RNA Processing A Practical Approach VOLUME II Ed ited hy S ." III G G lN S and B. D. II A \.1 ES
.
t·
.
m'G - - ....
GTP
G\"b
CD, • •
~f
AAUAAA
f-(Aln
Thl! Practical Approech Sl'fil!S
®
The Practical Approach Series SERIES EDITORS: D. RICKWOOD and B. D. HAMES In recent years. the importance of RNA processing in the regulation of eukaryotic gene expression has become abundantly clear and so, not surprisingly, it is now an area of intense research activity. These two volumes give detailed practical guidance on all major aspects of this subject. Step-by-step practical protocols from leading laboratories are presented for studies of the termination and end-processing, capping, methylation, splicing, and editing of mRNA as well as stud ies of mRNA stability and processing of rRNA and tRNA. The reader is led through all the key steps required for a successful experimental investigation. This includes full descriptions of the synthesis and purification of RNA substrates for in vitro work. the characterization of specific RNAs, and the isolation and analysis of ribonucleoprotein complexes.
RNA Processing: A Practical Approach, Volumes I and II. are invaluable laboratory companions for researchers working on mRNA, rRNA. or tRNA gene expression. Their lucid explanatory style and comprehensive coverage of all the key techniques for analysing RNA processing will be welcomed by both experienced researchers and workers embarking on such studies for the first time. Volume I 0199633444 spiral hardback o 19 963343 6 paperback Two volume set 0199634734 spiral hardback 0199634726 paperback
I S B N 0- 19-963470-X
OIRLPRESS
--,---at---OXFORD UNIVERSITY PRESS
Oxford New York Tokyo
RNA Processing Volume II A Practical Approach Edited by STEPHEN J. HIGGINS and B . DAVID HAMES Department of Biochemistry and Molecular Biology , University oj Leeds
= - at -
OXFORD UXI\'ERSITY PRESS O x fo rd
:\' ~ W
York Tok y o
11111111111111111 1111111111111111111111111111111111111111111
The Practical Approach Series SERIES EDITORS D. RICKWOOD Department of Biology. University oj Essex Wivenhoe Park. Colchester. Essex C04 3SQ, UK
B. D. HAMES Department of Biochemistry and Molecular Biology University of Leeds, Leeds L52 9JT. UK
Affinity Chromatography Anaerobic Microbiology Animal Cell Culture (2nd edition) Animal Virus Pathogenesis Antibodies 1 and 11 Behavioural Neuroscience Biochemical Toxicology Biological Data Analysis Biological Membranes Biomechanics-Materials Biomechanics-Structures and Systems Biosensors Carbohydrate Analysis Cell-Cell Interactions The Cell Cycle Cell Growth and Division Cellular Calcium Cellular Interactions in Development Cellular Neurobiology Centrifugation (2nd edition) Clinical Immunology Co mputers in Microbiology Crystallization of Nucleic Acids and Proteins Cytokines The Cytoskeleton Diagnostic Molecular Pathology 1 and 11 Directed Mutagenesis
DNA Cloning 1, 11 , and III Dro sophila Electron Microscopy in Biology Electron Microscopy in Molecular Biology Electrophysiology Enzyme Assays Essential Developmental Biology Essential Molecular Biology 1 and 11 Experimental Neuroanatomy Fermentation Flow Cytornetry Gas Chromatography Gel Electrophoresis of Nucleic Acids (2nd edition) Gel Electrophoresis of Proteins (2nd edition) Gene Targeting Gene Transcription Genome Analysis Glycobiology Growth Factors Haemopoiesis Histocompatibility Testing HPLC of Macromolecules HPLC of Small Molecules Human Cytogenetics 1 and 11 (2nd edition) Human Genetic Di sease Analysis
Imm obili zed Cells and Enzymes Immunocytochemistry In Situ Hybridization Iod inated Densit y Gradient Media Light Micros copy in Biology Lipid Analysis Lipid Modification of Proteins Lipoprotein Anal ysis Liposomes Lymp hocytes Mammalian Cell Biotechnology Mam ma lian Development Medical Bacteriolog y Medical Mycology Microcomputers in Biochemistry Microcomputers in Biolog y Microcomputers in Physiology Mitocho ndria Molecular Genetic Analysis of Populations Molecular Imaging in Neuroscience Molec ular Neuro biology Molecular Plant Pathology I and 11 Mo lecular Virology Mo nito ring Neuronal Activity Mutagenicity Testing Neural Transplantation Neurochemistry Neuronal Cell Lines NM R of Biological Macromolecules Nucleic Acid and Protein Sequenc e Analysis Nucleic Acid Hybridisation Nucleic Acids Sequencing Oligonucleotides and Analogues Oligonucleoti de Synthesis PCR Peptide Hormone Acti on
Peptide Hormone Secretion Photosynthesis: Energy Transduction Plant Cell Culture Plant Molecular Biolog y Plasmids (2nd editi on) Pollination Ecology Postimplantation Mammalian Embryos Preparative Centrifugation Prostaglandins and Related Substances Protein Architecture Protein Eng ineering Protein Function Protein Phosphorylation Protein Purification Applications Protein Purification Methods Protein Sequencing Protein Structure Protein Targeting Proteolytic Enzymes Radioi sotopes in Biology Receptor Biochemistry Receptor-Effector Coupling Receptor-Ligand Interactions Ribosomes and Protein Synthesis RNA Processing Signal Transduction Solid Phase Peptide Synthesis Spectrophotometry and Spectrofluorimetry Steroid Hormones Teratocarcinomas and Embryonic Stem Cells Transcription Factors Transcription and Translation Tumour Immunobiology Virolo gy Yeast
Oxford University Press, wa lton Street, Oxford OX2 6DP Oxford
New York Toronto Calculla Ma dras Karachi Petaling Jaya -Singapo re Hong Kong Tokyo Nai robi Da r es Salaam Cape Town Melbourne Auckland Delhi
Bombov
and associated companies in Berlin Ibadan Oxford is a trade mark of Oxford University Press A Practical A ppr oach ~ is a registered trade mark of the Chancellor, Masters, and Scholars of the University of Oxford trading as Oxford University Press Publish ed in the United States by Oxford University Press, New York @Oxf ord Unil'enilY Press , 1994
All rights reserved. No part of this publinJtion m ay be reproduced, sto red in a retrieval system, or transmitted, in any form or by any means, without the prio r permission in writing of Oxf ord Uni versity Press. Within th e UK, exceptions are allowed in respect of any fair dealing f or the pu rpose of research or priva te study, or crit icism or review, as pe rm itted under the Copyright, Designs and Patents Act, 1988, or in the case of reprographic reproduction in accordance with the terms of licences issued by the Copy right Licensing A gency. Enquiries concerni ng reproduction outside those terms and in ot her count ries shoufd be sent to the Rights Depart m ent, Oxford University Press, at the address above. This book is sold subject to the condition that it shall not , by way oj trade or otherwise, be lent, re-sold, hired out, or otherwise circulated withou t the publishe r's prior consent in any form of bindin g or cove r othe r than that in which it is publis hed and with out a similar condition including this cond ition being imp osed on the subsequent pu rchaser. Users of boo ks in the Practical Approach Series are ad vised that p rudent laboratory safety procedures should be followed at all times. Oxford Unive rsity Press m akes no representation, express or implied. in respe ct of the accuracy of the mate rial set forth in boo ks in this series and cannot accept any legal responsibility or liability for any errors or omissions that may be made. A catalogue record fo r this book is availab le from the British Libra ry Library of Congress Catalog ing in Publica tion Da ta RNA processing: a practical approach I edited by Step hen J. Higgins and B. David Ha m es. cm.-(The Practical appro ach series .. 135, /36) p. Includes bibliographical references and index. I. RNA-Research-Laboratory manuals. 2. Genetic regulationResearch-Labo ratory manuals. I. H iggins, S. J. (Steve J.) ll. Hames. B. D. Ill. Series. 93-230[9 QP623 .R58 1993 574.87 '3283-dc20 [SBN 0 19 963344 4 (hlb) Volume I ISBN 0 19963343 6 (plb) [SBN 0199634718 rhlb) ISBN 0 19 963470 X (plb)
J Volume JI
Typeset by Dobbie Typesetting Limited, Tav istock, Devon Printed and bound in Great Britain by Information Press, Ltd., Eynsham, Oxon
Foreword RNA processing: two helpfu l guides for cutting , pastin g, trimmin g, and editing RNA TOM MANIAT IS
During the past decade , the study of RNA processing ha s risen to the forefront of molecular biology. At one time , the RNA processing field was a relatively small group of investigato rs interested in the trimming and packaging o f prokaryotic transfer RNA and ribosomal RNAs . However , with the disco very of a bewildering array of RNA processing events in both prokaryotes and eukaryotes, the field has become a burgeoning enterprise. This change is rellected in the annual RNA processing meeting which has been rap idly transformed from a small informal gathering to one bursting the seam s of the expanded meeting facilities at the Col d Spring Harbor Laboratory. Due to the pace of this unprecedented growth in size and diversity, no one ha s stopped long enough to compile a detailed description of even the most basic techniques used to study RNA. Students in the field are therefore introduced to the laboratory with handwritten protocols pass ed fro m person to per son . The methods compiled in this book and its companion volume by leaders in the field should therefore contribute significantly to the tr aining of new st udent s and hopefully lead to the development of new tech niques. Volumes I and II of RNA processing: a practical approach cover virt ua lly all aspects of RNA processing , including ca pping, splicing of both pre-mR NA and tRNA, polyadeny lation, editing and ribosomal RNA processing. In addit ion, an overview of ribozy mes is provided. An understanding of the various types of RNA processing in vivo is an essential prerequisite for studying in vitro the ~echanisms involved. Therefore , the descrip tion o f methods for RNA mapping, mcluding nuclease S I ma pping , prime r extension analysis, and the use of th e polymerase chain reaction sho uld be very useful. An important technical adva nce in the analysis of RNA processing in vitro was the development of pla smid vecto rs containing bacteriophage-specific promoters for synthesizing labelled substra te RNA s. A descr iption of the inthesls and purificat ion of in vitro transcripts pro duc ed with SP6, T7, and . 3 polymerase IS there fo re an essential part of the coverage. Also imp ortant IS a description of the prep aration and opt imization of nuclear extracts for several
Foreword
For eword
types of in vitro RNA processing, 3 ' -end formation and polyadenylation, The manipulation and fractionation of the se extracts have led to the purifi cation of individual proce ssing components and the cloning and characterization of th e corresponding genes. Methods have also been developed for the selective inactivation of individual snRNPs using specific DNA oligonucleot ides, or the depletion or purification of sn RNP particles using oligo(2 ' -O-alkylribonucleotides). In addition, immunoaffinity purification methods for purifying snRNPs are presented as well as procedures for reconstituting individual snRNPs from purified protein s and RNAs. These tools for characterizing snRNPs will become increasingly important as the details of protein-protein and protein-RNA interactions in the spliceosome are unravelled. An area less intensely investigated but of increasing importance, is the turnover of mRNA. There are now man y examples in which gene expre ssion during development is regulated by the selective turnover of specific RNAs . In addition, it is now clear that the rap id decrease in the expression of cytokines after induction by extracellular inducers involves a specific recognition sequence in the 3 ' non-c oding sequence of mR NA. Th e chapter describin g methods for stud ying mRNA turn over in vitro will stimulate furt her studies of this important problem. The most recentl y discovered and most unexpected forms o f RNA processing are autocatalytic splicing and RNA editing. The discover y of group I intron self-splicing and the RNase P ribozyme has led rapidly to detailed characterization of the catalytic reactions involved. Dissection and manipulat ion of the group I intron ribozyme have led to the creation of new catalytic activities and an understanding of the role of RNA secondary and tertiary structure in ribo zyme function. Similarly, the elucidation of the structure and acti vity o f group II introns has provided important new insights into the nature of ribo zymes and has led to critica l insights into the mechani sms involved in pre-mRNA splicing. Both group II intro n and pre-mR NA splicing proceed through a similar branched RNA inter mediate. Alth ou gh th e former process is au tocatalytic a nd the latte r requ ires multiple components assembled into a spliceoso rne, the cat alytic events may be qu ite simila r. In fact, the parallels between the role of RNA struc t ure in group II intron splicing and the role of snRNA-pre-mRNA interac tions have increased by the recent demo nstratio n of dynamic interactions between snRNA and specific sequences in pre-mRNA during the pre-mR NA splicing reaction. Th e simple view is that the same cata lytic mechani sms are involved in bot h processes. However, in the case o f pre-mR NA processing th e RNA-R NA interactions are mediated by spliceosoma l proteins. More practical appli cation s of ribozymology are emerging from the study of hammerhead and hairpin ribozymes, a nother top ic discussed in the ribo zyme chapter. The manipulation of these interesting molecules and their use in ta rgeting and processing specific RNA tra nscripts are leading to a better understanding of RNA catalysis. RNA editi ng, the pro cess o f post-trans criptional inserti on, deletion or substitution of specific bases in mRNA, is the most pu zzling example o f RNA
processi ng. The discovery o f guid e RNAs complementary to edited portions of mRNA and the identification of putative intermediates in the editing processes have led to the proposal of a specific mechanism for RNA editing . Th is mecha nism invo lves an orderly cycling of the editing process in a 3 ' to 5 ' directi on. Specific catalytic mechanisms have been proposed based in part on compa risons to the group I and II self-splicing reactions . The detailed description in RNA processing: a practical approach of the biological systems and techniques used to study RNA editing should stimulate the development of new approaches to the study of the mechanisms involved. Although most of this two volume set focuses on the use of biochemical approaches to RNA processing, a description of genetic techniques used to study pre-mRNA process ing in yeast is also provided. The application of these techniques has led to the identification and cloning of genes encoding essential splicing factors (PRP genes). In addition, a variety of genetic tools, including gene 'knockouts'. targeted mutagenesis. interactive suppression, and conditional expression vecto rs have been used to study the function of a number of yeast splicing components including PRP proteins and specific snRNAs. These tools have been used in conjunction with in vitro studies using nuclear extracts prep ared from wild type yeast strains as well as strains lacking an essential splicing factor. Co mparison of the results of these studies with those of mammalian splicing has revealed a remarkable conservation in the mechani sms of the splicing reaction and spliceosome assembly, and in the role of snRNPs in splice site recognition . This comparison of yeast and mammalian splicing points to an additional benefit of collecting a variety of RNA processing methods together . It provides the opportunity to examine the similarities and differences in experime ntal approaches established for studying different types of RNA processing. Hopefull y, this will lead to previously unrecogni zed connections and new ideas.
viii
Ix
Preface This book arose ou t of the success of a book we edited for the Pr actical Approach series in 1984, entitled Transcription and translation: a practical approach. When the time came to consider organizing a second edition . it rapidly became clear that no one book o f th e desired size co uld include. in suf ficient detail, the myriad of important new techniques. parti cularly in the area of posttra nscriptional processing of RNA transcript s. that had arisen since the first edition. Thu s the logical decision was taken to produce several books. th e first of which. Gene transcription: a practical approach. has recently been published. RN A processing: a practical appro ach Volumes I and II are the next book s in this planned set and are compan ion volumes. RNA processing Volume II includ es contribution s fro m Walter Keller' s laborato ry on techniques for analysis of 3 ' end -processing of mR NA and fro m Aaron Shatkin's group on capping and meth ylation of mR NA . RNA editing and the analysis mRNA tu rnover are also included in this volume. with cha pters by Larry Simpson and his colleagu es and Jeff Ross. respecti vely. Barbar a Se llner-Webb and Cathy Enri ght describe methods for ribo somal RNA processing a nd Chris Greer similarl y covers transfer RNA processing. Finally, David Shub , Craig Peebles. and Arn old Hampel have combined their ef fort s to produce a very top ical chapter covering investigati on s o f detai led splicing reaction s. co mprehens ively covering group I intron slicing. group II intron splicing a nd hammerhead and hair pin ribozymes . Th e co mpanion volume. RNA processing Volume I , begins with Beno it Chabot describing the synthesis a nd purification of RNA substrates for RNA proce ssing investigations. Paula Grabowski then covers essential meth od s for the identification and analysis of spliced mRNA s. Next, Ian Eperon and Ad rian Krainer have collaborated to produce an import ant cha pter on the analysis o f th e splicing of mRNA precur sors in mammalian cells. Another key area . the analysis of RNP complexes and their inter action s, is covered in two chapters by An gus Lamond and Brian Sproat and by Reinhard U ihrmann and his colleague s. These are all centr al issues in RNA processing stu dies. Fina lly, Andrew Newma n has contributed a chapter on investigat ions of pre-mR NA splicing in yeast , which allows some approaches that a re not possible with mammalian cells, most not ably the isolation and analysis of splicing mutants. T he aim of both books remains the aim of thi s popular series; to present , in a clea r readable ma nner, the background to the range of techniques and experimental ap proaches available , to describe in precise detail a key selection of tried and tested proto cols, and to discuss pot ential pitfalls . data interpret ation, and a variety of ot her hint s and tips for the active scientist. It is a measure of the effort s of our contribu tors th at we believe these aims have been more than
Prefa ce met in each volume. We thank them for their diligence in writing texts which address these aims and, where we felt that editorial changes were essential, for graciously accepting these. We hope and believe that the end result will be seen as a comprehensive and valuable two volume compendium of the best of current methodology in th is subject area . Because of the scientific qualit y of the contributions and the deliberatel y explanatory style of writing, we ar e confident that, like their predecessor, these books will rightfully enjoy popularity among
Contents
both newcomers to the field and more experienced researcher s.
Stephen Higgins David Hames
Leeds February 1993
List of contr ibutors Abbreviati ons
xix
xxi
1. 3' end-processin g of mRNA Elmar Wahle and Walter Keller 1. Introduction: overview of 3' end-processing in mammalian cells
2. 3' -en d cleavage and polyadenylatio n in nuclear extracts Introdu ction RNA substrates for 3 ' end-proce ssing in vitro Preparation of nuclear extracts for 3 ' end-processing
Formation of polyadenylation complexes Assay of cleavage and polyadenylation in nucl ear extracts
3. Polyadenyl ation and 3' -cleavage factors
2 2
2 4
6 21
26
Overview of processing factors Purification of processing factors
26 26
4 . Analysis of po lyadenylation in viva
33
References
2. Capping and methylation of mRNA
33 35
Yasuhira Furuichi and Aaron J. Shatkin
II
1. Int roducti on
35
2 . Synth esis of capped mRNAs during in vitro transcription of viral genes
38
Introductio n
38
Strategy
38
Synthesis of capped mRNA s using insect cytoplasmic polyhedrosi s virus Synthesis of capped mRNAs using reoviru s Synthesis of capped mRNAs using vaccinia virus In vitro transcriptionand capping of cellular mRNAs using bacteriophage
39 41 42
RNA polymerase,
xii
43
Cont ent s
Cont ent s 3 . Site-specific radiolabelling of caps in ce llu lar mRNAs
45
Introduction
45
Radiolabelling mRNAs by decapping and recapping
45
Labelling of mRNA by period ate oxidation followed by reduction with
['H]sodium borohydride Labelling of mRNA by 2 ' -O-methylation 4 . Cha racte rizati on of caps Strategy
50 51
52 52
Enzymic analysis o f caps Ana lysis of caps follo wing enzymic digestio n
54 57
Acknowledgem ents
65
Refer en ces
65
3. RNA editi ng in mi tochondria
69
Larry Simpson. Agda M . Simpson. and Beat Blum 1. Introduction
69
Po st-transcripti onal mod ifications of eukaryot ic mRNAs RNA editing in kinetoplastid proto zoa
2. Growth and maintenance of kinetoplastid protozoa Choice of species for experimental work Growth and maintenance of L. tarentolae Cloning of kinetoplastid protozoan stock cultures Growth of cultures for production of kinetoplast components
3 . Kinetoplast DNA Introduction Isolation of kDNA
71 71 72 74 74
75 75
76
Schizodeme-typing of kinetoplastid proto zoan strains Isol ation of maxici rcle DNA
78 79
4 . Isolation of the kinetoplast-mitochondrion
81
5. [solation of transcripts from th e mitoch ondrial gen ome
85
Introductio n Preparation of mitochondrial extracts Terminal uridylyl transfe rase (TU Tase)
99
gRNA :mRNA chimaera-form ing activity
101
Acknowle dgement
[03
Refere nces
103
4. Analysis of messenger RNA turn over in cell-free
extracts from mammali an cells 1. Introduction
107
2. Ch oice of mRNA substrate
[08
Endogenou s substrates (free mRNP or polysome-associated mRNP) Exogen ous substrates
110
4 . Preparation of ce ll extracts for in vitro mRNA decay reactions
I II
5 . Methods for performing in vitro mRNA decay reacti ons
[ [9
Detergent-free extracts Lysolecithin extracts Reticulocyte translation extract s
6. mRNA detecti on, data interpretation, and troubles hooting
126 126 130
Troubleshooting
[3[
References
132
5. Ribosomal RNA processing in vertebrates
90
Introduction
91
Analysis of rRNA processing by hybrid selection and gel electrophoresis
xiv
119 123 125
mRNA detection and some typical in vitro mRNA decay experiments Data interpretation
chimaeric molecules Synthesis of gRNAs by transcription in vitro
94
109 110
3. Preparation of u ndegraded mRNA and polysomes
88 88
co mparison
107
Jef f Ross
Northern blot analysis of kRNA Hybrid selection of guide RNAs PCR amplification of partially-edited mRNAs and gRNA:mRNA
7 . Identification of gRNAs b y comp u te r-assisted sequence
96 97 97
[00
85 87
91
96
RNA ligase Cryptic RNase
Introduction and strategy Isola tion of kine topla st RNA
6. Run-on transcription in iso lated kin et opl ast mitocho ndria
II
69 69
8. Enzymatic activities in the kin etoplast-mitochondrion fraction whic h are involved in RJ\lA ed it ing
135
Cathleen Enright and Barbara Sollner- Webb 1. In trod u cti on
135
2 . Analysis of rR.'\!A processing in vivo
137 137
of cellular rRNAs
139
Use of psorale n cross-linking for the analysis of assoc iatio ns betwee n rRNAs in vivo Elect ron micro scope anal ysis o f rRNA synthesis (Miller spreading)
xv
142 143
Cont ents 3. Analysis of rRNA processing in vitro
Cont ent s 144
Introduction
144
Preparation o f synthetic rRNA substrates Preparation and use of mouse 5·1 00 processing extract Preparation and use o f nucleolar and cytoplasmic processing extracts from HeLa cells Cleavage of pre-rRNA using purified nucleolar endonuclease Use of deletion analysis in the investigation of rRNA processing
144 148
4. Analysis of protein -RNA associations in rRNA processing Gel-s hift analysis Analysis of rRNA associations by rate zonal centrifugation Use of UV cross-linking to detect polypeptides associated with rRNA in processing complexes
5. Assessmen t of the requirement for cellular RNA in rRNA processing Micrococcal nuclease digestion
I SO 155 155
155 157
Acknowledgements
164 164 165 169 170
170
173 173 173
2. Preparation of pre-tRNA pro cessing substrates
176
Preparation of crude cell extracts Partial purification of proce ssing activities Preparatio n of highly-purified processing components
xvi
206
7. Ribozymes
211
David A. Shub , Craig L. Peebles, and Arnold Hampel 211 211
1. Int rod uction General methods tRNA structure and function Overview of processing pathway s Biochemical analys is of tRNA processing
3. Preparation of tRNA pro cessing ext racts
Referen ces
2. Group I intron rib ozymes
Chris L. Greer
Introduction Preparation of nuclease-free carriers Preparation of tRNA substrates from whole cells Preparation of substrates by in vitro transcription
206
] 60
170
6. Processing of transfer RNA precursors
Acknow ledgements
211
170
References
195 195 196 199
1. Introduction
6. rRNA pro cessing associated with tra nscriptional ter mi nation 169 3' end -processing in the mouse 3 ' end-processing in Xenopus laevis
SITaregy Assa y of pre-tRNA splicing Assay of base mod ification s and tRNA splice junctions
157
Oligonucleotide-directed RNase H digestion of cellular RNAs involved in pre-rRNA processing Immunological analy sis o f polypeptides required for proce ssing
4. Pre-tRNA processing assays
173 175 176
176 176 177 181
The self-splicing activi ty of group I introns Detection o f self-s plicing intron s in cellular RNA Co nfirmatio n o f self-splicing by in vitro transcription of cloned DNA conta ining putative group I introns
3. Group II intron ribozymes Identification of group II introns Group II intron secondar y structure and functional anatomy Poten tial of grou p II intron ribozymes for practical use Self- splicing o f group II introns Other reactions of group II int ro n ribozyme s
4. Hammerhead and hairpin ribozym es Introduction Application s of hammerhead and hairpin ribozy mes in mole cular biology Use o f hammerhead and hairpin ribozymes to target specific RNA transcripts Use o f autocatalytic cassettes to generate transcripts with defined 5 ' or 3 ' termini
213 215 217 217 21 9 220 221 224 230 230 232 232 236
Acknowledgements
237
References
237
Appendix : Suppliers of specialist Conten ts of Volume I Index
185 187 189 191 xvii
items
241
245 247
Contributors BE AT BLUM Department of Biochemistry, University of Bern, Freiestrasse 3, CH-30l2 Bern, Switzerland. CAT HLEEN ENRIGHT Depa rt ment of Biological Chemistry, School of Medicine , The Johns Hopkins Un iversity, 725 North Wolfe Street, Baltimore, MD 21205-2185, USA. YAS U HI RO FURUICHI Biotechn ology Subdivision, Nippon Roche K.K., Nippon Roche Research Center, 200 Kajiwar a, Kamakura-Shi, Kanagawa 247, Japan . CHRI S L. GREER Department of Biologica l Chemistry, College of Medicine, D240, Med Sci I, University of California, Irvine , CA 92717, USA. AR NOL D HAMPEL Department of Biological Sciences, Northern Illinois University, De Kalb, Illinois 60115-2861, USA . B. DA VID H A ME S Department of Biochemistry and Molecular Biology, University of Leeds, Leeds LS2 9JT. STEPH E N 1. H IGG I NS Department of Biochemistry and Molecular Biology, University of Leeds, Leeds LS29JT. WALT E R KE L L E R Abteilun g Zellbiologie, Biozentrum der Universitat Basel, Klingelbergstrasse 70, CH-405 6 Basel, Switzerland . TOM MAN IATIS Departm ent of Biochemistry and Molecular Biology, Harvard University, 7 Divinity Avenue , Cambridge, MA 02138, USA . C RAIG L. PE EBLES Depart ment of Biological Sciences, 365 Crawford Hall, University of Pittsburgh , Pitt sbur gh, PA 15260, USA . JE F F ROSS MCArdle Laboratory for Cancer Research, University of Wiscon sin-Madi son, 1400 University Avenue , Madi son , Wisconsin 53706, USA.
Contributors DA VID A . S HUB Center for Molecular Genetics, Depart ment of Biologica l Scienc es 126, State University of New Yo rk at Alba ny, 1400 Washingto n Avenue, Albany NY 12222, USA. AARON 1. S HATK IN Center for Advanced Biotechnology and Medicine, 679 Hoes Lane, P iscataway, New Jersey, NJ 08854-5638, USA. LA R RY S I M PSON Howard Hughes Medical Institut e, Dep artment of Biology, University of California, Los An geles, 10833 Le Coate Avenue, Los Angeles, Ca lifornia, CA 90024-1662, USA. A GD A M . SIMP SO N Depart ment of Biology, University of California, Lo s Angeles, 405 Hilgard Avenu e, Los Angel es, Ca lifo rnia, CA 90024-1606, USA. BARBARA SO LL N ER -W EBB Depart ment of Biological Chemistry, School of Medicine, The J o hns Ho pki ns University, 725 North Wolfe Str eet, Baltimo re, MD 21205-2185, USA . E L M AR WAH L E Abteilung Zellbiologie, Biozen tru m der Universitiit Basel, Klingelbergstrasse 70, CH -4056 Basel, Switzerland .
I
Abbreviatio ns A
Am
A 260 A 600 Ado Hc y Ado Met AMP ASV ATP bp BHI BHK BSA C cD NA Ci CIP CM P c.p.rn . CPSF C PV CTP D DAP I dATP dd ATP DEAE DBAE DEPC DH FR DMEM DMSO DNase d .p .m . DTE
DTT EBS EDTA EGTA xx
ade nine /adenosine/adenylate a bso rba nce (259 nm ) absorba nce (260 nm) absorba nce (600 nm) S-ad enosylhom ocysteine S-adenosylm eth ion ine adenosi ne 5 ' -monophosphate avian sarcoma virus adenosine 5 ' -trip hosphate base pair(s) bra in hea rt infusion baby ham ster kidney bov ine serum albumin cytos ine/ cytidine/ cyt idylate co mplementa ry DNA curie calf intestinal alkaline phosp hatase cytidine 5 ' -mo no ph osph at e counts per minute cleavage and polyad en ylation specificity factor cyto plas mic polyhed rosis virus cyt idine 5 ' -triphosphate dih ydr o uridine 4 ' ,6 ' -diamidino-2- phenylindo le deoxyad enosine 5 ' -triphosphate dideoxyadenosine 5 ' -trip hospha te dieth yl amino ethyl dihydro xybo ryl ami noethyl diethyl pyrocarb onate dih yd ro folate reductase Dul becco ' s mo difie d Eagle's medium dimeth yl sulphoxide deoxyrib onuclea se disintegratio ns per minute dithioeryt hrito l dithiot hreitol exo n binding sequence ethyle nediamine-tet raacetic acid ethylene glycol-D ,D ' -bis(2-aminoethyl)-N,N,N ' ,N '-tetraacetic acid
Abbreviations elF ETS FPLC
kbp
eukaryotic initiation factor external transcribed spacer fast-per formance liquid chromatography gravity guanine! guanosine! guanylate guanosine 5 ' -monophosphate guide RNA guanosine 5 ' -triphosphate hour(s) hypoxanthine-guanine phosphoribosyltransferase human immunodeficiency virus high-performance liquid chromatography intron binding sequence internal guide (sequence) isopropyl-{3-D-thiogalactoside internal transcribed spacer apparent equilibrium constant kilobase pair(s)
kcat
turnover number
kDa kDNA
kilodaltons kinetoplast DNA Michaelis constant kinetoplast RNA lactate dehydrogenase I-methyl adenine 6-methyl adenine 2,2,7-trimethyl guanosine 7-methyl guanine minute(s) messenger RNA messenger ribonucleoprotein (particle) nucleotidc(s) reduced nicotinamide dinucleotide nucleoside diphosphate kinase sodium acetate, EDT A, SDS buffer nucleoside 5' -triphosphate origin of replication po ly(A)-binding protein polyacrylamide gel electrophoresis phosphate-buffered saline polymerase cha in reaction packed cell volume polyeth ylene glycol polyethylenimine
g G
GMP gRNA GTP h
HGPRT HIV HPLC IBS ig IPTG ITS k app
Km
kRNA LDH m'A m6A m3G m'G min mRNA mRNP nt NADH NDK NES NTP ori PABP PAGE PBS PCR PCV PEG PEl
xxii
Abbre viations PEP PER
phosphoenolpyruvate pre-edited region inorganic phosphate Pi pyru vate kinase PK PM SF phen ylmethyl sulphonyl fluoride PNV packed nuclear volume p.p.m. parts per million p.s.i, pounds per square inch PYA polyvinyl alcohol Poly(A) polyadenylic acid Poly(A) + RNA polyadenylated RNA Poly(C) polycytidyli c acid RNase ribonuclease RNP ribonucleoprotein (particle) rNTP ribonucleoside 5 ' -triphosphate r. p.m . revolutions per minute rRNA ribosomal RNA RSE Renografin, sucrose, EDT A buffer RSTE Renografin, sucrose, Tris, EDTA buffer RSW ribosomal salt wash S svedberg SDS sodium dodecyl sulphate SDS- PAGE SDS -polyacrylamide gel electrophoresis sec second(s) SER spliced exon reopening SET salt, EDTA, Tris buffer SJH splice junction hydrolysis snRNA small nuclear RNA snRNP small nuclear ribonucleoprotein (particle) SSC standard saline citrate STE sucrose, Tris, EDTA buffer STM sucrose , Tris, MgCl, buffer SV40 simian virus 40 T ribothymidine!ribothymidylate TAE Tris-acetate-EDTA buffer TBE Tris-borate-EDTA buffer TCA trichloroacetic acid TE Tris-EDTA buffer TEAB triethylammonium bicarbonate TE MED N, N,N ' ,N ' -tetramethylenediamine TLC thin layer chromatography Tm melting temperature TM A tetramethylammonium chloride TM N Tris , MgCl" NaCI buffer
xxiii
Abbreviations
W
tobacco mosaic virus tra nsfer RNA Tris saline Tr is, SOS, EOTA buffer term inal uridy lyl transferase uracil /uridi ne/uridylate urid ine 5 ' -diphosphate uridine 5 ' -monophosphate uridine 5 ' -triphosphate ultravio let volt(s) initial velocity resuspension vol ume watt(s)
X-gal
5 -bromo-4-ch lo ro-3-in do lyl - ~-D-galactos ide
'!'
pseudo uridine
TMV tRNA TS TSE TUTase U
UOP UMP UTP UV V V;
v,
3' end-processing of mRNA ELMAR WAHLE and WALTER KEL L E R
1. Introdu cti on : overview of 3' end-process ing in
mammali an cells The 3 ' -end of a eukaryotic messenger RNA is not formed by precisetermination of transcription. Rather, RNA polymerase II synthesizes a precursor RNA extending far beyond the 3' -end of the mature mRNA. The end of the mature RNA is then generated by post-transcriptional processing of the precursor RNA. Formation of the 3' -end is an essential step in mRNA synthesis; mutations in the RNA sequences that direct this process lead to a strongly reduced steady-state level of mRNA. In mammalian cells, almost all mRNAs receive their 3 ' -ends in a process consisting of two reactions: (a) The precursor RNA is first cleaved endonucleolytically at a particular site 3' (downstream) of the coding sequences. (b) The 5' (upstream) fragment is then extended by the addition of approximately 200 adenylate residues (polyadenylation), whereas the downstream fragment is degraded . The reactions depend on two sequences in the precursor RNA: the highly conserved sequence AAUAAA 10-30 nt upstream of the cleavage site and one or more sequence element s within approximately 50 nt downstream of this site. The downstream elements are not well defined; they may be either GU-rich or runs of U. In yeasts and plants, mRNAs are also polyadenylated. However, the sequence requirements and, possibly, mechanisms of 3 ' -end formation are different from those outlined abo ve (for a recent review, see ref. I) . Cleavage and polyadenylation can be carried out in vitro in a nuclear extract from tissue culture cells (2). A complex sex of protein factors has been identified that catalyses both reactions in a tightly coupled fashion . At least two of these proteins, and possibly more, are 'cleavage factors' invol ved only in the initial endonu cleolytic scission; they are dispensable for polyadenylation. Two other facto rs are also required for the cleavage reaction but, in addition, carry out polyadenylation. These are poly(A) polymerase, the enzyme that catalyses the Polymerization of the poly(A) tail from ATP, and a factor that binds to
xxiv
1: 3' en d-p rocess ing of m RNA
Elmar Wahl e and Walt er Kell er
AAUAAA and provides the polymerase with its specificity for RNAs containing this sequence. This latt er facto r has been described under different names; the auth ors will refer to it as cleavage and polyadenylation specificity factor (CPSF) . In addition, a poly(A)-bindin g pro tein is involved in the elongation pha se of poly(A) tail synt hesis. For recent reviews, see refs I, 3. In mammalian cells, the only known exceptions to the mRNA pro cessing pathway described above are the precursors to the major histone mRNAs. These undergo an endonucleolytic cleavage just downstream of a hairpin-loop structure . Poly(A) tails are not add ed. The react ion can be analysed in vitro (4). One of the factor s required is the U7 sn RNP . Messenger RNA sequences involved in this form of 3 ' end-processing include a region which basepairs with the U7 RNA. The ot her facto rs involved in the reaction have not been well
(a) There is some suggestion that the 5 ' -cap structure of an RNA transcript plays a role in 3 ' end-processing, though th is is still controversial. The cap certainly sta bilizes the RNA during the incubation in nuclear extra cts. In general, capping o f the RNA sub strate is, there fore, desirable . (Volume I, Chapter I, Section 2.2 pro vides a protocol for the synthesis of capped RNA substrates.)
characterized. For a review, see ref. 5.
This chapter describes methods used to study endo nucleolytic cleavage and pol yadenylation of mRNA precursors in vitro. Pr ocedu res tha t are of basic importance and general applicability have been included. References will be given for procedures which are beyond the scope o f thi s chapter or with which the authors have no personal experience. Section 2 deals with 3 ' end-processing in vitro in nuclear extracts from HeL a cells. Section 3 describes meth ods for the purification and assay of pro cessing factors. Section 4 gives advice on monitoring proces sing in vivo .
2 . 3 <end cleavage and polyadenylation in
nuclear extracts
(b) RNA substrates of different specific radioactivities are required for different assays (see Section 2.5). Make RNA of the desired specific radioactivity by adjusting the ratio of labelled to unlabelled UTP in the tran scription reaction (see Volume I, Chapter I, Section 2.2). (c) In vitro processing react ions should be ca rried out with known quantities o f the RNA substrate. Calculate the conc entration of radioactive RNA synthesized from the amount of radioact ivity incorporated (count an aliquot o f the RNA preparation in a scint illation counter), the specific radioactivit y o f the [32P JUTP used in the transcription reaction, and the number of UMP residues in the RNA . Further advice is given in Volume I, Chapter I, Section 2.3.4. (d) Store the RNA as an ethanol precipitate at - 20 °C. Brief vortexing of th is should be sufficient to generate a homogeneous suspension so that th e desired quantity of RNA for use in processing reactions can be withdrawn . If a homogeneous suspension can not be generated for accurate and reproducible samp ling, precipitate the RNA (see Protoco/ 5 , steps 1- 3), disso lve the pellet either in water or 50070 ethanol, and store it at - 20 °C. In these cases, lyophi lize an aliquot of the RNA prior to use.
2.1 Introdu ction Processing of the 3 ' -end of an RNA tran script may be studied in nuclear extracts with the use o f appropriate RNA substrates synthesized in vitro. This section describes the preparation of nuclear extract s active in 3 ' end-processing and their use in studies of the formation of polyadenylation complexes, as well as of th e cleavage and polyadenylation reactions th emselves.
2.2 RNA substrates for 3' end- proc essin g in vitro In vitro, RNA proce ssing reactions are most easily analysed by the use of rad iolabelled RNA substrates. The se precur sors are synthesized generally by 'run-off' transcription of appropriate gene sequences cloned downstream from prom oter s for bacteriophage-encoded RNA polymera ses (e.g . T3, T7, o r SP6 polymera ses). Protocol s fo r the synthesis of RNA substrates and their puri fication by dena turing polyacrylamide gel electro pho resis are described in Volum e I, Chapter I (Section 2). Th e followin g practi cal points should be observed when RNA substrates for in vitro proce ssing are synthesized: 2
(e) If the RNA is to be used as a 'precleaved ' substrate for polyadenylation (see Section 2.5.4), a free hydrox yl group at the 3 ' -end is essential. Both chemical hydrol ysis of RNA and degradation by the majority of nucleases leave a phosphate group at the 3' -end which prevent s polyadenylation. While RNA synthesized by the standard run-off procedure is generally a good substrate for polyadeny lation, the authors have encountered phage T7 RNA polymera se from one supplier which generated RNA that could not be extended by poly(A) polymerase. The quality o f the RNA 3 ' -ends can be checked by a preliminary nonspecific polyadenylation procedure. For this purpose, Mn' +-dependent extension with unlabelled ATP followed by gel electrophoresis (see Section 2.5.5) should be used for radiol abelled RNA. Extension with radiolabelled ATP followed by TCA precip itati on (see Protocol 8) sho uld be used for unlab elled RNA substrates. Many studies of 3 ' end-p rocessing in vitro have used two part icular polyadenylati on sites: the site in SV40 late mRNA and the L3 site o f adenovirus 2. 3
1: 3' end -processing
Elmor \Vahle and \Volter Kell er
of mRNA
If a polyadenylatio n site is to be used that has not been characterized before, the following points should be noted:
Protocol 1. Pre parat ion of HeLa cell nuc lear extract for 3' end-processing'
(a) Use a genomic fragment rath er than a cDNA fragment to generate the RNA substrate since the essential downstream elements will not be present in the latte r. Design the run-of f transcription so that at least 50 nt downstream of the polyadenylation site are included in the precursor RNA.
Equipm ent and reagents
(b) Some genes contain additional 3' end-processing signals upstream of the AAUAAA sequence. Therefore , include in the RNA substrate about ISOnt upstream of this sequence. (c) To test whether any processing event obse rved is genuine , use an RNA precursor with a point mutation in the AAUAAA sequence (e.g. AAGAAA) as a control.
e Doun c e
• High-salt buffe r C (buff er C contai ning
homogen izer (capacit y 120 ml for 1 5litres of Hela celt cultu re! with a type S pestle
1.2 M NaCl1 e Buffer Db (2 0 m M Hepes -KOH , pH 7 .se, 20% glycerol, 100mM KC!. 0 .2 mM
• Phase-contrast microsc ope • 1.0 M OTT in 10 mM sodium acet ate. pH 4 .8 . sto re at - 20 °C
EDTA, 0.5 mM
e Phosphate-buffered saline (PBSlb (40 9 NaC!. 1 9 xcr, 5 .75 9 Na 2HP0 4 • 1 g KH2P0 4 • H20 to B fitr esl. the pH should be 7 .2
• 0 . .2 M PMSF in absolute ethanol or
isopropano l. sto re at - 20 °C e Buff er
Ab
(10 mM
Hepes -KOH .
pH 7.9", 1.5mM MgCI" 10mM KCI, O.5mM
orr-i
e Hela cells grown in suspension culture to 4-5 x 105 cells/ml in standard tissue culture med ium (e.q . Jok lik's minimal essential med ium) plus 5 % newborn calf serum
• Buffer C b (25 % glycerol, 20 mM Hepes-
2.3 Preparation of nuclear extracts for 3' end-processing Because the substrate for most RNA processing assays is a small quantity of radiolabelled RNA, RNases pose a major problem. In practice. the high abu ndance of RNases in extracts of many tissues limits the choice of starting material for extract preparation to cultured cell lines. Nuclear extracts from HeLa cells are the stan dar d mate rial for the analysis of RNA processing in vitro because they are low in endogenous RNases and can be readily prepar ed . Th e following procedure (Prot oco! 1) uses freshly-harvested HeLa cells for the preparation of nuclear extract, but it also works with cells that have been frozen. However, fresh cells swell better in the hypotonic buffer and thu s can be more easily lysed. The use of a Doun ce homogenizer for cell lysis requires a certain minimal volume (approxi mate ly 2 ml for a small homogenizer). If small amo unts of cells are to be processed, they can be taken up in a syringe and forced repeatedly th rough a 25-gauge needle (7). Alternatively. cells are not broken mecha nically but lysed by the addit ion of lysolecithin (8). Usua lly, the nuclear extract is dialysed before use since the salt concentration limits the amount of non-dialysed extract that can be added to a processing reaction . However, the non-di alysed extract seems to be more active in the cleavage assay (see Protocol 6 and Section 2.5.3). Note that the buffers used to prepare the nuclear extract (see Protocol 1) contain 1.5 mM Mg2 + . whereas dialysis buffer (bu ffer D; see Protocol 1) does not. The amount of Mg2 + added to in vitro reactions may thus have to be adjusted. Alth ough Protocol 1 was developed for use with HeLa cells, extracts from ot her cell lines can generally be made by the same procedure . However, the protocol cannot be applied to the preparation of nuclear extract from solid tissue (e.r. calf thymus, see Section 3.2.3). 4
KOH, pH 7.g e , 1.5 mM MgCI2 • 0 . .2 mM EDTA. 20 mM NaCI, 0 .5 mM DTTd 0. 5 mM PMS F· )
orr-i
,
Method 1. Ha rv ~ st the HeLa suspension cells by centrifugation at 13009max for 10 min at room tempe rature. 2 . Resuspend the cells in a small volume of PBS and transfer the suspension to graduated Falcon tubes. Centrifuge as in step 1. 3 . Decant the supernatant and determine the packed cell volume (PCVI. Resuspend the cells in PBS to 5 x PCV and centrifuge as in step 1.
4 . Carry out this and all subsequent steps on ice . Resuspend the cells in icecold buffer A to 2 x PCV. Leave the suspension on ice for 10 min while the cells swe ll. 5 . H o m o ~ e n i z e the cells by 10 strokes in the Dounce homogenizer on ice, checking for complete cell lysis by phase-contrast microscopy'. 6 . Centrifuge the homogenate in graduated Falcon tubes for 15 min at 2000 x g m... 7 . Mea sure the total volume of the homogenate . Carefully take off the supernatant and measure its volume. Calculate the difference between total volume and supern atant volume. This difference is the packed nuclear vo lume IPNVI , 8 . Resuspend the nuclei in ice -cold buffer C t o 0 .5 x PNV , Slow ly add hig hsalt buffer C (0 .5 x PNV I drcpwise and with swirlingg . 9 . Homogenize the nuclei by 10 strokes in the Dounce homogenizer on ice. T r a n ~f e r the homogenate to a beaker in an ice bath and stir it gently for 30 min on a magnetic stirrer. 10 . Centrifuge the homogenate for 30 min at 16 000 X 9 max'
Continued 5
Elmar Wah le a nd Wa Iter Keller
1: 3' end-processing of m RNA Protocol 1 . Continued 11 . Rem ove the supernatant (nucl ear extract ). Dialyse (optiona l) the extrac t for 5 h agains t 21it res of buffer D . 12. Freeze t he nuclear extract in suitable aliquots in liquid nitrogen and sto re
them at - 70 DC. • A mod if ication of the original protocol of Dignam et al. (6 ) developed by Rob in Reed (Harv ard Medical School! and communicated to the authors by Ryszard Kole (University of Nort h Carolina). b All bu ffers ca n be stored f or several weeks in a refrig erat or. C A djust a 1.0 M st ock sol ution of this compound at room t emperature to the pH indicated . Us e t his stock t o m ak e up the buffer wit ho ut furthe r adj ustment of the p H. d Ad d OTT ju st prio r t o use f rom the 1.0 M stock so lu tion . e PMSF must be added dro pw ise wit h vigo rous stirring just prio r to use . f The number of st rokes suggested is mean t as a guide. Mo nito ring of cel l lysis by m ic roscopy is essen ti al. tI T he sus pension should become on ly slightly vis cous . If high -sal t buffer is added too fast , DNA will be released . making the ex tract too viscous for homogenization and centrifugation .
2.4 Formation of polyadenylation complexes 2 .4 .1 Introduction Th e endo nucleolytic cleavage at the site o f future poly(A) addition o nly occurs aft er the subs trate RNA ha s been assembled into a la rge co mplex th at co ntai ns so me o r all the trans-actin g proc essing facto rs . T he co m position o f th is complex is not known precisely since its for mation has so far been an alysed on ly in crude extracts. It is. ho wever, quite clear that binding of C PSF to th e AAUAAA seq uence is the key event for com plex fo rm atio n (9). O ne of the cleavage factors may interact with a dow nst rea m seq uence eleme nt a nd sta bilize the CPSF-AAUAAA interactio n (10) . It is not kn own whet her any of the other factors is stably associated with the process ing complex. A variety of experimental approaches can be made to investigate polyade nyla tion co mp lexes: (a) The formation of processing complexes may be detected by either glycerol gradient sedime ntation (I I) or, mor e co nveniently, no n-denatur ing gel electro phoresis (see Sect ion 2.4 .2). (b) RNA ca n be ext rac ted from the co mplexes . Analysis of th is RNA by denaturing gel electropho resis will revea l wh ether a co mplex co ntains substrates, intermedi ates, or products of th e reaction . (c) T he regio n of an RNA substra te associated with processing factors ca n be investigated by nucleas e digestion or chemical probi ng techniques (see Sections 2.4.3 a nd 2.4.4). 6
(d) P olyadenylatio n complexes separated by native gel electrophoresis may be analysed with respect to their protein content by sub sequent SDS pol yacrylamide gel electrophoresis. UV cross-linking to the substrate RNA can identify those proteins that are in close contact with the RNA (9). (e) Th e presence or absence of panicular antigens or nucleic acids (e.g. snRNAs) in the polyadenylarion co mplexes can be tested with blotting techniques (12).
2.4 ,2 Ba sic m ethod Protocol 1 describes the format ion of process ing (or polyadenylation) comp lexes using th e nucl ea r extract prepared in Protocol 1. Prot ocol 1 for ms th e bas is for man y of the subsequent procedu res. With slight modi fications, it can also be used to a nalyse th e bind ing of puri fied CPS F to AA UAA A-co ntain ing RNAs (see Sectio n 3.2.5) o r o ther RNA -protein interactions . Th e polyadenylation co mp lexes generated in Protocol 1 are visua lized by no ndenat uring (native) polyacrylamide gel electrophoresis . Complex fo rmation requ ires ATP and so creati ne phosphate is included in the reactio n to replenish the ATP pool that is depleted by ATPases in the nuclear extract. The extract contains adequate amou nts of creatine kinase. Polyvinyl alcohol (P VA) is included in the reaction because it stim ulates many macromolecular interactions, pro bab ly by increa sing the effective concentrations of the reacti ng molecu les ('excluded volume effe ct ' ). Aft er the polyad enylation com plexes have formed , heparin is added . Heparin is an anionic polymer that binds many nucleic acid-binding proteins and here serves to remove non-specific RNA-binding proteins fro m the precursor RNA. Protocol 2. Fo rma tion of pol yadenylation c o m plex e s a nd the ir an alysi s by e lec tro pho re s is in na t ive gel s
Equipm en t an d reagents e Equ i pm en t fo r polvacrvlamid e gel elect rophoresis
e l 0 mM MgCI 2
. 2.5 m g/ml tRNA . 12.5 % pol yv in y l alcoh ol (PV A; lo w
e Dialysed HeLa cell nu clear ex t ract (see Prot ocol 1)
mo lecu lar weight , col d water solub le; Sigma )
e 3.2P-labelled precursor RNA (approximate ly 500 - 10 0 0 c .p .m ./fmol ), sy nt hesized and capped in vitro {see Vol ume I. Chapter 1, Se ct ion 2 .2 1
e Heparin (2 5 mg /m l; Sigma . grade II e t.oadinq bu ff er 110 % glycerol , 0 .05 % xylene cvanol , 0 .05% bromophenol blue)
on
in 10 mM sodium acetate, e O.2 M pH 4 .8 , store at - 20 DC
e Nat ive gel electrophoresis buffer (2 5 mM Tris base . 25 mM boric acid , 1 mM
e 10 mM A TP. D issolve A TP in wa ter at app ro xim at ely 0 .1 M . Neutral ize with ammonia, checking t he pH w ith indicat or paper. Determi ne the ex act concentration by me asur in g the A 2 S9 (~259 = 15 4 0 0 ). Store t his st ock so lut ion at - 20 DC. D ilut e aliquots to 10 m M as required and stor e t hem also at - 20 DC
• 10 % am monium persulphate {freshly ma de )
e O.2 M creatine phosphate
e TEMED
EDTAJ' e Acrvlarrude stock solution (20% acrylamide , 0.25 % bisacrylam ide in nat ive gel electropho resis buffer )
7
Continue d
Elmor Wahl e a n d Wolt er Kell er
1: 3 ' end-proces sing of mR NA
Pro toc o l 2 . Continued
p rotocol 2 . Continued
7 . Transfer the rest of the reaction mi xture t o a 30 °C w at er bath. Tak e
Method
additional 25
1. Clean glass electrophoresis plates, space rs . and , most of all , the comb
a.
thoroughly. 2. Assemble the glass plates (20 em long ) w ith 1 mm thick spacers .
3 . Make up a 4 % pol yacrylam ide solution in nati ve gel electrophoresis buffer in a sufficient volume for the gel apparatus being u sed . For each
100 ml mix: • acry lami de stock solution
20 ml
• native gel electrophoresis buffer
80 ml
• TEM ED
0 .1 ml
3 . If t he RNA has alr eady been pr ec ipitated and stor ed in wa ter or 50 % etha nol, lyo ph iliz e th e appr opr iat e qu ant ity . 4 . To dissolve t he RNA pellet , add t he following reagents in the order given :
12 ~I 6 ~I 3 ~I 24~1 ~I
1 5 ~1
5. Vortex t o mi x and dissolve th e RNA. Place the tube on ice and add 60~1 of 12 .5 % PVA and 150 ~1 of nuclear extract ". Mix gentl y by pipetting . 6 . With draw a 25 1-11 aliquot into a fresh microcentrifuge t ube and freeze it on dry ice .
l
Con tin ued 8
C. Elec trop horesis of potvedenvtstion complexes 1 . Remov e the comb from the non-denarurinq po lya crylam ide ge l, assemb le t he apparatus and rin se the wells w ith native ge l electrophoresis
co mpatible with the size of the wells 15-30 ~ IJ . 4 . Mix an aliquot of substrate RNA , t hat has not been incubated with proteins, w ith loading buffer. Load this sample into an empty lane of the gele.
5. Run t he gel at 250V. The running time depends on th e lengt h of th e
'l , Prepare a reaction mixture sufficient t o enable 12 x 251-11 t im ed samples to be ta ken. To do this , dispense 120 fm ol of RNA pr ec urs or (equivalent to 10 ng of an RNA 250 nt long l into a m icr ocentrifuge tube . Calcu lat e th e vol u me of RNA sol ution to be use d fr om it s concent ratio n det erm ined as descr ib ed in Section 2.2. 2 . If th e RNA ha s be en stored as an et hanol pr ecipitate, ce nt rifuge it for , 5 min in a microcentrifuge at 4 ° C. Remo ve t he supernat ant as completely as possible sinc e it contains the urea from the gel that wa s u sed to purify the RNA Isee Volume I, Chapter 1, Section 2.3 .3) . Add 1 ml of 70 % ethanol to the RNA pellet, ce nt rifuge for 5 min , discard th e sup ernatant and dry the pellet . U se a han d -held Geig er c ount er t o control the recove ry of radioacti ve RNA in th is and all sub seq uent st eps. Proceed t o ste p 4 .
30
At t he end of t he incubation , take all the tubes fro m the dry ice . Add 51-11 of 25mg /ml heparin t o each and leave them for 10min at room t em perat ure.
3 . Load aliquots of all the samples into the wells of the gel. Use aliqu ot s
B. Preparation of the polyadenylation complexes
2.5 mg/ml tRNA H, O 0.2 M DTT 10 mM ATP 0.2 M crea tine phosp hate 10 mM MgCl,
aliquots after 1, 2, 5, 10, 20 , 30 , 40, 50, 60 , and 120 min.
buffer. 2. Prerun the gel for at least 1 h at 250 V.
. 10 % ammonium pe rsulphate a .75m! 4 . pour the acry lamide solution between the glass plates, insert a well -forming comb , and leave the gel to polymerize .
• . . . . .
~I
Freeze each on dry ice .
A . Preparation of the non-denaturing polyacrylamide gel
pr ecu rsor RNA ; for an RNA of 250 nt, run th e xyl ene cya no l marker dy e in the lan e cont aining fr ee sub strate RNA t o the bottom of the ge l. 6 . Remov e the ge l from the electrophoresis apparatu s and separate the glass plates. 7 . W ith the gel on one of the plates, wrap it in pla stic cling fil m le.g. Sar an wrap). Place an X- ray film in contact with the gel and u se a ma rking pen t o ma rk the orientation of the gel on th e film .
a.
Exp ose the film at - 70 " C for an appropriate tim e and then develop it according to the manufacturers instru ctions.
• Note that this buffer is different fr om the TBE buffer norm ally used fo r gel elect ro phores is . b Less nu cle ar extr act may be used . If so, add bu ffer D (see Protoco l 1) so tha t the combin ed volume of nu c lear extract and buffer D is one ha lf of the total react ion vo lume . C loading buffer is added only to the free substrate RNA but not to the aliquots from the react ion . These conta in sufficient glycerol f or loading .
The experiment sho wn in Figure JA dem on st rate s th e tim e cours e o f pol yaden ylation com plex formation . The fastest migrating complex (complex H) repre sent s the non-spec ific binding of proteins to the RNA . It s fo rmatio n d oes not requ ire any of the polyadenylation signal s. A seco nd complex (S), th e specific polyadenylatio n complex, forms o nly with RNAs th at contain all th e seq uences necessary fo r 3 ' end-processing. In t he presen ce of ATP, thi s co mplex fo rms rap idly even at 0 "C . RNA that was at firs t bound by non-specific co mpo nents ~complex H) is al so incorpo rat ed into co mplex S. Later, a th ird complex (P) pp~ars that co nt ains the cleaved and pol yadenylated RNA product (see SectIon 2.5 .2). g
Elmo r Wahle a nd Wa lter Keller
1: 3' end-processing of mRNA
A ..... x
Z
I 0 1 2
s
10 20 )0 1.0 50 60 120 m in ~... . - origin
. ..- . .. _ ......... "" ...
.......
.......... - s ••• - p
iH
•
- pre
B
This nuclease cleaves RNA on the 3 ' -side of GMP residues, pro ducing a pattern of oligonucleotides that is characteristic for every RNA and tha t can be predicted from its sequence. In Proto col 3, the polyad enylation co mplexes are tr eated with a low concentration of RNa se Tl and then separated on a non-denaturing gel. After the polyad en ylation compl exes ha ve been isolated from the gel, they are deproteinized and their prot ected RNA fragment s are further an alysed by denaturing polyacr ylamide gel electrophore sis and auto radiography. In the final stage o f th e procedure, each o f these fragments is isolated from the denaturing gel a nd subjected to a second RNase TI digestion and denaturing gel electrophore sis. RNase Tl cleavage sites tha t were pr otected during the first digestion will be accessible during the secon d digestion. A co mpa riso n of the oligonucleotides generat ed from the protected fragments during the second digestion to the pattern of oligonucleoti des generated fro m the nak ed substrate RNA will identify those sites that were protected during the first digestion and thu s the region s of the RNA involved in polyadenylation com plex formation. The polyadenylation co mplex is generated as describ ed in Protocol 2. However, fo r the efficient detecti on of RNA fragments a fter the vari ou s RNase digestions, the RNA subst rate mu st have a higher specific radioactivity.
\potylAI pre
Protocol 3 . Analysis of po lyadenylation c o mplex form ati on by RNase T1 protection Equipm ent and reagents
· 1 Time course of the format ion of po!yade nyla tion com plexes and poly adenyla ted Frqure . . 12 RNA (246 ntl RNA . Polyadenylatio n complexes were formed accord ing to Protoco on an conta ining the adenovirus L3 polyadenylation site . (A) Complexes were analysed on a non -denaturing 4 % polyacrylamide gel. For an explanation of the complexes formed. see the
e Non -denaturing polyac rylamide gel and electropho resis equipment (see Protocol2)
and analysed on a 6 % denaturing gel. Reproduced from ref. 13 .
e T w o denaturing polyacr ylamide (sequencing) gels' , one with 12 % ecrvl amid e and the other with 20% , TBE running buffer, and formamide loading buff er (see Volume L Chap ter "
text. Precursor RNA is indicated by 'pre' . A controllan~ with RNA in the absence of n~clear extract is labelled '.NX T ' . tBI RNA was purified from attquots of the same samples as In (A I
In the pro cedu re described in Prot ocol 2, RNA can be pur ified fro m a second aliquot of each sample and analysed on a denaturing gel so that a um e course of the processing reaction itself is obtained (for a detailed d~ scusslon, ~ee Section 2.5.2 and Protocol 6) , Although complex formation IS very rapi d , polyadenylated product appe ars onl y after a significant lag period (compare Figures IA and IB). The RNA present in each o f the vanous complexes can be anal ysed as described (ref. 13; see also Section 2.4 .3, Prot ocol 3).
2.4 .3 Analysis of polyadenylation complexes by RNase protection
•
Prot ocol 81
I
P·labelled RNA
substrate
(20 000
extract . and reagents for generating the polyadenylation comp lex (see
Protoco l 2 ) eTSE buffer (25 mM Tris base. 25 mM boric acid, 1 mM EDTA) conta ining 0.1 % sos e r E buff er 110mM Tris- He l. pH7.5 .
e RNase T1 (Calbiochem). Make a stock solution (50 units/lJll in 10 m M Tris-HCI, pH 7 .5 ,1 mM EDTA and dilute this in the sam e buff er as required e Proteinase K 110 mg /ml ; Boehringer M annheim ) e RNA elut ion buffer 10 . 7 5 M amm onium acetate , 10 mM magnesium acetat e, , % (v/v ) phenol, 0 .1 % SOS, 25IJ g /mi tRNAI
RNA sequences involved in the formation o f the polyad enylatio n complex are pro tected by the binding of processing factors against digestIOn WIth RNase Tl , 10
32
c.p.m.ztmon". dialysed Hela cell nuclear
1 mM EDTAl containing 2.5 mg/ml tRNA and 1 unit /~l RNase T 1
e Phenol .chloroforrn
(1 : 1)
. 32P·label1 ed DNA electrophoresis size markers covering the range between 20 and 100 nt Ie.q. pBR322 digested w ith Hpall and end -labelled with I )_ 3 2 p J A TP and polynucleotide kinase ; see Vo lume I, Chapter 2, Sect ion 3 .2 .2 )
Co n tinue d 11
1: 3' end-p rocessing
Elmor \Vahl e a nd WaIter Keller
of mRNA
Prot oco l 3 . Continued
Protocol 3 . Continued
16 . Incubat e eac h gel fragm ent separately overnigh t at 3 7 °C in 4 0 0 IJI of
Method 1. Set up a 25 1.11 polyadenylation react ion as in Protocol 28 . RTNoAd~ ~hi S . dispense app roximate ly 10 7 c .p. rn. of labelled precursor In 0 a microcentrifuge t ube. 2 . Depending on the way the RNA has been sto~ed . c.ent rifuge or lyophilize the RNA in a microcentrifuge tube as described In Protocol 28 . steps 2 or 3 . In this and all subsequent steps, control the reco very of RNA w ith a hand -held Geiger counter.
3. To the RNA sample Istep 21 add:
. 2.5 mg/ml tRNA 1 ~I • 10 mM A TP 2 ~I . 0 .2 M creatine phosphate 2.5 ~I . 70 mM OTT 0 .75 ~1 . 10 mM MgCl , 1 .25~1 4 . Vortex to dissolve the RNA. PLace the tube on ice and add 5 ~I of 12. 5% PVA and 12.5 ~ I HeLa cell nuclear extract. o
5 . M ix gen tly by pipetting and incubat e for 30 min at 30 e. 6 . Add 0 .1 unit of RNase Tl and incub at e the mixture for 10 m in at room
RNA elution buffer. 17 . Remove the buffer from th e gel fragments into a fre sh m icrocentrifuge tube . 18 . Extra ct th e eluted RNA by vortexing w ith an equal vol. of phenol :chloroform. 19 . Centri fug e in a mi cro centrifuge f or 30 sec t o separate th e ph ases. 20 . Remove th e upper (aqueous) phase into a fr esh microcentrifuge tu be. Add 2 . 5 vol. of eth anol , m ix, and centrifug e f or 15 m in in a microcent rif ug e. 2 1. D iscard the sup ern at ant . Add 1 ml of 70 % et hanol t o the t ube and centr ifu ge it again fo r 5 m in . Discard t he su pern ata nt and dr y the pelle t un der v acuum . 22. Dissolve the RNA oligonucleotide in 41J1 of TE buffer conta ining 2.5 mg /ml tR NA and 1 unit /nl of RNase T1. Sim ilarly set up a sample of the orig inal RNA substrate for RNase T 1 digestion as a cont rol.
23 . Digest the samples for 30 min at 30 °C. 24 . Lyophilize each sample and then di ssol ve t hem in f orm am ide load ing buffer. Boil eac h for 3 min and chill on ice . 25 . Load the sam ples on to the 20% denaturing polya cr ylamide gel. Run th e gel until th e bromophenol blue m arker is 10 cm fro m th e bott om.
26 . Autoradiograph the gel.
temperature . 7 . Add 25 mg /ml heparin t o 0 .2 mg /ml final concent ration. 8. Incubate for 10 mi n at room temperature then load an aliquot (~t least 5 ~ I) or all of the sample onto t he non -dena turing polyacrylam Ide gel. Separate the com plex by electrop horesis as in Pro t ocol 2C . 9 . After electrophoresis, w rap th e gel in cling film . Place an X -ray fil m .on the gel and use a fib re-tip pen to apply several guide m arks ext~ndmg from t he X -ray film t o the cling f ilm . Expose and develop t he film . 10 . Place t he gel on top of the autoradiograph and u~e the mar~s to align the film . Us ing the autorad iographic im age as a gUl~e.' and using a c1~an sc alpel, cut out a gel piece co nt aini ng the spec if ic polvadenvlation
complex".
• The 1 2 % gel should be of t he same th ickn ess as t he nat ive gel (1 mm l so that gel fragments fr om the nativ e gel ca n be placed on t op of the den atur ing gel. The 20 % gel shou ld be a norm al (th in ) sequenc ing gel. b Make RNA of t he req uired spec if ic act iv it y by usin g the appropr iate speci f ic act ivi ty of [ o:_ 3 2 p j UTP in th e transcription reaction . Obv iou sly , the inte nsity of the au to radiogr aphic band s in the gels (ste ps 9, 15 , and 26 ) depend s on the number of UMP residues in eac h RNase T1 fragment. Fragments cont aining no UMP resid ue s w ill not be visi bl e. It m ay t hus be more conveni ent to label the RNA w ith ( o:-32 PJGTP. Since ever y fragme nt contains only a sing le GMP residu e at its 3 ' -end . this will result in uniform label ling in te ns ity of all fragmen ts . /;The RNase T 1 digest ion m ay cha nge the po sition o f the specif ic co m pl exes in the gel. Non-specif ic complexe s may d isappear.
11. Soak th e gel piece in TBE buffer cont~ining 0.1 %. SDS foro1~ min at room t emperature, then incubate the gel shce for 30 min at 30 C In TBE buffer
containing 0.1% 50S and 0.5 mg/ml proteinase K. 12 . Using clean forceps , pla ce th e gel piece in a well of the 12 % denaturing
polyacrylamide gel. 13 . Mi x radioacti ve DNA size markers w ith f ormamide loading buffer. Heat the solution for 3 min to 95 °C , chill it on ice, and load it into one of th e wells of the denaturing gel. 14. Run the gel until the br om oph eno l blue mark er ha s reach ed th e bottom
of the gel. 15. Autorad iograph the gel and cu t out gel piec es cont aining the ma jor protected RNA fragment(sl as in st ep 10.
Continued 12
The RNase T1 digestion patt ern of an RNA can be predicted from its sequence and most of the larger digestion products can be identified unambiguou sly by their size. Thus, the protected RNA fragment obtained from the polyadenylation Complex can also be identified by the pattern o f oligonucleotides generat ed from it in the second RNa se T1 digestion . If a fragment obtained in this second digestion cannot be identified by its size alone, an additional digestion, e.g. with RNase T2, ma y be used to det erm ine its nucleotide composition (14). The size of the pro tected fragment before the second digestion gives an estimate of the total size of the bind ing site. Th is is significa nt because the distribution of GMP residues in the substrate RNA may not allow the identificati on of every oligonucleotide produ ced in the second digestion . 13
1; 3' end-processing of m RNA
A
Elmar Wohle and Walt er Kell er
c
B
2 .4 .4 Analysis of polyadenylation co m p lex formation by modification-interference
TI -
+
U
"
S-
L
1>J90-
I
-u
C -L
7"
67- _ _
NUL 21 - .
_ . - 21
15- e
1\-.
•
10 -
• •-10
19-.-.
H-
_ - 15
8- -
76- • • a-6 1-
pre - . 2
,
3
2 3
2 P otection of precursor RNA from RNase T1 d igestion by polvadenylati~n F' 19ure . r h RNA as In complex f orm at io n. Polyadenylation complexes were formed on t e same Figure 1 (20 m in incubationI and t reated wi t h RNase T1 (see Prot ocol 3 ). IA) T ~o such complexes , U (upper) and l (lowed , la ne 3 , w ere se para t ed on a non -den at uri ng 4 % poly acryl amide gel afte r RNase T1 digesti on (+) . The reason fo r the ap pe~rance of tw o specific com pl exes is unkno w n. Contro l co m plexe s (lane 2 ) w ere not Incu bat ed wit h RN ase T1 1-); 'p re ' indic at es pr ecursor not exp o sed to nuclear e~t ract . ' H' (lan.e. l1 repr esent s a non-specif ic complex (see Fig ure n. (B) RNA cliqonuclectides ":ere purified from complexes U and L and analysed on a denaturing 1 2% poly.acrylamld~ gel (CL The 67 nt oligonucleotide from U and l was eluted from t he gel, d igested with R .N ~ s e T 1 and analysed on a denatu ring 20 % polyacrylamide gel. The four fragments visible from U and L arise from both sides of the po lyadenylation site . N (lane 1) shows RNase T 1 digestion products f rom precursor RNA . The numbers indicat e sizes (nt) of RNase T1 fragments o r D NA size ma rkers . Reproduced from ref . 13 .
