (for family reasons and an opportunity to learn mouse methods in a more appropriate environment). Thus, although we briefly described the principal results in our 1996 review (Longtine et al., 1996), no full account of this work was ever published.

EARLY ADOLESCENCE Although a few other groups had happened upon septin genes, as described above, attempts to explore septin function were essentially confined to my laboratory for about 10 years, which of course meant that progress was rather slow and there was little interest by the outside world. Although a paper on cdc10 mutations and yeast budding patterns (Flescher, Madden and Snyder, 1993) was more important than was apparent at the time (see below), I think that the maturation of the septin field really began with the analysis of Drosophila septin function by Neufeld and Rubin (1994). In a complex screen for mutations affecting eye development, Tom Neufeld had recovered a mutation that he named pnut, following the traditional (and maddening to outsiders) practice in Drosophila genetics of giving whimsical names to genes; in this case, the name was based on the superficial resemblance of some cells that had failed cytokinesis, and thus had several nuclei, to peanuts in the shell (T. Neufeld, personal communication). When the gene was sequenced in early 1992, database searches revealed only the DIFF6 and H5 sequences

CH01 ORIGINS AND DEVELOPMENT OF THE SEPTIN FIELD

20 (a)

(b)

(c)

(d)

Figure 1.8 Some early observations on mammalian septins (M. Valencik and J. Pringle, unpublished; Longtine et al., 1996). (a) SEPT1 concentrates in the cleavage furrow (arrow) of at least some dividing mammalian cells (an EL4 lymphoid cell is shown; similar observations were made on NB41A3 neuroblastoma cells). Note that in the coloured originals, the propidium iodide-stained telophase chromosomes (*) were clearly distinguished from the Bodipy-labelled secondary antibody used in the SEPT1 staining. (b) SEPT1 is particularly concentrated in cells of the central nervous system. SDS-PAGE and immunoblotting were performed on protein extracts from various organs of a dissected mouse (1, thymus; 2, spleen; 3, cerebrum; 4, cerebellum; 5, spinal cord; 6, lung; 7, stomach; 8, liver; 9, pancreas; 10, kidney; 11, heart). (c) SEPT1 is concentrated in the growth cones of differentiating PC-12 neural cells in culture. As in (a), the red-fluorescent nucleus (*) and green-fluorescent antibody staining (around the tips of the multiple growth cones) were distinct in the original. (d) Evidence for differential expression and differential splicing of Sept1 and Sept4 in different tissues. A commercially obtained blot of poly A+ RNAs from various mouse organs (shown are 1, heart; 2, brain; 3, spleen; 4, lung; 5, testis) was hybridized to radiolabelled Sept1 (left) and Sept4 (right) cDNAs. Reproduced with permission from Longtine, M.L. et al . (1996) Curr. Opin. Cell Biol., 8, 106–19

and no connection to the yeast proteins or any biological function (see above). Nonetheless, Tom and Gerry were soon in contact with me. (We have been unable to reconstruct exactly how this happened; the path presumably led through Nottenburg and St. John, although George Miklos – who visited the Rubin lab at about this time – may also have been involved.) In March and April 1992 (by phone and mail, and when I visited Berkeley for a seminar that had been arranged months earlier for other reasons), I passed on everything that we knew at that point about the septins, including our progress with Drosophila Sep1. Additional exchanges of information and reagents followed, so that by the time the Pnut paper was submitted in early 1994, Tom and Gerry were able to provide extensive context that added to the impact of their own highly interesting results. The Pnut paper was important for several reasons. First, it was the first conspicuous public announcement that this family of proteins existed in animals and not

EARLY ADOLESCENCE

21

just in budding yeast. [As described above, we had presented this fact at meetings but not in any publication. And although Flescher, Madden and Snyder (1993) had been able to point out that Cdc10p-like proteins existed in flies and mammals, this paragraph was buried within their paper and unlikely to have been noticed by many.] Second, it presented both protein-localization and mutant-phenotype data that were highly informative about septin function in flies. In particular, it showed that the septins were involved in cytokinesis in animal cells as well as in yeast, while also suggesting strongly that they had other roles, notably in the nervous system. Third, it helped to get Chris Field and Tim Mitchison interested in the septins (Sanders and Field, 1994), paving the way for their later seminal contributions to the field (see below). The Rubin lab did not continue studies of the fly septins after Tom Neufeld’s graduation, but my lab carried on with the indispensable collaboration of Mark Peifer, a card-carrying fly geneticist who had joined UNC as an Assistant Professor within a few months of my own move there from Michigan. A combination of genetic, protein-localization, and biochemical studies (Fares, Peifer and Pringle, 1995; Adam, Pringle and Peifer, 2000; Shih et al., 2002; see Chapter 6 by Field et al.) produced additional important information, including (i) evidence for differentiation of function among the fly septins; (ii) evidence that the septins may be non-essential for cytokinesis in some cell types even though they are essential in others; (iii) evidence that Sep2, unlike Pnut, is dispensable for development to adulthood [although this issue is complicated by the possibility of functional redundancy with the very similar (73 % sequence identity) Sep5]; and (iv) evidence that Pnut, Sep1, and Sep2 are all particularly highly concentrated in non-dividing cells of the embryonic central nervous system, suggesting strongly (as did the preliminary data for mouse; see Figure 1.8) that the septins must have important roles in at least some cell types that are unrelated to cytokinesis. Unfortunately, despite this progress and the promise of much more, the fly septin project waned at UNC: Mark lost interest as the other projects in his lab gained momentum; Johnny Fares, Jenny Adam, and Hsin Shih all graduated and were not replaced by other students; and postdoc Karen Hales received an irresistible offer of a faculty position while she still had a year left on her fellowship (and when she was just on the verge of getting the sep5 mutant that would have allowed the critical sep2 sep5 double mutant to be constructed). [We never attempted to generate a sep1 mutant because of the (unrealized) expectation that the Miklos group would eventually publish something on this topic (see above).] I thought that the fly septin project might revive when I moved from UNC to Stanford in 2005, but with more and more of my own attention devoted to our new study of the dinoflagellate-cnidarian symbiosis (we have cloned two anemone septin genes, but it is unlikely that they will become a major focus of the project), I now think that this is unlikely. Thus, it seems to me that study of the septins in the genetically tractable Drosophila system, where septin mutations even give some strong phenotypes, represents an extraordinary opportunity for an enterprising investigator who wants to do something different and important.

22

CH01 ORIGINS AND DEVELOPMENT OF THE SEPTIN FIELD

Meanwhile, our labours in Drosophila had also allowed us to contribute in a small but important way (the anti-Sep1 and anti-Sep2 antibodies) to the first serious and successful effort at a biochemical analysis of the long-presumed septin complexes, the breakthrough study by Chris Field, Tim Mitchison, and co-workers (Field et al., 1996). As a research associate in Bruce Alberts’s lab, Chris had been using biochemical approaches to the study of Drosophila cytoskeletal proteins (discovering, among other things, anillin: Field and Alberts, 1995; and see Chapter 6), so that she was well positioned to take a similar approach to the fly septins once her (and husband Tim’s) attention had been drawn to these proteins. This study established or solidified the evidence for multiple important points, including (i) the several septins in a given cell type (in this case, at least Pnut, Sep1, and Sep2) do form a physical complex with each other; (ii) as suggested by their sequence motifs, the septins can bind and hydrolyse GTP; and (iii) the septin complexes can form at least short filaments in vitro (a point of particular interest given the paucity of evidence for septin filaments in vivo other than in S. cerevisiae). With Tim’s student Jen Frazier leading the way, the approach was soon extended to the S. cerevisiae septins (Frazier et al., 1998), leading to similar conclusions (except that much longer filaments could be formed in vitro) and a solid beginning for yeast septin biochemistry (for further discussion, see Chapters 3 and 6). Field et al. (1996) also extrapolated from their structural observations to propose a new model for septin-filament organization in yeast, in which the filaments would run longitudinally along the mother-bud axis rather than in a helix around the neck as Byers and Goetsch (1976a) had proposed and we all had thought ever since. At the time, I thought (and politely mentioned to Chris and Tim) that this model was not justified by the data and indeed was almost certainly totally wrong. However, I soon changed my tune when Mark Longtine showed that cells defective in the protein kinase Gin4p displayed a reorganization of the septins (into thick bars running through the neck) that I found (and, indeed, still find) easy to explain if the septin filaments are longitudinal and hard to explain if they are helical (Longtine, Fares and Pringle 1998). Although the jury is still out on the ultimate validity of this model (see the detailed discussion in Chapter 3), there can be no doubt that it, and the experiments and discussions it has spawned, have contributed greatly to the development of our understanding of septin organization. There were several other important developments in the septin field during the mid-1990s. First, studies by Beth DiDomenico and Yigal Koltin in Candida albicans (first communicated to me in late 1991; later published by DiDomenico et al., 1994), by Michelle Momany and co-workers in Aspergillus nidulans (Momany et al., 1995; Momany and Hamer, 1997), and by Michael Glotzer and Tony Hyman in Xenopus laevis [first communicated to me in October, 1994; abstract and poster presented in 1996 (Glotzer and Hyman, 1996); eventually published in part in 2002 (Mendoza, Hyman and Glotzer, 2002)], taken together with the S. pombe and Drosophila data, made it seem certain that the septins would prove to be ubiquitously present in both fungi and animals.

EARLY ADOLESCENCE

23

Second, Fatima Curckov´a in Kim Nasmyth’s lab performed a genetic screen to identify yeast proteins that were critical in cells deficient in G1 cyclins; she identified mutations both in CDC12 and in the novel gene CLA4 and showed that cla4 mutations affected septin organization (Cvrckov´a et al., 1995). Because Cla4p proved to be a protein kinase in the PAK family, activated by Cdc42p and/or other Rho-type GTPases, this was the first step in trying to understand how Cdc42p controls the spatially polarized localization of the septins (among other things), an effort that continues into the present (see Iwase et al., 2006, and references cited therein). Third, in mid-1995 I received an invitation to write a short review on ‘recent progress’ with the septins for Current Opinion in Cell Biology. Recognizing that enough other people were becoming interested that a true septin field was beginning to emerge, and acknowledging (even if tacitly and grudgingly) that some of our early work would probably never be published in detail, I decided that we should try to put the new field on a sound footing by providing a comprehensive review of what was known about the septins to that point (including our own unpublished results and personal communications from others as needed to make the state of the art clear). The resulting paper (Longtine et al., 1996) was rather different from (and about twice as long as!) what the commissioning editors had envisaged, but after some discussion they agreed to publish it anyway. In re-reading this review a dozen years after it was written, it still seems to me that it did quite well in summarizing both what was known at the time and the challenges that lay ahead. Fourth, some important general points about septin function were revealed by studies of the rather specialized process of yeast ascospore formation. To ask if there were additional septins in S. cerevisiae beyond Cdc3p, 10p, 11p, and 12p, Johnny Fares had used degenerate PCR. He recovered and began to study a fifth septin gene. The knockout strain had no detectable vegetative phenotype but had a defect in spore formation, which caused me to start calling friends in the sporulation field. From Mary Clancy [from whose lab Tom Pugh (see above) had come to mine], we learned that our gene was identical to the previously named SPR3 (Ozsarac et al., 1995; Fares, Goetsch and Pringle, 1996), which had been identified much earlier by Mary and Pete Magee (Clancy et al., 1983) as one of a set of genes with SPorulation Regulated patterns of gene expression. Because Mary’s (and collaborator Ian Dawes’s) focus was on the mechanisms of transcriptional regulation (Kao et al., 1989; Ozsarac et al., 1997), there was little information about the function of Spr3p. However, an important role for the septins during sporulation was suggested strongly by the very strong transcriptional induction of SPR3, CDC10 (Kaback and Feldberg, 1985; see above – previously these data had appeared to make no sense), and a sixth septin gene that was revealed by the yeast genome project and named SPR28 (to connect it to SPR3 and minimize the number of gene-name acronyms) when it also proved to be induced strongly during sporulation (De Virgilio, DeMarini and Pringle, 1996). Although the fascinating process of ascospore formation, and the role of the septins in this

24

CH01 ORIGINS AND DEVELOPMENT OF THE SEPTIN FIELD

process, remain poorly understood (Fares, Goetsch and Pringle, 1996; De Virgilio, DeMarini and Pringle, 1996; Moreno-Borchart et al., 2001; Neiman, 2005), these early studies provided the first solid evidence for two important general points, namely that an organism can express different subsets of its septin genes, and form complexes containing different sets of septin proteins, in different cell types. Because Maria Valencik’s early studies of the mouse septins had also pointed to at least the first of these conclusions (Figure 1.8), we felt comfortable in making this point quite strongly in Longtine et al. (1996). Fifth, several different lines of investigation led to development of the scaffold model for septin function. In retrospect, the first intimations of this mode of septin action had come much earlier. In the course of their long-running studies of septum formation, Enrico Cabib and his co-workers had reported, based on EM observations that I myself found rather difficult to evaluate, that ‘septum-like structures’ formed at ectopic locations in the cdc3, 10, 11, and 12 mutants (Slater, Bowers and Cabib, 1985). At about the same time, Alison Adams observed by fluorescence microscopy (using a specific dye) that these mutants displayed a diffuse deposition of chitin, in contrast to the normal tight localization of this cell-wall component to a ring around the neck and the primary layer of the septum (Adams, 1984). This observation suggested that Cdc3p, 10p, 11p, and 12p were involved in the localization of the chitin synthase(s), an hypothesis that was reinforced when we later observed that in cells changing shape in response to mating pheromone, the diffuse deposition of chitin (Schekman and Brawley, 1979) was correlated with the presence of the septins in a zone that was much more diffuse than the tight band seen in vegetative cells (Ford and Pringle, 1991; Kim, Haarer and Pringle, 1991). Alison had also observed that the normal concentration of actin at the neck during cytokinesis failed to occur in the cdc3, 10, 11, and 12 mutants (Adams and Pringle, 1984). However, this observation was difficult to interpret at the time (and indeed remained so until 1997, when we finally began to understand the various actin structures that form at the neck during cytokinesis). Thus, in the early 1990s, we had three dots, but two of them were a little fuzzy, and so far as I can reconstruct, we had not connected them into any general model for septin function. This situation might have changed when Mike Snyder and his co-workers analysed a cdc10 mutant that they had isolated in a screen for genes whose products interacted with Spa2p, which they were studying intensively at the time (Flescher, Madden and Snyder, 1993). On the basis of the mutant’s abnormal budding pattern, they proposed (correctly as it turned out) that the septins functioned to localize to the division site some protein(s) that marked future budding sites. However, at the time this paper had little impact (at least on me), partly because the putative marker protein(s) were purely hypothetical at this point and partly because the cell-polarization side of my lab (and of my own brain) felt that the methods used in Mike’s lab to score budding patterns made their conclusions difficult to evaluate. In any event, it was not long before work in my own and Ira Herskowitz’s laboratories identified some of the proteins involved in marking

LATE ADOLESCENCE

25

potential budding sites and showed that they indeed localized to the division site in a septin-dependent manner (Chant et al., 1995; Sanders and Herskowitz, 1996). In addition, at about this time, Jamie Konopka found that Afr1p, a protein induced by mating pheromone and involved in producing the change of shape in responding cells, interacts with Cdc12p and co-localizes with the septins both in mating cells and (when ectopically expressed) in vegetative cells (Konopka, DeMattei and Davis, 1995), providing another example of a protein whose localization appeared to depend upon the septins. Thus, by 1996, there were lots of dots. However, from re-reading our 1996 review, it does not appear that we had connected them into a coherent general model until a combination of genetic and protein-localization experiments by Doug DeMarini made a persuasive case that the chitin synthase Chs3p was anchored at the neck by a hierarchical scaffold in which Chs3p bound its activator Chs4p, which bound the newly identified Bni4p (necessary for Chs3p localization but not for its activity), which bound the septin complex through Cdc10p (DeMarini et al., 1997). Additional support for the scaffold model soon came from other studies, such as the demonstration by Erfei Bi (Bi et al., 1998) and by Lippincott and Li (1998) that the actin structures at the division site include an actomyosin contractile ring and that the myosin component of this ring is recruited to the budding site early in the cell cycle in a septin-dependent manner. However, I think that the result that really crystallized the scaffold model was the demonstration by Mark Longtine that when the septins reorganize in the absence of Gin4p (see above), the various proteins that display septin-dependent localization undergo a parallel reorganization (Longtine, Fares and Pringle, 1998). From this time forward, additional examples of septin-dependent localization accumulated rapidly, so that by the time Amy Gladfelter bravely led an effort to summarize the known information, there were at least 22 proteins that were known to localize to the neck, in a variety of temporal and spatial patterns, and with a very wide variety of functions, in a septin-dependent manner (Gladfelter, Pringle and Lew, 2001). A similar effort today would certainly at least triple this number, so I think it is well established that at least one role of the septins is to provide a scaffold for the recruitment of other proteins and probably also for their organization at the site to which they have been recruited. However, I also think that it remains a little embarrassing that we still know so little about the molecular details of this recruitment and about which proteins actually interact directly with the septins! (See additional discussion in Chapters 8 and 9).

LATE ADOLESCENCE Although new areas of cell biology are often pioneered by studies in model organisms, they are not usually considered mature (unless they are plant or microbe-specific) until they have captured the attention of mammalian cell biologists. For the septins, the first big step in this direction was the publication by

26

CH01 ORIGINS AND DEVELOPMENT OF THE SEPTIN FIELD

Makoto Kinoshita and his co-workers (Kinoshita et al., 1997); this first detailed exploration of mammalian septin function had required more than three years of work since the Noda laboratory began raising antibodies to ‘NEDD5’ (now SEPT2) in 1993 (see above). Among other things, the data presented indicated that septins are important for cytokinesis in mammalian cells as well as in yeast and flies, that GTP binding and perhaps hydrolysis are important for septin function, and that the septins have distinct functions in interphase cells that involve interactions with the actin cytoskeleton. (See further discussion in Chapters 8 and 10.) Other important developments soon followed and included the following. 1. The first links of the septins to human disease (another step toward the full maturity – and funding! – of a field) appeared. The first publication appears to have been the report that a translocation implicated in the genesis of acute myeloid leukaemia fuses the MLL gene to SEPT5 (Megonigal et al., 1998). However, other studies suggesting oncogenic roles also for SEPT9 and SEPT6 soon appeared (Osaka, Rowley and Zeleznik-Le, 1999; Kalikin, Sims and Petty, 2000; Russell et al., 2000; Sorensen et al., 2000; Borkhardt et al., 2001), as did a report that the septins are concentrated in tangled fibres in the senile plaques in the brains of Alzheimer’s disease victims (Kinoshita et al., 1998), suggesting a septin role in this disease. Meanwhile, Barbara Zieger, Jerry Ware, and their co-workers observed that SEPT5 was expressed at high levels in platelets, suggesting a possible role in haemostasis and hence in bleeding disorders (Zieger, Hashimoto and Ware, 1997; Yagi et al., 1998), a hypothesis that was later supported when alterations of secretory function were observed in platelets from a Sept5 knockout mouse (Dent et al., 2002). Relationships between the septins and human diseases are discussed further in Chapters 9, 11, 12, 14, and 15. 2. The very high concentrations of septins in the central nervous system cells of Drosophila (Neufeld and Rubin, 1994; Fares, Peifer and Pringle, 1995) had suggested strongly that the septins had roles unrelated to cytokinesis (which these cells do not do) and hinted that they might have roles in vesicle trafficking (of which these cells do a lot). These ideas gained strong support as papers began to appear documenting both the high concentrations of various septins in non-dividing cells of the mammalian brain and the association of these septins with structures and proteins involved in vesicle trafficking (Caltagarone et al., 1998; Hsu et al., 1998; Kinoshita et al., 1998; Yagi et al., 1998; Beites et al., 1999; Kinoshita, Noda and Kinoshita, 2000; Xue et al., 2000). Understanding exactly what the septins are doing in neurons and in relation to vesicle trafficking there and in other types of cells remain major challenges for the field, as discussed further in Chapters 8, 11, and 15. 3. Bill Trimble, a highly accomplished investigator of vesicle trafficking, became the first major recruit to the septin field from mainstream

LATE ADOLESCENCE

27

mammalian cell biology when he leaped in with both feet (and four papers) in 1999 (Beites et al., 1999; Trimble, 1999; Xie et al., 1999; Zhang et al., 1999). This initial burst of activity from Bill’s lab provided, among other things, evidence for differentiation of mammalian septin function in different tissues, high levels of expression of some septins in brain, involvement of the septins in vesicle trafficking and secretion, and direct binding of the septins to phospholipids as at least part of the mechanism for septin-membrane association. 4. Ian Macara, a highly accomplished investigator of the many functions of small GTPases, was led to study the septins when they were found as binding partners of the Borg proteins, which his lab had identified previously as apparent effectors of Cdc42 function during cell polarization (Joberty et al., 2001). Ian’s entry into the field had significance well beyond the direct (and substantial) impact of the research contributions from his lab. First, it further legitimized the septins as an object of study by mainstream cell biologists. Second, it was Ian, after he participated in the 2001 ASCB Meeting workshop (see below and Appendix C), who said ‘enough is enough’ to the nightmare of mammalian septin names and led the successful effort to rationalize and simplify the taxonomy and nomenclature of the genes and proteins (Macara et al., 2002). It would be difficult to overstate the value that this contribution has already had, and will continue to have, to the field, and it should be stressed that Ian’s leadership could not have been successful without the cooperation of essentially everyone in the field, all of whom should feel proud of their participation in this community-spirited action. An update on this nomenclature can be found in Appendix B. As the mammalian septin field was taking off, there were also important developments on several other fronts. First, Tri Nguyen, John White, and their coworkers discovered that the Caenorhabditis elegans unc-59 and unc-61 genes, which John had identified many years earlier on the basis of mutants with uncoordinated movements (White, Horvitz and Sulston, 1982), encoded the worm’s two and only two septins and that even a mutant carrying null mutations of both genes produced viable adult worms (Nguyen and White, 1996; Nguyen et al., 2000). This work raised important questions about septin evolution (why do yeasts, flies, and mammals have multiple septins and the worm just two?), structure (do these two septins form assemblies similar to those of the multi-protein assemblies in other organisms, and, if so, how?), and function (why are the worm septins non-essential for cytokinesis in most cells even though they localize to the cleavage furrows as in other cells?) that still engage the field. Second, studies by Yves Barral, Mike Snyder, and co-workers (Barral et al., 2000) and by Peter Takizawa, Ron Vale, and co-workers (Takizawa et al., 2000) established the important point that, at least in yeast, the septins function to restrict the mobility of integral membrane proteins and other cortical components and thus

28

CH01 ORIGINS AND DEVELOPMENT OF THE SEPTIN FIELD

allow the cell to maintain distinct mother-cell and bud polarized cortical domains. Whether the septins form the diffusion barrier themselves or by recruiting other proteins that form the actual barrier is still unclear, although some arguments favour the former model (see further discussion in Chapter 4). Third, as in so many other areas of cell biology, the ability to view the behaviour of the septins in real time in living cells by using GFP and other fluorescence tags opened up the study of septin dynamics (Cid et al., 2001; Lippincott et al., 2001). Continuation of these studies in many laboratories, and in other organisms as well as yeast, has provided many insights into septin assembly, organization, and function, as discussed further in various other chapters in this book.

CONCLUDING REMARKS Most of the program for the American Society for Cell Biology’s Annual Meeting is set by the society president and program committee, but a modest number of ‘member-initiated’ workshops can be organized for the Saturday afternoon at the beginning of the meeting. In the summer of 2001, Chris Field called me to suggest that there was enough activity and interest in the emerging septin field to justify organizing such a workshop for the December 2001 meeting. I agreed with enthusiasm, and the workshop was duly proposed to the ASCB, approved, and organized (with Chris doing most of the work). As the program shows (Appendix C), it brought together many of the people who had done interesting work on the septins. Not many things could have caused me to miss this event, but the invitation to Stockholm for the celebration of Lee Hartwell’s Nobel Prize was one of them. My absence did not appear to detract much from what proved to be a very valuable session with multiple beneficial consequences, including the attraction of other investigators into the field, and I think that this session really marked the emergence of the septins as a mature (if still growing) field. From this point forward, to summarize progress with the septins requires a book (i.e. this one) and not a single chapter, so I end my history at this point. Since the 2001 session, there have been two International Septin Workshops (see Appendix C), each of which has been delightful both for its science and for its camaraderie. The first of these workshops was conceived and catalysed by my two co-editors, Hilary Russell and Peter Hall. These workshops have illustrated the growth of the field, our progress, and also how much there is still to do before we really understand the structure, dynamic assembly, and multiple functions (and dysfunctions) of this class of proteins. The next Septin Workshop should be at least as exciting, and I am looking forward to it keenly. In the meantime, for my own part, I will continue, with my group, to try to understand better the question with which we began, and to which I think we still do not really have a very satisfying answer: what do the septins do in cytokinesis, and why is this role essential in some cell types and not in others?

REFERENCES

29

ACKNOWLEDGEMENTS It is a pleasure to thank the many members of my own group, our collaborators, and the other interested scientists who have helped to build the septin field over the years. To those who have suffered because various parts of the developing story were not published in a timely way (or at all), I can only apologize and express my relief that the field has developed so well despite these lapses. I would also like to thank the National Institute of General Medical Sciences, which has generously supported our work from a time long before it was clear whether the septins would ever turn out to be relevant to human health.

REFERENCES Adam, J.C., Pringle, J.R. and Peifer, M. (2000) Evidence for functional differentiation among Drosophila septins in cytokinesis and cellularization. Molecular Biology of the Cell , 11, 3123–35. Adams, A.E.M. (1984) Cellular Morphogenesis in the Yeast Saccharomyces cerevisiae. Ph.D. Dissertation, The University of Michigan, Ann Arbor. Adams, A.E.M. and Pringle, J.R. (1984) Relationship of actin and tubulin distribution to bud growth in wild-type and morphogenetic-mutant Saccharomyces cerevisiae. The Journal of Cell Biology, 98, 934–45. Barral, Y., Mermall, V., Mooseker, M.S. and Snyder, M. (2000) Compartmentalization of the cell cortex by septins is required for maintenance of cell polarity in yeast. Molecular Cell , 5, 841–51. Beites, C.L., Xie, H., Bowser, R. and Trimble, W.S. (1999) The septin CDCrel-1 binds syntaxin and inhibits exocytosis. Nature Neuroscience, 2, 434–39. Bi, E. et al. (1998) Involvement of an actomyosin contractile ring in Saccharomyces cerevisiae cytokinesis. The Journal of Cell Biology, 142, 1301–12. Borkhardt, A. et al. (2001) An ins(X;11) (q24;q23) fuses the MLL and the Septin 6/KIAA0128 gene in an infant with AML-M2. Genes, Chromosomes and Cancer, 32, 82–88. Byers, B. (1981) Cytology of the yeast life cycle, in The Molecular Biology of the Yeast Saccharomyces: Life Cycle and Inheritance (eds J.N. Strathern, E.W. Jones and J.R. Broach), Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, pp. 59–96. Byers, B. and Goetsch, L. (1976a) A highly ordered ring of membrane-associated filaments in budding yeast. The Journal of Cell Biology, 69, 717–21. Byers, B. and Goetsch, L. (1976b) Loss of the filamentous ring in cytokinesis-defective mutant of budding yeast. (Abstract). The Journal of Cell Biology, 70, 35a. Caltagarone, J., Rhodes, J., Honer, W.G. and Bowser, R. (1998) Localization of a novel septin protein, hCDCrel-1, in neurons of human brain. Neuroreport, 9, 2907–12. Chant, J. et al. (1995) Role of Bud3p in producing the axial budding pattern of yeast. The Journal of Cell Biology, 129, 767–78. Cid, V.J. et al. (2001) Cell cycle control of septin ring dynamics in the budding yeast. Microbiology, 147, 1437–50.

30

CH01 ORIGINS AND DEVELOPMENT OF THE SEPTIN FIELD

Clancy, M.J. et al. (1983) Isolation of genes expressed preferentially during sporulation in the yeast Saccharomyces cerevisiae. Proceedings of the National Academy of Sciences of the United States of America, 80, 3000–4. Clarke, L. and Carbon, J. (1980) Isolation of the centromere-linked CDC10 gene by complementation in yeast. Proceedings of the National Academy of Sciences of the United States of America, 77, 2173–77. Cvrckov´a, F. et al. (1995) Ste20-like protein kinases are required for normal localization of cell growth and for cytokinesis in budding yeast. Genes and Development, 9, 1817–30. Delaney, S.J. et al. (1991) Molecular cloning and analysis of small optic lobes, a structural brain gene of Drosophila melanogaster. Proceedings of the National Academy of Sciences of the United States of America, 88, 7214–18. DeMarini, D.J. et al. (1997) A septin-based hierarchy of proteins required for localized deposition of chitin in the Saccharomyces cerevisiae cell wall. The Journal of Cell Biology, 139, 75–93. Dent, J. et al. (2002) A prototypic platelet septin and its participation in secretion. Proceedings of the National Academy of Sciences of the United States of America, 99, 3064–69. De Virgilio, C., DeMarini, D.J. and Pringle, J.R. (1996) SPR28, a sixth member of the septin gene family in Saccharomyces cerevisiae that is expressed specifically in sporulating cells. Microbiology, 142, 2897–905. DiDomenico, B.J. et al. (1994) Homologs of the yeast neck filament associated genes: Isolation and sequence analysis of Candida albicans CDC3 and CDC10. Molecular and General Genetics, 242, 689–98. Fares, H.F., Goetsch, L. and Pringle, J.R. (1996) Identification of a developmentally regulated septin and involvement of the septins in spore formation in S. cerevisiae. The Journal of Cell Biology, 132, 399–411. Fares, H.F., Peifer, M.A. and Pringle, J.R. (1995) Localization and possible functions of Drosophila septins. Molecular Biology of the Cell , 6, 1843–59. Field, C.M. and Alberts, B.M. (1995) Anillin, a contractile ring protein that cycles from the nucleus to the cell cortex. The Journal of Cell Biology, 131, 165–78. Field, C.M. et al. (1996) A purified Drosophila septin complex forms filaments and exhibits GTPase activity. The Journal of Cell Biology, 133, 605–16. Flescher, E.G., Madden, K. and Snyder, M. (1993) Components required for cytokinesis are important for bud site selection in yeast. The Journal of Cell Biology, 122, 373–86. Ford, S.K. and Pringle, J.R. (1991) Cellular morphogenesis in the Saccharomyces cerevisiae cell cycle: localization of the CDC11 gene product and the timing of events at the budding site. Developmental Genetics, 12, 281–92. Frazier, J.A. et al. (1998) Polymerization of purified yeast septins: evidence that organized filament arrays may not be required for septin function. The Journal of Cell Biology, 143, 737–49. Gladfelter, A.S., Pringle, J.R. and Lew, D.J. (2001) The septin cortex at the yeast mother-bud neck. Current Opinion in Microbiology, 4, 681–89. Glotzer, M. and Hyman, A.A. (1996) The biochemistry of Xenopus septins and their role in cytokinesis. (Abstract). Molecular Biology of the Cell , 7, 212a.

REFERENCES

31

Haarer, B.K. and Pringle, J.R. (1987) Immunofluorescence localization of the Saccharomyces cerevisiae CDC12 gene product to the vicinity of the 10-nm filaments in the mother-bud neck. Molecular and Cellular Biology, 7, 3678–87. Hartwell, L.H. (1967) Macromolecule synthesis in temperature-sensitive mutants of yeast. Journal of Bacteriology, 93, 1662–70. Hartwell, L.H. (1971) Genetic control of the cell division cycle in yeast. IV. Genes controlling bud emergence and cytokinesis. Experimental Cell Research, 69, 265–76. Hartwell, L.H. (2002) Yeast and cancer in Les Prix Nobel (ed T. Fr¨angsmyr), Nobel Foundation, Stockholm, pp. 246–65. Hartwell, L.H., Culotti, J. and Reid, B.J. (1970a) Genetic control of the cell-division cycle in yeast, I. detection of mutants. Proceedings of the National Academy of Sciences of the United States of America, 66, 352–59. Hartwell, L.H., Culotti, J., Pringle, J.R. and Reid, B.J. (1974) Genetic control of the cell division cycle in yeast. Science, 183, 46–51. Hartwell, L.H., Hutchison, H.T., Holland, T.M. and McLaughlin, C.S. (1970b) The effect of cycloheximide upon polyribosome stability in two yeast mutants defective respectively in the initiation of polypeptide chains and in messenger RNA synthesis. Molecular and General Genetics, 106, 347–61. Hayward, D.C. et al. (1993) The sluggish-A gene of Drosophila melanogaster is expressed in the nervous system and encodes proline oxidase, a mitochondrial enzyme involved in glutamate biosynthesis. Proceedings of the National Academy of Sciences of the United States of America, 90, 2979–83. Hsu, S-C. et al. (1998) Subunit composition, protein interactions, and structures of the mammalian brain sec6/8 complex and septin filaments. Neuron, 20, 1111–22. Iwase, M. et al. (2006) Role of a Cdc42p effector pathway in recruitment of the yeast septins to the presumptive bud site. Molecular Biology of the Cell , 17, 1110–25. Joberty, G. et al. (2001) Borg proteins control septin organization and are negatively regulated by Cdc42. Nature Cell Biology, 3, 861–66. Kaback, D.B. and Feldberg, L.R. (1985) Saccharomyces cerevisiae exhibits a sporulation-specific temporal pattern of transcript accumulation. Molecular and Cellular Biology, 5, 751–61. Kalikin, L.M., Sims, H.L. and Petty, E.M. (2000) Genomic and expression analyses of alternatively spliced transcripts of the MLL septin-like fusion gene (MSF ) that map to a 17q25 region of loss in breast and ovarian tumors. Genomics, 63, 165–72. Kao, G. et al. (1989) Independence of inessential late gene expression on early meiotic events in S. cerevisiae. Molecular and General Genetics, 215, 490–500. Kato, K. (1990) A collection of cDNA clones with specific expression patterns in mouse brain. The European Journal of Neuroscience, 2, 704–11. Kilmartin, J.V. and Adams, A.E.M. (1984) Structural rearrangements of tubulin and actin during the cell cycle of the yeast Saccharomyces. The Journal of Cell Biology, 98, 922–33. Kim, H.B., Haarer, B.K. and Pringle, J.R. (1991) Cellular morphogenesis in the Saccharomyces cerevisiae cell cycle: localization of the CDC3 gene product and the timing of events at the budding site. The Journal of Cell Biology, 112, 535–44. Kinoshita, A. et al. (1998) Identification of septins in neurofibrillary tangles in Alzheimer’s disease. American Journal of Pathology, 153, 1551–60.

32

CH01 ORIGINS AND DEVELOPMENT OF THE SEPTIN FIELD

Kinoshita, A., Noda, M. and Kinoshita, M. (2000) Differential localization of septins in the mouse brain. The Journal of Comparative Neurology, 428, 223–39. Kinoshita, M. et al. (1997) Nedd5, a mammalian septin, is a novel cytoskeletal component interacting with actin-based structures. Genes and Development, 11, 1535–47. Konopka, J.B., DeMattei, C. and Davis, C. (1995) AFR1 promotes polarized apical morphogenesis in Saccharomyces cerevisiae. Molecular and Cellular Biology, 15, 723–30. Kumar, S., Tomooka, Y. and Noda, M. (1992) Identification of a set of genes with developmentally down-regulated expression in the mouse brain. Biochemical and Biophysical Research Communications, 185, 1155–61. Lippincott, J. and Li, R. (1998) Sequential assembly of myosin II, an IQGAP-like protein, and filamentous actin to a ring structure involved in budding yeast cytokinesis. The Journal of Cell Biology, 140, 355–66. Lippincott, J. et al. (2001) The Tem1 small GTPase controls actomyosin and septin dynamics during cytokinesis. Journal of Cell Science, 114, 1379–86. Longtine, M.L. et al. (1996) The septins: roles in cytokinesis and other processes. Current Opinion in Cell Biology, 8, 106–19. Longtine, M.S., Fares, H. and Pringle, J.R. (1998) Role of the yeast Gin4p protein kinase in septin assembly and the relationship between septin assembly and septin function. The Journal of Cell Biology, 143, 719–36. Macara, I. et al. (2002) Mammalian septins nomenclature. Molecular Biology of the Cell , 13, 4111–13. Maleszka, R., De Couet, H.G. and Miklos, G.L.G. (1998) Data transferability from model organisms to human beings: insights from the functional genomics of the flightless region of Drosophila. Proceedings of the National Academy of Sciences of the United States of America, 95, 3731–36. Megonigal, M.D. et al. (1998) t(11;22)(q23;q11.2) in acute myeloid leukemia of infant twins fuses MLL with hCDC rel , a cell division cycle gene in the genomic region of deletion in DiGeorge and velocardiofacial syndromes. Proceedings of the National Academy of Sciences of the United States of America, 95, 6413–18. Mendoza, M., Hyman, A.A. and Glotzer, M. (2002) GTP binding induces filament assembly of a recombinant septin. Current Biology, 12, 1858–63. Momany, M. and Hamer, J. (1997) The Aspergillus nidulans septin encoding gene, aspB, is essential for growth. Fungal Genetics and Biology, 21, 92–100. Momany, M., Morrell, J.L., Harris, S.D. and Hamer, J. (1995) Septum formation in Aspergillus nidulans. Canadian Journal of Botany, 73, A–D. Moreno-Borchart, A.C. et al. (2001) Prospore membrane formation linked to the leading edge protein (LEP) coat assembly. The EMBO Journal , 20, 6946–57. Nakatsuru, S., Sudo, K. and Nakamura, Y. (1994) Molecular cloning of a novel human cDNA homologous to CDC10 in Saccharomyces cerevisiae. Biochemical and Biophysical Research Communications, 202, 82–87. Neiman, A.M. (2005) Acospore formation in the yeast Saccharomyces cerevisiae. Microbiology and Molecular Biology Reviews, 69, 565–84. Neufeld, T.P. and Rubin, G.M. (1994) The Drosophila peanut gene is required for cytokinesis and encodes a protein similar to yeast putative bud neck filament proteins. Cell , 77, 371–79. Nguyen, T.Q., Sawa, H., Okano, H. and White, J.G. (2000) The C. elegans septin genes, unc-59 and unc-61 , are required for normal postembryonic cytokineses and

REFERENCES

33

morphogenesis but have no essential function in embryogenesis. Journal of Cell Science, 113, 3825–37. Nguyen, T.Q. and White, J.G. (1996) Identifying septin genes in C. elegans. (Abstract). Molecular Biology of the Cell , 7, 212a. Nottenburg, C., Gallatin, W.M. and St. John, T. (1990) Lymphocyte HEV adhesion variants differ in the expression of multiple gene sequences. Gene, 95, 279–84. Nurse, P.M. (2002) Cyclin dependent kinases and cell cycle control in Les Prix Nobel (ed T. Fr¨angsmyr), Nobel Foundation, Stockholm, pp. 308–20. Osaka, M., Rowley, J.D. and Zeleznik-Le, N.J. (1999) MSF (MLL septin-like fusion), a fusion partner gene of MLL, in a therapy-related acute myeloid leukemia with a t(11;17)(q23;q25). Proceedings of the National Academy of Sciences of the United States of America, 96, 6428–33. Ozsarac, N., Bhattacharyya, M., Dawes, I.W. and Clancy, M.J. (1995) The SPR3 gene encodes a sporulation-specific homologue of the yeast CDC3/10/11/12 family of bud neck microfilaments and is regulated by ABFI. Gene, 164, 157–62. Ozsarac, N., Straffon, M.J., Dalton, H.E. and Dawes, I.W. (1997) Regulation of gene expression during meiosis in Saccharomyces cerevisiae: SPR3 is controlled by both ABFI and a new sporulation control element. Molecular and Cellular Biology, 17, 1152–59. Pringle, J.R. (2006) Cytoskeleton and morphogenesis, in Landmark Papers in Yeast Biology (eds P. Linder, M. Hall and D. Shore), Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, pp. 211–41. Pringle, J.R. et al. (1986) Cellular morphogenesis in the yeast cell cycle, in Yeast Cell Biology (ed J. Hicks), Alan R. Liss, New York, pp. 47–80. Pringle, J.R. et al. (1995) Establishment of cell polarity in yeast. Cold Spring Harbor Symposia on Quantitative Biology, 60, 729–44. Russell, S.E.H. et al. (2000) Isolation and mapping of a human septin gene to a region on chromosome 17q, commonly deleted in sporadic epithelial ovarian tumors. Cancer Research, 60, 4729–34. Sanders, S.L. and Field, C.M. (1994) Septins in common? Current Biology, 4, 904–10. Sanders, S.L. and Herskowitz, I. (1996) The Bud4 protein of yeast, required for axial budding, is localized to the mother/bud neck in a cell cycle-dependent manner. The Journal of Cell Biology, 134, 413–27. Schekman, R. and Brawley, V. (1979) Localized deposition of chitin on the yeast cell surface in response to mating pheromone. Proceedings of the National Academy of Sciences of the United States of America, 76, 645–49. Shih, H-P., Hales, K.G., Pringle, J.R. and Peifer, M. (2002) Identification of septininteracting proteins and characterization of the Smt3/SUMO-conjugation system in Drosophila. Journal of Cell Science, 115, 1259–71. Slater, M.L., Bowers, B. and Cabib, E. (1985) Formation of septum-like structures at locations remote from the budding sites in cytokinesis-defective mutants of Saccharomyces cerevisiae. Journal of Bacteriology, 162, 763–67. Sloat, B.F., Adams, A.E.M. and Pringle, J.R. (1981) Roles of the CDC24 gene product in cellular morphogenesis during the Saccharomyces cerevisiae cell cycle. The Journal of Cell Biology, 89, 395–405. Sloat, B.F. and Pringle, J.R. (1978) A mutant of yeast defective in cellular morphogenesis. Science, 200, 1171–73.

34

CH01 ORIGINS AND DEVELOPMENT OF THE SEPTIN FIELD

Soll, D.R. and Mitchell, L.H. (1983) Filament ring formation in the dimorphic yeast Candida albicans. The Journal of Cell Biology, 96, 486–93. Sorensen, A.B. et al. (2000) Sint1, a common integration site in SL3-3-induced T-cell lymphomas, harbors a putative proto-oncogene with homology to the septin gene family. Journal of Virology, 74, 2161–68. Steensma, H.Y. and van der Aart, Q.J.M. (1991) Sequence of the CDC10 region of chromosome III of Saccharomyces cerevisiae. Yeast, 7, 425–30. Takizawa, P.A., DeRisi, J.L., Wilhelm, J.E. and Vale, R.D. (2000) Plasma membrane compartmentalization in yeast by messenger RNA transport and a septin diffusion barrier. Science, 290, 341–44. Trimble, W.S. (1999) Septins: A highly conserved family of membrane-associated GTPases with functions in cell division and beyond. The Journal of Membrane Biology, 169, 75–81. White, J.G., Horvitz, H.R. and Sulston, J.E. (1982) Neurone differentiation in cell lineage mutants of Caenorhabditis elegans. Nature, 297, 584–87. Xie, H., Surka, M., Howard, J. and Trimble, W.S. (1999) Characterization of the mammalian septin H5: distinct patterns of cytoskeletal and membrane association from other septin proteins. Cell Motility and the Cytoskeleton, 43, 52–62. Xue, J. et al. (2000) Phosphorylation of a new brain-specific septin, G-septin, by cGMP-dependent protein kinase. The Journal of Biological Chemistry, 275, 10047–56. Yagi, M., Zieger, B., Roth, G.J. and Ware, J. (1998) Structure and expression of the human septin gene HCDCREL-1. Gene, 212, 229–36. Zhang, J. et al. (1999) Phosphatidylinositol polyphosphate binding to the mammalian septin H5 is modulated by GTP. Current Biology, 9, 1458–67. Zieger, B., Hashimoto, Y. and Ware, J. (1997) Alternative expression of platelet glycoprotein Ibβ mRNA from an adjacent 5’ gene with an imperfect polyadenylation signal sequence. The Journal of Clinical Investigation, 99, 520–25.

2 Evolution and conserved domains of the septins Michelle Momany, Fangfang Pan and Russell L. Malmberg Plant Biology Department, University of Georgia, Athens, GA 30602, USA

SEPTIN EVOLUTION Septins were named for their earliest known role in formation of the yeast septum, the cross-wall that partitions mother and daughter cells. In the years since their discovery septins have been characterized in many other fungi and animals, where they are always members of multi-gene families. Most studies of fungal septins have focused on their roles in cell division, shaping emerging protrusions, forming diffusion barriers between compartments and coordinating nuclear division with cell division (Douglas et al., 2005; Gladfelter, 2006). Studies of animal septins have examined roles in cell and nuclear division; however, in contrast to fungal studies, they have also focused on septin roles in vesicle trafficking, apoptosis, cytoskeletal organization (Martinez and Ware, 2004) and, in the case of mammalian septins, links to cancer and neurodegeneration (Hall and Russell, 2004; Kinoshita, 2006; Spiliotis, Kinoshita and Nelson, 2005). In localization studies fungal septins largely form rings shaping new growth or partitions dividing existing cytoplasm (Lindsey and Momany, 2006). Animal septins also localize to division planes, however, they also are frequently seen as filament-like structures colocalizing with actin or microtubules or as punctate spots thought to be involved in vesicle trafficking (Lindsey and Momany, 2006). Fungal septins have been classified by phylogenetic analysis (Momany et al., 2001) and mammalian septins have been classified by primary sequence similarity (Martinez and Ware, 2004). In cross-kingdom phylogenies using sequences from two fungal yeast species and three animal species (Kinoshita, 2003), concluded that orthologous relationships existed within fungal or animal septins, but not The Septins Edited by Peter A. Hall, S.E. Hilary Russell and John R. Pringle © 2008 John Wiley & Sons, Ltd. ISBN: 978-0-470-51969-1

36

CH02 EVOLUTION AND CONSERVED DOMAINS OF THE SEPTINS

between fungal and animal septins. With the recent availability of more genome sequences, it is now clear that certain fungal and animal septins are orthologous (Pan, Malmberg and Momany, 2007). It is hoped that a better understanding of shared evolutionary history might allow lessons learned from septins in simple model organisms to be extended to mammalian septins.

SEPTINS FORM A DISTINCT SUBFAMILY OF P LOOP GTPases One of the defining characteristics of septins is the presence of G1, G3 and G4 core GTPase domains in the central region of the protein (Field and Kellogg, 1999; Leipe et al., 2002) (Figure 2.1). These well-defined domains form β strands that interact with the nucleotide triphosphate (Leipe et al., 2002; Saraste, Sibbald and Wittinghofer, 1990). Though myosins, kinesins and ras proteins contain closely related GTPase domains, in phylogenetic analyses septins clearly form their own clade, distinct from other GTPases (Figure 2.2).

Septins are in many eukaryotes, but not protists or plants Septins are found in fungi, microsporidia and animals (opisthokonts) (Pan, Malmberg and Momany, 2007). The basal eukaryote Giardia lamblia contains a septinlike sequence that is missing half of the core GTPase domain. Similarly, searches of the genomes of 22 other recently sequenced protists including Dictyostelium discoidium and Entamoeba histolytica returned septin-like sequences, but no true septins. In phylogenetic analysis with septins, myosins, kinesins and ras

Figure 2.1 Typical septin structure. Septin sequences range from about 300 to 600 amino acids. Septins contain the conserved GTP CDC binding domain with three motifs: G1, GxxxxGK[ST] (amino acids 126–135 in S. cerevisiae Cdc3); G3, DxxG (amino acids 204–209 in S. cerevisiae Cdc3) and G4, xKxD (amino acids 280–289 in S. cerevisiae Cdc3). The previously described polybasic region (amino acids 111–116 in S. cerevisiae Cdc3; Casamayor and Snyder, 2003; Versele et al., 2004) is shown as a black box and the previously described ‘septin unique element’ (amino acids 360–413 in S. cerevisiae Cdc3; Versele et al., 2004) is shown as a grey box. S1–S4 mark positions of new septin motifs (amino acids 237–242, 247–259, 261–268, 364–365 in S. cerevisiae Cdc3) and lines below diagram show conserved single amino acid positions (amino acids 117, 295, 300, 339, 360, 396 in S. cerevisiae Cdc3). Many septins also have a predicted coiled-coil domain at the C terminus (amino acids 476–507 in S. cerevisiae Cdc3; Versele et al., 2004) Pan, Malmberg and Momany (2007). Reproduced from Pan et al . (2007) Analysis of septins across kingdoms reveals orthology and new motifs. BMC Evolutionary Biology, 7, 103. Copyright  2007, BioMedCentral Ltd

SEPTINS FORM A DISTINCT SUBFAMILY OF P LOOP GTPases

37

CelMyo6

AniCdc42 SceCdc42p CelCdc42 HsaCdc42 HsaRhoF CelRac2 HsaRhoD CelRac1 AniRac1 AniRho2 HsaRhoG HsaRac3 SceRho2 HsaRhoB HsaRac2 HsaRac1 HsaRhoC HsaRhoA CelRho1 AniRhoA SceRho1 SceRho3

CelMyo3 CelMyoII

CelMyo2

HsaMyoVa

RAS

HsaMyoVb AniMyoV

SceSpr3 SceCdc12

SceMyo4p

AniAspC Hsasep4 Hsasep5 Hsasep1

MYOSIN

SceMyo2p

Hsasep2 HsaSept13 CelUnc59 Hsasep6

Hsasep7

Hsasep11 Hsasep8 Hsasep10

SEPTIN CelKrp95

CelUnc61

AniAspB HsaKif3c HsaKif3a

SceCdc3 CelOsm3 Hsasep12

Hsasep3 Hsasep9 AniAspA AniAspD SceCdc10 SceCdc11

HsaKiaa1590

KINESIN

HsaKif13b

SceShs1 AniKif1 SceSpr28 HsaCra HsaKif18A AniAspE

HsaKif4 HsaKif27

AniKip3 SceKip3 HsaKif19 HsaKif2

HsaKif1b

Figure 2.2 Phylogenetic tree of kinesin, myosin, ras and septin GTPases. Bayesian phylogenetic analysis was performed for 74 kinesin, myosin, ras and septin proteins from Aspergillus nidulans, Saccharomyces cerevisiae, Caenorhabditis elegans and Homo sapiens. Analysis was performed using the Mrbayes program run for 1.5 million generations, discarding 400 000 generations as burn-in. Protein designations and GenBank GI numbers follow. From A. nidulans: AniMyoV (67903750), AniRhoA (67539140), AniRho2 (67537566), AniCdc42 (67901000), AniRac1 (67537146), AniKif1 (67541386), AniKip3 (40741666), AniAspA (13398364), AniAspB (1791305), AniAspC (34811845), AniAspD (34148975), AniAspE (34811843). From Caenorhabditis elegans: CelMyo6 (170726), CelMyo2 (37859222), CelMyoII (15718184), CelMyo3 (127737), CelRho1 (17541992), CelRac1 (156424), CelCdc42 (51704309), CelRac2 (115532882), CelKrp95 (9800183), CelOsm3 (22532874), CelUnc59 (17509405), CelUnc61(32566810). From Homo sapiens: HsaMyoVa (119597854), HsaMyoVb (39932736), HsaRhoA (10835049), HsaRhoC (132543), HsaRhoB (4757764), HsaRac1 (249582), HsaRho (74355286), HsaRhoD (20379120), HsaRac3 (118138331), HsaRac2 (20379104), HsaRhoF (13633711), HsaCdc42 (46397381), HsaKif18A (71051935), HsaCra (119571949), HsaKif19 (126215730), HsaKif27 (30794488), HsaKif3a (116283753), HsaKif13b (46852172), HsaKif3c (120660366), HsaKif4 (29351664), HsaKif1b (66347734), HsaKif2 (33187651), HsaKiaa1590 (27529917), HsaSept1 (16604248), HsaSept2 (4758158), HsaSept3 (22035572), HsaSept4 (4758942), HsaSept5 (9945439), HsaSept6 (22035577), HsaSept7 (4502695), HsaSept8 (41147049), HsaSept9 (6683817), HsaSept10 (18088518), HsaSept11 (8922712), HsaSept12 (23242699), HsaSept13, (113418512). From Saccharomyces cerevisiae: SceMyo2 (6324902), SceMyo4 (6319290), SceRho1 (172420), SceRho3 (218474), SceCdc42 (6323259), SceRho2 (1988087), SceKip3 (1723958), SceCdc3 (6323346), SceCdc10 (6319847), SceCdc11 (6322536), SceCdc12 (6321899), SceShs1 (6319976), SceSpr28 (6320424), SceSpr3 (6321496)

38

CH02 EVOLUTION AND CONSERVED DOMAINS OF THE SEPTINS

proteins, the septin-like sequences from protists did not group with other clades (data not shown). In fact, they did not even group with each other suggesting that if these sequences are related they have diverged significantly. Neither septins nor septin-related sequences are found in plants.

Septins form five major groups Phylogenetic analysis of 162 septins using Bayesian and maximum likelihood methods showed that they form five major clades (Figure 2.3) (Pan, Malmberg and Momany, 2007). The Group 1 clade contains fungal and animal septins including Saccharomyces cerevisiae Cdc10 and human SEPT3, SEPT6, SEPT8, SEPT9, SEPT10, SEPT11 and SEPT12. The Group 2 clade contains fungal and animal septins including S. cerevisiae Cdc3 and human SEPT1, SEPT2, SEPT4, SEPT5, SEPT7 and SEPT13. The Group 3 clade contains fungal and microsporidial septins and includes S. cerevisiae Cdc11, Shs1 and Spr28. The Group 4 clade contains fungal and microsporidial septins and includes S. cerevisiae Cdc12 and Spr3. The Group 5 clade contains only fungal septins and includes Aspergillus nidulans AspE.

Septin evolution Based on their relationships with other GTPase families Leipe et al., 2002, suggested that septins likely evolved from a bacterial GTPase that was horizontally transferred from the ancestral pro-mitochondrial endosymbiont to eukaryotes. Based on phylogenetic analysis, the septins might have evolved as follows (Pan, Malmberg and Momany, 2007): Either the GTPase septin ancestor was lost early in the lineage leading to plants and retained in the lineage leading to animals and fungi (Figure 2.4), or the horizontal transfer occurred after the divergence of plants. The ancestral septin was likely duplicated before the animal/fungal split and the resulting duplicates became the founding members of Group 1 and Group 2 septin clades. The ancient Group 1 septin duplicated. One paralog lost its C-terminal extension giving rise to Group 1A septins in animals and fungi. The other paralog retained its C-terminal extension giving rise to Group 1B septins in animals, but was lost from fungi. In animals Group 1 septins have undergone significant duplication giving rise to multiple Group 1 paralogs. In contrast, there is a single Group 1 septin in all fungi so far examined. The ancient Group 2 septin gave rise to Group 2A septins in fungi and Group 2B septins in animals. In animals Group 2 septins have undergone significant duplication giving rise to multiple paralogs while a single Group 2 septin is found in all fungi so far examined. In the lineage leading to fungi and microsporidia, Group 1 and Group 2 septins duplicated giving rise to Group 3 and Group 4 septins. Subsequently, Group 1 and Group 2 septins were lost in the lineage leading to microsporidia. Group 3 and Group 4 septins duplicated and diverged, giving rise to multiple fungal paralogs. Because Group 5 septins are found only

SEPTINS FORM A DISTINCT SUBFAMILY OF P LOOP GTPases

39

Saccharomyces cerevisiae Myo Saccharomyces cerevisiae Cdc10 Candida glabrata Hyp3 Eremothecium gossypii Hyp3 0.96/100 0.99/97 Kluyveromyces lactis Hyp3 Candida albicans Cdc10 1.00/100 0.96/100 Debaryomyces hensenii Hyp3 Yarrowia lipolytica Hyp2 Schizosaccharomyces pombe Spn2 1.00/83 Coccidioides immitis Sep3 1.00/59 Aspergillus nidulans AspD 1.00/100 Gibberella zeae Hyp4 0.76/42 Neurospora crassa Hyp3 1.00/84 0.98/82 Magnaporthe grisea Hyp3 0.96/100 Cryptococcus neoformans Hyp3 1.00/90 Ustilago maydis Cdc10 Mus musculus Sept3 1.00/100 Rattus norvegicus Sept3 0.63 1.00/98 Homo sapiens Sept3 Danio rerio Msf 1.00/98 Homo sapiens Sept9 1.00/100 1.00/100 Mus musculus Sept9 1.00/93 Rattus norvegicus Sept9 Xenopus laevis Hyp1 Homo sapiens Sept12 1.00/100 0.55/85 Mus musculus Sept12 0.80/38 1.00/99 Rattus norvegicus Sept12 Drosophila melanogaster Sep5 0.99/92 Drosophila melanogaster Sep2 Anopheles gambiae Hyp2 1.00/78 1.00/100 Mus musculus Sept10a 1.00 /100 Rattus norvegicus Sept10a 1.00/100 Homo sapiens Sept10 Mus musculus Sept10b 1.00/100 0.57/79 Rattus norvegicus Sept10b 1.00/100 Danio rerio Hyp2 1.00/99 Homo sapiens Sept8 1.00/100 Mus musculus Sept8 0.99/89 Rattus norvegicus Sept8 1.00/92 1.00/100 Danio rerio Sept6 1.00/100 Homo sapiens Sept6 Mus musculus Sept6 1.00/100 1.00/97 Rattus norvegicus Sept6 1.00/93 Homo sapiens Sept11 Rattus norvegicus Sept11 1.00/100 0.74/69 1.00/91 Mus musculus Sept11 1.00/100 Caenorhabditis briggsae Hyp2 Caenorhabditis elegans Unc61 Schizosaccharomyces pombe Spn1 Yarrowia lipolytica Hyp1 Saccharomyces cerevisiae Cdc3 1.00/98 1.00/100 0.79/100 Candida glabrata Hyp1 1.00 Eremothecium gossypii Hyp1 1.00/100 0.83 Kluyveromyces lactis Hyp1 1.00/100 Candida albicans Cdc3 Debaryomyces hansenii Hyp1 1.00/60 Aspergillus nidulans AspB 1.00/100 Coccidioides immitis Sep2 Gibberella zeae Hyp1 0.68/54 Neurospora crassa Hyp1 1.00/99 1.00/52 1.00/97 Magnaporthe grisea Hyp1 Ustilago maydis Hyp1 1.00/96 Cryptococcus neoformans Hyp1 Drysophila melanogaster Pnut 1.00/100 Anopheles gambiae Hyp1 1.00/95 0.78/78 Homo sapiens Sept13 Danio rerio Hyp1 1.00/94 1.00/89 Mus musculus Sept7 0.68/71 Rattus norvegicus Sept7 0.97/70 Bos taurus Cdc10 0.85/95 Homo sapiens Sept7 Drysophila melanogaster Sep1 1.00/100 Anopheles gambiae Hyp3 1.00/100 Xenopus laevis SeptA 1.00/100 Danio rerio Nedd5 0.50/69 Rattus norvegicus Sept2 1.00/95 Mus musculus Sept2 1.00/97 0.65/72 0.99/68 Homo sapiens Sept2 1.00/97 1.00/100 Drysophila melanogaster Sep4 Anopheles gambiae Hyp4 Homo sapiens Sept4 1.00/100 Mus musculus Sept4 1.00/100 1.00/98 0.82/70 Rattus norvegicus Sept4 Danio rerio Hyp4 1.00/99 Homo sapiens Sept5 1.00/100 Macaca fascicularis Hyp1 1.00/99 0.99/58 Rattus norvegicus Sept5 1.00/98 1.00/97 Mus musculus Sept5 1.00/100 Homo sapiens Sept1 Mus musculus Sept1 0.81/38 1.00/100 Rattus norvegicus Sept1 Geodia cydonium Sep1 1.00/100 Suberites domuncula Septl Caenorhabditis briggsae Hyp1 1.00/100 Caenorhabditis briggsae Hyp3 0.77/76 Caenorhabditis elegans Unc59 0.97/98

0.97/100

Group 1A

(CDC10 + M_I) Animals + Fungi

Group 1B (M_II) Animals

Group 2A (CDC3) Fungi

Group 2B (M_III & IV) Animals

Figure 2.3 Overview phylogenetic tree of septin gene family. Half-compat consensus phylogram of 1.5 million generations of the MCMC Montecarlo Markov Chain analysis of the Bayesian phylogenetic analysis, discarding 400 000 generations as burn-in. Nodal numbers in front of the slash are posterior probabilities for Bayesian analysis. At the nodes where the maximum likelihood tree topology agrees with the Bayesian analysis, numbers after the slash are bootstrap percentages from maximum likelihood bootstrap analysis using 1024 replicates. Names in parenthesis under Group names indicate the best characterized fungal septin (CDC10, CDC3, CDC11 and CDC12, ASPE) or the mammalian septin classification of Martinez and Ware (2004) (MI, MII and MIII). Pan, Malmberg and Momany (2007). Reproduced from Pan et al . (2007) Analysis of septins across kingdoms reveals orthology and new motifs. BMC Evolutionary Biology, 7, 103. Copyright  2007, BioMedCentral Ltd

40

CH02 EVOLUTION AND CONSERVED DOMAINS OF THE SEPTINS Schizosaccharomyces pombe Spn5 Yarrowia lipolytica Hyp5 Schizosaccharomyces pombe Spn7 0.98/85 Saccharomyces cerevisiae Cdc11 1.00/100 Candida glabrata Hyp5 Eremothecium gossypii Hyp4 1.00/94 0.92 Kluyveromyces lactis Hyp5 0.85/67 1.00/100 Candida albicans Cdc11 0.66/36 Debaryomyces hansenii Hyp6 Encephalitozoon cuniculi Sep3 0.55/51 Kluyveromyces lactis Hyp6 0.52/47 1.00/100 Eremothecium gossypii Hyp7 Saccharomyces cerevisiae Shs1 0.77/56 1.00/99 0.52 1.00/78 Candida glabrata Hyp6 1.00/100 Candida albicans Sep7 Debaryomyces hansenii Hyp5 Yarrowia lipolytica Hyp6 Yarrowia lipolytica Hyp4 0.85/95 0.84/67 0.97/95 Coccidioides immitis Sep1 Aspergillus nidulans AspA 0.52 0.94/97 Pyrenopeziza brassicae Pbs1 Gibberella zeae Hyp3 0.71/74 Magnaporthe grisea Hyp4 0.63/72 0.85/86 0.61/74 Neurospora crassa Hyp4 Schizosaccharomyces pombe Spn3 Ustilago maydis Hyp3 0.63/48 1.00/95 Cryptococcus neoformans Hyp4 Candida albicans Spr28 1.00/91 Saccharomyces cerevisiae Spr28 1.00/100 1.00/96 Candida glabrata Hyp8 0.95/63 Eremothecium gossypii Hyp6 Kluyveromyces lactis Hyp7 0.53 Eremothecium gossypii Hyp2 Kluyveromyces lactis Hyp2 0.90/100 Candida albicans Spr28 0.88/100 Saccharomyces cerevisiae Cdc12 0.90/90 0.57 Candida glabrata Hyp4 0.99/100 Candida albicans Cdc12 0.81/99 Debaryomyces hansenii Hyp2 Yarrowia lipolytica Hyp3 1.00/100 Aspergillus nidulans AspC Neurospora crassa Hyp2 0.56/80 Magnaporthe grisea Hyp2 0.98/70 1.00/90 Gibberella zeae Hyp2 0.54/66 0.86/76 Mucor circinelloides SepA Ustilago maydis Hyp2 0.76/91 1.00/98 Cryptococcus neoformans Hyp2 0.55/69 Schizosaccharomyces pombe Spn4 0.56/97 Schizosaccharomyces pombe Spn6 Yarrowia lipolytica Hyp7 0.99/68 Candida glabrata Hyp7 0.99/87 Debaryomyces hansenii Hyp4 Kluyveromyces lactis Hyp4 1.00/96 0.65/41 1.00/57 Eremothecium gossypii Hyp5 0.90/52 Saccharomyces cerevisiae Spr3 Encephalitozoon cuniculi Sep1 0.98/85 Candida albicans Spr3 Encephalitozoon cuniculi Sep2 Gibberella zeae Hyp6 1.00/100 Neurospora crassa Hyp6 1.00/99 0.55/49 Magnaporthe grisea Hyp6 Cryptococcus neoformans Hyp5 0.67/92 Neurospora crassa Hyp5 0.89/98 Gibberella zeae Hyp5 1.00/100 0.90 Magnaporthe grisea Hyp5 0.81/76 Coccidioides immitis Sep4 Aspergillus nidulans AspE

Figure 2.3

1 2

+5

+3 +4 −1 −2

Group 3 (CDC11)

Fungi + Microsporidia

Group 4 (CDC12)

Fungi + Microsporidia

Group 5 (ASPE)

Fungi

(continued)

Plants

None

Animals

1A, 1B (M_I & II) 2B (M_III & IV)

Fungi

1A (CDC10) 2A (CDC3) 3 (CDC11) 4 (CDC12) 5 (ASPE)

3 (CDC11) Microsporidia 4 (CDC12)

Figure 2.4 Postulated septin evolution. Group 1 and Group 2 septins likely arose before the animal/fungal split. Group 3 and 4 septins likely arose before the fungal/microsporidial split and Group 5 septins likely arose in fungi after the fungal/microsporidial split. Numbers refer to septin Group and + or − represents their gain or loss. Names in parenthesis under Group names indicate the best characterized fungal septin (CDC10, CDC3, CDC11 and CDC12, ASPE) or the mammalian septin classification of Martinez and Ware (2004) (MI, MII and MIII)

SEPTIN DOMAINS

41

in filamentous fungi, they either arose early in fungal evolution and were lost in yeasts or arose more recently.

SEPTIN DOMAINS Motifs with known function Three septin domains have known functions associated with them, though the relevance of the domains for septin assembly, localization and function in vivo is not always clear. The GTPase domain mediates the binding and/or hydrolysis of GTP. The polybasic region mediates association of the septins with phospholipids found in plasma membranes. In some, but not all cases, the C-terminal coiled-coil domains mediate septin–septin interactions.

GTPase domains The G1 core GTPase domain (GxxxxGK[ST]) forms a flexible loop which interacts with the phosphate group of the nucleotide (Bourne, Sanders and McCormick, 1991; Leipe et al., 2002; Saraste, Sibbald and Wittinghofer, 1990). It is the most conserved motif among septins with the consensus conserved in 98 % of 162 septins examined (Pan, Malmberg and Momany, 2007). The residues immediately following the G1 core GTPase domain are also conserved in the septins with [TS][LF] appearing in 96–97 % of septins. The G3 core GTPase domain (hydrophobic residues followed by DxxG) binds Mg2+ and interacts with the β and γ phosphates of GTP (Bourne, Sanders and McCormick, 1991; Leipe et al., 2002; Saraste, Sibbald and Wittinghofer, 1990; Vetter and Wittinghofer, 1999). The G3 consensus for septins (DT[PV]GxG) is found in over 80 % of 162 septins examined (Pan, Malmberg and Momany, 2007). The G4 core GTPase domain (NKxD) determines GTP binding specificity (Dever, Glynias and Merrick, 1987; Leipe et al., 2002). In septins the first position in the consensus (N) is often replaced by A, S or G. However positions two and four of the G4 consensus (K and D, respectively) are found in over 90 % of 162 septins examined (Pan, Malmberg and Momany, 2007).

Polybasic region Septins associate with the plasma membrane and have been shown to bind the membrane phospholipids phosphatidylinositol 4, 5-bisphosphate and phosphatidylinositol 3, 4, 5-triphosphate (Zhang et al., 1999). Deletion and mutation analysis in mammalian cells and later in yeast identified a six amino acid region N terminal to the G1 core GTPase domain that is responsible for the binding of these membrane phospholipids (Casamayor and Snyder, 2003; Zhang et al., 1999) (Figure 2.1). Though individual residues are not conserved, a basic residue (H,

42

CH02 EVOLUTION AND CONSERVED DOMAINS OF THE SEPTINS LPNPRHRKSVKKG*FFQFNLM VVG+

Figure 2.5 Septin polybasic region consensus. Most common amino acid in polybasic region and surrounding area (corresponds to positions 105–126 in S. cerevisiae Cdc3) among 162 septins examined. Bold designates basic residue (H, K or R) found in 51–90 % of septins examined. Underline designates polybasic region identified by Zhang et al. (1999). Asterisk designates highly conserved glycines found in 90 % of septins examined. Plus sign indicates initial glycine in G1 GTPase core domain

K or R) is present in polybasic region positions 1, 2, 5 and 6 of 60–78 % of the 162 septins examined (Pan, Malmberg and Momany, 2007) (Figure 2.5). Glycine immediately follows the polybasic region in 90 % of the 162 septins examined.

Coiled-coil The coiled-coil motif mediates interactions between proteins. This structural motif forms a super helix with heptad repeats and is found in many proteins (Lupas, 1996; Mason and Arndt, 2004). Septins in Group 1A are truncated at the C terminus and so do not contain the region with the coiled-coil motif. In computational structural predictions all Group 1B, Group 2 and Group 4 septins were strongly predicted to form C-terminal coiled-coils (Pan, Malmberg and Momany, 2007). In contrast, 5 of 29 Group 3 septins and all Group 5 septins were not predicted to form C-terminal coiled-coils. Interestingly, previous work showed that the coiled-coil motifs at the C termini of the S. cerevisiae Group 2 septin Cdc3 and the Group 4 septin Cdc12 were absolutely required for septin association and function, while the C-terminal coiled-coil in Group 3 septin Cdc11 was dispensable (An et al., 2004; Casamayor and Snyder, 2003; Versele and Thorner, 2005). The absolute conservation of coiled-coils in all Group 2 and Group 4 septins suggests that interactions mediated by the C-terminal coiled-coils of these groups might also be essential for function across species.

Motifs with unknown function A number of conserved regions with no known function have been identified in septins. Though the functional significance of these motifs is not known, their conservation suggests important roles.

Septin unique element Versele and Thorner (2005) previously identified a conserved 53 amino acid region C-terminal to the G4 core GTPase domain and designated it the septin unique element or SUE (Versele et al., 2004; Versele and Thorner, 2005) (Figure 2.1). Analysis of the SUE in 162 septins showed that just over half of the residues are conserved across 50–93 % of septins (Figure 2.6). When the consensus for

CONCLUSIONS

43

R*XYPW*G*XXEV ENXXHCDFXX LRXXLIRTHX XDLXXXT*XXX HYEXYRXXXL XXX

Figure 2.6 Septin unique element consensus. Most common amino acids in septin unique element identified by Versele and Thorner (2005) (corresponds to positions 360–413 in S. cerevisiae Cdc3). Residues found in at least 50 % of 162 septins examined are shown. Bold designates residues conserved in at least 75 % of septins examined. Asterisks designate single R and T residues conserved in 93 % of septins examined and WG Sep4 motif found in 92 % of all septins examined (Pan, Malmberg and Momany, 2007). Underline designates regions which returned only septins in GenBank searches as described in text

the entire 53 residue SUE was used to query GenBank, no hits were returned. However, querying with SUE residues 5–11 as bait returned 32 hits, all of them septins, and SUE residues 37–46 returned 41 hits, all of them septins. So while there is variation across species, at least potions of the SUE appear to be unique to septins.

Sep1–4 and conserved single residues Alignments of 162 septins revealed four new highly conserved motifs (Sep1–4) and six highly conserved amino acid positions (Pan, Malmberg and Momany, 2007) (Figure 2.1). Each of the two consensus amino acids in the Sep1 motif (ExxxxR, located between the core G3 and G4 GTPase domains) were conserved in 96–98 % of septins examined. Each consensus amino acid in the Sep2 motif (DxR[VI]Hxxx[YF]F[IL]xP, located between G3 and G4 core GTPase domains) was conserved in 88–96% of septins examined. Each consensus amino acid in the Sep3 motif (GxxLxxxD, between the G3 and G4 core GTPase domains) was conserved in 86–96% of septins examined. Each consensus amino acid in the Sep4 motif (WG within the SUE) was conserved in 92 % of septins examined. When used as queries in searches of GenBank, the Sep1, Sep3 and Sep 4 motifs returned proteins in addition to septins. However the Sep2 motif returned 54 proteins, all of them septins. One of the highly conserved single amino acid positions was a glycine immediately following the PB region (Figure 2.5) and two others were within the SUE (Figure 2.6). The remaining highly conserved residues were between the G4 GTPase core domain and the SUE (Figure 2.1).

CONCLUSIONS Our better understanding of the relationships among septins has raised new questions. Perhaps the most obvious question is: What is the significance of the highly conserved SUE and Sep1–4 motifs? This question could be addressed, at least in part, through site-directed mutagenesis targeting these regions. A more difficult question to answer is: Does the shared evolutionary history of Group 1 and Group 2 septins from mammals and fungi translate into shared function or

44

CH02 EVOLUTION AND CONSERVED DOMAINS OF THE SEPTINS

structure? The answer to this question will likely only emerge over time as more researchers investigate septins in different species.

REFERENCES An, H., Morrell, J.L., Jennings, J.L. et al. (2004) Requirements of fission yeast septins for complex formation, localization, and function. Molecular Biology of the Cell , 15, 5551–64. Bourne, H.R., Sanders, D.A. and McCormick, F. (1991) The GTPase superfamily: conserved structure and molecular mechanism. Nature, 349, 117–27. Casamayor, A. and Snyder, M. (2003) Molecular dissection of a yeast septin: distinct domains are required for septin interaction, localization, and function. Molecular and Cellular Biology, 23, 2762–77. Dever, T.E., Glynias, M.J. and Merrick, W.C. (1987) GTP-binding domain: three consensus sequence elements with distinct spacing. Proceedings of the National Academy of Sciences of the United States of America, 84, 1814–18. Douglas, L.M., Alvarez, F.J., McCreary, C. and Konopka, J.B. (2005) Septin function in yeast model systems and pathogenic fungi. Eukaryotic Cell , 4, 1503–12. Field, C.M. and Kellogg, D. (1999) Septins: cytoskeletal polymers or signalling GTPases. Trends in Cell Biology, 9, 387–94. Gladfelter, A.S. (2006) Control of filamentous fungal cell shape by septins and formins. Nature Reviews Microbiology, 4, 223–29. Hall, P.A. and Russell, S.E. (2004) The pathobiology of the septin gene family. The Journal of Pathology, 204, 489–505. Kinoshita, M. (2003) The septins. Genome Biology, 4, 236. Kinoshita, M. (2006) Diversity of septin scaffolds. Current Opinion in Cell Biology, 18, 54–60. Leipe, D.D., Wolf, Y.I., Koonin, E.V. and Aravind, L. (2002) Classification and evolution of P-loop GTPases and related ATPases. Journal of Molecular Biology, 317, 41–72. Lindsey, R. and Momany, M. (2006) Septin localization across kingdoms: three themes with variations. Current Opinion in Microbiology, 9, 559–65. Lupas, A. (1996) Coiled coils: new structures and new functions. Trends in Biochemical Sciences, 21, 375–82. Martinez, C. and Ware, J. (2004) Mammalian septin function in hemostasis and beyond. Experimental Biology and Medicine (Maywood), 229, 1111–19. Mason, J.M. and Arndt, K.M. (2004) Coiled coil domains: stability, specificity, and biological implications. ChemBioChem, 5, 170–76. Momany, M., Zhao, J., Lindsey, R. and Westfall, P.J. (2001) Characterization of the Aspergillus nidulans septin (asp) gene family. Genetics, 157, 969–77. Pan, F., Malmberg, R.L. and Momany, M. (2007) Analysis of septins across kingdoms reveals orthology and new motifs. BMC Evolutionary Biology, 7, 103. Saraste, M., Sibbald, P.R. and Wittinghofer, A. (1990) The P-loop–a common motif in ATP- and GTP-binding proteins. Trends in Biochemical Sciences, 15, 430–34. Spiliotis, E.T., Kinoshita, M. and Nelson, W.J. (2005) A mitotic septin scaffold required for mammalian chromosome congression and segregation. Science, 307, 1781–85.

REFERENCES

45

Versele, M., Gullbrand, B., Shulewitz, M.J. et al. (2004) Protein-protein interactions governing septin heteropentamer assembly and septin filament organization in Saccharomyces cerevisiae. Molecular Biology of the Cell , 15, 4568–83. Versele, M. and Thorner, J. (2005) Some assembly required: yeast septins provide the instruction manual. Trends in Cell Biology, 15, 414–24. Vetter, I.R. and Wittinghofer, A. (1999) Nucleoside triphosphate-binding proteins: different scaffolds to achieve phosphoryl transfer. Quarterly Reviews of Biophysics, 32, 1–56. Zhang, J., Kong, C., Xie, H. et al. (1999) Phosphatidylinositol polyphosphate binding to the mammalian septin H5 is modulated by GTP. Current Biology, 9, 1458–67.

Section II Septins in model systems

The Septins Edited by Peter A. Hall, S.E. Hilary Russell and John R. Pringle © 2008 John Wiley & Sons, Ltd. ISBN: 978-0-470-51969-1

3 Biochemical properties and supramolecular architecture of septin hetero-oligomers and septin filaments Michael A. McMurray and Jeremy Thorner Division of Biochemistry and Molecular Biology, Department of Molecular and Cell Biology, University of California, Berkeley, CA 94720-3202 USA

IMPETUS AND SCOPE Septins are ‘filament-forming GTPases’, but until very recently little was known about how the organization of septin-containing filaments relates to the structure of each of its constituent subunits. How or even whether guanine nucleotide hydrolysis is involved in assembly of higher-order septin structures has also remained an open question. Now, however, more than 30 years of research have reached a watershed with elucidation of the arrangement of septin monomers within hetero-oligomeric complexes from several different species and of how these complexes are polymerized within filaments. These insights have come from intensive biochemical and ultrastructural analysis, especially high-resolution electron microscopy (EM) and X-ray crystallography. Significant progress has also been made in understanding the roles of particular structural elements within an individual septin, and of nucleotide binding and hydrolysis, in the formation of multi-septin complexes and filaments. Our aim here is to interpret the available data so as to present a unified and coherent picture of the biochemical and ultrastructural features of septin complexes and filaments, and to identify key unresolved issues. The Septins Edited by Peter A. Hall, S.E. Hilary Russell and John R. Pringle © 2008 John Wiley & Sons, Ltd. ISBN: 978-0-470-51969-1

50

CH03 BIOCHEMICAL PROPERTIES AND SUPRAMOLECULAR ARCHITECTURE

ARRANGEMENT OF SEPTIN SUBUNITS WITHIN COMPLEXES AND FILAMENTS FROM UNICELLULAR EUKARYOTES Core septins of budding yeast (Saccharomyces cerevisiae) Investigation of septin ultrastructure began with the observation of an array of ordered filaments at the mother-bud neck in S. cerevisiae cells (Byers and Goetsch, 1976a). These 10-nm-diameter filaments appeared to be aligned in parallel at ∼28-nm intervals and orthogonal to the mother-bud axis (Byers and Goetsch, 1976a). Of the temperature-sensitive cell division cycle (cdc) mutants isolated by Hartwell et al. (1974) these neck filaments were completely absent in cells carrying mutations in the CDC3 , CDC10 , CDC11 or CDC12 genes (Byers and Goetsch, 1976b). Antibodies against Cdc3, Cdc10, Cdc11 and Cdc12 (Haarer and Pringle, 1987), and later GFP fusions (Cid et al., 1998), demonstrated that all four proteins are located at the bud neck. Moreover, cdc3ts , cdc10ts , cdc11ts and cdc12ts mutants fail to undergo septation (cell division) at the restrictive temperature (Hartwell, 1971). Three related gene products – Shs1/Sep7, which is expressed in mitotically growing cells, and Spr3 and Spr28, which are expressed only in cells undergoing meiosis and sporulation – were identified via their sequence homology to the core septins (De Virgilio, DeMarini and Pringle, 1996; Fares, Goetsch and Pringle, 1996; Carroll et al., 1998; Mino et al., 1998). However, none of these is required for localization of Cdc3, Cdc10, Cdc11 and Cdc12 to the bud neck of a vegetatively growing cell or, presumably, for assembly of Cdc3, Cdc10, Cdc11 and Cdc12 into filaments (although, to our knowledge, the ultrastructure of the bud necks of shs1
, spr3
or spr28
cells at the EM level has not been reported). High-stringency biochemical fractionation of yeast cell extracts typically yields a septin-containing complex composed of roughly equal amounts of Cdc3, Cdc10, Cdc11 and Cdc12, having an apparent molecular mass in high salt consistent with a 2 : 2 : 2 : 2 stoichiometry (Frazier et al., 1998; Mortensen et al., 2002). Under conditions of low ionic strength, these purified native septin complexes polymerize into filaments of somewhat smaller diameter (∼5–7 nm) than the neck filaments (Frazier et al., 1998). Likewise, when co-expressed and purified from bacterial (Escherichia coli ) cells, Cdc3, Cdc10, Cdc11 and Cdc12 are necessary and sufficient to form complexes similar in composition and stoichiometry to the native complexes and that polymerize in low salt into filaments that have virtually identical characteristics to those generated by the purified native septin complexes (Versele et al., 2004; Versele and Thorner, 2004; Farkasovsky et al., 2005). Thus, the 2 : 2 : 2 : 2 complex formed by Cdc3, Cdc10, Cdc11 and Cdc12 is the essential building block of yeast septin filaments. Of course, in the cell, other proteins known to co-localize with the septins at the isthmus between a mother and its bud (Gladfelter, Pringle and Lew, 2001) may be responsible for the somewhat

SEPTIN COMPLEXES IN UNICELLULAR EUKARYOTES

51

greater thickness of the endogenous neck filaments and for dictating their spatial organization.

The yeast septin complex is a linear rod Ultrastructural analysis of the septins has focused on examining the features of purified and recombinant hetero-octameric complexes. In conditions of high ionic strength (e.g. 1 M KCl), septin complexes containing Cdc3, Cdc10, Cdc11 and Cdc12 purified from yeast cells (Frazier et al., 1998), or expressed in and purified from E. coli cells (Sirajuddin et al., 2007; Bertin et al., 2008), appear by negative-stain EM as beaded 32-nm-long rods (Figure 3.1a, top). Decreasing the ionic strength of the buffer (e.g. 50 mM KCl) promotes formation of very long filaments of the same diameter as that of the rod (Frazier et al., 1998; Bertin et al., 2008). Depositing septin complexes onto EM grids immediately after lowering the salt concentration yields apparent intermediates that are twice the length of the rod, or three or four times (Figure 3.1a, middle), consistent with the conclusion that the filaments arise from ionic strength-dependent end-to-end polymerization of the rods. Formation of the filaments seems for several reasons to be a highly cooperative process. First, unlike the basic rod itself, the double- and triple-length intermediates are transient and relatively rare. Second, once nucleated, the filaments that form are very long and invariably align in pairs (separated by a gap of 15–25 nm). Moreover, these paired filaments appear to be in register because the ends of each filament in the pair are flush with each other, not staggered (Figure 3.1a, bottom). Under some conditions, still not well defined (but definitely influenced by pH), yeast filament doublets tend to collapse into very large bundles, which are likely non-physiological aggregates. This behaviour may reflect aggregation that occurs upon partial pH-dependent subunit unfolding because an intermediate in the thermal unfolding of a human septin (SEPT4) forms similar amyloid-like fibrils in vitro (Garcia et al., 2007). This property of septins may have implications for human pathophysiology because septins have been found as components of the characteristic neurofibrillary tangles in the brains of patients suffering from Alzheimer’s disease (Kinoshita et al., 1998). In the brain, septins appear to compartmentalize the dendritic spines on neurons, similar to the way they demarcate the boundary between a mother cell and its bud in yeast (Barral and Mansuy, 2007). Averaging the images of very large numbers of classes of individual rods (Sirajuddin et al., 2007; Bertin et al., 2008) affords enhanced resolution that reveals that each rod is composed of eight globular densities of roughly equivalent size and shape, each about 4–5 nm in diameter (Figure 3.1b). These features and dimensions correspond well with those for small GTP-binding proteins (Stewart, Kent and McCoy, 1998), suggesting that the globular density observed reflects, in large part, the conserved GTP-binding domain common to Cdc3, Cdc10, Cdc11 and Cdc12. Moreover, the calculated molecular mass for eight protein spheres

CH03 BIOCHEMICAL PROPERTIES AND SUPRAMOLECULAR ARCHITECTURE

52

of these measured dimensions agrees well with that determined for native and recombinant septin complexes composed of these same four septins in which the apparent stoichiometry is 2 : 2 : 2 : 2. Thus, the appearance of eight globular densities in the 32-nm-long rod viewed by EM is fully compatible with the conclusion that each rod contains two molecules of each of the four different septin polypeptides.

32 nm

4 nm 4 nm

100 nm

(b)

100 nm

Cdc11

Spn3

Cdc12

Spn4

Cdc3

Spn1

Cdc10

Spn2

Cdc10

Spn2

Cdc3

Spn1

Cdc12

Spn4

Cdc11

Spn3

S. cerevisiae

S. pombe

Pnut

SEPT7

Sep2

SEPT6

UNC-59

Sep1

SEPT2

UNC-61

Sep1

SEPT2

UNC-61

Sep2

SEPT6

UNC-59

Pnut

SEPT7

C. elegans

D. melanogaster

100 nm

(a)

(c)

H. sapiens, R. norveticus, M. musculus

SEPTIN COMPLEXES IN UNICELLULAR EUKARYOTES

53

Arrangement of subunits within the octameric rod To verify the above conclusions and to pinpoint which septin occupies which position within the rod, various strategies can be applied. Use of the two-hybrid method (Fields and Sternglanz, 1994) to examine the hierarchy of interactions that a given yeast septin may have with the others is fraught with potential problems because septin complexes normally assemble in the cytosol (not in the nucleus) and because apparent associations in vivo might arise indirectly from ‘bridging’ through other members of the complex (Farkasovsky et al., 2005). Likewise, even when individual, bacterially expressed and purified yeast septins were examined via in vitro binding experiments, the results were, in some cases, misleading because apparent self-association of a given septin can occur in the absence of its preferred partner septin and the ability of a given septin to associate with itself or another is influenced by the presence of a third (Versele et al., 2004). Some of the non-physiological one-on-one associations observed might be attributable to the fact that the C termini of Cdc3, Cdc11 and Cdc12 each contain segments with 4-3 (heptad) repeats of hydrophobic residues (Figure 3.2) predicted to have the propensity to form a coiled coil with a partner heptad repeat protein (DeMarini et al., 1997; Versele et al., 2004). An analogy is c-Jun, which by itself readily forms homodimers, but forms only c-Fos-c-Jun heterodimers when both proteins are present (O’Shea et al., 1989). In this context, spurious homo-dimerization of septins with the capacity to form a coiled coil is perhaps not surprising. Such

Figure 3.1 EM images of negatively stained core septin complexes from budding yeast and models of the arrangement of subunits in septin complexes of various species. (a) Appropriate vectors were used to co-express Saccharomyces cerevisiae Cdc3, Cdc10, Cdc11 and (His)6 -Cdc12 in Escherichia coli and the resulting heteromeric multi-septin complex was purified in high-salt buffers to near-homogeneity in three chromatographic steps (immobilized metal ion affinity, size exclusion and ion exchange) (Bertin et al., 2008). The resulting complex contained stoichiometric amounts of all four septins in a particle of apparent molecular mass most compatible with a 2 : 2 : 2 : 2 hetero-octamer, as observed previously (Versele et al., 2004; Versele and Thorner, 2004). Samples of the solution of the purified complex were spread under the different conditions indicated on carbon-coated EM grids, negatively stained with uranyl acetate, and viewed in the EM. Top, 300 mM KCl; Middle, diluted from 300 to 50 mM KCl and then immediately prepared for examination in the EM; Bottom, diluted from 300 to 50 mM KCl, incubated for 2 h, and then prepared for viewing in the EM. (b) Individual particles (‘rods’), as in (a, Top), were classified by their appearance and computationally averaged. Image shown is the class average for 1986 particles of the most common class observed, with the dimensions indicated. (c) Models of the number, identity and order of subunits within the rod-like septin complexes from the indicated species. For S. cerevisiae, C. elegans and H. sapiens, the number, identity and order of subunits shown are based on direct experimental evidence generated as described in detail in the text. For other species, the order of subunits and the sizes of the rods are based on low-resolution EM analysis of purified complexes, subunit–subunit interaction data, sequence homology and/or phylogenetic relationships with respect to the corresponding budding yeast and/or human septins, as described in further detail in the text

54

CH03 BIOCHEMICAL PROPERTIES AND SUPRAMOLECULAR ARCHITECTURE +

+

+

+

+

+

+

Cdc3_CTE Cdc12_CTE Cdc12-6_CTE

LGIKQDNSVFKEFDPISKQLEEKTLHEAKLAKLEIEMKTVFQQKVSEKEKKLQKSETELFARHKEMKEKLTKQLKALEDKK ENTGEGNEDFT--LPAIAP-ARKLSHNPRYKEEENALKKYFTDQVKAEEQRFRQW----EQNIVNERIRLNGDLEEIQGKV ENTGEGNEDFT--LPAIAP-ARKLSHNPRYKEEENALKKYFTDQVKAEEQRFRQW----EQNIVNERIRLNGDLEEIQGKV

Cdc3_CTE Cdc12_CTE Cdc12-6_CTE

KQLELSINSASPNVNHSPVPTKKKGFLR. KKLEEQVKSLQVKKSHLK. KKTRRAGQKLASKKIPFKMMIN.

Cdc11_CTE Shs1_CTE

SVAAESIRPNLTKLNGSSSSSTTTRRNTNPFKQSNNINNDVLNPASDMHGQSTGENNETYMTREEQIRLEEERLKAF VANAEEIGPNSTKRQSNAPSLSNFASLISTGQFNSSQTLANNLRADTPRNQVSGNFKENEY(47)PDLPERTKLRNI

Cdc11_CTE Shs1_CTE

EERVQQELLLKRQELLQREKELREIEARLEKEAKIKQEE. SETV--PYVLRHERILARQQKLEELEAQSAKELQKRIQELERKAHELKLREKLINQNKLNGSSSSINSLQQ(23).

+

+

+

+ +

+

−

+ +

+

+

+ +

+

+ +

+

+

+

+ +

+

+

+

+

+

+

+

+ +

+ +

+ +

+ +

+ +

+ +

+ +

Figure 3.2 The C-terminal extensions (CTEs) of four mitotically expressed septins from budding yeast. As in Versele et al., 2004, the CTE of a septin was defined as that segment of the primary structure from the residue immediately following the ‘septin unique element or domain’ (SUE or SUD) to the C terminus. Gratifyingly, the C-side boundary of the SUE corresponds to the end of α6 helix (see Plate 3.3c, after p. 246), the last well-ordered structural element found in the crystal structures of all three human septins analysed by X-ray diffraction (Sirajuddin et al., 2007). As judged by sequence conservation, the α6 element is present in every septin known in every species that has been analysed to date (Pan, Malmberg and Momany, 2007). Shown, in the one-letter code, are the CTEs of four of the five yeast mitotic septins (Cdc10 lacks a CTE; see Plate 3.3c). Cdc3 and Cdc12 are grouped together, and separately from Cdc11 and Shs1, because the members of each of these pairs are somewhat more related to each other than they are to the other pair. Underlined residues are predicted by a variety of computational algorithms to have strong α-helix forming propensity. Positions marked above or below with a ‘+’ indicate hydrophobic residues (and other compatible side chains) that fit the 4-3 heptad repeat characteristic of sequences known to form α-helical coiled coils. Bold-face type indicates the amino acid substitutions caused by the frameshift mutation in the cdc12-6 allele; and, ‘–’ highlights the absence of a hydrophobic residue at the indicated position. Numbers in parentheses, ‘(47)’ and ‘(23)’, represent stretches of residues of the indicated length present in Shs1, but not found in Cdc11 (or the other two septins)

promiscuous binding seems to extend to certain septin subunits in other species. For example, it has been observed that individually expressed and purified human SEPT2 or SEPT7 homo-dimerize, whereas they only form heterohexameric complexes when they are co-expressed along with another one of their physiological binding partners, SEPT6 (Low and Macara, 2006; Sirajuddin et al., 2007). The ability to reliably image septin complexes in the EM has provided more incisive solutions to surmount these problems. First, one can determine whether any complex is formed, and the structure of that complex seen in the EM, when it is prepared in the absence of one (or more) of the septin subunits. Second, the rods can be decorated with specific antibodies against particular septins. However, when it binds, the antibody must become sufficiently immobilized; otherwise, its position cannot be visualized in the EM. Of course, the primary antigenic site recognized by the antibody must reside at the surface of the folded protein and must not become buried when contact is made with another subunit. To try to avoid the latter problem, a septin can be extended at its N- or C-terminal end with an epitope tag and, if it is competent for incorporation into complexes, they

SEPTIN COMPLEXES IN UNICELLULAR EUKARYOTES

55

can then be decorated with an appropriate anti-epitope antibody. To be seen, this antibody must also be immobile once bound. Third, a septin can be enlarged at its N- or C-terminal end by fusion to a small monomeric marker protein, such as GFP or E. coli maltose-binding protein (MalE), and if it is competent for incorporation into complexes and thereby immobilized, the location of the extra protein density can be visualized in the EM. It is impossible to conduct the first approach to examine the position of either Cdc3 or Cdc12 in native septin complexes isolated from yeast cells because cdc3
and cdc12
mutants are inviable in any strain background that has been tested. Thus, in the absence of Cdc3 or of Cdc12, the biological processes that require septin function cannot be properly executed. However, the nature of Cdc3-less and Cdc12-less complexes can be assessed by other means. In particular, complexes lacking Cdc3 or Cdc12 have been examined when the remaining three septins were co-expressed in bacterial cells. When Cdc10, Cdc11 and (His)6 -Cdc12 were co-expressed, stoichiometric (His)6 -Cdc12-Cdc11 binary complexes were recovered, but no significant amount of the Cdc10 was incorporated (Versele et al., 2004). The most parsimonious interpretations of this finding are, first, that Cdc11 interacts directly with Cdc12 and that Cdc3 is needed to recruit Cdc10. However, when (His)6 -Cdc3, Cdc10 and Cdc11 were co-expressed, no protein other than (His)6 -Cdc3 was recovered (Versele et al., 2004). Thus, this result suggests that, in the absence of interaction with Cdc12, Cdc3 is not competent to associate with Cdc10. Indeed, Cdc3 and Cdc12 associate avidly with each other and, when co-expressed, Cdc3, Cdc10 and (His)6 -Cdc12 form stoichiometric ternary complexes (Versele et al., 2004). Taken together, these findings already indicated that there is an intrinsic order to the assembly of the octameric rod, with Cdc12 serving as the linchpin since it associates directly with both Cdc11 and Cdc3. In most strain backgrounds, loss of Cdc11 is also lethal (Lee et al., 2002; Versele et al., 2004). However, two groups were able to isolate cdc11
derivatives, which are quite slow-growing and display grossly distended buds (Frazier et al., 1998; Casamayor and Snyder, 2003). When isolated from cdc11
cells, the Cdc11-less septin complexes form rods of somewhat heterogeneous length, but, in general, these are always detectably shorter than those formed by wild-type complexes (Frazier et al., 1998). Quite similarly, when co-expressed in bacteria in the absence of Cdc11, the remaining three septins form complexes containing stoichiometric amounts of Cdc3, Cdc10 and Cdc12 (Versele et al., 2004; Farkasovsky et al., 2005), as mentioned. In the EM, these complexes appear as rods with six clear-cut globular densities instead of the usual eight (Bertin et al., 2008). Thus, Cdc3, Cdc10 and Cdc12 are able to form a stable hexameric complex when no Cdc11 is present. This finding immediately suggests that Cdc11 normally occupies a terminal position in the rod – either a Cdc11-Cdc11 dimer at one end or, more likely, one Cdc11 at each end. Notably, Cdc11-less rods isolated from cdc11
cells (Frazier et al., 1998) or prepared from bacterial cells (Versele et al., 2004) are unable to polymerize into filaments in low salt (although one group

56

CH03 BIOCHEMICAL PROPERTIES AND SUPRAMOLECULAR ARCHITECTURE

reported that their Cdc11-less complexes assembled into filament-like structures, but that these were ‘less ordered’ than those generated by Cdc11-containing complexes (Farkasovsky et al., 2005)). Correspondingly, neck filaments cannot be detected in the EM after fixation and staining of cdc11
cells (Frazier et al., 1998). These observations are generally most consistent with the conclusion that Cdc11 molecules situated at the end(s) of octameric rods are necessary for and may directly mediate the efficient end-to-end polymerization required for normal filament formation. Given that the only direct interaction with another core septin detected for Cdc11 is with Cdc12, and that Cdc12 interacts with Cdc3 and that Cdc3, in turn, is able to interact with Cdc10 (and can do so, even when Cdc11 is absent, as long as Cdc12 is present) (Versele et al., 2004; Farkasovsky et al., 2005), already suggests that the arrangement in the rod must include the order: Cdc11-Cdc12-Cdc3-Cdc10. In most strain backgrounds, a cdc10
mutation also causes inviability; however, such a cell is able to pick up as yet unidentified unlinked suppressors or apparent epigenetic modifiers that allow such cells to survive and grow with increasing robustness as they are serially passaged due, perhaps in part, to observed increases in ploidy (McMurray et al., 2008). Nevertheless, for this reason, it has been possible to isolate native septin complexes from Cdc10-less cells (Frazier et al., 1998). Although the Cdc10-less complexes are also somewhat heterogeneous in length, the available images show that they are much shorter on average than the wild-type octameric rod or even the Cdc11-less hexameric rod (Frazier et al., 1998). Because these complexes were obtained by immunoaffinity enrichment using an anti-Cdc3 antibody (Frazier et al., 1998), and given that roughly stoichiometric amounts of Cdc3, Cdc11 and Cdc12 were obtained, these three subunits must be able to associate stably in vivo in the absence of Cdc10. Likewise, when co-expressed in bacterial cells, (His)6 -Cdc12 co-purifies with stoichiometric amounts of Cdc3 and Cdc11, even when Cdc10 is absent (Versele et al., 2004; Farkasovsky et al., 2005). The majority of the images for native Cdc10-less complexes seen on the EM grid seem to represent a Cdc11-Cdc12-Cdc3 trimer (but, because the presence of Cdc11 was somewhat sub-stoichiometric, the shortest forms visible are presumably Cdc3-Cdc12 dimers). When recombinant Cdc11-Cdc12-Cdc3 complexes are viewed in the EM, these are quite clearly trimeric assemblies (Bertin et al., 2008). Accordingly, these results immediately suggest that a Cdc10 homodimer is situated at the centre of the wild-type octameric rod and plays a central role in linking together two flanking Cdc3-Cdc12-Cdc11 trimers. Consistent with this conclusion, both native (Frazier et al., 1998) and recombinant (Versele et al., 2004) Cdc3-Cdc12-Cdc11 complexes are unable to polymerize into long filaments in low salt and, correspondingly, neck filaments cannot be detected in the EM after fixation and staining of cdc10
cells (Frazier et al., 1998). However, Cdc3-Cdc12Cdc11 complexes do form short forms, which likely represent Cdc11-mediated end-to-end polymerization of these trimers into hexamers, which then have a tendency to aggregate into non-specific bundles rather than into the clear-cut paired filaments formed by the wild-type octameric complex (Versele et al., 2004;

SEPTIN COMPLEXES IN UNICELLULAR EUKARYOTES

57

Farkasovsky et al., 2005). In any event, taken all together, an unambiguous order of septin subunits in the octameric rod could be inferred from these observations, namely Cdc11-Cdc12-Cdc3-Cdc10-Cdc10-Cdc3-Cdc12-Cdc11 (Figure 3.1c). The results of querying the structure of the octameric rod using the other two strategies mentioned above are fully consistent with this arrangement. When added, an anti-Cdc11 rabbit polyclonal antibody (Carroll et al., 1998) can be found bound to both termini of an individual octameric rod and, given their bivalent nature, such ‘Y’-shaped IgG molecules situated at each end often hold two octameric rods together side-by-side (Bertin et al., 2008). These observations definitively establish that Cdc11 is indeed the septin that occupies the outer-most position at both termini of the linear octameric rod. When complexes containing Cdc3, Cdc10, Cdc11 and (His)6 -Cdc12 are prepared, a commercial anti-(His)6 antibody most frequently decorates the second subunit in from one (or both) ends (Bertin et al., 2008), confirming that Cdc12 is the septin that occupies the penultimate position at each end of the octameric rod. It has also been possible to prepare by expression in bacteria complexes that contain Cdc3, Cdc10-GFP, Cdc11 and (His)6 -Cdc12, in agreement with the fact that expression of Cdc10-GFP in vivo is able to fully complement a cdc10
mutation (Cid et al., 1998). When such octamers are examined in the EM, extra density is readily observed and is juxtaposed to the centre of the rod and projects from the fourth subunit from one end and, occasionally, from both of the fourth subunits (i.e. counting in from each end) (Bertin et al., 2008). The fact that most of the Cdc10-GFP-containing rods yield only one extra density per octamer most likely reflects some degree of flexibility in the joint between GFP and Cdc10. Hence, when EM images of particles are grouped together on the basis of a common position for one GFP ‘blob,’ the second GFP presumably has swivelled through a variety of different angles, which tends to cancel out the second GFP density. Nonetheless, this observation confirms that Cdc10 is at the centre of the octameric rod. Similarly, co-expression in bacteria of Cdc10, Cdc11, (His)6 -Cdc12 and a MalE-Cdc3 fusion yields a stable octameric complex with the same overall structure as observed for complexes containing untagged Cdc3, with the exception of extra density juxtaposed to the third subunit from one end and, occasionally, from both of the third subunits (i.e. counting in from each end) (Bertin et al., 2008). Collectively, these findings made using the EM confirm unequivocally that the core yeast septins associate as a linear hetero-octameric rod with the order: Cdc11Cdc12-Cdc3-Cdc10-Cdc10-Cdc3-Cdc12-Cdc11 (Figure 3.1c). This arrangement demands that polymerization of octamers into filaments occurs via Cdc11-Cdc11 association. Indeed, both two-hybrid analysis (Lee et al., 2002; Casamayor and Snyder, 2003; Farkasovsky et al., 2005) and in vitro binding studies (Versele et al., 2004) showed that Cdc11 is capable of interacting with itself, consistent with the pivotal role of Cdc11-Cdc11 interaction at the octamer–octamer interface. Homotypic pairing of another septin, Cdc10, in the middle of the octamer is also central to the EM-derived structure. Curiously, however, this predicted Cdc10-Cdc10

58

CH03 BIOCHEMICAL PROPERTIES AND SUPRAMOLECULAR ARCHITECTURE

interaction was not detectable using the two-hybrid method (Lee et al., 2002; Farkasovsky et al., 2005) and only weak Cdc10-Cdc10 self-association was observed using purified, differentially tagged Cdc10 derivatives (Versele et al., 2004). Given the EM results and the requirements for hetero-octamer assembly deduced from simultaneous expression of various combinations of septins (already summarized above), the most reasonable explanation is that robust Cdc10-Cdc10 interaction requires a conformational change induced in Cdc10 by its prior binding to Cdc3, or, more likely, to a Cdc3-Cdc12 dimer. Such relationships provide further support for the conclusion that proper spatial organization of septin heterooctamers is achieved as the direct result of an ordered pathway for assembly.

Alternative arrangements and substitute subunits There is an additional mitotically expressed septin, Shs1/Sep7 (Carroll et al., 1998; Mino et al., 1998), and two septins expressed only during meiosis and sporulation, Spr3 and Spr28 (De Virgilio, DeMarini and Pringle, 1996; Fares, Goetsch and Pringle, 1996). Therefore, certain alternative arrangements of subunits within yeast hetero-octamers and/or filaments are likely to exist in vivo. To our knowledge, however, the positions of these three proteins within hetero-oligomeric complexes or in filaments have not been delineated. Shs1 is often considered an accessory septin because even null alleles (shs1
) have only a mild phenotype (Mino et al., 1998; Iwase et al., 2007) and because it is recovered in sub-stoichiometric amounts in purified septin complexes under certain conditions. When septin complexes are purified from wild-type yeast cells in buffers containing high salt (0.5–1 M ) by immunoaffinity absorption using either an anti-Cdc3 antibody (Frazier et al., 1998) or Protein A-tagged Cdc10 (Farkasovsky et al., 2005), or by co-immunoprecipitation with Gin4, a septin filament-associated protein kinase (Mortensen et al., 2002), Shs1 is present at ≤15 % of the amount of Cdc3, Cdc10 or any of the other septins. However, under conditions of intermediate ionic strength (100–250 mM salt), the complexes recovered contain higher amounts of Shs1, approaching equimolarity with the other septins, regardless of whether the method of purification involves affinity to Cdc3, Cdc10 or Cdc12 (Vrabioiu et al., 2004; Farkasovsky et al., 2005; Sung et al., 2005). This salt-sensitive association suggests that Shs1 interaction with other septins is mainly electrostatic. Perhaps revealing, even when purification is conducted at 100 mM salt, Shs1 is noticeably sub-stoichiometric when purification relies on affinity to Cdc11 (Sung et al., 2005). Conversely, in complexes prepared by immunoaffinity purification of (HA)3 -Shs1, the other septins are present in near stoichiometric amounts, except Cdc11 (Mortensen et al., 2002). In this regard, it is noteworthy that in native septin complexes purified in high salt from yeast cells (which contain Shs1) using either immunoaffinity adsorption to anti-Cdc3 antibody (Frazier et al., 1998) or via Protein A-tagged Cdc10 (Farkasovsky et al., 2005), Cdc11 is present at approximately half the level of the other three septins,

SEPTIN COMPLEXES IN UNICELLULAR EUKARYOTES

59

whereas when Cdc3, Cdc10, Cdc11 and Cdc12 are co-expressed in and purified from bacterial cells (which lack Shs1), the amount of Cdc11 is stoichiometric with the other three septins (Versele et al., 2004). The most parsimonious explanation to reconcile these observations is that Shs1 and Cdc11 compete for the same location in hetero-octamers. Compatible with this view, Cdc11 and Shs1 are closely related at the primary structure level and together occupy a phylogenetic group clearly distinct from the other mitotic S. cerevisiae septins (Pan, Malmberg and Momany, 2007). These considerations predict that there may be three classes of complexes in mitotic cells: Cdc11-Cdc12-Cdc3-Cdc10-Cdc10-Cdc3-Cdc12-Cdc11; Shs1-Cdc12-Cdc3-Cdc10-Cdc10-Cdc3-Cdc12-Shs1; and, the mixed oligomer, Cdc11-Cdc12-Cdc3-Cdc10-Cdc10-Cdc3-Cdc12-Shs1. Indeed, various assays to assess the capacity of Shs1 to interact with other septins demonstrated binding only to Cdc12 and Cdc11 (Mino et al., 1998; Versele et al., 2004; Farkasovsky et al., 2005). These latter observations are consistent with the structures proposed and with the idea that Shs1-capped hetero-octamers can interact with Cdc11-capped hetero-octamers, permitting their end-to-end co-assembly into filaments. Nonetheless, because the globular GTP-binding domain of Shs1 contains a large 30-residue insert that the other four mitotic septins lack, and because the C-terminal segment of Shs1 is nearly 100 residues longer than that of the next largest septin, it is possible that the hetero-octamers that contain Shs1 provide a building block that generates filaments with special features or the ability to recruit unique proteins as a means to fine-tune yeast cell morphogenesis. Furthermore, it is possible that the relative proportion of Shs1- versus Cdc11-containing hetero-octamers is tightly regulated, and coupled to events in the cell division cycle (including the status of septin collar assembly), via phosphorylation. This speculation is supported by the fact that, of all of the septins, Shs1 is the most highly phosphorylated and is the target of multiple protein kinases, including bud neck-associated Gin4 (Mortensen et al., 2002; Dobbelaere et al., 2003) and the CDK Pho85 (Egelhofer et al., 2008). Given the roles of septins in subcellular compartmentation and membraneassociated morphogenesis (reviewed in Longtine and Bi (2003); Douglas et al. (2005); Gladfelter (2006); Kinoshita (2006); Spiliotis and Nelson (2006); Barral and Mansuy (2007)), it is not surprising that formation of the novel membrane envelopes around the four meiotic nuclei during sporulation would involve specialized septins, Spr3 and Spr28, expressed only during this developmental event (De Virgilio, DeMarini and Pringle, 1996; Fares, Goetsch and Pringle, 1996). Global transcriptome profiling confirms that both the SPR3 and SPR28 genes are expressed only after the onset of sporulation, and expression of CDC3 and CDC10 is upregulated (Chu et al., 1998; Friedlander et al., 2006). Furthermore, Spr3 and Spr28 co-localize with a select subset of the mitotic septins (Cdc3, Cdc10 and Cdc11) in novel structures associated with the developing spore coats (De Virgilio, DeMarini and Pringle, 1996; Fares, Goetsch and Pringle, 1996), which are distinct from the filamentous septin collar at the bud neck. Although present and detectable throughout sporulation, Cdc12 and Shs1 are excluded

60

CH03 BIOCHEMICAL PROPERTIES AND SUPRAMOLECULAR ARCHITECTURE

from the spore-associated structures (McMurray and Thorner, 2008). Moreover, CDC12 and SHS1 (and also CDC11 ) expression is not upregulated detectably during sporulation. It is tempting to speculate (Versele and Thorner, 2005) that Spr3 replaces Cdc12, and that Spr28 replaces Shs1, given that the two subunits in each of these pairs are more closely related to each other at the sequence level than they are to any other septin (Pan, Malmberg and Momany, 2007). These observations suggest that there may be three classes of hetero-octameric complexes in meiotic cells: Cdc11-Spr3-Cdc3-Cdc10-Cdc10-Cdc3-Spr3-Cdc11; Spr28-Spr3-Cdc3-Cdc10-Cdc10-Cdc3-Spr3-Spr28; and, the mixed oligomer, Cdc11-Spr3-Cdc3-Cdc10-Cdc10-Cdc3-Spr3-Spr28. Furthermore, it has been reported that Spr28 is able to interact both with itself and with Cdc11 (De Virgilio, DeMarini and Pringle, 1996; Uetz et al., 2000; Drees et al., 2001), suggesting that the proposed rods would be able to undergo end-to-end polymerization to form filaments of distinctly different composition from those in mitotic cells. Thus, the sporulation-specific filaments presumably possess unique structural characteristics and the ability to associate uniquely with other proteins required for nuclear engulfment and spore membrane formation. Given that the mitotic septin hetero-octamer, once formed, appears to be exceedingly stable in vitro, resisting dissociation through multiple purification steps even in 1 M salt, how is subunit substitution accomplished in vivo? One possibility is that monomeric septins are subject to rapid proteolytic turnover. However, direct examination of two pivotal mitotic septins, Cdc10 and Cdc12, shows that, once synthesized, these individual subunits are practically indestructible, persisting through numerous successive cell divisions (McMurray and Thorner, 2008). Nonetheless, pulse-chase analysis shows that newly made Cdc10 and Cdc12 subunits can be incorporated into pre-formed octamers during vegetative growth (McMurray and Thorner, 2008). This suggests that there are cellular factors that promote subunit exchange, which presumably also allows for insertion of novel subunits for structure-specific purposes. In this regard, it is noteworthy that, even though they are excluded from the spore-associated structures, Cdc12 molecules persist during sporulation. Moreover, they are also precluded from re-assembling into mitotic structures when spores germinate, because only newly made Cdc12 is found at the necks of the first buds (McMurray and Thorner, 2008). This segregation of Cdc12 may be due to a sporulation-specific post-translational modification that cannot be ‘erased’ even in a mitotic cell.