RNase protection yields only information about the segment of RNA that is stably bou nd by pro cessing facto rs. In contrast, modification-int erference ana lyses indicate which pa rticular nucleotid es are essential for the formation of processing complexes (15) . In thi s procedure, end-labe lled substrate RNA is either modified chemically at purine bases with diethylpyrocarbonale or pyrimidine bases are removed with hydrazine, As in nucleotide sequencing by the chemical method, the extent of modification must be very low, so tha t less than one base is modified per RNA molecule. T he modified RNA is used to form processing complexes. After non-denaturing gel electrophoresis, com plexed RNA is recovered from the gel. Both this RNA aad substrate RNA that has not been incubated with protein are then cleaved at the sites of chemical modificatio n and the cleavage pro ducts are separated on a sequencing (dena tur ing) gel. The non -complexed RNA provi des gel tracks containing purin e-specific or pyrimi dine-specific cleavage lad ders with which the RNA fro m the polyad enylat ion co mplex ca n be compared. If the modification of any particular nucleotide has interfered with complex formation , the ban d corresponding to this nucleotide will be missing in the ladders generated fro m the RNA sample isolated from the complex. The same procedure can be used to invest igate the binding of purified processing factors , such as CPSF to AAUAAA-contai ning RNAs. It can also be applied to stu dy the for mation of any other RNA - protein co mplexes. T he method is not even limited to experiments in which a complex fo rmed between proteins and the modified RNA is separated by native gel electro phoresis. Other methods (e.g. denaturi ng gel electroph oresis) ca n also be used to isolate the RNA that is accepted as a subst rate in the react ion of interest and to separate it from the RNA that is excluded. The procedu re for modification-interference (see Protocol S) uses an RNA precur sor labelled at its 3 ' -end by a ligation procedure (see Protocol 4).
Prot oc ol 4. 3 ' e nd-labelling of RNA substrates by lig at ion An esti mation of the total binding site size fro m the sum of these secondary oligonucleotides may, the refore, not be possible. F igure 2 sho ws an exa mple of RNase Tl prot ~ct io n. In the case of the adenovirus L3 polyadenylation site used in this expenment, a fragment of 67 nt , co ntaining the AA UAA A sequence as well as downstream ~eque nces , IS protected (Figure 28) . A secon d RNase Tl digestion after purification of the protected 67 nt oligonucleotide generates several R ~ A fragmen ts of between 6-21 nt which lie on both sides of the polyadenylation site (Figure 2C). (The nucleotide sequence of the subst rate RNA used in this experiment and its pattern of RNase T l digestions pro ducts can be fo und In ref. 13.) 14
Equipment and reagents . Oenat uring 6 % po ly acryla mi de gel , electroph oresis equipment, TB E buffer . and f orm amide loadin g buffer (V olum e I, Chapter 1, Protocol 8 )
e 10 x RNA ligase buffer (0. 5 M Tris- He !.
• Un labelled capped substrate RNA sy nthesized as described in V olume I. Chapt er 1, Protocol 4
e O. 2 5 mM ATP , d iluted fr om a st ock solution (see Protocol 2 )
J 2PjpCp
(New Eng land Nuclear or • [ S,. Amer sham ; 3000 Cilmmol l
e T4 RNA ligase (Pharmacial . 1.0 M OTT (see Protocol 1)
pH 7.9 . 0 .15 M M9C 1, . 33 mM DTT" )
e OMSO (Sigm a) e O. 1 mg /ml eSA (nuc lease free )
Continued 15
Elmar Wahle and Walt er Kell er
1: 3' end-proce ssing of mRNA Protocol 4.
Protocol 5 .
Continued
Continued
Preparation of RNA substrates
Method
3. Cool the mixture to 4 °C and add 10-20 units of RNA ligase .
1. Use precursor RNA as the substrate, synthesized without radioactive label as des cribed in V olum e I. Chapter 1, Section 2 .2, and 3 ' end-labelled as Protocol 4 . Mix tw o separate aliquots of thi s RNA (approx im ate ly 10 6 c.p .m . each) with 12 . 5 ~g tRNA as carrier. Add 0 .1 vol. of 7.5 M ammonium acetate and 2 .5 vol. of absolute ethanol. V ort ex and chill th e mixture on ice for a fe w minutes.
4 . Inc ubat e at 4 °C for at least 10 h . 5. Purify full -length labelled RNA by denaturing pol ya crylamide gel electro -
2 . Centrifuge the mi xture for 15 min in a mi cro centrifuge. Discard the supernatant. In thi s and all subsequent ste ps, co nt rol the reco ver y of RNA with a hand -held Geiger counter.
1. In a mi cro centrifuge tube, mix 2 ~I each of 0 .25 mM ATP , 0 .1 mg lml BSA , 10 x RNA ligase buffer, and DMSO plus 10 ~I of [5 ' _32P JpCp. 2 . Add 5-10 pmol of substrate RNA (equiv alent to 80 -160 ng of RNA 50 nt in length) in a volume of up to 2 ~1.
phoresis (see V olume I, Chapter 1, Protocol 8 ). a
3 . Add 1 ml of 70 % ethanol to the pellet. Centri f uge for 5 min in a mi cro centrifuge . Discard the supernatant and dr y the pellet under vacuum.
Ad d OTT ju st before us e fr om t he 1.0 M stock.
Purine m odificatio n 4. Dissolve one RNA sample in 200 1 mM EDTA .
~l
of 50 mM sodium acetate , pH 4 .5 ,
5 . Add 3 ~I of fresh DEPC and in cubate the RNA for 5 min at 90 °C w ith the lid of th e tube open .
Pro to col
5 . Analys is of polyadenylation c o m p le x formation by mod ifi c atio n - interfe ren ee
Equipment and reagents e O.3 M sodiu m acet at e, adju st ed to pH 3. 8 wi t h acetic ac id
• No n- denaturing polyacryl am ide gel and electropho re sis bu ff er (see Protocol 2l, 12% dena tu ring po lyacr y lam ide ge l (seq uenci ng gel) , TBE bu ffer, and fo rmamide loading buffe r (Volume I. Chapter 1. Pro tocol B) tog et he r with the approp riate elec t rophores is eq ui pme nt
e nrethvlov rocerbonate (OEPC) (Si g mal ~ e Anhydrous hyd razine (Eastman-Koda k]" containi ng 0 . 5 M NaC!, prepare this solution by dissolving solid NaCI in hydrazine
• Elect ro bloning apparat us fo r wet blotting (e.g. Bio -Rad or Hoef er) e OEAE t ransfer mem br ane {Schleicher and Schu ell, NA45l
12. Dissolv e the RNA in 10 ~I of w at er and sto re it at - 20 °C .
e TBE buffer containing 0.1 % SOS (see
Protocol3) to
e 50 mM sodium acetate , pH 4.5, , mM EOTA , use the 1 M sodiu m acetate stock to make this solution wi thout adjusting the pH
11 . Dissolve and repre cipitate the RNA pellet as in st ep 7.
Complex formation and electrophoresis
. 0 . 5 M NaOH
e 2 .5 mg /ml tRNA e , .O M sodium acetate , adjusted pH 4 .5 with acetic acid
9 . Dissolve th e ot her RNA sam ple (st eps 1-3 ) in 20 ~I of anh ydrou s hydrazine co nt aining 0 .5 M NaCI and inc ubat e it for 70 min at 4 °C .
e 1. 0 M aniline in 0 .3 M sodi um acetate , pH 3.8, use redistille d aniline"
. 10 mM EDTA. pH 7.6
e 7 .5 M ammonium acetate
8. Dissol ve the RNA in 10 ~I of water. Store it at - 20 °C . Pyrim idine modification
10 . Add 200 ~I of 0 .3 M sodi um acet ate , pH 3 .8, and 900 ~ I et hano l, th en collect th e RNA pre cipitate as in st eps 1 and 2.
• Phenol :ch loroform (' : 1)
e 3 ' end-labelled substrate RNA (see Protocol 4)
7 . Dis solv e the RNA pellet in 200 ~I of 0 .3 M sodium acetate , pH 3.8 . Precipitate the RNA by adding 600 ~I of ethanol. Colle ct th e RNA precipitate as in steps 1 and 2.
e , .O% SOS
e Butan -J -01
e Hel a cell nuclea r ext ract and reagents for generating polya denylation complexes (see Protocol 2)
6 . Add 75 ~I of 1 M sodium acetate , pH 4 .5 , and 900 ~I of ethanol. Coll ect the RNA precipitate as in st eps 1 and 2 .
e NA 4 5 elution bu ffe r (55 % forrnamide". ' .8 M sodium acetate , 2 mM EOTA , 0.2 % SOS\, use solid sodium acetate, not a solutio n whose pH has been adjusted
13. Set aside one -te nt h of eac h samp le of mod if ied RNA (st eps 8 and 12 1 as control samples f or lat er anilin e cleav age and elec t ro phoresis st eps. Use th e rest of eac h sam ple separately f or co m plex f ormation as des cribed in Protocol 28 but use a reaction vol ume of 125 ~1. 14. Load each sam ple on th e non -denatu ring gel using severa l adjace nt lanes to acc om modate t he enti re samp le. 15. Run th e gel as in Protocol 2C .
e Uree load ing bu ffer (8 .0 M urea , 20 mM Tris - HCI , pH 7.9 , , mM EOTA, 0.05 % xylene cya nol , 0.05% bromophenol blue)
16. After elect rop horesis, tran sf er the RNA e onto an NA45 DEAE memb rane by elect roblott ing in a w et blo tting ce ll. To do t his, cut out one piece of NA45 membrane slight ly lar ger than t he gel and soa k it f or 10 mi n in 10mM EDTA , pH 7 .6, at ro om temp erat ure.
Continued
Continued
16
17
1: 3' en d-processing
of mRNA
Elm ar Wahl e and W alter Kell er
P rotocol 5 . Continued
Protocol 5 . Continued 17 . Transfer the membrane into O.5M NaOH and let it soak for 5min. 19 . Cut out four pieces of Whatman 3MM paper of the same size as the NA45 memb rane and we t them in TBE buffer containing 0 .1 % 50S .
20 . Place the NA45 memb rane onto a stack of two pieces of 3MM paper. Place the gel directl y onto the DEA E membrane and finally t wo shee ts of 3MM paper on top . Be careful not to trap air bubbles. 21 . A ssemble this sandwich in the blott ing cell according to the manufactu rer's inst ructions as for Western blott ing. For tran sfer , use TBE
buffer containing 0.1 % 50S. Blot for 6 h in a coldroom at 50 V. 22 . Aft er the transfe r is complete, wrap the DEAE membrane in cling film to prevent it drying and expose it to X·ray film using guide marks to align the memb rane on the film (see Protocol 3, step 9l. 23. Cut out strips from the membrane corre sponding to the specific polyadenylation complex . 24. Elute the RNA from the DEAE membrane by soaking the membran e st rips in NA45 elution buffer for 1 5 min at 70 °C . 25. Remove the elution buffer from the me mbrane into fresh tub es. Add 12 .5 ~g carrier tRNA to each RNA sample and ext ract once wi th an equ al vo lume of phenol :chloroform. 26 . Add 2 .5 vol. of ethanol. Collect the RNA precipitate and wa sh it as in
steps 1-3. 27. Purify full -length RNA on a denaturing polya crylamide gel (see V olum e
I, Chapter 1. Protocol 81Cleavage of modified RNA 28 . Dissolve each of the modified RNAs from step s 27 in 20 ~I of 1 M aniline in 0 .3 M sodium acetate , pH 3.8 . Incubate for 2 0 min at 60 °C in the dark . Similarl y treat samples of the cont rol RNA s not used to form polyadenylati on complexes (st ep 13 ).
29 . Add 1.4 ml of butan-l -oJ. 30 . Vortex and reco ver th e RNA precipitate by centrif ugation in a microce ntrifuge for 15 min.
31 . Redissolve the RNA in 150 ~I of 1% 50S and repeat step s 29 and 30. 32. Wash the RNA pellets by adding 1 ml of 70 % ethanol and centrifuging in a microcentrifuge for 5 min .
33 . Dissolve the RNA in 5 ~I of urea loading buffer. 3 4 . Det ermine the amount of radioacti vity recovered by Cerenkov counting. 35 . D ilute aliquots of each sample (containing equal amounts of radioactivity ) wi th urea loading buff er, heat for 90 sec at 90 °C and load on a 12% denaturing polyacrylam ide gel (sequencing gell' . 36. Run th e gel until the bromophen ol blue marker is 15 cm from the bottom and detect the RNA by autor adiography .
Continued 18
• Store opened bottles of DEPC in a refrigerator for no more than one week. Store hvdrazin e in a closed container with some desiccating material in a fume hood at room temperature. C Redistill the aniline under nitrogen and storeit in aliquot s at - 20 " C in the dark. Be careful : aniline is toxic and combustible. d Deionize the formamide with a mixed-bed ion-exchange resin and store it in the dark at - 20 -c. • The extent to which proteins contained in the complex are transferred to the DEAE membrane together with the RNA is unknown. Any protein transferred to the membrane and eluted in steps 21 - 24 will be removed by the phenol:chloroform extraction in step 25. f If the sample in urea loading buffer is heated for too long. urea may precipitate. Redissolve it by adding a small amount of water.
tJ
18 . Wash the membrane in 5 changes of water (2 min each ).
A few experimental points should be note d with respect to the modification int erference procedure (Protocol 5) : (a) Th e procedure does not tolerate much non -specific degradation of RNA since this will generate band s in the sequence ladder that do not stem from aniline cleavage of modified nucleotides. Therefore, exercise care to avoid RNase contamination (see Volume I, Chapter I, Section 1. 1). The shorter a substrate RNA is the easier it is to use in this protocol since sho rt molecules are obviously less susceptible to non-specific degradation than long molecules. RNase inhibitor from placenta may also be used (see Pr otocol 6, footnote b) . (b) Th e RNA substrate used for modification -interference can also be 5 ' endlabelled with [-y-32PlATP and polynucleotide kinase instead of with RNA ligase. However, it has been reported that 5 ' -labelled RNA does not give clean sequencing ladders with pyrimidin e modifications (15). (c) If, after anal ysis, the extent o f modificat ion is found to be too high or too low, var y the incubation times for the modifi cation reaction s (see Protocol 5, step s 5 and 9). (d) In Protocol 5, the sequence ladders to which the RNA from the specific polyadenylat ion complex is compared are generated from substrate RNA th at has never been incubated with nuclear extract. These control ladd ers may also be generated from RNA isolated from non- specific complexes. For unknown reasons, cleavage at G residues is more efficient with RNA that has been incubated with extract and eluted from the non-d enaturing gel. The best cont rol is thu s to use both substrate RNA without incubation and RNA from non-spe cific complexes. Figure 3 shows the results of a modification-interference experiment that was ca rried out with puri fied CPSF rather than nuclear extract. The nucleotid e seq uence of the substrate RNA can be read in the outer lanes (uncomple xed 19
1: 3' end-processing of mRNA
Elmar Wahle and Walter Keller
A+G C+U
In a variation of Protocol 5, the modification-interference method can also be used to identify the nucleot ides involved in the cleavage and polyadenylation reactions that follow complex format ion. To do this, rather than isolating RNA from polyadenylation comp lexes separated in native gels, carry out the chosen processing reaction (Section 2.5) with the chemically mod ified RNA and then separat e the cleavage products or pa lyadenylated RNA from unreacted substrate RNA using denaturing gel electrophoresis (15). Treat these RNAs according to Protocol 5 to analyse which nucleotides are essential for processing. For obvious reasons, polyadenylated products and 5 ' -cleavage products can only be analysed with 5 ' end-labelled RNA which may not always be a good substrate (see abo ve).
~~
•• ---
.... -.i: _
11::1
~
• ..:1\_
_ c __ c .._
B:
-
R: _
~ 1!1
15-
.-.-
·0 '0
o
• _
2.5 Assay of cleavage and polyadenylation in nuclear extracts
u
_
c-
__
u
..
1 2 34 Figure 3. Modification-interference analysis of the binding of CPSF to an AAUAAA-containing
RNA. A ' prec leaved ' RNA 165 ntl . containing the upstream polyadenylation signal of the adenovirus L3 site and ending close to the natural po lyadenylation site . was modified at either pyrimid ine or pur ine residues (see Protocol 5) . Complexes were formed between the modified
RNA and CPSF purified from calf thymus (see Section 3.2.3; ref . 23). Complexed and free RNA were further analysed as described in Protocol 5. Lanes 1 and 4 contain uncomplexed ANA, lanes 2 and 3 contain RNA from the complex . All six bands corresponding to the AAUAAA sequence are strongly decreased in the RNA purified from the complex , indicating that the modification of any of these nucleotides interferes with CPSF binding . In contrast , cleavage of the three nucleotides immediately following the AAUAAA signal is enhanced in the complexed RNA . Reproduced from ref . 9 .
2.5 .1 An overview Endonucleolytic cleavage of the substrate RNA in vivo and in vitro is very rapidly followed by poly(A) addit ion. Thus, the cleaved interme dia te is not normally present in detectable quantities . However, there are two major reasons why it is important to find conditions whereby cleavage can be separated from polyadenylation in vitro . First, if one wants to study each of th e two reactions, in term s of their RNA sequence and protein factor requirements or their biochemical mechanism, one has to have separate assays for each of them . Second, ther e is no mechanistic depende nce of polyadenylation on the cleavage event and no strict req uirement for a particular distance between the AAUAAA sequence and the 3 ' -end that is to be polyadenylated . Thus , polyadenylation can occur independently of cleavage by simple AMP addition to the 3 ' -end of an RNA, even if this RNA is a perfect substrate for cleavage. Consequently, the fact that an RNA has been polyadenylated does not imply that cleavage has occurred . Therefore, one cannot use polyadenylation as a coupled assay for cleavage; a direct assay is necessary. Nevertheless, it is sometimes advantageous to study coupled cleavage and polyadenylation in vitro. This is described in the following section. Modifications to this basic method that allow the separate assay of cleavage and of polyadenylat ion as well as of non-specific polyadenylation are described in Sections 2.5.3,2.5.4, and 2.5 .5, respectively.
2.5.2 Coupled cleavage and pol yadenylation
RNA) . The inner lanes contain RNA that was isolated from the comple x and show that each of the bands corresponding to the AAUAAA sequences is strongly decreased, indicating that the modification of any of these six nucleotides inhibits the binding of CPSF. No other band is decreased and thus no other nucleotide is critical for binding .
Proto col 6 describes how cleavage and polyadenylation are assayed in a nuclear extr act. Alternatively, purified or partially purified processing factors can be used in place of the nuclear extract. In these cases, include comme rcial creatine kinase at 10 ug/rnl in the reactio n. This may be necessary to maintain the ATP concentration since the endogenous enzyme of the extract may be remo ved during the course of protein purification, while ATPases are likely to contaminate partially purified factors. Also include Nonidet P-40 (0.01 '10)
20
21
Elmar Wahle and WaIter Keller
1; 3' en d-processing of mR NA
a nd RNase-free BSA (0.2-0.5 mg/ml) to stabi lize pur ified facto rs which
Protocol 6 . Continued
are used at low protein co ncen trations. A large white interphase after pheno l:chloroform extraction is due to PYA . If it is too large to perm it eff icient separation of the phases , increase the volumes of the aqueous phase and of the phenol:chloroform m ixture and increase the centr ifugation t ime. dWhen the protein content of the reaction mixture is low . the extraction can be omitted and the RNA can be precipitated directly with ethanol. Th is has to be tested under the particular conditions used. In this case it is import ant tha t not more t han 2.5 vo l. of ethanol are used for precipitation of the RNA . Otherwise . PYA w ill also precipitat e. ~ Choose the acrylamide content of the gel and the running time accord ing to t he size of the precursor RNA and the expected products (see Figure 4 for an example). C
Protocol 6. Coupl ed clea vage and polyad en ylat ion in nuclear ex t ract Equ ipment an d reag ents • HeLa cell nuclear extract, radiolabelled s ubstrate RNA ( lOO-lODOe.p.m .! fmol) and reagents for formation of polvadenvlaticn complexes as in Protocol 2
• Phenol:chloroform (1 : 1 J and reagents for ethanol precipitation of RNA (see Protocol 5) • Denaturing
polyacrylamide /urea
gel ,
formamide loading buffer, and T BE buffe r
(see Volume I, Chapter 1, Protocol 8) , approp riate electropho resis plus equipment
• Proteinase K (10 mg /ml in water)
. 2 x proteinase K digestion buffer (0.2 M Tris-Hel, pH 7 .5,25 mM EDTA, 0.3 M NaGI. 2% SDS)
• RNase inhibitor from placenta (Pharmacia
' RNAguard')
M eth od 1. Using 10-60 fmol of substrate RNA , set up a 2 5 ~ 1 processing reaction as in Pro tocol 3, steps 1-4". Include 4 -5 units of RNase inhibitor to minimize RNA degradation Io pt iona ll". 2. Incubate the mixtu re for at least 1 h at 30 O( to allow cleavage and polyadenylation to occur. 3. Te rminat e the reaction by the addition of 100 ~ I of 2 x proteinase K
digest ion buffer, 2 ~ I of 10 mg/ml prote inase K. 2 ~ I of 2.5 mg/ml tRNA and 711J1 of water. Incubate for 30 min at 30 "C . 4 . Extract the mixture once wi th an equal volume of ohenol.chtorofo rrn-".
5. Precipitate the RNA by the addition of 500 ~I of et hanol. Collect the precipitate by cent rifugation, w ash it with 70% ethanol, and dry it (see Pro tocol 5 , steps 1- 3 ). During these steps, monito r the recovery of RNA w ith a hand-held Geiger counter. 6. Dissolve the RNA in 3 1J1 of formamide loading buffer, boil it . and chill it on ice. Load the sample on a denaturing polyacry lamide gele.
7. Run and auto radiograph the gel to detect the RNA Isee Protocol3i. If the nuclear ext ract is replaced by purified factors, or if smaller volumes of ext ract are used , the incubation volume should be made up to one-half of the total reaction volume with buffer D (see Protocol 11 . If either the nuclear ext ract or the purified factor samples contain higher concentrations of KCI. buffer D lacking KCI can be used to make up the volume . The final KC' concent ration in the processing reaction should not be much higher than 50 mM. Note that the amount of Mg 2 + introduced into the reaction with nuclear extract or protein f ractions can vary (see Section 2 .3) so some adjustment may be necessary . b Ribonuclease inhibito r from placenta inactivates RNases by forming a tight complex with them. T his activity depends on the presence of OTT. Make sure that the OTT is not too old and has been stored under appropriate condit ions (see Protocol 1 I. If the inhibitor is inactivated , it will release any bound RNases which may then degrade the substrate RNA .
8
Continued
22
2. 5,3 Cleavage of RNA substrates in vitro Two different modifications of the coupled reactio n (see Protocol 6) are widely used to block polyadenylation and th us allow the cleavage reaction itself to be studied. One method is to omit Mg2+ (except the Mg? + introduced with the nuclear extra ct) and to include 1 mM EDT A in the reaction. This inhibits poly(A) polyme rase but permits cleavage . Not all precursor RNAs are processed well under these conditions . EDT A inhi bits nucleases in the extract so that the downstream cleavage product may also be detected. The othe r method is to replace ATP with one of two different classes of ana logues. Chai n-terminat ing analogues lack a 3 ' -hydroxyl group. Thus, only a single nucleotide can be inco rpo rated at the po ly(A) site so that the cleaved RNA can be detected. ATP analogues with a non-hydrolysable ,,-~ bond [e.g . (,, - ~ methylene)A TP ] cannot be used as substrates for po lyadenylation so that the 5 ' -cleavage product is not extended at all. ATP analogues are easier to use than EDTA . The preferred compound is 3 '-deoxy-ATP (cordycepin triphosphate). Thus, to examine the 3 ' -cleavage of RNA , replace ATP in Protocol 6, step 1, with 0. 5 mM 3 ' -dATP. However, note that the nuclear extra ct may still contain ATP, especially if it has not been dialysed . This endo genous ATP competes with 3 ' -dATP so that some elongation may still be observed. The downstream product of RNA cleavage is rapidly degraded and canno t be detected. The cleavage reaction must be incubated at 30 °C; it does not work at 37 "C.
2,5 ,4 Polyadenylation of RNA substrates in vitro Polyadenylation can be measured independently of cleavage simply through the Use of a 'precleaved ' substrate RNA that ends at. or close to . the natural poly(A) site, downstream of the AAUAAA sequence. For ' precleaved' RNA, there is no strict requirement either for the distance between AAUAAA and the 3 ' -end or for the sequence at the end itself. Use any con venient restriction site in the recombinant plasmid close to the AAUAAA
sequen ce to prepare the template for run-off transcription (see Volume 1, 23
1: 3' end-processing
of mRNA
Chapter I, Section 2.1). The natural distance between AAUAAA and the poly(A) site is 15- 30 nt. The method for assaying polyadenylation in the absence of cleavage follows Protoco l 6 except that a ' precleaved' substrate RNA is used. ATP is included as the substrate for poly(A) polymerase. Polyadenylation occurs at least as efficiently at 37 °C as at 30 "C . An incubation time of 30 min is usually sufficient. The specific polyadenylation reaction is critically dependent on Mg" ions. Lower concentrations of Mg' + (and also of KCI) may stimulate polyadenylation but, at least with purifie d factors, they also reduce the specificity (i.e. AAUAAA-dependence) of the reaction. The use of a substrate RNA with a point mutation in AAUAAA is an important control. Dialysed and non-dialysed extracts differ in their Mg' + concent ration (see Section 2.3). In any case, the Mg'+ content of the extract is unlikely to be known accurate ly. A titration of Mg' + with a wild-type and a mutant substrate may thus be useful.
2.5 ,5 Non-specific pol yadenylation As discussed in Section 2.5.4, the dependence of polyadenylation on the AAUAAA sequence in the substra te RNA is influenced by the reactio n conditions, especially by the presence of Mg' +. If Mg2+ is substituted by Mn' +, the specificity is lost and poly(A) polymerase can polyadenylate any RNA , independent of the AAUAAA sequence and of CPSF (Section 2.2). To carry o ut this non -specific reactio n, replace MgCl, in Protocol 6, step I, with an amount of MnCI, equimolar to the amount of ATP used (0.& mM). As a substrate, use an RNA with a point mutation in the AAUAAA sequence or any other RNA that does not contain the polyadeny lation signal. The non-s pecific polyadeny latio n reaction is useful for checking the quality of RNA 3' -ends (see Section 2.4. 4) and as an assay for poly(A) polymerase. Protocol 8 (Section 3.2.4) provides a modification of the non-specific polyadeny lation reaction that is particula rly useful for the latter purpose .
2.5.6 Interpretati on of the data
Figure 4 . Cleavage and polyadenylation assays in Hel a cell nuclear extract. fA) A polyadenylation assay was carried out in nuclear extract with the precleeved RNA described in Figure 3. Lane 1, precleaved RNA (pre) was incubated without nuclear extrac t. Lane 2, RNA afte r incubation with extract . Products, PolylAI. were analysed on a 12 % denaturing gel. No size markers were included in this gel. lB} A cleavage assay was carried out in Het.a cell nuclear extract in the presence of ddA TP with the L3 ANA described in Figure 1. Lane 1, incubation without extract . Lane 2, incubation with extract . RNA was analysed on a 6 % denaturing gel. 'Pre' indicates the RNA substrate. The arrowhead indicates the upstream cleavage product. Lane M contains DNA size markers with the sizes [n t} indicated on the left .
as described in Sections 2.5.3 and 2.5.4. Since the poly(A) tails of mRNAs are not of uniform size, but show a size distribution around a mean of approximately 200 nt, poly(A) + RNA forms a smear in the gel (Figure 4A) , migrating more slowly than the precursor RNA. The appearance of the polyadenylated product in a coupled cleavage/polyadenyiation assay (Figure 1B) is the same as in a polyadenylation assay with a precleaved substrate RNA (Figure 4A). However, as pointed out in Section 2.5.1, the poly(A) + RNA generated in a coupled assay may contain a mixture of molecules that have been correctly cleaved and then polyadenylated and others that have received
polY(A) at their original 3' -end without having been cleaved. Note that the latter reaction is also dependent on the AAUAAA sequence in the substrate RNA. It is thu s different from the non -specific polyadenylation reaction described in Sections 2.5.5 and 3.2.4. The cleavage reaction shown in Figure 4B was performed with dideoxy-ATP (ddATP) as the chain terminating analogue. (Note, however, that3 '-dATP is more efficient than ddATP.) Since the cleavage product migrates faster than the SUbstrate among all the fragments generated by non-specific nuclease action, the specificity of the cleavage reaction has to be verified not just by the size of the product but also by its dependence on the correct processing signal sequences. As a control for the cleavage reaction, use a precursor RNA with a point mutation in the AAUAAA sequence.
24
25
Figure 4 shows typical data from an experiment in which the polyadenylation reaction (Figure 4A) and the cleavage reaction (Figure 4B) were studied separately
1: 3 ' end-processing of mRNA
3. Polyad enylation and 3'-cleavage factors 3.1 Overview of processing factors The reac tions of 3 ' -cleavag e and po lyadenylatio n are carried out by at least five. possibly mor e, facto rs (see Section I ). Of the two or more cleavage factors, only one, a hete rot rimer called CStF or CF I , ha s been pur ified to homogeneity (16 , 17), and cDN As encoding one of its subunits have been cloned (18). Th e cleavage factors ca n be assayed only by recon stitut ion of the entire cleavage reaction (see Sectio n 2.5.3) since no partial reactions have been identified that are catalysed by these co mponents . Poly( A) polymerase has been purified to ho mogeneity (19, 20) an d cDNAs encoding the enzyme have been cloned (21, 22). Pol y(A) po lymerase is most easily assayed by the non-specific polyadenylation reaction as described in Section 3.2.4. The specificity factor CPSF has been puri fied to hom ogeneity (23). It ca n be assay ed by th e specific polyaden ylation assay when it is supp lemented with pu rified poly(A) polymerase or by specific binding to AA UAAA-cont aining RNAs (Section 3.2.5) . A poly(A) binding protein that causes poly(A) polymerase to elongate oligo(A) tails rapidly has also been purified to hom ogeneity (24; E. Wa hle, unpublished data). This factor can be assayed by its activity in stimulating specific polyadenylation or by its specific poly(A) binding activity. For details, consult ref. 24. A detailed description of protein purification procedures is beyond the scope of this article. However , Section 3.2 gives some general advice and describes methods for separa ting poly(A) polymerase from CPSF and for assaying these two facto rs duri ng purification .
3.2 Purification of processing factors 3.2 .1 General considerations Befor e a protein purification is undenaken, some consideration should be given to the choice of starting material. The almost exclusive use of HeLa cell nuclear extract for RNA processing experiments does not reflect a panicular wealth of processing factors in HeLa cells, but merely the fact that their low level of nucleases permits the pro cessing reactions to be readily observe d. For pur ification of processi ng facto rs, HeLa cells ha ve some disadvantages, on e of which is cost. If a protein has to be purified 10OOO-fo id with a yield of 10"10, 5 g of extract pro tein have to be processed to yield 50 ug of homogeneous materia l. This correspo nds to the nuclear extract fro m several hundred grams of H eLa cells. Mu ch more than this is consumed du ring the development o f the purification procedur e. These quantities may go beyond the financial capacities of many laboratories. Also, HeLa cell nuclear extract does not appe ar to be particularly rich in 3' -pro cessing factors . Both po ly(A) polymerase an d CPSF are as much as ten -fold more abundant in calf thymus extract than in nuclear extract from the same amount of HeLa cells. Thus, whereas I kg of 26
Elmar Wahl e and Walt er Keller cal f thymus yields approximately 0.5 rng of poly(A) polymerase (19 20) I 50 ug wo uld be expected from J kg of HeLa cells Even this I'S an' t.' on .y . . h . op Iffil StiC estl~.at e. smce t e ~m~ 1l quantities present in He La cells wou ld make th purificati o n more difficu lt . e If, despite these potential probl ems, HeLa cells are chosen as the so urce for the puri ficat ion of protein factors, one should remember that the sta d d nuclear extract preparation (see Protocol l) has not b " df n ar . " een opn rruze or the factor of mt eresl: Significant quantities of this particular factor may have leaked from the nu clei mto the cytoplasm during cell lysis or the salt concentration may not be opti mal for ItS extraction from the nuclei.
3.2.2 Separation of poly(A] pol ymerase and CPSF from a H L II nuclear extract e a ce Protocol Zdescribes the sepa rat ion of poly(A) po lymerase fro m C PSF by DEAE colum? chrom atograph y o f HeLa cell nucl ea r extract. Poly(A) polyme rase is ~OU~d m the flow-th rough along with at least one cleavage facto r, whereas CPSF m s to an d IS eluted fro m the column alon g with at least one ot her cleavage factor. Assays for these factors are given in Sections 3 2 3 and 3 2 4 Obvi I h frac ti . . . .. VIOUS Y t e same racnons may also be assayed for cleavage activity (see Section 2.5.3):
Protocol 7. Frac t ionat ion of HeLa cell nuclea r ex t rac t by DEAE c hroma tog raphy Equip m ent and reag ents • HeLa cell nuclear extract (see Protocol 1}1
• 1.0 M OTT (see Protoc ol 11
e Chrom atography c ol um n"
• Starting buffer (50 mM Tris- He l pH 7 9
• Perist alt ic pump, fra ction co llecto r (optionafj s, and c onduc t ivi ty met er
e OEAE chrom ato gr aphy matrix [ DEA ESepbarose fPharm acia) or equ ivalent ]
7S".:M KCI, O.SmM EDTA. ·O.S "; M DTT • 10 % glycerol) • Starting buffer w ithout KCI
• Elutio~ . bu ffe r las starting bu ff er. but cont aining 0 .5 M KCIl
M ethod 1. Make a slurry of DEAE S h . . Equilibra h ~ ep a~ose In starting buffer and pour the colu mn . ~~ t e cO I~ m n ~ n star t ing buffer unt il the con duc tivity and pH of th e was lngs are Identical to those of the starting butters.
;:tc~sethteo coOI~Cmn in a c~ldroom and carry out the
rest of the procedure as possible. . . 3 . Dilut e the Hela cell nu I KCl until it s cond tivi c ear extra ct With Ice-cold starting buffer w ithout p uc tvrtv equals that of the starting buffer', 2.
4 . Load the dil~ted nuclear extract onto the colum n at about 1 column vol /h a approximately 10 mg of protein/ml colum n vel. . .
Con tinued 27
1: 3' end-processing of mR NA Protocol 7. Con tinued 5 . Collect the flow -through , excluding as much as possible of the buffer preceding the protein peak. A fraction collector is helpful but not necessa ry.
6 . Wash the column w ith one volume of starting buffer and combine the wash wi th the flow -thr ough. Precipitat e the proteins wi th ammo nium
sulpha te (step 91. 7 . Elute the column with elution buffer . The beginning of the salt step can be identified by the increase in conductivity or in protein concentration. Direct activity assays may not be possible at this ste~ .
8 . Dialyse the eluate against start ing buffer . 9. To the combined flow -through and washings add solid ammonium sulphate to 80% saturation 10 . 5 16 9 of ammonium sulphate per mil . Stir the mixture on ice until the salt has dissolved and then for at least one additional hour. 10. Cent rifuge at 18 000 9 malt for 30 min. 11. Dissolve the precipitate in the smallest volume of sta rting buffer possible and dialyse against starting buffe r. 12. A fte r dialysis. check the conductivity to make sure that most of the ammonium sulphate has been removed . 13. Assay the two fractions (flow-through and eluate) for poly(A) polymerase
and for CPSF Isee Sections 3.2.4 and 3.2 .5 ), • Not rnetvsed (see Proto col I , step 11 1. Oialysed extracts are less active for cleav age factors. b Choose a column large enough to accommodate the protein in the nuclear extract Isee step 4) . The length should be approximately 5 time s the dia~eter . . . C A fraction collector w ith a UV monitor is convenient but can be misleading . Optical density at 280 nm may indicate the presence of protein or nucleic acids, or light scattering by particulate materia l. A direct protein assay (25 ) is more reliable. d Add OTT just before use from the 1.0 M stock solution. • The equilibration is faster if the OEAE slurry is first titra ted to the desired pH be fo~e it is pouted into the column. fDo not use a magnetic stirrer to stir the slurry as this may destroy the matrix. I f D ialysis of the nuclear extract instead of dilution is also possible tbut see footnote a l. In this case, ammon ium sulphate precipitation is not necessary if only the peak flow through fractions are collected wi thout dilution by the w ashings. g With a flow -through UV monitor, collect ion of the entire peak in a single fraction is possible. If the peak has to be identified by protein assays or conductivity measurements, collect fract ions of approx imatel y one -Quarter of the column volume.
The behaviour of a ny of the processing factors during chromatography depends on a number of variables, and so the separation may not be complete. If this is the case, try varying the salt concentration of the starting buffer or reducing the amount of protein loaded on the column . If the DEAE column is intended for a simple separation of poly(A) polymer ase and CPSF, a step elution as described in Proto col 7 (step 7) is adequate. If DEAE chromatography is used as the first step in a more complete purification of any processing factor, the column should be eluted with a gradient (e.g.
28
Elmar Wahl e and Walter Kell er 10column vol, from sta rting buffer to elution buffer). Under these conditions, CPSF and a cleavage factor elute in overlapping peaks around 0.15 M KCI. 3.2. 3 Purification of processing factors from calf thymus Alth ou gh a high nuclease content generally does not permit the observation of RNA processing react ions in tissue extract , polyadenylation is an exception. This is du e to the fact that the po ly(A) tail s are quite resistant to degradation . If the precursor RNA is shorter than the po ly(A) tail s (i.e . shorter tha n 200 nt) , any polyad enylated product, however badly degraded it may be, will still migrate more slowly than the precursor R A in a gel and will thus be distinguishable from the deg raded material. As long as it retains a single labelled nucleotide, it will be detectable by autoradiography. Very little RNA remains after incubation in the extra ct , but , in the course of the pur ification of processing factors , the situation improves with every column. Since cleavage assays are much more sensitive to RNases, howev er , th ey ca nno t be ca rried out with tissue extracts. Protocol 7 can be easily adapted as the first step in the pur ification of poly(A) polymerase an d CPSF fro m calf thy mus (23). In this case, a to ta l extract prepared in a Waring Blendor (20) is used as starting material rather tha n nuclear extract. Th e mai n practical problem is the presence of large amounts of particulat e material in the calf thymus extract. Thi s material can not be removed by centrifugation . The most convenient and generally applicable way of remo ving it is to apply the turbid extract directly to the DEAE column . In this case, a short , rather wide column pa cked with a material with goo d flow properties [e.g. DEAE -Sepharose FastFlow (Pharmacia) J is important. When the column becomes clogged, stir up the top layer and continue loa ding the sample . Thi s procedure may have to be repeated several times duri ng loa ding. The column will not clog during washing and elution . A major portion of the particulate material passes rapidly through the column . Apparent ly through a gel filtration effect, it starts to appear in the flow-through after only one-half of a column volume o f buffer has passed through. whereas polY(A) polymerase acti vity and total protein increase significantly only after almost one column volume. Thus, poly(A) polymerase is partially separated from the particulate material. The enzyme can be precipitated by ammonium sulphate (at 50"70 sat uration) . After dia lysis against phosphate buffer and dilution , the remaining part icles aggregate and can then be removed by centrifugation before the enzYme is applied to a Biorex column (Bio-Rad) . Six additional columns are needed after the initial DEAE co lumn to bri ng the protein to homogeneity . For detail s of the purification procedure, consult ref. 20. Note that the enzyme as described initIally (19,20) had suffered partial proteolysis (21, 22). Undegraded enzyme c~n. be obtained by essentially the same procedure, except that the protease inhtbltors PMSF (0.5 mM),leupeplin (0.4 ug/rnl), and pepstatin (0.7 ug/ rnl) should be Included throughout all purification steps (E . Wahle, unpublished data) . C PSF bound to the DEAE column and eluted with a salt gradient will be largel y. although not entirely, free of particulate material. Its purification 29
1: 3' end-processing of mR NA
Elmor Wahl e and Wol ter Keller
to near homo geneity require s five additi onal steps (23). A third facto r, poly(A) binding protein II, can be isolated from the DEAE-bound calf th ymus fraction s by three additional steps as described in ref. 24. The separated processing factors are assayed as describ ed in the following sections.
Pro to c ol 8 . Continued . PolylA) (Boe hringe r Ma nnheim .
e W ash sol ution (0 . 5 % TCA . so dium pyr ophosp hat e)
Pharm ac ia, or Sigma l l1 • Phen ol:chlor of o rm (1: 1)
3 .2.4 A convenien t a ssa y for poly(A) p ol ym er a se
The assay described in Proto col 8 is useful for monitoring the purificat ion of poly(A) polymerase becau se it is very fast and quantitati ve. In a non- specific, Mn? -dependent reaction, the enzyme add s radio labelled AMP residues from [a- 32Pl ATP to an unlabelled poly(A) primer. The radiolabelled reactio n product is separated from unincorporated ATP by TCA precip itation and filtration. As an alternative to TCA precipitation, the incorporati on o f AMP can be qu ant ified by adsorption to DEAE paper as described in ref. 20. Th is is necessar y if short oligonucleotides are used as primers . Th e non-specific extension of poly(A) has the additional advantage that this polynucleotide, which is both the primer and the product of the reaction , is insensitive to many nucleases. Thus, this assay can be used in extracts of almo st any tissue. Commercial preparations of homoribopolymers can have substantia l proportions of their 3 ' -ends blocked by pho sphate gro ups. The se polymers are inactive as primer s for poly(A) polymerase. Unless these phosphates are 2 ' - 3 ' cyclic phosphate s from alkalin e hydrolysis o f the poly(A), they can be removed with calf intestinal alkaline phosphatase. This will activate the polynucleot ide as a primer (20). However, it is easier and cheaper to try a different batch o f the poly(A) primer. Note that the activity of the enzyme with a given batch o f primer depends not only on the extent to which the 3 ' -ends are blocked but also on the length of the primer, since, at a given concentration of poly(A), the length determines the concentraction of 3 ' -ends. Results obtained with different batche s of poly(A) are, therefore, not necessaril y comp arable . Non-specific polyaden ylation can also occur in the presence of Mg" ions. However , the reaction is much less efficient du e to a low affinity of the polymerase for the 3 ' -end of the RNA pr imer (20). Protocol 8 . Measurement of poly(A) polymerase activity by a TeA -precipitation assay
e Gle ss fibre filters (W hatm an GF /C , 24 mm diameter)
Method 32P
1. Mix [ a - JATP. 0 .1 M ATP, and w ater to provide a 12 .5 mM 32P-labelled A TP working solut ion at 20 -100 c.p. m./pmol. Spot 1 ~I ont o a GF/C filter and count it in a sci nti llat ion co unte r. Calcu late the spec ific radioa cti v it y ignoring t he co nt ri bution of th e 32P-labelled AlP t o th e t otal AlP con centrat ion.
2 . M ake a s~ock solution of polyfA) ; dissolve co mmercial poly (A ) in wate r at approxim atelv 10 mg /ml , ex tract t his onc e w ith phenol :c hlor oform and dialyse th~ soluti on ex ha us tively against water. Det erm ine the exact con centranon f rom the A 258 (f258 = 9800; the ext in ction coeffic ient refers to the conc entr at ion of the AMP mon omers ).
3 . In a ~ i.cro c entrifuge tube in an ic e bath, set up a reacti on m ixture con taining the f ollowing (sufficient f or 10 assa y s ):
125 ~I
. 2 x polyadenylat ion buffer . 1 M KCI . . . .
1O~ 1
2 5 mM Mn Cl, 12. 5 mM [ a -32 PI AT P (ste p 1) 5 mg/m l poly( A)' 2 5 mM OTT
5 ~I 1O~1
15 ~I 5~1
• H,O 93 .5 ~I 4 . Disp ense 2 5 J.l1 aliquots of th is mixture int o 1.5 ml microcentrif uge tubes on Ice. To eac h tube , add an aliquot 10. 5 -3 J.l1l o f the colum n f rac t ion or enz y ~ e pr ep ara t ion to be assayed ' . Include a tube w ith col um n bu ffer repla cin g t he enz ym e t o determ ine t he backg ro und rad io activ it y and a sa~pl e of known po ly (A ) po ly merase ac tivity as a po sit iv e co nt ro l, if availabla ,
5. Incu bat e the t ubes at 37 °C for 10 mi n or long er, depending on the act ivity of t he fracti on s.
6. Stop t he react ion by add ing 100 ~I of TCA solut ion . Vo rtex and place on
Equipment and reagents
Ice .
e va cuu rn filtration device suitable for 24 mm diameter filtersr Ie. q . Sch leicher
and Schuen. Sart orius, Millipore) e 25 mM MnC12 e O. l M A TP (see Pro tocat
e TCA solution (10% Iw/vl tr ich loroacetic acid . 0 . 1 M sod ium pyroph osp hate )
10m M
. 2 x polyaden ylatio n buffet 120% gly cero l. 50 mM Tr is- He !. pH 8 .S b , 0 .1 mM
8 . Fill t he firs t ass~y tube w ith wash solu tion f ro m a sq ueeze bott le. Do th is
BSAI • (a · J 2 PJATP (A m ersham ; 4QQC i/mmo W
. , M KCI
glass f ibre fil te rs in wash solu t ion , place one on a vacuum filtration unit and ap ply the va cuum .
7.
EDTA , 0 .02 % Non id et P-40 , 0. 4 mg /m l
Zv
SO~k the
. 2 5 mM DTT (d ilu t ed f ro m a 1 M stock so lution ; see Protocol 11
gently, c t he rw iaa y ou ma y splas h the c onte nts of th e tube . Pour t he contents of th e tube on the filter . 9 . Wa sh th e t ube out t w ic e wi t h wa sh solut io n , allowing th e washi ngs t o pass t hro ugh t he f il t er.
Continued 30
Continued 31
Elmor Wah le and Wolter Keller
1: 3' end-processing of mRNA
reactio n volume is made up as necessa ry with buffer D (see Protocol I) . T he co mplex for mation assay (see Protocol 2 ) is modified as follows:
Protocol 8 . Continued 10 . Finally, wash the filter and the filtrat ion device w ith a few millilitres of wash solution followed by ethanol. 11 . Number the filter w ith a soft pencil and place it under a heat lamp.
• ATP and creatine phos phate may be omitted since binding of purified CPSF to the AAUAA A sequence is ATP- inde pendent .
12. Filter all other samples in the same man ner and count the radioactivity retained on each filter in a scintillation counter. 13. Calculate the amount of poly(A) polymerase activity. One unit of enzyme activity incorporates 1 pmol AMP per min .
• Th e C PSF -AAUAA A complex is sensitive to heparin . Thus , neither heparin nor tracking dyes are added . Tracking dye s are run in an empty lane .
• A unit holding a single filter is much easier to use than ~hose holding many .filt~rs . Also . since f iltration th rough glass fibre filte rs is very rapid. one does not gain t ime
• Samples are incubated for 10 min at 37 ' c.
• The sa mples are loaded direc tly o nto the native gel while it is run ning . (Be ca reful : do not electrocute you rself'.)
by performing seve ral filtrations at once . . " Ma ke up a 1.0 M stock solution of Tris buff er and adjust the pH a~ ro~m temperatur~ . Use this stock to make up the polvadenvlaticn buffer wit~out readju sting th e pH . This will result in a pH of 8 .3 unde r react ion condit ions , the optimum for polylA I polymerase
4. Analysis of po lyadenylation
in this assay (201, . ·f' · ·' r; Since the 32P_labelled AT P is mixed with unlabelled AT P, a high speer IC activity IS not necessarv . Also , the labelled ATP does not have to be fresh , d The ave rage size of the polylA ) varies from bat ch to batch , from less ~han ' .00 and up to several hund red nucleot ides . The only ~ommerical product of defined Size that
for pol yadenylatio n . Wh ereas the requireme nt fo r the AAUAAA seq uence and for the downstream eleme nts has been seen both in vivo and in vitro, upstream seq uences necessary for 3 ' end-processing have so far been demonstrated o nly in vivo (reviewed in ref. 26). Transient expression of transfected plasmid DNA is generally used as an assay. RNA is isolated from the cells (see Volu me I, Chapter I , Section 3) and analysed by No rt hern blotting, nuclea se SI a na lysis, o r RNase protection (see Volume I, Chapter 2, Section 3). Po lyadenylatio n is an essential step in mRNA synthesis. The result of a mut atio nal inactivation of the polyadenylation site is not generally the accum ulatio n of the precursor RNA but a strong reduction in the concentration of th e matu re RNA. Unprocessed precusor appears to be degraded . Since the result of such a mutation is the ab sence of the RNA of interest, it is essential tha t the experiment co ntains the proper controls to ensure that thi s ab sence is not the result of a non-specific effe ct. The transfection effi cienc y, RNA extract ion , blott ing, etc . shou ld all be controlled for by the inclusion of an unrela ted standard plasmid . However, the most useful control is to incl ude a kno wn strong polyadenylation site (e.g . from SV40 o r ad enovirus) in all the exper iment al pla smid co nstr uct io ns downstr eam o f the site that is being rested by mut atio n. Any RNA that fail s to be pr ocessed at the upstre am site under Investigatio n will be ' rescued ' by the downstream site. The efficiency of l?e site being analysed can thus be directly compared to the internal standard Site. Examples o f thi s strategy and experimental pr ocedures can be found in refs 27 and 28.
the authors are aware of is A m _1 BI (Ph a rm~cla l. . . • Change the volume of th is reage nt according to the ~oncentratlon of yo~r partl~ular stock solution . Also, t he amou nt given is only a suggestion. The concentratlo,n of pnmer required to saturate th e enzy me depe nds on th e length of t~ e polynucl eotide. the K m of poly(AI polymerase for the 3 '·end of t he prime r is aoproxirnatelv 4 ~M under these
. cond itions (20 ). f Dialysis of enzyme fra ct ions prior to the assay is nor~,al1 y not necessar y Since .the volume added is small and the react ion is not very sensrnve t~ the .salt concen~ratlon . If desired , the salt concentration in all assays can be mad e Identlc~I b~ varying th e amount of KCI added in step 3 to take into account the KCI concentration In t he sample to be assayed .
Under co nditions of primer saturation, HeLa cell nuclear extract should have 500- 1000 units of po ly(A) polymera se/m g protein by this assay. Ca lfthymus extract sho uld have 1500-2000 un its/ mg. P oly(A) polymerase purified to homogeneity has in excess of 10' units /rng (20). . . . Poly(A) polymerase can also be assayed by the specific polyadenylat~on reaction utilizing RNA substrates containing the AAUAAA polyadenylatlon signal (see Protocol 6) . However , this Mg? + -dependent reacnon also requires the addition of CPSF fro m the DEAE-bou nd fractions. The assay IS not very convenient for monitoring the purification of poly(A) po lymerase .
3, 2 ,5 Assay of CPSF . Pure or partially purified CPSF can be assayed by the specific po lyadcnylatie n reactio n (see Protocol 6) when it is supplemented with po ly(A) poly~erase~ Alternatively, complex format ion With a radiolabelled RNA .contamm g a AA UAAA sequence may be assayed , essentially as described m Proto col l In each case, the nuclear extract is rep laced by the CPSF fractio n and the 32
In
vivo
In vivo a nalyses are freq uently used to determ ine sequence requ irements
References I. Wahle, E. and Keller, W. (1992). Annu. Rev. Biochem., 61, 419. 2. Moore, C. L. and Sharp, P. A. (1985). Cell, 41, 845. 3. Wahle, E. (1992) . Bioessays, 14, 113.
33
1: 3 ' end-processing
of mRNA
4. Stauber, D., Soldati, D., Luscher, B., and Schumperli, D. (1990). In Methods in emymotogv. Vol. 181 (ed. 1. E. Dahlberg and J . N. Abelson), p. 74. Academic Press,
San Diego. 5. Birnstlel, M. L. and Schaufele, F. J . (1988) In Structure and junction oj major and minor small nuclear ribonucleoprotein particles {ed. ~1. L. Birnstiel), p. 155.
Springer. Berlin.
.
.
6. Dignam, J . D., Lebovitz, R. M.. and Roeder, R. G. (1983). Nucleic ACids Res. , II , 1475. 7. Lee, K. A. W. and Green, M. R. (1990). In Methods in enzymology , Vol. 181 (ed. J. E. Dahlberg and J . N. Abelson), p. 20 . Academic Press, San Diego. 8. Zerivitz, K. and Akusjarvi, G. (1989) . Gene Anal. Techn., 6, 101. 9 . Keller, W., Bienroth, S., Lang, K. 1.1 ., and Christofori, G. (199 1). EMBO J., 10, 424 1. 10. Weiss, E. A.. Gilmartin, G., and Nevins, 1. R. (1991). EMBO J. , 10,21 5. 11. Moore, C. L. , Skolnik-David, H., and Sharp, P. A. (1988). Mol. Cell. BioI., 8, 226. 12. Konarska, M. M. (1989) . In Methods in enzymology, Vol. 180 (ed. J. E. Dahlberg and J. N . Abelson), p. 442. Academic Press. San Diego. 13. Humphrey, T ., Christofori , G., Lucijanic, V., and Keller , W. (1987) . EM BO J. , 6 ,41 59. 14. Kramer, A. (1987). J . Mol. BioI.. 196, 559. 15. Conway, L. and Wickens, M. (1987). EMBO J .. 6, 4177. 16. Takagaki, Y., Manley, J . L., MacDonald, C. C.. Wilusz, 1., and Shenk, T. (1990). Genes Dey.. 4, 2112. 17. Gilmartin, G. and Nevins, J. R. (1991) . Mol. Cell. BioI., 11, 2431. 18. Takagaki, Y., MacDonal, C. c.. Shenk, T ., and Manley, J. L. (1992). Proc. Natl. A cad. Sci. USA , 89, 1403. 19. Tsiapalis, C. M., Dorson, J . W., and Bollum, F. 1. (1975). J. BioI. Chem., 250, 4486. 20. Wahle, E. (1991). J . BioI. Chem. , 266, 3131. 21. Raabe, T ., Bollum, F. 1., and Manley, J. L. (1991). Nature, 353 , 229. 22. Wahle, E., Martin, G., Schilz, E., and Keller, W. (1991). D ,fBO J.. 10,4251. 23. Bienroth, S.. Wahl, E., Suter-Crazzolara, c., and Keller, W. (1991). J . BioI. Chem. , 266, 19768. 24. Wahle, E. (199 1). Cell, 66, 759. 25. Bradford, M. M. (1976). A nal. Biochem.. 72, 248. 26. Proudfoot, N. (1991). Cell, 64 , 67 1. 27. Iwasaki, K. and Temin, H. M. (1990). Genes Dey.. 4, 2299. 28. Russnak, R. and Ganem, D. ( 1990). Genes Dey.. 4, 764.
34
Capping and methylation of mRNA YA S UHIR O F UR UI CHI and A AR O N J . S HAT KIN
1. Introduction A ' cap ' structure of the type m 7GpppN mpNmp (Figure l) is pre sent at th e 5' -term inus o f almost all euk ar yot ic mR NAs (1, 2). Cap s are formed on cellular mR NA precur so rs in the nucleus during the initial phases of tra nscri ptio n and before several other RNA pr ocessing events ta ke place, incl uding inte rnal N6A methylatio n, 3 ' -po ly(A) addition, and exon splicing. Ca p struc tur es with multiple meth yl grou ps on the terminal guanos ine residu e (i.e. m/ ·2,7GpppNm) are also present on th e small nucle ar RNA species (sn RNA s) whi ch apparently pro vide scaffolding fo r the va rious steps in mRNA splicing (3). With o ne possible excepti on (4), a ll capped cellul a r and viral mR NAs so fa r exam ined contain only a single methyl gro up o n the terminal G residu e, while the adjacent nucleo tides a re 2 ' -O-m et hylat ed to di fferen t extents, pro viding a ba sis for the follo wing cap nomenclatur e: m7GpppN (Cap 0), m 7GpppN m (Cap I) and m7GpppNmpNm(Cap 2). The meth yl groups in Ca p 0 and Cap I structures are add ed in the nucleu s while th e addition al 2 ' -O- methylation in Ca p 2 is a cYtoplasmic event, like the 2-am ino rnethylations on snRNAs (5). In many lower eukaryo tes, incl uding yeast, m RNAs contain mainly Cap 0, while higher organisms usuall y have more extensively meth ylated caps (1). It sho uld also be noted that trypanoso me mR NAs contai n a lra ns-splic ed , capped leade r RNA with several modified residues as part of the cap (6, 7), and in some mRNAs where base I is an ade nine resid ue, ring meth ylati on o n the N6 position can also occur. Despite th ese variations on the methylation theme, the important biological consequences of a cap structure appear to correlate only with the N-7 met hyl grou p on the 5 ' -terrni nal G. For exa mple, ca ps increase mR NA stability by Protecting the mR NA against 5' - 3 ' exo nucleolytic degradation (8). Unmethylated caps ca n also stabilize m RNA to a certain exte nt , but unl ess the 5 ' -ter rnina l GpppN is converted to m7GpppN, it is still capable of being hYd rolysed to (p)pN , resu lting in the subseq uent destabilization of the decapped mR NA . Sp licing accuracy and efficiency are both increased by the presence of the 5 ' -termi nal m7GpppN (9, 10), possibly d ue to the involvement of nuclear cap -binding protein(s) (11) ana logo us to th e cytoplasmic cap-binding
2: Capping and methylatian of mRNA
o
Yasuhiro Furuichi and Aaron ] . Shatkin
en,
pppN
NH~~·
I ~ .. ",II AN
J- N~
- - Nascent mRNA
Pi~
basel
I" :1' ~ ~ 5~ IA-::~\ CII,.o·i· o. rjo·l't",C a ",u/I ()3 ' 5' 1
0-
1pN2p
(primary transcript)
RNA tripho sphatase
0-
o
3'
o
I
mRNA guanylylt ransfera se
2' O-C IIJ
b us e
S'
O=I·O.II'C~
z
AdoHet) I
03'
2' ()
I I
O=P- (}- (R:'IOA
t
AdoBcy
o--CHJ
rn'G(S')pppN chain )
0-
Ad . "..
Ad oflcy
RNA (quanine-7 +)me t h y l transferase
)
1pN 2p
-
-
(CapO)RNA
I ,
m'G(S' )pppN
RNA
methyltransferase
1mpN2mp
_ _ (Cap2)RNA
Figure 1. Cap structure of the type m 7GpppN "'pN"'p (Cap 21 show ing the m 7 -guanosi ne linked to the 5 '-end of the primary transcript via a 5'-5' phosphodiester bond . The ribose moieties of the first two nucleotides (w ith bases 1 and 2) of the primary transcript are 2 '-0 methylated. Both of these 2 '-O-methyl groups are absent in Cap 0 while in Cap 1 only the first nucleotide is 2 '·O-methylated. Where the 5 ' nucleotide is an adenylate residue, the adenine (base 1l may be additionally methylated on the N 6 position.
Figure 2 . Schematic representation of the mechanism of cap formation (Cap 2) . Details of the scheme are given in the text . Abbreviations are pppG . GTP ; N ' and N 2 , first and second nucleosides of the RNA primary transcript; E, mRNA guanylyltransferase ; AdaMet . S~a den o s y l m et h i on i n e ; Ado'-lev, S ·adenosylhomocysteine; m , methyl group.
initiation factor eIF-4F that is required for eukaryotic protein synthesis (12). Int erestingly, the donor of the methyl group(s) in the cap is an activated form of t.-methionine, S-adenosylmethionine (AdoMet), and protein synthesis is also initiated with an activated L-methionine intermediate, initiator rnethionyltRNA. For most cellular and viral mRNAs, capping and subsequent methylation occurs by the scheme illustrated in Figure 2. The S' -terminal end of the nasc ent RNA product of RNA polymerase II is first modified by removal of the S' -l phosphate by RNA triphosphatase to yield a diphosphorylated end (ppNpNp-). This is then capped by GMP transferred from GTP by mRNA guan ylyltransferase (GpppNpNp-). Subsequent methylation then occurs in stages catalysed by RNA methyltransferases using AdoMet as the methyl donor to produce the S' cap structure, mJGpppNmpNmp_. RNA capping does not oc cur on the products of RNA polymerase I and III which synthesize rRNAs, and the tRNAs and SS rRNAs, respectively. This indicates that the enzyme(s) involved in mRNA capping are associated with RNA polymerase II and/or oth er component(s) of the transcription initiation complex (13).
Th e process of mRNA capping has been studied most extensively using vaccinia virus-derived proteins . A heterodimeric 6.5S protein complex, consisting of 90 kDa and 31 kDa subunits, was found to catalyse the reactions shown in Figure 2 (14). The 90 kDa subunit contains both the RNA triphosphatase and the RNA guanylyitransferase (the capping enzyme), while the 31 kDa subunit is req uired, but is not sufficient, for methyltransferase activity (IS, 16). Inte restingly, the capping enzyme complex is also responsible for termination of vaccinia virus transcription, although the presence of as ' -cap is nat a prerequisite for transcript termination (17), unlike the case of influenza virus mRNA synthesis (18) and histone mRNA 3 ' -processing (19). Vaccinia virus provides another example of a protein with effects on both S ' and 3 ' endprocessing; the S ' cap-specific viral 2 ' -O-methyltransferase also stimulates mRNA 3 ' -polyadenylation (20). Since transcript synthesis begins at the capping site, the identification of the precise species of cap [e.g. cap A (mJGpppAm) or cap G (mJGpppGm)] and their adjacent sequences in RNA transcripts provided the first accurate markers of tra nscriptio n start sites on polynucleotide templates. The key to the complete
36
37
2: Co pping and m ethylation of mRNA
Yasuh iro Furu ichi and Aaron]. Shatkin
characterization o f mRNA caps is their efficient isolatio n and the ident ification of the resulting nucleosides after cleavage of the 5 ' - 5 ' triphosphate linkage between the m 7G and N'", In this chapter emphasis is pla ced on the in vitro synthesis of high ly ra diolabelled capped RNA (Section 2), methods for po st-transcriptional decap ping and recapping of RNAs (Section 3) and the characterization of ca p structures (Section 4). Meth od s which were used for discovery of the cap st ruc ture (2 1-23) a re also desc ribed because they facilitated the isolation of capped m RNA s and the characterization of different type s of caps. Readers should consult the original publi cations for additional detail s.
incubated with [met hyl-Jl-l] Ad oMet yield [3H J meth yl-labelled caps . Alt ernatively, [or-32P ]GT P or [or- 32P ]ATP can be used to label either both of the 5' phosphates in m 7GpppG mor onl y the pA min m'GpppAm, respectively. Additi on o f inorganic pyrophosphate to the se reaction mixtures diminishes guanyl ylat ion of RNA and increases the proportion of ppN 5' -ends formed . In contrast , the presence of inorganic pyrophosphatase yields mostly 5' -ter minal GpppN which is co nverted to rrr'Gpppbl'" when AdoMet is also present (Table I and Figure 2) (27). T he follo wing sections contain protocols describing the preparation of a variety of viral mRNAs bearing ca ps radiolabelled in different ways. For the preparation of radi olabelled m'GpppAm (or GpppA) structures, C P V wo uld normally be the system of cho ice. Howev er, th is virus is usuall y purified from infected silkworms, a source not always readily availabl e. Vaccinia virus also produces m 7GpppA m-ended transcripts, and it can be obtained in relatively large am ounts from infect ed cell cultures, e.g. HeLa cells or chick embryo fibroblasts. For radio labelled G caps (rrr 'GpppG'", GpppG), reovirus m RNAs provide a good so urce , since milligram quantities of virions can be readily purified from a few litres of infe cted mou se L cells. Although the transcription and capping reactions for producing capped viral mRNAs are very similar, they have different temperature optima, etc ., and the protocols described are not interchangeable. Section 4.2 contains protocols for the enzymic removal of caps from vira l mR NAs for use as authentic cap markers. Non -radioactive caps and cap anal ogues have also been synt hesized chemically and are commercially available from Pharmacia- LKB and other suppliers. Th e synt hesis of radiolabelled capped mR NAs from cellular genes using bacteriophage promoters and RNA pol yrnerases is described in Section 2.5 . These product s can be especially useful for producing large amounts of mRNA for translation studies, model pre -mRNAs for matu ratio n studies, and capped probes for various analyses.
2. Synthesis of ca p ped mRNAs during in vitro
transcription of viral ge ne s 2.1 Introduction Several eukaryotic vir uses, including mammalian reovirus, vaccinia virus and insect cytop lasm ic polyhedrosis virus (C PV), contain virion-associated RNA polyme rase, guanylyltra nsferase and methy ltransferase(s). Consequently, in vitro transcription reactions mediated by purified viru s particles yield capped mRNA. This has proved invaluable fo r determining the mechanism of viral cap formation which is essentially identical for cellular mRNAs. Since the viral mRNAs, like most cellular m RNAs, initiate with either A or G, th ey can provide convenient sources of [3H]methyl- o r J2 P-Iabelled forms of cap A (m''Gppp.A '") and ca p G (m 7GpppGm) for use as aut hentic standard markers, e .g. in the a na lysis of cellula r mRNA caps. Table I sum marizes the type s of cap str uct ures that can be obtained from viral transcription reactions. Standard methods for th e propagation and purification of th ese viruses will be found in refs 24-26.
2.2 Strategy Select ive rad iolabelling of specific sites in caps can be de vised o n the ba sis of the mechanism of synt hesis. For example, tran scription reaction mixtures
2.3 Synthesis of cap ped mRNAs using insect cytoplasmic polyhedrosis virus [3H] Methyl-labelled CPV m RNA is synt hesized using [methyl-Il-l ] AdoMet as described in Protocol I. Note that all tubes, pipettes and solutions sho uld be free of RNase. This can be accomplished where necessar y by autoclaving or treatment with DE P C (further details will be found in Volume I, Chapter I , Section I.I).
Tabl e 1. mRNA c aps synthesized in vitro by v iral tr an scription
Prot oc ol 1 . In vitro sy nthesis of ca pped CPV m RNA Transcription co ndi tions"
--
+ Ad oMe t
- Ad oM et or + A doHc y
Insect cyto plas m ic poly hedrosis vi rus (CPV )
m 7G p p pA m
Reoviru s
m 7 G p p p Gm
GpppA GpppG GpppA, GpppG
Virus
Vaccinia v irus
Equipment and reagents e S -adenosyl-t *( m eth y l-JH I methionine
m 7 G pp pA m , m 7G pp pGm
lAda Met ; Am er sh am , Searle; - 70 ei l mrn ol , 1 mC i/ml in sol ution in He l, pH 2.0 -2 .5 1; this should be neutrali zed before us e by addi tion of an eq ual vo lume of 70 mM NH...OH
(1-2 mg /mll prep ared m eth ods desc ribed in
ref. 24 . 1.0 M Tris-HCl, pH 8 .0 e O. l M Mg CI2
Con tinued
• A do met . S -adenosy lm et hion ine : AdoHcy. S ·adenosyl hom ocyst eine.
38
_ CPV virions by stand ard
39
2: Capping and methylation of mR NA Protocol 1 . Continued e Sephadex G· 'O O col umn 120 em x 0 .6 em) equi lib rated in elution buffer
e St oc k solutions (40 mM I of ATP , CTP. GTP , and UTP prep ared in H 20 . neutralized to pH - 7 with Tris base . and sto red frozen
e Pnenolcbtorotorm (1: 1, v /vl
. 3 .0 M potassium aceta te . pH 4 .5
• Elu t ion bu ffe r (50 mM Tr is-He l.
pH8.0)
e Dry ice
Yasuhiro Furuichi and Aaron ]. Shatkin To prepare " P -Ia belled unmethylated GpppA , carr y out the RNA synthesis in a modified transcription reaction mixture (Protocot J, step I) containing 2 mM S-adenosylhomocysteine (AdoHey) in pla ce of [methyl-'H ]AdoMet and with th e addition of O.4mM [~-"P ]ATP in pla ce o f the ATP_ Prepare th e [~_J2P l ATP from [-y-"PlATP according to th e method of Furuichi and Shatkin (28) .
Method
2.4 Synthesis of capped mRNAs using reovirus
1 . Set up the transcr iption react ion by m ixing th e f ollow ing reage nt s in a st eri le m icrocentrifuge tube on ice :
Protocol 2 describes the preparation of washed reo viru s cores and the ir subsequ ent use in an in vitro tra nscript io n system with [m ethyl-'H 1Ad oMet and [a-"P 1GTP .
e 1.0 M Tris -HCI, pH 8.0 e 0 .1 M MgCl, e 40 mM ATP e 40 mM CTP e 40 mM GTP e 40 mM UTP e (methyl-' HIAdoMet 120 e c pv virions 1 1 0 - 2 0 ~ g l e H,O
5 ~I
12 ~I 1O~ 1
5
Prot ocol 2 . S ynt he s is of capped mRNAs us ing washed reovirus cores
~I
5~ 1
Equipm ent and reagents
5~1
~C i )
40
~I
1 -2~ 1
to 100
~I
final volume
2. Incubate the mixtu re at 31 °C for 2 h.
3. Add an equal volume (1 00
~I )
of phenol:chloroform I' :11 and vortex
for 1 min . 4 . Centrifuge the mixture at 10 000 9 for 5 min in a microcentrifuge to separate the phases . 5. Carefully remove the upper (aqueous ) pha se to a clean mi crocentrifuge
tube . 6. Separate th e radi olabelled CPV
mRNA s from
• Purif ied reovirus ( - 5 mg /m ll prepared by standard meth ods as described in ref . 2 5
. V iral core was hing bu ffer (50 m M Tr is-
• 10 mg /ml a-chym ot ry psi n (Wo rthingt on Bioc hem ical. crys tal line enz y m e fr eshly diss o lv ed in H 20 )
. 0 .2 5 M phosphoenol py ruvate
• , mg /m l in org anic pyrophosphatase (Boehrin qer Mannheim , 1 m g/m ll
e 3.0 M KCI . 1.0 M magnesium acetate , pH 7 .0 32 • [a· PJGTP (A m ersbam. Sear le; 4 10 Ci /mm ol , 10 m Ci/ml l
HCI. pH
8.0.
50
mM KCII
• Pyruv at e kinase lCalbiochem ,
- 8450
IU /mI J
• [ me t hy l· 3HjA do Me t . 1 M Tr is -H CI, pH 8.0 , 4 0 m M stock so lu ti ons of A TP. GTP , CTP, and UTP , phenol :ch lor of orm , elut io n bu f fe r, and Sephadex G-1 00 colum n as de scr ibed in Proto co l 1
th e uninc orporat ed
[meth yl-'H 1AdoMet by chromato graphy on Seph adex G-' 00. Collect
Method
0 .1 ml fra ct ion s and m on itor the sepa ration by sc intill ation co unt ing. 7. Pool th e fr act ion s co nt aining the CPV mRNAs (first peak ) in microcentrifuge tube s and ad d 3 M potas sium aceta te, pH 4 .5, t o 0 . 15 M f inal conce nt rat ion plu s, 2.5 v ol. of ethan ol. Plac e on dry ice for 15 m in t o pr ec ipitate th e RNA, and th en pellet th e precipitated RNA by ce ntrifugat ion in a mic roce nt rifuge f or 15 min . 8. St or e t he RNA as a we t pellet at - 20 °C or dry it and redissolv e it in autoclaved or OEPC -treat ed , nuclease-free distilled H 20 before freezing .
1. Using st ?Ck solutions, prepare reoviru s cores by mixing 2 mg (me asured as protein ) of purified reo viru s in buffer co nsisti ng of 7 0 mM Tris-HCI
pH 8, and 90 mM KCI {final conce ntrationsl . Add 10 mg/ml c hym otrypsi~
t o 0 .5 mg /ml and in cubate the mixture (tota l volume of 2 mil f or 45 m in
at 42 °C .
2 . Collect th e vi ral co res by c ent rif uga tion at 10 0 0 0 9 for 10 mi n at 4 0C . 3 . Resuspend the pellet in - O. 5 ml of wa shing buffer and ce nt rif ug e again as In ste p 2. Repeat. 4 . Resuspend the washed vi ral co res in 0 .2 m l of washing buffer.
Under the conditions descr ibed in Protocol J about 5"70 of the [' H] met hyl gro ups are inco rpo rated into m 7Gp ppA m from the [methyl-' H] Adoxtet prec ursor. To obtain ['H]methyl-Iabelled m 7G pp pA m, the m RNA products are digested with nu clease PI, and the caps are purified by high-voltage paper electro phoresis or chromatography o n polyethylenirnine (PEl) cellulose or paper (see Section 4). 40
5.
S ~t u p th e. transcr iption reac tion by mixing t he follow ing reagents in a rrucrocentr ifuqa tube on ic e: e 1.0 M Tris -HCI, pH 8.0 • 1.0 M magnesium acet ate , pH 7.0
e 40 mM ATP
Continued 41
Yosuhiro Furuichi and Aaron J. Shafkin
2: Capping and methylation of mRNA
Protoc ol 3. Continued
Protocol 2 . Continued e 40 m M CTP
50~1
e 4 0 m M UTP
50~1
e 40 m M GTP
50 ~I
e [ a -"PjGTP (100 ~Cil
10 ~I
Method 1 . Set up the transcript ion reaction by m ix ing the follow ing re ag ent s in a m ic ro c en t rif ug e t u b e on ice :
e 1.0 M Tr is - HCI, pH 8 .0
100 ~I
e [ met hyl-'HI A doMet 150~Cil e 0 .25 M phosphoenolypyru vate • pyruvate kinase ( - 8 units ) • washed reovirus cores (equivalent t o 500 lJ9 virus measured as protein ) • , mg /m l inorganic pyrophosphatase
10
~I ~I
e O. l M MgC I2
10
10 ~I
e 40 m M ATP
12~ 1
1 ~I
e 40 m M CTP
12~ 1
e 40 m M GTP
12~ 1
50~1
5~1
e 40 m M UTP
1 ~I to 0 .5 ml final vol ume
e H 20 6 . Incubate the m ixture at 45 °C for 2 h with shaking every 1 5 min to resuspend cores . 7 . Recover and store the capped . radlotabetled reovirus mRNAs as described in Protocol 1. steps 3-8 except in step 6 colle ct 0 .5 ml fractions from the Sephadex column .
10~1
e 1% NP-40 e O.2 M OTT
lO ~ t
e [ me thyl·'H ) Ado Met (20 ~Cil
40 ~I
e Purified vacci nia virus (100 jJg)
e H 20
50 jJl t o 200
~I
fina l volume
2 . In c u b at e the m ixture at 37 °C for 2 h . 3 . Centrifuge the mixture at 10 000 9 for 15 m in at room temperature in a microcentrifuge to pellet the v irions.
To obtain the [3H I methyl- and 32P-IabelJed m'GpppGffi cap s, the reovir us transcripts are digested with nuclease PI and the caps are isolated as described in Section 4.
2.5 Synthesis of capped mRN As using va ccinia virus In vitro transcription using purified vaccinia virus yields poly(A) ' mRN As containing m'GpppAffi and m'GpppGmin a 2: I ratio (29). These two cap for ms may be released from the transcrip ts by nuclease P I digestion and are r~ad l l Y resolved by paper chr omatography (see Section 4.3.2), The transc npn on procedure is described in Protocol 3. Pro tocol 3. In vitro synthesis of capped va cc inia virus poly(A) ' mRNAs
Equipment and reagents e Veccinia virions (2- 4 mg /ml calculate fr om the A 260 value where 1 mg/m l is equivale nt to - 15 .7 A 260 units) pur if ied by st andard methods as descr ibed in ref. 26 e 1 % tv/v) Nonidet P-4 0 (NP-40; GIBCO-
e O.2 M OTT tCalbioc heml
e 1.0 M Tr is-H CI. pH 8 .0 . 0 . 1 M Mg CI2 • 40 m M stock solutions of A TP. CTP. GTP. and UTP. I methyl -3HjA do Met. phenokch'o rof or rn . elution buffer. Sep hedex G- l 00 col umn" as descr ibed in Prot ocol 1
BRLl
Continued 42
4 . Ca refu lly remove the supernatant to a fresh m icrocentrifuge tube and purify t h e vi ral mRNAs as described in Protocol 1 steps 3_8 6 • • Since the vaccini a virus mRNA s are polvadenvt at ed . st andard aff inity chromato graphy on oligoldTl- cellulose described in ref. 30 may be used for th eir pur ifi cati on in place of Sephade x G·l 00 ch romatography .
2.6 In vitro transcription and capping of cellular mRNAs using bacteriophage RNA polymerases In vitro transcription systems based on SP6, T3 or T7 phage promoters and RNA polymerases make it po ssible to synt hesize biochemical amounts of RNA molecules of any desired sequence including cellular mRNA s (31). The transcript s have been widely used not only for expression cloning but also for the investigat ion of RNA processing, RNA stability, and protein synthesi s. Alth ough bacteri op hage polymerases transcribe the DNA sequence s placed downstream of their specific promoters very accurately and efficiently, the 5' -ends of the produ cts are not capped and remain triphosphorylated. However, for some purposes, such as studies on RNA splicing and /or translation , it is important to use capped transcripts. By including vaccinia virus-derived capping enzyme and AdoMet in th e trans cription incubation mixtu re, the 5 ' -triphosphorylated RNA can be readily capped and methylated according to the reaction sequence depicted in Figure 2. Alternatively, cap analogues such as GpppG or m'GpppG can be added to the tran scription mixture . By concomitantly decreasing the GTP concent ration, tr anscription by the phage polymerase then starts preferentiall y 43
Ya s uhiro Fu ru ic hi a n d Aaron
2: Ca p p ing and m ethylation of mRNA with GpppG o r m'GpppG. Chain elongation is still effective at th~ lower GTP con centration. A procedure for th e synthesis of capped mRNA usm g SP 6 po lymerase is described in Prot ocol 4.
J.
Sh atkin
3 . Site-sp eci fic radiolabelling of caps in
cellular mRNAs 3.1 Introduction
Protoco l 4 . S ynt he s is of c a ppe d cellular mRNAs us ing SP6 RNA polymerase
Equipm ent and reagen ts . ' .0 M Tris-He !. pH 7. 5 e O. l M MgCI 2 • Sephadex G· l00 column , elu tion bu ffer. phenol :chlo ro form as described in Protocol 1
• RNasin (Prom ega. 20 units/pll
e (a _32P 1UTP (A mersham. Searle; 400 Ci /mmol ,10mCi /ml l
. 20 m M spermi dine-He1
. 50 mM stock so lutions of ATP, CTP. and UTP
e 2 mg /m l BSA • SP6 RNA polymerase (Pro mega . 10 - 20
e 5 mM GTP
uni ts / ~l )
• Temp lat e D NA . line arized. 1 mg /ml
. 0 .1 M OTT tCalbiocheml
. 50 mM m 7Gp ppG IPha rmac ia- LKBl • RNase-fr ee DNase (Promega. 0.2 ~ g /~ 1I
1 U '~I •
Me thod t , Set up the transcription reaction by mixing in a microcentntuge t~be (at room temperature to avo id pre cipitation of the DNA by spermid ine ):
• DNA temp late . 1.0 M Tris - HCI. pH 7.5 . 0 .1 M MgCl , • O. 1 M OTT
. 2 mg/ml BSA • 50 mM A TP • 50 mM CTP • 50 mM UTP .5 mM GTP e [ a-32P I UT P I 20~C i l
. . . • •
50 mM m'GpppG RNasin 20 mM sp ermidine-HCI SP6 RNA polymerase H,O
10 ~I 4 ~I
6 ~I 10 ~I 5 ~I 1 ~I 1 ~I 1 ~I
1 ~I 2~1
1 ~I 5 ~I 10 ~I 4 ~I
to 100 ~ I final volume
2 . Inc ubate the mi xture at 40 °C for 1 h. 3 . Rem ove th e t emplate DNA by adding 2 ~I RNase -fr ee DNase (fi n al cone.
20
~g/ml l
and incuba te at 37 DC for 10 min.
4 . Reco ver the capped t ranscripts as des cr ibed in Protocol 1, st ep s 3- 8 .
44
Characterization o f the cap str uctu res in individual cellular mR NAs is often a difficult ta sk because of the limited amounts of material available. Hence. it is important to be able to labe l the cap and /or the adjacent nucleotide specifically and to a high specific radioactivi ty in o rder to esta blish the 5 ' -terminal struc ture of mRNAs. T his can be acco mplished by o ne of the following three approaches: e labelling with
32p
by decapping and recapping (Section 3.2)
e labelling with 3H by periodate oxidation a nd borohydride reduction (Sectio n 3.3) e labelling with 3H by 2' -O -methylation (Sect ion 3.4) T he first approach. radio labelling of mRNA by chemical removal o f the m'G followed by enzymatic recapping in the pre sence of (a- 32P JGT P . is
probably the most practical procedure since purifie d guany lyltransferase (capping enzyme) is now commerically available (Gibco -BRL) . In ad dition. purification procedures for guan ylyltransferases from rat liver (32), yeast (Saccharomyces cerevisiae) (33, 34) and brine shrimp (Anemia salina) (35) ha ve been described . Th ese enzymes can also be used for recapping . Radi olab elling by oxidation and lH -borohydride reduction. the second approach , has the disadvantage that the 5 ' -terrnina l m''G is converted to the trialc oh ol de rivative in this procedure. In addition , because the 3 ' -terminal residue in RNA a lso contains a free cis-diol , it becomes lH -labeIled by the oxidatio n and reduct ion scheme. Th e third technique, labelling with lH by 2 ' ·O-methylation, requires that th e mRNA contain a Cap 0 struc tur e, a limitation because most animal virus and cellular mRNAs contain Cap I or 2 struct ures .
3.2 Radiolabelli ng mRNAs by decapping and recapping This labelling strategy comprises the following steps: e che mical removal (decapping) of th e 5 ' -terrninal m'G from capped mRNA e recapp ing with guanylyltrans ferase and [a_ 32p 1GTP
3.2 .1 Dec apping mRNA Caps contain a fr ee 2 ' -3 ' cis-diol o n the rib ose rin g of the 5 ' -linked m' G . Th ese hydroxyl groups can be oxidi zed to the dialdeh yde form by treat ment with sodium metaperiodate (NaIO.) under co nditions th at lea ve th e RNA chain intact (with the exception of the 3' -terminal nucleotide which also contains an oxidizab le cis-diol). The decapping procedure, described in Protocol 5, 45
2: Cop pi ng and me thyla ti on of mRNA
Yasuhiro Furuichi and A aron ]. Shutkin
exploits this periodate oxidation followed by a ~ -el iminat ion react i o ~ with aniline . Th e oxidation an d ~ -e li mi nation reactio ns conv ert the mR NA 5 -caps to 5 ' -trip hosphor ylated ends with release of the trialeohol deriv ative of rn/G , referred to as m7Gt : :'\aIO~
m'G pppNm _ _
1\
O H OH
aniline
7
t
Proto c ol 5 . Contin ued 11 . Precipitate the RNA by adding 0 .1 ml of 2 .0 M ammonium acetate ,
pH - 7.0 , and 0 .55 mt of cold (- 20 · C) ethanol. 12. Recover the RNA as described in steps 5-8.
m
m'GpppNm_ _ m G + pppN -
~
13. Redissolve the deca pped RNA in 10 ~ I of distilled H 0 . 2
CHO C HO
3.2.2 Rec a p p in g mRNA
Protocol 5 . Chem ical decapping by oxidation and
~ -elimination rea ct ion
The deca pped triphosphorylated RNA is recapped by incub ati on with the vaccinia viru s capping enzyme complex which contai ns RNA triphos phat ase and meth yltra nsferase(s) as describ ed in Pro tocol 6. The reac tio n seque nce is:
Reagents e Capped cellular mRN A (- 100
purified as described in Vo lume L Chapter
r. Section
ppp Nm- _
. 20 % glyce rol
~ g /mll
3
• Metaperiodate reagent (fresh ly prepa red. 10 mM NatO,. in 0 .15 M sodium acetate , pH 5.3 1
_ 0 .3 M aniline. pH 5.3
ppNm-
[Q'-J2PJGTP
•
GpppNm_
I mefhyPH I Ado~fel x
•
m' GpppNm_
. 2.0 M am mon ium acetate, pH -7 .0 e l SE buffer (20 mM Tris- HCl. pH 7.4 . 0 .5 mM EDTA. 0 .5 % 50S )
e Canier yeast tANA (Si gm a) disso lved in H 2 0 at 2 m g /ml
where ' an d ' denote " P- and [ l H) methyl-la belled positions, respectively. Protoc ol 6 . Re c apping mRNA with [a-32P1 GTP Reag ents • Deeap ped mRNA ( - O. 5-1 .0}Jg in - 5-1 0 IJ I H 2 0 ) prepa red as in
Method
. 2 mM S ~ ade n os y l m e t h i o n i n e (Ado Met, Boehringer Mannh eim) dissolved in dilute HCI at pH - 2 and stored frozen ; just before use neutralize with an equal volume of 70 mM NH 40H
Protocol 5 ·
Periodate oxida tion
. 1.0 M Tr is -H CI, pH 7.9
1 . Mix the capped cell ular mRNA 1- , H 20 .
~g )
with 20 ~g of carrier tRNA in 10 ~ I
. O. 15 M KCI
2 . Add 1 ~I of periodat e reagent and incubate the mixture on ice for 1 h in the dark.
3. Add 10
~I
. 12 mM M gCl 2
of 20 % glycerol to st op the oxidatio n.
4 . Precipitat e the RNA by adding 0 .15 ml of 2. 0 M ammo nium acetat e,
G-l00 column, elution buffer, and 3 .0 M potass ium acetate , pH 4 .5. as in
. 50 0 mg /ml BSA
Proto col 1
• Guan ylylt ransferase from vaccinia virus (Gibeo - SR L ca t alog ue ** 8024SA .
1-5
pH -7 .0 and 0 .B5 ml of cold 1- 20 · C) ethanol.
• Phenol.c hlorofc rrn ( 1: 1, v/vl . Sephadex
• 0 .125 M OTT (Calbiocheml
U /~I )
. 0 . 2 5 M phosphoeno lpyruvate, pyruvate 32P JGTP as in Prot ocol 2 kinase, l a -
5. M ix well and leave at - 70 ° C for 15 min. 6 . Centrifuge the tube at 10 OOOg for 1 5 min at 4 °C. Carefully remove and discard the supernatant.
7. Redissolve the RNA pellet in O. IBm} TSE buffer . Add O.IB ml of 2.0 M ammo nium acetate, pH -7 .0, followed by 0 .9 ml ethano l.
B. Repeat steps 6 and 7. 9. D ry the RNA pellet under reduced pressure and redissolve it in 201-11 of
TSE buffer. {3-elimination reaction 10. Add 0 . 1 ml of 0 .3 M aniline, pH 5.3, mix and incubate the mixtu re for 5 min at 30°C .
Method 1. Set up the recapping reaction by mixing the following reagents in a microcentrifuge tube on ice:
. • . .
1.0 M Tris -H CI. pH 7.9 I 2 mM MgCI2 0 . 15 M KCI 0 . 125 M DTT
2.5
5 ~I 2 ~I , ~I
• 500 mg/ml BSA • [a -32P IGTP (lOO ~Ci)
5
~I
1O~1
Contin ued
46
~I
Con tinued
47
Yasuhiro Furuichi and Aaron j. Shatkin
2: Ga p pi ng a nd me thyla tion of mRNA
A
Protocol 6 . Continued e d ec apped RNA 10 . 5 - 1. 0 ~9)
3
4
1O~ 1
B
5
EB
5 ~I
e 2 mM Ad oMet
2
Pi
1 ~I
e gu anylylt ran sf erase (1- 5 un it s) eH 0
to 50 ~I final v olu m e
2
2 . Incubat e t he mi xt ure for 60 min at 3 7°C . One unit of gu.a~Yl yltransferase will fo rm _ 1 pm ol of ca p in 30 mi n under these co ndi tions . 3 . Ad d
an
equ al vo lume o f phenol :chloroform (1: 1) and v ortex f or 1 m in.
4. Reco ver the RNA by ethanol precip itation as described in Protocol 1, step s
1 .
4 -7 .
(,
m
m Gppp lm ' A
•
5 . Rinse th e RNA pel let wit h 50 ~I of 75 % eth anol at - 20 DC. 6. Dr y the pellet and redi ssol v e the radio labe lle d RNA in 1 0-20 ~I of H 2 0 .
", 'GpppG
7 . Store the RNA at - 70 °C or in liquid nitrogen.
\
/'2
I
• The de capped mRNA w ill als o conta in c arr ier t RNA.
-_ .
_ m 7Gpppm 'A IIl
m 7GpppA m
•
_
m 7GpppA III
G p pp" m
Gppp (m " )A m
After the recapping procedure. the recapped RNA may be subjected to cap analysis (Section 4) to do cument the presence of m''Gpppbl'". Figures 3 and 4 show typical cap analyses for a var iety of mRNA s. Unmethylated caps of the stru ctur e Gppp bl'" are formed if the recapp ing (see Protocol 6) IS earned out either in the absence of Ado Met or in the presence of 2 mM AdoH cy.
8
_ _ or igin
_____ -ori gin .-
2 3
non e TMVglobin CPV U5
Figure 4 . An alysis of decap ped and recapped 5' -t ermin i by DEBl pap er electrophoresis and t hin layer chr oma t ography on PEl- c ellulose. (A I Dec apped RNA s were rec apped by incubation w ith [a _32 P]GTP in t he pr esen c e of A da M et (see Prot ocol 6 ), dige st ed with nucle ase Pl (see Protoco/ 9 l. and analysed by elect rophoresis on DEBl paper (see Protocol 16 ). Lane 1. no mRNA ;
3
2 :;;
5
4
6
lane 2. tobacco mos aic virus (TMVI RNA. 1 ~g ; lane 3. rabbit globin mRNA . 0. 1 J.lg; lane 4, CPV mRNA . 0 .5 IJg ; lane 5. ch icken US snRNA . 0 .4 1-19 . (B) Identif ication of m 7 Gppp A M and m 7Gpppm 6A M cap st ruc tu res by th in-layer ch romatog raphy on PEl -cell ulose. After electro phoresis as sho wn in fA) , 32P-label1ed cap st ruc t ures were elu t ed fr om th e DEBl paper in 1 M triethylammonium bicarbon ate, pH 7.6 , and furt her analysed by chromatography on PEl- cellulose (see Protocol l SI. t he deve lop ing solvent was 0 .6 M LiCI. Lane 1, m 7Gp ppm 6 A m of glo bin mRNA (upper spo t f rom lane 3A I; lane 2. m 7GpppA mCPV m RNA cap (upper spot fr om lan e 4 A I; lane 7Gpp 3, m pA m of U 5 snRNA (upper spot f rom lane 5A I. (Dat a kindly provided by Dr K . Mizumoto.)
m
.. (;Pl'p'\
I
I
-noRNA
CPV mRNA
- a ppG-RNA
~
EIA mRNA CHela )
48
EIA t ransc r Ipt
< o r t z t n
ppG-RNA
Figure 3 . Rad iolabelling of 5 ' -eaps in viral m ANAs and analysis by DEBl paper electrophoresis . CPV mA NA and adenov irus E1A m AN As expr es sed in HeLa cells an d yeast, respectivel y , were decapped and rec app ed with Icx- 32P I GTP as descr ib ed in Protocols 4-6. Aft er digest ion with nu clea se Pl (see Protocol 9 ), the c aps w er e analy sed by el ec tro pho resis on DEB l pap er (see Protocol 16 ). l ane 1, n o m RNA ; lane 2, decapped and reca pp ed CPV m RNA ; lane 3 , CPV RNA (negati v e control, synthesized in th e presenc e of A doH c y and subjected to all th e same steps as CPV m AN A and, in add it ion , alkal ine ph ospha tase treatment t o remo ve the unprote ct ed 5 ' -d iphosphate); lanes 4 and 5, adenoviru s E1A mRNA s (t ransc ription starts w ith ei t he r A or G l : lan e 6 , CPV ANA {positive cont rol, synt hesi zed in th e presence of Ad oH c y and t hus containing ppG 5 '·ends, capped in a modified capping reac t ion w ith { cr-32P1GTP but w ithout Adcmetl . Positions labe lled with 32p are in dic at ed by '" , Dat a kind ly provid ed by Dr Kiyohisa Mizumoto (Kitasato Un iv ersit y , Toky o, J ap anl .
49
Yasuhiro Furuich i and Aaro n
2: Capping and methylation of m RNA
3.3 Labelling of mRNA by periodate oxidation followed by reduction with [3H 1sodium borohydride As mentioned in Section 3.2.1, the fr ee 2 ' - 3' cis-dial on the ribose ring of the 5 ' -linked m' G ca n be oxid ized to the dialde hyde form by treat men t with sodium metaperiod at e (Na J0 4 ) . The dialdehyde can then be ra dialabe lied by reduction to the dialco hol form (m''G ") with [lH ]NaBH 4 acco rding to the following scheme where * indicates radio labe lled positions : Nal04
m'GpppNm -
~
!\
OH O H
P H ) NaB H 4
m' GpppNm-
m' G ' ppp Nm-
.*~* * CH,O H
/\
CHO C HO
f. Shatkin
Protoc ol 7. Continued 7 . Wash the RNA pellet with ether, air dr y it and redissol ve it in 301-11 of
0 .5 M sodium phosphate buffer , pH 7.0 . 8 , Add 2 jJl of
for 1 h.
[ 3H J NaBH 4 (2 mC i) and incubate the mixture on ice
9 . Remov e unincorporated [ 3H I NaBH 4 by chromatog raphy on Sephadex G- l 0 0 and then recover the radiolabelled RNA by ethanol precipitation as described in Protocol' , steps 6 and 7. 10. St ore the labelled RNA frozen either as an ethanol pellet or dried and di ssolv ed in Btvase-fr ee dis ti lled H 2 0 (see Protoco l l step 8 ).
CH,O H
This labelling procedure is described in Protocol 7.
3.4 Labelli ng of mRNA by 2 '-O-methylation 2' -O- Methyltransferase catalyses the transfer of th e met hyl gro up from AdoMet to the 2 ' -OH of the nucleotide immed iately adjacent to th e cap nucleotide in
capped mR NAs: Protocol 7 . Labell ing of m RNA by re du cti on wi t h [3H] NaBH4 AdoMet
Equipment and reage nts • Capped mANA (- 100 IJg/mll iso lated as in Vo lume I, Chapter 1, Protocol 4 . 20 mM sodium acetate buffer ,
pH 5.0 . , 0% propylene glycol . 0 .3 M sodium acetate buffer , pH 5.0 e O.5 M sod ium phosphate buffer ,
rrr'Gppplv" - _
m' G pppNm-
• [3H ]NaBH 4 (New England Nuclear ; 8 .8 Ci /mmoJ, 10 mCi ll 01-11 in 50 mM
NaOH I • Elution buffer , Sephadex G-l 00 column , and 3 .0 M potassium acetate , pH 4 .5 , as in Protocol 1 and metape ricdate reagent
as in Protocol 5
pH 7.0
Me thod
The 2 ' -O -met hyltr an sferase has bee n purified from vaccirna virus and used for radi olab elling the RNA s of bra me mosaic virus and tobacco mosaic viru s by conversion of 5 ' -terminal m'GpppG to m'Gpp pGm (36). For efficient methyl gro up transfer, vaccinia virus 2' -O·methyltransferase requires a polyribonucleo tide substrate capped with a Cap 0 (m'GpppN) structure. Since mR NAs fr om lower eukaryotes and plants contain mainly Cap 0 structures, they are good substrates for the enzyme. The procedure for labelling mRNA by 2 ' -O-methylation with [rnethyl- Jl-l] Ado Met is described in Protocol 8.
1. Dissolve 2-5 ~ g of RNA in 0. 1 ml of 20 mM sodium aceta te buffer, pH 5.0. 2 . Add 10 J.l1 of metaperiodate reagent and incubate the mixture for 1 h at ° C in the dark .
o
3 . Destroy the excess Na l0 4 by addi ng 20 IJI of 10 % propylene glycol and inc ubat e for a further 10 min as in step 2 .
4. Add 0 .15 ml of 0 .3 M sodium acetate buffer , pH 5.0 , and 0 .6 ml of cold (- 20 °CI ethanol.
Proto col 8. La belling of m RNA by 2 '-O -methylation
Equipm ent and reagents . 1.0 M Tr is-HCI , pH 7.6
5 . Recover the oxidized RNA as described in Protocol 5 , steps 5 and 6 .
• RN A con ta in ing m 7GpppN (Cap 0 ) structu res ( 1 mg /mll
6 . To rem ov e propylene aldehyde , redissol ve the RNA pellet in 1501-11 of H 20 and reprec ipit at e it as in steps 4 and 5.
• [ methyl -3HjAdomet. phenol.chlorof orm ,
Continued
Con tinued 50
elution buffer , 3.0 M pot assium acetate , pH 4.5, and Sephadex G-100 col um n as in Protocol 1; 0.125 M OTT and vacci ni a capping enzym e (guany ly lt ransf erase) as in Protoc ol 6
51
2: Ca p pi ng an d me thyla tion of m RNA Protocol 8 . Continued
Method t . Set up the methyl transfe rase reaction by mixing in a microcentrifuge tube: 5 ~I
. ' .0 M Tris-HCI, pH 7.6 . 0 . 125 M DTT
10 ~t
• [methyl-' H1AdoMet 120 ~Cil
40 ~I
• m ' GpppN RNA • vaccinia capping enzyme (2 - 10 units)
10 ~t
• H2 0
2 ~I to 0 .1 ml final volume
2 . Incubate the mixture at 37 °C for 30 min . 3 . Purify the radiolabelled RNA as described in Prot ocol 1, steps 3 -8.
4. Characterization of caps 4. 1 Strategy Th e first stage in the analysis of cap structures in m RNA is the enzymic remova l of the cap from the mRNA chain . Th e m' Gpp pN a nd Gpp pN are resista nt to vario us nucleases th at hydrolyse phosphod iester bo nds in DNA and RNA . In add ition, 2 ' -O-methy lation of th e ribose moiety in an oligonucleotide render s the 3 ' - 5 ' phosphodiester linkage adjacent to the cap resista nt to RNa ses th at cleave RNA via 2 ' -3 ' phosphate cyclization . Co nsequently, Cap I and Cap 2 struct ures (i.e. m'GpppNmpNp and m'GpppNmpNmpNp) that are common in mRNA s of higher eu karyot ic cells, are resistant to RNase T2, a nuclease th at hydrolyses RNA withou t base specificity. Nuclease PI will also remove m' Gpp pNm from mRNA, but becau se nuclease P I hydro lyses RNA to 5 ' -mononucleotides it is unaffected by 2 ' -O-methyla tion . In additio n, du e to its mecha nism o f phos phodiester cleavage, nuclease P I digestion of cap ped mRNA yields m'Gpp pNm, i.e. cap s withou t a 3 '
Yosuhiro Furu ich i and Aaro n J. Sho tkin characterization of the nucleoside N(m) following the rn' G. Severa l enzymes are known which can cleave the cap pyro phosphate linkage. including the nucleotid e pyrop hosphatases fro m snake venom (22), tobacco (38), potato (39) and yeast (40) as well as a m'G-specific pyrophosphatase from HeLa cells (41, 42). The choice of enzyme is usually dictated by bot h pur pose a nd availab ility. Some enymes may also be used to prepa re RNA chains free of caps as indica ted below. (a) Snake venom nucleotide pyropho sphatase catalyses the conversion of m'GpppNmto m''Gp-t- Pi + pNm. It is often used for the compl ete digestion o f caps (17). However, because it is always contaminated with trace s of phosphodiesterase, snake venom pyrophosphatase cannot be used to obtain 'decapped ' RNA chai ns. (b) N ucleotide py ropho spha tases from tobacco tissue culture cells (38) and potato (39) are, in cont rast, nuclease-free which also makes them convenient enzymes for removing the cap fro m mRNA withou t dam aging the polynucleotide strand: m'GpppNm-
~
m' Gp + Pi + pNm_
Digestion with tobacco pyrophosphatase (followed by phosphatase treatment) is part icularl y valua ble for preparing th e 5 ' -end of mRNA for subseq uent radiolabelling by incubatio n with polynucleotide kinase and ['Y_J2 P] ATP (see Volum e I, Chapter 2, Protocol 2) . (c) m 'Grspecific pyro pho sphatase fro m HeLa cells recognizes the m -G in caps a nd cleaves the pyro phosphate linkage in capped short oligomers (41). This enzyme, m'G-specific HeLa nucleoti de pyrophospha tase , cat alyses the following reaction: m' Gp ppN mpNp -
~
m''Gp « ppNmpNp-
The enzyme preferent ially cleaves the pyrophosphate linkage adjacent to the m'G residue and leaves a 5 ' -diphosphate on the residu al oligonucleotide. Caps conta ining trimethyl G derived from U2 snRNA and unm ethylated GpppN structures are not substrates for the HeLa pyrophosphatase (42). Unlike to bacco pyrophosphatase, the HeLa enzyme apparently does not attack m' G-capped polyribonucleotides that are longer than about 10 residues. Th is enzyme is present mainly in the cytoplasmic fraction of He La cell extract s. In the intact cell, it presumably scavenges capped oligonucleotides resulting from mRNA turnover that might otherwise inhibit the initiation of translation. (d) Yeast m RNA -decapping enzyme is a pyrophosphatase that cleaves m'GpppN and Gpp pN but not pppN end groups in RNA; it has been partially purified from yeast (40). The nuclease-free preparation catalyses the release of m' Gpp and Gpp from mRNA and capped synthetic polyribonucleotides 53
Yo suhiro Furu ichi and Aaron [, Shatki n
2: Gapping and meth yla tion of mRNA at rate s directl y dependent on chain length. For instance. a 540mer was a lO-fold better substrate than a 5Omer. Predictably, pretrea tment with RNase A or nuclease P I renders capped RNAs inactive as substrates. Th e following flow chart illustr ates two different pat hways for ana lysis of caps by enzymatic digestion. Bath pathways permit the purification of nuclea se-resistant 5 ' -terrninal caps from the vast numbers of internal ribonucleotides which then simpli fies the subsequent analysis of cap con stituents .
)G ',"X """\
Reagents • Radiolabelled ca pped mRNA ( - , - 2 mg t mil prepared as in Protocols 1-4,
6-8
• RNase T2 (Celbioc hernl e O. l M sodium acetat e buffer . pH 6 .0 e O. l M sodiu m acetat e buffe r, pH 4 .5 , co nt aini ng 10 mM EDTA
e Nu clea se Pl fCalbiocheml
Meth od A. Diges tion w ith nuclease PI
Capped mRNA
"",,' 'rz
Proto col 9 . Re mova l of caps w ith nuclease P l or RNase T2
1. Dilut e the capped RNA to a fi nal concentration of - O. 5 mg /ml and 10 m M sodiu m ace tate buffer, pH 6 .0, using the 0 . 1 M stock solution of Nuclea se P]
0 .1 M sodium acetate , pH 6.0 . 2. Add nuclease PI to 50 ~ g / ml. 3 . Incubat e t he mixtu re at 37 °C for 30 min .
:=~::~T'"
·'T·"'··;;:~·
" " , : ::,. " '" " ' [ " ",
m"
c• • "'" " .
4 . Proceed to Pro tocol 16 (or another pu rification or analytical techn ique).
8 . Diges tion with RNase T2 1. Dilut e the capped mRNA to a final con centration of -0.5 mg /ml in 0 .1 m l
of 10 mMsodium acetate, pH 4.5. and 1 mM EDTA using the O.1 Mstock solution of 0 .1 M sodium acetate. pH 4.5 . 10mM EDTA. 2 . Add RNase T2 t o 50 U/ml final concentrat ion . 3 . Incu bat e the m ixture at 37 °C for 2 h. 4 . Proceed to Protocol 12 (or another purification or analytical m et hod),
pX'" + pym+ pZ + Pi
Note that init ial cleavage with RNase T2 retain s the methylated nucleosides (X'", Y'") and another residue (Z) in phosphodiester linkage as part of an extended cap structure. By contrast, nuclease PI treatment releases the ca p and y m and Z as 5 ' -mononucleotides; nuclease PI also has a 3 ' -phosphatase activit y. The con stituent nu cleot ides in the capped oligonucleotide ca n be identi fied at various stages by two-dimensional thin- layer cellulose chromatography (43).
4.2 Enzymic analysis of caps 4.2.1 Removal of ca p s from mRNA Protocol 9 describes the removal o f caps fro m mRNAs using either nuclease PI or RNase T2. This procedure may also be employed for prepa ring rad iolabelled cap structures for use as mark ers for analytical purpo ses, e.g. see Figures 5 and 6 (Section 4.3). 54
4 .2. 2 Enzymatic cl eavage of ca ps with nucleotide pyrophosphatases Pro toco ls 10 and 1I described the cleavage of caps with sna ke venom pyrophosphatase and tobacco pyrophosphatase, respectively. Prot ocol 10. Cleavage of ca ps with snake venom pyrophosphata se to ide ntify the Nm in m 7 GpppN m
Reagents e m 7 GpppNm caps removed fr om cap ped mR NA by nu clease P 1 dige stion (see Protacol 9 ) and isolated by pap er elect roph oresis as in Protoc ol 14 (Sect io n 4 .3.2 1. see also Figure 68
• Snak e venom nucleot ide py rophosphatase (Siqm al dissolved in H 20 at - 1 U /m l e 1.0 M Tris-H CI, pH 7 .5 e 10 mM MgCl,
M ethod
1. Dissolve the caps in 8 ~I of 20 mMTris-HCI, pH 7.5, and add 1 ~ I of 10 mM MgC I2 and 1 ~ 1 of nu cle ot ide pyrophosph ata se.
Continued 55
2: Capping and methylat ion of mRNA
Yasuhiro Furuichi and Aaron ] . Shatkin
Protoco l 10 . Con tinued
4.3 Analysis of caps following enzymic digestion
2. Incubate the mixture at 37 °C for 30 m in .
4.3.1 Chro matography on DBAE-cellulose and DEAE- cellulose
3 . Freeze and lyophilize the sample f or subsequent analysis by high-voltage paper electrophoresis (see Protocol 16 ) or two-dimensional chromatog raphy (see Protocol 17 and Figure 7) .
Dihydro xyboryl aminoethyl cellulose (DBAE) is a derivat ive of DEAE-cellulose cont aining covalently-bound dihy droxyboryl gro ups . Its use as an affinity mat rix depends up o n the selective a ffi nit y of 2' - 3 ' cis-diols for dih ydro boryl groups. DBA E- cellulose was developed originally for the isolat ion of 3 ' -te rrninal oligo nu cleotid es from hig h mo lecular weight RNA (45). H owever , it can also be used to iso late the 5 ' -terminal capped oligo nucleotide from RNA digests since the cap possesses a 2 ' - 3 ' cis-diol on the ribose mo iety of m/G . T he proced ure is described in Prot ocol /2. Cap s rem oved from m RNA by enzymic digest ion (Sectio n 3.2) may also be separat ed b y ch romatography on DEAE-cellulose. Protocol /3 describ es the initial separation of caps, which elute from DEA E- cellulose between 0.2 M and 0.25 M Na C!, and su bsequent concentration and desal ting of the caps.
T reat ment of cap ped mR NA wit h tobacco pyro ph osphatase is used to prepare uncapped m RNA (containing a 5 ' -termina l pNm) by releasing the cap constituents as m''G p an d P i. After removal of the 5 ' -monoph os p hat e by phosphatase treatment, the RNA can the n be 5 ' end-labelled by incubation with po lynucleotide kinase and [-y_l2p 1A T P (Vol ume I, Cha pte r 2, Protocol 2). Protocol 1 1. Remova l of ca ps from m RNA w it h to ba c co pyro phosp hat ase Equipm ent an d reagen ts . 0 .2 5 M sod ium acetate buffer , pH 6.0
one of the methods described in V olum e I, Chapter 1, Section 3
e O. 12 5 M DTT • Tobacco acid pyrophosphatase (Sigm a,
Equ ip m en t and reagents
• Phenol:chloro f orm ,
elution buffer , Sephadex G- l00 column , and 3 M potass ium acetate , pH 4 .5 , as described in Protoco l 1
-1000 U/mll e Capped
Prot o c o l 12. DBAE - c el lulose c h ro m atog raphy
mRNA (1 mg /ml l isolated by
e Capped mRNA digested with RNase T2
e Application
pH 4.5. 1 mM EDTA
pH 8.71
e Application buffer plus 1.0 M so rbitol
. 0 . 5 M morpholine buffer , pH 8 .7
Me thod
buffe r (1.0 M NaC!, 0 .1 M MgCI 2 , 20 % DMSO , 50 mM morpholine ,
as described in Protocol 9 ; th is will be -0.5 mg /ml in 10 mM sodium acetate ,
IJ . T. Bake r)
. 5.0 M NaCI
1 . Mix in a microcent rifuge t ube:
. OBA E- cellulose colum n (1 5 x 0 .6 ern ] equilibrated in application buffer
. 1.0 M MgCl 2
. 0 .2 5 M sod ium acetate buffer
1 IJI
. 0 .12 5 M DTT
1 ~I
• tobacco acid pyrophosphatase
1 jJl
• capped mRNA
1 ~I
• H 20 to 10 IJI final vo lume 2 . Incubate the mixture at 37 °C for 1 h. Stop th e reaction by placing th e tube on ice . 3 . Spot an aliquot of the reaction mixture on paper and analyse it accordin g to Protocolll, (Section 4 .3 .5; see also Figure 7) for confirmat ion of the release of rn /Gp (identified by scintillation co unt ing if radiolabelled or by fluorescence under UV light if present in sufficient amount ). 4. Purify the decapped RNA ac co rd ing to Protocol 1, steps 3 -8 .
e Otm et hvrs u tpho x toe HPLC , spectrograde l
(OMSO ;
Pier ce
Method 1. Adj ust t he com posit ion of the RNase T2 -treated starting sample to that of application buffer (buffer A ) as follows :
• • . . .
RNase T2-d igested capped mRNA 0.5 M morpholine , pH 8 .7 5.0 M NaCI 1.0 M MgCl, DMSO
250 ~I 100 ~I 200 ~I 100 ~I 200~1
• H 20 to 1 ml final volume 2. A pply the sample t o the equilibrated DBAE -cellulose column .
The decapped RNA can be l2P -Iabelled at the 5 ' -end by first removing the residual phosphate by phosphatase treatment (e.g. with 10 unit s of calf inte stin al alkaline phosphatase per millilitre in 50 mM Tris -HCI, pH 8.0, fo r 60 min at 37 °C) and then incubating the repurified RN A with polynucleotide kinase an d [y_32P] ATP (44) .
3 . Develop the column wit h application buffer at a flow rate of - 5 ml /h and collect 1 ml fra ctions .
56
57
Continued
Ya suhira Furuichi and Aaron J. Sha tki n
2: Capping and me thylation of mRNA Protocol 12. Continued
Pro t ocol 13. Continued
4 . Assay - 251J1 aliqu ots of each fra ction by sc intillation counting .
8 . Wash t he c o lu mn success ive ly w it h 5 m l ea c h of the fr eshl y prepared
5 . After most of the 3 ' -nucleotides (resulting from the digested mRNA ) have elu ted from the colum n, change the elution abruptly to app lication buffer plus 1.0 M sorbitol and co nti n ue th e eluti on at a fl ow rate of
9 . Finally elut e the caps wi t h 1 ml of 1 M NH 4HC0 3 and lyophilize the eluate .
- 6- 7 ml/h. 6 . Combine the fractions containing t he c aps (t hey elute within - 10 ml after changing to the sorbitol -containinq buffer) and proceed to the next procedure , e.g . DEAE -cellu lose chromatography as des cribed in
Protocol 13.
Protocol 13. DEAE- c e llulose c hromatography
Equipm ent and reagents and 1.0 x 0 .6 cm ) equ ilib rat ed w ith 20 mM 'rrts-H c t. pH 8.0, containing 7 M u rea
e Ceos isol at ed by DBAE - c ellu los e chromatog raph y (so rbitol- elu t ed fra cti on s from Pro to col 12 . st ep 6 1
. ' .0 M Tris-He l. pH 8.0
e Y east
tRNA dige st ed w it h p ancre atic RNa se ; inc ub ate 1 m g /ml t RNA '" w it h 0. 1 mg /ml o f pan cr eat ic RNase fo r 6 h at 37 °C in 0. 1 M Tr is -HCI, pH 8 .0
. 50 mM NaC! in 20 mM Tr is -HCI , pH 8 .0 , co ntain ing 7.0 M u rea e O.3 M N aCI in 20 mM Tr is -HCI , pH 8.0, co nt aini ng 7. 0 M u rea e DEAE - cell ulose col u m n s (4 0 x 0. 6 cm
e A ser ies o f fre sh ly-prepared am mon ium bi carbonate so lution s (50 mM , 75 m M, 0. 1 M an d 1.0 M )
Method 1 . D ilute th e com bined fra ctions of the sorbitol-eluted ca ps 10-fo ld wit h H 20 . 2. A dd , as a marker for the oligonucleotide eluti on po siti on , 0 .5 ml (- 20 A 260 un it s) of a pan creatic RNase digest of y east tRNA .
3. Appl y the mixture to the DEA E- cellulose colu mn and wash the sample in 20 mM Tris -HCI , pH 8 .0, conta ining 7 M ur ea. 4 . Elute the column w ith a linear salt gradient of NaCI (50 ml each o f 50 mM and 0 .3 M NaCI in 20 mM Tris-HCI. containing 7 M ureal. 5. Monitor the eluate c ont in uously at A 260 (or by scintillat ion co unting if radi olabelled y east tRNA has been used ) for m on onu cleotide s ( - 2 charge ), dinucleotides ( - 3 c ha rg e), et c .
6. Colle ct t he fra cti ons conta ini ng cap s (pos ition of - 4 t o - 6 charg e, depending on extent o f 2 ' -O-meth ylati on ).
7. Desal t and co ncent rat e the sam ples con t ain ing t he caps by DEA E c hro m at ography by diluting t he co mbined fracti ons 10 -fold wi t h H 2 0 and apply ing the po oled samp le t o a sma ll DEAE col u mn (0. 6 x 1 ern}.
Continued 58
50 mM. 75 mM, and 0 .1 M NH ,HC0 3 solutions . .. Hadiola belled t RNA m ay be u sed.
Figure 5 illustrates the combined use of DBAE - c ellulose and DEAE-cellulose chrom atography. In thi s example, J2P-labelled avian sarcoma virus RNA (ASY; 1.1 X 10' c.p.m., sp ecific radioactivity 1.5-3.0 x 10' c.p.m.zug) was isolated from purified virus (46), mixed with [lH] methyl-labelled reovirus mRNA synthesized in vitro (see Protocol 1) (10 ug , I X IO' c.p .m .) and digested with 50 unit s o f RNase T2 (see Protocol 9) . After incubation for 6 h at 37 °C , the mixture was applied to a DBAE-cellulose column (see Protocol 12). After the decapped J2 p _ labelled ASY RNA had been removed by washing the column with application buffe r, the column was eluted by applying buffer containing 1 M sorbito l (see Figure 5A) as described in Protocol /2, step 5. The J2P-labelled caps eluted in the same fractions as the [lH J methyl-labelled reovirus mRNA caps, immediately after th e buffer change . The combi ned fractions containing the caps were then analysed by DEAE - cellulose chromatography (Figure 58) as described in Protocol 13. A major peak of radioactivity with a net charge of about - 4.5, which corresponds to the Cap I structure (m'GpppGmpCp) , was resolved by this procedur e. A minor peak eluted later with a net negative charge of about - 5.5; it corresponds to Cap 2 structures containing the rin g-opened derivative of m'G (m" G) which has lost the + char ge (46). These values are based on theoretical charges of + I at th e N7 position of m''G, - 3 for Gppplv'", and approximately - 3 and - 4 for pNp and pNmpNp, respectively . As shown in Figure 68 (inset), m'GpppG melutes from DEAE-cellulose with a net charge between - 2 and - 3, based on the elution of nucleo tide markers . The di vergence from the theoretical charge of - 2 ( + I - 1- 1 - 1) may be due to the stacking effect of the nucleosides in m' GpppGm, resulting in stronger bindin g to the DEAE-cellulose matrix.
4.3.2 Chromatography on paper Cap structures and their derivative s can be resolved by descendin g paper chrom atogra phy on Whatmann 3MM paper in isobutyric acid : 0.5 M NH ,OH (10:6 v/v) as described in Protocol /4 . Th e R F values of cap s and their derivatives in this system are: m'GpppAm, 0.40; m'GpppGm, 0.36; GpppA, 0.25; Gppp C, 0.16; GpppG , 0.08; and GpppU, 0.08. It is important to note that m''G residue of the cap is sensitive to alkaline conditions and is degraded to products that include the 'ring-opened' form in which the 8-9 bond of the m''G imidazole :ing is cleaved with an accom panying loss of the positive charge. Therefor e, it IS not appropriate to u se alk aline solven ts f or t h e characte rizat io n o f caps .
59
2: Capping and methylatian
Yasuh iro Furu ich i a nd Aaron J. Sha tkin
of mRNA
Prot ocol 14 . Paper chr om atography of ca ps A Z4
.-, 0
r&".r
16
e
-
IZ-
Co
8 f'-
..!!.
4f'-
0-
I
RNase T2 di geste d ASV RNA
0
Cap s
20
0
.,
-
-
----
Protocol 9)
0
8
~
x
-
6
-
4 o
-
2
e
.=
2-r -
0 ~
Z
-.I
B -4
-5
I
-6
I I
m7 GpppGDlpCp
X
\.
E!
Co 1-
."
0 40
,
U
60
80
- K)
' pGmpCp m7Gpp
~
100
. I
. W hatm an 3MM paper and chromatography tank
Meth od 'l . M ix the radiolabelled cap sample wi th authentic marker com pound(s) in H20 .
4 . When the solvent front approaches the end of the paper ( - 40 ern}, remove the paper from the tank and air dry it. Locate the marker compounds under UV light and circle their positions using a pencil. 5 . Cut the paper into - 1 cm horizont al strips. Count the strips to locate the radiolabelle d cap .
- 6
-
4
2
~
x
e Co
u
.=
120
Fraction No.
Figure 5 . Sepa ration of capped oligonucleotides by column chromatog raphy on DBA Ecellulose and DEAE-cellulose . ASV genomic RNA (purified from B7 7-ASV that had been 32P-labelled in chick embryo fibroblasts) was mixed w ith reovirus mRNAs labelled in vitro using [me thyl-JHjAdoMet (see Protocol 1I. After digestion with RNase T2 (see Protocol 9 J. the caps were separated from the J2P-labelled nucleotides by chrom atography on DBAEcellulose (panel A l as described in Protocol 12. T he pooled cap fract ions indicated w ere further purified and resolved by D EAE-cellulose column chromatography (panel BI as described in Protocol 13 into Cap 1 st ruct ures (m 7G pppGmpCpl and the slightly more negati vely mpC 7G 7 charged Cap 1 struc tures which contai ned th e ring-opened form of m (m 'Gpp pG p). T he approximate net negative charge is indicated by the scale in panel B. Da ta taken from ref . 46. Open circles , 32P· labelled nucleotides and caps from ASV ANA ; closed circles. ( 3Hlmethyl.labelled reovirus mANA cap (m 1GpppGm pCpl .
60
marke rs
0
- 8
. }
U 0-
nucleo tide
3 . Develop the paper by descending chromatography in isobuty ric acid: 0.5 M NH.OH (10 :6 vlvl.
,
I
and
. 0 . 5 M NH ..OH
2 . Spot the mixtur e on Whatm an 3MM paper and also separately spot samples of the desired marker (s).
Fraction No .
.
• Isobutyric acid
. m 1G pppG . m 7GpppG m. m 1Gppp A . m 1Gp ppA m ma rkers IPharmacia-LK B) • Nu cleoside
(Boehringer Mannheiml as appropriate for t he partic ular separation
Co
80
60
40
• Aadiolabelled caps removed from capped mRNA by nucle ase Pl digestion (see
10 _
-
Buffer A. ~rbi t" ,
zo f'-,-,
~
x
,
I
A
Reagents
Figure 68 illustra tes paper chromatogra phy of cap structures using the procedure described in Protocol 14. 4.3.3 Thin-layer chromatography on PEl- cellulose Perry et al. (47) have described in detail the sepa ration of a mixture of m' GpppAp and m'GpppGp, prepared from RNase T2 digests of 32P -labelled Trypanosoma brucei poly(A) + RNA, by P El cellulose thin -layer chromatography using 2 M pyridinium for mate, pH 3.4, as the solvent. Protocol 15 describes this technique. Figure 48 shows typical data obtained by this method. Prot ocol 15 . Analy s is of caps on PEl- cellulose
Equipm ent and reag ents • Caps derived from mRNA by nuc lease digestion (see Protocol 9) preferably 32p. labelled for ease of detection . PEI- cellulose th in-layer plates (Sigma l and TlC tank . 0 .6 M LiCI or 2.0 M pyridinium formate . pH 3 .4 (see step 21
61
• Marker authentic caps. nucteosides . and nucleot ides (see Protocol 141 • UV light source
Continued
2: Copping and methylat ion of mRNA
Yasuh iro Furui chi and Aaron
J. Shatkin
A
Protocol 15. Continued
,
Method
--
pG
1. Spot a sample of t he caps nea r one end of the plate (oriqin). Separately spot marker co mpounds along t he same origin. 2. Develop the plate with 0 .6 M Liel as so lvent if the caps are de rived f rom nuclease P1 digestion and correspond to m 7 GpppNm structures (see Figure 48 ). A lterna tively . for RNase T 2 der ive d cap structu res (m 7 GpppNp) use 2 .0 M py ridi nium formate , pH 3.4, as the solve nt .
- 10
1
o ~
- 6
eo
the t ank and air dry it in a hood. 4 . Loc at e the m arke rs by UV light and the sample spots by auto radiograph y
- 8
3
E
3. W hen the solvent front is nea r t he end of t he plate, remove the plate from
zj-
0.
~
~
I f-
or co unting .
o o 4.3.4 Paper electrophoresis
- 12
pU
r
, 10
- 4
e,
o
'"
" -2
o
30
20
E
B
Nuclease PI digests of capped mR NAs may be ana lysed by electroph oresis on DE8l paper or What man 3MM paper. T he procedure is described in Protocol 16. Under the conditions descr ibed, the electro phoretic migrat ion positions of the different caps relative to 5 ' pU are: m/Gpp pA'", 0.38; m''GpprsG'", 0.40; Opp pA, 0.50 ; OpppO , 0.5 2; m 70pppA mpO, 0.26; and m 70 pp pOpAmpC, 0.24.
I '" , 3~ 0.1
"~
~
' -2
,
I
0 3
----
x
e'
-
0 2~
.
~ I
"-
"\ ; o. l-=:--:~!::0 10 20 '"
,I
b ....
Fraction no .
X
E
c,
-
. lI-
Protocol 16. Pa per elect ro phores is of caps on DE8l pape r
.,
I I ' -
0-
I'I'lG pwG '"
"
- 3 -4
E
u
'"
c.
"
U
Equipm ent and reagents 10 % acetic acid , pH 3.5 ,1 mM EDTA )
e Caps derived from m RNA by nuclease Pl digestion (see Protocol 9 ) preferably 32 P-labelled for ease of detection
• Marker compounds includ ing authentic caps (see Protocol 14 l
e whatman DEBl paper • Elec trophoresis buffer (1 % pyridine ,
• Apparatus f or one -dimen siona l pa per elect rophoresis
10
20
30
40
Fraction No . Figure 6 . Separation of cap structures by pape r electrophoresis and paper ch romatography . A
mixtur e of 5' -termina l fragments derived from 32P_labelled ASV RNA and [3H 1methyl -labelled reovirus mRNA was obtained by RNase T2 digestion and DBAE -cellulose chrom at oqraphv as in Pr otoc ol 12 and Figure 5A . A tter desa lting as descr ibed in Protocol 13 (st eps 7 ~ 9), the
Method
Figures 3 and 4A show the separation, by electrophoresis on DE81 paper, of cap structures obtained by decapping and recapping a variety of RNAs.
sample was successively digested w ith nuclease P1 and alkal ine phosphatase and analysed by high -voltage paper electrophoresis in pyr idine acetate buffer, pH 3.5 , at 50 V /cm for 50 min (see Prot oc ol 16 l (panel A ). Th e radioactive material m igrating in the pos ition of authentic 7G m pppGm, between the marker nucleotides 5 '·adeny lic ac id (pAl and 5' -guanylic acid lpG l. was extracted with H 20 from the dried paper , and chromatographed on Whatman 3MM pape r using iso buty ric acid : 0 .5 M NH 40H ( 10: 6 v/vl as des cribed in Protocol 14 {pan el Bl . The mater ial in the position of the marker m 7GpppG m was extracted with H 20 from the dried paper and an aliquot was analysed by DEAE -cellulose chrom at ography acco rding t o Protoco l 13 (pan el B, inset) . V isible amounts (A 26Onm - 0. 5) of pC , pA , pG , and pU (in pane l A ) and GpppG and m 7GpppG m (in panel B) were included as standard markers in t h e separ at ion . In addition , the rati o of 32p:3 H (c .p.m .) was plotted t o show the homogeneity of the sample (p anel B, solid circ les and line ).
62
63
1. Spot a sample of t he caps near one end of a 20 x 20 em sheet of the paper (originl with ma rker compounds spo t ted in adjacent positions . 2 . Run the electrophoresis at 500 V for 2 h using electrophoresis buff er.
3 . Locat e the compounds unde r UV light and by scintillation count ing of 1 em slices of the dried pape r.
2: Capping and m eth ylation of mR NA
Yasuhiro Furuichi a n d Aaron J. S ha tk in Prot ocol 17 . Continued e Merke r compounds (see Protocol 141
• Isob utyric acid:O. 5 M am monium hydro xide (5:3 v/v l
• e rooan-z -or.conc. HC t:w ate r (70: 1 5: 1 5 v/v/v) ...... pAm
Method
pem pA pm
1. Spot a sample of the radiolabelled digested cap near one corner of the cellulose thin-laye r plate (origin). Also apply a sample of markers to the
70.,) pC
origin .
pU m
pGm , p G '::::::
2 . Develop the plate in the first dimens ion by ascending chromatography using isobutyric acid:0 .5 M amm onium hydroxide (5: 3 v/vl as the solvent.
pU
3. D ry the plate in a hood at room tempe rature . 4 . Turn the plate through 90 0 and develop it in the second dimension by ascending chromatography wi th propan-2-o l:con. HCI :water (70: 15: 15 v/v /v l as the solvent. 5 . Locate the marker compounds under UV light and circle them w ith a pencil. 6. If the sample contains 32p, autoradiograph the plate .
Figure 7 . T w o-d imensiona l th in-layer cellulose chromatog raphy of cap constituent s. Relative mo bilities are shown fo r m 'Gp, 2 '-O -methy l·S '·ribonucleotides. and inorganic phosphate (PI) analysed on an A vicel -SF cellulose th in-layer plate (4 3) , The numb ers and arrows indic ate th e order and direct ion 01 the two chromatography system s as described in Proto col 17. The first dimension solvent was isobutyric acid:O.5 M NH 40H (5: 3 v /vl and the second dimen sion solv ent was propan -2 -ot:HC I:wate r (70: 1 5: 15 v/v /v) .
4.3 .5 Separation of ca p constituents by two-dimensional chromatography Cap constituents including nucleotides, m''Gp . and the penultimate 2 ' -0methyl nucleotides can be identified by two-dimensional chromatography on cellulose thin layers after digestion of the caps by snake venom pyrophosphatase or tobacco pyrophosphatase (see Section 4.2.2). The method, describ ed in Protocol / 7, can also be applied for the cap-containing oligonucleotide obt ained from capped mRNA by RNase T2 digestion (see Protocol Fi (48). Figure 7 shows typical data.
Protocol 17 . Two-dimensional separation of 2 '-0-methylated nucleotides us ing th in-layer cellulose chromatography
Equipm ent and reagen ts
1-4 . 6 -8 and digested w ith nucleotide de scribed
7 . If the sample is (methyl·3H] -labelled. scrape portions of the cellulose layer off w ith a fine spatula and count the radioactivity using a scintillation count er.
4.3 .6 Other m ethods
Capped mRNAs and caps have also been isolated by affinity chromatography on m-aminophenylboronate agarose (49) (as an alternative to the DBAE-cellulose chromatography described in Protocol /2) and with a m'G-specific antibody immobilized on Sepharose (ref. 50), a matrix which is not available commercially but ma y be prepared as described in Volume I, Chapter 5, Prot ocols / - 3 .
Acknowl edgements The authors thank Dr Kiyohisa Mizumoto (Kitasato University, Tok yo, Japan) for valuable discussion and Akiko Fujii and J anet Hansen for excellent secretarial assistance .
References
• Ca p constituents prepa red and labelled with 32 p or 3H as described in Protocols pyroph osphatases as Protocols 10 and 1 1
or
e Cellulc se th in-layer plat e (Avicelt-Sf cellulose thin -layer plate. 15 x 15 em l
and TLC tank
in
Continued 64
1. Shatkin, A. J . (t976). Ceil, 9. 645. Banerjee, A. K. (1980). Microbial. Rev. , 44. 175. Maniatis. T. and Reed , R. (1987). Nature. 325, 673. HsuChen, c.-C. and Dubin, D. T. (1976). Nature, 264, 190. Mattaj, 1. (1986). Cell. 46. 905.
2. 3. 4. 5.
65
2: Capping an d m eth ylation of mRNA 6. Agab ian , N. (1990). Cell, 61, 1157. 7. Bang s, 1. D., Crain, P . F., Ha shizume, T., ~lcClosk ey , J. A., and Boo thro yd, J . C. (1992). J. Bioi. Chem., 267, 9805. 8. Furui chi, Y., LaFiandra, A., and Shatkin, A. J. (1977). Nature, 266, 235. 9. Green, M. R., Maniati s, T ., and Sharp, P. A. (1984). Cell, 38, 731. 10. Kon arska , M. M. , Padge tt, R. A., and Sharp, P. A. (1984). Cell, 38, 731. I I. Ohno , M., Sakamoto, H., and Shimura, Y. (1987). Pr oc. Natl A cad. Sci. USA . 84,51 87. 12. Shat kin, A. J . (1985). Cell, 40, 223. 13. Ha gler, J . and Shuman, S. (1992). Science, 255, 983. 14. Shuman, S. (1985). J. Bioi. Chem., 2601, 9690. 15. Co ng, P . and Shuman, S. (1992). J. BioI. Cbem. , 267, 16 424. 16. Higman , M. A., Bou rgeois, N., and Niles, E. G. (1992). J. Bioi. Chem . , 267, 16430. 17. Shum an, S. and Moss, B. (1989). J. BioI. Chern. , 264, 21 356. 18. Beat on, A. R. and Krug, R. M. (1986). Proc. Nail A cad. Sci. USA , 83, 6282. 19. Georgi ev, 0 ., Moss, J ., and Birnstiel, ~1. L. (1984). Nuel. A cids Res. , 12, 8539. 20. Schnierl e, B.S., Gershon, P. D., and Moss, B. (1992). Proc. Natl A cad. Sci. USA , 89, 2897. 21 . Miura, K.-I. , Watanabe, K., Sugiura, M., and Shatkin, A. J . (1974). Proc. Natl A cad. Sci. USA, 72, 3979. 22. Furui chi, Y., Morgan, M., Muthuk rishnan, S., and Shatkin, A. J . (1975). Proc. Natt A cad. Sci. USA , 72, 362. 23. Furui chi, Y. and Miura, K.-1. (1975). Nature, 253, 374. 24. Fu ruichi, Y. and Miura, K. (1973). Virology, 55, 418. 25. Shatkin, A. J . and LaFiandra, A. 1. (1972) . J. Virol. , 10, 698. 26. Joklik, W. K. (1962). Biochim . Biophys. Acta, 62, 290. 27. Furui chi, Y. and Shatkin, A. J . (1976). Proc. Nat! Acad. Sci. USA, 73, 3448. 28. Furui chi, Y. and Shatkin, A. J . (1977). Nucl. Acids Res., 4, 3341. 29. Wei, C. M. and Moss, B. (1975). Proc. Natl Acad. Sci. USA , 72, 318. 30. Aviv, H . and Leder , P . (1972). Pro c. Natl Acad. Sci. USA , 69, 1408. 31. Kreig, P . A . and Melton, D . A. (1987). In Methods in enzymology , Vol. 155 (ed . R. Wu) , p.397, Academi c Pr ess, San Diego. 32. Mizumot o, K., Kaziro , Y., and Lipmann, F. (1982). Proc. Natl A cad. Sci. USA , 79, 1693. 33. ltoh , N., Mizumoro. K., and Kaziro, Y. (1984). J. Bioi. Chem . , 259, 13 923. 34. Wang, D. and Shatkin, A. J . (1984). Nuel. Acids Res., 12, 2303. 35. Yagi, Y., Mizumoro , K., and Kaziro. Y. (1 984). J. Bioi. Chem. , 259, 4695. 36. Barbos, E. and Moss, B. (1978). J. Bioi. Cbem ., 253, 7698. 37. Hand a, H ., Mizumoto, K., Oda, K., Okamoto, T. and Fukusawa, T . (1985). Gene, 33, 159. 38. Shinshi, H ., Miwa, M. , Sugimura, T., Shimctohno, K., and Miura, K. (1976). FEBS Lett. , 65, 254. 39 . Zan-Kowalczewska, M .; Bretner.M .• Sierakowska, H .• Szczesna, E., Filipowicz. W.. and Shat kin, A. J . (1977). N uc/. A cids Res. , 4, 3065. 40. Stevens , A. (1988). Mo lec. Cell. Bioi., 8, 2005. 41. Nuss, D. 1.., Furuichi, Y., Koch , G ., and Shatkin, A. 1. (1975). Cell, 6, 21. 42. Nuss. D. I.. and Furuichi, Y. (1977). J. Bioi. Chern. , 252, 2815.
66
Yasuhiro Furuichi and Aaron ]. Shatkin 43. Nishimu ra, S. (1979). In Transfer RNA : structure, prope rties, and recogni tion (ed. P. R. Schimmel , D. 5011, and J . N. Abel son ), Append ix Ill, Chromatographic mo bilities of modified nucleo tides, p. 551. Co ld Spri ng Ha rbor Laboratory, Cold Spring Harbor, NY. .t4. Efstrat iadis, A., Vournakis, J . N. , Donis-Ketler, H .; Chaconas. G ., Dougall, D. K., and Kafat os, F. C. (1977). Nu el. A cids Res. , 4, 4165. 45. Rosenberg, M. (1974). Nuel. Acids Res., 1, 653. 46. Furuichi, Y . , Shatkin , A. J ., Stavne zer, E., and Bishop, J . M. (1975). Nature, 257, 618. 47. Perry, K. K., Watki ns, K. P ., and Agabian , N. (1987). Proc, Natl A cad. Sci. USA , 84, 8190. 48. Furuichi, Y., Morga n, M., Shatkin, A. J ., Jelinek , W., Saldin-Georgief f, M., and Darnell, J . E. (1975). Proc . Natl A cad. Sci. USA , 72, 1904. 49. Wilk, H. -E., Kecskemethy, N., and Schaf er, K. P . (1982). Nuel. Acids. Res., 10, 7621. 50. Munns, T. W., Oberst, R. J ., Sims, H. F., and Liszewski, M. K. (1979). J. Bioi. Chem. , 254, 4327.
67
RNA editing in mitochondria LA RRY SIMPSON , AG DA M . SIMPSON, and BEAT BL UM
1. Int roducti on
1.1 Post-transcriptional modification of eukaryotic mRNAs Several examples of specific modifications of nucleoti de seq uences within coding region s of mRNAs, and , in some cases , within rRNAs, have been observed in recent years. This process, which is know n as RNA editi ng, ranges from the single C ~ U substitution in th e mamm alian apolipoprotein B (I, 2), to the multiple C ~ U substitutions found at specific sites in plant mitochond rial mRNAs (3, 4), the multiple C insertions fo und in mitochond rial mRNAs and rRNA s in Physarum polycephalum (5), and th e multiple U insertio ns and deletions fou nd in the mitochondria of kineto plastid protozoa (6). The use of the term , RNA editing, fo r these various types of modifications does not necessa rily imply th at th e mechani sms a re identical. In fact , the polym erase 'stuttering-induced' addition of extra G residues at specific sites in mRNAs from negative stran d RNA viruses is also terme d ' RNA editing' (7). Th e editi ng occurring in the mammalian apolipoprotein B and mRNA appea rs to be due to a site-specific enzymatic deamination of cytidine (8- 13), but nothing is known about the mechanism of the multiple C ~ U tra nsition s in plant mitoch ond ria. Likewise, nothing is known abo ut the mechani sm o f C-insertion in Physarum mitochondrial RNAs. Th e best understo od case is kinetoplastid RNA editing which is described below.