Septins and septin organization in fission yeast (Schizosaccharomyces pombe) As in S. cerevisiae, the S. pombe genome encodes seven septin genes. Four of these (SPN1 , SPN2 , SPN3 and SPN4 ) are expressed during vegetative growth, whereas the remaining three (SPN5 , SPN6 and SPN7 ) are expressed only during sporulation. Primary sequence similarity (Pan, Malmberg and Momany, 2007)

SEPTIN COMPLEXES IN UNICELLULAR EUKARYOTES

61

and other considerations (expression pattern, mutant phenotype, size, etc.) (An et al., 2004) permit identification of an apparent S. pombe counterpart for each S. cerevisiae septin. For the mitotic septins, clearly Spn1 ≈ Cdc3, Spn2 ≈ Cdc10, Spn3 ≈ Cdc11 and Spn4 ≈ Cdc12 (Pan, Malmberg and Momany, 2007). For the sporulation-specific septins, the relationships are less clear, but it seems that Spn5 ≈ Spr28, Spn6 ≈ Spr3 and Spn7 ≈ Shs1. Given the roles attributed to septins in morphogenesis, the fact that S. pombe is bullet-shaped and forms a linear tetrad of spores, whereas S. cerevisiae is ovoid and forms tetrads with tetrahedral geometry, presumably explains the biological basis of some of these differences. Unlike S. cerevisiae, loss of no single septin renders S. pombe inviable, and fission yeast cells remain viable even in the absence of all four of their mitotic septins (An et al., 2004). Nonetheless, absence of any one of the septins causes a pronounced delay in the completion of cell division, resulting in a distinctive morphological phenotype, namely formation of chains of cells. In any event, because every fission yeast septin is non-essential, the effects on septin complex assembly of deleting individual (or combinations of) subunits could be assessed in vivo. When the ability a tagged septin allele to co-immunoprecipitate each of the other septins was examined (An et al., 2004), the resulting data support with few exceptions (noted below) a model for organization of the mitotic septin complex in S. pombe that is reassuringly similar to that of the S. cerevisiae hetero-octamer (Figure 3.1c). First, salt-stable septin complexes from fission yeast contain two copies of each of the four subunits, and are thus hetero-octamers (An et al., 2004). Second, in the absence of Spn2, a stable Spn1-Spn3-Spn4 complex (presumably a 1 : 1 : 1 trimer) still forms and, in the absence of Spn3, a stable Spn1-Spn2-Spn4 complex (presumably a 2 : 2 : 2 hexamer) still forms; but, in the absence of Spn1, only a Spn3-Spn4 complex (presumably a 1 : 1 dimer) is observed and, in the absence of Spn4, only a Spn1-Spn2 complex (presumably a 2 : 2 tetramer) is found (An et al., 2004). These results are fully compatible with a linear hetero-octamer with the order Spn3-Spn4-Spn1-Spn2-Spn2-Spn1-Spn4-Spn3 (Figure 3.1c). Third, in this arrangement, each fission yeast septin occupies a position congruent with that of its budding yeast ortholog. Fourth, and also consistent with this model, Spn2 self-associates in the absence of the other three septins (An et al., 2004). Homotypic interaction of this sort was also observed for Spn1, as seen in vitro (albeit weakly) for its S. cerevisiae ortholog, Cdc3, when its preferred partners are absent (Versele et al., 2004). Likewise, interaction of differentially tagged Spn4 molecules can still be detected when either Spn3 or Spn2 is absent. Because these experiments were performed under relatively low stringency salt conditions (150 mM ) (An et al., 2004), and because the deduced structure of the S. pombe octamer predicts that polymerization should occur via Spn3-Spn3 association, the co-precipitation of differentially tagged forms of Spn4 when Spn2 is absent could be explained by end-on-end joining of Spn1-Spn4-Spn3 trimers to form Spn1-Spn4-Spn3-Spn3-Spn4-Spn1 hexamers. Indeed, under the same conditions, Spn3-Spn3 co-precipitation required Spn1 and Spn4, but not Spn2 (An et al., 2004). Finally, despite the potential pitfalls of the two-hybrid method, especially

62

CH03 BIOCHEMICAL PROPERTIES AND SUPRAMOLECULAR ARCHITECTURE

when conducted in a heterologous host (budding yeast) that contains closely related proteins, the observed interactions between the S. pombe septins assessed by this approach agree well, for the most part, with the order predicted above (Figure 3.1c). Spn3 interacts with Spn4, but not with Spn1 or Spn2; and, Spn1 interacts with Spn2 and Spn4, but not Spn3 (An et al., 2004). The sole departure from the predicted arrangement is that a robust interaction of Spn2 with Spn4 was observed (An et al., 2004). The corresponding S. cerevisiae orthologs, Cdc10 and Cdc12, also displayed a modest ability to associate when Cdc3 (Spn1 ortholog) was absent, but a much more robust interaction when all three septins were present (Versele et al., 2004). Even though it is now abundantly clear that the septin complexes from budding yeast (Frazier et al., 1998; Bertin et al., 2008), a nematode (John et al., 2007), an insect (Field et al., 1996) and human cells (Lukoyanova, Baldwin and Trinick, 2007; Sirajuddin et al., 2007) are linear rods of single-subunit width, this has not yet been formally demonstrated for the fission yeast septin complex. Furthermore, to our knowledge, it has not yet been formally demonstrated either in vitro or in vivo that fission yeast septin complexes can, in fact, polymerize into filaments of any sort.

ARRANGEMENT OF SEPTIN SUBUNITS WITHIN COMPLEXES AND FILAMENTS OF INVERTEBRATES The worm’s turn: subunit organization in a minimal septin complex Of all organisms whose genome has been completely sequenced, the one with the fewest number of recognizable septin genes is the nematode, Caenorhabditis elegans (Pan, Malmberg and Momany, 2007), which encodes only two septins, UNC-59 and UNC-61. The phenotype of worms lacking functional versions of either (or both) of the corresponding genes is virtually the same (Nguyen et al., 2000; Finger, Kopish and White, 2003), suggesting that, together, they represent a minimal septin complex whose function relies on both subunits. Biochemical and structural analysis have been performed on UNC-59-UNC-61 complexes expressed in and purified from bacterial cells (John et al., 2007). These behave in many ways like analogously prepared S. cerevisiae Cdc3-Cdc12 complexes (Versele et al., 2004; Farkasovsky et al., 2005). First, based on primary sequence similarity and size (Pan, Malmberg and Momany, 2007), UNC-59 most resembles Cdc12 and UNC-61 most resembles Cdc3. Second, when expressed individually, worm septins are insoluble, as observed especially for native Cdc12 (Versele and Thorner, 2004), but co-expression of UNC-59 and UNC-61 yields a soluble complex, as observed for co-expression of Cdc3 and Cdc12 (Versele et al., 2004). The UNC-59-UNC-61 complex is an equilibrium mixture of heterodimer and heterotetramer, each of which contains equimolar amounts of each septin and no bound nucleotide (John et al., 2007). However, unlike Cdc3-Cdc12 dimers, the

SEPTIN COMPLEXES IN INVERTEBRATES

63

UNC-59-UNC-61 complex is able to polymerize in vitro into filaments (often paired) with a diameter of 8–9 nm (John et al., 2007), although these structures are rather rudimentary and disorganized compared to the reproducibly long and luxuriant filaments formed from S. cerevisiae hetero-octamers (Bertin et al., 2008). High-resolution EM analysis reveals that the heterotetrameric UNC-59-UNC-61 complex is a linear 17–19-nm-long rod (John et al., 2007). Complexes composed of GFP-UNC-59 and native UNC-61 display extra density localized to the rod termini, whereas complexes composed of GFP-UNC-61 and native UNC-59 display extra density near the centre of the rod (John et al., 2007). Furthermore, in the GFP-UNC-61-UNC-59 complexes, two extra densities were often found at the centre of and on the same side of the heterotetrameric rod (John et al., 2007). Hence, the subunits in the heterotetrameric rod are arranged in the order UNC-59-UNC-61-UNC-61-UNC-59 (Figure 3.1c). Like Cdc3 and Cdc12 (and most septins), both UNC-59 and UNC-61 contain a C-terminal extension (CTE) that contains at its distal end a segment with predicted coiled coil–forming capacity. These elements from UNC-59 and UNC-61 are able to associate with each other to form a parallel heterodimeric coiled coil, as judged by cysteine cross-linking studies (John et al., 2007). Furthermore, in complexes composed of UNC-61-GFP (that is, marked where the end of the coiled coil presumably is) and native UNC-59, the extra density appears at various positions along an arc centred 10 nm from the core of the heterotetramer (John et al., 2007). This observation, combined with the fact that the N-termini of both UNC-61 molecules in a heterotetramer seem to project off the same side of a rod, suggests that coiled coils formed by the C-terminal tails of UNC-59 and UNC-61 project orthogonally off the opposite side of the rod. Thus, the worm heterotetramer, and presumably the budding and fission yeast hetero-octamers, possess twofold rotational symmetry about an axis that runs between and orthogonal to the central septin pair (Figure 3.1c), the only obvious difference being the number of subunits. How did C. elegans ‘lose’ four subunits per septin complex, relative to the yeast octamers? It seems just as likely that other organisms have gained subunits through gene duplication and divergence, generating derivatives that retained the capacity to be incorporated into the rods. For example, the fact that GTP-binding domains of Cdc10 and Cdc12 are more closely related to each other (40 % identity) than either is to any other mitotic septin suggests that Cdc10 could have arisen from duplication and divergence of an ancestral CDC12 -like gene followed by deletion of its C-terminal end. Since the C-terminal coiled coil of Cdc12 is not essential for Cdc12-Cdc12 interaction (Versele et al., 2004; Farkasovsky et al., 2005) (see below), Cdc10 retained the capacity for self-pairing and association with Cdc3, without interfering with the formation of coiled coil–mediated Cdc3-Cdc12 dimers. Of course, alternative scenarios are also possible. Based on phylogenetic relationships (rather than on one-on-one comparisons), and aside from its substantially longer C-terminal end, UNC-61 appears most related to Cdc10 (Pan,

64

CH03 BIOCHEMICAL PROPERTIES AND SUPRAMOLECULAR ARCHITECTURE

Malmberg and Momany, 2007). Thus, the Ur-septin to both Cdc10 and UNC-61 may have lacked a CTE and played the ‘central homodimer’ role, which has been conserved; however, over the course of time, UNC-61, but not Cdc10, underwent an illegitimate recombination event that fused it to a sequence with coiled coil–forming capacity, thereby further ‘accessorizing’ this worm septin, making it more like Cdc3. In the absence of capping by a Cdc11-like septin, the Cdc12-like UNC-59 may have retained an inchoate ability for self-association, allowing for end-to-end polymerization of UNC-59-UNC-61-UNC-61-UNC59 rods into filaments.

A word from the Pnut gallery The first Drosophila melanogaster septin gene identified was called pnut, on the basis of the vestigial peanut-shaped imaginal discs found in larvae defective in the corresponding gene (Neufeld and Rubin, 1994), and the genome of this fly encodes four other septin genes. Native multi-septin complexes have been isolated from D. melanogaster (Field et al., 1996) in a manner similar to that used to enrich for endogenous yeast septin complexes (Frazier et al., 1998). Fly embryo extracts incubated with a Pnut-specific antibody yield a salt-stable complex containing, in addition to Pnut, stoichiometric amounts of Sep1 and Sep2 (Field et al., 1996). When dialysed into low salt and examined in the EM, this complex forms filaments of 7–9-nm-diameter that tend to associate laterally and form large bundles at high concentration (Field et al., 1996). In high salt, the complex seen in the EM is a linear ∼26-nm long rod (Field et al., 1996). Based on the dimension of an individual yeast septin subunit (∼4 nm), this length corresponds well to that expected for a linear heterohexameric rod that presumably contains two molecules of each of the three co-purified Drosophila septins. The correspondence between fly septins and their fungal counterparts are not perfectly obvious from either one-on-one comparisons or phylogenetic analysis. In terms of sequence identity (40 %) and match length (435 residues), Pnut (539 residues) is most similar to Cdc3 (520 residues). For the same reasons (42 % identity over 322 residues), Sep1 (361 residues) most resembles Cdc10 (322 residues). In contrast, Sep2 (419 residues) is much more ambiguous. It bears nearly the same degree of similarity to all of the core mitotic septins of yeast; however, it is slightly closer in both size and identity to Cdc12 (407 residues; 34 %) than it is to Cdc11 (415 residues; 32 %). Thus, it is tempting, on these grounds, to speculate that the Pnut-containing septin complex has the order Sep2-Pnut-Sep1-Sep1-Pnut-Sep2. However, to our knowledge, there are no pair-wise interaction data to support this particular arrangement. A slightly different ordering of the subunits within the purified Drosophila hexamer will be proposed below (see also Figure 3.1c), based on structural information about a hexameric complex of human septin subunits that are perhaps more closely related to the fly septins.

SEPTIN COMPLEXES IN VERTEBRATES

65

ARRANGEMENT OF SEPTIN SUBUNITS WITHIN COMPLEXES AND FILAMENTS OF VERTEBRATES Onward and upward in evolution and resolution: mammalian septin complexes The preceding presentation of septin complex organization and ultrastructure in several model eukaryotes has set the stage for discussing mammalian septins, in all their resplendent complexity. As detailed elsewhere in this volume, the Homo sapiens genome encodes 14 septin genes, most of which produce an array of products via differential splicing and/or via use of alternative translation initiation sites (Macara et al., 2002; Kinoshita, 2003; Hall and Russell, 2004; Joo, Tsang and Trimble, 2005). This situation makes the combinatorial complexity of potential multi-septin complexes dauntingly high. Only a few such complexes have been analysed in a way that is informative about their ultrastructure. However, this includes the crystal structure of a heterohexameric complex of human SEPT2, SEPT6 and SEPT7 (Sirajuddin et al., 2007), which has provided new and important insights about the interfaces by which individual septin subunits interact with their neighbours, as discussed in detail below.

Rods are the rule: mammalian septins form linear complexes Immunoaffinity isolation of a given septin and its associated proteins from mammalian cells produces complexes containing different septin members, depending on the cell type and the subunit harbouring the epitope used for purification. In mouse brain lysates, anti-SEPT2 antibodies enrich for complexes containing nine different septins, among which SEPT2, SEPT4, SEPT6 and SEPT7 are the most prominent and, together, represent a complex of roughly 1 : 1 : 2 : 2 stoichiometry (Kinoshita et al., 2002). When extracted from a human cell line (HeLa), anti-SEPT2 antibodies pull down a simpler complex containing, as the other septins, only SEPT6 and SEPT7 (Kinoshita et al., 2002). Using anti-SEPT7 antibodies, lysates of rat brains yield a complex containing SEPT2, SEPT4, SEPT6 and SEPT7, but only three (SEPT4, SEPT6 and SEPT7) appear equimolar (Hsu et al., 1998). When examined in the EM, rods were observed in the preparations from the rat; the rods were ∼8 nm in width and ∼25-nm long with ∼10-nm-long thin ‘strands’ projecting laterally from one side of the rod, and were occasionally seen to be associated end-to-end (Hsu et al., 1998). The strands were evenly spaced ∼16 nm apart, with one or two strands per 25-nm rod (Hsu et al., 1998). The significance of these features was not appreciated at the time, but in light of the preceding discussion, it is clear that the strands represent the coiled coils formed between component subunits in the rod. Which of those subunits in the rod are able to make such associations was revealed by the structure determined by X-ray diffraction, described in detail below.

66

CH03 BIOCHEMICAL PROPERTIES AND SUPRAMOLECULAR ARCHITECTURE

In addition to recovery of native complexes, human SEPT2, SEPT6 and SEPT7 have been co-expressed heterologously both in insect cells (Kinoshita et al., 2002) and bacterial cells (Sheffield et al., 2003; Sirajuddin et al., 2007), and yield a stable complex with 1 : 1 : 1 stoichiometry. Notably, each full-length septin was insoluble when expressed individually in E. coli (Sheffield et al., 2003), as has also been observed for SEPT4 (Garcia et al., 2007), whereas any pair was soluble and formed a 1 : 1 stoichiometric complex (Sheffield et al., 2003). As judged by size exclusion chromatography (Kinoshita et al., 2002; Sirajuddin et al., 2007) or dynamic light scattering (Sheffield et al., 2003) in 100–500 mM salt, the apparent molecular size of the purified SEPT2-SEPT6-SEPT7 complex is most compatible with a 2 : 2 : 2 hexamer. When examined in the EM, this complex appears as a linear rod that is 5–10 nm in width (Kinoshita et al., 2002; Sheffield et al., 2003; Sirajuddin et al., 2007) and 250-nm long (Sirajuddin et al., 2007), with a substructure composed of six roughly globular subunits each ∼4 nm in diameter (Sirajuddin et al., 2007). Aside from being two subunits shorter, this ultrastructure is virtually identical to what is observed for the yeast hetero-octamer (Figure 3.1b) (Bertin et al., 2008). Also, its length (25 nm) is nearly identical to that of the Drosophila Sep1-Sep2-Pnut (∼26 nm) and rat SEPT4-SEPT6-SEPT7 (∼25 nm) complexes described earlier, suggesting these non-human septin complexes are also 2 : 2 : 2 hexamers. Decreasing the salt concentration allows the human SEPT-2-SEPT6-SEPT7 rods to self-assemble into a variety of structures, including filaments, filament bundles and even rings (Kinoshita et al., 2002; Sirajuddin et al., 2007).

Crystal structure of the human SEPT2-SEPT6-SEPT7 complex Near atomic level structures for crystals of an individual purified septin (SEPT2) ˚ resolution) and the physiologically relevant SEPT2-SEPT6-SEPT7 complex (3.4 A ˚ resolution) have been determined by X-ray crystallography introduced above (4 A (Sirajuddin et al., 2007). The first striking conclusion to be drawn from these two structures is that, both when alone and in the complex, SEPT2 homodimerizes; but, in the isolated SEPT2-SEPT2 homodimer crystal, one of the interfaces by which monomers are joined is completely different from the interface by which SEPT2 self-associates when it is located at the centre of the SEPT2-SEPT6-SEPT7 hexamer (Sirajuddin et al., 2007). This observation clearly illustrates that any given septin has the ability to interact with other septins in multiple ways that depend on the identity of the other available subunits, and implies that a variety of septin complexes differing in composition and subunit arrangement could be achieved in vivo simply by altering the expression, localization, or modification of the different subunits and/or their isoforms. Second, although the structure of the heterohexamer confirmed many aspects of the ultrastructural model for the yeast complex (Figure 3.1c), it also revealed unforeseen features vital for understanding how these structures assemble. That

SEPTIN COMPLEXES IN VERTEBRATES

67

the conformation of each septin subunit is dominated by its GTP-binding domain fold (‘G domain’) was expected, given the sequence conservation of this domain in all septins (Pan, Malmberg and Momany, 2007) and the dimensions (4-nm diameter) of the globular densities viewed in the EM (Figure 3.1b). The G domain in SEPT2 and in the other two human septins is very much like the canonical G domain in Ras, with five α-helices and six β-strands (Sirajuddin et al., 2007) (see Plate Figure 3.3a, after p. 246). However, septins have some important additional accoutrements that are critical for conferring the ability to form heteromeric complexes. At their N termini, septins have an additional helix (dubbed α0 so as to preserve the numbering of the first Ras-like α-helix as α1), which can be further extended by additional non-helical sequence (e.g. in SEPT6). Compared to Ras-like GTPases, the G domain of a septin also has an extra helix situated between α4 and β6, two more β-strands between β6 and helix α5 and another helix (α6) located just preceding the CTE present in most septins (Sirajuddin et al., 2007) (Plate Figures 3.3a and 3.3c). The CTE contains at its distal end the heptad repeat sequences predicted to be able to form an α-helical coiled coil. In fact, the sequence that encompasses the structural elements β7, β8 and α6 was already recognized as diagnostic of septins (as opposed to Ras-like GTPases or any other class of known GTP-binding protein) and was, for that reason, dubbed the ‘septin unique’ domain (SUD) or septin unique element (SUE) (Versele et al., 2004). As is necessary for a linear array of similar subunits that possesses twofold rotational symmetry about the axis running orthogonal to and between the central doublet, there must be two types of interfaces between any given pair of ˚ 2 ) comprises a G domain-G domain interaction subunits. One interface (1851 A dominated by contacts in and around the guanine nucleotide-binding site in each monomer, and was dubbed a G dimer. The other significantly more extensive ˚ 2 ) involves contacts involving the extra N-terminal (α0) and interface (2995 A C-terminal (α6) helices that distinguish septins from Ras-like small GTPases, and was dubbed an NC dimer. Individual residues identified as playing key roles in each of these septin–septin interfaces are, as expected, well conserved (Sirajuddin ˚ et al., 2007), although it should be noted that, at the resolution achieved (4 A), many side chains in the final refined structure were modelled as alanines, and several loops and other segments (most notably the region after α6) were either disordered or deleted in the derivatives used to prepare the crystals (Sirajuddin et al., 2007). The structure of the human septin heterohexamer was obtained in sufficient detail to determine that the order of subunits is SEPT7-SEPT6-SEPT2-SEPT2SEPT6-SEPT7. Crystals of SEPT2 alone show that it is able to self-associate in both the G dimer and NC dimer modes, whereas in the hexamer the SEPT2SEPT2 pair is clearly a dimer in the NC orientation. This arrangement forces the SEPT2-SEPT6 interface to be a G dimer and, as a result, the SEPT6-SEPT7 interface is an NC dimer (Plate Figure 3.3a). Hence, these two kinds of interfaces alternate along the rod. Thus, if the hexamer were extended by an additional subunit on each end, as seen in the yeast hetero-octameric rod, the interface

68

CH03 BIOCHEMICAL PROPERTIES AND SUPRAMOLECULAR ARCHITECTURE

between SEPT7 and that other septin (and presumably between Cdc12 and Cdc11; Plate Figure 3.3b) is predicted to be in the G-dimer mode. Therefore, when yeast hetero-octamers polymerize end-on-end, the Cdc11-Cdc11 interface will be an NC dimer, whereas when the human heterohexamer polymerizes end-on-end, the SEPT7-SEPT7 interface will be a G dimer. This latter difference could explain the more robust filament-forming capacity of the yeast complex (and the differences in the resulting structures formed) as compared to the human complex. Given the order of subunits within the human SEPT2-SEPT6-SEPT7 hexamer established by crystallography, the order of the orthologous subunits in the corresponding purified mouse and rat complexes is undoubtedly the same (Figure 3.1c). Phylogenetic analysis suggests that D. melanogaster Sep1 is in the same category as human SEPT2, that fly Sep2 is most related to SEPT6, and that Pnut is grouped with human SEPT7 (Pan, Malmberg and Momany, 2007). Hence, we favour, as a model for the fly hexamer, the order Pnut-Sep2-Sep1-Sep1-Sep2-Pnut (Figure 3.1c). It should be recalled that the rat complex subjected to EM analysis was a SEPT4-SEPT6-SEPT7 complex (Hsu et al., 1998). Given that SEPT4 is significantly more related (61 % identity) to SEPT2 than to either SEPT6 (40 % identity) or SEPT7 (53 % identity), this complex presumably represents a SEPT7-SEPT6-SEPT4-SEPT4-SEPT6-SEPT7 heterohexamer. Similarly, the mouse complex obtained using anti-SEPT2 antibodies had a apparent composition of SEPT2, SEPT4, SEPT6 and SEPT7 of 1 : 1 : 2 : 2 (Kinoshita et al., 2002). Thus, these preparations could represent an equimolar mixture of SEPT7-SEPT6-SEPT2SEPT2-SEPT6-SEPT7 and SEPT7-SEPT6-SEPT4-SEPT4-SEPT6-SEPT7 rods that have associated end-on-end to form mixed polymers.

Cellular asymmetry and septin filament polarity To date, every heteromultimeric septin complex analysed at sufficient resolution bears two-fold rotational symmetry about the axis that runs through and orthogonal to the centre of the rod. Consequently, each rod is non-polar in its long dimension and, thus, the two ends (if intact) present indistinguishable surfaces for end-to-end interaction. Therefore, filament assembly in solution should be able to occur by addition of a second complex to either or both ends of a rod, such that a filament can potentially grow in both directions. To our knowledge, however, kinetic studies of polymerization to confirm this prediction have not been performed (for example, using rods with a recognizable fiduciary mark as the seed to nucleate polymerization of an excess of unmarked rods). It is theoretically possible that addition of a second complex to an initial rod may constitute a ‘priming’ step in filament growth, if complex addition at one end of a rod causes a conformational change that is propagated through the rod to its other end and makes that unoccupied terminal subunit more competent to associate with another rod. In this way, once started, polymerization of rods into filaments would be greater than first order with respect to rod concentration. In principle, however, other

SEPTIN COMPLEXES IN VERTEBRATES

69

processes could differentially affect the otherwise identical subunits at each end of a rod and cause one end of a rod to be favoured for polymerization or, conversely, could cap one end of a rod and disfavour polymerization at that end. If so, filaments might be able to grow unidirectionally, especially in the cell, where the nature of the guanine nucleotide bound (see below), the state of post-translational modification (see also below), or the extent of association with anchoring proteins or other cellular factors might differ at each of a rod. If the two ends of a rod and/or a filament are somehow different in vivo, septin-containing cellular structures may acquire inherent asymmetry. Indeed, there are clues that such is the case, at least in S. cerevisiae. In budding yeast, a septin-containing ring assembles at the cell cortex and marks the incipient budding site (Longtine and Bi, 2003). During bud formation, this ring expands into a hollow hourglass-shaped collar of septin filaments that lines the mother-bud neck (Byers and Goetsch, 1976a) (Figure 3.4).The mechanism of this expansion and the direction of growth of the septin filaments in the collar are still matters of some controversy, as discussed further below. Fluorescence Recovery After Photobleaching (FRAP) studies of cells expressing GFP-tagged septins support the idea that the septin complexes in the ring are mobile, whereas those in the collar are fixed (Caviston et al., 2003; Dobbelaere et al., 2003). This result is consistent with the idea that, regardless of the length or orientation of the filaments present in the ring, there are free ends from which rods can be lost and onto which rods can be added, but in the collar there are no free ends, as would be expected if the filaments eventually circle around and form closed hoops, as suggested by the available EM images (Figure 3.4). In any event, many proteins bind asymmetrically to one or the other side of this collar (Gladfelter, Pringle and Lew, 2001). There are many possible mechanisms to explain this bipartite behaviour. However, interestingly, it has been reported that at least one protein (Bni4-CFP) is located exclusively on the exterior of the ring even before budding and remains on the mother side of the collar post-budding, whereas at least one other protein (Kcc4-YFP) associates only with the interior aspect of the ring even before budding and is brought to the bud side of the collar during budding (Kozubowski, Larson and Tatchell, 2005). This kind of observation suggests that, both prior to and following filament assembly, there is some sort of polarity imposed on the seemingly non-polar septin filaments. Contrary to the just stated assertion, it has been concluded that at least in terms of the relative arrangement of Cdc3, Cdc11 or Cdc12, the filamentous septin collar is symmetrical (Vrabioiu and Mitchison, 2007). This conclusion was reached as the result of examining the fluorescence anisotropy of GFP-tagged derivatives of each of these three septins in which the GFP was purportedly fused at a fixed rotational angle with respect to the predicted coiled coils of intact Cdc3, Cdc11, or Cdc12 (Vrabioiu and Mitchison, 2007). One hypothesis proposed by Vrabioiu and Mitchison (2007) to explain the apparent discrepancy between this conclusion and the asymmetric distribution observed for many collar-associated proteins, is that during the filament assembly necessary for collar formation (i) the newly synthesized septin complexes are added unidirectionally along the mother-bud axis

70

CH03 BIOCHEMICAL PROPERTIES AND SUPRAMOLECULAR ARCHITECTURE

3

5

12

10

28

(a)

10

5

28

(b)

5

SEPTIN COMPLEXES IN VERTEBRATES

71

to protofilaments composed of pre-existing complexes and (ii) Bni4 and Kcc4 are somehow able to bind preferentially to septin complexes of different ‘ages.’ However, this proposal seems unlikely because pulse-chase analysis shows that pre-existing and newly synthesized septins are distributed uniformly and symmetrically within the septin collar, at least at the same level of microscopic resolution at which Bni4 and Kcc4 were found to be differentially localized (McMurray and Thorner, 2008). How, then, is the inherent symmetry of the collar filaments ‘broken’ to permit asymmetric localization of septin-associated factors? In yeast, septin accumulation at specific cortical sites depends on the prior arrival of the Rho-type GTPase Cdc42 (Iwase et al., 2006), but assembly into the ring further requires Cdc42-dependent actin filament assembly (Gladfelter et al., 2002; Kozubowski, Larson and Tatchell, 2005; Iwase et al., 2006). Blocking the latter prevents establishment of the observed asymmetry in the localization of Bni4 and Kcc4 (Kozubowski, Larson and Tatchell, 2005). Thus, the ‘polarity’ of septin-containing structures that is responsible for the asymmetric localization of certain septinassociated proteins likely stems from non-septin factors, such as Cdc42-interacting proteins, polymerization of F-actin (which is a filament with inherent polarity) and/or the directed actin-mediated delivery of other molecules to one side or the other of nascent septin structures. According to this idea, this asymmetry would be established early in the maturation process of septin-containing structures (e.g. in non-filamentous septin rings), but would continue during formation of the collar at the bud neck, as long as the asymmetric association(s)/delivery is maintained. In essence, mother-bud asymmetry at the neck may derive from differential interactions of non-septin proteins with ‘anchored’ or ‘free’ ends of inherently non-polar

Figure 3.4 Structure-based interpretation of the arrangement of the septin filaments in the collar at the S. cerevisiae bud neck. In all the EM images and corresponding schematic depictions shown, the mother bud axis runs right-to-left. Arrows in the images in (a) indicate profiles of a single presumptive filament followed through serial sections; arrows in the right image in (b) indicate regions where the presumed filament density is resolved as two fine lines 5 nm apart. The accompanying schematic illustrations are not necessarily to scale. (a) Top left, a negatively stained thin-section of a sagittal section thorough the mother-bud neck of a mitotic S. cerevisiae cell. Bottom left, a schematic representation of the EM shown above. Top right, higher magnification EM image of the region boxed in the bottom left diagram. Bottom right, schematic representation of the EM image shown above. (b) Top left, a negatively stained thin-section of a grazing section of the mother-bud neck just tangential to the plasma membrane. Bottom left, schematic representation of the EM image shown above. Top right, higher magnification image of the central portion of the image shown in top left of (b). Bottom right, schematic representation of the EM image shown above. EM labels: ‘cw’, cell wall; ‘m’, mitochondrion; ‘f’, 10-nm filament; ‘mc’, membrane connection; ‘lc’, lateral connections. Symbols in schematic diagrams: Septin subunits or their G domains, circles or ovals; CTEs wavy lines; numbers, dimensions (in nm). All EM images taken from Byers and Goetsch (1976). Reproduced with permission of Rockefeller University Press, from The Journal of Cell Biology, 69, 717–21

72

CH03 BIOCHEMICAL PROPERTIES AND SUPRAMOLECULAR ARCHITECTURE

septin octamers that polymerize unidirectionally via the mechanisms proposed above.

Orientation and organization of the predicted coiled coil–forming domains Homotypic Cdc12-Cdc12 interaction is reproducibly detected using either the two-hybrid method (Lee et al., 2002; Farkasovsky et al., 2005) or in vitro binding of differentially tagged Cdc12 derivatives (Versele et al., 2004), and is eliminated or greatly diminished when its entire CTE (or the predicted coiled coil–forming region at the distal end of the CTE) is deleted. By contrast, deleting the C-terminal 58 residues of Cdc12, which includes the 40-residue segment with predicted coiled coil–forming capacity (Figure 3.2 and Plate Figure 3.3c), does not prevent interaction of Cdc12 with Cdc3, as judged by the two-hybrid method (Farkasovsky et al., 2005), suggesting that a different region of the CTE of Cdc12 must be responsible for its ability to associate with Cdc3. Indeed, deleting the entire CTE (C-terminal 68 residues of Cdc12), thereby removing a segment (‘α7’) (Figure 3.2 and Plate Figure 3.3c) that is strongly predicted to be α-helical, totally ablates the ability of Cdc12 to interact with Cdc3 (Versele et al., 2004). Moreover, when this exact portion (residues 339–407) of Cdc12 was fused to GST, the resulting non-septin chimera is capable of associating specifically with Cdc3 (Versele et al., 2004). These findings highlight, first, an important distinction between the entire CTE and its more distal portion, which contains the segment with potential coiled coil–forming propensity. Second, these findings also highlight that, at least in certain septins, regions in the CTE other than the α6 element (which was revealed by the crystal structure (Plate Figure 3.3a), is shared by all septins (Plate Figure 3.3c) and involved in the NC dimer interface (Plate Figure 3.3a)) can strongly influence specific septin–septin association. It is unfortunate, therefore, that notably absent from the crystallographic structure of the human septin rod (Sirajuddin et al., 2007) is any consistent electron density beyond α6 that would correspond to any part of the CTEs of these septins. Likewise, aside from the work of Hsu et al. (1998) on a septin complex isolated from rat brain, EM images of other septin rods and filaments from budding yeast (Frazier et al., 1998; Bertin et al., 2008), nematode (John et al., 2007) and fruit fly (Field et al., 1996) do not display any orthogonal strands visible by negative staining. Failure to visualize the CTEs is not due to their proteolytic loss because (i) the input full-length septins were recovered when crystals of the human septin complex were redissolved (Sirajuddin et al., 2007), (ii) full-length septins are the sole species present in the yeast preparations visualized by EM (Bertin et al., 2008) and (iii) the C-terminal GFP tag in the UNC-61-GFP fusion was detectable when a worm septin complex containing it was examined in the EM (John et al., 2007). However, as should be recalled, in the latter analysis, the density corresponding to GFP was at a number of positions distributed in an apparent arc

SEPTIN COMPLEXES IN VERTEBRATES

73

∼10 nm from the globular domain of UNC-61. This suggests that pairing of the CTE of UNC-61 and the CTE of UNC-59 to form the parallel coiled coil occurs, but that this element is sufficiently flexible that it does not assume any fixed position relative to the rod. Hence, the locations of these projections are impossible to discern by either EM or crystallography. Indeed, disorder in the CTE was clearly a source of difficulty in obtaining crystals of SEPT2 that diffracted well; the structure of the isolated SEPT2-SEPT2 homodimer was only solved when its C-terminal 46 residues were deleted (Sirajuddin et al., 2007). Nevertheless, in the crystal structure of the human septin hexamer (Sirajuddin et al., 2007), the α6 helices are well ordered and clearly point in a direction perpendicular to the long axis of the rod (Plate Figure 3.3a). Assuming the path of a CTE beyond that point follows in essentially the same direction, for each septin pair at an NC interface, their CTEs and C-terminal coiled coil–forming segments will project from the same face of the rod and be closely juxtaposed, permitting their interaction. As long as the orthogonal ‘strands’ observed in the EM images of Hsu et al. (1998) were not due to another rat brain protein that associates with the immunopurified septin complex, it seems that, compared to the rods of other species, something about these preparations and/or how they were spread, dehydrated and stained for viewing in the EM resulted in immobilization and accentuation of the CTEs so they could be visualized, even in raw (unaveraged) images. If so, these ∼10-nm-long projections confirm that the CTEs of septins bound in the NC mode do pair with each other. The fact that these appear at 16-nm intervals (i.e. four 4-nm subunits apart) suggests that, at least for the constellation of different septin monomers of which the rat brain complex was composed, the CTEs only pair up (or are only visible) for every other NC pair. Although the crystal structure of the human septin hexamer revealed nothing about the structure of the CTEs beyond α6, it was a surprise that stoichiometric complexes could still be prepared by the same expression and purification method using derivatives of all three septins lacking their CTEs (namely, SEPT2(
309-361), SEPT6(
311-434) and SEPT7(
306–437)) and that the hexameric rods so formed did not look significantly different in the EM from the rods prepared with the full-length version of each protein (Sirajuddin et al., 2007). Thus, for these human septins, the additional contacts provided by CTE–CTE interaction are dispensable for complex formation. Likewise, budding yeast Cdc10 lacks any CTE (Plate Figure 3.3c), and a Cdc10-Cdc10 dimer in the NC orientation presumably occupies the centre of the budding yeast rod (Figure 3.1c and Plate Figure 3.3b). The Cdc3-Cdc12 association in the budding yeast rod is also predicted to be a NC interface (Plate Figure 3.3b); but, removal of the CTE from either protein ablates their interaction and eliminates hetero-octamer formation both in vivo and in vitro (Versele et al., 2004). Thus, at least in budding yeast, and at least for this pair, contacts mediated by the corresponding CTEs are essential for this NC interaction, as mentioned above. The importance of the CTEs in these two septins in vivo is further emphasized by the fact that CTE-less alleles of either Cdc3 or Cdc12 act in a dominant-negative fashion to disrupt assembly

74

CH03 BIOCHEMICAL PROPERTIES AND SUPRAMOLECULAR ARCHITECTURE

and function of the filaments at the septin collar when each is expressed in an otherwise wild-type cell (Versele et al., 2004). How could the CTEs be so important for yeast septin complex formation, when the CTEs of the human septins are dispensable? In this and presumably other cases, the devil is likely in the details. Synthetic peptides corresponding to the heptad repeat segments of the nematode septins that are predicted to have coiled coil–forming capability do associate to form a stable, parallel coiled coil in vitro (John et al., 2007). Using a similar approach, the corresponding regions of Cdc3 and Cdc12 showed no evidence of any stable interaction (Versele et al., 2004). Indeed, as already mentioned above, the preceding helical regions (‘α7’) (Figure 3.2 and Plate Figure 3.3c) in the CTEs of both Cdc3 and Cdc12 must also be present for the interaction of these two septins to occur. Thus, this short extra helix is a key component of the NC interaction surface between Cdc3 and Cdc12. Aside from those septin–septin interactions, like that of Cdc3 with Cdc12, where the CTEs clearly provide contacts needed for forming the NC interface, what else, if anything, might CTE–CTE pairing contribute for septin rod and filament ultrastructure? Based on the orientations of the α6 helices in the human heterohexamer (Sirajuddin et al., 2007), there are, theoretically, three CTE–CTE pairs that could project off the same face of a rod (Plate Figure 3.3a). These elements could themselves act as an interaction surface. Such ‘molecular Velcro’ provides an obvious mechanism for recruitment of other non-septin proteins; in this manner, septin filaments could act as scaffolds to bind and organize other structural and/or enzymic functions at the specific subcellular locations where septin complexes and filaments assemble. In animal cells, the Borg family of small Cdc42-binding adapter proteins binds to the CTE–CTE element formed by SEPT6-SEPT7 in the NC orientation (Joberty et al., 2001; Sheffield et al., 2003). Also, it should be recalled that, regardless of the species, when rods polymerize, the filaments so formed have a strong tendency to form parallel ‘railroad tracks’ (Figure 3.1a, middle and bottom). Clearly, if the CTE–CTE interaction forms a parallel heterodimeric coiled coil, then the ‘ties’ in these ‘railroad tracks’ presumably represent interaction between the heterodimeric coiled coil from one filament with the heterodimeric coiled coil from the other filament, that is, formation of an anti-parallel heterotetrameric coiled coil (a four-helix bundle). Four characteristics of the filament pairs seen in the EM are consistent with the notion that the CTE–CTE elements serve as cross-bridges. First, filaments tend to align almost invariably in pairs (Figure 3.1a, middle), implying that the relevant lateral interactions occur between interfaces on only one side of each filament. Second, paired filaments are in register (Figure 3.1a, bottom), as expected if pairing only occurs via interaction at specific points along the filament. Third, the separation between filaments is reasonably uniform over rather long distances and this spacing is of sufficient width to be mediated by contact between CTE–CTE pairs if they extend orthogonally off each filament. Finally, at higher magnification, density spanning

SEPTIN COMPLEXES IN VERTEBRATES

75

the gap between two paired filament with the appropriate periodicity can sometimes be discerned in negatively stained EM images (Bertin et al., 2008). On the other hand, when separated from the context of their attachment to their cognate G domain and the rest of their CTE, the coiled coil segments of the worm and mammalian septins tested showed no evidence of anti-parallel tetramerization of the parallel heterodimeric coiled coils (Low and Macara, 2006; John et al., 2007). Moreover, the heptad repeats in the corresponding sequences are not of the 3-3-1 type that is typically diagnostic of those able to form four-helix coiled coil bundles (Deng et al., 2006). Thus, the disposition of the CTEs in septin rods and filaments remains ill-defined. The yeast septin octamer, despite having two additional subunits, has fewer possible CTE–CTE pairings than the mammalian and insect hexamers (compare Plate Figure 3.3b with 3.3a) because Cdc10 lacks a CTE, and the CTE of a terminal Cdc11 subunit is presumably unpaired unless and until it associates with its sister on another rod. Thus, the CTE of Cdc11 would seem well-positioned to play an important role in mediating rod polymerization via Cdc11-Cdc11 interaction (and via interaction of Cdc11-capped rods with Shs1-capped rods, given that a CTE-dependent interaction between Cdc11 and Shs1 has been observed between the purified proteins (Versele et al., 2004)). However, yeast cells expressing a CTE-less derivative of Cdc11 as the sole source of this septin are viable, whereas a cdc11
null mutation causes a much more severe phenotype (Casamayor and Snyder, 2003) or is outright lethal (Versele et al., 2004), depending on the strain background. On the other hand, based on arguments presented above (and earlier), these effects could reflect the degree to which Shs1 has the capacity in vivo to fully substitute for Cdc11, in a given strain background. Nonetheless, salt-stable octamers can be prepared when CTE-less Cdc11 is co-expressed with Cdc3, Cdc10 and Cdc12, and these are capable of forming filaments in vitro when the salt concentration is lowered (Bertin et al., 2008). Interestingly, whereas Cdc11, like Cdc3 and Cdc12, possesses in its CTE a stretch of residues preceding its presumptive coiled coil–forming segment that are predicted to have α-helix-forming propensity (Figure 3.2 and Plate Figure 3.3c), Shs1 seems to lack this feature. Instead, Shs1 has a predicted β-strand at this location and, moreover, has a CTE that is, overall, nearly 100 residues longer than that of the mitotic septin with the next longest CTE (Cdc11). Because it lacks the α7 helix, Shs1-Shs1 (or Shs1-Cdc11) association in the NC orientation may require interactions that are unique to its apparently highly specialized CTE. Important evidence for a functional role of the coiled coil–forming region in a CTE in helping to stabilize rod and filament structure in vivo comes from the properties of a particular allele (cdc12-6ts ) in the CDC12 gene. Upon shift to the non-permissive temperature, septin-containing structures in cells carrying the cdc12-6 allele rapidly disassemble, accompanied by complete delocalization of all mitotic septins (Ford and Pringle, 1991; Kim, Haarer and Pringle, 1991). These effects are not due to destruction of the mutant protein at the restrictive temperature because the steady-state level of Cdc12-6 is equivalent to that of wild-type

76

CH03 BIOCHEMICAL PROPERTIES AND SUPRAMOLECULAR ARCHITECTURE

Cdc12 under the same conditions (McMurray and Thorner, unpublished results). By contrast, in cells carrying another temperature-sensitive allele, cdc12-1 , further assembly of new structures does not occur, but preformed septin structures do not unravel, after shift to the non-permissive temperature (Dobbelaere et al., 2003). Revealingly, the cdc12-6 allele arises from a frameshift mutation that results in the last 16 residues at the C terminus of Cdc12 being replaced with twenty residues of abnormal sequence (Longtine et al., 2000). This alteration is located far from α7 and only affects the distal half or so of the predicted coiled coil–forming segment in the Cdc12 CTE, but, fortuitously, disrupts only partially the 4-3 heptad repeat in this region (Figure 3.2). It is very telling that such a seemingly subtle change at the extreme C-terminal end of the protein has such a profound affect on the stability of preformed septin complexes in the cell and, thus, supports the conclusion that the coiled coil between the CTEs of Cdc12 and Cdc3, although not essential for their interaction, nonetheless contributes greatly to the stability of this interface once formed. The mutational alteration in the cdc12-1 allele, by comparison, is a G227E substitution (corresponding to residue 241 in human SEPT2), which should affect contacts important for interaction of the G domain of Cdc12 with the terminal rod subunit, Cdc11, thereby explaining its phenotype.

Guanine nucleotide binding and hydrolysis in septin complex assembly The crystal structure of the human septin heterohexamer was of sufficient resolution to discern that GDP is bound to each SEPT2 monomer in the central SEPT2-SEPT2 N-C dimer, whereas the next subunit associated with SEPT2 via a G domain-G domain interaction, namely SEPT6, has GTP bound and the terminal subunit that interacts with SEPT6 in NC orientation, namely SEPT7, has GDP bound (Plate Figure 3.3a). The observed polymerization of hexamers into filaments is undoubtedly mediated by end-on-end, that is, SEPT7-SEPT7, interaction. Thus, at both homotypic septin–septin junctions (SEPT2-SEPT2 and SEPT7-SEPT7), GDP is the nucleotide present in both partners; but, at the heterotypic septin–septin junction (SEPT2-SEPT6), GTP is bound to one partner. These observations and others summarized below suggest that the nature of the nucleotide bound plays a role in both multi-septin complex formation and filament assembly. When expressed individually in bacterial cells, many purified septins are insoluble, aggregated and nucleotide-free (Versele and Thorner, 2004; Hu et al., 2006; Garcia et al., 2007). By contrast, when purified from their native source or when co-expressed with partner septins known to be components of the same physiological complex, the resulting septins are typically soluble and saturated with bound nucleotide (Farkasovsky et al., 2005; Sirajuddin et al., 2007). At the very least, these and other biochemical findings indicate that the presence of a bound guanine nucleotide stabilizes the folded state of every septin examined to date. However, some of the available biochemical data are conflicting and/or subject to multiple interpretations.

SEPTIN COMPLEXES IN VERTEBRATES

77

In agreement with the crystal structure, native septin complexes contain protein and nucleotide in equimolar amounts and a ratio of GDP-to-GTP of ∼2 : 1 (Field et al., 1996; Vrabioiu et al., 2004). One interpretation of these findings is that SEPT2 and SEPT7 bind GDP preferentially, whereas SEPT6 is selective for GTP. An alternative possibility is that all bind GTP, but SEPT2 and SEPT7 can hydrolyse the bound GTP to GDP, whereas SEPT6 is unable to do so. In this regard, it has been observed that both bacterially expressed Cdc10 (which occupies a position in the yeast hetero-octamer equivalent to that occupied by SEPT2 in the human heterohexamer) and bacterially expressed Cdc12 (which occupies a position in the yeast hetero-octamer akin to that occupied by SEPT7 in the human heterohexamer) are able to bind exogenously provided GTP and hydrolyse it (albeit at a slow rate) to GDP (Versele and Thorner, 2004; Farkasovsky et al., 2005). By contrast, when purified in the same way, neither Cdc11 nor Cdc3 (which occupies a position in the yeast hetero-octamer akin to that occupied by SEPT6 in the human heterohexamer) was able to bind exogenously supplied GTP (Versele and Thorner, 2004; Farkasovsky et al., 2005), presumably because each already contained a non-exchangeable guanine nucleotide bound to their G domains. Moreover, mutants of Cdc10 and Cdc12 unable to bind nucleotide still assemble, albeit much less efficiently (Bertin et al., 2008), into complexes with wild-type Cdc3 and Cdc11 and, in such complexes, GTPase activity is almost totally eliminated (Versele and Thorner, 2004), suggesting that, even though GTP may bind to all four of the core yeast septins initially, Cdc10 and Cdc12 are the only two septins in the yeast rod capable of exchanging and subsequently hydrolysing GTP to GDP. In this same regard, the interaction between SEPT2(GDP) and SEPT6(GTP) in the human heterohexamer is clearly more salt-resistant than the SEPT7(GDP)SEPT7(GDP) interaction required for the end-on-end polymerization of rods into filaments. The SEPT2-SEPT6 interaction occurs via a G domain-G domain interface (Plate Figure 3.3a). The parts of a GTP-binding protein that undergo the greatest conformational changes upon nucleotide binding, and are the most different depending on whether the nucleotide is GDP or GTP, are located in the G domain and are dubbed its ‘switch regions’ (Vetter and Wittinghofer, 2001). It was noted in the crystal structure that the switch regions at the SEPT2-SEPT6 interface are better defined than elsewhere in the human rod (Sirajuddin et al., 2007). Taken together, all of these data argue that the state of the bound nucleotide at a given interface could play a role in stabilizing certain subunit–subunit combinations, and also raise the idea that GTP hydrolysis by certain septins and not others is an important step in ordered complex assembly. There is at least some additional experimental evidence in support of these notions, although each is fraught with a potential caveat(s). Formation of a heterodimer from independently expressed and purified SEPT6 and SEPT7 monomers is associated with both binding and hydrolysis of GTP (Sheffield et al., 2003). These findings are at least compatible with the idea that GTP binding and hydrolysis by at least one of these two septins induces the conformational state that

78

CH03 BIOCHEMICAL PROPERTIES AND SUPRAMOLECULAR ARCHITECTURE

is competent for their interaction, even though isolated SEPT6 and SEPT7 presumably also interact, as they do in the hexamer, via a NC-interface (not via their G domains) (Low and Macara, 2006; Sirajuddin et al., 2007). Somewhat similarly, when expressed in and purified from bacteria, Xenopus laevis Sept2 is soluble and nucleotide-free; however, when provided with GTP (or a non-hydrolysable GTP analog), it polymerizes into filaments (Mendoza, Hyman and Glotzer, 2002). However, this filament assembly was preceded by a significant lag phase. Hence, even though the frog Sept2 prepared by bacterial expression was soluble, it could be argued, as others have (Mitchison and Field, 2002), that it was partially unfolded and that GTP merely promotes its stable folding. In other words, this same argument goes, the nature of the guanine nucleotide doesn’t matter for Sept2-Sept2 interaction and filament formation, but only that Sept2 be in its native conformation. Much evidence does suggest that continuous cycles of GTP binding, hydrolysis, product release and rebinding of fresh GTP by the septins themselves are not involved in either complex formation or filament assembly (or disassembly). For example, using liquid chromatography and tandem mass spectrometry to measure the rate of dilution of the 15 N label in septin-associated guanine nucleotides in vivo, it was concluded that the majority of the septin-bound guanine nucleotides do not turn over during the period of a one yeast cell cycle. On the other hand, rapid exchange into one or two of the subunits within an octamer, or into a small population of the rods in the total cellular pool, could have been missed by this kind of bulk measurement (Vrabioiu et al., 2004). In fact, in these in vivo studies guanine nucleotide exchange was only detected in Cdc10 and Cdc12 (Vrabioiu et al., 2004), in agreement with the aforementioned in vitro studies demonstrating that only Cdc10 and Cdc12 show appreciable GTP binding and hydrolysis (Versele and Thorner, 2004; Farkasovsky et al., 2005). The conclusion that neither hydrolysis nor exchange of bound nucleotide is likely to play a regulatory role in the dynamic changes in septin organization that occur during the cell cycle is also consistent with the very slow rate of GTP hydrolysis displayed by isolated Cdc10 and Cdc12 (or the hetero-octameric complexes that contain them) (Versele and Thorner, 2004; Farkasovsky et al., 2005). Nonetheless, such findings certainly do not rule out, as emphasized earlier in this section, that a one-time GTP-binding event is an important step in ensuring that septins are fully folded, in the correct conformational state for properly ordered septin complex assembly, and competent for rod–rod polymerization. It has been difficult, however, to design experimental strategies to nail down this latter point unequivocally. Unlike isolated native complexes, individual septins or septin sub-complexes expressed in and purified from heterologous systems contain variable amounts of guanine nucleotide (and different ratios of GDP-to-GTP), which do not correlate well with the expression system used or the septins examined. For example, ∼60 % of human SEPT2 molecules produced in bacteria contain bound nucleotide, with GDP-to-GTP in a ∼1 : 1 ratio (Huang et al., 2006), whereas, it should be recalled, bacterially expressed frog Sept2 is nucleotide-free

SEPTIN COMPLEXES IN VERTEBRATES

79

(Mendoza, Hyman and Glotzer, 2002)]. Moreover, when expressed in and purified from insect cells, essentially all the human SEPT2 molecules contain bound nucleotide, with GDP-to-GTP in a ∼10 : 1 ratio (Huang et al., 2006). Likewise, yeast septins expressed in E. coli can display different nucleotide- and septin-binding behaviours when expressed individually versus when they are co-expressed with their partner septins (Versele et al., 2004; Versele and Thorner, 2004; Farkasovsky et al., 2005). Also, as already noted above, measuring the ability of a septin or septin complex to bind and hydrolyse guanine nucleotide requires exchange of any bound nucleotide with the exogenously added labelled nucleotide. Thus, only purified septins that are either nucleotide-free, or have an appreciable rate of nucleotide exchange, can be analysed. The same is true for any given subunit within a complex. Conversely, the subunits that bind nucleotide the tightest are most likely to be already saturated with it, exchange it slowly, and thus be much harder to study with regard to their nucleotide binding and hydrolysis properties. With these considerations in mind, it seems plausible that, at least for the yeast hetero-octamer, Cdc10 and Cdc12 are the most reasonable candidates for subunits in which conversion of bound GTP to GDP may be involved in rod assembly. However, single amino acid substitutions in Cdc10 and Cdc12 that do not prevent GTP binding, but do drastically compromise hydrolysis of the bound GTP in vitro, have absolutely no phenotype when expressed as the sole source of either or both of these septins in vivo (Versele and Thorner, 2004). Nevertheless, when modelled on the state of the bound guanine nucleotides in the subunits of the human hexamer, and considering the >2 : 1 GDP : GTP ratio reportedly found in purified yeast octamers (Farkasovsky et al., 2005), it seems likely that, in a yeast rod, only Cdc3 is present in the GTP-bound state and the other three are GDP bound (Plate Figure 3.3b). Thus, even if Cdc10, Cdc11 and Cdc12 bind and hydrolyse GTP, once incorporated into the rod, these subunits do not readily release the resulting bound GDP. Consistent with the need for a bound nucleotide for optimal septin complex assembly and rod stability, single amino acid substitutions that prevent nucleotide binding to Cdc10 and Cdc12 (in contrast to the mutations that prevent hydrolysis of GTP once bound), interfere with efficient hetero-octamer formation when these Cdc10 and Cdc12 mutants are co-expressed with wild-type Cdc3 and Cdc11 in bacterial cells (Bertin et al., 2008) and show a readily detectable and synergistic temperature-sensitive phenotype when expressed in vivo as the sole source of Cdc10 and Cdc12. Specifically, a double mutant of two GTP binding-defective alleles, cdc10(S46N) cdc12(T48N), grows slowly and has drastically elongated buds at 30◦ and is inviable at 37◦ , whereas a double mutant of two GTP-binding competent, but GTP hydrolysis-defective alleles, cdc10(S41V) cdc12(S43V), grows well and shows no abnormality at either temperature (Versele and Thorner, 2004). Moreover, the recombinant doubly mutant rods that can be isolated, that is, Cdc11-Cdc12(T48N)-Cdc3-Cdc10(S46N)-Cdc10(S46N)-Cdc3-Cdc12(T48N)Cdc11 are much less stable than wild-type rods in solution and incapable of polymerizing into filaments (Versele et al., 2004; Bertin et al., 2008). Thus,

80

CH03 BIOCHEMICAL PROPERTIES AND SUPRAMOLECULAR ARCHITECTURE

optimal rod assembly and rod stability require guanine nucleotide binding to at least Cdc10 and Cdc12. However, the opposite conclusion was reached for mutations that putatively affect nucleotide binding by Cdc11 (Casamayor and Snyder, 2003); cdc11 alleles designed to incapacitate nucleotide binding had little discernible functional consequence in vivo (Casamayor and Snyder, 2003). However, it could not be convincingly verified experimentally that the mutations made had the anticipated effects on GTP binding. When modelled on the crystal structure of the human hexamer, these mutational alterations are predicted to perturb the G-domain interface of Cdc11, which would prevent its association with Cdc12. Cdc11-less rods are unable to polymerize into filaments (Versele et al., 2004; Bertin et al., 2008), probably explaining the phenotype of these mutants in vivo. The G domains of Cdc10 and Cdc12 do not participate directly at the interface required for end-on-end polymerization of rods. Thus, the fact that the Cdc11-Cdc12(T48N)-Cdc3-Cdc10(S46N)-Cdc10(S46N)-Cdc3-Cdc12(T48N)Cdc11 double mutant complex fails to polymerize suggests that guanine nucleotide-dependent conformational changes in Cdc10 and Cdc12 are propagated outward and influence the terminal subunits of the octamer. This does not seem far-fetched, at least for Cdc12, because this subunit directly abuts Cdc11 at each end of the rod. In any event, to address these same issues more incisively in the near future, it should be possible to pinpoint, based on the now available crystal structure, residues that when altered will compromise GTP binding (or hydrolysis) without unduly affecting contacts crucial for the G-domain interface.

Post-translational modifications of septins and identification of other important residues For the reasons discussed in the preceding section, septins are unlikely to behave as ‘molecular oscillators’ driven by repeated cycles of nucleotide binding and hydrolysis. Yet, septin-based structures clearly undergo a series of marked structural rearrangements during progression though each cell cycle that need to be properly regulated both spatially and temporally. What processes control these dynamic transitions? Post-translational modifications of the septins are good candidates to regulate the ultrastructure and biochemical properties of rods and filaments in signal-dependent manner. Indeed, septins are heavily modified, especially by both phosphorylation and SUMOylation (Table 3.1), and the locations of these modifications modelled on the crystal structure of the human hexamer are potentially revealing about their conceivable effects. Some of these positions are indicated in Plate Figure 3.3a, along with the locations of other residues implicated in septin function for other reasons (e.g. a cellular phenotype results from mutation of that residue). A few selected examples are discussed below. Phosphorylation of Cdc10 on Ser256 by the Cdc42-regulated protein kinase Cla4 is thought to promote filament assembly (Kadota et al., 2004; Versele and

α1, near NC interfaceb N/A Exposed disordered loop between α1 and β2

SEPT2 K30

SEPT2 S31

SEPT2 G44, G47, G49, K50

SEPT2 S51

SEPT2 N55

HsSEPT2 V/I67d

ScCdc11 R12, K13, R14, K15, H16

ScCdc3 S113

ScCdc11 G29, G32, G34, R35

ScCdc10 S46, ScCdc12 T48

ScCdc11 N40

RnSEPT3 S91

P-loop, near G interfaceb

P-loop, G interface

α0, NC interfaceb

Phosphorylation (cGMP-dependent protein kinase-I

N/A

Phosphorylation (PAK family kinase Cla4) N/A

Extreme N terminus Phosphorylation (disordered)b (unknown) N/Ac Extreme N terminus (‘polybasic motif’)b

SEPT2 Y17

HsSEPT2 Y17

Modification (modifier)

Equivalent/ Location representative residue(s) in human septin of solved structure

Residue(s)a

Subcellular localization? (phospho-formspecific localization)

(continued overleaf )

(Casamayor and Snyder, 2003) (Xue et al., 2004)

(Versele and Thorner, 2004; Farkasovsky et al., 2005)

(Zhang et al., 1999; Casamayor and Snyder, 2003) (Versele and Thorner, 2004) (Casamayor and Snyder, 2003)

Membrane interaction (phospholipid binding in vitro) Unknown Nucleotide binding/hydrolysis (behaviour of purified mutant) Nucleotide binding, filament assembly (behaviour of purified mutant, cellular phenotype) Unknown

(Rush et al., 2005)

Reference(s)

Unknown

Predicted function (evidence)

Table 3.1 Locations of residues in septin proteins known or predicted to be post-translationally modified, and those with effects on septin structure or function when mutated

SEPTIN COMPLEXES IN VERTEBRATES 81

SEPT2 H140

SEPT2 F156

SEPT2 A182

SEPT2 D185

HsSEPT2 T186, T188

SEPT2 G241

ScCdc3 S244

HsSEPT2 F156

ScCdc10 G179

ScCdc10 D182

ScCdc12 T183, T185; ScShs1 S221

ScCdc11 G230, ScCdc12 G227

N/A

Phosphorylation (Mec1,Tel1, and/or Rad53 for ScShs1 S221) N/A

G interfaceb

Between β5 and α4, G interface

Near G interfaceb

N/A

Phosphorylation (PAK family kinase Cla4) N/A

Modification (modifier)

G interface

Disordered residue near NC interface G interfaceb

Equivalent/ Location representative residue(s) in human septin of solved structure

(continued )

Residue(s)a

Table 3.1

Hetero-oligomer assembly (chromatographic behaviour of purified mutant) Hetero-oligomer assembly (cellular phenotype)e Hetero-oligomer assembly (cellular phenotype)e DNA damage response for ScShs1 S221 (cellular phenotype) Filament assembly but not stability (cellular phenotype)f

Unknown

Predicted function (evidence)

(Casamayor and Snyder, 2003; Dobbelaere et al., 2003)

(Smolka et al., 2006; Smolka et al., 2007)

(Hartwell, 1971)

(Cid et al., 1998)

(Versele and Thorner, 2004) (Sirajuddin et al., 2007)

Reference(s)

82 CH03 BIOCHEMICAL PROPERTIES AND SUPRAMOLECULAR ARCHITECTURE

SEPT2 S218

Not conserved, near SEPT2 S218 SEPT2 T228, SEPT7 T228

SEPT2 W260

SEPT2 G261

SEPT2 H270

SEPT7 Y319

SEPT7 S334

HsSEPT2 S218, RnSEPT2 S218, MmSEPT4 S325

ScCdc10 T216

MmSEPT7 T227; RnSEPT2 T228

HsSEPT2 W260

ScCdc10 S256

HsSEPT2 H270

HsSEPT7 Y319

HsSEPT7 S334; MmS333

Phosphorylation (PAK family kinase Cla4) N/A

β6, G interfaceb

Disordered, between Phosphorylation α6 and coiled coil (unknown) Disordered coiled coil Phosphorylation (unknown)

G interfaceb

N/A

G interfaceb

(Hoffert et al., 2006, 103, p. 7159; Dai et al., 2007) (Sirajuddin et al., 2007)

(Beranova-Giorgianni et al., 2006; Hoffert et al., 2006; Huang et al., 2006; Nousiainen et al., 2006; Olsen et al., 2006; Dai et al., 2007; Molina et al., 2007) (Li et al., 2007)

Unknown

(continued overleaf )

(Olsen et al., 2006; Munton et al., 2007)

Hetero-oligomer assembly (chromatographic behaviour of purified mutant) Filament assembly (Versele and Thorner, (cellular phenotype) 2004) Hetero-oligomer (Sirajuddin et al., 2007) assembly (chromatographic behaviour of purified mutant) Unknown (Rush et al., 2005)

Unknown

Unknown

Phosphorylation (casein Nucleotide binding kinase II) (behaviour of purified mutant)

Disordered loop Phosphorylation between α4 and α5’ (unknown) Phosphorylation Exposed α5’, near (unknown) G interfaceb

Disordered loop between α4 and α5’, near NC interface

SEPTIN COMPLEXES IN VERTEBRATES 83

(continued )

Not conserved

Not conserved

Not conserved

Not conserved

Not conserved

Not conserved

ScCdc3 S503g

ScCdc3 S509g

ScCdc10 S312

ScCdc10 H314

ScCdc10 S316

ScCdc11 S305, S316, S318, S319, S327

Phosphorylation (ATR/ATM?) SUMOylation (SUMO-protein ligase Siz1) Phosphorylation (cyclin-dependent kinase) Phosphorylation (cyclin-dependent kinase) Phosphorylation (PAK family kinase Cla4)

Extreme C terminus, disordered Extreme N terminus

Phosphorylation (unknown) Exposed C terminus Phosphorylation (unknown) Between α6 and Phosphorylation predicted coiled coil (unknown)

Exposed C terminus

Exposed C terminus

Extreme C terminus

Extreme C terminus

Phosphorylation (unknown)

Modification (modifier)

Extreme C terminus, disordered

Equivalent/ Location representative residue(s) in human septin of solved structure

HsSEPT7 S423, T426; SEPT7 S423, T426 MmSEPT7 T425; RnSEPT7 T425 HsSEPT6 S408, S411, SEPT6 S408, S411, S416, T418; S416, T418 MmSEPT6 S416 ScCdc3 K4, K11, Not conserved K30, K63

Residue(s)a

Table 3.1

(Hoffert et al., 2006; Dai et al., 2007; Molina et al., 2007) (Matsuoka et al., 2007; Molina et al., 2007; Munton et al., 2007) (Johnson and Blobel, 1999)

Reference(s)

Unknown

Unknown

(Chi et al., 2007; Li et al., 2007; Smolka et al., 2007)

(Smolka et al., 2007)

Filament disassem(Tang and Reed, 2002) bly/reorganization (cellular phenotype) Filament disassem(Tang and Reed, 2002) bly/reorganization (cellular phenotype) Unknown (Versele and Thorner, 2004; Smolka et al., 2007) Unknown (Chi et al., 2007)

DNA damage response (role of modifier) Unknown

Unknown

Predicted function (evidence)

84 CH03 BIOCHEMICAL PROPERTIES AND SUPRAMOLECULAR ARCHITECTURE

Not conserved

HsSEPT9 S82, S85, S89

Not conserved

Not conserved

Not conserved

HsSEPT9 T42

Not conserved

ScShs1 S519, S525, S539, S545, S548 MmSEPT8 A2; HsSEPT11 A2 MmSEPT8 S10

Not conserved

Extreme N terminus

Not conserved

ScShs1 S408, S416

Not conserved

Extreme N terminus

Not conserved

ScShs1 S447, S460

HsSEPT9 S30; MmSEPT9 S30 RnSEPT9 T31i

Extreme C terminus

Not conserved

ScShs1 K426, K437

Phosphorylation (unknown) Unknown Phosphorylation (unknown) Unknown Phosphorylation (unknown) Unknown Phosphorylation (unknown) Near residues affected Phosphorylation (unknown) in HNj

Extreme C terminus

Extreme C terminus

Extreme C terminus

Extreme C terminus

Not conserved

ScCdc11 K412

SUMOylation (SUMO-protein ligase Siz1) SUMOylation (SUMO-protein ligase Siz1) Phosphorylation (unknown) Phosphorylation (unknown) Phosphorylation (unknown) Acetylation (unknown)

Between α6 and N/A predicted coiled coil

ScCdc12 K351, E368h Not conserved

(Olsen et al., 2006; Molina et al., 2007)

Unknown

(continued overleaf )

(Beausoleil et al., 2006)

(Beausoleil et al., 2006; Dai et al., 2007) (Hoffert et al., 2006)

(Gruhler et al., 2005; Dai et al., 2007) (Dai et al., 2007; Smolka et al., 2007) (Dai et al., 2007; Smolka et al., 2007) (Gevaert et al., 2003; Trinidad et al., 2006) (Trinidad et al., 2006)

(Johnson and Blobel, 1999)

(Johnson and Blobel, 1999)

Unknown

Unknown

Unknown

Unknown

Unknown

Unknown

Unknown

Unknown

Unknown

Interaction with chaperone/disaggregase Hsp104? (genetic suppression) Unknown

SEPTIN COMPLEXES IN VERTEBRATES 85

Not conserved

Not conserved

Not conserved

HsSEPT9 R88, S93F

HsSEPT9 T142

RnSEPT9 T150

Phosphorylation (unknown) Phosphorylation (unknown)

N/A

Modification (modifier)

Unknown

Septin–septin interaction (cellular co-localization phenotype)j Unknown

Predicted function (evidence)

(Hoffert et al., 2006)

(Molina et al., 2007)

(Sudo et al., 2007)

Reference(s)

a Amino acid residues are identified using the standard one-letter abbreviation. The identity of the septin protein in question is given with a two-letter prefix identifying the host species (e.g. ‘Sc’, Saccharomyces cerevisiae). b Indicates the (predicted) location of this residue is shown in Plate Figure 3.3. c N/A, not applicable. d The sequence of HsSEPT2 presented in (Sirajuddin et al., 2007) includes an isoleucine at position 67 rather than the valine predicted by all available sequences in public databases. e Mutation of ScCdc10 at G179 (in the cdc10-11 allele) causes a temperature-sensitive phenotype, and the inability of the mutant to assemble at the bud neck in the presence of wild-type ScCdc10, even at the permissive temperature (Cid et al., 1998). Both phenotypes also apply to the cdc10-1 allele, isolated by Hartwell (1971) and determined by sequencing to harbour the substitution D182N (M. A. McMurray and J. Thorner, unpublished data). f The temperature-sensitive cdc12-1 allele identified by Hartwell (1971) encodes a mutant ScCdc12 with the substitution G227E (Casamayor and Snyder, 2003). When shifted to the non-permissive temperature, yeast cells carrying this mutation cannot assemble new septin structures, but existing structures are not affected (Dobbelaere et al., 2003). Mutation at the equivalent position in ScCdc11 (G230) also renders cells temperature-sensitive (Casamayor and Snyder, 2003). g Residue is conserved in HsSEPT10. h ScCdc12 mutations K351N or E368Q eliminate the adverse effects on septin organization and function of overexpression of Hsp104G17S T499I , an allele of the chaperone/disaggregase harbouring mutations in both the coiled coil–forming middle region, and in the nucleotide-binding domain (Schirmer et al., 2004). i Equivalent to HsSEPT9 T49. j Mutations R88W and S93F in the uniquely extended N terminus of SEPT9 are associated with hereditary neuralgic amyotrophy (HN) (Kuhlenbaumer et al., 2005). The mutant forms appear to interact with certain other septins, whereas the wild type does not (Sudo et al., 2007).

Unknown

Unknown

Extreme N terminus

Equivalent/ Location representative residue(s) in human septin of solved structure

(continued)

Residue(s)a

Table 3.1

86 CH03 BIOCHEMICAL PROPERTIES AND SUPRAMOLECULAR ARCHITECTURE

SEPTIN COMPLEXES IN VERTEBRATES

87

Thorner, 2004). Moreover, the effect of this modification is thought to function in conjunction with the effect that nucleotide binding has on Cdc10 and Cdc12 for three reasons (Versele and Thorner, 2004). The phenotype of cdc10(S46N) cdc12(T48N) double mutant cells is ameliorated by overexpression of CLA4 specifically. Conversely, the phenotype of cdc10(S46N) cdc12(T48N) double mutant cells is markedly exacerbated by a cla4
mutation. Finally, a cdc10(S46N S256A) double mutant has a more severe phenotype that either a cdc10(S46N) or a cdc10(S256A) single mutant. Interestingly, Ser256 (equivalent to G261 in human SEPT2; see Table 3.1), while not conserved as a phosphorylatable residue in any of the three human subunits whose structures have been solved, is predicted to be located near the G-domain interface and at a position where phosphorylation could conceivably influence the strength of the G-dimer interaction (Plate Figure 3.3a). Accordingly, the severe consequences on septin function of combining the nucleotide binding mutations with the non-phosphorylatable substitution at S256A probably stem from additive effects that weaken the G-dimer interaction between Cdc10 and Cdc3. Another phosphorylation event that potentially affects a G-dimer interface is Ser221 in Shs1 (equivalent to T186 in human SEPT2; see Table 3.1) and its counterparts (T183 and/or T185) in Cdc12. Phosphorylation of Shs1 on S221 occurs in response to DNA replication stress and appears to play a role in this Rad53 (human ortholog is CHK2)-dependent checkpoint (Smolka et al., 2006). Hence, Shs1 is presumably the target of one of the kinases activated in response to the DNA damage that occurs when DNA replication is compromised (Smolka et al., 2007). Based on its location, it is tempting to speculate that phosphorylation of Shs1 at this position diminishes its ability to interact with Cdc12 and that phosphorylation of Cdc12 at the equivalent position reinforces this effect, perhaps by charge repulsion between the two negatively charged phosphate groups. By extension of this idea, phosphorylation of SEPT2 at T186 may affect its interaction with SEPT7, and might be a consequence of the DNA damage response in human cells, in which septins have recently been shown to function (Kremer, Adang and Macara, 2007). Many similar, intriguing and readily testable predictions can be generated by comparing the locations and functions of specific septin residues between different species, now made possible by our new level of understanding of septin structure and about the organization of septin rods.

Lipophilia and the structural basis for septin–membrane interaction The 10-nm filaments at the yeast bud neck are closely apposed to the plasma membrane (Figure 3.4a, right) (Byers and Goetsch, 1976a). Moreover, it has been noted that, upon fractionation of cell extracts, septins often co-purify with the membrane-containing material under conditions in which it would be expected that interaction with other factor(s) that might mediate septin-membrane interaction would be disrupted (Xie et al., 1999). Indeed, during a typical preparation

88

CH03 BIOCHEMICAL PROPERTIES AND SUPRAMOLECULAR ARCHITECTURE

from animal cells, the soluble septin complexes are retained, whereas those in the membrane fraction are removed. For example, in the preparation of Pnutcontaining septin complexes from D. melanogaster (Field et al., 1996), ∼50 % of the total Pnut was in the particulate fraction and was discarded. It is at least conceivable that the membrane-associated septin complexes could be different in a significant way from those in the soluble fraction. Nonetheless, it is generally accepted that septin complexes and filaments interact directly with membranes. This interaction could influence, and certainly places constraints on, the ultrastructure of septin filaments and may facilitate or preclude interactions between any given septin in a filament and other proteins. Conversely, as has been pointed out by others, such membrane-associated septin ‘fences’ can act as barriers to the diffusion of integral membrane proteins (Takizawa et al., 2000; Dobbelaere and Barral, 2004). At first glance, the curvature of the plasma membrane itself seems an ideal candidate to act as a template figure to direct assembly of otherwise linear septin rods and filaments into ring-shaped structures. However, some purified septin complexes display ring-forming ability even in the test tube (Kinoshita et al., 2002; Farkasovsky et al., 2005). This property of self-closure may be related to those septin rods that possess a ‘kink’ observable by EM, such as seen in purified preparations of the recombinant human SEPT2-SEPT6-SEPT7 hexamer (Sirajuddin et al., 2007). If the kink were always on the same side of each rod in a filament, it would bend in a curved arc even in the absence of membrane structures. Furthermore, septins can form similarly shaped rings in cells that have drastically different membrane curvatures. For example, the septin rings formed at the convex plasma membranes of ovoid S. cerevisiae cells are indistinguishable from those formed at a concave membrane when budding occurs at the isthmus of a dumbbell-shaped zygote (Cid et al., 2001). Indeed, at least in part by virtue of their recruitment of membrane- and wall-remodelling enzymes, septins sculpt the cell envelope, and not vice versa (Gladfelter, Pringle and Lew, 2001; Schmidt et al., 2003; Gladfelter, 2006). How do septins contact the plasma membrane? Eukaryotic membranes contain various phosphoinositides that often display discrete localization patterns and facilitate the assembly of higher-order protein complexes at the membrane via the specific binding of membrane-associated proteins. Direct phospholipid binding has been reported for human SEPT4 (Zhang et al., 1999) and budding yeast Cdc3, Cdc10, Cdc11 and Cdc12 (Casamayor and Snyder, 2003). In the case of SEPT4, binding is specific for phosphatidylinositol 4,5-bisphosphate (PtdIns[4,5]P2 ) (Zhang et al., 1999), which is enriched at sites of septin function (e.g. the cleavage furrow of dividing animal cells) (Saul et al., 2004; Field et al., 2005) and required for normal SEPT4 localization (Zhang et al., 1999). PtdIns[4,5] P2 is also enriched at the yeast bud neck (Garrenton, 2007); C. Stefan and S.D. Emr, personal communication; L.E. Stolz and J.D. York, personal communication) and is required for normal septin localization there (Rodriguez-Escudero et al., 2005). It has been reported that GST-yeast septin fusions expressed in and

SEPTIN COMPLEXES IN VERTEBRATES

89

purified from yeast cells bind with highest affinity to filter-immobilized PtdIns[4]P and PtdIns[5]P, and not at all to PtdIns[4,5]P2 (Casamayor and Snyder, 2003). However, in studies assessing binding to phosphoinositide-containing liposomes, yeast septins individually expressed in and purified from bacteria bind better to vesicles containing PtdIns[4,5]P2 than to those containing PtdIns[4]P (T. Allyn, V. Votin and J. Thorner, unpublished results). Additional studies are clearly needed to clarify the lipid-binding specificities of septin subunits, septin rods and septin filaments. Protein-membrane interaction is often mediated by motifs in the protein that contain multiple basic residues that can interact electrostatically with the negatively charged head groups of phosphoinositides and/or other acidic glycerophospholipids. Most septins have a polybasic motif located just N-terminal to the P-loop of the G domain (Plate Figure 3.3c), and mutations in this motif reportedly disrupt lipid binding in vitro (Zhang et al., 1999; Casamayor and Snyder, 2003) and interfere with normal septin function in vivo (T. Allyn, V. Votin and J. Thorner, unpublished results). However, the crystal structure of the human septin hexamer reveals that the polybasic motif falls squarely within the α0 helix and is substantially buried at each NC dimer interface. Nonetheless, one or two of the basic side chains might be sufficiently exposed so as to contribute to membrane interaction. It should be appreciated that, in the context of the complex, even just one or two such contacts contributed by each subunit would represent 8-16 basic residues all projecting off essentially the same face of a rod (Plate Figure 3.3b). Indeed, this arrangement could provide a zipper-like effect that would lock a rod or filament all along its length to the surface of the membrane, consistent with the EM images showing that the 10-nm neck filaments are closely and continuously apposed to the plasma membrane (Byers and Goetsch, 1976a) (Figure 3.4a, right). Also, it is noteworthy that of the seven septins encoded in the S. cerevisiae genome, the three subunits known (Cdc11) or deduced (Shs1 and Spr28), as described earlier, to be at the ends of a rod have the most highly basic α0 sequences: Cdc11, -RKRKHLKR-; Shs1, -RRKKEHKR-; and, Spr28, -RRRKGYKK- (as compared, for example, with the corresponding segment of Cdc12, -RYKIVNEE-). At the ends of the hetero-octamer (a potential, but unfilled, NC interface), the α0 helix is completely exposed (Plate Figure 3.3b). Therefore, these basic ‘fingers’ at each end of a rod could provide a robust mechanism by which rods are recruited to the membrane as soon as they are assembled or, alternatively, to ensure that assembly of rods (or filaments) is nucleated only in association with membranes.

Byers and Goetsch revisited We feel that, as a fitting conclusion to this chapter, we must return to where we began, with the 10-nm filaments at the bud neck of S. cerevisiae cells (Byers and Goetsch, 1976a; Frazier et al., 1998). How has what we’ve learned more

90

CH03 BIOCHEMICAL PROPERTIES AND SUPRAMOLECULAR ARCHITECTURE

recently about septin ultrastructure and biochemistry improved our ability to interpret those EM images? Remarkably, despite the advances described in this article, a long-standing question in the field has remained how the filaments at the bud neck are aligned. Byers and Goetsch (1976a) interpreted their images as reflecting continuous filaments (∼10 nm in cross-section) (Figure 3.4a) wrapping circumferentially around the bud neck, either as separate bands or as a helix with a wide gyre, spaced ∼30 nm apart (Figure 3.4b). However, others have suggested (Field et al., 1996) that yeast septin filaments actually run in the orthogonal direction, extending from the mother cell into the bud, and that the pattern observed by Byers and Goetsch (1976a) arises artifactually. Specifically, it was proposed that (i) the filaments run parallel to the mother-bud axis, (ii) the filaments contain some regularly repeated structure that occurs at intervals of ∼30 nm, (iii) for some reason, this periodic structure stains more intensely than any other part of the filament and (iv) the filaments are in register, so that the heavily staining blobs on one filament are aligned with the blobs of the filaments to its right and to its left. In this way, it was posited, the more densely stained striations made up of all the blobs would appear like filaments running circumferentially. Some, albeit weak, evidence for this highly speculative alternative model of filament organization at the bud neck comes from the fact that certain mutants (e.g. gin4
) cause aberrant septin structures to be formed and these include aggregated filament bundles with their long dimension parallel to the mother-bud axis (Longtine, Fares and Pringle, 1998). More recently, using a strategy designed to tag septins with GFP in a ‘fixed’ orientation so as to take advantage of the fluorescence anisotropy of the chromophore in this fluorescent probe, valiant attempts were made to examine filament orientation in vivo (Vrabioiu and Mitchison, 2006). The GFP molecules in such septin-GFP fusions appeared to be oriented all in the same direction, confirming that in the filaments at the bud neck the septins themselves are uniformly and linearly arrayed (Vrabioiu and Mitchison, 2006). Intriguingly, whatever direction this represents (parallel to the bud neck or orthogonal to it), the orientation of the GFP shifts by 90◦ when the septin collar at the bud neck is split into two distinct rings during cytokinesis (Vrabioiu and Mitchison, 2006). One interpretation of this observation (if one assumes that each septin-GFP chimera is immutably rigid) is that every septin (or rod or filament) rotates by 90◦ during this structural transition, a feat of molecular gymnastics truly astounding to envision. Of course, some other change in the septins (or rods or filaments), like association with another protein or some post-translational modification, could cause the GFP to rotate its position by 90◦ relative to the septin to which it is attached. In any event, to answer the original question – the orientation of filaments in the collar in relation to the mother-bud axis – the fluorescence polarization of the septin-GFP probe at the bud neck must be compared with a septin-GFP-containing structure of known orientation. When bundles of filaments formed in vitro by purified yeast complexes are used (visualizing the polarized fluorescence of an individual filament requires a level of sensitivity that cannot yet be achieved for

SEPTIN COMPLEXES IN VERTEBRATES

91

obvious technical reasons), the results obtained were interpreted to indicate that the septin filaments in the filamentous collar run parallel to the mother-bud axis (Vrabioiu and Mitchison, 2006). The subsequent 90◦ shift observed would then mean that, at cytokinesis, the filaments rotate to become perpendicular to the mother-bud axis. However, there are many important caveats to the use of this approach. One is that the preparation used as the known standard for orientation was a bundle of septin aggregates. Typically, the formation of such aggregates reflects some degree of denaturation or unfolding of the constituent septin(s). Hence, the analysis may not reflect the true situation when septins are in their native state. Furthermore, another defining feature of such septin bundles is a greatly reduced spacing between the filaments within it (Bertin et al., 2008). If the normal spacing between unbundled filaments pairs (15–25 nm) reflects cross-filament bridges composed of the CTE–CTE elements projecting orthogonally off each filament, as we have argued, then it is conceivable that the collapse of the spacing observed when filaments bundle is accompanied by a ∼90◦ rotation, so that the CTE–CTE elements that were formerly perpendicular to the filament axis are now approximately parallel to it. Thus, the standard used as the ‘known’ may be the exact opposite of what was assumed. In any event, the rotation observed in vivo during cytokinesis might reflect a difference in the orientation of the CTE–CTE elements rather than any change in the filaments themselves. Another potential concern is that GFP was fused to the C terminus of either Cdc3 or Cdc12, with the assumption that the orientation of the GFP relative to the septin would be fixed by virtue of the putative α-helical nature of the presumptive Cdc3-Cdc12 coiled coil and the known α-helical structure of the N-terminal segment of GFP (Vrabioiu and Mitchison, 2006). Accordingly, the crux of the interpretation of the polarized fluorescence experiments is that the orientation of the coiled coils in the collar at the bud neck remains rigidly fixed and the same as in the standard (the bundles of filaments formed in vitro) . However, as described earlier, both EM and crystal structure analyses clearly demonstrate that, at least in vitro, the coiled coils can be ‘floppy’, occupying a variety of positions relative to the filament axis of nearly 180◦ (John et al., 2007). Consequently, the jury is still out on the issue of the orientation of the septin filaments in the collar relative to the mother-bud axis. There are many reasons to be reassured that the images of the neck filaments observed by Byers and Goetsch (1976a) actually do represent continuous circumferential hoops (or the gyres of a continuous long helix filament) (Byers and Goetsch, 1976a). Based on what we now know about the ultrastructure and atomic level structure of the rods, which constitute the fundamental building block of filaments, the molecular significance of the dimensions observed and described by Byers and Goetsch (1976a) fall neatly into place, but only if the septin filaments in the collar are continuous and circumferential around the bud neck. First, the 10-nm diameter of neck filaments is approximately double the width of the yeast octameric rod (∼4–5 nm) as measured by high-resolution EM (Sirajuddin et al., 2007; Bertin et al., 2008) or

92

CH03 BIOCHEMICAL PROPERTIES AND SUPRAMOLECULAR ARCHITECTURE

of the human hexameric rod (4–5 nm) from either high-resolution EM or X-ray crystallography (Sirajuddin et al., 2007). This suggests that the neck filaments could be pairs of filaments, adjoined via contacts between two filaments mediated by the face of the G domains opposite to that from which the α6 helix and CTEs project (Figure 3.4b and Plate Figure 3.3b). Indeed, in bud neck sections in which the filaments can been viewed longitudinally, the 10-nm thickness can occasionally be resolved as two fine lines ∼5 nm apart (Byers and Goetsch, 1976a) (Figure 3.4b, right). These fine lines are likely the ‘spines’ of each of the individual ∼5-nm-wide filaments, visible only near the edges of the cylindrical bud neck because elsewhere the sections slice through the thicker filament cores (Figure 3.4b, left). G-domain-mediated pairing of filaments might even explain the otherwise incongruous homotypic interactions observed when differentially tagged versions of the same yeast septin were tested for the capacity to self-associate in vitro (Versele et al., 2004) or by the two-hybrid method (Farkasovsky et al., 2005). The 28-nm interval between filaments (presumably measured from centre to centre) translates into 18 nm of physical separation (given a 5-nm thickness of each filament within the pair). This distance is quite similar to the spacing between the paired filaments formed in the test tube (e.g. 15–25 nm (Frazier et al., 1998; Bertin et al., 2008). The most parsimonious interpretation of this spacing is that it represents the separation imposed by extension of the CTEs from one member of each of two opposed filament pairs (Figure 3.4a (right) and 3.4b (right)), assuming that the CTE–CTE elements (calculated to be ∼10 nm if α-helical and fully extended) only overlap/conjoin via contacts within their very distal coiled coil–forming segments (Figure 3.2). The coiled coils (and/or proteins associated with them in vivo) would, according to this model, represent the 3-nm-thick lateral extensions interconnecting adjacent filaments seen by Byers and Goetsch (1976a) (Figure 3.4a, right). A final consideration is the relative position of the filaments with respect to the plasma membrane, from which they appear slightly separated by density that Byers and Goetsch dubbed the ‘connecting zone’ (Byers and Goetsch, 1976a). It does not seem likely that residues from the polybasic motif could protrude away from the G domain far enough to explain this gap (Figure 3.4a, right). It should be recalled, however, that the N terminus of Cdc3 is 80-100 residues larger than those of the other three core yeast septins. Thus, one possibility is that the N terminus of Cdc3 is specialized to ensure membrane contact and occupies the connecting zone. Alternatively, the connecting zone could be occupied by any of the numerous non-septin proteins at the bud neck that interact both with septins and the plasma membrane.

FUTURE DIRECTIONS Although only a small fraction of the last 30 years of research on septins has focused on their ultrastructure and biochemistry, the recent major accomplishments summarized and analysed in this chapter reveal that we now know more

REFERENCES

93

about the organization and architecture of these polymer-forming proteins than we do about their physiological functions and the molecular basis for how they contribute to the cellular processes in which they participate. Hence, what remains is to bridge this gap and connect the features of septin structure to the demands of the biological processes in which septin-containing structures participate. In this regard, the application of modern EM techniques, such as 3-D tomography, in combination with mutant or tagged septin alleles, designed on the basis of our improved understanding of rod and filament architecture, should permit analysis of septin ultrastructure that will ultimately solve unequivocally the riddle of the orientation and organization of the septin filaments within the collar at the yeast bud neck. Furthermore, the actual mechanisms that control septin rod assembly and regulate the dynamic cell cycle-dependent transitions that septin-containing structures undergo also remain largely ill-defined. However, the newly acquired structural information now provides a set of basic tools with which to attack all of these remaining questions and to move forward, with the promise of achieving the kind of profound understanding accomplished by others who have studied historically more cytologically prominent cytoskeletal polymers.

ACKNOWLEDGEMENTS We want to acknowledge in particular the stimulating and fruitful collaboration we have had with our colleague, Prof. Eva Nogales, and various members of her group, specifically Dr Aurelie Bertin, Dr Patricia Grob and Galo Garcia. We also thank another colleague, Prof. Thomas C. Alber and his associate, Dr Ho-leung Ng, for very helpful discussions and technical assistance, Dr Mark S. Longtine, Dr Douglas R. Kellogg, Dr John D. York and Dr Scott D. Emr for the communication of unpublished information, and Raymond E. Chen for useful conversations and comments on a draft of the manuscript. We also gratefully acknowledge the support of a Postdoctoral Research Fellowship (#61-1295) from the Jane Coffin Childs Memorial Fund for Medical Research (to M.A.M.) and of NIH Research Grant GM218421 (to J.T.).

REFERENCES An, H., Morrell, J.L., Jennings, J.L. et al. (2004) Requirements of fission yeast septins for complex formation, localization, and function. Molecular Biology of the Cell , 15, 5551–64. Barral, Y. and Mansuy, I.M. (2007) Septins: cellular and functional barriers of neuronal activity. Current Biology, 17, R961–63. Beausoleil, S.A., Villen, J., Gerber, S.A. et al. (2006) A probability-based approach for high-throughput protein phosphorylation analysis and site localization. Nature Biotechnology, 24, 1285–92. Beranova-Giorgianni, S., Zhao, Y., Desiderio, D.M. and Giorgianni, F. (2006) Phosphoproteomic analysis of the human pituitary. Pituitary, 9, 109–20.

94

CH03 BIOCHEMICAL PROPERTIES AND SUPRAMOLECULAR ARCHITECTURE

Bertin, A., McMurray, M.A., Grob, P. et al. (2008) Assembly and supramolecular organization of the septin hetero-oligomers and filaments of the yeast, Saccharomyces cerevisiae. Proceedings of the National Academy of Sciences of the United States of America, 105, 8274–9. Byers, B. and Goetsch, L. (1976a) A highly ordered ring of membrane-associated filaments in budding yeast. Journal of Cell Biology, 69, 717–21. Byers, B. and Goetsch, L. (1976b) Loss of the filamentous ring in cytokinesis-defective mutants of budding yeast. Journal of Cell Biology, 70, 35. Carroll, C.W., Altman, R., Schieltz, D. et al. (1998) The septins are required for the mitosis-specific activation of the Gin4 kinase. Journal of Cell Biology, 143, 709–17. Casamayor, A. and Snyder, M. (2003) Molecular dissection of a yeast septin: distinct domains are required for septin interaction, localization, and function. Molecular and Cellular Biology, 23, 2762–77. Caviston, J.P., Longtine, M., Pringle, J.R. and Bi, E. (2003) The role of Cdc42p GTPase-activating proteins in assembly of the septin ring in yeast. Molecular Biology of the Cell , 14, 4051–66. Chi, A., Huttenhower, C., Geer, L.Y. et al. (2007) Analysis of phosphorylation sites on proteins from Saccharomyces cerevisiae by electron transfer dissociation (ETD) mass spectrometry. Proceedings of the National Academy of Sciences of the United States of America, 104, 2193–98. Chu, S., DeRisi, J., Eisen, M. et al. (1998) The transcriptional program of sporulation in budding yeast. Science, 282, 699–705. Cid, V.J., Adamikova, L., Cenamor, R. et al. (1998) Cell integrity and morphogenesis in a budding yeast septin mutant. Microbiology, 144, 3463–74. Cid, V.J., Adamikova, L., Sanchez, M. et al. (2001) Cell cycle control of septin ring dynamics in the budding yeast. Microbiology, 147, 1437–50. Dai, J., Jin, W.H., Sheng, Q.H. et al. (2007) Protein phosphorylation and expression profiling by Yin-yang multidimensional liquid chromatography (Yin-yang MDLC) mass spectrometry. Journal of Proteome Research, 6, 250–62. DeMarini, D.J., Adams, A.E., Fares, H. et al. (1997) A septin-based hierarchy of proteins required for localized deposition of chitin in the Saccharomyces cerevisiae cell wall. Journal of Cell Biology, 139, 75–93. Deng, Y., Liu, J., Zheng, Q. et al. (2006) Antiparallel four-stranded coiled coil specified by a 3-3-1 hydrophobic heptad repeat. Structure, 14, 247–55. De Virgilio, C., DeMarini, D.J. and Pringle, J.R. (1996) SPR28, a sixth member of the septin gene family in Saccharomyces cerevisiae that is expressed specifically in sporulating cells. Microbiology, 142 (Pt 10), 2897–905. Dobbelaere, J. and Barral, Y. (2004) Spatial coordination of cytokinetic events by compartmentalization of the cell cortex. Science, 305, 393–96. Dobbelaere, J., Gentry, M.S., Hallberg, R.L. and Barral, Y. (2003) Phosphorylationdependent regulation of septin dynamics during the cell cycle. Developmental Cell , 4, 345–57. Douglas, L.M., Alvarez, F.J., McCreary, C. and Konopka, J.B. (2005) Septin function in yeast model systems and pathogenic fungi. Eukaryotic Cell , 4, 1503–12. Drees, B.L., Sundin, B., Brazeau, E. et al. (2001) A protein interaction map for cell polarity development. Journal of Cell Biology, 154, 549–71.

REFERENCES

95

Egelhofer, T.A., Vill´en J., McCusker D., Gygi S.P. et al. (2008) The septins function in G1 pathways that influence the pattern of cell growth in budding yeast. PLoS ONE , 3, e2022.1–e2022.14. Fares, H., Goetsch, L. and Pringle, J.R. (1996) Identification of a developmentally regulated septin and involvement of the septins in spore formation in Saccharomyces cerevisiae. Journal of Cell Biology, 132, 399–411. Farkasovsky, M., Herter, P., Voss, B. and Wittinghofer, A. (2005) Nucleotide binding and filament assembly of recombinant yeast septin complexes. Biological Chemistry, 386, 643–56. Field, C.M., al-Awar, O., Rosenblatt, J. et al. (1996) A purified Drosophila septin complex forms filaments and exhibits GTPase activity. Journal of Cell Biology, 133, 605–16. Field, S.J., Madson, N., Kerr, M.L. et al. (2005) PtdIns(4,5)P2 functions at the cleavage furrow during cytokinesis. Current Biology, 15, 1407–12. Fields, S. and Sternglanz, R. (1994) The two-hybrid system: an assay for proteinprotein interactions. Trends in Genetics, 10, 286–92. Finger, F.P., Kopish, K.R. and White, J.G. (2003) A role for septins in cellular and axonal migration in C. elegans. Developmental Biology, 261, 220–34. Ford, S.K. and Pringle, J.R. (1991) Cellular morphogenesis in the Saccharomyces cerevisiae cell cycle: localization of the CDC11 gene product and the timing of events at the budding site. Developmental Genetics, 12, 281–92. Frazier, J.A., Wong, M.L., Longtine, M.S. et al. (1998) Polymerization of purified yeast septins: evidence that organized filament arrays may not be required for septin function. Journal of Cell Biology, 143, 737–49. Friedlander, G., Joseph-Strauss, D., Carmi, M. et al. (2006) Modulation of the transcription regulatory program in yeast cells committed to sporulation. Genome Biology, 7, R20. Garcia, W., de Ara´ujo, A.P., Lara, F. et al. (2007) An intermediate structure in the thermal unfolding of the GTPase domain of human septin 4 (SEPT4/Bradeion-beta) forms amyloid-like filaments in vitro. Biochemistry, 46, 11101–9. Garrenton, L.S. (2007) Spatial and temporal regulation of the archetypal MAPK scaffold protein, Ste5, in the budding yeast Saccharomyces cerevisiae, Ph.D. Thesis, University of California,Berkeley. Gevaert, K., Goethals, M., Martens, L. et al. (2003) Exploring proteomes and analyzing protein processing by mass spectrometric identification of sorted N-terminal peptides. Nature Biotechnology, 21, 566–69. Gladfelter, A.S. (2006) Control of filamentous fungal cell shape by septins and formins. Nature Reviews. Microbiology, 4, 223–29. Gladfelter, A.S., Bose, I., Zyla, T.R. et al. (2002) Septin ring assembly involves cycles of GTP loading and hydrolysis by Cdc42p. Journal of Cell Biology, 156, 315–26. Gladfelter, A.S., Pringle, J.R. and Lew, D.J. (2001) The septin cortex at the yeast mother-bud neck. Current Opinion in Microbiology, 4, 681–89. Gruhler, A., Olsen, J.V., Mohammed, S. et al. (2005) Quantitative phosphoproteomics applied to the yeast pheromone signaling pathway. Molecular and Cellular Proteomics, 4, 310–27. Haarer, B.K. and Pringle, J.R. (1987) Immunofluorescence localization of the Saccharomyces cerevisiae CDC12 gene product to the vicinity of the 10-nm filaments in the mother-bud neck. Molecular and Cellular Biology, 7, 3678–87.

96

CH03 BIOCHEMICAL PROPERTIES AND SUPRAMOLECULAR ARCHITECTURE

Hall, P.A. and Russell, S.E.H. (2004) The pathobiology of the septin gene family. The Journal of Pathology, 204, 489–505. Hartwell, L.H. (1971) Genetic control of the cell division cycle in yeast. IV. Genes controlling bud emergence and cytokinesis. Experimental Cell Research, 69, 265–76. Hartwell, L.H., Culotti, J., Pringle, J.R. and Reid, B.J. (1974) Genetic control of the cell division cycle in yeast. Science, 183, 46–51. Hoffert, J.D., Pisitkun, T., Wang, G. et al. (2006) Quantitative phosphoproteomics of vasopressin-sensitive renal cells: regulation of aquaporin-2 phosphorylation at two sites. Proceedings of the National Academy of Sciences of the United States of America, 103, 7159–64. Hsu, S.C., Hazuka, C.D., Roth, R. et al. (1998) Subunit composition, protein interactions, and structures of the mammalian brain sec6/8 complex and septin filaments. Neuron, 20, 1111–22. Hu, H., Yu, W.B., Li, S.X. et al. (2006) Crystallization and preliminary crystallographic studies of human septin 1 with site-directed mutations. Acta Crystallographica Section F: Structural Biology and Crystallization Communications, 62, 128–32. Huang, Y.W., Surka, M.C., Reynaud, D. et al. (2006) GTP binding and hydrolysis kinetics of human septin 2. FEBS Journal , 273, 3248–60. Iwase, M., Luo, J., Bi, E. and Toh, E.A. (2007) Shs1 plays separable roles in septin organization and cytokinesis in Saccharomyces cerevisiae. Genetics, 177, 215–29. Iwase, M., Luo, J., Nagaraj, S. et al. (2006) Role of a Cdc42p effector pathway in recruitment of the yeast septins to the presumptive bud site. Molecular Biology of the Cell , 17, 1110–25. Joberty, G., Perlungher, R.R., Sheffield, P.J. et al. (2001) Borg proteins control septin organization and are negatively regulated by Cdc42. Nature Cell Biology, 3, 861–66. John, C.M., Hite, R.K., Weirich, C.S. et al. (2007) The Caenorhabditis elegans septin complex is nonpolar. EMBO Journal , 26, 3296–307. Johnson, E.S. and Blobel, G. (1999) Cell cycle-regulated attachment of the ubiquitinrelated protein SUMO to the yeast septins. Journal of Cell Biology, 147, 981–94. Joo, E., Tsang, C.W. and Trimble, W.S. (2005) Septins: traffic control at the cytokinesis intersection. Traffic, 6, 626–34. Kadota, J., Yamamoto, T., Yoshiuchi, S. et al. (2004) Septin ring assembly requires concerted action of polarisome components, a PAK kinase Cla4p, and the actin cytoskeleton in Saccharomyces cerevisiae. Molecular Biology of the Cell , 15, 5329–45. Kim, H.B., Haarer, B.K. and Pringle, J.R. (1991) Cellular morphogenesis in the Saccharomyces cerevisiae cell cycle: localization of the CDC3 gene product and the timing of events at the budding site. Journal of Cell Biology, 112, 535–44. Kinoshita, M. (2003) Assembly of mammalian septins. Journal of Biochemistry, 134, 491–96. Kinoshita, M. (2006) Diversity of septin scaffolds. Current Opinion in Cell Biology, 18, 54–60. Kinoshita, M., Field, C.M., Coughlin, M.L. et al. (2002) Self- and actin-template figured assembly of mammalian septins. Developmental Cell , 3, 791–802.

REFERENCES

97

Kinoshita, A., Kinoshita, M., Akiyama, H. et al. (1998) Identification of septins in neurofibrillary tangles in Alzheimer’s disease. American Journal of Pathology, 153, 1551–60. Kozubowski, L., Larson, J.R. and Tatchell, K. (2005) Role of the septin ring in the asymmetric localization of proteins at the mother-bud neck in Saccharomyces cerevisiae. Molecular Biology of the Cell , 16, 3455–66. Kremer, B.E., Adang, L.A. and Macara, I.G. (2007) Septins regulate actin organization and cell-cycle arrest through nuclear accumulation of NCK mediated by SOCS7. Cell , 130, 837–50. Kuhlenbaumer, G., Hannibal, M.C., Nelis, E. et al. (2005) Mutations in SEPT9 cause hereditary neuralgic amyotrophy. Nature Genetics, 37, 1044–46. Lee, P.R., Song, S., Ro, H.S. et al. (2002) Bni5p, a septin-interacting protein, is required for normal septin function and cytokinesis in Saccharomyces cerevisiae. Molecular and Cellular Biology, 22, 6906–20. Li, X., Gerber, S.A., Rudner, A.D. et al. (2007) Large-scale phosphorylation analysis of alpha-factor-arrested Saccharomyces cerevisiae. Journal of Proteome Research, 6, 1190–97. Longtine, M.S. and Bi, E. (2003) Regulation of septin organization and function in yeast. Trends in Cell Biology, 13, 403–9. Longtine, M.S., Fares, H. and Pringle, J.R. (1998) Role of the yeast Gin4p protein kinase in septin assembly and the relationship between septin assembly and septin function. Journal of Cell Biology, 143, 719–36. Longtine, M.S., Theesfeld, C.L., McMillan, J.N. et al. (2000) Septin-dependent assembly of a cell cycle-regulatory module in Saccharomyces cerevisiae. Molecular and Cellular Biology, 20, 4049–61. Low, C. and Macara, I.G. (2006) Structural analysis of septin 2, 6, and 7 complexes. Journal of Biological Chemistry, 281, 30697–706. Lukoyanova, N., Baldwin, S.A. and Trinick, J. (2008) 3D reconstruction of mammalian septin filaments. Journal of Molecular Biology, 376, 1–7. Macara, I.G., Baldarelli, R., Field, C.M. et al. (2002) Mammalian septins nomenclature. Molecular Biology of the Cell , 13, 4111–13. Matsuoka, S., Ballif, B.A., Smogorzewska, A. et al. (2007) ATM and ATR substrate analysis reveals extensive protein networks responsive to DNA damage. Science, 316, 1160–66. McMurray, M.A. and Thorner, J. (2008) Septin stability and recycling during dynamic structural transitions in cell division and development. Current Biology, 18, 1205–1208. McMurray, M.A., Votin, V., Moellmann, M.K. and Thorner, J. (2008) Requirement for septin Cdc10 in budding yeast cell division is masked by spontaneous epigenetic suppression and results in polyploidy. Genetics, manuscript in preparation. Mendoza, M., Hyman, A.A. and Glotzer, M. (2002) GTP binding induces filament assembly of a recombinant septin. Current Biology, 12, 1858–63. Mino, A., Tanaka, K., Kamei, T. et al. (1998) Shs1p: a novel member of septin that interacts with spa2p, involved in polarized growth in Saccharomyces cerevisiae. Biochemical and Biophysical Research Communications, 251, 732–36. Mitchison, T.J. and Field, C.M. (2002) Cytoskeleton: what does GTP do for septins? Current Biology, 12, R788–90.

98

CH03 BIOCHEMICAL PROPERTIES AND SUPRAMOLECULAR ARCHITECTURE

Molina, H., Horn, D.M., Tang, N. et al. (2007) Global proteomic profiling of phosphopeptides using electron transfer dissociation tandem mass spectrometry. Proceedings of the National Academy of Sciences of the United States of America, 104, 2199–204. Mortensen, E.M., McDonald, H., Yates, J.R. and Kellogg, D.R. (2002) Cell cycledependent assembly of a Gin4-septin complex. Molecular Biology of the Cell , 13, 2091–105. Munton, R.P., Tweedie-Cullen, R., Livingstone-Zatchej, M. et al. (2007) Qualitative and quantitative analyses of protein phosphorylation in naive and stimulated mouse synaptosomal preparations. Molecular and Cellular Proteomics, 6, 283–93. Neufeld, T.P. and Rubin, G.M. (1994) The Drosophila peanut gene is required for cytokinesis and encodes a protein similar to yeast putative bud neck filament proteins. Cell , 77, 371–79. Nguyen, T.Q., Sawa, H., Okano, H. and White, J.G. (2000) The C. elegans septin genes, unc-59 and unc-61, are required for normal postembryonic cytokineses and morphogenesis but have no essential function in embryogenesis. Journal of Cell Science, 113 (Pt 21), 3825–37. Nousiainen, M., Sillje, H.H., Sauer, G. et al. (2006) Phosphoproteome analysis of the human mitotic spindle. Proceedings of the National Academy of Sciences of the United States of America, 103, 5391–96. Olsen, J.V., Blagoev, B., Gnad, F. et al. (2006) Global, in vivo, and site-specific phosphorylation dynamics in signaling networks. Cell , 127, 635–48. O’Shea, E.K., Rutkowski, R., Stafford, W.F. III and Kim, P.S. (1989) Preferential heterodimer formation by isolated leucine zippers from fos and jun. Science, 245, 646–48. Pan, F., Malmberg, R.L. and Momany, M. (2007) Analysis of septins across kingdoms reveals orthology and new motifs. BMC Evolutionary Biology, 7, 103. Rodriguez-Escudero, I., Roelants, F.M., Thorner, J. et al. (2005) Reconstitution of the mammalian PI3K/PTEN/Akt pathway in yeast. Biochemical Journal , 390, 613–23. Rush, J., Moritz, A., Lee, K.A. et al. (2005) Immunoaffinity profiling of tyrosine phosphorylation in cancer cells. Nature Biotechnology, 23, 94–101. Saul, D., Fabian, L., Forer, A. and Brill, J.A. (2004) Continuous phosphatidylinositol metabolism is required for cleavage of crane fly spermatocytes. Journal of Cell Science, 117, 3887–96. Schirmer, E.C., Homann, O.R., Kowal, A.S. and Lindquist, S. (2004) Dominant gain-of-function mutations in Hsp104p reveal crucial roles for the middle region. Molecular Biology of the Cell , 15, 2061–72. Schmidt, M., Varma, A., Drgon, T. et al. (2003) Septins, under Cla4p regulation, and the chitin ring are required for neck integrity in budding yeast. Molecular Biology of the Cell , 14, 2128–41. Sheffield, P.J., Oliver, C.J., Kremer, B.E. et al. (2003) Borg/septin interactions and the assembly of mammalian septin heterodimers, trimers, and filaments. Journal of Biological Chemistry, 278, 3483–88. Sirajuddin, M., Farkasovsky, M., Hauer, F. et al. (2007) Structural insight into filament formation by mammalian septins. Nature, 449, 311–15. Smolka, M.B., Albuquerque, C.P., Chen, S.H. and Zhou, H. (2007) Proteome-wide identification of in vivo targets of DNA damage checkpoint kinases. Proceedings

REFERENCES

99

of the National Academy of Sciences of the United States of America, 104, 10364–69. Smolka, M.B., Chen, S.H., Maddox, P.S. et al. (2006) An FHA domain-mediated protein interaction network of Rad53 reveals its role in polarized cell growth. Journal of Cell Biology, 175, 743–53. Spiliotis, E.T. and Nelson, W.J. (2006) Here come the septins: novel polymers that coordinate intracellular functions and organization. Journal of Cell Science, 119, 4–10. Stewart, M., Kent, H.M. and McCoy, A.J. (1998) The structure of the Q69L mutant of GDP-Ran shows a major conformational change in the switch II loop that accounts for its failure to bind nuclear transport factor 2 (NTF2). Journal of Molecular Biology, 284, 1517–27. Sudo, K., Ito, H., Iwamoto, I. et al. (2007) SEPT9 sequence alternations causing hereditary neuralgic amyotrophy are associated with altered interactions with SEPT4/ SEPT11 and resistance to Rho/Rhotekin-signaling. Human Mutation, 28, 1005–13. Sung, H., Chul Han, K., Chul Kim, J. et al. (2005) A set of epitope-tagging integration vectors for functional analysis in Saccharomyces cerevisiae. FEMS Yeast Research, 5, 943–50. Takizawa, P.A., DeRisi, J.L., Wilhelm, J.E. and Vale, R.D. (2000) Plasma membrane compartmentalization in yeast by messenger RNA transport and a septin diffusion barrier. Science, 290, 341–44. Tang, C.S. and Reed, S.I. (2002) Phosphorylation of the septin cdc3 in g1 by the cdc28 kinase is essential for efficient septin ring disassembly. Cell Cycle, 1, 42–49. Trinidad, J.C., Specht, C.G., Thalhammer, A. et al. (2006) Comprehensive identification of phosphorylation sites in postsynaptic density preparations. Molecular and Cellular Proteomics, 5, 914–22. Uetz, P., Giot, L., Cagney, G. et al. (2000) A comprehensive analysis of protein-protein interactions in Saccharomyces cerevisiae. Nature, 403, 623–27. Versele, M., Gullbrand, B., Shulewitz, M.J. et al. (2004) Protein-protein interactions governing septin heteropentamer assembly and septin filament organization in Saccharomyces cerevisiae. Molecular Biology of the Cell , 15, 4568–83. Versele, M. and Thorner, J. (2004) Septin collar formation in budding yeast requires GTP binding and direct phosphorylation by the PAK Cla4. Journal of Cell Biology, 164, 701–15. Versele, M. and Thorner, J. (2005) Some assembly required: yeast septins provide the instruction manual. Trends in Cell Biology, 15, 414–24. Vetter, I.R. and Wittinghofer, A. (2001) The guanine nucleotide-binding switch in three dimensions. Science, 294, 1299–304. Vrabioiu, A.M., Gerber, S.A., Gygi, S.P. et al. (2004) The majority of the Saccharomyces cerevisiae septin complexes do not exchange guanine nucleotides. Journal of Biological Chemistry, 279, 3111–18. Vrabioiu, A. and Mitchison, T.J. (2006) Structural insights into yeast septin organization from polarized fluorescence microscopy. Nature, 443, 466–69. Vrabioiu, A.M. and Mitchison, T.J. (2007) Symmetry of septin hourglass and ring structures. Journal of Molecular Biology, 372, 37–49. Xie, H., Surka, M., Howard, J. and Trimble, W.S. (1999) Characterization of the mammalian septin H5: distinct patterns of cytoskeletal and membrane association from other septin proteins. Cell Motility and the Cytoskeleton, 43, 52–62.

100

CH03 BIOCHEMICAL PROPERTIES AND SUPRAMOLECULAR ARCHITECTURE

Xue, J., Milburn, P.J., Hanna, B.T. et al. (2004) Phosphorylation of septin 3 on Ser-91 by cGMP-dependent protein kinase-I in nerve terminals. Biochemical Journal , 381, 753–60. Zhang, J., Kong, C., Xie, H. et al. (1999) Phosphatidylinositol polyphosphate binding to the mammalian septin H5 is modulated by GTP. Current Biology, 9, 1458–67.

4 Yeast septins: a cortical organizer Yves Barral Institute of Biochemistry, ETH Zurich, Schaffmatstrasse 18, 8093 Zurich, Switzerland

INTRODUCTION Hartwell and colleagues identified the septins in the early 1970s as four yeast cell division cycle (cdc) genes (CDC3 , CDC10 , CDC11 , CDC12 ) that caused cytokinesis defects when inactivated (Pringle Chapter 1, this book; Hartwell, Culotti and Reid, 1970; Hartwell, 1971; Hartwell et al., 1974; Longtine et al., 1996). Genetic analysis showed that septin mutants were also defective in cell morphogenesis, leading to the formation of elongated buds. These four septin genes encode related proteins, characterized by the presence of a Ras-related GTPase domain followed by a coiled-coil domain (for review see Faty, Fink and Barral (2002)). The C-terminal coiled-coil indicates that unlike Ras small GTPases septin proteins stably associate with binding partners. All four proteins localize together and in an interdependent manner to the yeast bud neck (Haarer and Pringle, 1987; Kim, Haarer and Pringle, 1991). Moreover, septin mutants lacked the so-called bud-neck filaments, an ordered structure apparent at the bud neck in Electron Microscpy (EM) images (Byers and Goetsch, 1976). Accordingly, septins assemble together into a novel cytoskeletal structure made of filaments, and recombinant complexes composed of Cdc3, Cdc10, Cdc11 and Cdc12 form filaments in vitro (Frazier et al., 1998). Over the past twenty years, the yeast septin collar at the bud neck have been implicated in a number of cellular processes, ranging from cell division and cell cycle control, to bud site selection, cell polarity and the compartmentalization of organelles. Here, I will review these different functions, and their known molecular mechanisms. I will subsequently try to build on this knowledge to evaluate different models about what the actual molecular function of septins might be. However, before discussing the details of septin function in yeast, let’s consider first the structures they form. The Septins Edited by Peter A. Hall, S.E. Hilary Russell and John R. Pringle © 2008 John Wiley & Sons, Ltd. ISBN: 978-0-470-51969-1

102

CH04 YEAST SEPTINS: A CORTICAL ORGANIZER

SEPTIN STRUCTURES IN YEAST Budding yeast septins are involved in two major structures. During vegetative divisions, the five septins Cdc3, Cdc10, Cdc11, Cdc12 and Shs1/Sept7 (Mino et al., 1998; Kim, Haarer and Pringle, 1991; Haarer and Pringle, 1987; Ford and Pringle, 1991) form the bud-neck filaments. Upon completion of meiosis, another complex comprising Cdc3, Cdc10, Cdc11, Spr28 (De Virgilio, DeMarini and Pringle, 1996) and Spr3 (Fares, Goetsch and Pringle, 1996) contribute to spore wall formation. Our knowledge of this second process remains fragmentary. This review focuses on vegetative septin structures.

Three septin structures at the bud neck Septins form at the bud neck three successive structures during the budding cycle (Faty, Fink and Barral, 2002). At bud emergence, septins form a cap that covers the forming budding site (Versele and Thorner, 2004). This cap structure is transient and has not been observed identically by all authors. Some have described a mesh of randomly oriented filaments (Cid et al., 2001), while others reported diffuse staining throughout the presumptive bud site (Versele and Thorner, 2004). Yet other investigators documented the formation a ring encircling the future bud site (Haarer and Pringle, 1987; Kim, Haarer and Pringle, 1991). The reason for these differences are unclear, but probably reflects the method of visualization used. Green fluorescent protein (GFP)-tagged septins revealed rather the diffuse cap, while septin visualization by indirect immunofluorescence on fixed cells showed either a filament mesh or a ring. Fixation probably stabilizes the structural intermediates that are the least dynamic, trapping septins in their filamentous form. GFP-tagged septins might also shift the population towards the most dynamic intermediates. Thus, diffuse and filamentous structures, might correspond to two intermediates that coexist in the reality. Very soon after bud emergence, septins form a sharp ring identical by all visualization techniques. This ring, also called septin collar, is formed of septin filaments tightly apposed to the plasma membrane and corresponds to the bud neck filaments observed by electron microscopy (Byers and Goetsch, 1976). Following the curvature of the plasma membrane at the bud neck, this ‘ring’ takes an hourglass shape (Cid et al., 2001; Lippincott et al., 2001); it is a cylinder constricted in the middle (review Gladfelter, Pringle and Lew, 2001; Longtine and Bi, 2003). How the filaments are organized in this cylinder was solved only recently. The bud-neck filaments observed by EM run parallel to the plane of the bud neck (Byers and Goetsch, 1976). However, these filamentous densities might correspond to cross-links between filaments. Indeed, in vitro and under low salt conditions recombinant septin filaments assemble into sheets (Frazier et al., 1998), where

SEPTIN STRUCTURES IN YEAST

103

stripes become visible at regular intervals perpendicular to the actual filaments. These stripes very much resemble the neck filaments. Furthermore, mutants lacking the septin-dependent kinase (SDK) Gin4 form irregular rings where filaments seem to become visible as bundles that align with the mother-bud axis (Longtine and Bi, 2003). Therefore, in the septin collar the septin filaments might not assemble in a spring shape, as originally expected, but seem to orient parallel to the mother-bud axis like the boards of a barrel (Frazier et al., 1998). The current, and probably definitive solution to this discussion was provided by the Mitchison lab in a set of very innovative experiments, using polarized light microscopy (Vrabioiu and Mitchison, 2006; Vrabioiu and Mitchison, 2007). These data show that septin filaments run parallel to the mother-bud axis (Vrabioiu and Mitchison, 2006). Thus, septins filaments form a sort of gaze through lateral interactions at the plasma membrane. The septin collar stays at the bud neck throughout bud growth and mitosis, i.e. for half to three quarters of the cell cycle. Upon mitotic exit it then splits into two rings, one on each side of the bud neck (Lippincott et al., 2001). Polarized light microscopy indicates that during this transition the filaments turn 90◦ and align with the plane of the bud neck (Vrabioiu and Mitchison, 2006). Simultaneously, the actomyosin ring starts to contract, and the curvature of the plasma membrane inverses to close the bud neck (Figure 4.1, IV). Therefore, the septin rings must shift from a nearly cylindrical organization (septin collar) to two flat rings (Figure 4.1, IV). This shape change could explain the reorientation of the filaments (see below). During these changes, the diameter of the septin collar and the septin rings remains constant (Lippincott and Li, 1998a). The septin ring(s) does not contract at cytokinesis.

Septin dynamics Cytoskeletal networks generally are highly dynamic, such as microtubules and actin microfilaments. Continuous alternation between growth and shrinkage promotes rapid reorganization and reshaping of the cell in response to intra and extracellular signals. In contrast, septin filaments show very little dynamics during most of the cell cycle. In fluorescence recovery after photobleaching (FRAP) experiments, septin turnover is observed only during short intervals prior to bud emergence (phase I, Figure 4.1), at ring splitting (phase III), and as the ring disassembles in the next G1 phase (Dobbelaere et al., 2003; Caviston et al., 2003). These intervals correspond to rapid reorganization steps. During phase II (septin collar) and phase IV (split rings), little to no recovery is observed, indicating that the septin filaments do not turnover. During the dynamic intervals, fluorescence recovery appears to take place at the cost of the unbleached part of the ring, suggesting that the ring is in a ‘fluid’

CH04 YEAST SEPTINS: A CORTICAL ORGANIZER

104

I

II

III

IV

Figure 4.1 The different forms adopted by the septin cytoskeleton at the yeast bud neck throughout the budding process. During bud emergence (stage I) the filaments assemble into a cap or flat ring morphology structure around the site of bud emergence. As the bud emerges (stage II) the filaments adopt a ring like structure that follows the outward curvature of the plasma membrane at the bud neck. In stage III, which corresponds to the period of bud growth, the ring now forms an hourglass shaped collar at the transition between mother and bud. This collar is tightly apposed to the inner surface of the plasma membrane. Upon cytokinesis onset, the collar splits into two rings (Stage IV) that are remodeled again to follow the changing curvature of the plasma membrane as the neck closes. Thereby, the rings probably adopt back the flat ring morphology. In stage II and III, the filaments are aligned with the mother-bud axis. We suggest that they open in an iris-like manner at each extremities of the collar, where it increases in diameter. In the stage I and IV, the filaments slide along each other to accommodate the flat topology of the ring.

SEPTIN STRUCTURES IN YEAST

105

state, where subunits move mainly inside the ring. Accordingly, Dobbelaere et al. observed that completely photobleached rings did not recover, indicating that there is little or no exchange of subunits between the ring and the cytoplasm (Dobbelaere et al., 2003). This observation was, however, not reproduced by Caviston and coworkers, who observed some recovery of fully photobleached rings (Caviston et al., 2003). This discrepancy is possibly due to these experiments focusing on different stages in phase I. Indeed, Dobbelaere et al. focused on cells undergoing bud emergence, while Caviston et al. focused on cells prior to bud emergence. Therefore, we suggest that during the first half of phase I the newly recruited septins do rapidly exchange with the cytoplasm. During the second half, the filaments start to be more stable, and movement is restricted within the ring. Thus, we distinguish three different dynamic states for septin higher-order structures: (i) During assembly and disassembly rings are dynamic and exchange subunits with the cytoplasm. (ii) Concomitant with changes in the curvature of the plasma membrane at bud emergence and cytokinesis, the septin ring is in the fluid state. It adapts to the new topology of the cell, and exchanges are due to filament movements within the ring. (iii) During the rest of the time, which is most, the septin ring is in a frozen state. The subunit do neither move within the ring nor exchange with the cytoplasm.

A model for filament organization in the septin ring Thus, septin structures adapt remarkably well to the complex topological changes of the plasma membrane at bud emergence and cytokinesis, without loosing membrane interaction. How this behaviour is supported by the structure of the collar is not known. It is unclear how the filaments, which are parallel to the bud neck axis, support the hourglass shape of the collar. It is unlikely that the filaments are either compressed in the middle of the cylinder, or separate from each other at its extremities. Instead, we propose that the filaments might slide along each other as the barrel becomes widens, like in an iris (Figure 4.1, X). At ring splitting, this movement might reach the extreme, such as to fit the ring in a disc, on the cleavage plane (Figure 4.1, X). This model is attractive because it sheds light onto what the fluid phase might correspond to. Indeed, in late phase I the septin ring rearranges from a disc-crown to a cylinder, requiring extensive sliding of the filaments along each other. This would explain the fluorescence recovery observed in FRAP experiments, despite the lack of exchange with the cytoplasm. Likewise, the recovery observed during phase III might be due to similar sliding events as the membrane curvature changes again. Thus, our current picture of septin organization reveals a sophisticated system of filaments that covers the topologically complex surface of the plasma membrane at the bud neck and precisely follows its changes.

106

CH04 YEAST SEPTINS: A CORTICAL ORGANIZER

THE CELLULAR FUNCTIONS OF SEPTINS: VARIETY AND RECURRENCE Yeast septins have been implicated in a broad panel of functions in cytokinesis, cell cycle control, and cellular compartmentalization. These processes involve septins in both structural and regulatory functions. This structural/regulatory duality has long represented an important difficulty in septin research.

Septins in cell division In yeast, division plane determination is subordinated to the budding process and cytokinesis always takes place at the bud neck. Thus, the division plane is set at bud emergence, early in the cycle, even before DNA replication and mitosis (for review on cytokinesis in budding yeast (Tolliday, Bouquin and Li, 2001; Bi, 2001)). Therefore, yeast cells need (i) to memorize the position of the cleavage plane up until cytokinesis, and (ii) to position the spindle perpendicular to this plane. This second condition ensures that spindle elongation correctly segregates one daughter nucleus into the mother and one into the bud. The yeast septin collar is involved in the solution to these problems. Its position serves as a memory for the position of the bud neck until cytokinesis, and as a spatial cue to direct spindle positioning.

Septins in spindle positioning The septin ring at the bud neck plays at least three distinct roles in spindle positioning to the bud neck and alignment with the mother-bud axis. Its first role is to organize capture sites for microtubule plus-ends at the cortex of the bud neck. Upon capture, force is generated on attached microtubules to pull the spindle towards the bud neck (Kusch et al., 2002). Septin contribution to microtubule attachment is mainly indirect. Indeed, the septin ring on the inner side of the plasma membrane at the bud neck is in fact not in direct contact with the cytoplasm, but shielded by a sheet of endoplasmic reticulum (ER, see below) (Byers and Goetsch, 1976; Luedeke et al., 2005). Thus, microtubules are unlikely to directly interact with septins at the bud neck. Furthermore, microtubule attachment at the bud neck also depends on Bud6. Bud6 is a peripheral-membrane protein that localizes to the bud neck, probably to the endoplasmic membrane, in a septin-dependent manner (Amberg et al., 1997; Luedeke et al., 2005; Huisman et al., 2004). Thus, microtubule attachment sites might be located on the surface of the ER at the bud neck. There, Bud6 might directly interact with microtubule ends during microtubule capture (Huisman et al., 2004). The second role of the septin collar is in the regulation of microtubule dynamics at the bud neck, downstream of attachment. There, septin function depends on the SDKs Hsl1 and Gin4 (Barral et al., 1999; Altman and Kellogg, 1997; Kusch et al.,

THE CELLULAR FUNCTIONS OF SEPTINS: VARIETY AND RECURRENCE

107

2002; Tanaka and Nojima, 1996; Ma, Lu and Grunstein, 1996). In the absence of SDKs, microtubules attached to the bud neck show a reduced catastrophe frequency and grow longer. They also fail to stay attached once they finally start to shrink. As a consequence the spindle pole body (SPB) (functionally equivalent to the centrosome of animal cells) at the other end of the attached microtubule is pushed back into the mother cell instead of being pulled towards the cleavage plane (Kusch et al., 2002). The nature of the substrates regulated by the SDKs remains unknown. Finally, a third process depending on the septin collar is the regulation of dynein distribution. Indeed, in wild-type cells dynein accumulates at the tip of the microtubules emanating from only one SPB, the SPB directed towards the bud (and hence towards the bud neck), here called SPBbud (Shaw et al., 1997; Segal et al., 2000; Grava et al., 2006). Cells lacking SDKs localize dynein to microtubules emanating from both SPBs, and dynein level on microtubules remains weak (Grava et al., 2006). How SDKs control the distribution of dynein is still unknown, but time-lapse analysis indicate that dynein recruitment to the SPBbud and emanating microtubules depends on contact events between the microtubule plus end and the bud neck (Grava et al., 2006). In turn, dynein asymmetry ensures that only one SPB is finally pulled into the bud upon anaphase onset (Grava et al., 2006). Thus, septins contribute to spindle positioning mainly by providing a spatial cue for spindle organization and the control of spindle movements. In these functions, the septins provide a scaffold for the assembly and regulation of other structures, such as the ER and SDKs.

Septins and the assembly of the cytokinetic machinery Cytokinesis is a highly complex process, requiring the coordination of multiple events. Like in animal cells, yeast cytokinesis proceeds in two steps: furrow ingression, which reduces the cytoplasmic connection between the daughter cells to a narrow bridge, and abscission, which resolves the plasma membrane and definitely separates the two daughters (for review on cytokinesis see Guertin, Trautmann and McCollum, 2002; Glotzer, 2005). Furrow ingression depends on myosin II, Myo1, which is recruited in a septin-dependent manner to the bud neck shortly after bud emergence, but does not function before late anaphase, when it contributes to actin recruitment and assembly of the actomyosin contractile ring (Lippincott and Li, 1998b; Bi et al., 1998). Upon mitotic exit, the kinase Mob1/Dbf2, a downstream effector of the mitotic exit network (MEN) triggers ring contraction (Lippincott and Li, 1998b; Lippincott et al., 2001; Luca et al., 2001; Menssen, Neutzner and Seufert, 2001). MEN is equivalent to the septation initiation network (SIN) in fission yeast (for review see Krapp, Gulli and Simanis (2004). Although septins are involved in the recruitment of myosin II to the bud neck, they are not part of the actomyosin ring itself, which contracts between the two split septin rings (Lippincott et al., 2001; Dobbelaere and Barral, 2004).

108

CH04 YEAST SEPTINS: A CORTICAL ORGANIZER

Remarkably, myosin II is not essential for cytokinesis in yeast. Thus, septins must play additional roles besides myosin II recruitment. One such role is the recruitment of other cytokinetic factors, such as the proteins Hof1 (Lippincott and Li, 1998a), Cyk3 (Korinek et al., 2000; Vallen, Caviston and Bi, 2000; Iwase et al., 2007). Hof1, related to fission yeast cdc14, localizes like Myo1 to the bud neck throughout most of bud growth and focuses between the split septin rings at cytokinesis (Vallen, Caviston and Bi, 2000; Lippincott and Li, 1998a; Dobbelaere and Barral, 2004). Its exact function is not yet clear, but is independent of and overlapping with that of Myo1. While both myo1
and hof1
single mutant cells are alive, the myo1
hof1
double mutant is dead (Vallen, Caviston and Bi, 2000). Thus, in the absence of the Myo1 Hof1 is able to support furrow ingression and vice versa, while co-inactivation of both prevents ingression. The molecular function of Cyk3 is also unclear. It is recruited to the bud neck only in late mitosis (Korinek et al., 2000) and functions independently of both Myo1 and Hof1. Indeed, while each of the cyk3
, hof1
and myo1
single mutants are alive in most genetic backgrounds, each double mutant combining two of these mutations is dead (Korinek et al., 2000). Thus, Cyk3 can support cytokinesis in the absence of any one of Hof1 or Myo1. Together, these data indicate that at least three parallel pathways contribute to yeast cytokinesis. Each of these pathways depends on septin function for their proper recruitment to the cleavage plane. Thus, one aspect of septin function in yeast cytokinesis is to serve as a scaffold for the assembly of the cytokinetic machinery prior to cell cleavage.

Septins during cytokinesis The role of septins in cytokinesis is however not restricted to pre-cytokinetic events. Disruption of the split septin rings during cytokinesis does not affect the localization of Hof1 and Myo1 anymore but still leads to cytokinetic failure (Dobbelaere and Barral, 2004). During cytokinesis, these rings located on each side of the abscission site function as diffusion barrier to confine numerous membrane-remodeling factors to the abscission site (Dobbelaere and Barral, 2004). Such factors comprise the exocyst, a complex involved in the targeting of exocytic vesicles to the plasma membrane (Finger, Hughes and Novick, 1998) and the polarisome (Sheu et al., 1998), involved in actin-cable nucleation at the cell cortex (Dobbelaere and Barral, 2004). The exocyst is absolutely required for proper furrowing and abscission. The implication of several independent molecular pathways in cytokinesis clearly provides robustness to the process. In turn, their involvement in all of these pathways explains why yeast septins are so crucial for cytokinesis. These data also establish that septins provide at least two independent functions: a scaffolding function prior to cytokinesis and a confinement role during cell cleavage. As we will see, this second duality, scaffold/barrier, is also recurring in septin function.

THE CELLULAR FUNCTIONS OF SEPTINS: VARIETY AND RECURRENCE

109

Conservation of septin functions It is not clear whether the cell division functions of septin are conserved. Septins do not appear to mediate spindle positioning in another cell type than yeast. However, septins certainly interact with and contribute to the regulation of interphase and mitotic microtubules in mammals (Surka, Tsang and Trimble, 2002; Spiliotis, Kinoshita and Nelson, 2005; Kremer, Haystead and Macara, 2005). Thus, a conserved connection between septin and microtubules might exist, yet be implicated in different processes in different cell types. Similarly, while septins are not necessary for cytokinesis in many cells. For example, C. elegans embryos and fission yeast assemble and contract their actomyosin ring and complete cytokinesis fine in the absence of septins (Nguyen et al., 2000; An et al., 2004; Tasto, Morrell and Gould, 2003; Berlin, Paoletti and Chang, 2003; Wu et al., 2003). Still, septins localize to both sides of the cleavage plane in all tested cell types (for review see Kinoshita and Noda, 2001; Joo, Tsang and Trimble (2005)). Furthermore, recent data established that insect, nematode and mammalian septins interact with anillins (Field and Alberts, 1995) and contribute to the proper regulation and functional robustness of myosin II (Oegema et al., 2000; Kinoshita et al., 2002; Hime, Brill and Fuller, 1996; Maddox et al., 2007). Anillins form a family of peripheral-membrane proteins that interact with both actin- and septin-based structures. In tissue culture, septins are required for cytokinesis events posterior to furrow ingression, suggesting that they contribute to abscission (Kinoshita et al., 1997; Joo, Tsang and Trimble, 2005). Thus, the molecular functions of septin in both actomyosin-related and abscission processes are likely to be conserved, although the impact of these functions on the cleavage process is variable. The nature and strength of this impact probably depends on the biology of the cell. For example, the role of septins in ‘memorizing’ the location of the bud neck during bud growth might play a determining role in making budding yeast septins so central. By contrast, the size of the nematode oocyte is probably the reason why emphasis is rather put on actomyosin ring contraction in these cells.

Septins and cell morphogenesis Septins contribute in two ways to budding. First, they contribute to bud-site selection in late G1. Second, they mediate proper bud morphogenesis.

Septins and bud site selection Yeast buds are not positioned randomly on their mothers, but according to selection programs that depend on the plo¨ıdy of the cell. Haploid cells bud axially, i.e. adjacent to the previous budding site. Diploids cells display a bipolar budding pattern. They bud alternatively adjacent and distal to the original cite of cytokinesis (see Casamayor and Snyder (2002) for a review on bud-site selection). In both cases, the septin rings set the spatial cues that the cell use to position new buds. In

110

CH04 YEAST SEPTINS: A CORTICAL ORGANIZER

haploid cells, the septin rings of telophase cells, which persist during G1, activate a signalling cascade comprising the anillin-related protein Bud4 (Sanders and Herskowitz, 1996; Park and Bi, 2007; Gladfelter et al., 2005), the Bud4-interacting protein Bud3 (Chant et al., 1995), the membrane proteins Axl1 and Axl2 (Fujita et al., 1994; Roemer et al., 1996), the ras-related GTPase Rsr1/Bud1 (Bender and Pringle, 1991), and its guanine nucleotide exchange factor (GEF), Bud5 (Bender, 1993) (reviewed in Park and Bi (2007)). Although we do not understand this cascade in detail, its activation is initiated by the septin-dependent recruitment of the Bud4/Bud3 complex to the cortex (Sanders and Herskowitz, 1996; Chant et al., 1995). This initial step depends on septin function and appears to involve a direct interaction between Bud4 and septins. Subsequent activation of Bud1 by this cascade at the previous site of cytokinesis triggers the recruitment of the Rho-like GTPase Cdc42, and the assembly of the new bud site (Kozminski et al., 2003). In turn, a novel septin collar is assembled at the new bud neck. In diploid cells, septins similarly contribute to the recruitment of the cortical proteins Bud8 and Bud9, involved in the spatial regulation of Bud5 and Bud1 (Park and Bi, 2007; Schenkman et al., 2002). Thus, in haploids and diploids the role of septins in bud-site selection involves their scaffolding function.

Septins and bud morphogenesis The involvement of septins in bud morphogenesis was first evidenced by the elongated buds formed by septin-defective cells (Hartwell, 1971). This morphology first suggested a role for septins, which are located at the bud neck, in the spatial control of growth deposition within the bud. Normal bud morphogenesis takes place in two steps. During bud emergence, buds grow apically; the localization of growth, i.e. plasma membrane expansion and cell wall remodelling, is restricted to the bud tip. Subsequently, buds grow isotropically, i.e. growth becomes deposited uniformly over the bud surface (Lew and Reed, 1993; Adams and Pringle, 1984). Because of its rigidity, the extension of the cell wall is determined locally by the targeted delivery of new material and cell wall remodelling enzymes by exocytosis (reviewed in Pruyne and Bretscher, 2000a; Pruyne and Bretscher 2000b). Bud morphology is hence determined by the pattern of vesicle delivery, first to the bud tip during bud emergence, apical bud growth and to the entire bud periphery during isotropic bud growth. Vesicles delivery to the plasma membrane is ensured by myosin V-dependent transport along actin cables (Johnston, Prendergast and Singer, 1991; Liu and Bretscher, 1992; Govindan, Bowser and Novick, 1995) reviewed in Pruyne et al., (2004)). Because Myosin V moves its cargos unidirectionally towards the barbed end of microfilaments, vesicles are delivered to the sites of actin-cable nucleation by formins (Pruyne et al., 2002; Evangelista et al., 2002; Sagot et al., 2002), enclosed in the polarizome (Sheu, Barral and Snyder, 2000). Further spatial regulation is provided by the fact that exocytosis of vesicle content also requires the activity of a multiprotein complex, the exocyst (TerBush

THE CELLULAR FUNCTIONS OF SEPTINS: VARIETY AND RECURRENCE

111

et al., 1996) and of its regulators the Rho-like GTPases Rho1 and Cdc42 (Zhang et al., 2001; Roumanie et al., 2005). Thus, the localization of the exocyst and the polarizome determines the distribution of cell growth. Accordingly, at bud emergence exocyst and polarizome co-localize to the bud site, accumulate at the bud tip during apical growth, and redistribute to the entire bud cortex at the shift to isotropic bud growth (reviewed in Pruyne et al. 2004). In addition, bud morphogenesis is coordinated with cell cycle progression (Lew and Reed, 1993). Bud emergence and apical bud growth is initiated in late G1 by Cdk1 and the G1 cyclins Cln1 and Cln2. Subsequently, the B-type cyclins Clb1 and Clb2 trigger the shift to isotropic bud growth during the S phase. Consequently, cells that fail to activate Clb1 and Clb2 keep growing apically and form elongated buds. These observations suggested two models to explain how septins contribute to bud morphogenesis. They might help control the spatial pattern of vesicle delivery inside the bud, or participate to the control of cell cycle progression and the activation of Clb1,2/Cdk1 complexes. The second model is the most accurate.

The morphogenetic checkpoint Beside septin-defective cells, a number of other mutants form elongated buds (Blacketer et al., 1993; Blacketer, Madaule and Myers, 1995; Kellogg and Murray, 1995). Among them, the nap1
, gin4
, hsl1
, hsl7
, and elm1
cells also delay anaphase onset, indicating that cell cycle progression is affected (Barral et al., 1999; Bouquin et al., 2000; Altman and Kellogg, 1997). Both the morphological and cell cycle phenotypes of these mutants are suppressed by disrupting the SWE1 gene, the ortholog of the wee1 kinase in Schizzosccharomyces pombe (Booher, Deshaies and Kirschner, 1993), and by mutation of the tyrosine 19 of Cdk1/Cdc28 to phenylalanine (Barral et al., 1999; Shulewitz, Inouye and Thorner, 1999; Longtine et al., 2000). Tyrosine 19 of Cdk1 is the residue that is phosphorylated by Swe1 in vivo and the Y19F mutation prevents Cdk1 inhibition by Swe1. Thus, the anaphase delay and the elongated bud phenotype of these mutants is caused by Swe1-dependent repression of Clb1,2/Cdk1. Together, these and others data have established a model in which Hsl1, Gin4, Hsl7 and Elm1 act in a cascade that control activation of the kinase Cdk1/Clb1,2 through inhibition of its negative regulator Swe1 (see Figure 4.2). Remarkably, this pathway links Cdk1 activity to the status of septin assembly. Indeed, Hsl1, Hsl7, Gin4, Elm1 and Swe1 all localize to the bud neck during bud growth, in a septin-dependent manner (Carroll et al., 1998; Barral et al., 1999; Shulewitz, Inouye and Thorner, 1999; Moriya and Isono, 1999). Among them, both Hsl1 and Gin4 are SDKs; they directly interact with and depend on septins for activity (Carroll et al., 1998; Barral et al., 1999; Bouquin et al., 2000). Accordingly, anaphase onset is delayed in septin-defective cells and both this delay and the elongated bud phenotypes of septin mutants are suppressed upon disruption of SWE1 (Barral et al., 1999). Thus, the abnormal bud morphology of

CH04 YEAST SEPTINS: A CORTICAL ORGANIZER

112

septin collar

Elm1 ?

Gin4

Hsl1 Hsl7

? Swe1

Cdc28/Clb1,2 Y19

Cdc28/Clb1,2 Y19

anaphase isotropic bud growth

P

Mih1

Slt2/Mpk1

Mkk1, 2

Bck1

Rho1

actin defects

Figure 4.2 Schematic organization of the two branches of the morphogenetic pathway. On the top, the Hsl1/Hsl7 cassette responds to proper septin organization by targeting the Swe1 kinase to degradation. Thereby, it links activation of the Cdk complex Cdc28/Clb1,2 by the Mih1 phosphatase to proper assembly of the septin collar. Cdc28/Clb1,2 can in turn trigger the shift of bud growth from apical to isotropic, as well as the activation of the anaphase promoting complex. From the bottom, the Slt2/Mpk1 MAP kinase pathway prevents Mih1 activation in response to actin defects. Thereby it ensures that no adaptation from Swe1-dependent cell cycle arrest can take place before the bud has reached the proper size.

septin mutants is due to a cell cycle response that maintains the cells in apical bud when the septin collar is disrupted. What is the physiological relevance of linking Swe1-control, cell cycle progression and septins in budding yeast? There are two possible interpretations. On one hand, the septin collar might provide a scaffold to organize and control the

THE CELLULAR FUNCTIONS OF SEPTINS: VARIETY AND RECURRENCE

113

Swe1 pathway. On the other hand, the Swe1 pathway might monitor septins to delay anaphase in response to septin organization defects. Current data are consistent with a Swe1/Mih1 cassette monitoring both actin and septin organization, independently of each other (Figure 4.2). Indeed, cells where actin organization is disrupted early in bud growth arrest prior to anaphase in a Swe1-dependent manner for hours, much longer than septin-defective cells do (McMillan, Sia and Lew, 1998). This arrest depends on the Rho1-Mpk1/Slt2 pathway (Harrison et al., 2001), which acts to inhibit the phosphatase Mih1. Mih1, the homolog of fission yeast cdc25 counteracts Swe1 by removing the inhibitory phosphate that Swe1 puts on Cdk1 (Russell, Moreno and Reed, 1989). Thereby, it stimulates the resumption of cell cycle progression after Swe1-dependent arrest. Thus, actin defects occurring early in bud growth, when the septin collar is not yet fully stabilized, prevent adaptation from Swe1-dependent arrest and impair anaphase onset. Accordingly, actin defects no longer cause a cycle arrest if the cells have progressed beyond the time at which the Hsl1/Hsl7 regulatory cassette becomes active and represses Swe1, thanks to proper septin assembly. In contrast, septin-defective cells are fully proficient for the induction of a Swe1-dependent arrest, yet adapt and escape the arrest after 40 to 50 minutes. Thus, on one hand, the SDK/Swe1 pathway monitors septin organization and prevents activation of late Cdk1 complexes when the septin ring is not properly assembled. On the other hand, the Rho1-Mpk1/Slt2-Mih1 pathway monitors actin function and bud growth and allow adaptation to Swe1-dependent arrest only if the bud has already reached a decent size. In the first pathway, SDKs function as sensors for proper septin organization (Barral et al., 1999; Bouquin et al., 2000; Hanrahan and Snyder, 2003; Carroll et al., 1998). An interesting twist to this story is that the Hsl1/Hsl7 module can therefore be viewed as a sensor for bud emergence. Indeed, septin reorganization following the changes of plasma membrane curvature during bud emergence appears to be necessary for proper activation of the Hsl1 kinase (Theesfeld et al., 2003). However, it must be pointed out that the two lines of thinking (i) Hsl1 monitors proper septin organization or (ii) the Hsl1/septin module monitors the topology of the plasma membrane at the bud neck, are only semantically different. Molecularly speaking, they probably correspond to one same mechanism, in which bud emergence is required for the proper formation of a septin collar that is the only septin structure supporting SDK activation. In any case, the morphogenetic checkpoint derived from these mechanisms ensures that anaphase onset depends on the presence of a bud, of a properly assembled cleavage apparatus and on the cell cortex and the cell wall being correctly organized.

Septin and cell polarity Although preventing anaphase completion in response to defects of the cytokinetic machinery makes biologically sense, it is not immediately clear what the advantages are of preventing the shift to isotropic growth in response to septin

114

CH04 YEAST SEPTINS: A CORTICAL ORGANIZER

assembly defects. The phenotypic characterization of cells lacking both Swe1 and septin function, i.e. of cells undergoing the shift to isotropic bud growth in the absence of the septin collar, helped answer this question. Indeed, although at restrictive temperature the swe1
cdc12-6 double mutant cells fail to form elongated buds, they are still morphologically abnormal. The buds remain small, and the mother cells swell (Barral et al., 2000), indicating that septin defective these cells fail to restrict growth to the bud. Similarly, the swe1
cdc12-6 cells fail to confine the polarizome and the exocyst to the bud only (Barral et al., 2000). Thus, an important function of the septin ring is to help restrict the localization of polarity factors and cell growth to the bud. Thus, the septin collar appears to lay at the bud neck a boundary for the organization of cortical polarity. A major mechanism by which the septin collar bounds the distribution of cortical components is through the formation of a diffusion barrier in the plasma membrane. This has been clearly established for the case of Ist2, an integral membrane protein localized to the bud plasma membrane. FRAP experiments indicated that Ist2 diffuses rapidly within the plasma membrane, yet fails to equilibrate between mother and bud. Upon disruption of the septin ring at the bud neck, however, the distribution of Ist2 equilibrates rapidly between mother and bud, indicating that septins contribute to preventing Ist2 from diffusing through the bud neck (Takizawa et al., 2000). The phenotype of the swe1
cdc12-6 cells indicate that this diffusion barrier helps maintaining the asymmetric distribution of polarity markers. There are at least two possible models for how septins function in the assembly of this diffusion barrier. First, septins might form this barrier themselves (Faty, Fink and Barral, 2002). Septins might directly interact with phospholipids such as phosphatidylinositol 4, 5-bisphosphate (PtdIns(4,5)P(2)) and phosphatidylinositol 3,4,5-trisphosphate (PtdIns(3,4,5)P(3)) (Zhang et al., 1999), which in yeast are required for their localization to the plasma membrane at the bud neck (Casamayor and Snyder, 2003; Rodriguez-Escudero et al., 2005). The many interactions linking septins subunits and phospholipids underneath a single septin filament might therefore lead to two simultaneous outcomes. The filaments would be tightly apposed to the cytoplasmic face of the plasma membrane, impairing the passage of integral plasma membrane proteins with any sizable cytoplasmic domain. These septin-lipid interactions would also recruit and immobilize phospholipids assembling a specialized lipid domain underneath the filament. Such a domain would in turn limit the diffusion of other lipids and the passage of membrane proteins poorly soluble in this lipid phase. Second, septins might not compose the barrier itself but rather be involved in its assembly and maintenance. In this model, other proteins must form the barrier itself. However, we failed so far to identify other bud neck proteins besides septins for being required for the restriction of polarized cortical proteins to the bud. Extensive studies will be required to test both types of models and determine their relative contribution to the control of lateral diffusion of plasma membrane proteins at the bud neck.

MOLECULAR FUNCTIONS FOR YEAST SEPTINS

115

Cellular compartmentalization Cell polarity consists in the asymmetric distribution of cellular factors between two poles of the cell. We recently described a related process, which consists in the compartmentalization of cellular organelles into mother and bud domains. A well-characterized example concerns the ER. While the yeast ER is a single continuous organelle extending throughout the cell cortex and the nuclear envelope, a lateral diffusion barrier restricts the exchange of ER-membrane proteins between mother and bud. This barrier does not seem to play a direct role in cell polarity. First, the proteins subjected to compartmentalization are not polarized. Second, disruption of the barrier in sur2
(Buvelot Frei and Barral, unpublished data), shs1
and bud6
(Luedeke et al., 2005) cells causes no obvious defect in cell polarity. However, ER compartmentalization might help maintaining or reinforcing cellular asymmetry downstream of polarity establishment and bud growth. For example, this barrier might help confining polarized mRNAs (Shepard et al., 2003; Takizawa et al., 2000) to the bud cortex by preventing their diffusion back into the mother cell. Thereby, it might contribute to the bud-specific expression of cell fate determinants such as the transcription factor Ash1 (Bobola et al., 1996). How do septins influence ER organization? Most of the surface of the yeast ER is covered with ribosomes, indicating that yeast cells contain mostly rough ER. At the bud neck, however, a continuous ring of smooth ER covers the septin collar. This sheet disappears in septin defective cells, where it is replaced with rough ER tubules as in the remainder of the cortex. Furthermore, the smooth ER at the bud neck lacks many ER-membrane proteins, such as the translocon and Hmg1 (Luedeke et al., 2005). Thus, smooth ER formation involves the exclusion of these proteins from that ER-domain. The peripheral membrane protein bud6 though appears to be enriched on the bud-neck ER. This protein, which localizes to the bud neck in a septin-dependent manner, is required for the restriction of lateral diffusion in the ER-membrane at the bud neck. Yet, it is not involved in the exclusion of ER-membrane proteins from the bud-neck ER (Luedeke et al., 2005). Thus, barrier formation might be a two steps process, where the septins first direct the formation of a ring of smooth ER at the bud neck, to which Bud6 binds to restrict diffusion. How septins direct smooth ER formation is not yet known. However, cells lacking the SDKs Hsl1 and Gin4 fail to assemble a barrier (Luedeke et al., 2005) and to exclude rough ER markers from neck region. Thus, formation of the barrier throughout the entire ER-membrane appears to be independent of direct septin interactions with the membrane. Rather, it seems to involve the phosphorylation of yet unidentified ER factors by SDKs.

MOLECULAR FUNCTIONS FOR YEAST SEPTINS While we start to know much about septin functions, we still know very little about septin biochemistry. Genetic and cytological studies indicate that septins

116

CH04 YEAST SEPTINS: A CORTICAL ORGANIZER

fulfil their cellular functions in either of two ways: through the recruitment of specialized proteins to the bud neck, or through formation of lateral diffusion barriers in membranes passing the bud neck. In all these processes the main function of septins boils down to establishing, maintaining and signalling the position of the bud neck, defined as both a boundary and a functionally specialized area of the cortex. This septin collar forms a stable anchor point around which the cell organizes. As we will discuss below, these observations already make predictions about what septin’s function(s) must be at the molecular level. Septin function mostly depends on the formation of higher-order structures. First, this appears to be the condition for them to stay at the bud neck and not diffuse away. Second, the individual septins show no functional specialization, and so far no function could be genetically separated from collar assembly. Thus, collar assembly is a pre-requisite for all cellular aspects of septin function. Consequently, most septin-interacting proteins might bind septin filaments or sheets of filaments rather than individual subunits, explaining the difficulty of purifying septin-interacting proteins using traditional biochemistry techniques. Therefore, one fundamental challenge for future studies in septin research is to understand the mechanisms governing septin dynamics, the assembly and disassembly of higher order structures, and the interaction of septin-binding proteins with individual septins and with the septin collar. As mentioned earlier, a key feature of septin sheets resides in their ability to precisely cover the plasma membrane at the bud neck, despite its complex topology and the complex changes it undergoes during bud emergence and cytokinesis. We suspect that the plasticity of septin assemblies lies at the heart of their functions. Particularly, it is very possible that the reorganization that the septin gaze undergoes during the cell cycle serves not only to maintain contact with a changing plasma membrane, but also to guide and control these changes and to coordinate with them the ability of septins to interact with and modulate the activity of other proteins, such as SDKs.

CONCLUSION AND PERSPECTIVES In light of what we start to understand about septin function in yeast, septins form a cytoskeletal system that is fundamentally different from microtubules and microfilaments. Septin structures are poorly dynamic, do not appear to serve as tracks for motor-driven and directional transport of cargos, and are non-polar (John et al., 2007; Sirajuddin et al., 2007). Instead, they are able to mark a specific location for an extended period of time, to tightly interact with membranes, which they help to organize and to serve as a complex scaffold that can recruit and activate proteins in a manner that seem to be coordinated with the topological changes of the underlying membrane. Thus, the yeast septin cytoskeleton forms structures that are more reminiscent of an organizer like centriols and centrosomes, rather than of a force generator as microtubules are. To keep with the skeletal metaphore,

REFERENCES

117

the functional characterization of the microfilaments and the microtubules showed that they are the muscle of the cell. Septins now emerge as the missing bones on which these muscles anchor to articulate the forces that they generate. This pivotal position of septins in the biology of the yeast cell makes the septin collar an outstanding candidate to contribute to many long-term processes in the cell. In that sense, further research on septin biology is likely to provide us many insights about the long process through which the yeast bud emerges and differentiates itself from its mother. Thus, the septin collar has the perfect properties and is exactly at the right place to not only help us understand how cells divide but also how old cells generate successive new-born daughters while they continue to age. In years to come, we predict that septins will turn out to take a central role in cell biology well beyond yeast.

ACKNOWLEDGEMENTS I would like to express my deep gratitude to Stanie Buvelot Frei for her critical reading of the manuscript, her numerous and helpful comments, and her help with establishing and formating the final version of the manuscript. I would also like to acknowledge the members of the septin field for their comments and criticisms throughout our work, and particularly Peter Hall, Hilary Russel and John Pringle for taking on the challenge to edit this book.

REFERENCES Adams, A.E. and Pringle, J.R. (1984) Relationship of actin and tubulin distribution to bud growth in wild-type and morphogenetic-mutant Saccharomyces cerevisiae. The Journal of Cell Biology, 98, 934–45. Altman, R. and Kellogg, D. (1997) Control of mitotic events by Nap1 and the Gin4 kinase. The Journal of Cell Biology, 138, 119–30. Amberg, D.C., Zahner, J.E., Mulholland, J.W. et al. (1997) Aip3p/Bud6p, a yeast actin-interacting protein that is involved in morphogenesis and the selection of bipolar budding sites. Molecular Biology of the Cell , 8, 729–53. An, H., Morrell, J.L., Jennings, J.L. et al. (2004) Requirements of fission yeast septins for complex formation, localization, and function. Molecular Biology of the Cell , 15, 5551–64. Barral, Y., Mermall, V., Mooseker, M.S., Snyder, M. (2000) Compartmentalization of the cell cortex by septins is required for maintenance of cell polarity in yeast. Molecular Cell , 5, 841–51. Barral, Y., Parra, M., Bidlingmaier, S. and Snyder, M. (1999) Nim1-related kinases coordinate cell cycle progression with the organization of the peripheral cytoskeleton in yeast. Genes and Development, 13, 176–87. Bender, A. (1993) Genetic evidence for the roles of the bud-site-selection genes BUD5 and BUD2 in control of the Rsr1p (Bud1p) GTPase in yeast. Proceedings of the National Academy of Sciences of the United States of America, 90, 9926–29.

118

CH04 YEAST SEPTINS: A CORTICAL ORGANIZER

Bender, A. and Pringle, J.R. (1991) Use of a screen for synthetic lethal and multicopy suppressee mutants to identify two new genes involved in morphogenesis in Saccharomyces cerevisiae. Molecular and Cellular Biology, 11, 1295–305. Berlin, A., Paoletti, A. and Chang, F. (2003) Mid2p stabilizes septin rings during cytokinesis in fission yeast. The Journal of Cell Biology, 160, 1083–92. Bi, E. (2001) Cytokinesis in budding yeast: the relationship between actomyosin ring function and septum formation. Cell Structure and Function, 26, 529–37. Bi, E., Maddox, P., Lew, D.J. et al. (1998) Involvement of an actomyosin contractile ring in Saccharomyces cerevisiae cytokinesis. The Journal of Cell Biology, 142, 1301–12. Blacketer, M.J., Koehler, C.M., Coats, S.G. et al. (1993) Regulation of dimorphism in Saccharomyces cerevisiae: involvement of the novel protein kinase homolog Elm1p and protein phosphatase 2A. Molecular and Cellular Biology, 13, 5567–81. Blacketer, M.J., Madaule, P., Myers, A.M. (1995) Mutational analysis of morphologic differentiation in Saccharomyces cerevisiae. Genetics, 140, 1259–75. Bobola, N., Jansen, R.P., Shin, T.H., Nasmyth, K. (1996) Asymmetric accumulation of Ash1p in postanaphase nuclei depends on a myosin and restricts yeast mating-type switching to mother cells. Cell , 84, 699–709. Booher, R.N., Deshaies, R.J., Kirschner, M.W. (1993) Properties of Saccharomyces cerevisiae wee1 and its differential regulation of p34CDC28 in response to G1 and G2 cyclins. The EMBO Journal , 12, 3417–26. Bouquin, N., Barral, Y., Courbeyrette, R. et al. (2000) Regulation of cytokinesis by the Elm1 protein kinase in Saccharomyces cerevisiae. Journal of Cell Science, 113, 1435–45. Byers, B. and Goetsch, L. (1976) A highly ordered ring of membrane-associated filaments in budding yeast. The Journal of Cell Biology, 69, 717–21. Carroll, C.W., Altman, R., Schieltz, D. et al. (1998) The septins are required for the mitosis-specific activation of the Gin4 kinase. The Journal of Cell Biology, 143, 709–17. Casamayor, A. and Snyder, M. (2002) Bud-site selection and cell polarity in budding yeast. Current Opinion in Microbiology, 5, 179–86. Casamayor, A. and Snyder, M. (2003) Molecular dissection of a yeast septin: distinct domains are required for septin interaction, localization, and function. Molecular and Cellular Biology, 23, 2762–77. Caviston, J.P., Longtine, M., Pringle, J.R. and Bi, E. (2003) The role of Cdc42p GTPase-activating proteins in assembly of the septin ring in yeast. Molecular Biology of the Cell , 14, 4051–66. Chant, J., Mischke, M., Mitchell, E. et al. (1995) Role of Bud3p in producing the axial budding pattern of yeast. The Journal of Cell Biology, 129, 767–78. Cid, V.J., Adamikova, L., Sanchez, M. et al. (2001) Cell cycle control of septin ring dynamics in the budding yeast. Microbiology, 147, 1437–50. De Virgilio, C., DeMarini, D.J. and Pringle, J.R. (1996) SPR28 , a sixth member of the septin gene family in Saccharomyces cerevisiae that is expressed specifically in sporulating cells. Microbiology, 142 (Pt 10), 2897–905. Dobbelaere, J. and Barral, Y. (2004) Spatial coordination of cytokinetic events by compartmentalization of the cell cortex. Science, 305, 393–96.

REFERENCES

119

Dobbelaere, J., Gentry, M.S., Hallberg, R.L. and Barral, Y. (2003) Phosphorylationdependent regulation of septin dynamics during the cell cycle. Developmental Cell , 4, 345–57. Evangelista, M., Pruyne, D., Amberg, D.C. et al. (2002) Formins direct Arp2/3independent actin filament assembly to polarize cell growth in yeast. Nature Cell Biology, 4, 260–69. Fares, H., Goetsch, L. and Pringle, J.R. (1996) Identification of a developmentally regulated septin and involvement of the septins in spore formation in Saccharomyces cerevisiae. The Journal of Cell Biology, 132, 399–411. Faty, M., Fink, M. and Barral, Y. (2002) Septins: a ring to part mother and daughter. Current Genetics, 41, 123–31. Field, C.M. and Alberts, B.M. (1995) Anillin, a contractile ring protein that cycles from the nucleus to the cell cortex. The Journal of Cell Biology, 131, 165–78. Finger, F.P., Hughes, T.E. and Novick, P. (1998) Sec3p is a spatial landmark for polarized secretion in budding yeast. Cell , 92, 559–71. Ford, S.K. and Pringle, J.R. (1991) Cellular morphogenesis in the Saccharomyces cerevisiae cell cycle: localization of the CDC11 gene product and the timing of events at the budding site. Developmental Genetics, 12, 281–92. Frazier, J.A., Wong, M.L., Longtine, M.S. et al. (1998) Polymerization of purified yeast septins: evidence that organized filament arrays may not be required for septin function. The Journal of Cell Biology, 143, 737–49. Fujita, A., Oka, C., Arikawa, Y. et al. (1994) A yeast gene necessary for bud-site selection encodes a protein similar to insulin-degrading enzymes. Nature, 372, 567–70. Gladfelter, A.S., Kozubowski, L., Zyla, T.R. and Lew, D.J. (2005) Interplay between septin organization, cell cycle and cell shape in yeast. Journal of Cell Science, 118, 1617–28. Gladfelter, A.S., Pringle, J.R. and Lew, D.J. (2001) The septin cortex at the yeast mother-bud neck. Current Opinion in Microbiology, 4, 681–89. Glotzer, M. (2005) The molecular requirements for cytokinesis. Science, 307, 1735–39. Govindan, B., Bowser, R. and Novick, P. (1995) The role of Myo2, a yeast class V myosin, in vesicular transport. The Journal of Cell Biology, 128, 1055–68. Grava, S., Schaerer, F., Faty, M. et al. (2006) Asymmetric recruitment of dynein to spindle poles and microtubules promotes proper spindle orientation in yeast. Developmental Cell , 10, 425–39. Guertin, D.A., Trautmann, S. and McCollum, D. (2002) Cytokinesis in eukaryotes. Microbiology and Molecular Biology Reviews, 66, 155–78. Haarer, B.K. and Pringle, J.R. (1987) Immunofluorescence localization of the Saccharomyces cerevisiae CDC12 gene product to the vicinity of the 10-nm filaments in the mother-bud neck. Molecular and Cellular Biology, 7, 3678–87. Hanrahan, J. and Snyder, M. (2003) Cytoskeletal activation of a checkpoint kinase. Molecular Cell , 12, 663–73. Harrison, J.C., Bardes, E.S., Ohya, Y. and Lew, D.J. (2001) A role for the Pkc1p/ Mpk1p kinase cascade in the morphogenesis checkpoint. Nature Cell Biology, 3, 417–20.

120

CH04 YEAST SEPTINS: A CORTICAL ORGANIZER

Hartwell, L.H. (1971) Genetic control of the cell division cycle in yeast. IV. Genes controlling bud emergence and cytokinesis. Experimental Cell Research, 69, 265–76. Hartwell, L.H., Culotti, J., Pringle, J.R. and Reid, B.J. (1974) Genetic control of the cell division cycle in yeast. Science, 183, 46–51. Hartwell, L.H., Culotti, J. and Reid, B. (1970) Genetic control of the cell-division cycle in yeast. I. Detection of mutants. Proceedings of the National Academy of Sciences of the United States of America, 66, 352–59. Hime, G.R., Brill, J.A. and Fuller, M.T. (1996) Assembly of ring canals in the male germ line from structural components of the contractile ring. Journal of Cell Science, 109, 2779–88. Huisman, S.M., Bales, O.A., Bertrand, M. et al. (2004) Differential contribution of Bud6p and Kar9p to microtubule capture and spindle orientation in S. cerevisiae. The Journal of Cell Biology, 167, 231–44. Iwase, M., Luo, J., Bi, E. and Toh, E.A. (2007) Shs1 plays separable roles in septin organization and cytokinesis in Saccharomyces cerevisiae. Genetics, 177, 215–29. John, C.M., Hite, R.K., Weirich, C.S. et al. (2007) The Caenorhabditis elegans septin complex is nonpolar. The EMBO Journal , 26, 3296–307. Johnston, G.C., Prendergast, J.A. and Singer, R.A. (1991) The Saccharomyces cerevisiae MYO2 gene encodes an essential myosin for vectorial transport of vesicles. The Journal of Cell Biology, 113, 539–51. Joo, E., Tsang, C.W. and Trimble, W.S. (2005) Septins: traffic control at the cytokinesis intersection. Traffic, 6, 626–34. Kellogg, D.R. and Murray, A.W. (1995) NAP1 acts with Clb1 to perform mitotic functions and to suppress polar bud growth in budding yeast. The Journal of Cell Biology, 130, 675–85. Kim, H.B., Haarer, B.K. and Pringle, J.R. (1991) Cellular morphogenesis in the Saccharomyces cerevisiae cell cycle: localization of the CDC3 gene product and the timing of events at the budding site. The Journal of Cell Biology, 112, 535–44. Kinoshita, M., Field, C.M., Coughlin, M.L. et al. (2002) Self- and actin-templated assembly of Mammalian septins. Developmental Cell , 3, 791–802. Kinoshita, M., Kumar, S., Mizoguchi, A. et al. (1997) Nedd5, a mammalian septin, is a novel cytoskeletal component interacting with actin-based structures. Genes and Development, 11, 1535–47. Kinoshita, M. and Noda, M. (2001) Roles of septins in the mammalian cytokinesis machinery. Cell Structure and Function, 26, 667–70. Korinek, W.S., Bi, E., Epp, J.A. et al. (2000) Cyk3, a novel SH3-domain protein, affects cytokinesis in yeast. Current Biology, 10, 947–50. Kozminski, K.G., Beven, L., Angerman, E. et al. (2003) Interaction between a Ras and a Rho GTPase couples selection of a growth site to the development of cell polarity in yeast. Molecular Biology of the Cell , 14, 4958–70. Krapp, A., Gulli, M.P. and Simanis, V. (2004) SIN and the art of splitting the fission yeast cell. Current Biology, 14, R722–30. Kremer, B.E., Haystead, T. and Macara, I.G. (2005) Mammalian septins regulate microtubule stability through interaction with the microtubule-binding protein MAP4. Molecular Biology of the Cell , 16, 4648–59.

REFERENCES

121

Kusch, J., Meyer, A., Snyder, M.P. and Barral, Y. (2002) Microtubule capture by the cleavage apparatus is required for proper spindle positioning in yeast. Genes and Development, 16, 1627–39. Lew, D.J. and Reed, S.I. (1993) Morphogenesis in the yeast cell cycle: regulation by Cdc28 and cyclins. The Journal of Cell Biology, 120, 1305–20. Lippincott, J. and Li, R. (1998a) Dual function of Cyk2, a cdc15/PSTPIP family protein, in regulating actomyosin ring dynamics and septin distribution. The Journal of Cell Biology, 143, 1947–60. Lippincott, J. and Li, R. (1998b) Sequential assembly of myosin II, an IQGAP-like protein, and filamentous actin to a ring structure involved in budding yeast cytokinesis. The Journal of Cell Biology, 140, 355–66. Lippincott, J., Shannon, K.B., Shou, W. et al. (2001) The Tem1 small GTPase controls actomyosin and septin dynamics during cytokinesis. Journal of Cell Science, 114, 1379–86. Liu, H. and Bretscher, A. (1992) Characterization of TPM1 disrupted yeast cells indicates an involvement of tropomyosin in directed vesicular transport. The Journal of Cell Biology, 118, 285–99. Longtine, M.S. and Bi, E. (2003) Regulation of septin organization and function in yeast. Trends in Cell Biology, 13, 403–9. Longtine, M.S., DeMarini, D.J., Valencik, M.L. et al. (1996) The septins: roles in cytokinesis and other processes. Current Opinion in Cell Biology, 8, 106–19. Longtine, M.S., Theesfeld, C.L., McMillan, J.N. et al. (2000) Septin-dependent assembly of a cell cycle-regulatory module in Saccharomyces cerevisiae. Molecular and Cellular Biology, 20, 4049–61. Luca, F.C., Mody, M., Kurischko, C. et al. (2001) Saccharomyces cerevisiae Mob1p is required for cytokinesis and mitotic exit. Molecular and Cellular Biology, 21, 6972–83. Luedeke, C., Frei, S.B., Sbalzarini, I. et al. (2005) Septin-dependent compartmentalization of the endoplasmic reticulum during yeast polarized growth. The Journal of Cell Biology, 169, 897–908. Ma, X.J., Lu, Q. and Grunstein, M. (1996) A search for proteins that interact genetically with histone H3 and H4 amino termini uncovers novel regulators of the Swe1 kinase in Saccharomyces cerevisiae. Genes and Development, 10, 1327–40. Maddox, A.S., Lewellyn, L., Desai, A. and Oegema, K. (2007) Anillin and the septins promote asymmetric ingression of the cytokinetic furrow. Developmental Cell , 12, 827–35. McMillan, J.N., Sia, R.A. and Lew, D.J. (1998) A morphogenesis checkpoint monitors the actin cytoskeleton in yeast. The Journal of Cell Biology, 142, 1487–99. Menssen, R., Neutzner, A. and Seufert, W. (2001) Asymmetric spindle pole localization of yeast Cdc15 kinase links mitotic exit and cytokinesis. Current Biology, 11, 345–50. Mino, A., Tanaka, K., Kamei, T. et al. (1998) Shs1p: a novel member of septin that interacts with Spa2p, involved in polarized growth in Saccharomyces cerevisiae. Biochemical and Biophysical Research Communications, 251, 732–36. Moriya, H. and Isono, K. (1999) Analysis of genetic interactions between DHH1, SSD1 and ELM1 indicates their involvement in cellular morphology determination in Saccharomyces cerevisiae. Yeast, 15, 481–96.

122

CH04 YEAST SEPTINS: A CORTICAL ORGANIZER

Nguyen, T.Q., Sawa, H., Okano, H. and White, J.G. (2000) The C. elegans septin genes, unc-59 and unc-61, are required for normal postembryonic cytokineses and morphogenesis but have no essential function in embryogenesis. Journal of Cell Science, 113 (Pt 21), 3825–37. Oegema, K., Savoian, M.S., Mitchison, T.J. and Field, C.M. (2000) Functional analysis of a human homologue of the Drosophila actin binding protein anillin suggests a role in cytokinesis. The Journal of Cell Biology, 150, 539–52. Park, H.O. and Bi, E. (2007) Central roles of small GTPases in the development of cell polarity in yeast and beyond. Microbiology and Molecular Biology Reviews, 71, 48–96. Pringle, J. (2008) The History of the Septin Field. This volume. Pruyne, D. and Bretscher, A. (2000a) Polarization of cell growth in yeast. Journal of Cell Science, 113 (Pt 4), 571–85. Pruyne, D. and Bretscher, A. (2000b) Polarization of cell growth in yeast. I. Establishment and maintenance of polarity states. Journal of Cell Science, 113 (Pt 3), 365–75. Pruyne, D., Evangelista, M., Yang, C. et al. (2002) Role of formins in actin assembly: nucleation and barbed-end association. Science, 297, 612–15. Pruyne, D., Legesse-Miller, A., Gao, L. et al. (2004) Mechanisms of polarized growth and organelle segregation in yeast. Annual Review of Cell and Developmental Biology, 20, 559–91. Rodriguez-Escudero, I., Roelants, F.M., Thorner, J. et al. (2005) Reconstitution of the mammalian PI3K/PTEN/Akt pathway in yeast. The Biochemical Journal , 390, 613–23. Roemer, T., Madden, K., Chang, J. and Snyder, M. (1996) Selection of axial growth sites in yeast requires Axl2p, a novel plasma membrane glycoprotein. Genes and Development, 10, 777–93. Roumanie, O., Wu, H., Molk, J.N. et al. (2005) Rho GTPase regulation of exocytosis in yeast is independent of GTP hydrolysis and polarization of the exocyst complex. The Journal of Cell Biology, 170, 583–94. Russell, P., Moreno, S. and Reed, S.I. (1989) Conservation of mitotic controls in fission and budding yeasts. Cell , 57, 295–303. Sagot, I., Rodal, A.A., Moseley, J. et al. (2002) An actin nucleation mechanism mediated by Bni1 and profilin. Nature Cell Biology, 4, 626–31. Sanders, S.L. and Herskowitz, I. (1996) The BUD4 protein of yeast, required for axial budding, is localized to the mother/BUD neck in a cell cycle-dependent manner. The Journal of Cell Biology, 134, 413–27. Schenkman, L.R., Caruso, C., Page, N. and Pringle, J.R. (2002) The role of cell cycle-regulated expression in the localization of spatial landmark proteins in yeast. The Journal of Cell Biology, 156, 829–41. Segal, M., Clarke, D.J., Maddox, P. et al. (2000) Coordinated spindle assembly and orientation requires Clb5p-dependent kinase in budding yeast. The Journal of Cell Biology, 148, 441–52. Shaw, S.L., Yeh, E., Maddox, P. et al. (1997) Astral microtubule dynamics in yeast: a microtubule-based searching mechanism for spindle orientation and nuclear migration into the bud. The Journal of Cell Biology, 139, 985–94. Shepard, K.A., Gerber, A.P., Jambhekar, A. et al. (2003) Widespread cytoplasmic mRNA transport in yeast: identification of 22 bud-localized transcripts using DNA

REFERENCES

123

microarray analysis. Proceedings of the National Academy of Sciences of the United States of America, 100, 11429–34. Sheu, Y.J., Barral, Y. and Snyder, M. (2000) Polarized growth controls cell shape and bipolar bud site selection in Saccharomyces cerevisiae. Molecular and Cellular Biology, 20, 5235–47. Sheu, Y.J., Santos, B., Fortin, N. et al. (1998) Spa2p interacts with cell polarity proteins and signaling components involved in yeast cell morphogenesis. Molecular and Cellular Biology, 18, 4053–69. Shulewitz, M.J., Inouye, C.J. and Thorner, J. (1999) Hsl7 localizes to a septin ring and serves as an adapter in a regulatory pathway that relieves tyrosine phosphorylation of Cdc28 protein kinase in Saccharomyces cerevisiae. Molecular and Cellular Biology, 19, 7123–37. Sirajuddin, M., Farkasovsky, M., Hauer, F. et al. (2007) Structural insight into filament formation by mammalian septins. Nature, 449, 311–15. Spiliotis, E.T., Kinoshita, M. and Nelson, W.J. (2005) A mitotic septin scaffold required for Mammalian chromosome congression and segregation. Science, 307, 1781–85. Surka, M.C., Tsang, C.W. and Trimble, W.S. (2002) The mammalian septin MSF localizes with microtubules and is required for completion of cytokinesis. Molecular Biology of the Cell , 13, 3532–45. Takizawa, P.A., DeRisi, J.L., Wilhelm, J.E. and Vale, R.D. (2000) Plasma membrane compartmentalization in yeast by messenger RNA transport and a septin diffusion barrier. Science, 290, 341–44. Tanaka, S. and Nojima, H. (1996) Nik1: a Nim1-like protein kinase of S. cerevisiae interacts with the Cdc28 complex and regulates cell cycle progression. Genes Cells, 1, 905–21. Tasto, J.J., Morrell, J.L. and Gould, K.L. (2003) An anillin homologue, Mid2p, acts during fission yeast cytokinesis to organize the septin ring and promote cell separation. The Journal of Cell Biology, 160, 1093–103. TerBush, D.R., Maurice, T., Roth, D. and Novick, P. (1996) The Exocyst is a multiprotein complex required for exocytosis in Saccharomyces cerevisiae. The EMBO Journal , 15, 6483–94. Theesfeld, C.L., Zyla, T.R., Bardes, E.G. and Lew, D.J. (2003) A monitor for bud emergence in the yeast morphogenesis checkpoint. Molecular Biology of the Cell , 14, 3280–91. Tolliday, N., Bouquin, N. and Li, R. (2001) Assembly and regulation of the cytokinetic apparatus in budding yeast. Current Opinion in Microbiology, 4, 690–95. Vallen, E.A., Caviston, J. and Bi, E. (2000) Roles of Hof1p, Bni1p, Bnr1p, and myo1p in cytokinesis in Saccharomyces cerevisiae. Molecular Biology of the Cell , 11, 593–611. Versele, M. and Thorner, J. (2004) Septin collar formation in budding yeast requires GTP binding and direct phosphorylation by the PAK, Cla4. The Journal of Cell Biology, 164, 701–15. Vrabioiu, A.M. and Mitchison, T.J. (2006) Structural insights into yeast septin organization from polarized fluorescence microscopy. Nature, 443, 466–69. Vrabioiu, A.M. and Mitchison, T.J. (2007) Symmetry of septin hourglass and ring structures. Journal of Molecular Biology, 372, 37–49.

124

CH04 YEAST SEPTINS: A CORTICAL ORGANIZER

Wu, J.Q., Kuhn, J.R., Kovar, D.R. and Pollard, T.D. (2003) Spatial and temporal pathway for assembly and constriction of the contractile ring in fission yeast cytokinesis. Developmental Cell , 5, 723–34. Zhang, X., Bi, E., Novick, P. et al. (2001) Cdc42 interacts with the exocyst and regulates polarized secretion. The Journal of Biological Chemistry, 276, 46745–50. Zhang, J., Kong, C., Xie, H. et al. (1999) Phosphatidylinositol polyphosphate binding to the mammalian septin H5 is modulated by GTP. Current Biology, 9, 1458–67.

5 Septins in four model fungal systems: diversity in form and function Amy S. Gladfelter Department of Biology, Dartmouth College, Hanover, NH, USA

Peter Sudbery Department of Molecular Biology and Biotechnology, Sheffield University, Western Bank, Sheffield S10 2TN, UK

INTRODUCTION After their initial discovery by the analysis of genes found in the screen for cell-division-cycle mutants in the yeast Saccharomyces cerevisiae (Hartwell, 1971; Longtine et al., 1996), septin proteins have been identified based on sequence similarities in a variety of fungal genomes and indeed appear to be ubiquitous in the fungal kingdom (Lindsey and Momany, 2006; Pan, Malmberg and Momany, 2007). Septin orthologues clearly emerge in the amino acid sequences based on characteristic regions such as guanosine triphosphate (GTP)-binding and coiled– coiled domains. However, a remarkable feature of the fungal septins is the apparent diversity of functions, localizations and higher-order structures found in different cell types. In this chapter, we focus on septin function in four fungal species in which septins have been analysed in some depth. We begin with the fission yeast Schizosaccharomyces pombe, which grows exclusively as a unicellular, uninucleate yeast; then we consider the pathogen Candida albicans which switches between yeast, pseudohyphal, and true hyphal

The Septins Edited by Peter A. Hall, S.E. Hilary Russell and John R. Pringle © 2008 John Wiley & Sons, Ltd. ISBN: 978-0-470-51969-1

126

CH05 SEPTINS IN FOUR MODEL FUNGAL SYSTEMS: DIVERSITY IN FORM AND FUNCTION

Table 5.1

Summary of septin proteins in four fungi

Organism

Septins

Localization

Function

S. pombe

Spn1p-4p Spn5p-7p (sporulation specific)

Spn1-4 : Double ring (apparently derived from initial single ring) assembles at division plane in anaphase. Spn5-7: pro-spore membrane. Yeast cells and pseudohyphae: assemble at prebud site and form collar at mother-bud neck Hyphae: Band of septin ‘bars’ transiently present at base of germ tube (lacks Cdc3p), cap at hyphal tips, rings within germ tubes Always assembled as discrete, filamentous bars of variable length depending on maturity of septin ring. Appear at hyphal tips, along hyphae, and at the bases of lateral branches. AspB: localizes post-mitotically as a single ring at septation sites; matures to a double ring; then the ring on apical side of septa persists while ring on basal side disassembles at septation.

Efficient septation (non-essential). Orderly pro-spore membrane extension.

C. albicans

Cdc3p, 10p, 11p, 12p, Sep7p, Spr3p, Spr28p

A. gossypii

Cdc3p, 10p, 11ap, 11bp, 12p, Sep7p, Spr3p, Spr28p

A. nidulans

AspA, B, C, D, E

Cell separation, morphogenesis, exocytosis (Cdc3p and Cdc12p essential).

Mitosis control, sporulation, hyphal morphogenesis, septum formation (non-essential).

Chitin deposition in septa, branching pattern, conidiophore development (AspB non-essential).

morphologies; and we conclude with two fungi, Ashbya gossypii and Aspergillus nidulans, that grow exclusively as multinucleate filaments (summarized in Table 5.1). The connections and comparisons between these different species will illustrate the plasticity of form and function displayed by septins within the fungal kingdom.

SCHIZOSACCHAROMYCES POMBE

127

SCHIZOSACCHAROMYCES POMBE: A UNINUCLEATE FISSION YEAST S. pombe has rod-shaped cells formed through polarized growth at the cell ends followed by division in the middle of the cell. The molecular mechanisms driving this reproducible morphogenetic program have been studied intensively, and much detail is known about the basis for polarized growth and cytokinesis in S. pombe (Brunner and Nurse, 2000). Notably, fission yeast cells polarize growth and execute cytokinesis using mechanisms that appear to have some fundamental differences from those in the similarly well-studied S. cerevisiae, as might have been expected given their significant evolutionary divergence (on the order of 900 million years) (Heckman et al., 2001). Likewise, the higher-order structure, assembly, and function of septins in S. pombe share some striking similarities but also display many differences from those of S. cerevisiae. The S. pombe genome contains seven genes encoding septins. Spn1p, 2p, 3p and 4p are expressed in vegetatively growing cells, whereas Spn5p, 6p and 7p are expressed exclusively during sporulation where they (along with Spn2p but not Spn1p, Spn3p, or Spn4p) are important for the normal formation of the forespore membrane and spore wall (Longtine et al., 1996; Mata et al., 2002; Rustici et al., 2004; M. Onishi, Y. Fukui and J. Pringle, personal communication). Notably, there is no Shs1p/Sep7p orthologue apparent in the genome or present in septin complexes when purified from vegetatively grown cells (An et al., 2004). Intriguingly, Spn7p, which in overall sequence is most similar to septins in the Cdc11 group, appears to have lost all GTP-binding motifs and lacks any coiled-coil domain but nonetheless still seems to localize and function together with the other septins during sporulation (J. Pringle, personal communication), calling into question the absolute importance of these domains for septin function. Spn1p-4p share relatively low amino-acid-sequence identity with their orthologues in ¡i>S. cerevisiae¡/i>, which are ScCdc3p, 10p, 11p and 12p respectively (Longtine et al., 1996); strikingly, Spn2p, like Cdc10p, also lacks a predicted coiled-coil domain. Despite amino acid identities of only 42–52 %, an antibody raised against ScCdc3p recognized Spn1p, and some S. pombe septins can complement the function of S. cerevisiae septins inactivated by temperature-sensitive mutations (H.B. Kim. T. Pugh, and J. Pringle, personal communication). Purified Spn1p-4p assemble into a predicted linear array composed of subunits of homodimers of each septin protein associating in the order Spn3-4-1-2 (An et al., 2004), similar to the predicted arrangement of S. cerevisiae septins (Versele et al., 2004; Bertin et al., 2008) and mammalian septins (Sirajuddin et al., 2007). Although these data indicate some fundamental structural conservation in the septin assemblies, there are also some striking differences in the localization and function of septins in fission yeast compared to S. cerevisiae. In the latter, septins form a ring at the prebud site (the future site of cytokinesis) during late G1 of

128

CH05 SEPTINS IN FOUR MODEL FUNGAL SYSTEMS: DIVERSITY IN FORM AND FUNCTION

Nucleus Septin ring

Figure 5.1 Septin-ring assembly during the S. pombe cell cycle. The septin ring in S. pombe assembles in mitosis in an apparent single ring that then matures into a double ring upon the onset of anaphase. The double ring persists until cell separation

the cell cycle and arrive at this site before other components of the cytokinetic machinery such as actin and type II myosin (Kim, Haarer and Pringle, 1991; Bi et al., 1998; Iwase et al., 2006). In contrast, in S. pombe, the septins coalesce into a ring around the middle of the cell only in mitosis after many other cytokinesis proteins have already gathered at that site (Berlin, Paoletti and Chang, 2003; Tasto, Morrell and Gould, 2003; An et al., 2004); Figure 5.1). Notably, Spn1p-4p do not appear to be functionally equivalent, presumably reflecting their differential positions and roles in septin filament assembly (see above). Deletion of spn1 or spn4 leads to the loss of all organized septins at the division site; whereas in the absence of Spn2p or 3p, the remaining septins still mostly find the cell middle, although additional ectopic, septin assemblies appear in these mutants in other locations in the cell cortex, and the medial ring does not mature normally (An et al., 2004); (John Pringle, personal communication). Thus, Spn1p, 4p and at least one other vegetative septin protein are required for normal assembly and maturation of the ring. Although initial assembly may be as a single ring, it is soon apparent that the septins form a double ring that brackets the medial actomyosin ring and persists until septin-ring disassembly at cell separation (Berlin, Paoletti

SCHIZOSACCHAROMYCES POMBE

129

and Chang, 2003; An et al., 2004). The septin rings do not contract but instead maintain a constant diameter at all times while other cytokinetic factors invaginate within the domain bounded by the septin ring (Berlin, Paoletti and Chang, 2003); (J.Q. Wu and J. Pringle, personal communication). Moreover, fluorescence recovery after photobleaching (FRAP) of Spn4-GFP, suggests that the septin subunits are stably incorporated within the ring rather than rapidly exchanging with free subunits in the cytosol (Berlin, Paoletti and Chang, 2003). The septin ring retains its original diameter through cytokinesis in both S. cerevisiae and S. pombe, yet is essential in the former case but not in the latter. Thus, the ability of the septin structure to invaginate with the cleavage furrow (as seen in many animal cells) is not a reliable predictor of the essentiality of septins in cytokinesis. Interestingly, the initial accumulation of septins in the cell middle appears to be distinctly regulated from the later steps of ring organization and maturation to a double ring. The septins can congregate at the cortex in the middle of the cell in the absence of the protein Mid2p, an anaphase-specific protein that is weakly similar in sequence to the S. cerevisiae Bud4p (involved in bud-site selection and perhaps in septin organization at the bud site); however, Mid2p is required to form the compact septin rings (Berlin, Paoletti and Chang, 2003; Tasto, Morrell and Gould, 2003). In cells lacking Mid2p, the dynamics of the septin assembly also change dramatically such that there is a 30-fold increase in the exchangability of septins in the ring compared to the stable rings of wild-type cells (Berlin, Paoletti and Chang, 2003). Additionally, instead of the clear double ring, the septins assemble into a ‘disc’ or ‘washer’-like higher-order structure that seems to track the cleavage furrow rather than maintaining its dimensions as acquired at assembly (Berlin, Paoletti and Chang, 2003; Tasto, Morrell and Gould, 2003). Similarly, overexpression of Mid2p leads to persistence of the septin rings (Tasto, Morrell and Gould, 2003; An et al., 2004). Thus, fission yeast septin-ring organization and function late in mitosis are tightly regulated either directly or indirectly by Mid2p. Despite this carefully coordinated assembly and maturation of the septin rings at the site of cytokinesis, the septins are not essential for cytokinesis in fission yeast. Loss of Spn1p or Spn4p produces cells with a delay in cell separation that is more severe than that in cells lacking Spn2p or Spn3p, which only show defects under certain stress conditions (Longtine et al., 1996); (O. Al-Awar, M. Valencik and J. Pringle, personal communication). This difference in the severity of phenotype parallels the differential effects on septin localization as seen in these mutants and further supports the idea that not all septin subunits are functionally equivalent in fission yeast (An et al., 2004). The mild phenotypes are not due to functional redundancy among the septins, because even a sextuple mutant lacking Spn1p-5p and Spn7p has a phenotype indistinguishable from an spn1 or spn4 mutant (J.-Q. Wu, J. B¨ahler, and J. Pringle, personal communication). Septin mutants form normal primary and secondary septa based on transmission electron micrographs however hydrolytic enzymes (Eng1p and Agn1p) are mislocalized to a broad disk instead of a tight ring in spn4 mutants (Martin-Cuadrado et al., 2005). These data suggest that septin mutants may have

130

CH05 SEPTINS IN FOUR MODEL FUNGAL SYSTEMS: DIVERSITY IN FORM AND FUNCTION

defects in targeting the hydrolytic enzymes that help execute cell separation. The delay in cell separation leads to an appearance of chains of cells linked together with prominent but not separated septa. The extent of cell-separation delay varies from cell to cell and septum to septum in septin mutant populations. Based on genetic interactions between the S. pombe septin mutations and mutations effecting factors in the cell-wall-integrity pathway (which appears to be similar to the well-studied pathway in S. cerevisiae (Levin, 2005)), it has been hypothesized that the variable kinetics of cell separation may be due to variable rates of repair of partially defective septa formed in the absence of a normal septin assembly (J.Q. Wu and J. Pringle, personal communication).

CANDIDA ALBICANS: A POLYMORPHIC HUMAN PATHOGEN C. albicans is a fungus that is usually found growing harmlessly in the human gastrointestinal and genitourinary tract. However, it can also be an important pathogen, causing conditions that range from painful mucosal infections of the vagina in otherwise healthy women to severe, life-threatening, blood-stream infections among vulnerable groups such as newborns and certain intensive-care patients, especially those undergoing cancer chemotherapy, immunosuppressant therapy or catheterization (Kibbler et al., 2003). A striking feature of C. albicans pathogenesis and biology is its ability to switch among several morphological forms (Figure 5.2). These range from unicellular budding yeasts to true hyphae with parallel-sided walls. Between these two extremes, C. albicans can exhibit a variety of growth forms that are filamentous but retain a constriction between adjacent cellular compartments. These are

(a)

(b)

(c)

Figure 5.2 Morphological forms of C. albicans. Pseudohyphae (a), yeast (b), hyphae (c). Like hyphae, pseudohyphal cells are polarized, but they have constrictions at the sites of septation (arrows), Hyphae are thinner than pseudohyphae, and there are no constrictions at sites of septation (arrowed and inset)

CANDIDA ALBICANS

131

collectively referred to as pseudohyphae (Sudbery, 2001; Berman and Sudbery, 2002; Sudbery, Gow and Berman, 2004). In yeast form cells, daughter buds separate from the mother cell after cytokinesis. In contrast, the daughters remain associated after cytokinesis in pseudohyphae, but they are readily separated by brief sonication or by the mild mechanical shearing forces present in shaken liquid cultures. In hyphae, the cellular compartments remain firmly attached to each other and are not separated even by vigorous shaking or prolonged sonication. The differences between yeast, pseudohyphae and hyphae have recently been reviewed in detail Sudbery, Gow and Berman, (2004). Notably, septins assemble into different higher-order structures depending on the cell type. Therefore, the same set of septin proteins can be differentially regulated and have different appearances and functions, depending upon cell fate.

Septin organization The C. albicans genome contains seven septin genes, CDC3, CDC10, CDC11, CDC12 , SEP7, SPR3 and SPR28 (Warenda and Konopka, 2002). The gene designations follow the S. cerevisiae nomenclature, where SEP7 is the homologue of S. cerevisiae SHS1/SEP7. The similarities with the S. cerevisiae homologues are 54 % for Cdc3p, 69 % for Cdc10p, 48 % for Cdc11p, 61 % for Cdc12p and 33 % for Sep7p. CDC3 and CDC12 are essential for viability, but cdc10
/
and cdc11
/
mutants are viable although they show conditional morphological and cell-wall defects (Warenda and Konopka, 2002). (C. albicans is an obligate diploid, hence both copies of a gene must be deleted to generate a null mutant.) The homology is lower in the case of SPR3 (34 %) and SPR28 (33 %). In S. cerevisiae, expression of SPR3 and SPR28 is meiosis-specific, and their function in C. albicans is unclear since meiosis has not yet been demonstrated in this organism. The organization of septin structures is different in yeast and pseudohyphae compared to hyphae, and this reflects a fundamental difference in the regulation of the cell cycle in these morphologies (Figure 5.3; Plate 5.1 [see p. 246 for Plates]) (Sudbery, 2001; Warenda and Konopka, 2002; Sudbery, Gow and Berman, 2004). In yeast and pseudohyphae, a septin patch forms at the site of bud evagination composed of Cdc3p, Cdc10p, Cdc11p and Cdc12p. As the bud forms, the septin complex remains at the neck, forming first a ring and then a collar covering the bud neck, similar to what occurs in S. cerevisiae. Mitosis takes place across the plane of the bud neck (Colour plate 5.1a and b). At cytokinesis the septin collar splits into two as it organizes the formation of the primary and secondary septa. A hyphal germ tube evaginates from a mother yeast cell before the Start of the cell cycle (Figure 5.3). Evagination is accompanied by the formation of a septin patch which remains at the germ-tube neck (Sudbery, 2001; Warenda and Konopka, 2002). However, this septin complex does not mature into a ring; rather, it forms a band of longitudinal septin bars (Figure 5.4; Colour plate 5.1c). This structure is known as the basal septin band (Berman and Sudbery, 2002) and is

132

CH05 SEPTINS IN FOUR MODEL FUNGAL SYSTEMS: DIVERSITY IN FORM AND FUNCTION

Yeast

Pseudohyphae

Hyphae

Septin collar

Basal septin band

Septin ring

Septin Cap

Nucleus

Primary septum

Figure 5.3 Septin organization in C. albicans yeast, pseudohyphae and hyphae. In yeast and pseudohyphae, a septin ring forms just prior to bud emergence. As the bud forms, the septins spread to form a collar or hourglass structure at the bud neck. Mitosis takes place across the plane of the septin collar, which splits into two during cytokinesis. In hyphae, a septin patch forms at the same time as, or just before, the emergence of the hyphal germ tube. As the germ tube elongates, the septin complex becomes a disorganized band of longitudinal bars its base and a cap of septins is present at the hyphal tip. As the cell cycle commences, a septin ring forms from the septin cap and remains in place as the germ tube continues to elongate while the basal septin band fades and disappears. At the end of the first cell cycle, the nucleus migrates out of the mother cell into the germ tube, and mitosis takes place across the septin ring. After anaphase, one nucleus migrates back into the mother cell. The septin ring then splits into two as the primary and secondary septa form

CANDIDA ALBICANS

(a)

(b)

133

(c)

(d)

Figure 5.4 Formation of the septin ring during germ-tube emergence in C. albicans. Frames taken from a time-lapse movie. (a) A young germling displays a basal septin band and septin cap. (b) A faint ring forms from the cap. (c) The ring becomes brighter and the cap and basal band fade. (d). The septin ring remains stationary as the germ tube elongates, and the cap and basal band completely disappear (Images courtesy of Laura Jones)

similar to the structures seen at the base of S. cerevisiae mating projections (shmoos) (Ford and Pringle, 1991) and vegetative cells lacking Gin4p, Nap1p, or Cla4p (Longtine, Fares and Pringle, 1998; Longtine et al., 2000). In addition to the basal band, a septin cap is visible at the germ-tube tip (Figure 5.4; Colour plate 5.1c). As the germ tube elongates, a septin ring forms from this cap so that septin ring is present within the germ tube rather than at its neck (Figure 5.4, Colour plate 5.1d). This septin ring is not sharply defined and is probably equivalent to the septin collar at the base of a yeast or pseudohyphal bud (Colour plate 5.1d, inset). The basal band and septin cap disappear after the septin ring forms (Figure 5.4), possibly due to competition between the different structures for septin subunits. The band and cap persist longer in gin4
/
mutants, which can form the basal septin band and cap but not the septin ring within the hypha (Wightman et al., 2004) (see below). The nucleus migrates out of the mother cell, mitosis takes place within the germ tube, and one daughter nucleus then migrates back into the mother cell (Colour plate 5.1d). After anaphase, the septin ring splits into two sharply defined rings and the primary septum forms between the two rings (Colour plate 5.1e).

Control of septin organization by the protein kinases Gin4p and Ccn1p-Cdc28p The role of the basal septin band in hyphal cells is unclear, but it is possible that it is involved in determining the shape of the germ-tube base, which lacks a constriction in contrast to that present between the mother cell and a bud in yeast or pseudohyphae. As noted above, the basal septin band resembles the septin array in S. cerevisiae mutants lacking Gin4p suggesting that the formation of the band does not require Gin4p. Indeed, in Gin4p-depleted hyphae, the basal band formed normally but the septin ring failed to assemble (Wightman et al., 2004). Thus, the basal band and septin ring have different genetic requirements for their formation and maintenance. Interestingly, in Gin4p-depleted cells, nuclear migration and mitosis occurred normally within the germ tube. Thus, the septin ring is not the

134

CH05 SEPTINS IN FOUR MODEL FUNGAL SYSTEMS: DIVERSITY IN FORM AND FUNCTION

spatial marker that determines where mitosis will occur. Furthermore, in wild-type hyphae the septin rings form before nuclear migration is initiated; therefore nuclear position cannot be the spatial cue for septin ring formation. Thus, there must be at least one independent spatial marker determining the location of the septin ring and position where mitosis will occur in C. albicans hyphae. What is the substrate of Gin4p that is crucial for septin-ring formation? Recent work has demonstrated that Gin4p can directly phosphorylate the septin Cdc11p at Ser395 (Sinha et al., 2007). This phosphorylated Cdc11p is now ‘primed’ to be a substrate for Cdc28p (a cyclin-dependent kinase) in complex with the Ccn1p cyclin, which phosphorylates the adjacent Ser394 residue. If phosphorylation of Ser394 is prevented, either by deletion of CCN1 or by a cdc11 S394A mutation, germ tubes evaginate normally and septins localize to a ring. However, after the septin ring forms, growth becomes isotropic, resulting in a swollen hyphal tip. Tip swelling in a ccn1
/
strain is rescued by cdc11 S394D S395D phosphomimetic mutations, showing that phosphorylation of Cdc11p is the only Ccn1p function required for normal hyphal development. Cdc11p phosphorylation by Gin4p is a cell cycle regulated event occurring during the mitosis and cytokinesis phases of the previous cell cycle. Thus, Ser395 is already phosphorylated when unbudded yeast cells are induced to form hyphae. After hyphal induction, Ccn1p-Cdc28p associates with the septin complex within 5–10 minutes. This rapid response appears to be independent of the signal-transduction pathways that promote the hyphal-specific pattern of transcription, because it occurs in an efg1
/
cph1
/
mutant. This suggests that at least one early event of hyphal morphogenesis is independent of the hyphal-specific transcriptional response. Under the alias CLN1 , CCN1 had previously been shown to be essential for the maintenance but not the establishment of hyphal growth (Loeb et al., 1999), consistent with the conclusions of Sinha et al. (2007). However, it should be noted that phylogenetic analysis has shown that Ccn1p is not an orthologue of S. cerevisiae Cln1p and indeed is not a member of the S. cerevisiae G1 (Cln1-3p) or G2 (Clb1-6p) cyclin families (unpublished observations). The phosphorylation of Cdc11p does not appear to be important for septin organization but clearly impacts septin-ring function. How does the phosphorylation at early times after hyphal induction promote normal hyphal morphogenesis much later after the septin ring has formed? Polarized growth in fungi involves the directed flow of secretory vesicles to the site of growth to generate new cell wall and membrane upon exocytosis. Septins have been implicated in directing exocytosis in several systems (Beites et al., 1999; Beites, Peng and Trimble, 2001; Beites, Campbell and Trimble, 2005; Spiliotis and Nelson, 2006). It has been suggested that septins play a role in generating the characteristic hourglass shape of the bud neck of S. cerevisiae yeast cells by targeting secretion to the bud base, the resulting lateral growth producing the characteristic swelling on the daughter side of the neck (Gladfelter et al., 2005). Although C. albicans yeast cells have a similar hourglass shape at the bud neck, there is no constriction at the site of the septin ring in hyphae. Thus, the propensity of the septin ring to target secretion

CANDIDA ALBICANS

135

must be suppressed in hyphae. Perhaps phosphorylation of Ser395 of Cdc11p by Ccn1p-Cdc28p is a critical modification that limits septin–exocyst interactions. The exocycst is a multi-protein structure to which secretory vesicles dock before fusion with the plasma membrane. In S. cerevisiae, Sec3p acts as the spatial cue for the formation of the exocyst: localization of other exocyst components depends on Sec3p, but Sec3p localization is autonomous of the other components (Finger and Novick, 1997; Finger, Hughes and Novick, 1998). Support for a mechanism involving regulation of septin–exocyst associations in hyphae is provided by the observation that the exocyst component Sec3p co-immunoprecipitates with the septin Cdc11p in C. albicans (Li et al., 2007). Furthermore, sec3
/
mutants could initiate apparently normal germ tubes, but when the septin ring formed, growth became isotropic and the tip became swollen – a phenotype reminiscent of that of the ccn1
/
and cdc11 S394A mutants. However, this tip swelling did not occur in a cdc10
/
sec3
/
double mutant (Li et al., 2007). A model that accomodates all of these data is as follows. At early stages in hyphal development, growth is directed towards the tip. When the septin ring forms, continued polarized growth depends on tip localization of Sec3p. However, Sec3p affinty for Cdc11p in the ring complex leads to competition for Sec3p between the tip and the septin ring. Phosphorylation of Cdc11p Ser394, which depends on the formation of a septin complex that includes Cdc10p, weakens the affinity of Sec3p for Cdc11p. Thus, the tip localization of Sec3p dominates, thereby promoting polarized growth at the tip at the expense of lateral growth directed from the septin ring.

Control of septin organization by Cdc42p GDP/GTP cycling In S. cerevisiae, the Cdc42p Rho-type guanosine triphosphatase (GTPase) controls many aspects of morphogenesis including the formation of the septin ring at the bud neck (Johnson, 1999; Park and Bi, 2007). Like other GTPases, Cdc42p cycles between GTP- and guanosine diphosphate (GDP)-bound forms. Hydrolysis of GTP is enhanced by guanosine triphosphatase-activating proteins (GAPs) that bring Cdc42p to the GDP-bound form. The C. albicans genome encodes two Cdc42 GAPs, Rga2p and Bem3p (named by homology to their S. cerevisiae orthologues). Because the GTPases are normally active in the GTP-bound form, Cdc42p is expected to be hyperactive in cells lacking Rga2p or Bem3p. In rga2
/
bem3
/
cells grown under pseudohyphal-promoting conditions, the septin ring was found to form within the germ tube instead of at the bud neck, suggesting that the position of the septin ring is determined by the level of Cdc42p activation (Court and Sudbery, 2007). Furthermore, such cells also assumed a more hyphal-like shape, and a Spitzenk¨orper (a cluster of vesicles characteristic of the tips of hyphae) was present, indicating that Cdc42p also promoted the extreme polarized growth characteristic of hyphae. However, when Cdc42p was locked into the GTP-bound form by the conditional expression of a cdc42 G12V allele, cells grown in pseudohyphal-promoting conditions were swollen, and the septins formed prominent bars at the bud neck. The differences in phenotype between

136

CH05 SEPTINS IN FOUR MODEL FUNGAL SYSTEMS: DIVERSITY IN FORM AND FUNCTION

cells with Cdc42p activated either through loss of the GAPs or through being locked in the GTP-bound state suggest that Cdc42p cycling between GDP- and GTP-bound forms is required for the maturation of the septin collar. This conclusion is consistent with studies in S. cerevisiae that suggest that formation of the septin ring is a two-stage process: Cdc42p-GTP recruits septin subunits to form a patch, but maturation into a collar requires Cdc42p cycling between the GDP- and GTP-bound forms together with the action of Gin4p, Nap1p, Cla4p, Gic1p and Gic2p (Gladfelter et al., 2002; Caviston et al., 2003; Gladfelter, Zyla and Lew, 2004; Iwase et al., 2006). In C. albicans, germ-tube evagination from the mother cell occurs before the cell cycle is initiated (Hazan, Sepulveda-Becerra and Liu, 2002). Presumably, Cdc42p-GTP at the site of evagination will recruit septin subunits, but in the absence of Rga2p and Bem3p these subunits do not mature into a collar. When the cell cycle is initiated, the expression of RGA2 and BEM3 facilitates Cdc42 GDP/GTP cycling to promote maturation of the septin subunits into the collar that forms within the growing germ tube. In summary, septins clearly play important roles in generating the different morphological forms of C. albicans. As in S. cerevisiae, they organize the formation of primary and secondary septa during cytokinesis and cell cycle regulated phosphorylation may modify the septin ring in hyphae to prevent it from programming the lateral growth that produces the constriction in yeast and pseudohyphal buds.

Sep7 modifies the septin ring in hyphae to abrogate the cell-separation program Cell separation occurs after cytokinesis in C. albicans yeast and pseudohyphae, but not in hyphae. In the yeast form, Cdc14p localizes to the site of septation and is required for the localization and activation of the Ace2p-dependent transcriptional program that results in the synthesis of hydrolytic enzymes that will destroy the primary septum (Clemente-Blanco et al., 2006). In contrast, Cdc14p does not localize to the septation site in hyphae and the resulting absence of the Ace2p-dependent transcription program explains the failure of hyphal compartments to separate (Clemente-Blanco et al., 2006). FRAP experiments show that the septin subunits Cdc3p, Cdc12p and Sep7 are stable in both hyphae and yeast, that is, once the septin ring forms; it does not exchange subunits with the free cytoplasmic pool. However, in hyphae, but not yeast, the septin Cdc10p does exchange subunits with the cytoplasmic pool (Gonzalez-Novo et al., 2008). This dynamic property of Cdc10p in hyphae depends on the septin Sep7p. In sep7
/
mutants, the septin ring assumes the properties of the yeast ring – Cdc10p in the septin ring is stabilized, Cdc14p is recruited and adjacent compartments separate after cytokinesis. The hyphal properties of the septin ring also depend on Hgc1p, a G1 cyclin that is only expressed in hyphae. The phosphorylation pattern of Sep7p is different, and total Sep7p levels are reduced in an hgc1
/
mutant. This suggests that Hgc1p-Cdc28p phosphorylates and stabilizes Sep7p, allowing it to modify the

ASHBYA GOSSYPII

137

properties of the septin ring to abrogate the cell-separation program. This is one of the first examples in fungi where the post-translational modification of a single septin subunit has been shown to change both the dynamics and function of the septin ring in a normal developmental program.

ASHBYA GOSSYPII: A MULTINUCLEATE, FILAMENTOUS HEMIASCOMYCETE A. gossypii is an ascomycete that unlike C. albicans does not switch between different morphologies but rather is constitutively filamentous. Based on their genome sequences, A. gossypii is predicted to have diverged from a common ancestor with S. cerevisiae approximately 100 million years ago. About 90 % of the A. gossypii genome is homologous and syntenic to S. cerevisiae (Dietrich et al., 2004) so that A. gossypii genes are named according to their syntenic homologues in S. cerevisiae. While the overall sets of proteins are similar, the sequence identity between orthologous proteins is often less than 50 %, consistent with the observed divergence in morphology and lifestyles between the two organisms. A. gossypii grows exclusively as filaments in which nuclear division occurs in the absence of cytokinesis, thereby producing large syncytia in which many nuclei share a common cytoplasm (Gladfelter, Hungerbuehler and Philippsen, 2006). It has never been observed to form uninucleate yeast in which a single nuclear division is coupled to a single cell division. Thus, A. gossypii is an excellent model for understanding septin functions unrelated to cytokinesis. Eight septin genes are found in the A. gossypii genome, including homologues to all of the S. cerevisiae vegetative and sporulation-specific septins. An additional CDC11 (called b) gene lies adjacent to the ancestoral CDC11 (called a) and is probably a product of a tandem duplication occurring after the divergence from S. cerevisiae (Dietrich et al., 2004). Preliminary expression data suggest that CDC11b may only be expressed under specific growth conditions (Peter Philippsen, personal communication). Identity between the S. cerevisiae and A. gossypii septins ranges from 58 % (for Cdc3p) to 78 % (for Cdc12p), and typical septin domains including the GTP-binding domain, predicted phosphoinositidebinding elements and coiled-coil domains in the C-termini are clearly identifiable in the A. gossypii sequences.

Morphologically diverse septin organizations assemble within one cell Septins assemble and mature through morphologically distinct cortical organizations in A. gossypii hyphal cells (Figure 5.5, Colour plate 5.2). Septins initially assemble into a loose, elongated band of filaments at the growing tips of hyphae (Colour plate 5.2, panels a and b). Upon an unidentified signal, this band coalesces into a more compact and apparently organized ring that becomes fixed to

138

CH05 SEPTINS IN FOUR MODEL FUNGAL SYSTEMS: DIVERSITY IN FORM AND FUNCTION

1

2

5

3

Spore

4 Nucleus Septin rings

Figure 5.5 Morphologically diverse septin organizations assemble in A. gossypii. In the cortex of a single, multinucleate A. gossypii cell septins appear in a variety of different organizations. Diffuse filamentous septins appear at the growing tips of hyphae (1) and then the filaments compress into more compact rings made of bars of septin protein (2). The consolidation of the septin filament structure is accompanied by an ‘anchoring’ of the ring at the cortex and detachment from the growing tip. The bars of septin protein persist for variable times before the bars elongate and the single ring transforms into two adjacent rings that persist (3). Septins also congregate in a cap at emerging branches (4); this pool of protein remains at the cortex at the base of the branch, forming a ring made of bars. This ring then appears to split into two rings while the growing branch tip assembles new septin filaments (5). These two rings at the branch base are asymmetric, with very faint septin bars visible at the side of the main hyphae and more compact and bright bars of septins on the branch side (5). (Schematic based on observations of Sep7p-GFP in living cells and Cdc11p by immunofluorescence; see Colour plate 5.2)

ASHBYA GOSSYPII

139

one location on the cortex while tip growth continues (Colour plate 5.2, panels c and d). Unlike the smooth septin collars observed to date in other fungi, this newly ‘anchored’ ring remains in a filamentous organization of discrete short bars that dot the cortex, reminiscent of the transient basal septin band seen in C. albicans germlings (Figures 5.4 and 5.5, Colour plate 5.2, panels c, d, f and g). Meanwhile, the growing tip assembles a new band of filaments. In some cases, the bars split when a septum is built and this doublet ring then persist. Diffuse clouds of short septin filaments begin to coalesce prior to branch emergence on the sides of hyphae, and these organize into a ring as branch hyphae emerge (Colour plate 5.2, panels e and f). These rings persist at the base of the branch and develop an asymmetric organization, while new filaments assemble at the growing branch tips (Colour plate 5.2, panels g and h). In any single septin mutant (cdc3
, cdc10
, cdc11a
or cdc12
), Sep7p does not assemble into any of these organized structures (Helfer and Gladfelter, 2006); (B. DeMay and A. Gladfelter, unpublished data). Septin-ring assembly and maturation occur simultaneously in multiple spots on the cortex of A. gossypii hyphae. Thus, a common cytoplasm underlies these apparently independent septin-organization events. Current questions about septin localization and organization that are under investigation in this system include: is there a maturation program built into the organization of septin rings that essentially acts as a timer to execute the changes in appearance without additional input? Do the rings respond to internal and/or external cues to modulate their organization? What are the differences in septin-subunit composition and dynamics of the different organizations? What regulatory proteins act on the septin cortex to direct changes in organization? Do these regulatory factors respond to signals from the nuclear division cycle? Answers to these questions will greatly expand our understanding of how different molecular signals are translated into diverse higher-order septin structures.

Septin function in spatial control of mitosis Although deletion of any septin gene except CDC11b perturbs assembly of an organized septin ring, A. gossypii cells lacking any one septin are viable and show only a 30 % reduction in radial growth rate (Helfer and Gladfelter, 2006). Hyphal morphology is somewhat aberrant in cells lacking septins, and the phenotypes of septin null mutants are especially pronounced in mature mycelia, in which hyphae become kinked and ‘wavy’ and the incomplete septa that normally form between multinucleate compartments are absent. In addition, deletion of CDC3, CDC10, CDC11a or CDC12 led to sporulation defects suggesting these septins contribute to normal spore morphogenesis (H. Helfer, P. Philippsen and A. Gladfelter, unpublished data). The viable phenotypes of septin mutants, combined with a cell cycle lacking cytokinesis, make A. gossypii a powerful organism for uncovering novel cellular roles for septins.

140

CH05 SEPTINS IN FOUR MODEL FUNGAL SYSTEMS: DIVERSITY IN FORM AND FUNCTION

One novel function of the septins in A. gossypii has emerged from studies of the spatial control of mitosis in these hyphal cells. In multinucleate hyphae, mitosis is asynchronous such that neighbouring nuclei divide independently despite residing in a common cytoplasm (Gladfelter, 2006). Notably, however, the locations where nuclei most frequently divide do not appear to be random. In particular, it was shown recently that the septin cortex can direct the position of mitosis such that nuclear division preferentially occurs near assembled septin rings, particularly at branch points (Helfer and Gladfelter, 2006). In wild-type cells, 45 % of mitoses occur near branch points, whereas in septin mutants only 12 % of mitoses occur near branches and instead nuclei appear to divide randomly throughout hyphae independent of morphological features. How does the septin cortex create mitosis-promoting zones within the continuous cytoplasm of hyphae? Genetic evidence and localization data support a model in which the septins locally reduce the activity of the cyclin-dependent kinase antagonist, Swe1p (Wee1 homologue) by concentrating its negative regulators (Helfer and Gladfelter, 2006). This localized inhibition of Swe1p then favours division of nuclei in the vicinity of the septin cortex. This signalling network of septins-Swe1p-regulators and Swe1p is also responsive to external nutrient status, and we hypothesize that the septins therefore integrate both internal and external signals to promote localized mitoses. The septins enable A. gossypii cells to control precisely where new nuclei are made, which may be key for responding to localized nutrient gradients in the natural environment. Thus, in this system, the septin cortex links the cell interior and exterior by transmitting and responding to a variety of signals. In summary, A. gossypii , despite its close relationship on the genome level to S. cerevisiae, shows distinct septin organizations and seemingly unique septin functions. These features of the septin cortex probably reflect the multinucleate, hyphal architecture of these cells. Not only do these studies show how different septins assemblies can appear between species but they also indicate how diverse higher-order septin structures can assemble and mature within a single cell. Future analyses of the bases for the varied septin rings in A. gossypii hold much promise for understanding how specific post-translational modifications to septins lead to changes in septin organization.

ASPERGILLUS NIDULANS: A TRUE FILAMENTOUS ASCOMYCETE A. nidulans is a filamentous ascomycete that shows diverse morphologies during vegetative growth and the formation of asexual spores by conidiation (Harris and Momany, 2004). After germination of a spore, several rounds of mitosis occur while a hyphal germ tube emerges by polarized growth. The first septation site is formed after the third mitosis, and as the hypha grows, compartments containing 2–6 nuclei will be generated that are demarcated by additional septa. Only the

CONCLUSIONS

141

compartments at the tips or from which branches emerge contain mitotically active nuclei. Powerful genetics and well-developed cell biological tools have allowed A. nidulans to make major contributions to understanding of the eukaryotic cell cycle, polarized growth, septation and metabolism. Analysis of the septins in A. nidulans has begun to impact nearly all of these diverse areas of cell biology, and the unique morphology of A. nidulans has placed it in a central position for understanding septin function outside of the uninucleate yeasts. A. nidulans has five genes encoding septins, which correspond to the core vegetative septins in S. cerevisae plus an additional septin that is unique to other filamentous fungi although is not present in A. gossypii. The septins (and their orthologues) are named AspA (Cdc11p), AspB (Cdc3p), AspC (Cdc12p), AspD (Cdc10p) and AspE (filamentous-fungus specific) (Lindsey and Momany, 2006; Pan, Malmberg and Momany, 2007).

Localization and function AspB has been most intensively investigated for localization and function (Momany and Hamer, 1996; Westfall and Momany, 2002). AspB was shown to localize to newly forming septa as a single ring that colocalizes with the actin ring (Figure 5.6). As in S. pombe, the AspB ring does not contract but maintains a constant diameter over time. As the septum matures, two rings of AspB become visible, and then only a single ring on the apical (growth tip) side of the septum persists while the ring on the basal side disassembles. Thus, intriguingly, the septin rings react to the overall spatial organization of these cells and display different patterns of organization relative to the axis of polarity. Based on analysis of nuclear-cycle stages in the apical compartment, AspB ring assembly appears to occur post-mitotically. In contrast, AspB assembles into rings that ‘anticipate’ the sites of branching in subapical compartments, before the onset of mitosis. Thus, depending on the morphogenetic context, septin-ring assembly can be either preor post-mitotic in A. nidulans (Figure 5.6). Although AspB was initially thought to be essential, cells lacking this protein are viable, although they grow slowly (R. Lindsey and M. Momany, personal communication). aspB deletion mutants and mutants harbouring a conditional allele of aspB are hyperbranching, form septa deficient in chitin, and have aberrant conidiophores (Momany and Hamer, 1996; Westfall and Momany, 2002). Thus, even in these multinucleate cells in which a single cell division is not coordinated with a single nuclear division, septins are integral to normal growth and morphogenesis.

CONCLUSIONS Studies to date in these four morphologically diverse fungi have demonstrated a tantalizing spectrum of septin organizations and functions. One unifying principle is that septins assemble where septa form. However, notable differences

142

CH05 SEPTINS IN FOUR MODEL FUNGAL SYSTEMS: DIVERSITY IN FORM AND FUNCTION

4

3

2

1 Nucleus arrested in interphase (basal) Post-mitotic nucleus (apical) 5 Septin rings

Chitin ring, mature septum

Figure 5.6 Septin rings assemble post-mitotically in the apical compartment and pre-mitotically at branching sites in A. nidulans. The thick disk (1) represents a mature septum that separates the growing apical compartment from the basal compartments and this lacks any septin proteins. The ring adjacent to this (2) is the youngest septum in the growing apical compartment and the septins are shown here as they initially assemble as a single ring. This single septin ring splits into a double ring as the septum matures (3). This double ring then is dismantled asymmetrically, so that only the half of the ring closest to the tip persists, while the more basal ring dissipates (shown as the discontinuous ring in 4). Septin rings also assemble at sites of lateral branch emergence(s) before the nuclei in this compartment re-enter the cell cycle (5). (From data in Westfall and Momany, 2002)

REFERENCES

143

are seen in how the assembly of septin rings coordinates with the timing of the cell cycle, and in some cases, in how septins actually influence the timing of the nuclear cycle. The exact higher-order structures vary both between species and also between single cell types or even single cells of the same species. This suggests a tremendous regulatory flexibility in how the septin subunits may interact. Furthermore, whether or not the septins are essential varies from organism to organism, and this does not simply reflect whether the cells are uninucleate or filamentous or if the septin assemblage contracts or remains a fixed diameter at cell division. Finally, the septins contribute to diverse processes in these different fungi, including regulation of the timing of mitosis, the polarized growth of hyphae, hyphal morphology and the formation of spores. Thus, work in these four diverse fungi demonstrates that septin function varies as much as their forms of assembly and each system holds great promise for addressing the mechanisms of septin function and organization.

REFERENCES An, H., Morrell, J.L., Jennings, J.L. et al. (2004) Requirements of fission yeast septins for complex formation, localization, and function. Molecular Biology of the Cell , 15, 5551–64. Beites, C.L., Campbell, K.A. and Trimble, W.S. (2005) The septin Sept5/CDCrel-1 competes with α-SNAP for binding to the SNARE complex. The Biochemical Journal , 385, 347–53. Beites, C.L., Peng, X.R. and Trimble, W.S. (2001) Expression and analysis of properties of septin CDCrel-1 in exocytosis. Methods in Enzymology, 329, 499–510. Beites, C.L., Xie, H., Bowser, R. and Trimble, W.S. (1999) The septin CDCrel-1 binds syntaxin and inhibits exocytosis. Nature Neuroscience, 2, 434–39. Berlin, A., Paoletti, A. and Chang, F. (2003) Mid2p stabilizes septin rings during cytokinesis in fission yeast. Journal of Cell Biology, 160, 1083–92. Berman, J. and Sudbery, P.E. (2002) Candida albicans: a molecular revolution built on lessons from budding yeast. Nature Reviews. Genetics, 3, 918–30. Bertin, A., McMurray, M.A., Grob, P. et al. (2008) Saccharomyces cerevisiae septins: supramolecular organization of heterooligomers and the mechanism of filament assembly. Proc Natl Acad Sci USA, 105 (24), 8274–9. Bi, E., Maddox, P., Lew, D.J. et al. (1998) Involvement of an actomyosin contractile ring in Saccharomyces cerevisiae cytokinesis. Journal of Cell Biology, 142, 1301–12. Brunner, D. and Nurse, P. (2000) New concepts in fission yeast morphogenesis. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences, 355, 873–77. Caviston, J.P., Longtine, M., Pringle, J.R. and Bi, E. (2003) The role of Cdc42p GTPase-activating proteins in assembly of the septin ring in yeast. Molecular Biology of the Cell , 14, 4051–66. Clemente-Blanco, A., Gonz´alez-Novo, A., Mach´ın, F., Caballero-Lima, D., Arag´on, L., S´anchez, M., de Aldana, C.R., Jim´enez, J., Correa-Bordes, J. (2006) The Cdc14p phosphatase affects late cell-cycle events and morphogenesis in Candida albicans. Journal of Cell Science, 119 (6), 1130–43.

144

CH05 SEPTINS IN FOUR MODEL FUNGAL SYSTEMS: DIVERSITY IN FORM AND FUNCTION

Court, H. and Sudbery, P. (2007) Regulation of Cdc42 GTPase activity in the formation of hyphae in Candida albicans. Molecular Biology of the Cell , 18, 265–81. Dietrich, F.S., Voegeli, S., Brachat, S. et al. (2004) The Ashbya gossypii genome as a tool for mapping the ancient Saccharomyces cerevisiae genome. Science, 304, 304–7. Finger, F.P., Hughes, T.E. and Novick, P. (1998) Sec3p is a spatial landmark for polarized secretion in budding yeast. Cell , 92, 559–71. Finger, F.P. and Novick, P. (1997) Sec3p is involved in secretion and morphogenesis in Saccharomyces cerevisiae. Molecular Biology of the Cell, 8, 647–62. Ford, S.K. and Pringle, J.R. (1991) Cellular morphogenesis in the Saccharomyces cerevisiae cell cycle: localization of the CDC11 gene product and the timing of events at the budding site. Developmental Genetics, 12, 281–92. Gladfelter, A.S. (2006) Nuclear anarchy: mitosis in multinucleated cells. Current Opinion in Microbiology, 9 (6), 547–52. Gladfelter, A.S., Bose, I., Zyla, T.R. et al. (2002) Septin ring assembly involves cycles of GTP loading and hydrolysis by Cdc42p. Journal of Cell Biology, 156, 315–26. Gladfelter, A.S., Hungerbuehler, A.K. and Philippsen, P. (2006) Asynchronous nuclear division cycles in multinucleated cells. Journal of Cell Biology, 172, 347–62. Gladfelter, A.S., Kozubowski, L., Zyla, T.R. and Lew, D.J. (2005) Interplay between septin organization, cell cycle and cell shape in yeast. Journal of Cell Science, 118, 1617–28. Gladfelter, A.S., Zyla, T.R. and Lew, D.J. (2004) Genetic interactions among regulators of septin organization. Eukaryotic Cell , 3, 847–54. Gonzalez-Novo, A., Correa-Bordes, J., Labrador, L., S´anchez, M., V´azquez de Aldana, C.R., Jim´enez, J. (2008) Sep7 is essential to modify septin ring dynamics and inhibit cell separation during Candida albicans hyphal growth. Molecular Biology of the Cell , 19 (4), 1509–18. Harris, S.D. and Momany, M. (2004) Polarity in filamentous fungi: moving beyond the yeast paradigm. Fungal Genetics and Biology, 41, 391–400. Hartwell, L.H. (1971) Genetic control of the cell division cycle in yeast. IV. Genes controlling bud emergence and cytokinesis. Experimental Cell Research, 69, 265–76. Hazan, I., Sepulveda-Becerra, M. and Liu, H. (2002) Hyphal elongation is regulated independently of cell cycle in Candida albicans. Molecular Biology of the Cell , 13, 134–45. Heckman, D.S., Geiser, D.M., Eidell, B.R. et al. (2001) Molecular evidence for the early colonization of land by fungi and plants. Science, 293, 1129–33. Helfer, H. and Gladfelter, A.S. (2006) AgSwe1p regulates mitosis in response to morphogenesis and nutrients in multinucleated Ashbya gossypii cells. Molecular Biology of the Cell , 17, 4494–512. Iwase, M., Luo, J., Nagaraj, S. et al. (2006) Role of a Cdc42p effector pathway in recruitment of the yeast septins to the presumptive bud site. Molecular Biology of the Cell , 17, 1110–25. Johnson, D.I. (1999) Cdc42: an essential Rho-type GTPase controlling eukaryotic cell polarity. Microbiology and Molecular Biology Reviews, 63, 54–105. Kibbler, C.C., Seaton, S., Barnes, R.A. et al. (2003) Management and outcome of bloodstream infections due to Candida species in England and Wales. Journal of Hospital Infection, 54, 18–24.

REFERENCES

145

Kim, H.B., Haarer, B.K. and Pringle, J.R. (1991) Cellular morphogenesis in the Saccharomyces cerevisiae cell cycle: localization of the CDC3 gene product and the timing of events at the budding site. Journal of Cell Biology, 112, 535–44. Levin, D.E. (2005) Cell wall integrity signaling in Saccharomyces cerevisiae. Microbiology and Molecular Biology Reviews, 69, 262–91. Li, C.R., Lee, R.T., Wang, Y.M. et al. (2007) Candida albicans hyphal morphogenesis occurs in Sec3p-independent and Sec3p-dependent phases separated by septin ring formation. Journal of Cell Science, 120, 1898–907. Lindsey, R. and Momany, M. (2006) Septin localization across kingdoms: three themes with variations. Current Opinion in Microbiology, 9, 559–65. Loeb, J.D., Sepulveda-Becerra, M., Hazan, I. and Liu, H. (1999) A G1 cyclin is necessary for maintenance of filamentous growth in Candida albicans. Molecular Cell Biology, 19, 4019–27. Longtine, M.S., DeMarini, D.J., Valencik, M.L. et al. (1996) The septins: roles in cytokinesis and other processes. Current Opinion in Cell Biology, 8, 106–19. Longtine, M.S., Fares, H. and Pringle, J.R. (1998) Role of the yeast Gin4p protein kinase in septin assembly and the relationship between septin assembly and septin function. Journal of Cell Biology, 143, 719–36. Longtine, M.S., Theesfeld, C.L., McMillan, J.N. et al. (2000) Septin-dependent assembly of a cell-cycle-regulatory module in Saccharomyces cerevisiae. Molecular Cell Biology, 20 (11), 4049–61. Martin-Cuadrado, A.B., Morrell, J.L., Konomi, M. et al. (2005) Role of septins and the exocyst complex in the function of hydrolytic enzymes responsible for fission yeast cell separation. Molecular Biology of the Cell , 16, 4867–81. Mata, J., Lyne, R., Burns, G. and Baehler, J. (2002) The transcriptional program of meiosis and sporulation in fission yeast. Nature Genetics, 32, 143–47. Momany, M. and Hamer, J.E. (1996) The Aspergillus nidulans septin encoding gene, aspB, is essential for growth. Fungal Genetics and Biology, 21, 92–100. Pan, F., Malmberg, R.L. and Momany, M. (2007) Analysis of septins across kingdoms reveals orthology and new motifs. BMC Evolutionary Biology [Electronic Resource] , 7, 103. Park, H.-O. and Bi, E. (2007) Central roles of small GTPases in the development of cell polarity in yeast and beyond. Microbiology and Molecular Biology Reviews, 71, 48–96. Rustici, G., Mata, J., Kivinen, K. et al. (2004) Periodic gene expression program of the fission yeast cell cycle. Nature Genetics, 36, 809–17. Sinha, I., Wang, Y.M., Philp, R. et al. (2007) Cyclin-dependent kinases control septin phosphorylation in Candida albicans hyphal development. Developmental Cell , 13, 421–32. Sirajuddin, M., Farkasovsky, M., Hauer, F. et al. (2007) Structural insight into filament formation by mammalian septins. Nature, 449, 311–15. Spiliotis, E.T. and Nelson, W.J. (2006) Here come the septins: novel polymers that coordinate intracellular functions and organization. Journal of Cell Science, 119, 4–10. Sudbery, P.E. (2001) The germ tubes of Candida albicans hyphae and pseudohyphae show different patterns of septin ring localization. Molecular Microbiology, 41, 19–31. Sudbery, P., Gow, N. and Berman, J. (2004) The distinct morphogenic states of Candida albicans. Trends in Microbiology, 12, 317–24.