1.2 RNA editing in kinetoplastid protozoa An unusu al type of RNA processing occurs in the single mitoc ho ndrion of kinetoplastid protozoa. Man y of the mRNA tran script s of structura l genes encoded in the maxicircle DNA molecules are modi fied within coding regions in a post-transcriptional process characterized by the insertion and, less frequent ly, the deletion of urid ylate (U) residues (14). This process , known as RNA editing, involves small RNA molecules , guide RNAs (gRNAs), which Specify th e sequence information required. Two models have been proposed for the mechanism of this process (15, 16). Both mode ls prop ose an initial
3: RNA editing in mitoch ondrio
Larry Simpson , A gda M. S im pson . and Beat Blum
base-pairin g inte ract ion betwee n the 5 ' -end of a specific gRNA and the mRNA immediately 3 ' to the pre-edited region (PER) to form the 'anchor' hybrid . Th e enzyme cascade model then proposes a specific cleavage at the first mismat ched nucleotide of the mRNA, followed by the addition o f a U resid ue to th e liberated 3 ' -end, and an RNA ligation to seal the mRNA backbone. Each U residue added extends the initial hybrid between the mRNA and gRNA by forming an add itional base pair to either A or G 'guiding' nucle otides of the gRNA. Thi s process is repeated until a complete hybrid between gRNA and mRNA is ach ieved. The trans-esterification model (16, 17) proposes that the cleavage-ligatio n occurs by mean s of two successi ve trans-e sterification steps . Th e first of these involves the 3 ' -term inal oligo(U) tail of the gRNA as the so urce of the Us for transfer. Formation of gRNA :mRNA chimaeric molecule s attached within the PER ha s been ob served both in vivo among kinetoplast RNA (kR A) molecules (16) and in vitro using mitochondrial extracts (18, 19). Formation in vitro is dependent on complementary anchor sequences in th e gRNA and mRNA (20). However, the available evidence does not allow a distinction to be made between the chimaeric molecules formed by transesterifi cation or by cleavage -ligation . Th e trans-esterificat ion model is attractive due to its clo se analogy with RNA splicing . RNA editing is restricted to genes encoded by the mito chondrial genome which consi sts of a network of catenated DNA minicircles and maxicircles . T his network, referred to as kinetoplast DNA (kDNA), is contained within the single mitochondrion which retic ulates throughout the entire cell and is loc ated in close pro ximity to the basal body of the flagellum (21). The genes for the mitochondrial rRNAs an d a set of mitochondrial structural genes, several of which undergo RN A editing , are encoded on the maxicircle DN A . The maxicircle DNA of Leishmania tarentolae co nsists of 20-50 copies of a 30 kbp circu lar molecule . H owever, the bu lk of kDNA comprises minicircles, which ar e pre sent in high cop y number (approximately 10' molecules per net work) a nd display multiple sequence classes. In L. tarentolae each copy of the DNA minicircle is orga nized into a conserved region of approximately 170 bp and a variabl e region of approximately 700 bp which defin es the sequence class. Each seq uence class encode s a single unique gRNA located app roximately 150 bp from the end o f th e conserved region (22). In other kinetoplastid species, th e nu mber of co nserved region s varies from one to fou r. In Trypanosoma brucei th ere is a sing le conserved region, but the var iable region encodes three gRNA genes (23). A region of 'bent' DNA of unknown funct ion exists either adjace nt to the co nserve d region in the minicircles of L. tarentolae and T. brucei , or at 90 0 to the con served region in the larger minicircles of Crithidia fascicu lataThe presence of this bend decre ases the electrophoretic mo bility of the DNA in po lyacry lami de gels as compa red to agarose gels a nd is pa rt ially responsible fo r the increased complexity o f the polyacrylamide gel profiles of restr iction enzyme-digest ed kD NA (24). Th e complexity of minicircle DNA varies between kineto pla stid species. So me species, which lack the ability to live in an insect
host , ha ve minic ircles o f a single sequence class. The kD NA of T. brucei con ta ins approxima tely 300 differ ent minicircle sequence clas ses; the kDNA of L . tarentolae contains at least 17 dif ferent sequence classes in di ffer ing abundances (25). The complexity of restr iction enzyme-digested kDNA, as visuali zed by electrophoresis in gradient polyacrylam ide gels, can be used to clas sify and type different str ains with in a species, at least fo r Leishmania and Trypanosoma cruzi. The term , schizo derne, was coi ned to ind icate orga nisms with simila r kDNA restriction profiles (26, 27 ). T he kDNA restri ction pro file is a very useful molecular marker since strains are frequently mislabelled, even some ava ilable throu gh the American Type Culture Collectio n (28). In thi s chapter, procedures are described fo r the growth of kinetoplastid cells, isolation of th e kinetoplast-mit ochondrion fr actio n, isolat ion of kDNA and kRNA, identification of gRNAs by computer a na lysis, isolation of gRNAs by hybrid selection, and assays for several mitochondrial enzymes possibly involved in RNA editin g including the termina l urid ylyl tran sferase (TUTase), an RNA ligase, a cryptic RNas e, and a gRNA :m RNA chimaeric-forming activit y.
The kineto plastid protozoa comprise a larg e group of parasitic flagellated cells with a single mu ltilobular mitochondrion containing the kDNA network of caten ated circular DNA molecules (29). Th e kin etoplastid flagellates belong to 8- 10 genera, several of which are digenetic with a life cycle that involves successive vertebrate (or plant) and invert ebrate hosts, and several of which are monogenetic with a life cycle in a single invertebrate host : The monogenetic species such as Crithidia or Leptomonas have simpler nutritional requirement s than the digenetic species such as Trypanosoma or Leishmania. In the latter case, the stage of the cycle in the insect vector is genera lly ea sier to culture th an the stage in th e vertebrate host. Phytom onas is a digenetic species which inhabits a pla nt host and an insect vecto r. Unlik e the other digenetic species, the stage from the plant host grow s readil y in simple medi a and may provide a model system for the study of metabolic changes occurring during the life cycle (F. Opperdoes, per sonal comm unication). RNA editin g has been studied in three species to date: T. brucei, L. tarentolae, ~nd C. Iascicukna (6, 14, 29-31). The maj or di fferences between these protozoa invol ve the complexity of the minicircle DNA populat ion and the correlated extent o f edit ing of the three cryp togenes-ND7, CO lli , and MURF4. In add'uion, . regulation of RNA edi ting has been ob ser ved to occur during the life Cycle of T. brucei (32. 33). Th e author s have chos en to use the saur ia n leiShmania , L. tarentolae, as a mod el system to st udy RNA editing for the follow ing reaso ns:
70
71
2. Growth and maintenance of kinetoplastid protozoa 2.1 Choice of species for experimental work
3: RNA edit ing in mitochondria
Lorry Sim pson, Ag da tvl . Simpson , and Beat Blum
(a) The cells are not pathogenic for humans, which is a major advantage in term s of growing large amounts of cells in the laboratory for the isolation of mitochondria. nucleic acids . and enzymes. (b) The cells grow rapidly (6-9 h division time) in brain heart infusion (BHI ) medium . an easily prepared rich medium. without the need for seru m supplementation. (c) The cells can be frozen for storage.
Prot ocol 1. Growth and storage of L. tarentolae s tock c ulture s
Equipment an d reagents . 2 mgfml haemin: add 2 mg haemin (Sigma H ~2375 1 per ml of 50 mM NaOH , stir for 30 min. and sterilize by filtrat ion through a 0 .22 urn membrane filter (Nalgel; store at - 20 °C
(d) Cell concentrations of 4 x 1()8 cells/ml can be obtained at stationary phas e.
• Tissue-cu tt ura grade water (prepared using Barnstead Nano-pure or Millipore M illi-Q cartridge filtra tion system ) (Barnstead Co .; Milli pore Co rp.)
(e) The cells can be ruptured, after swelling in hypotonic medium, to yield an intact kinetoplast -mitochondrion fraction which is active in transcription and po ssibly also in RNA editing.
e Pyrex bottles . 500 ml (Fisher. "064 14- 1C ) w ith screw caps (Fisher. "06-414-2A I. for auto claving solutions
(f) Several recently-isolated strains exist for comparative studies.
(g) The minicircle DNA and. ther efore, the gRNA complexity is limited, making it feasible to determine a complete list of gRNAs for all th e genes kno wn to be the subjects of RNA editing.
e Brain heart infusion (SHU medium": dissolve 37 9 SHI po wder (Di fco) per
• 66 0 08-3091
4 . Every 2-4 days, aseptically remove most of the culture , leaving about 0. 0 5- 0 . 1 ml and add 5 ml of fresh BHI medium.
B. Storage and recovery of cells 1. Dispense 0 .5 ml aliquots of health y log-phase cultures (approximately 8 10 cells /ml) into sterile 2 ml freezer vials.
2. Add 0 .5 ml of BHI medium containing 20 % glycerol.
~I a c e the vials in a 1 litre beaker filled with cotton wool and put the beaker In a - 70 ° C freezer. This allows slow freezing of the cells.
4 . The next day transfer the vials to liquid nitrogen for long-te rm storage .
~ ecove r. live cells from s~ored cultures by th aw ing the vials rapidly and inoculating the c.ontents Into 2 -5 ml of BHI medium . Check the cells by phase-contrast microscopy; they should be motile immediately after tha win q .
5.
describes the maintenance and storage o f stock cultures. To avoid accumulation
72
eF reezer vials. ster ile. polypr opylene 2 ml (V an W aters and Rogers
3 . C~ ec k the growth of the cultures daily using an inverted . phase-contrast mlcroscop.e . Cultures ca~ be monitored more accurately for the absence of. bacteria by screening we t-mou nt slides under phase-contrast microscopy at 400 x or 1000 x magnification . Leishmania cultures have a characteristic wav e-like appearance to the eye that is quite different from the homogeneous turbid appearance of bact erial cultures. Also the smell of the healthy culture is fruity and quite characte ristic.
3.
of genotypic changes by prolonged culture, begin a new stock culture from frozen stocks every few months.
e ln vert ed. phase-cont rast micros cope
1. Inoculate healthy (motile) L. terentotse cells at 0 .6-1 .5 x 10 6 cells/m l in 5- 10 ml BHI medium in 25 ern? tissue culture flasks. 2 . Grow the cultures at 27 "C.
(b) The main disadvantage of C. fasciculata is that the cells swell poorl y in hypotonic buffers and do not break by the use of shear forces alon e. On the other hand , they grow extremely easily and to high den sities in BHI medium and also in a defined medium . They also plat e on agar surfaces mo re easily than either L. tarentola e or T. brucei.
Initial stocks of L. tarentolae (UC strain) can be obt ain ed from Dr L. Simpson, Howard Hughe s Medical Institute, UCLA, Los Angeles, CA 90024. Prot ocol I
e Tissue culture flasks (qlass: Co rning " 25 100; 25 cm 2 )
A . Maintenanc e and monitoring of cultures
(a) The proc yclic cells of T. brucei require complex media with seru m and do not reach high cell densities . Th e blood stream cells must be grown either in a rod ent host or in a complex medium in which they only reach low cell den sities (34).
2.2 Growth and maintenance of L. tarentolae
e SH I medium cont aining 20 % glycerol. sterilized by auto claving in Pyrex bottles
Method
In co ntrast. the other two protozoa in which RNA editing is kno wn to occur have some disadvantages:
The maj or disadvantage of work ing with L. tarent olae is th at the biology of the parasite within the natural lizard host is essentially unknown. Therefore, one cannot study the vertebrate stage of th e life cycle. It is clear. however. that the leishmania which infect lizard s form a sub gro up of the genus Leishmania (28. 35), which also cont ains th e mamm alian path ogenic leishm an ia such as L. major. L. mexicana, and L. brasiliensis.
lit re of tissue culture grade water; auto clave in Pyrex bottles at 120 °C for 30-40 mine; w hen cool. asept ically add haemin to 10 mg /ml ; the med ium is stable for several months at 5 ° C
:
~~
antibi.otics are necessary if as~Ptic tech niques are used and sterility is preserved. e medium can be autoclaved directly in glass culture bott les or in glass bott les for
storeqe.
73
3: RNA editing in mitochondria
Larry Simpson . Agda M. Simpson . and Beat Blum
2.3 Cloning of kinetoplastid protozoan stock cultu res
Pro t o co l 2 . Continued
. .id stoc k s are unc Ion ed and may contain" several Most laboratory kinetoplast different strains . Soon after receipt, cells should always be cloned either by limiting dilution in microtitre plates or by growth on 0.7-.'.0'70 agar / BH Ihaemin plates and selecting sing le co lo nies . O ne sho uld verify the identity of the strain or even the species by a combination of light microscope morphology and molecular characteristics . Schizodeme analysis by companson of kDNA sequences (see Protocol 4) is an easy method to use for verifying the Identity of a strain of T. cruz} or Leishmania.
2.4 Growth of cultures for preparation of kinetoplast components
,
Cultures up to I litre in volume can be grown in 3.8 litre rolle r bottles usmg a standard ro ller-boule cu lture apparatus at 16 r.p .rn. H arvest the cells b y centrifugation at 2500 g for 10 min at 5°C. Grow larger cu ltures .m a ~llrrer culture apparatus with forced aeration and harvest the cells by filtration as described in Protocol 2. Ce lls can also be grown in a microbiol.ogical fermento r with aeration and stirring. H owever, the addition of silicon antifoarn (Antifoarn B, Sigma) is required to prevent foaming .
4 . Harvest the cells when they have reached a density appropriate for the subsequen t preparat ion Isee later Pro tocols ). To do this , cool the culture by immersing the cul tu re bottle in ice water whi le sti rring the cu ltu re . Co nce nt rate the c ulture t o app roximately 100-200 ml by filtration using the M illipore Pellicon transverse fil ter system .
5 . Wash the cells by addit ion of the desired wash medium to the concentrated cell suspension and reconcen trat e the cells .
6 . Flush the washed cells out of the filter system by a f inal rinse w ith wash medium .
7. Concentrate the cells again and then pellet them by centrifugation at 2500g for 10min at 5 °C . 8 . For kDNA iso lation. resu spend t he cells in SET medi um and fr eeze them at - 70 ° C (see Protocol 3) . For isolation of the kin etoplast.mitochondrion fract ion , process t he cells immedia te ly as described in Protocol 6 .
3, Kinetoplast DNA 3.1 Introduction
Protocol 2. Growth and ha rves ting of large-scale cultures
Equipment and reagents • Sti rrer culture apparatus (Bellea 15 litre or 36 litre bottles, 111964·15000 or 36000, with overhead stirrer # 7664-001 10 and sta inless steel impeller. # 19 64 ·6 00 1 51 • A ir pump or source of compressed air fitted w ith a 0 .2 IJm Gilson
microbiological filter on input and ex it lines
e Compressed O 2 (medical grade ); required for larger cu ltu res e Transverse filter system (Mlllipore Pe1licon with 0.45 IJm Durapore filter casset~~ l
with 8 1itres/min peristaltic pump (MIIIIpore Masterflex XX80 ELO 01 ) e t.oq-phese culture of L. tarentolae and BHI medium as in Proto col 1
Method Inoculate BHI medium with L. ts rentolee at approximately 1.2 x 10 6 cells /rnl 1 . usi ng a log-phase cu lture. For 151it re bottles use 180 ml of cul ture at a ce ll density of 10' cells /m!. 2 . Stir the culture at maximum rate at 27 °C . Blow filtered air int o the
?~tt l e
(but not through the m edium ) at 6-8Iitres /m in . Filter the exhaust air Into a hood exhaust.
3. If larger cell yields are des ired, substitute O2 for air after a~ir~x~::~e~~
50 h. This will allow cells to grow in toq -phase up to ~ ce e .h approximately 4 x 108 cells/rnl , which IS twice the density reached Wit air alone .
Continued 74
Kineto plast DNA is present in the form of a single giant network per cell Composed of approximately 10" catenated min icircles and 20-50 maxici rcles (2 1). T he kDNA netwo rk has a sedimentation coefficient of 4000 S units and is relati vely resistant to shear forces d ue to its co mpactness (36, 37). It is easily isol ated from a sheared to ta l cell lysate by sedimentation through CsCI (27). Th e maxicircle DNA represents 5"70 of the kDNA and can be iso lated on the basis of its relatively higher A + T content (84"70 A + T vers us 550/0 A + T for minicircle DNA ) after its release from the network by digestion with a restriction enzyme that cut s only once or infrequently (38) . The complete sequence of the 23 kbp maxicircle of T. brucei is known , and 21 kbp of the 30 kbp maxicircle genome of L. tarentolae has been sequenced (Genbank entry LEIKPMAX). The st ructural genes are clustered in -17 kbp and t he remainder, which is known as t he divergent region , represents tandem repeats of differing complexities. Mi nicircles and maxi circles replicate once per cell cycle in a n S phase which is synchronous with t he nuclear DNA S phase . Minicircles are randomly dec atenated , replicate the network and are recatenated at two sites at the ends of the kDNA nucle oid body (39). Replicated minicircles are 'nicked' or 'ga pped ' , and are sim ulta neo usly covalently clo sed early in the G 2 phase of the cell cyc le. Maxicircle DNA replicates by the rolling circle me chanism and a low concentration o f linearized maxicircle intermediate molecules can be f o un d fre e of the network (40). Therefore, to maximi ze yield s and to obtain covalently clo sed circular DNA mol ecules, stationary phase cells mu st be used for kDNA
isolation.
75
Lorry Simpson . Agda M . Simpson and Beat Blum
3: RNA editing in mitochondria I . of the minicircle DNA population is characteristic of each The camp exit y . urthermore in Leishmania and T. cruzi, a specifi c kinetoplastId species. F 'thin the minicircle population defines each particular seq~ence h~ter~genell~i;:yrenc et al. (41) ha ve suggested that the genetic reaso n stram ? r sc lZO em~. " the long-term genetic isolation of each schizoderne h In the ca se of T cruzi the y have fo r thi s heterogenelt) IS " . I al line without gene exc ange . or . c o n , loner ' (Tibayrenc and Ayala, personal communicatIo n) C01~:~c;~ee t;~:e ~ines of parasites which have been clOnal~y de rived ~nd to millions of years. In any case, the term schizod eme IS ISolated for perhaps. I term to describe T. cruzi or Leishmania strains t hat an appropnate operatIona differ in k DNA mini circle sequences.
\
13 d
Me thod 1 . Harvest the c ells from rol ler cult ures by cen t rifugati on (10 m in at
2000-3000 gat 5 DCI or by filtration for large cult ures (s ee Protocol 2 ). 2 . Wa sh the cells by resuspend ing them in at least 50 v ol. of SET buffer and repeating step 1 .
3 . Resuspend the cells at 1.2 x 109 cells /ml by st irring and pipetting in SET buffer ensur ing th ere are no cl umps . Add 0 .02 v ol. of 10 mg /m l prona se and 0.1 vol. of 30% sarkosvl . Incubate at 60 °C for 1-3 h with occasional shaking until the lysate clarifies .
4 . Pass the viscous ly sat e t hro ugh an 18-gaug e sy ringe needle at 25 p .s .i. to shear nuclear DNA selectively and decrease the viscosity. For small
volumes (10 -100ml) use a 12ml syringe fitted with an 18-gaug e
3 2 Isolation of kDNA .
Pro to co l 3 . Continued
ibes the isolation of kDNA from stationary phase cultures of escn
Protoco L. larenrolae .
needle and force the lysate through it by hand. For large volumes , use a dispensing pressure ve ssel wi th the pressure supplied by compressed air .
5 . Centrifuge the lysate for 1.5 h in an SW28 Beckman rotor at 22000 r.p .m . at 5°C . This pellets the crude network DNA.
Protocol 3 . Is olat ion of kDNA
6 . Add 1-2 ml of 1010.1 TE buffer to each gelatinou s pellet. Resuspend
Equipm en t and reagents . SET buffer10. 15 M NaC!.0.1 MEDTA . 10mM Tris -Hel . pH 7.5} • Stationar y pha se culture of L. tarentofae;
each pellet by shaking and pour t he pellets from all the t ubes into . , 0 /0. 1 TE buff er (10 mM Tris-HC !.
pH 7.9.0.1 mM EDTAI • '0" TEbuffer (10 mM Tris-HCI, pH 7.9.
1 mM EDTAI
about 4 x 10 8 ce llsfml for large aer ated cult ures (s ee Proto col 2 1 or 1.5-2.0 x 10 8 cells /ml for roller-bott le cult ures
. , 0 mg lml pronase iCalbiochem)~ dissolve the pronase in SET and pred..g es~ the solution for 30 min at 37 DC to tnactivete nucleases
. 30 % Sarko syl: stir 30 9 of sodium sercoslnate in 60 mt of water at 60 ~ C ; add w ate r to 100 ml; store the solution in afiquots at - 20 °C • Hypodermic syringe (12 mil fitted with an 18-gauge needle or . Dispensing pressure vessel (M i~1ipore XX67 OOP 05 or 10. for the .5 htre or 10 litr e sizes. respectively): this has an 18-gauge needle fitte d via a Luer-Iock adapter to one outlet and a pressure release valve attachedto anotheroutlet; use a quick release tube adaptor t? attach the vessel to a compressed air supply e Beckman ultracentrifuge, SW28 rotor, and 36 ml polyallomer tubes (or equivalent) e CsCI Idenaltv gradient or optical grade)
a 250 ml polycarbonate screw-cap flask. Use 6 ml of 1010 . 1 TE buffer per 1-2litre of orig inal culture. Shake the pellets vigorously for 30 m in t o dissolve the crude network D NA . The crude network D NA can be stored at 5 ° C.
e 250 ml polycarbonate screw -capped flask e 10 mg/ml ethidium bromide
7 . Prepare CsCI step grad ients in SW28 polyallomer tubes . Slowly introduce
. Upper CsC' solution: 37 .629. Cs.CI in 62 ml of lO ll TE buffer [refractive Index at 25 °C = 1.37051
8 . Layer 6 ml of the resuspended crude kDNA solution (f rom step 6) on each gradient and centrifuge the gradients f or 15 min at 20000 r.p.m . at 20 °C in an SW28 roto r. Allow the rotor to slow down w ith the brake on . The kDNA networks will sed iment to the int erf ace between the lower and upper CsCI solutions while nuclear DNA , RNA , and pro tein will rem ain in t he upper CsCI solution . If the shearing of the lysate was insuff ici ent. some nuclear DNA w ill also sediment close to the interface .
e l.ower CsCI solution : 29.2 9 CsCI in 20 ml of 10" buffer plus 0 .14 ml .of 10 mg/ml athidlum bromide (refracti ve index at 25 °C = 1.4040): note only the lower CsCI solution contains dye
9 . Visualize the kDNA band at the inte rface with the UV lamp . Seal the top of t he tube with a rubber stopper f itted w ith a needle attached to a 20 ml syringe. Puncture the bottom of the tube and use the syringe to collect the kONA band into a 12 ml polystyrene centrifuge tube .
e Peristaltic pump • UV microscope e u v lamp (254 nm) e Butan-1-01 saturated with water • 12 ml polystyrene centrifuge tube • Sec-butanol e Phenol:chloroform (1: 1, v/vl e DAPI (4', 6'-diamidino-2 phenylindole. Sigma " 013 88) e 2.0 M NaCI
76
6 ml of t he lower CsCl solu tion below 2 4 ml of the upper CsCI solution using a pe rista ltic pump to give a sharp interface.
10. Remove the ethidium bromide by extracting twice w ith an equal v ol. of
water-saturated butan-I-cl . 11. Dialyse the kDNA against 41itres of 1010.1 TE buffer. 12 . Con centrate the kDNA solution to approximately 400 IJI by several extractions with sec -butanol. Use an equal vol. of sec -butanol for each extraction and the volume will be halved each time . Centrifuge for 2 min in a clinical centrifuge to break the emulsion . Transfer th e 400 IJI kONA solution to a microfuge tube.
Continued
Continued 77
3 : RNA editing in mitochondria
Lorrj- Sim pson . Agdo M. Simpson an d Bea t Blum
Protocol 3 . Continued
Protocol 4. Continued
13. Extract the kDNA by vortexing with an equal vol. of pheno l:chloroform (1: 1). Separate the phases by centrifugation at 10 000 9 for 1 min . Excess centrifugation will result in loss of network DNA at the interface . 14. Remove the aqueous (upper) phase and remove traces of phenol from it by vortexing it 2 -4 times with an equal vol. of water-saturated ether. Allow the phases to separate by standing and then aspirate off the ether.
15 . Add 0 .01 vol. of 2 M NaCI and 2 vol. of ethanol. Incubate in dry icealcohol bath for 15 min or at - 20 °C overnight . 16. Recover the precipitated kDNA by centrifugation at 12 000 9 f or 14 m in at 5 ° C in a m icrocentrifuge .
17 . Resuspend the kONA peliet at 1 mg/ml in 1011 TE buffer. 18 . Check the int egrit y of the network DNA by diluting a sample ten -fold in
10/1 TE buffer and adding OAPlto 1 ~g /ml final concentration. Observe the stained kDNA using an UV microscope at 10 0 0 x magnification . The size and shape of the liberated kDNA networks are distinctive for each species of kinetoplastid . If the shearing (step 4) was too harsh, the networks will be fragmented . L. tarentolae networks often break into half- or quartersized networks. C. fasciculata networks are more st able to shear forces.
19. Store the kONA in aliquots at - 20 °C .
• 10 m g /ml et hidi um br om ide
• Silver nitrate solut ion {I 9 in 10 ml w ater)
. 50 % me t hanol. 10 % tr ich loroacet ic acid . 5 % acet ic aci d , 10% methan ol e Concentra ree NH 40H
• 0 .01 % cit ric acid contai ni ng 0 .4 5 ml of f ormaldehyde per 300 m l; mak e t his solu t io n fresh
• 20 % et hanol . 5% acetic aci d
Method 1. Digest 4 J..Ig aliquots of the kDNA with several different restr iction enzym es.
2 . Separ~te the restrict~d kDNA on the 4.5-10 % polya cr y lam ide gradient gel usmq TBE buffer In th e electrode reserv oirs as des cr ibed in Volu me I Chapter 1, Prot ocol 8 or V olume I, Chapter 4 , Protocol 11. ' 3 . Visualize th~ DNA bands by st aining the g el with either ethidium bromide (0 .5 J..Ig /ml final c~ncentration) or silver . For ethidium bromide sta ining see ref. 42 . For silver staining , see steps below",
4 . Fix t~e gel in 200ml of 50 % methanol , 10 % trichloroacetic acid for Sn rn!n.
5 . Wash ~he gel twice in 250 ml 5 % acetic acid . 10% ethanol for 20 min each n m e, then wash twice in 250 ml 10 % ethanol for 10 min each .
Using the procedure described in Protocol 3, the yield of kDNA is 0.5-1.0 mg/litre of culture. Figure / shows the kDNA isolated from L. tarentolae as in Protocol 3 . If de sired, nuclear DNA can be isolated from the crude DNA preparation (step 6). Add 2 vol. of cold (- 20 °C) ethanol and spool the nuclear DNA onto a glass rod. Dissolve the DNA in 10 ml of 10/1 TE buffer. Deproteinize it by phenol extraction and recover the DNA by ethanolprecipitation as described in Proto col 3, steps 13-16.
6. Mix 40 ml of 0.1 M NaOH. 3 ml of conce ntrated NH,OH , and 10 ml of silver nitrate solution. St ir th e mixture until the black pre cip itate diss olves.
Add water to 200 ml.
7 . Stain the gel in the silver solut ion (st ep 6 ) for 40 min . 8 . Wash the gel with 300 ml of water for 5 m in. 9 . Develop the gel with 300 ml of freshly made 0 .01 % cit ric acid contain ing 0.~5 ml formaldehyde . Stop the development in 20 % ethanol, 5% acetic acid for 10 min . 10 . Wash th e gel three t im es in 250 ml of 20 % ethanol f or 30 min each t im e.
3.3 Schizodeme-typi ng of kine toplastid protozoan strains Schizoderne-typing is used to verify the strain of protozoan used . It relies on the use of specific restriction enzymes to cut the kDNA mini circles and maxicircles. For L. tarentolae, HaelII , Mspi , Taql , and Rsal a re suitable enzymes which liberate most of th e min icircles and have multiple sites in some of the minicircle DNA sequence classes . The procedure is described in Protocol 4.
• Based on ref s 4 3 , 44 .
The complex restriction pa ttern obtained, which is mainl y due to the digested mllllclrcle DNA, is characteri stic o f a particular schizodeme. Maxicircle DNA ca n be seen as minor high molecular weight bands. Patterns of severa l T. cruzi schtzodemes are shown in refs 26, 27, and 45-47 . Figure 2 shows the schizod eme ana lysts of T. cruzi strains.
Protocol 4 . Schizodeme-typing of strains
3.4 Isolatio n of ma xi circle DNA
Equipment and reagents • kDNA prepared as described in Protocol 3
• Appropria te re strictio n en zymes and bu ffets (see su pplier 's ins t ructio ns ) • Poly ac rylamide gel (1 mm thi ck . 4 .5- 10 % linea r grad ient of acrvtamid e, 4 % st ack)
in TBE buffer (see Vo lume I, Chapter 1. Pro tocol 8 o r Vo lume I, Ch ap ter 4 . Protocol 11 1; use a stand ard grad ient f orm er and pump and add 15% gly cer ol t o the 10 % acrylam id e sol ut ion w he n f or m ing th e g rad ien t
Continued 78
In all kinetoplastid species analysed, maxicircle DNA , which ran ges in size from has a relatively high A + T content as compared to minicircle DNA . hIS property ca n be exploited to allow the separation of maxicircle DNA on B oechst 33258-CsCI den sity gradients (38). The Hoechst dye bind s preferentially :~ (A + T)-nch sequences, thereby ~ecreasing the buoyant den sity. The technique described III Protocol 5. The linearized maxicircle DNA prepared by th is method from L. tarentolae IS 30 kbp in size.
~ to 36 kbp,
79
Larry S impson . Agda ,',,1. Simpson . and Beat Bl um
3: RNA ed iting in mit och ondria
Pro to col 5 . Continued 2 . Monitor the extent of release of the 30 kbp max icircle DNA by running
10 ~I samples on a 0.7 % agarose gel (Volume I. Chapt er 1, Protocol 6 or Vo lume I, Chapter 4 , Protocol 12 ).
3. M ix the 2 ml digested kDNA with 1011 T E buffer to 12 ml final volume. Add 18.5 g CsCI and stir until dissolved. 4. Add 0 .5 mg/ml Hoechst 33258 dye drcpwise with mixing to approximately 1 I-Ig dye per I-Ig DNA . Stop the addition if the solution become s cloudy since this can result in the precipitat ion of the DNA . 5 . Adjust the density of the solution to a refractive index of 1.3950 at 25 °C by adding eithe r more CsCI or more 1011 TE buffer. 6 . Centrifuge the mixtu re at 4 0 000 r.p.m . in the Beckmann 50.2 or Ti60 rotor for 48 h at 40 000 r.p .m. at 25°C to establish the density gradient. 7. V isualize the blue-fluorescent DNA bands with 3000 A ° UV illumination and recover the minor upper band. T he lower band contains the undigested kDNA networks and released rninicir c le DNA .
8 . Re-adjust the refractive index of the upper band to 1_3950-1 .3935, and centrifuge 6 .5 ml per tube in the Beckmann 50 rotor at 39 000 r.p.m . for 48 hat 25 ° e . Again recover the upper band. This rotor provides a better separation and the second centrifugation completely removes any contaminating mlnlcircle or kDNA . 9 . Remove the Hoechst dye by extract ion w ith an equal vot. of propan -Zvol.
I
1 1
10. Dialyse the maxici rcle D NA against 10/1 TE buffer and concen trate the DNA using sec -butano l (see Protocol 3 , step 12l. t d kDNA stain ed w ith OA P' and
F~gUr~. 1'd Ki~~t~~~:t ::;~i~~c~~:n t:;hLu~r~~t~ i~:·t i:~ ~s:e Prot ocol 31. T~e.se cup -like
11 . Recover the DNA by ethanol precipitation (see Protoco/3, steps 15 and 16 ).
vrsu a rze a locular sheets a t cat enated mtniclrcles and ma xicirctes . structures represent manama
4. Isolation of th e kinetop last-mitocho n drion Prot ocol 5. Isolation of ma xicircle DNA
Equipment and reagents Chapter 1 . Proto col 6 or Vol ume I, Chapter 4 , Protocol 12 1
. kONA , 10/1 TE buffer. sse -butano l . 2 .0 M NaC \, and UV lamp 12 5 4 om ) as described in Protocol 3
eesc, (optic al g rade I
• Appropr iate restr.ictio~ enzy m.e s, a~d buffers (as descn bed In supplier S inst ructions) ; use an enzyme that cut s rnaxicircle D NA once or infreq uently. f eaR 1 is suit able tor L. tarento/ae kDN A
e O. 5 mg/ml Hoe chst dye L3 3 2 5 8 (Sigm a.
• 0.7 % agarose gel in TBEbuffer (Volume I.
• Propan -2 -o1
S 28831 • Beckman ultra cent rifuge. 50 and 50.2 or T i60 rot ors and tubes (or eq uivalent)
Method
Dilute 1 mg of kD NA into 2 ml of f eoR1 r.estriction enzyme b~ffer. Add 1. 1000 units of f coR 1 and incubate the mixture for 3 h at 37 C.
Continued 80
Kinetoplastids contain a single complex mitochondrion per cell. The port ion of th e mitochondrion that contains the kDNA nucleoid body is called the kinet oplast. Most kinetoplastid cells are very resistant to shea r for ces in isot onic media . However . in hypotonic media. most swell and consequent ly can be ruptured by shear forces. th ereb y releasin g the swollen kinetoplastmitochondrion (48) . Due to the multilobular nature of the single mitochondrion . isolati on of the ent ire organelle is probably not possible. When the swollen cell is ru ptured . the single mito chondrion breaks at the narrow portion adjacent to th e nucleu s. After resealing, however. th e mitochondrion portion cont ainin g the kDNA can be isolated by its relativel y high buo yant den sity (1.2 g/ml) (49) . Often the kinetoplast remains associat ed with th e basal body of the flagellum , indicating some type of attachment, which is also apparent in vivo. Th e kinetoplast portion shrinks to a cretenated disc upon addition o f 0.25 M sucrose and is resistant to exogenous DNase I. Digestion with DNase I to remo ve nuclear
DNA followed by den sity gradient cent rifugation are then used to recover the kinetoplast-mitochondrion fra ction as described in Protocol 6. Th e use of 81
Larry Simpson, Agda M. Simpson , and Beat Blum
3: RNA editing in mitochondrio
-' u
1=1
IS
I:'!
-' o
-' u
-' u
U
9
~
o-'
-' u
U
U
:c
£,'
e
-' o
-'
-' o
o
Protocol 6 . Continued
~
0
C\J
-' u
-' u
e STM buffer 18 5. 6 9 su crose, 100 ml of 0 .2 M Tr is-He l, pH 7 .9 .3 m l of 1. 0 M MgCl z• H20 to 1 litre) . 60 m l plastic syr inge fitted w ith a 26 gauge need le and a piston dr iven by compressed air at 100 p.s.i ., or a d ispensing pressure vessel (see Protocot S ) fitt ed w it h a 26 -g auge needle • Bec km an ultracen trifuge and SW28 t ubes, Sarva11 RC·5 centri fuge and 5534 or GS3 rot or s, plus appro priate tubes (or equivalen ts)
- 1.75 M suc rose _ SET bu ff er, CAP I, and UV m ic ro sco pe as in Protocol 3
. L. tarentolae cells har vested at lat e log- phase (50 - 1 50 x 106 celis /m Il fr om eithe r roller cult ures (1 lit re) or aerated stirred cultures (13 litre ] as descr ibed in
Protoco l 2 • DNase I (Sigm a, electrephoretically pure) . 2 mg /ml in 10 m m T ris-H e l. pH 7. 4 . 10 mM CaC l l • 50 % glycerol
For Renografin gradients . 7 6 % Aenografin (Squ ibb Pharmaceutical Co . : 66 % dtetrizcate meglumin e and 10% diatr izoate sod ium )
- 76 % RSE buffer (8. 56 9 sucr ose, 50 IJI of 0 .2 M EDTA , 76 % Renografin to
lOOml) - 20 % RSTE buffer {2 6 .3 ml of 76 % Renografi n, 8 .56 9 sucrose, 10 mt of 0 .2 M Tr is -Hel . pH 7.9 . 50"1of 0 .2 M EDT A , H 2 0 to 100 mil; ad just the de nsit y to 1. 14 g /m l - 35 % RSTE buffer (46.1 m l of 76 %
Renografin , 8 .56 g sucrose . 10 ml of 0 .2 M Tris - Hel, pH 7 .9 . 50 ~I of 0 .2 M EDTA . H 2 0 to 100 m l); adjust the density to 1.26 g /ml . 20 - 3 6 % RSE gradients. Layer 16 m l of 35 % ASTE buffer in a series of SW28 tubes; gently ov erlay with 16 ml of 20 % ASTE . Freeze the tubes at - 20 "C. Thaw the tubes overnight at 4 ° C bef ore using . The grad ient is c reat ed by the freezing and thawing . The advantage is that many tubes can be easily prepared and stored frozen until use
For Percall gradien ts Figure 2 . Schizodeme analysis . Acrylamide grad ient gel profiles of f eaR 1 digests of kD NA from sever al T. cruz; CL st rain cultures and clo nes. I, CL cells kept in cult ure for 2 ye ars (197 8- 80 ) and har ves ted in the stat ionar y ph ase: II. CL cells harvested in log phase ; III, CL cells kept at - 70 °c for 2 yea rs (1978 -80): IV . CL cells kept in mice for 2 years (19 7 8 - 80 ) and grown f or 10 passag es in cu lture; CL 10 . 11. 14 , 12, 16 . 18 , and 20 . are cl ones from t he p are ntal Cl str ain . Zymodeme gr oup s (B. C l are given in parentheses . Note the pr esenc e of t w o sch izodeme s in the st oc k Cl stra in . Repri nt ed fro m ref . 26 with permission .
Renografin gradients gives better results than sucrose or Percoll gradients, and flotation is better than sedimentation from the top. However mitochondria from Percoll gradients are more active in trans cription assays (50) (see Protocol II ) than are mitochondria from Renografin gradients . Protocol 6 . Is olat ion of kinetoplast -mitochondrion fraction
Equipment and reagents e ST E buffer (8 5.6 9 suc rose , 100ml of 0 .2 M Tris-Hel. pH 7 .9 , 10 ml of 0 .2 M EDTA , H;P to 1 lit re)
• OlE buffer; a 1: 100 dilut ion of 10/1 TE buffe r (see Prot ocol 3 1
- 9 3 % Percot l (Pharmacia] in 0 .25 M sucrose. 10mM Tris - He!. pH 7 .9 . 0 . 1 mM EDTA . 0 .01 mM z -merceoto ethanol ; si m il ar sol ut ions cont aini ng 5 2 . 5 % Perc 011 and 16% Percotl
Method 1. Wash the cells by resuspending them in 50- 100 vol. of ice-cold SET buffer and centrifuging at 4000 9 for 10 min at 5 ° C. Repeat th is washing. A lt ernat ively , add SET to the cell filt rat e when the Millipore Pellicon system (see Protoco l 21 is used. 2 . Thoroughly resuspend the cells in X vol. of cold DTE buffer and leave them on ice. X is calculated by dividing the total cell number by the fac to r, 1. 2 x 10 9 . 3 . Monitor the cells f or swelling by phase- contrast microscopy . At 10 0 0 x magnification the swollen kine to plast-mitochondrion is readi ly seen. It often ' pops' out of the cell and rema ins at tac hed by t he f lagellum. Be ca ref ul not to induce ruptu re by shea ring when app lying t he covers lip . If insufficient swelling has occurred , add mo re DTE buffer; 5 - 10 min on ice is generally sufficient for co mplete swelling .
Continued
Continued 82
. 16 - 52 . 5 % Percell gradients fe rmed as f or Reneg raf in gradients (see abov el using 16 % Percell and 52 .5 % Percel l solution s
83
Larry Simpson . Agda M . Simpson . and Beaf Blum
3: RNA e d it ing in mitochondria
Pro toco l 6 . Continued
Protocol 6 . Continued 4 . Break the cells by passage through a 26 -gauge needle . For 10-100 ml of c ell suspensi on , use a 60 ml syringe with air -driven piston . For , 00 ml-3litre v olu m es. use the disp ensing pressure v essel. When th e needle becomes clogged , close the shut -off val ve and repla ce the needle . Monitor the extent of breakage by phase -contrast microscopy at 400 x magni fication . There should be no more than ~ne intact ~ell every ~ -4 field s. If th ere are more intact cell s, they WIll c ont amin at e the final fr action , as v isualized by the presence of cvtosouc rRNA " .
5 . Add 0 .125 vol. of 1.75 M sucrose to the lysate directly after rupture . This causes the intact kinetoplast -mitochondrion to shrink back into a highly refractile c ret enat ed disc . 6 . Centrifuge the lysate at 5 ° C for 10min in the 5534 rot or at 11500 r.p.m . or in the G53 rotor for 15 min at 9000 r.p.m . 7 . Aspirate the supernatant carefully since the pellet is loosely pa cked .
Resuspend the pellet in STM buffer (50 ml for each 2litres of the lysate in step 4 ). 8 . Add 0 .005 vol. of 2 mg /ml DNase I and incubate for 1 h on ice to digest the nuclear DNA which will otherwise cause aggregation of the material in the density gradients (steps 13-151. 9 . Add an equal vol. of STE buffer to stop the reaction . 10. Centrifuge the solution in the SS34 rot or for 10 min at 11 500 r.p.rn . or in the GS3 rotor for 15 min at 9000 r.p.m. at 5 °C. 11 . Aspirate the supernatant; the pellet should be w ell packed if the DNase digestion was suc c essf ul. Now pro ceed to sep arate the kinetoplast-mit ochondrion fra ct ion using either Renografin or Percell gradients.
check the puri t y by phase mi cro scopy at 1000 x magnification . Vis ualize the kDNA by st aini ng with 1 1-11 DAPI and obse rving by pha se-c on t rast and UV microscopy . 7 . W ash th e pell et by resuspend ing it in 50 vo l. of SrE buffer and cent ri -
fuging as in step SA. Finally wash the pellet with STM buffer prior to kRNA isolat ion. 8 . Resuspend th e mit oc hondrial pe llet in an appropriate medium (see Protoco ls 7, 9 , 11. lS I. U se it im me di at ely or st ore it froz en at - 70 °C in small aliquot s.
B. For Percoll gradients 'l . Resuspend the pellet in 93 % Percoli luse 4 ml of Percoll for th e pellet from a 1 litre c ult u re ). 2 . Layer the mixture under 16-52.5% Perc 011 gradient (use 2 grad ients for material from 1 lit re culturel. Centrifuge f or 45 min at 24 000 r.p .m. in a Beckman SW28 rotor at 5 °C . 3 . Locate and remove the kinetopl ast band (see st ep 4A l. 4 . Dilute the kinetoplast band w it h 200 ml of 5re buffer. Recov er kinet opla st m itochondrion fra ction by c ent rif ugat ion at 12 000 g for 30 min at 5 ° C . Wash the preparation on ce in sre buffer and resuspend it in sre bu ffer. 5. Store the fraction as for Renografin gradients (step SA). • A quantitat ive estimate of the contaminatio n can be obtained by measuring the relative amount of cvtosouc rRNA in the final preparation of kRNA tsee Prot ocol 7 ). The cell lysate can be used to prepare cytosolic rRNA substrate for Protocol 12 . b The RSE gradient procedure is the rate-Hrrutinq step in larpe-scale preparat ions. The lysate from up to tw o Htres of culture can be loaded on each Renografin gradient and six gradients can be run in the SW 28 rot or.
A. For Renografin gradients 'l . Add cold 76 % RSE buffer to the pellets (4 ml of 76 % RSE per litre of original culturel. Vortex well .
2. Layer 4 -5 ml of the mixture beneath each 20-35 % RSE gradient ' us ing a 12 ml syr inge w ith polyethylene tubing attached t o an la-gauge needle. If the first drop of lysate floats . add more 76 % RSE buffer to the lysate until the lysate remains at bottom of gradient .
3 . Centrifuge the gradients for 2 hat 24000 r.p.m. at 5 °C in a Beckma n SW28 rotor. 4. Visualize the k inet oplast -mi t ochondrio n band just abo ve the 76% -35 % RSE interface by side illumination with a UV lamp . Puncture the side of the tube and remove the band using a syringe and needle. 5. Dilute the mixture with 2 vol. of 5rE buffer and centrifuge the preparat ion in an 5S34 rotor f or 15 min at 11 500 r.p .m. at 5 °C. 6 . Check the purity of the preparation at this point . jf d esired ; in the next step, the us e of STM buffer will cau se agg regation of the mitochondria. To c heck the purity , resuspend a sam ple of th e pell et in STE buffer and
Continued 84
Figure 3 illustrates the use of Renografin gradients to prepare the kinetoplastmitochondrio n fraction from L. tarentolae.
5. Isolati on of transcripts from the mitochondrial genome 5.1 Introduction and strategy Isolation of intact RNA (kRNA) from the purified kinetoplast-mitochondrion fraction (prepared as in Protocol 6) is easil y accomplished due to the lack o f nuclease activity in this fraction (49). The major steady-state components of the kRNA are the 98 and 12S mitochondrial rRNAs and tRNAs (51). All ~he mitochondrialtRNAs appear to be nuclear-encoded and must be transported Into the mitochondrion by a mechanism as yet undetermined (52, 53). Unedited transcripts of maxicircle stru ctural gene s can be detected by Northern blot 85
Larry Simps on, Agdo M. Simpson , and Beat Blu m
3: RNA editing in mitochondria 16
TOP
A
\2
~
:£ 2
4
8
2
4
u
Transcr ipt ion
'"
e Zp' GTP)
a-
u
2
'"
a-
B.
2
4
'a-"
2
u
(54, 55). The ratio of pre-edited tran script s to mature edited mRNA varies from gene to gene (56). Partially-edited mRNA s are usually less abundant than edited or pre-edited RNAs and can only by detected by selective PCR amplification in L . tarentolae , Th e gRNA transcripts which are involved in RNA editing co-migrate with tRNA in agarose gels and migrate ahead of t RNA in polyacrylamide gels (52). Th e 3 ' oligo(U)-taiis of gRNAs vary in length from 5- 28 nt and are responsibl e for th e chracteristic fam ily of electrophoretic bands each differing by a single nucleotide (57). The abundance of gRNAs is relatively low compared to mitochondrial tRNAs and so the y can not be visualized in gels stained with ethidium brom ide unless 4-5 ug of kRNA is used per lane .
5.2 Isolation of kinetoplast RNA
RNA Ligose + KRNA)
C~zpCP
Protocol 7 describes the isolation of kRNA from the kinetop last-mitochondrion fraction prepared by Renogra fin grad ient centrifugatio n (see Protocol 6). Protocol 7 . Iso lat ion of kinetoplast RNA C
90
:£
~ o·
02
0
01
.,
~
"5. 0 60 :;:
Adenylole Kina se
0
~
.= 0
30
5
15
10
PG
£
Frac t ion (1 m1)
e T M N buffe r (10m M Tr is-H Cr. pH 7 .5 . 10 mM MgCI 2 • 5 mM NaCll . 2 mg/ml D Nase I (RNase-tree; BRL .
25
PG
fugation (see Prot ocol 6 1 and with the mitoc hond ria from 1-2 lit res of original cell culture resuspended in 5 -1 0 ml of
TMN buffer ; may be froze n at
,80475AI
~
70 °C
until use
e Phe nol.chlorotorm ( 1: 1, v/vl • 10 mg/ml et hidium bromide
20%
76°/. 35%
PG
20
.';
Equ ip m ent and reagents
• Kinetopla st -mitochondrion fract ion : prepared by Aenograf in grad ient c entri-
. 1% agarose gel in TBE buffer (V olume I, Chapter 1, Prot ocol 6 or Vol ume I, Cha pter 4 , Prot oco l 12 1
. 10 % SDS
M ethod Figure 3 . Renograf in isopvcnic density grad ient fract ionation of the kinetop last-m itochond rion fract ion from L. tare nrolae . Grad ients were fract ionated and fract ions washed with STE buffer (Pro tocol 6 1. A liquots w ere assayed for (A) TUTase (Protocol 72) and run-on transcription (Prot ocol 9 ) act ivities; tSI RNA ligase activity (Prococol 131: Ie) adeny late kinase act ivity and protein . Aden ylate kinase is a marke r enzyme for the m itochond rial inner membrane space. RG. Renog rafin . Reprinted from ref. 50 with perm ission.
hybridizaton using specific oligonucleotide prob es. Edited transcripts can be detected using genomic DNA probes if the editing is not extensive. However, mature pan-e dited mRNAs , such as the L. tarentolae RPSI 2 RNA (54), can be detected only by know ledge of the edited sequence. In this case, prim ary transcripts containing pre-edited regions can be monitored using genom ic DNA probes . Furthermore, one can take advantage of the 3' - 5 ' polarity of editin g to detect partially-edited mRNAs which still contain genom ic 5 ' sequences, by selective PC R amplification using an oligo(dT) 3 ' primer for the first strand synthe sis and a genomic 5 ' primer for the subsequent PCR 86
1. Quickly thaw the kinetcptast-mitochcndrion fraction (if necessary). Add 0 .01 vol. of 10 % 50S and incubate the mixture on ice for 5 min. 2 . Extract the mixture by vortexing w ith an equal vol. of phenol:chloroform (1: 1) as described in Protoc ol 3 , step 13 . Remove the aqueous (upper) phase to a fresh tube . 3 . Be -ext ract the interface with one vol. of water and pool this with the upper phase. 4 . Ethanol-precipitate the nucleic acids, (see Protocol 3 . steps 15 and 16 ). 5 . Resuspend the nucleic acids in TMN buffe r using 1 ml TMN per 2 -3litres of original cell culture. 6 . Add 0.005 v ol. of 2 mg /ml DNas e I and digest the mixture for 30 m in
at 37 °c.
7 . Extract the solution w ith phenol:chloroform and precipitate the kRNA wit h ethanol as described in steps 2 -4.
Continued 87
3: RNA editing in mitochondria
La rry Si m pson , A gda M. Sim pson , and Beat Blum
Protocol 7 . Continued
Protocol 8 . Continued
8 . Wash the pellet in 70 % ethanol. Resuspend it in water (1 ml per 4 1itres of original c ult ure yields approximatel y 2 ~g RNA /~I) . 9. Monitor the purity of the RNA by electrophoresing 2-5 1Jg in a 1.5 % agarose gel in TB E buffer, and staining the gel with ethidium bromide (O.51l9 /mll. Any contaminating cytosolic rRNA will be appa rent as three bands migrating abo ve the mitochondrial 125 rRNA band (the cy toso lic large rRNA has a break and gives rise to two bands in gel electrophoresis). Also visible shouid be the 9S rRNA and the tRNA bands 151 1.
• rm m unoPure Ag /A b immobilizat ion kit (Pierce 448 90) toget her w it h im mobilizat io n buffe r as provided by the m anu facturer • Centr ic on , 0 centrifugal centrat er (Arnicon 4205 l
m icr oc on -
. 50 m l ce ntrifuge t ubes w ith plug seal cap s (Corn ing 2 5 3 3 1-5 0 ) • Rot at in g pl atfo rm sti rrer
10. St or e the kRNA in aliquots at - 20 vC .
NaOHI
Northern blot anal ysis is described in Volume I, Chapter 2, Section 3.6 and Volume I, Chapter 4, Section 7.5 . Electrophorese the kRNA (see Protocol 7) in a formaldehyde-agarose gel and blot the gel onto a nylon filter. H ybrid ize the filter with an appropriate radi olabelled probe complementary to the kRNA sequences under investigat ion . For internal size standards, strip and reprobe the kRNA blot with a nicktrans lated pLtl20 probe, which represents a 6.6 kbp fragment of the maxicircle DNA containing the 9S and 12S rRNA genes, and the ND7, CO llI, and Cyt 6 genes (58). The resulting bands in the autoradiograph shou ld be 320 nt, 610 nt , 1100 nt , and 1200 nt in size .
. 0 . 1 x SSC buffer . 0. 1 % SOS (prepa red u sing 20 x SSC buffe r stock and 10 %
505 1 • Hybr idization bu ffer (1 .0 M NaCI. 0.1 66 M Hepes-K OH, pH 7. 5, 1 m M EDTA . 0. 1%
50 S) . W ash bu ffer ( 10 mM Tr is -H CI. pH 8 .0,
. 20 x SSC bu ff er stock (3. 0 M NaCI , 0 .3 M trisodium citrate, pH 7.0 w it h
5.3 Northern blot analysis of kRNA
. 6 x SSC bu ffe r and 1 x SSC bu ffe r (prepared us ing 20 x SSC bu ffer st ock }
. 5 .0 M t et ram et hyl am mo nium chlo ride stock solu tion (Fisc her /I 04 6 40- 500) ; adju st th e density t o a refra ct ive index of 1.4 21 at 25 ° C . 1.0 M Tr is - HCI, adj u st t he pH t o 8. 0 us ing HCI
e O.5 M EDTA, pH 8.0 • TMA buff er 13.0 M t etramethylammonium chloride, 50 mM Tr is-HCI , pH 8.0. 2 m M
ED TA, 1% 50 51
1 mM EOTAI . 2 ml ag arose beads together wit h t he appr opr iat e chro m at ography co lumn (as pro vided b y the Immun oPur e Ag JA b im mobilizatio n kit , Pierce 44890 ) • Carri er nu clei c acid (5 '- am ino oliq onu cle otide, not c omplementary t o th e gRNA t o be hybrid sele cted) ; 5 ' -amino link ed oligonu cleot ide s are con veni ent c arriers becaus e th ey do not interfere w it h t h e 5 ' labell ing reac t ion w hic h follows the hy br id select ion step; circul ar DNA molecules or gl ycogen ma y work as well
Me thod 1 . ~ttach 100 J..lg of the aminolink 2 ~0ligonucleotide to 2 ml of agarose beads In 2 ml of the recommended immobilization buffer .
5.4 Hybrid selection of guide RNAs A hybrid selection procedure can be emp loyed (57) to isolate specific gRNAs. The method makes use of synt hetic oligonucleotides which contain a reactive amino group at th e 5 ' -end. The amino group is covalently reacted with the hydro xyl groups of agarose in the presence of a strong reducing agent. This step imm obil izes th e olig onucleotides to the agarose which can then be used for a ffinity purificati on of specific RNA species by column or batch proc edures. Proto col 8 is a modification of the filter hybridization method of Wood et al. (59). A stringent wash removes non- specifically bound RNA s before the selected RNA s are eluted by using a low-salt bu ffer at an elevated temperature.
2 . At the end of the required incubation period pack th e agarose into the column and wash it twice w ith 5 ml of wash buffer and store the washed col um n at 4 " C until required . 3. For hybrid sele cti on of gRNA , equilibrate the column in hyb ridizati on butler . 4 . Load appr oximatel y 500 ~g of kRNA in 1 ml h ybridization buffer onto the co lum n. 5 . Seal th e colum n w it h th e cap and sto ppe r pro vided ; incubate it f or 18 h
at 37 -c,
6 . Drain th e col umn and add 5 ml of ic e-c old 6 x
Pro tocol 8 . Hybrid selection of guide RNA s
Equipment and reagents • Sy nt het ic DNA oli gon ucleotide: 10 0 JJg; 20 -30 nt complementary t o t h e gR NA
to be is ola te d; at t he last s tep of the synt hesis, coup le the reagent , Aminolink 2 (Applied Bio svstems! t o t he 5 ' -en d, f ol lo w ing t h e supplier's ins t ructions
• kRNA from 5 lit res of a log-phase Leishmania culture at a cell density of 0.7 - 1. 5 x 10 8 cel ls/ml fo r t he hy brid selectio n of each gRNA (see Pro tocol 7 )
.1 0 % 50S
a.
buf fer .
Drain the column and equilibrate it w ith 5 rnl of TMA at room temperat ure.
9 . Rep eat step 8.and pla ce the sealed co lumn into a 50 ml ce nt rifu ge tu be. Seal the c ent rifuge tube and pla ce it into a 3 7 " C w at er bath f or 30 m in.
10. Drain th e col u m ~ , th en add 5 ml TMA buffer (preheat ed at 60 "C]. Again seal the column Into a 50 m l centrifu ge t ube and plac e into a 60 " C water bath f or 30 min ".
Continued 88
sse
7. Re.pea.t ste p 6 t w ice , th en incub at e the column at 4 vC f or 2 h w ith gentle aqitatron by placing the c? lumn into a 50 ml cent rifu ge tube and rot ating It hor izont ally on a rotating pl atform st irre r w hic h is set t o 50 r.p .rn .
Continued 89
3: RNA edit ing in mitocho nd ria
La rry Simpson . Agda 1\ 1. Simpson. a n d Beat Bl um
5.6 Synthesis of gRNAs by transcription in vitro
Protocol 8 . Continued 11 . Wash the column successively with 10 ml of TMA buffer ~ t room t em per at ure, 10 ml of ice-cold 6 x buffer, and 10 ml of Ice-cold
sse
1 x SSC buffer. 12. Elute the gRNA at 70 °C tor 20 minusing 2 ml of 0 .' x SSC buffer. O·'I%d 5 0S . preheated t o 70 ° C. This is done by placing the column Into a sea e 50 ml centrifuge tube and immersing it into a 70 ° C wate r bath " . Collect t he colum n eluate . Repeat this twice with fresh buffer. 13. A dd 2 1-19 of carrier nucleic acid . Concentrate t he ~ooled eluat e with t ~e Cent rico n 10 cent rif ugal ml croconcentrator followmg t he manuf act ur er 5 instructions.
14 . Store the hybrid-selec ted gRNA at -20 °C. • Note that the exact wash ing temperature must be varied according to the ~;2~ of the hybrid 15 91. Highly st ringent washing oc curs at Tm - 4 °C . The Tm of a hybrid In TMA bu ffer is solely dependent on its length and can be calc ul at ed from ref. 59 . b Do not heat t he co lumns co ntaining the agarose beads t o mo re than 70 °C . The aqarose co uld melt above this temperature.
5.5 peR amplification of partially-edited mRNAs an d gRNA :mRNA chimaeric molecules Pa rtially-edited RNA s ca n be amplified by the use of an oligo(dT) 3' pri mer complementary to th e oligo(U)-taii of the gRNA. for the reverse transcnptase step (60). This is followed by the addition of a 5' primer consisting of a gen omic seq uence com plementary to the unedited 5' sequence of the mRNA and Taq po lyme rase fo r the P CR amplificat ion cycles. In a modified p rocedur e: it is po ssible to select for pa rt ially-edited molecules covering a limited regIO n of the m RNA by choosi ng th e a ppro priate set of primers (55. 61). In this case. the 3 ' pri mer should cover the edited sequence while the 5 ' pnmer sho uld cover the up stream pre-edited sequence . PCR amplificatio n selects for RNA molecules part ially edite d in the sequence between the two primers . This approac h allows for fine an alysis of partially-edited molecules at single sites. In vivo chi maeric gRNA:m RNA molecules have been detected In kR NA by Northern blot hybridization and by PCR amplification using a 3 ' primer comple mentary to mR NA downstr eam of a PER (16). The po~ymerizat i ~~ of eD NA using reverse transcriptase is followed by the addition of a ) primer with the 5' ter mina l sequence from the corresponding gRNA . PC R am plification reveals the presence of chimaeric molecules. The existence of such chimaeras is ind icative for the trans-esterification model. However, the y co uld . also arise as by-p roducts of a cleavage -ligation process. Detailed pro cedures for P CR will be found in Volume I, Chapter 2, Secu on 3.4 a nd Volume I, Cha pter 3. Section 2.2.3 and in ref. 62. P CR pr oducts both from pa rtially-edit ed and chimaeric molecules may be checked by gel electrophoresis and their nucleotide sequences determined. Figure 4 showS partially-edited sequences in the maxi circle Gerich region 6 of L. tarentolae90
Using bacter iop hage pro moters and purified ph age RNA polymera se. gRNAs ma y be transcribed in vitro . For inst ance, a T7 phage prom oter seq uence can be incor po rated int o the 5 ' P C R primer for gRNA amplificat ion (see Section 5.5). The amplified DNA is th en transcribed using T7 RNA polymer ase. The transcription prod ucts will have heterogeno us 3' termini du e both to the prop ensity of the pol ymer ase to ad d an extra non-encod ed A or C and to premature terminat ion (63), and so the desired species mu st be purified by gel electro phoresis. Such tra nscripts ha ve been used to study the fo rmation in vitro of gRNA :mR NA chimaeric molecules in a mito cho ndrial extract (18-20). Proced ures for the tra nscription of RNA in vitro using ph age pr om oters and polymerase> will be fo un d in Volum e I, Ch apte r I (Pr ot ocol 4 ) an d Chapter 2, thi s volume (proto col 4 ).
6. Run-on transcription in isolated kinetoplast mitochondria Kineto plast mito chond ria isolated by th e proced ure desc ribed in Pro tocol 6 are active in fun-on tr an script ion whereby RNA chains init iat ed in vivo ar e completed in vitro. As explai ned in Section 4 , mitochondr ia prepared in Percoll gradient s a re generally more active in run -on tr an scription than those prep ared using Renogra fin gradie nts . Neverth eless, mitoc ho nd ria prep ar ed by th e latt er pr ocedures are adequate for studying ru n-on tran scrip tion. Run-on transcription ma y also be stu died in mito chondrial extr acts, pr epared as described in Sectio n 8. Proto col 9 describes ru n-on transcription using intact mitoc hondria . Using this method , the rate of incorporatio n of [",_J2 p I GT P increases for 10- 15 min and then decr eases. Protocol 9 . Run-on t ra nsc ript ion in iso lat e d mitochondria
Reagents
6'
• [0' . 32P1GTP (New England Nuclear; 3000C i/mmol.l0mCi /m l ' • M itochond rial fraction isolat ed by Reno grafin or Percoll gradients as described . 30 mM potassium phosphate , pH 6.8 in Protocot B, resuspende d at 2-3 IJg/1J 1 in STE buffer for immediate use or e 2 M KCI stored frozen in STE with 8 % DMSO at e O.2 5 M Hepes-KOH. pH 7 .6 - 70 °C ; run-on transcription activity is stable for several weeks in frozen . 0 . 2 5 M z -rnercaotoerbaoor m it ochondria . D EB l filter discs (W hat m an, 2 .5 em e lO mM AT P diameter)
. 5 x STE buffer {see Protocol
e l 0 mM CTP
. 0 .5 M sodium phosphate buffer, pH 6 .8 , 0 .5 % sod ium pyrophosphat e
. 10 m M UTP
Continued 91
A Ii G
DNA:
A
RNo\:
AUI.AAlUIJIJGuGlJlJI.A.UJG LUG
Si t e:
26
25
TlG TTTGTTTTTG
24
23
G
22
Ii eHG
GAl. e n A
UGuu.JGU\..IUUlAJGAAuC
21
20
AuuuG C
19 16 17
A A
UG~uA
16 15
Ii
GAATTTTTTGG eG cc
GuuGAA
13
14
A
GA
Cet ACGAG AAAG A
TTTCACG
GAATTGTTTTCGT
GGuGCuGGuuuuAuuGAuuuGCC AGGAGuMAGuAw..uJCACGUI.JI.JlAAAIUUGAA UG
12
11 10
9
8
7
6
5
4
3
CGU
2
Clone: 1
A
Ii Ii
TTGTTTGTTTTT G
c
GAl. en A
Ii c r r c
A A
Ii
GMTT TTTTGG cc cc
A
iiI.
GCC AGGAG MAG A
lTTCACG
2
A
G
c
TTGTTTGTTl TTG
Ii
GAA en A
c ctrc
A A
c
GM TTTTT TGG c c c c
A
ii I.
GCCAGGAG MAG A
TT TCACG
GAA_C$.':~WCGt
3
A
Ii G
TTGTTTGT TTTTG
G
GAl.
cr r a
e er-e
A It
Ii
GAATTTTTTGG
cc ec
A
GA
CetA CGAC MAG A
TTTCACG
tt~~thGM aG]:;l ntGt
4
A
G
c
TTGTTTGTTTTTG
Ii
GAA eTTA
Ii er r s
A A
G
GAATTTTTTGG li li lili
A
GA
CetACGAG MAG A
TTT CA CGt:tthnhllM ::lG :::~,t~!~:CGt
5"
A
Ii Ii
TTGTT TGTTTTTC
Ii
GAA eTTA
G cr r c
A A
Ii
GAATTTTTTGG GG cc
A
GA
GCCAGCA GctAAA~t.AJPn~AC~,~ .Ht.ft.hj,c;M ;) (fftC~t
GAATTCTTTTCGT
A Ge TTGTTTGTTTTTG G GAl. eTTA e crrc A A c GAA TT TT TTGG~.~.c,~ ~~.H,[email protected]:eCM~l::A C;,tMA~ t.A.n~ nCAt G t{t.ft.(fft GA'(ic':~'?t ct A GG TTGTTTG_TTTTG G GAA c A Gte TG tA Att G G.u:+:\~tL~!G (GGt'GG.'tt{h'; tJA.it:lt(;~ t.~ C"(;A G t,W~'t,~.(~n J.CAC Gf.f(tI(f.(t GM::: T C}~'~::;::;· c c; r: A CG TTGTTTCTT TT1 G G GM CTTA G CTTG A A cu CAA ~t:httcc t GC t CC t u t "' n tiAfHtctAGCAt twq~Ah r.u eAC c t .t.tt tt ft it.M ,J (j, ;*-"$, CG ~ A GC TT CTTlCTT TTT G C CAA CTlA G~ C A A cti:w :r~:~:':::\?~l:: t.(;G t GGU ttAHGAt't'tGctAt:G.Acd.AA'GtA ttlTT tACct"ftt td:'t'tCAA trG '%WC' t A CG TTCTTlC_TT TT G C GAA C A tCte TG tA A t G $IGM?tttt:~~,~~.qf~~ t.n,t~lt.~MiiG.Ct~.~'G~G1:.AM'G't,A.ft}n ¢At,Gft t t.t t.tft.GAA ::rG d~;rcqt A GC TTGTT TCT TTTTG C CAA C A C c_~J~ At~jM t, t w::#~:~~~;~;:;r /;Ct G c t'CC; ttftAn,~~it GCCACCAc t AAA G t At't ~ TTC ACG t ft t t t' t" ~ t CiAA :: T C "'~::'cGt, A G G_ _ rcsrrCt:;:::: Tct' t tc;~It·s'.t·t c.AAtc;~::: Act.tc :~,:c trGi~ "~·tA,$.G w ·( t t t ?l::ciCGtCic t't ~ ~AH4ftJ ~'CCCAGGA.e tMA.GtAtt T1TcAeett t ~ t.'tt tt GAA :1.cg ::? C1t A CC TTG= :=,itG} :==P 'ttiI:i~G t1:I:'f t't,G~ t P =::,A.t t.t~:=:'c\JGt.t iM. A }~::~ i t.eM} :~:?:/:~::~"GtGC, t GO t'i ft A.tJ~Aftt CC ~AG.eA~ tM.A.G fA.~t, TJrCACG,tt \'~- t'ti t i:'G"A/ T,G:W::ntqt ,AJtft t ftc t Gt t t ttl G" ~;.T/G.:ii~iJc.t tGt t t .~i(~'AA ci; At ttG ,;c:::::ttt i:tAtA;%:,Ct t CMi:) ii~~: ~': ' [;G t cct'GGt t A"ftC-A t t'i:tcCAGGAGtillGtitj t}TCACG~' t ft, '~,i.t.~ ~GAA:iJ ~:~fi.ii~~j; '~.tt t.n t t GtGt t t t uc:=;tt~::~mfr GU'~'l::Hn.fi: GAA.:£.I. At t tG: 1:,) ~ tttA.tA. .',' CttGAA,:= ? :f :tt ,:GCteG tC~ t t t t At t GA t t t GCCACCACt W Gt At t TtTcACc tt t t i ft t t cA," ') C :."eel
10 11 12 13
14 15 16 ,7 18- 21
n,
t
t
tt
22 23 24
A A
CG TlCTTTGll TTT G G GAA CTTA G eTTG A A G CAATTT TT TGC GC GC A CA CeCACGAG AMG A TT TCACG CAATTCTTTT eCT t t Hata t t tTfG?'ttG::&KTCi't tGt t'f t,tJ'c",AtC'::;:'AHiG:c.aGtftAtA ?' Ctt GM:t ~:;::::::::' GC t GG t GC t t itAt t.GA t tt ccCAGGACrWGtAt tfT TeACCtt t t t t tt'tGAA"'TG cut ~A; hH t t ~G t Gt t t rtr ~::,t t,G::~:LtT G t, ~ t Gft,ft.ti:(;A~. t,C:i= :=At,I:,t C::.C:J ,~.tftA t A. '~< G (tGM}~t)~ '{GCt CGt GG t t tt A'tttiA tI t t CCACGACt A A"G ~ A t tf TT tAcet t t t t.t n tGM:,:TG:,:=:;:: CcI
25
ttA G C tt tTIC TTC TTGttt C t ttt( .t~MtC ::~;' At tt G' :C;=J~ft,~ A.i:A7(ji: iGAA. ::~: ~~H{GG t G ct C C t fH A tt: GA t ft G CCA GCA c tW G l: A i t T TT CAC G t t t t tt't ft GM. T G ACA t t t t t Ct t C TTCTTTcTTTTTe c GAA e TA tC q.rG.fo..A. t AJt:~'t t C.A ,( :I}t\G G t c ~ t c G ~ H.~ A tJ ~A U t G c,~A e CA G t W G ~A t t T TT CA C C t t t t t t t t t CAA 1(0
26
cct COT
Identi ca l cl one s : - - 6 , 7, 8, 9 .. . 19, '20
B OIllA :
G
RillA:
G
S i r e:
55
C I~ :
TC A GTTTC CTT TTC lTATTTTA C GC TA A AAG CCACGA ACCTACTT eCCc AATCCACG C A ATTTCCAA A G A A C loA A AC CuuuAuC UC uuC CUUIJl.(JUA UAUUUUuGuuuCCUUIJJAuAuuuAACueCACGA uuACCUAGUUeeGCuAAUCCACCuuCuuuAuA UCCM uAuGuuAuAuCuuuAAuAuuuuAuG 54' · ·52 51 50 49 48 47 46 45 104 43 42 41 10 0 393837 J6 35••. • •• •31 30 29 26 27
17 18
A CTTTC CTTT TC A CTTTC CTT TTe A C TTCttttC C C C - TG .:.-lY C C
I. ZO Z1
zz
~ A C-IQ
Z3 Z4 25 26
A C G G
TO TO
1&
A CTTTC A
crrrc
TTATTTTA TTATTTTA TlATTTTA ~
TA
,
.
l
.!i
M
l
G
M TA
c eTTTTc CTTTTC
,
A ATTTGeAA A G A A G A ATT TGCAA A C A A C
AA A M A
AC AC
iotA , ':::JGtAA. (~A t,(jHA.~At..G,fhM.(ln{ t,~.t';
TTA TTTTA TTATT TTA
G
GG
T' ,
G
GG
TA'
Aloe eCACCA AAC CCACGA
ACCTAGTTCCCC AATeCAce C A A TCCAA A e t A AtttG ttt tAtAtAtAC acct ACT TeCGG M TCGt t tt t tACGtt t t t t ct t CAA t CCAtAACAt tttAt GtA t AAtt t
Figure 4 . Partially edite d m axi eir cl e G ·ric h reg ion 6 ( = RPS 12 ) eD NA sequences fr om L. tarentolae . kRN A was amplified u sin g a 3 ' oligoCT) prime r and a 5 ' geno m ic prime r. Th e peR pro duc ts were clo ned and t he clones sequenc ed . Seq ue nces are alig ned to show t he 3 ' -. 5 ' prog ression of exp ec te d edi ti n g eve n t s. Un ex pect ed ed iti ng events are un de rlined . Pattern 5 w as seen in 5 clon es. p att ern 18 was seen in 3 clo nes . Th e unedit ed domain -connecti on seque nces are shaded . (A) 3 ' t erm in al port ion s o f th e cDNA sequen ces . (Sl 5 ' te rm in al po rti on s o f th e seque nces . Clones 1-1 6 are not sho w n si n ce t heir sequ enc es, lik e c lone 17. are un ed it ed in this regi on . Reprint ed from re f. 54 wit h permission .