146

CH05 SEPTINS IN FOUR MODEL FUNGAL SYSTEMS: DIVERSITY IN FORM AND FUNCTION

Tasto, J.J., Morrell, J.L. and Gould, K.L. (2003) An anillin homologue, Mid2p, acts during fission yeast cytokinesis to organize the septin ring and promote cell separation. Journal of Cell Biology, 160, 1093–103. Versele, M., Gullbrand, B., Shulewitz, M.J. et al. (2004) Protein-protein interactions governing septin heteropentamer assembly and septin filament organization in Saccharomyces cerevisiae. Molecular Biology of the Cell , 15, 4568–83. Warenda, A.J. and Konopka, J.B. (2002) Septin function in Candida albicans morphogenesis. Molecular Biology of the Cell , 13, 2732–46. Westfall, P.J. and Momany, M. (2002) Aspergillus nidulans septin AspB plays preand postmitotic roles in septum, branch, and conidiophore development. Molecular Biology of the Cell , 13, 110–18. Wightman, R., Bates, S., Amornrrattanapan, P. and Sudbery, P. (2004) In Candida albicans, the Nim1 kinases Gin4 and Hsl1 negatively regulate pseudohypha formation and Gin4 also controls septin organization. Journal of Cell Biology, 164, 581–91.

6 Septins in the metazoan model systems Drosophila melanogaster and Caenorhabditis elegans Christine M. Field Department of Systems Biology, Harvard Medical School, Boston, MA 02115, USA

Amy Shaub Maddox Institute for Research in Immunology and Cancer, Department of Pathology and Cell Biology, University of Montr´eal, Montr´eal H3C 3J7, Canada

John R. Pringle Department of Genetics, Stanford University School of Medicine, Stanford, CA 94305, USA

Karen Oegema Ludwig Institute for Cancer Research, Department of Cellular and Molecular Medicine, University of California, San Diego, CA 92093, USA

INTRODUCTION The fruit fly Drosophila melanogaster and the nematode worm Caenorhabditis elegans have made enormous contributions to genetics, cell biology, and developmental biology. The key role of these organisms has resulted from their numerous experimental advantages, including particularly their tractability for genetic analysis. It also reflects the intensive attention that has been focused on these organisms over the past 100 (Drosophila) or 30 (C. elegans) years; this attention has led to the development both of broad and deep knowledge of the biology of the organisms and of powerful methodologies for further study. In the hope (and The Septins Edited by Peter A. Hall, S.E. Hilary Russell and John R. Pringle © 2008 John Wiley & Sons, Ltd. ISBN: 978-0-470-51969-1

148

CH06 METAZOAN MODEL SYSTEMS

expectation) that these metazoan model systems will prove equally important in the further elucidation of the properties and biological roles of the septins, we have attempted to summarize here what is currently known of the septins in these organisms.

THE GENE AND PROTEIN FAMILIES Drosophila has five known septins, Pnut, Sep1, Sep2, Sep4, and Sep5, as summarized in Table 6.1 (The ‘Sep3’ position was left free for Pnut, but that name now seems unlikely to change.) pnut was found serendipitously in the course of a genetic analysis of photoreceptor development (Neufeld and Rubin, 1994). (Because the gene was originally identified by a recessive mutation, Drosophila convention dictates that its symbol begins with a lower-case letter.) Sep1 and Sep2 were found in deliberate searches by PCR using degenerate primers based on the sequences of previously known septins (Fares, Peifer and Pringle, 1995; Al-Awar, 1996; Longtine et al., 1996; Field et al., 1996; see also Chapter 1 for more historical context). For Sep2 , a PCR product amplified from genomic DNA was used to recover cDNA clones, which were sequenced (GenBank entry U28966, June 1995); the sequences matched the coding sequence predicted from the subsequently released genome sequence. Sep4 was identified independently by A. Maeland and N. Brown and by J. Kramer and S. Hawley in 1999 (personal communications); each group noticed a septin-encoding gene very near the site of a mutationally defined gene that it was trying to clone (see Kramer and Hawley, 2003; Estrada et al., 2007), as noted first in FlyBase in 1999. At least one cDNA was sequenced and matched the coding sequence predicted from the genome sequence, but contemporary annotation of the databases appears to have been absent or incomplete, so that Sep4 had to be ‘rediscovered’ by Mu˜noz-Soriano and Paricio (2007). Sep5 was found by searching the expressed sequence tag (EST) sequences released by the Berkeley Drosophila Genome Project. Full sequencing of one of these clones (LD12056) revealed what appeared to be a full-length septin gene (K. Hales, C. Brown, J. Pringle, and M. Peifer, unpublished results; GenBank entry AF167578, July 1999); this sequence matched the coding sequence predicted from the genome sequence. Because some significant regions of the Drosophila genome remain unsequenced (notably in the heterochromatic regions), it remains possible that an additional septin gene(s) will be found there. Remarkably, given the larger number of septins in every other animal and fungal species that has been examined carefully, C. elegans appears to have just two septins, UNC-59 and UNC-61 (Nguyen et al., 2000; Table 6.1). The genes encoding these proteins were identified by a deliberate search of C. elegans EST and genomic sequences for septin-like genes, but they proved to correspond to unc-59 and unc-61 , which had been defined (genetically) much earlier by the isolation and analysis of uncoordinated mutants (White, Horvitz and Sulston, 1982).

BIOCHEMISTRY OF THE DROSOPHILA AND C. ELEGANS SEPTINS

149

BIOCHEMISTRY OF THE DROSOPHILA AND C. ELEGANS SEPTINS Studies in both Drosophila and C. elegans have been central to our current understanding of septin biochemistry.

Isolation of a three-septin complex from Drosophila embryos Co-immunoprecipitation and density-gradient centrifugation experiments had suggested that Sep1 and Pnut formed a complex and that this complex probably contained additional polypeptides (Fares, Peifer and Pringle, 1995). These conclusions were confirmed and greatly extended when Field et al. (1996) achieved the first purification of a septin complex. This complex was isolated in a single step from Drosophila embryo extracts by binding to a polyclonal antibody, directed against the C-terminal 14 amino acids of Pnut, that had been immobilized on a resin, and then eluting with a buffer containing the peptide against which the antibody had been made. The isolated complex contained three-septin polypeptides, Pnut, Sep1 and Sep2, in a stoichiometry of ∼1 : 1 : 1. Analysis by gel filtration and density-gradient centrifugation indicated that the complex had a Stokes radius of 9.9 nm and sedimented at ∼8 S, suggesting a native molecular mass of ∼340 kDa. The large Stokes radius relative to the molecular weight indicated that the complex was highly asymmetric. Given the predicted molecular weights of the three septins and the stoichiometry of the complex, it was proposed to contain two copies of each polypeptide (Field et al., 1996). It is interesting that neither Sep4 nor Sep5 appeared to be present in the isolated septin complexes, particularly given that Sep4, at least, appears to co-localize with other septins in the cleavage furrows of at least some cells (Table 6.1, footnote i). Similarly, although Shih et al. (2002) observed multiple interactions of Pnut, Sep1, and Sep2 with each other during two-hybrid screening, as well as interactions of these three proteins with various other proteins, neither Sep4 nor Sep5 was recovered. It is possible that the failure to detect Sep4 and Sep5 reflects low levels of expression, or expression only in certain cell types (as has been shown for mammalian septins: Hall et al., 2005; see also Chapters 7, 8 and 14), and thus poor representation in the extracts and cDNA libraries used. Further exploration of this issue should be very interesting.

Binding, slow exchange, and hydrolysis of GTP by the isolated septin complex Nearly all known septins contain a predicted nucleotide-binding sequence, or P-loop (Saraste, Sibbald and Wittinghofer, 1990), and additional consensus sequences that define the GTPase superfamily (Bourne, Sanders and

FBgn0026351 CG2916

Sep5

1B

2B

CC

CC

Cellularization, cytokinesis (L).g

At least some Cytokinesis (L).i cleavage furrows, various cell-cortex regions.i No information. No information.

FBgn0026362 CG9699

Sep4

1B

PB, CC

NT,h PB, CC

FBgn0014029 CG4173

Sep2

2B

FBgn0011710 CG1403

Sep1

Mutants

None reported.

None reported.

sep2

pnut Cellularization, cytokinesis (L,M)e . Nervous system (L,M). Cellularization, None reported. cytokinesis (L).

Functionc

Cleavage furrows, some intercellular bridges, neurons, various cell-cortex regions. Cleavage furrows, some intercellular bridges, neurons, various cell-cortex regions. Cleavage furrows, some intercellular bridges, neurons, various cell-cortex regions.g

2B

FBgn0013726 CG8705

Pnut

Localization

NT, PB, CC

Classa Featuresb

Properties of fly and worm septin genes and proteins

Protein FlyBase or WormBase ID

Table 6.1 Referencesd

73 % identical to Sep2.

73 % identical to Sep5. Sep2-GFP-expressing flies available. Can localize in some (but not all) contexts even when Pnut is absent.

Fails to localize when Pnut is absent

9

14, 15

3–10, 22

2–8, 10, 13f

Zygotic lethal when 1–8, 10–12, 22 homozygous; germ-line clones have been analyzed.

Comments

150 CH06 METAZOAN MODEL SYSTEMS

WBGene00006795

UNC-61

1B

2B

Cortical contractile patches during polarity establishment. Contractile rings during polar body extrusion and cytokinesis. Cortex of cellularizing gonad. Same as UNC-59

PB, CC

NT,h CC

Asymmetric furrow ingression during early embryonic cytokinesis (R). Increases robustness of embryonic cytokinesis (R). Post-embryonic cytokinesis (M). Neuronal migration, distal-tip-cell migration (M). Same as UNC-59 e228 and n3169 (point mutations)

e261 and e1005 (premature stops)

17–21

16–21

b

As defined by Pan, Malmberg and Momany (2007); see also Chapter 2. All seven proteins have the signature motifs of a GTP-binding site, and Pnut, Sep1, and Sep2 have been shown biochemically to bind GTP. NT, an extended N-terminus relative to other septins; PB, a cluster of three (UNC-59), four (Pnut), or five (Sep1 and Sep4) basic residues lying just N-terminal to the GTP-binding domain that may be involved in phosphoinositide binding (see Chapter 3); CC, predicted coiled-coil domain. c Evidence for function can be derived from localization (L), mutant phenotypes (M), and/or RNAi (R). d (1) Neufeld and Rubin, 1994; (2) Fares, Peifer and Pringle, 1995; (3) Longtine et al., 1996; (4) Field et al., 1996; (5) Hime, Brill and Fuller, 1996; (6) Al-Awar, 1996; (7) Adam, Pringle and Peifer, 2000; (8) Sisson et al., 2000; (9) Hales, Peifer and Pringle, 2000; (10) Shih et al., 2002; (11) Somma et al., 2002; (12) Field et al., 2005; (13) Maleszka, de Couet and Miklos, 1998; (14) Kramer and Hawley, 2003; (15) Mu˜noz-Soriano and Paricio, 2007; (16) White, Horvitz and Sulston, 1982; (17) Nguyen et al., 2000; (18) Finger, Kopish and White, 2003; (19) Maddox et al., 2005; (20) Maddox et al., 2007; (21) John et al., 2007; (22) Echard et al., 2004.

a

WBGene00006793

UNC-59

BIOCHEMISTRY OF THE DROSOPHILA AND C. ELEGANS SEPTINS 151

(continued )

f Considerable

text for discussion of the apparent nonuniformity of the Pnut requirement for cytokinesis. information about the cytogenetics (the endpoints of relevant deletions, etc.) of the Sep1 region is available in reference 13 and earlier publications from the Miklos group that are cited therein. However, it does not appear possible to deduce the phenotype (if any) resulting from loss of Sep1 (referred to as innocent bystander in reference 13) from any of the published information. See Chapter 1 for a summary of the historical context. g Reflecting the similarity in sequence between Sep2 and Sep5, the anti-Sep2 antibodies used in both biochemical (Field et al., 1996; Al-Awar, 1996) and protein-localization (Al-Awar, 1996; Adam, Pringle and Peifer, 2000; Figure 1c) studies were found later to interact weakly with Sep5. However, multiple lines of evidence, including peptide sequencing of immunoprecipitated protein and corroboration of many central immunofluorescence results with observations on flies expressing Sep2-GFP (Figure 1a and b), suggest that the conclusions reached with respect to Sep2 are valid (Field et al., 1996; J. Adam, K. Hales, G. Dillard, H.-P. Shih, O. Al-Awar, M. Peifer and J. Pringle, unpublished results). h The N-terminal extensions on Sep4 and UNC-61 are considerably shorter than those on some septins such as Pnut and S. cerevisiae Cdc3p. i In immunoblots, the antibody generated against Sep4 also recognized a second protein that was expressed throughout embryogenesis and in some adult tissues (C. Field, J. Adam and K. Oegema, unpublished results), precluding the successful use of this antibody in immunofluorescence experiments on embryos. However, this cross-reacting protein did not appear to be expressed in adult ovaries, where staining of both cleavage furrows (in dividing germline stem cells) and the cortical regions of germline cells was observed (J. Adam, C. Field, J. Pringle and M. Peifer, unpublished results). Interestingly, the cortical regions of follicle cells, which are rich in Pnut, Sep1, and Sep2 (Fares, Peifer and Pringle, 1995; J. Adam, C. Field, K. Oegema, J. Pringle and M. Peifer, unpublished results), did not stain with the Sep4 antibodies.

e See

Table 6.1

152 CH06 METAZOAN MODEL SYSTEMS

BIOCHEMISTRY OF THE DROSOPHILA AND C. ELEGANS SEPTINS

153

McCormick, 1991). Thus, septins were predicted to bind and perhaps hydrolyse GTP. GTP binding was confirmed for the isolated Drosophila complex by using HPLC ion-exchange chromatography to analyse the nucleotide released following denaturation of the complex with urea (Field et al., 1996). On average, 1.1 mol of guanine nucleotide was released per mole of septin polypeptide (corresponding to ∼6 per native complex), with an average GDP/GTP ratio of 2.6. These data suggested that each septin polypeptide in the complex binds one molecule of guanine nucleotide and also indicated tight binding, given that the isolation procedure involved extensive washes in nucleotide-free buffer. Subsequently, 1 : 1 binding of guanine nucleotide to septin polypeptide has been observed for complexes isolated from all other species examined, with the possible exception of the bacterially expressed C. elegans complex (see below). The ability of the purified Drosophila septins to exchange and hydrolyse GTP was also measured. In experiments in which [α-32 P]GTP was added to septin complexes attached to antibody-coated beads, approximately 10 % of the total nucleotide exchanged after 18 h at room temperature and ∼75 % of the exchanged nucleotide was hydrolysed within 2 h (Field et al., 1996). These data indicate that the Drosophila septin complex is relatively inert in exchange assays, either because it does not exchange GTP in cells or because the purified septins lack appropriate exchange factors. It remains unclear whether nucleotide exchange or hydrolysis regulates septin function in vivo. Many investigators favour a role similar to that in small GTPases, although there is no direct support for such a model. Mitchison and Field (2002) have argued that GTP binding may play a structural role in protein folding and perhaps complex formation, akin to the role of GTP bound to the non-exchanging site on α-tubulin, but that GTP exchange/hydrolysis does not regulate septin function once the complex is assembled (see also Vrabioiu et al., 2004).

Filament formation by septin complexes in vitro Early studies of wild-type and mutant yeast by electron microscopy (EM) and immunofluorescence led to the hypothesis that septin polypeptides are the major structural components of the 10-nm ‘neck filaments’ observed in the neck region of budding cells (Byers and Goetsch, 1976; Longtine et al., 1996; and see Chapter 3). It was thus satisfying to find that when dialyzed into a buffer of physiological ionic strength, the immunopurified Drosophila septin complex assembled into filaments in vitro. The assembled filaments were approximately 7 nm in diameter and of variable length. Length histograms revealed that filaments were multiples of an ∼26-nm subunit, presumably corresponding to the 2 : 2 : 2 septin complex (Field et al., 1996). The distribution of lengths, with short filaments co-existing with longer ones, suggested that pure septin complexes polymerize in vitro by an isodesmic mechanism, without a nucleation step (see, for example, Romberg, Simon and Erickson, 2001). In this mechanism, assembly and disassembly are not

154

CH06 METAZOAN MODEL SYSTEMS

limited to polymer ends. Adjacent polymer subunits in the middle of a filament can dissociate from each other, breaking the filament in two, and, in a converse reaction, two filaments can also anneal to form a longer one. Analysis of septin complexes isolated from other organisms demonstrated that the polymerization behaviour exhibited by the Drosophila complex is typical (Frazier et al., 1998; Hsu et al., 1998; Kinoshita et al., 2002). Characterization of yeast and human septin complexes also revealed additional features suggesting that in addition to linear polymers, septin filaments can associate laterally to form higher-order structures in cells. For example, yeast septin filaments tend to align side by side (Frazier et al., 1998), whereas human septin filaments tend to roll up into torus-shaped assemblies (Kinoshita et al., 2002). The 26-nm periodicity of the Drosophila septin filaments led to the speculation that the yeast neck filaments might not run circumferentially around the mother-bud axis, as suggested by the early EM studies, but instead be oriented axially, parallel to the mother-bud axis (Field et al., 1996), with the alignment between monomers in adjacent filaments generating the circumferential striations seen in the EM images. Although still controversial (see Chapters 3 and 4), this model has recently been supported by a new technique utilizing polarized fluorescence microscopy of GFP-septin fusions (Vrabioiu and Mitchison, 2006, 2007).

Assembly of a two-septin C. elegans complex Septins are found in fungi and animals, but not in plants. Based on phylogenetic comparisons, all septins can be partitioned into five groups, with the animal septins falling exclusively into groups 1 and 2 (Pan, Malmberg and Momany, 2007; Chapter 2). The genomes of most sequenced animals encode between four and thirteen septins, with all animals containing at least one, and typically multiple, septins from each group. In contrast, the C. elegans genome encodes just two septins, one in group 1 (UNC-61) and one in group 2 (UNC-59) (Nguyen et al., 2000; Pan, Malmberg and Momany, 2007). Depletion of either one or both septins gives essentially identical phenotypes (Nguyen et al., 2000; Finger, Kopish and White, 2003; Maddox et al., 2007), and each protein is required for the other’s localization (Nguyen et al., 2000), suggesting that both proteins are required to form a functional complex in vivo. This relative simplicity has made the C. elegans septins an attractive system for functional characterization and biochemical reconstitution. The C. elegans septin complex has been assembled after expression either in E. coli or in insect cells (John et al., 2007). In both cases, the two proteins were insoluble when expressed individually but formed a soluble complex when co-expressed. The predominant soluble complex was a tetramer with a 2 : 2 UNC-59:UNC-61 stoichiometry, but dimers with a 1 : 1 stoichiometry of the two subunits were also observed. EM of negatively stained tetramers revealed an

BIOCHEMISTRY OF THE DROSOPHILA AND C. ELEGANS SEPTINS

155

elongated complex ∼7-nm wide and 17–20-nm long. After classification and averaging, these elongated particles resolved into four linearly arranged globular densities. By performing EM on preparations in which either UNC-59 or UNC-61 was tagged with GFP, the heterotetramer was shown to exist in an UNC-59/UNC-61/UNC-61/UNC-59 configuration, suggesting that the UNC-59/ UNC-61 dimers form a non-polar tetramer via the association of their UNC-61 subunits. It should be noted that in contrast to previously purified septin complexes, these heterologously expressed C. elegans complexes exhibited low nucleotide content (<0.1 GDP and no GTP per septin monomer). In addition, whereas the C. elegans septin complexes purified from insect cells were able to polymerize and assemble into filamentous sheets and bundles when dialyzed into low salt, only a few filaments were formed when the complex purified from E. coli was treated similarly, indicating that the polymerization efficiency of the E. coli -expressed complex is very low. More work is required to determine if the C. elegans septins are really different from other septins in terms of GTP binding and whether GTP content is related to polymerization efficiency. Interestingly, when the C. elegans septins were expressed in yeast, they localized to the bud neck in the majority of cells examined. Localization required both C. elegans septins but not the endogenous yeast septins (John et al., 2007). These remarkable observations suggest that there is some highly conserved property(ies) of septins that cause bud-neck targeting, for example, binding to a specific lipid that is enriched in this region. The nature of this targeting mechanism clearly warrants further investigation.

Interactions between the septins and the actin and microtubule cytoskeletons When Drosophila embryo extracts were passed consecutively over F-actin and microtubule affinity columns, the three-protein septin complex (discussed above) was among the proteins found to bind to both, suggesting either direct or indirect associations with both filament systems (Sisson et al., 2000). Consistent with this idea, mammalian septins have been shown to co-localize with both the actin and microtubule cytoskeletons in vivo (Kinoshita et al., 1997, 2002; Surka, Tsang and Trimble, 2002; Spiliotis, Kinoshita and Nelson, 2005; Spiliotis et al., 2008); these associations appear to affect the stability both of actin stress fibres (Kinoshita et al., 1997, 2002) and of microtubules (through an interaction with MAP4: Kremer, Haystead and Macara, 2005). The septins have not been shown to bind directly to either F-actin or microtubules. However, there is excellent evidence, both in vitro and in vivo, for a functionally important interaction between the septins and actin filaments, mediated by the actin-binding and -bundling protein anillin, that is important during both cell division and development (see below).

156

CH06 METAZOAN MODEL SYSTEMS

Other septin-interacting proteins The molecular mechanisms of septin function in animal cells are largely unknown, and identification of additional septin-interacting proteins should be highly informative. Such identification has been attempted by two-hybrid screening. Shih et al. (2002) screened using Pnut, Sep1, and Sep2 as baits. In addition to the other septins, they recovered both activating (Uba2) and conjugating (Ubc9) enzymes of the Drosophila sumoylation system. However, despite considerable effort, Shih et al. were unable to demonstrate either that the Drosophila septins are subject to sumoylation (as they are in yeast) or that the septins serve as a scaffold for the sumoylation system. Shih et al. (2002) also identified four other proteins (Sip1-Sip4) that appeared to interact strongly with one or more of the septin baits. However, to our knowledge, there has been no further investigation of the possible functional significance of the septin–Sip interactions. In addition, a large-scale two-hybrid screen identified a variety of proteins that appeared to interact with the Drosophila septins (Giot et al., 2003; see FlyBase for details). However, the biological significance of these interactions has not yet been demonstrated. We are not aware of any other attempts to identify septin-interacting proteins in either Drosophila or C. elegans.

ROLES OF THE SEPTINS (AND ANILLIN) IN CYTOKINESIS AND CELLULARIZATION Ordinary cytokinesis In Drosophila, Pnut, Sep1, and Sep2 are all concentrated in most if not all cleavage furrows during cytokinesis and in many (although not all) of the intercellular bridges that sometimes persist (depending on the cell type) after cytokinesis (Neufeld and Rubin, 1994; Fares, Peifer and Pringle, 1995; Al-Awar, 1996; Hime, Brill and Fuller, 1996; Figure 6.1b). Sep4 also appears to be concentrated in at least some cleavage furrows (Table 6.1, footnote i). Similarly, in C. elegans, both UNC-59 and UNC-61 are concentrated in the cleavage furrows that have been examined (Nguyen et al., 2000). These localization data, the genetic and RNAi data discussed below, and protein-localization, antibody-microinjection, and RNAi data in mammalian cells (Kinoshita et al., 1997; Surka, Tsang and Trimble, 2002) all suggest that the septins play a role in animal-cell cytokinesis, as they do in budding yeast. However, the analyses done to date suggest that the function of the septins during cytokinesis in animal cells is quite different from that in yeast. In budding yeast, the septins form an ordered array at the bud neck long before cytokinesis begins (see Chapters 3 and 4) and are required to recruit and localize the majority of proteins, including myosin II and actin, that

ROLES OF THE SEPTINS (AND ANILLIN) IN CYTOKINESIS AND CELLULARIZATION

157

(a)

(b)

(c)

Figure 6.1 Localization of Sep2 during embryogenesis (Al-Awar, 1996; K. Hales, O. Al-Awar, H.-P. Shih, G. Dillard, J. Adam, M. Peifer, and J. Pringle, unpublished results). (a and b) Time-lapse microscopy of embryos expressing Sep2-GFP. The panels show selected images from the time-lapse sequences; the numbers indicate the order of the images but not their actual timing. (a) Cellularization of a syncytial-blastoderm embryo. Sep2 co-localizes with Pnut and Sep1 (not shown) at the leading edge of the invaginating membrane. (b) Cell division in a mitotic domain of a gastrulation-stage embryo. Sep2 begins to concentrate at the incipient cleavage furrow (arrowheads in image 4) and becomes more concentrated as the furrow invaginates. Because other cells observed in the same way displayed a clear Sep2-GFP signal persisting at the intercellular bridge that connects the recently divided cells, this bridge is probably out of the focal plane in image 10. (c) Localization to the embryonic central nervous system (CNS) and its dependence on Pnut. Flies heterozygous for the pnut XP mutation (Neufeld and Rubin, 1994) and wild-type pnut (i.e. pnut XP /+) were allowed to mate, and the resulting embryos were examined by immunofluorescence after the CNS had formed using double staining either with a monoclonal anti-Pnut (Neufeld and Rubin, 1994) and an anti-Sep2 antibody (Field et al., 1996; Al-Awar, 1996) or with the anti-Sep2 and the monoclonal antibody BP102, which is a marker for the CNS (Seeger et al., 1993). In about 75 % of the embryos, the embryonic CNS stained well with both anti-Pnut and anti-Sep2 (data not shown) or with both anti-Sep2 and BP102 (images 1 and 2, taken at the same focal plane). These embryos are presumably those homozygous or heterozygous for wild-type pnut , as previous studies have indicated that maternally supplied Pnut is exhausted by the time of CNS formation (Fares, Peifer and Pringle, 1995). The remaining 25 % of the embryos (presumably those homozygous for the pnut mutation) failed to stain either with anti-Pnut or with anti-Sep2 (data not shown) or stained with BP102 but not with anti-Sep2 (images 3 and 4, taken at the same focal plane). Similar results were obtained for Sep1 (Fares, Peifer and Pringle, 1995).

158

CH06 METAZOAN MODEL SYSTEMS

are implicated in the execution of cytokinesis (Gladfelter, Pringle and Lew, 2001; Dobbelaere and Barral, 2004; see also Chapter 4). In contrast, the septins appear to be downstream in the assembly pathway of cytokinesis components in animal cells (Field et al., 2005; Maddox et al., 2005), and they have not yet been shown to be required for the targeting of any other cytokinesis protein. Moreover, in contrast to budding yeast, the septins appear to be nonessential for cytokinesis in many types of animal cells. As noted above, the available genetic and RNAi data support a role for the septins in cytokinesis in flies and worms, but they do not present a simple picture. In their initial characterization of the Drosophila pnut mutant, Neufeld and Rubin (1994) observed that flies homozygous for the mutation die as pupae with small imaginal discs, apparently reflecting reduced cell proliferation in the discs. [The survival of pnut mutant flies to the pupal stage apparently results from the maternal contribution to the egg of Pnut protein (from the heterozygous mothers), as shown by later studies in which germline-clone flies that lacked this maternal contribution died much earlier, during gastrulation (Adam, Pringle and Peifer, 2000).] In addition, large, multinucleate and/or polyploid cells were observed in all larval tissues examined in the mutants, again indicating defects in cytokinesis. However, although a later reexamination confirmed the presence of polyploid cells in larval brains, it found them to be in the minority, suggesting that most cytokinesis events were completed successfully during the development of this organ (Somma et al., 2002). In addition, analysis of pnut germline clones indicated that stem-cell and cystoblast divisions in the female germ line can apparently proceed in the complete absence of Pnut (Adam, Pringle and Peifer, 2000). Moreover, the embryos derived from such germline clones also showed a lack of Sep1 (although, interestingly, not of Sep2) at the cellularization front, suggesting that the localization of Sep1 might generally depend on Pnut (consistent with findings by Fares, Peifer and Pringle, 1995), in which case Sep1 would also be missing from the cleavage furrows during the stem-cell and cystoblast divisions. Thus, division of these cell types is probably not dependent on the septins. However, the possibility that Sep2, Sep4, and/or Sep5 could still be localized to the stem-cell and cystoblast cleavage furrows, and providing an essential septin function there, makes it impossible, at present, to be certain of this conclusion. Further complicating the picture is that flies homozygous for what appeared to be a null mutation of sep2 survived and developed until just before eclosure of adult flies, at which point most died but some survived (Hales, Peifer and Pringle, 2000). However, the very similar sequences of Sep2 and Sep5 (Table 6.1) raise the possibility that functional redundancy between Sep2 and Sep5 might allow successful cell division in the absence of Sep2. RNAi studies in cultured Drosophila S2 cells also fail to give an entirely clear picture. Targeted knockdown of Pnut did not result in a strong block in cytokinesis (Somma et al., 2002), and although one genome-wide screen identified both Pnut and Sep2 knockdowns as producing cytokinesis defects, these were not among the strongest phenotypes seen (Echard et al., 2004; Eggert et al., 2006).

ROLES OF THE SEPTINS (AND ANILLIN) IN CYTOKINESIS AND CELLULARIZATION

159

However, interpretation of these results is complicated both by the possibility of functional redundancy among the septins and by the possibility that knockdown of the proteins in question was not sufficiently complete to show its role clearly. Some of these complications do not pertain in C. elegans, where animals null for the function of both septin genes have been analysed. Such animals exhibit a spectrum of developmental defects but can survive to adulthood and reproduce, indicating that the septins are not generally essential for cell division. Nonetheless, the septins appear to make cytokinesis more robust. Variable failure of the post-embryonic divisions of the ventral-nerve-cord-precursor cells in septin mutant worms was documented early (White, Horvitz and Sulston, 1982), and a subsequent detailed examination of the Q/V5 neuroblast precursor revealed that its division fails ∼30 % of the time in the mutants (Nguyen et al., 2000). Although Nguyen et al. (2000) did not detect a defect in embryonic cytokinesis, a subsequent examination of a large number of embryos at the eight-cell stage indicated that cytokinesis fails in ∼0.6 % of cells in embryos depleted of the septins by RNAi (Maddox et al., 2007). A significant role in embryonic cytokinesis was further suggested by the observation that depletion of Rho kinase, which normally does not prevent successful completion of the first embryonic cytokinesis, led to 63 % cytokinesis failure when the septins were also absent (Maddox et al., 2007). Cumulatively, these results suggest that although the septins are not strictly essential for cytokinesis, they do provide a function(s) that makes cytokinesis more robust to stochastic errors, transient inhibition of contractility, and the like. What could this function(s) be? Work in C. elegans has also shed some light on this question. In both Drosophila (Field et al., 2005) and C. elegans (Maddox et al., 2005), concentration of septins in the cleavage furrow depends on the actin-binding protein anillin (ANI-1 in C. elegans). Consistent with this, depletion of either anillin or the septins results in a very similar effect on cytokinesis: both depletions increase the frequency of stochastic failure of the embryonic divisions and dramatically increase cytokinesis failure in a Rho-kinase-depleted background (Maddox et al., 2007; Maddox and Oegema, unpublished results). Interestingly, although their depletion does not affect the rate of contractile-ring closure, depletion of either anillin or the septins dramatically alters the geometry of closure within the division plane. During the first division of wild-type embryos, the contractile ring normally closes asymmetrically, resulting in a furrow that cuts unidirectionally across the division plane from one side of the embryo towards the other. This asymmetric furrowing results from an intrinsic symmetry-breaking mechanism within the contractile ring that promotes the coalescence of ring components, including actin and myosin II, on one side of the ring. In embryos depleted of either anillin or the septins, this symmetry-breaking mechanism is lost, and both the structure of the contractile ring and its closure within the division plane become symmetric (Maddox et al., 2007). Taken together, these data indicate that the septins function together with anillin to coalesce the actomyosin structures. A model for how this might occur is suggested by the filament-cross-linking properties of anillin and the observation that

160

CH06 METAZOAN MODEL SYSTEMS

the septins can form two-dimensional filament assemblies that can associate with membranes (Rodal et al., 2005; Vrabioiu and Mitchison, 2007). In addition to a C-terminal region that has a PH domain and confers the ability to associate with the septins (Oegema et al., 2000; Kinoshita et al., 2002), the N-terminal region of anillin has previously been shown to bind and bundle actin filaments and also to associate directly with activated myosin II (Field and Alberts, 1995; Oegema et al., 2000; Straight, Field and Mitchison, 2005). A reasonable model is therefore that anillin promotes the coalescence of the actomyosin structures by cross-linking membrane-associated arrays of septin filaments to the underlying actomyosin assemblies. This activity of the septins and anillin probably underlies their role both in promoting asymmetric furrow ingression and in making furrow ingression robust to mechanical challenges.

Cellularization of the syncytial Drosophila embryo The Drosophila embryo undergoes a specialized form of cytokinesis after the 13th cell cycle, when the nuclei of the syncytial blastoderm are simultaneously partitioned into individual cells by synchronous ingression of plasma membrane furrows. This dramatic remodelling of the embryo membrane requires many proteins also required in ordinary cytokinesis, such as actin, myosin II, RhoA, and anillin, and it has served as a useful model system for furrow ingression (Crawford et al., 1998; Royou et al., 2004; Strickland and Burgess, 2004; Field et al., 2005). The three Drosophila septins that constitute the complex present in early embryos (Pnut, Sep1, and Sep2) all localize to the ingressing cellularization furrows (Neufeld and Rubin, 1994; Fares, Peifer and Pringle, 1995; Figure 6.1a). In the absence of Pnut (in germline clones), cellularization appears to begin normally but becomes progressively more abnormal as the process proceeds (Adam, Pringle and Peifer, 2000). Sep1 fails to localize to the cellularization front although Sep2 is still present, and actin organization becomes abnormal as cellularization proceeds. Similarly, anillin is recruited to the early furrows, but dissociates prematurely. Overall, the pnut-mutant cellularization phenotype is similar to that observed with anillin loss-of-function maternal-effect alleles (Field et al., 2005). This phenotypic similarity is consistent with the biochemical data (see above) indicating that a septin complex and anillin physically associate and function together both during furrow ingression and in stabilization of ingressed furrows. It is not yet known which parts of the septin molecule are required for the interaction with anillin. However, the studies of Drosophila cellularization have helped to map the region of anillin required for interaction with the septins. Previous work had shown that the C-terminal ∼378 amino acids of anillin, which contains its PH domain, were required for interaction with the septins as judged both by co-localization to ectopic cortical foci in vivo (Oegema et al., 2000) and by co-sedimentation in vitro (Kinoshita et al., 2002). Characterization of EMS-induced anillin point mutations causing changes in conserved amino acids

SEPTIN FUNCTIONS DURING DEVELOPMENT AND IN ADULT ANIMALS

161

in the PH domain showed that they can inhibit the interaction between anillin and the septins (Field et al., 2005). Interestingly, the interaction between the septins and anillin-like proteins is very broadly conserved, being found not only in flies, worms, and mammals (see discussion and references above), but even in the fission yeast Schizosaccharomyces pombe, where the distant anillin relative Mid2p plays an important role in generating the normal organization of septins at the division site (Berlin, Paoletti and Chang, 2003; Tasto, Morrell and Gould, 2003; see also Chapter 5). Thus, this interaction between the proteins must be ancient in evolutionary origin, suggesting that it may also be a pervasive aspect of septin function even in processes other than cytokinesis, as discussed further below.

SEPTIN FUNCTIONS DURING DEVELOPMENT AND IN ADULT ANIMALS A variety of data suggest that the Drosophila and C. elegans septins have various roles unrelated to cytokinesis. However, as none of these roles is very well understood, we review them only briefly here.

A role for the septins (and anillin) in neuronal migration and other aspects of worm development The fact that a loss of septin function does not result in extensive defects in cytokinesis in C. elegans has allowed characterization of other developmental roles of the septins in mutant animals. Worms mutant for each of the two C. elegans septin genes unc-61 and unc-59 were originally identified based on their uncoordinated movements. Although this phenotype was originally proposed to be due to a failure of post-embryonic neuroblast divisions (White, Horvitz and Sulston, 1982), a more recent study indicated that septin mutant larvae were already uncoordinated before these divisions occur (Finger, Kopish and White, 2003). Both C. elegans septins are expressed in neurons in embryos and larvae but are absent from most neurons in adults, suggesting a developmental role (Finger, Kopish and White, 2003). Examination of a GFP marker for the GABAergic DD neurons in the ventral nerve cord in newly hatched larvae, before any post-embryonic cell divisions, revealed mis-positioned axons, missing commissures, and discontinuities in the ventral cord that suggest defects in axon guidance and migration (Finger, Kopish and White, 2003). Both C. elegans septins are also expressed in the distal-tip cell, and defects in distal-tip-cell migration have been observed in mutant animals, suggesting that the septins have a more general role in cell migration (Finger, Kopish and White, 2003). This hypothesis is consistent with localization data suggesting a role for the septins in dorsal closure of the Drosophila embryo (Fares, Peifer and Pringle, 1995). The worm septins are also required for the normal morphology

162

CH06 METAZOAN MODEL SYSTEMS

of the vulva and the male tail, although there are no clues as yet to the basis for these morphological defects (Nguyen et al., 2000). Interestingly, many of the developmental functions of the worm septins, like their role in cytokinesis, appear to be performed in conjunction with anillin. In contrast to loss of septin function, depletion of anillin by RNAi results in embryonic lethality, probably because its depletion, unlike that of the septins, results in failure of polar-body extrusion, thereby increasing the ploidy of the resulting embryos. The observation that anillin retains some function in septin-deficient embryos is consistent with the observation that it can target to the contractile ring, and to other cortical contractile structures, in the absence of the septins (Maddox et al., 2005). If the embryonic lethality associated with anillin depletion is bypassed by performing the RNAi after hatching, at the L1 larval stage, the resulting adult worms exhibit a spectrum of developmental defects, including a lack of coordination and aberrant vulval development, that are essentially identical to those seen in septin-deficient worms (Figure 6.2). These results suggest that anillin and the septins may function together in developmental processes outside the context of cell division. Further studies will be needed to determine if this is the case and to determine which functions of the septins in other animal systems are performed in conjunction with anillin or anillin-related proteins as opposed to other possible partners.

A possible role for the septins in neural function Early studies showed that both Pnut and Sep1 are very highly concentrated in neurons of the embryonic central and peripheral nervous systems (Neufeld and Rubin, 1994; Fares, Peifer and Pringle, 1995), and the same is true of Sep2 (Al-Awar, 1996; Figure 6.1c, panels 1 and 2). Both Sep1 (Fares, Peifer and Pringle, 1995) and Sep2 (Al-Awar, 1996; Figure 6.1c, panels 3 and 4) are missing from the nervous system in pnut mutants, suggesting that whatever role the septins play there probably involves the same three-septin complex that has been found in embryos and appears to function during both ordinary cytokinesis and cellularization. Interestingly, the absence of all three septins does not grossly disrupt the development of the embryonic central nervous system (CNS) (Figure 6.1c, panels 1 and 3), suggesting that the septin role in flies may relate to neural function and not only to development (in possible contrast to worms: see above). Nonetheless, a role for the septins in at least some aspects of nervous-system development in flies is strongly indicated by the genetic interaction (with a gene involved in photoreceptor development) that led to the original identification of the pnut mutant (Neufeld and Rubin, 1994). Strikingly, certain septins are also expressed at high levels in the mammalian CNS (as reviewed in Chapters 11 and 14), suggesting that the septin role(s) in neurons is ancient and conserved. Although there are several indications that this role may involve one or more aspects of vesicle trafficking, the precise nature of the septin role in neurons remains unclear and a tantalizing

SEPTIN FUNCTIONS DURING DEVELOPMENT AND IN ADULT ANIMALS (a)

L1 larvae soaked in dsRNA for 24h

Transferred to food for 72h

163

Protruded vulva Uncoordinated movement and protruded vulva Uncoordinated movement Exploded Immobile Non-adult, defective Non-adult, normal Normal adult

(b)

ani-1 (RNAi ) n = 520

Vulva

Adult Hermaphrodite

unc-61 (e228) n = 534

Control

(c)

Control n = 546

unc-61 (e228) Anillin (ANI-1) (septin mutant) depleted

Immature, exploded

Gravid, protruded vulva, unc

Gravid, protruded vulva, unc

Figure 6.2 C. elegans hermaphrodites depleted of anillin (ANI-1) after embryogenesis display a similar array of developmental defects to worms null for septin function. (a) Schematic of the soaking method for RNAi. (b) Wild-type worms were depleted of ANI-1 by soaking in dsRNA directed against ani-1 , or treated with RNA directed against a yeast gene (NDC10 ) as a control. CB228 is a strain homozygous for the septin null mutation unc-61(e228). For CB228, starved L1 larvae were plated onto food for 72 h. Pie charts of the percentages of worms exhibiting the indicated phenotypes are shown. Terminal phenotypes were scored by eye using a dissection microscope. (c) DIC images of control, ani-1(RNAi), and septin mutant [unc-61(e228)] worms illustrating the protruded vulva and general growth defect/exploded phenotype resulting from loss of anillin or the septins.

164

CH06 METAZOAN MODEL SYSTEMS

object for further investigation. Perhaps the recent report that overexpression of the Parkin substrate Sep4 is toxic to dopaminergic neurons in Drosophila, suggesting that the fly might provide a model for the understanding of human Parkinsonism (Mu˜noz-Soriano and Paricio, 2007), will help to spur further interest in this problem.

Other possible roles for the Drosophila septins Careful examination of Pnut, Sep1, and Sep2 localization in various larval and adult tissues has revealed distinctive patterns of cortical and noncortical localization that suggest, among other things, roles in the maintenance of cell polarity in various epithelial tissues (Neufeld and Rubin, 1994; Fares, Peifer and Pringle, 1995; Al-Awar, 1996; J. Adam, J. Pringle, and M. Peifer, unpublished results). These possibilities await more incisive investigation by genetic and/or biochemical methods.

CONCLUDING REMARKS As apparent from the discussion above and from the other chapters in this volume, we have a long way to go to elucidate the full range of septin functions and mechanisms in animal cells, and a variety of systems and methods will no doubt contribute to this goal. However, given the apparent variety of septin functions, the complexity of the septin gene and protein families in many animals (such as mammals), and the complexity and subtlety of septin interactions with other proteins, it is difficult to imagine that major progress will come without effective exploitation of genetic approaches. Mutants of mice have already been of considerable value (see elsewhere in this volume, particularly Chapter 15) and will no doubt continue to be so, and there will be more clues to come from analysis of human diseases (see Chapter 14), as there have been in other areas of cell biology. However, the extraordinary genetic tractability of Drosophila and C. elegans means that they should have a major role to play in the elucidation of septin function both in cytokinesis and in other contexts such as the central nervous system. The experimental advantages of flies and worms are amplified in the case of the septins by the relative simplicity of the gene and protein families in these organisms and because they also offer the opportunity for facile RNAi studies to complement traditional genetic studies. Drosophila should be a particularly attractive model organism for septin studies because an absence of septins produces strong phenotypes (in particular, the lethality of pnut mutants), so that possible issues of functional redundancy with other types of proteins should not be intractable. In this context, we find it rather surprising that (to our knowledge) there is so little effort being devoted to study of the Drosophila septins. Perhaps this chapter and the rest of this volume will help to change that situation!

REFERENCES

165

REFERENCES Adam, J.C., Pringle, J.R. and Peifer, M. (2000) Evidence for functional differentiation among Drosophila septins in cytokinesis and cellularization. Molecular Biology of the Cell , 11, 3123–35. Al-Awar, O.S. (1996) The role of septins in the morphogenesis of Schizosaccharomyces pombe and Drosophila melanogaster, Ph.D. Dissertation, The University of North Carolina, Chapel Hill. Berlin, A., Paoletti, A. and Chang, F. (2003) Mid2p stabilizes septin rings during cytokinesis in fission yeast. Journal of Cell Biology, 160, 1083–92. Bourne, H.R., Sanders, D.A. and McCormick, F. (1991) The GTPase superfamily: conserved structure and molecular mechanism. Nature, 349, 117–27. Byers, B. and Goetsch, L. (1976) A highly ordered ring of membrane-associated filaments in budding yeast. Journal of Cell Biology, 69, 717–21. Crawford, J.M., Harden, N., Leung, T. et al. (1998) Cellularization in Drosophila melanogaster is disrupted by the inhibition of Rho activity and the activation of Cdc42 function. Developmental Biology, 204, 151–64. Dobbelaere, J. and Barral, Y. (2004) Spatial coordination of cytokinetic events by compartmentalization of the cell cortex. Science, 305, 393–96. Echard, A., Hickson, G.R.X., Foley, E. and O’Farrell, P.H. (2004) Terminal cytokinesis events uncovered after an RNAi screen. Current Biology, 14, 1685–93. Eggert, U.S., Mitchison, T.J. and Field, C.M. (2006) Animal cytokinesis: from parts list to mechanisms. Annual Review of Biochemistry, 75, 543–66. Estrada, B., Maeland, A.D., Gisselbrecht, S.S. et al. (2007) The MARVEL domain protein, Singles Bar, is required for progression past the pre-fusion complex stage of myoblast fusion. Developmental Biology, 307, 328–39. Fares, H.F., Peifer, M.A. and Pringle, J.R. (1995) Localization and possible functions of Drosophila septins. Molecular Biology of the Cell , 6, 1843–59. Field, C.M. and Alberts, B.M. (1995) Anillin, a contractile ring protein that cycles from the nucleus to the cell cortex. Journal of Cell Biology, 131, 165–78. Field, C.M., Al-Awar, O., Rosenblatt, J. et al. (1996) A purified Drosophila septin complex forms filaments and exhibits GTPase activity. Journal of Cell Biology, 133, 605–16. Field, C.M., Coughlin, M., Doberstein, S. et al. (2005) Characterization of anillin mutants reveals essential roles in septin localization and plasma membrane integrity. Development, 132, 2849–60. Finger, F.P., Kopish, K.R. and White, J.G. (2003) A role for septins in cellular and axonal migration in C. elegans. Developmental Biology, 261, 220–34. Frazier, J.A., Wong, M.L., Longtine, M.S. et al. (1998) Polymerization of purified yeast septins: evidence that organized filament arrays may not be required for septin function. Journal of Cell Biology, 143, 737–49. Giot, L., Bader, J.S., Brouwer, C. et al. (2003) A protein interaction map of Drosophila melanogaster. Science, 302, 1727–36. Gladfelter, A.S., Pringle, J.R. and Lew, D.J. (2001) The septin cortex at the yeast mother-bud neck. Current Opinion in Microbiology, 4, 681–89. Hales, K.G., Peifer, M. and Pringle, J.R. (2000) Genetic analysis of the Sep2 and Sep5 septins. Abstracts of the 41st Annual Drosophila Research Conference, a158 (Abstract), Pittsburgh, Pennsylvania.

166

CH06 METAZOAN MODEL SYSTEMS

Hall, P.A., Jung, K., Hillan, K.J. and Russell, S.E.H. (2005) Expression profiling the human septin gene family. The Journal of Pathology, 206, 269–78. Hime, G.R., Brill, J.A. and Fuller, M.T. (1996) Assembly of ring canals in the male germ line from structural components of the contractile ring. Journal of Cell Science, 109, 2779–88. Hsu, S.-C., Hazuka, C.D., Roth, R. et al. (1998) Subunit composition, protein interactions, and structures of the mammalian brain sec6/8 complex and septin filaments. Neuron, 20, 1111–22. John, C.M., Hite, R.K., Weirich, C.S. et al. (2007) The Caenorhabditis elegans septin complex is nonpolar. EMBO Journal , 26, 3296–307. Kinoshita, M., Field, C.M., Coughlin, M.L. et al. (2002) Self- and actin-templated assembly of mammalian septins. Developmental Cell , 3, 791–802. Kinoshita, M., Kumar, S., Mizoguchi, A. et al. (1997) Nedd5, a mammalian septin, is a novel cytoskeletal component interacting with actin-based structures. Genes and Development, 11, 1535–47. Kramer, J. and Hawley, R.S. (2003) The spindle-associated transmembrane protein Axs identifies a membranous structure ensheathing the meiotic spindle. Nature Cell Biology, 5, 261–63. Kremer, B.E., Haystead, T. and Macara, I.G. (2005) Mammalian septins regulate microtubule stability through interaction with the microtubule-binding protein MAP4. Molecular Biology of the Cell , 16, 4648–59. Longtine, M.L., DeMarini, D.J., Valencik, M.L. et al. (1996) The septins: roles in cytokinesis and other processes. Current Opinion in Cell Biology, 8, 106–19. Maddox, A.S., Habermann, B., Desai, A. and Oegema, K. (2005) Distinct roles for two C. elegans anillins in the gonad and early embryo. Development, 132, 2837–48. Maddox, A.S., Lewellyn, L., Desai, A. and Oegema, K. (2007) Anillin and the septins promote asymmetric ingression of the cytokinetic furrow. Developmental Cell , 12, 827–35. Maleszka, R., De Couet, H.G. and Miklos, G.L.G. (1998) Data transferability from model organisms to human beings: Insights from the functional genomics of the flightless region of Drosophila. Proceedings of the National Academy of Sciences of the United States of America, 95, 3731–36. Mitchison, T.J. and Field, C.M. (2002) Cytoskeleton: what does GTP do for septins? Current Biology, 12, R788–90. Mu˜noz-Soriano, V. and Paricio, N. (2007) Overexpression of Septin 4, the Drosophila homologue of human CDCrel-1, is toxic for dopaminergic neurons. The European Journal of Neuroscience, 26, 3150–58. Neufeld, T.P. and Rubin, G.M. (1994) The Drosophila peanut gene is required for cytokinesis and encodes a protein similar to yeast putative bud neck filament proteins. Cell , 77, 371–79. Nguyen, T.Q., Sawa, H., Okano, H. and White, J.G. (2000) The C. elegans septin genes, unc-59 and unc-61 , are required for normal postembryonic cytokineses and morphogenesis but have no essential function in embryogenesis. Journal of Cell Science, 113, 3825–37.

REFERENCES

167

Oegema, K., Savoian, M.S., Mitchison, T.J. and Field, C.M. (2000) Functional analysis of a human homologue of the Drosophila actin binding protein anillin suggests a role in cytokinesis. Journal of Cell Biology, 150, 539–51. Pan, F., Malmberg, R.L. and Momany, M. (2007) Analysis of septins across kingdoms reveals orthology and new motifs. BMC Evolutionary Biology, 7, 103. Rodal, A.A., Kozubowski, L., Goode, B.L. et al. (2005) Actin and septin ultrastructures at the budding yeast cell cortex. Molecular Biology of the Cell , 16, 372–84. Romberg, L., Simon, M. and Erickson, H.P. (2001) Polymerizatin of FtsZ, a bacterial homolog of tubulin: is assembly cooperative? Journal of Biological Chemistry, 276, 11743–53. Royou, A., Field, C., Sisson, J.C. et al. (2004) Reassessing the role and dynamics of nonmuscle myosin II during furrow formation in early Drosophila embryos. Molecular Biology of the Cell , 15, 838–50. Saraste, M., Sibbald, P.R. and Wittinghofer, A. (1990) The P-loop – a common motif in ATP- and GTP-binding proteins. Trends in Biochemical Sciences, 15, 430–34. Seeger, M., Tear, G., Ferres-Marco, D. and Goodman, C.S. (1993) Mutations affecting growth cone guidance in Drosophila: genes necessary for guidance toward or away from the midline. Neuron, 10, 409–26. Shih, H.-P., Hales, K.G., Pringle, J.R. and Peifer, M. (2002) Identification of septininteracting proteins and characterization of the Smt3/SUMO-conjugation system in Drosophila. Journal of Cell Science, 115, 1259–71. Sisson, J.C., Field, C., Ventura, R. et al. (2000) Lava Lamp, a novel peripheral Golgi protein, is required for Drosophila melanogaster cellularization. Journal of Cell Biology, 151, 905–17. Somma, M.P., Fasulo, B., Cenci, G. et al. (2002) Molecular dissection of cytokinesis by RNA interference in Drosophila cultured cells. Molecular Biology of the Cell , 13, 2448–60. Spiliotis, E.T., Hunt, S.J., Hu, Q. et al. (2008) Epithelial polarity requires septin coupling of vesicle transport to polyglutamylated microtubules. Journal of Cell Biology, 180, 295–303. Spiliotis, E.T., Kinoshita, M. and Nelson, W.J. (2005) A mitotic septin scaffold required for mammalian chromosome congression and segregation. Science, 307, 1781–85. Straight, A.F., Field, C.M. and Mitchison, T.J. (2005) Anillin binds nonmuscle myosin II and regulates the contractile ring. Molecular Biology of the Cell , 16, 193–201. Strickland, L.I. and Burgess, D.R. (2004) Pathways for membrane trafficking during cytokinesis. Trends in Cell Biology, 14, 115–18. Surka, M.C., Tsang, C.W. and Trimble, W.S. (2002) The mammalian septin MSF localizes with microtubules and is required for completion of cytokinesis. Molecular Biology of the Cell , 13, 3532–45. Tasto, J.J., Morrell, J.L. and Gould, K.L. (2003) An anillin homologue, Mid2p, acts during fission yeast cytokinesis to organize the septin ring and promote cell separation. Journal of Cell Biology, 160, 1093–103. Vrabioiu, A.M., Gerber, S.A., Gygi, S.P. et al. (2004) The majority of the Saccharomyces cerevisiae septin complexes do not exchange guanine nucleotides. Journal of Biological Chemistry, 279, 3111–18.

168

CH06 METAZOAN MODEL SYSTEMS

Vrabioiu, A.M. and Mitchison, T.J. (2006) Structural insights into yeast septin organization from polarized fluorescence microscopy. Nature, 443, 466–69. Vrabioiu, A.M. and Mitchison, T.J. (2007) Symmetry of septin hourglass and ring structures. Journal of Molecular Biology, 372, 37–49. White, J.G., Horvitz, H.R. and Sulston, J.E. (1982) Neurone differentiation in cell lineage mutants of Caenorhabditis elegans. Nature, 297, 584–87.

Section III Septins in mammals

The Septins Edited by Peter A. Hall, S.E. Hilary Russell and John R. Pringle © 2008 John Wiley & Sons, Ltd. ISBN: 978-0-470-51969-1

7 The genomics and regulation of the human septin genes S.E. Hilary Russell Ovarian Cancer Research Laboratory, Centre for Cancer Research and Cell Biology, Queen’s University Belfast, A floor, Belfast City Hospital, Lisburn Road, Belfast, Northern Ireland, UK

The ever growing catalogue of cellular processes involving septin gene products has been paralleled by the need to understand how septin gene information is expressed and regulated. Broadly, control of gene expression can be considered at three levels; the genome (genes and alleles), the transcriptome (mRNA transcripts) and the proteome (protein level). With respect to the genome, regulation of gene expression is influenced by the presence or absence of responsive elements within promoters and also promoter modification for example by methylation or acetylation. Transcriptional control of the rate of mRNA production is influenced by the availability of relevant transcription factors. Post transcriptionally, there is regulation by processing of the initial transcript including nuclear export and mRNA localization, alternative splicing and transcript stability. Translation can be regulated either globally within the cell in response to environmental cues or specifically by the mRNA itself. Finally, protein levels will also be subject to regulation by post translation modification and degradation. For the human septin genes, there is now evidence for considerable complexity at least at the genomic and transcript levels. Perhaps the septin which is best understood in these terms is septin 9 but evidence in the literature on other family members suggests that such features are common.

LESSONS FROM SEPT9 Interest in the SEPT9 gene developed in the late 1990s when it was identified as a fusion partner of the MLL gene in patients with acute myeloid leukaemia. MLL maps to chromosome 11q23 and is frequently rearranged in patients with therapy-related acute myeloid leukemia who previously were treated with DNA The Septins Edited by Peter A. Hall, S.E. Hilary Russell and John R. Pringle © 2008 John Wiley & Sons, Ltd. ISBN: 978-0-470-51969-1

172

CH07 THE GENOMICS AND REGULATION OF THE HUMAN SEPTIN GENES

topoisomerase II inhibitors. Osaka, Rowley and Zeleznik-Le (1999) identified a fusion partner of MLL in a 10-year-old female who developed therapy-related acute myeloid leukemia 17 months after treatment for Hodgkin disease. Leukemia cells of this patient had a t(11;17)(q23;q25), which involved MLL. The partner gene was cloned from cDNA of the leukemia cells by use of a combination of adaptor reverse transcriptase-PCR, rapid amplification of 5-prime cDNA ends (RACE), and BLAST database analysis to identify ESTs. The full-length cDNA of 2.8 kb was found to be a member of the septin family, and Osaka, Rowley and Zeleznik-Le (1999) therefore designated the gene MSF for ‘MLL septin-like fusion gene’. Indeed other septin family members have also been defined as MLL fusion partners although the significance of such fusions is not clear (see Hall chapter 14). In parallel, studies by groups interested in the genetic aberrations associated with the development of breast and ovarian tumours had shown allelic imbalance at distal chromosome 17q (17q25.3) and by fine deletion mapping identified SEPT9 as the target gene of this region (Kalikin, Sims and Petty, 2000; Russell et al., 2000). Through positional cloning of a contig across the region, DNA sequence was obtained and BLAST searches identified 1502 nucleotides of a 4-kb septin cDNA (then called Ov/Br septin) lying within the contig. From the sequence of the 4-kb cDNA (which was originally named the β transcript but is now SEPT9 v4 ), the genomic structure of the first six exons of this gene was predicted. A splice variant of this transcript (which was named the ‘α’ transcript and is now SEPT9 v1 ) was identified that differs at the 5
end because of an alternative exon 1. The alternative first exon was located 57 kb upstream from exon 1b and was 833 bases in length. PCR product generated from exon 2 and the beginning of exon 3 and thus common to both the α and β transcripts was used as a probe in Northern blot experiments against a variety of normal tissue RNAs. A 4-kb and 4.3-kb transcript were evident in 10 tissues (spleen, thymus, prostate, testis, ovary, small intestine, colon, peripheral blood lymphocyte, heart and pancreas), and there was variation in their relative amounts. In the remaining tissues, only the 4.3-kb transcript was detected. An additional genomic and expression analysis of this 17q25 septin gene, called MSF by this group, had identified three alternatively spliced transcripts (Kalikin, Sims and Petty, 2000). The MSF transcripts, included two transcripts (MSF-A and MSF-B), which were found to share nine coding exons. The intron/exon boundaries of these transcripts were described and six polymorphic variants identified. A third transcript was identified by database searches. So the idea was quickly established that more than one transcript was generated from this gene and it soon became clear that the picture was even more complex. A follow up study of the genomic structure of SEPT9 integrating new experimental and database information with that already published detailed 17 exons, asymmetrically distributed over ∼240 kb of sequence McIlhatton et al. (2001) (Figure 7.1). Alternative splicing is used to shuffle various combinations of eight 5
exons, six of which are unique to their respective mRNAs. This results in the generation of a number of splice variants that differ dramatically in the nature

LESSONS FROM SEPT9 D17S1790

D17S937

S

N

15 kb 13 kb 1ε 1ζ

2

173 D17S939 AFM203wc5

N

S

56 kb

1α

S

52 kb

27 kb 1γ

S N N

21 kb+

33 kb 1δ

3

1β

S

45678 9 11 12 10

10 kb

Figure 7.1 The genomic architecture of the SEPT9 gene showing the relative location of the various exons and introns and patterns of splicing which link different exons in the splice variants. The microsatellite markers (denoted by ↓) relate to the chromosomal position on 17q25.3 and were those used initially in the mapping studies of this gene. Distances given between exons are approximate and where this is unknown, a dashed line has been used. S and N denote Sal1 and Not1 sites respectively. ATG

SEPT9_v1 1ε

3

2

ATG

SEPT9_v2 3

1γ ATG

SEPT9_v3 3

1α

ATG

4

SEPT9_v4 3

1β

ATG

SEPT9_v4* 1ζ

2

SEPT9_v5 1δ

Figure 7.2 The experimentally verified splice variants of SEPT9 generated by shuffling of alternate 5
and 3
ends. All transcripts share the common exons 4–11 and also exon 12. However, 3
splice variants are produced by intra-exonic splicing within exon 12 at non-consensus splice sites (b and c) to access an open-reading frame at the end of the exon. Open boxes represent UTRs, ATG is the start codon for translation and * is the stop codon

and extent of their 5
UTR sequences but share the common exons 4–11 (Figure 7.2). The genomic organization of these exons, and the fact that some transcripts appear to demonstrate tissue specificity, suggests that multiple promoters probably feature in the transcription of SEPT9 . The ORFs of SEPT9 v1 , v2 and

CH07 THE GENOMICS AND REGULATION OF THE HUMAN SEPTIN GENES

174

v3 begin within their respective exon 1 and conceptual translation yields proteins that are very similar in size. SEPT9 v4 and v4* mRNAs are remarkable because, even though they differ over the first half of their 5
UTRs, both ORFs initiate at the same position within exon 3 and encode the same protein. Exon 3 embodies a further twist in the complex transcriptional arrangements of the SEPT9 gene in that its translational context is partially non-coding within SEPT9 v4 and v4* , where it comprises part of the 5
UTR, but wholly coding within SEPT9 v1 – v3 , where it contributes to the ORF. Exon 2 exhibits a comparable flexibility in its translational context. All of the standard intron/exon boundaries within the SEPT9 gene are demarcated by conserved splice donor/acceptor (GT/AG) dinucleotides (Senapathy, Shapiro and Harris, 1990), unlike the splicing events that occur within exon 12. As indicated in Figure 7.3, intervening non-coding segments of either 1801 bp (SVb) or 1849 bp (SVc) are removed through alternative splicing at cryptic sites which mediates the juxtaposition of the terminal 193 nucleotides of exon 12 with the remaining upstream sequences. This facilitates access to a small second ORF of 44 amino acids and generates a potential further twelve transcripts with altered 3
UTRs, corresponding to SEPT9 v1–5 b and c. Thus the various SEPT9 isoforms either differ minimally at their amino and carboxy termini or are equivalent to truncated versions of larger isoforms. THMQNIKDITSSIHFEAYRVKRLNEGSSAMANGVEEKEPEAPEM

4455 bp

2519 bp

(a) SSIPALRCFTFLLCNQTLKNCSFHQALFLNGLLLRASREICLFV

THMQNIKDITSSIHFEAYRVKRLNEG

2462 bp GGGC AGCAGCG

1801 bp spliced out

4263 bp GGGC AGCAGCATCC

(b) THMQNIKDIT

2414 bp C AGCAGCATCC

SSIPALRCFTFLLCNQTLKNCSFHQALFLNGLLLRASREICLFV

1849 bp spliced out

4263 bp GGGC AGCAGCATCC

(c)

Figure 7.3 Intra-exonic splicing within exon 12 of SEPT9 uses non-consensus splice sites in splice variants (b) and (c) to access a short discrete downstream open-reading frame maintained within the 3
UTR of splice variant (a)

EVIDENCE FROM OTHER SEPTIN GENES

175

EVIDENCE FROM OTHER SEPTIN GENES Following on from this complex genomic architecture described for SEPT9 , an obvious question is thus to determine if this is broadly representative of the septin gene family. It must first be stated that no other family member has received the same level of attention with respect to genomics and splice variants but interrogation of any of the databases suggests each septin gene has more than one transcript. Considerable caution must however be applied to interpretation of such data. Rarely has any of it been verified experimentally and given the high degree of homology between the various SEPT9 transcripts it is crucial that cloning artefacts or partial cDNAs are distinguished from bona fide full-length mRNAs. With these caveats in mind though, there are still important lessons to be learned. SEPT6 is a good example and another septin family gene identified as a fusion partner of MLL. Several patients have now been reported with complex chromosomal abnormalities involving 11q23 (MLL) and Xq24(SEPT6) (Borkhardt et al., 2001; Ono et al., 2002; Slater et al., 2002; Fu et al., 2003; Kim et al., 2003). As with SEPT9 , the fusion product contained most of the septin moiety and considerable deletion of the C terminal region of MLL. The gene (SEPT6 , Unigene Hs.496666, NCBI accession number NM 145802; D50918; Entrez GeneID: 23157) spans 85 kb and is composed of 12 exons. With respect to the transcripts from SEPT6, searches of the databases at NCBI (http://www.ncbi.nlm.nih.gov/), Ensembl (http://www.ebi.ac.uk/ensembl) and ASG gallery (http://statgen.ncsu.edu/asg/) indicates 5 predicted transcripts (I–V) which potentially encode four isoforms since transcripts I and III encode the same protein and differ only in their 3
UTRs. Verification of these transcripts has recently been completed (Todd et al., submitted) and highlights the importance of such a rigorous approach. It was predicted that a forward primer in exon 6 and reverse primer in exon 9 would generate products of 513 bp from transcripts I–III and V, and a 211 bp fragment from transcript IV. However there was no evidence of this smaller product in a panel of cell line cDNAs nor indeed was isoform C identified in the NCBI database and its existence cannot be confirmed. Additional PCR reactions with a forward primer in exon 9 and reverse primer in either exon 11 or 12 in a panel of cell line cDNAs confirmed the existence of the other splice variants. While both reactions gave rise to fragments of the predicted size based on the known SEPT6 splice variants, additional bands were also seen. When excised and sequenced these indicated previously unrecognized splicing events. For example, splicing from exon 10 into exon 11B and hence a transcript encoding the same isoform as transcripts I and III. In addition, the same forward primer but using a reverse primer in exon 12 indicated a transcript where exon 9 was spliced (in frame) into exon 12 at position 182 bp. Thus exon 12 has been designated as 12A and 12B (from 182 bp). This transcript encodes a novel polypeptide which includes 12 amino acids at the COOH terminus from sequence in exon 12B. The splice variants for SEPT6 can thus be redefined (Figure 7.4) and renamed according to the standard nomenclature (Macara et al., 2003). As

176

CH07 THE GENOMICS AND REGULATION OF THE HUMAN SEPTIN GENES

SEPT6_v1 6

7

8

9

12b

6

7

8

9

6

7

8

9

6

7

8

9

10

6

7

8

9

10

SEPT6_v2 11b

SEPT6_v3 12

SEPT6_v4 11b

SEPT6_v4** 11

SEPT6_v4* 6

7

8

9

10

12

SEPT6_v5 6

9

12

Figure 7.4 The splice variants of the SEPT6 gene demonstrating the extensive splicing involving the 3
exons. Open boxes represent UTRs, ATG is the start codon for translation and
is the stop codon

observed for other septin family members, multiple SEPT6 transcripts encode one isoform such that six transcripts encode four protein isoforms. Despite its presence in several databases and after exhaustive attempts, no experimental evidence for the existence of SEPT6 v5 could be provided. In addition to these studies RACE was performed and no additional SEPT6 transcripts were identified. Thus experimentally (and in keeping with entries in the various databases) it seems that all splicing for SEPT6 is 3
. And what of the other septin genes? Is this emphasis on 3
splicing as observed in SEPT6 typical or the exception? As already stated such studies are far from complete but a useful resource has recently come on line which may help. Fast DB (http://www.fast-db.com) is a new interactive database of 22,218 splice variants with a number of tools which allows their analysis (De la Grange et al., 2007). Eleven septin genes are listed and a simple comparison suggests that only 3
spliced transcripts are found for septins 3, 6, 10 and 11. For SEPT1 few splice

THE ROLE OF THE UTRs

177

variants were listed but all appear to have internal exons deleted. Both 5
and 3
splice variants are listed for the remaining genes. Some supporting experimental validation is found in the literature for SEPT2 (Mori et al., 1996); SEPT3 (Methner et al., 2001); SEPT4 (Paavola et al., 1999; Zieger et al., 2000; Larisch et al., 2000; Tanaka et al., 2001); SEPT5 (McKie et al., 1997; Yagi et al., 1998; Zieger et al., 2000); SEPT10 (Sui et al., 2003). In addition, it is also clear from this database that some septins (eg. SEPT1 , 2 and 4) promote deletion of internal exons in generation of some of their splice variants. Various tools have been applied to this database including one which predicts whether a particular splice variant is a target for the nonsense-mediated decay (NMD) pathway and hence is rapidly degraded (Lewis, Green and Brenner, 2003; Lejeune and Maquat, 2005). NMD is a quality-control mechanism that selectively degrades mRNAs harbouring premature termination (nonsense) codons. If translated, these mRNAs can produce truncated proteins with dominant-negative or deleterious gain-of-function activities. Based on well defined criteria, at least one splice variant of septins 1, 4, 5, 7 and 9 are predicted to be targets for this regulatory pathway. Another interesting strategy is employed by SEPT5 . Two transcripts of this gene, a major one of 2.2 kb and a minor one of 3.5 kb, have been observed. The 2.2 kb form which encodes this 369 amino acid protein, results from the utilization of a non-consensus polyA signal (AACAAT). In the absence of polyadenylation at this imperfect site, this gene uses the consensus polyA signal within its 3
neighbour gene (GP1BB; platelet glycoprotein Ib), resulting in the 3.5 kb read-through transcript (Zieger, Hashimoto and Ware, 1997). A transcript variant with a different 5
end arising by alternative splicing, and possibly also use of an alternative promoter, has also been identified; however, its full-length nature is not known (Yagi et al., 1998).

THE ROLE OF THE UTRs One of the earliest observations on SEPT9 genomics and expression was the definition of two transcripts which encoded the same polypeptide i.e. had the same ORF but differed in their 5
UTRs. These were SEPT9 v4 and v4* encoding the SEPT9 i4 protein with the v4 transcript being the usual form in normal cells but v4* becoming predominant in tumours (Burrows et al., 2003). Both have long 5
UTRs (∼700 bp). Most eukaryotic mRNAs have short 5
UTRs, of the order 100–150 bp; however, in the subset with long 5
UTRs, the majority represents genes involved in growth control or other regulatory mechanisms and often contain motifs involved in regulation of translation of that mRNA (Kozak, 1986; 1991). In fact the 5
UTRs of v4 and v4* have distinct 5
ends encoded by exons 1β (v4) and 1ζ and 2 (v4*) and a common 3
region and initiating ATG encoded within exon 3. There is now compelling evidence that these 2 mRNAs are not only translated with different efficiencies (the v4* mRNA more efficiently that

178

CH07 THE GENOMICS AND REGULATION OF THE HUMAN SEPTIN GENES

the v4 mRNA) but also that translation of the v4* mRNA is refractory to cellular stresses such as treatment with the drug etoposide or hypoxia (McDade, Hall and Russell, 2007). A putative internal ribosome entry site was identified in the common region of the v4 and v4* 5
-UTRs and translation modulated by a short upstream open-reading frame in the unique region of the v4 5
UTR. Indeed the evidence suggests that translation of the v4* transcript continues under cellular conditions where cap-dependant translation is inhibited while that of the v4 transcript does not. The availability of two transcripts (probably under control of two different promoters) encoding the same isoform and where one is essentially regulatable at the level of translation and the other not is consistent with an important role for that isoform perhaps in regulation of septin function as part of an overall septin scaffold. Clearly it will be important to define the normal physiological role and expression of the v4* (i.e. the unregulatable) transcript and why the cell requires such careful regulation of this isoform and its role in stress response mechanisms. Further tantalizing evidence for the importance of the 5
UTRs of SEPT9 comes from an alternative interpretation of data on germline mutations in families with hereditary neuralgic amyotrophy (HNA) (Kuhlenbaumer et al., 2005). In this condition affected individuals experience episodic pain and wasting of muscles in the limbs and especially the arms, often triggered by inter-current infection, prolonged exercise and other stress (Chance, 2006). Kuhlenbaumer et al. carried out a mutation screen of SEPT9 coding sequence and UTRs on 10 families with disease known to be linked to chromosome 17q25. Although three different germline mutations where identified in six of the families the authors reported that only two of these (mapping to exon 3) actually caused coding changes. One of these mutations, R88W, was identified in four families and since a founder effect was excluded it seems likely this represents a mutation hotspot. Curiously though, because of the complexities of the alternative splicing in this gene, this mutation is only within an ORF of the larger SEPT9 isoforms (i1-3). Careful sequence analysis shows that it maps to the common region of the 5
UTRs of SEPT9 v4 and SEPT9 v4* and is in fact within the region McDade et al., demonstrate necessary for translational control of this transcript and which is the putative internal ribosome entry site. They then introduced this mutation (262C→T) into reporter constructs of SEPT9 v4 and v4* for comparison with wild type under various conditions. The HNA mutation had no effect on the translatability of the SEPT9 v4* mRNA sequence under any condition nor on the v4 mRNA under normoxia. However, under hypoxic stress, the HNA mutation led to a greatly enhanced translation of v4 mRNA. This may explain the episodic nature of the disease and also the various stressful triggers associated with this – under normal conditions the rate of translation of the SEPT9 v4 mRNA is unaffected by the mutation but when the translation machinery is altered as a stress response is mounted, the mRNA is rendered sensitive to such changes by this mutation in a key regulatory motif. When similar experiments were carried out introducing the other exon 3 mutation (278C→T) similar but less dramatic effects were observed (data not shown). The third germline mutation in HNA families

CONTROL OF SEPTIN GENE TRANSCRIPTION

179

maps to the 5
UTR of SEPT9 v3 . This is the only other SEPT9 transcript with a long 5
UTR (812 bp) and although no experimental analysis has been carried out, it seems possible that some regulatory function must be attributed to such a long UTR. It is unlikely to be just coincidence that a germline HNA mutation maps to such a UTR. There is one report of functional effects of the exon 3 mutations on the larger SEPT9 isoforms but these are not particularly striking (Sudo., 2007). When all of this data is considered together and with the clinical picture, it seems more likely that HNA is associated with regulatory mutations in key SEPT9 mRNAs to which the cell is particularly sensitive under stress. With respect to the 3
splice variants of SEPT9 , it is worth noting that the predominant transcript in normal cells is variant a in which the 3
UTR is long (1936 bp). Most of this region is deleted in variants b and c by virtue of the unusual splicing at non-consensus splice sites within exon 12 and also results in amino terminal differences in the proteins. It is increasingly recognized that the 3
regions of mRNAs are important in regulating for example, mRNA stability/degradation, nuclear export, subcellular localization and translation and the 3
UTR is critical in governing mRNA interactions with both cis and trans acting factors (Chen, F´erec and Cooper, 2006). It is tempting to speculate that the different 3
regions of the SEPT9 transcripts represent a further mechanism for post-transcriptional regulation of this gene and may result in altered levels, subcellular localization or even post-translational modification by trans acting factors. As described above for SEPT6 , there is also complex splicing involving 3
UTRs and several transcripts encode the same isoform. It may also be for SEPT6 that the differences in the 3
UTRs of the SEPT6 v4 , v4* and v4** transcripts have implications for mRNA stability, location or translation. The use of alternate splicing to regulate gene expression by diverse means is increasingly recognized (Hughes, 2006) and the role of the 3
UTR is of particular significance including their potential regulation by microRNAs (De Moor, Meijer and Lissenden, 2005). It is noteworthy that the 3
regions of SEPT6 have several potential miRNA binding sites. A further feature of the fast DB data (De la Grange et al., 2007) is the prediction of miRNA target sites and consideration of the 11 septin genes listed intriguingly shows that each of them has at least one 3
splice variant potentially regulated by this pathway.

CONTROL OF SEPTIN GENE TRANSCRIPTION Little is known of the promoters of any of the human septin gene family but given the complexity of splicing and various patterns of splice variant expression in tissues, it is probable that each gene is under control of more than one promoter. For SEPT9 the close proximity of the first exons of the SEPT9 v2 and v4 transcripts might predict a single promoter regulating their expression and a putative CpG island was indeed identified upstream of the first of these exons. Interestingly, treatment of tumour cell lines where there are low or undetectable

180

CH07 THE GENOMICS AND REGULATION OF THE HUMAN SEPTIN GENES

levels of these transcripts with the demethylating agent 5
azacytidene led to their upregulation suggesting a role for promoter methylation in control of these SEPT9 splice variants (Burrows et al., 2003). An alternative approach namely identification of regions with high CpG content associated with first exons of SEPT9 splice variants and subsequent development of cloned bisulphite sequencing has proved useful (Lofton-Day et al., 2007). At least two CpG islands in putative promoters have been identified upstream of exons 1α (SEPT9 v1 ) and 1γ (SEPT9 v2 ) and their methylation observed in colorectal tumour samples but not in healthy whole blood samples. Thus despite the complexity of the genome and transcript regulation of this gene there is real optimism for a septin genome assay which has clinical application. Some clues as to factors which may be associated with expression of the septin genes can be found in the literature although some are little more than anecdotal and most to date have not been pursued. An early report regarding SEPT1 expression suggested it was up-regulated in cell lines expressing high levels of cell surface glycoprotein gp90MEL-14 and co-regulation of these two gene products was proposed (Nottenburg, Gallatin and St John, 1990). The use of retinoic acid to induce differentiation of a human teratocarcinoma cell line led to upregulation of SEPT3 and by a quantitative RTPCR approach, upregulation of this gene was also noted during neuronal differentiation (Methner et al., 2001). One interesting report of putative p53 responsive genes identified SEPT4 as a potential target of this transcription factor. Given recent reports indicating septin involvement in various stress response pathways including DNA damage (Kremer, Adang and Macara, 2007) and hypoxia (Amir et al., 2006), it would not be surprising if p53 was an important influence in regulating septin gene expression given its central role in maintenance of the genome. Finally, it is clear that expression of one septin gene can influence expression of another. The generation of SEPT5 null mice (which were phenotypically normal) was accompanied by decreased levels of SEPT7 and increased levels of SEPT2 – the other two septins which co-immunoprecipitated from brain (Peng et al., 2002). The lack of a phenotype was unexpected since previous experiments with dominant-negative SEPT5 had shown enhanced neurotransmitter secretion. One explanation to account for these observations is that disruption of SEPT5 transcription (in the knockout animals) led to generation of other septin complexes which had functional compensation. Thus the possibility should also be considered that cells have evolved a molecular circuitry by which there is co-regulation of particular septin genes as required for generation of a scaffold with certain functions and the ability to respond and compensate for altered levels of any one component.

EXPRESSION PROFILING Initial attempts to catalogue expression patterns of mRNA of a particular septin gene were by Northern blotting, but it soon became clear (in parallel with evolving

EXPRESSION PROFILING

181

knowledge of the genomics) that this would not be straightforward. Clearly with so many splice variants and often a high degree of homology between them make it difficult if not impossible to generate a transcript specific probe. Also, the similarity of many of the splice variants of a given gene means resolution on the basis of molecular weight is all but impossible. Various means for a global expression analysis exist for example SAGE, EST databases, cDNA microarrays etc and while some useful information can be deduced from these approaches all will ultimately require confirmation by transcript specific RTPCR and (probably) microdissected tissues. Many of the databases provide some attempt at transcript profiling in various tissues. For instance, fast DB contains a tissue distribution histogram of all gene transcripts lodged within its database to enable the expression pattern of a given gene to be visualized clearly. A comprehensive study of expression of 13 of the septin family was carried out by DNA microarray methods on over 10 000 samples including normal, disease and tumour material (Hall et al., 2005). This analysis highlighted a number of key points and suggested that most septins are expressed in all tissue types but some show enhanced expression in lymphoid or CNS tissues. Moreover, the presence of several probe sets on the microarray and careful mapping of these to known splice variants of a gene provided evidence for specific expression patterns for certain splice variants. However, important issues arose from the data; considerable care must be taken to assign a probe set to the gene in question since it emerged that some had been wrongly assigned and in fact did not map to any septin gene. Also, on occasion, apparently conflicting results with several probe sets for one gene emerged. Most likely the explanation for this is the existence of as yet unrecognized splice variants and their differential expression. So while this has been a useful resource it will not reach its full potential until we have a more precise estimate of all transcripts for a septin and their interrelationship. However, in general it is clear that SEPT1 , 6, 9 and 12 show high expression in lymphoid tissue and SEPT2 , 3, 4, 5, 7, 8 and 11 in brain. For a given septin, some isoforms are highly expressed in the brain while others are not. Thus SEPT8 v2 and v1, 1* and 3 are highly expressed in the brain and cluster with SEPT2 , 3, 4, 5, 7 and 11. However, a probe set specific for SEPT8 v1 with low brain expression clusters away from this set. Similarly in SEPT4 there are lymphoid and non-lymphoid forms; SEPT2 has lymphoid and CNS forms; and SEPT6 and SEPT9 show elevated expression in lymphoid tissues but also have forms which cluster away from such lymphoid forms. There is experimental data available for some of these observations which adds credence to them; both quantitative RTPCR and Northern analysis of SEPT3 demonstrated its almost neurospecific pattern of expression (Xue et al., 2000; Methner et al., 2001); abundant expression of SEPT4 in brain by Northern blotting (Zieger et al., 2000; Tanaka et al., 2001); Northern blot data for SEPT9 indicated some transcripts ubiquitously expressed but one found in haematopoietic tissues (Osaka, Rowley and Zeleznik-Le, 1999) while

182

CH07 THE GENOMICS AND REGULATION OF THE HUMAN SEPTIN GENES

a transcript specific RTPCR approach suggested the pattern was more complex across a panel of tissues (McIlhatton et al., 2001). A further interesting point from this expression data was the frequency with which disease (tumour or non-tumour) in a tissue was associated with up- or down regulation of a septin splice variant. Although outwith a discussion in this chapter and detailed more exhaustively in Chapter 14 it is once again an indication that maintenance of levels of septin splice variants is important for normal cellular physiology. Intriguingly, the microarray data for tumour and non-tumour disease samples of CNS showed that SEPT5 and SEPT7 were down-regulated while SEPT2 was up-regulated. Thus as with the SEPT5 null mouse described above (Peng et al., 2002) where SEPT7 was down-regulated and SEPT2 up-regulated alterations in expression of one septin gene (or splice variant) may be a consequence of changes in another.

FUTURE DIRECTIONS AND KEY AREAS To fully understand the genomic architecture of the entire family of human septin genes is a daunting task yet one which requires to be addressed with some urgency. It is clear that each septin gene has evolved in such a way that it can encode several proteins with a high degree of homology but apparently somewhat diverse functions. Some family members exhibit complex splicing confined to the 3
ends of the mRNAs, for others the variation is both 5
and 3
. Although not verified experimentally in each case, a common trend for all septin genes is apparently the capability of generating a particular isoform from several splice variants. We do not know the significance of this particular isoform of each gene but such is the process of evolution that it must have a crucial role in septin function or at least one which sets it apart from related isoforms. The data emerging from control of septin gene expression suggest that there is a good deal of tissue specificity with respect to these genes and indeed of individual splice variants. We might speculate that this is concerned with modifying activity of a septin scaffold but have no idea how the cell achieves it. Features of the various septin mRNAs suggests that virtually all cellular strategies known to contribute to control of levels of gene expression are employed by this family. So why is there such an emphasis on ‘fine tuning’ of activity of septin gene products? Our knowledge of septin function has been guided by over-expression, siRNA or gene knockout studies – effectively an ‘all or nothing’ approach. But, ‘science is the art of the doable’ (to paraphrase Nobel Laureate Sir Peter Medawar). It may be that to fully understand the intricacies of septins in cellular physiology, we must develop more subtle approaches.

ACKNOWLEDGEMENTS The author is grateful to Dr Simon McDade for assistance with the figures of SEPT9 and SEPT6 transcripts. This laboratory is funded by WellBeing of Women,

REFERENCES

183

The Pathological Society of Great Britain and Ireland, the NI Leukaemia Research Fund, the Samaritan Trust and HPSS R&D Office.

REFERENCES Amir, S., Wang, R., Matzkin, H. et al. (2006) MSF-A interacts with hypoxia-inducible factor-1alpha and augments hypoxia-inducible factor transcriptional activation to affect tumorigenicity and angiogenesis. Cancer Research, 66, 856–66. Borkhardt, A., Teigler-Schlegel, A., Fuchs, U. et al. (2001) An ins(X;11)(q24;q23) fuses the MLL and the Septin 6/KIAA0128 gene in an infant with AML-M2. Genes, Chromosomes and Cancer, 32, 82–88. Burrows, J.F., Chanduloy, S., McIlhatton, M.A. et al. (2003) Altered expression of the septin gene, SEPT9, in ovarian neoplasia. The Journal of Pathology, 201, 581–88. Chance, P.F. (2006) Inherited focal, episodic neuropathies: hereditary neuropathy with liability to pressure palsies and hereditary neuralgic amyotrophy. Neuromolecular Medicine, 8, 159–74. Chen, J.M., F´erec, C. and Cooper, D.N. (2006) A systematic analysis of diseaseassociated variants in the 3
regulatory regions of human protein-coding genes I: general principles and overview. Human Genetics, 120, 1–21. De la Grange, P., Dutertre, M., Correa, M. and Auboeuf, D. (2007) A new advance in alternative splicing databases: from catalogue to detailed analysis of regulation of expression and function of human alternative splicing variants. BMC Bioinformatics, 8, 180–94. De Moor, C.H., Meijer, H. and Lissenden, S. (2005) Mechanisms of translational control by the 3
UTR in development and differentiation. Seminars in Cell and Developmental Biology, 16, 49–58. Fu, J.F., Liang, D.C., Yang, C.P. et al. (2003) Molecular analysis of t(X;11)(q24;q23) in an infant with AML-M4. Genes, Chromosomes and Cancer, 38, 253–59. Hall, P.A., Jung, K., Hillan, K.J. and Russell, S.E. (2005) Expression profiling the human septin gene family. The Journal of Pathology, 206, 269–78. Hughes, T.A. (2006) Regulation of gene expression by alternative untranslated regions. Trends in Genetics, 22, 119–22. Kalikin, L.M., Sims, H.L. and Petty, E.M. (2000) Genomic and expression analyses of alternatively spliced transcripts of the MLL septin-like fusion gene (MSF) that map to a 17q25 region of loss in breast and ovarian tumors. Genomics, 63, 165–72. Kim, H.J., Ki, C.S., Park, Q. et al. (2003) MLL/SEPTIN6 chimeric transcript from inv ins(X;11)(q24;q23q13) in acute monocytic leukemia: report of a case and review of the literature. Genes, Chromosomes and Cancer, 38, 8–12. Kozak, M. (1986) Point mutations define a sequence flanking the AUG initiator codon that modulates translation by eukaryotic ribosomes. Cell , 44, 283–92. Kozak, M. (1991) Structural features in eukaryotic mRNAs that modulate the initiation of translation. The Journal of Biological Chemistry, 266, 19867–70. Kremer, B.E., Adang, L.A. and Macara, I.G. (2007) Septins regulate actin organization and cell-cycle arrest through nuclear accumulation of NCK mediated by SOCS7. Cell , 130, 837–50. Kuhlenbaumer, G., Hannibal, M.C., Nelis, E. et al. (2005) Mutations in SEPT9 cause hereditary neuralgic amyotrophy. Nature Genetics, 37, 1044–46.

184

CH07 THE GENOMICS AND REGULATION OF THE HUMAN SEPTIN GENES

Larisch, S., Yi, Y., Lotan, R. et al. (2000) A novel mitochondrial septin-like protein, ARTS, mediates apoptosis dependent on its P-loop motif. Nature Cell Biology, 12, 915–21. Lejeune, F. and Maquat, L.E. (2005) Mechanistic links between nonsense mediated mRNA decay and pre-mRNA splicing in mammalian cells. Current Opinion in Cell Biology, 17, 309–15. Lewis, B.P., Green, R.E. and Brenner, S.E. (2003) Evidence for the widespread coupling of alternative splicing and nonsense-mediated mRNA decay in humans. Proceedings of the National Academy of Sciences of the United States of America, 100, 189–92. Lofton-Day, C., Model, F., DeVos, T. et al. (2007) DNA-methylation biomarkers for blood-based colorectal cancer screening. Clinical Chemistry, 54, 414–23. McDade, S.S., Hall, P.A. and Russell, S.E. (2007) Translational control of SEPT9 isoforms is perturbed in disease. Human Molecular Genetics, 16, 742–52. McIlhatton, M.A., Burrows, J.F., Donaghy, P.G. et al. (2001) Genomic organization, complex splicing pattern and expression of a human septin gene on chromosome 17q25.3. Oncogene, 20, 5930–39. McKie, J.M., Sutherland, H.F., Harvey, E. et al. (1997) A human gene similar to Drosophila melanogaster peanut maps to the DiGeorge syndrome region of 22q11. Human Genetics, 101, 6–12. Macara, I.G., Baldarelli, R., Field, C.M. et al. (2002) Mammalian septins nomenclature. Molecular Biology of the Cell , 13, 4111–13. Methner, A., Leypoldt, F., Joost, P. and Lewerenz, J. (2001) Human septin 3 on chromosome 22q13.2 is upregulated by neuronal differentiation. Biochemical and Biophysical Research Communications, 283, 48–56. Mori, T., Miura, K., Fujiwara, T. et al. (1996) Isolation and mapping of a human gene (DIFF6) homologous to yeast CDC3, CDC10, CDC11, and CDC12, and mouse Diff6. Cytogenetics and Cell Genetics, 73, 224–27. Nottenburg, C., Gallatin, W.M. and St John, T. (1990) Lymphocyte HEV adhesion variants differ in the expression of multiple gene sequences. Gene, 95, 279–84. Ono, R., Taki, T., Taketani, T. et al. (2002) SEPTIN6, a human homologue to mouse Septin6, is fused to MLL in infant acute myeloid leukemia with complex chromosomal abnormalities involving 11q23 and Xq24. Cancer Research, 62, 333–37. Osaka, M., Rowley, J.D. and Zeleznik-Le, N.J. (1999) MSF (MLL septin-like fusion), a fusion partner gene of MLL, in a therapy-related acute myeloid leukemia with a t(11;17)(q23;q25). Proceedings of the National Academy of Sciences of the United States of America, 96, 6428–33. Paavola, P., Horelli-Kuitunen, N., Palotie, A. and Peltonen, L. (1999) Characterization of a novel gene, PNUTL2, on human chromosome 17q22–q23 and its exclusion as the Meckel syndrome gene. Genomics, 55, 122–25. Peng, X.R., Jia, Z., Zhang, Y. et al. (2002) The septin CDCrel-1 is dispensable for normal development and neurotransmitter release. Molecular and Cellular Biology, 1, 378–87. Russell, S.E., McIlhatton, M.A., Burrows, J.F. et al. (2000) Isolation and mapping of a human septin gene to a region on chromosome 17q, commonly deleted in sporadic epithelial ovarian tumors. Cancer Research, 60, 4729–34.

REFERENCES

185

Senapathy, P., Shapiro, M.B. and Harris, N.L. (1990) Splice junctions, branch point sites, and exons: sequence statistics, identification, and applications to genome project. Methods in Enzymology, 183, 252–78. Slater, D.J., Hilgenfeld, E., Rappaport, E.F. et al. (2002) MLL-SEPTIN6 fusion recurs in novel translocation of chromosomes 3, X, and 11 in infant acute myelomonocytic leukaemia and in t(X;11) in infant acute myeloid leukaemia, and MLL genomic breakpoint in complex MLL-SEPTIN6 rearrangement is a DNA topoisomerase II cleavage site. Oncogene, 21, 4706–14. Sui, L., Zhang, W., Liu, Q. et al. (2003) Cloning and functional characterization of human septin 10, a novel member of septin family cloned from dendritic cells. Biochemical and Biophysical Research Communications, 304, 93–98. Sudo, K., Ito, H., Iwamoto, I. et al. (2007) SEPT9 sequence alternations causing hereditary neuralgic amyotrophy are associated with altered interactions with SEPT4/ SEPT11 and resistance to Rho/Rhotekin-signaling. Hum Mutat., 28, 1005–13. Tanaka, M., Tanaka, T., Kijima, H. et al. (2001) Characterization of tissue- and celltype-specific expression of a novel human septin family gene, Bradeion. Biochemical and Biophysical Research Communications, 286, 547–53. Yagi, M., Zieger, B., Roth, G.J. and Ware, J. (1998) Structure and expression of the human septin gene HCDCREL-1. Gene, 212, 229–36. Xue, J., Wang, X., Malladi, C.S. et al. (2000) Phosphorylation of a new brain-specific septin, G-septin, by cGMP-dependent protein kinase. The Journal of Biological Chemistry, 275, 10047–56. Zieger, B., Hashimoto, Y. and Ware, J. (1997) Alternative expression of platelet glycoprotein Ib(beta) mRNA from an adjacent 5
gene with an imperfect polyadenylation signal sequence. The Journal of Clinical Investigation, 99, 520–25. Zieger, B., Tran, H., Hainmann, I. et al. (2000) Characterization and expression analysis of two human septin genes, PNUTL1 and PNUTL2. Gene, 261, 197–203.

8 The functions of septins in mammals Carol D. Froese and William S. Trimble Program in Cell Biology, Hospital for Sick Children, Toronto, Canada

INTRODUCTION Septins are conserved from yeast to mammals, but their functions are far better understood in yeast than in animal cells. This is in part due to the powerful genetic approaches that can be applied to their study in yeast cells, and to some extent to the increased gene complexity in mammals. However, the greatest challenge in defining a general function for septins in mammals has come from the almost bewildering variety of different roles to which they have been ascribed to date. With a view to the assimilation of these diverse ideas, this chapter will first describe the features and properties of mammalian septins in general. We will then broadly discuss the many and varied septin-protein interactions and functions that have been proposed for them in the context of these biochemical properties. With this background we will attempt to provide some coherence to the plurality of roles that septins may play, with a focus on integrating pathways and cytoskeletal networks within the mammalian cell. Many of these proposed functions and interactions will be described in greater detail in the subsequent chapters of this section. Septins are a family of filamentous guanosine adenine triphosphatases (GTPases) first discovered to be important in the cell division cycle in budding yeast (Hartwell, 1971). They have since been shown to figure in various other processes in fungi as well as in animals. In budding yeast, septins have been reported to function as diffusion barriers, cell cycle checkpoint regulators and dynamic, regulated scaffolds. They are critical in such processes as bud site selection and cytokinesis (Chapters 3, 4, 5 and 6). Nevertheless, the precise functions of the The Septins Edited by Peter A. Hall, S.E. Hilary Russell and John R. Pringle © 2008 John Wiley & Sons, Ltd. ISBN: 978-0-470-51969-1

188

CH08 THE FUNCTIONS OF SEPTINS IN MAMMALS

septins, and whether they function as monomers, complexes or filaments is not yet clear, and this is particularly true for animal septins. The first animal septins were described in differential display and random cloning studies (Nottenberg, Gallatin and John, 1990; Kumar, Tomooka and Noda, 1992) and it was not until 1994 that functional studies revealed a role for septins in animal cell division in Drosophila (Neufield and Rubin, 1994). However, new and unexpected roles continue to be revealed. As described in Chapter 7, based on genome sequencing projects, we now know that there are at least 14 septin genes in mammals. These genes can be multiply spliced to produce a complicated array of different polypeptides. The most extreme case of alternative splicing is in the SEPT9 gene, which can encode 18 different messanger ribonucleic acids (mRNAs) and 15 different polypeptides (McIlhatton et al., 2001 and see Chapter 7). In dividing cells, those septins that are expressed appear to be necessary for cell division. Interestingly however, most of the septins identified in mammals are highly expressed in the brain in post-mitotic neurons (unpublished observations), suggesting that the roles of septins in mammals may extend far beyond those in budding yeast. Below we will discuss the general properties of septin proteins that may help to elucidate their varied functions.

PHYSICAL PROPERTIES OF SEPTINS Septin complexes Mammalian septins are similar in sequence and overall structure to the yeast septins. They have a variable sequence amino terminal domain, followed in most septins by a polybasic motif adjacent to the GTPase domain, and then a series of highly conserved amino acids termed the septin unique element. Neither the polybasic motif nor the septin unique element are found in all septins (Pan, Malmberg and Momany, 2007), but both are found in most mammalian septins. In most septins the septin unique element is followed by a coiled-coil domain of variable length (Versele and Thorner, 2005). The septin proteins are generally unstable when expressed alone, and a complex of septins appears to be their favoured and probably functional configuration. In vitro binding and yeast 2-hybrid studies have allowed an interaction map to be developed for the mitotic budding yeast septins (Versele and Thorner, 2005) allowing the development of a model which suggests that septins associate in an ordered manner to form a complex containing stoichiometric ratios of each subunit. These complexes would then serve as the building blocks for the assembly of much larger septin filaments. Based on published interaction data for mammalian septins, and on classification of the 14 mammalian septins into four subfamilies using sequence similarity, the presence or absence of conserved coiled-coil domains, and the length of N and C terminal extensions, Versele and Thorner (2005) predicted that each subfamily of mammalian septins would have an orthologous relationship with one of the four

PHYSICAL PROPERTIES OF SEPTINS

189

main yeast septins. Mammalian SEPT6, 8, 10, 11 and 14 (Peterson et al., 2007) are predicted to be most like Saccharomyces cerevisiae Cdc3p, and SEPT3, 9 and 12 are most like Cdc10p. Whereas SEPT1, 2, 4 and 5 are most like Cdc11p, and SEPT7 and 13 are analogous to Cdc12p. This has allowed Versele and Thorner to propose a universal assembly model for septins which predicts that septin complexes are octomers comprised of two copies of a protein from each family. This attractive model is consistent with the idea that different septins have different tissue expression patterns, and would lead to the prediction that combinatorial mixtures of septins could account for a huge number of different complexes. In support of this mix-and-match assembly scheme, it seems clear that many of the mammalian septins are functionally redundant, as mouse knockouts of Sept3, 4, 5 or 6 appear to have subtle phenotypes despite the fact that these septins are highly or solely expressed in the brain (Peng et al., 2002; Ihara et al., 2005; Kissel et al., 2005; Ono et al., 2005; Fujishima et al., 2007; unpublished observations). Even double knockouts of Sept4 and 6, or Sept3 and 5 do not appear to have gross morphological defects (Ono et al., 2005; unpublished observation). Not surprisingly, this has complicated the analysis of their functions. The idea that the septin complex needs to be comprised of stoichiometric assemblies of septins from four families does not appear to hold up entirely. First, septin filaments can assemble from a single septin in vitro or when overexpressed in cells (Mendoza, Hyman and Glotzer, 2002; Nagata et al., 2003; Schmidt and Nichols, 2004a; Huang et al., 2006). Moreover, when SEPT2, 6 and 7 are co-expressed either in bacterial or insect cell systems, they can assemble into complexes that are capable of forming filaments (Sheffield et al., 2003; Kinoshita et al., 2002). In addition, many studies have examined endogenous septin complexes. The first mammalian septin complex was identified by co-purification with the exocyst complex from rat brains (Hsu et al., 1998). The septins identified in that study were SEPT2, 5, 6 and 7. Interestingly, when mouse brain lysate was used to precipitate SEPT2-containing complexes, SEPT1, 2, 4, 5, 6, 7, 8 and 9 coimmunoprecipitated with an ambiguous stoichiometry (Kinoshita et al., 2002). This heterogeneity suggests that several different complexes are being extracted from brain tissue. In a simpler system, when an antibody against SEPT2 is used for immunoprecipitation from HeLa cells it coprecipitates SEPT6, 7, 8 and 9 (Kinoshita et al., 2002). Similar results were obtained when antibody to SEPT9 was found to immunoprecipitate SEPT2, 6 and 7 in near equimolar amounts from HeLa cells (Surka, Tsang and Trimble, 2002), supporting the notion that these septins form a unit complex. In human platelets SEPT5 can be immunoprecipitated in a complex with SEPT2, 6, 7, 8 and 9 as judged by western blotting and mass spectrometry (Martinez et al., 2006). In rat embryonic fibroblast cells, SEPT7 antibody was able to immunoprecipitate a complex pattern of polypeptides that were visualized on a Coomassie blue-stained gel. The septins in this complex, identified by western blotting, were SEPT2, 8, 9 and 11, and there may have been others not identified (Nagata et al., 2004).

190

CH08 THE FUNCTIONS OF SEPTINS IN MAMMALS

In complex tissues such as the brain, different cell types could express different sets of septins, but in pure cell types or cell lines, such as platelets, rat embryo fibroblasts or HeLa cells most studies report five or more different septins in immunoprecipitates and often more than one member from a subfamily. Thus, while the in vitro data supports the concept that a minimal set of septin proteins can participate in complex formation, the fact that a single septin can co-precipitate more than the expected three septins suggests several possibilities. It is feasible that octomeric septin complexes could contain substoichiometric mixtures of different members of the same subfamily. Alternatively, a single cell could have more than one type of octomeric septin complex, each of which is comprised of a stoichiometric mixture. In addition, some septins may function individually outside of the unit complex. For example, in interphase HeLa cells, SEPT9 i1 is primarily associated with microtubules, apparently in the absence of other septins, while SEPT9 i2 and SEPT9 i4 are found on actin stress fibres with SEPT2 (Surka, Tsang and Trimble, 2002). A more complicated possibility is that the size and composition of the building block subunit complex could differ depending upon the pattern of septins expressed in a particular cell, and that septin filaments may be far more heterogeneous than has been hitherto proposed. Thus the ordering of septins and septin-containing structures within mammalian cells remains incompletely understood. Recent advances have begun to address the issue of septin organization and led to the suggestion that septin filaments assemble from symmetric complexes in a non-polarized fashion. Studies by John et al. (2007) have revealed that the C. elegans septins UNC-59 and UNC-61 assemble into a unit complex of UNC-59/UNC-61/UNC-61/UNC-59 which would then assemble into filamentous structures end on end. This concept is supported by the recent crystallographic study of human SEPT2 alone and a complex of SEPT2, 6 and 7 (Sirajuddin et al., 2007). The crystal structure reveals that septins oligomerize by interacting at two major surfaces, one comprising the N and C termini of the protein, and the other involving a face of the GTPase domain. In the case of the SEPT 2/6/7 complex, it appears that the septins organize into a symmetric hexamer with the order 7/6/2/2/6/7. Once again, it would appear that polymerization into longer filaments would occur in a non-polarized fashion.

Septin filaments When endogenous septin complexes are purified from yeast (Frazier et al., 1998) or Drosophila (Field et al., 1996), or when recombinant mammalian septin proteins are incubated at appropriate low salt conditions (Kinoshita et al., 2002; Sheffield et al., 2003), septin complexes can polymerize into very long filaments in vitro. Surprisingly, when SEPT2 alone from Xenopus laevis (Mendoza, Hyman and Glotzer, 2002) or humans (Huang et al., 2006) is maintained at high concentrations in vitro, it is capable of polymerization into similar filaments,

PHYSICAL PROPERTIES OF SEPTINS

191

although there may be differences between these species in the contribution of GTP binding and hydrolysis to this polymerization. While SEPT2 alone appears capable of polymerization, the near stoichiometric ratios seen from immunoprecipitated septin complexes suggest that in vivo there is a strong preference for heterotypic over homotypic septin oligomerization. Septins indeed have a filamentous appearance in many cell types, but their association with other cellular filaments, such as actin stress fibres and microtubules, has left open the caveat that they may simply appear filamentous by association. However, it is likely that they exist as long filaments in many cell types, since the once straight actin-associated structures gradually curve to form ring-like structures by end-to-end annealing, following latrunculin B-induced disassembly of actin filaments (unpublished observations). Based on the observation that the rings formed have a diameter of approximately 0.5–0.6 µm (Kinoshita et al., 2002), the filaments that gave rise to them would have to be more than 1.5 µm long. In addition, septin filaments can assemble laterally to form thick bundles of filaments (Kinoshita et al., 2002).

Lipid binding Septins interact directly with lipid and appear to be involved in membrane dynamics. SEPT4 was shown to interact with phosphoinositol 4,5 bisphosphate (PI(4,5)P) via a conserved polybasic domain in the amino-terminus half of the protein. Sequestering this lipid leads to a loss of SEPT 4-containing filaments (Zhang et al., 1999). Most of the other septins also have a region of basic amino acids in their primary sequences in the same position, although these are less dramatic or absent in the SEPT6, 8, 10 and 11 subgroup. In protein– lipid overlays, SEPT2 and SEPT12 interact most strongly with monophosphorylated phosphoinsositols and phosphatidylserine (Steels et al., 2007) although the lipid-binding specificity of these septins has not been observed when using liposomes as a more realistic model of cellular membranes. Yeast septins also interact preferentially with monophosphorylated phosphatidylinositols when protein– lipid overlays are performed, and interestingly they do not assemble properly during mitosis when PI(4)P synthesis is disrupted (Casamayor and Snyder, 2003). In addition, depletion of yeast PI(4,5)P2 by overexpression of mammalian PI3-kinase results in disruption in septin architecture in the cells (Rodriguez-Escudero et al., 2005). Binding of septins to membranes may be important for septin targeting or, in case of a diffusion barrier, for their function. It is not clear if the septins alter the availability of the phosphoinositides, but they could serve to sequester it and affect downstream signalling by this important second messenger.

Curvature Mammalian septins are frequently found in ring-like structures, and this has led to suggestions that the ring form may be an intrinsically favoured structure. Upon

192

CH08 THE FUNCTIONS OF SEPTINS IN MAMMALS

treatment of mammalian cells with agents that disrupt the actin cytoskeleton, the filamentous appearance of SEPT2 is lost and SEPT2-containing rings with a diameter of approximately 0.6 µm in diameter become apparent (Xie et al., 1999; Kinoshita et al., 2002). When actin filaments are restored the SEPT2-containing rings disappear and SEPT2 resumes its filamentous appearance and relocalizes along actin filaments (Kinoshita et al., 2002), prompting speculation that these rings may represent a storage form or default structure for septins. Purified septin complexes consisting of SEPT2, 6 and 7 dialysed into low salt buffers also form rings as well as spirals (Kinoshita et al., 2002). These rings can also form spontaneously under conditions of serum starvation and can be found in neurons and a neuroblastoma cell line (unpublished observation). SEPT2-containing rings have also been reported in normal rat kidney epithethial (NRK) cells during cell ruffling, and when green fluorescent protein (GFP)-tagged septins are overexpressed, fluorescent recovery after photobleaching suggests that these rings are highly dynamic (Schmidt and Nichols, 2004a). The formation of septin rings is consistent with rings seen in other organisms. For example, in Drosophila, septins make up stable intercellular bridges known as ring canals in male germ cells (Hime, Brill and Fuller, 1996). In mammalian spermatozoa, a specialized ring-shaped structure called the annulus is comprised of SEPT1, 4, 6, 7 and 12 and this structure separates the middle and the principal pieces of the sperm flagellum (Ihara et al., 2005; Kissel et al., 2005; Steels et al., 2007). As discussed below, the annulus is necessary for the structure and the function of the flagellum.

SEPTIN FUNCTIONS A wide array of functions have been proposed for septins, and assimilation of these disparate roles is not straightforward. We have therefore considered the proposed functions within general categories and discuss the possible overlaps between them. For a summary of the putative functions for mammalian septins see Table 8.1.

Mechanical stability Perhaps the most basic function for a filamentous septin complex would be that it provides some structural integrity to the cell. This is indeed the case in the sperm tail septin annulus where septins contribute to flagellar function. The annulus is a fibrous structure seen to separate the middle and principal pieces of the sperm tail, and in spermatozoa from SEPT4 knockout mice the annulus is absent. This results in spermatozoa with flagella that are structurally fragile, as manifested by flagella which are bent and broken (Ihara et al., 2005; Kissel et al., 2005), with resultant male infertility.

SEPTIN FUNCTIONS

193

Table 8.1 Septin functions. The specific septins observed to be involved in each function with the appropriate references are listed Function

Septin

Reference

Mechanical stability

SEPT1, SEPT4, SEPT6, SEPT7, SEPT9, SEPT12 SEPT2, SEPT6, SEPT7

Tooley et al. 2007; Steels et al. 2007 Kissel et al. 2005; Ihara et al. 2005

Scaffolding platforms Actin dynamics

SEPT4, SEPT2, SEPT9, SEPT6/7

Microtubule regulation

SEPT9 i1 SEPT 2/6/7 SEPT2

Membrane trafficking

SEPT2, SEPT4, SEPT5, SEPT9

Membrane protein regulation Diffusion barrier Viral replication

SEPT1 SEPT2 SEPT2 SEPT2 SEPT6 SEPT4 SEPT2/6/7 complex SEPT4 i2 (formerly ARTS)

DNA repair Apoptosis

Spiliotis, Kinoshita and Nelson, 2005 Xie et al. 1999; Kinoshita et al. 1997; Kinoshita et al. 2002; Surka, Tsang and Trimble, 2002; Nagata et al. 2004; Ito et al. 2005; Schmidt and Nichols, 2005; Joo, Surka and Trimble, 2007; Nagata and Inagaki, 2005 Surka, Tsang and Trimble, 2002 Nagata et al. 2003 Vega and Hsu, 2003 Kremer, Haystead and Macara, 2005 Beites et al. 1999; Ihara et al. 2001; Hsu et al. 1998; Dent et al. 2002; Tanaguchi et al. 2007 Nottenburg, Gallatin and John, 1990 Kinoshita et al. 2004 Schmidt and Nichols, 2004b Kim et al. 2007 Lin et al. 2007 Kremer, Adang and Macara, 2007 Larisch et al. 2000 and Gottfried et al. 2004

Another example where septins appear to provide a structural role is in the uropod of the T cell. Krummel and colleagues have found that septins formed an annular collar around the base of the uropod in migrating T cells and septin depletion caused the uropod to bend and lose shape. However, despite having normal migratory rates, the depleted cells were capable of migrating through smaller pores than the control T cells. This data is consistent with a role of septins in providing a rigidifying structure to the uropod. (M. Krummel, personal communication). Septins may also have structural roles in platelets where they localize to the periphery near the circumferential band of microtubules. This band is implicated in maintaining the shape and integrity of platelets and the co-localization of septins suggests that they may contribute to this property (Martinez et al., 2006), although this remains to be confirmed.

194

CH08 THE FUNCTIONS OF SEPTINS IN MAMMALS

Scaffolding platforms The numerous combinations of mammalian septins in unit complexes, coupled with the myriad of splice variants make the study of these proteins difficult. But this very diversity makes them good candidates for a role as a platform for interactions between other proteins. This was an early suggestion for a general function of septins, and it has been convincingly demonstrated for yeast septins where numerous proteins require septins at the mother-bud junction for their proper localization during cell division (Douglas et al., 2005). The number of septin-interacting proteins identified in mammals continues to grow and so it is likely that the mammalian septins also function as scaffolds for other proteins. One example of this platform function is that septins may coordinate the timing of chromosome congression and segregation with the onset of anaphase and cytokinesis by forming a platform on which the mitotic kinesin, centromereassociated protein E (CENP-E) and other effectors can interact (Spiliotis, Kinoshita and Nelson, 2005). This was demonstrated by depleting Madin–Darby canine kidney(MDCK) cells of SEPT2, which resulted in lower levels of SEPT6 and SEPT7 (Kinoshita et al., 2002). Septin depletion also resulted in mislocalized CENP-E, which is normally found at the kinetochore ends of disassembling microtubules. Interestingly, unattached kinetochores accumulated the protein Mad2, which halted progression of mitosis. Thus, these septins appear to have a role in the activation of a mitotic checkpoint. SEPT2 or SEPT2–containing filaments also appear to provide a platform for modification of non-muscle myosin II by Rho kinase and citron kinase as dissociation of the interaction between Sept2 and non-muscle myosin II significantly decreases diphosphorylation of myosin II light chain (Joo, Surka and Trimble, 2007). It is possible that the process of modifying proteins by the small ubiquitin-like modifier (SUMO) may also require septins as a platform on which to hold the required enzymes of its conjugation pathway, in some circumstances. Drosophila septins associate with both the E1 and E2 of this pathway (Nguyen et al., 2000) and mammalian septins associate with SUMO itself, the E2 and members of the protein inhibitor of activated STAT (PIAS) family, which have been shown to act as E3s in this pathway. Despite the association with these components, animal septins do not appear to be sumoylated in vivo (Shih et al., 2002; unpublished observations). SEPT9 i1 (the protein product of the SEPT9 v1 transcript) was recently shown to associate with hypoxia-inducible factor-1α (HIF-1α) in the nucleus of PC-3 cells, a human prostatic cancer cell line, which forestalled HIF-1 ubiquitination and subsequent degradation (Amir et al., 2006). SUMO and ubiquitin have been shown to compete for specific lysine residues (Ulrich, 2005) and it is tempting to speculate that HIF-1α may be stabilized by a septin-dependent addition of SUMO.