3: RNA edit ing in mitoch ondria
A modified local homology alignment program (BESTFIT, Uni versit y of Wisconsin Genetics Computer Group package, version 6) is used in the search for gRNA sequences. The scoring mat rix is altered so as to score positivel y for complementary base pairs (including G -V) rather than for matches. Moreover, it is possib le to appl y different weight s for each possible base pair.
Proto col 9 . Continued
Method 1 . In a microcentri fuge tube on ice mix:
e 30 mM potassium phos phate. pH 6.8 e 2 M KC I e O.25 M Hepes -KOH, pH 7.6
5 ~I 1 . 5~ 1
1 ~I
Protoc a l 10 . Identification of gRNA sequences by co mputer analysis
1 ~I
e O.2 5 M 2-mer capto ethan ol
A. Preparation of the files
2.5 ~I 1 ~I
e 5 x STE buffer e [".32p) GTP e lO mM ATP e lO mM CTP e W mM UTP
1. Dow nload the kDNA genomic sequences from the Genbank databa se. 2. Reformat the files to GCG format.
5 ~1
3 . Create a GCG sequence file w ith the determined edited mRNA sequence
5 ~I
Isee SEQED, UWGCG) .
5 ~1
10
~I
to 50
~I
e mltoch ondria l fraction
e H20
Larry Simpson , Agda M. Simpson, and Beat Blum
4. Reverse the mRNA sequen ce , using REVERSE (reverse only , UWGCG I.
final volume
2 . Incubate the mixture for 1 5 min at 27 ° C . 3 . Remove 10 -20 IJI aliquot s and spot the m onto DEBl filter discs. 4 . Dry the discs and then w ash them for 30 min in 100 ml of 0. 5 M potassium
phosph ate buffer, pH 6 .8, 0. 5% sod ium pyrophosphate. Repeat the washing three times.
B. Modification of the scoring matrix
5. Edit the scor ing matrix SWGAPDNA .CMP IFETCH this fileI using an appropriate editor to score for base pairing inst ead of matches . The follow ing table SWGAPDNA .CMP is an example of we ighted scores :
A
- o.s
5. Wash the discs wit h 100 ml of ethanol as in step 4.
C 0.01 - 0 .9
6 . Dry the discs and count th e ret ained radioactivity in a scintillati on counte r.
The run-on tra nscripts, synt hesized by the method given in Protocol 9. migrate in aga rose gels as a smear, suggesting that they represent nascent RNAs elongated in vitro dur ing the labelling period (50). The labelled RNAs hybr idize to all regions of the maxicircle, even to the divergent region which show s very low levels of steady-state transcripts, an d to minicircle DN A.
7. Identification of gRNAs by co mputer-assisted seque nce comp arison In the editing of mRNAs, gRNA molecules specify the editing by fanning duplexes with the edited mRNA (15). Such duplexes display G-U base pairs in addit ion to the canonical G-C and A-U base pairi ng. A search for gRNAs can be carr ied out with the help of a computer (see also ref. 64). The edited mRNA sequences are compared with the kinetoplast genome (consisting of the maxicircle an d the known minicircle DNA sequences) taking into account such non-canonical base pairing . P rior to the computer search, the sequence of all or part of the edited mRNA must be obtained by direct RNA sequen cing. Reference 65 gives details of RNA sequencing methods. Protocol 10 descr ibes the computer analysis. 94
G
- n. s 1.0 - 0. 9
T 0 .5
- o.s - 0. 9 - 0 .9
U 0.5 - 0 .9 0.25 -0.9 - 0 .9
A C G T U
Note: T his table does not ' punish' for C -A base pairing; C-A pairs have been found in putati ve gRNA :mRNA duplexes from C . fasciculata (66 ).
C. Running the program
6. Type: 8ESTFIT /DA TA1 = SWGAPDNA.CMP/PAI= 0.01 . 7. Select a high value for gap weight 1"100" ) and gap weight length I" 2.00") to avoid any alignments with gaps . 8. In order to find the gRNA sequences corresponding to an edit ed mRNA sequence , it may be necessary to vary the size and borders of the input mRNA sequence (e.g , at the sites of two overlapping gRNAs) . Known
gRNAs help to define the borders for the search of additional gRNAs. The smaller the size of the gRNA:m RNA duplex, the more difficult it becomes to find the corresponding gRNA (64) . In the case of L. tarento lae , each minicircle encodes a single gRNA located 150 nt from the end of the conserved region (22) . The DNA sequence of this region can be tested for the presence of gRNA sequences against known edited mRNA sequences. In the case of T. brucei, each minicircle encodes at least three gRNAs located between l 8mer inverted repeats (2 3 ). One could locate these conserved repeats and computer search the intervening DNA sequences .
Continued 95
3: RNA editing in mitochondria Protocol 10 . Continued D. Search for edited m RNA 9. In some cases putative gRNA sequences (fo r instance. as indicated by the presence of a 3 ' oligo(UI-taili have been found prior t o the identification of the corresponding edited mRNA (5 2). In this case the method described in Protocol 9 can be used to search for the correspond ing mANA . However, when genomic kDNA is scanned for the presen ce of t he correspond ing m RNA. t he search is complicated by the fact t hat pre -edited m RNA does not yet contain the U residues t o be inserted or dele te d .
8. Enzymatic activities in the kinetoplast-mitochondrion fraction which are involved in RNA editi ng
Larry Simpson , A gda M. Simpson. an d Beat Blu m mRNA at a normal editing site, with the downstr eam editing sites being full y edited . Recently, mitochondrial ext rac ts have been prepared from T. brucei (18, 19) which show chimaera-forming activities in vitro, For L . tarentolae the chimaeric molecules synt hesized in vitro are similar but distinct from the on es observed in vivo (20). in that attachment also occu rs predominantly at editin g sites, but no editing co uld be detected downstream of the att achment site.
8.2 Prepa ration of mitochondrial extracts Several of the activities involved in mitochondrial RNA editing mentioned in Section 8. 1 may be assa yed in either intact isolated mitochondria or various mitochondrial extra cts. Protocol II describe s the preparation of deter gent lysates of kine toplast mitochondria isolat ed in Ren ografin gradients (see Proto col 6 ). These are the TL . TS , and S-I OO extracts.
8.1 Introduction
Protoc ol 11 . Preparation of Trit on X- 100 Iys a tes
Severa l enzymatic activities which are thought to be involved in editing maxicircle tran script s have been identified in purified kinetoplast-mitochondrion fractions. Th ese include a terminal uridylyl transferase (TUTase) (50), an RNA ligase (50), a site-specific cryptic RNase (67) and a gRNA :mRNA chimaer a-forming act ivity
Equip m ent and reagents
(18- 20) (Figu re 3).
• Ki netoplast 'mitoch ondr ion fr ac t ion iso lated b y flotation in Renogr af in density gradien t s (see Protocol 61 . Resuspend the washed m itochondrial pe llet at 5 mg proteinl ml in 20 mM Hepes-KOH. pH 7.5 . 0 .' M KCI. 0 .2 mM EDTA. 20 % glycerol.
It may be stored fr ozen in 200 j.ll aliquots at - 70 DC . 10 % Trit on X·100 (Pierce) . Pellet pestle mixer econree Glass Co . # 749520. motor driven)
The TU Tase ad ds U residues to the 3 ' hyd roxyl group of RNA molecules with no apparent sequence specifity. Th e ro le the TUTase is thought to be the re-addition of U residues to the 3 ' oligo(U) -tail of the gRNA which has been used as the source of UMP for transfer to the editing site of th e mRNA in two successive trans-esterifications. The function of the RNA ligase in the mitochondrion is unclea r; in the original enzyme cascade model (15) for RNA editing. a ligase is required for joining together the cleaved m RNA fragments after the addition of UMPs to the 3 ' hydroxyl group by the TUTase, but there is no requirement for an RNA ligase in the trans-esterification model (16) . A cryptic RNase in the kineto pla st-mitochondrion fraction can be activated by the add ition of hepar in or by pred igestion of the extract with proteinase K or pronase (67) . Th e autho rs have suggested that the cleavage acti vity involved in the activation is actually a site-specific hydrol ysis catalysed by the same enzyme which normall y trans-esterifies gRNA and the mRNA, and that th is hydrolysis is induced by inhibiting or destroying the TUTase. It is also possible that this acti vity is an enzyme involved in RNA turnover. According to the trans-esterification model of RNA editing (16 .17) , und ine resid ues a re directly transferred from the oligo(U) 3 ' -rail of the gRNA into the m RNA via a gRNA: m RNA chimaeric intermediate. Such chimaeric molecules have been fo und in steady-state mitochondrial RNA from L. tarentola e (16) and T. brucei (68-70). In L. tarentolae, these in vivo chimaeras generally con sist of gRNA covalently linked via the 3 ' ol igo(U) sequen ce to the corresponding
The kineto plast-mitochondrion frac tion isolated from Renografin gradients . . Contams a TUTase . ThiIS activity, 3' h whiIC h a dd s multiple uridine residues to th e . ydroxyl group of RNA molecules, can be solubilized by hom ogen izat ion With 0.3"10 Triton X-IOO. The solubilized TUTase activit y may be assayed by the mcorporation o f UM P residues fro m UT P . Since it onl y requ ires one
96
97
Meth od 1. If necessar y, th aw the mi toc hondrial fr action in ice .
2. Add 6 ~I of 10 % Triton X-lOa to 200 ~ I of the mitoc hondrial fraction in a mi cr ocent rifug e t ube. M ix gently.
3.
H ~m og en ize the mi tochond ria for 15 sec at 5 " C usin g th e pellet pes tle m ixer . This ho mogenate is th e TL ex t ract .
4 . eent~ifu g e t he TL ex t ract f or 30 mi n at 12 000 g at 5 °e in a m icro centnfuge Ito obtain the TS extract) or for 1 h at 100 000 9 at 5 DC (to obtain the 5 -100 extract ). In each case. carefully rem ove and reta in th e supern atant . 5 . If nece ssary, sto re t he extrac ts in aliqu ots at _ 70 0e .
8.3 Terminal uridylyl transferase (TUTase)
3: RN A editing in mitochondria
Larry Simpson , Agda M . Simpson . and Beat Blum
nucleoside triphosphate in addition to UT P , TUTase ma y be assayed in the absence o f the run-on transcription activity, which would also result in nucleot ide incorporation . TUTase activity may be studied in either intact mitochond ria or mitochondrial extracts. To assay TUTase activity in ext rac ts , subst ra te RNAs must be provided by adding cytosolic R A from L. tarentolae. This has five small rRNA components (215 nt , 195 nt, 175 nt, 140nt, and 110 nt) that migrate in po ly. acrylamide gels between the 9S rRNA and the tRNA regions (51) . No subst ra te RNA need be added with intact mitochondria, in which endogenous mitochondrial RNAs a re labelled with [a·J2 p 1UT P, including the 9S and 12S rR NA s and the gRNAs. The mitochondr ial t RNAs do not label well with this enzyme acti vity. Hepa rin (5 ug/rnl) inhibits the TUTase activity of extracts, but does not a ffect the en dogenou s TUTase activity in isolated intact kinetoplast mitochondria, probably due to t he lack of penetration through the mitochondrial membra ne. Pro tocol 12 describes the assa y of TUTase in mitochondrial extracts an d the preparation of the cytosolic RNA sub strate. Under these conditions, the incorporation of UTP occurs linearl y for at lea st 30 min a t 27 ' c. To assa y TUTase in intact mitochondria, Prot ocol 12 should be modified by omitt ing the addition o f the cytosoli c R NA substr ate.
Proto co l 12. Continued 4 . Et hanOl-precipit at e the nuc leic acids Isee Protoco l 3 , steps 15 . 16 ).
5 . Resu spend th e pellet in TMN buffer (1 ml TMN per 2-3 litres of original cell cult ure). Add 0 .005 vol. of 2 mg /ml DNase I and incubat e the mixture for 30 m in at 37 °C. 6 . Ext ract w ith ph eno l :chloroform and ethanol precipitate th e RNA (steps 4 , 5). Wash th e pellet three times with 70 % ethanol and spin vac dry. 7 . Resuspend t he RNA f rom four litres of original cu lture in 1 ml redistilled
ste rile wate r (0 .5 - 1.2~g R NA /~ I I and sto re it at - 20 ' C. B. TU Tase assay
1. In a microcent rifuge tube on ice , mix: 1 ~I
. 1 M DTT
• 30 mM potassium phosphate, pH 7.0 • 10 mM GTP . 0 .2 5 M Hepes-KOH , pH 7.5 e 2 M KCI . 60 mM magnesium acetate
Protocol 12. As s a y of terminal u rid ylyl transferase (TUTase) in isolated m itoc ho ndrial extracts Equip m ent and reagen ts
• cvtosolic RNA substrate (prepared as above )
• Cell ly sate prepared as in Protocol 6 , steps 1-4 • Mitochondrial ext ract (Tt.. TS or 5 -100, see Protoco l 11 l (2 - 3 1J9 protein/pll
e O.2 5 M Heoes -Kcf-t . pH 7 .5 .2 M KCI. 10mM AlP , and DEBl discs as in Proto c ol 9
. 60 mM magn e s ium acetate e 1.0 M DTT
• 0.5 M sodium pho sph at e buffer, pH 6 .8, 0. 5% sod ium pyroph osphate w ith and w ith out 0.1 % 5DS
•
5 ~I 1. 5 ~ I 5J,l1
5 ~I 1 ~I 10 IJI
• 10 mM A TP • I a · 32 p J UTP • mitochondrial extra ct
. 30 mM potassiu m ph osphate , pH 7 .0
5 ~I 5 ~I
1 ut
to 50 ~ I final volume
H2 0
2 . Incubat e the mixture for 40 min at 27 0C.
3 . St op the reaction by adding 50~1 of 0 .5 M sodium phosphate buffer , pH 6 .8 ,0.5 % sodium pyrophosphate , 0 .' % SDS.
e TMN buff er, 10 % 50 S. 2 mg /ml DN ase I sol ut ion s as in Pro to co l 7
4 . Remove 80 IJI aliquots onto DEBl filt er discs.
. 2.0 M NaCI and ph enol :ch loroform as in Protocol 3
5. Process th e discs and determine the reta ined radi oact ivity as described in Protoc ol 9 , ste ps 4-6.
e lO mM GTP
• [a- J 2PJUTP (New Eng land Nuclear; 800 Ci /m mo l, 10 IJCi/lJ1)
8.4 RNA ligase
Method A. Preparation of c vtosolic RNA substrate 1. Centrifuge the cel l lysate (prepared as per Protocol 6 . ste ps 1- 4 ) at 16 000 9 for 10 min at 5 ' C. 2 . Remove the supernatant and add 0 .01 vet . of 10 % 50S . 3 . Deproteinize with an equal vo l. of phenol :chloroform (1: 1) as described in Pro tocol 3 , step 13 . Remove the upper (aqueous) phase to a fresh tu be. Re-ext ract the phenol layer and interface with one vel . of w at er and pool it with the first upper phase.
Continued
98
An RNA ligase activity was first observed in to tal cell extracts fro m T. brucei (71 ). The relationship o f the total cell RNA ligase acti vity to the m ito chondrial activity descr ibed in L. tarent olae is not clear. T he L. tarentolae RNA ligase activit y ~o·sediments with the kinetoplast·mitochondrion fraction (50) . It is so lubilized In Triton X ·lOO and thu s can be assa yed in mitochond rial extracts . Incubation of cyt o so lic RNA with mitochondrial TL extract and [a . J2p I UTP under the condition s required for TUTase in th e presence of A TP and M g 2 + ions yield s a .180 nt labelled RNA product whose relative electrophoretic mobility varies wuh the gel concentration, a behaviour chara cteristic o f circular RNA molecules. RNA liga se is quantitati vely assa yed by th e addition o f [a· J2p J pCp to th e 3 ' term ini o f RNA molecules as d escr ibed in Protocol 13. 99
Lorry S im pson, Agda M. Simpson, and Beat Bl um
3: RNA editing in mi tocho ndri a
Protoco l 13. Assay of RNA ligase in mitochondrial extracts
Prot ocol 14. Assa y for a sequence - or structure-specific cryptic RNase act ivity in mit o c hond ria l ex tr ac ts
Reagents
Equipment and reagents
• Mitochondrial extract (TL. T5 . or 5 -10 0 ; see Pro tocol 11 ) e Cyt osolic RNA substrate (500 ~g /mll pre pared as descr ibed in pro tocol 12
. 10 mM ATP
e l .0 M
MgCl, e O.25 M Hepes-KOH . pH 7 .9
• (a_32P j pC p (New E ngl ~nd Nuclea r; 3000Ci /m imol , 10mC l/mll
se e Protocol 11 J
e OMSO
• Heparin (Sigm a)
e RNA guard (pharrnacia. #27~0815·01 .
• Prot einase K (BRLl: dissolve the pro teinase
33 ~ lmll
K at 10 mg /ml in 50 m M Tris- He!.
• DE8l discs and 0.5 M sodium phosphat e buffer. pH 6. 8. w ith 0 . 5% sodi um pyr oph osphat e (see Protocot 9 1
pH 8 .0. 1
• lO x buffer (30 mM MgCI 2 • 0.1 M Tr is He l, pH 7.5 . 5 0 ~ g !m l he parin]
Method
• mitochondrial extract
e H20
. pNB2 RNA substrate (108 c .p .m .flJg l : t ranscribe Bam HI-digest ed pNB2 plas mid DNA in vitro usi ng T7 RNA pol yme rase and [ a· 32PI UTP as desc ribed in Volume I, Chapter 1 (Prot ocol 4 ) or Chapter 2 , th is volu me (Protocol 4 ) • Ph enol :chlorof or m (1 : 1, v /v l and 2 M NaC I as Protocol 3
. 8% polyacry lami del7 M ur ea dena tu ring
• pNB 2 plasmid (a 198 bp A cc llRsal res tr ic t ion fra gm ent from th e pLt 120 rna xicircle region con taini ng th e 5 '·end
1 . In a micro centrifuge tube on ice, m ix:
e 1.0 M MgCI 2 e O.1 M OTI e O.25 M Hepes-KOH, pH 7.9 e l 0 mM ATP e DMSO e [a .32 p ] pCp e RNAguard
mm cscr,
. 10 mM AlP
e O. l M OTT
• cytoso 1ic RNA sub st rat e
Smal site of pB luescr ipt SK I - I vector f ro m St rat agene 1
• M it ochondrial extract ITl . T5 , or 5 -100.
1-5lJ\
of cytochrome b gene cloned into the
2 ~I 3.3~
(sequencing) gel . lo adin g and electroph oresis bu ff er s and equipme n t Isee Vo lu m e I, Ch apter 1. Protocol 8 or Volume I, Ch apte r 4 , Pro tocol 11 1
M eth od
20~1 1 ~1 10 ~I 2~1 1 ~I
A ctivation of the cr yp tic RNase 1 . Eit her add prote inase K ( l aO I-Ig /m l f inal concent rati on) to the mito chondrial extract and incubate t he extract for 5 min at 37 °C or add heparin t o 5 I-Ig /m l. A ctivat ion by protease predigestion and heparin are synergist ic and t ogether result in better cl eav age.
10 1-1 1 to 50 IJI final volume
2. In a microcentrifuge tube on ice , mi x:
2 . Incubate t he mixture at 4 ° C fo r 15 h .
.
h addition of [a_32p]pCp to the RNA substrate. by SpoltIOg 3 . ~i~~sO~~e ~f et he mixture on D E8l filter discs and processing them as described in Protocol 9 , steps 4 -6.
e 10 x buffer e lO mM ATP . 32P-labelled pNB2 RNA substrate (10' c.p.m.)
5 ~1
• acti vated mitochondrial extract (from st ep 1)
1O ~1
e H2 0
5 ~I 1 ~I
t o 501-11 fina l v olum e
3 . Incubat e the m ixture for 1 h at 27 °C .
8.5 Cryptic RNase
.
:rc
A se uence- or structure-specific cryptic RNase activit y can be dete cted mitO~lOndrial extracts (TL , TS , or 5-100 extra:ts, see Proto~ol ~~~)c~rP bV RNase can be activated either by the addition of heparin ug . predig estion of the lysate with proteinase K. . . uent Protocol 14 describe s the activation of the cryptic RNase and us SUbs.eqd bv . . 200 nt RNA synrhesi ze , assay The RNA substrate used for thi s assay IS a . T7 RNA in vi;ro transcription from a T7 bacteriophage promoter using h me 01 merase. Th e template DNA consists of the 22 nt PER of th e c~oc rOn a ~ :ene together with S6 nt of S' nanking sequence, 186 nt of 3 nank\~ nt of Bluescript vecto r sequence at the S' end. Cleavage R' sequence, and 73 . . ithi the PE . the crypt ic RNa se occurs at one maj or site a nd four mmor sites W I In
g).
100
4 . Phenol ext ract the RNA and reco ver the RNA by etha nol precipitation as described in Protocol 3 . st eps 13-16.
5. Add gel loading buffer 11 0 ~I of 10M urea l and analyse the cleavage produ ct s by gel electroph ores is and aut orad iogra ph y as des cribed in
Volume I, Chapter 1, Pro tocol 8 or Volume I. Chapter 4, Protocol 11.
8 .6 gRNA:mRNA chima era-for ming activit y In order to study the first step of the proposed trans-esterificati on model for RNA editing (16, 17), the authors developed a cell-f ree system using a n extract from mitochondria prepared by sonication followed by salt extraction . Chimaera formation is monitored by the covalent transfer of uniforml y-labelled " P-gRNA 101
3: RNA editing in mitochondria
Larry Sim pson , Agda M. Si mpson, an d Beat Blu m
to a higher molecu lar weight non-radioactive test mRNA by gel electrophoresis. Both the J2 P-gRNA and the test mRNA are synthesized by T7 transcription from PCR-derived DNA templates (63) . Sequence anal ysis of th e reaction products by selective PCR amplifi cation and cloni ng revealed gRNA:mRNA chimaeric molecules as the most prominent products (20). Thi s in vitro system has been used to confirm the importance of the ancho r sequence between the gRNA and mRNA , and should be useful fo r developing conditions for complete in vitro editing (20). The system is described in Protocol 15.
Protocol 15. Continue d
Method 1 . Resuspend the ki netoplast -mitoc ho ndrion fra cti on fr om one lit re of cell cul t ure in 2 ml of ic e-cold hyp oton ic buffer . Keep th is on ic e for 10 m in t o allow th e m ito ch ondria t o swe ll .
2 . Disrupt the sw ollen mit ochondria by sonication at 5 ° C using thre e 20 sec peri od s at 100 wa tts.
3 . Imm ediat ely add 2 ml of 2 x salt extract ion buffer and gentl y agitate t he solut ion w it h a small magnetic sti rrer for 30 m in on ice .
4 . Clarify the extract by centrif ugin g at 50000 9 f or 30 min at 4 0C.
Protocol 15. gRNA:m RNA chimae ra-fo rming act ivity
5 . Con centrate the ext ract at 4 ° C on t w o Cent ric on 10 microcon centrators to an approximate vol , of 100 J.l1 eac h.
Equipm ent and reagents e Centric on 10 centrifugal c ent rat ora (A m ico n 42 05 )
6 . Change the buffer at 4 ° C using the same micro concentrators by appl ying mi crocon -
e Kinetoplast-rnitoch cnd rion fra ct ion isol at ed f rom one litre o f L. terentotee acc or ding t o Prot oc ol 6 . Keep the
kinet op las t -mitoch ondr ion fraction a t - 70 ° C un til extra ct ion is pe rform ed • Uniformly-labelled 32p-gRNA prepared by in vitro trans cr iption with { a _32 P j UTP and T7 RNA polymerase (Vo lum e I, C hap t er 1. Protoco l 4 or Cha pte r 2 , t his vo lume, Protocol 4 ) and isolat ed by ge l elec tr oph or esis (V olume I, Chapte r 1, Protoco l 8 o r Vo lu m e I, Chapter 4 ,
Protocol 1 11 e Test m RNA co nt ain ing t he PER plus fla nki n g (anc ho r sequ enc e) reg ion ; pre pare t his by in vitro tr ans cr ipt ion with T7 RNA polym eras e f oll o w ed by gel ele ct rophore sis (see abo ve ) . 1.0 M Hepes-N aOH , pH 7 .9
two 2 m l lots of extract buffer to each .
a O. l M DTT • O. 1 M PMSF (Sigm a It P-7626 ) in isopr opan ol • Hu m an plac ental ANas e inhi bit or (SAL # 5 518SA) . 20 m g /ml proteinase K (SAL # 55 30U A ) stock • Hy potonic buf f er (10 mM Hepes-NaOH, pH 7 .9 . 0.5mM EDTA I
e Ext rect
bu ff er (20 mM Hepes -N aO H, pH 7.9, 20 % glycerol. 0 . 1 M KCI, 0 .2 mM EDTA . 0. 5 m M PMS F. 0 . 5 m M DTT . prepared fr esh fr om st ock so lutions )
• 2 x salt extraction buffer (40 mM Hepes, pH 7 .9, 50 % glyc erol, 0.84 M NaC!' o.a mM EDTA . 1 mM PMSF . 1 mM DTT . prep ared fr esh fr om st ock so lutions)
e Glvcerol
e Proteinase K sol ut ion
(Q.2 5% N -Iauryl' sarcosine , 25 mM EDT A , 0 .25 m g /m l proteinase K )
• Sonic at or (Br aunsoni c m ic rotip ; Braun )
. 0 . 1 M MgCI2 .1 3 % (w!v) polyethy lene glyc ol 8000 (PEG 8000 ; Sigma # P·2139 1 e N -laurylsarc osin e (Sigm a tI L ~ 5 1
2 5)
9 . Add 8 J.l1 of reaction buffer and 15 J.l1 of thawed mitochondrial extract (from step 7) to the ann ealed RNA s. Incubate for 15 -120 min at 27 0C.
10. Stop the reaction with 100 J.l1 of proteinase K so lut ion and incubate at
37 °C for 20 min.
11 . Reco ver the RNA by phenol extraction and ethanol precipitation as des c ribed in Protocol 3 , steps 13 -16 . 12 . Analyse the resu lting RNA chimaeric produ ct s on denat uring acrylamid el urea gel s f ollowed by aut oradiography (Volum e I, Chapter 1, Protocol 8. or Vo lume I, Chapt er 4 , Protoco l t t l .
pH 7 .9 . 0 . 1 M KCI, 1 mM EDTA I
. 0 . 5 M EDTA (adj ust the pH to 8 .0 w ith NaOH )
a 2 .0 M NaCI a l .0 M KC I
8. Anneal equ imolar amounts of the 3 2P-labelled gRNA and the test mRNA in 2 .5 J.l1 of annealing buff er. Den ature the RNA for 3 min at 70 0C then anneal them at 37 ° C and 25 DC for 10 m in each .
e Ann ealin q bu ffer (20 mM Hepe s- NaOH, • React ion buf fer (8% PEG 8000, 12 .5 m M MgCI 2 , 2 .5 mM A TP , 1 uni t /1J1 ANase inhibi t or)
(ultrapure ; SRL 55 14UA )
7 . Quick f reeze 200 J.l1 aliquots of the poo led extracts in dry ice lethanol and store them at - 70 ° C. Chimaera-f orming activit y is maintained f or up to 3 months .
# 4 5 10
with
• Mate rials and equ ipment fo r p olyac rylamide /urea gel ele ctr opho re s ts as described in V olume I. Chapter 1, Protocol 8 or Vo lume I, Ch apter 4 ,
. 20 mM ATP (Pharma cia # 27 -2056-01 )
Protoc ol 11
Acknowledgement This work was support ed in part by NIH research grant A 109102.
References Continu ed
102
Th e RNA products from Protocol 15 may also be analysed by sequence determ ination following RT -PCR and cloning . In order to amplify specifically the exogenous RNA , a 'tag' sequence ma y be added to the test RNA . Further details of RT - PCR can be found in Volume I, Chapter 2, Section 3.4 and Volume I, Chapter 3, Section 2.2.3
1. Hodges, P . E ., N avarat nam , N ., Gr eeve, J. C. and Scott , J . (1991). N ucleic A cids Res. , 19, 1197. 103
3: RNA editing in mito ch ondria 2. Lau , P . P. , Xion g, W., Zhu, H. -J .. Chen , S. -H . , and Chan , L. (1991). J. BIOI. Chem., 266, 20550 . 3. Gualberto. J . M.• Lamattina, L. . Bonnard , G.• Weil, J . H., and Greinenberger, J. M. (1989). Nature, 341, 660. 4. Cov ello , P . S. an d Gray , M . W . (1989). Nature, 341, 662. 5. Mahendran, R.. Spottswood . ~1. R. , and Miller , D. L. (199 1). Na ture , 349 . 434. 6. Simpson L. and Shaw , J . (1989). Cell, 57, 355. 7, Vidal, S" Curra n, 1. , and Kolakofsky, D. (1990) . EMBO J" 9, 2017. 8. Dri scoll , D. , Wynne , 1. , Walli s, S., and Scott, J . (1989). Cell, 58, 519. 9. Chen, S.-H ., u , X., Liao , W . S. L. , Wu, J , H .. and Chan. L. (1990). J . Bioi. Chern., 265,6811. 10, Lau , P , P. , Chen, S.-H ., Wan g, J . C .. and Chan , L. (1990), Nucl. A cids Res.. 18, 5817. II. Smith, H. C .. Kuo , S.-R.. Backu s, J . W ., Harris, S. G ., Spar ks, C. E., and Sparks, J . D. (1991) Proc. Na tl Acad. Sci. USA , 88, 1489. 12. Greeve, L. Navaratnam, N ., and Seen, J . (199 1). Nuc/ . Acids Res. , 19, 3569. 13. Backu s, J. W . and Smit h , H . C. (1991). Nucl. Acids Res.. 19, 6781. 14. Simpson , L. (1990) . Science, 250, 512, 15. Blum, B., Bak alara, N" and Simpson, L. (1990) . Cell, 60 , 189 . 16. Blum, B., Sturm, N . R.. Simpson, A . M ., an d Simpson, L. (199 1). Cell, 65 , 543. 17. Cech , T. R. (1991). Cell, 64, 667. 18. Koslo wsky, D. J. , Gri nger , H . U. , Mor ales, T. H .. and Stuart, K. (1992). Nature, 356, 807. 19. Harri s, M. E. an d Hajduk , S, L. (1992). Cell, 68 , 109 1. 20 . Blum, B, and Simpson, L. (1992). Proc, NaIl Acad. Sci. USA , 89 , 11 944. 21. Simpson, L. (1987). Annu. Rev. Microbial.. 41, 363. 22. Stur m, N. R. and Simpson , L. (1991). N uel. A cids Res.. 19, 6277. 23. Pollard , V. W .. Rohr er , S. P. , Michelotti , E. F., Han cock, K., and Hajduk. S. L. (1990) . Cell, 63, 783, 24. Kidan e, G ., H ughes, D" and Simpson, L. (1984). Gene, 27. 265. 25. Maslov, D. A. an d Simpso n, L. (1992). Cell, 70 , 459 .
26. Morel , C., Chia ri, E., Camargo , E., Mattei, D., Romanha, A., and Simpson, L ( 1980). Proc. No ll A cad. Sci. USA , 77, 6810. 27. St urm, N. R.; Degrave, \\'.• Mo rel. C., an d Simpson, L. (1989). Alai. BiochemParasitol., 33,205. 28. Go mez-Eic helma nn, M, C .; Holz, G ., Jr. , Beach, D.. Simpson. A . M.. and Simpso n, L. (1988). Mal . Biochem . Parosirol., 27, 143. 29. Simpso n, L. (1972). Int . Rev. Cyto l. , 32, 139. 30. Stuart, K. (1991). Annu. Rev, Microbiol., 45, 327. 31. Benne, R. (1989). Biochim. Biophys. A Cla, 1007, 131. 32. Feag in, 1. and Stua rt , K. (1988). Mol. Cell Bioi" 8, 1259 . 33. Feagin, J ., Ja smer, D., and Stuart , K. (1987). Cell, 49, 337, 34. Duszenk o . M., Ferguson, M ., Lam o nt, G ., Rifk in , M.• and Cross. G . (1985). J . Exp . Med. 162, 1256. 35. Brion es. M. R. S., Nelson, K. • Beverley, S. M., Affonso. H . T ., Camargo , E. P., and Floeter-Winter, L. M. (1992) . Mot. Biochem . Parasitol. , 53, 121. 36. Simpson, L. and Berlin er , 1. (1974). J . Prorozool., 21, 382, 37. Simpso n. L. and Simp son , A . (1974). J . pr ot ozoot. , 21, 774. 104
Larry Simpson, Agda M. Simpson , and Beat Blum 38. Simpson, L. ( 1979). Proc. NaIl Acad. Sci. USA, 76, 1585. 39. Ryan, K. A., Shapi ro, T . A ., Rau ch, C. A ., and Englund , P , T . (1988). Annu. Rev. Microbiol. , 42, 339. 40. Hajduk, S. , Klein , V., and Englund, P . (1984), Cell, 36 , 483. 41. Tibayrenc, M. , Ward, P ., Moya , A ., and Ayala , F. (1986). Proc. Nat l A cad. Sci. USA, 83, 115. 42. Bru nk, C. and Simpson , L. (1977). Anal. Biochem. , 82, 455. 43. Go ncal ves, A . M., Nehme, N . S., and Morel, C. M. (1984). In Genes and antigens of parasites (ed . C . !\1. Mo rel), p. 95. Fundacao Oswa ldo Cr uz, Rio de Janeiro . 44. Beidler , J . L., Hill iard, P , R., and Rill , R. L. (1982). Anal. Biochem., 126, 374. 45. Lope s, U. , Mornen, H ., Grimaldi, G ., Mar zochi , M. , Pac heco , R., an d Morel, C. (1984). J . Parasitol. , 70, 89. 46. Morel , C. and Simpson, L. (1980). Am. J. Trap. Med. Hyg.. 29, 1070. 47. Avila, H ., Gon cal ves, A . M., Nehm e, N. S., Morel, C. M., and Simp son , L. (1990). Mol . Biochem. Parasitot., 42, 175. 48. Simp son, L. (1968). J. Protozoot. , 15, 132. 49. Braly, P ., Simp son , L. , and Kretzer, F. (1974). J. Protozool . , 21, 782. 50. Bakalara, N., Simp son, A . M., and Simp son , L. (1989). J. Bioi. Chem., 264,1 8 679. 51. Simpson , L. and Simpson , A . (1978). Cell, 14 , 169. 52. Simpson, A . M" Suyama, Y. , Dewes. H . , Ca mpbell, D., and Simpson, L. (1989). Nu cl . Acids Res. , 17, 5427. 53. Hancock, K. and Hajduk , S. L. (1990). J. Bioi. Chem .. 265 , 19 208. 54. Maslov. D. A ., Sturm, N . R., Niner, B. M., G ruszyns ki, E. S., Peris, M., and Simpson , L. (1992) . .Wol. Cell. Bioi. , 12, 56. 55. Sturm , N. R. an d Simpson, L. (1990). Cell, 61, 871. 56. Shaw, J., Ca mpbell, D., and Simpso n, L. (1989). Proc. Nat, Acad. Sci. , 86, 6220, 57. Blum , n. an d Simp son , L. (1990). Cell, 62, 391. 58. Masuda , H ., Simpson , L. , Rosenblatt. H . , and Sim pson , A. (1979). Gene, 6 , 51. 59. Wood , W., Gitschier, J ., Lasky, L., and Lawn, R. (1985). Proc. Natl Acad. Sci. USA, 82, 1585. 60. Abrah am, J ., Feagin, J ., an d Stu art, K. (1988). Cell, 55, 267. 61. Decker, C . J . and So llner-We bh, B. (1990). Cell, 61, 1001. 62. McP herso n, M, J . , Quirke, P. , and Taylor, G . R. (ed.) (1991). PCR: a practical app roach. IRL Press, Oxford . 63. Milligan, J . F.. Groebe, D. R., Wit herell, G. W " and Uhlenhec k, O . C. (1987) . Nuel . Acids Res., 15, 8783. 64. Von Haeseler. A., Blum, B., Simpson , L. , Sturm , N. and Waterman, M. S. (1992) . Nuel. Acids Res. , 20, 2717. 65. Sha w, J. , Feagin, J . E" Stuart, K., and Simpson, L. (1988), Cell, 53, 40 1. 66. Van der Spek, H ., Arts, G .-l ., Zwaal , R. R., Van den Burg, J ., Sloo f', P .. and Benn e, R. (1991) . EM BO J" 10, 1217, 67. Simp so n, A. M., Bak alar a , N. , and Simp so n, L. (1992). J. Bioi. Chem ., 267, 6782 . 68. Koslowsky, D. J ., Bhat, G . J., Read, L. K., and Stuart, K. (1992). Cell, 67, 537. 69. Read , L. K., Myler, P , J .. an d Stuart, K. (1992). J . Bioi. Chem ., 267, 1123. 70. Read , L. K. , Cor ell, R. A ., and Stuart , K. (1992) . Nucl. A cids Res.. 20, 2341. 71. White, T . and Borst , P , (1987), Nucl, Acids Res. , 15 , 3275. 105
Analysis of messenger RNA turnover in cell-free extracts from mammalian cells JEFF ROSS
1. Introduction Messenger RNA turnover influences gene expression in most or all cells, from bacteria to mammals 0, 2). Two sorts of observations support this notion. First, the steady-state levels of ma ny mRNAs are determined more by their half-lives t han by the rat es at which their genes are transcribed. Second , the half-lives of some mRNAs are not fixed but vary in response to the nutritional status of the cell, its position in the replication cycle, or its stage of differentiation . Ma ny impo rtant questions need to be answered to understa nd how mRNA turnover is regulated and how it influences cell growth, differentiation, and neop last ic tra nsfo rmation. How man y mR NA-degradin g enzymes exist per cell? Which are endo nucleases and which exon ucleases, and at what site in a mRNA molecule do they first begin the degradation process? What is the pathway of mRNA decay, i.e. what are the degradation interm ediates? What sequences in a mRNA determi ne its half-life? How does translation affect mR NA turnover? What trans-acting factors regu late mRNA turn over and how do they function? These questions illustrate the need to use in vitro systems to investigate mRNA turnover, particularly in cells of higher organisms in which genetic anal ysis is difficult. In vitro systems offer several advantages (Table I). Two of these are particularly important. (a) mR NA decay intermediates that are difficult or impossible to detect in intact cells can somet imes be identified in cell-free systems. The average int racellula r half-life of an mRNA might be 20 h, but, once each individual mR NA molecule begins to be destroyed, the reaction is apparently complete in minutes or seconds. If degradation occurred slowly, it would be easy to detect mRNA decay intermediates in cells, which, with few exceptions, is not the case. Since degradation rates are slower in vitro than in cells, it is possible to ident ify intermediates.
4: mRNA turn over in cell-free extrocts
'eff Ross
Table 1. A dvant ages of analysing mANA stability in cell-free ex tracts e mRNA decay interm ed iate s wh ich are diff icult or impossible t o observe in cells can be dete ct ed in vitro .
• Mess enger ribonucleases and trans·acting facto rs w hich stabilize mRNA can be assayed and purifie d. • Relative mRNA turno ver rates can be measured w ithout using inhibitors of transcription , most or all of which are toxic to cells and not specific for transcr iption. • The eff ects of mRNA seque nces and structures on half-li fe can be readily investigated.
Tab le 2 . Use of endogenous and exogenous substrates for in vitro mRNA turnover studies Endogenous substrate (mRNP. eithe r free or polyso me-associated)
Advantages • ' authentic' substra te . made by cellular transcription /modification enzymes • know n to be respons ive to at least some trans-acting regulatory factors in vitro
Disadvantages • diffi cult t o det ect if the mRNA is scarce • pron e t o degradat ion during isolation • cum bersome when varian ts need to be analysed
(b) In vitro mRNA deca y systems, like in vitro tra nscription systems, permit the investigator to detect and purify enzymes, co-facto rs, an d trans-acting regulators . Even if prac tical schemes are available to select mutations in genes involved in mRNA turnover, some of the interesting mutants might be leth al, making it necessary to devise in vitro assays to char acterize the relevant gene products. Thi s chapter focuse s on three in vitro methods currently used to investigate mR NA tu rnover in cells o f higher organisms. It is intended as a guide and by no mean s presents all possible assay schemes. Improvements on the current systems will surely continue to be made. The ad vant ages and disadvantages of each system will need to be considered by the investigator in the co ntext of the experiments. In choosing a suitable in vitro system, it is important to remember that relative mRNA half-lives in vitro should ret1ectthose in cells: mRNAs that a re stable in cells should also be stab le in vitro and , conversely, unstable mRNAs should be unstab le in both . Half-lives in cells and in vitro are unlikel y to be identical , since mRNA decay occur s more slowly in vitro. Howe ver, if two mRNAs are degraded at rates dif fering by 20-foJd in cells but by 3-fold or less in vitro, then the extract or reaction conditions should be modified . For example, mRNA -specific regulat or y factors might have been inactivated during extract preparation. The protocols in this chapter suggest some options for mod ifying each system .
2, Cho ice of mRNA substrate Th e choice of substra te (endogenous or exogenous) depends on the experimental question being asked and on various parameters such as the numb er of available cells, th e am ount per cell of the mRNA of intere st, and the acti vity of ' nonspecific' cellular RNa ses (Table 2). NB. The term non-specific RNase denotes an activity capable of degrading all RNAs (rRNA, tRNA , or mRNA) at similar rates in vitro. The term is not meant to imply lack of function for such an RNase, only that it is not yet und ersto od what its function might be. 108
Exogenous subst rate (protein -free cellular mRNA or in vitro synthesized ANA l
Advantages • eas y to detect . espec ially if radiolabelled • easy to prepare va riant s
Disad vantages • translated w ith low efficiency in some extract s • might not respond to regulatory factors
2,1 Endo~enous substrates (free mRNP or polysomeassociated mRNP) Polysomes cont ain the substrate messenger ribonucleoprotein (mRNP) to be ana ly.sed and often the messenger RNases as well. Th ere a re several ad vant ages to usmg endogenous mR NP substrates. First, the nature o f the substrate is relatively clear, i.e, the mRNA was synt hesized by the cell and is competent for tran slation , since it is polysome-associated . Th erefore, within the constraints Imposed on any subcellular fra ction prep ared in the laboratory, an endogenous SUbstrate could be considered more or less authentic. Second , some endogen ou s SUbstr ates respond in vitro to factors involved in regulating mRNA stab ility (3-7). In contrast, the turnover o f some exogeno us, depr oteinized mRNA substrates IS una ffected by such regulatory factors (see Section 2.2). . There a re three potential disad vantages of endogenous substrates. First If the mR NA is scarce, large amounts of poly somes might be requi red to detect It. Second , polysomes containing undegraded mRNA can be difficult pr.epa re, especiall y from some prim ar y animal tissues, unless special echmques. are devised to reduc e non-specific RNase activity (8). Th ird, it might be difficult or Impossible to engineer cells to express interesting modified SUbstrates, for exam ple, mRNA that has a discrete 3 ' terminus and th at lacks a poly(A) tract.
:0
109
4: mRNA turn over in ce ll -free extrac ts
2.2 Exogenous substrates The assay is relatively easy with an exogenou s substrate, especially if it is radiolabelled . The substrate is simpl y incubated with a cell extract , purifie d , and a nalysed by gel electrophoresis. Exogenous substrates can be modifi ed mor e easily than endo genou s mRNA s, without concern for the effect s o f the modi fication s on transcription, capping, or tra nsport to the cyt oplasm . For exa mple, substrates of almos t any size and sequence can be generated rap idly by cutting a cDNA clone at differ ent sites and transcribing each template with an appropriate bacteriophage RNA polymera se. Ho wever, there are two important disadvantages o f exogenous substrates. First, mRNA turnover and translat ion ar e often linked 0, 9, 10), but it is diffi cult to translate exogenous mRNA efficientl y in extracts from nuclea ted mammalian cells. Second , some exogenou s substrates, unlike their endogen ous counterparts, fail to respond to regulatory factors in vitro. For example, polysome-associated (endogenous) c-myc mRNP is regulated in vitro by cyto solic destabilizers (3, 7) but deproteinized (exogenous) c-myc mRNA is not (G . Brewer and J . Ross, unpublished observations).
Jeff Ross
(e) Filter all solutio ns a nd store them at 4 °C or at - 20 °C. Magnesium ions a nd carbon sou rces (e.g. sucrose) might permit growt h of certain micro orga nisms, all of which contain RNases. The best way to han dle these solutions is to sterilize them by filtration through a micro pore (.;; 0.45 prn) membra ne (e.g . Millipore HAWP) in a tissue culture hood. Store them in a refrigerator, only open the stock bottles in a sterile hood, and dispense them ~si ng sterile pipettes. Tape a note on each bottle stating that the solution IS sterile and must be opened only in a hood. Some laboratories pretreat their glass-distilled water by adding DEPC to a final concentration of 0.020/0 and then autoc laving HPLC-grade water or Water prepare d by deionization a nd double distillat ion might be suitable for use without sterilization or DEP C treat ment . (I) Carry out all procedu res at 2-4 °C, from the time the cells are harvested
unt il the extrac ts are placed in the - 70 °C freezer.
Prot ocol I describes th e treatment of glassware with DEP C and by ba king. Protocol 1. Treatment of gla ssware to ina cti vat e RN ase s .
3. Preparation of undegraded mRNA and polysomes When using an endogenous substr ate, it is important that the mRNA and polysornes prepa red are undegraded in order to obtain meaning ful data. Contamination of glassware or reagents with RNases , which seem to be ubiqu itous , can ruin a preparation . T he following precauti ons, coupled with a large dose of com mo n sense, will help maintain RNA integrity: (a) Choose the proper cell or tissue . Some cells contain more non-spe cific RNa ses than others, so even the most careful investigator might have diffi culty isolating intact polysornes from them . In general, tissue culture cell lines contain less RNase activity than animal tissues. If necessary, various RNase inhibitors are availabl e. Two of them , placental RNase inhibitor a nd diethylpyrocarbonate (DEPC), are discussed in this chapter. Va nadyl ribo nucleoside complex at 4 mM and heparin at 0.2 mg/ml have also been exploited to block RNase activity during preparation of extracts from animal tissues (8). (b) Use sterile plasticware whenever possible. (c) Wear disposa ble plastic gloves at all times, since body surfaces and fluids contain active RNase(s). (d) Pre-r inse all glassware with DEP C at least one day before use (see Protocol l ). DEP C treat ment is recomme nded , because autoclavi ng might no t inactivate all RNase con taminants in the glassware. More over, DEPC is an excellent sterilizing agent. 110
1 . Add
a small volume of DEPC to each ite m of glasswa re.
2 . Rotat e the glassware so that all inner surfac es are ex posed to the DEPC.
3. Ca~tion : DEPC can form a carcinogenic compound by reacting with primar y
~mlnes . Therefore , after rinsing the glassware, pour the DEPC into a beaker
In
a hood, add excess water , and then let th e liquid evaporate to dryn ess.
4 . Rinse the glasswar e thoroughl y wi th distilled wa t er unt il all tra ces of the distincti ve DEPC odour have disappeared (a minimum of t en rinses should do ). 5 . Bake the glasswar e in a 22 0 °C oven for at least 1 h, prefer ably overnight .
4, Preparation of cell extracts for in vitro mRNA decay reactions There is no single, universally accepted standard method for a nalysing mRNA turnover in vitro, nor is one likely to evolve . Different extracts and reaction condition s a re required to account for variables which can influence mRNA turnover within the cell. These variables, in tu rn , depend on the horm onal and nmnuon al .'tatus of the cell, its position in the cell cycle, and its stage of dIfferentIatIon . For example , the half-lives of some hepatic cell mRNAs can fluctua te by 30-fold or more when certai n hormones a re added to or removed from the Culture medium . The following questions are relevant to the stabilit y of mo st mRNA s and must be considered in decid ing how best to carry out cell-free mRNA deca y reaction s. 111
4 : mR NA turn ov er in cell-free extracts • Is th e mRNA translated in cells and , if so, at what rate? • Does the cell contain trans-acting regulatory factors which influence the mRNA half-life? If so, are the se factors freely diffusible throughout the cyto sol or are they associated with particular cyto plasmic components such as ribosomes, endoplasmic reticulum, the mRNA itself, etc .? • Is the mRNA itself localized as a result of its association with microfilaments, end oplasmic reticulum, etc.?
Jeff Ross to micrococcal nuclease , which eliminates th e mRNA but not the ribosomes or other compo nents necessary for translation (see ProtocoI5). After inactivating the micrococcal nuclease, the extract can be incubated with exogenous mRNA under conditions permitting acti ve translation and the half-life of mRNA which ha.d become polysome-associated can be determined (see Protocols 10 and 11) , Mlc roc~cca l n~cl~ase treatment of detergent-free extracts is fruitles s, because tra nslation iruuanon occurs so inefficientlv. It is important to stress once again that these methods should be used as guides and ~a~ need t? be modified to suit specific experimental parameters. Flexibil ity IS essential and extracts can , in theory, be mixed in various combinations. For example, if an investigator wishes to analyse the turn over of a mRNA which is being actively translated, radiolabelled mRNA can be incubated .In nuclease-treated reticulocyte extract. However, suppose that the mvesngarn- suspect s that the intracellular half-life of this mRNA IS influenced by a cytosolic trans-acting regulatory factor, which is likely to be. pres~nt In SUO . Since reticulocytes are end-stage cells whose SI30 contains primarily haemoglobin, they are unlikely to have retained all of the SI30 regulatory factors which are present in undifferentiated , nucleated cells. Therefore, the best a pproach might be to mix the mRNA to be tested with reticulocyte extract plu s SI30 isolated by detergent -free lysis of nucleated cells.. The multlcomponent reac tion mix might reflect the int racellular conditions to which the mRNA is normall y exposed (efficient translation pl us a soluble regulatory factor) more accurately than reactions containing on ly detergent-free polysomes (minimal translation, no soluble factors) or detergent-free polysomes plus SI30 (soluble factors pre sent, but minimal tr anslatio n].
Three methods are described for preparing extracts. Each has ad vantages and di sadvantages, depending on the mRNA to be analysed and the variab les cited above. The investigator must choo se the appropriate extract or combinati on of extracts, substr ates, and in vitro conditions so as to refl ect as clo sely as possible the circumsta nces under which th e mR NA is degraded wit hin the cell. The first method exploits ultracentrifugation to prepare pol ysomes and post-polysornal supernata nt (5130) from cells homogenized in a detergent-free, low-salt bu ffer (see Proto col 2) . A major advantage of th is method is its flexibil ity. Since polysome s contain mRNA and messenger RNa se(s), the turnover of polysome-associ ated m RNAs can be ana lysed simply by incubat ing polysom es under a ppropriate conditions (see Prot ocol 6) . The 5 130 seem s to have little messenger RNase activity but does contai n soluble, trans-acting facto rs which regulate th e half-lives of certain mR NAs. Th ese facto rs ca n be assayed in reac tions containing mixtures o f 5130 and pol ysomes (see Protocol 6). Exogenousl y add ed, radiolabelled mRNA can also be assa yed in reactions cont ainin g polysomes or polysomes plus 5130 (see Proto col 7). Ribosom e-bound messenger RNases can be separated from polysomes by exposing the polysornes to high salt, pelleting them in an ultracent rifuge, and har vesting the supernatant (see Protocol 3). Th e high-salt ext ract can then be used to a na lyse the decay of radiolabelled mR NA (see Prot ocol B). On e disadvantage of using detergentfree polysomes is a very low tra nslation rate , bot h for endogenous and exogenous m RNAs. The seco nd method for pr eparing extracts uses lysolecith in to perrneabilize ad herent tissue culture cells, which are the n scraped fro m the culture dish, brok en , and briefly centrifuged to remove nuclei (see Protocol-tv. These extracts a re easy to prepare, and the y suppo rt translation of endog eno us and exogen o us mR NA s more efficient ly than detergent-free extracts (see Prot ocols 9 and 10 ). One di sadvantage is a lack of informati on on wheth er lysolecith in ext racts retain their translationa l and mR NA tu rnover activit ies when they are fractionated or frozen. The third method uses a lysate prepared from reticulocytes, which are immatur e, enucleated red blood cells containing abu ndant polysornes (see Protocol I I). This system is the most active of the three fo r tran slation, and reticulocyte extracts ca n be purchased at mod est cost. Th e lysolecithin an d reticul ocyte extracts ca n be made mR NA-dep endent by expo sing them br iefly
More~ver, bu ffer s can be design ed to contain combina tio ns o f inhi bitors which might be more effecti ve than a single inhibitor. The composition o f on e ?uffer used to isolat e intact polysom es from primary cultures of amphibian liver IS given below (8):
112
113
. Prot~col 2 and th e following protocol s pertain to tissue culture cells, since limited mformation is available on the use of extracts from animal tissue s for cell-free mRNA decay. In fact, it is unclear whether the se protocols can be used with pn~ary animal tissues . Investigators wishing to use animal tissues sho uld take spe cial care to redu ce non-speci fic RNa se ac tivi ty. Th e following RNa se inhibirors might be useful: • Pla cental or liver RNase inhibitor (II), available from Amer sham , Boehring er Mannheim, Calbiochern, ICN Biomedicals, Promega, Stratagene. Useful amounts must be determined empirically, but an appropriate starting point IS 100 umts per ml. • Rib onucle?side-vanadyl complexes (12), a va ila ble fro m Bethe sda Research Lab orat ories, New England Biolabs, Sigm a . Starting point : 2.5 mM.
4: mR NA turnover in cell-free extra cts
Jeff Ross
. 0.2 M sucrose
Protocol 2. Continued
. 70 mM KCl
6. Resuspend the cells in solution A . Th e v olume wil l depend pr imarily on the c ell number and the cell t ype , since so me ce lls do not ly se eas ily by mechanica l homogenization , especially if they are too co nce ntrate d (see step 8 below ). The ma ximum con centrati on of c ells f or high effi cien cy homogenization (9 5% or greater lysis ) must be determin ed in a pilot experiment by c ounti ng the per centage of broken c ells us ing a haemocytometer and microscope . In m ost cas es , lO Bc ells/ml is a reasonable starting po int, but the optimum number mi ght be up t o 'l Ovfold lowe r.
.1 0 mM MgCl, . S mM NaCI .4 mM va nadyl- ribonucleoside complex . l mM EGTA . 0.2 mg/ml heparin • 1 mM OTT
7. Transfer the cells to the prechilled Pott er -El vehjem homogenizer.
• 100 units /ml placental RNase inhibitor • SO mM Hepes- NaOH , pH 7.8
Protocol 2. Prep arat ion of polysomes and post -poly som al supernat ant (5 13 0) fr om ex po ne nt ially-g ro w ing t issue cult ur e c ells ly sed w it ho ut det erqent"
Equipmen t and reagents NB A ll buffers and tubes sho uld be at 2-4 DC p rior to the expe riment pes tle w it h a clea ran c e of app rox imately 0.1 m m ; use a vessel w ith a c apac ity 2t o 3-fold g reate r th an t he vo lume of So lution A used for c ell ly sis (see step 6 l ; a Dou nce -type glass homogen izer wi th a t igh t- f itt ing p est le (cle arance 0.06 mm ] may also be requi red (see step 17 )
• Expon ent ially-growing t issue cultu re cells
• So luti on A 11 mM po t assium aceta t e, 2 mM magnes ium ace tate , 2 mM OTT ,
10 mM Tria-acetate . pH 7 .6 at room t emperature )
• Solution B (sol ut ion A cont aini ng 30% Iw/vl RNase -free sucrose ] • Tissue cult ure m edium without serum or antibio tics; med ium Ham's F 12 (Flow, Gfbc o, H vmonej has been use d in the aut hor ' s lab oratory , but any isoto ni c gr ow t h medium sh ould do
• Pcttet-Elvebje m homogenizer wit h Teflon
• Lo w -speed cen trifuge , u lt racen t rifu g e, and app ro p riate tubes ; oorvanomer ultracent rifuge tubes are prefe rable ; the y need no t be baked but should be th or oug hly rin sed wit h RNase -free w at er and t hen drained dry im m ed iat ely pr ior t o use
8. Place the homogenizer v essel in an ice-w at er bath and break the ce lls by mo ving the pestle rapidl y up and down and twisting it left and right at the same time . The number of up and down strok es requi red for efficient cell ly sis w ill var y , depending upon the ce ll type and the ir conc ent rat ion . A s a st arti ng poi nt , 20 strokes should do (see ste p 6 above ). A void gen erating exc ess bubbles . 9. Centrifuge the homogenate for 10min at 8300g (av erage), w hic h corresponds to 9000 r.p .m. in a swinging bucket rotor , radius 9 .1 em . 10. Rem ove and save the supernatant {5 10 l, tak ing c are t o av oid th e pelleted material (nucl ei and membranes ). It is best to leav e behind a small amount of supernatant so that none of the pelleted mat eri al con t aminates the 510. 11 . Gently layer the 510 ove r a cushi on of solution B in an ultracentrifuge tube . The v olum e of solution B should be approximately 25 % of the tube capac it y. If the tube is insufficiently filled wit h 510 , add soluti on A.
12. Centrifuge at 130000g (average). 2 °C , 2.5h , which corresponds to 36000 r.p .m . in a Beckman 5W60 rotor. If yo u intend to save the 5130, allow the rot or t o de celerat e without the brake. 13. Car efully rem ov e the supernatan t (5130 ) abo ve the su crose c ushion, taking care not t o har v est the w hiti sh material located on top of the cushion. Leave a small amount of S 130 abov e the cus hion, if nec essary . Do not be conce rned if some of the lipid at the t op of the tube is harvest ed along w it h the 51 30. Set aside t he 5130 in an ice bu cket. 14. A sp ir ate t he remain in g liqui d , inclu ding the suc rose , and di sc ard it. A blu ish -white pell et of pol ys omes shou ld be vi sible in th e cent re of t he tub e bottom .
Method 1. Coun t t he c ells. 2 . Centrifuge t he cult ure fo r 4 min at 34 9 (av erage), cor respondi ng to 500 r.p. m . in a low-speed ce nt rifug e wit h a swing ing bucke t radiu s of 8 em . Not e : Th is and all subsequent pr oced ur es sho uld be perfo rmed at 1-2 ° C. All bu ffers sho uld be kept ice co ld th rough out th e ma nipulation s.
15. Lay t he cent rif uge tube on it s side and cut it ac ross wi t h a razor bla de appr oximately two-third s fr om t he t op. Discard the top piec e.
4 . Gently resuspend the cells in ice -cold t issue cult ure meaium wit hout serum or antibiotics . Cen tri fuge as in step 2 . 5 . Repeat steps 3 and 4 .
16. Gentl y add approximately 1 ml of sol uti on A down th e sides of th e bottom part of t he ce nt rif uge tube. Gentl y move th e tube back and f ort h t o was h any co nta min at ing mat eri al, incl uding suc ros e, fr om the uppe r surface of the polysome pellet . A spirat e or pour off th e liquid. Repeat t his washing st ep once . It is not nec essary to perf orm the se w ashin g st eps quickly , since the poly some s will no t dissolve in th e buff er, but it is necessary t o add the buf f er slow ly and gently , so as not t o di slodge th e poly some pellet f rom th e bottom of t he tu be.
Con tinu e d
Continued
3 . Aspi rate and disca rd t he sup ern ata nt .
114
115
4: mRNA turnover i n cell-free extracts
Jeff Ro ss
Protocol 2. Continued
Protocol 3. Continued
17. Resuspend the polysomes in a small volume of solution A . The volume will depend on the quantity of polysomes recovered and t he subsequent experiments . As a rule , use a volume such that the polysomes from 10 6 cells will be resuspended per microlitre. Pol ysom es can be difficult to resuspend . They tend to be ' st icky', forming clumps that are difficult to disperse . First, try resuspending them by pipettinq the liqu id up and down , using an automatic pipettor with a 2001-11 or 1 ml disposable tip. If visible clumps remain , transfer the material to a tight-fitting Dounce type glass homogenizer with a clearance of 0.06 mm . Several up and down strokes with this homogenizer should break up the clumps .
4 . Stir gently for an additional 15 m in in the cold . Avoid bubbles .
18. Measure the optical density of the polysomes and S 13 0 at 260 nm and 280 nm and store bo th preparations in mu ltiple aliquots at - 70 °C. As a guide , po lysomes isolated from a logarit hm ically growing human erythroleukaemia cell line (K562) resuspended at 106 cells ' polysomes per microlitre give an absorbance reading of 220 -260 through a 1 em
5 . Layer the material over an appropriate volume of solution B in an ultracentrifuge tube and centrifuge for 2 .5 hat 130 OOOg Iaveraqe ], 2 °C . 6. Carefully harvest the RSW located above the sucrose cushion. Retain the salt-washed polysome pellet (see step 8 ). 7 . Mix it gently but thoroughly, determine the protein concentration " (15), and store the RSW in small aliquots at - 70 ° C. 8 . If desired , the salt -washed polysomes in the pellet at the bottom of the tube (step 6) can also be harvested , as per Protocol 2 , steps 16-19. " From ref. 14. b Kits available from Bio-Rad and Pierce con tain reagents for the Bradford prot ein determination assay.
light path at 260 nm. 19. To assess po lysome integrity, RNA from a small amount of the pre paration can be extracted and analysed by aga rose gel electrophoresis (13, 14). The rRNA can be visualized by ethidium bromide staining (13, 14) , The rRNA is intact if the 28S and 18S rRNA bands are discrete and
it the band intensity appears to be approximately 2:11285 :1851. ~
from refs 13 , 14 .
Protocol 3 . Preparation of ribosom al salt wash IRSWj from detergent -free polvsornes-
Equipmen t and reagents e Pclvsomes isolat ed as in Protocol 2 , steps 1-16
Protocol 4. Prepa ration of crude lysate by homogenization with lysolecithin'
Reagents - Non-confluent, adherent tissue culture cells in 100 mm diam . tissue culture dishes ; suspension cells can perhaps be used by substituting low-speed centri fugation and gentle resuspension for aspiration (see step 4 below ) e Solution C lO.15 M sucrose , 33 mM NH 4Cl , 7 mM KCI, 4 .5 mM magnesium acetate , 30 mM Hepes -KOH , pH 7.41 - Solution D t 100 -300 J,lg /mllysolecithin (L-o-l vsophosphati dvlc hotin e, palm itoyl) in so lution C lb
e 4 .0 M potassium acetate
e Solutions A and B, and ultracentrifuge equipment as in Protocol 2
e Sotution E (0. 2 M NH 4CI, 20 mM magnes ium acetate , 7 mM KCI, 1 mM DTT, 1 mM ATP (dipot assium salt ), 1 mM GTP (sodium salt ), 40 J,lM each of the common 20 amino acids , 0.1 mM S -adenosyl meth ionine, 1 mM spermidine , 10 mM creatine phosphate (dipotassium sem. 40 units per m l of creatine kinase , 100 units of placental RNase inhi bit o r, 0.1 M Hepes -KOH, pH 7.41c
Method 1 . Aspirate the medium from each dish of cells .
Me thod 1. Resuspend the polysomes in solution A . The polysomes need not be full y dispersed , as described in step 17 of Pro tocol 2 , but they should be m or e concentrated . A reasonable starting point is 2-5 x 10 9 cells' polysome s per rnlllilitre , but higher co nce ntrations can be used , if desired. 2. Transfer the polysomes to a small test tube or beaker containing a stirring bar , in the cold room . Prior to use , the stirring bar should be soaked for 3 -5 min in DEPC and rinsed thoroughly with water . 3. Add sufficient 4 .0 M potassium acetate drcpwise and with gentle stirring to bring the final concentration to 0 .5 M .
2. Place the dish on ice and quickly wash the cells twice with 2-3 ml ice cold solution C , removing the liquid by aspiration . 3. Add 1-2 ml of solution D to t he dish and leave fo r 60 sec on ice . Be sure that the entire surface of the dish is covered with liquid. 4. Aspirate the solution , and gently invert the dish for 60 sec to drain the remaining liquid to one corner . 5 . Aspirate off the remaining so lution D. 6 . Put t he dish back on ice and add 200 JJI of ice -cold solution E. 7. Scrape the cells from the dish with a standard rubber-tipped cell scraper and transfer the cell suspension to a microcentrifuge tube.
Continued 116
Continued 117
Jeff
4: mRNA turnover in ce ll-free ex tracts Protocol 5 . Continued
Protocol 4 . Continued 8 . Disrupt the cells by passing them 10 tim es th roug h a 25 -gauge needle attached to anappropriately-sized hypodermic sy ringe . Monitor cell lysis by m icroscopy, at least forthe first few times the extrac tion is performed . If aspiration through a needle appears to be t oo har sh (i.e . if the nuclei are disrupted)cells can also be lysed by asp irating them repeatedly w ith
2. M ix th e tu be con tents th orough ly and t rans fer t he tube t o a 20 ° C w ate r
bath. 3 . Incu bat e f or 15 m in .