SEPTIN FUNCTIONS

195

Relationship with the actin and microtubule cytoskeletons Actin dynamics Many mammalian septins have been shown to co-localize with actin structures in several cell types (Kinoshita et al., 1997; Kinoshita et al., 2002; Surka, Tsang and Trimble, 2002; Nagata et al., 2004; Ito et al., 2005). Indeed, disruption of the actin cytoskeleton results in the loss of septin–actin co-localization and the appearance of septin rings in the cytosol (Xie et al., 1999; Kinoshita et al., 2002; Schmidt and Nichols, 2004a). Moreover, proteins involved in regulation of actin dynamics also affect septin organization. The small GTPase Cdc42 has been shown to be important for septin assembly in yeast (Richman, Sawyer and Johnson, 1999; Tjandra, Compton and Kellogg, 1998) and in mammalian cells a Cdc42 effector known as Borg3 interacts with the SEPT6/SEPT7 heterodimer (Joberty et al., 2001; Sheffield et al., 2003). Overexpression of a single domain of Borg3, the BD3 domain, or full length Borg3 with a large amino terminal tag, results in septin filament aggregation and subsequent disassembly of actin filaments. Conversely, the disruption of septin filaments by siRNA results in the disappearance of central actin stress fibres, indicating that septin structures and actin structures are dependent on one another (Kinoshita et al., 2002; Schmidt and Nichols, 2004a; unpublished observations). Actins and septins have never been demonstrated to interact directly, but it is believed that this interaction requires a protein mediator. Anillin, a component of the contractile ring of mitosis and an actin-binding protein, has been shown to interact with septins (Kinoshita et al., 2002), but anillin is nuclear during most of the cell cycle (Suzuki et al., 2005), suggesting that there is another protein acting as an adaptor between septins and actin during at least part of interphase. One candidate for this is non-muscle myosin II, an actin-based motor protein responsible for cross-linking actin stress fibres. In support of this, SEPT2 binds directly to the coiled-coil domain of the heavy chain of myosin II and this interaction appears to be important in controlling phosphorylation of myosin light chain (MLC) and myosin activation (Joo, Surka and Trimble, 2007). Importantly, a septin– myosin interaction has been shown, by yeast 2-hybrid screens, between yeast type II myosin and a yeast septin (Drees et al., 2001). Thus, septins are dependent on actin for their distribution, but in turn also affect actin filament bundling and cell contraction via myosin. In a related discovery, septins were also found to interact with proteins involved in regulation of the actin cytoskeleton. One protein shown to bind to SEPT9 was the septin-associated or SA-Rho guanine nucleotide exchange factor (GEF) (Nagata and Inagaki, 2005), a GEF for Rho. Rho is a small GTPase responsible for the bundling of central actin stress fibres. Active Rho activates Rho kinase (ROCK) which in turn phosphorylates both MLC to promote actin binding, and MLC phosphatase to prevent dephosphorylation of MLC (Matsumura,

196

CH08 THE FUNCTIONS OF SEPTINS IN MAMMALS

2005). Although it was shown that SEPT9 binding inhibited SA-Rho GEF, it is tempting to speculate that this type of interaction would fit with the scaffolding model for septin functions where the myosin substrate is held in close proximity to the kinase responsible for its activation. In addition to SA-Rho GEF, rhotekin, a downstream effector of Rho co-localizes with SEPT9 i3 (see Appendix B) although this binding is not direct (Ito et al., 2005). Unfortunately, the function of rhotekin and its role in this process remain unclear.

Microtubule regulation Besides the actin cytoskeleton, septins also associate with microtubules. SEPT9 i1 is dependent on the microtubule cytoskeleton for filament stability, and disruption of microtubules results in an alteration of the association of SEPT9 i1 with microtubules. Under these circumstances SEPT9 i1 co-localizes with actin filaments (Surka, Tsang and Trimble, 2002; Nagata et al., 2003). The association of SEPT9 i1 with tubulin is direct and it appears to be specific to the longest isoform of SEPT9, as SEPT9 i3 associates with the actin cytoskeleton (Surka, Tsang and Trimble, 2002). Interestingly, SEPT9 i1 is associated with acetylated microtubules in HeLa cells (Surka, Tsang and Trimble, 2002), which are more stable than non-acetylated microtubules. It is worth remembering that these three SEPT9 isoforms are almost identical differing in the N terminal 25, 18 and 7 residues only (see Chapter 7). Under certain circumstances, SEPT2 may also associate with tubulin (Vega and Hsu, 2003; Spiliotis, Kinoshita and Nelson, 2005). Septins are also involved in regulating the stability of the microtubule cytoskeleton. The SEPT2/6/7 heterotrimer can bind to MAP4, one of the microtubuleassociated proteins involved in microtubule binding and stability, and the interaction between septins and MAP4 was shown to be specifically to SEPT2. It has been proposed that septin filaments sequester MAP4 resulting in the destabilization of microtubules. In support of this, knockdown of SEPT7 or 2, which also depletes SEPT6, increased the levels of acetylated tubulin indicating increased microtubule stability (Kremer, Haystead and Macara, 2005, unpublished observations).

Regulation of membrane traffic SNAREs In addition to binding directly to lipids some septins have been shown to bind to proteins involved in membrane remodelling. SEPT2, 4 and 5 have been reported to interact with syntaxin1 and SEPT5 can also interact with syntaxin2 (Beites et al., 1999; Dent et al., 2002; Ihara et al., 2007). Syntaxins are a family of proteins which are involved in vesicle fusion. Syntaxin makes up a component of the soluble N-ethylmaleimide-sensitive fusion proteins attachment receptor (SNARE) complex, which is also composed of vesicle associated membrane protein (VAMP) and soluble N -ethylmaleimide-sensitive

SEPTIN FUNCTIONS

197

fusion proteins attachment protein-25 (SNAP25). This complex is extremely stable and is central to the process of membrane fusion. Indeed, septins were originally identified in the brain by generating monoclonal antibodies against proteins which were immunoprecipitated with synaptophysin, a synaptic vesicle protein (Honer, Hu and Davies, 1993; Caltagarone et al., 1998). The other proteins identified in this study were all components of the SNARE complex. The dissociation of the SNARE complex is necessary for the SNARE components to become available for subsequent membrane fusion events. This process requires two proteins: N -ethylmaleimide-sensitive fusion protein (NSF) and α-SNAP. SEPT5 competes with α-SNAP for binding with syntaxin, suggesting that SEPT5 has a regulatory role of controlling syntaxin availability for participation in forming the SNARE complex (Beites, Campbell and Trimble, 2005). SEPT5 phosphorylation of serine 17 by cyclin-dependent kinase 5 (CDK5), decreases the binding between SEPT5 and syntaxin. This suggests that CDK5, a kinase important in brain development as well as synaptic transmission, may play a role in regulating the inhibition of the SNARE complex through SEPT5 (Tanaguchi et al., 2007). Overexpression of SEPT5 has been shown to inhibit secretion in HIT cells (Beites et al., 1999), a hamster pancreatic islet β cell line that secretes insulin. There is no obvious morphological alteration in the central nervous system of SEPT5 knockout mice (Peng et al., 2002), but there is an increase in serotonin secretion in platelets from these mice (Dent et al., 2002). Platelets normally secrete molecules from storage granules. In fact, they contain several SNARE components which have been characterized as having defined roles in neuron exocytosis (Martinez and Ware, 2004). In platelets, SEPT5 localizes with α-granules and interacts with syntaxin 4 (Dent et al., 2002). Interestingly, SEPT5 is located in a chromosomal region that is deleted in patients with DiGeorge syndrome, or velocardiofacial syndrome, which has variable symptoms most commonly including a cleft palate and other characteristic facial features, heart defects, recurrent infections and aplasia or hypoplasia of the thymus and parathyroid (Kirby and Bockman, 1984). These patients also exhibit an increased probability of having schizophrenia (Murphy, Jones and Owen, 1999). A rat model of schizophrenia has been shown to have lower levels of SEPT5 in regions of the brain (Barr et al., 2004), suggesting a role for septins in the regulation of neurotransmission. Most of the 14 mammalian septins are highly expressed in the brain and SEPT2, 3, 5, 6 and 7 are associated with presynaptic or postsynaptic membranes (Kinoshita, Noda and Kinoshita, 2000; Walikonis et al., 2000; Satoh et al., 2002).

The exocyst complex Septins have also been shown to interact with two components of the exocyst complex, sec8 and Exo70 in rat brain and PC12 cells (Hsu et al., 1998; Vega and Hsu, 2003). This complex is made up of eight subunits and is involved in

198

CH08 THE FUNCTIONS OF SEPTINS IN MAMMALS

the transport of vesicles to specific areas of the plasma membrane (Munson and Novick, 2006; Wang and Hsu, 2006). It is thought that the exocyst complex may be divided into two subcomplexes, one associating with secretory vesicles and the other associated with specific regions of plasma membrane, and that these components associate when the vesicles reach the plasma membrane (Munson and Novick, 2006; Wang and Hsu, 2006). The exocyst complex is believed to function in association with microtubules and seems to play roles in both tethering secretory vesicles and vesicle docking at the plasma membrane. It has been speculated that the exocyst complex may be involved in transferring vesicles from microtubules to the actin at the cortex of the cell (Wang and Hsu, 2006). It does not seem that the exocyst complex has a role in fusion of synaptic vesicles at mature synapses as judged by the fact that exocyst component mutations in Drosophila do not have a synaptic vesicle fusion defect, but rather have defects in trafficking pathways (Andrews et al., 2002; Murthy et al., 2003). Furthermore, the mammalian exocyst complex appears to be similar to the Drosophila complex in that its expression decreases as synapses mature in primary hippocampal cultures (Hazuka et al., 1999), whereas septin expression remains high. Nevertheless, septins and components of the exocyst complex co-localize at growth cones and at synapse assembly domains as synapses mature (Hsu et al., 1998) where they may cooperate with some of the syntaxins to play a central role in the directed targeting of membrane to regions of active growth.

Diffusion barrier There is convincing evidence that yeast septins serve as a diffusion barrier (Finger, 2004; Douglas et al., 2005). During yeast cytokinesis the septin collar, which accumulates at the mother-bud neck, splits and at the same time becomes more fluid. This is believed to be a result of a rotation in septin filaments likely due to dephosphorylation (Vrabioiu and Mitchison, 2006). The two rings of septin filaments compartmentalize the cortex to concentrate the factors required for cytokinesis between them (Dobbelaere and Barral, 2004). Despite the fact that mammalian septins are required for the completion of mitosis (Surka, Tsang and Trimble, 2002; Kinoshita et al., 1997) and that septins accumulate at the cleavage furrow during cytokinesis, as yet there is no evidence to suggest that mammalian septins function in the same manner as the yeast septins do in cytokinesis. However, it appears that in mammalian systems there may be a restricted movement of membrane proteins with cytoplasmic domains and cytoplasmic leaflet lipids, but not those of proteins or lipids embedded in the outer leaflet or soluble GFP, across the contracted cleavage furrow (Schmidt and Nichols, 2004b). It is tempting to speculate that septins could serve as this diffusion barrier.

OTHER FUNCTIONS

199

SEPT7 has also been proposed as a candidate for a diffusion barrier during the development of dendritic spines on hippocampal neurons in culture (Tada et al., 2007). Thus far this proposal is based only on the subcellular localization of this septin, along with its binding partners SEPT5 and 11, forming an arc at the base of the developing dendritic spine (Tada et al., 2007; Xie et al., 2007). The annulus of mammalian spermatozoa appears to act as a diffusion barrier for an integral plasma membrane protein, CE9 (Cesario and Bartles, 1994). This suggests that SEPT4 (Ihara et al., 2005; Kissel et al., 2005), SEPT12 (Steels et al., 2007) and the other annulus septins may be playing a role in this context.

OTHER FUNCTIONS Membrane protein regulation The first mammalian septin identified, DIFF6, was found in a differential RNA expression screen from lymphoid cells selected for elevated levels of the cell adhesion molecule gp90 (Nottenburg, Gallatin and John, 1990). Since septins are associated with membranes and could have scaffolding and structural roles, it is possible that septins participate in some aspect of adhesion. Another membrane protein associated with septins is the astrocyte glutamate-aspartate transporter (GLAST). SEPT2 binds to the C-terminus of GLAST in a guanosine diphosphate (GDP) dependent manner and may regulate the internalization (or conversely prevent the externalization) of GLAST proteins from the cell surface, thereby potentially affecting glutamate uptake by astrocytes (Kinoshita et al., 2004).

Viral replication It is possible that septins may also serve as a scaffold for interactions necessary for viral reproduction. SEPT6 was shown to interact with the RNA-dependent RNA polymerase, NS5B from hepatitis C virus and the host cell hnRNP A1. SEPT6 was immunoprecipitated with both of these proteins and SEPT2 was also detected in this complex. Interestingly, depletion of SEPT6 appeared to significantly reduce viral replication (Kim et al., 2007). It was suggested that the septins may be instrumental in directing NS5B and hnRNP A1 to a membranous compartment of infected cells in which the virus replicates. In addition, SEPT4 has been shown to interact with the herpes virus protein, kaposin A (Lin et al., 2007) although the significance of this for viral reproduction is unknown as yet.

DNA repair Recent data from the Macara lab has suggested that the SEPT2/6/7 complex interacts with both SOCS7, a suppressor of cytokine signalling, and with Nck,

200

CH08 THE FUNCTIONS OF SEPTINS IN MAMMALS

an actin-associated adaptor protein. Following the depletion of septins by siRNA, both SOCS7 and Nck enter the nucleus. Interestingly, this nuclear localization is necessary for efficient activation of DNA damage checkpoints and growth arrest (Kremer, Adang and Macara, 2007), indicating that septins may participate in regulating DNA damage responses through scaffolding of effectors in the cytoplasm.

Apoptosis One isoform of SEPT4 has been implicated in an apoptotic response (see Chapter 13). This variation of SEPT4 lacks a functional P loop and the truncation of this septin variant, SEPT4 i2, previously known as apoptosis-related protein in TGF-β signalling pathway (ARTS), is thought to reveal a mitochondrial targeting sequence, as it is localized to the mitochondria (Larisch et al., 2000). SEPT4 i2 has been shown to bind to X-linked inhibitor of apoptosis protein (XIAP) resulting in the activation of caspase 3 and possibly caspase 9 in response to tumour grouth factor-β (TGF-β) (Larisch et al., 2000; Gottfried et al., 2004). This appears to leads to a release of Smac/Diablo and cytochrome c from the mitochondria resulting in apoptosis. The loss of SEPT4 i2 expression may have functional consequences as many patients with lymphoblastic leukaemia have lost SEPT4 i2 expression. Another SEPT4 isoform has also been reported to be localized to the mitochondria in neurons and these are associated with collapsin response mediator protein/collapsin response mediator protein-associated molecule proteins (CRMP/CRAM) (Takahashi et al., 2003). In contrast to SEPT4 i2, this SEPT4 variant does not lead to an apoptotic response and as yet the significance of the localization of this SEPT4 is not known. Nevertheless, CRMP/CRAM proteins have a role in axon growth during brain development, and overexpression of CRMP2 in primary hippocampal cultures results in supernumerary axons. This SEPT4 variant is up regulated during differentiation of P19 cells, a teratocarcinoma cell line which can be induced to differentiate into neuronal and glial cells. The functional significance of these mitochondrial SEPT4 variants in organisms is difficult to determine as SEPT4 knockout mice, which should be lacking both of these variants, show a mitochondrial phenotype only in sperm. Here the mitochondria are variable in size, when compared to other tissues within the knockout mouse or to wild type mouse sperm, and have defects in their internal membranes (Kissel et al., 2005). The roles played by SEPT4 in mitochondria in these knockout mice may be compensated by other septins as has been suggested for SEPT5 knockout mice (Peng et al., 2002). Differences between man and mouse also complicate interpretation.

COORDINATION OF CELLULAR EVENTS Most processes in the cell require the careful coordination of different reactions. Given the diversity of functions attributed to septins, it seems likely that septins

COORDINATION OF CELLULAR EVENTS

201

Figure 8.1 Representation of three processes in which septins are involved. The functions influencing these processes are boxed, and the septins involved are listed below the box with the cell types. The localization of the septins is indicated by an arrow. For details, with references, see text

may be involved in co-ordinating several functions in a given process. The following are three examples of processes in which septins may play a role in more than a single capacity. These processes, with the septins involved, and the cell types in which the evidence was collected are represented in Figure 8.1.

Polarity A characteristic cellular process which requires coordination of the actin and microtubules as well as membrane remodelling is polarity. Microtubules are thought to establish the proper position for cortical polarity by delivering proteins to the cortex that activate and help maintain the activated forms of Cdc42 and Rac1. These include the microtubule, plus end-associated protein CLIP-170’s association with IQ motif-containing GAP1 (IQGAP1) and the local deliveries of GEFs. Active Cdc42 is also considered to be the key upstream regulator of PAR6, one of the components of the highly conserved polarity complex (Macara, 2004). Both Cdc42 and Rac1 are involved in remodelling the actin cytoskeleton and extensive changes to the membrane (Rossman, Der and Sondek, 2005;

202

CH08 THE FUNCTIONS OF SEPTINS IN MAMMALS

Briggs and Sacks, 2003; Siegrist and Doe 2007). This remodelling requires delivery of lipid to specific regions of the plasma membrane for the development and maintenance of cellular polarity (Rodriguez-Boulan, Kreitzer and Musch, 2005). Indeed, the exocyst complex was first identified for its essential role in polarized exocytosis in yeast (Novick, Field and Schekman, 1980) and more recently the exocyst complex has been shown to participate in the polarity of epithelial cells in Drosophila. Interestingly, polarized exocytosis also involves recycling endosomes in Drosophila (Langevin et al., 2005). As yet, there is limited evidence to link septins to polarity, although they are likely candidates to participate given their association with the actin and microtubule cytoskeletons, with the exocyst complex, and their putative role in regulating SNARE-mediated exocytosis. One study that indicated septins may play a role in polarity showed that overexpression of the SEPT9 v4 transcript results in the formation of projections and increased motility in human breast adenocarcinoma (MCF7) cells (Chacko et al., 2005). Similarly, a GTPase mutant of SEPT2 overexpressed in PC12 cells showed an increased number of neurites growing from the cell body. In addition, SEPT2 exhibited a pattern of redistribution towards the growth cones of developing neurites upon nerve growth factor (NGF)-induced differentiation of these cells (Vega and Hsu, 2003). However, in differentiating N18 cells, a neuroblastoma cell line, SEPT2 was observed through the length of the neurite along with actin (Kinoshita et al., 1997). In developing neurons, an isoform of SEPT4 interacts with CRMP/CRAM which is believed to be part of one of the major signalling pathways in neuronal polarization (Takahashi et al., 2003; Yoshimura, Arimura and Kaibuchi, 2006). Conversely, depletion of septins in T lymphocytes did not result in any defects in cell polarity (M. Krummel, personal communications, Tooley et al., submitted).

Cytokinesis Another event in which actin, tubulin and membrane remodelling are all coordinated is cytokinesis, the first process in which septins were shown to be required. Cytokinesis is the final step in cell division in which the two cells separate from one another (for more detail one septin involvement in cytokinesis see Chapter 17). This process includes physical ingression of the plasma membrane mediated by actin and myosin, and this ingression is followed by disassembly of the ring formed by actin and myosin, concomitantly, there is disassembly of microtubules, one of the components of the midbody. There is a growing body of evidence that the addition of membrane, seemingly necessary for sealing the gap between the two cells, is a requirement for cell separation (Joo, Tsang and Trimble, 2005). Components of the SNARE complex localize to the midbody and are required for this process (Low et al., 2003). In addition, the exocyst complex also localizes to the midbody at this stage of the cell cycle (Wilson et al., 2005) and de

SUMMARY

203

novo synthesized PIP2 is believed to accumulate at the cleavage furrow (Emoto et al., 2005; Field et al., 2005). The septins also localize to the cleavage furrow during cytokinesis. Surprisingly however, SEPT2 co-localizes with the actomyosin ring, whereas SEPT9 co-localizes with the spindle microtubules at the midbody (Surka, Tsang and Trimble, 2002). SEPT2 and PIP2 were both found on vesicles which had accumulated after failed cytokinesis due to overexpression of a Rab35 dominant mutant, which is believed to be responsible for a fast recycling step (Kouranti et al., 2006), suggesting that PIP2, and not actin, is responsible for the localization of SEPT2 at the cleavage furrow. Thus septins, with their actin and tubulin co-localization as well as interaction with syntaxin, the exocyst complex, and direct interaction with PIP2, appear to be in an ideal physical position to coordinate and regulate the final steps in this process.

Phagocytosis Recent studies are emerging that link septins with phagocytosis. In one study, phagosomes were isolated from an epithelial cell line that had been exposed to latex beads coated with the internalization proteins InlA and InlB of Listeria monocytogenes. Early phagosomes were subjected to proteomic analysis and one protein prominently enriched on these phagosomes was SEPT9 (Pizarro-Cerd´a et al., 2002). Interestingly, we have also found that septins associate with phagosomes at early stages in phagocytosis in a variety of phagocytic cell types. Moreover, siRNA depletion of SEPT2 inhibits phagocytosis, indicating an essential role for septins in phagosome formation (Huang et al., 2008). Phagosome formation is known to require new membrane addition, as well as microtubule and actin remodelling. The hypothesis that septins are necessary for coordination of these processes has yet to be evaluated.

SUMMARY The structure and biochemical properties of mammalian septins strongly support the concept that they can assemble into large filament-like structures that associate with cellular membranes, raising the possibility that they could serve as physical diffusion barriers. Such barriers could regulate the local environment or mobility of membrane proteins and lipids. As such, they may also provide important structural stability to cells at critical sites, such as the sperm annulus. Their association with membrane trafficking machinery, both exocytic and endocytic, could play a role in a wide variety of signalling cascades. In addition, their hetero-oligomeric nature and combinatorial assembly properties would make them ideal candidates to serve as multimolecular scaffolds, co-coordinating the recruitment and interaction of many potential binding partners. Several of these interaction partners play important roles in the regulation of the actin and microtubule cytoskeletons, and septins

204

CH08 THE FUNCTIONS OF SEPTINS IN MAMMALS

may therefore function as a nexus between these two superstructures. Given their central participation in so many important cellular pathways, the many diverse functions that have been ascribed to mammalian septins are unsurprising. Full elucidation of septin physiology surely awaits the results of much future research.

REFERENCES Amir, S., Wang, R., Matzkin, H. et al. (2006) MSF-A interacts with hypoxia-inducible factor-1alpha and augments hypoxia-inducible factor transcriptional activation to affect tumorigenicity and angiogenesis. Cancer Research, 66, 856–66. Andrews, H.K., Zhang, Y.Q., Trotta, N. and Broadie, K. (2002) Drosophila sec10 is required from hormone secretion but not general exocytosis or neurotransmission. Traffic, 3, 906–21. Barr, A.M., Young, C.E., Sawada, K. et al. (2004) Abnormalities of presynaptic protein CDCrel-1 in striatum of rats reared in social isolation: relevance to neural connectivity in schizophrenia. The European Journal of Neuroscience, 20, 303–7. Beites, C.L., Campbell, K.A. and Trimble, W.S. (2005) The septin Sept5/CDCrel-1 competes with alpha-SNAP for binding to the SNARE complex. The Biochemical Journal , 385, 347–53. Beites, C.L., Xie, H., Bowser, R. and Trimble, W.S. (1999) The septin CDCrel-1 binds syntaxin and inhibits exocytosis. Nature Neuroscience, 2, 434–39. Briggs, M.W. and Sacks, D.B. (2003) IQGAP proteins are integral components of cytoskeletal regulation. EMBO Reports, 4, 571–74. Caltagarone, J., Rhodes, J., Honer, W.G. and Bowser, R. (1998) Location of a novel septin protein, hCDCrel-1 in neurons of human brain. Neuroreport, 9, 2907–12. Casamayor, A. and Snyder, M. (2003) Molecular dissection of a yeast septin: distinct domains are required for septin interaction, localization, and function. Molecular and Cellular Biology, 23, 2762–77. Cesario, M.M. and Bartles, J.R. (1994) Compartmentalization, processing and redistribution of the plasma membrane protein CE9 on rodent spermatozoa. Relationship of the annulus to domain boundaries in the plasma membrane of the tail. Journal of Cell Science, 107, 561–70. Chacko, A.D., Hyland, R.L., McDade, S.S. et al. (2005) SEPT9 4 expression induces morphological change, increased motility and disturbed polarity. The Journal of Pathology, 206, 458–65. Dent, J., Kato, K., Peng, X.R. et al. (2002) A prototypic platelet septin and its participation in secretion. Proceedings of the National Academy of Sciences of the United States of America, 99, 3064–69. Dobbelaere, J. and Barral, Y. (2004) Spatial coordination of cytokinetic events by compartmentalization of the cell cortex. Science, 305, 393–96. Douglas, L.M., Alvarez, F.J., McCreary, C. and Konopka, J.B. (2005) Septin function in yeast model systems and pathogenic fungi. Eukaryotic Cell , 4, 1503–12. Drees, B.L., Sundin, B., Brazeau, E. et al. (2001) A protein interaction map for cell polarity development. The Journal of Cell Biology, 154, 549–71. Emoto, K., Inadome, H., Kanaho, Y. et al. (2005) Local change in phospholipid composition at the cleavage furrow is essential for completion of cytokinesis. The Journal of Biological Chemistry, 280, 37901–907.

REFERENCES

205

Field, C.M., Al-Awar, O., Rosenblatt, J. et al. (1996) A purified Drosophila septin complex forms filaments and exhibits GTPase activity. The Journal of Cell Biology, 133, 605–16. Field, S.J., Madson, N., Kerr, M.L. et al. (2005) PtdIns(4,5)P2 functions at the cleavage furrow during cytokinesis. Current Biology, 9, 1407–12. Finger, F.P. (2004) Reining in cytokinesis with a septin corral. BioEssays, 27, 5–8. Frazier, J.A., Wong, M.L., Longtine, M.S. et al., (1998) Polymerization of purified yeast septins: evidence that organized filament arrays may not be required for septin function. Journal of Cell Biology, 143, 737–749. Fujishima, K., Kiyonar, H., Kurisu, J. et al. (2007) targeted disruption of Sept3, a heteromeric assembly partner of Sept5 and Sept7 in axons, has no effect on developing CNS neurons. Journal of Neurochemistry, 102, 77–92. Gottfried, Y., Rotem, A., Lotan, R. et al. (2004) The mitochondrial ARTS protein promotes apoptosis through targeting XIAP. The EMBO Journal , 23, 1627–35. Hartwell, L.H. (1971) Genetic control of the cell division cycle in yeast. IV. Genes controlling bud emergence and cytokinesis. Experimental Cell Research, 60, 265–76. Hazuka, C.D., Foletti, D.L., Hsu, S.C. et al. (1999) The sec6/8 complex is located at neurite outgrowth and axon synapse-assembly domains. The Journal of Neuroscience, 19, 1324–34. Hime, G.R., Brill, J.A. and Fuller, M.T. (1996) Assembly of ring canals in the male germ line from structural components of the contractile ring. Journal of Cell Science, 109, 2779–88. Honer, W.G., Hu, L. and Davies, P. (1993) Human synaptic proteins with a heterogeneous distribution in cerebellum and visual cortex. Brain Research, 609, 9–20. Hsu, S.C., Hazuka, C.D., Roth, R. et al. (1998) Subunit composition, protein interactions, and structures of the mammalian brain sec6/8 complex and septin filaments. Neuron, 20, 1111–22. Huang, Y.-W., Surka, M.C., Reynaud, D. et al. (2006) GTP binding and hydrolysis kinetics of human septin 2. The FEBS Journal , 273, 3248–60. Huang, Y.-W., Yan, M. Collins, R.F., et al. (2008) Mammalian septins are required for phagosome formation. Molecular Biology of the Cell , 19, 1717–26. Ihara, M., Kinoshita, A., Yamada, S. et al. (2005) Cortical organization by the septin cytoskeleton is essential for structural and mechanical integrity of mammalian spermatozoa. Developmental Cell , 8, 343–52. Ihara, M., Yamasaki, N., Hagiwara, A. et al. (2007) Sept4, a component of presynaptic scaffold and Lewy bodies, is required for the suppression of alpha-synuclein neurotoxicity. Neuron, 53, 519–33. Ito, H., Iwamoto, I., Morishita, R. et al. (2005) Possible role of Rho/Rhotekin signaling in mammalian septin organization. Oncogene, 24, 7064–72. Joberty, G., Perlungher, R.R., Sheffield, P.J. et al. (2001) Borg proteins control septin organization and are negatively regulated by Cdc42. Nature Cell Biology, 3, 861–66. John, C.M., Hite, R.K., Weirich, C.S. et al. (2007) The Caenorhabditis elegans septin complex is nonpolar. The EMBO Journal , 26, 3296–307. Joo, E., Surka, M.C. and Trimble, W.S. (2007) Mammalian Sept2 is required for scaffolding non-muscle myosin II and its kinases. Developmental Cell , 13, 677–90. Joo, E., Tsang, C.W. and Trimble, W.S. (2005) Septins: traffic control at the cytokinesis intersection. Traffic, 6, 626–34.

206

CH08 THE FUNCTIONS OF SEPTINS IN MAMMALS

Kim, C.S., Seol, S.K., Song, O.K. et al. (2007) An RNA-binding protein hnRNP A1 and a scaffold protein septin 6 facilitate HCV replication. Journal of Virology, 81, 3852–65. Kinoshita, M., Field, C.M., Coughlin, M.L. et al. (2002) Self- and actin-templated assembly of mammalian septins. Developmental Cell , 3, 791–802. Kinoshita, N., Kimura, K., Matsumoto, N. et al. (2004) Mammalian septin Sept2 modulates the activity of GLAST, a glutamate transporter in astrocytes. Genes to Cells, 9, 1–14. Kinoshita, M., Kumar, S., Mizoguchi, A. et al. (1997) Nedd5, a mammalian septin, is a novel cytoskeletal component interacting with actin-based structures. Genes and Development, 11, 1535–47. Kinoshita, A., Noda, M. and Kinoshita, M. (2000) Differential localization of septins in the mouse brain. The Journal of Comparative Neurology, 428, 223–39. Kirby, M.L. and Bockman, D.E. (1984) Neural crest and normal development: a new perspective. The Anatomical Record , 209, 1–6. Kissel, H., Georgescu, M.M., Larisch, S. et al. (2005) The Sept4 septin locus is required for sperm terminal differentiation in mice. Developmental Cell , 8, 353–64. Kouranti, I., Sachse, M., Arouche, N. et al. (2006) Rab35 regulates an endocytic recycling pathway essential for the terminal steps of cytokinesis. Current Biology, 16, 1719–25. Kremer, B.E., Adang, L.A. and Macara, I.G. (2007) Septins regulate actin organization and cell cycle arrest through SOCS7-mediated nuclear accumulation of NCK. Cell , In press. Kremer, B.E., Haystead, T. and Macara, I.G. (2005) Mammalian septins regulate microtubule stability through interaction with the microtubule-binding protein MAP4. Molecular Biology of the Cell , 16, 4648–59. Kumar, S., Tomooka, Y., Noda, M. (1992) Identification of a set of genes with developmentally down-regulated expression in the mouse brain. Biochemical and Biophysical Research Communications, 185, 1155–61. Langevin, J., Morgan, M.J., Rosse, C. et al. (2005) Drosophila exocyst components Sec5, Sec6 and Sec15 regulate DE-cadherin trafficking from recycling endosomes to the plasma membrane. Developmental Cell , 9, 355–76. Larisch, S., Yi, Y., Lotan, R. et al. (2000) A novel mitochondrial septin-like protein, ARTS, mediates apoptosis dependent on its P-loop motif. Nature Cell Biology, 2, 915–21. Lin, C.W., Tu, P.F., Hsiao, N.W. et al. (2007) Identification of a novel septin 4 protein binding to human herpesvirus 8 kaposin A protein using a phage display cDNA library. Journal of Virological Methods, 143, 65–72. Low, S.H., Li, X., Miura, M. et al. (2003) Syntaxin 2 and endobrevin are required for the terminal step of cytokinesis in mammalian cells. Developmental Cell , 4, 753–59. Macara, I.G. (2004) Parsing the polarity code. Nature Reviews in Molecular Cell Biology, 5, 220–31. Martinez, C., Corral, J., Dent, J.A. et al. (2006) Platelet septin complexes form rings and associate with the microtubular network. Journal of Thrombosis and Haemostasis, 4, 1388–95.

REFERENCES

207

Martinez, C. and Ware, J. (2004) Mammalian Septin Function in Hemostasis and Beyond , Society for Experimental Biology and Medicine, pp. 1111–19. Matsumura, F. (2005) Regulation of myosin II during cytokinesis in higher eukaryotes. Trends in Cell Biology, 15, 371–77. McIlhatton, M.A., Burrows, J.F., Donaghy, P.G. et al. (2001) Genomic organization, complex splicing pattern and expression of a human septin gene on chromosome 17q25.3. Oncogene, 20, 5930–39. Mendoza, M., Hyman, A.A. and Glotzer, M. (2002) GTP binding induces filament assembly of a reconstituted septin. Current Biology, 12, 1858–63. Munson, M. and Novick, P. (2006) The exocyst defrocked, a frame work of rods revealed. Nature Structural and Molecular Biology, 13, 577–81. Murphy, K.C., Jones, L.A. and Owen, M.J. (1999) High rates of schizophrenia in adults with velo-cardio-facial syndrome. Archives of General Psychiatry, 56, 940–45. Murthy, M., Garza, D., Scheller, R.H. and Schwarz, T.L. (2003) Mutations in the exocyst component Sec5 disrupt neuronal membrane traffic, but neurotransmitter release persists. Neuron, 37, 433–47. Nagata, K., Asano, T., Nozawa, Y. and Inagaki, M. (2004) Biochemical and cell biological analyses of a mammalian septin complex, Sept7/9b/11. The Journal of Biological Chemistry, 279 (53), 55895–904. Nagata, K. and Inagaki, M. (2005) Cytoskeletal modification of Rho guanine nucleotide exchange factor activity: Identification of a Rho guanine nucleotide exchange factor as a binding partner for Sept9b, a mammalian septin. Oncogene, 24, 65–76. Nagata, K., Kawajiri, A., Matsui, S. et al. (2003) Filament formation of MSF-A, a mammalian septin, in human mammary epithelial cells, depends on interactions with microtubules. The Journal of Biological Chemistry, 278, 18538–43. Neufield, T.P. and Rubin, G.M. (1994) The Drosophila peanut gene is required for cytokinesis and encodes a protein similar to yeast putative bud neck filament proteins. Cell , 77, 371–79. Nhyuen, T.Q., Sawa, H., Okano, H. and White, J.G. (2000) The C. elegans septin genes, unc-59 and unc-61 are required for normal morphogenensis but have no essential function in embryogenesis. Journal of Cell Science, 113, 3825–37. Nottenberg, C., Gallatin, W.M. and John, T.S. (1990) Lymphocyte HEV adhesion variants differ in the expression of multiple gene sequences. Gene, 95, 279–84. Novick, R., Field, C. and Schekman, R. (1980) Identification of 23 complementation groups required for post-translational events in the yeast secretory pathway. Cell , 21, 205–15. Ono, R., Ihara, M., Nakajima, H. et al. (2005) Disruption of Sept6, a fusion partner gene of Mixed Lineage Leukemia (MLL), does not affect the ontogeny, leukemogenesis induced by MLL-SEPT6, or the phenotype induced by the loss of September 4. Molecular and Cellular Biology, 25, 10965–78. Pan, F., Malmberg, R.L. and Momany, M. (2007) Analysis of septins across kingdoms reveals orthology and new motifs. BMC Evolutionary Biology, 7, 103–19. Peng, X.R., Jia, Z., Zhang, Y. et al. (2002) The septin CDCrel-1 is dispensable for normal development and neurotransmitter release. Molecular and Cellular Biology, 22, 378–87.

208

CH08 THE FUNCTIONS OF SEPTINS IN MAMMALS

Peterson, E.A., Kalikin, L.M., Steels, J.D. et al. (2007) Characterization of a testisspecific septin, SEPT14, a novel testis-specific septin. Mammalian Genome, 18, 796–807. Pizarro-Cerd´a, J., Jonqui`eres, R., Gouin, E. et al. (2002) Distinct protein patterns associated with Listeria monocytogenes InlA- or InlB- phagosomes. Cellular Microbiology, 4, 101–15. Richman, T.J., Sawyer, M.M. and Johnson, D.I. (1999) The Cdc42p GTPase is involved in a G2/M morphogenetic checkpoint regulating the apical-isotropic switch and nuclear division in yeast. The Journal of Biological Chemistry, 274, 16861–70. Rodriguez-Boulan, E., Kreitzer, G. and Musch, A. (2005) Organization of vesicular trafficking in epithelia. Nature Reviews in Molecular Cell Biology, 5, 233–47. Rodriguez-Escudero, I., Roelants, F.M., Thorner, J. et al. (2005) Reconstitution of the mammalian PI3K/PTEN/Akt pathway in yeast. The Biochemical Journal , 390, 613–23. Rossman, K.L., Der, C.J. and Sondek, J. (2005) GEF means go: turning on RHO GTPases with guanine nucleotide-exchange factors. Nature Reviews in Molecular Cell Biology, 6, 167–80. Satoh, K., Takeuchi, M., Oda, Y. et al. (2002) Identification of activity regulated proteins in the postsynaptic density fraction. Genes and Cells, 7, 187–97. Schmidt, K. and Nichols, B.J. (2004a) Functional interdependence between septin and actin cytoskeleton. BMC Cell Biology, 5, 43–55. Schmidt, K. and Nichols, B.J. (2004b) A barrier to lateral diffusion in the cleavage furrow of dividing mammalian cells. Current Biology, 14, 1002–6. Siegrist, S.E. and Doe, C.Q. (2007) Microtubule-induced cortical cell polarity. Genes and Development, 21, 483–96. Sheffield, P.J., Oliver, C.J., Kremer, B.E. et al. (2003) Borg/septin interactions and the assembly of mammalian septin heterodimers, trimers, and filaments. The Journal of Biological Chemistry, 278, 3483–88. Shih, H.P., Hales, K.G., Pringle, J.R. and Peifer, M. (2002) Identification of septininteracting proteins and characterization of the Smt3/SUMO-conjugation system in Drosophila. Journal of Cell Science, 15, 1259–71. Sirajuddin, M., Farkasovsky, M., Hauer, F. et al. (2007) Structural insight into filament formation by mammalian septins. Nature, 449, 311–15 Spiliotis, E.T., Kinoshita, M. and Nelson, W.J. (2005) A mitotic septin scaffold required from mammalian chromosome congression and segregation. Science, 307, 1781–1785. Steels, J.D., Estey, M.P., Froese, C.D. et al. (2007) Septin 12 is a component of the sperm annulus. Cell Motility and the Cytoskeleton, 64, 794–807. Surka, M.C., Tsang, C.W. and Trimble, W.S. (2002) The mammalian septin MSF localizes with microtubules and is required for completion of cytokinesis. Molecular Biology of the Cell , 13, 3532–45. Suzuki, C., Daigo, Y., Ishikawa, N. et al. (2005) ANLN plays a critical role in human lung carcinogenesis through the activation of RHOA and by the involvement in the phsophoinsitide-3-kinase/AKT pathway. Cancer Research, 65, 11314–25.

REFERENCES

209

Tada, T., Simonetta, A., Batterton, M. et al. (2007) Role of septin cytoskeleton in spine morphogenesis and dendrite development in neurons. Current Biology, 17, 1752–58. Takahashi, S., Inatome, R., Yamamura, H. and Yanagia, S. (2003) Isolation and expression of a novel mitochondrial septin that interacts with CRMP/CRAM in the developing neurones. Genes and Cells, 8, 81–93. Tanaguchi, M., Taoka, M., Itakura, M. et al. (2007) Phosphorylation of adult type Sept5 (CDCrel-1) by cyclin-dependent kinase 5 inhibits interaction with sytaxin-1. The Journal of Biological Chemistry, 282, 7869–76. Tjandra, H., Compton, J. and Kellogg, D. (1998) Control of mitotic events by the Cdc42 GTPase, the Clb2 cyclin and a member of the PAK kinase family. Current Biology, 8, 991–1000. Ulrich, H.D. (2005) Mutual interactions between the SUMO and the ubiquitin systems: a plea of no contest. Trends in Cell Biology, 15, 525–32. Vega, I.E. and Hsu, S.C. (2003) The septin protein Nedd5 associates with both the exocyst complex and microtubules and disruption of its GTPase activity promotes aberrant neurite sprouting in PC12 cells. Neuroreport, 14, 31–37. Versele, M. and Thorner, J. (2005) Some assembly required: yeast septins provide the instruction manual. Trends in Cell Biology, 15, 414–24. Vrabioiu, A.M. and Mitchison, T.J. (2006) Structural insights into yeast septin organization from polarized fluorescence microscopy. Nature, 443, 466–69. Walikonis, R.S., Jensen, O.N., Mann, M. et al. (2000) Identification of proteins in the postsynaptic density fraction by mass spectrometry. The Journal of Neuroscience, 20, 4069–80. Wang, S. and Hsu, S.C. (2006) The molecular mechanisms of the mammalian exocyst complex in exocytosis. Biochemical Society Transactions, 34, 687–90. Wilson, G.M., Fielding, A.B., Simon, G.C. et al. (2005) The FIP3-Rab11 protein complex regulates recycling endosome targeting to the cleavage furrow during late cytokinesis. Molecular Biology of the Cell , 16, 849–60. Xie, H., Surka, M., Howard, J. and Trimble, W.S. (1999) Characterization of the mammalian septin H5: distinct patterns of cytoskeletal and membrane association from other septin proteins. Cell Motility and the Cytoskeleton, 43, 52–62. Xie, Y., Vessey, J.P., Konecna, A. et al. (2007) The GTP-binding protein septin 7 is critical for dendrite branching and dendritic-spine morphology. Current Biology, 17, 1746–51. Yoshimura, T., Arimura, N. and Kaibuchi, K. (2006) Signaling networks in neuronal polarization. The Journal of Neuroscience, 26, 10626–30. Zhang, J., Kong, C., Xie, H. et al. (1999) Phosphatidylinositol polyphosphate binding to the mammalian septin H5 is modulated by GTP. Current Biology, 9, 1458–67.

9 Septin-interacting proteins in mammals Brandon E. Kremer and Ian G. Macara Department of Microbiology, University of Virginia School of Medicine, Charlottesville, VA, USA

INTRODUCTION Septins can associate with guanine nucleotides, with phospholipids and with other proteins. The association with guanine nucleotides appears to play a structural role rather than functioning as a molecular switch, and might facilitate the formation of oligomers or filaments, at least in the case of SEPT2. Phospholipid binding may be important in budding yeast to organize the septin ring at the budneck, and to enable the ring to act as a diffusion barrier, but it is not yet clear what functional role it plays in mammals. Finally, septins bind to other proteins, including other septins. For budding yeast, as described in Chapter 4, septins segregate the bud from the mother cell, and different proteins associate asymmetrically with the septin ring. Again, no similar organization has yet been reported in mammalian cells. Nonetheless, a number of proteins have been identified that do bind to mammalian septins, and functions for several of these interactions have been elucidated. These will be discussed in the following sections and functions for several of these interactions have been elucidated (Table 9.1). Yet, given the assumption that septins act as scaffolds, it seems likely that multiple other binding partners for septins have yet to be found.

SEPTINS INTERACT WITH OTHER SEPTINS To date, some of the best-characterized septin-binding proteins are other members of the septin family. As described elsewhere for yeast septins, the mammalian The Septins Edited by Peter A. Hall, S.E. Hilary Russell and John R. Pringle © 2008 John Wiley & Sons, Ltd. ISBN: 978-0-470-51969-1

CH09 SEPTIN-INTERACTING PROTEINS IN MAMMALS

212 Table 9.1

Septin-interacting proteins

Septin

Binding partner

SEPT1

Aurora B kinase SEPT2/6/7 Mts1 SEPT2 SEPT2

GLAST Anillin

SEPT2

MAP4

SEPT2

P85 (PI3K subunit) PKG-I

SEPT3

SEPT4 XIAP (ARTS) SEPT4 α-synuclein SEPT4 SEPT5

Kaposin A (KSHV) Parkin

SEPT5

Syntaxin

Reference

Role

(Qi et al., 2005)

Chromosome segregation/cytokinesis?

(Koshelev, Kiselev and Unclear Georgiev, 2003) (Kinoshita et al., 2004) Regulates glutamate uptake in astrocytes (Kinoshita et al., 2002) May mediate actin-septin binding; may play role in cytokinesis (Kremer, Haystead and Destabilizes microtubules; important for Macara, 2005) cell division (Garcia et al., 2006) Recruits SEPT2 to cleavage furrow (Xue et al., 2004a; Xue Phosphorylates SEPT3; ay regulate et al., 2004b) SEPT3 localization (Gottfried et al., 2004) Antagonizes anti-apoptotic role of XIAP (Ihara et al., 2003) (Lin et al., 2007) (Choi et al., 2003; Zhang et al., 2000)

Inhibits synuclein aggregation and neurotoxicity Inhibits apoptosis in KSHV-infected cells Downregulates SEPT5 in dopaminergic neurons; effective in Juvenile Parkinson’s Disease Inhibits exocytosis by blocking SNAP-syntaxin binding Regulates septin organization

(Beites et al., 2005; Beites et al., 1999) SEPT6/7 Binder of Rho (Joberty et al., 2001; GTpases Sheffield et al., 2003) (BORGs) SEPT6 SOCS7 (Kremer et al., CELL) Regulates SOCS7 localization; complex important for actin arrangement and DNA damage response SEPT6 hnRNP A1 (Kim et al., 2007) Replication of Heptatitis C Virus NS5b (HCV) SEPT8 PFTAIRE1 (Yang et al., 2002) Unclear SEPT9 Rhotekin (Ito et al., 2005) May regulate septin organization SEPT9 SA-RhoGEF (Nagata and Inagaki, Activation of Rho small GTPase 2005) SEPT9 HIF-1α (Amir et al., 2006) Inhibits HIF degradation and enhances HIF-induced transcripts SEPT9 Internalin B (Pizarro-Cerda et al., Invasion/migration of Listeria (Listeria) 2002)

SEPTINS INTERACT WITH OTHER SEPTINS

213

form heteromeric complexes. However, there are more mammalian septins from which to assemble complexes, and numerous splice variants, so the combinatorial possibilities are much larger. Four distinct septin–septin complexes have been identified to date: SEPT2/6/7 (Joberty et al., 2001; Low and Macara, 2006; Sheffield et al., 2003), SEPT 7/9b/111 Nagata et al., (2004), SEPT4/5/8 (Blaser et al., 2004; Blaser et al., 2002; Martinez et al., 2004) and SEPT3/5/11 (Fujishima et al., 2007). In addition, SEPT 9 can bind to the SEPT2/6/7 complex in a non-stoichiometric manner (Joberty et al., 2001). Other septin–septin interactions have also been identified, including SEPT5–SEPT11 (Blaser et al., 2006) and SEPT2–SEPT9 (Surka, Tsang and Trimble, 2002). As has been describe for the yeast proteins, mammalian septins can self-assemble into filaments in vitro and in vivo (Kinoshita, 2003; Kinoshita et al., 2002; Kinoshita et al., 1997; Lindsey and Momany, 2006; Low and Macara, 2006; Sheffield et al., 2003). Importantly, the structure of the SEPT 2/6/7 complex has recently been determined, and reveals the mechanism of septin–septin interactions (Sirajuddin et al., 2007). Unexpectedly, the septins associate through their G-domains, rather than via the C-terminal coiled-coil regions. SEPT2 can homodimerize through residues adjacent to the nucleotide binding site, but in SEPT 2/6/7 hexamers it dimerizes through residues on the opposite face where α-helices contributed by the N- and C-terminal sequences in the guanosine triphosphate (GTP)-binding domain are situated (Sirajuddin et al., 2007). The structural data reveal a surprising flexibility in the modes of septin–septin interaction, but are not yet of sufficient resolution to reveal the rules that govern these interactions (see Chapter 3). The identification of septins as a major class of septin-binding proteins raises a number of questions. First, are different complexes found within individual cells, and, if so, are different complexes found in different subcellular regions? Do these different complexes perform different functions? The septin family is composed of at least 14 individual gene products, and these can be classified into 4 subfamilies based on sequence homology: the ‘SEPT2 Group’, composed of septins 1, 2, 4 and 5; the ‘SEPT3 group’, composed of septins 3, 9 and 12; the ‘SEPT6 group’, composed of septins 6, 8, 10, 11 and 14; and the ‘SEPT7 Group’, which contains SEPT7 and SEPT13 (Hall et al., 2005; Kinoshita, 2003). The complexes described above typically include one member each from the SEPT2, SEPT6 and SEPT7 group, and some data suggest that different members within the SEPT2 group or SEPT6 group can substitute for one another within an oligomer (Kinoshita, 2003). Whether these different septin–septin interactions are important, however, remains to be determined. Further study of the interaction between individual septin proteins will likely provide insight into the processes mediated by these proteins. 1 SEPT9b is a high molecular weight isoform of SEPT9, and is also designated SEPT9 i3, the product of the SEPT9 v3 transcript. The three high molecular weight isoforms of SEPT9 are almost identical differing only by 7, 18 and 25 unique N terminal residues.

214

CH09 SEPTIN-INTERACTING PROTEINS IN MAMMALS

Second, is the biologically relevant unit the septin monomer, or do the proteins function only in heteromeric septin complexes? In vitro and in vivo data exist to support the latter idea. Septin proteins tend to be unstable unless expressed together in vitro, at which point they form oligomeric complexes that are extremely stable (Joberty et al., 2001; Sheffield et al., 2003; Sirajuddin et al., 2007). Furthermore, depletion by RNA interference of a single septin in the SEPT2/6/7 complex can drive the loss of the other septin proteins within the complex (Kinoshita et al., 2002; Kremer, Haystead and Macara, 2005; Kremer, Adang and Macara, 2007). Nonetheless, it remains possible that the oligomers represent a storage form of septins, and that in some circumstances they act alone. Third, if septins usually function as a heteromeric complex, what is the role of individual subunits within the complex? Do most septin-binding partners recognize only the oligomeric form of the proteins? If not, why do the septins form oligomers at all? As described below, some proteins, such as Borgs, appear to bind only to a septin complex while others, such as anillin or MAP4, bind to an isolated septin. Possibly, septin heteromers recruit different proteins in order to bring them into close proximity, either to drive an enzymatic reaction or to assemble a protein complex. Further work will be required to test this idea.

SEPTINS INTERACT WITH ACTIN-ASSOCIATED PROTEINS One of the most striking aspects of septin distribution is the frequent colocalization with actin microfilaments (Kinoshita et al., 1997). SEPT2 (Joberty et al., 2001; Kinoshita et al., 2002; Kinoshita et al., 1997), SEPT 6 (Joberty et al., 2001), SEPT7 (Joberty et al., 2001), SEPT9 (Nagata et al., 2004; Surka, Tsang and Trimble, 2002) and SEPT11 (Hanai et al., 2004; Nagata et al., 2004) each co-localize at least partially with actin stress fibres during interphase. The pattern of co-localization varies with cell type, for reasons that are not apparent. For example, in HeLa cells the septins tend to associate primarily with stress fibres underneath the nucleus, which suggests that there is something unique about the actin filaments in this region that can recruit septins, but what confers this ability is not known. Two groups have demonstrated that septin filaments ‘template’ onto actin stress fibres (Kinoshita et al., 2002; Kinoshita et al., 1997; Schmidt and Nichols, 2004). If actin stress fibres are eliminated by drug treatment, the septin filaments reorganize into cytoplasmic rings with a diameter of 600 nm (Kinoshita et al., 2002). However, normal septin arrangement is re-established following wash-out of the actin-inhibiting drug, and septin reorganization is concomitant with stress fibre reassembly (Kinoshita et al., 2002; Kinoshita et al., 1997). In addition, there appears to be some level of reciprocal regulation of septins and actin, as septin depletion leads to disruption of the actin cytoskeleton (Kinoshita et al., 2002; Kremer, Adang and Macara, 2007). The regulation of the actin cytoskeleton by mammalian septins will be discussed further in Chapter 10.

SEPTINS INTERACT WITH ACTIN-ASSOCIATED PROTEINS

215

How do septins interact with the actin cytoskeleton? To date, this question has not been fully answered. Recombinant septins alone do not co-sediment with purified actin stress fibres, nor do they bind to actin bundles in vitro. To identify actin-associated proteins that might mediate the septin–actin interaction, Kinoshita et al. co-incubated recombinant SEPT2/6/7 complex with phalloidin-stabilized actin and several actin bundling proteins. They found that co-incubation with anillin caused the septin complex to co-localize with actin stress fibres, while α-actinin, fascin and filamin, which also bind to actin, had no effect (Kinoshita et al., 2002). Anillin is a nuclear protein that, during cell division, localizes to the cleavage furrow (Field and Alberts, 1995) and is necessary for the localization of myosin II and RhoA to the cleavage furrow (Field and Alberts, 1995; Zhao and Fang, 2005). Importantly, in the absence of anillin, these proteins mislocalize, resulting in a failure of cytokinesis (Field et al., 2005; Zhao and Fang, 2005). Like septins, anillin is a conserved protein, with homologues in Schizosaccharomyces pombe (Tasto et al., 2003), Caenorhabditis elegans (Maddox et al., 2005), Drosophila melanogaster (Field and Alberts, 1995) and humans (Oegema et al., 2000). In addition to binding to the septin complex in vitro (Kinoshita et al., 2002), anillin co-localizes to the cleavage furrow with SEPT7 (Oegema et al., 2000). In Drosophila, anillin is required for septin recruitment to the cleavage furrow (Field et al., 2005), though this has yet to be demonstrated in mammalian cells. These data indicate that anillin binds to actomyosin as well as septins at the cleavage furrow of dividing cells. However, there are no published data on the direct interaction of anillin and septins. In addition, anillin is a nuclear protein, and only enters the cytoplasm following nuclear envelope breakdown during cell division (Field and Alberts, 1995; Kinoshita et al., 2002). Thus, while anillin acts to recruit septins to actin during cell division, it is unlikely to do so during interphase. Actin assembly and dynamics are controlled by a large number of regulators. One important class of such factors is the small GTPases of the Rho family (For a general review, see (Etienne-Manneville and Hall, 2002; Jaffe and Hall, 2005; Schmidt and Hall, 2002)). Rho and its homologues act as molecular switches that can activate downstream effector molecules in their GTP-bound states. The intrinsic GTP hydrolysis and exchange rates are very low, and the switch mechanism is operated by factors that induce hydrolysis (guanosine triphosphatase activating proteins, GAPs), or catalyse nucleotide exchange (guanosine nucleotide exchange factors, GEFs). The GAPs and GEFs are subject to regulation by many other factors. In most cases they are subject to auto-inhibition, which is relieved by a signalling event such as phosphorylation or protein binding. In a two-hybrid screen for SEPT9-interacting proteins, Nagata and Inagaki identified a Rho exchange factor they called SA-RhoGEF (for septin associated RhoGEF, also named ARHGEF18, and p114-RhoGEF), (Nagata and Inagaki, 2005). SA-RhoGEF is broadly expressed and activates Rho both in vitro and in vivo (Nagata and Inagaki, 2005). SEPT9 i3 binds directly to SA-RhoGEF, and inhibits activation of endogenous Rho (Nagata and Inagaki, 2005). In addition, over-expression of SA-RhoGEF disrupted SEPT9 i3 filaments (Nagata and

216

CH09 SEPTIN-INTERACTING PROTEINS IN MAMMALS

Inagaki, 2005), but this effect might be an indirect result of actin reorganization caused by Rho activation. While these discoveries provide a potential link between septins and Rho-dependent regulation of the actin cytoskeleton, it is unclear what the biological significance or function performed by endogenous SA-RhoGEF might be. This GEF also binds and is activated by heterotrimeric G-protein βγ subunits (Niu et al., 2003), and it remains possible that SEPT9b inhibits the GEF indirectly, by blocking Gβγ binding. In addition to its interaction with SA-RhoGEF, SEPT9 v2 also interacts with the Rho effector rhotekin (Ito et al., 2005). Rhotekin was first discovered as a binding partner of activated Rho (Reid et al., 1996). While the biological role of rhotekin remains unclear, it can bind to cell polarity domains, inhibit apoptosis, and enhance signalling in response to serum treatment of cells (Ito et al., 2006; Liu et al., 2004; Reynaud et al., 2000). Ito et al. demonstrated that rhotekin co-localizes along actin stress fibres with SEPT9 i2; it also co-purifies in a complex with septins, and over-expression of the septin-binding fragment of rhotekin induces septin, but not actin, reorganization (Ito et al., 2005). Much work remains to be done, however, to determine the physiological role of both rhotekin and the septin–rhotekin interaction. In addition to their association and regulation by rhotekin, mammalian septins are under the control of another Rho-family GTPase, Cdc42. While Rho is thought to play a role in stress fibre formation, Cdc42 is critical for structures at the periphery of cells called filopodia (Nobes and Hall, 1995). Macara and colleagues found a novel class of Cdc42 effectors, termed Borgs, that bind to the GTP-bound form of Cdc42 (Joberty, Perlungher and Macara, 1999). These proteins also interact with the SEPT2/6/7 complex at the interface between Septs 6 and 7, and can co-purify a stoichiometric complex of the SEPT2/6/7 trimer (Joberty et al., 2001; Sheffield et al., 2003). Over-expression of Borg3 induces a dramatic shift in septin architecture, as the filaments collapse into a tight mass at the centre of the cell, co-localized with the Borg protein (Joberty et al., 2001). This process is negatively regulated by the interaction between Cdc42-GTP and Borg3, which sequesters the Borg protein away from the septin complex, allowing for septin filaments to form along the actin cytoskeleton (Joberty et al., 2001). However, while this provides one of the few known mechanisms of regulation of septin filaments, the physiological significance of Borg–septin interactions remains to be determined. Recently, we found that the SEPT2/6/7 complex co-precipitates with suppressor of cytokine signaling-7 (SOCS7) (Kremer, Adang and Macara, 2007), one of a family of SOCS proteins. SOCS proteins are traditionally associated with attenuation of cytokine signalling, and SOCS7 has been shown to limit insulin stimulation (Banks et al., 2005; Howard and Flier, 2006). SOCS7 has also been linked to the actin cytoskeleton, as it binds to the actin-associated adapter protein NCK and to the focal adhesion protein vinexin (Martens et al., 2004; Matuoka et al., 1997). SOCS7 is a nucleocytoplasmic shuttling protein, and it carries NCK into the nucleus in a process regulated by septins (Kremer, Adang and Macara, 2007).

SEPTIN INTERACTION WITH MICROTUBULES

217

Septin depletion by siRNA leads to the nuclear accumulation of both SOCS7 and NCK, leading to disruption of the actin cytoskeleton and a loss of cell polarity (Kremer, Adang and Macara, 2007). Unexpectedly, this process is involved in the DNA damage response. The nuclear accumulation of NCK is essential for efficient activation of the checkpoint kinase Chk2 and the tumour suppressor p53, as well as for cell cycle arrest and growth arrest (Kremer, Adang and Macara, 2007). This process is also regulated by septins. Septins, therefore regulate the actin cytoskeleton via a SOCS7-NCK pathway, as well as regulate the DNA damage response by the same mechanism.

SEPTIN INTERACTION WITH MICROTUBULES, AND THE MICROTUBULE-ASSOCIATED PROTEIN MAP4 In addition to their physical interaction with the actin cytoskeleton, various septin proteins, including SEPT9 and SEPT11, partially co-localize with microtubules (MTs) in interphase cells (Nagata et al., 2003; Spiliotis, Kinoshita and Nelson, 2005; Surka et al., 2002). SEPT2 is not associated with MTs in interphase, but does decorate spindle MTs in mitosis (Spiliotis, Kinoshita and Nelson, 2005), and localizes to the midbody with MTs in cytokinesis (Figure 9.1). However, the molecular basis for none of these interactions is known. There is no evidence to date that septins can associate directly with tubulin, so we assume that the MT interaction is indirect. However, only one MT binding protein, MAP4, has so far been shown to bind septins, and this association is mutually exclusive with MT binding. Macara and colleagues reported that cytoplasmic SEPT2 interacts with the microtubule-associated protein MAP4 (Kremer et al., 2005). MAP4 is a type

Figure 9.1 Immunostaining of SEPT2 in HeLa cells at different stages of the cell cycle. Cells were fixed in paraformaldehyde, permeabilized with Triton X-100 and stained with a rabbit polyclonal antibody against SEPT2

218

CH09 SEPTIN-INTERACTING PROTEINS IN MAMMALS

II MAP that binds tubulin, nucleates microtubule formation and bundles MTs into larger assemblies (Andersen, 2000; Mandelkow and Mandelkow, 1995). This MT–MAP4 interaction stabilizes the MTs (Andersen, 2000; Drewes, Ebneth and Mandelkow, 1998; Ebneth et al., 1999), which leads to an increase in acetylated tubulin (Kremer et al., 2005), a known marker of stabilized microtubules (Piperno, LeDizet and Chang, 1987). The kinetics of several MT motor proteins, and the transport of mitochondria along the MTs, are inhibited by MAP4 decoration of MTs (Ebneth et al., 1998; Mandelkow et al., 2004; Trinczek et al., 1999), suggesting that MAP4–MT binding may play a role in the regulation of directed transport of vesicles or organelles. Furthermore, MAP4 has also been implicated in the regulation of cell migration, cell adhesion and cell division (Ebneth et al., 1999; Mandelkow et al., 2004; Ookata et al., 1995), though the mechanisms by which MAP4 affects these processes have yet to be determined. MAP4 interacts directly with SEPT2, and specific co-precipitation can be detected from whole cell lysates as well as recombinant, bacterially expressed protein (Kremer et al., 2005). The binding of SEPT2 to MAP4 inhibits the ability of MAP4 to interact with MTs, causing destabilization of MTs throughout the cell (Kremer et al., 2005). Septin protein also inhibits the ability of MAP4 to nucleate new MTs from tubulin and bundle the resulting MTs in vitro (Kremer et al., 2005). Furthermore, this process is involved in the regulation of cell division, as MAP4 depletion can reverse the multinucleation phenotype observed following depletion of septins by siRNA (Kremer et al., 2005). These data suggest a novel mechanism for MAP4 regulation, which may be important for cellular division. To date, phosphorylation of MAP4 by members of the Par1/MARK (Microtubule Affinity-regulating Kinase) family of kinases, and by cdc2, have been described to inhibit MAP4–MT interaction (Chang et al., 2001; Drewes et al., 1997; Ebneth et al., 1999; Ookata et al., 1995; Tombes, Peloquin and Borisy, 1991; Trinczek et al., 2004). These phosphorylations, and the subsequent MT destabilization, are the best-characterized method of MAP4 regulation. Many questions, however, remain to be answered. Septins co-localize with MTs, but septin–MAP4 interaction appears to block MAP4–MT binding. How, then, do septins template along MTs? Furthermore, how is the septin–MAP4 interaction regulated, both spatially and temporally? What processes are affected by septin inhibition of MAP4? Do other septins interact with other type II MAPs, including MAP2 and tau, and what roles might these interactions regulate?

SEPTIN-BINDING PARTNERS AND REGULATION OF SECRETION In addition to the ubiquitously expressed septins that appear to be involved in regulation of actin and MTs, a subset of septins is expressed almost exclusively

SEPTIN-BINDING PARTNERS AND REGULATION OF SECRETION

219

in the brain (Hall et al., 2005; Takehashi et al., 2004), while others are highly expressed in post-mitotic cells such as neurons and platelets (Caltagarone et al., 1998; Dent et al., 2002; Martinez et al., 2006). This distribution suggested that septins might play a role in exocytosis. Neuronal SEPT2 and SEPT5 localize to the plasma membrane and can be co-purified with the SNARE protein syntaxin (Beites et al., 1999). Syntaxin forms a complex with membrane-associated proteins on both the plasma and vesicular membranes, and the formation of this complex is critical for vesicle docking and fusion (Beites et al., 1999; Weber et al., 1998). SEPT5 interacts with the region of syntaxin 1A that binds to other secretory proteins (including vesicle-associated membrane proteins, VAMP and soluble N -ethylmaleimide-sensitive factor attachment protein-25, SNAP-25 (Beites, Campbell and Trimble, 2005)) and inhibits the binding of syntaxin to these proteins. Since the binding of syntaxin is critical for exocytosis, SEPT5 would be predicted to act as a negative regulator of secretion. In support of this idea, over-expression of SEPT5 in a neuronal cell line inhibited secretion of human growth hormone (HGH) in response to potassium stimulation (Beites et al., 1999). Furthermore, recent data suggest that SEPT5 is phosphorylated by cyclindependent kinase 5 (Cdk5) on Ser17, and that this phosphorylation reduces the binding of SEPT5 to syntaxin-1 (Taniguchi et al., 2007). This represents the only known mechanism by which septin–syntaxin interaction is regulated, and provides a potential mechanism to fine-tune this interaction. Recently, William Trimble and coworkers generated SEPT5-null mice, but found no deficits in neural development or neurotransmitter release, nor any inability for the mice to develop to adulthood (Peng et al., 2002). While this argues for the dispensability of septins in the exocytic process, it remains possible that the expression of other septins in the SEPT5-null mice was sufficient to perform the role of the missing septin protein in neurons (Peng et al., 2002). To explore this possibility, platelet function was examined in wild type and SEPT5-null mice. SEPT5 is expressed in platelets, purifies with members of the exocyst complex, and inhibits agonist-induced secretion. In platelets from SEPT5-null mice, platelet secretion was enhanced as compared to wild type (Dent et al., 2002). This observation provides direct evidence of septin regulation of secretion, and lends support to the possibility of functional substitution of other septins in the neurons of SEPT5-null mice (Peng et al., 2002). In addition to SEPT5, 4 other septins (SEPT1, 4, 6 and 7) have been co-purified from rat brains with the exocyst complex (Hsu et al., 1998). The exocyst complex is critical for vesicle targeting and neurite outgrowth (Vega and Hsu, 2001). While it has not been demonstrated directly, it is possible that these septins either regulate the complex, or else tether and recruit other factors and/or vesicles to sites of active exocytosis (Kartmann and Roth, 2001). In support of this hypothesis, SEPT2 has been demonstrated to re-localize from the cytoskeleton to peripheral exocyst complexes upon nerve growth factor (NGF) stimulation of rat neural cells (Vega and Hsu, 2003).

220

CH09 SEPTIN-INTERACTING PROTEINS IN MAMMALS

SEPTIN-BINDING PARTNERS AND NEUROLOGY In addition to regulating exocytosis, septins have been found to alter conductance through the glutamate transporter Glutamate-aspartate transporter (GLAST) (Kinoshita et al., 2004). GLAST is widely expressed throughout the brain in astrocytes, and it is important for proper brain development and function (Matsugami et al., 2006). Kinoshita et al. reported that SEPT2 interacts directly with the C-terminus of GLAST, and that the two proteins can be co-precipitated from mouse brain extracts (Kinoshita et al., 2004). This interaction inhibits GLASTmediated glutamate uptake, perhaps by causing the internalization of GLAST from the plasma membrane (Kinoshita et al., 2004). SEPT2 is expressed in neurons, and is enriched during embryonic development (Hall et al., 2005; Kinoshita et al., 1997). Whether this SEPT2–GLAST interaction is important during the development of the brain remains to be determined. Septin–protein interactions have been implicated in the neurodegenerative Parkinson’s disease. SEPT4, but not septins 2, 5, 6, 7 or 8, have been found in Lewy bodies, which are cytoplasmic inclusions associated with some variants of this disease (Ihara et al., 2003). Within these bodies, SEPT4 is associated with α-synuclein, a protein largely responsible for Lewy body formation and pathology (Chua and Tang, 2006; Ihara et al., 2003). While the relationship between SEPT4 and Lewy bodies is not fully clear, recent data suggest that SEPT4 expression is neuroprotective, and that proper function of this molecule is necessary to prevent α-synuclein precipitation and neurotoxicity (Ihara et al., 2007). More data will certainly shed light on this relationship, and provide insight into the progression of Parkinson’s disease. Recent data have also emerged that implicate septins in a familial form of Parkinson’s disease. This inherited type of dopaminergic neurodegeneration is due to a loss of the protein Parkin, an E2 ubiquitin ligase that targets specific neuronal proteins for degradation (Chung et al., 2001; Goldberg et al., 2003; Tsai et al., 2003). SEPT5 is a Parkin substrate, and individuals with the mutated Parkin gene have increased expression of SEPT5 (Choi et al., 2003; Zhang et al., 2000). Intriguingly, this accumulation of SEPT5 appears to be responsible for the pathology of familial Parkinson’s disease. Over-expression of SEPT5 leads to dopamine-dependent neurotoxicity that can be reversed by concomitant Parkin over-expression (Dong et al., 2003; Son et al., 2005). More research will be necessary, however, to conclusively link septins to Parkinson’s disease.

SEPTIN-BINDING PARTNERS AND INFECTIOUS DISEASES Recently, various infectious agents septins have been found to hijack septins. Listeria monocytogenes is an intracellular bacterium that is internalized through phagocytosis, and subsequently escapes to live and proliferate within the cytoplasm (Reviewed in Cossart and Lecuit, (1998); Cossart, Pizarro-Cerada and

SEPTIN-BINDING PARTNERS AND INFECTIOUS DISEASES

221

Lecuit (2003)). Listeria uses the internalin family of proteins to mediate cellular entry. Internalin A (InlA) binds to E-cadherin and uses cofilin and cellular catenins to alter the actin cytoskeleton and mediate the bacterium’s passage through the epithelial layer and into the cell (Pizarro-Cerda et al., 2004). Internalin B (InlB) binds to a number of different proteins, including the hepatocyte growth factor receptor c-Met, the complement receptor gClq-R, and glycosaminoglycans (Pizarro-Cerda et al., 2004). Intriguingly, however, Internalin B also interacts with SEPT9 (Pizarro-Cerda et al., 2002). SEPT9 is recruited specifically to vesicles induced by InlB-coated beads and is absent in InlA-induced vesicles (Pizarro-Cerda et al., 2004). While these results are intriguing, many questions remain. Does SEPT9 bind directly to InlB? Is SEPT9 essential for Listeria escape from the phagosome? If so, what role does SEPT9 play in internalization and escape? Is Inl-cytoskeletal interaction a general theme of Listeria pathogenesis? In addition to roles in bacterial pathogenesis, the septins have been linked to several virally encoded proteins. The hepatitis C virus (HCV) is a positive-strand RNA virus that causes chronic hepatitis (Kim et al., 2007). Novel copies of the HCV genome are generated by the virally encoded NS5b, an RNA-dependent RNA polymerase that uses the RNA template to generate new copies of the viral RNA (Cheney et al., 2002). However, virally encoded proteins alone are insufficient to generate new copies of the virus. To replicate, HCV requires a number of cellular factors to synthesize new copies of its RNA. One of these proteins is hnRNP A1, an RNA-binding protein that is normally involved in mRNA processing and transport (reviewed in Weighardt, Biamonti and Riva, (1996)). Recently, it was found that the replication of the hepatitis C virus genome relies on a quaternary complex of viral RNA, NS5b, host-encoded hnRNP A1 and host-encoded SEPT6 (Kim et al., 2007). SEPT6 interacted with both NS5b and hnRNP A1, both by yeast two-hybrid and immunoprecipitation. Importantly, the loss of any of these host components resulted in deficits in viral reproduction (Kim et al., 2007). While the exact role of SEPT6 in this complex is unclear, it is tempting to speculate that the septin acts as a scaffold to bring the other factors together. Future research will shed light on this process and detail the role of septins in a disease that kills millions of individuals each year. Recently, it was reported that SEPT4 binds to the Kaposi’s sarcoma-associated herpesvirus (KSHV) protein Kaposin A (Lin et al., 2007). KSHV is a virus associated with several distinct cancers in immuno-compromised patients. Kaposin A is expressed during the latent phase of the viral life cycle, and is involved in tumorogenic transformation and alteration of cellular kinase activity (Muralidhar et al., 1998; Muralidhar et al., 2000). However, the mechanism by which Kaposin A mediates cellular proliferation and transformation is not clearly understood. Lin et al identified SEPT4 as a Kaposin A-associated protein (Lin et al., 2007). A short splice variant of SEPT4 also appears to play a role in promoting apoptosis (Larisch, 2004; Larisch et al., 2000), and the Kaposin A-SEPT4 interaction apparently blocks this pro-apoptotic effect, which may account in part for the

222

CH09 SEPTIN-INTERACTING PROTEINS IN MAMMALS

transformed phenotype induced by the virus (Lin et al., 2007). While more work remains to determine the importance and mechanism of the Kaposin–SEPT4 interaction in KSHV-infected cells, these data suggest that septins may play a role in the tumourigenesis of KSHV.

CONCLUSIONS There is a consensus view emerging that septins function primarily as scaffolds to which other proteins are attached, either to localize them to a particular place in the cell, or to sequester them away from other proteins, or perhaps to increase the local concentration of multiple proteins to facilitate an enzyme reaction or complex formation. Yet surprisingly few septin-binding proteins have to date been discovered, as compared to other scaffold ‘nodes’ in signalling networks, some of which have scores of partners. Why is this? We suggest several possible reasons. First, two-hybrid screens using individual septins as bait will not identify partners that only recognize septin heteromers or filaments. For example, Borg3 associates with the SEPT6/7 dimer, but not with either monomer alone. Thus, standard two-hybrid screens might miss the majority of important partners. Second, binding affinities to soluble oligomers might be very low, so the complexes will disassociate during purification. Perhaps one reason for the assembly of septins into filaments is to increase the avidity of the septins for their partners, by providing a very high local concentration of binding sites. Proteomics screens that disrupt filament organization would in this case not detect low affinity partners. Third, the septins might bind key partners only when arrayed along actin microfilaments or along microtubules, or when bound to cell membranes, and cell lysis disrupts this higher-order organization. Finally, key partners might only bind to post-translationally modified septins that form a small fraction of the total septin pool. Together, these problems argue that new approaches might be required to identify key binding partners. Perhaps fluorescence resonance energy transfer (FRET) screens in intact cells might be devised, or reversible, cell-permeable cross linkers could be employed to stabilize protein–protein associations. Whatever tools are used, it seems certain that there are still many interesting new septin-binding proteins awaiting discovery.

ACKNOWLEDGEMENTS This work was supported in part by NIH grants GM070902 and GM66306.

REFERENCES Amir, S., Wang, R., Matzkin, H. et al. (2006) MSF-A interacts with hypoxia-inducible factor-1 alpha and augments hypoxia-inducible factor transcriptional activation to affect tumorigenicity and angiogenesis. Cancer Research, 66, 856–66.

REFERENCES

223

Andersen, S.S. (2000) Spindle assembly and the art of regulating microtubule dynamics by MAPs and Stathmin/Op18. Trends in Cell Biology, 10, 261–67. Banks, A.S., Li, J., McKeag, L. et al. (2005) Deletion of SOCS7 leads to enhanced insulin action and enlarged islets of Langerhans. The Journal of Clinical Investigation, 115, 2462–71. Beites, C.L., Campbell, K.A. and Trimble, W.S. (2005) The septin Sept5/CDCrel-1 competes with alpha-SNAP for binding to the SNARE complex. The Biochemical Journal , 385, 347–53. Beites, C.L., Xie, H., Bowser, R. and Trimble, W.S. (1999) The septin CDCrel-1 binds syntaxin and inhibits exocytosis. Nature Neuroscience, 2, 434–9. Blaser, S., Horn, J., Wurmell, P. et al. (2004) The novel human platelet septin SEPT8 is an interaction partner of SEPT4. Thrombosis and Haemostasis, 91, 959–66. Blaser, S., Jersch, K., Hainmann, I. et al. (2002) Human septin-septin interaction: CDCrel-1 partners with KIAA0202. FEBS Letters, 519, 169–72. Blaser, S., Roseler, S., Rempp, H. et al. (2006) Human endothelial cell septins: SEPT11 is an interaction partner of SEPT5. The Journal of Pathology, 210, 103–10. Caltagarone, J., Rhodes, J., Honer, W.G. and Bowser, R. (1998) Localization of a novel septin protein, hCDCrel-1, in neurons of human brain. Neuroreport, 9, 2907–12. Chang, W., Gruber, D., Chari, S. et al. (2001) Phosphorylation of MAP4 affects microtubule properties and cell cycle progression. Journal of Cell Science, 114, 2879–87. Cheney, I.W., Naim, S., Lai, V.C. et al. (2002) Mutations in NS5B polymerase of hepatitis C virus: impacts on in vitro enzymatic activity and viral RNA replication in the subgenomic replicon cell culture. Virology, 297, 298–306. Choi, P., Snyder, H., Petrucelli, L. et al. (2003) SEPT5 v2 is a parkin-binding protein. Brain Research. Molecular Brain Research, 117, 179–89. Chua, C.E. and Tang, B.L. (2006) alpha-synuclein and Parkinson’s disease: the first roadblock. Journal of Cellular and Molecular Medicine, 10, 837–46. Chung, K.K., Zhang, Y., Lim, K.L. et al. (2001) Parkin ubiquitinates the alphasynuclein-interacting protein, synphilin-1: implications for Lewy-body formation in Parkinson’s disease. Nature Medicine, 7, 1144–50. Cossart, P. and Lecuit, M. (1998) Interactions of Listeria monocytogenes with mammalian cells during entry and actin-based movement: bacterial factors, cellular ligands and signaling. The EMBO Journal , 17, 3797–806. Cossart, P., Pizarro-Cerda, J. and Lecuit, M. (2003) Invasion of mammalian cells by Listeria monocytogenes: functional mimicry to subvert cellular functions. Trends in Cell Biology, 13, 23–31. Dent, J., Kato, K., Peng, X.R. et al. (2002) A prototypic platelet septin and its participation in secretion. Proceedings of the National Academy of Sciences of the United States of America, 99, 3064–69. Dong, Z., Ferger, B., Paterna, J.C. et al. (2003) Dopamine-dependent neurodegeneration in rats induced by viral vector-mediated overexpression of the parkin target protein, CDCrel-1. Proceedings of the National Academy of Sciences of the United States of America, 100, 12438–43. Drewes, G., Ebneth, A. and Mandelkow, E.M. (1998) MAPs, MARKs and microtubule dynamics. Trends in Biochemical Sciences, 23, 307–11.

224

CH09 SEPTIN-INTERACTING PROTEINS IN MAMMALS

Drewes, G., Ebneth, A., Preuss, U. et al. (1997) MARK, a novel family of protein kinases that phosphorylate microtubule-associated proteins and trigger microtubule disruption. Cell , 89, 297–308. Ebneth, A., Drewes, G., Mandelkow, E.M. and Mandelkow, E. (1999) Phosphorylation of MAP2c and MAP4 by MARK kinases leads to the destabilization of microtubules in cells. Cell Motility and the Cytoskeleton, 44, 209–224. Ebneth, A., Godemann, R., Stamer, K. et al. (1998) Overexpression of tau protein inhibits kinesin-dependent trafficking of vesicles, mitochondria, and endoplasmic reticulum: implications for Alzheimer’s disease. The Journal of Cell Biology, 143, 777–94. Etienne-Manneville, S. and Hall, A. (2002) Rho GTPases in cell biology. Nature, 420, 629–35. Field, C.M. and Alberts, B.M. (1995) Anillin, a contractile ring protein that cycles from the nucleus to the cell cortex. The Journal of Cell Biology, 131, 165–78. Field, C.M., Coughlin, M., Doberstein, S. et al. (2005) Characterization of anillin mutants reveals essential roles in septin localization and plasma membrane integrity. Development (Cambridge, England), 132, 2849–60. Fujishima, K., Kiyonari, H., Kurisu, J. et al. (2007) Targeted deletion of Sept3, a heteromeric assembly partner of Sept5 and Sept7 in axons, has no effect on developing CNS neurons. Journal of Neurochemistry, 102, 77–92. Garcia, Z., Silio, V., Marques, M. et al. (2006) A PI3K activity-independent function of p85 regulatory subunit in control of mammalian cytokinesis. The EMBO Journal , 25, 4740–51. Goldberg, M.S., Fleming, S.M., Palacino, J.J. et al. (2003) Parkin-deficient mice exhibit nigrostriatal deficits but not loss of dopaminergic neurons. The Journal of Biological Chemistry, 278, 43628–35. Gottfried, Y., Rotem, A., Lotan, R. et al. (2004) The mitochondrial ARTS protein promotes apoptosis through targeting XIAP. The EMBO Journal , 23, 1627–35. Hall, P.A., Jung, K., Hillan, K.J. and Russell, S.E. (2005) Expression profiling the human septin gene family. The Journal of Pathology, 206, 269–78. Hanai, N., Nagata, K., Kawajiri, A. et al. (2004) Biochemical and cell biological characterization of a mammalian septin, Sept11. FEBS Letters, 568, 83–88. Howard, J.K. and Flier, J.S. (2006) Attenuation of leptin and insulin signaling by SOCS proteins. Trends in Endocrinology and Metabolism, 17, 365–71. Hsu, S.C., Hazuka, C.D., Roth, R. et al. (1998) Subunit composition, protein interactions, and structures of the mammalian brain sec6/8 complex and septin filaments. Neuron, 20, 1111–22. Ihara, M., Tomimoto, H., Kitayama, H. et al. (2003) Association of the cytoskeletal GTPbinding protein Sept4/H5 with cytoplasmic inclusions found in Parkinson’s disease and other synucleinopathies. The Journal of Biological Chemistry, 278, 24095–102. Ihara, M., Yamasaki, N., Hagiwara, A. et al. (2007) Sept4, a component of presynaptic Scaffold and Lewy bodies, is required for the suppression of alpha-synuclein neurotoxicity. Neuron, 53, 519–33. Ito, H., Iwamoto, I., Morishita, R. et al. (2005) Possible role of Rho/Rhotekin signaling in mammalian septin organization. Oncogene, 24, 7064–72. Ito, H., Iwamoto, I., Morishita, R. et al. (2006) Identification of a PDZ protein, PIST, as a binding partner for Rho effector Rhotekin: biochemical and cell-biological characterization of Rhotekin-PIST interaction. The Biochemical Journal , 397, 389–98.

REFERENCES

225

Jaffe, A.B. and Hall, A. (2005) Rho GTPases: biochemistry and biology. Annual Review of Cell and Developmental Biology, 21, 247–69. Joberty, G., Perlungher, R.R. and Macara, I.G. (1999) The Borgs, a new family of Cdc42 and TC10 GTPase-interacting proteins. Molecular and Cellular Biology, 19, 6585–97. Joberty, G., Perlungher, R.R., Sheffield, P.J. et al. (2001) Borg proteins control septin organization and are negatively regulated by Cdc42. Nature Cell Biology, 3, 861–66. Kartmann, B. and Roth, D. (2001) Novel roles for mammalian septins: from vesicle trafficking to oncogenesis. Journal of Cell Science, 114, 839–44. Kim, C.S., Seol, S.K., Song, O.K. et al. (2007) An RNA-binding protein, hnRNP A1, and a Scaffold protein, septin 6, facilitate hepatitis C virus replication. Journal of Virology, 81, 3852–65. Kinoshita, M. (2003) Assembly of mammalian septins. Journal of Biochemistry, 134, 491–96. Kinoshita, M., Field, C.M., Coughlin, M.L. et al. (2002) Self- and actin-templated assembly of mammalian septins. Developmental Cell , 3, 791–802. Kinoshita, N., Kimura, K., Matsumoto, N. et al. (2004) Mammalian septin Sept2 modulates the activity of GLAST, a glutamate transporter in astrocytes. Genes to Cells, 9, 1–14. Kinoshita, M., Kumar, S., Mizoguchi, A. et al. (1997) Nedd5, a mammalian septin, is a novel cytoskeletal component interacting with actin-based structures. Genes and Development, 11, 1535–47. Koshelev, Y.A., Kiselev, S.L. and Georgiev, G.P. (2003) Interaction of the S100A4 (Mts1) protein with septins Sept2, Sept6, and Sept7 in vitro. Doklady. Biochemistry and Biophysics, 391, 195–97. Kremer, B.E., Adang, L.A. and Macara, I.G. (2007) Septins regulate actin organization and cell cycle arrest through SOCS7-mediated nuclear accumulation of NCK. Cell , 130, 837–50. Kremer, B.E., Haystead, T. and Macara, I.G. (2005) Mammalian septins regulate microtubule stability through interaction with the microtubule-binding protein MAP4. Molecular Biology of the Cell , 16, 4648–59. Larisch, S. (2004) The ARTS connection: role of ARTS in apoptosis and cancer. Cell Cycle (Georgetown, Tex), 3, 1021–23. Larisch, S., Yi, Y., Lotan, R. et al. (2000) A novel mitochondrial septin-like protein, ARTS, mediates apoptosis dependent on its P-loop motif. Nature Cell Biology, 2, 915–21. Lin, C.W., Tu, P.F., Hsiao, N.W. et al. (2007) Identification of a novel septin 4 protein binding to human herpesvirus 8 kaposin A protein using a phage display cDNA library. Journal of Virological Methods, 143 (1): 65–72. Lindsey, R. and Momany, M. (2006) Septin localization across kingdoms: three themes with variations. Current Opinion in Microbiology, 9: 559–65. Liu, C.A., Wang, M.J., Chi, C.W. et al. (2004) Overexpression of rho effector rhotekin confers increased survival in gastric adenocarcinoma. Journal of Biomedical Science, 11, 661–70. Low, C. and Macara, I.G. (2006) Structural analysis of septin 2, 6, and 7 complexes. The Journal of Biological Chemistry, 281, 30697–706.

226

CH09 SEPTIN-INTERACTING PROTEINS IN MAMMALS

Maddox, A.S., Habermann, B., Desai, A. and Oegema, K. (2005) Distinct roles for two C. elegans anillins in the gonad and early embryo. Development (Cambridge, England), 132, 2837–48. Mandelkow, E. and Mandelkow, E.M. (1995) Microtubules and microtubule-associated proteins. Current Opinion in Cell Biology, 7, 72–81. Mandelkow, E.M., Thies, E., Trinczek, B., Biernat, J. and Mandelkow, E. (2004) MARK/PAR1 kinase is a regulator of microtubule-dependent transport in axons. Journal of Cell Biology, 167, 99–110. Martens, N., Wery, M., Wang, P. et al. (2004) The suppressor of cytokine signaling (SOCS)-7 interacts with the actin cytoskeleton through vinexin. Experimental Cell Research, 298, 239–48. Martinez, C., Corral, J., Dent, J.A. et al. (2006) Platelet septin complexes form rings and associate with the microtubular network. Journal of Thrombosis and Haemostasis, 4, 1388–95. Martinez, C., Sanjuan, M.A., Dent, J.A. et al. (2004) Human septin-septin interactions as a prerequisite for targeting septin complexes in the cytosol. The Biochemical Journal , 382, 783–91. Matsugami, T.R., Tanemura, K., Mieda, M. et al. (2006) From the cover: indispensability of the glutamate transporters GLAST and GLT1 to brain development. Proceedings of the National Academy of Sciences of the United States of America, 103, 12161–66. Matuoka, K., Miki, H., Takahashi, K. and Takenawa, T. (1997) A novel ligand for an SH3 domain of the adaptor protein Nck bears an SH2 domain and nuclear signaling motifs. Biochemical and Biophysical Research Communications, 239, 488–92. Muralidhar, S., Pumfery, A.M., Hassani, M. et al. (1998) Identification of kaposin (open reading frame K12) as a human herpesvirus 8 (Kaposi’s sarcoma-associated herpesvirus) transforming gene. Journal of Virology, 72, 4980–88. Muralidhar, S., Veytsmann, G., Chandran, B. et al. (2000) Characterization of the human herpesvirus 8 (Kaposi’s sarcoma-associated herpesvirus) oncogene, kaposin (ORF K12). Journal of Clinical Virology, 16, 203–13. Nagata, K., Asano, T., Nozawa, Y. and Inagaki, M. (2004) Biochemical and cell biological analyses of a mammalian septin complex, Sept7/9b/11. The Journal of Biological Chemistry, 279, 55895–904. Nagata, K. and Inagaki, M. (2005) Cytoskeletal modification of Rho guanine nucleotide exchange factor activity: identification of a Rho guanine nucleotide exchange factor as a binding partner for Sept9b, a mammalian septin. Oncogene, 24, 65–76. Nagata, K., Kawajiri, A., Matsui, S. et al. (2003) Filament formation of MSF-A, a mammalian septin, in human mammary epithelial cells depends on interactions with microtubules. The Journal of Biological Chemistry, 278, 18538–43. Niu, J., Profirovic, J., Pan, H. et al. (2003) G-protein beta gamma subunits stimulate p114RhoGEF, a guanine nucleotide exchange factor for RhoA and Rac1. Circulation Research, 93, 848–56. Nobes, C.D. and Hall, A. (1995) Rho, rac, and cdc42 GTPases regulate the assembly of multimolecular focal complexes associated with actin stress fibers, lamellipodia, and filopodia. Cell , 81, 53–62. Oegema, K., Savoian, M.S., Mitchison, T.J. and Field, C.M. (2000) Functional analysis of a human homologue of the Drosophila actin binding protein anillin suggests a role in cytokinesis. The Journal of Cell Biology, 150, 539–52.

REFERENCES

227

Ookata, K., Hisanaga, S., Bulinski, J.C. et al. (1995) Cyclin B interaction with microtubule-associated protein 4 (MAP4) targets p34cdc2 kinase to microtubules and is a potential regulator of M-phase microtubule dynamics. The Journal of Cell Biology, 128, 849–62. Piperno, G., LeDizet, M. and Chang, X.J. (1987) Microtubules containing acetylated alpha-tubulin in mammalian cells in culture. The Journal of Cell Biology, 104, 289–302. Peng, X.R., Jia, Z., Zhang, Y. et al. (2002) The septin CDCrel-1 is dispensable for normal development and neurotransmitter release. Molecular and Cellular Biology, 22, 378–87. Pizarro-Cerda, J., Jonquieres, R., Gouin, E. et al. (2002) Distinct protein patterns associated with Listeria monocytogenes InlA- or InlB-phagosomes. Cellular Microbiology, 4, 101–15. Pizarro-Cerda, J., Sousa, S. and Cossart, P. (2004) Exploitation of host cell cytoskeleton and signalling during Listeria monocytogenes entry into mammalian cells. Comptes Rendus Biologies, 327, 115–23. Qi, M., Yu, W., Liu, S. et al. (2005) Septin1, a new interaction partner for human serine/threonine kinase aurora-B. Biochemical and Biophysical Research Communications, 336, 994–1000. Reid, T., Furuyashiki, T., Ishizaki, T. et al. (1996) Rhotekin, a new putative target for Rho bearing homology to a serine/threonine kinase, PKN, and rhophilin in the rho-binding domain. The Journal of Biological Chemistry, 271, 13556–60. Reynaud, C., Fabre, S. and Jalinot, P. (2000) The PDZ protein TIP-1 interacts with the Rho effector rhotekin and is involved in Rho signaling to the serum response element. The Journal of Biological Chemistry, 275, 33962–68. Schmidt, A. and Hall, A. (2002) Guanine nucleotide exchange factors for Rho GTPases: turning on the switch. Genes and Development, 16, 1587–609. Schmidt, K. and Nichols, B.J. (2004) Functional interdependence between septin and actin cytoskeleton. BMC Cell Biology, 5, 43. Sheffield, P.J., Oliver, C.J., Kremer, B.E. et al. (2003) Borg/septin interactions and the assembly of mammalian septin heterodimers, trimers, and filaments. The Journal of Biological Chemistry, 278, 3483–88. Sirajuddin, M., Farkasowsky, M., Hauer, F., Macara, I.G., Weyand, M., Stark, H. and Wittinghofer, A. (2007) Structural insights into filament formation by mammalian septins. Nature, 449, 311–15. Son, J.H., Kawamata, H., Yoo, M.S. et al. (2005) Neurotoxicity and behavioral deficits associated with Septin 5 accumulation in dopaminergic neurons. Journal of Neurochemistry, 94, 1040–53. Spiliotis, E.T., Kinoshita, M. and Nelson, W.J. (2005) A mitotic septin scaffold required for mammalian chromosome congression and segregation. Science, 307, 1781–85. Surka, M.C., Tsang, C.W. and Trimble, W.S. (2002) The mammalian septin MSF localizes with microtubules and is required for completion of cytokinesis. Molecular Biology of the Cell , 13, 3532–45. Takehashi, M., Tanaka, S., Stedeford, T. et al. (2004) Expression of septin 3 isoforms in human brain. Gene Expression, 11, 271–78. Taniguchi, M., Taoka, M., Itakura, M. et al. (2007) Phosphorylation of adult type Sept5 (CDCrel-1) by cyclin-dependent kinase 5 inhibits interaction with syntaxin-1. The Journal of Biological Chemistry, 282, 7869–76.

228

CH09 SEPTIN-INTERACTING PROTEINS IN MAMMALS

Tasto, J.J., Morrell, J.L. and Gould, K.L. (2003) An anillin homologue, Mid2p, acts during fission yeast cytokinesis to organize the septin ring and promote cell separation. The Journal of Cell Biology, 160, 1093–103. Tombes, R.M., Peloquin, J.G. and Borisy, G.G. (1991) Specific association of an M-phase kinase with isolated mitotic spindles and identification of two of its substrates as MAP4 and MAP1B. Cell Regulation, 2, 861–74. Trinczek, B., Brajenovic, M., Ebneth, A. and Drewes, G. (2004) MARK4 is a novel microtubule-associated proteins/microtubule affinity-regulating kinase that binds to the cellular microtubule network and to centrosomes. The Journal of Biological Chemistry, 279, 5915–23. Trinczek, B., Ebneth, A., Mandelkow, E.M. and Mandelkow, E. (1999) Tau regulates the attachment/detachment but not the speed of motors in microtubule-dependent transport of single vesicles and organelles. Journal of Cell Science, 112 (Pt 14), 2355–67. Tsai, Y.C., Fishman, P.S., Thakor, N.V. and Oyler, G.A. (2003) Parkin facilitates the elimination of expanded polyglutamine proteins and leads to preservation of proteasome function. The Journal of Biological Chemistry, 278, 22044–55. Vega, I.E. and Hsu, S.C. (2001) The exocyst complex associates with microtubules to mediate vesicle targeting and neurite outgrowth. The Journal of Neuroscience, 21, 3839–48. Vega, I.E. and Hsu, S.C. (2003) The septin protein Nedd5 associates with both the exocyst complex and microtubules and disruption of its GTPase activity promotes aberrant neurite sprouting in PC12 cells. Neuroreport, 14, 31–37. Weber, T., Zemelman, B.V., McNew, J.A. et al. (1998) SNAREpins: minimal machinery for membrane fusion. Cell , 92, 759–72. Weighardt, F., Biamonti, G. and Riva, S. (1996) The roles of heterogeneous nuclear ribonucleoproteins (hnRNP) in RNA metabolism. BioEssays, 18, 747–56. Xue, J., Milburn, P.J., Hanna, B.T. et al. (2004a) Phosphorylation of septin 3 on Ser-91 by cGMP-dependent protein kinase-I in nerve terminals. The Biochemical Journal , 381, 753–60. Xue, J., Tsang, C.W., Gai, W.P. et al. (2004b) Septin 3 (G-septin) is a developmentally regulated phosphoprotein enriched in presynaptic nerve terminals. Journal of Neurochemistry, 91, 579–90. Yang, T., Gao, Y.K. and Chen, J.Y. (2002) KIAA0202, a human septin family member, interacting with hPFTAIRE1. Sheng Wu Hua Xue Yu Sheng Wu Wu Li Xue Bao Acta Biochimica et Biophysica Sinica, 34, 520–25. Zhang, Y., Gao, J., Chung, K.K. et al. (2000) Parkin functions as an E2-dependent ubiquitin-protein ligase and promotes the degradation of the synaptic vesicleassociated protein, CDCrel-1. Proceedings of the National Academy of Sciences of the United States of America, 97, 13354–59. Zhao, W.M. and Fang, G. (2005) Anillin is a substrate of anaphase-promoting complex/cyclosome (APC/C) that controls spatial contractility of myosin during late cytokinesis. The Journal of Biological Chemistry, 280, 33516–24.

10 Septin functions in the mammalian cytoskeleton Elias T. Spiliotis Departments of Bioscience and Biotechnology, Drexel University, 3141 Chestnut St. Philadelphia, PA 19104, USA

W. James Nelson Department of Biology, Stanford University, The James H Clark Center, 318 Campus Drive, Stanford, CA 94305, USA

INTRODUCTION The major components of the mammalian cytoskeleton comprise three types of filamentous polymers: actin microfilaments, microtubules and intermediate filaments. Based on their fibrous organization, septins have been identified as a fourth type of cytoskeleton. However, septin fibres colocalize extensively with actin filaments and microtubules, both of which are required for septin fibre integrity. Do septins comprise a unique cytoskeleton with functions distinct from those of the other three types of filaments, or do they regulate functions of the cytoskeleton elements with which they are associated? The properties of septins indicate that they provide a skeleton-like scaffold that structurally supports membranes, however they also appear to function as regulatory scaffolds in the organization of the actin and microtubule cytoskeleton. In this chapter, we review septin interactions with actin and microtubules, and discuss the role of septins as integral, functional components of the cytoskeleton. In addition, we consider how these cytoskeleton-based functions may contribute to the etiology of diseases, in which septins are abnormally expressed.

The Septins Edited by Peter A. Hall, S.E. Hilary Russell and John R. Pringle © 2008 John Wiley & Sons, Ltd. ISBN: 978-0-470-51969-1

230

CH10 SEPTIN FUNCTIONS IN THE MAMMALIAN CYTOSKELETON

SEPTIN INTERACTIONS WITH THE ACTIN AND MICROTUBULE CYTOSKELETON For nearly three decades since their identification in the budding yeast Saccharomyces cerevisiae (Hartwell, 1971), septins were thought to be an independent filamentous system that was exclusively bound to the cell membrane with little interaction with the actin or microtubule cytoskeleton. In 2000, a biochemical screen for new actin and microtubule-associated proteins identified Drosophila septins (Sisson et al., 2000). Subsequently, genetic studies in budding yeast showed that septins are required for the interaction of microtubules with the cell membrane during mitosis and formation of a ring-like actin structure upon actin over-expression (Kusch et al., 2002; Norden, Liakopoulos and Barral, 2004). However, it was not until the identification and characterization of a growing number of mammalian septins that the cytoskeleton-binding properties of septins became fully appreciated. SEPT2 (previously called Nedd5; see Appendix B) was the first mammalian septin characterized as an actin-interacting protein (Kinoshita et al., 1997). SEPT2 fibres were originally observed to colocalize with actin stress fibres at focal adhesions (Kinoshita et al., 1997), but the extent of colocalization varies with cell type and growth phase (proliferating vs quiescent). However, some septins bind to different populations of actin filaments: for example, in certain cells SEPT2 fibres are absent from regions of rapid actin turnover (e.g. cortical actin; Kinoshita et al., 1997; Schmidt and Nichols, 2004b; Xie et al., 1999), whereas SEPT4 binds extensively to both cortical actin and stress fibres (Xie et al., 1999). Nevertheless, as a generalization, most mammalian septins (Table 10.1) localize to intracellular regions of actin enrichment such as the fillopodia, lamellopodia, stress fibres and cortical actin bundles of fibroblasts and epithelia (Kinoshita et al., 1997; Zhang et al., 1999; Xie et al., 1999; Joberty et al., 2001; Surka, Tsang and Trimble, 2002; Schmidt and Nichols, 2004b; Hanai et al., 2004; Robertson et al., 2004; Nagata et al., 2004; Nagata and Inagaki, 2005). The fibrillar morphology of the mammalian septins that colocalize with filamentous actin (F-actin) depends on the structural integrity of the actin cytoskeleton. Septin filaments (e.g. SEPT2, SEPT4, SEPT9, SEPT11) are lost upon F-actin disassembly induced by serum starvation, phosphoinositide modifying agents, and Rho guanosine tri phosphatase (GTPase) inhibitors (Kinoshita et al., 1997; Zhang et al., 1999; Xie et al., 1999; Kinoshita et al., 2002; Schmidt and Nichols, 2004b). Interestingly, direct depolymerization of F-actin with cytochalasin D or latrunculin B results in the curling of linear septin fibres into arc-shaped structures and rings of 0.6–1.4 µm in diameter indicating that the linear organization of septin bundles is dynamic and depends on F-actin integrity (Kinoshita et al., 2002; Schmidt and Nichols, 2004b). That septin fibres are dynamic was further shown by fluorescence recovery after photobleaching (FRAP); both linear and circular septin bundles are dynamic structures, albeit less mobile and diffusible than F-actin (50–80 % vs

SEPTIN INTERACTIONS WITH THE ACTIN AND MICROTUBULE CYTOSKELETON

231

Table 10.1

Septin localization in the mammalian cytoskeletona

Septin

Celltypeb

Actin

Microtubules

IFs

Centrosomes

References

SEPT1 SEPT2

HeLa HeLa

n/d Yes

Yes Scarce

n/d n/d

Yes n/d

MDCK

Yes

Yes

n/d

n/d

COS7 NIH3T3

Yes Yes

n/d n/d

n/d n/d

n/d n/d

PC12

Yes

n/d

n/d

Yes

NRK

Yes

Scarce

n/d

n/d

ERC neurons

Yes Yes

n/d n/d

n/d n/d

n/d n/d

SEPT3

neurons

No

No

n/d

n/d

SEPT4

COS7 NIH3T3 HUVEC Platelets

Yes Yes Scarce No

n/d n/d n/d Yes

n/d n/d n/d n/d

n/d n/d n/d n/d

HeLa

n/d

Scarce

n/d

n/d

MDCK

Yes

Yes

n/d

n/d

Platelets

n/d

Yes

n/d

n/d

MDCK

Yes

Yes

n/d

n/d

REF52

Yes

Scarce

n/d

n/d

HeLa

Scarce

Yes

n/d

n/d

HMEC

No

Yes

No

No

Qi et al., 2005 Surka, Tsang and Trimble, 2002; Spiliotis, Kinoshita and Nelson, 2005 Joberty et al., 2001; Spiliotis, Kinoshita and Nelson, 2005 Xie et al., 1999 Kinoshita et al., 1997, 2002; Xie et al., 1999 Vega and Hsu, 2003 Schmidt and Nichols, 2004b Ahuja et al., 2006 Kinoshita et al., 1997 Fujishima et al., 2007; Xue et al., 2004 Xie et al., 1999 Xie et al., 1999 Blaser et al., 2006 Martinez et al., 2006 Spiliotis, Kinoshita and Nelson, 2005 Joberty et al., 2001; Spiliotis, Kinoshita and Nelson, 2005 Martinez et al, 2006 Joberty et al., 2001 Nagata et al., 2004 Surka, Tsang and Trimble, 2002 Nagata et al., 2003

SEPT5

SEPT6

SEPT7

SEPT9 i1

(continued overleaf )

232

CH10 SEPTIN FUNCTIONS IN THE MAMMALIAN CYTOSKELETON

Table 10.1

(continued)

Septin

Celltypeb

Actin

Microtubules

IFs

Centrosomes

References

MCF7

n/d

Yes

n/d

n/d

CaOv3

Yes

Yes

n/d

n/d

HeLa

Yes

Scarce

n/d

n/d

COS7

No

No

No

n/d

REF52

Yes

No

No

n/d

MCF7

n/d

Yes

n/d

n/d

HeLa HMEC REF52 HUVEC

Yes Scarce Yes No

Yes Yes Scarce n/d

No No No n/d

n/d n/d n/d n/d

Chacko et al., 2005 Robertson et al., 2004 Surka, Tsang and Trimble, 2002 Nagata et al., 2004 Nagata et al., 2004 Chacko et al., 2005 Hanai et al., 2004 Hanai et al., 2004 Hanai et al., 2004 Blaser et al., 2006

SEPT9 i3

SEPT11

a Working summary of immunofluorescence studies reporting on the localization of endogenous septins (see also Lindsey and Momany, 2006). Because degree of colocalization varies with cell cycle phase, growth conditions and subcellular locale, the list is neither exhaustive nor generalizable of all intracellular pools of actin and microtubules. b Cell types: HeLa, human cervical carcinoma; MDCK, Madin–Darby canine kidney; COS7, African green monkey kidney fibroblasts; NIH3T3, mouse embryonic fibroblasts; PC12, rat pheochromocytoma; NRK, normal rat kidney; ERC, embryonic rat cardiomyocytes; HUVEC, human umbilical vein endothelial cells; HMEC, human mammary epithelia; MCF7, human mammary epithelia; CaOv3, human ovarian carcinoma; REF52, rat embryo fibroblasts.

∼100 % recovery, 10–70 s vs <10 s half times of recovery; Schmidt and Nichols, 2004b). Significantly, a direct role for F-actin as a template for septin fibre organization was shown by in vitro reconstitution experiments. Recombinant SEPT2/6/7 complexes assembled into linear fibres in the presence of F-actin, but in the absence of F-actin they formed circular rings of a diameter similar to those observed upon actin depolymerization in vivo (Schmidt and Nichols, 2004b). However, septin fibres do not have an intrinsic affinity for F-actin; their organization along actin bundles depends on adaptor proteins that can bind both actin and septins (Kinoshita et al., 2002). Although the actin bundling protein anillin can act as an adaptor protein between septin fibres and F-actin, the nuclear localization of anillin would appear to preclude a direct role in septin organization during interphase (Kinoshita et al., 2002). New studies suggest that the non-muscle myosin II anchors SEPT2 to actin stress fibres by interacting directly with SEPT2 and F-actin (see Chapter 8).

SEPTIN INTERACTIONS WITH THE ACTIN AND MICROTUBULE CYTOSKELETON

233

Irrespective of how septins are anchored to F-actin, septin fibre organization depends on Rho-family GTPase signalling pathways that also regulate the actin cytoskeleton. For example, association of SEPT9 i3 (previously known as SEPT9b: see footnote in Chapter 9) with the Rho-specific guanine nucleotide exchange factor (RhoGEF) SA-RhoGEF (septin-associated RhoGEF) is required for the fibrillar morphology and colocalization of SEPT9 i3 with actin stress fibres (Nagata and Inagaki, 2005). Similarly, Rhotekin, a downstream effector of Rho, interacts with SEPT9 i3 influencing the filamentous organization of actin stress fibres and septin fibres (Ito et al., 2005). The Cdc42-binding protein Borg3 also associates with SEPT7 molecules causing changes in septin fibrillar organization (Joberty et al., 2001). It remains unclear, however, whether or not the Rho and Cdc42 pathways influence septin assembly in different intracellular regions of actin organization (e.g. stress fibres, lamelopodia, fillopodia). In contrast to their interaction with F-actin, mammalian septins associate with microtubules directly. In protein overlay and cosedimentation assays, recombinant SEPT9 i1 (previously known as MSF-A or SEPT9a) or SEPT2/6/7 bind to unpolymerized tubulin and microtubules (Nagata et al., 2003; our unpublished observations). Mutation analysis has mapped this interaction to the GTP-binding domain (amino acids 284–561) of which is common to all known isoforms of SEPT9 (Nagata et al., 2003). Consistent with their intrinsic affinity for tubulin in vitro, septins colocalize with microtubules in vivo. Colocalization of septins (specifically SEPT2, SEPT5, SEPT9, SEPT11) with microtubules has been documented in many cell types including platelets, neurons and epithelia (Joberty et al., 2001; Surka, Tsang and Trimble, 2002; Nagata et al., 2003; Spiliotis, Kinoshita and Nelson, 2005; Martinez et al., 2006; Vega and Hsu, 2003; Hanai et al., 2004). Interestingly, colocalization of some septins (SEPT2, SEPT9, and SEPT11) with microtubules is more often observed in epithelial cells (e.g. human mammary epithelia, HMEC; Madin–Darby canine kidney, MDCK) than fibroblasts (e.g. NIH3T3, COS7, REF52; Hanai et al., 2004; Nagata and Inagaki, 2005; Nagata et al., 2004; Kinoshita et al., 1997). It is unclear if this is due to different roles of septins in specific cell types, or to differences in the relative abundance of F-actin and microtubules between epithelial cells and fibroblasts. Similar to F-actin-anchored septins, microtubule-bound septins have a characteristic fibrillar appearance. In interphase cells, these fibres dissipate upon microtubule depolymerization (Surka, Tsang and Trimble, 2002; Nagata et al., 2003; Spiliotis, Kinoshita and Nelson, 2005; Martinez et al., 2006; Vega and Hsu, 2003); however, in mitotic MDCK cells, some fibrillar elements persist even when microtubules are completely depolymerized (brief nocodazole treatments at room temperature; Spiliotis, Kinoshita and Nelson, 2005). Currently it is unknown if some microtubule-bound septin bundles are more stable than microtubules, although time-lapse imaging and FRAP indicate that their interaction is dynamic (our unpublished observations). Similar to the behaviour of intermediate filament particles and squiggles in live cells (Helfand, Chang and Goldman, 2004), septin fibrillar elements of variable length have been observed to extend, shorten and

234

CH10 SEPTIN FUNCTIONS IN THE MAMMALIAN CYTOSKELETON

move along microtubules, while thicker and longer ‘bundles’ remain relatively static. FRAP of either ‘bundles’ or finer fibrillar elements showed that septin structures recover fluorescence from the cytosol with no apparent treadmilling suggesting a dynamic exchange between cytosolic and microtubule-bound septins (our unpublished observations). Overall, fibrillar septin elements are less abundant than microtubules, and binding to the latter does not occur randomly in the cell. Some septins also localize to the centrosome (e.g. SEPT4 and SEPT7; Vahan Indjeian, personal communication), and during mitosis, SEPT1 localizes to the microtubule spindle poles and translocates to the midbody during cytokinesis (Qi et al., 2005). This latter redistribution resembles the midbody accumulation of γ-tubulin and centriolin (Shu et al., 1995; Gromley et al., 2003), which are integral components of the centrosome. Interestingly, colocalization of SEPT2 with γ-tubulin has been reported in neuronal PC12 cells (Vega and Hsu, 2003). While further work is required to establish septins as bona fide centrosomal components, the possibility may not be so remote. The centrosomal matrix is characterized by a unique filamentous system of ‘extended thin, rigid transverse structures (or a disc)’ that are attached to centrioles (Paintrand et al., 1992). It is unknown if mammalian septins could be part of this filamentous network.

SEPTIN REGULATION OF ACTIN, MICROTUBULES AND THEIR CROSSTALK While septins self-assemble into fibres in vitro, their fibrillar appearance in vivo is largely dependent on the integrity of actin microfilaments and microtubules. The dynamic turnover of septins on these structures suggests a regulatory role in the organization of actin and microtubules, their crosstalk and their interaction with cell membranes. A septin requirement for the organization of F-actin was first demonstrated in fibroblasts (Kinoshita et al., 2002). Expression of the septin-binding domain of anillin, which normally anchors septin fibres to a subset of actin bundles around the nucleus of NIH3T3 cells, resulted in the dissolution of the actin-bound septin fibres and the loss of subnuclear actin bundles, although cortical F-actin was less affected (Kinoshita et al., 2002). Similarly, disruption of septin organization by expression of the septin-binding domain of Borg3, which induces a perinuclear accumulation of septin elements, or by septin depletion with RNAi led to the disruption of subnuclear stress fibres (Kinoshita et al., 2002). Depletion or over-expression of SEPT2 has also been reported to decrease F-actin in normal rat kidney (NRK) cells (Schmidt and Nichols, 2004b). However, it is unclear if septin fibres and their colocalization with actin are necessary prerequisites for regulating actin organization. For example, perturbation of the distribution of SEPT9 i3, which colocalizes with actin stress fibres, by expression of exogenous SEPT7 and SEPT11 had little or no effect on actin stress fibres (Nagata et al., 2004). In addition, over-expression of SEPT2 domains that do not form fibres can lead to the disruption of F-actin (Schmidt and Nichols, 2004b).

SEPTIN REGULATION OF ACTIN, MICROTUBULES AND THEIR CROSSTALK

235

Perhaps, the best evidence for the requirement of fibrillar septins in actin organization comes from their interactions with the Rho signalling effectors SA-RhoGEF and Rhotekin. SEPT9 i3 and the Rho activator SA-RhoGEF colocalize with actin stress fibres. This localized interaction between SEPT9 i3 and SA-RhoGEF is critical for SA-RhoGEF inhibition. SA-RhoGEF interacts with the N-terminal domain of SEPT9 i3 and in the absence of SEPT9 i3, SA-RhoGEF is activated and triggers the formation of stress fibres (Nagata and Inagaki, 2005). Significantly, SA-RhoGEF activity cannot be suppressed by the N-terminus of SEPT9 i3 alone, but requires the full-length SEPT9 i3, which forms fibres along F-actin (Nagata and Inagaki, 2005). Hence, SEPT9 i3 fibre-dependent inhibition of SA-RhoGEF along actin bundles may limit further formation of actin stress fibres. Conversely, the fibrillar organization of SEPT9 i3 is regulated by Rhotekin, a downstream effector of Rho signalling, which codistributes with SEPT9 i3 along stress fibres (Ito et al., 2005). Upon prolonged activation of Rho signalling with lysophosphatidic acid (LPA), Rhotekin induces the reorganization of SEPT9 i3 from an F-actin-like filamentous pattern to a cytosolic vesicular pattern (Ito et al., 2005). Taken together, these data suggest that septin fibres may act as transient scaffolds for the spatial organization of F-actin and regulators of Rho signalling. Cytosolic, non-fibrous septin complexes may also control the actin cytoskeleton. Recent evidence shows that cytosolic septins sequester components of the nucleocytoplasmic shuttling pathway of the p21-activated kinase PAK1 away from the nucleus and thus, may effect actin organization by maintaining/modulating the cytoplasmic presence of PAK1 (Kremer, Adang and Macara, 2007); PAK1 activity is known to influence the stability of focal contacts, stress fibre organization and the formation of filipodia and membrane ruffles (Bokoch, 2003). Perhaps, the relative distribution of different septin proteins between cytoplasmic and filamentous assemblies defines their function in actin regulation (termed diversity of septin scaffolds; Kinoshita, 2006). Regulation of microtubules by septin fibres is not well understood. In HMEC epithelia cells, loss of microtubules is observed upon RNAi depletion of SEPT9 i1 (Nagata et al., 2003). In contrast, SEPT7-depleted HeLa cells appear to have more microtubules, which are highly acetylated and resistant to depolymerization with nocodazole (Kremer, Haystead and Macara, 2005). These seemingly contradicting data may reflect septin- and/or cell line-specific differences in microtubule organization or differences in the roles of cytoplasmic (non-filamentous) and microtubule-bound (filamentous) septins. Note that in contrast to the fibrillar SEPT9 i1, HeLa cells scarcely contain any fibrillar SEPT7/2/6 that colocalizes with microtubules (some SEPT2 coalignment with microtubules is reported during mitosis; Surka, Tsang and Trimble, 2002). Although little is known about the functional significance of septin-binding to interphase microtubules, in mitotic MDCK cells SEPT2 fibres are positioned within the metaphase plate for the proper capture and alignment of chromosomes by spindle microtubules (Spiliotis, Kinoshita and Nelson, 2005). Here, SEPT2 fibres act as a scaffold for the localization of the centromere associated

236

CH10 SEPTIN FUNCTIONS IN THE MAMMALIAN CYTOSKELETON

protein E (CENP-E; a mitotic microtubule motor and checkpoint protein) to the kinetochores of congressing chromosomes. Fibrillar SEPT2 and SEPT9 are also detected in the central spindle during anaphase and in the midbody during cytokinesis (Surka, Tsang and Trimble, 2002; Spiliotis, Kinoshita and Nelson, 2005). On these microtubule structures, septins may scaffold chromosome passenger proteins that coordinate chromosome segregation with cytokinesis. In addition, they may support the delivery and fusion of membrane vesicles with the plasma membranes of the two separating daughter cells during abscission (Joo, Tsang and Trimble, 2005). Although these functions are consistent with the range of mitotic phenotypes observed upon septin depletion or injection of septin-blocking antibodies (chromosome misalignment, lack of spindle elongation and chromosome segregation, cleavage furrow regression, failed abscission; Surka, Tsang and Trimble, 2002; Spiliotis, Kinoshita and Nelson, 2005), it is unclear how septin scaffolds may coordinate chromosome congression and segregation with the positioning and activation of the plane of cytokinesis. An interaction between spindle microtubules and the plasma membrane is most likely involved, but it is unknown if such crosstalk between microtubule- and membrane-bound septins exists. The role of septins in the crosstalk between the actin and microtubule cytoskeleton that underlies many processes (cell division, wound healing and cell motility) is not understood. Our knowledge of the signalling networks supported by the cytoskeleton-bound septin scaffolds is incomplete, but septins clearly regulate F-actin and microtubule organization raising the possibility that they may also mediate the structural and regulatory interactions that occur between these two cytoskeleton systems. Do various septin complexes exhibit differential association with actin and microtubules? If so, can this differential binding provide a code for a structural and regulatory crosstalk between actin and microtubules? Some evidence for the preferential association of specific septin isoforms with microtubules or actin has emerged from immunolocalization studies. In HeLa and possibly other cell lines (MCF7, HMEC), filaments that contain SEPT9 i1 exhibit strong colocalization with microtubules while other SEPT9 isoforms and SEPT2/6/7 complexes colocalize predominately with actin (Surka, Tsang and Trimble, 2002; Robertson et al., 2004; Nagata et al., 2003). Notably, depolymerization of microtubules results in the redistribution of filamentous SEPT9 i1 (in HeLa cells) and SEPT2 (in MDCK cells) from microtubules to actin stress fibres (Surka, Tsang and Trimble, 2002; our unpublished observations). Although this relocalization might be brought about by Rho activation, which is triggered by microtubule depolymerization, septins appear to respond to the signals of an actin– microtubule crosstalk. Redistribution of septin complexes in response to changes in cytoskeleton organization may in turn affect their assembly into regulatory scaffolds. Consistent with this possibility, exogenous expression of recombinant SEPT7, SEPT9 or SEPT11 in COS7 cells has different effects on the fibrillar organization of the other septins (Nagata et al., 2004). While coexpression of SEPT9 and SEPT11

THE SEPTIN MEMBRANE SKELETON

237

leads to the bundling of both septins, coexpression of SEPT7 and SEPT 11 disrupts both types of septin fibres (Nagata et al., 2004). Similar results have been reported for SEPT9 i4, whose exogenous expression disrupts the fibrillar organization of endogenous SEPT9 (Chacko et al., 2005). Taken together with their role in actin and microtubule organization, septins may be part of a feedback regulatory loop that connects actin and microtubules by responding to cell signalling and cytoskeleton dynamics.

THE SEPTIN MEMBRANE SKELETON On the cell membrane of the budding yeast S. cerevisae, septin filaments scaffold the localization of a large network of proteins including components of various signalling pathways (Gladfelter, Pringle and Lew, 2001). In parallel, septins create a diffusion barrier that compartmentalizes the bud neck cortex (plasma membrane and smooth endoplasmic reticulum) limiting the free exchange of membrane proteins and mRNAs (Barral et al., 2000; Dobbelaere and Barral, 2004; Luedeke et al., 2005; Takizawa et al., 2000). Mammalian septins have conserved the scaffolding functions of their yeast counterparts by adapting them within the cytoskeleton, where they interact with signalling effectors and other proteins. Little is known about the interactions of mammalian septins with the plasma membrane and the membrane-associated portion of the cytoskeleton (membrane skeleton), but recent work suggests that septins are required for the organization of mammalian membranes (Ihara et al., 2005; Ihara et al., 2007; Kissel et al., 2005). Mammalian cell membranes are patchy; they consist of distinct microdomains, which are chiefly formed by the anchoring of membrane proteins to the underlying cytoskeleton (Edidin, 1992). Membrane–cytoskeleton interactions constrain the diffusion of many proteins, which through their cytoplasmic domains collide against the filamentous mesh of the membrane skeleton, and through their membrane domains bump against the cytoskeleton-bound protein ‘fences’ (Kusumi and Sako, 1996). In addition to this corralling effect, membrane–cytoskeleton interactions contribute to the elasticity of cell membranes (Sheetz, 2001). By adjusting its mesh size and interaction with cell membranes, the cytoskeleton can modulate the surface tension of a lipid bilayer, which by itself is intrinsically inelastic. This principle is most saliently exemplified by the spectrin– actin cytoskeleton of the erythrocyte membrane (Bennett and Gilligan, 1993), but other proteins such as clathrin and dynamin, and the filamentous tropomyosin and myosin contribute to the mechanochemical characteristics of cell membranes. Mammalian septins interact directly and indirectly with membrane bilayers. A direct interaction with the phospholipids phosphatidylinositol 4,5-biphosplate [PtdIns(4,5)P2] and phosphatidylinositol 3,4,5-triphosphate [PtdIns(3,4,5)P3] has been demonstrated for SEPT4 (Zhang et al., 1999). The phospholipid-binding domain was mapped to a highly conserved stretch of basic amino acids at the N-terminal end of the GTP-binding domain (Zhang et al., 1999). Since yeast

238

CH10 SEPTIN FUNCTIONS IN THE MAMMALIAN CYTOSKELETON

septins bind phosphotidylinositol (4)-phosphate [PtdIns(4)P] and phosphotidylinositol (5)-phosphate [PtdIns(5)P] through this domain (Casamayor and Snyder, 2003), it is thought that phospholipid-binding must be a common property of all septins. However, it is unknown if different septin molecules and/or complexes have different affinities for specific phosphoinositides, which in turn might be responsible for their differential distribution in subcellular membranes. Recent data show that SEPT2 and SEPT11 are enriched in early and late endosomes, respectively (Stasyk et al., 2007), but it is uncertain if this is due to their selective binding to phosphatidylinositol 3-phosphate, which is the major phosphoinositide on early endosomes, or phosphatidylinositol 3,5-bisphosphate, which is predominately found on late endosomes (Roth, 2004). Alternatively, septins can bind to cell membranes indirectly via binding partners such as anillin that contains a PtdIns(4,5)P2-binding PH domain, and components of the machinery for vesicle docking and fusion such as the mammalian exocyst complex (Hsu et al., 1998; Vega and Hsu, 2003) and the soluble N-ethylmaleimide-sensitive fusion (NSF) protein attachment SNARE receptors (Beites et al., 1999). Coincidentally, SNARE proteins cluster in membrane microdomains that correspond to ‘hot spots’ of vesicle fusion (Lang et al., 2001; Keller et al., 2001), but it is unknown whether a fibrous septin skeleton corrals SNAREs and vesicles into a confined region for fusion with the cell membrane. From a morphological standpoint, the best evidence for a septin-membrane skeleton comes from studies in budding yeast. Electron microscopy of yeast spheroplasts revealed a gauze-like mesh of crosslinked septin filaments ‘aligned into flat arrays with occasional septin filaments laid across’, and ring-like arrangements in which a few septin filaments were circumferentially aligned and crosslinked by fibres positioned in a radial orientation (Rodal et al., 2005). Polarized fluorescence microscopy revealed that septin filaments in these meshworks can rearrange by rotating 90◦ (Vrabioiu and Mitchison, 2006). Remarkably, this molecular choreography is accompanied by a shift in septin organization and dynamics; the highly immobile, ‘frozen’, hourglass septin corset that wraps around the bud neck during S, G2 and M converts into two highly dynamic rings in telophase (Dobbelaere et al., 2003). This transition resembles the type of rearrangements that could occur within the mammalian membrane skeleton in response to local changes in membrane tension or due to cell signalling. In mammalian cells, SEPT2 molecules are arranged as clusters beneath the plasma membrane and as strings that run parallel and in close apposition to the cortical actin meshwork (Kinoshita et al., 1997). However, the lack of high-resolution ultrastructural studies leaves unresolved a number of questions about how SEPT2 or any other mammalian septins are organized under the membrane bilayer. For example, do mammalian septins form a membrane-bound filamentous network similar to that of the yeast septin skeleton? If yes, do they contribute to the diffusion barriers of the actin/spectrin skeleton? The evidence is at best circumstantial. Sequestration of PtdIns(4,5)P2 or reduction of its overall intracellular levels results in the dissolution of SEPT4 filaments from the ventral membranes of fibroblast

SEPTINS AND THE CYTOSKELETON IN HUMAN DISEASE

239

cells (Zhang et al., 1999). Although this indicates that membrane-bound septins maintain a fibrous organization, SEPT4 localization persists at peripheral cell membranes indicating that the observed effect may pertain septin filaments and F-actin at focal adhesions. In a similar vain, it is uncertain if septin fibres act as diffusion barriers. In mitotic cells, transmembrane and inner leaflet membrane proteins do not diffuse across the plane of cytokinesis and during telophase, they are excluded from the midbody (Schmidt and Nichols, 2004a). Since septins are enriched at these membrane regions, they may function to corral the mitotic cell cortex. Interestingly, SEPT4 is involved in the organization of the sperm cortex. The plasma membrane of the mammalian spermatozoon is compartmentalized in at least four biochemically distinct domains (Friend, 1989). Within their respective domains, membrane proteins are highly mobile, and free exchange between domains is constrained by regional diffusion barriers (Nehme et al., 1993). Between the anterior and posterior domains of the sperm tail, the annulus is an electron-dense ring structure that lies beneath a membrane domain of ‘densely packed collections of intramembranous particles’ (Friend, 1989; Nehme et al., 1993). Although the exact relationship between the annulus and the overlying membrane is unknown, together they are thought to function as a diffusion barrier. Remarkably, SEPT4 and other septins (SEPT1/6/7) are the major structural components of the annulus (Ihara et al., 2005; Kissel et al., 2005). In spermatozoa from Sept4 knockout mice, the annulus is completely absent affecting not only cortical organization, but also kinesin-mediated intraflagellar transport in the underlying axoneme (Ihara et al., 2005). Additional studies have shown that SEPT4 is required for the membrane/cytoskeleton organization of the distal processes of Bergmann glia and the presynaptic terminals of dopaminergic neurons (Ihara et al., 2007). Thus, mammalian septins are in the interface of cell membrane–cytoskeleton interactions in various cell types including spermatozoa, neurons and platelets. However, more work is required to determine how they interact with lipid bilayers and membrane skeletons to structurally support cell membranes.

SEPTINS AND THE CYTOSKELETON IN HUMAN DISEASE In recent years, abnormal septin expression has been linked to many human diseases (see also Chapter 14). The significant advances made in understanding mammalian septin structure and function are being translated into mechanisms underlying the etiology of septin-related diseases. For example, loss of the SEPT4 annulus has been reported in the spermatozoa from patients with idiopathic asthenospermia (Ihara et al., 2005). Moreover, in patients with Parkinson’s disease (PD), SEPT4 downregulation from the presynaptic membranes of dopaminergic neurons correlated with SEPT4-dependent mislocalization and

240

CH10 SEPTIN FUNCTIONS IN THE MAMMALIAN CYTOSKELETON

aggregation of α-synuclein, a presynaptic protein whose mutant form (α-synucleinA54T ) is responsible for familial PD (Ihara et al., 2007). While these studies connect reproductive and brain diseases with septin functions in cell membrane organization, septin roles in cytoskeleton organization might be best manifested in the progression of cancer. Formation of neoplastic tumours and their development to malignant cancers involves a series of alterations in cell physiology enabled by a high rate of genetic mutations (genomic instability; Cahill et al., 1999; Hanahan and Weinberg, 2000). Tumour formation is initiated by cells that grow independently of their extracellular environment, and is further enhanced by their ability to circumvent molecular pathways that signal antigrowth and apoptosis (cell death; Hanahan and Weinberg, 2000; Kinzler and Vogelstein, 1996). Progression toward malignancy ensues with the clonal expansion of tumour cells that have acquired a set of mutations that enables them to replicate endlessly while maintaining a relatively stable, albeit grossly altered, genomic material (Cahill et al., 1999; Nowell, 1976). Cells with an inevitably compromised cytoarchitecture can then migrate to distant tissues (metastasis), where further proliferation depends on their interaction with the local microenvironment and their ability to stimulate blood vessel growth (angiogenesis) for continuous supply of oxygen and nutrients (Thiery, 2002; Chambers, Groom and MacDonald, 2002). In human cancers and mouse tumour models, septin alterations occur at the level of transcriptional and protein expression (Scott et al., 2005; Burrows et al., 2003; Montagna et al., 2003). This is in contrast to the mutations that occur in gene products (e.g. oncogenes, tumour suppressors) that control cell growth (Sager, 1997). Hence, if septins have a causative role in the progression of cancer, it might be that they most likely contribute to the increasing genomic instability, which leads to malignacy, and/or the demise of cell identity and architecture, which underlies tumour metastasis. Mechanistically, these processes involve the microtubule and actin cytoskeleton. During mitosis, aberrant chromosome segregation by the microtubule spindle apparatus can result in large-scale chromosome rearrangements (chromosome losses, gains, deletions, breaks, non-reciprocal translocations) yielding mutations in individual genes (Storchova and Pellman, 2004). In turn, demise of the actin and microtubule cytoskeleton can affect the cytoarchitecture of tumour cells and consequently, their adhesive and migratory properties (Thiery, 2002; Ingber, 2002). Based on the phenotypes caused by septin depletion/over-expression, septins could be directly involved in mechanisms of genetic instability and metastasis. In epithelial cells, SEPT2 depletion results in polyploidy (Spiliotis, Kinoshita and Nelson, 2005), which in subsequent cell divisions can result in chaotic mitosis, aneuploidy and severe chromosome rearrangements (Storchova and Pellman, 2004). Over-expression of SEPT9 i4 in epithelial cells induces actin-based protrusions and increased cell motility (Chacko et al., 2005). Given the enhanced expression of SEPT9 i4 and other SEPT9 isoforms in breast and ovarian carcinomas as well as the ectopic expression of brain-specific SEPT4 isoforms in

CONCLUSIONS

241

colorectal cancer (Burrows et al., 2003; Scott et al., 2006; Tanaka et al., 2003; Tanaka et al., 2001), septins may contribute to the epithelial-to-mesenchymal transitions that characterize invasive carcinomas. Cytoskeleton-based mechanisms of septin function also underlie diseases other than cancers in which septin genes are mutated. Recently, two missense variants of SEPT9 i3 were found to be associated with hereditary neuralgic amyotrophy; these variants differed from wild type SEPT9 i3 in both their intracellular localization but also their fibrillar organization in response to Rho signalling (Sudo et al., 2007). These mutions also alter the 5’ UTR of other SEPT9 transcripts and so other interpretations may exist (see Chapters 7 and 14). It is possible that septin roles in cytoskeleton organization may contribute to a variety of diseases ranging from infectious pathogens like Listeria monocytogenes, which propel their way through cell hosts by high-jacking the actin cytoskeleton, to brain disorders like Alzheimer’s disease, which involves the regulation of cytoskeleton interactions with microtubule-associated proteins (e.g. tau; Pizarro-Cerda et al., 2002; Kinoshita et al., 1998).

CONCLUSIONS Nearly 10 years ago, perplexed by the multifunctional properties of yeast and fly septins, Field and Kellogg posed the question ‘septins: novel polymers or signalling GTPases?’ (Field and Kellogg, 1999). Today, mammalian ‘septinologists’ are grappling with the same question. Are septins a novel cytoskeleton or do they regulate the microtubule and actin cytoskeleton they are associated with? Definitive answers might be premature, but three fundamental trends are beginning to emerge. First, septins have both regulatory and structural functions. Regulatory functions are best understood in the context of the cytoskeleton, where septin fibres scaffold protein networks that regulate microtubule and actin organization, their crosstalk and their binding to other effectors. Structural functions are best understood in the context of cell membranes, on which septins fulfil a cytoskeleton-like support for the compartmentalization of the lipid bilayer in distinct microdomains. Second, septins have distinct and overlapping intracellular distributions. While septins work together in complexes, different septins are found in different complexes, which localize to various intracellular regions (e.g. membranes, actin stress fibres, microtubules, cytoplasm) in a tissue-specific manner. Third, septin complexes are highly dynamic and vary in their ability to form high order fibrillar structures. Conceivably, fibrillar septin complexes can vary in function from their soluble, cytoplasmic counterparts, but this needs to be further explored. On these premises, septins appear to constitute a unique cytoskeleton-like network, whose plasticity in response to intracellular signalling may coordinate many cytoskeleton- and membrane-based functions in a spatiotemporal fashion.

242

CH10 SEPTIN FUNCTIONS IN THE MAMMALIAN CYTOSKELETON

ACKNOWLEDGEMENTS Work from the Nelson lab was supported by NIH grant GM35527, and Elias Spiliotis was also supported by a post-doctoral fellowship from the Jane Coffin Childs Memorial Fund for Medical Research. We thank Drs. William Trimble, Brandon Kremer, Ian Macara, Vahan Indjeian, and Brian Dynlacht for sharing unpublished observations and data, Serge Mostowy for comments on the manuscript, and many of our colleagues at the 2nd International Meeting on Septin Biology for stimulating discussions.

REFERENCES Ahuja, P., Perriard, E., Trimble, W. et al. (2006) Probing the role of septins in cardiomyocytes. Experimental Cell Research, 312, 1598–609. Barral, Y., Mermall, V., Mooseker, M.S. and Snyder, M. (2000) Compartmentalization of the cell cortex by septins is required for maintenance of cell polarity in yeast. Molecular Cell , 5, 841–51. Beites, C.L., Xie, H., Bowser, R. and Trimble, W.S. (1999) The septin CDCrel-1 binds syntaxin and inhibits exocytosis. Nature Neuroscience, 2, 434–39. Bennett, V. and Gilligan, D.M. (1993) The spectrin-based membrane skeleton and micron-scale organization of the plasma membrane. Annual Review of Cell Biology, 9, 27–66. Blaser, S., Roseler, S., Rempp, H. et al. (2006) Human endothelial cell septins: SEPT11 is an interaction partner of SEPT5. The Journal of Pathology, 210, 103–10. Bokoch, G.M. (2003) Biology of the p21-activated kinases. Annual Review of Biochemistry, 72, 743–81. Burrows, J.F., Chanduloy, S., McIlhatton, M.A. et al. (2003) Altered expression of the septin gene, SEPT9 , in ovarian neoplasia. The Journal of Pathology, 201, 581–88. Cahill, D.P., Kinzler, K.W., Vogelstein, B. and Lengauer, C. (1999) Genetic instability and Darwinian selection in tumours. Trends in Cell Biology, 9, M57–60. Casamayor, A. and Snyder, M. (2003) Molecular dissection of a yeast septin: distinct domains are required for septin interaction, localization, and function. Molecular and Cellular Biology, 23, 2762–77. Chacko, A.D., Hyland, P.L., McDade, S.S. et al. (2005) SEPT9 v4 expression induces morphological change, increased motility and disturbed polarity. The Journal of Pathology, 206, 458–65. Chambers, A.F., Groom, A.C. and MacDonald, I.C. (2002) Dissemination and growth of cancer cells in metastatic sites. Nature Reviews. Cancer, 2, 563–72. Dobbelaere, J. and Barral, Y. (2004) Spatial coordination of cytokinetic events by compartmentalization of the cell cortex. Science, 305, 393–96. Dobbelaere, J., Gentry, M.S., Hallberg, R.L. and Barral, Y. (2003) Phosphorylation dependent regulation of septin dynamics during the cell cycle. Developmental Cell , 4, 345–57. Edidin, M. (1992) Patches, posts and fences: proteins and plasma membrane domains. Trends in Cell Biology, 2, 376–80.

REFERENCES

243

Field, C.M. and Kellogg, D. (1999) Septins: cytoskeletal polymers or signalling GTPases? Trends in Cell Biology, 9, 387–94. Friend, D.S. (1989) Sperm maturation: membrane domain boundaries. Annals of the New York Academy of Sciences, 567, 208–21. Fujishima, K., Kiyonari, H., Kurisu, J. et al. (2007) Targeted disruption of Sept3, a heteromeric assembly partner of Sept5 and Sept7 in axons, has no effect on developing CNS neurons. Journal of Neurochemistry, 102, 77–92. Gladfelter, A.S., Pringle, J.R. and Lew, D.J. (2001) The septin cortex at the yeast mother-bud neck. Current Opinion in Microbiology, 4, 681–89. Gromley, A., Jurczyk, A., Sillibourne, J. et al. (2003) A novel human protein of the maternal centriole is required for the final stages of cytokinesis and entry into S phase. The Journal of Cell Biology, 161, 535–45. Hanahan, D. and Weinberg, R.A. (2000) The hallmarks of cancer. Cell , 100, 57–70. Hanai, N., Nagata, K., Kawajiri, A. et al. (2004) Biochemical and cell biological characterization of a mammalian septin, Sept11. FEBS Letters, 568, 83–88. Hartwell, L.H. (1971) Genetic control of the cell division cycle in yeast. IV. Genes controlling bud emergence and cytokinesis. Experimental Cell Research, 69, 265–76. Helfand, B.T., Chang, L. and Goldman, R.D. (2004) Intermediate filaments are dynamic and motile elements of cellular architecture. Journal of Cell Science, 117, 133–41. Hsu, S.C., Hazuka, C.D., Roth, R. et al. (1998) Subunit composition, protein interactions, and structures of the mammalian brain sec6/8 complex and septin filaments. Neuron, 20, 1111–22. Ihara, M., Kinoshita, A., Yamada, S. et al. (2005) Cortical organization by the septin cytoskeleton is essential for structural and mechanical integrity of mammalian spermatozoa. Developmental Cell , 8, 343–52. Ihara, M., Yamasaki, N., Hagiwara, A. et al. (2007) Sept4, a component of presynaptic scaffold and lewy bodies, is required for the suppression of alpha-synuclein neurotoxicity. Neuron, 53, 519–33. Ingber, D.E. (2002) Cancer as a disease of epithelial-mesenchymal interactions and extracellular matrix regulation. Differentiation, 70, 547–60. Ito, H., Iwamoto, I., Morishita, R. et al. (2005) Possible role of Rho/Rhotekin signaling in mammalian septin organization. Oncogene, 24, 7064–72. Joberty, G., Perlungher, R.R., Sheffield, P.J. et al. (2001) Borg proteins control septin organization and are negatively regulated by Cdc42. Nature Cell Biology, 3, 861–66. Joo, E., Tsang, C.W. and Trimble, W.S. (2005) Septins: traffic control at the cytokinesis intersection. Traffic (Copenhagen, Denmark), 6, 626–34. Keller, P., Toomre, D., Diaz, E. et al. (2001) Multicolour imaging of post-Golgi sorting and trafficking in live cells. Nature Cell Biology, 3, 140–49. Kinoshita, M. (2006) Diversity of septin scaffolds. Current Opinion in Cell Biology, 18, 54–60. Kinoshita, M., Field, C.M., Coughlin, M.L. et al. (2002) Self- and actin-templated assembly of mammalian septins. Developmental Cell , 3, 791–802. Kinoshita, A., Kinoshita, M., Akiyama, H. et al. (1998) Identification of septins in neurofibrillary tangles in Alzheimer’s disease. American Journal of Pathology, 153, 1551–60.

244

CH10 SEPTIN FUNCTIONS IN THE MAMMALIAN CYTOSKELETON

Kinoshita, M., Kumar, S., Mizoguchi, A. et al. (1997) Nedd5, a mammalian septin, is a novel cytoskeletal component interacting with actin-based structures. Genes and Development, 11, 1535–47. Kinzler, K.W. and Vogelstein, B. (1996) Lessons from hereditary colorectal cancer. Cell , 87, 159–70. Kissel, H., Georgescu, M.M., Larisch, S. et al. (2005) The Sept4 septin locus is required for sperm terminal differentiation in mice. Developmental Cell , 8, 353–64. Kremer, B.E., Adang, L.A. and Macara, I.G. Septins regulate action organization and cell cycle arrest through SOC7-mediated nuclear accumulation of NCK. (2007) Cell 130, 837–50 (in press.) Kremer, B.E., Haystead, T. and Macara, I.G. (2005) Mammalian septins regulate microtubule stability through interaction with the microtubule-binding protein MAP4. Molecular Biology of the Cell , 16, 4648–59. Kusch, J., Meyer, A., Snyder, M.P. and Barral, Y. (2002) Microtubule capture by the cleavage apparatus is required for proper spindle positioning in yeast. Genes and Development, 16, 1627–39. Kusumi, A. and Sako, Y. (1996) Cell surface organization by the membrane skeleton. Current Opinion in Cell Biology, 8, 566–74. Lang, T., Bruns, D., Wenzel, D. et al. (2001) SNAREs are concentrated in cholesteroldependent clusters that define docking and fusion sites for exocytosis. The EMBO Journal , 20, 2202–13. Lindsey, R. and Momany, M. (2006) Septin localization across kingdoms: three themes with variations. Current Opinion in Microbiology, 9, 559–65. Luedeke, C., Frei, S.B., Sbalzarini, I. et al. (2005) Septin-dependent compartmentalization of the endoplasmic reticulum during yeast polarized growth. The Journal of Cell Biology, 169, 897–908. Martinez, C., Corral, J., Dent, J.A. et al. (2006) Platelet septin complexes form rings and associate with the microtubular network. Journal of Thrombosis and Haemostasis, 4, 1388–95. Montagna, C., Lyu, M.S., Hunter, K. et al. (2003) The Septin 9 (MSF) gene is amplified and overexpressed in mouse mammary gland adenocarcinomas and human breast cancer cell lines. Cancer Research, 63, 2179–87. Nagata, K., Asano, T., Nozawa, Y. and Inagaki, M. (2004) Biochemical and cell biological analyses of a mammalian septin complex, Sept7/9b/11. The Journal of Biological Chemistry, 279, 55895–904. Nagata, K. and Inagaki, M. (2005) Cytoskeletal modification of Rho guanine nucleotide exchange factor activity: identification of a Rho guanine nucleotide exchange factor as a binding partner for Sept9b, a mammalian septin. Oncogene, 24, 65–76. Nagata, K., Kawajiri, A., Matsui, S. et al. (2003) Filament formation of MSF-A, a mammalian septin, in human mammary epithelial cells depends on interactions with microtubules. The Journal of Biological Chemistry, 278, 18538–43. Nehme, C.L., Cesario, M.M., Myles, D.G. et al. (1993) Breaching the diffusion barrier that compartmentalizes the transmembrane glycoprotein CE9 to the posterior-tail plasma membrane domain of the rat spermatozoon. The Journal of Cell Biology, 120, 687–94. Norden, C., Liakopoulos, D. and Barral, Y. (2004) Dissection of septin actin interactions using actin overexpression in Saccharomyces cerevisiae. Molecular Microbiology, 53, 469–83.

REFERENCES

245

Nowell, P.C. (1976) The clonal evolution of tumor cell populations. Science, 194, 23–28. Paintrand, M., Moudjou, M., Delacroix, H. and Bornens, M. (1992) Centrosome organization and centriole architecture: their sensitivity to divalent cations. Journal of Structural Biology, 108, 107–28. Pizarro-Cerda, J., Jonquieres, R., Gouin, E. et al. (2002) Distinct protein patterns associated with Listeria monocytogenes InlA- or InlB phagosomes. Cellular Microbiology, 4, 101–15. Qi, M., Yu, W., Liu, S. et al. (2005) Septin1, a new interaction partner for human serine/threonine kinase aurora-B. Biochemical and Biophysical Research Communications, 336, 994–1000. Robertson, C., Church, S.W., Nagar, H.A. et al. (2004) Properties of SEPT9 isoforms and the requirement for GTP binding. The Journal of Pathology, 203, 519–27. Rodal, A.A., Kozubowski, L., Goode, B.L. et al. (2005) Actin and septin ultrastructures at the budding yeast cell cortex. Molecular Biology of the Cell , 16, 372–84. Roth, M.G. (2004) Phosphoinositides in constitutive membrane traffic. Physiological Reviews, 84, 699–730. Sager, R. (1997) Expression genetics in cancer: shifting the focus from DNA to RNA. Proceedings of the National Academy of Sciences of the United States of America, 94, 952–55. Schmidt, K. and Nichols, B.J. (2004a) A barrier to lateral diffusion in the cleavage furrow of dividing mammalian cells. Current Biology, 14, 1002–6. Schmidt, K. and Nichols, B.J. (2004b) Functional interdependence between septin and actin cytoskeleton. BMC Cell Biology, 5, 43. Scott, M., Hyland, P.L., McGregor, G. et al. (2005) Multimodality expression profiling shows SEPT9 to be overexpressed in a wide range of human tumours. Oncogene, 24, 4688–700. Scott, M., McCluggage, W.G., Hillan, K.J. et al. (2006) Altered patterns of transcription of the septin gene, SEPT9 , in ovarian tumorigenesis. International Journal of Cancer Journal international du cancer, 118, 1325–9. Sheetz, M.P. (2001) Cell control by membrane-cytoskeleton adhesion. Nature reviews. Molecular Cell Biology, 2, 392–96. Shu, H.B., Li, Z., Palacios, M.J. et al. (1995) A transient association of gamma-tubulin at the midbody is required for the completion of cytokinesis during the mammalian cell division. Journal of Cell Science, 108 (Pt 9), 2955–62. Sisson, J.C., Field, C., Ventura, R. et al. (2000) Lava lamp, a novel peripheral golgi protein, is required for Drosophila melanogaster cellularization. The Journal of Cell Biology, 151, 905–18. Spiliotis, E.T., Kinoshita, M. and Nelson, W.J. (2005) A mitotic septin scaffold required for mammalian chromosome congression and segregation. Science, 307, 1781–85. Stasyk, T., Schiefermeier, N., Skvortsov, S. et al. (2007) Identification of endosomal epidermal growth factor receptor signaling targets by functional organelle proteomics. Molecular and Cellular Proteomics, 6, 908–22. Storchova, Z. and Pellman, D. (2004) From polyploidy to aneuploidy, genome instability and cancer. Nature reviews. Molecular Cell Biology, 5, 45–54. Sudo, K., Ito, H., Iwamoto, I. et al. SEPT9 sequence alternations causing hereditary neuralgic amyotrophy are associated with altered interactions with SEPT4/SEPT11

246

CH10 SEPTIN FUNCTIONS IN THE MAMMALIAN CYTOSKELETON

and resistance to Rho/Rhotekin-signaling. (2007) Human Mutation, 28 (10), 1005–13. Surka, M.C., Tsang, C.W. and Trimble, W.S. (2002) The mammalian septin MSF localizes with microtubules and is required for completion of cytokinesis. Molecular Biology of the Cell , 13, 3532–45. Takizawa, P.A., DeRisi, J.L., Wilhelm, J.E. and Vale, R.D. (2000) Plasma membrane compartmentalization in yeast by messenger RNA transport and a septin diffusion barrier. Science, 290, 341–44. Tanaka, M., Tanaka, T., Kijima, H. et al. (2001) Characterization of tissue- and cell-type-specific expression of a novel human septin family gene, Bradeion. Biochemical and Biophysical Research Communications, 286, 547–53. Tanaka, M., Tanaka, T., Matsuzaki, S. et al. (2003) Rapid and quantitative detection of human septin family Bradeion as a practical diagnostic method of colorectal and urologic cancers. Medical Science Monitor, 9, MT61–68. Thiery, J.P. (2002) Epithelial-mesenchymal transitions in tumour progression. Nature Reviews. Cancer, 2, 442–54. Vega, I.E. and Hsu, S.C. (2003) The septin protein Nedd5 associates with both the exocyst complex and microtubules and disruption of its GTPase activity promotes aberrant neurite sprouting in PC12 cells. Neuroreport, 14, 31–37. Vrabioiu, A.M. and Mitchison, T.J. (2006) Structural insights into yeast septin organization from polarized fluorescence microscopy. Nature, 443, 466–69. Xie, H., Surka, M., Howard, J. and Trimble, W.S. (1999) Characterization of the mammalian septin H5: distinct patterns of cytoskeletal and membrane association from other septin proteins. Cell Motility and the Cytoskeleton, 43, 52–62. Zhang, J., Kong, C., Xie, H. et al. (1999) Phosphatidylinositol polyphosphate binding to the mammalian septin H5 is modulated by GTP. Current Biology, 9, 1458–67.

11 Septins and the synapse Jing Xue Vascular Biology Centre, Medical College of Georgia, 1459 Laney Walker Blvd, Room CB 3330, Augusta, GA 30912, USA

Victor Anggono and Phillip J. Robinson Cell Signalling Unit, Children’s Medical Research Institute, Locked Bag 23, Wentworthville, NSW 2145, Australia

INTRODUCTION In the brain billions of neurons make thousands of connections to each other. They use the connections to communicate with each other through a highly specialized structure known as synapse. Each synapse is comprised of a presynaptic membrane responsible for the transmission of signal from the signalling neuron and a postsynaptic membrane which acts as a receiver of the signal. Both the presynaptic and postsynaptic membranes are separated by a narrow extracellular space of approximately 15 nm known as the synaptic cleft. When an action potential reaches the presynaptic nerve terminal, synaptic vesicles (SVs) filled with neurotransmitters fuse with the membrane, releasing their contents into the synaptic cleft, a process termed SV exocytosis. The neurotransmitters diffuse across the synaptic cleft and activate receptors on the postsynaptic neurons to complete the process of synaptic transmission. Since there are a finite amount of SVs in presynaptic terminals, the SV membrane is rapidly retrieved and recycled (SV endocytosis) in order to maintain the continuity of synaptic transmission. The ability of the neurons to change the strength of their synaptic connection is known as synaptic plasticity, an important part of the mechanisms underlying learning and memory. The strength of synaptic transmission can be modulated through changes in the quantity of neurotransmitter released into a synapse and changes in the efficacy of cells responding to those neurotransmitters (Gaiarsa, Caillard and Ben Ari, The Septins Edited by Peter A. Hall, S.E. Hilary Russell and John R. Pringle © 2008 John Wiley & Sons, Ltd. ISBN: 978-0-470-51969-1

248

CH11 SEPTINS AND THE SYNAPSE

2002). Synaptic transmission and plasticity are crucial for all aspects of nervous system function and necessary for proper development of the central nervous system (CNS). Abnormalities in synaptic transmission and plasticity contribute to the pathology of many neurological disorders (e.g. schizophrenia, autism, mental retardation) and neurodegenerative diseases (e.g. Alzheimer’s disease (AD) and Parkinson’s disease (PD)). Septins are cytosolic proteins first discovered in yeast as important regulators of membrane partitioning such as cytokinesis. To date, 14 septins, have been identified in human cells (Martinez et al., 2006; Lindsey and Momany, 2006; see Appendix B). They are ubiquitously expressed in mammalian tissues, including brain, and have been associated with a plethora of different biological functions (Trimble, 1999; Kinoshita, Noda and Kinoshita, 2000; Joo, Tsang and Trimble, 2005). In the brain, septins play various roles such as in vesicle trafficking (Hsu et al., 1998; Beites et al., 1999), neurite outgrowth (Vega and Hsu, 2003), and transporter-mediated glutamate uptake by astrocytes (Kinoshita et al., 2004). Septins have also been implicated in the pathology of such neurological disorders as AD (Kinoshita et al., 1998; Takehashi et al., 2004a), PD (Zhang et al., 2000; Ihara et al., 2003), schizophrenia (Barr et al., 2004) and epilepsy (Yang et al., 2006). This chapter reviews the distribution, localization and functions of septins expressed in the brain, as well as their role in the pathology of various neurodegenerative diseases.

DISTRIBUTION OF SEPTINS IN BRAIN Mammalian septins are present in many different tissues such as lung, spleen, heart, testis, pancreas, brain, liver and kidney. Among them, SEPT2, SEPT3, SEPT4, SEPT5, SEPT6, SEPT7, SEPT8 and SEPT11 are highly expressed in CNS (Hall et al., 2005). SEPT3 and SEPT5 are the only two septins that seem to be primarily expressed in the brain (Yagi et al., 1998; Beites et al., 1999; Xue et al., 2000). However, SEPT5 is also present (Figure 11.1) in megakaryocytes, particularly in platelets (Yagi et al., 1998; Dent et al., 2002), suggesting that SEPT3 is the only septin to date that is specific to brain as it is not found in platelets or cell lines derived from non-neuronal origin. In general, SEPT3 expression is high in all adult brain regions, and is reduced in those enriched in white matter (Xue et al., 2004b). It is present in the cerebral cortex, hippocampus, cerebellum, thalamus, hypothalamus, olfactory bulb and is especially abundant in the mossy fibres of the hippocampus CA3 region (Xue et al., 2004b; Fujishima et al., 2007). The alternatively spliced variants of SEPT3, SEPT3A and SEPT3B, are also expressed in all brain regions, with the highest level in the temporal cortex and hippocampus, and the lowest level in the brainstem regions (Takehashi et al., 2004b). There is a higher level of SEPT3B detected in the cerebellum. (NB the nomenclature of these isoforms is not yet finalized: see Appendix B.)

DISTRIBUTION OF SEPTINS IN BRAIN

249

Adrenal Gland

Aorta

Spleen

Heart

Liver

Lung

Pancreas

Skeletal muscle

Kidney

Stomach

Testis

Intestine

Brain

Sept 3 and 5 are highly expressed in CNS

Testis

Pancreas

Kidney

Liver

Brain

Heart

Spleen

Sept3

Sept5 Figure 11.1 Brain-specific septin proteins. Protein extracts from various rat (top, 60 µg) and mouse (bottom, 6 µg) organs were blotted with sheep polyclonal anti-SEPT3 and rabbit polyclonal anti-SEPT5, respectively. Both SEPT3 and SEPT5 are predominantly expressed in the brain. [Figure in the top panel was reproduced with permission from Xue J. et al (2004), J Neurochem, 91, 579–590, Copyright  2004 Blackwell Publishing, figure in the bottom panel was reproduced with permission from Beites et al. (1999) Nature Neuroscience, 2,, 434–439, Copyright  1999 Macmillan Publishers Ltd

SEPT5 is also widely distributed in the brain and is particularly abundant in ventral pallidum, globus pallidus, entopeduncular nucleus, substantia nigra pars reticulata and deep cerebellar nucleus (Caltagarone et al., 1998; Kinoshita, Noda and Kinoshita, 2000). The expression and distribution patterns of SEPT3A and SEPT3B are similar to SEPT5. They form a complex in the frontal cortex of human brain as shown by immunoprecipitation assay (Takehashi et al., 2004b). Prominent immunoreactivity was observed diffusely in the neocortices in association with neuropils and punctate structures suggestive of synaptic junctions (Takehashi et al., 2004b). SEPT4 is distributed diffusely in the neuropil with prominent staining in the thalamus, ventral pallidum, globus pallidus and cerebellar molecular layer (Kinoshita, Noda and Kinoshita, 2000; Ihara et al., 2007). On the other hand, SEPT7 is distributed evenly in most of the brain subregions, including the cerebral white matter, such as the corpus callosum (Kinoshita, Noda and Kinoshita, 2000). SEPT6 also has a broad distribution but is especially dense in the olfactory bulb, cerebral and cerebellar cortices, thalamus and hippocampus (Kinoshita, Noda and Kinoshita, 2000). SEPT2 expression is prominent in the molecular layer of the cerebellum (Kinoshita et al., 2004). Interestingly, SEPT8 displays isoform specific expression in certain areas of the brain as shown by Northern analyses. SEPT8 i1 mRNA presents at low level in areas such as the cerebellum, occipital

CH11 SEPTINS AND THE SYNAPSE

250

Sept 2,3 and 5 expression are developmentally regulated E10

E18 E21 P7

P14 P21 P28 P35 P42 Adult

Sept3 E17

P5

P7

P10

P15

Adult

Sept2 Sept5 Figure 11.2 Developmental expression of septins in the brain. Immunoblot analysis of SEPT3 (top), SEPT2 (middle) and SEPT5 (bottom) during rat brain development. SEPT3 and SEPT5 expression is extremely low in embryos and dramatically increase during postnatal life and reaches a plateau. In contrast, SEPT2 expression is down-regulated during development, then remains relatively stable in the adult. [Figure in top panel reproduced by permission from Wiley-Blackwell, Xue J., et al. (2004) J Neurochem 91, 579–590, Copyright  2004. Figures in bottom and middle panels reproduced from Peng, X.-R. et al. (2002) Molecular and Cellular Biology, 22, 378–387. Copyright  2002, American Society for Microbiology.]

pole, frontal lobe, temporal lobe and putamen; whereas the other SEPT8 transcripts (SEPT8 i2, SEPT8 i3 and SEPT8 i1*) are highly expressed throughout the brain (Blaser et al., 2003). Expression of some septins is developmentally regulated (see Figure 11.2). The differential expression and distribution of these septins indicates that a subset of neuron or glial cell expresses a specific set of septin monomers and that the resulting septin complexes with distinct compositions may play distinct roles in the brain.

LOCALIZATION OF SEPTINS IN NEURONS In the brain, SEPT3 expression is restricted to neurons and is not present in glial cells (Xue et al., 2004b). It shows a punctate distribution along the neurites of hippocampal neurons and co-localizes with the presynaptic marker protein synaptophysin as well as the endocytic protein dynamin I (Xue et al., 2004b). SEPT3 is also expressed in other types of neurons, including cerebellar granule neurons and cortical pyramidal neurons (Fujishima et al., 2007). In cultured hippocampal neurons, SEPT3 is enriched at the tips of growing neurites during differentiation and co-localizes with other septins, SEPT5 and SEPT7 at nerve terminals (Plate 11.3, see p. 246 for Plates; Fujishima et al., 2007). In cultured hippocampal neurons, SEPT5 also displays a punctate staining which partly co-localizes with another SV marker VAMP-2 (Peng et al., 2002), This

SUBCELLULAR DISTRIBUTION OF SEPTINS IN NERVE TERMINALS

251

septin directly interacts with the synaptic proteins synaptophysin and syntaxin (Honer, Hu and Davies, 1993; Caltagarone et al., 1998; Beites et al., 1999). It has also been reported to be associated with SVs in the presynaptic nerve terminals in inhibitory neurons (Kinoshita, Noda and Kinoshita, 2000). SEPT7 also displays similar punctate distribution throughout cultured hippocampal neurons and it appears perisynaptic but not presynaptic, since it does not co-localize with synaptophysin (Hsu et al., 1998; Kinoshita, Noda and Kinoshita, 2000; Fujishima et al., 2007). Rather, it co-localizes with postsynaptic density (PSD) 95 kDa protein (PSD-95) along the dendrites of hippocampal neurons (Walikonis et al., 2000). SEPT4 is abundant in the presynaptic terminals of dopaminergic (DA) neurons projecting from substantia nigra pars compacta to the striatum, where it co-localizes with the dopamine transporter (DAT) and α-synuclein (Ihara et al., 2007). In addition to their localization in neurons, SEPT2, SEPT4 and SEPT7 are also present in glia cells (Kinoshita, Noda and Kinoshita, 2000; Ihara et al., 2003; Kinoshita et al., 2004).

SUBCELLULAR DISTRIBUTION OF SEPTINS IN NERVE TERMINALS In agreement with its presynaptic localization in cultured neurons, SEPT3 expression is highly enriched in nerve terminals (Xue et al., 2000; Xue et al., 2004b). It is also associated with a peripheral membrane extract, consistent with its role in filament formation or superficial membrane attachment (Xue et al., 2004b). SEPT5 is also present in synaptosomes (isolated nerve terminals) wherein it forms a complex with SEPT2 and SEPT7 (Peng et al., 2002). It co-purifies with SNAP-25 and synaptophysin containing synaptosomes from human brain (Caltagarone et al., 1998). SEPT5 and 6 have also been specifically localized to subpopulations of nerve terminals shown by immuno electron-microscopy (EM) (Kinoshita, Noda and Kinoshita, 2000; Phillips et al., 2001). SEPT5 localizes mainly around the presynaptic vesicles, but not along the presynaptic membrane, and appears to be restricted to the inhibitory presynaptic terminals associated with GABAergic vesicles (Kinoshita, Noda and Kinoshita, 2000). Although it is present in nerve terminals, it is not specifically enriched there. Interestingly, it is relatively low in the peripheral membrane fraction and high in soluble and membrane fractions, in contrast to the subcellular localization of SEPT3 (Xue et al., 2004b). On the other hand, SEPT7 is not associated with SVs but localizes just beneath the axon terminal membranes (Kinoshita, Noda and Kinoshita, 2000). SEPT7 is particularly enriched in the PSD fraction (Walikonis et al., 2000), consistent with the proteomics analysis that shows SEPT7 association with the synaptic plasma membrane containing such protein as PSD-95/SAP-90 complex and neuromodulin (GAP-43) (Witzmann et al., 2005).

252

CH11 SEPTINS AND THE SYNAPSE

REGULATION OF SEPTIN EXPRESSION IN BRAIN The expression levels of several members of the septin family of proteins are developmentally regulated in the brain. This is significant because early appearance implicates a function in neuronal outgrowth or migration while late appearance implies a function of the mature neurons or glia like synaptic transmission. SEPT3 expression in rat brain is not detectable in early embryos and is first detectable at a late stage of embryonic development (embryonic day 18) and increases until postnatal day 7 into adulthood (Xue et al., 2004b). A similar expression pattern is also observed for dynamin I which correlates with synapse maturation (Powell and Robinson, 1995). This expression profile correlates with other mRNA studies suggesting a role for SEPT3 in neuronal development or differentiation (Xiong, Leahy and Stuhlmann, 1999; Methner et al., 2001). SEPT5 also exhibits a similar expression pattern in the brain (Peng et al., 2002). The expression level of SEPT5 is extremely low in embryos and dramatically increases during postnatal life reaching a plateau at approximately postnatal day 10. This expression profile is similar to VAMP-2, the expression of which is known to increase during postnatal synaptogenesis (Shimohama et al., 1998). Two alternate splice forms of SEPT5 in rats have been described, one which is most abundant in early postnatal life (SEPT5 i2) and a second that is prevalent in adults (SEPT5 i1) (Toda et al., 2000). In contrast to SEPT3 and SEPT5, SEPT2 expression is down-regulated during the differentiation of adult brain. Expression of SEPT2 decreases during development from the embryo to postnatal day 1, then remains relatively stable throughout adulthood (Kinoshita et al., 1997; Peng et al., 2002; Xue et al., 2004b). Interestingly, a recent study revealed that septin expression in the hippocampus is regulated in an estrous cycle-dependent manner in female rats (Diao et al., 2007a). SEPT8 expression is higher during proestrous, which shows the highest estrogen and progesterone levels across the estrous cycle, in comparison with other phases of estrous cycle or with male rats. In contrast, SEPT3 expression increases during the metestrous phase (Diao et al., 2007a). Alterations in septins expression during the female estrous cycle may contribute to mechanisms underlying the estrous cycle-dependent changes in synapses and spine morphogenesis, and in-turn synaptic plasticity in the hippocampus (Woolley, 2007). Further studies showed that changes between early and late metestrous in the hippocampus of female rats induce SEPT5, SEPT6 and SEPT11 (Diao et al., 2007b). This suggests a correlation between septin expression and cognitive changes between early and late metestrous in rats.

SEPTIN FILAMENTS IN BRAIN Most septins have a predicted coiled-coil domain at their carboxyl termini, which has been shown to mediate the intra- and inter-molecular interaction with other

SEPTIN FILAMENTS IN BRAIN

253

septins (Sheffield et al., 2003; Versele and Thorner, 2004; Nagata et al., 2004; Low and Macara, 2006). Septins commonly form stable heteromeric complexes as functional units. Purified septins oligomerize and connect tandemly to form longer polymers or filaments, which can align laterally to form thicker polymers (bundles) and rings (Kinoshita, 2006). These filamentous structures have been recognized as cytoskeletal components associated with the actin cytoskeleton (Kinoshita et al., 1997; Kinoshita et al., 2002; Schmidt and Nichols, 2004; Kinoshita, 2006) or microtubules (Surka, Tsang and Trimble, 2002; Nagata et al., 2003; Kremer, Haystead and Macara, 2005) and act as scaffolds, recruiting proteins related to cell division (Field and Kellogg, 1999; Spiliotis, Kinoshita and Nelson, 2005), or as diffusion barriers for asymmetric protein distribution in dividing cells (Takizawa et al., 2000; Dobbelaere and Barral, 2004). The structure of the human SEPT2-SEPT6-SEPT7 complex shows that the polymer building block is bipolar– involving interactions between adjacent nucleotide binding sites or the N- and C-terminal tails (Sirajuddin et al., 2007). Surprisingly the coiled coils are not required. This is a unique filament that is fundamentally unrelated to all other cytoskeletal structures. The coiled coils of C. elegans septin filaments, comprising only two septins, extend laterally away from the core filaments (John et al., 2007). Septin complexes immunoprecipitated from mouse brain lysates with antiSEPT2 antibodies contain at least eight different septin monomers (SEPT1-2 and SEPT4-9), but the exact composition of a single complex is still unclear (Kinoshita et al., 2002). A similar group of rat brain septins (SEPT1-2, SEPT4-7) form heterocomplexes with sec6/8 complexes, the components of the exocytic machinery, suggesting a role in vesicle trafficking (Hsu et al., 1998). Indeed, SEPT5 and SEPT2 form a complex with syntaxin to negatively regulate SV exocytosis (Beites et al., 1999). SEPT5 binds to SEPT8 with a remarkably high affinity (Blaser et al., 2002). Consistent with its role in exocytosis, SEPT5 has also been shown to form a heteromeric complex with SEPT4 and SEPT8 to regulate secretion in platelets (Dent et al., 2002; Blaser et al., 2004). SEPT3 does not have the coiled-coil domain at the C-terminus. It has been found to form a septin complex via it phosphoinositide- and GTP-binding domains (Fujishima et al., 2007). Overexpressed SEPT3 in HEK293 cells (Fujishima et al., 2007) or cultured cerebellar granule neurons (Xue and Robinson, unpublished observation) forms filamentous aggregates that are independent of actin or microtubules, although no endogenous SEPT3 filament has yet been found at the level of fluorescence microscopy in neurons. Immunoprecipitation studies show that SEPT3 form a heteromeric complex with SEPT5 and SEPT7 in the brain. Consistently, overexpression of SEPT3 recruits SEPT5 and SEPT7 to filamentous aggregates in the somatodendritic area (Fujishima et al., 2007). Quick-freeze deep-etch EM studies of the presynaptic terminal has revealed filament-like strands linking vesicles to each other as well as strands between SVs and the plasma membrane (Hirokawa et al., 1989). Synapsin I, having a tadpole-like structure, appears to form short strands that link SVs to actin filaments and microtubules. However, the long strands that link SVs to plasma membrane

254

CH11 SEPTINS AND THE SYNAPSE

are still yet to be defined. Septins have been suggested to be attractive candidates for the protein composition of these filaments (Beites, Campbell and Trimble, 2004). In studies involving the giant synaptic terminals of goldfish retinal bipolar neurons, newly arrived vesicles from the reserve pools stop approximately 20 nm away from the cell membrane (Zenisek, Steyer and Almers, 2000), suggesting that these vesicles are physically restricted from the plasma membrane. Interestingly, septin filaments isolated from rat brain have a subunit size of 8.25 nm and filament lengths in multiples of 25 nm (Hsu et al., 1998) and these septin complexes interact with exocyst proteins that target vesicles to the plasma membrane (see reviews in Hsu et al. (1999); Hsu et al. (2004)). These septin filaments may bridge the space between SVs and plasma membrane, confining the movement of the vesicles until the proper releasing signal occurs. Additionally, SEPT5 co-localizes with the syntaxin in PC12 cells (Beites et al., 1999). Immunogold labelling of mice brain terminals reveals the presence of SEPT5 around SVs near the presynaptic terminals indicating that SEPT5 might be the linkage between the vesicles and the plasma membrane (Kinoshita, Noda and Kinoshita, 2000). In a study using chromaffin cells, the movement of granules towards and away from the plasma membrane was analysed (Johns et al., 2001). As granules approached the membrane, they moved more slowly. This restriction in movement was not due to the actin cytoskeleton, VAMP-2 or SNAP25, since treatment of the cells with Latrunculin A, tetanus toxin or botulinum toxin, respectively, caused little change in the restricted motion of granules toward the plasma membrane. This restriction may be controlled by septin filaments as SEPT5 binds to the SNARE complex composed of syntaxin, SNAP-25 and VAMP-2. Our unpublished data showed that SEPT3 is a binding partner for synapsin I (Xue and Robinson, unpublished data), suggesting that septin filaments may be part of the filamentous strands that are linking the SVs to the synaptic membrane and may act as diffusion barriers in order to keep SVs constrained to sites of vesicle release. Therefore, understanding how the formation of septin macromolecular complexes is temporally regulated during septin assembly is an important step in understanding septin function in neurons.

SYNAPTIC FUNCTIONS OF SEPTINS Vesicle secretion The identification of a direct interaction between a septin complex and the sec6/8 complex (also called the exocyst) provided the first insight into septin roles in polarized vesicle secretion (Hsu et al., 1998). Since then, SEPT5 has been found to interact with various SV markers, SNAP-25 and synaptophysin in the brain (Caltagarone et al., 1998). The link between septins and neurotransmitter release became stronger when SEPT5 was demonstrated to bind the SNARE protein, syntaxin and regulate vesicle secretion in neuroendocrine cells (Beites et al.,

SYNAPTIC FUNCTIONS OF SEPTINS

255

1999). Overexpression of SEPT5 into HIT-T15 cells inhibits the release of human growth hormone, while dominant-negative SEPT5 enhances secretion, suggesting that Sept5 is a negative regulator of vesicle release (Beites et al., 1999). However, SEPT5–/–mice appear normal with respect to neurotransmitter release and neuronal development (Peng et al., 2002). The functional redundancy of SEPT5 may be accounted for by the change in expression of other septins in the knockout (KO) mice. Nevertheless, platelets from SEPT5 KO mice resulted in an enhanced platelet secretion response, confirming the negative regulatory role of SEPT5 in exocytosis (Dent et al., 2002). SEPT5 may tether the vesicle to the plasma membrane or act directly as a physical barrier, hence restricting the movement of vesicles towards the plasma membrane. Since the interaction of α-SNAP and SEPT5 to the 7S complex (VAMP, SNAP-25 and syntaxin) is mutually exclusive, α-SNAP may then displace SEPT5 from the complex, priming the vesicle fusion event (Beites, Campbell and Trimble, 2004). Clearly these studies linking septins to exocytosis are tantalizing but incomplete.

Neurite outgrowth A number of septins are found in postmitotic neurons and have been shown to co-localize and/or interact with proteins required for neurite outgrowth or axon pathfinding. SEPT2 co-immunoprecipitates with the other septins, tubulin and the exocyst complex which is required for neurite outgrowth from rat brain lysates (Hsu et al., 1998; Vega and Hsu, 2001; Vega and Hsu, 2003). Endogenous SEPT2 is enriched at the perinuclear region in undifferentiated PC12 cells and radiates outward toward the growth cone, upon neuronal differentiation in NGF-induced PC12 cells (Vega and Hsu, 2003). Overexpression of GTPase-defective SEPT2 mutant results in excessive and aberrant neurite sprouting in differentiating PC12 cells. This is in contrast with the effect observed upon overexpressing the exocyst complex subunit sec10 mutant, which abolishes neurite outgrowth in differentiating PC12 cells (Vega and Hsu, 2001). These suggest that SEPT2 is required for neurite sprouting, rather than promoting neurite outgrowth in PC12 cells undergoing neuronal differentiation, perhaps by guiding the exocyst and microtubule redistribution towards designated plasma membrane domains to facilitate polarized neurite outgrowth (Vega and Hsu, 2003). A mitochondrial septin, M-septin, is a major alternatively spliced variant of the Sept4 gene in developing mouse brain and its expression is up-regulated during the neuronal differentiation of embryonal carcinoma P19 stem cells (Takahashi et al., 2003). M-septin interacts with the complex of collapsing response mediator protein (CRMP) and CRMP-associated molecules (CRAM), which are implicated in axon guidance and outgrowth during neuronal development (Byk et al., 1996; Inagaki et al., 2001; Takahashi et al., 2003). M-septin also partially co-localizes with CRAM in neurites and growth cones of dorsal root ganglion (DRG) neurons and differentiating P19 cells (Takahashi et al., 2003).

256

CH11 SEPTINS AND THE SYNAPSE

In addition, overexpression of M-septin induces cytoplasmic CRAM translocation to mitochondria. Thus M-septin, together with CRMP/CRAM, may play an important role in the neuronal differentiation and axon guidance through the control of mitochondrial function. A role of septins in neurite outgrowth and axonal migration has also been observed in C. elegans. UNC-59 and UNC-61 are the only two septins found in worms (Nguyen et al., 2000). Although unc-59 and unc-61 mutants were initially characterized as having normal embryonic development, defects were observed in post-embryonic development affecting the morphogenesis of the vulva, male tail (including its sensory neurons), and gonad (Nguyen et al., 2000). Subsequently, the worm septin mutants were found to display defects in locomotory behaviour due to embryonic defects in axonal migration and guidance of motor neurons (Finger, Kopish and White, 2003). Many gene products required for axonal migration are also required for the migration of distal tip cells (Hedgecock et al., 1987), which are leaders for gonad arms extension (Lehmann, 2001). Septins are expressed in migrating distal tip cells and septin mutants affect morphology of the distal tip cells, as well as their migration and guidance during gonadogenesis (Finger, Kopish and White, 2003). Taken together, septins may be generally required for developmental migration and pathfinding by neurites (Finger, Kopish and White, 2003).

Motor performance Proteomic analysis has revealed reduced expression of at least 50 proteins in the cerebellum of an inbred mouse strain 129X1/SvJ which exhibits poor performance on the rotarod test, the standard test for motor coordination, in comparison with two other inbred strains, C57BL/6J and nNOS WT (Gerlai, 2001; Pollak et al., 2005). These include SEPT5, syntaxin binding protein 1 and the calcium binding protein, calbindin (Pollak et al., 2005). A strong association of impaired motor function with altered calbindin expression is in agreement with previous observations of motor deficiencies in a calbindin KO mouse (Airaksinen et al., 1997). Reduction in SEPT5 occurs in parallel with that of syntaxin binding protein 1, another binding protein for syntaxin, thought to control membrane fusion by providing a platform for SNARE complex assembly (Misura, Scheller and Weis, 2000). Altogether, reduction of these three proteins may well indicate partial disruption of the exocytic machinery in the cerebellum of 129X1/SvJ mice and may partly account for the observed deficits in motor performance.

Astrocytes SEPT2 is ubiquitously expressed and in the brain it is found mainly in the Bergmann glial cells, which are essential for the maturation and survival of Purkinje cells in the cerebellum (Kinoshita et al., 1997; Kinoshita et al., 2004). It

PHOSPHORYLATION OF SEPTINS

257

co-localizes with GLAST, an astrocyte glutamate transporter, in the brain, and directly interacts with GLAST in a GDP-bound dependent manner (Kinoshita et al., 2004). Expression of constitutive GDP-bound SEPT2 mutant promotes the internalization of surface GLAST hence reducing glutamate uptake activity in Bergmann glial cells (Kinoshita et al., 2004). Interestingly, SEPT4 also co-immunoprecipitates with SEPT2 and GLAST, suggesting that SEPT2 may form complexes with other septins to assemble into filaments and regulate GLASTmediated glutamate uptake by astrocytes, which is important for appropriate neurotransmitter signalling in the cerebellum (Kinoshita et al., 2004).

PHOSPHORYLATION OF SEPTINS Phosphorylation is one of the fundamental mechanisms governing a plethora of cellular functions in cells, including SV trafficking, synaptic transmission and synaptic plasticity. Abnormal phosphorylation of synaptic proteins often leads to the pathology of several neurodegenerative diseases; most notably being the AD and the PD. Hyperphosphorylation of microtubule-associated tau protein by cyclin-dependent kinase 5 (Cdk5) has been linked to the formation of neurofibrillary tangles (NFTs) associated with AD (reviewed in Tsai, Lee and Cruz (2004)). Phosphorylation of Aβ peptides by Cdk5 also promotes their aggregation, leading to deposition of amyloid plaque, another hallmark of AD pathology (Tsai, Lee and Cruz, 2004). Phosphorylation of α-synuclein enhances its self-aggregation in vitro, which may be related to the formation of Lewy bodies in sporadic PD (Fujiwara et al., 2002). In budding yeast, septin phosphorylation by Cla4 protein kinase is essential for septin filament assembly and the formation of septin collar during emergence of the bud and cytokinesis (Versele and Thorner, 2004). Additionally in yeast, phosphorylation of Bni5 controls its interaction with cdc11 and mediates its delocalization from septin filaments at late mitosis, thereby regulating cytokinesis (Nam et al., 2007). SEPT3 is a specific substrate for cGMP-dependent protein kinase (PKG) in vitro (Xue et al., 2000). It is constitutively phosphorylated on Ser-91 by PKG in intact nerve terminals and is increased in response to stimulation with cGMP analogues (Xue et al., 2004a). While SEPT3 is mostly associated with the plasma membrane in nerve terminals, phospho-SEPT3 is exclusively cytosolic, suggesting phosphorylation might induce translocation (Xue et al., 2004a). Therefore SEPT3 phosphorylation on Ser-91 may play a key role in nerve terminals by regulating its subcellular localization during stimulation (Xue et al., 2004a). Interestingly, increased PKG activity facilitates the induction of long-term potentiation (LTP) in cultured hippocampal CA1 neurons (Son et al., 1998; Arancio et al., 2001), where SEPT3 immunoreactivity is particularly high (Xue et al., 2004b). This forms a potentially interesting link between SEPT3, cGMP and synaptic plasticity in the hippocampus and future studies should evaluate if SEPT3 might be involved in some of these plasticity changes.

258

CH11 SEPTINS AND THE SYNAPSE

SEPT2 is phosphorylated by protein kinase C (PKC) in vitro and by cAMPdependent protein kinase (PKA) to a lesser extends (Xue et al., 2000). It is also shown to be phosphorylated by casein kinase II (CK2) on Ser-218 as revealed by mass spectrometry analysis of purified SEPT2 from Sf9 cells (She et al., 2004). (Note this site was initially called Ser-248 due to an N-terminal tag (Huang et al., 2006)). Targeted mutation of Ser-218 to alanine residue abrogates its phosphorylation (She et al., 2004). Ser-218 in SEPT2 was shown to be phosphorylated by CK2 in vivo which significantly affected SEPT2 nucleotide binding affinity in vitro (Huang et al., 2006). These results suggest that phosphorylation of SEPT2 may regulate septin assembly through effects on nucleotide binding. SEPT5 is phosphorylated in the N-terminal polybasic region in response to agonist stimulation such as thrombin, collagen or phorbol 12-myristate 13-acetate in platelets (Dent et al., 2002). It was thought that SEPT5 is phosphorylated by PKC since PKC inhibitor, Ro31-8220, inhibits its phosphorylation (Dent et al., 2002). However, Ro31-8220 inhibits the activity of Cdk5 with much higher potency than inhibiting PKC activity (Tan et al., 2003) and potently inhibits GSK3. Indeed, SEPT5 v1 has recently been shown to be phosphorylated by Cdk5 on Ser-17 both in vitro and in mouse brain (Taniguchi et al., 2007). SEPT5 v2 is also phosphorylated by Cdk5 at Ser-327. Although the phosphorylation sites are different, both phosphorylations reduce binding to syntaxin (Taniguchi et al., 2007). These results indicate the role of SEPT5 in exocytosis is modulated by Cdk5 phosphorylation in the brain. Large scale phosphoproteomic studies in mouse liver identified multiple phosphorylation sites in vivo for SEPT2, 4, 5, 8 and 9 (Villen et al., 2007). This must still be considered as a preliminary and partial list that requires independent validation. Interestingly, all these sites were in an acidic amino acid environment indicative of a CK2 specificity.

SEPTINS AND NEURODEGENERATIVE DISEASE Alzheimer’s disease (AD) At least three septins, SEPT1, SEPT2 and SEPT4 are found in NFTs in postmortem human brain tissues from AD patients (Kinoshita et al., 1998). Double immunofluorescence staining shows largely overlapping staining for SEPT4 and SEPT1 in the hippocampus and entorhinal cortex of AD brains, but the septins do not always co-localize in the same cells. A small population of neurons that are SEPT1-positive, are SEPT4-negative. In NFTs, septins co-localize with tau protein, which is a major constituent of NFTs. Similarly, most NFTs contain both tau and one of the septins, but a small number of neurons contain either tau or SEPT1. These SEPT1-positive, phospho-tau-negative NFTs are potentially very interesting because they indicate for the first time that some NFTs might not contain hyperphosphorylated tau, thus such septin accumulation in NFTs might be a previously unrecognized component of neurodegeneration. Interestingly, SEPT7

SEPTINS AND NEURODEGENERATIVE DISEASE

259

is present in post-mortem human brain yet does not localize to the NFTs but is down-regulated in a presenilin 1-hypomorphic mouse line, a mouse model for familial AD (Liauw et al., 2002). Another study has also linked the polymorphism of SEPT3 gene with the susceptibility of the brain to neurodegenerative changes specific for AD, but not for PD, suggesting a determinative role in the pathogenesis of AD (Takehashi et al., 2004a). SEPT4 may accumulate in two types of amyloid filaments–tau-based filaments in AD and cytoplasmic inclusions in PD. SEPT4 was recently shown to undergo temperature induced folding transitions to a form rich in beta sheets that have amyloid properties and form filaments (Garcia et al., 2007). The ability of SEPT4 to form fibrillary aggregates may be of significance to the amyloidosis that characterizes these diseases.

Parkinson’s disease (PD) SEPT4, but not five other septins (SEPT2, SEPT5, SEPT6, SEPT7 and SEPT8), co-aggregates with α-synuclein in cytoplasmic inclusion bodies known as Lewy bodies in sporadic PD and dementia with Lewy bodies (Ihara et al., 2003). SEPT4 co-immunoprecipitates with α-synuclein, and when co-expressed in culture cells, they form detergent-insoluble complex and induces proteasome inhibitor-induced cell death (Ihara et al., 2003). Interestingly, SEPT4 co-localizes with the DAT and α-synuclein, suggesting physiological association of SEPT4 with these presynaptic proteins in DA neurons (Ihara et al., 2007). Mice lacking SEPT4 exhibit diminished DA neurotransmission due to scarcity of these presynaptic proteins (Ihara et al., 2007). These data demonstrate an important role for septin scaffolds in the brain. In transgenic mice expressing human α-synucleinA53T , a mutant protein responsible for familial PD, loss of SEPT4 significantly enhances neuropathology and locomotor deterioration (Ihara et al., 2007). In this PD model, insoluble deposits of Ser-129-phosphorylated α-synucleinA53T are negatively correlated with the dosage of SEPT4. SEPT4 directly protects α-synuclein against self-aggregation and Ser-129 phosphorylation in vitro. Together, these data show that SEPT4 may be involved in PD as a dual susceptibility factor, as its insufficiency can diminish DA neurotransmission and enhance α-synuclein neurotoxicity (Ihara et al., 2007). SEPT5 is a known substrate for a ubiquitin ligase, parkin (Zhang et al., 2000). Mutations in parkin gene cause autosomal recessive juvenile parkinsonism (ARJP) (reviewed in Mizuno et al. (2001)). Parkin interacts with the E2 ubiquitinconjugating enzyme and has E3 ubiquitin-protein ligase activity, which covalently attaches ubiquitin moieties onto substrate proteins destined for proteasomal degradation. The human SEPT5 splice variant that is most abundant in developing neurons, SEPT5 v2, is both a target for parkin-mediated ubiquitination and a parkin binding protein (Choi et al., 2003). Both SEPT5 v1 and SEPT5 v2 accumulate in substantia nigra region of brains of AR-JP patients, suggesting that an important relationship exists between parkin and septins (Choi et al., 2003). Loss of Parkin enzymatic activity results in aberrant accumulation of its specific

260

CH11 SEPTINS AND THE SYNAPSE

target substrates that leads to DA dysfunction and cell death in AR-JP (McNaught et al., 2001; Dawson and Dawson, 2003). Acute accumulation of elevated levels of SEPT5 in well characterized DA SN4741 cell causes cytotoxic cell death that can be rescued by overexpressing functional parkin (Son et al., 2005). A transgenic mouse model expressing a parkin dominant-negative mutant shows moderately increased levels of SEPT5 in substantia nigra DA neurons (Son et al., 2005). The mice develop DOPA responsive motor dysfunction but no neuronal loss, suggesting a role for SEPT5 in the motor deficits of the disease. Parkin KO mice are another PD model that mimics the presymptomatic disease but not the neuronal degeneration. SEPT5 and 7 are specifically up-regulated in the striatum of these mice, suggesting they might compensate to protect the DA neurons in these mice (Periquet et al., 2005). In Drosophila two septins are also substrates for the fly parkin and together they appear to regulate p53-inducedd apoptosis (Bae, Kang and Park, 2007). This suggests a mechanism for loss of DA function. SEPT5 overexpression in brain has been found to exert dopamine-dependent neurotoxicity, suggesting that inhibition of dopamine secretion by SEPT5 may contribute to the early development of AR-JP (Dong et al., 2003).

Other neurological disorders While septins have been implicated in the pathogenesis of PD and AD, recent advances in proteomic or genetic analyses of patients or post-mortem brain samples also link septins to the etiology of several human neurodegenerative disorders such as epilepsy, schizophrenia, Down’s syndrome and various human brain tumours. SEPT3 expression is increased in the hippocampus from a mesial temporal lobe epilepsy patient (Yang et al., 2006), while other septins (SEPT3-7, SEPT9 and SEPT11) are found in a variety of astrocytoma and medulloblastoma cell lines, and several solid brain tumour specimens (Kim et al., 2004). In addition, SEPT2 and SEPT5 are also found early on in fetal brain with Down’s syndrome (Cheon et al., 2001). Three mutations have been identified in SEPT9 gene in six families with hereditary neuralgic amyotrophy, which exhibits painful brachial plexus neuropathies and muscle atrophy; making it the first monogenetic disease caused by mutations in a gene of the septin family (Kuhlenbaumer et al., 2005). Moreover, SEPT5 gene is mapped to chromosome 22q11.2, a region commonly deleted in DiGeorge syndrome and velo-cardio-facial syndromes, which are associated with increased prevalence of schizophrenia (McKie et al., 1997). Indeed, SEPT5 expression is lower in the striatum of isolation-reared animals, of which behaviours are reminiscent of schizophrenic patient (Stopkova et al., 2004). Taken together, abnormal expression of septins, which have functions ranging from signalling to cellular scaffolding, could contribute to or reflect brain dysgenesis in various neurological disorders.

REFERENCES

261

CONCLUSIONS The septins are a family of GTPases involved in diverse cellular processes including cytokinesis, remodelling the cytoskeleton, apoptosis and vesicle trafficking. Most septins are present in brain and they form a heterocomplex which assembles end-to-end as filaments in nerve terminals. The large number of septin proteins are likely to be functionally redundant, as targeted deletion of SEPT3, SEPT5 or SEPT6 show no obvious CNS abnormalities (McNaught et al., 2001; Peng et al., 2002; Ono et al., 2005), except SEPT4, which specifically affects only DA neurons (Ihara et al., 2007). Although septins have been widely implicated in various neurodegenerative disorders, it is still not clear whether septins play a causative role or are a result of disease manifestation. Based on the current literature, a theme has emerged for a role of septin as scaffolding proteins that serve as diffusion barriers for exocytosis in presynaptic nerve terminals. Future studies of the complexity of septin biology will facilitate even more understanding of the role of septins in the physiology and pathology of the brain.

REFERENCES Airaksinen, M.S., Eilers, J., Garaschuk, O. et al. (1997) Ataxia and altered dendritic calcium signaling in mice carrying a targeted null mutation of the calbindin D28k gene. Proceedings of the National Academy of Sciences of the United States of America, 94, 1488–93. Arancio, O., Antonova, I., Gambaryan, S. et al. (2001) Presynaptic role of cGMPdependent protein kinase during long-lasting potentiation. Journal of Neuroscience, 21, 143–49. Bae, Y.J., Kang, S.J. and Park, K.S. (2007) Drosophila melanogaster Parkin ubiquitinates peanut and septin1 as an E3 ubiquitin-protein ligase. Insect Biochemistry and Molecular Biology, 37, 430–39. Barr, A.M., Young, C.E., Sawada, K. et al. (2004) Abnormalities of presynaptic protein CDCrel-1 in striatum of rats reared in social isolation: relevance to neural connectivity in schizophrenia. European Journal of Neuroscience, 20, 303–7. Beites, C.L., Campbell, K.A. and Trimble, W.S. (2004) The septin Sept5/CDCrel-1 competes with alpha-SNAP for binding to the SNARE complex. Biochemical Journal , 385, 347–53. Beites, C.L., Xie, H., Bowser, R. and Trimble, W.S. (1999) The septin CDCrel-1 binds syntaxin and inhibits exocytosis. Nature Neuroscience, 2, 434–39. Blaser, S., Horn, J., Wurmell, P. et al. (2004) The novel human platelet septin SEPT8 is an interaction partner of SEPT4. Thrombosis and Haemostasis, 91, 959–66. Blaser, S., Jersch, K., Hainmann, I. et al. (2002) Human septin-septin interaction: CDCrel-1 partners with KIAA0202. FEBS Letters, 519, 169–72. Blaser, S., Jersch, K., Hainmann, I. et al. (2003) Isolation of new splice isoforms, characterization and expression analysis of the human septin SEPT8 (KIAA0202). Gene, 312, 313–20.

262

CH11 SEPTINS AND THE SYNAPSE

Byk, T., Dobransky, T., Cifuentes-Diaz, C. and Sobel, A. (1996) Identification and molecular characterization of Unc-33-like phosphoprotein (Ulip), a putative mammalian homolog of the axonal guidance-associated unc-33 gene product. Journal of Neuroscience, 16, 688–701. Caltagarone, J., Rhodes, J., Honer, W.G. and Bowser, R. (1998) Localization of a novel septin protein, hCDCrel-1, in neurons of human brain. Neuroreport, 9, 2907– 12. Cheon, M.S., Fountoulakis, M., Dierssen, M. et al. (2001) Expression profiles of proteins in fetal brain with Down syndrome. Journal of Neural Transmission. Supplementum, 61, 311–19. Choi, P., Snyder, H., Petrucelli, L. et al. (2003) SEPT5 v2 is a parkin-binding protein. Brain Research. Molecular Brain Research, 117, 179–89. Dawson, T.M. and Dawson, V.L. (2003) Molecular pathways of neurodegeneration in Parkinson’s disease. Science, 302, 819–22. Dent, J., Kato, K., Peng, X.-R. et al. (2002) A prototypic platelet septin and its participation in secretion. Proceedings of the National Academy of Sciences of the United States of America, 99, 3064–69. Diao, W.F., Hoger, H., Chen, W.Q. et al. (2007a) Estrous-cycle-dependent hippocampal levels of signaling proteins. Hippocampus, 17, 563–76. Diao, W.F., Hoger, H., Chen, W.Q. et al. (2007b) Hippocampal signaling protein levels are different in early and late metestrus in the rat. Biochimica et Biophysica Acta, 1774, 1044–51. Dobbelaere, J. and Barral, Y. (2004) Spatial coordination of cytokinetic events by compartmentalization of the cell cortex. Science, 305, 393–96. Dong, Z., Ferger, B., Paterna, J.C. et al. (2003) Dopamine-dependent neurodegeneration in rats induced by viral vector-mediated overexpression of the parkin target protein, CDCrel-1. Proceedings of the National Academy of Sciences of the United States of America, 100, 12438–43. Field, C.M. and Kellogg, D. (1999) Septins: cytoskeletal polymers or signalling GTPases? Trends in Cell Biology, 9, 387–94. Finger, F.P., Kopish, K.R. and White, J.G. (2003) A role for septins in cellular and axonal migration in C. elegans. Developmental Biology, 261, 220–34. Fujishima, K., Kiyonari, H., Kurisu, J. et al. (2007) Targeted disruption of Sept3, a heteromeric assembly partner of Sept5 and Sept7 in axons, has no effect on developing CNS neurons. Journal of Neurochemistry, 102, 77–92. Fujiwara, H., Hasegawa, M., Dohmae, N. et al. (2002) [alpha]-Synuclein is phosphorylated in synucleinopathy lesions. Nature Cell Biology, 4, 160–64. Gaiarsa, J.L., Caillard, O. and Ben Ari, Y. (2002) Long-term plasticity at GABAergic and glycinergic synapses: mechanisms and functional significance. Trends in Neurosciences, 25, 564–70. Garcia W., de Ara´ujo AP., Lara F., Foguel D., Tanaka M., Tanaka T., Garratt RC. (2007) An intermediate structure in the thermal unfolding of the GTPase domain of human septin 4 (SEPT4/Bradeion-beta) forms amyloid-like filaments in vitro. Biochemistry. 46, 11101–9. Gerlai, R. (2001) Gene targeting: technical confounds and potential solutions in behavioral brain research. Behavioural Brain Research, 125, 13–21. Hall, P.A., Jung, K., Hillan, K.J. and Russell, S.H. (2005) Expression profiling the human septin gene family. The Journal of Pathology, 206, 269–78.

REFERENCES

263

Hedgecock, E.M., Culotti, J.G., Hall, D.H. and Stern, B.D. (1987) Genetics of cell and axon migrations in Caenorhabditis elegans. Development, 100, 365–82. Hirokawa, N., Sobue, K., Kanda, K. et al. (1989) The cytoskeletal architecture of the presynaptic terminal and molecular structure of synapsin 1. Journal of Cell Biology, 108, 111–26. Honer, W.G., Hu, L. and Davies, P. (1993) Human synaptic proteins with a heterogeneous distribution in cerebellum and visual cortex. Brain Research, 609, 9–20. Hsu, S.C., Hazuka, C.D., Foletti, D.L. and Scheller, R.H. (1999) Targeting vesicles to specific sites on the plasma membrane: the role of the sec6/8 complex. Trends in Cell Biology, 9, 150–53. Hsu, S.C., Hazuka, C.D., Roth, R. et al. (1998) Subunit composition, protein interactions, and structures of the mammalian brain sec6/8 complex and septin filaments. Neuron, 20, 1111–22. Hsu, S.C., TerBush, D., Abraham, M. and Guo, W. (2004) The exocyst complex in polarized exocytosis. International Review of Cytology, 233, 243–65. Huang, Y.W., Surka, M.C., Reynaud, D. et al. (2006) GTP binding and hydrolysis kinetics of human septin 2. FEBS Letters, 273, 3248–60. Ihara, M., Tomimoto, H., Kitayama, H. et al. (2003) Association of the cytoskeletal GTP-binding protein Sept4/H5 with cytoplasmic inclusions found in Parkinson’s disease and other synucleinopathies. Journal of Biological Chemistry, 278, 24095–102. Ihara, M., Yamasaki, N., Hagiwara, A. et al. (2007) Sept4, a component of presynaptic scaffold and lewy bodies, is required for the suppression of alpha-synuclein neurotoxicity. Neuron, 53, 519–33. Inagaki, N., Chihara, K., Arimura, N. et al. (2001) CRMP-2 induces axons in cultured hippocampal neurons. Nature Neuroscience, 4, 781–82. John, C.M., Hite, R.K., Weirich, C.S. et al. (2007) The Caenorhabditis elegans septin complex is nonpolar. The EMBO Journal , 26, 3296–307. Johns, L.M., Levitan, E.S., Shelden, E.A. et al. (2001) Restriction of secretory granule motion near the plasma membrane of chromaffin cells. Journal of Cell Biology, 153, 177–90. Joo, E., Tsang, C.W. and Trimble, W.S. (2005) Septins: traffic control at the cytokinesis intersection. Traffic, 6, 626–34. Kim, D.S., Hubbard, S.L., Peraud, A. et al. (2004) Analysis of mammalian septin expression in human malignant brain tumors. Neoplasia, 6, 168–78. Kinoshita, M. (2006) Diversity of septin scaffolds. Current Opinion in Cell Biology, 18, 54–60. Kinoshita, M., Field, C.M., Coughlin, M.L. et al. (2002) Self- and actin-templated assembly of mammalian septins. Developmental Cell , 3, 791–802. Kinoshita, N., Kimura, K., Matsumoto, N. et al. (2004) Mammalian septin Sept2 modulates the activity of GLAST, a glutamate transporter in astrocytes. Genes to Cells, 9, 1–14. Kinoshita, A., Kinoshita, M., Akiyama, H. et al. (1998) Identification of septins in neurofibrillary tangles in Alzheimer’s disease. American Journal of Pathology, 153, 1551–60. Kinoshita, M., Kumar, S., Mizoguchi, A. et al. (1997) Nedd5, a mammalian septin, is a novel cytoskeletal component interacting with actin-based structures. Genes and Development, 11, 1535–47.

264

CH11 SEPTINS AND THE SYNAPSE

Kinoshita, A., Noda, M. and Kinoshita, M. (2000) Differential localization of septins in the mouse brain. The Journal of Comparative Neurology, 428, 223–39. Kremer, B.E., Haystead, T. and Macara, I.G. (2005) Mammalian septins regulate microtubule stability through interaction with the microtubule-binding protein MAP4. Molecular Biology of the Cell , 16, 4648–59. Kuhlenbaumer, G., Hannibal, M.C., Nelis, E. et al. (2005) Mutations in SEPT9 cause hereditary neuralgic amyotrophy. Nature Genetics, 37, 1044–46. Lehmann, R. (2001) Cell migration in invertebrates: clues from border and distal tip cells. Current Opinion in Genetics and Development, 11, 457–63. Liauw, J., Nguyen, V., Huang, J. et al. (2002) Differential display analysis of presenilin 1-deficient mouse brains. Brain Research. Molecular Brain Research, 109, 56–62. Lindsey, R. and Momany, M. (2006) Septin localization across kingdoms: three themes with variations. Current Opinion in Microbiology, 9, 559–65. Low, C. and Macara, I.G. (2006) Structural analysis of septin 2, 6 and 7 complexes. Journal of Biological Chemistry, 281, 30697–706. Martinez, C., Corral, J., Dent, J.A. et al. (2006) Platelet septin complexes form rings and associate with the microtubular network. Journal of Thrombosis and Haemostasis, 4, 1388–95. McKie, J.M., Sutherland, H.F., Harvey, E. et al. (1997) A human gene similar to Drosophila melanogaster peanut maps to the DiGeorge syndrome region of 22q11. Human Genetics, 101, 6–12. McNaught, K.S., Olanow, C.W., Halliwell, B. et al. (2001) Failure of the ubiquitinproteasome system in Parkinson’s disease. Nature Reviews. Neuroscience, 2, 589–94. Methner, A., Leypoldt, F., Joost, P. and Lewerenz, J. (2001) Human septin 3 on chromosome 22q13.2 is upregulated by neuronal differentiation. Biochemical and Biophysical Research Communications, 283, 48–56. Misura, K.M., Scheller, R.H. and Weis, W.I. (2000) Three-dimensional structure of the neuronal-Sec1-syntaxin 1a complex. Nature, 404, 355–62. Mizuno, Y., Hattori, N., Mori, H. et al. (2001) Parkin and Parkinson’s disease. Current Opinion in Neurology, 14, 477–82. Nagata, K.I., Asano, T., Nozawa, Y. and Inagaki, M. (2004) Biochemical and cell biological analyses of a mammalian septin complex, Sept7/9b/11. Journal of Biological Chemistry, 279, 55895–904. Nagata, K., Kawajiri, A., Matsui, S. et al. (2003) Filament formation of MSF-A, a mammalian septin, in human mammary epithelial cells depends on interactions with microtubules. Journal of Biological Chemistry, 278, 18538–43. Nam, S.C., Sung, H., Kang, S.H. et al. (2007) Phosphorylation-dependent septin interaction of Bni5 is important for cytokinesis. Journal of Microbiology, 45, 227–33. Nguyen, T.Q., Sawa, H., Okano, H. and White, J.G. (2000) The C. elegans septin genes, unc-59 and unc-61 , are required for normal postembryonic cytokineses and morphogenesis but have no essential function in embryogenesis. Journal of Cell Science, 113, 3825–37. Ono, R., Ihara, M., Nakajima, H. et al. (2005) Disruption of Sept6, a fusion partner gene of MLL, does not affect ontogeny, leukemogenesis induced by MLL-SEPT6,

REFERENCES

265

or phenotype induced by the loss of Sept4. Molecular and Cellular Biology, 25, 10965–78. Peng, X.-R., Jia, Z., Zhang, Y. et al. (2002) The septin CDCrel-1 is dispensable for normal development and neurotransmitter release. Molecular and Cellular Biology, 22, 378–87. Periquet, M., Corti, O., Jacquier, S. and Brice, A. (2005) Proteomic analysis of parkin knockout mice: alterations in energy metabolism, protein handling and synaptic function. Journal of Neurochemistry, 95, 1259–76. Phillips, G.R., Huang, J.K., Wang, Y. et al. (2001) The presynaptic particle web: ultrastructure, composition, dissolution, and reconstitution. Neuron, 32, 63–77. Pollak, D., Weitzdoerfer, R., Yang, Y.W. et al. (2005) Cerebellar protein expression in three different mouse strains and their relevance for motor performance. Neurochemistry International , 46, 19–29. Powell, K.A. and Robinson, P.J. (1995) Dephosphin/dynamin is a neuronal phosphoprotein concentrated in nerve terminals: evidence from rat cerebellum. Neuroscience, 64, 821–33. Schmidt, K. and Nichols, B.J. (2004) Functional interdependence between septin and actin cytoskeleton. BMC Cell Biology, 5, 43. She, Y.M., Huang, Y.W., Zhang, L. and Trimble, W.S. (2004) Septin 2 phosphorylation: theoretical and mass spectrometric evidence for the existence of a single phosphorylation site in vivo. Rapid Communications in Mass Spectrometry, 18, 1123–30. Sheffield, P.J., Oliver, C.J., Kremer, B.E. et al. (2003) Borg/septin interactions and the assembly of mammalian septin heterodimers, trimers and filaments. Journal of Biological Chemistry, 278, 3483–88. Shimohama, S., Fujimoto, S., Sumida, Y. et al. (1998) Differential expression of rat brain synaptic proteins in development and aging. Biochemical and Biophysical Research Communications, 251, 394–98. Sirajuddin, M., Farkasovsky, M., Hauer, F. et al. (2007) Structural insight into filament formation by mammalian septins. Nature, 449, 311–15 (advance online publication 18 July 2007, doi:10.1038/nature06052). Son, J.H., Kawamata, H., Yoo, M.S. et al. (2005) Neurotoxicity and behavioral deficits associated with Septin 5 accumulation in dopaminergic neurons. Journal of Neurochemistry, 94, 1040–53. Son, H., Lu, Y.F., Zhuo, M. et al. (1998) The specific role of cGMP in hippocampal LTP. Learning and Memory, 5, 231–45. Spiliotis, E.T., Kinoshita, M. and Nelson, W.J. (2005) A mitotic septin scaffold required for mammalian chromosome congression and segregation. Science, 307, 1781–85. Stopkova, P., Vevera, J., Paclt, I. et al. (2004) Analysis of SYNJ1, a candidate gene for 21q22 linked bipolar disorder: a replication study. Psychiatry Research, 127, 157–61. Surka, M.C., Tsang, C.W. and Trimble, W.S. (2002) The mammalian septin MSF localizes with microtubules and is required for completion of cytokinesis. Molecular Biology of the Cell , 13, 3532–45. Takehashi, M., Alioto, T., Stedeford, T. et al. (2004a) Septin 3 gene polymorphism in Alzheimer’s disease. Gene Expression, 11, 263–70.

266

CH11 SEPTINS AND THE SYNAPSE

Takahashi, S., Inatome, R., Yamamura, H. and Yanagi, S. (2003) Isolation and expression of a novel mitochondrial septin that interacts with CRMP/CRAM in the developing neurones. Genes to Cells, 8, 81–93. Takehashi, M., Tanaka, S., Stedeford, T. et al. (2004b) Expression of septin 3 isoforms in human brain. Gene Expression, 11, 271–78. Takizawa, P.A., DeRisi, J.L., Wilhelm, J.E. and Vale, R.D. (2000) Plasma membrane compartmentalization in yeast by messenger RNA transport and a septin diffusion barrier. Science, 290, 341–44. Tan, T.C., Valova, V.A., Malladi, C.S. et al. (2003) Cdk5 is essential for synaptic vesicle endocytosis. Nature Cell Biology, 5, 701–10. Taniguchi, M., Taoka, M., Itakura, M. et al. (2007) Phosphorylation of adult type Sept5 (CDCrel-1) by cyclin-dependent kinase 5 inhibits interaction with syntaxin-1. Journal of Biological Chemistry, 282, 7869–76. Toda, S., Kajii, Y., Sato, M. and Nishikawa, T. (2000) Reciprocal expression of infant- and adult-preferring transcripts of CDCrel-1 septin gene in the rat neocortex. Biochemical and Biophysical Research Communications, 273, 723–28. Trimble, W.S. (1999) Septins: a highly conserved family of membrane-associated GTPases with functions in cell division and beyond. Journal of Membrane Biology, 169, 75–81. Tsai, L.H., Lee, M.S. and Cruz, J. (2004) Cdk5, a therapeutic target for Alzheimer’s disease? Biochimica et Biophysica Acta, 1697, 137–42. Vega, I.E. and Hsu, S.C. (2001) The exocyst complex associates with microtubules to mediate vesicle targeting and neurite outgrowth. Journal of Neuroscience, 21, 3839–48. Vega, I.E. and Hsu, S.C. (2003) The septin protein Nedd5 associates with both the exocyst complex and microtubules and disruption of its GTPase activity promotes aberrant neurite sprouting in PC12 cells. Neuroreport, 14, 31–37. Versele, M. and Thorner, J. (2004) Septin collar formation in budding yeast requires GTP binding and direct phosphorylation by the PAK, Cla4. Journal of Cell Biology, 164, 701–15. Villen, J., Beausoleil, S.A., Gerber, S.A. and Gygi, S.P. (2007) Large-scale phosphorylation analysis of mouse liver. Proceedings of the National Academy of Sciences of the United States of America, 104, 1488–93. Walikonis, R.S., Jensen, O.N., Mann, M. et al. (2000) Identification of proteins in the postsynaptic density fraction by mass spectrometry. Journal of Neuroscience, 20, 4069–80. Witzmann, F.A., Arnold, R.J., Bai, F. et al. (2005) A proteomic survey of rat cerebral cortical synaptosomes. Proteomics, 5, 2177–201. Woolley, C.S. (2007) Acute effects of estrogen on neuronal physiology. Annual Review of Pharmacology and Toxicology, 47, 657–80. Xiong, J.W., Leahy, A. and Stuhlmann, H. (1999) Retroviral promoter-trap insertion into a novel mammalian septin gene expressed during mouse neuronal development. Mechanisms of Development, 86, 183–91. Xue, J., Milburn, P.J., Hanna, B.T. et al. (2004a) Phosphorylation of G-Septin on Ser-91 by cyclic GMP-dependent protein kinase-I in nerve terminals. Biochemical Journal , 381, 753–60.

REFERENCES

267

Xue, J., Tsang, C.W., Gai, W.P. et al. (2004b) Septin 3 (G-septin) is a developmentally regulated phosphoprotein enriched in presynaptic nerve terminals. Journal of Neurochemistry, 91, 579–90. Xue, J., Wang, X., Malladi, C.S. et al. (2000) Phosphorylation of a new brain-specific septin, G-septin, by cyclic GMP-dependent protein kinase. Journal of Biological Chemistry, 275, 10047–56. Yagi, M., Zieger, B., Roth, G.J. and Ware, J. (1998) Structure and expression of the human septin gene HCDCREL-1. Gene, 212, 229–36. Yang, J.W., Czech, T., Felizardo, M. et al. (2006) Aberrant expression of cytoskeleton proteins in hippocampus from patients with mesial temporal lobe epilepsy. Amino Acids, 30, 477–93. Zenisek, D., Steyer, J.A. and Almers, W. (2000) Transport, capture and exocytosis of single synaptic vesicles at active zones. Nature, 406, 849–54. Zhang, Y., Gao, J., Chung, K.K. et al. (2000) Parkin functions as an E2-dependent ubiquitin-protein ligase and promotes the degradation of the synaptic vesicleassociated protein, CDCrel-1. Proceedings of the National Academy of Sciences of the United States of America, 97, 13354–59.

12 Septins and platelets Jerry Ware Department of Physiology and Biophysics, University of Arkansas for Medical Sciences, Little Rock, AR, USA

Constantino Mart´ınez Centro Regional de Hemodonaci´on, University of Murcia, Murcia 30003, Spain

Barbara Zieger Department of Paediatrics and Adolescent Medicine, University Hospital Freiburg, Germany

HISTORICAL OVERVIEW The possibility of septins in higher eukaryotic cells was dismissed for some time partly due to the bias that asymmetric cell division, or budding, was a process unique to yeast and likely not relevant to the symmetric cell division common in higher eukaryotes (Chant, 1996). However, based on cDNA sequences characterized in the mid-1990s a limited number of publications did identify septin proteins in higher eukaryotic cells and, in one case, the protein was functionally similar to yeast septins with a role in cytokinesis (Neufeld and Rubin, 1994; Nottenburg, Gallatin and St.John, 1990; Nakatsuru, Sudo and Nakamura, 1994). In the past few years the field has seen a rapid increase in citations with a current recognition of at least 14 different mammalian septin proteins (Macara et al., 2002). Surveying the literature makes it difficult to assign a single function to the mammalian septin family but themes are emerging, including cytokinesis, exocytosis, tumorigenesis and the development of neurological syndromes (Hall and Russell, 2004; Kojima et al., 2004; Ihara et al., 2003; Choi et al., 2003; Dong et al., 2003; Barr et al., 2004). Our interest in this area started with the identification of a human septin protein present in circulating blood platelets (Zieger, Hashimoto and Ware, 1997). The Septins Edited by Peter A. Hall, S.E. Hilary Russell and John R. Pringle © 2008 John Wiley & Sons, Ltd. ISBN: 978-0-470-51969-1

270

CH12 SEPTINS AND PLATELETS

At the time of its identification this represented only the third identification of a septin protein outside of the yeast septin protein family. The platelet participates in a myriad of biological processes but foremost is its role in blood coagulation and thrombosis. As an anucleate cell of blood, the platelet is produced from a precursor cell, the megakaryocyte, present in bone marrow. The megakaryocyte is unique in that it undergoes a complex series of maturation steps including polyploidization. During the final stages of maturation DNA content within the megakaryocyte can exceed 128N. Via events that are still poorly understood, the megakaryocytic cytoplasm fragments to release several thousand platelets per megakaryocyte into the bloodstream. Normal human blood contains approximately 3 × 105 platelets per microlitre of blood while mouse platelets are slightly smaller and the mouse platelet count is typically three times greater than in human blood. Simply put, the circulating blood platelet monitors the endothelial cell lining of blood vessels and via a series of surface membrane receptors can recognize a damaged or inflamed vascular surface where it will tether and recruit additional platelets. The net result, in combination with blood coagulation enzymes, is the generation of a platelet-rich plug stabilized by insoluble fibrin strands which either prevent blood loss (haemostasis) or restrict normal blood flow to a tissue or organ (thrombosis).

A PROTOTYPIC PLATELET SEPTIN The repertoire of platelet membrane receptors that support haemostasis and thrombosis is fairly extensive (Ware, 2004). The physiologic relevance of many of the receptors was recognized by careful characterization of bleeding disorders and their association with an absent receptor. A family pedigree with severe bleeding and fatal intracranial haemorrhage led two French physicians to describe the Bernard-Soulier syndrome approximately 60 years ago (Bernard and Soulier, 1948). Today, the molecular basis of the Bernard-Soulier syndrome is well-defined and known to be an absent platelet membrane complex, glycoprotein (GP) Ib-IX, that is critical in the initial events that lead to haemostasis or thrombosis (Lopez et al., 1998; Konstantinides et al., 2006). The complex is composed of three distinct gene products, the α- and β-subunits of GP Ib and GP IX. Of interest in this work was the β-subunit of GP Ib (GP Ibβ) seemingly encoded by a 1 kb mRNA whose expression appeared to be restricted to megakaryocytes and platelets (Yagi et al., 1994). Somewhat forgotten in this analysis was another mRNA species of ∼3.5 kb hybridizing strongly with cDNA probes encoding the β-subunit of GP Ib. Kelly et al. (1994) reported the larger variant could be a unique form of the β-subunit leading to speculation that endothelial cells contain a variant form of the GP Ib-IX complex. This idea was never confirmed by others and we demonstrated the larger 3.5 kb transcript was the product of an adjacent gene with an imperfect polyadenylation sequence (Zieger, Hashimoto

A PROTOTYPIC PLATELET SEPTIN CDCrel-1/Septin 5/SEPT5 E

H

H X XH

B

271

Platelet GP Ibβ Gene

B

B H

X

E

5’

3’

I

II

Human chromosome 22q11.2

1 kb

Figure 12.1 A complete characterization of the platelet GP Ibβ gene identified a second megakaryocytic/platelet gene product residing 250 nucleotides 5 to the GP Ibβ gene. A detailed characterization of this gene and its septin protein product, SEPT5, has been a focus of studies to define the relevance of SEPT5 to platelet biology

and Ware, 1997). The larger transcript contained open reading frames for a septin protein fused to coding sequence for platelet GP Ibβ. The 3 end of the newly identified gene and initiating methionine codon of GP Ibβ were approximately 250 nucleotides apart in the larger transcript (Figure 12.1). The newly identified gene residing closely to platelet GP Ibβ could be translated into a polypeptide of 369 residues (Zieger, Hashimoto and Ware, 1997). Sequence comparisons with GenBank deposits revealed the protein had similarities with eukaryotic septins that had been recently characterized. The highest similarity, at the time, was a protein product in Drosophila designated, PNUT (Neufeld and Rubin, 1994). An absence of Drosophila PNUT is embryonically lethal and immunohistochemistry placed the protein in the cleavage furrow of dividing cells similar to that described for their yeast counterparts. Any relevance to the anucleate platelet was not obvious but, nevertheless, based on sequence similarity with PNUT, we designated the human gene located adjacent to GP Ibβ, CDCrel-1 (Cell d ivision cycle rel ated-1 ). This designation would prove to be less than ideal since we later observed the highest levels of CDCrel-1 protein were found in neurons, heart tissue and megakaryocytes/platelets typically thought of as post-mitotic organs (Caltagarone et al., 1998; Zieger et al., 2000). Indeed, this would represent one of the first clues that septin function in higher eukaryotic cells might not be restricted to cytokinesis, a theme that has continued to emerge and making it likely that no single function can be assigned to the family of mammalian septins. A standardized nomenclature for mammalian septins was adopted in 2002 and the designation of CDCrel-1 (also identified as PNUTL-1) was standardized to SEPT5 (Macara et al., 2002 and see Appendix B). In the meantime an ever-expanding recognition of mammalian septin proteins had proceeded with our recognition of a gene/protein highly similar to CDCrel-1 and designated CDCrel-2/PNUTL-2 but later renamed SEPT4. Shortly thereafter, SEPT8 was identified with a nearly similar gene expression pattern to SEPT5, that is, highest levels in neurons and megakaryocytes (Blaser et al., 2003). The recognition of SEPT5 in platelets lead to an experimental strategy in mice to genetically ablate the SEPT5 gene and determine the phenotype, if any, of circulating platelet devoid of SEPT5 protein.

272

CH12 SEPTINS AND PLATELETS

SEPT5 AND PLATELET FUNCTION Biochemical data suggested SEPT5 would have relevance for platelet function. First, we showed SEPT5 is phosphorylated in platelets in the presence of known platelet agonists, such as collagen or thrombin (Dent et al., 2002). A role for SEPT5 in platelet secretion was suggested by the observation that immunopurification of SEPT5 from platelet lysates co-purified syntaxin 4, a component of the SNARE complex involved in the exocytic pathway in platelets (Dent et al., 2002; Flaumenhaft et al., 1999). Upon activation, platelets release a plethora of pro-thrombotic agonists and ligands that are essential for normal platelet function in haemostasis. A role for SEPT5 in platelet secretion was supported with observations that SEPT5 had a link to exocytosis or neurotransmitter release in neurons (Beites et al., 1999; Beites, Peng and Trimble, 2001; Beites, Campbell and Trimble, 2005)). Indeed, the molecular similarities between the platelet release reaction and neurotransmitter release have long been recognized to share mechanistic features (Reed, Fitzgerald and Polg´ar, 2000). A definitive link between the platelet release reaction and SEPT5 was established with characterization of platelets obtained from the SEPT5Null mouse (Dent et al., 2002). We performed platelet studies on animals deficient in SEPT5 and obtained evidence that SEPT5 may participate in the regulation of the platelet secretory response. Real time release studies showed a 2- to 3-fold increase in ATP release in stirred aggregation (Martinez et al., 2006). Moreover, SEPT5 is phosphorylated in response to platelet agonists and the SEPT5 knockout mouse platelets have an enhanced platelet secretion response. In the absence of SEPT5, platelets were observed to aggregate using subthreshold levels of agonist. We have also shown that mouse platelets devoid of SEPT5 are hyper-responsive in shear-induced platelet aggregation (Figure 12.2), a process that requires competent platelet receptors, GP Ib-IX and GP IIb/IIIa, and an active release/secretion process. Is SEPT5 an inhibitor of platelet secretion? The absence of SEPT5 would suggest so but such a conclusion could be wrong if SEPT5 is somehow dynamically regulating secretion. For example, does the phosphorylation state of SEPT5 change its ability to regulate secretion? If so, a simple conclusion just based on studies where SEPT5 is absent could be biased as the complete absence of the protein does not provide any insights into the dynamic regulation that might be critical to a protein’s function. Figure 12.1 illustrates the close spatial proximity of the SEPT5 and GP Ibβ genes. Indeed, it was the initial description of a variant GP Ibβ transcript that led to the identification of the SEPT5 gene (Zieger, Hashimoto and Ware, 1997). Both genes are expressed by megakaryocytes and seem to share some promoter elements. When we have used the short fragment (∼250 bp) separating the two genes as a promoter cassette it has very low level of activity in megakaryocytic-like cell lines. Thus, we assume there exists promoter and enhancer elements 5 to both genes that influence expression. In addition, we originally described the variant GP Ibβ transcript containing both coding sequences for SEPT5 and GP Ibβ

PLATELET SEPTIN COMPLEXES

273

108 dynes/cm2

SEPT5Null

Light Transmittance

SEPT5WT

0

1

2 3 4 Time (min)

5

6

Figure 12.2 Shear-induced platelet aggregation of SEPT5-deficient mouse platelets is shown. Platelet-rich plasma was prepared from SEPT5 knockout (Null) and SEPT5 wild-type (WT) littermates produced from het (×) het crosses. Platelet-rich plasma is from pooled animals and the aggregation curve shown is representative of several experiments. Shear-induced platelet aggregation is dependent upon plasma von Willebrand factor, the platelet GP Ib-IX receptor, the GP IIb-IIIa receptor, activation, and active platelet secretion of stored α-granule contents

(Zieger, Hashimoto and Ware, 1997). The expression of SEPT5 from this transcript was possible but the expression of GP Ibβ would require an internal ribosome binding site. The later seemed unlikely. In the process of characterizing the platelet GP Ib-IX complex, we also generated a mouse with a complete deletion of the GP Ibβ gene (Kato et al., 2004). Besides the obvious generation of a Bernard-Soulier phenotype we observed the GP Ibβ knockout platelets had fewer α-granules but those α-granules present were larger in size. A similar analysis of another mouse model of the Bernard-Soulier syndrome, deletion of GP Ibα (Ware, Russell and Ruggeri, 2000), did not yield differences in α-granule size. We suspected the larger α-granules were related to the GP Ib locus and not a direct consequence of an absent GP Ib-IX receptor. Thus, we checked SEPT5 protein levels since this gene is spatially close to GP Ibβ to determine if genetic ablation of the GP Ibβ gene has affected expression of SEPT5. As shown in Figure 12.3, SEPT5 levels were increased in the GP Ibβ knockout mouse. Indeed, the over-expression was tissue-specific and not observed in brain. Thus, we have evidence linking SEPT5 levels to the maintenance of normal platelet α-granule morphology.

PLATELET SEPTIN COMPLEXES In our early analysis of SEPT5 it became apparent that septin–septin interactions, analogous to those described in yeast did occur with mammalian septins. For

274

CH12 SEPTINS AND PLATELETS

(a)

(b)

Figure 12.3 SEPT5 levels in platelet lysates. (a) Protein concentrations in mouse platelet lysates samples were determined and similar protein amounts were analysed by SDS-PAGE. After electrophoresis, the proteins were transferred to nitrocellulose and immunoblotted with an anti-SEPT5 monoclonal antibody (LJ-33). The predominant SEPT5 signal present in platelet lysates is approximately 45 kDa. Quantitation revealed a 2- to 3-fold increase in SEPT5 levels in platelets from GP IβNull animals. (b) The same membrane was reacted with an anti-14-3-3ζ polyclonal antibody to document the approximate protein load for each lane. This research was originally published in Blood . Kato et al., K., Martinez, C., Russell, S. et al. (2004) Genetic deletion of mouse platelet glycoprotein Ibβ produces a Bernard-Soulier phenotype with increased α-granule size. Reproduced with permission from Kato et al. (2004) Blood , 104, 2339–2344, Copyright  The American Society of Hematology 2004

example, using SEPT5 as bait in a yeast-2 hybrid experiment would isolate other septin proteins (Blaser et al., 2002). We have now expanded that characterization and know that platelet septin complexes composed of SEPT5 also contain SEPT6, SEPT7 and SEPT9 (Figure 12.4) (Martinez et al., 2006). Immunoprecipitation of platelet lysates with an anti-SEPT5 monoclonal antibody we developed, LJ-33, simultaneously purifies SEPT6, SEPT7 and SEPT9. A similar immunoprecipitation using an anti-SEPT6 antibody co-precipitates SEPT5 and SEPT7 (Figure 12.4). To identify other potential septin partners included in the complex(es) identified above, we performed additional immunoprecipitations of platelet SEPT5 using LJ-33 and subjected the eluates to nanoLC MALDI MS/MS. This revealed the presence of SEPT5, as expected, along with SEPT1, SEPT2, SEPT6, SEPT7, SEPT8 and SEPT11. Overall the results provide direct evidence that SEPT5 co-immunoprecipitates a minimum of seven additional septin proteins. Thus, the constituents and stoichiometries within a septin complex may vary requiring a coordinated-regulation for their assembly.

PLATELET SEPTIN COMPLEXES

275

(a)

(b)

Figure 12.4 Co-immunoprecipitation of multiple platelet septins. (a) SEPT5 from platelet lysates of resting (R), thrombin-activated (Thr) and collagen-activated (Col) platelets was immunoprecipitated with an anti-SEPT5 monoclonal antibody, LJ-33, and subjected to immunoblotting using anti-SEPT6 or anti-SEPT7. In platelets, SEPT5 is associated at least with SEPT6 and SEPT7. The complex formed by these three septins is not disrupted during platelet activation. Results from a control immunoprecipitation using a non-specific IgG is shown for comparison (IgG). (b) SEPT6 from resting platelets was immunoprecipitated with a non-specific IgG (lane 1) or an anti-SEPT6 antibody (lane 2). Septins were further analysed by immunoblotting. This research was originally published in Journal of Thrombosis and Haemostasis. Martinez, C., Corral, J., Dent, J.A. et al. (2006) Platelet septin complexes form rings and associate with the microtubular network. Reproduced with permission from Martinez et al. Journal of Thrombosis and Haemostatis, 4, 1388–1395, Copyright  2006, Blackwell Publishing

What is the relevance of septin–septin complex formation? We performed experiments where SEPT5 was tagged with green fluorescent protein (GFP). Transfection of this construct into heterologous cells, COS-7, resulted in a diffuse GFP signal throughout the cytoplasm (Martinez et al., 2004). Similar results were obtained with the transfection of SEPT8 into COS-7 cells. However, cotransfection with cDNAs encoding both SEPT5 and SEPT8 resulted in a localization of both proteins to cytosolic vesicles (Martinez et al., 2004). We concluded that septin–septin interactions can dictate where a particular septin complex goes in the cell and it is the stoichiometry of the complex that may target a complex for a specific cellular task. This work also included a detailed study on the domains of SEPT5 that support the intermolecular interactions. In the Martinez et al. (2004) manuscript we tested the importance of a GTP-binding motif within SEPT5.

276

CH12 SEPTINS AND PLATELETS

The results demonstrated the GTP motif is essential for proper folding and structure. Any mutation we have made, to date, within this motif has produced an aggregated protein that is not able to function or assemble with other septins. We originally proposed a regulatory role for the GTP motif but our data supports a structural or folding role for the SEPT5/GTP interaction.

IMMUNOLOCALIZATION OF PLATELET SEPTIN COMPLEXES AND THE PRESENCE OF A TUBULIN-ASSOCIATED SEPTIN ‘RING’ The results described above demonstrate the intracytoplasmic association of two different septins, SEPT5 and SEPT8 but as shown in Figure 12.5 the situation in the platelet is more complex with multiple septins. SEPT8 is also an interaction partner of SEPT4 (Blaser et al., 2004). For platelets, the fluorescence produced by labelling with the anti-SEPT5 monoclonal, LJ-33, revealed SEPT5 predominantly in the periphery of the platelet along with a punctate pattern in the cytoplasm. Co-labelling with an anti-von Willebrand factor polyclonal antibody clearly marked the interior of platelet α-granules and was distinct from the pattern of SEPT5 protein (Figure 12.5). Immunofluorescence of SEPT6 antigen also revealed strong staining in a ring-like structure similar to that observed for SEPT5 (Figure 12.5). The localization of both SEPT5 and SEPT6 to the platelet periphery places the majority of these septins near the tubulin rich platelet microtubule ring. Indeed, immunofluorescence of platelet tubulin demonstrated co-localization with SEPT6 as revealed by the merged images (Figure 12.5). Following the localization of SEPT5 and SEPT6 in the resting platelet we investigated their localization as the platelet is allowed to spread on fibrinogen-coated glass. The SEPT5 ring-like structure seen in resting platelets was not disrupted during the formation of lamellipodia or pseudopodia (Figure 12.5). SEPT5 did not co-localize with actin before or during active platelet spreading (Figure 12.5). In order to determine if a septin/tubulin interaction can be validated biochemically, we isolated the microtubule coil from resting platelets. Tubulin was stabilized with taxol and extracted after a non-ionic treatment by sequential ultracentrifugation. Tubulin was fractionated into soluble and polymerized fractions and analysed by immunoblotting using available antibodies for individual septin proteins. Tubulin extracted from resting platelets was mostly represented in the pelleted fraction corresponding to the microtubule circumferential coil (Kenney and Linck, 1985; Schwer et al., 2001). As expected from the microscopic results, SEPT5, SEPT6 and SEPT7 were all represented in the insoluble fraction together with tubulin (Martinez et al., 2006). This result confirmed the co-purification of a septin complex composed of SEPT5/6/7 and the association of the complex with the platelet tubulin-containing fraction.