4 . Add 2 1 ~ I of 0 .1 M EGTA and
9. Centrifuge thecell lysate in a microcentrifuge in the co ld room for 1 min to pellet the nucl ei. 10 . Remo ve t he supernat ant, taking care not to con taminate it with pelleted material. 1 1. St ore the supernatant on ice in preparation for in vitro mRNA decay reactions. It maybe possible to store the lysate in aliquots at - 70 °C but the effects of freezing andthawing on the t ranslat ional or mA NA turnover activity of theseext racts have not been repo rte d . Therefor e, preliminary studies mig ht be required t o determine whether freezing is f easible. From ref. 5.
The concentration of lysolecit hin t o be used depend s on the type of cells and m ust be determined empirically, since excess lysoleci thin w ill reduce t ranslat ional activity, wh ile insuffici ent lysolecithin will not perm eabili ze t he cells . Brown et al . (16 1 used 100 IJg/m l f or PK cells, a pig kidney tine, and 250 IJg/m l for BHK cells, a baby hamster kidney line . Krikorian and Read (5 ) used 300 IJg/m l for Hela cell s. Tr y di fferent conce ntrations and compare the capacity of each extract to translate endogenous mRNA , using the method described in Sec tion 5.2 C The op timum concentrations of magnesium , potassium , and ammoni um ions f or translation or mRNAturnover seem to vary consi de rably depending on the so urce of the cell extract and soshould be determined f or eac h cell line (discussed in ref . 16 ). T he concen tr ations described here have been op t imi zed for Hel a cell ext racts ( 5 ). Op timum co ncentrations for mANA t urnover m igh t di ffe r from those for trans latio n (5).
b
Protocol 5 . Treatment of lysolecit hin e xtract w ith micr ococcal nuclease"
Rea gents e Crude
lysol ecit hin extract (prepared as in Protoco l 4 )
e M icroc occal nuclease (1 mglml in distilled w at er); st ore at - 20 " C: t he specif ic act ivity should be atleast 6000 units per m illigram of protein
e O. l M CaCI2 e O. l M EGTA dissolved in w at er and neu tra lized with KOH e Calf liver tANA lBoehrin ger Mannheim , Sigma ; 20 mg /ml in w at er)
Method 1. To a reaction tube in ice add the following : • lysole cith in extract
e O. l M csc i, • micrococc al nuclease
1O~ 1
of calf li ver tRNA and mix
thoro ug hly.
a Pasteur pipette.
~
Ross
5 . This ex t ract can now be use d f or in vitro m ANA turn o v er re action s. It m ight be possi ble to fr eeze the ext ract in liquid nit rogen, bu t pre lim in ary ex perime nts shou ld be perfor me d t o determine the effe ct of fr eezin g and t haw ing on translat ion and mRNA turno ver. ~
From ref. 5 .
5. Methods for performing in vitro mRNA decay reactions 5.1 Detergent-free extracts Three variables, the source of messenger RNa se activit y, the combination of subcellular fractions, and possible modifications of the basic system, need to be considered . (a) Reaction mixtures must contain either crude post-nuclear supernatant (510), po lysomes or R5W as a source of messenger RNase activity. Deproteinized, exogenous mRNA can be included with any of these, as desired. (b) Various combinations of fractions, for example, po lysomes or R5W plus 5 DO, with or witho ut exogenous mRNA substrate, can be used . The ability to determine how various fractions influence mRNA turnover is essential for identifying and characterizing mRNa ses and regulatory factors (3, 6, 7). (c) With regard to possible modifications, the conditions described below should not be considered rigid (see Section 4). For example, the author has identified a mRNA destabilizer activity in 5130 from cells infected with Herpes simplex viru s type 1 (6). The destabilizer was relati vely inacti ve under the standard reaction conditions described below but was fully acti ve in reactions co ntaining approximately one -third lower potassium and ma gnesium ion concentrations. Therefore, the investigator should be prepared to modify th e standa rd system as necessar y.
0. 98 ml
10 ~I
5.1.1 Using endogenous mRNA as subs trate
10
Th e sta nda rd system for this assay is described in Protocol 6 using polysomes as the so urce of endogenous mRNA.
~,
Continued
118
119
Jeff Ross
4: mRNA turnover in cell -free extracts
Prot ocol 6. Turnover of endogenous polysomal mRNA in detergentfree ex t rac ts"
Reagents e Soluti on F: prepare th is as fo llows and store it in sma ll aliquots at - 70 ° C :
Prot ocol 6 . Continued 3 . Remove the t ubes and either store them at - 70 " C or extract the RNA immediat ely f or analysis . A ny RNA ext ractio n method can be used, as long as care is t aken t o avoid non -specific RNase contamination. RNA extraction methods are given in Volume I, Chapter 1. , From refs 13 , 14.
Final conce ntrat ion in 25 reaction mixtures
~I
Comp onent
water
200
1.0 M c reatine phosphate
200
10mM
40
2mM
lMDlT 0 .2 M AlP (Tris salt)
100
1 mM
0 .2 M GTP (Tris salt!
20
0 .2mM
2.0 M potassium acetate
~l
As mentioned in Sectio ns 4 and 5. 1, various combinations of subcellular fractions can be assayed in this system, which is one of its important features. The components of the reaction mixture given in Protocol 6 (step I) can be modified accordingly. For example:
0 .1 M
1000
2 .5 mM spermine
800
0.1 mM
1.0 M Tria-acetate . pH 7.5
200
10 mM
• Solution A and polysomes at a con centration of 10 15 cells ' polvsornes/ut prepared as described in Protocol 2
solution F must be added be fore th e RNase inhibitor; the vo lumes gi ven are for each 25 IJI reaction requ ired in step 1:
• Creatine koase (4 mg /ml in 50 % glyce rol)
solution F
3.2
~I
• 0.1 M magnesium acetate
4 mg /ml creatine kinase
0 .25
~I
• Placenta l RNase inhibit or (40 un its!1..l11
RNase inhibitor
0 .25
~I
• Soluti on G: prepare this so lution on ic e just pr ior to performing the reaction ; the author does not know if the order of addit ion of components is important, but
so lution A
16 ~ I
0 .1 M magnesium acetate
0 .2
~I
water
1. 1
~I
Method 1. Add the following to the required number of microcentrifuge tubes on ice:
• solution G
21 ~ I
• polysomes
4 ~l
2 . Mix gently without vortexing and incubat e the reactions at an appropriate temperatu re, usually 37°C, for the required time . No single reaction temperature can be consid ered optimal f or all mA NA decay react ions. Therefore, whi le sta nda rd reactions are usua lly incubated at 37 ° e . situations may arise in which it may be advantageous to incubate at lower temperatures . For example, cell -free mANA turnover reactions are particularly use ful for identi f ying m RNA decay products . The nature of these products provides a pic tu re of the dec ay pathway : its direction (5' ..... 3' or 3 ' - 5 ' ) and the sorts of enzymes involved (exo- or endo ribonucleases) . How ever, decay products of some mRNAs might be too unstable to be detected even in vitro at 37 " C. If so, try incubating the reactions at a temperature between 20 " C and 32°C , rather than 37 ° C. Reaction rates are slower at lower temperatu res, permitting easier detection of decay products (17) .
(a) Some trans-acting factors that regulate mRNA stability are located in the S130 fraction and can be identified by mixingS130 and polysomes. Simply substitute SI 30 for an equivalent volume of solution A when making up solution G. To assay the effect of 10 ~I of SI30 on the decay of polysomal mR NP , prepare solution G with 6 ~ I per reaction of solution A, instead of 16 ~I , and set up reactions conta ining I I ~ I of solution G, 10 ~I of S130, and 4 ~I of polysomes. (b) To analyse mRNP decay in SIO, simply substitute SIO for polysornes and solut ion A . Care sho uld be taken to maintain the magnesium ion concentration at 2 mM. 5.1.2 Using exogenous RNA substrates A system for assaying the turnover of exogenous RNA in detergent-free extracts is described in Pro tocol 7. The method is similar to that described in Protocol 6, but the substrate is prote in-free (pheno l-extracte d) RNA, not endogenous polysomal mRNP. The two major sources of substrate are cellular mRNA or RNA synthesized in vitro. Cellular RNA should be prepared from a cytopl asmic extract and can be pre-purified by affinity chro matography to select poly(A)+ mRNA. If synthe tic mRNA is used as a substrate, some important variables must be considered . Should the mRNA be synthesized in the presence of a cap analogue (m'GpppG)? Should it be polyadenylated? Should it be radiolabelled , and, if so, to what specific activity? As a general rule, it seems reaso nable initially to use a substra te as close as possible to authentic, i.e . both capped and polyadenylated . Protocol 7 . Turnover of exogenous mRNA substrates in detergentf ree ex t racts-
Reagents • Solution A and polysomes 110 6 cells ' polvsomes/ul as descr ibed in Pro -
tocol 2
e Creatme kinase . RNase inhibit or, solution F. magnesium acetate as described in Protocol 6
Continued
Continued 120
121
4: mRNA turnover in cell-free extroct s
Jeff Ross
Protocol 7 . Continued ( - 10 7 c .p .m ./1J9 . 5000-20000 c .p.m.r
f ormula be low is f or eac h 25 JJI react ion requ ired in step 1:
1-11
so lution F
• Either a J2P· labelled RNA
substrate
in water ) or unlabe lled RNA 10,1-2 .0 1-19 /jJl in water; see V ol um e I. Chapter 1 (Pro tocols 3 and 4 ) and Chapter 4 , th is volume (Prot oc ol 4 1
4 mg /m l creati ne k ina se
RNase inhi bito r solution A
• Solution H : p repare th is on ice ; the
3.2 " I 0 .25 " I 0 .25"1
Rega rdless of the so urce of enzyme, it is important to ensure that the relative d~ca y rates observed with endogenous substra tes ar e maint ain ed in react ion s with exogenou s substra tes. As discussed in Section 2.2, exogenous substrates might faJ! to rellec t mRNA decay pro cesses as accurately as their polysomeassociated co unterpa rts .
16 "I
0 . 1 M m agnesium ace tate
0 .2 " I
RNA subst rate
1.1 " I
Protocol 8. Turnover of exogenous mRNA substrates usi ng ribosomal salt wash (RSWj a
Reagen ts Method 1. Ad d the follow ing t o t he requi red number of microcentrifug e tubes on ice:
e solution H
2 1 "I
• po lysomes
4 IJI
2. Incubate the react ions at 37 °C for t he desired time. 3. Eit her sto re the react ion mixtures at immediately (see Protocol 6 , step 3 1.
- 70 ° C or ex tract the RNA
• Creatine kinase, 0. ' M ma gn esiu m ace ta te , RNase inhi bit o r, sol ution A as in Pr otocol 6 • RNA substrate as descr ibed in Protocol 7 • RSW (see Pro tocol 3) • Solut ion I: prepa re t his as follows and st ore at - 20 °C in multiple aliqvots : water 1.2 ml 1.0 M creatine phosphate
" From refs 13, 14 .
1.0 M OTT 0 .2 M A TP !T ris salt I
Th e reaction mixtur es described in Protocol 7 can be mo dif ied to include SIO, S130, or combinations thereof. When doing so, th e com position of solution H must be changed, decrea sing the volum e of solution A to accommodate that o f the add ed subcellula r fraction . If it is necessary to add more th an 2"g o f RNA substrate per 25"' reactio n, preliminary assays should be performed at the higher substrate levels, since excess RNA non- specifically reta rds in vitro mRNA degrad ati on and , at sufficient ly high levels, sto ps it completely. A system fo r assaying turnover of exogeno us mR NA substrates using RSW is described in Protocol 8. This involves a minor mod ification of Protocols 6 a nd 7, prima rily to accou nt fo r the high salt in the RSW. Since some messenger RNases are a ppare ntly inactive at salt concent rations above 0.3 M and lose activi ty ab ove 0.2 M (14), care should be take n to keep the salt concentration below 0. 2 M an d prefer ably at 0.1 1>1 or lower. It is unclear whether differential mRNA sta bility with exogenou s lORNA substr ates is better st udied using polysome s or RSW as a sour ce of messenger RNa se activity. RSW has two potential advantages. • Whereas polysom es contain an activity which rap idly degrades un capped substrates, RSW does not (18). e RSW contai ns solub le RNases and thus can serve as starting material for the purificat ion of messenger RNases. 122
0 .2 M GTP (T ris salt)
200"1 40"1 100 "I 20"1
2 . 5 m M spermine
8 00 IJI
1.0 M Tris- acet at e, pH 7.5
20 0 IJI
• Solu tion J: prep are t his on ice just bef or e perfo rm ing the reac t ion s; the vo lu mes giv en below are fo r each 2 5 1J1 react ion requi red in step 1: solution I 4 mg lml creatine kinase RNas e inhi bi to r /) so lu tion A
3. 2 " I 0 .2 5 "I 0. 2 5 "1
0 . 1 M ma gnesium aceta te
15 " I 0 .2 " I
RNA subs tra te
1.1 " I
Method 1. Add the fo llo wing to the requ ired number of microcentrifuge tubes on ice : e solution J 20 ~I
e RSW
5"1
2 . Incubat e th e react ion s at th e desired temperature f or the desired ti m e lsee
Protocol 6 1.
3 . Either store t he react ion m ixtures at im medi ately (see Protoc ol 6 . st ep 3 ).
- 70 °C or extract t he RNA
" From ref . 14 .
01
.
t ~lght be necessary to inc rease t he amount of RNase inhi bitor, If so , the amount of solu tion A should be decreased to ma intain the 2 5 1-11 reaction volu me .
5,2 Lysolecithin extracts As discussed i~ Section 4, lORNA
is translated more efficiently in lysolecithin extracts than In extracts prepared by lysing cells in detergent-free buffer. In ord er to exploit th is prop ert y of the system, it might be useful to perform 123
Jeff Ross
4: mRNA tu rn over in ce ll-free extracts
the following preliminary experiments to determine optimal translation
Prot ocol 10 . Turnover of exogenous mRNA in lysolecithi n ex t rac ts
parameters:
Reagents
(a)
Incubate the extract at 37 °C for different times (e.g. between 30 min and 4 h) in the presence of a radioactive amino acid (see Pro t.ocol l I v. Determine the kinetics of incorporation of amino ac~ds m~o prot~m by a~l d precipitation. This experiment will determine the optimal time of incubatio n for mRNA decay assays.
(b) If exogenous mRNA is to be assayed in nuclease-treated extract (see Protocol 10), add I- lO~g per reaction of mR NA and measure theammo
acid incorp oration. This experiment is import ant both to deter mme the opti mal mRNA level for translation and to ensure ~ hat reacuo ns do not contain excess mRNA, which might slow translation or mRNA decay rates, or both. (c) Having determined the optimal time and mRNA level, compare transl~tio n rates at 25, 30, and 37°C, to optimize the temperatur e for translation. 5.2.1 Using endogenous RNA as substrate Anal ysis of the turnover of endogeno us mRN A in lysolecithin extracts is described in Prot ocol 9.
Prot ocol 9. Turnover of endogenous mRNA in lysolecith in e xt ract s'
Reagen ts • Post-nuclear supernatant (crude Iysate l prepared as in Protocol 4
Method 1. Pipette the required amount of ext rac t 1100-200 ~Il into microcentrifuge tu bes on ice. 2. Incubate the reactio ns in a wate r bath at the desired t em perat ure and fo r
• Lysolecithin extract; either untreated (see Protoco l 4 ) or mi crococcal nucleasetreated (see Protoco l 5 )
• Exogenous mRNA substrate in water (1- 5 1-19 IIJII. see Protoco l 7
Method 9 9 - 19 8 ~ 1 of extract and 1-2~1 of m RNA subs tra te into microcentrifuge tubes on ice. As discussed in Section 5.2, the amount of m RNA per reac tion wi ll depend on t w o f act ors, the capacity of the system t o tr anslat e a given quanti ty of m RNA and th e pos sibility t hat excess exogenous RNA m ight slow tr anslation or m RNA decay. Both factors should be investigated prior to performing reac tions.
1. Pipette
2 . Incubate the mixture at an appropriate t em peratu re f or the desi red time . 3 . Eith er store the react ion mixtures at - 70° C or ex tract th e RNA im me diately (see Protocol 6 . step 31.
As discussed in Section 5.1 with reference to the detergent-free extracts, reactions in lysolecith in extrac ts as described in Prot ocols 9 and 10 may also be modified in various ways. For example : (a) The role of trans lation in mRNA turnover is well-docum ented but poorly understood (I ). Therefore, the addition of reticulocyte lysate might be useful if it stimulates or prolongs translation of the mRNA und er study. (b) The association of immunoglobulin mRNA with membra nes might affect its half-life during B-cell differentiation (19, 20). Therefore, adding membranes to the system might be necessary to identify regulator y factors for certain mRNAs. The se possible effects can be tested by add ing rabb it reticulocyte lysate and canin e pancreatic membranes, both of which are commercially ava ilable (21). Reaction mixtur es (see Protocols 9 and 10 ) would then include 200 ~I of lysolecithin extract , with or without exogenous mRNA, 20 ~I of reticulocyte lysate, and 4 ~I of membranes .
an appropriate time . 3 . Eit her store the reac tion mixtures at - 70 °C or extract the RNA immediately (see Protocol 6, step 3) .
5.3 Reticulocyte translation extr acts
5.2.2 Using exogenous RNA substrates The use of exogenous RNA substra tes with lysolecithin extracts is described in Proto col 10. See Section 5.1.2 for a description of alternative substrates.
An alysis of exogenous mRNA turn over in reticulocyte lysate is described in Pro tocol I I . Rabbit reticulocyte extract treated with micrococcal nuclease (obtained from a suitable commercial source) is incubated with phenol-extracted RNA from cells or from in vitro transcriptio ns. As discussed in Sections 4 and 5.2 modifications of the reticulocyte system should be considered for particular applications. For examp le, S130, polysomes, and/or membranes can be added .
124
12 5
• From ref . 5.
Jeff Ross
4: mRNA turnover in cell-free extracts Protocol 11. Turn over of e xog en ous mRNA in a n mRNA·de penden t rab bit retic uloc yte translatio n svsrerrv
Reagents • Comme rc ial rabbit reticulocyte lysate pre treated with mic rococc al nu clease (A m ersh am, Boehringer Mannheim , DuPont. Gibeo BRL. Promegal
e mRNA dissolved in water at 0 .1 -5 .0 IJg l 1-1', either from cells or from in vitro transcr iptions . If using cellular RNA . it m ight be necessary to prepa re poly(A )* mRNA . In vitro transcription of RNA is des cr ibed in Vo lume I. Chapter 1 (Pro toc ols 3 and 4) and Chapter 2 , this volume.
• Place ntal RNase in hibito r. 4 0 uni t s/pl (see Proto col 6) • 1 mM amin o acid m ixture; various am ino acid m ixtures are ava ilab le from the c om panies that sell the lysat e; som e are complete ; others lack on e o f the 20 am ino acids . so that it can be rep laced by it s rad ioa ct ive ana logue in order t o measure translation eff ic iency
ampli ficat ion method ology might be required. Rad iolab elled mRNA substrates are the simplest to analy se, requiring only electrophoresis in a denaturing polyacrylamide gel (see Volume I, Chapter I , Protocol 8 and Volum e I, Chapter 4, Protoc ol ll). Three examples of mRNA deca y experiment s from the author's laboratory are presented below to illustrate these and other point s. 6.1.1 Relative d eca y rates of e n d og enous, polysome-a sso ciated mRNPs in ce ll- free mRNA deca y rea ctions The experiment shown in Figure J was performed to compare the in vitro decay rates of three endogenous mRNAs. The experiment involved the following:
• Prepa ration of detergent-free polysomes from tissue culture cells (human erythroleukaemia) as per Proto col 2. • Incubation of the polysome s for the indicat ed times essentiall y as described in Protocol 6.
(Protocol 4 )
• Isolation of total RNA from each reaction by phenol extraction .
Method Add th e reagent s in t he order list ed below t o microcentri fuge tubes in an ice bat h. Th e final reac tion volume is 501-l\' • nuclease-t reated lys at e
• An nealing of a portion of the extracted RNA separately to each of three 31P· labelled DNA hybridization probes for human "'(-globin, c\-globin, or c-myc mRNAs. • Treatment of each hybridization reaction with nuclease 51 and electroph oresis of the nuclease-re sistant DNA in denaturing polyacrylamide/urea gels.
• RNase inhi bitor • amino acid mixt ure • radi oacti ve amino acid and /or wa te r • mRNA subst rate in w ate r 2 . Incubate at 30 °C for the required tim e.
as nec essary as nec essary
• Autoradiography of the gels and quantification of the amount of each DNA fragment using a laser densitometer. The quantity of nuclease-resistant DNA indicates the amount of each mRNA remaining at each time po int.
3 . Either st ore th e sam ples at - 70 ° C or extract th e RNA im mediately (see
Protocol 6 , ste p 3 1. ~
A.
1' 100 ~
From refs 22. 23 .
~---_ .. _--- o------
0
Q)
E 75 •
.;=
6. mRNA detection, data interpretation, and troubleshooting 6.1 mRNA detection and some typical in vitro mRNA decay experiments
•
~ 50 ~
Z 25
cr E
100
"<.. 200
10 0
200
300
Man y method s a re availab le for ana lysing and qua ntifying endogenous mRNA levels, including Northern blottin g, nuclease 5 I mapping and RNase prote ction, primer extension ana lysis, and amplification by PCR. The method of choice will depend in large measure on the quantity per cell of the particular mRNA. Blotting and nuclease protection, which are described in Volume I, Chapters 2, 3, and 4 and Chapter I, this volume , are preferred for moderately abundant mRNAs (at least 100 molecules per cell). For very scarce mRNAs , PCR
Figure 1. Co m pariso n of in vitro st abili ty o f th ree endogeno us , polysome- associ ated mRNA s. Polvsorne s isol ate d fr o m ery t hrol euk aemia ti ssu e cul t ure cell s were incu bated as de scr ib ed in the t ext , and t he stabilities of three mR NA s were co m pa red by nuclease S 1 pr ote c t io n and qu an t if ic at ion of the autoradiograms. Panel A ; l'-g lo bi n mR NA (un filled ci rcl es) and 6-glo bi n m RNA (filled circ les). Panel B; l' -glo bin mRNA (unfilled circ les) and c-m yc mRNA (fill ed triang les). Reprinted with perm ission f rom ref. 13.
126
127
Time (min)
Jeff Ross
4: mRNA turnover in cell-free ex tracts
deadenyla ted (i- Glo bin
(i- Globin time (min)
A ( -I ~ -
- PABP 1
• Preparation of three 32P-labelled substrates, polyadenylated human ~ -globi n mR NA, deadenylated huma n ~- gl o b in mRNA, and huma n H4 histo ne mRNA, by transcribing appropriate cDNA clones with SP6 RNA polymerase. H isto ne mRNA is one of the few mammalian mRNAs without a po ly(A) trac t , lt served as a control in some reactions.
. PABP
- PABP
· PABP
• Preparation o f RSW fro m tissue culture cells (human eryt hro leukaemia) as per Protocol 3 and then depletion of the RSW of endogenous PARP.
0 20 40" 0 20 40" 0 20 40 " 0 20 40 I
-
-
• Incubation of substrates with RSW as per Protocol 8 for the times indicated in Figure 2. Some reactions also contained highly purified PARP from yeast. • Extraction of total RNA from each reaction and electrophoresis of each RNA in a denaturing polyacrylamide/urea gel.
439 -
-
H
370 -
2 3 4
5 6
7 8 9 10 11 12 13
Figure 2. Comparisonof the stabilities of exogenous polya~enylated and deadenyl,ated mRNAs in RSW either lacking or containing poly(A)-binding protein (P~B Pl: ASW was first depl ete~ of PABP . It was then incubated as described in Protocol 8 With either of thr~e substrat es. polyadenylated p-globin mRNA. f!; deadenylated a-globin mANA . A( -:-H~ ; o~ ~lstone mR NA . H, which rcrmaltv lacks polylAI . Some reactions were supplemented With pun~led yeast PABP: Lane 1. DNA size markers (sizes of two fr agm ent s are given in base pairs . on the lett ): lanes 2 -7 , polyadenylated a-globin , P, and deadenylated histone . H; mRNAs Jnc~~ated for the times indicat ed with PABP·depleted RSW or with PAB P-depleted RSW plus PUrlfled ~east PABP ; lanes 8-13 , deadenylated p-globin mRNA , A(
-lS incubated as per lanes
2-7 . Rep nnted
with permission from ref. 24 .
Figure I shows the decay of -y-globin mRNA, whose intracellular hal f-life is approximately 17 h , Ii-globin mRNA (intracellular hal f-life 4 h) and c-myc mRNA with an intr acellula r half-life of 0. 5 h. This expenrn ent Illustrates an essential property of this in vitro mRNA decay system: it reflects known intracellular mRNA ha lf-lives.
• Autoradiograp hy of the gel.
Figure 2 shows the rapid decay of polyadenylated globin mRNA and of histone mRNA in the PA RP-depleted RSW. In contrast , ~-globin mRNA is at least 10-fold more stable tha n histo ne mRNA in reactions supplemented with purified PA RP . Therefore, PA BP contributes to the long half-life of polyadenylated ~-globin mRNA in these reactions. However, PARP does not stabi lize deadenylated ~-globin mRNA . T herefore, PABP probably helps to stabilize mRNA by interacting with the 3' poly(A) tract. 6 .1 .3 Effect o f a tra ns-acti ng mRNA stability regulato ry fa ct or in the post-polysomal supernatant (51 30) The experiment of Figure 3 was performed to determine whether the destabilization of -y-globin and other host cell mRNAs following Herpes simplex virus infection could be reproduced in vitro and, if so, in what subcellular fraction the destabilizing factor resided . The following experiment was performed: • Two sets of tissue culture cells were used. One was infected with Herpes simplex virus, the other was 'mock infected' , meaning that it was treated in the same way but no virus was added . . At 60, 180, and 360 min following virus inoculation (mixing virus and cells), cells were harvested and post-polysomal supernatant (S 130) was prepared, as per Protocol 2.
(a) To compare the hal f-lives of two radiolabelled globin mRNA substrates. one with an 80 nt 3 ' poly(A) tract, the other lacking poly(A) . (b) To determine whether po ly(A)-binding protein (PABP) influences th: in vitro decay rate of either substrate. The expenment involved th following:
. In vitro mRNA decay reactions were performed by mixing polysomes from uninfected cells with S130 from either mock-infected or virus-infected cells, as per Protocol 6. • Each reaction was incubated for 0 to 150 min; then total RNA was isolated and annealed with a 32P-labelled DNA probe for human -y-globin mRNA. The amount of undegraded -y-globin mRNA in each reaction was determined by incubation with nuclease SI and electrophoresis of nuclease-resistant DNA in a gel.
128
129
6 .1.2 Radiolabelled substrate incubated with ribosomal salt wash The experiment of Figure 2 was performed for two reasons:
4: mRNA turnover in ce ll-free ex trac ts
Mock I
o
I 15 30 90 150
r in vitro
< Incubation
~ Time (mi n)
Y-Globin -
Wild-type } Post -Inoculation
Time (min)
60 r --"-''----, I
a
180
360
in vitro I I I I I 15 30 90 150 a 15 30 90 150 0 15 30 90 150 { Incubat ion Time (mi n)
Y·Globi n-
Figure 3 . Ident if icat ion of an mRNA destabilizing factor in the S 130 of cells infected wi th Herpes simplex v irus . T issu e cultu re cells were eithe r mock -infected or infected with He rpe s simplex virus for up to 6 h. as desc ribed in the text. S 130 was prepa red fr o m each cultu re flask and was added to polvscrnes from uninfected cells . The reactions were incubat ed as per Protocol 6 . The stability of po lysome -associated )' -globin m ANA was determ ined by nuclease 51 mapping . T o p panel , polysomes plus S, 30 from mock -infected cells, i.e . c ells not exposed to v irus. Bottom panel, polysomes plus S 130 from cells infected for the ind icated times (wild -type, post-inoculation time lminl] with Herpes simplex vi rus . The t ime co urse in min of each incubation is shown lin vitro in c u bat ion time (min)) above each lane . Reprinted with permission from ref. 6.
Figure 3 shows tha t polysomal, endogenous "'( -globin mRNA is very stable in reactions supplemented with 5130 from mock -infected cells (no virus) but that 5 130 from cells infected with the virus for 60, 180, or 360 min contains a potent mRNA destabilizer. This destabilizer is now known to be a protein encoded by the virus and an integral part of the virion itself. 8y an unknown mechanism, this protein accelerates the degradation of almost all cellular mRNAs, even those like "'( -globin which are normally very stable in unin fected cells.
6.2 Data interpretation As discussed in Section 1 an d Protoco/ 6, mRNA decay rates are slower in vitro than in intact cells. Relatively slow reaction rates are also observed in cell-free 130
Jeff Ross tran scripti on and translation systems. Therefor e, it seems reason able to exploit cell-free mRNA decay systems to measure relative, as opposed to absolute, decay rate s. To assess whether the system is functio ning properl y, the decay rat es of mRNAs which a re kno wn to ha ve different int racellular hal f-lives sho uld be determined. mRNAs that are un stabl e in cells sho uld also be unstable in vitro relative 10 sta ble mR NAs. ' Some ca ution mu st be exercised in plottin g an d interpret ing in vitro mR NA decay da ta . In cells, mRNA is degraded in a logar ithm ic fash ion, accord ing to the fo rmula:
where C a nd Co are mRNA levels at time t a nd tim e 0, respecti vely, and k d is the decay consta nt (reviewed in ref. 25). The probl em with apply ing simila r par ameters to in vitro dat a is that decay rates may decrease as the rea ction tim e increases. For example, the author has found that the c-myc mRNA decay rate decrea ses at least two-fold after several hours of incubation, presumably because critical reaction components becom e rate-limiting (J. Ross and G. Brewer, unpublished observations). Therefore, It seems reasonable to plot the in vitro data as either a linear or logarithmic function, provided that each mRNA is a nalysed by the sa me criteria (see Figure I as an example) .
6. 3 Troubleshooting Several potential obstacles and trouble spots which might arise during the course of cell-free mRNA decay experiments are listed below, as well as some suggestion s for solving them. 6 .3 .1 Excessive RNA degradation during preparation of polysomes and extracts
Pre~aut ions for redu cing endogenou s, non -speci fic RNase activity while pre panng polysomes and duri ng in vitro incub ation s are discussed in Prot ocol 2. If the se are inadeq uate, it might be helpful to avoi d using primar y anima l cells and to use instead a tissue culture cell line synthesizing the mRNA(s) of interest. Some cells lines contain less ' non-specific' nuclease activity than ot hers. Tr y screening several cell lines to find o ne which allows intact polysomes to be isolated . 6.3.2 All mRNAs in the mRNA decay reactions are degraded at similar rates Again , non -specific RNase activity might be ove rwhelming the system . Several modifica tions sho uld be tried : • Add more RNa se inhi bitor or try various combinations of inhibitors. 131
4: m RNA turn over in ce ll-free extra cts • Modify the reaction conditions. For example, non -specific RNa se activity might be depressed by changing the monovalent o r divalent cation concentrations, the pH, or the reaction temperature. • Add carrier single-stra nded RNA , for example, pol y(C) at 1-2 ~g/25 ~I reaction (see Protocols 6-8). Avoid higher amounts of exogenou s RNA , which can slow mRNA decay processes. 6 .3.3 No
mRNA degradation is observed
The simplest explanation for thi s result would be inacti vation of messenger RNases (and other RNases) during extract preparation. If no RNA is degraded during in vitro reactions, the composition and pH of all extraction buffers should be monitored carefully. An unlikely explanation is that all RNase activity in the extract is blocked by the placental RNase inhibitor. Th erefore, reac tions should be carried out without inhibitor.
6.3.4 Regulated decay is no t observed For example, a particular mRNA might be known to be degraded five-fold less rapidly in cells treated with a translational inhibitor, as compared with untreated cells. However, the mRNA is found to be degraded in vitro at the same rate with polysomes isolated from treated and untreated cells. There could be many explanations for this result, but the following can be investigated : • Add Sl30 (Protocol 2) to the reactions. Perhaps the regulatory factor is in the Sl30 and is not polysome-associated.
Jeff Ross 4. 5. 6. 7. 8. 9. 10. II.
12. 13. 14. 15. 16. 17. 18. t9. 20. 21. 22. 23. 24. 25.
Wager, R. E. and Assoian, R. K. (1990). Mol. Cell. Bioi.. 10, 5893. Krikorian. C. R. and Read, G. S. (1990). J. Viral., 65, 112. Sorenson, C. M ., Hart, P. A., and Ross, J . (1991). Nuel. Acids Res. , 19,4459. Brewer, G. (1991). Mol. Cell. Bioi.. II , 2460. Blume, 1. E. and Shapiro, D. J. (t989). Nuel . Acids Res.. 17, 9003 . Graves, R. A., Pandey, N. B., Chodchoy, N., and Mar zlu ff', W. F. (1 987). Cell, 48,615 . Capasso, 0., Bleecker, G. c.. and Heintz, N. (1987). EMBO J., 6, 1825. Blackburn, P. and Jailkhani, B. L. (1979). J. Bioi. Chem., 254,12488. Berger, S. L. and Birkenmeier, C. S. (t979). Biochemistry, 18, 5143 . Ross, 1. and Kobs, G. (1986) . J. Mol. Bioi., 188,579. Ross, J., Kobs, G., Brewer, G. , and Peltz, S. W. (1987). J. Bioi. Chern., 262 , 9374. Bradford, M. (1976). A nal. Bio chem. , 72, 248. Brown, G. D., Peluso, R. W., Moyer, S. A., and Moyer, R. W. (1983). J. Bioi. Chem., 258, 14309. Peltz, S. W. and Ross, J. (1 987). M ol. Cell. Bioi. , 7, 4345. Peltz, S. W.. Brewer, G ., Kobs, G ., and Ross, J. (1987). J. Bioi. Chem ., 262, 9382. Mason, J. 0 ., Williams, G. T ., and Neuberger, M. S. (1988). Genes Develop., 2. 1003. Jack, H·M. and Wabl, M. (1988). EMBO J. , 7, 1041. Baker, S. C. and Lai, M. M. C. (1990). EMBO J. , 9, 41 73. Wreschner, D. H. and Rechavi, G. (1988). Eur. J. Btochem., 172, 333. Hepler, J. E., Van Wyk, J . J., and Lund, P. K. (1990). Endocrinology, 127, 1550. Bernstein, P., Peltz, S. W., and Ross, J. (1989). M ol. Cell. Bioi. , 9, 659. Hargrove, J . L., Hulsey, M. G.. and Beale, E. G. (1991) . BioEssays, 13,667 .
• Alter the reaction con ditions, as per Section 6.3 .2. • Prepare a lysolecithin extract (Protocols 4 and 9) rather than a detergentfree extract from both sets of cells, to encourage translation of the mR NA. since the regulation of some mRNAs might be linked to their translation. In this case , the mRNA might be unstable under conditions of effi cient translation elongation, but stabilized when elongation is retarded by an inhibitor. The in vitro system would then reflect the intracellular situ ation only by comparing the stability of elongating versus non-elongating mRNA. Some translation does occur in the lysolecithin extract but little is obser ved in detergent-free extract. To determine whether the act of translational elongation itself inlluences the stability of the mRNA, compare hal f-lives in lysolecithin extracts with or without added translational inhibi to r.
References I. Peltz, S. W., Brewer , G.. Bernstein , P., Hart, P. A., and Ross, J. (l99t).ln Critical reviews in eukaryotic gene expression (ed. G . S. Stein , J. L. Stein, and J . B. Lian),
Vol. 1, p. 99, CRC Press, Ann Arbor. 2. Belasco, J . G. and Higgins, C. F. (1988). Gene, 72, 15. 3. Brewer, G. and Ross, J. (1989). Mol. Cell. Bio/., 9, 2006. 132
133
Ribosomal RNA processing vertebrates
.
In
CATHLEEN ENRIGHT and BARBARA SOLLNER-WEBB
1. Introduction Ab out half of the transcription of an actively growing eukaryotic cell is devoted to the production of the large ribosomal RNA precursor (475 or - 14kbp in mammals) that gives rise to the 185, 5.85, and 285 mature rRNAs of the ribosome (see Figure I). Thus a major processing commitment of the euk aryotic cell is devoted to the maturation of rRNA. This rRNA processing occurs in the nucleolus and involves:
• cleavages at the 5' - and the 3' -ends of each of the mature rRNA segments • cleavages at a still undetermined number of sites within the tra nscribed spacer region s • numero us methylation eve nts
Due to the large cellular abundance of the mature rRNAs and a number of rRN A processi ng intermediates, the cleavages that give rise to the 5' - and the 3' -ends of the 18S and 285 rRNAs were able to be identified prior to the recombinant DNA revolution, making rRNA processing the first eukaryotic transcript maturation pathway in which the major stages were described . Despite thi s early progress, a great deal still remains to be learned about rRNA pro cessing. For a recent review of the field of eukaryotic rRNA processing, see ref. 1. This chapter focuses on rRNA processing in vertebrates and describes lab or ato ry techniques available for its experi mental investigation. Since most is known about the pri mary processing event that occurs within the 5' external transcribed spacer of mouse rRNA (5 ' ETS, see Figure I ), most attention will be given to: • re-creation of this processing event in vitro using synthetic rRNA substrates and cell extracts po ssessing processing activities
• ident ificatio n of the processing signa ls within the rRNA
5: Ribosomal RNA processing in vert ebrat es
Cothleen Enright ond Borbara Sollner-Webb
PROCESSING OF VERTEBRATE PRE·rRNA
5' ErS
ITS 1 ITS 2
3' ErS
•
•
18 S
5' ErS
5.8 S
T able 1 . Metazoan rRNA processing events currently accessible to study in systems amenable to experimental manipulation
47 S
Processing event
Specie s
Method of study
5 ' ETS cleavage
mOUSe human
mouse
in $-100 extract of mouse cells in $ - 10 0 extract of mo use cells in $ -100 extract of human or mouse cells using nucleolar endonuclease
human mouse mouse
in nucleolar extract of human cells in nucleolar extract of mouse cells by transfection into rat cells
12
mouse
using nucleolar endonu clease
14'
by following oocyte endogenous rRNA processing in mouse ce ll ex t racts
16 17
by oocyte injection
18
28 S
peocassmq
Xenopus
•
46 S
•
--
18 S
18 S 18S
\
• 41 S
5 ' -end of 18S region
3 '-end of 18S region
Clea vage at 5 '-end of 5 .85 region Xenopus
•
36 S
..
34S
32S
32S 20S 28S
""-
-
18S
• 32S
I •
11
13
Cle avage at alternative
• •
9 , 10 10 40
Cleavage at
45S
/
Reference
mouse term inat ion-associat ed 3 '-en d processing Xenopus .. T. Nilsen. personal communication. b But see reservation in ref. 15.
28S
5.8S
Figure 1. Diag rammatic representa tion of rRNA processing in verteb rates . The comp osite diagram draws on data from mouse, human , frog, and other ve rtebrate spec ies; the sedimentation coefficients of t he processing interm ediat es are show n for mouse rRNA . T w o alternative pathw ays are show n. The re may well be other mino r pathways and it is not certain w hethe r every rRNA precursor mu st pass through each of the processing steps in the order shown. 5 ' ETS , 3' ETS ; exte rnal transcribed spac er regi ons at the 5 '- and 3 '-ends of the rRNA
transcription unit. respectively. tTS 1 and ITS 2; internal transcribed spacers. The arrow head shows the position of the cleavage site in the 5' ETS referred to in the text. (Adapted from ref. 1; where further information can be tou nd.l
• analysis of the nucleopr ot ein complex that for ms on the processing signal • identi ficati on of the polypeptides th at bind specifically to this rRNA region
describ ed in Section 3.3 . Table I summarizes the major metazoan rRNA processing events whose study has thus far been reported . This chapter does not cover : • genetic techniques that are used to analyse rRNA processing in yeast (2-6) . the removal of the rRNA intro n that is peculiar to certain unicellular eukaryotes (see Chapter 7) • the cleavages that yield fragmented rRNA in certain species (described and reviewed in ref. 7) • processing of nucleolar sma ll nuclear RNAs (reviewed in ref. I) • ana lyses of rRNA methylation (reviewed in ref. 8)
2. Analysis of rRNA processin g in vivo
• identificatioo an d analysis o f the require ment for U3 snRNP in this pro cessing reaction
2,1 Introduction
Recent evidence indicates tha t such a 5 ' ETS proces sing complex is fou nd in man y if not all euka ryotes, Relat ively littl e is known abo ut most of the ot her co nserved eukaryo tic rRNA processing event s, with the exception of the processing of the 5 ' -end of the 18S rRNA region . Thi s 5 ' processing has been reproduced in vitro as a two step (cleavage and trimming) maturatio n, as
Clearly. the first step in analysing an RNA proce ssing event is to define the process in vivo. Significant data on rRNA processing have been derived through the identification of proce ssing intermediates and mature rRNA s by size and structural features. thereby allowing processing sites to be located within the pri ma ry transcript and processing steps to be temporall y ordered . The early
136
1 37
Cothlee n Enright and Barbara So llner-Web b
5: Ribosomal RNA processing in vertebrates analyses of processing involved fractionation of rRNA by velocity gradient ultracentrifugation and the electron microscopic analysis both of the secondary structures of rRNAs and of their heteroduplex for mation with clone d rRNA genes (see refs 19-2 1 for proce dures). The majo r vertebrate rR NA processing intermediates and mature rRNAs were thu s deduced to be cleavage fragment s of the primary transcript, without the discontinuities that would arise by splicing, RNA editing, etc. This conclusion has since been confirmed by nuclease SI mapping and nucleotide sequence analysis. These early stu dies also showed that there is a preferential but not obligatory order to the cleavage s at the ends of the mature rRNA regions (see Figure I) .
1
2
3
2.2 Analysis of rRNA processing by h ybrid selection and gel electrophoresis of cellular rRNAs Precursor RNA and proce ssing products are now generally analysed by gel electrophoresis rather than by ultracentrifugation. In a striking application o f this, the initial processing events in mou se rR NA matu rati on ha ve been or dered temp or ally by electr ophoretic sizing o f different populations of rRNA intermediates, isolated by hybrid selection (22). This approach is illustrated in Figure 2 where the data show that the cleavage within the mouse 5 ' ETS precedes the removal o f th e 3 ' ETS, which in turn occurs prior to th e cleavages at the 5 ' - and 3 ' -end s o f the 18 and 5.8S regions an d at the 5 ' -end of the 28S region (within the 45S rR NA ; see Figure I). Pro/acol J describes the procedure for la belling rRNA in vivo, the isolation of rR NA processing intermediates by hybrid selection and their sepa ration by a nalytical gel electrophoresis. An alternati ve approach would be to use unlabelled rRNA a nd then detect the rRNA intermediate s after electrophor esis by Northern blotting with appropriate rRNA probes. Proced ures for Northern blotting and pro be preparat ion are given in Volume I, Chapter 2 (Section 3.6), Volume I, Chapter 4 (Section 7.5), a nd ref. 23.
47 S .... 46S 45S ~
Protocol 1. Anal ysis of rRNA intermediate s by hybrid selection and ge l electrophoresis' Equipm ent and reagents
32S . .
28S
• SV 40·t ransfo rmed Balb/c mouse cell line SVT2
e Vacuum dot blot app arat us (Schleicher and Schuelll
e Growth med ium (DM EM with 5 % new born c alf serum) and t issue culture dishes IB cm 2 or 20 ern"}
_ Nitrocel lulose ISc hlei cher and Schnell!
3H]
• [ 5· ur id ine ( 18 Cl /mmol. 1 m Ci/m l, ICN l or [ 5 ,6- 3 H] uri di ne (40 Ci /mmol , 1 mC i/m l. ICN ) . PBS / 20 mM potass ium phosphate , pH 7.4 .0.13 M NaC I) e t.v sis buffer 10. 15 M NaC!' 50 mM TrisHCI. pH 8 . 1. 25 mM EOTA . 0 .5 % 505 1
Figure 2 . Gel electrophoretic analys is of the early processing reactions of mouse pre ·rRNA . RNA was labelled in vivo by growi ng mouse tissue culture cells lo r 1 h in the prese nce of ( 3H]uridine. RNA was then extr acted and fracti onated by hybri d selection using a cloned rONA segment corresponding to resid u es 1 - 29 2 of the rRNA transcrip t ion un it (lane? L or t o the 282 nt sequence that beg ins 58 nt beyond the 3 ' -end of the 285 rRNA reg ion (lane 3 ). Unfractionated rRNA is shown in lane t . Follow ing electrophoresis. [ 3H1RNA was detected by tluo roqrephv. g es'dues 1-2 92 of th e 5' ETS are present on ly in t he largest rRNA (47S ) whereas seq uences o f the 3' ETS are pr esent in both t h e 4 7S and 4 6S spec ies bu t not i~ smaller ~RNA S such as the 32S rRNA Isee Proto col 1 for details} . Adapted fr om ref. 22 w ith permis src n -
138
e Pronase (Sigm a; 20 m g /m l in 10 mM Tr is -HC I, pH 7. 5 , 10 mM NaC ]). seudigest the pronase by in cubat in g the solution at 37 °C for 1 h. Store the solution in atiquots at - 20°C e Phenol (red istilled , buffer saturated ): chl oro f orm ( 1: 1 v/ vl • Hybridization buffer (0. 4 M NaCl , 0 .1 M Pipes-KOH . pH 6 .8 . 10 mM EOTA . 70 % v tv tc rrnemide"!
• Denaturat ion bu ffer / 50% vtv for memide". 2.5 m M EDT A , 2 M f orm alde hy de , 40 m M t ri ethano lam ine- HCI , pH 7.51 . 10 x SSC / 1. 5 M NaCI , 0 .15 M trisodium ci t rat e) _ Plasmid DNA contain ing the desi red rONA sequence, line arize d at an app ro priate restriction enzyme site . D issolve 100 IJg li n earized DNA into 2 ml of denaturation buffer and incub at e it at 55 °C for 15 m in . Di lute the denatured DNA to 2 IJg/m l in lO x SSC . Store it at -20°C • Pun ch m ade f rom sharpened sta in less steel tubing (in t ern al diameter 3 mm corresponding to the we lls of the dot blot apparatus) e Glass fib re fil te rs (Whatman GF/Cl
Con tin ued 139
Cath leen Enrigh t and Barbara Sollner-Webb
5; Ribosomal RNA processing in vertebrates Proto col 1.
Pro t o c o l 1
Continued
e Filte r holder (Hoef er) e t.ow -salt wash buffer: 10 mM T ris - He !. pH 7 .5 .1 mM EDTA . 0 . 1% tw /wl SOS e E n d ~o v er · e n d rotating wheel shaker e Modified Oenhardt ' s buffer : 0 .1 5 M NaCI, 30 mM Tris- HCI, pH 7 .5 , 2 .5 mM ED TA , 25 mM sod ium phosphate , 0 .22 mM sodium pyrophosphate , 25 ~g /ml sonic ated , heat-dena tured salmon sperm DNA , 1 mg/ml Fico1l40 0 , 1 mg /ml dextran sul phate SOD 000, 1 mg lm l BSA , 1 % (w /wl 50S . ~r!o r ~o use , inacti v at e the RNase acuvuv In D enhardt ' s buffer by sti rri ng t he buffer w it h 0. 5 % OEPC. Destr oy th e DEPC by inc ubation at 90 ° C f or 20 min . Aft er cooling , adjust the pH t o 7 .2
Continued
e Elut ion bu ffer (9 1% v /v formamide~ , 30 mM Pipes- K OH , pH 6 .8 , 2 mM EOTA )
10. Dr y t he RNA pellet unde r vacu um and resuspend it at - 1 mg /ml in hybridization buffer. The specific radioact ivit y of t he RNA shou ld be
e Horizontal submarine gel appa ratus for agarose gels (1 4 cm long x 19 cm wide x 0,4 cm deep)
Hybrid selection us ing DNA nitrocellulose filters
e Co n cent r at ed formaldeh yde (3 7 % . _ 12 .3 M) ; store in a fu me hood
12. Apply 0 . 5 ml aliquots of t he linearized denatu red DNA (2 ~g /ml in 10 x SSC ) t o the we lls o f the dot bl ot app arat us. Appl y a slow vacuum t o dra w t he so lutio n through t he nitrocellulose .
e Formaldehyde elec tropho resis bu ffe r (2 .0 M formaldehyde, 2 mM EOTA , 40 mM tr iethanolamine-HC!, pH 7 . 5) • Reagent s for fluo rog raphy Ie. q. Kodak G8X fil m , X· ray film develope r, Kodak rap id fix)
- 1 x 10 ' c.p .m. ruq.
11 . A ssem ble t he do t blot appa rat us with a sheet of ni t rocellulose accord ing to t he manufacturer 's ins t ructio ns .
13. Dismant le the apparatus. Dry the fil ter under vacuum and bake it for 2.5 h at 80°C in a vac uum oven . 14. Using the st ainless-steel pun ch , c ut out ci rc les (3 mm in diam et er ) from t he nitrocellul ose sheet cor respo ndi ng t o th e posit ions of t he wells of the dot bl ot apparatu s. , 5 . Incubat e each circle of DNA -n it rocellul ose wi t h 5IJI of RNA solution (from ste p 10 ) f or 3 hat 52 -54 °C in a sealed microc ent rif uge tube .
Method NB Formaldeh yde is tox ic . Store and use all solutions con taining formaldehyde in a fume hood. Labelling o f cells and extraction of RNA Seed the SVT2 cells in 8 cm 2 or 20 em? .tissue cUltu r~ dishes o~ 1. 2 x 10 5 c ells /cm 2 and grow them as monolayers In gro~th ":,edlum at 37 in 95 % air: 5% CO , for 18-24 h until the cells are rn mid -log pha se. 2
Add 3H-labelled uridine precursor to each dish (2? IJCi per 20 cm dish or 2 . 10 IJ Ci per 8 cm 2 dish ) and incubate f or an additional 1 -2 h.
16. After hybrid ization, w ash the fi lt ers in f ou r successive ste ps t o redu ce backg ro und : (a) Lo w -salt w ash on GF/C filter. Remo ve t he filt er fr om the tube (ste p 15 ) with f or cep s and bend the filter with th e f orcep s t o incr ea se wa shing efficiency . Place 1-5 bent filters on a 2.5 cm Whatman GF/C fi bre-glass filter lying on the filter hold er. Rin se the GF/C f ilt er 20 ti mes w it h low-salt was h buffe r (5 m l each t ime). (b) Low-salt rotary wash . Remo ve the D NA f ilters from the GF/C filter .
Place them in a test tube with 0 .5 ml per filt er of low-salt wash buffer. Cap the tube and rotate it for 1 h on th e end -over -end sha ker.
3 . Chi ll the cul tu res rap idly to min imi ze R~A2degralda.ttio h ~. ~~o~~::i~~~e~:~t the growth medium and replace it [within sec ,!"I . . Rin se the cells twice more (5 sec per rinse) wi th Ice -cold PBS .
Ic) Oenherdt 's 55 °C rotar y wa sh . Place the DNA filte rs in a test tube w it h 0 .5 ml per filter of modified Denhardt ' s buffer. Cap and rotat e for 2h at 55 °C .
Lyse th e cell s by add ing 1-2 ml of lysis buffer containing 100 ~glml self 4 . digested pronas e (l.e. 5 or 10 ~I of 20 mg lm l sto ck ] at 20 °C and Incubat e
(d) Reh vbridiretion , In a clean m icrocent rifu ge tube, rinse the DNA filt ers twic e wi th 0 .5 ml of low-salt bu ffer per f ilt er and once with 0 .1 ml o f hybridization buffer per filter. Centrifuge th e tube br iefl y and remove the hybridizat ion buffer. Add 2 ~I per filt er of fresh hybridization buffer and heat it for 10 min at 70 °C. T hen rehybridize the filt er s f or 2 h at 49 °C wit h rot ati on.
th e cells at 20 ° C . 5 . Scrape the lysed cell s int o a test tube and incubate the lysate for 30 m in at 45 °C . 6 . Ext ract t he lys ate by vo rtexi ng wi t h an equal volu me of phenol :chloroform at room temperature . Separate the phase s by c ent rifugation for 2 min at 10000 g . Add 2.5 v el. 7 . o f absolut e et hanol t o the aqueo us ph ase and inc ubate th e mixtu re In an
17 . Elut e the rRNA fr om the was hed filte rs at 90 ° C f or 10 m in in 10 elution buffer.
~I
of
Forma ldehydel agarose gel electrophores is
eth anol ice bath for 15 min. . 8 . Coll ect th e RNA pre cipitat e by ce ntrif ugati on at 12000 g fo r 20 min at
18 . Cast a horizont al 0 .7% agaro se gel in f ormald eh y de elec tro pho resis bu ff er.
'f . 4 °C . 9 . W ash the RNA pellet by resus pending it in 70% ethanol and ce nt n uglng for 1 min (as in step 8 ).
NB Formaldehyde is toxic. Cast the formaldehyde gels and run them in a fume hood .
Continued
Continued
140
141
5: Ribosomol RNA proce ssing in vertebrates Cathleen Enright and Borbara Solln er-Webb Protocol 1 . Continued
Protoc ol 2 . Continued
19 . Bring the eluted RNA Istep 17) to 55 °C and add concentrated formaldehyde to a final concentration of 2 M . Incubate the RNA f or
15min at 55 °C. 20 . Load the RNA on to the 0 .7 % agarose gel and electrophorese the gel for 650 -1000 volt -hours at room temperatu re using electrophoresis
buffer. 21 . Impregnate the gel w ith an enhancer fluor. using th e method recommended by the manufacturer.
e A.M T ps ora len (H Al A s so ci at es. Inc .) dissolved in TS bu ff er
• Long-w ave (36 5 om) UV transillum inator le.g. Mo del C- 6 2 . Ul tra-Vi olet Product s fnc .I .
e TE-5DS buff er (10 mM Tris- He l pH 7 2 5 mM EDTA. pH 8 .0 .0. 1 % 50S ) "
• Proteina se K
. Phe nol, buffer sa turated e Ph enol :c h lo ro f orm :iso am yl alcohol
12 5:2 4: 1 vlv)
• 4 .0 M so d ium ac e tat e, pH 5 .4
22. Visualize the RNA bands by autoradiography. From refs 22 and 24. De ion ize t he fo rmamide using a mi xed bed jon exch ang e res in. How ever . th ere is no need t o deion ize SRL Ultrapure f ormamide before use . St ore the formamid e at - 20 °C .
~ b
Method 1. Grow the cells to a densi ty of 0 .5-1 x 10 6 ce lls/m l. 2. Rinse th e cells t wi ce with PBS, - 5 m l eac h rin se.
2.3 Use of psoralen cross-linking for the analysis of associations between rRNAs in vivo Cross-linking with psoralen has been successfully employed to investigate the potential juxtapositioning of two rRNA species during processing, by creating covalent cross-link s between ba se-paired or otherwise clo sely-positioned nucleotide residues . Th is technique is a powerful one for the characterization of RN P complexes, because it allows detection of RNA associations that are stabilized by proteins as well as ones that are formed directly between RNAs. Pso ralen cross-linking has been used to show that the human and rat U3 snRNAs are specifically associated with the pre-rRNA within the few hundred nueleo tides surr o unding the 5 ' ETS processing site (25,26). The U3 sequence that is cr osslink ed to the rat 5 ' ETS was then identified by rever se transcribing the cro sslink ed RNA. The transcriptase is unable to transcribe past the cross-lin ked nueleotides (see ref. 26 for procedures) . Protocols for cross-linking and for the analysis of the cross-linked species are described in Volume 1, Cha pter 5 (Section 7) and refs 25 and 26. How ever, Protocol Z briefly describ es a procedure for in vivo cro ss-linking with psoralen using rat hepatoma cells. This procedure can be adapted easily for an alternative cell line .
Protoc ol 2 . Analysis of snRNA and rRNA inte rac ti ons by psora le n cross-linking in vivo
3 . Resusp end t he ce lls in 0 . 1 vol. of ice-cold TS buffer.
4 . Add AMT psora len (final concentration 20 ~g /m ll . 5 . Place the ce lls in an open Petri dish on ice . 6 . Irradiate th e. c ells for - 15 min (in 5 m in interv als coo ling on ic e in ~~~~e:~/~~~.g 365 nm light from a long-wa ve UV transilluminator at 7 . Brief ly pe llet the cells in a m ic roce nt rifuge. Rem o ve the bu ffer .
8 . Resusp end the cells in TE-SDS buffer. 9 . Pass t he suspe ns ion t w ice thro ug h a 23 -gau ge needle.
10 . Ad~ proteinase K to 0. 4 m g/m l and diges t the cells for 2- 3 hat 65 0C w it
occasrona ! gentle mi xing.
'
11 ..Extra ct the mi xture twi ce wit h phen ol and once w it h Phenol.chloroform : l,s~~~b' al~Oh~1. I.n eac h case separate th e phases by ce nt;ifu gation a~ 9 or min . Rem ov e and ret ain t he upper (aqueous) phase .
12 . Add 0 .1 vol. of 4 .0 M sodium ac etate , pH 5.4. 13 . Add 1 vol. of abs olute etha nol. 14. Sp ool ou t high m olec ular w eigh t D NA and discard ' rem aining nu cl eic ac id wit h th e additi on of a second
..
vJ~i:~:C~I~~~~~~I~
15 . Collect the precipitated RNA by centrifugation at 10000 9 for 10 min. 16 . Determine the positi o f th UV . . if d A n 0 e cross-link by primer ex tensio n of t he P ~ rI re r NA (st e~ 15 ) wi t h rev erse transcriptase and the app ro riate ~ end-~abelled oligon ucl eot id e prime rs and ana lysis of the RJA on C~n aturln2g PsOIY~ CrYlamide gels (see Vo lume I, Cha pte r 1, Sect ion 2 3 ' . , apter , ectron 3.5; Chapter 4 , Sect ion 7.2 1.
Equipment and reagents e PBS (see Protocol 1)
e Hat Nov ikoff hepat oma cells. (HGPRT
NISI-67) e Gro wth medium (OM EM wi th 10% foe t al c alf se ru m and 2 mM glutamine)
_ T5 bu ffer fO. 1 5 M NaCI. 10mM TrisHe !. pH 7.5 )
Continued
2.4 Ele,ctron micro scop e analysis of rRNA synth ' (MIlle r sprea ding) esis Another valuabl~ technique for an alysing rRNA proce ssing in vivo is the chromatin spreading and electron microscopic visuali zation me thod pione ered
142 143
5: Rib osomol RNA process ing in vertebrates
Cathleen Enrig ht and Barbara Solln er-lVeb b
by Oscar Miller's laboratory (27) over two decade s ago. These investigators demonstrated that actively transcribed rRNA genes appear as structures that have been called ' Christmas trees' , with the DNA axis fo rming the tru nk of the tree, the nascent rRNA molecules and associated proteins form ing the branches, and a biochemically undefined ' terminal knob' for ming a ball at the 5 ' -end of each branch. Recent work (reviewed in ref. 41) has shown that this 'te rminal knob' is the processing complex that forms at the 5 ' ETS pro cessing site. Spliceosome complexes have also been visualized on nascent pre-mRNAs by Miller spreading (28). While far from being simple protocols (27, 28), for th ose skilled in electron microscopy this spreading technique is a very powerful method for studying processing because it provid es spatial and temporal data that are not readily obtained from biochemical analy ses.
Protocol 3 . In vitro preparation of internally-labelled rRNA substrate us ing T7 RNA polymerase' Equip m ent and reagents • Reagents for in vitro synthesis of rad io labell ed rRNA subst rat es u sing T7 RNA polym era se and rONA tem plat e (see
Volume I. Chapter 1. Protocol 4 a nd Cha pter 2. th is volume . Prot ocol 4 ) • Dena turing polyacr ylam ide (sequencing) gel, rea gents an d equipmen t (Volume I Cha pt er 1. Pr otoco l 8 or Volume " Chapter 4 , Pro tocol 11) • 32
• r a_ p ] CTP {le N; 800 Ci/mmo l 10 "C iI" " or [ 5·'H I CTP IICN : 25 Ci/mm ol, 10 jJCi/jJll '
• Ge r el ut ion buffer (0 . 5 M am m onium aceta t e, pH 6 .5 , 0 .1 % 50S , 0 .1 m M
EDTA I
• Glycog en (m olec ular bio logy grad e; Boehrin ger M annheim l
. 0 . 25 M NaCI • Oiethy lp yrocarbonate IOEPC) -t reated H 20 , add DEPC to 0 .01 % lv/v) and allow the water t o sta nd for 1 h, t hen au t oclave it t o inactivate D EPC
3, Ana lysis of rRNA processing in vitro Me thod
3,1 Introduction In order to study the events of rRNA processing, it has proved extrem ely ad vantageous to have a system in which the processing ca n be re-creat ed in vitro, using fractionated processi ng components and synthetic rRNA substrates. Processing extract s have been relative ly easy to obtain fro m mou se and Xenopus for the 5 ' ETS cleavage reaction, and the reactio ns that excise the 18S region have also been a menable to in vitro ana lysis. The other metazoan rRNA pro cessing event s, however, have thus far been refractory to re-creation in vitro. Table I summarizes the availab le systems for in vitro analysi s of proces sing at recognized metazoan rRNA proce ssing sites. In yeast , the pro cessing that occurs at the 3 ' -end of the 25S region has been re-created in vitro using a whole cell extract and synth etic substrate rRNA (29). An alternative approach has been to study the proc essing o f rRNA transcripts in yeast transformed by the introduction of rRNA genes that have been 'tagged' by the introduction of heterogenous sequences (2, 3).
1. Synthesize rANA regions of int erest using T 7 ANA polymerase as described ~ n Vo lume I, Chapter 1, Protocol 4 or Chapter 2 , this volume Protoc ol -a In the presence of 800C i/mmol [a · J2PjCTP and 230 ng pl~smid rO NA template per 4~ J.l1reacno n . To synthesize large r amounts of longer-lived
subs trate that
Substrate rRNA s for in vitro rRNA processing are most commo nly synthesized using a bact erio phage RNA polymerase and a plasm id templ ate in which the desired rR NA gene region has been cloned adj acent to the corres ponding bacteriophage promoter (ref. 23 and Volume I, Chapter I, Section 2.2 and this volum e, Chapter 2, Section 2.6). Protocols 3-5 describe the synt hesis of rad iolabelled rRNA substrates using the T7 phage promoter and T7 RNA polymerase. 14 4
not J2P·labelled, use 25Ci/mmol [5· ' HJCTP and 1 ~ g
2 . Fract ionat e the rANA by elec trophoresis on a sequenc ing gel (Volum e I Cha~Jer 1, Protocol 8 or Volume I, Chapter 4 , Pro tocol 1 1) and vjsualiz~ t he P-Iabelled bands by a 2 min exposure to X-ray film . 3 . To is? 'at.e the ~NA , cu~ the appropriate band from the gel and crush th e ge l slice In a mlcroc.ent~lfuge tube us ing a disposable pipette t ip. Incubate the crus hed gel slice In 0 .5 m l of gel elution buffer conta in ing 20 jJg
glycogen for at least 2 h at 37 0C.
4 . Add 2.5 vol. of ethanol and collect the precipitate (Prot ocol t , steps 7-91 . 5. Resuspend the rRNA in 0.25 ml of 0.25 M NaCI by vortexing vigorously. 6.
3,2 Preparation of synthetic rRNA substrates
IS
plasmid rONA tem plate per 100 J..l 1react ion. To determine the yie ld of RNA (see below) cou nt a 2 J..l1 aliquot and calcu lat e th e to t al radioact ivit y .
Centrifu~e t he suspension at 12 000 9 at room temperature for 10 min and aqain ethano l precipitate the rANA from the supernatant (step 4 1.
7. Wash the rRNA pellet in 70 % ethanol and resuspend it in 50 ~I of DEPCtreated H20 .
Determination of the RNA yield
8.
Calcula~e th e t ota l inpu t c. p.m . of CTP (or other labell ed nu cleotide as appropriate ) from th e tota l rad ioact ivit y in the tran scr iption reaction (see step 1 abo ve ). Continued 145
Cathleen Enright and Barbara Salln er-Webb
5: Ribosomal RNA processing in vertebrates Protocol
Proto col 4 , Continued
3 . Continued
9 . To calculate the specific activity of the C TP uSse)db inhthe r e~cti o~ (c. p. m .l pm oll, divide the t ot al radioactivity (step y . t e num er. a prnol of CTP in the reaction (pmol of CTP = ~ M C!? In t he reaction m ult ip lied by t he volu m e in microlit res of t he reaction ). 10. Count an aliquot of the pur if ied RNA (f rom st ep 7 1and calculate the t otal c .p .rn. in the iso lat ed RNA . 11 . Div ide calculatio n 10 by calculat ion 9 t o ca lc ulate t he total numbe r of prn ol of CTP in corpo rat ed int o the isolated RNA . 12 . To calc ulat e th e pmol o f isolat ed RNA, divid e calculat io n 11 by the number of eM? resi due s in the RNA seque nce .
3. Incubate the mixture at 37 °C for 20 m in to dephosphorylate the 5 ' term inus of the rRNA .
4 . Inactivate the CIP by add ing 5
~I of 0 .1
6 . Extract the rRNA with phenol :chloroform an d ethanol precipitate it (see Protocol 1 steps 6 -91. 7 . L ab el the rR N A by m ixing the following in a microcentrifuge tube: volume containing 1-50 pmol of de phos pho rylated RNA 5 ' ·ends
• rRNA
. 10 x kinase buffer
3
• [a- 32 P1A TP 1l50~C i)'
4. In vitro preparation of 5 ' end -labe lled rRN A s u b st ra te
Equipment and reagents • Reag en ts for in vit ro sy nthesis of unlabelled rRNA substr ate using T7 RNA polymerase and rONA templ at e (see Volume I. Chapte r 1, Protoco l 3 1
e Calf intes tinal alkaline phosphat ase (CIPI (Gibeo - BRL, 24 unit s/utl • React 2 buffer (Gibeo - SRL; 50 mM Tris-
HCI, pH B.O, 10mM MgCI, . 50 mM NaCIi e Denaturing polyacrylamide (sequencing) gel , reagent s fo r gel isolat ion of RNA , and DEPC-tr eated H 20 as for int ernallylabelled substrate s (see Prot oco l 3 above)
e O. l M EDlA e 20 % 50S • 10 mg /ml proteinase K
• 3.0 M sodi um acetate , pH 5. 2 e Phenol:ehloroform (1: 1, v /vl e O.2 M OTT ; sto re at - 20
-c
e l 0 x kinase buffer (0. 5 M Tris - HCI. pH 7.5, 0 .1 M MgCI" 50mM OTT, 1 mM EDTA , pH 8.0 1or 10 x One-Phor All Buffer Plus (pharmacial e
~I
15~1
• 1 mg /ml BSA • T4 po ly n u cl e o t id e kinase (20 units )
1.5 ~I 2 IJI
• DEPC-treated H 2 0 to 30 ~ I fina l volume 8 . Incubate the mixture for 1 h at 37 "C. 9 . Isolat e th e labelled rRNA by el ec trophoresis and etha nol precipitation as descri bed in Protocol 3 (steps 2 -7) for internally-labe lle d rR NA . A lo nger exposure will be autorad iography.
necessary
to
v isu alize end-la belled
rRNA during
• To generate un labelled but 5 ' terminetlv-phosphory lated tran scripts , add 1 mM ATP in place 01 the radiolabell ed A TP.
i o3 2 p I ATP (ICN; 3000 Ci/m mol, 10 IJCi/lJ1)
e TE buff er t 10 m M Tris -H CI, pH 8.0 .
1 mM EOlAI . 1.0 mg /ml BSA (BRL. acet vlated BSA dilu ted in TE buff er) e T4 polynu cleotid e kinase (Pharmacia. 10 units/1J1l
Method 1. Synthes ize unlabelled rRNA by in V olum e I. Chapter 1, Protocol 3.
vitro tran scription as des c ri b ed in
2 . In a m ictocentrifuge tube mix :
• rRNA substrate (ste p 1 )
~I
5 . Dilute the m ixture to 250 ~I w ith 25 ~I of 3 M so diu m acetate , pH 5.2, and 125 ~I of H 2 0 .
• Steps 2 -7 a re adap te d from re f. 30 .
Protocol
M EDlA , 2.5 ~I of 20 % SDS , 2
of 10 m g /m l proteinase K and H 20 to 1 0 0 ~ 1 , and heat the m ixture at 6 5 °e fo r 30 min.
<;; 2 50 prnol
2 ~I
• React 2 buffer . CIP
0 .25
~I
. D EPC ·t reat ed H 20
to 20
~I
146
Pr o to c o l
5 . In vitro
preparation of
3'
end-labelled r RNA substrate
Equip m ent and reagents e Reagents for in vi tro synt hesis of unlabelled rRNA sub st rate us ing T7 RNA polymerase (see Volume I, Chapter 1, Protoco l 3 1
e O. l mg /ml BSA
e Oenaturing polyacrylam ide (sequencing) gel and reagen ts for isolation of RNA as for int ernally -labelled subst rates (see above )
e ( 32pJpCp (ICN. 3000Ci/mmol. 1 0~ i /lJ1)
e O.2 M OTT; st ore at - 20 " C
e OMSO (Sigm a) e T4 RNA ligase tpha-mac!a: 20 unit s/pl!
e 10 x ligat ion buff er (0 .5 M 'r ns-Hc f. pH 7.5, 0 .1 M MgCI" 50 mM OTT), NB add OTT just before use from the 0 .2 M stock solution
e O.25 M AlP
Continued
Continued
147
5: Ribosomol RNA processing in vertebro tes
Cathleen Enright and Barbara Sollner-Webb
Protocol 5. Continued
Pro tocol 6 . Continued
Me thod
e Low-salt buffer (10 mM Hepes-KOH, pH 7.9 , 10mM KCI. 1.5mM MgC I" 1 mM DTTb, 0 . 1 mM EDTA I
1. Synt hesize unlabelled rRNA by in vitro tran scr ipti on as d escr ibed in Volume I, Chapter 1, Protoc ol 3 .
. 5 x extract buffer 10 . 2 5 M Hepes-KOH . pH 7 .9 . 0 .75 M KCI , 17 .5 mM MgCI" 5 mM DTT', 0 .5 mM EDTA . 50 % glycerol)
• Oounee homogenizer with type 8 pestle
2 . In a m icrocentrifug e tube resu spend 2 0 pm ol of th e un lab ell ed rRNA in a mixt ure of 2 ul of 0 .2 5 M ATP and 2 ul of 0 . 1 mg lml BSA .
Method
3 . To t he rRNA sus pension add :
e DMSO e [ " p ] pCp (100 uCi)'
2 ul 10ul
• l O x ligation buffer
2 IJI
e T4 RNA ligase
2 ul
4 . Incubat e the mixture fo r
~
Carry out all steps at 0-4 °C 1 . Grow mouse L 12 10 cells in D ME M growth medium to m id -log phase (5 x 10 5 ce lis /mil in spinner flasks . 2 . Harv est the ce lls by pelle ting t hem in pla stic cen tr ifuge bottles at 2000 9 for 5 min .
8 h at 16 ° C .
5 . Isol at e t he lab elled rRNA by elec t ro phor esis and et ha nol precip it at ion as described in Protoc ol 3 (st eps 2-7 1for int ernally -labelled rRNA. A longer ex pos ure will be neces sar y to vis ualize end- lab elled rRNA duri ng autoradiog raphy .
4 . Mea sure th e packed ce ll volume . Resu spend the was hed cells in 8 vet. of lo w-sa lt buff er and ce ntrif uge t hem at 2000 9 f or 5 m in .
..Fo r the preparation of unlabelled bu t 3 ' terminally -phosphorylated t ranscripts . add un labe lled ceo in pla ce of the radiolabened pCp (ref. 1 2 ).
5 . Qui ck ly discard the supernatant . Resuspen d the ce lls in 3 vo l. (as in step 4 ) of low-salt buffer and allow the m t o swell for 15 m in at a vC.
3.3 Pre paration and use of mouse S-100 proces sing extract
6 . Homogenize the ce lls in the Oounce homogenizer using 10 strokes of th e B pestle . En sure that virtually all the cells and a sizeable fract ion of the nuclei have been broke n by examining a drop of the lysate, spread on a glass slide and covered w ith a cove rslip, under a m icr oscope.
The mou se S-IOO cell extract. whose preparat ion is describe d in Protocol 6, is commo nly used fo r transcription of mou se rRNA genes (9) but also efficiently directs the processing of pre-rR NA at the 5 ' ETS site. The processing facto rs involved appear to be more evolution arily conserved th an a re the RNA polymerase I tran scription compone nts , since the mouse S-lOO extract also processes the 5' ETS site of hum an pre-rRNA (10). Int ernally labelled rR NA substrates (see Protocol 3) are genera lly used for this analysis. An efficient processing substrate need contain only the 200 nt conserved processing signal 3 ' to the actual cleavage site (31) of mouse pre-rR NA, but processing is more efficient when the sequence is extended by an additional 250 nt. The use o f the S- IOO extract for processing pre-rRNA is described in Protocol 7. The Dignam extract (32) also provides an efficient system in which to study the mouse 5 ' ETS processing.
3 . Wa sh t he c ell pelle t t wi ce w ith 10 vet . of iso to nic bu ffe r.
7 . Add 0.2 5 vol. o f 5 x ext rac t buffer and cent rifuge t he mixt ure at 100 000 9 for 60 m in. 8 . Ca ref ully rem ov e the supernata nt and divi de it into 50 - 100 IJI aliquots in m icro cent rifu ge t ubes. Freeze th e tub es in an et hanol /d ry -ice bat h and sto re them at - 70 "C . • Based on ref . 9 . o Add the OTT just pr ior t o use f rom t he 0 .2 M OTT st ock soluti on .
Protocol 7 . Pro c ess in g of pre-rRNA su bs t rates in t he mous e S-' 0 0
ex t rac t-
Equipm ent and reagen ts Protocol 6. Prep arat ion of th e m ou se c ell
S-' 00 pro cessin g ex t ract"
Equipm ent and reagents • Mou se lymphoma cell line L121 0
172
e OM EM growth medium w ith 10% heatinac tivat ed (56 ° C for 30 m in ) horse serum , 3.02 gIl sodium bicarbonate. 172 unlt arml pen ic illin G sod ium,
~g /ml
streptom ycin sulphate. 3.4 mM
t -qtut amine _ 0. 2 M on; store at - 20 °C et soronic buffer (30 mM Hepes -KOH . pH 7 .9 .0.12 M NaCI . 1 mM DTTb )
e 5 -100 processing ext rac t and 0 .2 M OTT (see Protocol 6) e 5 x 5 -10 0 processing buffer ISO m M Hepes- KOH. pH 7.9 . 0.45 M Kel, 5 mM MgCI " 3 mM DTT' , 0 .5 mM EDTA , 35 % glycerol) e 10 mM A TP (neut ralized t o pH 7 . 5) e tntemeuv-rebenec [
32
strete (10 fmol in 2 ~I of H ) svn 20 thesized as in Protocol 3 e 3 .0 M sod ium acetate . pH 5.4 - 0 . 1 M EDTA e 10 mg /ml tANA
p I pre ·r ANA sub -
e Phenol :chloroform f 1: 1 v /vl
Continued 148
Continued 149
5: Ribosoma l RN A praces sing in vert ebrates
Cath lee n Enrigh t and Barbara Sollner-Webb
Protocol 7. Continued e Fo rm am id e/d ye m ixture (9 5% 10r memid e", 5 mM EDTA , 0 .03 % xy lene cv anol. 0 .03% bromophenol blue)
• 10 x TBE electrophoresis buffer (50 mM 'r na- bor at e, pH 8.3 . 1 mM EOTA I
A
1 2
e Denat urinq 4 % polya crylam ide /9 M urea (seque nci ng) gel 120 em lo ng , 0.4 m m t hic k . wi dt h dependent on sam ple number ) in TB E buffer . and electro ph ore sis equ ipment. See Volume L Chapter 1. Proto col 8 or Vol um e I• Chapter 4 , Proto col 11
p rec u rs o r p rod uct
--
Method
1 2
1 . In a microcentrifuge tube on ice mix :
p recu rso r
. 5 x 5 -100 processing buffer
. 10 mM ATP • [ 3 2 P] pre-,RNA substrate (10 fmol) . 5 -100 processing ex t ract . H2 0
'0 25 ~1 oC.
2 . Incubate the reaction fo r 25 -45 min at 30 3 . Dilute t he mixture to a fina l volume of 0 .2 rnl with 20 IJI of 3 M sodium acetate, 20 ~I of O. I M EDTA , 2 ~I of 10 mg/ml ,RNA Icarrierl and H 2 O. 4 . Ext ract t he sample with an equal volume of phenol :chloroform , eth~nol precipitate the rRNA , wash t he RNA pellet in 7 0 % ethano l. and dry It as described in Protocol 1. steps 6-10. 5 . Resuspend the rRNA in 5 ~I of formamide /dye mixture . 6 . Pre-run the seq ue nci ng gel f or 10 mi n or until warm using TBE electrophoresis buffer. Rinse the we lls of t he gel w ith TBE buffer t o remove
urea. 7 . Heat the rRNA sample at 95 °C for 3 min and load it on to the p~e -ru.n (w arm) sequencing gel. Run the gel until the requi red separation IS achieved; th is will depend on the sizes of the RNA s concerned .
8. Vacuum dry t he gel. T here is no need t o pretreat t he gel prior t o drying .
a
"I.
bcf
-
34
B
/ - .." -\.-f pbro d uc t
Figure 3 . In vitro rRNA pro cess ing . (A) S' ETS process ing in th e mouse 5 ·1 0 0 ex tract . J2 p . la bell ed pre -rRNA co nta in ing the mouse 5 ' ETS processing sequen ce (lane 11 was used as a sub strate in t he mouse $ -100 processing rea ction (lane 2 1; the RNA precursor and produ ct were resol ved by electrophoresis and visualized by auto radiography . The proced ure is descr ibed in Pro t ocol 7, A dapted fr om ref . 10. lS I Proc ess ing at the B'vend of th e 185 reg ion in HeLa cel l ex t racts . J2P·labelled pre -rANA cont aining t he 5 ' -en d of t he human 185 region Ilene 11 wa s used as su bst rat e in a processing react ion in vo lving the HeLa cell nucleo lar extract (lane 21. Band s a, b . and c represent rANA mo lecu les cl eaved, respe ct ive ly . 8 nt before , 3 nt before, or 1 nt after t he 5 '·end of the 185 region generated in vivo . Fr agm ent b was pur ified by electr o ph oresis (lane 31 and used as t he su bs t rate in a seco nd pr ocessin g reaction in vol v ing the HeLa cell P-10 0 cyto plasmic extr act (lane 4 ). Adapted fr om ref . 1 2 wi t h perm ission .
9 . V isualize the rRNA products by autoradiography (see Vo lume I, Chapter
1. Protocol 8 1.
Processing at the 5 ' bord er o f the 18S region, which cannot be achieved using the mouse 5-100 extrac t , occurs efficientl y in an alternative two-stage procedu re.
Thu s, this processing can be produced in vitro with hum an pre-rRblA by first using a HeLa cell nucleolar extr act which ma inly cleaves the pre-rRNA at a position either 3 0 ' 8 nt upstream of th e 5 ' -end of th e 18S region (12). Then , using a P-IOO cytoplasmic extract, the 3 nt exten sion of the form er intermediate can be trimmed off to generate the correct (in vivo ) 5 ' -end o f the 18S region (12). In addition, the nucleola r extract also catal yses the processing of the human 5 ' ETS. A nucleolar extr act prepared in the same way from mou se cells appears to have similar proce ssing activities to the HeLa cell nucleolar extract (T. Nilsen, personal communication; S. Kass and B. Sellner-Webb , unpublished data). Th e preparation of these two extracts and their use in the processing reaction are described in Proto cols 8 and 9. Successful processing requ ires a 300 nt synth etic pre-rRNA substrate contai ning the 18S ,RNA processing site.
150
1 51
• Adapted from refs 10 . 3 1. 3 3 , 34 . . b Add the OTT just prior to use fr om the 0 .2 M OTT st oc k solut ion. C Dei on ize the formam ide as in fo otnote b , Protocol 1.
Figure 3A illustrates the use of the mo use S-I 00 proces sing extract for the cleavage of the 5 ' ETS in mou se pre-rRNA.
3.4 Preparation and use of nucle olar and cyt op lasmic processing extracts from HeLa ce lls
Cathleen Enright and Barbara Sallner-Webb
5 : Ribosomol RNA processing in vertebrates
Pro t ocol 8 . Continued Protocol 8 . Preparation of nucleolar and P-l 00 cytoplasmic processing extracts from HeLa cells'
12. Layer this suspension of lysed nuclei over an equal vo lum e of solution C and cent rifuge the grad ient at 3500 9 for 15 m in to recov er the nucl eoli.
Equipm ent and reagents
13 . Resuspend the purified nucleolar pellet in an equal volume of solution
• Ultracentr ifuge. rotor. and (Beck man Ti50 or equiv alent)
tubes
• Virtis 60 homogen izer (or equiv alent ) e Ooun ce homogenizers (wi t h a B t ype pest le and a T eflon pest le) e HeLa cell suspe nsi on 5 x 1O~ cells/mil
culture
14 -
e Grow th medium (SM EM; M . A . Bio products ) supplemented w ith 7 .5 % horse seru m e p BS (20 mM potassium pho sphate.
pH 7.4. 0 .'3 M NaCI) e O. l M PMSF (fr eshly made in ethanol)
e O.2 M OTT e 3 M KCI
e Solution A (10 mM Tris-HC I, pH 7.5 , 1.5 mM Mg CI 2 , 10 mM NaC!, 50 ~g fml dextran sulphate) e Sojution A ' containing 0 .5 % Triton
X·l00
e Solution
B (0. 2 5 M sucrose , 10 mM
MgCI, '
e Solu tion
C (0. 8 8 M suc rose, 50 mM
MgCI, )
e Solution 0
t10 m M Tris -HC !' pH 7.5 . 50 mM MgC I2, 10 mM NaCl, 50 IJg /m l dextran sulp hate! e Solution E 120mM Tris -HCI . pH 7.5 , 0 .2 mM EOTA. 0.5 mM PMSFb • 0. 5 mM OTT c • 20 % glycerol) e Prote in assay kit (Bio -Had]
E. Add 3 M KCI to a final concentration of 0 .5 M . 14 . Incubate the nucleoli on ice for 10 m in with gentle agitat ion . 15. Dilute the nucleolar extract with solution E so that the final concentration of KCI is 0 .12M .
16 . Clari fy the extract by centrifugation for 10 min at 225 000 9 in a Beckman Ti50 rotor. 17 . Carefully recover the supernatant. Determine it s protein concentration using the Bio -Rad protein assay kit d . 18. Store the extract in aliquots {containing - 0.5 mg of protein per mi l at
-70
-c.
B. Preparation of the HeLa cell P·100 cytoplasmic extract 19. Clarify the supernatant from step 7 by centrifugation at 20000 9 for 10min.
e Dext ran sulphat e
20 . Carefully recover the supernatant and centrifuge it at 100000 9 for 60 min in a Beckman Ti50 rotor.
Method
21 . Discard the supernatant and quickly rinse the surface of the pellet with so lution E. 22. Resuspend the pellet in 0.2 vol. (calculated f rom the original cell volume as in step 3) of solution E by gentle homogenization in a 1 ml Dounce homogenizer fitted with a Teflon pestle.
NB All centrifugation steps are at 4 °C A. Preparation of the HeLa cell nucleolar extract
gr~wth moegdium t,o a de~sitdY of 4 x 10' to 6 x , 0' cells/m!. A total of 2 x , 0 - 4 x 1 cells rs requrre .
1 . Grow the Hela cells in suspension culture in
2 . Harvest the cells by centrifugation at 1000 9 for 5 min. 3 . Measure the packed-cell volume and wash the cells with 100 vol. of PBS. 4 . Resuspend the washed cells in 5 vol. of solution A and incubate the cel l suspension for 10 min on ice . 5 . Homogenize the cells in the Daunce homogenizer, usinq the B pestle , until lysis is > 95 % complete . as visualized under a microscope.
6 . Add an equal volume of solution A ' (containing 0 .5 % Triton X·, 00) and collect the nuclei by centrifugation at 1500 9 for 5 min. 7 . Carefully remove the supernatant and save it for'9the2dlmmed iate
preparation of the P-l00 cytoplasmic extract Isteps
-
.
8. Resuspend the nuclei in 10 vol . (referring to the ~riginal cell volume . see step 3 1 of solution B and layer this over a cushion of an equal volume of solution C in a Beckman Ti50 centrifuge tube .
9 . Centrifuge the gradient at 2500g for 10min. 7
10. Resuspend the pur ified nuclei at 2 x 10 nuclei /ml in solut ion 0 (count nuclei in a haemocytometer). 11 . Lyse the nuclei by shearing the solution for 11 min at 5000 r.p .m . in a Virtis 60 homogenizer fitted with the macroblade . .
23. Store the resulting extract in small aliquots at - 70 °C . Ty pically the protein concentration of the P-l00 extract is - 10 mg /ml. ~
Based on ref . 12 . Add PMSF just befor e us e fr om t he 0 .1 M stock solution. C Add OTT just before use from the 0.2 M sto ck solution. " S ee also ref. 35 . b
Prot oc ol 9 . Use of HeLa cell nucleolar and P-l 00 cytoplasmic extracts for processing of the 5 ' terminus of human 185 rRNA '
Equipm en t and reagents e 32P-labelled pre-rRNA subst rate ext end ing from - 70 to + 165 (corresponding t o restr iction sites Neil and Xbal) relative to th e B' <end of 185 rANA . both 5 ' and 3 ' end -labelled RNA s are requ ired (see Prot ocols 4, 5 1. • HeL a
cell
nu c leolar
and
P-l0 0
cy to plasmi c extracts and solution E as in Protocol 8 . e 5 x nucleolar pro cessing buffer 10.32 M KCl, 40mM Tris-HC l. pH 7 .5 , 10mM M gCI2 , 1 mM spermidlnel . OEPC-treat ed H 2 0 . gel elution buffer , and gly cogen ca rrier as in Protoco l 3
Continued
Contmued
152
153
5: Rib osomal RNA p ro cessing in verte brotes Protocol 9 . Continued e Ph en o l :c h l o f o f o r m . 3 .0 M sod iu m acetate , pH 5.4, 0.1 M EDTA . 10 mg /ml tANA as in Protocol 7
_ S e q ue n c ing gels . e lectrophoresis equ ipment . fo rmamide /dye mixture , TBE buffer as in Protocol 7
Cathleen Enrigh t and Barbaro Sollner-Webb Typical da ta obtained using the HeLa cell nucleolar an d P-l OO cytop lasmic extr acts are illustrated in Figure 3B . RNase Tl fingerprint an alysis (described in ref. 12) can be em ployed to verify the precision of the processing reactio ns by examining the termini generated (see also Volume I, Chapter 5, Section 8).
3.5 Clea vage of pre-rRNA using purified nucleolar endonuclease Me th od A . For the ini ti al cl eavage Set up a cle avage react ion for each o f th e end -lab elled pre ·rRNA subs t rate s (5·- or 3· -labe lled l 'l . M ix the fo llo w ing in a microcent rif uqe tube on ice: e 5 · or 3 · end-lab elled pre-rRNA substrat e
volume con taini ng up t o 500 nq 4 ~1
• 5 x nucleolar pro cessing buffer
2
• HeLa ce ll nu cl eolar ex tract
~I
t o 2 0 ~1 fin al vo lume
eDEPC-t reat ed H 2 0
2. Incu bate t he samp les f or 2 -3 hat 30
°c.
3. Recover RNA products as in Protocol 7 ste ps 3- 4 . 4 . Ana lys e RNA pr oducts on 40 em lo ng, 6 % acryla mid el7 M ure a sequencing gel as in Protocol 7, steps 5 -9 .
B. Fina l trimming of the rRNA in term ediate with the 3 n t ex tension 5 . Isola te th e appropriate sized 3' end -labelled RNA interme d i a~e (l.e. th e RNA with t he 3 nt 5 ' ex te nsi on) after the gel elect ro phor esIs (step 4 ) as de scr ibed in Prot ocol 3 , step 3 . 6. Recover t he rRNA int ermediat e by etha nol precipi t at ion (Pro tocol 1, ste ps
7 -9 1. 7. Dilut e the P-l 00 cytoplas m ic ex t ract wi t h 9 v el. of sol utio n E. 8. Set up t he second stage processing react ion as in step 1 by m ixing : e isol at ed rRNA inte rmediate (st ep 6 )
1 1 ~I
e diluted P-100 extract (ste p 7 1
5 ~I
. 5 x nucleolar pr oc essing buffer
411 1
9 . Incubate the m ixture at 37 °C f or - 1 h . 10. Dilute th e mixture to 100 ~I wi t h 10 ~I of 3 M sod ium acet at e. pH 5.4 . 10 ~ I of 0 . 1 M EDTA . 20 ~g glycogen carrier. and H 20 . 11. Phenol:chloro f or m ex tr act and pr eci pitate th e RNA as in Protocol 1, step s
6-9 . • Based on ref . 12 .
1 54
An end onuclease which can be isolated fro m mou se nucleol i ha s been sho wn to cleave mou se pre-rRNA at a limited number of positions (approximately one cleavage every 50-75 nt). One of the se preferential clea vage sites is identical to the position of the 5 ' ETS pro cessing site used in vivo. The relevant procedure s are described in ref. I I. Th e relationship of thi s cleavage reaction to the 5 ' ETS pro cessing which takes place in the various extracts described in Section s 3.3 and 3.4 rema ins to be determined . It doe s not involve the U3 snRNP . It may be th at there are alternative mean s of ach ieving thi s pro cessing, or tha t the processing in extracts may involve this endonuclease with additional facto rs being requ ired to achieve site specificity. Alternatively, the cleav ages by the endonuclease (which is single-strand specific) occur at the single-stranded loops which are predicted to occur in the secondary structur e of the pre-r RNA .
3.6 Use of deletion analysis in the investigation of rRNA processing To determine the nucleotide sequences (or structur al dom ains) of the pre-rRNA that a re necessary to dire ct the in vitro process ing reactions, a systematic series of alt ered rRNA sub str ate s ca n be tra nscribed from appropriatel y co nstructed mutant rO NA templ ates. T hese can be assa yed for processing efficiency by the methods described in Sectio ns 3.3 and 3.4 . P rocedures for pr eparing mu ta nt tem plates containing 5 ' or 3 ' deletion s, block substitution mutations , or point mutations, will be found in standard molecular biolog y/laboratory manuals (e.g . ref. 23). Illustrations of the use of such mutant substrates to anal yse the sequence requ irements for th e 5 ' ETS pro cessing of mou se pre-r RNA a re to be fo und in refs 31 and 33.
4. Analysis of protein-RNA associations in rRNA processing Since all the exper imental evidence available ind icates that metazoan r RNA processing events are not au tocatalytic, ther e is consi derab le int erest in lea rning about the cellular comp onents that are responsible for carrying out the processing reactions. A number of method s are availab le for exploring the a ssociation of proteins with pre-rRNA substrates. These are described in the following sections as ap plied to the 5 ' ETS processing in mous e. 155
Cath leen Enright and Barbara SolIner- Webb
5: Ribos omal RNA processing in ve rte bra tes
4 .1 Gel-shift analys is Protocol 10. Gel -shift assay for analysis of rRNA pro cessing comple xes" Equip m ent and reagents _ 5 -' 00 processing extrac t, 5 x 5 -100 pro ces sing buffer , J2P-label1ed pre-rRNA subst rat e, 10 mM AlP , TBE buffer , and fo rm am ide/dye m ixture (see Protocol 71 . 20 m g /ml hep arin (Si gm a) in OEPe · t reated H 20 (see Protocol 3 1
• Non -denaturing 4 % po ly acr y lam ide gel (2 .0- 2. 5 mrn t hick : 25 em lon g ;
acr y lam ide : bisacr y la m id e. 65 : 11 i n 0 .5 x TBE bu ffer, electrop ho resis equip m ent (pre p arat ion o f no n -dena t urin g po lyac ry lam ide gels f oll ow s t hat of de nat ur ing gels described in Volum e I. Chapter 1, Protocol 8 . or V olume I,
Cha pt er 4 , Protocol 11. e xce pt that t he ur ea is om it t ed)
Method 1. Set up the proces sing react ion as des crib ed in Protocol 7. step 1.
2 . Incubat e the mixture for 25 -45 min at 30 °C. and then place the tub es on ice .
3 . Dilute the st ock 20 mg lml hepar in solution t o 0 .35 mg lml w ith H 2 0 . Gently add 2 ~I of 0 .35 mg fml heparin to th e incubat ion m ixture (st ep 2) to remo ve loosely asso ciated proteins. Do not mix the co nt ents of th e microcentrifuge tube after the heparin addition . 4 . Incub at e th e reaction for an add it ional 10 min at 30 vC and retu rn it t o ice. 5 . Dur ing the processing rea ct ion pre -run the non -denaturing polyacrylamide gel at 13 V fcm f or 1 h using 0.5 x TBE buffer as the electrode
buffer . 6. Load 12 .5 - 15 ~I of the react ion mix t ure onto t he pr e-run gel. Load an equi valent amount of 32P-labelled pre -rRNA subst rat e that ha s no t been exposed to S -100 processing extract in an adjacent well. Also load a sam ple of formamide fdye mixtu re into an empty w ell. 7 . Electrophore se the samp les at no mo re than 13 V fcm (t he gel should never be w arm t o the touch ) until the xyl ene cyanol dye has trav elled the lengt h of the gel. 8 . Dry the gel onto Whatman chromatography paper and vi s u a l i z ~ the prerRNA substrate and the rRNA pro cess ing c om plex by autoradi ography .
Facto rs present in the mouse S-IOO processing extra ct (see Protocol 6 ) that catalyse th e 5 ' ETS proces sing can be identified by their ability to form a specific complex with the pre -rRNA subst rates. but not with processing-incompetent RNA which does not po ssess the necessar y processin g signals, Th e complex is detected by the electro phoretic retardation of the RNA subst rate (gel shift or mo bility retardation). The gel-shi ft assay is described in Protocol 10 using a rad iolab elled RNA substrate and visualizing the complexes di rectly by au to radiography. The result s o f a typical gel-shift assay are shown in Fi gure 4. The specificity of the gel-shift was demonstrated by competition analysis (Figure 4B) . Th e extra ct was pre-incubated with an unlabelled RNA- either a processingcom petent rRNA (sp) containing residues 645-668 o f the rRNA transcription unit which comprise the 5 ' ETS proc essing site or a processing-incompetent rR NA (n sp) devoid of these nucleotides (Figure 4A) . The former . but not the latt er, competito r rRNA competes with the labelled RNA substrate ad ded sub seq uent ly for the assembly of 5 ' ETS ribonucleoprotein processing complex
components. The processing complexes formed on pre-rRNA substrates ca n also be visualized by carrying out Northern blott ing analysis after the electrophoresis, pr o bing th e blotted co mplexes with 32P -labelled nucleic acid probes complementary to specific rRNA sequences. In th is case , perform the processing reaction (Protocol 10, step I) with 80 fmo l of ' H-Iabelled rRNA synt hesized as described in Protocol 3. Volume I, Chapter 2 (Section 3.6), Volume I, Chapter 4 (Section 7.5), and ref. 23 describe protocols for Northern analysis. Negligibl e am o unts o f endogenous rRNA from the mouse S-IOO processing extract appear in the gel and are transfe rred to th e Northern blot, so ba ckground signals du e to endogenous rRNA are minimal. Figure 5 illu strates the use of Northern blot analysis following the gel-shift assay. Here the presence of small nuclear RNA s was shown using an unlabelled rR NA sub strat e and a radi olabelled hybrid ization pro be for the snRNA . On e example is the U3 snRNP which is pre sent in complexes formed on the rR NA that extends from the 5 ' ETS processing site to resid ue - 1060 of the mou se rRNA transcription unit. Becau se the U3 snRNA is non-specifically associated with extract proteins fractionated throughout the gel lane, probing th e complexes formed on different sized rRNAs , which have distinctive mo bilities, permits assurance that the snRNA is indeed specifically associated with the input rRNA .
4.2 Analysis of rRNA associations by rate zonal centrifugation II
Bas ed on ref . 34 .
Th e 5' ETS processing complex can also be resolved by rate zonal centrifugation as described in Protocol ll . 156
157
Cathl een Enright and Barbara Solln er-lVebb
5: Ribosoma l RNA processing in vertebrates
A
Pra tocol 11. Analysis of the 5' ETS processing complex by rate zonal centrifuqation -'
188
588
288
Equipm ent and reagenrs e s-r 00
processing extract. 10 mM A TP 5 x S-1 00 processing buffer . and 32p_ labelled pre- rRNA substrate as in
l
850
Protocol 7
($$e*,
e
• S-1 00 processing buffer lack ing gl y cero l p lus t he same buffer c ontaining 10 % glycerol and conta in ing 30 % gl ycero l
645 - - - - - - - 669-------
e Grad ient former (e.g. Hoefer) and den sity gradient trec nonator (e. q. ISCQ I
tntracentrttuqe. tubes. and rotor (e.g. Beckman SW55 or eq uivalent !
e 20 mg lm l he p ar in
sp nsp Method
B
2
3
4
2
c
3
3 . Place the reac tions on ice , add heparin , and continue the incubation as described in Protocol 10 (st eps 3-4).
250 -
110 -
95 85 75-
s pec If Ie complexes
1. Set up a processing react ion as described in Protocol 7 (st ep 1) but scale up the volume of the reaction to 75 ul, 2 . Incubate the reaction for 50 min at 30 oe.
-
4. Dilut e the processing mixture with an equal vol. of S· 100 processing buffer lacking glycerol. 5 . Set up as ml, 10 -30% glycerol gradient using S-100 processing buffer containing 10% glycerol and 5 -100 processing buffer containing 30% glycerol using the gradient former according to the m anuf ac turer ' s instructions . 6 . Load t he processing reaction (step 4 1onto th e gradient and centr ifuge it
at 40000 r.p.m. in a SW55 rotor for 5 h at 19 0C.
Conrinued COlftpet It or -
•
sp nsp
Gel-shift and UV cross-linking analysis of the mouse 5 ' ETS processing co m ple x • . (AI The mouse 5' ETS processing region . The top of the dlagr~m s h.OWs the .5' ETS_processing 0- 8 50 l site in the mouse rRNA transcription unit . The 200 nt processmq signal l resld.u.e s 65 nonis shown expanded in the middle of t h e diagram . Below are shown the specific (gp.l and 645 specific (nsp ) competitor rRNAs used in parts Band C. which start at rRNA re~l~ues 5' and 669, respec tively. (B) Gel-shift analysis. 31P·label1ed m~use pre -rRNA containing th e nd ETS processing signal (residues 546 -8751 was incubated with the mouse S-100 extract a
th e resu lt ant nucleoprotein complexes (specific complexes] were resolved electrop horetically (lane 2). l ane 1 contains the pre-rRNA substrate alone . lanes 3 and 4 demonstrate the sp ec if ic it y of the ge l-shift . As indicated in the table below the autoradiograph , the extract in lane 3 was f irst pre -incubated w ith a 12-fold molar excess of unlabelled specific competitor rRNA teo: processing-competent rRNA extending from res idues 645-875; see pane l A ). In lane 4 . unlabelled non -specific competitor was used (nsp ; processing-incompetent rRNA ext end ing from residues 669-875 ; see panel A). Bot h extracts were then incubated w it h 32P·labelled pre -rRNA (hot RNAl . (CI UV c ross -linking . The polypeptides labelled when 32P·labelled. 4 -S-U MP substituted pre -rRNA was incubated with the mouse 5 -100 processing ex t rac t . subjected to UV cross-l ink ing and liberated after RNase d igestion are shown in lane 1 . A s indicated at the bottom of the eutoradicqraph, the specific ity of the UV cross-linking Was judged by the effects of pre -incubating the extract w ith either the specific (sp) or non spe c if ic (nsp ) competitior rRNA (see panel A) . Polypept ides in the size range 75-250 kOa Were eliminated by pre-incubation with the specific rRNA {lane 2 1but not with the non -specific rRNA (lane 3 ), indicating that they are specifically associated with the rRNA . In contrast , this pre-inc u bat io n had no effect on polypeptides smaller than 75 kOa. (Adapted from ref . 34 .)
158
159
• sp• nsp •
extr-ect
competi tor hot RNA
+
+
+
+
Hqure 4
5: Ribosomo l RNA p rocessing in ve rtebrates Pro tocol 11 . Continued 7 . Fract io nate t he gradient using the gradie nt fr act ionat or. Collect 25 0 IJI frac tio ns. 8 . Co unt an aliquot of eac h fra ction in a sci nt illat ion counter t o determine t he distributi on of radioact ivit y ac ros s th e gradient. 9 . Use t he rema inder of eac h f ract ion for subse quen t analysis b y gel -sh ift assay (see Prorocol 10 ) or f or iso lat ion of rRNA use one of t he me thod s described in Vol ume I. Ch apt er 1, Sect ion 3 . • From ref . 34 .
Cathl een Enright and Barbaro Sollner-Webb sub strate, 4-S-UT P , is prepared by phosphorylation of 4-S-UDP as described in Protocot tz , The substituted rR NA substrate is then used to form a processing complex with a m?u se S-IOO processing extract (see Protocol Hv, The processing complex IS then irra diated with UV ligh t to cross -link region s of the rR NA su bstra te to thei r assoc iated prot eins in th e complex. Th e rest of the rR NA is th en removed by digestion with RNase A and RNa se Tl , Fin ally, SDS- PAGE is used to sepa rate the cross-linked prOlein: RNA complex. Protocol 12 . S ynthe s is of 4 -$ -UTpa Reag ents
4.3 Use ofUV cross-linking to detect polypeptides associated with rRNA in processing complexes To lab el the polypeptides that are specifically associated with the rRNA 5 ' ETS processing signal, UV cros s-linking has been emp loyed (34). Th e experimental pr ocedure involves th e synt hesis of pr e-rRNA substrate sub stituted with a thi onucleotide, 4-S-UMP . The pr ecursor for the in vitro tra nscription o f this
· 1.0 M Tris-HCI , pH 7 .6
• O. 2 M phosphoenolpy ruvate (PEP)
e 3.0 M KC I
e 20 mM NADH
. 1.0 M MgCI 2
e Lactata dehydrogenase {LOHI
• 0 ,2 M and 0 .1 M OTT stock sol ut ions • 17 mM 4 -t hio·UO P
• Pyruvate kinase (PK l
e O.1 M ATP
• Nucleoside diphosphate kinase (NOK I
M e/h od 1 . Make up reaction buffer by mixing:
•
e 1.0 M Tris- HC!. pH 7.6 400 ~I e 3.0 M KC I 5 3~ 1 e 1.0 M MgCl, 48 ~I e O.2 M DTT 5 ~I e H, O 3.49 4 ml 2 . Make PKIl DHmix by dissolving PK and LDH in reaction buffer (_ O.75 U /~I PK and - 1.0 U /~I LDH). 3 . In a m icrocentrifuge tube m ix the following :
Figure 5 . Ana lysis of U3 RNA in electrophoretica lly separated complexes. Process ing react ions containing the indicat ed RNAs [ rRNA substrates 13 1 2 - 12 90 and 645 -1290) and th e pro cessing -in competent RNAs IrRNA 669-1290) and a 900 nt pGEM transcr ipt ) we re incubated as in Protocol 7, electrophoreticall y resolved o n non -denaturing gels as in ProtoCOl 10 . blotted, and analysed with a probe for U3 sequences . The solid arrow represent s the posit ions of specific shifted complexes .
160
e 1.0 M Tris-HCI, pH 7.6 e 3.0 M KCI e O. 1 M DTT e 1.0 M MgCl, e 17 mM 4-thio-UDP e O. l M ATP e O.2 M PEP e 20 mM NA DH e PK/LD H mix e 4 ~I NDK e H2 0
100 ~I 13.3 ~1 2 . 5~1 12~1
100 ~ I 30 ~1 40~ 1
64 ~I 4 .3 ~ I 4 ~I to 1.0 ml fina l volume
Continued 161
Cathleen Enright and Barbara Sollner-Webb
5: Rib osomal RNA processi ng in ve rte brates Protocol 12.
Protocol 13.
Continued
4 . Incu bate th e mixtur e at 37 °C f or 1 h. kee ping as ligh t ·free as possible. S
Extra ct th e mi xture wi th ph enol and t hen wit h eth er. Store t h e so lution at _ 70 0C, in foil -covered tubes. Under these conditi ons about 70% of the 4 -th io-UDP is phosphorylated to 4 -thi o-UTP .
Continued 2 ~I
e RNasin
5~1
. 5 mM ATP /GTP solution
4. 5 ~I
e 10 mM CTP
5 ~I
. 2 mM 4-$-UTP
• ra-
32
P j CTP 1120 ~Cil
12 ~I
e template DNA (0 . 7 ~g l
• Adapted from ref . 36 .
3 ~I
2 . Add 1.51J1 of T7 RNA po lymerase and incubate the reaction at 37 °C for 90 min, keeping it as light-free as possible.
Protocol 13. Analysis of the proteins associated with processing complexes us ing UV cross-l inking'
Equipment and reagents buffer 10.2 M TrisHCI. pH 7.5. 30mM MgCI, . 10mM
e Gel eluti on buffer and glycogen as in Protoco l 3
sp erm id ine )
e S -10 0 processing extract (see Pro tocol 61 e 5 x S · 10 0 processin g buff er ts ee Protocol 7) with NaC' replacing the KCI • Extract buff er (50 mM Hepes-KOH, pH 7.9. 0.15 M NaCI. 3.5 mM MgCI, . 1 mM OTT) e 20 IJgfml heparin e Transiliuminator (A=300 nm, 7 mW/cm2; FotoPrep 1 from Fotodyne or equivalent )
. 5 x transcription
e A T P/GTP solut io n (5 m M each), neutral-
ized to pH 7 .5 ; s tore at - 20 °C • Separate 5 mM ATP and 10 mM CTP solut ions . neutralized t o pH 7 .5 ; store
- 20° C • (o:.32P1CTP (leN; 3000 Ci/m mol,
10oCItoil e 2 m M 4 ·S -UT P (see Pro toco l 12)
e O.2 M OTT
e RNase A (SRl 10 mg/mll e RNase T 1 (BRL 200U /ol) e 3 x SDS-gel loading buffer to.i 8 M Triscloned adjacent to the T7 phage Hel, pH 6 .8 . 1.5 M 2-mercaptoethanol, promoter: 0 . 25 mglmll 30 % glycerol. 6 % SDS, 0 .1 % bromo• T7 RNA polymerase tSAl; 50 Ul IJI) phenol blue) • RNa sin (Promegal e SDS- PAGE gel (10% acrvl amid e prepared according to Volume I, Chapter e Denaturing polyacrylamide (sequencing) 5. Protocol 7 or ref. 37 ) gel. formamide/dye mixture. TBE buffer as in Prot ocol 7
• Lin earized template DNA (pla smid con taining th e appr op riat e rRNA sequenc e
Method Preparation of the pre-rRNA subsrrate substituted w ith 4 ·S -UMP 1. Synthesize the pre -rRNA substrate substituted with 4-S-UMP .by setting up the f oll owing react ion (in the order given ) in a micro cent nf u ge t ube on ice : volume to achieve 48 . 51J1 final vo lume . H, O 1O~1
• 5 x transcr iption buffer
2
eO.2 M DTT
~I
3. Isolat e the 3 2P-labelled. substituted rRNA product by electrophoresis on a sequencing gel. as described in Protocol 3. steps 2-7. 4 . Calculate the yield of rRNA product as described in Protocol 3. steps 8-12. using UT P as the marker nucleotide .
Prep aration of processing complexes using 4-5 -UMP substituted pre -rRNA substrate NB Protect the processing react ions from light in steps 5-7 to m in imize spurious cross -link ing .
5 . Set up a 50 IJI processing reaction . as described in Protocol 5 , (step 1) by mixing:
e 4 -$ -UMP substituted rRNA (step 3)
3 ~I
. 5 - 10 0 processing extract
. 5 mM ATP
15
~I
7 ~I
• extract buffer e 5 x 5·100 processing bulter (w ith NaCI in place of KC"
10
~I
to 50 ~I fina l volume 6 . Incubate the reaction for 50 min at 30 °C and then pla ce it on ice.
• H, 0
7 . Add 2 IJI of 20 IJg/ml heparin and incubate the mixture for an additional l Omln at 30 · C. Cross-Iink inq of processing proteins to rRNA and SDS-PAGE ana lys is 8. Transfer the sample to one of the wells in a 96·well microtitre plate and place the plate in a Pyrex dish containing just enough cold water to allow the rnicrotitre dish to float . 9 . Place the Pyrex dish on top of the UV transilluminator and irradi at e the sample for 15 min using the ana lytica l power setting . 10 . In a m icrocentrifuge tube . mix 40 IJg of RNase A and 200 units of RNase Tl in H 20 to a final volume of 20 IJI. 1 1 . Transfer 20 IJI of the irradiated processing sample (step 9 ) to the microcentrifuge tube conta ining the RNase A and RNase T1 (st ep 10) and inc ubat e the mixture f or 15 min at 30 °C .
Continued
Continued 162
volume containing 10 tmol
163
5: Ribosomal RNA processing in vertebrotes Protocol 13. Continued 12. Add 20 ~I of 3 x 50S -gel loading buffer to the digested sample and heat the m ixture for 3 min at 95 °C .
Ca th lee n Enright and Barbaro Sollner-Webb The length of time required for microco ccal nuclease digestion o f the extract should be determined in a pilot experiment , but the time in Protocol 14 has been used successfully in the au thor ' s laboratory.
13. Load the sample onto the 50S-PAGE gel and separate the proteins as described in Volume I, Chapter 5, Protocol 7, or ref. 37 . 14. Identify the proteins cross-linked to the nuclease-resistant 32P-labelled RNA residues by autorad iography. • Based on ref , 3 4 ,
Protocol 14. Treatment of the mouse S -1 0 0 processing extract with micrococcal nuc lease -
Equipm ent and reagents
Representative data from a UV cross- linking experiment are shown in Figure 4C. Polypeptides that become labelled when using a radiolabe lled proce ssingcompetent rRNA subst rat e, but not when using a radi olabelled proce ssingincompetent rR NA subs trate, are deemed to be specifically associ ated. Pre-incubatio n of the processing extract with unlabelled processing -competent or processing-incompetent rR NA substrates before the addition of the labelled 4-S-UMP substituted substrate af fords confirmation of specificity. The subsequent labelling of specifically associated polypeptides will be ab olished by the former but not by the latter competitor RNA as shown in Figure 4C.
5, Assessment of the requirement for cellular RNA in rRNA processing Section 4 described meth od s whereby processing polypeptides associated with pre-rRNA could be identified . Pr ocessing o f pre-rR NA may involve the participation of several cellular RNA s, e.g. U3 snRNA. Thi s section describes micrococcal nuclease treatment and oligonucleotide-directed RNase H digestion of RNAs in a process ing extract which can be used to pro vide evidence for the requirement of cellular RNA in the processing reaction .
• 5-100 processing ext ract (see Protocol 6 )
processing of pre -rANA as descr ibed in
. 10 mM CaCI2
Protoco! 7
• M icro coccal nuc lease lBRl ; 1 m g /m ll e 60 mM EGTA. pH 8.0 • Equipm ent an d reag ents f or stu dyi ng
• Mod if ied 5 x S-l 00 proc essi ng buffer
10 .4 2 M KC I. 40 mM Hepes-KOH. pH 7.9 , 16 mM M gCI 2 • 33 % glycerol. 2.6 mM DTT . 0 .46 mM EDTA I
Met hod 1. Mix 6 ~I of the 5 -' 00 processinq extract with 1 ~ I of '0 mM CaCl, and Incubat e th e mi xture for 30 sec at 30 °C . Cool the react ion m ixture on ice . 2 . Add 2 ~I of 1 mg /ml micrococcal nuclease and incubate the ex tract for
90 sec at 30 °C. Coo l the tube on ice . 3. Add 1 ~I of 60 mM EGTA and incubate for 30 Sec at 30 °C. Again cool th e extract on ice . 4. Test th e pro cessing act ivit y of t he t reat ed ext ract as describ ed in Protoc ol
7, excep t replace the 5 ~I of 5 -100 process ing extrac t wit h the 10 ~I of
t he nuclease·treate d ex t ract (step 3 above) and subs ti t ute modified 5 x S-1 0 0 processing buffer.
' From ref. 34 .
5.1 Mic rococc a l nuclease di gestion D ata from an exper iment in which the mouse S-IOO extract was treated with micrococcal nuclease and then tested for its ability to process pre-r RNA are shown in Figure 6A.
If treatment of a processin g extr act with micrococcal nuclease (but not a 'mock' treatment) destroys the processing activity, and this is not restored by the ad dit ion of non-specific RNA , it can be pres umed that the processing reaction has a requirement fo r one or more specific nucleic acids . Lack of sensitivity to DNase I can furthe r suggest that the nucleic acid is RNA. Note that an absence of an effect of micrococcal nuclease is not proof that an RNA is not required; a required RNA may not be accessible to the nuclease. Proto col 14 describes the tr eat ment of the mou se S- IOO processing extract with micrococcal nuclease. Th e advantage of micrococcal nuclease for this type of analysis is that it is a Ca' +-depe nde nt nuclease. This allows it to be inactivated once digestion of a putative nucleic acid is complete. using an excess of the Ca' +-chelator EGTA, so that the processing activity of the treated extract can then be tested .
OligonUcleotide-directed RNase H treat ment can be employed to determi ne wh ether a particular RNA is requ ired for a processing event. RNase H digests RNA on ly in an RNA :DNA hybrid . Therefore, hybridization of an ? "godeoxyn bonu cleotide com plementary to a specific RNA thoug ht to be Involved in processing cou ld render that RNA susceptible to RNase H digestion ~ f thi s RNA is digested by RNase H (checke d by Northern blot analysis) and s an essential component of the processing extract, the extract will then be unable
164
165
5.2 Oligon~cleotide:directed RNase H digestion of cellular RNAs involve d 10 pre-rRNA processing
5; Rib osomol RN A process ing in ve rtebrates
A
-
product
,-,
..
ce
,-. , t
- 12
Bg
Ca.E
~
Co
~
UJ 01. os ~
.0
. ~ ~
~ ~
-- ~
"'"
<;
2
R.....
Cathleen Enright and Barbara Sollner-lVebb
~
, •
0
~
5
Figure 6 . Sens itivity of the mo use S' ETS processing reaction to micrococcal nuclea se or RNase H treatment of the mouse $., 00 processing extract . (AI M ic rococca l nu cl ease sensitivity . Samples of the mouse $ -100 processing extract were treated as indicated in the
table below the autoradiograph . At the start (( =01, samples received CaC12 lea; lanes 3 and 5) or CaC12 and excess EGT A (Ca ,E; lane 41. or no addition (lane 21. After 4 min It = 4) ,micrococcal nuclease (MN) was added (lanes 4 and 51. After another 4 min It= 8), excess EGTA was added to all samples lacki ng this reagent . Finally after a furt her 4 min (t = 12 1. 32p . labelled pre-r RNA substrate was adde d to all the samp les to test t heir pr oce ssing ab ility . Th e pre ·rRNA (prec ursor) and processed pr oduct (product) were sepa rated elect rophoreti celfv . Lane 1 shows the untrea te d pre- rRNA. (A dapted from ref. 381( 8) Oligonucleotide-d irect ed RNase H sensitivity . Th e 5-100 extract was pre-incub ate d with RNA se H in t he pr esence of no olig on ucleotide (-oligo; lane 2 ), an oligonu cleot ide co mp lementa ry to U2 sn RNA (U2-L 15; lane 3), an oligo nuc leotide compl em entary to residues 64 - 79 of U3 snRNA (6 4-79 ; lane 4l, or an oligonucleoti de co mplem entary to residu es 80 -9 6 of U3 (80 -96; lane 51. All sam ples were then t reat ed with DNase I followed by incubation wi t h radiolabelled pre-rRNA proc essing substrate. Only the oligonucl eoti de ta rgeted on residu es 64 -79 of U 3 snRNA result ed in RNase H digestion and loss of processing activity as show n by the absence of the (smaller ) rRNA product . Neither the non -specific U2 oligonucleotide nor t he U3 oligonucleotide target ed to the double-stranded reg ion of U3 snRNA (residues 8D -961 supported RNase H dig est ion. (Adapted from ref . 38 ).
to process the pre-r RNA. Oligonu cleotides complementary to prominent singlest rande d regio ns of the RNA are th e most likely to be successful in direct ing digestion of this RNA in the presence of cellular co mponents. For insta nce, of eight anti-V3 oligonucleotides tested, only one, corresponding to residues 64-79 of mouse V3 RNA , has been successful in promoting RNase H directed removal of V3 RNA activity from the mouse S-IOO extract (38).
5.2.1 RNase H digestion of cellular RNAs in processing extracts The method is described in Protocol 15 and in Volume I, Chapter 4 (Section 6). Following RNase H treatment, assay an aliquot of the extract by Northern blot analysis to monitor the RNA digestion and test another aliquot for alt ered processing efficiency after incubation with a l2P-labelled pre-rRNA substrate. It is important to demonstrate that the loss of processing activity occurs following 1 66
digestion of the RNA of interest, but not following parallel treatments with con trol oligonucleotides. It should also be remembered that an oligonucleotide may fortuitously allow destruction of an unanticipated RNA and that an ap parent ly negative result is hard to interpret unless one can be certain that the RNA destruction was sufficiently complete to assure detection of any potential effect on processing .
Protocol 15. Oligonucleotide-directed RNase H digestion of RNA in the mouse S - l 00 processing extract'
Equipm ent and reagents e 5 -1OOpr ocess ing extract and D.2 M OTT stock so lution , Protocol 6
e 5 x RNase H bu ffer (D. 5 M KC I, 40 % g lycero l, 9 .76 mM OTTb ) • RNase-free DNase (Wort hi ngt on; 2-5 U /~I)
• Oligodeox yribonucleotide complementary to the target RNA sequence {1 jJg/jJll
e Equipment and reagents for studying pr oce ss ing of pre-rRNA as described in Protocol 7
. RNase H (S Rl; 2 U/IJI)
Meth od 1. Set up the following react ion mixture in a microcentri fuge t ube on ice: • S·l 00 processing extract 10 ~ I • oligonuc leo ti de
1 ~1
e 5 x RNase H buffer e RNase H
3 ~I 1 ~I
2 . In cu bat e the mixtu re for 10 min at 3 7 °C , then fo r 20 min at 30 °C and th en place it on ice. 3 . Add 1 ~I of RNase-free DNase and incubate the mixture fo r 3 min at 37 0C
followed by 7 min at 30 °C.
4 . Rem ov e 6.41J1 f or Northern blot analysis of the t arget ed sn RNA (see
Volume I, Chapters 2 (Section 3.61 and 4 (Section 7.511. 5 . Use the remainder of the reaction mixture from step 3 (9. B ~I ) t o test the process.ing activity of t he extract as described in Protocol 7. For the processing assay . dilu te the remainder of the reaction to 151J1 with 10 fmol J2P-lab ell e.d pre- rRNA and Al P to yield a free magnesium ion ~once~tratlon o.t - 0 .5 mM. Although processing does not require AlP , mcludinq AlP In the reaction as opposed to adjusting the MgCI to 0 . 5 mM results in sharper gel bands . 2 • From ref . 38 . b Add OTT just before use from the OTT stock solution .
167
Cath lee n Enright and Barba ra SoJJner-\\'ebb
5: Ribo somal RNA proce ssing in vertebrates
Figure 68 shows the results of an experi ment in which RNaseH ~vas used to demonst rate that U3 snRNA is required for 5 ' ETS processmg m t e mouse S-IDO extract (38).
5.2.2 Oligonucleotide-directed RNase H digestion of RNAs in micro-injected oocytes
• Modif ied Barth ' s buffer tSB m M NaC!.
1 mM KCI. 0 .8 2 mM MgSO •• 0 .3 3 mM calcium nitr at e, 0 .4 1 mM CaCI2 , 10 mM Tr is -He !. pH 7 . 5. 20 mg /l pen ici llin and 20 mg /l streptomycin )
• I a _J 2p I UTP (leN: 800 Ci /m mo l; 10
~Cj / ~I)
(0. 2 M
Tr is- He l,
pH 7.5 . 0 .6 M NaCI. 20 mM EDTA . 2% 5 0 S, 0 .2 mg lm l proteinase K ) • Phenol :ch loroform
( 1: 1 ,
v /v l
as
m
Pro to col 1 e M icro-inject ion equ ipmen t for Xe nopus oocvtes (see ref. 16 )
Method 1 . Inject 25 -50 nl of the oli gonucleotide solut ion (50 ng l int o the nucl ei of at least 20 oocyt es. 2 . Incubat e the injecte d ooc ytes for 11 -20 hat 20 ° C in mod ified Bart h' S buffer .
3. Repeat step 1. 4 . Incubate the oocyt es f or 6-14 h as in step 2 .
Continued 16 8
9 . Incubate the homogenate for 30 min at 50 °C. 10 . Ext ract t he nucleic acids wit h phenol:chloro f orm and then w it h chlorof orm atone ; finally ethan ol pre cipitate the extracted nu cleic acids
Isee Protocol 1 steps 6-9 1. 11 . Dissolve th e nucleic acid in H 20 and sto re it at - 70 ° C . 12. A ssay th e RNA prepar ed f rom the m icro-inje ct ed oocyt es f or alte rations in th e abundan ce of ind ividual rRNA pr ocessing int ermediates . Th is is co nvenient ly don e by elec t rop horesis (see Protoc ol 7, ste ps 5- 9 ). Alt ern at ively, primer exte nsion , nuclease S 1, or Northern blot analyses (as descr ibed in Vo lume I, Ch apter 2, Section 3 ) can be us ed to assess the ex t ent of remo val of the targ et ed snRNA .
5.3 Immunological analysis of polypeptides re quired for pro cessin g
Equipm ent and reagents bu ffer
7. Incubate the oocyte s f or 6 -14 h as in ste p 2 . 8 . Homogenize th e oocytes in ext raction buffer .
.. Based on ref . 16 .
Protocol 16. Analysis of nucleic acid components necessary for rRNA processing us ing oocytes micr o-inject ed wit h oligon uc leotides'
• Extract ion
5 . Mi x sufficient [ a _32p ] UTP wi th the 1- 2 mg /m l oligo nucl eoti de solut ion t o give a final concentr ation of 5 ~C i!~ 1 in t he 32P·labelled UTP /oligo· nucleotide m ixture. 6 . Inject 50 nl of the 32P-labelled UTP/oligonucleotide mi xture into the nu cleus of each ooc yte .
The endogenous RNase H present in Xen opus laevis oocytes has been exploited to demonstrate the involvement of U3 snRNA on the processmg of t.he cell ~lar pre-r RNA (16). In this case, the U3-specific oligonucleotide was injected into the oocyte nucleus and the endogenous RNase H performed the selective digestion . It was concl uded that the processing which occurs. at the 5 ' -end of the 5.8S rRNA region is depend ent on the presence of U3. in the ooc yte. Preparation of the oocytes an d micro-i njection of material into the oocyt e nucleus are specialized techniques which cannot be covered m the necessary d: ta'i in this chapter. For detailed instructions, refer to ref. 16. Protocol 16 descn bes the subsequent incubati on and processing of the oocytes .
. Oligodeo xyribo nucleot ide ( 1- 2 mg /m ll complement ary t o the ta rget snRNA
Protocol 16. Continued
In addition to detecting sn RNAs involved in processing, the presen ce and role of snRNP prot eins can also be mon itor ed and analysed. At present , there are no reported antibodies specific to individua l nucleolar snRN Ps. Ho wever, antibodies to fibrillarin , an - 34 kDa polypeptid e common to the U3, US, U 13, and UI4 family o f snRNPs, are ava ilable (reviewed in ref. I ). Th ese ant ibodies have been used both to immunodeplete the mou se S- IDO extract of 5 ' ETS processing activity and to irnmunoprecipitate the processing complex that forms on pre-rRNA add ed to a processing reaction (34). For details of these procedur es, see ref. 34 and Volume I, Ch apter 5, Section 6.
6. rRNA processing associated with transcriptional terminati on Mention sho uld also be made of two other cases o f rRNA proce ssing which have been found associated with , or part o f, transcription termination in mou se an d Xenopus laevis , Thes e ar e summarized below (Sections 6.1 and 6.2), but proto cols for their investigation are not reprod uced here beca use these events seem better studied in the context of polymerase I transcription termination than of rRNA processing itself. Furthermore. little is yet known about the 169
5: Rib osomal RNA processing in vertebrates proc essing aspect of these events, and so it seems advisable that the investigator sho uld consult the entire original article prior to co nd ucting stu d ies on th ese topics .
6.1 3 ' end-processing in the mouse Coupled to the actual polymerase I transcript ional termination event is a very rapid processing of the 3 --end of the rRNA (17), which yields what is generally reco gnized as the primary transcript. It appears that the two events can be pa rtly uncoupled when cat alysed in vitro using fractionated nucleolar extract s or mutated rRNA templates. The fa ct that the 3 ' -end o f the processed prod uct appears to map to the beginning of the fir st stem-loop stru ctur e preceding the termination signal rather than to a particular nucleotide sequence (39) suggest s the invol vement of a 3 - exonuclease in the processing.
6.2 3 ' end-processing in Xenopus Jaevis In Xenopus laevis, the expected site of termination for rRNA gene transcription differs by one nucleotide from the funct ional termination site in X. borealis (18). This mutated X. laevis site , called T2 , directs formation of the actual 3 ' -end by degradation of downstream r RNA sequences (18). There appears to be a requirement for a functional downstream termination site in order that T2-directed 3' end-formation occurs. Further studies are needed to determine the relationship, if any, of thi s processing to the terminat ion-associated 3 ' endprocessi ng in the mouse desc ribed in Section 6.1
Acknowledgements The authors thank Drs Ted Gurney and Tim Nilsen for allowing them to present their data and Ms Linda Mi ller and ~Is Mary Bonkowski for preparatio n of the manuscript. This work was supported by ASC grant N-731B (to B. 5.-\\ ' .) and NIH postdoctoral fello wship (to C.E .).
References 1. Soliner-Webb, B., Tyc, K., and Steitz, 1. (1993). In Ribosomal RNA : structure, evolution. processing and fun ction in protein sy nthesis (ed . R. Zimmerman and A. Dahlberg), CRC Press, Boca Raton, FL. (In press). 2. Musters. \V., Venema. J ., van der Linden, G., Van Heerikhuizen, H.; Klootwijk. J., and Planta, R. J . (1989) . Mol . Cell. Bioi. . 9, 551. 3. Musters, W.• Planta , R. J ., van Heerikh uizen, H., and Raue, A. (1990). In The ribosome-s-structure, fun ction and evolution , (ed . W. E. Hill et af.), p. 3435. Am. Soc. Microbiol ., Washington DC. 4. Li, H. V.. Zagorski, J ., and Fournier, ~1. J. (1990). Mol . Cell. Bioi.. 10, 1145. 5. Hughes, J. ~ 1. X. and Ares, M. Jr. (1991). £MBO J .. 10, 4231. 6. Tollervey, D., Lehtonen, H., Carmo-Fonseca, M., Hurt, E. C. (1991). £ MBOJ., 3, 573 .
170
Cathleen Enright and Barbara Sollner-lI'ebb Ware, V. c.. Renhawirz, R., and Gerbi, S. A. (1985). Nilci. A cids Res. , 13, 3581. Maden, B. E. H. (1990). Prog. Nilci. Acid Res. and M ol Bioi 39 241 ~I i ller , K. and Sollner-Webb, B. (1981). Cell, 27, 165. ' ., , . Kass, S. , Craig, N. , and Sollner-Webb, B. (1987). Mol. Cell Bioi 7 ' 891 Schumard, C. M. and Eichler, D. C. (1988). J . Bioi. Chem ' 263 "193 ;6 . 12. Hann~n , G. J ., Maroney, P . A . , Branch, A ., Henen field , B. ROben son 'H D and NIlsen, T. W. (1989). Mol. Cell. Bioi. , 9, 4422. ' . ., 13. Vance, V. B., Thompson. E. A.. and Bowman, L. H. (1985) Nucl Acids Re 13, 7499. '. s.. 7. 8. 9. 10. II.
i'-,
14. Schumard, C. M., Torres, c.. and Eichler, D. C. (1990). Mol. Cell. Bioi. , 10. 3868. 15. Schumard , C. M., Torres, C., and Eichler, D. C. (l99t). ,\[01. Cell Bioi 11 593 16. Savino, R. and Gerbi, S. A. (1 990). £MBO J 9 2299 .." . t7. Kuhn, A. and Grummt, I. (1989). Genes and 'De~ 3 i 24 18. Labha n , P. and Reeder, R. H. (1990). Genes and' 4: 269. 19. Wellauer, P. K. and Dawid, I. B. (1974). J . Mol. Bioi. , 89, 379. 20. Wellauer, P. K. and Dawid, I. B. (1974) . J . Mo l. Bioi. , 89, 397. 21. Forsheit, A. B., Davidson, N.. and Brown, D. D. (1974). J . Mo l. Bioi 90 301 22. Gurney, T. Jr. (1985). N ucleic A cids Res., 13, 4905. ., , . 23. Ausabel, F. M., Brent, R., Kingston, R. E., Moore, D. D. Seidman J G Smith' .J ' A ., an d Stru hi, K. (I 991). Current protocols in molecu"lar biologv. Wiley . ., . Interscience, New York, NY. 24. Gurney, T. Jr ., Sorenson, D. K., Gurney. E. G., and Wills, N. M. (1982) Anal Biochem . , 125, 80. . .
'0;,..,
25. 26. 27. 28 · 29. 30.
Maser, R. L. and Calvet, J . P. (1989). Proc. Natl Acad. Sci. USA 86 6523 Stroke, I. L. and Weiner, A. M. (1989). J . Mol. Bioi., 210, 497. " . Miller, O. L. Jr. and Beatty, B. R. (1969). Science 164 955 0h ' Y , , . ~ elm, . N., Miller, O. L. Jr., and Beyer, A. L. (1985). Cell 43 t43 YIp, M. T. and Holland, M. J. (1989). J . Bioi. Chem., 264, 4045. ' . Peanie, D. A. (1979). Proc. Natl Acad. Sci., USA , 76, t760. i~ ' Craig, N., Kass, S., and Sollner-Webb, B. (1991). Mol. Cell. Bioi. , 11,458. · Dignam, J. D., Lebowitz, R. M., and Roeder, R. G. (1983). Nucleic Acids Res II , 1475. .. 33. Craig, N., Kass, S.. and Sollner-Webb, B. (1987) . Proc Natl A cad S . . CI . , USA., 84, 629. ., 34. Kass, S. and Sollner-Webb, B. (1990). Mol. Cell. Bioi. , 10, 4920. 35. Brad ford, M. M. (1976). Anal. Bioehem 72 248 36 8 ., , . · artholomew. B.. Dahmus, M. E., and Meares, C. F (1986) J Bioi Chern 261 14 226. . .. . '" 37. 38. 39. 40.
Laemmli, U. K. (1970). Nature, 270, 680. Kass, S., Tyc, K.. Steitz, J . A., and Soliner-Webb, B. (1990). Cell, 60, 897 Henderson, S., Ryan, K., and Sollner-Webb, B. (1989). Genes and Dev 3 '2 12 Mougey. E. B., Pape, L. K. and Sollner-Webb B. (1993) '1101 Cell Bi~i 1'3 (I . press). • ., . . ., n
41.
~~ougey, E. B., O'Reilly, ~I., Osheim, Y., Miller, O. L.. Beyer, A. and SollnerII ebb, B. (1993). Genes and Dev., 7 (In press).
171
Processing of transfer RNA precursors CHRI S L . G RE E R
1. Introduction
1.1 General methods Th is chapter covers practical strategies and methods for studying the biochemistry of transfer RNA processing. Although processing of nuclearencoded t RNAs in yeast provides th e primary focus for the chapter, essentially all of the strategies and many of the experimental methods apply equally well to the tRNAs of other organisms. Thi s subject is generally amenable either to biochemical or genetic ana lyses. The methods in thi s chapter emphasize the biochemical approach and include the development of in vitro assay systems and characterization of reaction intermediates and products. Genetic methods for studying pre-t RNA substrate requirements are reviewed in ref. I and genetic app roaches to the identification of trans-acting components are reviewed in ref. 2. Cer tain genera l meth od s applicabl e to tRNA processing, including fra ctionat ion of cellular RNAs, characte rization of RNAs, and analysis of RNApro tein interactions, are described elsewhere in this book (Volume I , Chapters I, 2, and 5, respectively) . Th e study of t RNA processing is, in many ways, unique within the RNA processing field. This assertion is based on both the unique physical and function al pro perties of t RNAs and the degree of complexi ty characteri stic of tRNA biosynthetic pathways . Amo ng the unique properties of transfe r RNAs are their small size, stability, and relative intracellular ab unda nce. These features make tRNAs attractive experimental subjects among cellular RNAs and the result has been the accumulation of a wealth of information regarding tRNA structure, function, and biosynthesi s (see refs 3 and 4 for general reviews, and the compilation of tRNA and tDNA sequen ces in ref. 5).
1.2 tRNA structure and function Th e tR NAs function pri ma rily in mR NA translation. Evolutionary cons traints imposed by th is crucial function are respon sible in part for the high degree of conservation observed among tRNAs (reviewed in ref. 6). With the exception of certain mitochondrial examples, tRNA s share extensive co mmon features
Chris L. Greer
6: Processing of tron sfer RNA precursors
,.
e-
8 -Leeder - -8-,
=~
li
-
70
-
••
@-Tr a -i l er @
Arm
t
8
D Acm
@
B
'-'0
@@
("i'4'L ~I";"";'\Q\.!.J ~~~ II
@
@<§l@@@]
I
I
,
I
I
®
I
@@@@§]
I
_(22'x23)@@
®§l ~
@
~ §]
T Ar m
§]
@eS@)-@@J @@e\) @
@
@ @
@- @
8
® -@
Var i ab l e
@ -@
®
AC
Arm
Arm
@ ®
§]
@@~ Intervening 0,..
@l -
-
-a
Sequence Figure 1. Consensus pre-tRNA structu re. The tRNA number ing scheme sho wn he re.is base~ on t hat described in ref. 3 . Si tes of va riation in nuc leo tide sequence leng th occu r In t he 5 leader , 3 ' trailer . and intervening sequences . Invariant bases are represented by. boxes and semi- invariant positions by ellipses. Pos itions w ith in the intervening sequence , 5 lead er and 3 ' trai ler segments are indicat ed by a co lon . minus or plus sign . respect ive ly , and are nu m bered from the adjacent axon posit ion . Solid lines denote segments o f v aryi ng length, N. The AA arm and AC arm mark t he aminoacyl acceptor and anticodon arms , respectiv ely. A rms common ly containing dihydrou ridine and riboth ymidine mo difications are marked D arm and
T arm , respectively .
struct ure greatly increases the number of distinct processing reactions required for the synt hesis of an y single tR NA species and also requires a rather large nu mber of diverse proce ssing pathways to produce the complete cellular repertoire of tRNAs . At the level of secondary structu re, all of these tRNAs for m a conserved set of stem and loop structures which ca n be a rranged in the traditiona l clove rlea f configuration . Thi s stru cture and the corresponding nomencl ature a re summarized in Figure 1. Significant length variation occurs only within the D-Ioop and variable arm segment s with the lengths of the remaining segments being highly conserved. At the level of tertiary structure, all tRNAs are thought to adopt an L-shaped con figuration (reviewed in refs 9, 10). The sma ll arm of the L-shaped structure is formed by the aminoac yl acceptor (AA; Figure 1) and T-stem helices, while the long arm consists of the D-stem and the anticodon stem and loop (AC; Figure 1). T his conserved tertiary structure is stabilized both by stacking interaction s between conse rved helices and by tert iary interactions between conserved bases of the 0- and T-Ioop segments, and the variable arm and D-stem region s. This high degree of conservation in both primary sequence and threedimensional structure has two important implications for tRNA pro cessing. First , general proces sing enzym es may recogni ze their sub strates through conserved features within the mature , tRNA- like portion of the precursor, potentially at a distance from processing sites. Second, unique proce ssing step s may require fine distinctions among tRNAs, pot entiall y involving discr ete positive and negative discriminators analogous to the determinants of tRNA ident ity in aminoacy l synthetase interactions (1 1).
1.3 Overview of processing pathways The primary transcript s of tRNA genes in both prokaryotes and eukarvotes must und ergo a number of processing react ions to prod uce the mature, f unctional form of the molecu le (12, 13). These reactions fall into three general categori es: • removal of extraneous sequences • addition of non-encoded bases • introduction of modifications to the base and ribose moieties of specific nucleotides
of primary, secondary, a nd tert iary struc ture. The primary sequence.of mature t RNAs varies from 60- 95 nt in length with 13 inva riant and 12 sernl-invar tant positions. Extensive post-transcriptional modification is comn:on wit~ .as many as 20 070 of the nucleotides of a mature tRNA species present In modified form (reviewed in refs 7, 8). A total of 52 different mod ifications have been char acte rized thus far, with some co mmo n to a majority of tRNAs and others unique to one or a few species (see the compilatio n in ref. 5). This facet of tR NA
The temporal order of these processing events has been examined in a number of prokaryotic and euka ryotic systems. In all cases , there appears to be a pre ferre d seque nce of events for a given precu rsor since on ly a subset of the possible intermediates has been detected. For example , processing in Xenopus oocytes is tho ught to start with trimming of the 5 ' -end followed by maturation of the 3' -end, splicing, and finally tra nspo rt out of the nucleus (14). The addition of individual base modifi cation s occurs at distinct stages in this size-maturation process (15). While a similar order may apply to some precursors in yeast, others
174
175
6: Processing of transfer RNA p recu rsors may ut ilize diffe rent path ways (16). Whether substrate specificity , functio?al compartment at ion. kinetic constraints , or a co mbinat ion of these facto rs de f~ ne the processing pathways for individual tRNAs remain among the most pressing questions in tRNA biosynthesis.