IMMUNOLOCALIZATION OF PLATELET SEPTIN COMPLEXES Actin

277

Merged

SEPT5

(a) SEPT5

vWF

Merged

(b) SEPT6

α-tubulin

Merged

(c) Actin

SEPT5

Merged

(a)

(b)

(c)

(d)

(e)

(f)

(d)

Figure 12.5 Immunolocalization of a platelet septin ring. (a) Platelets fixed in suspension were immunostained for actin and SEPT5 protein: two localizations were observed, a peripheral and cytosolic punctate distribution. The actin and SEPT5 did not co-localize. (b) Immunostaining for von Willebrand factor (vWF), a platelet α-granule specific marker, and SEPT5 is shown. (c) SEPT6 and tubulin immunostaining are shown along with the merged image suggesting co-localization. (d) Platelets were spread on a fibrinogen-coated surface for 20 min and immunostained for actin and SEPT5. Septins conserved the ring-like structure that is centralized in spreading platelets (a, b, c). A more detailed localization is shown in the magnified images (d, e, f). This research was originally published in Journal of Thrombosis and Haemostasis. Martinez, C., Corral, J., Dent, J.A. et al. (2006) Platelet septin complexes form rings and associate with the microtubular network. Reproduced with permission from Martinez et al. Journal of Thrombosis and Haemostatis, 4, 1388–1395, Copyright  2006, Blackwell Publishing

278

CH12 SEPTINS AND PLATELETS

CONCLUDING REMARKS Although we have focused on platelet septins the family of proteins is associated with a wide variety of cellular events from cytokinesis to secretion and impacting cancer, neurological disorders and haemostasis. Is there a unifying aspect for such diversity? Our view from studies of platelet septins is that septins actively participate in cellular events where active membrane movement or cytoplasmic partitioning occurs. To fulfil this function we hypothesize the participation of a septin in any of these processes resides in the makeup of a specific septin complex composed of a unique combination of septin proteins. Thus, an individual cell type may have specialized septin function participating in exocytosis, cell division, or both. In the case of megakaryocytes and platelets the potential impact of septin function may exist in both the bone marrow and in the circulating platelet. To date, we have (i) defined the makeup of the platelet repertoire and the composition of platelet septin complexes (ii) illustrated the importance of septin–septin interaction for cytosolic localization and (iii) defined the relationship between septins, platelet secretion and the maintenance of normal α-granule morphology. For the future it will be imperative to more precisely define the platelet septin repertoire and understand functional redundancies among members of the family. Hopefully, understanding septin function in normal platelet biology will also contribute to a wide array of biological events beyond haemostasis and thrombosis.

REFERENCES Barr, A.M., Young, C.E., Sawada, K. et al. (2004) Abnormalities of presynaptic protein CDCrel-1 in striatum of rats reared in social isolation: relevance to neural connectivity in schizophrenia. The European Journal of Neuroscience, 20, 303–7. Beites, C.L., Campbell, K.A. and Trimble, W.S. (2005) The septin Sept5/CDCrel-1 competes with alpha-SNAP for binding to the SNARE complex. The Biochemical Journal , 385, 347–53. Beites, C., Peng, X-R. and Trimble, W.S. (2001) Expression and analysis of properties of septin CDCrel-1 in exocytosis. Methods in Enzymology, 329, 499–510. Beites, C.L., Xie, H., Bowser, R. and Trimble, W.S. (1999) The septin CDCrel-1 binds syntaxin and inhibits exocytosis. Nature Neuroscience, 2, 434–39. Bernard, J. and Soulier, J.P. (1948) Sur un nouvelle variete de dystrophie thrombocytaire hemorragipare congenitale. Semaine des Hˆopitaux de Paris, 24, 3217–33. Blaser, S., Horn, J., Wurmell, P. et al. (2004) The novel human platelet septin SEPT8 is an interaction partner of SEPT4. Thrombosis and Haemostasis, 91, 959–66. Blaser, S., Jersch, K., Hainmann, I. et al. (2002) Human septin-septin interaction: CDCrel-1 partners with KIAA0202. FEBS Letters, 519, 169–72. Blaser, S., Jersch, K., Hainmann, I. et al. (2003) Isolation of new splice isoforms, characterization and expression analysis of the human septin SEPT8 (KIAA0202). Gene, 312, 313–20. Caltagarone, J., Rhodes, J., Honer, W.G. and Bowser, R. (1998) Localization of a novel septin protein, hCDCrel-1, in neurons of human brain. Neuroreport, 9, 2907–12.

REFERENCES

279

Chant, J. (1996) Septin scaffolds and cleavage planes in Saccharomyces. Cell , 84, 187–90. Choi, P., Snyder, H., Petrucelli, L. et al. (2003) SEPT5 v2 is a Parkin-binding protein. Brain Research. Molecular Brain Research, 117, 179–89. Dent, J., Kato, K., Peng, X-R. et al. (2002) A prototypic platelet septin and its participation in platelet secretion. Proceedings of the National Academy of Sciences of the United States of America, 99, 3064–69. Dong, Z., Ferger, B., Paterna, J.C. et al. (2003) Dopamine-dependent neurodegeneration in rats induced by viral vector-mediated overexpression of the Parkin target protein, CDCrel-1. Proceedings of the National Academy of Sciences of the United States of America, 100, 12438–43. Flaumenhaft, R., Croce, K., Chen, E. et al. (1999) Proteins of the exocytotic core complex mediate platelet alpha-granule secretion. Roles of vesicle-associated membrane protein, SNAP-23, and syntaxin 4. The Journal of Biological Chemistry, 274, 2492–501. Hall, P.A. and Russell, S.E. (2004) The pathobiology of the septin gene family. The Journal of Pathology, 204, 489–505. Ihara, M., Tomimoto, H., Kitayama, H. et al. (2003) Association of the cytoskeletal GTP-binding protein Sept4/H5 with cytoplasmic inclusions found in Parkinson’s disease and other synucleinopathies. The Journal of Biological Chemistry, 278, 24095–102. Kato, K., Martinez, C., Russell, S. et al. (2004) Genetic deletion of mouse platelet glycoprotein Ibβ produces a Bernard-Soulier phenotype with increased α-granule size. Blood , 104, 2339–44. Kelly, M.D., Essex, D.W., Shapiro, S.S. et al. (1994) Complementary DNA cloning of the alternatively expressed endothelial cell glycoprotein Ibβ (GPIbβ) and localization of the GPIbβ gene to chromosome 22. The Journal of Clinical Investigation, 93, 2417–24. Kenney, D.M. and Linck, R.W. (1985) The cystoskeleton of unstimulated blood platelets: structure and composition of the isolated marginal microtubular band. Journal of Cell Science, 78, 1–22. Kojima, K., Sakai, I., Hasegawa, A. et al. (2004) FLJ10849, a septin family gene, fuses MLL in a novel leukemia cell line CNLBC1 derived from chronic neutrophilic leukemia in transformation with t(4;11)(q21;q23). Leukemia, 18 (5), 998–1005. Konstantinides, S., Ware, J., Marchese, P. et al. (2006) Distinct antithrombotic consequences of platelet glycoprotein Ibβ and VI deficiency in a mouse model of arterial thrombosis. Journal of Thrombosis and Haemostasis, 4, 2014–21. Lopez, J.A., Andrews, R.K., Afshar-Kharghan, V. and Berndt, M.C. (1998) BernardSoulier syndrome. Blood , 91, 4397–418. Macara, I.G., Baldarelli, R., Field, C.M. et al. (2002) Mammalian septins nomenclature. Molecular Biology of the Cell , 13, 4111–13. Martinez, C., Corral, J., Dent, J.A. et al. (2006) Platelet septin complexes form rings and associate with the microtubular network. Journal of Thrombosis and Haemostasis, 4, 1388–95. Martinez, C., Sanjuan, M.A., Dent, J.A. et al. (2004) Human septin-septin interactions as a prerequisite for targeting septin complexes in the cytosol. The Biochemical Journal , 382, 783–91.

280

CH12 SEPTINS AND PLATELETS

Nakatsuru, S., Sudo, K. and Nakamura, Y. (1994) Molecular cloning of a novel human cDNA homologous to CDC10 in Saccharomyces cerevisiae. Biochemical and Biophysical Research Communications, 202, 82–87. Neufeld, T.P. and Rubin, G.M. (1994) The Drosophila peanut gene is required for cytokinesis and encodes a protein similar to yeast putative bud neck filament proteins. Cell , 77, 371–79. Nottenburg, C., Gallatin, W.M. and St.John, T. (1990) Lymphocyte HEV adhesion variants differ in the expression of multiple gene sequences. Gene, 95, 279–84. Reed, G.L., Fitzgerald, M.L. and Polg´ar, J. (2000) Molecular mechanisms of platelet exocytosis: insights into the “secrete” life of thromboycytes. Blood , 96, 3334–42. Schwer, H.D., Lecine, P., Tiwari, S. et al. (2001) A lineage-restricted and divergent beta-tubulin isoform is essential for the biogenesis, structure and function of blood platelets. Current Biology, 11, 579–86. Ware, J. (2004) Dysfunctional platelet membrane receptors: from humans to mice. Thrombosis and Haemostasis, 92, 478–85. Ware, J., Russell, S. and Ruggeri, Z.M. (2000) Generation and rescue of a murine model of platelet dysfunction: the Bernard-Soulier syndrome. Proceedings of the National Academy of Sciences of the United States of America, 97, 2803–8. Yagi, M., Edelhoff, S., Disteche, C.M. and Roth, G.J. (1994) Structural characterization and chromosomal location of the gene encoding human platelet glycoprotein Ibβ. The Journal of Biological Chemistry, 269, 17424–27. Zieger, B., Hashimoto, Y. and Ware, J. (1997) Alternative expression of platelet glycoprotein Ibβ mRNA from an adjacent 5’ gene with an imperfect polyadenylation signal sequence. The Journal of Clinical Investigation, 99, 520–25. Zieger, B., Tran, H., Hainmann, I. et al. (2000) Characterization and expression analysis of two human septin genes, PNUTL1 and PNUTL2 . Gene, 261, 197–203.

13 Septins and apoptosis Marie-Jeanne Carp and Sarit Larisch Apoptosis and Cancer Research Laboratory, Department of Pathology, Rambam Medical Center, Haifa 31096, Israel

INTRODUCTION Playing major roles in the cell, septin proteins are phylogenetically conserved among diverse eukaryotes ranging from fungi to C. elegans all the way to mammals (Kinoshita, 2003 and see Chapter 2). All septin family members share a preserved P-loop GTP-binding domain, yet contain large sequence variability and diverse motifs (Kinoshita, 2003; Saraste, Sibbald and Wittinghofer, 1990; Bourne, Sanders and Mccormick, 1991; Hall and Russell, 2004). Accordingly, septin proteins are shown to participate in many key cellular functions including vesicle trafficking, cytoskeletal and filamental formation, membrane remodelling and exocytosis (Lindsey and Momany, 2006; Hall et al., 2005; Spiliotis and Nelson, 2006; Hall and Russell, 2004; Kartmann and Roth, 2001). Though septins are involved in a large variety of fundamental cell processes, a direct involvement of a family member in programmed cell death, apoptosis, was reported only in 2000, with the discovery of SEPT4 i2 (also known as ARTS protein; Larisch et al., 2000; Larisch-Bloch et al., 2000)1 . In this chapter we will describe the cellular and biochemical characteristics of apoptosis and concentrate on the so far only reported pro-apoptotic septin family member SEPT4 i2. We will illustrate the genetic and biochemical basis for involvement of the SEPT4 i2 in apoptosis.

1 In line with HGNC guidelines and the emerging consensus on septin nomenclature (see Appendix B) the SEPT4 isoform formerly known as ARTS will be designated hence forth SEPT4 i2, the product of the SEPT4 v2 transcript of the SEPT4 gene.

The Septins Edited by Peter A. Hall, S.E. Hilary Russell and John R. Pringle © 2008 John Wiley & Sons, Ltd. ISBN: 978-0-470-51969-1

282

CH13 SEPTINS AND APOPTOSIS

APOPTOSIS: A PROGRAMMED CELL DEATH MECHANISM Essentially all animal cells have the ability to kill themselves by activating an intrinsic cell suicide program when they are no longer needed or have become seriously damaged (Vaux and Korsmeyer, 1999; Jacobson, Weil and Raff, 1997). The execution of this program leads to a morphologically distinct form of cell death termed apoptosis (Kerr, Wyllie and Currie, 1972; Wyllie, 1980). It is now generally accepted that apoptosis is of central importance for development and homeostasis of metazoan animals. The roles of apoptosis include the sculpting of structures during development, deletion of unneeded cells and tissues, regulation of growth and cell number, and the elimination of abnormal and potentially dangerous cells (Jacobson, Weil and Raff, 1997). In this way, apoptosis provides a stringent and highly effective ‘quality control mechanism’ that limits the accumulation of harmful cells, such as self-reactive lymphocytes, virus-infected cells and tumour cells (Naik, Karrim and Hanahan, 1996; Reed, 1995; Thompson, 1995; White, 1996). On the other hand, inappropriate apoptosis is associated with a wide variety of diseases, including AIDS, neurodegenerative disorders and ischemic stroke (Martinou et al., 1994; Pettmann and Henderson, 1998; Thompson, 1995; Raff, 1998). The main executioners of apoptosis are a set of cysteine proteases called caspases (for cysteine aspartase), that are widely expressed as inactive zymogens (Nicholson and Thornberry, 1997). These caspase zymogens are converted to the active protease as cells are selected to die. Once activated, caspases are thought to cleave a variety of important structural proteins, enzymes and regulatory molecules which are essential for the proper function of the cell (Thornberry and Lazebnik, 1998). Generally, caspases are divided into two classes based on their function; effector caspases which cleave protein substrates and execute the apoptosis program (such as caspases 3, 6 and 7) and initiator caspases which cleave inactive pro-forms of effector caspases, thereby activating the effector caspases, leading to the death of the cell (such as caspases 8, 9, 10 and 2) (Kumar, 1995; Salvesen and Dixit, 1997; Thornberry, Rosen and Nicholson, 1997). Many different signals that can originate either from within the doomed cell or from its extracellular environment can trigger apoptosis (Steller, 1995). These signals include steroid hormones, peptide survival factors, cell adhesion, specific cell surface receptors, viral infection, oxidative stress, excitotoxicity, ischemia, unfolded proteins and unrepaired DNA breaks (such as caused by ionizing radiation) (Truman,Thorn and Robinow, 1992; Oppenheim, 1991; Raff, 1992; Pettmann and Henderson, 1998; Nagata, 1997; Bergmann et al., 1998). Two main signalling pathways transmit the death signals leading to the programmed cell destruction; the ‘extrinsic and the intrinsic pathway’. The extrinsic pathway is triggered by binding of ligands to specific cell surface death receptors

APOPTOSIS: A PROGRAMMED CELL DEATH MECHANISM

283

such as Fas, TNF1 and TRAIL (Wallach et al., 1997; Srinivasula et al., 1996; Golstein, 1997). The second pathway and most abundant in cells, is the intrinsic or mitochondrial pathway which when stimulated by various stress or cytotoxic cellular signals results in the release of apoptosis promoting factors including cytochrome c from the mitochondria (Green and Kroemer, 2004). The Bcl-2 (B-cell lymphoma 2) family of proteins govern mitochondrial outer membrane permeabilisation and can be either pro-apoptotic (Bax, Bak and Bok among others) or anti-apoptotic (including Bcl-2, Bcl-xL and Bcl-w, among an assortment of others). There are a number of theories concerning how the Bcl-2 gene family exerts their pro- or anti-apoptotic effect. A most common hypothesis states that this is achieved by activation or inactivation of an inner mitochondrial permeability transition pore, which is involved in the regulation of matrix Ca2+ , pH and voltage. It is also thought that some Bcl-2 family proteins can induce (pro-apoptotic members) or inhibit (anti-apoptotic members) the release of cytochrome c into the cytosol which, once there, activates caspase-9 and caspase-3, leading to apoptosis. Due to their ability to receive, coordinate and dispatch death signals, mitochondria serve as a central junction of cellular decisions ranging between cellular survival and demise. In healthy normal cells, unwanted apoptosis is prevented through the action of a set of proteins termed inhibitors of apoptosis proteins (IAPs), which bind active caspases during non-apoptotic conditions thereby inhibiting their function (Salvesen and Duckett, 2002). IAPs were first identified as baculovirus proteins that inhibit apoptosis in infected insect cells (Clem, Fechheimer and Miller, 1991; Clem and Miller, 1994) and were later found to be largely distributed in metazoan cells. All IAP proteins contain between one to three baculovirus repeat domains (BIR) which directly interact with caspases resulting in inhibition of their protease activity. Some of the IAP proteins also contain a RING domain bearing an E3-ubiquitin ligase function (Hay, Wassarman and Rubin, 1995; Roy et al., 1997; Rothe et al., 1995; Roy et al., 1995; Duckett et al., 1996). Under apoptotic conditions active caspases become available for promoting cell destruction through their release from binding to IAPs and activation within the apoptosome complex. The apoptosome is a cytosolic complex created once cytochrome c exits the mitochondria and binds an adaptor protein termed APAF-1, which in the presence of dATP recruits multiple pro-caspase 9 and induces its processing into active caspase 9 molecules. This high molecular weight complex can then activate the effector caspases-3 and -7 leading to the final disintegration of the cell (Thornberry and Lazebnik, 1998). Importantly, during apoptosis the caspase inhibition exerted by IAPs is lifted by a set of proteins termed IAP-antagonists which are released from mitochondria enabling them to bind IAPs in the cytososl and unleash caspase activity. The best known IAP-antagonists are Smac/Diablo, Omi/HtrA2 and SEPT4 i2 (Du et al., 2000; Suzuki et al., 2001; Gottfried et al., 2004).

284

CH13 SEPTINS AND APOPTOSIS

DIRECT INVOLVEMENT OF SEPTINS IN APOPTOSIS SEPT4 transcripts and isoforms The SEPT4 v2 transcript encodes SEPT4 i2 (formerly known as ARTS ). This is so far the only septin directly involved in promoting apoptotic cell death. SEPT4 i2 was discovered following a non-biased genetic screen using retroviral insertional mutagenesis approach on NRP-154 rat prostate epithelial cells specifically responding in apoptosis to TGF-β treatment. The screen yielded a 1.8 kb cDNA, that mapped to the human chromosome 17q22-23 and was revealed to be a splice variant of the SEPT4 gene (Larisch-Bloch et al., 2000; Larisch et al., 2000). Several structural differences exist between SEPT4 i2 and the other splice variants of the SEPT4 gene; First, SEPT4 i2 lacks 20 amino acids at its N-terminus, which are part of most other SEPT4 splice variants. Second, SEPT4 i2 lacks the coiled-coil domain at its C
-terminus which is thought to play a role in intermolecular interactions (Sheffield et al., 2003). Though sharing the conserved GTP-binding domain, SEPT4 i2 lacks the G4 domain which may result in its inability to exchange GDP/GTP (Larisch et al., 2000). Most importantly, SEPT4 i2 contains a unique stretch of 27 amino acid caboxy terminus not found in any other septin, or any other yet reported protein. This unique sequence of SEPT4 i2 is responsible for its unusual pro-apoptotic function which is atypical to the septin family members. This unique sequence became a part of the SEPT4 i2 splice variant following an intron retention event occurring in the SEPT4 gene. Intron retention is defined by the presence of a transcript-confirmed intron within a transcript-confirmed exon. Therefore, mRNAs for a functional protein include a retained intron code which differ from the protein isoforms translated from the non retained mRNA (Kondrashov and Koonin, 2003; Galante et al., 2004). In the case of SEPT4 i2, intronic sequences originally spaced between exon VI and VII of the SEPT 4 gene, turned into exonic sequence delineating the new SEPT4 i2 isoform containing unique C-terminus sequence (Larisch et al., 2000). SEPT4 i2 also differs from most other septin in its cellular localization. SEPT4 i2 is mainly localized in mitochondria of normal cells (Larisch et al., 2000), as opposed to the typical localization of other septins to actin stress fibres and cytoskeletal structures (Kinoshita et al., 1997; Joberty et al., 2001; Peng et al., 2002; Hsu et al., 1998; Kinoshita, 2003; Kinoshita, Noda and Kinoshita, 2000). One other mouse septin, M-septin, was reported to be localized in mitochondria, yet no indication for involvement of M-septin in apoptosis has been reported (Takahashi et al., 2003). Though four different splice variants of the human SEPT4 gene are described, careful analysis of the data reveals that only two experimentally validated transcripts are reported; variant one, (SEPT4 v1 also known as; PnutL2, H5, hCDCREL-2, MART, CE5B3, Bradeion beta) and variant two, SEPT4 v2 that encodes SEPT4 i2 (or ARTS). The third variant, SEPT4 v3 shown in NCBI Genbank is found thus far merely as a computational prediction

DIRECT INVOLVEMENT OF SEPTINS IN APOPTOSIS

285

with no experimental confirmation of its existence. A fourth hypothetical isoform is Bradeion α (Tanaka et al., 2002; Tanaka et al., 2001; Tanaka et al., 2003). Two versions of Bradeion were reported; Bradeion Alpha and Bradeion Beta. While Bradeion Beta cDNA sequence is identical to the sequence of all other SEPT4 v1 transcripts, Bradeion Alpha (NM 080417.1) which appeared in the past as GenBank accession number AB002110.1 was recently deleted from NCBI sequences since it was found to be a nonsense-mediated mRNA decay (NMD) candidate. NMD transcripts are degraded quickly and most likely do not encode a protein product. Figure 13.1a shows a schematic representation of the currently identified human SEPT4 gene splice variants as presented in the NCBI database and substantiated by scientific reports. Taking into consideration the current experimental knowledge available, it seems that the human SEPT 4 gene encodes only two distinct protein variants. Independent and thorough corroboration of this is however required.

Mechanism of SEPT4 i2 induced apoptosis Evidence that SEPT4 i2 plays a role in apoptosis has come both from gain and loss of function studies. In certain cells, elevated levels of SEPT4 i2 is sufficient to induce apoptosis (Gottfried et al., 2004; Lotan et al., 2005). More generally, expression of SEPT4 i2 can promote apoptosis in response to a variety of pro-apoptotic stimuli such as, Fas, TGF-beta, cytosine arabinoside, etoposide and staurosporine. Conversely, down-regulation of endogenous SEPT4 i2 by anti-sense expression was shown to protect cells against TGF-beta-induced apoptosis (Larisch et al., 2000). In all these cases, SEPT4 i2-mediated cell killing leads to caspase activation. SEPT4 i2 is found to induce activation of initiator caspase-9, as well as activation of the main effector caspase-3 (Gottfried et al., 2004; Lotan et al., 2005; Elhasid et al., 2004). Because SEPT4 i2 is implicated in a wide variety of apoptotic paradigms, it seems to function at a central apoptotic junction where different upstream apoptotic inputs converge to mediate caspase activation. The main mechanism by which SEPT4 i2 exerts its apoptotic activity is through direct binding and inhibition of IAPs (inhibitors of apoptosis proteins). Upon induction of apoptosis, SEPT4 i2 is released from mitochondria and co-localizes with XIAP (X-linked-IAP) in the cytosol. Binding of SEPT4 i2 to XIAP is direct, as recombinant SEPT4 i2 and XIAP proteins can bind to each other in vitro (Gottfried et al., 2004). SEPT4 i2 binding to XIAP is specific and related to its pro-apoptotic function, as mutant forms of SEPT4 i2 and other related but non-apoptotic septins fail to bind XIAP and fail to induce apoptosis. In addition, binding of SEPT4 i2 to XIAP causes a significant reduction in XIAP levels and leads to caspase activation and cell death (Gottfried et al., 2004). The mitochondrial localization of SEPT4 i2 and its ability to bind XIAP are shared with other mammalian IAP-antagonists, most known ones are Smac/Diablo

CH13 SEPTINS AND APOPTOSIS

286

17q22-23

53964410

53952615

Sep4_V2/ARTS Sep4_V3 Sep4_V1/PnutL2, H5, hCDCREL-2, MART; CE5B3, Bradeion beta - coding translated region

- coding untranslated region

(a) MDRSLGWQGNSVPEDRTEAG GESGLGKST G1; P-LOOP Sep4_V1/PnutL2, H5, hCDCREL-2, MART; CE5B3, Bradeion beta 1 151 Sep4_V3

DTPG G3

AKAD G4

Coil-Coil domain

208

289

417-479

270

458

1

132

189

1

132

189

Sep4_V2/ARTS 247-274

YGPSLRLLAPPGAVKGTGQEHQGQGCH

(b) Dapi

SEPT4_i1

SEPT4_i2

Lymphocytes from healthy donor

Lymphocytes from ALL patient

(c)

Figure 13.1 (a) Schematic illustration of the different splice variants of septin 4. The boxes indicate transcribed regions. *SEPT4 i3 isoform awaits experimental validation. Current data support the existence of only two distinct isoforms for the human SEPT4 gene. (b) Alignment of the different human septin 4 isoform sequences. Location of the conserved G1 (P-loop) G3, G4 and coil–coil motifs are denoted by the hedged boxes. Unique sequences are marked as black boxes. (c) SEPT4 i2 is the only human septin4 isoform shown to directly induce and promote apoptosis. SEPT4 i2 functions as a tumour suppressor protein. Immunofluorescence staining presents lymphocytes isolated from healthy donor (upper panel) and from acute lymphoblastic leukemia (ALL) patient (lower panel). Staining with Dapi, showing nuclei of cells (blue), the non-apoptotic isoform of SEPT4; SEPT4 i1 (red), and pro-apoptotic protein SEPT4 i2 (green). Whereas all lymphocytes in healthy control contain similar levels of SEPT4 i2 and SEPT4 i1 (upper panel), the sample from leukemia patient exhibits loss of SEPT4 i2 staining in all tumour lymphoblasts (only remaining normal cells are stained), indicating selective loss of the pro-apoptotic protein SEPT4 i2 in leukemia patients (adapted from Elhasid et al., 2004). A colour version of (c) appears as Plate 13.1, see p. 246

DISCUSSION

287

and Omi/HtrA2 (Du et al., 2000; Verhagen et al., 2000; Verhagen et al., 2002; Hegde et al., 2002). Yet, SEPT4 i2 exhibits a unique mode of action in activating caspases through antagonizing IAPs; SEPT4 i2 lacks any recognizable IBM (IAP-Binding-Motif), a short sequence that is necessary for IAP-binding and inhibition and that is conserved amongst all other known IAP-antagonists (Shi, 2002; Huang et al., 2001; Tittel and Steller, 2000; Vaux and Silke, 2003). Instead, the binding to XIAP requires the unique C-terminus of SEPT4 i2, same stretch of 27 amino acids which became exonic sequences only in the SEPT4 i2 isoform as a result of intron retention phenomenon. Consistent with this idea, deletion of the C-terminus of SEPT4 i2 results in loss of XIAP-binding (Gottfried et al., 2004; Larisch et al., 2000). SEPT4 i2 is the only protein shown to function as an IAP-antagonist in vivo; as increased XIAP levels were found in sperm from Sept4 deficient mice (Kissel et al., 2005).

SEPT4 i2 functions as a tumour suppressor protein Studies both in human patients and in mice have shown that SEPT4 i2 plays an important role in cancer, particularly in haematopoietic cancers, leukemia and lymphoma. Expression of SEPT4 i2 is lost in all lymphoblasts of more than 70 % of childhood acute lymphoblastic leukemia (ALL) patients. The loss of SEPT4 i2 is specific, as the related non-apoptotic septin protein SEPT4 i1, bearing 83 % identity to SEPT4 i2, is unaffected (Elhasid et al., 2004). During remission, SEPT4 i2 expression is detected again in almost all patients. Two leukemic cell lines, ALL-1 and HL-60 lacking SEPT4 i2, were resistant to apoptotic induction by cytosine arabinoside. Transfection of SEPT4 v2 into these cells restored their ability to undergo apoptosis in response to this chemotherapeutic agent (Elhasid et al., 2004). Moreover, a large percentage of the Sept4 deficient mice exhibit extremely large spleens, and develop spontaneous haematopoietic malignancies (Steller, H. personal communication, unpublished results). Altogether, it seems that a combination of serendipity along with rare genetic and biochemical events have turned SEPT4 i2 into an exceptional protein with unusual features both as a septin family member and as an IAP-antagonist protein. First, the rare intron retention phenomenon occurring in the SEPT4 gene, which generates a novel stretch of amino acids not found in any other septin or known protein. Second, this unique sequence which does not share any resemblance to other known IBM (IAPs binding motif) bears the ability to bind and antagonize IAPs, releasing caspases and leading to apoptotic death of the cell.

DISCUSSION Although generation of large number of splice variants is characteristic of the septin family of genes (see Chapter 7), only one splice variant of the SEPT4 gene, SEPT4 v2 /SEPT4 i2 is currently known to directly promote apoptosis. Nevertheless, the association of several septins with two pathological phenomenon, cancer

288

CH13 SEPTINS AND APOPTOSIS

and neurodegenerative diseases may indicates that other septins may be involved, at least indirectly, in apoptosis. But at present direct evidence for the role of septins in apoptosis in these, or other conditions, is lacking. Therefore, within the septin family, SEPT4 i2 could have a unique function that may have arisen as an evolutionary ‘accident’. Through coincidental retention of an intron, a novel open reading frame (ORF) was generated at the SEPT 4 locus that resulted in addition of a unique stretch of 27aa to the SEPT4 i2 isoform. This unique SEPT4 i2 sequence either had or evolved the capacity to bind IAPs, thereby endowing SEPT4 i2 with the ability to activate caspases and cell death (Gottfried et al., 2004; Larisch et al., 2000; Larisch-Bloch et al., 2000). The other SEPT4 splice variants, SEPT4 i1 which share 83 % sequence homology with SEPT4 i2, cannot bind XIAP and promote apoptosis (Gottfried et al., 2004; Elhasid et al., 2004; Larisch et al., 2000). Moreover, loss of SEPT4 i2 in ALL patients is specific and related to its pro-apoptotic function, as levels of the SEPT4 i1 non-apoptotic septin in these samples remained unaffected (Elhasid et al., 2004). All this data indicates that the unique C
-terminus of SEPT4 i2, which is not shared by any other septin, is critical for its pro-apoptotic and tumour suppressor function. Does the P-loop GTPase domain, septins most conserved and characteristic sequence which is also present in SEPT4 i2, contribute to its pro-apoptotic function? First, SEPT4 i2 contains a non-typical GTP-binding domain which does not include the G4 motif (Figure 13.1b). One might argue that this lack of G4 motif might contribute in any way to its pro-apoptotic function. We therefore conclude that the P-loop of SEPT4 i2 may be necessary, but not sufficient for the induction of apoptosis. Furthermore, Lee et al. analysed the DNA sequences encoding the P-loop domains of SEPT4 i2 and Nod1 proteins in the tissues of colorectal carcinomas, gastric carcinomas, non-small cell lung cancers, and hepatocellular carcinomas and compared them to the P-loop sequence of normal cells originating from the same patients. The data analysis indicated that mutational events in both the P-loop domain of SEPT4 i2 and Nod1 genes do not contribute to the development of these cancers (Lee et al., 2006; Lee et al., 2006a). These studies support our model that the P-loop GTPase domain is not directly involved in apoptotic function, but perhaps plays a regulatory role for SEPT4 i2. The pro-apoptotic function of SEPT4 i2 is highly correlated with its function as a tumour suppressor protein. Expression of SEPT4 i2 is lost in all lymphoblasts of more than 70 % of childhood ALL patients (Elhasid et al., 2004). A screen for proteins interacting with Kaposin A, a human herpes virus involved in Kaposi Sarcoma, and other types of cancers, revealed possible functional interaction of Kaposin A protein with SEPT4 i2. The authors suggest that expression of Kaposin A protein could inhibit the apoptotic effect of SEPT4 i2 leading to reduced apoptosis and transformation of infected cells to cancer cells (Lin et al., 2007). Most septins are localized at the cytosol although nuclear localization is seen with some. In contrast SEPT4 i2 localizes to mitochondria in healthy cells. Another splice variant of the SEPT 4 gene, M-septin, is also reported to be localized in mitochondria of developing mouse neurons. M-septin was not shown

REFERENCES

289

to be involved in apoptosis (Takahashi et al., 2003). Therefore, it seems the localization of septins in mitochondria per se does not necessarily indicate an apoptotic function. In the case of SEPT4 i2, its mitochondrial confinement serves as a compartmentalization mechanism used to prevent its interaction with IAPs which reside in the cytosol, interaction that could induce unwanted apoptosis via a C
-terminus unique stretch of 27aa which is not present in any other septin or any other yet known protein.

ACKNOWLEDGEMENT We thank Dr Yael Mandel-Gutfreund for her contribution regarding description of the intron retention phenomenon included in this chapter. We want to acknowledge the support of ISF (Israel Science Foundation) and BSF (Bi-national Science Foundation US-Israel) to S. L. which enabled the work on SEPT4 V2/ARTS presented in this chapter.

REFERENCES Bergmann, A., Agapite, J., Mccall, K. and Steller, H. (1998) The Drosophila gene hid is a direct molecular target of Ras-dependent survival signaling. Cell , 95, 331–41. Bourne, H.R., Sanders, D.A. and Mccormick, F. (1991) The GTPase superfamily: conserved structure and molecular mechanism. Nature, 349, 117–27. Clem, R.J., Fechheimer, M. and Miller, L.K. (1991) Prevention of apoptosis by a baculovirus gene during infection of insect cells. Science, 254, 1388–90. Clem, R.J. and Miller, L.K. (1994) Control of programmed cell-death by the baculovirus genes P35 and Iap. Molecular and Cellular Biology, 14, 5212–22. Du, C., Fang, M., Li, Y. et al. (2000) Smac, a mitochondrial protein that promotes cytochrome c-dependent caspase activation by eliminating IAP inhibition. Cell , 102, 33–42. Duckett, C.S., Nava, V.E., Gedrich, R.W. et al. (1996) A conserved family of cellular genes related to the baculovirus iap gene and encoding apoptosis inhibitors. EMBO Journal , 15, 2685–94. Elhasid, R., Sahar, D., Merling, A. et al. (2004) Mitochondrial pro-apoptotic ARTS protein is lost in the majority of acute lymphoblastic leukemia patients. Oncogene, 23, 5468–75. Galante, P.A., Sakabe, N.J., Kirschbaum-Slager, N. and De Souza, S.J. (2004) Detection and evaluation of intron retention events in the human transcriptome. RNA, 10, 757–65. Golstein, P. (1997) Cell death: TRAIL and its receptors. Current Biology, 7, R750–53. Gottfried, Y., Rotem, A., Lotan, R. et al. (2004) The mitochondrial ARTS protein promotes apoptosis through targeting XIAP. EMBO Journal , 23, 1627–35. Green, D.R. and Kroemer, G. (2004) The pathophysiology of mitochondrial cell death. Science, 305, 626–29.

290

CH13 SEPTINS AND APOPTOSIS

Hall, P.A., Jung, K., Hillan, K.J. and Russell, S.E. (2005) Expression profiling the human septin gene family. The Journal of Pathology, 206, 269–78. Hall, P.A. and Russell, S.E. (2004) The pathobiology of the septin gene family. The Journal of Pathology, 204, 489–505. Hay, B.A., Wassarman, D.A. and Rubin, G.M. (1995) Drosophila homologs of baculovirus inhibitor of apoptosis proteins function to block cell death. Cell , 83, 1253–62. Hegde, R., Srinivasula, S.M., Zhang, Z. et al. (2002) Identification of Omi/HtrA2 as a mitochondrial apoptotic serine protease that disrupts inhibitor of apoptosis protein-caspase interaction. Journal of Biological Chemistry, 277, 432–38. Hsu, S.C., Hazuka, C.D., Roth, R. et al. (1998) Subunit composition, protein interactions, and structures of the mammalian brain sec6/8 complex and septin filaments. Neuron, 20, 1111–22. Huang, Y., Park, Y.C., Rich, R.L. et al. (2001) Structural basis of caspase inhibition by XIAP: differential roles of the linker versus the BIR domain. Cell , 104, 781–90. Jacobson, M.D., Weil, M. and Raff, M.C. (1997) Programmed cell death in animal development. Cell , 88, 347–54. Joberty, G., Perlungher, R.R., Sheffield, P.J. et al. (2001) Borg proteins control septin organization and are negatively regulated by Cdc42. Nature Cell Biology, 3, 861–66. Kartmann, B. and Roth, D. (2001) Novel roles for mammalian septins: from vesicle trafficking to oncogenesis. Journal of Cell Science, 114, 839–44. Kerr, J.F., Wyllie, A.H. and Currie, A.R. (1972) Apoptosis: a basic biological phenomenon with wide-ranging implications in tissue kinetics. The British Journal of Cancer, 26, 239–57. Kinoshita, M. (2003) The septins. Genome Biology, 4, 236. Kinoshita, M., Kumar, S., Mizoguchi, A. et al. (1997) Nedd5, a mammalian septin, is a novel cytoskeletal component interacting with actin-based structures. Genes and Development, 11, 1535–47. Kinoshita, A., Noda, M. and Kinoshita, M. (2000) Differential localization of septins in the mouse brain. The Journal of Comparative Neurology, 428, 223–39. Kissel, H., Georgescu, M.M., Larisch, S. et al. (2005) The Sept4 septin locus is required for sperm terminal differentiation in mice. Developmental Cell , 8, 353– 64. Kondrashov, F.A. and Koonin, E.V. (2003) Evolution of alternative splicing: deletions, insertions and origin of functional parts of proteins from intron sequences. Trends in Genetics, 19, 115–19. Kumar, S. (1995) Ice-like proteases in apoptosis. Trends in Biochemical Sciences, 20, 198–202. Larisch, S., Yi, Y., Lotan, R. et al. (2000) A novel mitochondrial septin-like protein, ARTS, mediates apoptosis dependent on its P-loop motif. Nature Cell Biology, 2, 915–21. Larisch-Bloch, S., Danielpour, D., Roche, N.S. et al. (2000) Selective loss of the transforming growth factor-beta apoptotic signaling pathway in mutant NRP-154 rat prostatic epithelial cells. Cell Growth and Differentiation, 11, 1–10.

REFERENCES

291

Lee, J.W., Soung, Y.H., Kim, S.Y. et al. (2006a) Mutational analysis of P-loop domains of proapoptotic Nod1 and ARTS genes in colon carcinomas. Acta Oncologica, 45, 101–2. Lee, J.W., Soung, Y.H., Young Kim, S. et al. (2006b) Mutational analysis of proapoptotic ARTS P-loop domain in common human cancers. Pathology –Research and Practice, 202, 67–70. Lin, C.W., Tu, P.F., Hsiao, N.W. et al. Identification of a novel septin 4 protein binding to human herpesvirus 8 kaposin A protein using a phage display cDNA library. (2007) Journal of Virological Methods, 143, 65–72. Lindsey, R. and Momany, M. (2006) Septin localization across kingdoms: three themes with variations. Current Opinion in Microbiology, 9, 559–65. Lotan, R., Rotem, A., Gonen, H. et al. (2005) Regulation of the proapoptotic ARTS protein by ubiquitin-mediated degradation. Journal of Biological Chemistry, 280 (27), 25802–10. Martinou, J.C., Dubois-Dauphin, M., Staple, J.K. et al. (1994) Overexpression of BCL-2 in transgenic mice protects neurons from naturally occurring cell death and experimental ischemia. Neuron, 13, 1017–30. Nagata, S. (1997) Apoptosis by death factor. Cell , 88, 355–65. Naik, P., Karrim, J. and Hanahan, D. (1996) The rise and fall of apoptosis during multistage tumorigenesis: down-modulation contributes to tumor progression from angiogenic progenitors. Genes and Development, 10, 2105–16. Nicholson, D.W. and Thornberry, N.A. (1997) Caspases: killer proteases. Trends in Biochemical Sciences, 22, 299–306. Oppenheim, R.W. (1991) Cell death during development of the nervous system. Annual Review of Neuroscience, 14, 453–501. Peng, X.R., Jia, Z., Zhang, Y. et al. (2002) The septin CDCrel-1 is dispensable for normal development and neurotransmitter release. Molecular and Cellular Biology, 22, 378–87. Pettmann, B. and Henderson, C.E. (1998) Neuronal cell death. Neuron, 20, 633–47. Raff, M.C. (1992) Social controls on cell survival and cell death. Nature, 356, 397–400. Raff, M. (1998) Cell suicide for beginners. Nature, 396, 119–22. Reed, J.C. (1995) Regulation of apoptosis by bcl-2 family proteins and its role in cancer and chemoresistance. Current Opinion in Oncology, 7, 541–46. Rothe, M., Pan, M.G., Henzel, W.J. et al. (1995) The TNFR2-TRAF signaling complex contains two novel proteins related to baculoviral-inhibitor of apoptosis proteins. Cell , 83, 1243–52. Roy, N., Deveraux, Q.L., Takahashi, R. et al. (1997) The c-IAP-1 and c-IAP-2 proteins are direct inhibitors of specific caspases. EMBO Journal , 16, 6914–25. Roy, N., Mahadevan, M.S., Mclean, M. et al. (1995) The Gene for neuronal apoptosis inhibitory protein is partially deleted in individuals with spinal muscular-atrophy. Cell , 80, 167–78. Salvesen, G.S. and Dixit, V.M. (1997) Caspases: intracellular signaling by proteolysis. Cell , 91, 443–46. Salvesen, G.S. and Duckett, C.S. (2002) IAP proteins: blocking the road to death’s door. Nature Reviews. Molecular Cell Biology, 3, 401–10.

292

CH13 SEPTINS AND APOPTOSIS

Saraste, M., Sibbald, P.R. and Wittinghofer, A. (1990) The P-loop – a common motif in ATP- and GTP-binding proteins. Trends in Biochemical Sciences, 15, 430–34. Sheffield, P.J., Oliver, C.J., Kremer, B.E. et al. (2003) Borg/septin interactions and the assembly of mammalian septin heterodimers, trimers, and filaments. Journal of Biological Chemistry, 278, 3483–88. Shi, Y. (2002) A conserved tetrapeptide motif: potentiating apoptosis through IAPbinding. Cell Death and Differentiation, 9, 93–95. Spiliotis, E.T. and Nelson, W.J. (2006) Here come the septins: novel polymers that coordinate intracellular functions and organization. Journal of Cell Science, 119 (pt 1), 4–10. Srinivasula, S.M., Ahmad, M., Fernandes-Alnemri, T. et al. (1996) Molecular ordering of the Fas-apoptotic pathway: the Fas/APO-1 protease Mch5 is a CrmA-inhibitable protease that activates multiple Ced-3/ICE-like cysteine proteases. Proceedings of the National Academy of Sciences of the United States of America, 93, 14486–91. Steller, H. (1995) Mechanisms and genes of cellular suicide. Science, 267, 1445–49. Suzuki, Y., Imai, Y., Nakayama, H. et al. (2001) A serine protease, HtrA2, is released from the mitochondria and interacts with XIAP, inducing cell death. Molecular Cell , 8, 613–21. Takahashi, S., Inatome, R., Yamamura, H. and Yanagi, S. (2003) Isolation and expression of a novel mitochondrial septin that interacts with CRMP/CRAM in the developing neurones. Genes to Cells, 8, 81–93. Tanaka, M., Kijima, H., Itoh, J. et al. (2002) Impaired expression of a human septin family gene Bradeion inhibits the growth and tumorigenesis of colorectal cancer in vitro and in vivo. Cancer Gene Therapy, 9, 483–88. Tanaka, M., Tanaka, T., Kijima, H. et al. (2001) Characterization of tissue- and cell-type-specific expression of a novel human septin family gene, Bradeion. Biochemical and Biophysical Research Communications, 286, 547–53. Tanaka, M., Tanaka, T., Matsuzaki, S. et al. (2003) Rapid and quantitative detection of human septin family Bradeion as a practical diagnostic method of colorectal and urologic cancers. Medical Science Monitor, 9, MT61–68. Thompson, C.B. (1995) Apoptosis in the pathogenesis and treatment of disease. Science, 267, 1456–62. Thornberry, N.A. and Lazebnik, Y. (1998) Caspases: enemies within. Science, 281, 1312–16. Thornberry, N.A., Rosen, A. and Nicholson, D.W. (1997) Control of apoptosis by proteases, in Apoptosis (ed S.H. Kaufmann), Academic Press, San Diego. Tittel, J.N. and Steller, H. (2000) A comparison of programmed cell death between species (Reviews). Genome Biology, 1, 0003. Truman, J.W., Thorn, R.S. and Robinow, S. (1992) Programmed neuronal death in insect development. Journal of Neurobiology, 23, 1295–311. Vaux, D.L. and Korsmeyer, S.J. (1999) Cell death in development. Cell , 96, 245–54. Vaux, D.L. and Silke, J. (2003) Mammalian mitochondrial IAP binding proteins. Biochemical and Biophysical Research Communications, 304, 499–504. Verhagen, A.M., Ekert, P.G., Pakusch, M. et al. (2000) Identification of DIABLO, a mammalian protein that promotes apoptosis by binding to and antagonizing IAP proteins. Cell , 102, 43–53.

REFERENCES

293

Verhagen, A.M., Silke, J., Ekert, P.G. et al. (2002) HtrA2 promotes cell death through its serine protease activity and its ability to antagonize inhibitor of apoptosis proteins. Journal of Biological Chemistry, 277, 445–54. Wallach, D., Boldin, M., Varfolomeev, E. et al. (1997) Cell death induction by receptors of the TNF family: towards a molecular understanding. FEBS Letters, 410, 96–106. White, E. (1996) Life, death, and the pursuit of apoptosis. Genes and Development, 10, 1–15. Wyllie, A.H. (1980) Glucocorticoid-induced thymocyte apoptosis is associated with endogenous endonuclease activation. Nature, 284, 555–56.

14 Septins and human disease Peter A. Hall Institute of Pathology, School of Medicine, Queen’s University Belfast, Belfast, Northern Ireland UK BT12 6BL

Fern P. Finger Department of Biology and Center for Biotechnology and Interdisciplinary Studies, Rensselaer Polytechnic Institute, 110 Eighth Street, Troy, NY 12180-3590, USA

INTRODUCTION: CORRELATION IS NOT CAUSATION A range of stimuli have led to the growth in interest in septin biology. Elsewhere in this monograph compelling evidence for the diverse and complex roles of septin proteins in key cellular processes in animal and fungal cells have been described. Here we focus on another major aspect of the growth in interest in septins over the past decade: their possible role in human disease. The figure in the introductory chapter of this work shows a rapid increase in publications in which the term septin is used. A large proportion of these publications are studies in which alterations in septin expression have been observed in disease, whether this be by proteomic, immunochemical or gene expression analyses. Much of the data is correlative in nature and in only very few situations are there data that suggest that septins have a pathogenetic role in disease. The purpose of this chapter is to consider the development of the idea that septins may have a role in disease and to endeavour to assess critically the data supporting this hypothesis (see Table 14.1). It may be that this is but the tip of an iceberg since, hitherto, lack of critical tools has prevented the study of septins in pathology. With the exception of disease-associated mutations in hereditary neuralgic amyotophy (HNA) and specific gene rearrangements involving septin genes and MLL in leukaemias, all of the other alterations of septins reported in human diseases are, at present, phenomena of unproven significance. Some may indeed prove to be of The Septins Edited by Peter A. Hall, S.E. Hilary Russell and John R. Pringle © 2008 John Wiley & Sons, Ltd. ISBN: 978-0-470-51969-1

CH14 SEPTINS AND HUMAN DISEASE

296

Table 14.1 Human Septins and Disease. Data from non-human systems are not included. Entries were compiled from PubMed searches for terms including “septin”, “septins”, and each individual human septin gene. While we have endeavored to make this table comprehensive, we apologize to any colleagues whose work we have inadvertantly omitted Disease

Type

Hereditary Neuralgic Amyotrophy Cancer

Leukemia and lymphoma

Septin(s)

Evidence

References

SEPT9

Diseaseassociated mutations in exons Translocation partner of MLL

(Kuhlenbaumer et al., 2005)

SEPT2

SEPT5 SEPT6

SEPT9

Translocation partner of MLL

SEPT11

Translocation partner of MLL Allelic imbalance in sporadic cancers; altered expression pattern in tumors. Proteomic data indicates upregulation, changes in form expressed Proteomic data indicates upregulation Ectopic expression in melanomas Upregulation in melanoma metastases

Ovarian and breast epithelial cancers

SEPT9

Renal cell carcinoma

SEPT2

SEPT11

Melanoma

Translocation partner of MLL Translocation partner of MLL

SEPT4 SEPT6

(Cerveira et al., 2006; van Binsbergen et al., 2007) (Tatsumi et al., 2001) (Borkhardt et al., 2001; Fu et al., 2003; Kim et al., 2003; Ono et al., 2002; Shih et al., 2006; Slater et al., 2002; Strehl et al., 2006) (Kreuziger et al., 2007; Osaka et al., 1999; Yamamoto et al., 2002) (Kojima et al., 2004) (Gonzalez et al., 2007; Kalikin et al., 2000; Russell et al., 2000; Scott et al., 2005)

(Craven et al., 2006a; Craven et al., 2006b)

(Craven et al., 2006b)

(Tanaka et al., 2001) (Jaeger et al., 2007)

INTRODUCTION: CORRELATION IS NOT CAUSATION Table 14.1 Disease

(continued) Type

Septin(s)

Evidence

References

Colorectal SEPT1 cancer SEPT4

Proteomic data indicates upregulation Ectopic expression in melanomas

Oral cancer Brain tumors

Proteomic data indicates upregulation Proteomic data indicates upregulation in glioblastoma; variable expression in brain tumors and brain tumor cell lines. RT-PCR indicates loss of expression in testicular cancer cell lines Found in Lewy bodies

(Ying-Tao et al., 2005) (Tanaka et al., 2001) (Kato et al., 2007) (Khalil, 2007; Kim et al., 2004; Sakai et al., 2002)

SEPT1 SEPT2

Testicular SEPT14 cancer Parkinson’s Disease

SEPT4 SEPT5

Alzheimer’s Disease

Down’s syndrome

297

SEPT1

Binds to and is a substrate for ubiquitination by Parkin, resulting in its degradation Found in tangles

SEPT2

Found in tangles

SEPT3 SEPT4

Polymorphisms of exon 11 associated with AD Found in tangles

SEPT7

Not expressed in PD brains

SEPT3

Proteomic data indicates decreased expression in fetal Down’s syndrome cerebral cortex Proteomic data indicates decreased expression in fetal Down’s syndrome brains Proteomic data indicates decreased expression in fetal Down’s syndrome cerebral cortices.

SEPT6

SEPT7

(Peterson et al., 2007) (Ihara et al., 2003) (Choi et al., 2003; Zhang et al., 2000) (Kinoshita et al., 1998) (Kinoshita et al., 1998) (Takehashi et al., 2004) (Kinoshita et al., 1998) (Kinoshita et al., 1998) (Pollak et al., 2003)

(Cheon et al., 2001) (Engidawork et al., 2003)

(continued overleaf)

CH14 SEPTINS AND HUMAN DISEASE

298 Table 14.1

(continued)

Disease

Schizophrenia

Type

Septin(s)

Evidence

References

SEPT9

Proteomic data indicates increased expression in fetal Down’s syndrome cerebral cortices. Decreased expression in brains of schizophrenic patients Proteomic data indicates increased expression of three forms in dorsolateral prefrontal cortex Proteomic data indicates increased expression of one form in dorsolateral prefrontal cortex Proteomic data indicates increased expression of five forms in dorsolateral prefrontal cortex Proteomic data indicates increased expression in dorsolateral prefrontal cortex Proteomic data indicates increased expression of two forms in dorsolateral prefrontal cortex Proteomic data indicates increased expression in hippocampus

(Engidawork et al., 2003)

SEPT4 SEPT5

SEPT11

Bipolar Disorder

SEPT5

SEPT6

SEPT11

Mesial Temporal Lobe Epilepsy Systemic lupus erythematosus Listeria infection Human Herpes Virus 8 Hepatitis C

SEPT3

SEPT2

SEPT9 SEPT4 SEPT6

Autoantigen in patients with SLE associated with presence of psychiatric symptoms Recruited onto phagosomes internalizing Listeria Binds to HHV-8 kaposin A

(Gottfried et al., 2007) (Pennington et al., 2007) (Pennington et al., 2007) (Pennington et al., 2007) (Pennington et al., 2007) (Pennington et al., 2007) (Yang et al., 2006)

(Margutti et al., 2005)

(Pizarro-Cerda et al., 2002) (Lin et al., 2007) Binds to HCV N5b protein and to (Kim et al., 2007) the 5
and 3
NTRs of the HCV RNA containing elements required for replication; knockdown by siRNA or overexpression of N-terminally truncated SEPT6 blocks HCV replication in cell culture.

pathogenetic relevance in disease but for the most part we cannot as yet exclude the possibility that they reflect epiphenomena that are a consequence of the disease process rather than the cause. This would not make them wholly uninteresting events (indeed the characterization of septin changes in disease may tell us much of the normal physiological roles of septins in cells), but of course

SEPT9 MUTATIONS IN A HUMAN GENETIC DISEASE

299

epiphenomena would be much less significant in terms of the understanding of human disease, or in their diagnosis, monitoring or therapy. We will return to this broad issue at the end of this chapter when we consider how to best move forward with the difficult business of attributing cause and effect to septin perturbations in disease. Before this, we will overview the accumulating descriptions of septin alterations in human disease, beginning with disease-associated mutations in SEPT9 in a rare neurological disease and the abundant data on septins in neoplasia.

SEPT9 MUTATIONS IN A HUMAN GENETIC DISEASE The identification of mutations in a human septin gene and the segregation of those mutations with diseased individuals was an important observation (Kuehlenbaumer et al., 2005). The disease in question HNA is a very rare autosomal dominant condition that causes episodic pain and paresis in the distribution of the nerves of the brachial (and to a lesser extent sacral) plexus. This condition can be exacerbated by stresses including repeated use of the limbs, strenuous exertion, childbirth and inter-current infection. In some individuals there is an association with facial malformations but this is very variable and the exact relationship with the neurological phenotype is unclear (Chance, 2006). An international consortium endeavoured to map the gene(s) responsible for HNA and indeed had at one time excluded the SEPT9 locus on chromosome 17q (Meuleman et al., 2001). However further analyses reopened the possibility that SEPT9 might indeed be the disease associated locus. Kuhlenbaumer et al. (2005) provided clear data of disease-associated mutations in 10 kindreds with HNA and mapped the mutations to exons of SEPT9 . As discussed in Chapter 7 by Hilary Russell, although those exons are used as open reading frames in some SEPT9 transcripts, in others they are part of the 5
UTR (Kuhlenbaumer et al., 2005; McDade, Hall and Russell, 2007). Sudo et al. (2007) have suggested that the HNA associated missense mutations alter interactions of SEPT9 protein with SEPT4 and SEPT11 and perturb Rho/rhotekin signalling but the evidence is not compelling. An alternate view is presented in Chapter 7 and in McDade, Hall and Russell (2007) where it is suggested that the mutations alter the translational efficiency of transcripts encoding SEPT9 i4, particularly under conditions of stress (and the clinical features of HNA are potentiated by stress: Chance, 2006). This protein is over-expressed in neoplasia and also can alter microtubule dynamics (Russell and Hall, 2005). It is intriguing that the features of HNA are associated with long axons of the brachial plexus that rely on efficient microtubule function. This should be amenable to experimental test. However it should also be pointed out that we cannot yet be certain that the primary pathology is in neurones as opposed to the surrounding Schwann cells. Whatever the pathogenetic

300

CH14 SEPTINS AND HUMAN DISEASE

mechanism and cell type involved, it is certain that HNA is associated with SEPT9 mutations.

SEPTINS IN NEOPLASIA MLL fusions: smoking guns? The very first diseases associated with human septins were leukaemia where balanced translocations involving the MLL locus on chromosome 11 and the SEPT9 locus on chromosome 17 (Osaka, Rowley and Zeleznik-Le, 1999) gave rise to a fusion gene that could produce chimeric fusion proteins where the N terminus of MLL was linked, in frame, to almost the entire open reading frame of the long forms of SEPT9 . Since that first report, other septins (notably SEPT2, SEPT5, SEPT6 and SEPT11 ) have been found to form similar fusion proteins with MLL: again with the N terminal moiety of MLL fused to almost the entire open reading frame of the partner septin (Megonigal et al., 1998; Taki et al., 1999; Borkhardt et al., 2001; Ono et al., 2002; Yamamoto et al., 2002; Slater et al., 2002; Kim et al., 2004; Fu et al., 2003; Kojima et al., 2004; Cerveira et al., 2006; Kadkol et al., 2006; van Binsbergen et al., 2007; Kreuziger et al., 2007; Strehl et al., 2006). MLL is a remarkably promiscuous gene, forming in-frame chimeras with more than 50 other genes (van Binsbergen et al., 2007). These fusion partners fall into two groups: those that possess potential oligomerization motifs and those with a potent transactivation domain (So et al., 2003; Slany, 2005). The septins are not known to possess a transactivation domain but can oligomerise via their GTP-binding regions (see Salluhadin et al., 2007 and Chapter 3). There is data to support the idea that oligomerization by the septin moiety of MLL fusions is important (Ono et al., 2005) and this oligomerization may occur via the septin GTP-binding domain. A further point of note is the existence of predicted nuclear localization signals (NLS) in several septins, notably SEPT9 i1, all SEPT6 isoforms and in SEPT11. However other data make the situation more confusing since neither SEPT5 nor SEPT2 have NLS in their currently known isoforms. Moreover, complex splicing events lead to the production of heterogeneous fusion transcripts (Strehl et al., 2006). In addition at least in the case of SEPT6, wild type protein is not essential for MLL-SEPT6 fusion protein function (Ono et al., 2005). Thus present data indicate that MLL-septin fusions do have a pathogenetic role in at least some human haematological neoplasias. The simplest interpretation is that the septin moiety merely represents an oligomerization domain that can either directly (via an NLS in the septin fusion) or indirectly (by carriage with another NLS containing septin) mediate nuclear transport of MLL and its activation by dimerisation. This might suggest that septins per se do not have oncogenic activity, at least in this context.

SEPTINS IN NEOPLASIA

301

Linking SEPT9 and cancer As well as being seen in translocations with MLL, SEPT9 was initially linked to neoplasia by two other observations. First, the murine Sept9 locus is an integration site for the SL-3 retrovirus in T-cell lymphomas (Sorensen et al., 2000). Second, the human SEPT9 locus at 17q25.3 was identified as a common site for allelic imbalance in sporadic ovarian and breast cancer (Russell et al., 2000; Kallikin et al., 2001). Both observations would be expected to be associated with perturbed expression of SEPT9 isoforms. Indeed, while mutations have not been observed in the known open reading frames of SEPT9 , there is abundant evidence pointing to altered expression of SEPT9 in diverse tumour types (Scott et al., 2005) and particularly in ovarian and breast tumours (Burrows et al., 2003; Montagna et al., 2003; Robertson et al., 2004; Scott 2006; Gonzalez et al., 2007). Of note is the observation that neoplasia is associated not just with altered expression of SEPT9 but also with alterations in the expression of specific SEPT9 transcripts, with the SEPT9 v4 transcript being predominant in normal tissues but being replaced by SEPT9 v4 in tumours (see Chapter 7; Burrows et al., 2003; Scott et al., 2005). These transcripts encode the same polypeptide but differ in their 5
UTR sequences. The SEPT9 v4 transcript appears to be translated more efficiently than the SEPT9 v4 transcript (McDade, Hall and Russell, 2007) and thus this change in transcript profile has a profound effect on the level of the SEPT9 i4 isoform and its regulation (see Chapter 7). Chacko et al. (2005) had previously shown that over-expression of this protein could perturb epithelial cell morphology, motility and polarity. This is perhaps not surprising given that nematode septins can influence the directional movement of developing neurons and also the distal tip cells of the developing gonad (Finger, Kopish and White, 2003), and the role of septins in determining polarity in yeast (Irazoqui and Lew, 2004). Alterations in SEPT9 v1 and SEPT9 i1 have also been reported in neoplasia (Scott et al., 2005, 2006; Amir et al., 2007; Gonzalez et al., 2007). Using in vitro models Gonzalez et al. (2007) have shown that SEPT9 i1 expression can stimulate cell proliferation, induce altered cell motility and promote genomic instability. Furthermore siRNA mediated knock down of this isoform abrogates such responses: observations consistent with an involvement of septins in the neoplastic phenotype. It is also of note that Amir and Mabjeesh (2007) have reported that SEPT9 i1 expression promotes resistance to drugs that disrupt microtubules. The same group had previously made a crucial observation in the mammalian septin field: the identification of a complex of SEPT9 i1 and hypoxia induced factor 1 alpha (HIF1α). HIF1α is a key transcriptional regulator of genes involved in cellular response to stresses but in particular to hypoxia. SEPT9 i1 stabilises HIF1α preventing its ubiquitin-mediated degradation, even in the absence of hypoxic stress. As a consequence, the programme of genes activated by HIF1α in response to hypoxic stress is inappropriately activated (Amir et al., 2006). These data place SEPT9 in a quite new perspective and introduce the idea of mammalian septins being involved in signalling events and pathways, either

302

CH14 SEPTINS AND HUMAN DISEASE

directly or as components of scaffolds (as had been suggested before including in Hall and Russell, 2004). How do SEPT9 proteins function in this context? Is it as a component of a septin filament or is it as part of a non-filamentous complex? What associated moieties are involved and to what extent might the nature of the SEPT9 isoforms and their relative levels contribute to these activities? At present we simply do not know, but the complexity of spatial localization of septins in cells, and in particular in mammals (Lindsey and Momany, 2006), makes simple answers to such questions unlikely. The details of the stoichiometry and dynamics of such complexes and their interactions will be of great interest. While the original search for cancer associated genes on 17q by the Russell and the Petty groups focussed on the proposition that loss of heterozygosity (LOH) was seen, which presupposed that the loci harboured classical Knudsen style tumour suppressor genes, the currently available data are not consistent with that view. Re-examination of the original LOH data suggests that a more correct description of the observed changes at 17q25.3 in ovarian tumours is allelic imbalance (see Hall and Russell 2004; Russell and Hall, 2005). The situation is as yet unresolved, but it may be that septins, as exemplified by SEPT 9, are involved in neoplasia as a consequence of altered expression levels and should considered as class II cancer genes, as defined by Sager (1997), affecting phenotype as a consequence of changes in expression. Such mechanisms might be particularly relevant in the context of genes whose products are involved in complex formation and where stoichiometry and the detail of interactions are critical for subtle control of function. Certainly it is the case that in global screens of cancer associated mutations septins are vanishingly rare (Wood et al., 2007), however this strategy examined specifically mutations in known exons. Can it be confidently assumed that polymorphisms and/or somatic mutations in non-coding regions do not contribute to alterations in expression of key genes (including septins)? Since it is perturbations in the levels of transcripts that seems to be a feature of septin expression in neoplasia, the hypothesis that disease-associated genetic alterations may not be restricted to exons deserves serious consideration. Moreover, while it is the case that mutations are crucial to the role of many oncogenes and tumour suppressor genes in neoplasia, it is clear that the neoplasia can be a consequence of altered gene expression with haplo-insufficiency being an increasingly common theme (Bartek, Lukas and Bartkova, 2007) coupled with environmental factors, often having a complex spatial and temporal interplay (Orimo et al., 2005; Cook et al., 2005). Furthermore the surprising revelation of how relatively few genes are present in the human genome, and of the extent to which alternate splicing occurs, highlights the possibility that the range of genes whose products will have a role in neoplasia will continue to grow as we develop a more detailed understanding of the molecular events that regulate cells. Moreover, in some situations the distinction between oncogenes and tumour suppressor genes are blurred with the diverse products of the one gene having different and potentially opposing functions.

SEPTINS IN NEOPLASIA

303

Other septins and neoplasia Can these ideas be relevant to other human septins and neoplasia? Our current catalogue of septin expression changes seen in neoplasia (and other disease states) remains far from complete and the complex splicing events seen in the septins makes progress difficult. Nevertheless the available data suggests that at least some septins can be implicated in human (and murine) neoplasia. Proteomic studies have found over-expression of SEPT2 in 12 of 16 renal cell carcinomas and in 1 of 3 there was an abnormal mobility species suggesting an altered modification of SEPT2 (Craven et al., 2006a). Similar changes had also been seen in cell lines (Craven et al., 2006a, 2006b). SEPT2 was shown to be required for cytokinesis and to bind actin and associate with focal adhesions (Kinoshita et al., 1997). Spiliotis, Kinoshita and Nelson (2005) reported that SEPT 2 has a role in chromosome congression and segregation and that its altered expression might lead to disordered chromosomal dynamics and underlie the development of aneuploidy. SEPT6 expression has been reported as part of a ‘signature’ for tumour progression in melanoma (Jaeger et al., 2007) and has been associated with malignancy in lymphoid tissues (Todd, Russell, Hall et al., unpublished observations). A problem with the former data however is that it is underpowered (19 primary tumours and 22 metastases) given the multiple assessments made and the lack of discrimination of the complex transcripts of SEPT6 (see Chapter 7). Another approach of uncertain relevance is the use of proteomics methods for tumour profiling. A number of reports describe such approaches (Khalil and James, 2007) and while differences can be identified in the proteome (and some of these differences are in septins), it is hard to be confident of the biological relevance of such observations. SEPT5 expression has been reported in pancreatic tumours (Capurso et al., 2005) and differences in septin profile have been suggested between primary and metastatic tumours (Ying-Tao et al., 2005). SEPT1 mRNA and protein have been reported to be over-expressed in oral cancer (Kato et al., 2007) and aurora kinase has been shown to interact with SEPT1 protein (Qi et al., 2005). Several reports suggest that altered expression of SEPT4 isoforms (referred to as Bradeion) may be associated with the neoplastic phenotype (Tanaka et al., 2001, 2002, 2003) and several septins have altered expression in brain tumours (Sakai et al., 2002; Kim et al., 2005) and other tumours (Hall et al., 2005). Larisch et al. (2000) independently observed that the SEPT4 v2 transcript (encoding the SEPT4 i2 isoform formerly known as ARTS) could promote TGFβ mediated apoptosis (see Chapter 13). This isoform binds to and can modulate the function of XIAP and thus promote apoptosis (Gottfried et al., 2004). This SEPT4 protein (which is distinct from those reported by Tanaka et al ) might function as a tumour suppressor since expression is lost in most cases of childhood acute lymphoblastic leukaemia (ALL); Elhasid et al., 2004). The apparent absence of a tumour-prone phenotype in Sept4 null mice (Ihara et al., 2005) does not necessarily exclude a role for SEPT4 in neoplasia, as alterations in expression of specific Sept4

304

CH14 SEPTINS AND HUMAN DISEASE

transcripts have not yet been explored, and this further underscores the need for transcript specific analysis of septins. Alternatively, there may be significant differences between man and mouse. Finally, it is of note that the SEPT4 gene (and possibly other septins) can be regulated by p53. Using a subtraction hybridization approach in which a p53 allele was introduced into a p53 null cell, differentially expressed mRNAs were cloned, one of which was SEPT 4 (Kostic and Shaw, 2000). Given the existence of a G2/M checkpoint regulated by p53 this observation is provocative and deserves further study. It may link the p53 tumour suppressor to diverse aspects of cell biology via the septins. In this context it should be noted that the actin-based phenotypic effects of Cdc42 such as filopodia formation, cell polarization, spreading and motility can be regulated by p53 (Gad´ea et al., 2002). Septins have also been reported to associate with the metastasis associated gene product S100A4 (Koshelev, Kiselev and Georgiev, 2003) and it is well recognised that the cytoskeleton is implicated in cell transformation (Pawlak and Helfman, 2001). An exciting recent observation has connected septins and the DNA damage response (see Chapter 9 by Kremer, Haystead and Macara, (2005); Kremer, Adang and Macara, (2007); Kinoshita and Takeda, (2007)). This is of great interest since morphological changes are important in stress responses and are a characteristic feature of neoplasia. Elucidation of the links between septins and other oncogene and tumour suppressor gene pathways is likely to be a fertile area of study in the coming years.

SEPTINS AND NEUROPATHOLOGY A detailed consideration of septins and the brain is presented in Chapter 11 and in the following we merely describe some of the general points relating to septins and neuropathology. A role for septins in neurological disorders has emerged from several observations including the brain specific expression of some septins, the differential regulation of septins in neural development and the association of septins with some disease states (Tada et al., 2007; Caltagarone et al., 1998; Dong et al., 2003; Xiong, Leahy and Stuhlmann, 1999; Kinoshita, Noda and Kinoshita, 2000; Methner et al., 2001; Hall et al., 2005; Kinoshita et al., 1998). With the exception of HNA, the most abundant evidence for a connection between septins and neuropathology is for Parkinson’s disease (PD), the secondmost prevalent neurodegenerative disorder, clinically manifested by symptoms including slowness of movement, muscular rigidity and tremors (Gandhi and Wood, 2005). PD pathology is characterized by loss of dopaminergic (DA) neurons and the accumulation of cytoplasmic inclusions, Lewy bodies, containing protein aggregates of which the major constituent is α-synuclein (Wood-Kaczmar, Gandhi and Wood, 2006). The initial connection to PD came from the demonstration that SEPT5 binds in vitro and in cultured cells to Parkin, a ubiquitin ligase mutated in some forms of autosomal-recessive juvenile parkinsonism. Moreover, SEPT5 was demonstrated to be a substrate for ubiquitination by Parkin, resulting

SEPTINS AND NEUROPATHOLOGY

305

in SEPT5 degradation (Zhang et al., 2000). This suggested the hypothesis that the failure to degrade SEPT5 could promote PD. Subsequent studies confirmed that SEPT5 i2 in lysates from human substantia nigra bound to Parkin (Choi et al., 2003), and that over-expression of SEPT5 in rat DA neurons (Dong et al., 2003) or cultured DA neuronal cell lines (Son et al., 2005) caused cytotoxicity, which could be relieved in the cultured cells through co-expression of Parkin (Son et al., 2005). Transgenic mice expressing a dominant negative Parkin mutant suffered motor dysfunction and had dopamine levels 25 % higher than normal, similar to observations in Parkin null mice (Son et al., 2005). However, although there were moderate accumulations of SEPT5, there was no apparent loss of DA neurons (Son et al., 2005). Taken together with the observations in transgenic mice that direct over-expression of high levels of SEPT5 in DA neurons resulted in embryonic lethality, SEPT5 accumulation over a threshold level was suggested to be directly neurotoxic (Son et al., 2005). Over-expression of septin 4 , encoding the Drosophila SEPT5 homolog was found to be similarly toxic to DA neurons in vivo, and the Drosophila homologs of Parkin and this septin also physically interact in vitro (Mu˜noz-Soriano and Paricio, 2007), providing additional support for the hypothesis that SEPT5 may be a bona fide Parkin substrate, and that its over-expression directly contributes to PD. A second human septin, SEPT4 , is also implicated in PD. Uniquely among the human septins, SEPT4 was found in Lewy bodies associated with the sporadic form of Parkinson’s disease and in the cytoplasmic inclusions found in two other human synucleinopathies (Ihara et al., 2003). Elegant work by the Kinoshita group demonstrated that Sept4 null mice are defective in DA neurotransmission caused by deficiencies in various pre-synaptic proteins, and that Sept4 is itself a component of the machinery for dopamine turnover. In particular, levels of syntaxin 1A, SNAP25, Munc18, and α-synuclein (all components of the exocytic machinery) were all reduced, as were levels of tyrosine hydroxylase (the enzyme responsible for dopamine synthesis) and the dopamine transporter. Moreover, the expression of Sept4 protein seems to prevent the self-aggregation of α-synuclein that underlies Lewy body formation (Ihara et al., 2007). Thus, Sept4 appears to have a dual role in DA neurons: as part of a scaffold for localization of key pre-synaptic components, and in preventing α-synuclein accumulation and its associated toxicity. Interference with these functions would disrupt DA neuron function by interfering with synaptic transmission and causing accumulation of neurotoxic forms of α-synuclein. The most common neurodegenerative disease, Alzheimer’s disease (AD), is a dementia with a characteristic neuropathology of extracellular plaques enriched in aß and intracellular neurofibrillary tangles, of which the major component is hyperphosphorylated forms of the microtubule-associated protein tau (Goedert and Spillantini, 2006). The discovery that three human septin proteins, SEPT1, SEPT2 and SEPT4 are found in the plaques and tangles and associated with AD, including in tangle precursors (Kinoshita et al., 1998), provided the earliest connection between septins and neurodegeneration. In that study, SEPT7 was not

306

CH14 SEPTINS AND HUMAN DISEASE

associated with AD tangles (Kinoshita et al., 1998), consistent with the subsequent finding that SEPT7 expression is significantly reduced in presenilin-1 null mice, a model for AD (Liauw et al., 2002). Unfolding intermediates of SEPT4 are capable of forming amyloid-like filaments in vitro (Garcia et al., 2007), suggesting that SEPT4 could directly contribute to the aberrant structures observed in Alzheimer’s brains. Additionally, SEPT3 polymorphisms in exon 11 are correlated with AD (Takehashi et al., 2004). In the light of the association of changes in septin expression with AD, the implication of septins as differentially expressed in Down’s syndrome (Cheon et al., 2001; Engidawork et al., 2003; Pollak, Fountoulakis and Lubec, 2003), the most common birth defect associated with mental retardation (Antonarakis, 1998), is interesting to consider. Down’s syndrome pathology includes both developmental failures and neurodegeneration occurring at later ages (Mark and Griffin, 2004). Although the underlying cause of Down’s syndrome is trisomy of chromosome 21 (Antonarakis, 1998), the molecular mechanisms underlying the alterations of the brains of Down’s syndrome patients are not well understood. Down’s syndrome-associated neurodegeneration is characterized by tangles and plaques similar to those seen in Alzheimer’s disease (Wisniewski, Wisniewski and Wen, 1985). While no septin genes are located on chromosome 21 (Hall et al., 2005; Hall and Russell, 2004; Peterson et al., 2007), decreased levels of three human septins (SEPT3, SEPT6 and SEPT7) and increased levels of a fourth (SEPT9) have been observed in fetal Down’s syndrome cerebral cortices (Cheon et al., 2001; Engidawork et al., 2003; Pollak, Fountoulakis and Lubec, 2003), and it is possible that alterations in septin levels may contribute to the developmental and/or neurodegenerative phenotypes. Could it be that genes that regulate septin gene expression are to be found on chromosome 21? Finally, altered expression of septins has been reported by a variety of methods (and in particular proteomics) in a range of diseases including schizophrenia (Barr et al., 2004; Pennington et al., 2007, Gottfried et al., 2007), epilepsy (Yang et al., 2006), and cerebral tumours (Kim et al., 2004; Sakai et al., 2002). At present the significance of such observations is unclear.

SEPTINS AND INFECTIOUS DISEASE Septins have been implicated in pathogenesis by diverse infectious agents including bacteria, fungi and viruses. Several lines of evidence suggest that understanding septin biology is relevant to the pathogenesis and treatment of infectious disease with both the involvement of host and pathogen septins. The process by which intracellular pathogens enter and move around in cells depends on interactions with the cytoskeleton (Gruenheid and Finlay, 2003; Cossart and Sansonetti, 2004; Smith and Helenius, 2004; Sibley, 2004). Cossart and co workers have shown that Sept9 is a component of the machinery involved in the movement of Listeria monocytogenes around their host cells (Cossart, Pizarro-Cerada

OTHER DISEASES

307

and Lecruit, (2003); Pizarro-Cerda et al., 2002) and it may be that other pathogens including viruses (see below and Smith and Helenius, 2004) and bacteria such as Shigella and Rickettsia may utilise septin dependent processes (van Nhieu et al., 1999; Frischknecht and Way, 2001). Indeed, the critical and dramatic cytoskeletal re-organizations that occur in association with phagocytosis and with the specific immune response may involve septin function (see Chapter 8; May and Machesky, 2002; Miletic et al., 2003). It is also worth noting the association of septins with syntaxins, proteins involved with phagocytosis (Collins et al., 2002). Furthermore, microbes may have evolved mechanisms to modify or even subvert the functions of endogenous septins as they have with other cytoskeletal components (Gruenheid and Finlay, 2003). Fungal growth and virulence are compromised in strains of the pathogenic yeast Candida albicans defective in septin function (Warenda et al., 2003) and thus different septin alleles might be associated with different levels of virulence. More recently C. albicans Cdc10 was shown to be a virulence factor in a mouse infection model, affecting multiple aspects of infection, including adhesion to mouse cells, the ability to colonise organs and the ability to evade the immune response (Gonz´alez-Novo et al., 2006). It is conceivable that drugs that interfere with yeast septin function could be developed as potential novel antifungal agents. Viruses utilise diverse cellular systems for their own ends and so it would not be surprising if septins were targets of viral proteins. Recent evidence suggests this is the case. Human Herpes Virus 8 is associated with Kaposi’s sarcoma and other human tumours. A key transforming protein in this virus is Kaposin A, which has recently been shown to interact with SEPT4 (Lin et al., 2007). In Hepatitis C virus replication the virally encoded RNA-binding protein hnRNP A1 has been reported to interact with SEPT6. That this has biological relevance is suggested by the inhibition of viral replication by either siRNA mediated SEPT6 knock down or the use of an N terminally truncated SEPT6 transgene (Kim et al., 2007). We await with interest further reports of this kind and one would predict that septins may provide insights into diverse aspects of viral infection and vice versa!

OTHER DISEASES Elsewhere in this volume there are discussions of septins and platelet function (Chapter 12) and other aspects of septins and pathobiology. A number of reports have linked altered expression of septin genes (or proteins) to a range of disease states. For example, using proteomic approaches perturbed expression of SEPT2 has been seen in rodent models of pulmonary hypertension (Laudi et al., 2007) and anti SEPT2 antibodies have been reported in psychiatric disorders (Margutti et al., 2005). The significance of such observations is totally unclear at present. More convincingly, impaired fertility was reported in the male homozygous null Sept4 knock out mice with a phenotype reminiscent of some male human infertility syndromes (Kissel et al., 2005, Ihara et al., 2005). Sept4 localizes along with Septs

308

CH14 SEPTINS AND HUMAN DISEASE

1, 6 and 7 to the sperm annulus (Ihara et al., 2005), as does SEPT 12 (Steels et al., 2007). Furthermore, sperm from a subset of patients with asthenospermia lacked septin rings/annuli. SEPT14 is primarily expressed in testis (Peterson et al., 2007). Without question we will see more reports of altered septin expression in human (and animal disease) as tools for studying septins become more widely available. The crucial question will be the pathogenetic significance of such observations.

THE FUTURE: THE NEED FOR NEW MODELS OF SEPTINS AND DISEASE As described elsewhere in this volume, our burgeoning knowledge of septin proteins in diverse organisms paints a complex picture: we clearly only have parts of this picture, and much of it is poorly illuminated. We remain remarkably ignorant of much of the detail concerning the diverse roles of septins in many aspects of cellular physiology, and additional roles for septins are likely to be uncovered. Complete illumination of this picture is currently hindered by the enormous complexity of human septin genes and their expression and our fragmentary knowledge of this (see Chapter 7). Why are there so many septin genes and (in the case of mammals and possibly other vertebrates) so many transcripts and isoforms? Simply taking the examples of SEPT 9 and SEPT6 (see Chapter 6), we see multiple transcripts encoding many protein isoforms with different patterns of expression and the potential for diverse interactions. Moreover the subtlety to regulation is manifested by the truly remarkable existence of multiple transcripts that can encode a single polypeptide: a phenomenon seen in not only SEPT 9 and SEPT6 but in SEPT8 and perhaps in other human septins. Traditional pathological analysis of proteins encoded by a given gene usually does not take into account such enormous structural and functional diversity. Consequently, a complete understanding of septins in normal cellular physiology, as well as pathological conditions, will require new analytical approaches and tools as well as the ability to reconcile biochemistry with cellular and organismal pathobiology. At present we are merely at the beginning of the cataloguing of the potential alterations in septin expression and function in normal and diseased states: a low resolution first pass at what will require a much more detailed and sophisticated study. Furthermore the crucial issue of ‘cause or effect’ is almost totally unresolved. Certainly mouse models (see following chapter) have been important and will continue to provide new insights and clues. However to date the ‘knock -out’ technologies employed provide information only about null genotypes. Such amorphic genotypes (Muller, 1932) do not reflect the phenomena so far seen in human disease states and consequently cannot be viewed as adequately testing the idea that septins are (or are not) involved in disease pathogenesis. A range of hypo- and hyper-morphic alleles will need to be tested in mouse systems to replicate the subtle alterations in expression seen in disease states. Furthermore we

REFERENCES

309

remain only partially informed about the parallels and differences between septins in man and mouse. It follows that we should remain cautious about inferring possible roles in man from as yet incompletely characterised mouse septin biology. Indeed it may be that we need to use a diversity of metazoan systems (mouse, chicken, zebrafish, fly, worm etc.) to fully model the potentiality of deregulated septins expression and function. Given these caveats, what issues need to be addressed? With regard to human physiology and pathology, a detailed catalogue of the expression patterns of transcripts and isoforms is needed for each septin. This has to be coupled with an understanding of the regulation of these patterns of gene expression. Furthermore, modifications of the isoforms and their interactions with other septins and with other proteins (and perhaps also with lipids) need to be defined in detail. It is then only by integrating the pathophysiology with the biochemistry and cell biology that a detailed understanding of this family of proteins be obtained. As we have argued before (Hall and Russell, 2004), in many respects septins represent a model system for developing such approaches and a central challenge of post-genomic biology is to integrate such information into the broader context.

REFERENCES Amir, S., Mabjeesh, N.J. (2007) SEPT9 V1 protein expression is associated with human cancer cell resistance to microtubule disrupting agents. Cancer Biology and Therapy, 6 (12), 1926–31. Amir, S., Wang, R., Matzkin, H. et al. (2006) MSF-A interacts with hypoxia-inducible factor-1 alpha and augments hypoxia-inducible factor transcriptional activation to affect tumorigenicity and angiogenesis. Cancer Research, 66 (2), 856–66. Antonarakis, S. (1998) 10 years of genomics, chromosome 21 and Down syndrome. Genomics, 51, 1–16. Bae, Y.J., Kang, S.J. and Park, K.S. (2007) Drosophila melanogaster Parkin ubiquitinates peanut and septin1 as an E3 ubiquitin-protein ligase. Insect Biochemistry and Molecular Biology, 37 (5), 430–39. Barr, A.M., Young, C.E., Sawada, K. et al. (2004) Abnormalities of presynaptic protein CDCrel-1 in striatum of rats reared in social isolation: relevance to neural connectivity in schizophrenia. The European Journal of Neuroscience, 20 (1), 303–7. Barral, Y. and Mansuy, I.M. (2007) Septins: cellular and functional barriers of neuronal activity. Current Biology, 17 (22), R961–63. Bartek, J., Lukas, J. and Bartkova, J. (2007) DNA damage response as an anti-cancer barrier: damage threshold and the concept of ‘conditional haploinsufficiency’. Cell Cycle (Georgetown, Tex), 6 (19), 2344–47. van Binsbergen, E., de Weerdt, O. and Buijs, A. (2007) A new subtype of MLLSEPT2 fusion transcript in therapy-related acute myeloid leukemia with t(2;11) (q37;q23): a case report and literature review. Cancer Genetics and Cytogenetics, 176 (1), 72–75.

310

CH14 SEPTINS AND HUMAN DISEASE

Blaser, S., Horner, J., Wurmell, P. et al. (2004) The novel platelet septin SEPT8 is an interaction partner of SEPT4. Thrombosis and Haemostasis, 91, 959–66. Borkhardt, A., Teigler-Schlegel, A., Fuchs, U. et al. (2001) An ins(X;11)(q24;q23) fuses the MLL and the Septin 6/KIAA0128 gene in an infant with AML-M2. Genes, Chromosomes and Cancer, 32 (1), 82–88. Burrows, J.F., Chanduloy, S., McIlhatton, M.A. et al. (2003) Altered expression of the septin gene, SEPT9, in ovarian neoplasia. Journal de Psychologie Normale et Pathologique, 201, 581–88. Caltagarone, J., Rhodes, J., Honer, W.G. and Bowser, R. (1998) Localization of a novel septin protein, hCDCrel-1, in neurons of human brain. Neuroreport, 9, 2907–12. Capurso, G., Crnogorac-Jurcevic, T., Milione, M. et al. (2005) Peanut-like 1 (Septin 5) gene expression in normal and neoplastic human endocrine pancreas. Neuroendocrinology, 81 (5), 311–21. Cerveira, N., Correia, C., Bizarro, S. et al. (2006) SEPT2 is a new fusion partner of MLL in acute myeloid leukemia with t(2;11)(q37;q23). Oncogene, 25 (45), 6147–52. Chacko, A., Hyland, P.L., McDade, S.S. et al. (2005) SEPT9 v4 expression induces morphological change, increased motility and disturbed polarity. The Journal of Pathology, 206 (4), 458–65. Chance, P.F. (2006) Inherited focal, episodic neuropathies: hereditary neuropathy with liability to pressure palsies and hereditary neuralgic amyotrophy. Neuromolecular Medicine, 8 (1-2), 159–74. Cheon, M.S., Fountoulakis, M., Dierssen, M. et al. (2001) Expression profiles of proteins in fetal brain with Down syndrome. Journal of Neural Transmission. Supplementum, 61, 311–19. Choi, P., Snyder, H., Petrucelli, L. et al. (2003) SEPT5 v2 is a parkin-binding protein. Brain research. Molecular Brain Research, 117, 179–89. Collins, R.F., Schreiber, A.D., Grinstein, S. and Trimble, W.S. (2002) Syntaxins 13 and 7 function at distinct steps during phagocytosis. Journal of Immunology, 169, 3250–56. Cook, J.D., Davis, B.J., Cai, S.L. et al. (2005) Interaction between genetic susceptibility and early-life environmental exposure determines tumor-suppressor-gene penetrance. Proceedings of the National Academy of Sciences of the United States of America, 102, 8644–49. Cossart, P., Pizarro-Cerda, J. and Lecruit, M. (2003) Invasion of mammalian cells by Listeria monocytogenes: functional mimicry to subvert cellular functions. Trends in Biological Sciences, 13, 23–31. Cossart, P. and Sansonetti, P.J. (2004) Bacterial invasion: the paradigms of enteroinvasive pathogens. Science, 304, 242–48. Craven, R.A., Hanrahan, S., Totty, N. et al. (2006a) Proteomic identification of a role for the von Hippel Lindau tumour suppressor in changes in the expression of mitochondrial proteins and septin 2 in renal cell carcinoma. Proteomics, 6 (13), 3880–93. Craven, R.A., Stanley, A.J., Hanrahan, S. et al. (2006b) Proteomic analysis of primary cell lines identifies protein changes present in renal cell carcinoma. Proteomics, 6 (9), 2853–64. Dong, Z., Ferger, B., Paterna, J.C. et al. (2003) Dopamine-dependent neurodegeneration in rats induced by viral vector-mediated overexpression of the Parkin-targetprotein, CDCrel-1. Proceedings of the National Academy of Sciences of the United States of America, 100, 12438–43.

REFERENCES

311

Elhasid, R., Sahar, D., Merling, A. et al. (2004) Mitochondrial pro-apoptotic ARTS protein is lost in the majority of acute lymphoblastic leukemia patients. Oncogene, 23, 5468–75. Engidawork, E., Gulesserian, T., Fountoulakis, M. and Lubec, G. (2003) Aberrant protein expression in cerebral cortex of fetus with Down syndrome. Neuroscience, 122, 145–54. Finger, F.P., Kopish, K.R. and White, J.G. (2003) A role for septins in cellular and axonal migration in C. elegans. Developmental Biology, 261, 220–34. Frischknecht, F. and Way, M. (2001) Surfing pathogens and the lessons learned for actin polymerisation. Trends in Cell Biology, 11, 30–38. Fu, J.F., Liang, D.C., Yang, C.P. et al. (2003) Molecular analysis of t(X;11)(q24;q23) in an infant with AML-M4. Genes, Chromosomes and Cancer, 38, 253–59. Fujishima, K., Kiyonari, H., Kurisu, J. et al. (2007) Targeted disruption of Sept3, a heteromeric assembly partner of Sept5 and Sept7 in axons, has no effect on developing CNS neurons. Journal of Neurochemistry, 102 (1), 77–92. Gad´ea, G., Lapasset, L., Gauthier-Rouvi`ere, C. and Roux, P. (2002) Regulation of Cdc42-mediated morphological effects: a novel function for p53. The EMBO Journal , 21 (10), 2373–82. Gandhi, S. and Wood, N.W. (2005) Molecular pathogenesis of Parkinson’s disease. Human Molecular Genetics, 14, 2749–55. Garcia, W., de Araajo, A.P., Neto Mde, O. et al. (2006) Dissection of a human septin: definition and characterization of distinct domains within human SEPT4. Biochemistry, 45 (46), 13918–31. Garcia, W., Araujo, A.P., Lara, F. et al. (2007) An intermediate structure in the thermal unfolding of the GTPase domain of human Septin 4 (SEPT4/Bradeion-beta) forms amyloid-like filaments in vitro. Biochemistry, 46, 11101–9. Goedert, M. and Spillantini, M.G. (2006) A century of Alzheimer’s disease. Science, 314, 777–81. Gonz´alez-Novo, A., Labrador, L., Jimanez, A. et al. (2006) Role of the septin Cdc10 in the virulence of Candida albicans. Microbiology and Immunology, 50 (7), 499–511. Gonzalez, M.E., Peterson, E.A., Privette, L.M. et al. (2007) High SEPT9 v1 expression in human breast cancer cells is associated with oncogenic phenotypes. Cancer Research, 67 (18), 8554–64. Gottfried, Y., Rotem, A., Klein, E. and Larisch, S. (2007) The pro-apoptotic ARTS/ Sept4 protein is significantly reduced in post-mortem brains from schizophrenic patients. Schizophrenia Research, 96, 257–66. Gottfried, Y., Rotem, A., Lotan, R. et al. (2004) The mitochondrial ARTS protein promotes apoptosis through targeting XIAP. The EMBO Journal , 23, 1627–35. Gruenheid, S. and Finlay, B.B. (2003) Microbial pathogenesis and cytoskeletal function. Nature, 422, 775–81. Hall, P.A., Jung, K., Hillan, K.J. and Russell, S.E. (2005) Expression profiling the septin gene family. The Journal of Pathology, 206, 269–78. Hall, P.A. and Russell, S.E.H. (2004) The pathobiology of the septin gene family. The Journal of Pathology, 204, 489–505. Ihara, M., Kinoshita, A., Yamada, S. et al. (2005) Cortical organization by the septin cytoskeleton is essential for structural and mechanical integrity of mammalian spermatozoa. Developmental Cell , 8, 343–52.

312

CH14 SEPTINS AND HUMAN DISEASE

Ihara, M., Tomimoto, H., Kitayama, H. et al. (2003) Association of the cytoskeletal GTPbinding protein Sept4/H5 with cytoplasmic inclusions found in Parkinson’s disease and other synucleinopathies. The Journal of Biological Chemistry, 278, 24095–102. Ihara, M., Yamasaki, N., Hagiwara, A. et al. (2007) Sept4, a component of presynaptic Scaffold and Lewy bodies, is required for the suppression of alpha-synuclein neurotoxicity. Neuron, 53 (4), 519–33. Irazoqui, J.E. and Lew, D.J. (2004) Polarity establishment in yeast. Journal of cell science, 117 (Pt 11), 2169–71. Jaeger, J., Koczan, D., Thiesen, H.J. et al. (2007) Gene expression signatures for tumor progression, tumor subtype, and tumor thickness in laser-microdissected melanoma tissues. Clinical Cancer Research, 13 (3), 806–15. Kadkol, S.S., Bruno, A., Oh, S. et al. (2006) MLL-SEPT6 fusion transcript with a novel sequence in an infant with acute myeloid leukemia. Cancer Genetics and Cytogenetics, 168 (2), 162–67. Kalikin, L.M., Sims, H.L. and Petty, E.M. (2000) Genomic and expression analyses of alternatively spliced transcripts of the MLL septin-like fusion gene (MSF) that map to a 17q25 region of loss in breast and ovarian tumors. Genomics, 63, 165–72. Kang, H.J., Koh, K.H., Yang, E. et al. (2006) Differentially expressed proteins in gastrointestinal stromal tumors with KIT and PDGFRA mutations. Proteomics, 6 (4), 1151–57. Kato, Y., Uzawa, K., Yamamoto, N. et al. (2007) Overexpression of Septin1: possible contribution to the development of oral cancer. International Journal of Clinical Oncology, 31 (5), 1021–28. Khalil, A.A. and James, P. (2007) Biomarker discovery: a proteomic approach for brain cancer profiling. Cancer Science, 98 (2), 201–13. Kim, C.S., Seol, S.K., Song, O.K. et al. (2007) An RNA-binding protein, hnRNP A1, and a scaffold protein, septin 6, facilitate hepatitis C virus replication. Journal of Virology, 81 (8), 3852–65. Kim, D.S., Hubbard, S.L., Peraud, A. et al. (2004) Analysis of mammalian septin expression in human malignant brain tumours. Neoplasia, 6, 168–78. Kim, H.J., Ki, C.S., Park, Q. et al. (2003) MLL/SEPTIN6 chimeric transcript from inv ins(X;11)(q24;q23q13) in acute monocytic leukemia: report of a case and review of the literature. Genes, Chromosomes and Cancer, 38 (1), 8–12. Kinoshita, A., Kinoshita, M., Akiyama, H. et al. (1998) Identification of septins in neurofibrillary tangles in Alzheimer’s disease. American Journal of Clinical Pathology, 153, 1551–60. Kinoshita, A., Noda, M. and Kinoshita, M. (2000) Differential localization of septins in the mouse brain. The Journal of Comparative Neurology, 428, 223–39. Kinoshita, M., Kumar, S., Mizoguchi, A. et al. (1997) Nedd5, a mammalian septin, is a novel cytoskeletal component interacting with actin-based structures. Genes and Development, 11, 1535–47. Kinoshita, M. and Takeda, S. (2007) Connecting the dots between septins and the DNA damage checkpoint. Cell , 130 (5), 777–79. Kissel, H., Georgescu, M.M., Larisch, S. et al. (2005) The Sept4 septin locus is required for sperm terminal differentiation in mice. Developmental Cell , 8 (3), 353–64. Kojima, K., Sakai, I., Hasegawa, A. et al. (2004) FLJ10849, a septin family gene, fuses MLL in a novel leukemia cell line CNLBC1 derived from chronic neutrophilic leukemia in transformation with t(4;11)(q21;q23). Leukemia, 18, 998–1005.

REFERENCES

313

Koshelev, Y.A., Kiselev, S.L. and Georgiev, G.P. (2003) Interaction of the S100A4 (Mts1) protein with septins Sept2, Sept6, and Sept7 in vitro. Doklady. Biochemistry and Biophysics, 391, 195–97. Kostic, B. and Shaw, P.H. (2000) Isolation and characterisation of sixteen novel p53 response genes. Oncogene, 19, 3978–87. Kremer, B.E., Adang, L.A. and Macara, I.G. (2007) Septins regulate actin organization and cell-cycle arrest through nuclear accumulation of NCK mediated by SOCS7. Cell , 130 (5), 837–50. Kremer, B.E., Haystead, T. and Macara, I.G. (2005) Mammalian septins regulate microtubule stability through interaction with the microtubule-binding protein MAP4. Molecular Biology of the Cell , 16 (10), 4648–59. Kreuziger, L.M., Porcher, J.C., Ketterling, R.P. and Steensma, D.P. (2007) An MLLSEPT9 fusion and t(11;17)(q23;q25) associated with de novo myelodysplastic syndrome. Leukemia, 31 (8), 1145–48. Kuhlenbaumer, G., Hannibal, M.C., Nelis, E. et al. (2005) Mutations in SEPT9 cause hereditary neuralgic amyotrophy. Nature Genetics, 37 (10), 1044–46. Larisch, S., Yi, Y., Lotan, R. et al. (2000) A novel mitochondrial septin-like protein, ARTS, mediates apoptosis dependent on its P-loop motif. Nature Cell Biology, 2, 915–21. Laudi, S., Steudel, W., Jonscher, K. et al. (2007) Comparison of lung proteome profiles in two rodent models of pulmonary arterial hypertension. Proteomics, 7 (14), 2469–78. Liauw, J., Nguyen, V., Huang, J. et al. (2002) Differential display analysis of presenilin 1-deficient mouse brains. Molecular Brain Research, 109, 56–62. Lin, C.W., Tu, P.F., Hsiao, N.W. et al. (2007) Identification of a novel septin 4 protein binding to human herpesvirus 8 kaposin A protein using a phage display cDNA library. Journal of Virological Methods, 143 (1), 65–72. Lindsey, R. and Momany, M. (2006) Septin localization across kingdoms: three themes with variations. Current Opinion in Microbiology, 9, 559–65. Margutti, P., Sorice, M., Conti, F. et al. (2005) Screening of an endothelial cDNA library identifies the C-terminal region of Nedd5 as a novel autoantigen in systemic lupus erythematosus with psychiatric manifestations. Arthritis Research and Therapy, 7 (4), R896–903. Mark, R.E. and Griffin, W.S. (2004) Trisomy 21 and the brain. Journal of Neuropathology and Experimental Neurology, 63, 679–85. May, R.C. and Machesky, L.M. (2002) Phagocytosis and the actin cytoskeleton. Journal of Cell Science, 114, 1061–77. McDade, S.S., Hall, P.A. and Russell, S.E. (2007) Translational control of SEPT9 isoforms is perturbed in disease. Human Molecular Genetics, 16 (7), 742–52. McIlhatton, M.A., Burrows, J.F., Donaghy, P.G. et al. (2001) Genomic organization, complex splicing pattern and expression of a human septin gene on chromosome 17q25.3. Oncogene, 20, 5930–39. Megonigal, M.D., Rappaport, E.F., Jones, D.H. et al. (1998) t(11;22)(q23;q11.2) In acute myeloid leukemia of infant twins fuses MLL with hCDCrel, a cell division cycle gene in the genomic region of deletion in DiGeorge and velocardiofacial syndromes. Proceedings of the National Academy of Sciences of the United States of America, 95, 6413–18.

314

CH14 SEPTINS AND HUMAN DISEASE

Methner, A., Leypoldt, F., Joost, P. and Lewerenz, J. (2001) Human septin 3 on chromosome 22q13.2 is upregulated by neuronal differentiation. Biochemical and Biophysical Research Communications, 283, 48–56. Meuleman, J., Kuhlenbaumer, G., Audenaert, D. et al. (2001) Mutation analysis of 4 candidate genes for hereditary neuralgic amyotrophy (HNA). Human Genetics, 108, 390–93. Miletic, A.V., Swat, M., Fujikawa, K. and Swat, W. (2003) Cytoskeletal remodelling in lymphocyte activation. Current Opinion in Immunology, 15, 261–68. Montagna, C., Lyu, M.S., sHunter, K. et al. (2003) The Septin 9 (MSF) gene is amplified and overexpressed in mouse mammary gland adenocarcinomas and human breast cancer cell lines. Cancer Research, 63, 2179–87. Moran, L.B., Croisier, E., Duke, D.C. et al. (2007) Analysis of alpha-synuclein, dopamine and parkin pathways in neuro-pathologically confirmed parkinsonian nigra. Acta Neuropathologica, 113 (3), 253–63. Muller, H.J. (1932) Further Studies on the Nature and Causes of Gene Mutations. Proceedings of the 6th International Congress of Genetics, 213–55. Mu˜noz-Soriano, V. and Paricio, N. (2007) Overexpression of Septin 4, the Drosophila homologue of human CDCrel-1, is toxic for dopaminergic neurons. The European Journal of Neuroscience, 26 (11), 3150–58. van Nhieu, T., Caron, E., Hall, A. and Sansonetti, P.J. (1999) IpaC induces actin polymerisation and filopodia formation during Shigella entry into epithelial cells. The EMBO Journal , 18, 3249–62. Ono, R., Ihara, M., Nakajima, H. et al. (2005) Disruption of SEPT6, a fusion partner gene of MLL, does not affect ontogeny, leukemogenesis induced by MLL-SEPT6, or phenotype induced by the loss of SEPT4. Molecular and Cell Biology, 25 (24), 10965–78. Ono, R., Taki, T., Taketani, T. et al. (2002) SEPT6, a human homologue to mouse Septin6, is fused to MLL in infant acute myeloid leukemia with complex chromosomal abnormalities involving 11q23 and Xq24. Cancer Research, 62, 333–37. Orimo, A., Gupta, P.B., Sgroi, D.C. et al. (2005) Stromal fibroblasts present in invasive human breast carcinomas promote tumor growth and angiogenesis through elevated SDF-1/CXCL12 secretion. Cell , 121, 335–48. Osaka, M., Rowley, J.D. and Zeleznik-Le, N.J. (1999) MSF (MLL septin-like fusion), a fusion partner gene of MLL, in a therapy-related acute myeloid leukemia with a t(11;17)(q23;q25). Proceedings of the National Academy of Sciences of the United States of America, 96, 6428–33. Pawlak, G. and Helfman, D.M. (2001) Cytoskeletal changes in cell transformation and tumorigenesis. Current Opinion in Genetics and Development , 11 (1), 41–47. Pennington, K., Beasley, C.L., Dicker, P. et al. Prominent synaptic and metabolic abnormalities revealed by proteomic analysis of the dorsolateral prefrontal cortex in schizophrenia and bipolar disorder. (2007) Molecular Psychiatry. Periquet, M., Corti, O., Jacquier, S., Brice, A. (2005) Proteomic analysis of parkin knockout mice: alterations in energy metabolism, protein handling and synaptic function. The Journal of Neurochemistry, 95 (5), 1259–76. Peterson, E.A., Kalikin, L.M., Steels, J.D. et al. (2007) Characterization of a SEPT9 interacting protein, SEPT14, a novel testis-specific septin. Mammalian Genome, 18 (11), 796–807.

REFERENCES

315

Pizarro-Cerda, J., Jonquieres, R., Gouin, E. et al. (2002) Distinct protein patterns associated with Listeria monocytogenes InlA- or InlB-phagosomes. Cellular Microbiology, 4, 101–15. Pollak, D., Fountoulakis, M. and Lubec, G. (2003) The molecular synaptosomal deficit in Down syndrome starts early in life. Neuroembryology, 2, 143–46. Qi, M., Yu, W., Liu, S. et al. (2005) Septin1, a new interaction partner for human serine/threonine kinase aurora-B. Biochemical and Biophysical Research Communications, 336 (3), 994–1000. Robertson, C., Church, S.W., Nagar, H.A. et al. (2004) Properties of SEPT9 isoforms and the requirement for GTP binding. Journal de Psychologie Normale et Pathologique, 203, 519–27. Russell, S.E., Hall, P.A. (2005) Do septins have a role in cancer? British Journal of Cancer, 93 (5), 499–503. Russell, S.E., McIlhatton, M.A., Burrows, J.F. et al. (2000) Isolation and mapping of a human septin gene to a region on chromosome 17q, commonly deleted in sporadic epithelial ovarian tumors. Cancer Research, 60, 4729–34. Sager, R. (1997) Expression genetics in cancer: shifting the focus from DNA to RNA. Proceedings of the National Academy of Sciences of the United States of America, 94 (3), 952–55. Sakai, K., Kurimoto, M., Tsugu, A. et al. (2002) Expression of Nedd5, a mammalian septin, in human brain tumors. Journal of Neurooncology, 57, 169–77. Scott, M., Hyland, P., McGregor, G. et al. (2005) Multimodality expression profiling shows SEPT9 to be over-expressed in a wide range of human tumours. Oncogene, 24 (29), 4688–700. Scott, M., McCluggage, W.G., Hillan, K.J. et al. (2006) SEPT9 transcript expression profiling in epithelial ovarian neoplasia defines borderline tumours as a distinct category. International Journal of Cancer, 118 (5), 1325–29. Shih, L.Y., Liang, D.C., Fu, J.F. et al. (2006) Characterization of fusion partner genes in 114 patients with de novo acute myeloid leukemia and MLL rearrangement. Leukemia, 20, 218–23. Sibley, L.D. (2004) Intracellular parasite invasion strategies. Science, 304, 248–53. Sirajuddin, M., Farkasovsky, M., Hauer, F., K¨uhlmann, D., Weyand, M., Stark, H. and Wittinghofer, A. (2007) Structural insight into filament formation by mammalian septins. Nature, 449 (7160), 311–15. Sirajuddin, M., Farkasovsky, M., Hauer, F. et al. (2007) Structural insight into filament formation by mammalian septins. Nature, 449, 311–15. Slany, R.K. (2005) When epigenetics kills: MLL fusion proteins in leukemia. Hematological Oncology, 23 (1), 1–9. Slater, D.J., Hilgenfeld, E., Rappaport, E.F. et al. (2002) MLL-SEPTIN6 fusion recurs in novel translocation of chromosomes 3, X, and 11 in infant acute myelomonocytic leukaemia and in t(X;11) in infant acute myeloid leukaemia, and MLL genomic breakpoint in complex MLL-SEPTIN6 rearrangement is a DNA topoisome. Oncogene, 21 (30), 4706–14. Smith, A.E. and Helenius, A. (2004) How viruses enter animal cells. Science, 304, 237–42. So, C.W., Lin, M., Ayton, P.M. et al. (2003) Dimerization contributes to oncogenic activation of MLL chimeras in acute leukemias. Cancer Cell , 4, 99–110.

316

CH14 SEPTINS AND HUMAN DISEASE

Son, J.H., Kawamata, H., Yoo, M.S. et al. (2005) Neurotoxicity and behavioral deficits associated with Septin 5 accumulation in dopaminergic neurons. Journal of Neurochemistry, 94 (4), 1040–53. Sorensen, A.B., Lund, A.H., Ethelberg, S. et al. (2000) Sint1, a common integration site in SL3-3-induced T-cell lymphomas, harbors a putative proto-oncogene with homology to the septin gene family. Journal of Virology, 74, 2161–68. Spiliotis, E.T., Kinoshita, M. and Nelson, W.J. (2005) A mitotic septin scaffold required for mammalian chromosome congression and segregation. Science, 307, 1781–85. Steels, J.D., Estey, M.P., Froese, C.D. et al. (2007) SEPT12 is a component of the mammalian sperm tail annulus. Cell Motility and the Cytoskeleton, 64 (10), 794–807. Strehl, S., K¨onig, M., Meyer, C. et al. (2006) Molecular dissection of t(11;17) in acute myeloid leukemia reveals a variety of gene fusions with heterogeneous fusion transcripts and multiple splice variants. Genes, Chromosomes and Cancer, 45 (11), 1041–49. Sudo, K., Ito, H., Iwamoto, I. et al. (2007) SEPT9 sequence alternations causing hereditary neuralgic amyotrophy are associated with altered interactions with SEPT4/ SEPT11 and resistance to Rho/Rhotekin-signaling. Human Mutation, 28 (10), 1005–13. Tada, T., Simonetta, A., Batterton, M. et al. (2007) Role of septin cytoskeleton in spine morphogenesis and dendrite development in neurons. Current Biology, 17 (20), 1752–58. Takehashi, M., Alioto, T., Stedeford, T. et al. (2004) Septin 3 gene polymorphism in Alzheimer’s disease. Gene Expression, 11 (5-6), 263–70. Takehashi, M., Tanaka, S., Stedeford, T. et al. (2004) Expression of septin 3 isoforms in human brain. Gene Expression, 11 (5-6), 271–78. Taki, T., Ohnishi, H., Shinohara, K. et al. (1999) AF17q25, a putative septin family gene, fuses the MLL gene in acute myeloid leukemia with t(11;17) (q23;q25). Cancer Research, 59, 4261–65. Tanaka, M., Kijima, H., Itoh, J. et al. (2002) Impaired expression of a human septin family gene Bradeion inhibits the growth and tumorigenesis of colorectal cancer in vitro and in vivo. Cancer Gene Therapy, 9, 483–88. Tanaka, M., Tanaka, T., Kijima, H. et al. (2001) Characterization of tissue- and cell-type-specific expression of a novel human septin family gene, Bradeion. Biochemical and Biophysical Research Communications, 286, 547–53. Tanaka, M., Tanaka, T., Matsuzaki, S. et al. (2003) Rapid and quantitative detection of human septin family Bradeion as a practical diagnostic method of colorectal and urologic cancers. Medical Science Monitor, 9, 61–68. Tatsumi, K., Taki, T., Taniwaki, M. et al. (2002) The CDCrel-1 gene fused to MLL in de novo acute myeloid leukemia with t(11;22)(q23;q11.2) and its frequent expression in myeloid leukemia cell lines. Genes, Chromosomes and Cancer, 30, 230–35. Vega, I.E. and Hsu, S.C. (2003) The septin protein Nedd5 associates with both the exocyst complex and microtubules and disruption of its GTPase activity promotes aberrant neurite sprouting in PC12 cells. Neuroreport, 14, 31–37. Warenda, A.J., Kauffman, S., Sherrill, T.P. et al. (2003) Candida albicans septin mutants are defective for invasive growth and virulence. Infection and Immunity, 717, 4045–51.

REFERENCES

317

Wisniewski, K., Wisniewski, H. and Wen, G. (1985) Occurrence of neuropathological changes and dementia of Alzheimer’s disease in Down syndrome. Annals of Neurology, 17, 278–82. Wood, L.D., Parsons, D.W., Jones, S. et al. (2007) The genomic landscapes of human breast and colorectal cancers. Science, 318 (5853), 1108–13. Wood-Kaczmar, A., Gandhi, S. and Wood, N.W. (2006) Understanding the molecular causes of Parkinson’s disease. Trends in Molecular Medicine, 12, 521–28. Xiao, G.H., Shoarinejad, F., Jin, F. et al. (1997) The tuberous sclerosis 2 gene product, tuberin, functions as a Rab5 GTPase activating protein (GAP) in modulating endocytosis. The Journal of Biological Chemistry, 272, 6097–100. Xie, Y., Vessey, J.P., Konecna, A. et al. (2007) The GTP-binding protein Septin 7 is critical for dendrite branching and dendritic-spine morphology. Current Biology, 17, 1746–51. Xiong, J.W., Leahy, A. and Stuhlmann, H. (1999) Retroviral promoter-trap insertion into a novel mammalian septin gene expressed during mouse neuronal development. Mechanisms of Ageing and Development, 86, 183–91. Xue, J., Tsang, C.W., Gai, W.P. et al. (2004) Septin 3 (G-septin) is a developmentally regulated phosphoprotein enriched in presynaptic nerve terminals. Journal of Neurochemistry, 91, 579–90. Yamamoto, K., Shibata, F., Yamaguchi, M. and Miura, O. (2002) Fusion of MLL and MSF in adult de novo acute myelomonocytic leukemia (M4) with t(11;17)(q23;q25). International Journal of Hematology, 75, 503–7. Yang, J.W., Juranville, J.F., Hager, H. et al. (2005) Molecular diversity of rat brain proteins as revealed by proteomic analysis. Molecular Diversity, 9, 385–96. Yang, J.W., Czech, T., Felizardo, M. et al. (2006) Aberrant expression of cytoskeleton proteins in hippocampus from patients with mesial temporal lobe epilepsy. Amino Acids, 30, 477–93. Ying-Tao, Z., Yi-Ping, G., Lu-Sheng, S. and Yi-Li, W. (2005) Proteomic analysis of differentially expressed proteins between metastatic and non-metastatic human colorectal carcinoma cell lines. European Journal of Gastroenterology and Hepatology, 17, 725–32. Zhang, Y., Gao, J., Chung, K.K.K. et al. (2000) Parkin functions as an E2-dependent ubiquitin-protein ligase and promotes the degradation of the synaptic vesicleassociated protein CDCrel-1. Proceedings of the National Academy of Sciences of the United States of America, 97, 13354–59.

15 Insight into septin functions from mouse models Makoto Kinoshita Cell Biology and Biochemistry Unit, Kyoto University Graduate School of Medicine and CREST, Japan Science and Technology Agency, Yoshida Konoe, Sakyo, Kyoto 606-8501, Japan

INTRODUCTION The roles of mammalian septins other than mitosis/cytokinesis have been speculated upon because of their structural diversity and paradoxical abundance in post-mitotic cells. Mouse reverse genetics has begun to provide unparalleled insights, not only into this specific speculation but also into general biological themes, such as how post-mitotic cells differentiate. This chapter reviews in vivo research conducted to elucidate the physiological roles of mammalian septins and their pathological connections to neoplastic, neurological or reproductive disorders.1 In this interim review of the past and ongoing projects using septin knockout mice, it would be expedient to subdivide the mouse septin genes into two groups by their spatial expression patterns: ubiquitous (Sept2, 6, 7, 8, 9, 10, 11, 14 ) and tissue specific (Sept1, 3, 4, 5, 12 ). In vitro studies with tissue culture cells have consistently demonstrated that ubiquitous sub-units (represented by Sept2, 7, 9) are required for mitosis/cytokinesis (Kinoshita et al., 1997; Surka, Tsang and Trimble, 2002; Nagata et al., 2003; Spiliotis, Kinoshita and Nelson, 2005). In contrast, tissue-specific sub-units (e.g. Sept3, 4, 5) are expressed predominantly in post-natal mouse brain, pointing to unexplored roles for the septin 1 In this chapter, human and mouse septin genes are referred to as SEPT1–14 and Sept1–14, respectively, and the cognate polypeptides are represented collectively as Sept1–14 (See Appendix B).

The Septins Edited by Peter A. Hall, S.E. Hilary Russell and John R. Pringle © 2008 John Wiley & Sons, Ltd. ISBN: 978-0-470-51969-1

320

CH15 INSIGHT INTO SEPTIN FUNCTIONS FROM MOUSE MODELS

system in non-dividing cells. However, in vitro assays of post-mitotic cells are often technically difficult and/or physiologically irrelevant. Thus a reverse genetic approach is essential, not only to validate the mitotic functions of septins in physiological terms but also to explore undiscovered functions of this protein family in postmitotic, terminally differentiated cells. Indeed, there has been substantial progress in uncovering physiological roles of mammalian septins through work done on mutant mice that lack one or more septin sub-units (Table 15.1).

UBIQUITOUS SEPTINS Sept6 Background SEPT6 is the only X chromosome-linked septin gene, and was found at the breakpoints of t(X;11) chromosomal translocations in infantile acute myeloid leukemia (AML). The resultant in-frame fusions of SEPT6 with MLL (mixed lineage leukemia) gene, which encodes a trithorax-like histone methyltransferase, had generated leukemogenic chimeric proteins, MLL–SEPT6 (Ono et al., 2002; Kim et al., 2003; Kojima et al., 2004; Ono et al., 2005a). As has been proposed for other fusion partners of the human MLL gene (So et al., 2003), Sept6 seems to be exploited to dimerize MLL protein and potentiate its activity. To test whether the simultaneous loss of SEPT6 function also contributes to the leukemogenesis as a tumour suppressor gene, Sept6 was disrupted in mice. It had been predicted, however, that functional analysis of the highly redundant Sept6 sub-family (Sept6/8/10/11/14 ) would be extremely difficult.

Phenotype of null mutants Predictably, Sept6 −/− female and Sept6 −/Y male mice with a mixed genetic background of 129 and C57BL/6 strains were healthy and fertile, and their hematopoietic system was normal. In two models of myeloproliferative disorders, the loss of Sept6 did not alter the pathology induced by exogenous expression of an MLL–SEPT6 chimera. Extrapolating the data from the mouse models, the authors concluded that loss of SEPT6 alone does not confer susceptibility to leukemia in humans (Ono et al., 2005b). In the same study, double knockout mice were generated for the first time in the septin family. However, the dual loss of Sept4 and Sept6 did not give additional defects as compared with the single knockout of each septin gene (see section on Sept4 ). Thus septin sub-units that belong to different structural sub-families may not be functionally overlapping but complementary. These findings are compatible with a structure-based combinational hypothesis for the formation of the canonical

UBIQUITOUS SEPTINS Table 15.1

321

List of septin knockout mice

Phenotype of null mutants

Genes disrupted

References

chr. loci Brain-enriched septins

Sept3

15E2

Sept4

11C

No CNS anomaly 1. Hypodopaminergic anomaly 2. Enhancement of neurotoxicity by α-synuclein 3. Mild cerebellar anomaly 4. Asthenospermia 5. Enhancement of liver fibrosis

Ubiquitous septins

Sept5

16A3/16 1. No CNS anomaly 2. Elevated platelet sensitivity

Sept6

XA2

1. No systemic/CNS anomaly (even in DKO with Sept4) 2. No additional effects on MLL-Sept6-induced leukemia

Sept7

9A4

Embryonic lethal

Sept9

11 E2/11 Embryonic lethal

Fujishima et al. 2007 Ihara et al. 2007 Ihara et al. unpublished Ihara et al., Kissel et al. 2005 Iwaisako et al. 2005 Peng et al. 2002 Dent et al. 2002

Ono et al. 2005 Ware et al. unpublished Fuchtbauer et al. unpublished

Six of the 13 septin genes in the mouse genome have been disrupted conventionally, to the author’s knowledge as of September 2008. Chromosomal locations of other septin genes with cytoband positions are as follows: Sept1 (7F4), Sept2 (1D), Sept8 (11B1.3|11), Sept10 (10B4), Sept11 (5E2|5), Sept12 (16A1) and Sept14 (9C|9). (No ortholog for human SEPT13 has been found in the mouse genome.) Although future studies necessitate multi-gene disruptions within and beyond the septin family, those involving multiple loci on the same chromosome are difficult to obtain: for example, Sept4 , Sept8 and Sept9 on Chromosome 11; Sept7 and Sept14 on Chromosome 9. See text and references for details.

(i.e. Sept2/6/7 type) septin hetero-oligomers (Kinoshita, 2003), which predicts that Sept4 may substitute for Sept2, but not for Sept6 or Sept7.

Sept7 Background Human SEPT7 (Nakatsuru, Sudo and Nakamura, 1994) and mouse Sept7 (Soulier and Vilotte, 1998) encode the pivotal sub-units in the ubiquitous septin complexes of Sept2/6/7 (Joberty et al., 2001; Sheffield et al., 2003) and Sept7/9b/11 (Nagata et al., 2004). The significance of Sept7 is indicated by the fact that RNAi-mediated

322

CH15 INSIGHT INTO SEPTIN FUNCTIONS FROM MOUSE MODELS

acute depletion of Sept7 from mammalian cells also depletes the other sub-units (Kinoshita et al., 2002; Kremer, Haystead and Macara, 2005). This is perhaps because the formation of tight complexes stabilizes the septin sub-units, hence extending the lifetime of each. Although the human genome harbours SEPT13 as a duplicate for SEPT7 (Hall et al., 2004), it appears from current evidence that Sept7 stands alone in the mouse genome (personal communication with Dr. Elspeth Bruford of the HUGO Gene Nomenclature Committee). Thus, the hypothesis in the last part of the section on Sept6 predicts that Sept7 is irreplaceable in the canonical septin complexes in the mouse and that loss of Sept7 should totally abolish them, causing mitotic defects and severe deterioration in embryonic development.

Phenotype of null mutants Sept7 +/− mice generated from two independent gene trap lines were fertile and appeared healthy, although latent haploinsufficiency might be present. Sept7 −/− mice were never born from heterozygous pairs, verifying that damage to fundamental cellular and/or developmental processes by the loss of Sept7 sub-unit is too severe to be compensated. Although Sept7 −/− embryos have not been examined, the lethality seems due most simply to mitotic failure, considering the extensive expression of Sept7 in the early developmental stages (personal communication with Dr. Jerry Ware at the University of Arkansas).

Sept9 Background Sept9, along with Sept3 and Sept12, comprises the most arcane septin sub-family. Human SEPT9 gene can generate up to 15 polypeptides via complex variable splicing (Robertson et al., 2004 and see Chapter 7). In tissue culture cells, acute perturbation or depletion of major Sept9 isoforms interfered with mitosis, probably due to disorganized mitotic and/or cytokinetic apparatus (Surka, Tsang and Trimble, 2002; Nagata et al., 2003). Sept9 is also implicated in the invasion of an intra-cellular pathogen Listeria monocytogenes by interacting directly or indirectly with bacterial proteins for internalization (Pizarro-Cerd´a et al., 2002; Mostowy et al., unpublished ). As with SEPT6 , MLL–SEPT9 fusion genes formed by chromosomal translocations of t(11;17) were found in AML (Osaka, Rowley and Zeleznik, 1999) and myelodysplastic syndrome (Kreuziger et al., 2007). SEPT9 is often involved in chromosomal deletions at 17q25 as found in human ovarian and breast cancer (Russell et al., 2000; Kalikin, Sims and Petty, 2000). These findings have raised the possibility that SEPT9 is a tumour suppressor or a susceptibility gene. Conversely, a variety of human/mouse malignancies exhibit dysregulated expression of SEPT9/Sept9 , including over-expression caused by copious gene

TISSUE-SPECIFIC SEPTINS

323

amplification (Montagna et al., 2003; Scott et al., 2005). The tumour-promoting potential of Sept9 has also been indicated by a cell biological study which found that a major Sept9 isoform can associate with a transcription factor HIF1α and potentiate its growth-promoting and angiogenic activities (Amir et al., 2006). Aside from the somatic gene rearrangements found in malignancy, germline mutations in human SEPT9 are responsible for hereditary neuralgic amyotrophy (HNA), which is a recurrent peripheral neuropathy triggered by various stresses (Kuhlenb¨aumer et al., 2005). Although the pathogenic mechanism is as yet unclear, one of the mutations that cause HNA enhances the translational efficiency of SEPT9 mRNA (McDade, Hall and Russell, 2007). Thus it is possible that the resulting Sept9 overload disturbs neural homeostasis, as demonstrated by exogenous over-expression of Sept5 (Section 15.3.3) (Beites et al., 1999; Dong et al., 2003). The above findings, although fragmentary and partly inconsistent, seem to indicate the multi-functionality and complexity of Sept9.

Phenotype of null mutants Sept9 +/− mice were fertile and appeared healthy, while homozygous loss of Sept9 caused mid-gestational embryonic lethality. Although the details may differ by genetic background, embryonic growth was retarded typically from 8 dpc (days post coitus) onward. Development of the blood islands in the yolk sac, the larger blood vessels and the heartbeat appeared normal in Sept9 −/− embryos, even with obvious growth retardation. Around 9 dpc, degenerative changes with apoptosis developed in the mesenchymal tissues of Sept9 −/− embryos (Plate 15.1, see p. 246), and all of them died by 10.5 dpc (personal communication with Dr. Ernst-Marin F¨uchtbauer at the University of Aarhus). Although cell biological studies have consistently indicated a role for Sept9 in mitosis/cytokinesis, the apparently normal early embryogenesis could argue against absolute requirement of Sept9 for cell division. Additionally, the midgestational developmental failure could be due in part to an insufficiency of HIF-1α-mediated cell proliferation and/or angiogenesis.

TISSUE-SPECIFIC SEPTINS Sept3 Background Sept3, a Sept9 paralog, is expressed exclusively and extensively in the developing brain (Xiong, Leahy and Stuhlmann, 1999), localized to neurites and pre-synaptic terminals (Xue et al., 2000, 2004b; Takehashi et al., 2004; Fujishima et al., 2007). Pre-synaptic co-localization with dynamin 1 and Serine-91 phosphorylation by cGMP-dependent protein kinase I indicate involvement of Sept3 in

324

CH15 INSIGHT INTO SEPTIN FUNCTIONS FROM MOUSE MODELS

neurotransmission (Xue et al., 2000; Xue et al., 2004a; Xue et al., 2004b). However, the biological significance and molecular function of Sept3, including whether Sept3 constitutes typical heterotrimeric septin complexes, are still unknown.

Phenotype of null mutants Sept3 −/− mice were fertile and appeared to have normal brain architecture. Sept3 −/− hippocampal neurons in primary culture developed dendrites and axons with normal appearance. Although Sept3 was co-localized and coimmunoprecipitated with Sept5 and Sept7, the amounts of Sept5 and Sept7 were unaltered by the absence of Sept3 (Fujishima et al., 2007). The loss of Sept3 may be compensated by the ubiquitous paralog Sept9 , or tolerated due to other homeostatic mechanisms. Generating Sept3 −/− Sept9 +/− mice, or crossbreeding Sept3 −/− mice with those lacking functionally relevant genes such as dynamin 1 and/or 2 (Ferguson et al., 2007), might give a synthetic neural phenotype that provides clues to the molecular function of Sept3.

Sept4 Background Sept4, along with Sept1 and Sept5, belongs to the Sept2 sub-family. Besides the unique amino-terminal extension, Sept4 is distinct among the septin family owing to its spatiotemporal expression pattern: Sept4 is abundantly present in post-mitotic Bergmann glia in the cerebellar cortex, oligodendrocytes in the cerebral white matter (Kinoshita, Noda and Kinoshita, 2000; Hagiwara et al., unpublished ), dopaminergic (DA) neurons projecting from the midbrain, motor neurons in the spinal cord (Ihara et al., 2007), retinal horizontal cells (Pache et al., 2005), hepatic stellate cells (Iwaisako et al., 2008) and spermatozoa (Ihara et al., 2005; Kissel et al., 2005). These findings suggest unknown post-mitotic and post-meiotic roles for the septin system in disparate cell types. Sept4 and its closest paralog, Sept5, are implicated in neurodegenerative disorders. PARK2 (autosomal recessive early-onset Parkinson’s disease) is caused by loss of function of parkin, an E3 ubiquitin ligase. A major hypothesis for the pathogenesis of PARK2 is that some of the parkin substrates, including Sept4 and Sept5, accumulate and damage DA neurons (Zhang et al., 2000; Choi et al., 2003). Although acute over-expression of Sept4 or Sept5 can cause proteotoxic stress (Dong et al., 2003; Ihara et al., 2003), it is unclear whether DA neurons in PARK2 patients or parkin-deficient mice are chronically overloaded with these septins. In sporadic Parkinson’s disease, pathological cytoplasmic aggregates of αsynuclein (αS) involve Sept4 as one of accessory components (Ihara et al., 2003). In most cases, however, Sept4 is depleted from the striatum where DA neurons form pre-synaptic terminals and release dopamine (Figure 15.1a) (Ihara et al.,

TISSUE-SPECIFIC SEPTINS

(a)

(b)

325

(c)

Figure 15.1 Sept4 function in dopaminergic nerve terminals in the striatum. (a) A scanning laser confocal microscopic image of a mouse striatal section, immunostained for DNA, the dopamine transporter and tyrosine hydroxylase, a rate-limiting enzyme for dopamine synthesis. Note fine network of highly branched axon terminals that synthesize, release and reuptake dopamine. The nucleated cell bodies belong to the post-synaptic GABAergic neurons. A colour version of (a) appears as plate 15.2, see p. 246 (b) A brightfield microscopic image of a mouse striatal section, immunostained for Sept4 (DAB method, and a higher magnification than (a). Note the granular enrichment of Sept4, which corresponds to the dopaminergic pre-synaptic apparatus termed varicosities. Genetic loss of Sept4 reduced the content of the dopamine transporter, tyrosine hydroxylase and other proteins, without affecting axon terminal morphology. The attenuated dopaminergic neurotransmission was pinpointed in comprehensive behavioural screening. See text and Ihara et al., (2007) for details. (c) An immuno-EM image of Sept4 −/− striatum doubly labelled for tyrosine hydroxylase (DAB deposit) and type 2 dopamine receptor at post-synapse (immunogold). Note dopamine vesicles in tyrosine hydroxylase-positive axon terminals. Morphometric analyses based on these and other data indicated that the Sept4 sub-unit is dispensable as a cytoskeletal component in DA axon terminals. So far, however, even our best Sept4 antibody (used in b) has been unable to give high-contrast signals in immuno-EM. (a) and (c), unpublished data provided by courtesy of Dr. Akari Hagiwara

2007). Thus, mild up-regulation of SEPT4 mRNA found in DA neurons in some cases of Parkinson’s disease (Moran et al., 2007) may be a compensatory mechanism for the loss of Sept4 protein, rather than overloading Sept4. Intriguingly, these findings appear to be specific to Sept4 but not the case for Sept5 and other septins. As with SEPT9 , SEPT4 has been implicated in human malignancy both positively and negatively. A group reported up-regulation of SEPT4 in colon cancer and that ribozymes directed against SEPT4 suppressed cell proliferation (Tanaka et al., 2002). Another group have reported down-regulation of a pro-apoptotic isoform of SEPT4 in acute lymphoblastic leukemia (Chapter 13). However, the significance of the 32kDa variant is not testable in model organisms as it appears unique to human SEPT4 .

Neural phenotype of null mutants It had been anticipated that loss of Sept4 might give an uninformative phenotype as seen in other neural septin knockout mice. Alternatively, Sept4 −/− mice might

326

CH15 INSIGHT INTO SEPTIN FUNCTIONS FROM MOUSE MODELS

exhibit a complex phenotype, considering the fact that Sept4 mRNA is expressed in embryonic brain (Kato, 1990), and far more abundantly in post-natal brain in disparate neuronal and glial lineages (Section 15.3.2.1). In fact, Sept4 −/− mice of C57BL/6J strain exhibit a mild developmental anomaly in the cerebellum, the mechanism of which is currently unclear (Ihara, Hagiwara et al., unpublished ). Nevertheless, some of the potential problems have been overcome by a comprehensive physical and behavioural screening (Takao and Miyakawa, 2006). The unbiased screening pinpointed a specific augmentation in the pre-pulse inhibition (PPI) of the acoustic startle response in Sept4 −/− mice, indicating their attenuated nigrostriatal DA transmission. When compared with wildtype controls, Sept4 −/− DA neurons appear normal in morphology. However, they contain less dopamine and less proteins for dopamine metabolism, that is dopamine synthesis (tyrosine hydroxylase), release (syntaxin-1A, SNAP-25, Munc18), and reuptake (the dopamine transporter, αS). Sept4 and these molecules are co-localized at DA axon terminals (Figure 15.1), and co-immunoprecipitated from striatal homogenate (except tyrosine hydroxylase). In contrast, transgenic (Tg) mice that express the largest (54 kDa) isoform of Sept4 in the brain exhibit a diminished PPI, indicating a hyper-dopaminergic state. Together, the Sept4 knockout and Tg mice concordantly indicate that the septin system at DA axon terminals has a positive role in the maintenance of the pre-synaptic machinery for dopamine turnover. The detailed molecular mechanisms underlying the hypo-dopaminergic phenotype of Sept4 −/− mice, including which defect is the direct effect by the loss of Sept4 and which are adaptive responses, are being investigated. Another point needs to be considered is that the defect is limited in DA neurons despite the widespread expression of Sept4 in the nervous systems. This intriguing form of specificity may be due to a possible scarcity of redundant septin species, and/or based on a special requirement for Sept4, but not its paralogs, in DA neurons.

Neural phenotype of null mutants harboring α-synuclein transgene Sept4 is commonly involved in αS-based cytoplasmic aggregates formed characteristically in sporadic Parkinson’s disease and related disorders, collectively termed α-synucleinopathies (Ihara et al., 2003). To test the significance of the pathological Sept4-αS association in vivo, Sept4 was deleted from Tg mice that express human αSA53T by the prion gene promoter (Giasson et al., 2002). αSA53T is a mutant protein responsible for PARK1, a dominant trait of familial Parkinson’s disease. In the mouse model of α-synucleinopathy, neurons in the brainstem and spinal cord degenerate bearing αSA53T -based amyloid deposits, and half of the biallelic Tg mice die by 18 months after birth. Although this model does not reproduce

TISSUE-SPECIFIC SEPTINS

327

degeneration of DA neurons per se, it recapitulates human α-synucleinopathy in that endogenous Sept4 is co-aggregated with the αSA53T -based deposits. Loss of Sept4 from αSA53T Tg mice significantly enhances the amyloid deposits, neuropathology and locomotor deterioration, shortening the lifespan by three to four months. Intriguingly, Sept4 has turned out to be dispensable for, and rather suppressive against, the formation of αS-aggregates. One of the most significant biochemical findings is that the amount of amyloid deposit of Ser129 -phosphorylated αSA53T is negatively correlated with the dosage of Sept4, because pSer129 αS is highly correlated with the severity of Parkinson’s disease pathology in humans (Fujiwara et al., 2002). Another point is that αSA53T overload alone never forms pSer129 αS aggregates in DA neurons, but additional loss of Sept4 does (Ihara et al., 2007). Thus, the insufficiency of Sept4 could increase the vulnerability of DA neurons in Parkinson’s disease. In vitro data indicated conformation-sensitive interaction of αS with its protein partners including Sept4 (Woods et al., 2007). The interaction of Sept4 protects αS against Ser129 -phosphorylation and self-aggregation, which are most simply explained by steric hindrance (Ihara et al., 2007). The above data from Sept4 −/− mice, Sept4 Tg mice and αSA53T Tg × Sept4 −/− mice, and post-mortem brain analyses of Parkinson’s disease indicate that Sept4 may be involved as a dual susceptibility factor, because its insufficiency can diminish DA neurotransmission and enhance αS modifications and neurotoxicity. They have also reconfirmed that mouse genetics is a powerful tool to investigate pathophysiological mechanisms underlying human diseases.

Reproductive phenotype of null mutants Sept4 −/− male mice are unexpectedly sterile, without any signs of developmental or hormonal abnormality in their reproductive organs. In the testis, Sept4 is expressed exclusively in the post-meiotic (i.e. haploid) germ cells. Sept4-null spermatozoa are defective in flagellar morphology and motility. The annulus, a cortical ring that separates the proximal and distal parts of the flagellum, is replaced by a fragile segment without cortical material. Sept4 and several other septins are co-localized in the annulus, indicating that the annulus is a septin ring (Figure 15.2). The fragile segment which lacks the annulus snaps and/or falls apart when spermatozoa are maturated to start flagellar motions in the epididymis. Thus, the septin ring is likely to serve here as a local cytoskeleton to reinforce the flagellar cortex against the mechanical stress, which may be compared to a truss hoop or a corset. The lack of Sept4 causes other defects in spermatozoa, which indicate poor compensatory capacity of the haploid cells against various insults: They exhibit possible abnormalities in mitochondrial morphology, kinesin-mediated intra-flagellar transport, and protein phosphorylation associated with capacitation (functional maturation). Although the details and causality are as yet unknown, the pleiotropic

328

CH15 INSIGHT INTO SEPTIN FUNCTIONS FROM MOUSE MODELS

(a)

(b)

(c)

Figure 15.2 Spermiogenesis defect observed in Sept4 knockout mice. (a) Immunofluorescence signals superimposed on a DIC image of wildtype spermatozoa. Nuclear DNA, Sept4, and mitochondria, respectively, are labelled by DAPI, an anti-Sept4 antibody, and MitoTracker Red (Molecular Probes). Note that a septin-based ring (the annulus) is localized to the caudal end of the middle piece, a segment of flagellum covered with a mitochondrial sheath. A colour version of this figure appears as plate 15.3, see p. 246 (b and c) Scanning electron microscopy images of spermatozoa from wildtype (b) and Sept4 −/− mice (c) shown at the same magnification (head to the left). Note that the annulus/septin ring is completely abolished. (b and c, unpublished data provided by courtesy of Dr. Gary Hunnicutt.) For details, see Ihara et al. (2005) and Kissel et al. (2005)

TISSUE-SPECIFIC SEPTINS

329

anomalies have provided a glimpse into the multi-functionality of the septin system in vivo (Ihara et al., 2005; Kissel et al., 2005). A phenocopy of Sept4-null spermatozoa, that is, defective annulus organization and capacitation, is caused by the genetic loss of Slc26a8, an annulus-localized hypothetical anion transporter which belongs to the solute carrier protein family (Tour´e et al., 2007). Although the predicted anion transporter function of Slc26a8 and its significance remain to be tested, it appears that physical contact with this membrane-bound protein ensures the proper sub-membranous positioning and organization of the septin ring as the annulus. The intriguing set of mutants will help build a mechanistic model of the structural and functional organization of the annulus during spermiogenesis. It may also provide a hint of the pathogenesis of human asthenospermia (male infertility syndrome due to immotile spermatozoa) (Ihara et al., 2005; Sugino et al., in press).

Sept5 Background Human SEPT5 is another fusion partner of the MLL gene in AML with t(11;22) translocations (Megonigal et al., 1998; Tatsumi et al., 2001). Sept5 is present in megakaryocytes and blood platelets, but most abundantly in the brain (Yagi et al., 1998; Caltagarone et al., 1998; Toda et al., 2000; Martinez and Ware, 2004; Martinez et al., 2006). Sept5 is down-regulated in the striatum of rats reared in social isolation, suggesting its association with neural activity (Barr et al., 2004). Immuno-EM highlighted Sept5 surrounding synaptic vesicles in GABAergic nerve terminals in the striatum (Kinoshita, Noda and Kinoshita, 2000) and α-granules in blood platelets (Dent et al., 2002). Biochemically, Sept5 is co-purified from a synaptosomal fraction of rodent brain with SNAP-25, synaptophysin, VAMP-2, syntaxin-1A and -4, and is able to interact with the syntaxins (Caltagarone et al., 1998; Dent et al., 2002). The Sept5–syntaxin interaction is negatively regulated by α-SNAP (Beites et al., 2005) or Cdk5-mediated phosphorylation of Sept5 (Taniguchi et al., 2007; Amin et al., 2008). Acute overload of Sept5 can interfere with exocytosis in tissue culture cells (Beites et al., 1999) and damage neurons in the rat (Dong et al., 2003). BAC transgenic mice that harbour a human chromosome segment of 22q11.2, which contains five genes including SEPT5 , exhibit psychosis-like hyperactive behaviour (Hiroi et al., 2005).

Brain phenotype of null mutants Sept5 −/− mice, generated as the first knockout mice among the mammalian septin gene family, were fertile and appeared normal. They did not show obvious abnormalities in gross brain architecture, morphology of primary cultured

330

CH15 INSIGHT INTO SEPTIN FUNCTIONS FROM MOUSE MODELS

hippocampal neurons and their electrophysiological properties in the hippocampal CA1 region (i.e. short-term synaptic plasticity, size of the releasable pool of synaptic vesicles and long-term potentiation). Unlike Sept3, loss of Sept5 caused upregulation of a paralog Sept2 and down-regulation of Sept7. It has been speculated that loss of Sept5 is compensated by these adaptive mechanisms (Peng et al., 2002). Given that Sept2 is normally expressed in glial lineages (Kinoshita et al., 2004), whether and how the Sept2 gene is turned on in Sept5 −/− neurons needs to be examined.

Platelet phenotype of null mutants Blood platelets collected from Sept5 −/− mice released serotonin-containing granules and aggregated in response to a sub-threshold level of collagen (Dent et al., 2002). Sept5-null platelets released 2.5-fold more ATP than wild type control to the same dose of the agonist (Martinez and Ware, 2004). Although the platelet hypersensitivity was the first aberrant phenotype caused by the genetic loss of a mammalian septin gene, it is not severe enough to cause thrombosis in vivo. Intriguingly, Sept5-null platelets showed normal sensitivity to ADP, which is another agonist that directly induces aggregation without secretion. These findings, along with the perigranular localization of Sept5 and the data on neural Sept5, implicate platelet septins in the vesicle fusion machinery.

CONCLUDING REMARKS: WHAT HAS BEEN LEARNT FROM THE SEPTIN KNOCKOUT MICE? Previous studies with tissue culture cells have indicated that mammalian septins are associated with diverse cellular processes operated mainly by the actin, microtubule and vesicle transport systems. As is often the case with molecules involved in fundamental cellular processes, disruption of a mouse septin gene has caused embryonic lethality (Sept7 and Sept9 ) or given little or no abnormality (Sept3 , Sept5 and Sept6 ) due to compensation by functionally redundant paralogs and/or other adaptive mechanisms. Exceptionally, Sept4 −/− mice gave obvious but unexpected defects in behaviour and reproduction. These facts illustrate that phenotype by the loss of a given septin gene is unpredictable. Thus, it would be necessary not only to assess upcoming septin mutants thoroughly but also to reassess the existing non-lethal mutants by unbiased methods; for example, comprehensive and quantitative physical, histological and behavioural examination for subtle abnormalities (Crawley, 2007) blood/urine biochemistry, DNA micro-array analysis for aberrant gene expression profiles, crossbreeding each mutant with another to generate synthetic phenotypes, and so on. The embryonic lethality observed in Sept7 −/− mice had been more or less expected from the phenotype of Drosophila Pnut (a Sept7 homologue) mutant

REFERENCES

331

(Neufeld and Rubin, 1994). Although the early embryonic lethality of Sept7 −/− or Sept9 −/− mice is most likely due to mitotic defects, it remains to be established whether or not other septin-mediated processes are also responsible for the lethality. Future studies should overcome the limitation of the embryonic lethality caused by conventional gene disruption. Genetic analysis with conditional gene disruption, hypomorphic alleles, transgenic mice that express mutant septins or short hairpin RNAs, or combination of these will give a variety of informative phenotypes. For instance, intriguing in vitro phenotypes observed in RNAi studies await validation in vivo: Depletion of Sept7 or other sub-units affects dendritic spine morphology in primary cultured neurons (Xie et al., 2007; Tada et al., 2007), cortical integrity of T cells (Tooley et al., in press) and host– pathogen interaction (Mostowy et al., unpublished ). Therefore, conditional disruption of the essential septin sub-units, or spatiotemporally regulated septin perturbation in these and other systems should reveal the diversity and complexity of the physiological and pathological processes mediated by the septin system.

ACKNOWLEDGEMENTS The author is grateful to Dr Jerry Ware (Arkansas Cancer Research Center, University of Arkansas for Medical Sciences) and Dr Ernst-Martin F¨uchtbauer (Department of Molecular Biology, University of Aarhus) for generously sharing information on the unpublished knockout mice, Dr Gary R. Hunnicutt (Population Council, Center for Biomedical Research, Rockefeller University) and Dr Akari Hagiwara (Kyoto University, presently at Harvard University) for the unpublished high-resolution microscopy images, Dr Elspeth Bruford (HUGO Gene Nomenclature Committee) for the updated genomic data.

REFERENCES Amin, N.D., Zheng, Y.L., Kesavapany, S. et al. (2008) Cyclin-dependent kinase 5 phosphorylation of human septin SEPT5 (hCDCrel-1) modulates exocytosis. J Neurosci , 28, 3631–43. Amir, S., Wang, R., Matzkin, H. et al. (2006) MSF-A interacts with hypoxia-inducible factor-1alpha and augments hypoxia-inducible factor transcriptional activation to affect tumorigenicity and angiogenesis. Cancer Research, 66, 856–66. Barr, A.M., Young, C.E., Sawada, K. et al. (2004) Abnormalities of presynaptic protein CDCrel-1 in striatum of rats reared in social isolation: relevance to neural connectivity in schizophrenia. The European Journal of Neuroscience, 20, 303–7. Beites, C.L., Campbell, K.A. and Trimble, W.S. (2005) The septin Sept5/CDCrel-1 competes with alpha-SNAP for binding to the SNARE complex. The Biochemical Journal , 385, 347–53.

332

CH15 INSIGHT INTO SEPTIN FUNCTIONS FROM MOUSE MODELS

Beites, C.L., Xie, H., Bowser, R. and Trimble, W.S. (1999) The septin CDCrel-1 binds syntaxin and inhibits exocytosis. Nature Neuroscience, 2, 434–39. Caltagarone, J., Rhodes, J., Honer, W.G. and Bowser, R. (1998) Localization of a novel septin protein, hCDCrel-1, in neurons of human brain. Neuroreport, 9, 2907–12. Choi, P., Snyder, H., Petrucelli, L. et al. (2003) SEPT5 v2 is a parkin-binding protein. Brain Research. Molecular Brain Research, 117, 179–89. Crawley, J.N. What’s Wrong With My Mouse?: Behavioral Phenotyping of Transgenic and Knockout Mice. (2nd ed.) Wiley-Liss. Dent, J., Kato, K., Peng, X.R. et al. (2002) A prototypic platelet septin and its participation in secretion. Proceedings of the National Academy of Sciences of the United States of America, 99, 3064–69. Dong, Z., Ferger, B., Paterna, J.C. et al. (2003) Dopamine-dependent neurodegeneration in rats induced by viral vector-mediated overexpression of the parkin target protein, CDCrel-1. Proceedings of the National Academy of Sciences of the United States of America, 100, 12438–43. Ferguson, S.M., Brasnjo, G., Hayashi, M. et al. (2007) A selective activity-dependent requirement for dynamin 1 in synaptic vesicle endocytosis. Science, 316, 570–74. Fujishima, K., Kurisu, J., Hirano, T. and Kengaku, M. (2007) Targeted disruption of Sept3, a heteromeric assembly partner of Sept5 and Sept7 in axons, has no effect on developing CNS neurons. Journal of Neurochemistry, 102, 77–92. Fujiwara, H., Hasegawa, M., Dohmae, N. et al. (2002) alpha-Synuclein is phosphorylated in synucleinopathy lesions. Nature Cell Biology, 4, 160–64. Giasson, B.I., Duda, J.E., Quinn, S.M. et al. (2002) Neuronal alpha-synucleinopathy with severe movement disorder in mice expressing A53T human alpha-synuclein. Neuron, 34, 521–33. Hall, P.A., Russell, S.E., (2004) The pathobiology of the septin gene family. J Pathol., 204, 489–505. Hiroi, N., Zhu, H., Lee, M. et al. (2005) A 200-kb region of human chromosome 22q11.2 confers antipsychotic-responsive behavioral abnormalities in mice. Proceedings of the National Academy of Sciences of the United States of America, 102, 19132–37. Ihara, M., Kinoshita, A., Yamada, S. et al. (2005) Cortical organization by the septin cytoskeleton is essential for structural and mechanical integrity of mammalian spermatozoa. Developmental Cell , 8, 343–52. Ihara, M., Yamasaki, N., Hagiwara, A. et al. (2007) Sept4, a component of presynaptic scaffold and Lewy bodies, is required for the suppression of α-synuclein neurotoxicity. Neuron, 53, 519–33. Iwaisako, K., Hatano, E., Taura, K., Nakajima, A., Tada, M., Seo, S., Tamaki, N., Ikai, I., Uemoto, S., Kinoshita, M. (2008) Loss of Sept4 exacerbates liver fibrosis through the dysregulation of hepatic stellate cells. Journal of Hepatology, available online 15 July 2008. Joberty, G., Perlungher, R.R., Sheffield, P.J. et al. (2001) Borg proteins control septin organization and are negatively regulated by Cdc42. Nature Cell Biology, 3, 861–66. Kalikin, L.M., Sims, H.L. and Petty, E.M. (2000) Genomic and expression analyses of alternatively spliced transcripts of the MLL septin-like fusion gene (MSF) that map to a 17q25 region of loss in breast and ovarian tumors. Genomics, 63, 165–72. Kato, K. (1990) A Collection of cDNA Clones with specific expression patterns in mouse brain. The European Journal of Neuroscience, 2, 704–11.

REFERENCES

333

Kim, H.J., Ki, C.S., Park, Q. et al. (2003) MLL/SEPTIN6 chimeric transcript from inv ins(X;11)(q24;q23q13) in acute monocytic leukemia: report of a case and review of the literature. Genes, Chromosomes and Cancer, 38, 8–12. Kinoshita, M. (2003) Assembly of mammalian septins. Journal of Biochemistry (Tokyo), 134, 491–96. Kinoshita, M., Field, C.M., Coughlin, M.L. et al. (2002) Self- and actin-templated assembly of Mammalian septins. Developmental Cell , 3, 791–802. Kinoshita M, Kumar S, Mizoguchi A, Ide C, Kinoshita A, Haraguchi T, Hiraoka Y, Noda M. (1997) Nedd5, a mammalian septin, is a novel cytoskeletal component interacting with actin-based structures. Genes Dev . 11, 1535–1547. Kinoshita, N., Kimura, K., Matsumoto, N. et al. (2004) Mammalian septin Sept2 modulates the activity of GLAST, a glutamate transporter in astrocytes. Genes Cells, 9, 1–14. Kinoshita, A., Noda, M. and Kinoshita, M. (2000) Differential localization of septins in the mouse brain. The Journal of Comparative Neurology, 428, 223–39. Kissel H, Georgescu MM, Larisch S, Manova K, Hannicutt GR, Steller H. et al. (2005) The Sept4 septin locus is required for sperm terminal differentiation in mice. Dev Cell . 8, 353–364. Kojima, K., Sakai, I., Hasegawa, A. et al. (2004) FLJ10849, a septin family gene, fuses MLL in a novel leukemia cell line CNLBC1 derived from chronic neutrophilic leukemia in transformation with t(4;11)(q21;q23). Leukemia, 18, 998– 1005. Kremer, B.E., Haystead, T. and Macara, I.G. (2005) Mammalian septins regulate microtubule stability through interaction with the microtubule-binding protein MAP4. Molecular Biology of the Cell , 16, 4648–59. Kreuziger, L.M., Porcher, J.C., Ketterling, R.P. and Steensma, D.P. (2007) An MLLSEPT9 fusion and t(11;17)(q23;q25) associated with de novo myelodysplastic syndrome. Leukemia Research, 31, 1145–48. Kuhlenb¨aumer, G., Hannibal, M.C., Nelis, E. et al. (2005) Mutations in SEPT9 cause hereditary neuralgic amyotrophy. Nature Genetics, 37, 1044–46. Ihara M., Tomimoto H., Kitamaya H. et al. (2003) Association of the cytoskeletal GTP-binding protein Sept4/H5 with cytoplasmic inclusions found in Parkinson’s disease and other synucleinopathies. J Biol Chem. 278, 24095–24102. Martinez, C., Corral, J., Dent, J.A. et al. (2006) Platelet septin complexes form rings and associate with the microtubular network. Journal of Thrombosis and Haemostasis, 4, 1388–95. Martinez, C. and Ware, J. (2004) Mammalian septin function in hemostasis and beyond. Experimental Biology and Medicine, 229, 1111–19. McDade, S.S., Hall, P.A. and Russell, S.E. (2007) Translational control of SEPT9 isoforms is perturbed in disease. Human Molecular Genetics, 16, 742–52. Megonigal, M.D., Rappaport, E.F., Jones, D.H. et al. (1998) t(11;22)(q23;q11.2) in acute myeloid leukemia of infant twins fuses MLL with hCDCrel, a cell division cycle gene in the genomic region of deletion in DiGeorge and velocardiofacial syndromes. Proceedings of the National Academy of Sciences of the United States of America, 95, 6413–18. Moran, L.B., Croisier, E., Duke, D.C. et al. (2007) Analysis of alpha-synuclein, dopamine and parkin pathways in neuropathologically confirmed parkinsonian nigra. Acta Neuropathol (Berl), 113 (3), 253–63.

334

CH15 INSIGHT INTO SEPTIN FUNCTIONS FROM MOUSE MODELS

Montagna, C., Lyu, M.S., Hunter, K. et al. (2003) The Septin 9 (MSF) gene is amplified and overexpressed in mouse mammary gland adenocarcinomas and human breast cancer cell lines. Cancer Research, 63, 2179–87. Nagata, K., Asano, T., Nozawa, Y. and Inagaki, M. (2004) Biochemical and cell biological analyses of a mammalian septin complex, Sept7/9b/11. The Journal of Biological Chemistry, 279, 55895–904. Nagata, K., Kawajiri, A., Matsui, S. et al. (2003) Filament formation of MSF-A, a mammalian septin, in human mammary epithelial cells depends on interactions with microtubules. The Journal of Biological Chemistry, 278, 18538–43. Nakatsuru, S., Sudo, K. and Nakamura, Y. (1994) Molecular cloning of a novel human cDNA homologous to CDC10 in Saccharomyces cerevisiae. Biochemical and Biophysical Research Communications, 202, 82–87. Neufeld, T.P. and Rubin, G.M. (1994) The Drosophila peanut gene is required for cytokinesis and encodes a protein similar to yeast putative bud neck filament proteins. Cell , 77, 371–79. Ono, R., Ihara, M., Nakajima, H. et al. (2005b) Disruption of Sept6, a fusion partner gene of Mixed Lineage Leukemia (MLL), does not affect the ontogeny, leukemogenesis induced by MLL-SEPT6, or the phenotype induced by the loss of Sept4. Molecular and Cell Biology of Human Diseases Series, 25, 10965–78. Ono, R., Nakajima, H., Ozaki, K. et al. (2005a) Dimerization of MLL fusion proteins and FLT3 activation synergize to induce multiple-lineage leukemogenesis. The Journal of Clinical Investigation, 115, 919–29. Ono, R., Taki, T., Taketani, T. et al. (2002) SEPTIN6, a human homologue to mouse Septin6, is fused to MLL in infant acute myeloid leukemia with complex chromosomal abnormalities involving 11q23 and Xq24. Cancer Research, 62, 333–37. Osaka, M., Rowley, J.D. and Zeleznik-Le, N.J. (1999) MSF (MLL septin-like fusion), a fusion partner gene of MLL, in a therapy-related acute myeloid leukemia with a t(11;17)(q23;q25). Proceedings of the National Academy of Sciences of the United States of America, 96, 6428–33. Pache, M., Zieger, B., Bl¨aser, S. and Meyer, P. (2005) Immunoreactivity of the septins SEPT4, SEPT5, and SEPT8 in the human eye. The Journal of Histochemistry and Cytochemistry, 53, 1139–47. Peng XR, Jia Z, Zhang Y, Ware J, Trimble WS. et al. (2002) The septin CDCrel-1 is dispensable for normal development and neurotransmitter release. MolCell Biol . 22, 378–387. Pizarro-Cerd´a, J., Jonquieres, R., Gouin, E. et al. (2002) Distinct protein patterns associated with Listeria monocytogenes InlA- or InlB-phagosomes. Cellular Microbiology, 4, 101–15. Robertson, C., Church, S.W., Nagar, H.A. et al. (2004) Properties of SEPT9 isoforms and the requirement for GTP binding. The Journal of Pathology, 203, 519–27. Russell, S.E., McIlhatton, M.A., Burrows, J.F. et al. (2000) Isolation and mapping of a human septin gene to a region on chromosome 17q, commonly deleted in sporadic epithelial ovarian tumors. Cancer Research, 60, 4729–34. Scott, M., Hyland, P.L., McGregor, G. et al. (2005) Multimodality expression profiling shows SEPT9 to be overexpressed in a wide range of human tumours. Oncogene, 24, 4688–700.

REFERENCES

335

Sheffield, P.J., Oliver, C.J., Kremer, B.E. et al. (2003) Borg/septin interactions and the assembly of mammalian septin heterodimers, trimers, and filaments. The Journal of Biological Chemistry, 278, 3483–88. So, C.W., Lin, M., Ayton, P.M. et al. (2003) Dimerization contributes to oncogenic activation of MLL chimeras in acute leukemias. Cancer Cell , 4, 99–110. Soulier, S. and Vilotte, J.L. (1998) Sequence of murine CDC10 cDNA, gene organization and expression analysis. Biochimica et Biophysica Acta, 1442, 339–46. Spiliotis, E.T., Kinoshita, M. and Nelson, W.J. (2005) A mitotic septin scaffold required for Mammalian chromosome congression and segregation. Science, 307, 1781–85. Sugino, Y., ichioka, K., Soda, T., Ihara, M., Kinoshita, M., Ogawa, O., Nishiyama, H.. Septins as diagnostic markers for a subset of human asthenozoospermia. Journal of Urology, in press. Surka, M.C., Tsang, C.W. and Trimble, W.S. (2002) The mammalian septin MSF localizes with microtubules and is required for completion of cytokinesis. Molecular Biology of the Cell , 13, 3532–45. Tada, T., Simonetta, A., Batterton, M. et al. (2007) Septin cytoskeleton stabilizes dendritic protrusion and supports dendrite development. Current Biology, 17, 1752–58. Takao, K. and Miyakawa, T. (2006) Investigating gene-to-behavior pathways in psychiatric disorders: the use of a comprehensive behavioral test battery on genetically engineered mice. Annals of the New York Academy of Sciences, 1086, 144–59. Takehashi, M., Tanaka, S., Stedeford, T. et al. (2004) Expression of septin 3 isoforms in human brain. Gene Expression, 11, 271–78. Tanaka, M., Kijima, H., Itoh, J. et al. (2002) Impaired expression of a human septin family gene Bradeion inhibits the growth and tumorigenesis of colorectal cancer in vitro and in vivo. Cancer Gene Therapy, 9, 483–88. Taniguchi, M., Taoka, M., Itakura, M. et al. (2007) Phosphorylation of adult type Sept5 (CDCrel-1) by cyclin-dependent kinase 5 inhibits interaction with syntaxin-1. The Journal of Biological Chemistry, 282, 7869–76. Tatsumi, K., Taki, T., Taniwaki, M. et al. (2001) The CDCREL1 gene fused to MLL in de novo acute myeloid leukemia with t(11;22)(q23;q11.2) and its frequent expression in myeloid leukemia cell lines. Genes, Chromosomes and Cancer, 30, 230– 35. Toda, S., Kajii, Y., Sato, M. and Nishikawa, T. (2000) Reciprocal expression of infant- and adult-preferring transcripts of CDCrel-1 septin gene in the rat neocortex. Biochemical and Biophysical Research Communications, 273, 723–28. Tooley, A., Jacobelli, J., Trimble, W.S., Kinoshita M, Krummel MF. The septin cytoskeleton is essential for cortical integrity and persistent motility in amoeboid T lymphocytes, submitted. Tour´e, A., Lhuillier, P., Gossen, J.A. et al. (2007) The testis anion transporter 1 (Slc26a8) is required for sperm terminal differentiation and male fertility in the mouse. Human Molecular Genetics, 16, 1783–93. Woods, W.S., Boettcher, J.M., Zhou, D.H. et al. (2007) Conformation-specific binding of alpha-synuclein to novel protein partners detected by phage display and NMR spectroscopy. The Journal of Biological Chemistry, 282, 34555–67.

336

CH15 INSIGHT INTO SEPTIN FUNCTIONS FROM MOUSE MODELS

Xie, Y., Vessey, J.P., Konecna, A. et al. (2007) The dendritic spine-associated GTPbinding protein Septin7 is a regulator of dendrite branching and dendritic spine morphology. Current Biology, 17, 1746–51. Xiong, J.W., Leahy, A. and Stuhlmann, H. (1999) Retroviral promoter-trap insertion into a novel mammalian septin gene expressed during mouse neuronal development. Mechanisms of Development, 86, 183–91. Xue, J., Milburn, P.J., Hanna, B.T. et al. (2004a) Phosphorylation of septin 3 on Ser-91 by cGMP-dependent protein kinase-I in nerve terminals. The Biochemical Journal , 381, 753–60. Xue, J., Tsang, C.W., Gai, W.P. Malladi, C.S., Trimble, W.S., Rostas, J.A., Robinson, P.J. (2004b) Septin 3 (G-septin) is a developmentally regulated phosphoprotein enriched in presynaptic nerve terminals. Journal of Neurochemistry, 91, 579–90. Xue, J., Wang, X., Malladi, C.S. et al. (2000) Phosphorylation of a new brain-specific septin, G-septin, by cGMP-dependent protein kinase. The Journal of Biological Chemistry, 275, 10047–56. Yagi, M., Zieger, B., Roth, G.J. and Ware, J. (1998) Structure and expression of the human septin gene HCDCREL-1. Gene, 212, 229–36. Zhang, Y., Gao, J., Chung, K.K. et al. (2000) Parkin functions as an E2-dependent ubiquitin- protein ligase and promotes the degradation of the synaptic vesicleassociated protein, CDCrel-1. Proceedings of the National Academy of Sciences of the United States of America, 97, 13354–59.

Section IV Envoi

The Septins Edited by Peter A. Hall, S.E. Hilary Russell and John R. Pringle © 2008 John Wiley & Sons, Ltd. ISBN: 978-0-470-51969-1

16 Septins: 2008 and beyond Peter A. Hall Institute of Pathology, School of Medicine, Queen’s University Belfast, Belfast BT12 6BL, Northern Ireland, UK

S.E. Hilary Russell Ovarian Cancer Research Laboratory, Centre for Cancer Research and Cell Biology, Queen’s University Belfast, A floor, Belfast City Hospital, Lisburn Road, Belfast BT9 7AB, Northern Ireland, UK

John R. Pringle Department of Genetics, MC 5120, 300 Pasteur Drive, M-322 Alway Building, Stanford University Medical Center, Stanford, CA 94305-5120, USA

Like people in the dead of night looking at some vista, each with a pen torch, we all see different parts of the picture that are septins. Slowly, at times imperceptibly, the dawn comes and our view broadens. What we see and how we interpret the meaning of what we see will depend on many things. We all have our different perspectives, areas of interest and background knowledge, and we have different appreciations of others positions. As outlined in this monograph the dawn is certainly rising and our view of septins and the breadth of their role in biology is fast improving. But we still cannot see much that is of importance. Another aspect of the evolution of any field is that there is a property, much like inertia, that keeps a field moving in a given direction: the genesis of a field in a given area or discipline often means that a particular perspective has certain influence. The history of the septin field (as outlined in Chapter 1) is an important cameo in the story of the cell cycle and in particular cytokinesis. There can be no doubt of the importance of septins in this key cellular process and our understanding of the role of septins in cytokinesis continues to develop. But a crucial point that is fast emerging is the diversity of roles that septins have taken during evolution: an evolution that has (as described in Chapter 2) involved duplications and reduplications of septin genes, their gain and their loss, and the creation of extraordinary diversity of transcripts and isoforms particularly in vertebrates. Structural The Septins Edited by Peter A. Hall, S.E. Hilary Russell and John R. Pringle © 2008 John Wiley & Sons, Ltd. ISBN: 978-0-470-51969-1

340

CH16 SEPTINS: 2008 AND BEYOND

data that was published as this monograph was being prepared (Sirajuddin et al., 2007 and see Chapters 3 and 4) does suggest that there is a common theme in much of septin biology: the creation of heteroligomeric filamentous structures that do not have polarity. However they clearly have ‘sidedness’ with the C terminal extensions (often, but not invariably with coiled coil domains) extending out of one aspect (or side) of the filament. Such a sidedness would be able to create asymmetry, a requirement of many biological processes. Lindsey and Momany (2006) summarized the diversity of appearance of septin localization across fungi and animal cells. This important contribution allowed a synthesis from a plethora of diverse information and suggests a number of classes of septin function. They suggested that three broad classes of sub cellular localization could be defined. 1. Septins could localize to projections and thus shape and compartmentalize growth processes. In this context it is notable that septins are particularly associated with curved membranes and may have key roles in such physiological events. For example, Sept7 is required for neuronal projection formation and this can be compared to the formation of the bud neck in yeast (Barral and Mansuy, 2007). 2. Septins could localize to partitions and thus contribute to the compartmentalisation of pre-existing cellular materials. 3. Finally other septins might associate with the whole cell and be involved in aspects of membrane trafficking and cytoskeletal organisation. This last category is in fact very diverse in itself since a range of patterns of expression have been reported ranging from membrane associated, actin and/or tubulin associated, diffuse cytoplasmic and even nuclear. Recent observations in mammalian systems indicate that septins can be involved in signalling pathways including stress induced pathways and that nuclear functions of septins are indeed important (Amir et al., 2006). How might the structural principles outlined by Sirajuddin et al. (2007) and by McMurray & Thorner (Chapter 3) allow for the diversity of appearance (and in all likelihood function) of septins in diverse organisms? Certainly it is now clear that there are a huge range of non-cytokinesis associated functions for septins, particularly in animal cells. Another issue of importance is why there are so many septin genes and proteins, and why is there such diversity of alternative splice variants in man and mouse (and probably other vertebrates)? Do these represent an array of modules that cells can pick and chose from for differing physiological purposes, but with common structural features and guiding principles? Or are there different ways of using septin proteins? The parsimonious nature of evolution would suggest that the basic modular structures are the same but that the diversity of N and C termini provide for a wide range of different uses for these structures. The plethora of septin

REFERENCES

341

genes and transcripts leads to important questions about control and regulation of expression. This area was considered in Chapter 7 and it is interesting to consider why cells have invested so much genetic capital in generating not only a diversity of transcripts encoding different septin isoforms, but also have (in all cases studied in detail) multiple transcripts that encode the same polypeptide. Furthering our understanding in this area will require progress in many areas but five can be identified as of particular importance. First, the genomics and detailed characterization of other septins genes will be required: a detailed catalogue of their transcripts and encoded proteins with definition of their regulation and control. Second the tools to define the expression of these transcripts and proteins in physiological and pathological contexts. Third an insight into the detailed nature of the diverse septin complexes present in cells and the dynamics of their component moieties including post translational and other modifications. Fourth, we need a detailed understanding of the protein, and other macromolecular interactions, of septins under different conditions and cellular states. Many of these are likely to be hard to study because of the need for higher-order septin structures to be present for effective interactions and low affinity and transient interactions to potentially be of biological significance. Finally we will need an understanding of the structural biology of isoforms and variants and how they contribute (indeed if they do!) to the core filament structure. Only then will we really have significant insights into the biological roles of the diversity of septin in animals. Many of the chapters in this monograph have touched on these issues and we are well aware that the preceding is a perspective that others would take issue with. Moreover the history of science tells us that making predictions is a potentially foolish activity! However we have no doubt that coming years will provide more light on septins and their diverse roles in biology, that there will be many surprises, that some of the things we think we know and understand may prove to be less than certain, and that advances in other areas will illuminate septins. It is our hope that this monograph will provide a useful resource for those who, like us, have become fascinated by this intriguing family of genes.

REFERENCES Amir, S., Wang, R., Matzkin, H. et al. (2006) MSF-A interacts with hypoxia-inducible factor-1 alpha and augments hypoxia-inducible factor transcriptional activation to affect tumorigenicity and angiogenesis. Cancer Research, 66, 856–66. Barral, Y. and Mansuy, I.M. (2007) Septins: cellular and functional barriers of neuronal activity. Current Biology, 17, R961–63. Lindsey, R. and Momany, M. (2006) Septin localization across kingdoms: three themes with variations. Current Opinion in Microbiology, 9, 559–65. Sirajuddin, M., Farkasovsky, M., Hauer, F. et al. (2007) Structural insight into filament formation by mammalian septins. Nature, 449, 311–15.

Appendix A Septin and septin-like sequences Michelle Momany, Fangfang Pan and Russell L. Malmberg Plant Biology Department, University of Georgia, Athens, GA 30602, USA

GIa

Name in Chapter 3b

Gene Aliasd c symbol

gi|2244629 gi|31198659

AbiSep AgaHyp1

/ Pnut

gi|31202059

AgaHyp2

Sep2

A

gi|31206631

AgaHyp3

Sep1

A

gi|31204715

AgaHyp4

Sep4

A

gi|13398364

AniAspA

Cdc11

F

gi|1791305

AniAspB

Cdc3

F

gi|34811845

AniAspC

Cdc12

F

gi|34148975

AniAspD

Cdc10

F

gi|34811843

AniAspE

AspE

F

gi|45645169 gi|39580970

BtaCdc10 CbrHyp1

/ Unc59b CBG20268

A A

gi|39589843

CbrHyp2

Unc61

CBG04550

A

gi|39584450

CbrHyp3

Unc59a CBG19777

A

gi|17509405

CelUnc59

/

A

sepA

F A

Clade/speciese

Groupf

Gg

Ch

Agaricus bisporus Anopheles gambiae Anopheles gambiae Anopheles gambiae Anopheles gambiae Aspergillus nidulans Aspergillus nidulans Aspergillus nidulans Aspergillus nidulans Aspergillus nidulans Bos taurus Caenorhabditis briggsae Caenorhabditis briggsae Caenorhabditis briggsae Caenorhabditis elegans

tr 2B

+ +

+ +

1B

+

+

2B

+

+

2B

+

+

3

+

+

2A

+

+

4

+

+

1A

+

−

5

+

−

2B 2B

+ +

+ +

1B

+

+

2B

+

+

2B

+

+

The Septins Edited by Peter A. Hall, S.E. Hilary Russell and John R. Pringle © 2008 John Wiley & Sons, Ltd. ISBN: 978-0-470-51969-1

APPENDIX A

344

SEPTIN AND SEPTIN-LIKE SEQUENCES

GIa

Name in Chapter 3b

Gene Aliasd symbolc

gi|32566810

CelUnc61

/

A

gi|729090 gi|729064 gi|21435770 gi|21435778 gi|21435802 gi|46442449 gi|46444553 gi|50286825 gi|50284895 gi|50288449 gi|50289749 gi|50287113 gi|50288341 gi|50291973 gi|50286937 gi|18476091

CalCdc3 CalCdc10 CalCdc11 CalCdc12 CalSep7 CalSpr28 CalSpr3 CglHyp1 CglHyp2 CglHyp3 CglHyp4 CglHyp5 CglHyp6 CglHyp7 CglHyp8 CimSep1

/ / / / Shs1 / / Cdc3 Cdc12a Cdc10 Cdc12b Cdc11 Shs1 Spr3 Spr28 Cdc11

F F F F F F F F F F F F F F F F

gi|24637104

CimSep2

Cdc3

F

gi|24637108

CimSep3

Cdc10

F

gi|24473756

CimSep4

AspE

F

gi|50257384

CneHyp1

Cdc3

F

gi|50259101

CneHyp2

Cdc12

F

gi|50258769

CneHyp3

Cdc10

F

gi|50257720

CneHyp4

Cdc11

F

gi|50260201

CneHyp5

AspE

F

gi|41055580 gi|32822794 gi|41152396 gi|40538786

DreHyp1 DreHyp2 DreHyp4 DreMsf

Sept7 Sept8 Sept5 Sept9

gi|45709377 gi|47086783 gi|50420949

DreNedd5 DreSept6 DhaHyp1

Sept2 / Cdc3

gi|50426961

DhaHyp2

Cdc12

F

gi|50425027

DhaHyp3

Cdc10

F

gi|50426163

DhaHyp4

Spr3

F

gi|50418421

DhaHyp5

Cdc11

F

SHS1

zgc:56383 wu:fb22a03 zgc:73218 MLL septin-like fusion zgc:63587 zgc:66071

A A A A

A A F

Clade/speciese

Groupf

Gg

Ch

Caenorhabditis elegans Candida albicans Candida albicans Candida albicans Candida albicans Candida albicans Candida albicans Candida albicans Candida glabrata Candida glabrata Candida glabrata Candida glabrata Candida glabrata Candida glabrata Candida glabrata Candida glabrata Coccidioides immitis Coccidioides immitis Coccidioides immitis Coccidioides immitis Cryptococcus neoformans Cryptococcus neoformans Cryptococcus neoformans Cryptococcus neoformans Cryptococcus neoformans Danio rerio Danio rerio Danio rerio Danio rerio

1B

+

+

2A 1A 3 4 3 3 4 2A 4 1A 4 3 3 4 3 3

+ + + + + + + + + + + + + + + +

+ − + + + − + + + − + + + + + +

2A

+

+

1A

+

−

5

+

−

2A

+

+

4

+

+

1A

+

−

3

+

+

5

+

−

2B 1B 2B 1A

+ + + +

+ + + −

Danio rerio Danio rerio Debaryomyces hansenii Debaryomyces hansenii Debaryomyces hansenii Debaryomyces hansenii Debaryomyces hansenii

2B 1B 2A

+ + +

+ + +

4

+

+

1A

+

−

4

+

+

3

+

+

APPENDIX A SEPTIN AND SEPTIN-LIKE SEQUENCES GIa

Name in Chapter 3b

Gene Aliasd symbolc

gi|50414330

DhaHyp6

Shs1

F

gi|730352

DmePnut

/

A

gi|17647925

DmeSep1

/

A

gi|17738071

DmeSep2

/

A

gi|24642597

DmeSep4

/

gi|21356243

DmeSep5

/

A

gi|38047705

DyaSep2

/

A

gi|19075150

EcuSep1

Spr3a

ECU01 1370

M

gi|19074995

EcuSep2

Spr3b

ECU11 1950

M

gi|19173204

EcuSep3

Cdc11

ECU09 0820

M

gi|45198629

EgoHyp1

Cdc3

F

gi|45190841

EgoHyp2

Cdc12

F

gi|45184824

EgoHyp3

Cdc10

F

gi|45191046

EgoHyp4

Cdc11

F

gi|45199089

EgoHyp5

Spr3

F

gi|45201271

EgoHyp6

Spr28

F

gi|45185071

EgoHyp7

Shs1

F

gi|14041182 gi|29249771 gi|46121875 gi|46126005 gi|46135811 gi|46123315 gi|46128665 gi|46122029 gi|46139179 gi|16604248

GcySep1 Gla GzeHyp1 GzeHyp2 GzeHyp3 GzeHyp4 GzeHyp5 GzeHyp6 GzeHyp7 HsaSept1

/ / Cdc3 Cdc12 Cdc11 Cdc10 AspE AspE2 / SEPT1

gi|4758158

HsaSept2

SEPT2

gi|22035572

HsaSept3

SEPT3

CG9699

PNUTL3, DIFF6 KIAA0158, hNedd5, Pnutl3 DIFF6, NEDD5 SEP3

345

Clade/speciese

Groupf

Gg

Ch

3

+

+

2B

+

+

2B

+

+

1B

+

+

2B

+

+

1B

+

+

tr

−

+

4

+

+

4

+

+

3

+

−

2A

+

+

4

+

+

1A

+

−

3

+

+

4

+

+

3

+

−

3

+

+

A P F F F F F F F A

Debaryomyces hansenii Drosophila melanogaster Drosophila melanogaster Drosophila melanogaster Drosophila melanogaster Drosophila melanogaster Drosophila yakuba Encephalitozoon cuniculi Encephalitozoon cuniculi Encephalitozoon cuniculi Eremothecium gossypii Eremothecium gossypii Eremothecium gossypii Eremothecium gossypii Eremothecium gossypii Eremothecium gossypii Eremothecium gossypii Geodia cydonium Giardia lamblia Gibberella zeae Gibberella zeae Gibberella zeae Gibberella zeae Gibberella zeae Gibberella zeae Gibberella zeae Homo sapiens

2B slk 2A 4 3 1A 5 5 slk 2B

+ − + + + + + + + +

+ − + + + − − − + +

A

Homo sapiens

2B

+

+

A

Homo sapiens

1A

+

−

A

APPENDIX A

346

SEPTIN AND SEPTIN-LIKE SEQUENCES

GIa

Name in Chapter 3b

Gene Aliasd symbolc

Clade/speciese

Groupf

Gg

Ch

gi|4758942

HsaSept4

SEPT4

A

Homo sapiens

2B

+

+

gi|9945439

HsaSept5

SEPT5

A

Homo sapiens

2B

+

+

gi|22035577

HsaSept6

SEPT6

A

Homo sapiens

1B

+

+

gi|4502695

HsaSept7

SEPT7

A

Homo sapiens

2B

+

+

gi|41147049

HsaSept8

SEPT8

A

Homo sapiens

1B

+

+

gi|6683817

HsaSept9

SEPT9

gi|18088518

HsaSept10

gi|8922712

HsaSept11

gi|23242699

HsaSept12

Gi|113418512

HsaSept13

SEPT 10/ SEPT 11 SEPT 12/ SEPT 13/

Gi|153252197

N/A

gi|50306547

KlaHyp1

SEPT 14+ Cdc3

gi|50309827

KlaHyp2

Cdc12

F

gi|50311269

KlaHyp3

Cdc10

F

gi|50303889

KlaHyp4

Spr3

F

gi|50304439

KlaHyp5

Cdc11

F

gi|50311965

KlaHyp6

Shs1

F

gi|50311291

KlaHyp7

Spr28

F

gi|13358928

MfaHyp1

Sept5

A

H5, CE5B3, hucep-7, bradeion, ARTS, hCDCREL2, MART, PNUTL2 HCDCREL-1, H5, PNUTL1 KIAA0128, SEP2, SEPT2, MGC16619, MGC20339 CDC3, SEPT7A, CDC10 KIAA0202, SEP2 MSF1, KIAA0991, PNUTL4, AF17q25, SeptD1, OvBr septin, MSF FLJ11619

A

Homo sapiens

1A

+

−

A

Homo sapiens

1B

+

+

FLJ10849

A

Homo sapiens

1B

+

+

FLJ25410

A

Homo sapiens

1A

+

−

DKFZp 313J1114, SEPT7B, SEPT7P +

A

Homo sapiens

2B

+

+

A

Homo sapiens

1B

+

+

F

Kluyveromyces lactis Kluyveromyces lactis Kluyveromyces lactis Kluyveromyces lactis Kluyveromyces lactis Kluyveromyces lactis Kluyveromyces lactis Macaca fascicularis

2A

+

+

4

+

+

1A

+

−

4

+

+

3

+

+

3

+

+

3

+

−

2B

+

+

APPENDIX A SEPTIN AND SEPTIN-LIKE SEQUENCES GIa

Name in Chapter 3b

Gene Aliasd symbolc

gi|38110101

MgrHyp1

Cdc3

F

gi|38106951

MgrHyp2

Cdc12

F

gi|38109157

MgrHyp3

Cdc10

F

gi|38100755

MgrHyp4

Cdc11

F

gi|38100224

MgrHyp5

AspE

F

gi|38110686

MgrHyp6

AspE2

F

gi|6453576

MciSepA

/

sepA

F

gi|8567344 gi|6754816 gi|6755468

MmuSept1 MmuSept2 MmuSept3

Sept1 Sept2 Sept3

gi|6755120 gi|6685763

MmuSept4 MmuSept5

Sept4 Sept5

Diff6 Nedd5 mKIAA 0991, G septin M-Septin, H5 Cdcrel-1, Pnutl1

gi|20178348 gi|9789726 gi|39930477 gi|28204888

MmuSept6 MmuSept7 MmuSept8 MmuSept9

Sept6 Sept7 Sept8 Sept9

gi|26345492

MmuSept10a MmuSept11 MmuSept12

Sept10

Sept14

gi|32417050

MmuSept10b NcrHyp1

gi|32404966

347

Clade/speciese

Groupf

Gg

Ch

2A

+

+

4

+

+

1A

+

−

3

+

+

5

+

−

5

+

−

4

+

+

A A A

Magnaporthe grisea Magnaporthe grisea Magnaporthe grisea Magnaporthe grisea Magnaporthe grisea Magnaporthe grisea Mucor circinelloides Mus musculus Mus musculus Mus musculus

2B 2B 1A

+ + +

+ + −

A A

Mus musculus Mus musculus

2B 2B

+ +

+ +

A A A A

Mus musculus Mus musculus Mus musculus Mus musculus

1B 2B 1B 1A

+ + + +

+ + + −

A

Mus musculus

1B

+

+

A A

Mus musculus Mus musculus

1B 1A

+ +

+ −

A

Mus musculus

1B

+

+

Cdc3

F

2A

+

+

NcrHyp2

Cdc12

F

4

+

+

gi|32404320

NcrHyp3

Cdc10

F

1A

+

−

gi|32422439

NcrHyp4

Cdc11

F

3

+

+

gi|32417420

NcrHyp5

AspE

F

5

+

−

gi|32411577

NcrHyp6

AspE2

F

5

+

−

gi|32411845

NcrHyp7

slk

−

−

gi|5725417

PbrPbs1

Neurospora crassa Neurospora crassa Neurospora crassa Neurospora crassa Neurospora crassa Neurospora crassa Neurospora crassa Pyrenopeziza brassicae

3

+

+

gi|26324430 gi|20891621 gi|38082026

Sept11 Sept12

Cdc10 mKIAA0202 Sint1, E-septin, SLP-a

D5Ertd606e 4933413B09Rik 1700016K13Rik

F Cdc11

pcd1

F

APPENDIX A

348

SEPTIN AND SEPTIN-LIKE SEQUENCES

GIa

Name in Chapter 3b

Gene Aliasd symbolc

gi|34859284 gi|16924010 gi|9507085 gi|32423788

RnoSept1 RnoSept2 RnoSept3 RnoSept4

/ / / /

gi|16758814

RnoSept5

/

gi|34932994 gi|12018296 gi|34870727 gi|13929200 gi|34882181 gi|34872099 gi|34876531 gi|34868752 gi|6323346

RnoSept6 RnoSept7 RnoSept8 RnoSept9 RnoSept10a RnoSept10b RnoSept11 RnoSept12 SceCdc3

/ / / / / / / / /

gi|6319847

SceCdc10

/

F

gi|6322536

SceCdc11

/

F

gi|6321899

SceCdc12

/

F

gi|6319976

SceShs1

/

gi|6320424

SceSpr28

/

F

gi|6321496

SceSpr3

/

F

gi|19115666

SpoSpn1

/

F

gi|19114071

SpoSpn2

/

F

gi|13638491

SpoSpn3

/

F

gi|19114478

SpoSpn4

/

F

gi|19114952

SpoSpn5

/

F

gi|19075714

SpoSpn6

/

SPCC584.09

F

gi|15214304

SpoSpn7

/

SPBC 19F8.01c

F

gi|20177379

SdoSeptl

/

A

gi|33302067 gi|46099680 gi|46099269 gi|46099354

UmaCdc10 UmaHyp1 UmaHyp2 UmaHyp3

/ Cdc3 Cdc12 Cdc11

F F F F

LOC293507 G-septin LOC287606, EG3RVC, EG3-1RVC Gp1bb, CDCrel-1, PNUTL1ai, CDCrel-1AI LOC298316 Cdc10 LOC303135 Slpa, E-Septin LOC309891 LOC288622 LOC305227 LOC363542

Sep7

Sep3

Clade/speciese

Groupf

Gg

Ch

A A A A

Rattus norvegicus Rattus norvegicus Rattus norvegicus Rattus norvegicus

2B 2B 1A 2B

+ + + +

+ + − +

A

Rattus norvegicus 2B

+

+

A A A A A A A A F

Rattus norvegicus Rattus norvegicus Rattus norvegicus Rattus norvegicus Rattus norvegicus Rattus norvegicus Rattus norvegicus Rattus norvegicus Saccharomyces cerevisiae Saccharomyces cerevisiae Saccharomyces cerevisiae Saccharomyces cerevisiae Saccharomyces cerevisiae Saccharomyces cerevisiae Saccharomyces cerevisiae Schizosaccharomyces pombe Schizosaccharomyces pombe Schizosaccharomyces pombe Schizosaccharomyces pombe Schizosaccharomyces pombe Schizosaccharomyces pombe Schizosaccharomyces pombe Suberites domuncula Ustilago maydis Ustilago maydis Ustilago maydis Ustilago maydis

1B 2B 1B 1A 1B 1B 1B 1A 2A

+ + + + + + + + +

+ + + − + + + − +

1A

+

−

3

+

+

4

+

+

3

+

+

3

+

+

4

+

+

2A

+

+

1A

+

−

3

+

+

4

+

+

3

+

+

4

+

+

3

+

−

2B

+

+

1A 2A 4 3

+ + + +

− + + +

F

APPENDIX A SEPTIN AND SEPTIN-LIKE SEQUENCES GIa

Name in Chapter 3b

Gene Aliasd symbolc

gi|34784614 gi|12003372 gi|50551445

XlaHyp1 XlaSeptA YliHyp1

Sept12 Sept2 Cdc3

gi|50549207

YliHyp2

Cdc10

F

gi|50551749

YliHyp3

Cdc12

F

gi|50553330

YliHyp4

Cdc11a

F

gi|50549013

YliHyp5

Spr28

F

gi|50547965

YliHyp6

Cdc11b

F

gi|50557032

YliHyp7

Spr3

F

Gi|13940377

ZroCDC

/

a Genbank

MGC68931

er001-c

A A F

F

349

Clade/speciese

Groupf

Gg

Ch

Xenopus laevis Xenopus laevis Yarrowia lipolytica Yarrowia lipolytica Yarrowia lipolytica Yarrowia lipolytica Yarrowia lipolytica Yarrowia lipolytica Yarrowia lipolytica Zygosaccharomyces rouxii

1A 2B 2A

+ + +

− + +

1A

+

−

4

+

+

3

+

+

3

+

+

3

+

+

4

+

+

tr

+

−

identification numbers. first three letters represent genus and species names. The last letters represent current septin protein name. c Proposed names based on first- or best-characterized septin in each clade. d Alias designations from Genbank. e A represents animals; F represents fungi; M represents microsporidia. f Group names are assigned according to phylogenetic analysis shown in Figure 2.3 (Chapter 2). tr, truncated; slk, septin-like. g Presence of full length GTP CDC detected by the SMART program. h Predicted coiled-coil at C terminus. Adapted from Pan, F. et al. 2007. b The

Appendix B Mammalian septin nomenclature Peter A. Hall Institute of Pathology, School of Medicine, Queen’s University Belfast, Belfast BT12 6BL, Northern Ireland, UK

Elspeth Bruford Human Genome Nomenclature Committee, European Bioinformatics Institute, Wellcome Trust Genome Campus, Hinxton, Cambridge CB10 1SD, UK

S.E. Hilary Russell Ovarian Cancer Research Laboratory, Centre for Cancer Research and Cell Biology, Queen’s University Belfast, A floor, Belfast City Hospital, Lisburn Road, Belfast BT9 7AB, Northern Ireland, UK

Ian G. Macara Department of Microbiology, University of Virginia School of Medicine, Charlottesville, VA 22908-0577, USA

John R. Pringle Department of Genetics, MC 5120, 300 Pasteur Drive, M-322 Alway Building, Stanford University Medical Center, Stanford, CA 94305-5120, USA

Effective communication is critical for scientific progress but is often impeded by suboptimal nomenclature that reflects the serendipitous and disseminated nature of discovery. As genes have been discovered in diverse organisms by groups working on different aspects of biology and with different favoured perspectives, so the names used for them in publications have often been idiosyncratic if not frankly confusing or misleading. In the area of the mammalian septin genes and their products, a Babel-like state existed until Macara et al. (2002) brought order to what was chaos! The diversity of published names (see Table B.1 and Hall and Russell, 2004) was such that considerable confusion existed and unambiguous communication was often difficult. The proposals in Macara et al. (2002) were based upon the rules of the Human Genome Nomenclature Committee (HGNC; see http://www.genenames.org/) and Mouse Genomic Nomenclature Committee (MGNC, see http://www.informatics. The Septins Edited by Peter A. Hall, S.E. Hilary Russell and John R. Pringle © 2008 John Wiley & Sons, Ltd. ISBN: 978-0-470-51969-1

352

APPENDIX B MAMMALIAN SEPTIN NOMENCLATURE

Table B.1 The human septin genes Approved gene symbol

Approved gene name

Location

Sequence accession IDs

Aliases

SEPT1

Septin 1

16p11.1

PNUTL3, DIFF6

SEPT2

septin 2

2q37.3

AF308288 NM 052838 D28540

SEPT3

septin 3

22q13.2

SEPT4

septin 4

17q23

SEPT5

septin 5

22q11.2

SEPT6

septin 6

Xq24

SEPT7

septin 7

7p14.2

SEPT8

septin 8

5q31

SEPT9

septin 9

17q25

SEPT10

septin 10

2q13

SEPT11

septin 11

4q21

SEPT12

septin 12

16p13.3

SEPT13

septin 13

7p13

SEPT14

septin 14

7p11.2

NM 006155 AF285107 NM 145734 AF073312

NM 080417 Y11593 NM 002688 D50918 NM 145802 S72008 NM 001788 AF179995 XM 034872 AB023208

NM 006640 AF146760 NM 144710 AK001711 NM 018243 AK058139 NM 144605 AL133216 AK126048 NM 207366

KIAA0158, hNedd5, Pnutl3 DIFF6, NEDD5 SEP3 H5, CE5B3, hucep-7, bradeion, ARTS, hCDCREL-2, MART, PNUTL2 HCDCREL-1, H5, PNUTL1 KIAA0128, SEP2, SEPT2, MGC16619, MGC20339 CDC3, SEPT7A, CDC10 KIAA0202, SEP2 MSF1, KIAA0991, PNUTL4, AF17q25, SeptD1, OvBr septin, MSF FLJ11619 FLJ10849 FLJ25410 DKFZp313J1114, SEPT7B,SEPT7P FLJ44060

jax.org/mgihome/nomen/). At the 2nd Septin Workshop (2007), a satellite session discussed aspects of septin nomenclature and made several clarifications and additions to the 2002 proposals, again following the HGNC guidelines. The issues requiring attention included a clarification of how to designate transcripts and the protein isoforms they encode, the terminology for promoters, and the terminology for pseudogenes. As originally proposed in 2002 and approved by the HGNC, the human septin genes are denoted as SEPTX (NB: italicized), where X is the number of the septin gene, and the corresponding proteins are designated by the same symbol

APPENDIX B

MAMMALIAN SEPTIN NOMENCLATURE

353

but nonitalicized. Although the 2002 paper did not attempt to distinguish between splice-variant transcripts and splice-variant proteins, the discussants at the 2nd Septin Workshop agreed that this would be useful. Thus, they proposed that although transcripts would continue to be denoted as SEPTX vY (where v indicates a variant and Y is the numeric designation of that variant), protein isoforms would be designated similarly except nonitalicized and with an ‘i’ (for isoform; as recommended by the Human Genome Variation Society, http://www. hgvs.org) instead of the ‘v’. In general, the transcript(s) encoding the longest isoform would be given the number ‘1’, the next longest ‘2’, and so on, and the protein isoform numbers would correspond to those of the transcripts encoding them. Thus, using the example of human septin 9, the gene symbol is SEPT9 ; the transcripts include SEPT9 v1 , SEPT9 v2 , and so on; and the protein products of these transcripts are SEPT9-i1, SEPT9-i2, and so on. If more than one transcript encodes the same polypeptide (as is seen with SEPT6, SEPT8 and SEPT9 for example, see Chapter 7 by Russell), then the different transcripts are denoted by an asterisk. Thus SEPT9 v4 and SEPT9 v4* are distinct transcripts that encode the same polypeptide SEPT9 i4. If more than two transcripts encode the same polypeptide (as has been seen with SEPT6 : see Chapter 7) then additional asterisk are added (e.g. SEPT6 v4 , SEPT6 v4*, SEPT6 v4**, SEPT6 v4*** etc.). The need for consistent use of this systematic approach is emphasized by the extent and diversity of splicing events actually found in mammalian septin genes, as well as by the use of specialized names such as ‘ARTS’, which denotes one specific isoform encoded by (in this case) SEPT4 (see the full list of aliases in TableB.1). Although the proper systematic name of this protein is SEPT4 i2, it may in some contexts facilitate understanding by referring to it as ‘SEPT4 i2 (ARTS)’. Some septin genes also use multiple promoters. In such cases, the promoters should be designated in a similar way to splice variants, but using the lower case letters ‘pr’ (e.g. SEPTx pr1 ; note the italics appropriate for the genotypic symbol). Possible pseudogenes have also been suggested to exist for mammalian septins. These should be denoted on a case-by-case basis and the designation approved by the HGNC prior to publication. Usually this will involve assigning the gene symbol of the structural or functional gene followed by ‘P’ (or ‘ps’ in rodents) and a number. The HGNC currently recommended in their nomenclature guidelines (http://www.genenames.org/guidelines.html) that potential pseudogenes identified from sequence information should be assigned such a symbol if they show a minimum of 50 % predicted amino-acid identity, across 50 % of the open reading frame, to the named gene. It should be noted that, despite the fact that expression data exist, human SEPT13 may actually represent a transcribed pseudogene of SEPT7 . At present we do not have sufficient information to be certain and thus would propose it remain as SEPT13 with the alias SEPT7P until we can finally resolve this matter. (In favour of this being a transcribed pseudo gene is the failure to date of identifying a murine orthologue.) Note that as the numbering

APPENDIX B MAMMALIAN SEPTIN NOMENCLATURE

354

Table B.2 The mouse septin genes Symbol, name

Chr

Sept1 , septin 1 Sept2 , septin 2 Sept3 , septin 3 Sept4 , septin 4 Sept5 , septin 5 Sept6 , septin 6 Sept7 , septin 7 Sept8 , septin 8 Sept9 , septin 9 Sept10 , septin 10 Sept11 , septin 11 Sept12 , septin 12 Sept14 , septin 14

7 1 15 11 16 X 9 11 11 10 5 16 5

a Note

cM syntenic

11.42 syntenic 28.75 79.2 syntenic 52 syntenic

Genome coordinates 127005595-127009646 (−) 95309370-95340666 (+) 82102657-82119825 (+) 87385098-87406734 (+) 18535374-18543501 (−) 33342771-33423239 (−) 25002201-25058011 (+) 53363169-53387518 (+) 117015751-117178415 (+) 58536984-58617204 (−) 94168655-94251641 (+) 4902179-4913173 (−) 129982979-130013770 (−)

that a mouse Sept13, orthologous to human SEPT13, has not yet been identified.

of pseudogenes will usually be species-specific, orthology relationships cannot usually be inferred from the numbering of pseudogenes in other species, The currently known human septin genes with their chromosomal locations, representative mRNA accession numbers, and diverse aliases are shown in Table B.1. Corresponding information for the mouse genes are shown in Table B.2; note that the numbering of the mouse genes corresponds to that of their human orthologues. Essentially the same nomenclature guidelines apply (see http://www.informatics. jax.org/mgihome/nomen/gene.shtml). The only significant difference from the HGNC’s human nomenclature guidelines is in the use of case in gene names: human gene names are all upper case, whereas for mouse only the first letter is upper case (i.e. SEPT9 vs. Sept9 ). In closing, we note that the septin field is indeed fortunate that the importance of systematic nomenclature was recognized, and dealt with by effective community action, at an early stage in the development of the field. Future workers will have many occasions to be grateful for the simple and rational nomenclature embodied in Tables B.1 and B.2. It is to be hoped that parallel nomenclatures will be employed for other vertebrate species including important model systems such as the zebrafish Danio rerio. Finally, it will be the responsibility of all, but especially journal editors and reviewers, to ensure the widespread use of these simple rules.

REFERENCES Hall, P.A. and Russell, S.E. (2004) The pathobiology of the septin gene family. The Journal of Pathology, 204, 489–505. Macara, I.G., Baldarelli, R., Field, C.M. et al. (2002) Mammalian septins nomenclature. Molecular Biology of the Cell , 13, 4111–13.

Appendix C Septin meetings and workshops Peter A. Hall Institute of Pathology, School of Medicine, Queen’s University Belfast, Belfast BT12 6BL, Northern Ireland, UK

John R. Pringle Department of Genetics, MC 5120, 300 Pasteur Drive, M-322 Alway Building, Stanford University Medical Center, Stanford, CA 94305-5120, USA

It can be argued that in the development of any field, a key watershed is reached when there are enough scientists interested to warrant the organization of meetings! In the septin field, this has occurred only in the past few years. A session devoted to septins was held in association with the 41st ASCB Annual Meeting in Washington, DC, in December 2001. This pre-meeting ‘Special Interest Group’ session was organised by Christine Field and John Pringle, and the excerpt from the ASCB Programme is shown in Box 1 (with the permission of ASCB).

Box 1 The American Society for Cell Biology 41st Annual Meeting 8–12 December 2001, Washington, DC, [email protected], www.ascb.org Saturday, 8 December 2001 Septin Structure and Function Organizers Christine M. Field , Harvard Medical School and John Pringle, University of North Carolina, Chapel Hill Septins are a family of conserved proteins that have been implicated in a variety of cellular functions involving specialized regions of the cell cortex, such as cytokinesis, cell shape change, and vesicle fusion. Septins form heteromeric complexes that bind and hydrolyze GTP, polymerize in vitro, form The Septins Edited by Peter A. Hall, S.E. Hilary Russell and John R. Pringle © 2008 John Wiley & Sons, Ltd. ISBN: 978-0-470-51969-1

356

APPENDIX C

SEPTIN MEETINGS AND WORKSHOPS

neck filaments in vivo (yeast), and bind phophoinositide lipids. Septins are implicated in regulating contractility, exocytosis, membrane compartmentation, and apoptosis, but their precise molecular functions are not known, and there are many basic questions to explore. This session brings together researchers working on a variety of systems to discuss progress in understanding molecular architecture, biochemical activities, and cellular functions. Speakers John Pringle, University of North Carolina, Chapel Hill. Septin history (their discovery) and a mystery (their role in fission yeast cytokinesis) Michelle Momany, University of Georgia. Septins in Aspergillus nidulans are involved in septation, branching and asexual development. Mark Longtine, Oklahoma State University. Septin structure and function in S. cerevisiae Erica Johnson, Thomas Jefferson University. An E3-like factor that promotes SUMO conjugation to the yeast septins Erfei Bi , University of Pennsylvania Medical School. The Role of Cdc42p GAPs in septin organization in yeast Ian Macara, University of Virginia. Borgs Effectors of Cdc42 that control septin organization Makoto Kinoshita, Harvard Medical School. Biochemical and ultrastructural analysis of purified and reconstituted mammalian septin complexes William Trimble, Hospital for Sick Children, Toronto. Functional analysis of mammalian septins. Although the ASCB Programme Committee has yet to acknowledge the septin field with a dedicated session in the main programme, the investigators in the field did not sit around waiting for this (or similar societal recognition) to occur. Instead, a first full meeting devoted to septins was organized and held in Denmark in May 2005. Thirty-five scientists (Figure C.1) met at the Fuglsøcentret Conference Centre, Nr Aarhus, Denmark. The idea for this workshop initially came from Hilary Russell and Peter Hall (Belfast) and Finn Skou Pedersen and Ernst-Martin F¨uchtbauer (Aarhus) and who also managed to obtain significant sponsorship from Aarhus University, the Pathological Society, Genentech, and other sources. The meeting was highly successful and stimulated discussion, debate, and new collaborations. Because no other account of this meeting exists, we reproduce its programme here (see Box 2); because of space limitations, we cite only the actual presenter although many abstracts had additional authors.

Figure C.1

Participants at the First International Septin Workshop, May 2005

APPENDIX C SEPTIN MEETINGS AND WORKSHOPS 357

APPENDIX C

358

SEPTIN MEETINGS AND WORKSHOPS

Box 2 Programme of the First Septin Workshop Session 1 Birth and childhood of the septins, John Pringle Expression profiling of the human septin gene family, Peter Hall Session 2 – Assembly of Septin Structures Molecular architecture of septin complexes and regulation of septin assembly, Jeremy Thorner Septin GTP binding regulates septin-septin interactions in S. cerevisiae, Mark Longtine Global analysis of protein phosphorylation in yeast, Jason Ptacek (Michael Snyder lab) A direct role of a Cdc42p effector pathway in the recruitment of the septins to a discrete site in cell cortex, Erfei Bi Differential genetic control of septin rings and basal septin bands in hyphae of the human fungal pathogen Candida albicans, Peter Sudbery. Platelet septin-septin interaction, Barbara Zieger. Session 3 – Focus on SEPT9 and its Interactors Possible role of Rho/Rhotekin signaling in mammalian septin organization, Koh-ichi Nagata Septins activate Rho kinase to mediate myosin light chain phosphorylation, William Trimble Septin9 is required for embryonic development in mice, Ernst-Martin F¨uchtbauer The role of Septin 9 as an oncogene/tumor-suppressor gene in lymphomagenesis by murine leukemia virus, Finn Skou Pedersen Genomic amplification and overexpression of the Septin 9 oncogene occurs in mouse models for breast cancer, Cristina Montagna (Thomas Ried lab) Functional analysis of human septins: lessons from SEPT9, Hilary Russell Mammalian septins regulate microtubule stability through interaction with the microtubule binding protein, MAP4, Ian Macara Session 4 – Diversity of Septin Function Septins control vacuolar escape of intracytosolic bacteria St´ephanie BoissonDupuis (Pascale Cossart lab)

REFERENCES

359

The mitochondrial protein ARTS promotes apoptosis through binding and antagonizing XIAP, and functions as a tumor suppressor in acute lymphoblastic leukemia (ALL), Sarit Larisch. ‘The Lord of the Rings: The Return of the Ring’, Makoto Kinoshita Platelet septins and their role in hemostasis and thrombosis, Jerry Ware Functions of the C. elegans septins in axonal migration and organogenesis, Fern Finger Session 5 – Septins and Cell Division Distinct roles for two C. elegans anillins in the gonad and the early embryo, Karen Oegema Septins in cytokinesis, Christine Field The xeptins and the actomyosin ring in cytokinesis, John Pringle Septin function is a key downstream target of the ‘NoCut’ pathway involved in the coordination of cytokinesis with chromosome segregation, Yves Barral A mitotic Sept2 scaffold required for mammalian chromosome congression and segregation, Elias T. Spiliotis It was also agreed in Denmark that a second septin workshop would be held in May 2007, and Yves Barral (Z¨urich) and Christine Field (Boston) took on the mantle. The venue for ‘The Molecular Biology and Biochemistry of Septins and Septin Function’ was the Monte Verit`a Conference Centre near Ascona in Switzerland. Like the first meeting, the second workshop was enthusiastically supported and attended, but it had now more than doubled in size to 78 attendees (Figure C.2). Sponsorship came from many sources, and the Workshop was designated an EMBO Meeting. A summary account of this meeting has been published (Gladfelter and Montagna, 2007) The success of this meeting indicated that the biannual cycle of Workshops should continue, and planning for a third workshop is underway.

REFERENCE Gladfelter, A.S. and Montagna, C. (2007) Seeking truth on Monte Verita. Workshop on the molecular biology and biochemistry of septins and septin function. EMBO Reports, 8, 1120–26.

APPENDIX C

Figure C.2 Participants at the Second International Septin Workshop, May 2007

360 SEPTIN MEETINGS AND WORKSHOPS

Index α-synuclein 259, 326–7 actin-associated proteins 214–17 actins crosstalk regulation 234–7 mammalian septins 192, 193, 195–6, 229–37 metazoan model systems 155, 156, 159 septin interactions 230–4 actomyosin 109, 128–9 acute lymphoblastic leukaemia (ALL) 286–8, 303 acute myeloid leukaemia (AML) 26, 171–2, 320, 322, 329 ALL see acute lymphoblastic leukaemia Alzheimer’s disease 26, 241, 248, 258–9, 297, 305–6 AML see acute myeloid leukaemia anillins Caenorhabditis elegans 161–3 conservation of septin function 109 cytoskeleton 232 Drosophila melanogaster 155, 159–62 mammalian septins 214, 215 annulus 192, 199, 239, 327–9 antibodies historical development 11–13, 15–16, 18–19 supramolecular architecture 54–5, 65–6, 68 apoptosis 3 human septins 303 intron retention 284 mammalian septins 193, 200, 216, 240, 281–93

mechanism 282–3, 285–7 SEPT4 284–9 septin function 284–7 signalling pathways 282–3 tumour suppressor proteins 287 apoptosome complexes 283 Ashbya gossypii 3, 126, 137–40 Aspergillus nidulans historical development 3, 22 P loop GTPases 37, 38 septin organization and function 126, 140–1 astrocytes 256–7 bacterial septins 55–7 baculovirus repeat domains (BIR) 283 basal septin band 131, 133 Bcl-2 gene 283 bem3 135–6 Bernard-Soulier syndrome 270 bipolar disorder 298 BIR see baculovirus repeat domains BLAST 172 blood coagulation 270 Bni 69–71 Borg proteins 27, 214, 216 Bradeion alpha/beta 284–5 brain tumours 297 breast epithelial cancers 296, 301 Bud 106, 109–10, 114, 129 bud bite selection 109–10 bud morphogenesis 110–11 budding yeast see Saccharomyces cerevisiae

The Septins Edited by Peter A. Hall, S.E. Hilary Russell and John R. Pringle © 2008 John Wiley & Sons, Ltd. ISBN: 978-0-470-51969-1

362

C-terminal extensions (CTEs) 53–4, 63–4, 72–6, 91–2 cadherins 221 Caenorhabditis elegans 147–68 coiled coil-forming domains 253 conservation of septin functions 108 cytokinesis 150–1, 156–60 gene and protein families 148–51 historical development 2–3, 27 neuronal migration 161–2 P-loop GTPases 37 septin-interacting proteins 215 supramolecular architecture 62–4, 72–3 synapse 253, 256 two-septin complex 154–5 cancer progression 240–1 Candida albicans historical development 3, 22 human septins 307 septin organization and function 125–6, 130–6 cap formation 102, 104, 133 caspases 282 Ccn1p-Cdc28p 133–5 cdc see cell division cycle CDKs see cyclin-dependent kinases cDNA see complementary DNA cell division 105–9 cell division cycle (cdc) genes cortical organization 101–2, 109–13 fungal septins 126–7, 133–41 historical development 8–14, 19, 23–5, 27 mammalian septins 189, 201, 216 supramolecular architecture 50–3, 55–64, 69–76, 79–87, 91–2 cell morphogenesis fungal septins 126, 141 Saccharomyces cerevisiae 109–15 cell polarity 113–14 cell separation 136 cellular asymmetry 68–72 cellular compartmentalization 114–15, 239 cellularization 160–1

INDEX

central nervous system (CNS) 3, 157, 161–4 see also synapse chitin 126, 142 Cla4 23, 80, 87 CNS see central nervous system coiled coil-forming domains 42 fungal septins 125, 127, 137 invertebrates 63–4 mammalian septins 72–6, 188 Saccharomyces cerevisiae 101 synapse 252–3 collapsing response mediator protein/associated molecules (CRMP/CRAM) 200, 202, 255–6 collar assemblies fungal septins 133, 135–6 Saccharomyces cerevisiae 103, 104, 115 supramolecular architecture 69–72, 74, 90–1 colorectal cancer 297 compartmentalization 114–15, 239 complementary DNA (cDNA) 175, 181 conserved single residues 43 coordination of cellular events 200–3 cortical organization 101–23 COS-7 cells 275 CRMP/CRAM see collapsing response mediator protein/associated molecules crosstalk regulation 234–7 CTEs see C-terminal extensions cyclin-dependent kinases (CDKs) 111–12, 197, 219, 258 cyclins 110 cyk3 107–8 cytokinesis 2, 3 Caenorhabditis elegans 150–1, 156–60 Drosophila melanogaster 149–51, 156–60 fungal septins 129, 132, 134, 137 historical development 8–10, 15–16, 21, 26

INDEX

mammalian septins 198, 201, 202–3, 215, 269 platelets 269 Saccharomyces cerevisiae 105–9, 116 supramolecular architecture 90 cytoskeleton 229–46 crosstalk regulation 234–7 human diseases 239–41 intermediate filaments 229, 231 membrane skeleton 237–9 septin interactions 230–4 septin localization 231–2 see also actins; microtubules Dictyostelium discoidium 36 Diff6 19, 199 diffusion barriers 193, 198–9 DiGeorge syndrome 197, 260 DNA repair 193, 199–200 dopamine 251, 324–5 double ring assemblies 128–9, 198 Down’s syndrome 260, 297–8, 306 Drosophila melanogaster 147–68 actin and microtubule skeletons 155 cellularization 160–1 cytokinesis 149–51, 156–60 exocytosis 198, 202 filament formation 153–4, 160 gene and protein families 148–51 historical development 3, 17, 19–23, 26–7 neural function 162–4 platelet septins 271 septin-interacting proteins 215 supramolecular architecture 64, 68, 88 three-septin complex 152–4, 155, 160, 162 two-hybrid method 156 effector caspases 282 endocytosis 247 endothelial reticulum (ER) 106–7, 114–15 Entamoeba histolytica 36

363

epilepsy 248, 260, 298 epithelial cancers 296, 301–2 ER see endothelial reticulum Escherichia coli 50–1, 55 EST see expressed sequence tags eukaryotes 36–8, 50–62, 88–9 evolution of septins 35–41 classification 38–40 eukaryotes 36–8 historical development 27 P loop GTPases 36–41 phylogenetic analysis 35–40 septin domains 41–3 exocytosis fungal septins 134–5 human septins 329 in vitro models 329 mammalian septins 197–8, 202–3, 219 synapse 253 expressed sequence tags (EST) 181 expression profiling 180–2 extrinsic signalling pathway 282–3 F-actin 230–5 filament formation Drosophila melanogaster 153–4, 160 fungal septins 130, 137–41 invertebrates 62–4 mammalian septins 190–1, 214, 252–4 Saccharomyces cerevisiae 102–3, 104–5, 114 synapse 252–4 unicellular eukaryotes 50–1, 55–7, 59 vertebrates 68–72, 74–5, 77–84, 88–92 First Septin Workshop 1 fission yeast see Schizosaccharomyces pombe Fluorescence Recovery After Photobleaching (FRAP) fungal septins 136 mammalian septins 230–4

364

Fluorescence Recovery After Photobleaching (FRAP) (continued ) Saccharomyces cerevisiae 103–5, 113 supramolecular architecture 69 four-helix bundles 74 FRAP see Fluorescence Recovery After Photobleaching fruit fly see Drosophila melanogaster fungal septins 125–45 Ashbya gossypii 126, 137–40 Aspergillus nidulans 126, 140–1 Candida albicans 125–6, 130–6 cap formation 133 cell separation 136 collar assemblies 133, 135–6 cytokinesis 129, 132, 134, 137 evolution 35–41 exocytosis 134–5 filament formation 130, 137–41 historical development 15–18 hyphae 126, 130–4, 136, 137–41 mitosis 126, 139–40, 142 pseudohyphae 126, 130–3, 136 ring assemblies 128–9, 133, 136, 138–9, 141–2 Schizosaccharomyces pombe 125–30 GAPs see guanine triphosphatase activating proteins GDP see guanine nucleotides GEFs see guanine nucleotide exchange factors gene transcription 179–80 Giardia lambdia 36 gin4 115, 133–6 glutamate-aspartate transporter (GLAST) 199, 220 glycoproteins (GP) 270–3 GTP see guanine nucleotides GTP-binding domains apoptosis 281, 288 human septins 300 platelets 275–6 septin-interacting proteins 213

INDEX

synapse 253 GTPase domains 36–8, 41 guanine nucleotide exchange factors (GEFs) 109, 195–6, 215, 233, 235 guanine nucleotides apoptosis 284 Drosophila melanogaster 152–3 mammalian septins 199 supramolecular architecture 63, 67, 76–80 guanine triphosphatase activating proteins (GAPs) 215 haemostasis 26 HCV see hepatitis C virus HeLa cells 189–90, 214, 217, 235–6 hemostasis 3 hepatitis C virus (HCV) 221, 298 hereditary neuralgic amyotrophy (HNA) 178, 295, 296, 299, 323 hetero-oligomeric complexes invertebrates 62–4 mammalian septins 188–92, 211–28, 273–6 platelets 273–6 unicellular eukaryotes 50–62 vertebrates 65–8, 76–83 heterohexameric complexes 64, 67, 77 heterotetrameric complexes 63, 74 hgc1 136 HGH see human growth hormone HHV-8 see human herpes virus 8 HIF1α see hypoxia induced factor 1 alpha highly conserved residues 43 historical development 7–34 antibodies 11–13, 15–16, 18–19 cdc mutants 8–14, 19, 23–5, 27 cytokinesis 8–10, 15–16, 21, 26 fungal septins 15–18 insect septins 17–23 mammalian septins 17–19, 25–8 neck filament loss 10–11 nomenclature and taxonomy 27 publications 1–2, 14–15, 17–21, 23–5

INDEX

scaffold model 24, 25 ts mutants 7–8, 9 yeast septins 2–3, 7, 15–18, 22–3 HNA see hereditary neuralgic amyotrophy hof1 107–8 Hsl1 113, 115 human breast adenocarcinoma 202 human growth hormone (HGH) 219 human herpes virus 8 (HHV-8) 298 human septins 3, 295–317 apoptosis 303 cytoskeleton 239–41 disease correlation 295–9 evolution 37, 38–41 exocytosis 329 expression profiling 180–2 gene transcription control 179–80 genomics and regulation 171–85 historical development 26 in vitro models 319–36 infectious diseasess 298, 306–7 neoplasia 296, 300–4 neurology 297–8, 304–6 oncogenes 302 research models 308–9 SEPT1 176–7, 180, 303, 305 SEPT2 181–2, 303, 305, 307 SEPT3 323–4 SEPT4 303–6, 324–9, 330 SEPT5 177, 180, 304–5, 329–30 SEPT6 175–7, 179, 303, 308, 320–1 SEPT7 321–2, 330–1 SEPT9 171–4, 177–80, 181, 299–302, 308, 322–3 septin function 319–26 supramolecular architecture 65–8, 74–5, 77, 79, 88 tumour suppressor genes 302, 322 5
-untranslated region 172–4, 176, 177–9 hyphae 126, 130–4, 136, 137–41 hypoxia induced factor 1 alpha (HIF1α) 301

365

IAPs see initiators of apoptosis proteins immunogold labelling 254 immunolocalization 276–7 in vitro models 319–36 exocytosis 329 SEPT3 323–4 SEPT4 324–9, 330 SEPT5 329–30 SEPT6 320–1 SEPT7 321–2, 330–1 SEPT9 322–3 tissue-specific septins 321, 323–30 ubiquitous septins 320–3 infectious diseases 220–2, 241, 298, 306–7 initiator caspases 282 initiators of apoptosis proteins (IAPs) 283, 285–7, 289 innocent bystanders 18 insect septins 17–23, 26–7 intermediate filaments 229, 231 internalins 221 intrinsic signalling pathway 282 intron retention 284 Kaposi’s sarcoma-associated herpesvirus (KSHV) 221–2, 288, 307 Kcc 69–71 kinesins 36–8, 194 KSHV see Kaposi’s sarcoma-associated herpesvirus Latrunculin A 254 linear rod complexes 51–60, 65–6, 77–80, 88–92 lipid binding 191 lipophilia 87–9 Listeria monocytogenes 220–1, 241, 298, 306–7, 322 LJ-33 274–5 loss of heterozygosity (LOH) 302 mammalian septins actin-associated proteins 214–17 actins 192, 193, 195–6, 229–37

366

mammalian septins (continued ) apoptosis 193, 200, 216, 240, 281–93 cancer progression 240–1 coordination of cellular events 200–3 crosstalk regulation 234–7 cytokinesis 198, 201, 202–3, 215, 269 cytoskeleton 229–46 diffusion barriers 193, 198–9 distribution 248–50, 251 DNA repair 193, 199–200 evolution 35–41 exocytosis 197–8, 202–3, 219 filament formation 190–1, 214, 252–4 functions 187–209, 229–46, 254–7, 284–7 hetero-oligomeric complexes 188–92, 211–28, 273–6 historical development 17–19, 25–8 human diseases 239–41 immunolocalization 276–7 infectious diseases 220–2, 241 intermediate filaments 229, 231 lipid binding 191 localization 250–1 mechanical stability 192–3 membrane skeleton 237–9 membrane trafficking 193, 196–8 microtubules 193, 195–6, 217–18, 229–37 neurology 220, 239–40, 241, 258–60 organization 188–92 phagocytosis 201, 203 phosphorylation of septins 257–8 platelets 269–80 polarization 201–2 protein regulation 193, 199 regulation of septin expression 252 ring assemblies 191–2, 198, 276–7 scaffold platforms 193, 194, 235–6 secretion regulation 218–19 SEPT4 284–9 SEPT5 272–7

INDEX

septin-interacting proteins 211–28 septin–septin complexes 211–14 supramolecular architecture 65–93 synapse 247–67 two-hybrid method 215–16 viral replication 193, 199 see also human septins; in vitro models MAP4 see microtubule-associated protein mechanical stability 192–3 megakaryocytes 270 melanoma 296 membrane skeleton 237–9 membrane trafficking 193, 196–8 membrane-associated morphogenesis 59–60 MEN see mitotic exit network messenger RNA (mRNA) 171, 172–3, 177–9, 182 microsporidia 36, 40 microtubule-associated protein (MAP4) 217–18 microtubules crosstalk regulation 234–7 mammalian septins 193, 195–6, 229–37 metazoan model systems 155 platelets 276–7 septin interactions 230–4 septin-interacting proteins 217–18 Mid2p 129 Mih1 111–13 mitosis, fungal septins 126, 139–40, 142 mitotic exit network (MEN) 107 mixed lineage leukaemia (MLL) 175, 295, 296, 300–1, 320, 329 MLC see myosin light chain MLL see mixed lineage leukaemia model systems 2 morphogenetic checkpoint 111–13 motor performance 256 mRNA see messenger RNA myelodysplastic syndrome 322 myo1 107–8

INDEX

myosin light chain (MLC) 195–6 myosins 36–8, 156, 159 N -ethylmaleimide-sensitive fusion protein (NSF) 197, 238 NCK 216–17 Nedd5 19, 26 nematode worms see Caenorhabditis elegans neoplasia 296, 300–4 nerve growth factors (NGF) 202, 219 nerve terminals 251 neural function 162–4 neurite outgrowth 255–6 neurofibrillary tangles (NFTs) 257, 258 neurology cytoskeleton 239–40, 241 human septins 297–8, 304–6 septin-interacting proteins 220 synapse 258–60 see also synapse neurons 161–2, 250–1 NFTs see neurofibrillary tangles NGF see nerve growth factors NLS see nuclear localization signals nonsense-mediated mRNA decay (NMD) 177, 285 Northern blotting 172, 180–1, 249 NSF see N -ethylmaleimide-sensitive fusion protein nuclear localization signals (NLS) 300 octameric complexes 189–90 oncogenes 302 open reading frames (ORFs) 173–4, 178, 288 oral cancer 297 ORFs see open reading frames ovarian epithelial cancers 296, 301–2 P-loop GTP-binding domains apoptosis 281, 288 Drosophila melanogaster 152 supramolecular architecture 89 P-loop GTPases 36–41 p21-activated kinase (PAK1) 235

367

p53 180, 260, 304 PAK1 see p21-activated kinase Parkinson’s disease 220, 239–40, 248, 259–60, 297, 304–5, 324–7 phosphoinositides cortical organization 114 cytoskeleton 237–9 fungal septins 137 mammalian septins 191, 203 supramolecular architecture 88–9 synapse 253 phosphorylation of septins 257–8, 327–9 phylogenetic analysis 35–40 PKs see protein kinases plants 38 plasticity 248 platelets 269–80 cytokinesis 269 hetero-oligomeric complexes 273–6 immunolocalization 276–7 microtubules 276–7 protoypic septin 270–1 ring assemblies 276–7 SEPT5 272–7 pnut Drosophila melanogaster 148–9, 156–8, 160, 162 mutation 19–21 platelets 271 supramolecular architecture 64, 88 polarization 201–2 polarizomes 113–14 polybasic regions 41–2 post-translational modifications 80–7 postsynaptic density (PSD) 251 postsynaptic membranes 247 presynaptic membranes 247 primary sequence similarity 35–6 protein kinases (PKs) 257–8 protein regulation 193, 199 protists 38 PSD see postsynaptic density pseudohyphae 126, 130–3, 136 publication figures 1–2 pulmonary hypertension 307

368

rapid amplification of 5-prime cDNA ends (RACE) 172, 176 Ras-like GTPases 67, 101 Ras proteins 36–8 renal cell carcinoma 296 reverse transcriptase PCR (RTPCR) 172, 181 rga2 135–6 Rho kinase (ROCK) 159, 195 Rho-like GTPases cytoskeleton 233, 235 fungal septins 135 mammalian septins 215–16 Saccharomyces cerevisiae 109–11 rhoketin 216 ring assemblies 69, 88–9 fungal septins 128–9, 133, 136, 138–9, 141–2 mammalian septins 191–2, 198, 276–7 platelets 276–7 Saccharomyces cerevisiae 102–5, 106 ring canals 192 ROCK see Rho kinase Saccharomyces cerevisiae bud bite selection 109–10 bud morphogenesis 110–11 cap formation 102, 104 cell division 105–9 cell morphogenesis 109–15 cell polarity 113–14 cellular compartmentalization 114–15 cellular functions of septins 105–15 collar assemblies 103, 104, 115 conservation of septin functions 108–9 cortical organization 101–23 cytokinesis 105–9, 116 exocytosis 134–5 filament formation 102–3, 104–5, 114 historical development 2, 7, 15–17, 22–3

INDEX

molecular functions of septins 115–16 morphogenetic checkpoint 111–13 ring assemblies 102–5, 106 septin dynamics 103–5 septin evolution 36–8, 42 septin structures 102–5 spindle positioning 106–7 supramolecular architecture 50–60, 69, 73, 88–92 SAGE 181 scaffold platforms 74–5 historical development 24, 25 mammalian septins 193, 194, 235–6 schizophrenia 248, 260, 298, 306 Schizosaccharomyces pombe cell morphogenesis 111 historical development 3, 15–18, 22 septin organization and function 125–30 septin-interacting proteins 215 supramolecular architecture 60–2 SDKs see septin-dependent kinases Sec3p 135 Second Septin Workshop 1 secretion regulation 218–19 SEPT2/6/7 complex cytoskeleton 232–3, 235–6 mammalian septins 189–90, 196, 199–200, 213–17 supramolecular architecture 66–8, 88 see also human septins; mammalian septins septation initiation network (SIN) 107 septin domains 41–3 septin dynamics 103–5 septin evolution 35–41 classification 38–40 eukaryotes 36–8 historical development 27 P loop GTPases 36–41 phylogenetic analysis 35–40 septin domains 41–3 septin splice variants 182 septin unique elements (SUE) 42–3, 54, 67, 188

INDEX

septin-dependent kinases (SDKs) 103, 106, 115–16 septin-interacting proteins 211–28 actin-associated proteins 214–17 infectious diseases 220–2 microtubules 217–18 neurology 220 secretion regulation 218–19 septin–septin complexes 211–14 septin–membrane interactions 87–9 septin–septin complexes 211–14 Shs1 58–60, 75, 85, 87, 114 SIN see septation initiation network Sip proteins 156 small ubiquitin-like modifier (SUMO) 194 SNAP25 see soluble N -ethylmaleimide-sensitive fusion proteins attachment protein-25 SNAREs see soluble N -ehtylmaleimide-sensitive fusion proteins attachment receptors SOCS7 see suppressor of cytokine signalling-7 soluble N -ehtylmaleimide-sensitive fusion proteins attachment receptors (SNAREs) 196–7, 202, 219, 238, 254–5 soluble N -ethylmaleimide-sensitive fusion proteins attachment protein-25 (SNAP25) 196–7, 219, 254–5, 329 SPB see spindle pole bodies sperm flagella 192, 199, 239, 327–9 spindle pole bodies (SPB) 106–7 spindle positioning 106–7 Spn 60–2, 126–7, 129 Spr 23–4, 126, 131 stress fibres 214 subcellular compartmentalization 59–60 SUE see septin unique elements SUMO see small ubiquitin-like modifier suppressor of cytokine signalling-7 (SOCS7) 199–200, 216–17 supramolecular architecture 49–100 antibodies 54–5, 65–6, 68

369

Caenorhabditis elegans 62–4, 72–3 cellular asymmetry 68–72 coiled coil-forming domains 63–4, 72–6 collar assemblies 69–72, 74, 90–1 filament formation 50–1, 55–9, 62–4, 68–75, 77–84, 88–92 guanine nucleotides 63, 67, 76–80 hetero-oligomeric complexes 50–68, 76–83 invertebrates 62–4 linear rod complexes 51–60, 65–6, 77–80, 88–92 lipophilia 87–9 post-translational modifications 80–7 ring assemblies 69, 88–9 Saccharomyces cerevisiae 50–60, 69, 73, 88–92 Schizosaccharomyces pombe 60–2 septin–membrane interactions 87–9 substitute subunits 58–60 subunit arrangements 53–93 switch regions 77 two-hybrid method 53–4, 58, 61–2, 92 unicellular eukaryotes 50–62, 88–9 vertebrates 65–93 SVs see synaptic vesicles Swe1 111–13, 140 switch regions 77 synapse 247–67 astrocytes 256–7 endocytosis 247 exocytosis 253 filament formation 252–4 motor performance 256 nerve terminals 251 neurite outgrowth 255–6 neurology 258–60 neurons 250–1 phosphorylation of septins 257–8 regulation of septin expression 252 septin distribution 248–50, 251 septin functions 254–7 septin localization 250–1 vesicle secretion 254–5

370

Synapsin I 254 synaptic vesicles (SVs) 247 syntaxin 203, 219 systemic lupus erythematosus 298 T cells 193 testicular cancer 297 thrombosis 270 ts mutants 7–8 tubulins 217–18, 233–4, 276–7 tumour suppressor genes 287, 302, 322 two-hybrid method Drosophila melanogaster 156 mammalian septins 215–16 supramolecular architecture 53–4, 58, 61–2, 92 unc59/unc61 see Caenorhabditis elegans unicellular eukaryote 50–62, 88–9

Index prepared by Neil Manley

INDEX

5
-untranslated region (5
-UTR) 172–4, 176, 177–9 VAMPs see vesicle associated membrane proteins velocardiofacial syndrome 197, 260 vesicle associated membrane proteins (VAMPs) 196–7, 219, 250, 254, 329 vesicle secretion 254–5 vesicle trafficking 26–7 viral replication 193, 199 Western blotting 189 X-linked-IAP 285–7 Xenopus laevis 22, 78 yeast septins see Saccharomyces cerevisiae; Schizosaccharomyces pombe