1,4 Biochemical analysis of tRNA processing The rest of this chapter is divided into three experimental sections intended to reflect the approximate tem po ral order one might follow in initiating a biochemical study of a t RNA processin g reaction. Section 2 describes the preparation of tRNA precur sors for use as substrates in processing reaction s. Precursors may be purified from whole cells or be prepared by In vitro transcription . Section 3 includes the preparation of cell-free extracts active in transcription and specific processing reactions . Examples of the preparation of complex nuclear extracts and purified splicing enzyme fra ction s are provided. Section 4 describes the development of tRNA processing assays a nd prov ides example s of assays for measuring changes in the size and sequence of the precursor or for detecting the introduction of specific nucleotide modificatio ns.
Chris L. Greer Proto col 1. Preparation of nuclease-free RNA a nd glycogen ca rriers Reag ents . 2.0 M so dium acetate buffer. pH 5.0 e TE buffer (10 m M Tris-He!. pH 7.5 . 1 mM EDIA )
- 20 % 50S e S.O M NaOH e Proteinase K (20 U /mg protein; Sigma .
P 65561
• Pheno l : ch lo ro fo rm : isoam y l alcoho l
125: 24 : 11 - St eril e w at er; t reat deion ized . d ist ille d water with 2 drop s pe r lit re of DEPC, autoclave 1 h , 1 20 ° C. 16 p.s .i.
• Either gl ycogen (oyst er glycog en Sigm a Type II, G 8751 J or unfract ionat ed RNA fr om yeast (Sigm a T ype VI R66 2 5 )
M etho d 1. Dissolve the glycogen or RNA in TE buffer to 20-100 mg lml final concent ration . Note th at thi s fo rm of ANA must be carefu lly neutralized to dissolv e it ; add 5 M NaO H dropwis e with stirring . Use ind icator st rips t o check th at th e solu tion is at pH 7. 5 .
2. Prep aration of pre-tRNA processing substrates
2 . Add SDS to 0 .5% . pro t einase K to 1 m g /m l and incubate the solut ion for 30 mi n at 55 °C.
2.1 Introduction
3 . Extract th e solutio n wit h 1 vo l. of ph enol : ch loroform : isoam y l alcohol. Centrifu ge at 2000 9 f or 5 min to sep arate ph ases. Repeat t he extracti on of th e aqueou s pha se twi ce more and fi nally ex tract it w ith 1 vot. of eth er.
A biochemical approach to the study of tRNA processing relies heavily on the provision of substrates of high purity and defined structure, and usually in large amo unts or at high specific radioactivity. Additional requirements are the facility for manipulating substrate sequence or structure, the pote ntial for site-specific labelling, and the ability to include or omit one or more of the specific processing steps in preparing the substrat e. Two genera l approaches are availab le for the preparation of processing substrates: • expression in vivo followed by isolation from whole cell RNA fra ctio ns • expression in vitro using homo logous or heterologous transcription/processing systems Each of these approac hes has distinct advantages and limitation s which should be considered when choosing a n appropriate experimental strategy. The se are described along with the methods themsel ves in the sections below .
4. Add 0 .1 vo l. of 2 M sodi um ac eta te buff er and 2 vo l. of ethan ol. Keep th e m ixture on crushed dr y ice f or 1 h t o pr ecip it at e th e RNA or glyc ogen . 5 . Centrifug e f or 15 min at 12 OOO g , 4 0C. 6 . A ir dry th e pellet , resu sp end it in 0 . 1 vo l, of 2 M sodium ac etate buffer and repeat th e et hano l precipitat ion (steps 4 and 5) t w ice. ' 7 . Resu spend th e dried pellet in ste rile wa t er t o 20 mg/ml f inal concentration and store t he sample at - 20 °C in 1 ml aliquo ts .
2.3 Prepara tion of tRNA substrates from whole cells
Labelled RNA substrates must often be concentrated from dilute solutions by precipitation with ethanol. At RNA co ncentrations less than 5 pg/ ml, precipitation may be ineffici ent un less an appropriate carrier is added. For samples to be subjected to nuclease digestion (see Section 4.3) , the addition o f a know n amount of RNA carrier is appropriate. For samples to be used In processing or post- labelling reaction s, glycogen shou ld be added as carrier. Pr otocol I describes the prep aration of nuclease-free car riers for these purposes.
Th is a pproach is especially ap plicable where o nly limited info rmation exists regard ing the part icular subst rates or processing reactions to be stu died and where an initial characterization is required . Substrates prepared in this way may have already undergone some processing steps, such as the introduction of base modifi cations, which might be required for the processing reaction to ~e studied. If so, analysis of the partially-processed species may prov ide insights mto the nature of the naturally-occurring substrates . Disadvant ages to this approach include the difficulties in obtaining highly purified fractions th e limited strategies available for specific radio labelling of substrates either during In ViVO synthesis or after their isolation, a nd the low levels of tRNA precursors nor mally present in cells. This latter difficulty may be partially overcome through th e use of mutant cell lines or alte red growth condition s which result in the
176
177
2.2 Preparation of nuclease-free carriers
6: Pro cessing of transfer RNA precursors
Chris L. Greer
accumulation of partially-processed transcripts, An example of this approach relies on the phenotype of the rnal-I mutation in yeast (17) : the accumulation of intron-containing tRNA precursors upon shift to the restri ctive temperature. This method (ba sed on ref. 18) is described here in two parts : growth an d labelling of the yeast cells (Protocol 2) and the extraction and isolation oftRNA precursors (Protocol 3) , T he procedure relies on phenol extraction of nucleic acids and their pu rification by gel electropho resis; gene ral reviews of the se met hods can be fou nd in refs 19-2 1. Alternative procedures for t RNA purification incl ude conve ntional chromatog rap hy and HPL C as sum marized in refs 22 and 23, Prot oc ol 2 . Grow t h and labell ing of c e lls
#42245 1 (
3 . Collect th e cells by centrif uging at 7000g f or 5 min at 4 °C . Wash th e cell pellet by res us pen di ng the cells in 50 m l of low-phosphate labelling me di um and centri fuging as bef ore. 4 . Resu spen d the w ashed cells in 200 ml of fresh low -phosp hat e labelling m edium and inc ubat e w ith shakin g for 3 h at 24 0C.
5. Add 50 mCi of [" P]orthophosphate . Incubate for 5 min at 24 "C . 6. Carry out a temperature shif t by transf erring the cult ure t o a shaking water bath at 3 7 °C . b In cubat e w it h shaking for 30 min . 7 . Ch ill the c ult ure f or 5 m in in an ice-wat er bath . re mo ve all th e supernat an t from t he cell pell et.
e Saccharom yces cerevisiee, st rain M304; mal /mal , adel lADEl. ade2 /ADE2 . his7 lHIS7 , l ys 21L YS 2. tyrl lTYRl fA m eric an Typ e Cu lt ur e Collect ion •
2 . ~dd t he 5 ml overnight culture to 200 ml of fresh YPD medium and Incubate at 24 °C to A 6 60 ~ 1.0.
8. Collect the cells by centrifuging at 7000 9 for 5 min at 4 "C , Carefully
Reagents
32
Protocol 2 . Continued
P j -ortho phosphat e IHCI-f ree. c arrier-
free . 40 mCi/ml; Amersham cat . no . PBS
431 e Bectopeprone. yeast extract . and yeast nit rogen base with amino acids (Oifeo; cat alogue numbe rs 0118-01 -8. 0 127 -0 1-7 , and 0392- 15 , respect ively )
. 20 % glucose ; ste rilize 4 5 min . 120 ° C , 16 p.s .i. • Succin ic acid (solid)
. ' .0 M Mgel,
• Sterile water (Protocol 11 e YPD (rich med ium )", For 1 lit re cult ure. add 109 yeast extract , 20 g Bacto peptone t o 900 m l o f de ion ized water; sterilize by autoclaving 11 h, 120 oC , '6 p.s .i. ): add 100 m l o f 20 % glucose just before inoculati ng w ith yeast
e VPD agar plates (as f or li quid med iu m abo ve w ith the addition o f 20 9 agar per litre of med ium ) e 10 x succinate sto ck sol ution. Disso lve 100 g succ inic acid . 50 g NaOH in 1 lit re of de ion ized water , adjust the pH to 5 .7 and autoclave (1 h, 120 ° C. 16 p. s .i.} e 10 )( YP stock so lution. D issolve 100 g Bactopeptone , 50 g yeast ext ract , 67 9 yeast nitrogen base with amino acids in 750 ml of deionized water . Add cone . NH 4 0 H (14 .8 Ml slowly to pH 10. Add 100 m l of 1 M M gCI2 and incubat e at 4 ° C, 8- 10 h to precipi ta te fr ee phosphate . Add 50 ml of 1 M M gCl 2 and incubate 4 h at 4 ° C. Remove the precipi tate by centri fugat ion at 3000 9 f or 15 min . Carefu lly decant the supernatant and add conc . HCI to pH 6 .5 . Ad just th e final volume t o 1 li t re w ith wate r and autoclave 11 h, 120 ° C, 16 p .a.l .I e Low -phosphat e labelling medium . M ix 25 m l of 10 )( suc cinate st oc k so lut ion, 25 ml of 10 x YP st ock solut ion, 25 ml of 20 % glucose , and 175 ml of ster ile wa ter
Me th od NB The labelling involves the use of a large amount of care in its handling and disposal.
32 p
radioactivity. Take
1 . Sta rt a cultu re of yeas t st rain M 304 by inoculating 5 ml of YPD medium w ith a colo ny fr om a fr esh YPD plate . Gro w overnight with shaki ng at
24 -c .
" A dditi o nal in fo rmat ion rega rding yeas t growth media can be found in ref 24 b Use of a water bath is essential to ach ieve the rap id inc rease in temperature necessary for induction.
Protocol 3 . Extr a c t ion and purifi cat ion of la be lle d t RNA precursors from yeast cells
Equipm ent and reagents e D EA E- c ellu lose co lum n ; W h at m an DE- 5 2 (40 57 · 200), 0. 5 ml eq ui librated in co lumn loadi ng bu ffe r and packed into a disposable colum n or sy ringe e O ufk -Sep columns (disposable. lsolab. co de OS -V) e Sterile water !protocol 1 ) • NES buffer (0 .1 M sodium acetate , pH
5.7, 20 mM EOTA, 1% 5051
e First -d im ensional ge l( 10% acr ylam id e, 0.3 3% bisac rylamide , 4 M urea in TB E buff er; 31 x O.2 cml and sec ond-dimen sion al gel 120 % acrylam ide, 0.66 % bisacrylam ide , 4 M urea in TBE buffer; 3 1 x 4 1 xO.2 cm) plus extra second di mensiona l gel m ixtu re. and electro phoresis equipment (see Volume I Chapter 1, Protoco l 8 1 .
e 2.0 M ammon ium acetate , pH 7.0
e Ammon ium persu lphate 110% aqueou s solution , mad e fresh da ily )
e Redistilled phenol , saturated with 20 mM Tris-H CI. pH 7 .4
• TEMEO IIBI Inc.. 701201
e Colum n loading buffer (0 .2 M NaC I. 10mM Tr is- HCI. pH 7 .4 )
• Glycogen carri er. prepa red as in Protocol 1
e Column elution buffe r (1 .5 M NaC l, 10mM T ris- HC I, pH 7 .4 )
e Get elution buffer (0 .3 M NaCI. 20 mM Hepes- KOH , pH 7 .4 , 1 mM EDT A , 1% phenoll
e TBE buffer 190 mM Tr ia- bo rate. pH 8 ,3 ,
2.5 mM EDTA)
. 10 m M Tris - HCI, pH 7. 4
e Gel sample bu ffe r (20% suc rose , 7 M urea , 0 .1 % bromp henol blue , 0 .1 % xylene cvanon
e Yeast ce lls labelled with [ J2 Pj or thophosphate from Protoco l 2
Continued
Contin ued 178
179
6: Processing of transfer RNA pr ecursors
Chris L. Greer
Protocol 3 . Continued
Prot ocol 3 . Continued
Me thod
19 . Disassemble the electrop hor esis apparatus leavi ng th e gel on t he glas s backing plate. Wrap the gel and plate in plastic cli ng film and mar k the sides and corners w it h rad ioactiv e or fluore scent ink fo r late r alignment
1. Resuspend the yeast cell pellet in 4 ml of NES buffer. Add 4 ml of phenol and mi x by vortexing . Incubate at 55 °C for 30 min , vortexing at 10 min int erv als.
2. Separ at e the aqueous and phenol layers by centrif ugat ion (20 00 g , 5 m in at room temperature ). 3. Set aside the aqueous fract ion . Add 4 ml of fr esh NES buffer to the phen ol lay er and repeat the extra ction as in st ep s 1 and 2. 4. Combine both the aqueous layers and ext ract them w ith 8 ml of fres h ph eno l f or 20 m in at room temp erature (st eps 1 and 2). 5 . Transfer the re -extra cted aqueous layer to an appropriately sized
centrifuge tube . add 12 ml ethan ol and keep for 4-8 h at - 20 DC. 6 . Collect the RNA precipitate by centrifuging at 17000 9 for 30 min at 4 ° C. Resuspend the pellet in 4 ml of column load ing buffer . 7 . Appl y the redisso lved RNA to the equ ilibrated D EAE-cellulo se colum n. Wash the column with 10 m l of colum n loading buffer. 8 . Elute the column with 4 rnl of column elution buffer, coll ecting the entire eluate in an appropriately sized centrifuge tube . 9 . Add 3 ml of water , 1 ml of 2.0 M ammonium acetate buffer and 20 ml of ethanol. Keep the mixture on c rushed dry ic e for 1 h. 10. Centrifuge at 17 OOOg for 30 min at 4 °C. Discard the superna tant and
wash the RNA pellet by adding 5 ml of 70 % ethanol and centrifuging at 12000g, 15 min. 1 1 . Air -dry the pellet and resuspend it in 101.11 water .
12. Add 20
~I
of gel sample buffer.
13. Load the sample on to the first-dimensional gel . 14 . Using TBE as th e electrode buffer, run the gel at 40 rnA (constant curre nt )
until the xylene cy anol marker dye has migrated 25 -30 cm 116-20 hi. 15 . Using a sterile razor blade , cut out a 20 cm str ip of the sample lane fro m 5 cm ahead of the xy lene dye to 15 cm behind it. 16. Lay this strip horizontally across the top of the second -dimensional gel.
17. Add 2 ml of ammonium persulphate and 50
~1
of TEMED to 250 ml of
sec ond-di m ens ional gel mixture. Use thi s to cas t t he strip into place on t op of the second -dimensional gel. 18. Electrophorese the second-dimensional gel at 250V until the xy lene cy anol marker reaches the bottom of th e gel ( - 20 h). At th is point, RNA s 30 - 70 nt in leng t h should be well resolved in th e middle of th e gel. For better reso lution of products 70-120 nt in length , add a fresh aliqu ot 12 5 1J1l of gel sample buffer t o the top of the gel to pro vide a tra ck ing dy e and continue the electrophoresis until th e xy lene cyan ol marker again reache s the bottom of the gel.
Continued 180
of gel and film. 20 . Visualize the RNA bands by autoradiography (e.g. 8 -12 h exp osur e at - 70 °C us ing Hyperfilm-MP and an int ensif yi ng sc reen) . Develop t he film accord ing t o the manufacturer 's inst ructio ns. 2 1 . Consult ref s 25 and 26 for a com parison of patt erns , but , in genera l, pre-tRNAs and the ir pro cessed produ ct s migrate off th e diag ona l f ormed by rand om RNA fragments. Cut out indivi du al band s us ing a cl ean razor blade and elute the RNA by c ru sh ing eac h gel slice in a sterile glass vial using a glass rod and soaking the gel fragment in 2 m l of gel elution buffer
at 37 DC for 10- 12 h. 22 . Remove the gel fragments by filtering through a Ou ik -Sep column. 23 . Recover the RNA by adding 0 .4 ml of 2 M ammonium acetate buffer, 20 IJg gl y cogen carrier , and 7.5 ml ethan ol and then proceeding as in
steps s- 1 1. 24. Resuspend the final pellet in 20-50~ 1 of 10mM Tris-HCI. pH 7.4, and store it at - 20 °C .
Labelled pre-t RNA substrates prepa red as in Protocol 3 have a high specific rad ioactivity (5-6 x 106 d .p.m .z'pg) for use in assays requiring very high sensitivity . For applications which tolerate reduced sensitivity, the amount of label used can be reduced accor dingly. Unlabe lled pre-tRNA can also be prepared by th is method, visualizing the products in the gel (step 20) by staining with eth idium bromide (0.5 ug/ rnl in TBE buffer) . In this case, remove the ethid ium bromid e from the eluted RNA (step 23) by extracting twice with an equa l volume of propan-2-ol. P re-tRNA substrates prepared by thi s meth od contain many, but not all, of the base modifications characteristic of the mature tRNA product. For example , pseudouridine (.p), ribothymidine (T), and dihydrouridine (D) are present at high yield in pre-tRNATy" while only 50"70 of the expected l -methyladenosine (mIA) is detected (ref. 27). Thus the characteri zation of each new substrate prepared by thi s method should include an a na lysis of base modifications (see Section 4 below) .
2.4 Preparation of substrates by in vitro transcription This approach is applicable where a cloned tRNA gene is available and where requirements for other processing step s are either defined or can be taken int o account in the express ion system chosen . There are a number of distinct advantages to this approach including: • the abilit y to label the RNA substrate in defined nucleotide po sitions and to selected specific radioactivities 181
6: Processing of tronsfer RNA precursors
• the ease o f preparing pure substrates with defin ed structure • the abilit y to produce rare int ermediates, ab errant products, and substrate analogues by manipulating the structu re o f the transcription tem plate • the ability to incorporate nucleotide analogues, such as ph otor eactive Or biotin ylated deri vati ves, for use in cross-linking or rapid affinity-based purification procedures Both homologou s and heterologou s in vitro tRNA gene expre ssion systems are availabl e. The homologous systems rely on expre ssion by RNA pol ymerase III and intermediates can be produced by proce ssing activities present in extracts of whole cells or nuclei. This latt er feature, including the presen ce of certai n base-m odifying activitie s in so me extracts, represents one advantage to this approach . A significant limitation is that tRNA gene sequences requ ired for transcript ion and other processing step s cannot readil y be altered . Examples o f such coupled transcription /processing system s are described in Section 3 below. Het erologous systems are primarily based on the use of a plasmid DNA templat e containing a bacteriophage promoter fused to a tRNA gene and pur ified phage or bacterial pol ymerase for in vitro transcript ion reactions. Examples from the literature of the use of SP6 , lambda, and T7 bacteriophage promoter fusions to produce tRNA processing substrates are provided in Table J and these should be consulted in choo sing a specific expression system. These heterologous systems ha ve a number of dist inct advantages: (a) The presence and compos ition of 5 ' -leader and 3 ' -trailer tRNA sequences can be controlled . In the fo rmer case, this is achie ved by appropriate positioning of the phage promoter in th e ori ginal gene con stru ction an d in the latte r by the choice of the restriction site used for transcription 'r un-off' .
Ch ris L. Greer initiating nucleotide (usually GTP) . Th e low UT P concent rat ion used in Protocol 4 redu ces the reaction yield but results in high specific rad ioactivities ( - 5 x JOl d .p.m ./fmol).
Table 1. Examples of phage promoter/tRNA gene fusion s for the produ ction of semisyntheti c processing substrates
Promoter type /tRNA gene
Applic ati o ns -
References
Saccharom yces cerevis iae tRNA Glu (mitochondrial)
B
54
Sch izosaccharomyces pombe tRNA5er
A A
55 56 57
C, 0 , E
28 ,58
0, E
Phage SP6 promot er
Saccharom yces
cerevisiee tRNAleu
A
Escherichia coli tRNAPhe Phage A prom oter
Saccharom yces cerevisiae tRNATy, Phage T7 promoter
Sa ccharom yces cerevisise tRNAPhe Mycoplasma mycoides tRNAGly
C
59 ,60 61
Schizosaccharom yces pombe tRNA5er
A
62
"Uses made of the sem isyntheti c substra tes in the referen ces listed include: (AI formation of matu re 5 '-ends; (B) processing of 3 '-ends; (C} addition of one or more specifi c nucleotide modifications; (Dl presence of intron endonuclease and /or tRNA ligase acti vity ; and IE) structu ral analysis.
(b) Tr anscribed sequences can be altered at will without concerns for expression requirements characteristic of homologou s systems. (c) Transcript s can be prepared in high yield with th ese efficient transcription systems.
(d) Incorporati on of labelled or derivatized nucleot ides can be controlled through the selection o f appropriate nucl eot ide substrate mixtures. Limitation s to this approach are the ina bility to modify bases and the for ma tion of sub strates with inappropriate conformations under the in vitro transcription reaction cond itions. The former problem can be ad dressed in some instan ces by incubation of th e substrates in processing extracts (see Volume I, Chapter I). Th e latt er probl em may be alleviat ed by reannealing the pur ified substrates prior to their use in processing reactions (28). An example of the use of T7 polymerase to produce pre-tR NA Ph, subst rates labelled using [a_J2 p 1UTP is described in Protocol 4. Efficient tra nscrip tion requi res ;;,50 ~M (final concentration) of each of the elongati ng nucleotides and > 100 ~ M of th e 182
Protocol 4 . In vitro tR NA gene t ransc ript ion using T7 RNA polyme rase Equipment and reage nts • Phage T7 RNA polyme rase (60 Pharm acia 27 -0801 -01 )
U /~ I,
e Ribonucleotide stock solutions; separate 0.1 M stoc k solutions in 10 mM Tria HCI. pH 7.5 (ATP. 27·2056·0 1: CTP, 27·2066·01 : GTP. 27·2076·01 : UTP, 27 -2086-0 1; Pharmacia l e [a_3 2P j UTP (10mCi /ml , 3000Ci /mmo l; Amersham , PB. 10200-PB . 10203 1
e Plasmid DNA conta ining the tRNA gene cloned next to a T7 phage promoter ; prepared using Q IAGEN cartridges (Qiagen 510031 and the procedure recommended by the manufacturer
e 5 x transcription buffer (0 .2 M Tris- HCI, 50 mM MgC I 2 , 50 mM NaCl l e T EO.1 buffet 110 mM Tris -HCI , pH 7.5 , 0 .1 mM EDTA I
Continued 183
Chris L. Greer
6: Pro ce ssing of transfer RNA precursors Protocol 4. Contin ued
Protocol 4 . Continued . 50 mM OTT . St ore at - 20 ° C for up to 1 week e St op m ixture (0. 1 M EDTA . pH 8 .0 . 2 % 50S . 2 mg /ml prote inas e K (Sigm a. P
65561 • Phenol : chloroform : isoamyl alcohol
e Gtvcoqen carrier 120 mg /m ll and st erile water prepared as in Protocol 1 e TBE buffer , 2.0 M ammoniu m acetat e, pH 7 .0 . gel sample buffer . and gel elution buffer as in Protocol 3 • Appropriate restriction enzyme and buffer
12 5: 24: 11 • Polyacrylamide gel (1 5 . 5 % acrvt amide . 0 . 5 % bisacr ylam ide . 8 M urea , 1 x TBE buffer ; 19 x 22 x 0 .075 em ); and electroph oresis equ ipm ent (see Protocol 3 )
. 5 x rNTP m ixture 12.5 mM each of A TP• GTP . and CTP plus 0 .05 mM UTP I .l . 0 .4 M ammonium acetate buffer . pH 7 .0
Method A . Restriction of plasmid DNA template
1. Linearize the plasm id DNA by rest riction digestion at the appropriat e transcription ' run-off ' site .
6 . Add 5 ~I of gel sample buffer and incubate tor 5 min at 65 °C. 7 . Separate the RNA produ ct by electrophoresis in the 15 .5 % polyacrylamide gel at 20 rnA until the bromophenol blue dye has migrated 17 cm (2.0-2 .5 hi. Note that unincorporated nucleotides comigrate with bromophenol blue and that pre -tRNA transcripts migrate behind the xylene cyanol marker. Dispose of the used electrode buffer as radioactive waste. 8 . Disassemble the electrophoresis apparatus leaving the gel on the glass back ing plate . Wrap the gel and plate in cling film and mark the sides and corners with radioactive or fluo rescent ink for later alignment of gel and film . 9 . Visual ize the RNA products by autoradiography (e.g . 5-15 min exposure using Hyperfilm-MP and an intensifying screen) . Develop the film according to the manufacturer 's inst ructi ons.
10. Align the gel and film on a light box. With a razor blade , cut out the
2 . Extract the DNA by vortexing with an equal volume of phenol: chloroform : isoamyl alcohol. Centrifuge (2 mini in a mic rocent rifuge at room temperat ure to separate the phases. 3 . Precipitate the DNA from the upper (aqueous) phase by adding an equ al volume of cr open-z -o! and leaving the mixture at - 20 °C for 0.5 h.
4 . Rec over the DNA by centrifugation in a microcentrifuge for 15 m in . 5. Air dry the pellet, redissolve it in sterile water at 1 J,lg fJ,l1 and store it at
appropriate region of the gel and elute the RNA from each gel slice by soaking it for 8-12 hat 37°C in 2 ml of gel elution buffer as described
in Protocol 3 Istep 211. 11 . Add 0 .5ml of 2.0M ammonium acetate buffer ,
2 .5~1
of 20mg/ml
glycogen car rier, 7.5 ml of ethanol , and prec ipitate the RNA on crushed dry ice for 1 h.
12 . Centrifuge for 20 min. 15000 9 at 4 °C . Dry the pellet and resuspend it in 0 .35 ml of 0.4 M ammonium acetate buffer.
- 20 °C . B. In vitro transcrip tion of DNA template 1. Prepare a 50 IJI transcription reaction in a microcentrifuge tube on ice by
add ing: e ste rile H20 • 5 x t ranscription buffer e DTT e 5 x rNTP mixture e plasmid DNA template e (a · 32 p ) UTP
5 . Centrifuge in a microcentrifuge at 4 ° C for 15 min . Wash the pellet once with 70 % ethanol at 4 °C . Air dr y the pellet and dissolve it in 51J1of sterile water .
13. Repeat the ethanol precipitation as in steps 9 and 10 . 14 . Wash the pellet in 70 % ethanol. Resuspend the tinaldried pellet in TEO. 1 buffer 110 000 c.e .m .zui) and store at - 20 °C . · W hen lab ell ing w it h a nu c leos ide triph osphate ot he r than I cr_32 p JUTP. modif y the composit ion of th is 5 x rNTP m ixture accord ing ly .
12 ~I 10 J,ll
10 ~I 10 ~I 1 ~I 5 ~I
3. Preparation of tRNA processing extracts
2. Start the reaction by adding 21J1 of T7 polymerase and incubate f or 1 h at 37 °C . 3. Stop the reaction by the addition of 5 IJI of stop mixture and incubate f or
, 5 min at 50°C . 4. Add 40 ~I of 2.0 M ammonium acetate. pH 7.0 , 105 ~I sterile water , 0.5 ml ethanol and keep the mixture on crushed dry ice for 1 h to preci pit at e th e nucleic acids.
Continued 184
In addition to a method for preparing appropriate substrates, a bio chemical approach to tRNA processing requires the ability to prepare active processing extracts . Initially, the characterization of a processing reaction is often carried out with crude cell extracts . Long-term goals should include reconstitution of the reaction from highly purified components, obtained by fractionation of extracts from the original cell source or through the use of heterologous expression systems. The protocols described in this section represent examples from each of these three categories: preparation of a crude nuclear extract co ntaining a variety of activities , partial purification of a processing enzyme from yeast, and the use of a bacterial expression system. 185
Chris L. Greer
6: Processing of transfer RNA precursors
3.1 Preparation of cru de cell extracts Table 2 . Sources of active tRNA processing extracts
Extract source Eubacteria Escherichia coli
Activities II
Referen c es
A, B,C
63-65 66 67 27
A
Bacillus subtitis
A, B
Thermus therrnophuus
o
survey of diverse
bac terial sources Archaebacteria Halobacterium halobium Sulfolobu5 solfa taricus Mitochondria Saccharom yces cerevisiae
Rat lIiver) Human (Hela cells) Chloroplasts Euglena gracilis Sp inach
Maize Eukarvotes
A, 0 A
6B 69
A, B A,B A
54 70 71
A, B, C A A, B
72
A, B, C A 0 A, B,C ,D A, B A, B A, B,C , 0 A, B, C, 0
Saccharomyces cetevisiee Schizosaccharom yces pombe Chlam ydom onas
Wheat Drosoph ila m elanogaster Bom b yx mod X enopus laevis Hum an
72 , 73
74 31 55 75 76 - 7B 79
BO B1 - B3 B4- B8
(Hel.a. KB, leukaemia cells)
Crude or partially-puri fied extracts conta ining tR NA processi ng activities have been prepa red fro m a wide variety of sources and the litera ture provides an excellent starting point in designing an experimenta l strategy. Table 2 provides an overview of thi s literature with exampl es from eubacterial, archaebacterial, eukaryotic nuclear, a nd eukaryotic organellar sources . Many of these extracts are active for transcripti on as well as certain processing reactions allowing their use for stu dying coupled synthes is and maturati on reactions . Adva ntages and disadvantages in the use of the se extracts a re both rooted in their compl exity. Thus, one adva nta ge is the ability to exa mine processing steps requ iring severa l components or multiple processing reactions. A clear disad vant age is the presence of competing activities, including nuclea ses, proteases, pho sphatases, etc. Ho wever, with ap propriate rega rd for the inherent limitation s, these complex extr act s can be used effectively to obtain useful info rmation regarding specific tR NA pro cessing reactions. Protocol 5 describes the preparation o f a crud e yeast nuclear-extr act frac tion and is ad apted fro m th e procedur e described in ref. 29.
Protocol 5 . Pre par at ion of a nuc lear e xtrac t from yeast
Equipm ent and reag ents e Dounce homo genizer (40 ml size) with large (A) and small clearance t8l pest les (Kontes , 885300-00401 • Oak Ridge t ype plastic centrifuge tubes (30 ml , capped I • Saccharom yces cerevisiee strain JH AY20-2CO 1 (protease-deficient", MATa. his3D200, ura3-52 , leu 2-3. 112 , pep4-0 1 ::U AA3 ; other protease def icient strains can be substituted Ie.q . 20 8 -12. American Type Culture Col lec t ion . 206261
. 1% SOS e SI and SII spheroplast ing buffers (100 ml each ). Prepare 200 ml of 1.0 M sorbitol , 10 mM MgCI2, 50 mM Tris-HC I, pH 7.8 . Div ide this into 2 ahquots of 100 ml. Just prior to use add 3 ml of 1.0 M OTT to one aliquot to make SI buffe r and 0 .3 ml of 1.0 M OTT to the other aliquot to make SII buffer. • Sil l lysis buffe r (100 ml; 15 mM KC!. 5 mM MgC I2, 0 .1 mM EDT A . 10 mM Hepes -KOH pH 7.8 1
e 1.0 M OTT (10 m l); mak e up fresh using sterile water (see Protoco l 1)
e Oialysis tubing and dialys is buffer l3 titres : 0.1 M KCI, 10 % glycerol, 0 .1 mM EOTA , 20 mM Hepes -KOH , pH 7.8 1. Autoclave f1 h. 120 °C . 16 p. s.Ll : add 3 ml of 1.0 M OTT just before use
e y PO med ium (Pro t oc ol 2 1
e Solid (NH.)2S0. and 4 .0 M (NH4)2S04
e Yeast lytic enzyme (10 ~ U/g ; ICN Bio rnedicats . 190-123)
Me thod NB Carry ou t steps 1-3 at room temperature and all other steps at 2 - 4 DC.
.. .. r d incl ude: tAl f o rm at ion of mature steps .examm,ed In . t hc e PdUd~I.l catl~ns e l~:~ore speci fic. nucl eot ide m od ifications; S'-ends: tBI processing of 3 -ends: ( ) a men a on . and to) presen ce of intron endonuclease and /or tRNA ligase act wrtv .
~ Pro cessing
186
1 . Grow the yeast cells in 1 litre of Y PD medium at 30 °C with aeration until A eoc = 6 .0 -1 0 .0. H ar v e st the cells by centrifuging at 3 500 9 f o r 5 min.
Continued 187
6: Processing of tronsfer RNA precursor s Protocol 5 . Continued 2 . Resuspend the cells in 33 ml of S I buff er and divide th e suspension into t w o Oak Ridge-type centrifuge tubes. Centri fuge at 3500 9 for 5 min.
3. Resusp end each pellet in 16 .5 ml of 511 buffer and add 1.6 mg of yeast lytic enzyme to each tube . Incubate the cells at 30 °C w ith gentle agitat ion until spheroplasting is complet e. Mon itor the spheroplasting as follow s. Before addition of the lytic enzyme and at 10 min intervals thereafter remove a 10).11 sample of cell suspension and add it t o 2 ml of 1 % 50S . After 2 min, mea sure the A 600' Spher oplasting is com plete w hen the absorban ce falls to 20 % of t he time zero value (usually afte r 30-70 min). Excessive or inco mplete spheroplasting greatly affect s the final activity of the extra ct .
Chris L. Gree r meth ylated deri vati ves (when supplemented with 0.1-1 mM S-adenosylmethion ine). The extract also contains a sma ll amount of tRNA endonuclease and RNA ligase, but splicing is inhibited at the sa lt concentrations typically used in transcription reactions (30). Thus end -mature, unspliced precursors accumulate in reactions using intron-eont aining tRNA gene templates. Accumula tion of 5 ' a nd 3 ' end-extended precur sor s can be accomplished by carrying out tr an scripti on reactions at a reduced temperature ( 12- 18 0C).
I
4 . Collect the spheroplasts by centrif uging at 2000 g for 5 min at 4 °C . Resuspend them gent ly w ith a glass rod in Sli buff er 116 .5 ml per tube) and repeat the centrifugation .
5. Resuspend each pellet in 10 .5 ml of 5111 lysis buffer. Combine the contents of the two tubes into the Doun ce homo genizer (prechilled) and homogen ize once gent ly wi th the A pestle . Incubate the suspension for 20 min on ice .
6 . Lyse the spheroplasts by homogenizing vigorously 3 times wi th the B pestle.
7 . Collect the particulate fraction (including th e nucle i) by centrifuging at 17000 g , 10 min at 4 ° e and carefull y pour off the supernat ant .
8 . Resuspend th e pellet in 8 ml of Sill buffer . tran sfe r the suspension t o th e homogenizer , and use an additional 2 ml of Sill buffer to rinse out th e centrifuge tube . Homogenize the combined suspension once gent ly with th e A pestle .
9. Calculate the volume of the ' breakage pellet ' (f or use in step 11 below ) by measuring the final volume of homogenate in step 8 and subtracting
10 ml (the total volume of buffer added) . Add 0.2 vol. of 4 .0 M INH4 12 S0 4 . Extract for 10 min on ice with occasional mixing. 10. Remove particulate material by centrifugin g at 150 000 9 f or 1 hat 4 " C. Rem ove the supernat ant carefully and mea sure the volum e. Add 0 .25 g/ml of solid INH. I,SO• . Mi x gently and incubate for 10 min on ice.
11. Collect the precipitate by centr ifuging at 17 000 g for 30 min at 4 °C. Resuspend the pellet in 0 .5 time s the volume of the ' break age pellet '
(calculated in ste p 9). 12 . Dialyse the suspension against one litre batches of dialysis buffer (making 3 changes of buffer at 1 h intervals ) at 4 °C w ith stirring. 13 . Transfer 50 ~ I aliquots of the dialysed extract to microcentrif uge tu bes and sto re them frozen at - 70 °C .
3.2 Partial purification of processing activities Determinat ion of the reaction kinet ics o f RNA processing, the struc tures of pro cessing intermediates and products, and the concent ration dependence o f sub strat e, enzyme, or co-factors is often not feasible in crude extracts du e to the presence of competing activities and the poorly defined enzyme, co-facto r, and RNA content of the extra ct. Although reconstitution from pure comp onent s may be the ideal way to investigate these aspects, the use of part ially purified fractions provides a realistic intermediate step in the characteri zation of a processing system. Puri fication procedures are necessarily idio syncrati c and are developed empirically. However , a common strategy is the initial application of conventional protein fractionation techniques which rely on the general physical properties of the proteins of interest, followed by specific a ffinity methods which take advantage of distinct funct ional properties. Co nventi ona l techniques often ap plied to tRNA processing activities include diffe rential amm onium sulphate precipitation , ion-exchange chromatography (particularl y using heparin -agarose , phosphocellulose, DEAE-cellulose, hydroxyapatite, and certain FPLC resins, including Mono Q and Mono S from Pharmacia), and molecular sieving columns. Affinity chromatography methods include the use o f immobilized tR NA or synthetic RNA resins, and tRNA-mediated elution from conventional resins (32, 33). Protocol 6 describes a partial purification of yeast tRNA endonuclea se. A more extensive pur ification proc edure ca n be found in ref. 34. The method described here leads on from the preparation of the nuclear extract described in Prot ocol 5 which should be consulted before proceeding . This method takes advantage of the observation that yeast end onuclease behaves as an integral mem bra ne protein and is adapted , in part, from the method developed by Peebles et 01. (35). Note that in so me experimental systems, such as X enopus oocyte s (36), the endonuclease ma y be fo und in the soluble frac tion and so this purification procedure should be modi fied accordingly.
" See ref. 30 .
Typical yields from this procedure are 2-3 ml of an extract cont aining 8- 12 rng total protein per millilitre. This extract is effi cient in tR NA gene transcription (29) and is active in a number of pr ocessing reaction s including 5 ' -end maturation , 3 ' -end trimming and CCA addition , and base mo dificatio ns such as the fo rmatio n o f pseudouridine, dih ydr ouridin e, riboth ymidine, an d cert ain 188
Protocol 6 . Preparation of partially-purified endonuclease from yeast
Reagents • Yeast nucle ar pellet prepared as described in Protocol 5 . st eps 1 - 7
. 1.0 M
on
(see Protocol 5 1
Continued 1 89
Chris L. Gre er
6 : Pro cessing of tran sfer RNA precursors
This procedure yields an endonuclease prepara tion which is ap proxi mately equ ivalent to Fraction V of ref. 35. Endonuclease activity averages 100 U/ml where 1 Unit of endonuclease cleaves 20 pmol pre-tR NA substrate min - J under standard reactio n conditions (35). Signifi cant ly, the preparation contains onl y traces of contaminating phosphatase and nuclease activities and is suitable for most stu dies o f processing reactions includ ing their subst rate requirements, rea ct ion products, and general reaction k in et ic s.
Protocol 6 . Continued • 20 % lvlvl aqueo u s Trit on X-l 0 0 IBioRad . 161 -0407 1: equilibrat ed ov erni gh t prio r t o use • O. , 5 M PMSF (Sigma P 76 26 ), prep ared fr esh in prc pen- z -ol and added fresh t o buffers just pr ior t o use; NB This prote ase inhibi tor is t ox ic and shou ld be handled w ith care. • Heparin- agar ose (0 .8- ' m g h eparin /m l bed volume ; Bio-Rad 15 3 -6173 1. T w o co lu m ns are requ ired . one co ntai nin g 10 ml bed volume per litre of sta rt ing cell cul ture (see Protocol 5 1and one co nt ain in g 1 ml bed vol ume (in a disposable colum n or syr inge} each equi librat ed in ME buffer. e O.5 M sperm idine (phosphate salt. Sigma.
• HSM buffer ( ' .0 M KCI. 0 . 1 M Tris- He l• pH 8 .0. 10 % gl y cerol. 2 mM EDTA , 5 mM sperm id ine. 1 m M OTT ". 1 mM
PMSF' ) e LOM buffer (20 mM T ris -HCI, pH 8. 0 . 10 % gl ycerol, 5 m M 2-m ercap to ethan ol,
, mM EDTA. 0 .2 mM PMSF · . 0. 02% Tr it on X-' OOl
e ME
3.3 Preparation of highly-purified processing components
e ME buff er w it h 2 % Tr it on X~ 1 00 , 0 .1 M
Idea lly, compre hensive bioc hemical character izat ion of a tR NA processing step should be carried out thro ugh its reconstitution fro m purified compone nts. Often, this is accomplished by extensive purification of processing enzymes from hom ologou s cell extracts. This ap pro ach is pa rticularly a pplicab le to reactions requi ring multi ple compo nents or complex assem blies, does not include a requirement for cloned genes or cDNAs, and allows for the inclusion of essential post -translational modifi cations. However, this approach ca n be experimentally chall enging or even impracticable when the cell or tissue source is di fficult [0 obt ain in large quant ities, o r when the processi ng components are present at only trace levels in cells or are unsta ble in homol ogou s extracts . An alt ernative approach is expre ssion at high levels in a heterologous system followed by pur ificat ion by conventi onal or affinity meth od s. A prerequi site for this approach is an appro priat e cloned DNA seq uence encoding the desir ed product. Advantages include the ability to produce altered processing component s in vitro by modifying th e expressed gene (particularly important where the gene encodes an essential product) and the beneficial effect of efficient and high-level expression on both the degree of pur ifica tion requ ired to obtain a homogeneou s product and to produ ce lar ge quantit ies of the purified product . Disadvantages in thi s approa ch are those th at fav our homologous extracts: limited post-translational proce ssing of the expressed protein facto r and difficulty in recon stituting multicomponent or co mplex systems. Protocol 7 prov ides an example of a bacter ial expression system which allow s the rap id and efficient purificati on o f a yeast proce ssing enzyme (RNA liga se). The bacterial expre ssion vector used in thi s procedure, pDAKC (37), allows expre ssion of yeast ligase as a translational fusion protein with bacterial dihydrofolate redu cta se (DHF R). Th is vector, derived from plasmid pKK223-3 (38), includes both the tac promoter for high-lev el expressi on on induction with isopropyl-Bvn-thiogalactoside (IPTG), and a transcription terminator, rrnB Tl /T2. The fusion protein can be purified from bacterial cell extr acts using a rap id and sim ple affi nity method ba sed on DHFR bindin g to methotrexa te-aga rose (39). The a ffini ty-purified protein Can be used directly in pr ocessing studies, or contaminating nucleic acids and enzyme-bound folat e can be removed usin g a DEA E-cellulose flow-through pr ocedure as describ ed in Protocol 7.
buff er ILD M buff er but wi th 0 . 9 % Tr it on X · 100 and 0 . 1 M INH 4 12 S0 4 t (NH 4 J2 S0 4
e Storaqe bu ffer A (20 mM Tr is -HC I,
pH B.O. 0 .2 mM EDTA. 0.5 mM DTT'. 0 . 5 % Tr it on
x-too. 3 5 % gly cerol)
S 0381 ). pH to 7.5 with He l: stor e froz en at - 20 °C
Method NB Carry out all the following procedures at 2-4 OCt
t. High -salt wash ofnuclei. Resuspen.dthe veast nuclear pelltetin, 7 ,5 .ml of4H~CM buffer. Collect the nuclei by centrifuqation at 17 000 g or
5 m in at
.
2. Low-detergent wash. Resuspend the pell et in 7 .5 ml of LDbM bufffer ~nd incubate the suspension for 10 min on ice. Recover the mem rane racnc n by centrifuging at '50000 g lor 60 min at 4 " C. 3 . Membrane extraction . Resuspend the membrane pellet in 10lml 0 ~E buffer containing 2 % Trit on X -l00.lncubate the membranes or 3 0f ~In on ice with occasional mi xing. Remove any rema ining parti culate material by centriluging at 150000g lor 60min at 4 °C . 4. Apply the supernatant to the large heparin-agarose c olum n . Collec 2 ml fract ions th roughout each of the fo llowing wash and elution procedtur es. Wash the co lum n w ith 20 ml 01 ME buffer . 5 . Elute the column stepwise with 20 ml of ME buffer co ntaHiniln gs oo. 2 ~ ~ INHI,I,SO , . and th en 20 ml 01 ME buffer plu s 0.45 M IN " ,. e endonuclease elutes in the latter tracti ons . 6. Dialvse the elut ed endo nuclease again st 3 x 1 litr e of fraction buff er f or 1 h each time at 4 °C w ith st irr ing . 7 . Con centrate the dial ysed endonu clea se b y apply ing it t o th e 1 m l heparin -agarose column . Elute the endonuclea se w it h 2 m! of ME buffer plus 0.45 M \NH,I ,S0 4· 8 . Dialyse th e eluat e ove rnig ht at 4 ° C against one litre of s.torage buffer. St ore th e di alysed endo nuclease solu tion at - 20 ° C In 0 .1 - 0 .5 ml aliquots . , A dd t hes e reagent s just prior to use from stock solutions.
190
191
Chris L. Greer
6: Processing of transfer RNA precursors Protoc o l 7 . Continued Protocol 7 . Puri fic ation of DHFR - lig ase fu sion protein f rom E. coli
Equipm en t and reagents • Disposable chromatography columns
IOS-Y. lsclab l e E. coli. stra in SW1061 (pDAKC) : C60 0 Dlon 100. hsdR - . hsdM ~ . 5up E. thr- l . SQ: leu B6 . th i-v , lac Y l , (F' /ac I : TN5 1: tran sf ormed wi th the lig ase expression vecto r. pOAKe
47 .5mM K, HPO. I. 0.16M KCI. 10% glycer ol) • MBE buffer to. 1 M pota ssium borate . pH 9.0 . 1.0 M KC!. 1 mM EDTA . 2 mM folic acid 1 m M 2-mercaptoethanoll . Add th e fola~e and merc apt oeth anol just prior t o
use
• LKA growth medium (8 gIl NaC!' 5 g fl yeast extra ct, 10 gIl Bacto-trvptone (Di teD Laboratories) plus 50 ~g fml each of kanamycin and ampicillin }
• MB buffer (50 mM potassium phosphate . pH 6 .0 Ii.e. 44 mM KH 2P0 4. 6 mM K2HPO. 1. 0 .2 M KCI. 0 .5 m M EOTA . 1 mM 2-mercaptoethanol)
e Lvsozvme stock solution (17 mg /ml in '0 mM Tris- He l, pH 8.0 ; Sigma L 6876)
• Methotr exate-agarose. Equilibrate methotrexate - agarose slurry (Pierce 20206 HI in MB buffer . Use 20 ml of slurry. cont aining 10 ml (sett led volume I of methotr exate -agar ose and 10 m l of MB buffer
• Brij 58 solution (poly oxyet hyl ene. 20 cetyl ether ; 50 mg /m l aqu eous solution : Sigma P 58841 e ONase I solu t ion ( 13 mg /ml stock in 10 mM Tris-He l. pH 8 .0 ; Sigma DN-2 51
e O. l M IPTG • O. 2 M PMSF IP 7626 Sigma). prepared fr esh in propan ·2-01 and added fresh to buffers just prior to use. NB This protease inhi bi tor is toxic and should be handled with car e
e O.5 M EDTA (pH 6.0 with NaOHI . ' .0 M MgCI 2 . 40 mM Tris-HCI . pH 8 .0 • Solid INH4)2S0•• finely ground • DEAE buffer (50 mM potassium phosphate. pH 8 .0 tl.e. 2.5 mM KH 2P0 4•
. D EA E- c ell u lo se (W hat m an DE- 5 2 ; 4057 -2001. Equilibra te 3 ml (settle d volume) per litre of start ing cell cult ure in DEAE buffer • MW buffet 1200 mM potass ium phosphate . pH 6.0 Ii.e. 176 mM KH 2P0 4• 24 mM K 2HPO. ). 1 M KC!, 1 mM 2 -mercapt oethano ll • FA buffer (100 mM potassium phosphate, pH 8.0 J (l.e . 5 mM KH 2P0 4 and 95 mM
K 2 HP0 4 ) • Storag e buffer B ISO mM pot assium phosphate . pH 8.0, 0 .' M KC!, 5 m M z -mercaot oemen ol. 35 % glycerol )
M ethod NB Steps 4-13 are carried out at 2 -4 °C with prechilled equipment and buffers. Growth of bacterial cells
\
1 . Inoc ulate 50 0 ml of LKA growth med ium with bacteria and grow ov ern ight at 37 °C with aeration until t he A 600 is approximately 3.0 . 2 . Inoculate 10 litres of LKA med ium w ith 500 ml of the overnight cult ure and grow at 37 °C until the A 600 is app ro ximately 1.0 (3 .5-4 .0 h). 3 _ Add IPTG to 0 .5 mM and continue the growth for 45 min (A 600 shoul d reach 1.2-1 .6 ).
4 . Harve st the cells by centrifuging at 4200 9 fo r 10 min at 4 " C. Resuspend the cell pe llet in 500 ml of 40 mM Tr is -HCI . pH 8 .0, and centrifuge at 6000 9 for 15 min at 4 ° C in a preweighed centrif ug e tube . Record th e w eight o f th e ce ll pellet {typic ally 1.8-2 .8 9 per lit re of origi nal cu ltu re} . Store the cells at - 70 °C ; th ey are m or e read ily ly sed (in ste p 6 be low) after a freeze/thaw cy cle .
Ex trac tion of fu sion p rotein 5 . Add 2.14 ml of 40 mM Tris -HC I. pH 8 .0 , per gram of cell paste and measure the volum e (V, ) of the resuspended cells . Add 0 .005 x v, of 0 .2 M PMSF , 0 .0 1 x V, of 0.5 M EDTA and 0.04 x V. of lys o zym e st oc k solu ti on . Incuba t e th e mi xture at 2 - 4 ° C for 30 min w ith slo w stirring .
6 . Ad d 0 .04 x V r of Brij 58 solution. Incubate f or 30 min at 2 -4 ° C with slow st irri ng . The viscosity should inc rease wi t h cell ly sis. 7 _ Add 0 .082 x V. of 1.0 M MgC I" and 0 .004 x V. of DNase I solution. and incubate for 30 min . Highly viscous samp les may require the addition of a second aliquot o f DNase I sol ution and fu rt her incubat ion for 30 m in. 8 . Remo v e t he c ell debris by ce nt ri fu ging at 2 3 000 9 for 1 h at 4 °C . Carefully reco v er the sup ernat ant and m easure it s v olume . Add 0 .226 g of solid (NH4)2S04 per ml of supern at ant ; make th e additions in sm all amounts wai ti ng until t he (NH4)2S04 has fully dissolved bef or e further add iti on .
9 . Centrifuge the suspension at 12 000 9 fo r 40 m in at 4 °C . Recover the sup erna t an t and measure its vo lum e. A dd a furthe r 0 .212g of solid (NH4)2S0 4 per ml of su perna t ant slo w ly wit h mixing . Centrifuge at 12000 9 f or 40 m in . 10. Resuspend th e prot ein pellet in 18-20 m l o f MB buffer. Measure th e protein concent ration by the m ethod of Brad f or d {40 1. Ad ju st t he f ina l co nc entr ati on to 20 mg prot ein /rnl wi t h MB buffer (t ypic al yi elds are - 75 mg protein per lit re st arti ng culture ). 11 . A dd 1 ml of methotrexa te-agarose slurry per lit re of starting culture . Place the resu lt ing su spen sion in a dialysis bag and dialys e it again st two 4 litre bat ches of MB buffer f or 12 -16 h (m ix the suspen sion occasionally by inv erti ng th e dial ys is bag) .
12 . Mix t he dialys ed sample thor oughly and transfer it t o a disposable chrom atography column , allowing th e column to flo w at - O. 5 ml /min . Collect 7.5 ml fra ct ions . 13. W ash the co lumn at 0.75 ml/m in wi t h MW buffer until th e eluate reach es A zao < O. 5 (ty pic ally after 20-30 c olumn vo lumes) . 14 . Elute the bound protein with M BE buff er at 0 .5 ml /min , collecting 1.5 ml fr act ions . Measure the prote in content (ste p 101in the elute d fra ct ions and poo l those with protein con centrati ons ~ 20 % o f that of the peak fract ion . Dialy se the pooled fractions agains t 3 cha nges of FR buffer (1 litre each time ) over 10-12 h.
1\ Co ntinued
Continued
1 92
193
6: Pro cessing of transfer RNA precursors
Chris L. Greer
A.
Protocol 7 . Continued 15. Con c entrat e the eluted protein by pre cipitation with (NH 41 2 S0 4 as
17 . Apply the dial ysed solution to the DEAE-cellulo se col umn. Maintain th e flow rat e at 1.5 mllmin and collect 2 ml fract ion s. Wa sh th e column w it h 1 column volume of DEAE buffer, collecti ng fractions as before . Mea sure the protein concentrations of the fract ions and pool those fracti ons w it h v alues ~ 10% of the peak fraction .
lB. Concentrate the pur if ied protein by precipitation wit h INHa)2S0a as
--
0 .8
follow s: m easure the v olume of th e po ole d fra ction and add 0.5 16 g (NH 41 2 S0 4 per ml. Colle ct th e prec ipitate by centr ifuging at 12 000 9 for 45min.
16. Resuspend the pellet in 5 m l of DEAE buff er and dialy se the solution aga inst 3 changes of th is sam e buffer (1 lit re each t ime over 12-15 h).
8. r ~oclion~ Poo le d
-
-------j
97~
"' "'
..
0.6
C7>
66 -
~
C
'"0
0.4
~
Q.
45 -
described in step 15.
19 . Resuspend the prec ipitated protein in 1 ml of storage buffer Band di aly se it against 3 changes of this same buffer, 1 lit re each time over 15 -20 h. 20 . Store the dial ysed preparation in aliquots at - 20 vC.
0 .2
31
2
Final preparations obtained using Protocol 7 typically contai n 1- 2 rng protein/ml, corresponding to yields of 0. 1-0.2 mg per litre of starting cell culture. An examp le of the purification of the DHF R/ ligase fusion pro tein is shown in Figure 2, including the methotrexarc-agarose affinity chro matog raphy (Figure 2A) a nd SDS- PAGE analysis of poole d fractions obtained by affinity chromatography (Figure 2B). Ligase activity in the final pur ified fract ion is typically 1000- 3000 U/ml, where I Unit of enzyme activity ligates 1 pmol tRNA substrate per min (41). Other important features of the procedure include the following: r
'
(a) The peptide linker in the fusion protein includes a Factor X cleavage site which allows separation of the yeast and bacterial protein segments by endoproteinase treatment (42). However, the author has found th at the activity of the ligase is not affected by the pre sence of the DHFR segment.
4
6
8 10 12 I . 16 18 20 22 M
Fracti on Numb e r
DAKC
DAC
Figure 2 . Purificat ion of DHFR- ligase fusion protein . (A) Elut ion profile of methotrexat e- agarose affin ity chromatog raphy (Protocol 7) . The protein conte nts of the elution fractions were measured by th e meth od of Bradf ord (4 0). Th e A S9 5 values shown are fo r 20 IJI aliquo ts in 1 ml assays. Fract ions 9 -19 we re pooled f or su bsequen t processing (see Protocol B ). (B] Samp les of t he pool ed elut ion fra ct ion s in fA) w ere conce nt rated and th en analysed by electrophores is in SDSpolyacrylamide gels and stai ned with Coomassie blue (891. Lanes co nt ained - 10 IJg of an extract from cells expressi ng t he 11 0 kDa DHFR -l igase fusion prote in (DA KC) or a 9 1 kDa delet io n derivat ive lDAC, lacking ami no acids 397 -59 5; ref. 37) as indicated at the bott om of the panel. Molecular weights (kDal of pro tein mar kers (lane Ml are indica ted on the left.
(d) While native gel electrophoresis to measure pre-tRNA binding could not be carried out with fractions prior to the DEAE-cellulose step, the enzymatic act ivities of th e methotrexate-agarose and DEAE- cellulose fractions were similar (37) and an abbreviated purification (omitti ng steps 16-1 8 of Protocol 7) is suitable for many applications.
(b) Induction of the cell cultures with IPTG for long periods ( ;;,2 h) results in the accumulation of subfragments rather than full-length fusion protein. In fact , for many deletion derivatives, recovery of the intact fusion prot ein is better with uninduced cells. Thus, a comparison of induced and uninduced cell extracts should be made before sett ling on a final protocol.
4.1 Strategy
(c) The ionic strength of the DEAE buffer in the DEAE chromatography is intended to allow retention of contaminating folate and polynucleotides without retention of the fusion protein. Recover y in this step is sensitive to small changes in ionic strength and so the conductivity of buffer , eluant , and sample should be ca refully mon itored .
In addition to procedures for providing RNA substrates and active processing extracts , a third essential element in a biochemical approach to tRNA processing IS an accurate, sensitive, and practical assay method . Important considerations in the design and implementation of an appropriate assay fall into two categories. These are:
1 94
195
4. Pre-tRNA processing assays
6: Pro cessing of tronsfer RNA precursors • the choice of reaction conditions • the basis for the analyt ical method used to distinguish between substrates, intermediates, and reaction products. Reaction conditions chosen for initial characterizations are often intended to mimic, within practical limitations, the normal cellular environment. Primary variables include pH, temperature, ionic stre ngth, the nature of counterions, incubati on times, and the concent rations of co-factors, substrate, and processing components. The effec ts of stabilizing agents, including sperm idine and Mg2 . ions which bind to a nd stabilize tR NA structure (10), an d reducing agents should also be exami ned. An example of a completely recon stitu ted splicing assay in which reaction co nditions have been optimized is described in Protocol 8 (Section 4.2). Th e design of an analytical method for measu ring reaction products depends on the nat ure of the processing reaction . As described in Section 4.2, tR NA processi ng reactions are of three types: removal of extraneous sequences, addition of non -encoded bases, and introduction of base and ribose modifications. Reaction s in the first two categories result in substrates and products which differ both in size and sequence, so assay methods are generally designed to exploit these differences. Thu s electropho retic separation (20, 21) and different ial hybridiza tio n to oligo nucleotide prob es (16) are com monly applied to mon itor the first two types o f reaction. Electr ophor etic sepa ration is incorporated in Protocol 8. Reactions in the third cat egory, the int roduction of modifications, genera lly do not alter the electrophor etic mobilit y suffi ciently to serve as the bas is for an assay . Instead , indirect methods must be used , most of which rely on an an alysis of the base co mpo sition of subst rates and reaction products. An example of an indirect meth od for a nalysing base modifications is described in Protocol 9.
4.2 Assay of pre-tRNA splicing The presence of an intron is characteristic of a limited subset of euk aryotic nuclear t RNA genes (43). Splicing of the tra nscripts of these genes requ ires two distinct activiti es: a site-specific endo nuclease and an RNA ligase. The splicing reaction can be reconstituted in vitro using pre-t RNA substrates , prep ared as desc ribed in Sections 2.3 an d 2.4 , and splicing enzyme fractions, prep ared as described in Sectio n 3.2 a nd 3.3. Optimized co nditions for in vitro splicing and the use of gel electropho resis as a basis for the sepa ration and quantification of the reaction produ cts are illustr ated in Protocol 8. Importa nt features of this assay include the following: (a) T he yeast endonuclease preparation is a detergent extract o f a membra ne fraction and so detergent (Tri ton X-lDO) is required in this assay (in the lOx splicing cocktail) to maintain enzyme solubility. 196
Ch ris L. Gree r
(b) The stabilizing effect of spermidine varies a mong pre -IRNA su btrates with a mode st effect for some and a strict requi rement for added sper midi ne fo r others (cf. ref. 35). (c) Joining by RNA ligase involves two NTP-dependent reaction s; a GT Pdependent substrate pho sphorylation and an ATP-dependent ade nylylat ion of protein and RNA intermediat es (Belford, H. G. , Westaway, S. K., Abel son , J ., and Greer, C. L., in preparation). The nucleotide mixture in the assay provides both NTPs at concentrations 2-5 times their respective K app (NTP concentrat ion required for half-maximal velocity) values. (d) The splicing reaction is opt imal over a relative ly narrow range of NaCI and KCI concentra tions (5- 50 mM). Howeve r, glutamate, a more relevant physiological anion than CI - (45), can be used at concentrations up to 0.5 M with little deleterious effect on splicing and at the sa me time redu cing the non- specific binding of ligase by RNA (46). Thu s, as has been suggested for other protein/nucleic acid interactions. the use of glutamate as a counterion may. under certain conditions, enhance specificity. (e) Digestion of protein components with proteinase K (in the stop mixture, see Protocol 8, step 3) prior to precipitation with ethanol improves recover y and enhances subsequent electrophoretic resolution of the reaction products . This is especially apparent for reactions containing concentrated or compl ex protein mixtures.
Protocol 8 . Pre -IRNA s plicing assa y Equip m ent an d reag ents •
32 P-labelled pre-tANA subst rate (- 5000 d.p.m.lfmol ; 5 fmol/lJll prepared as in Protoco/3 or 4 ; each assay requires
• 10 x nucleotide mixture (1 mM ATP and 0 . 1 mM GTP in TEO. 1 buffer); store at - 20 DC
5imal e Yeast end onuclease prepared as in Protocol 6 ; each assay requires
O. 1-1 .0 xlO - ·units • DHFR- ligase prepared as in Protocol 7;
each as s ay requi res 0 .1- 1.0 x 10 - 3 units
. , M OTT (see Prot ocol 5 ) · 0 .25 M spermidine (Protocol B ), pH 7.5 w ith He l • Sterile w ater 'Protoc ol 1) . 10 x splicing cocktail (0.25 M NaCI, 0 .2 M Hepes-KOH . pH 7.5 , 50 mM MgCI 2 , 25 mM spermidine, 10 mM OTT , 4 % Triton X· 100 )
• Dilution buffer (2 5 mM NaCI, 20 mM Hepes -KOH . pH 8 .0 . 1 mM OTT . 0 .5 mM EOTA . 0 .5 % Triton X -1 00. 20 % glycerol) • Polyacrylamide gel ( 1 2% acrvlamide. 0 .4 % bisacrylamide. 8 M urea. 1 x TBE buffer; 20 x 20 x O. 1 ern): see Protocol 3 . Gel sample buffers. elution buffer, 2.0 M ammonium acetate, pH 7. 0 , TBE buffer, Quik-Sep columns as in Protocol 3 . • Stop mixture and TEO.1 buffer as in Protocol 4 . 2.0 M sodium acetate, pH 5.0
Continued 197
6: Pro ces sing of tran sfer RNA precursors Protocol 8. Continued
A.
Method
1.
2.
B. 60 I~-~-~-~~r--,
Onqm -
Dilute enzyme stock solution s, where nec essary, using dilution buffer and leav e to equilibr ate for 0 .5 -1 h on ice (equili brat ion aft er dilution is essential for detergent-solubilized fractions ). Dilute end onuclease fractions to 0 .1 - 1 x 10 - 3 Units/ul . Dilute ligase fra ctions to 0 .1 -1 x 10 - 2 Units /pl . To a microcentrifuge tube on ice , add the foll owing in the order give n:
• 50
• pre-tR NA - . . . .. . . . .
• st erile wate r . 10 x sp licing cocktail
5 jJl 1 1-11
t RNA -
. 10 x nucleotide m ixture
1 IJI
Xyl en e -
• yeast endonuclease
1 IJI
• DHFR -ligase
1 ~I 1 IJI
• 32P-labelled pre -tRNA substrate 3.
Ch ris L. Greer
- - ..,-
'0
w
40
c ...,o ~
30
••
Hnlve s ]" •
20
Incubate the mixture at 30 °C for 5-20 min . Stop th e reaction by adding 0 .1 vol. of stop mixture . Incubate for 10 min at 50 °c.
4.(a) For procedures requ iring fine elec trophoretic reso lution Ii .e. molecu les differing by less than 10 % in chain length ) Concentrate t he react ion samples prior to electrophoresi s by ethan ol precipitation : adjust the volume to 80 IJI with sterile water , add 20 IJ I of 2 M sodium acetate, pH 5.0 , 20 IJg of gly cogen carrier (Protocol 1L and 0 .25 ml of ethanol. Leave on crushed dry ice for 30 min . Reco v er the precipitated RNA by centrifugation at 12000 g for 20 min at 4 ° C. Resuspend the dried pellet in 51.11 of gel sample buff er. (bl For procedures requ iring limited reso lution (i.e . mol ecules differing by more than 10 % in c hain length ). Analy se the RNA pr odu ct s dire ctl y by ele ctrop hor esis by adding 1 assa y v olume (10 IJIl o f gel sample buffer. 5.
Denatu re th e samples immediate ly prior t o elect rophoresis by inc ubati ng fo r 5 min at 65 DC.
6.
Load t he samp les ont o th e pol ya c ry lam id e and elec t ro pho rese as descr ibed in Protocol 4 (st ep 7 ) allowing th e bromophenol blue m arker to reac h th e bott om of the gel 12 . 5 - 3 hat 20 mAl.
7.
Disassembl e th e electrophoresis apparatus leav ing the gel on th e glass backing plate . Wrap the gel and plate in cling film and mark th e sides and corne rs w it h radioa ct ive or fluorescent ink f or lat er alignme nt of gel
8.
Vis ualize th e RNA band s by auto radiog raphy Ie.q . 8 - 12 h exposure at - 70 °C using Hyperfilm -MP and an inte nsifying screen) . Dev elop t he X- ray f ilm accordin g t o th e manufacturer ' s instructions .
9.
Al ign the gel and the autoradiograph on a ligh t box. With a razor blade, cut out slices fr om th e gel cor respon ding t o t he tRNA precu rso r and its react ion produc ts and dete rmi ne th e amount of radioactivity present by me asuring the Cere nkov radiation. Recover the RNA produc ts from the gel slices by elution as described in Proto col 3 (steps 2 1-2 4).
and film .
10.
198
10
[GTP]
o 0.1
05
1
25
5
2
[GTPJ
3
4
5
( uM)
Figure 3 . Pre-tRNA spl icing as.sa~ . lA) Assays (10 IJI; see Protocol 8 ) contained pre- tRNA Ph. IJM A TP, and the Indicated concentration (IJMl of GTP. React ions were incubated with endonuclease for 10 min at 30 °C follow ed by the addition of ligase and furt her incubation f or 20 r:' in at 30 °C. The products were sepa rated by gel elect rophoresis and visualized by aut oradIography (Protocol 8 , steps 7 and 8 1.The ident it ies of labelled products and the pos ition of the x yl~ne cvano! marker are indicat ed on the left as follows : pre-tRNA , pre -tRNA substrate ' tR NA , sp liced tRNA produ ct ; halves, ex on produ cts of cleavage by tRNA endonuc lease lS ; The reaction prod.uc.ts s~pa rated by gel electrophoresis in panel A we re quantified by measurfnq th e Ceren kov radiation In gel slices (see Protocol 8 , steps 9 and 101. Ligase activity (% Joined ) w as calculated as des cribed in ref . 44 .
s~bstrate , 10
Th e results of a series of splicing assays using labelled pre·t RNA s" as substrate are shown in Figure 3. P anel A shows the autoradiog ra ph of the gel Th e endonuclease cleavage pro ducts (tRNA 5 ' and 3· exons and the intervenin~ seq~ence) mig rate between the dye mar kers in this gel system. The joined product (spliced tR NA) mig rat es behind the xylene cya nol marker and is well resolve d from endo nuclease products and from residual precursor.
4. 3 Assay of base modifications and tRNA splice junctions 4 .3 . 1 Strategy Tr an sfer RNA biosynthesis includes the post-transcrip tion al int rodu ction of a complex and extensive set of specifi c base an d rib ose mod ifications (reviewed 199
6: Processing of transfer RNA precursors
Chris L. Greer
in refs 7, 8). These modifications are essential for normal tRNA functio n, influencing decoding potential, translational efficiency and fidelity, and conformational stability . Reactions required to generate the full range of modifications found in a mature tRNA include simple methylation, reduction
J2p] [a_ GTP and subsequent digestion with nuclease PI (to generate 5 ' mononucleotides) would be effective in detecting modifications at G residues but would not allow the detection of other modified bases . The tRNA products of the yeast splicing enzymes have a characteristic junction structure consisting of a 2 ' -phosphomonoester-3' ,5 ' -phosphodiester linkage (41). This linkage is resistant to most cleavage agents and therefore yields a characteristic dinucleotide upon nuclease PI digestion of spliced tRNA. However, the nucleaseresistant dinucleotide can be isolated by TLC and subsequently digested with snake venom phosphodiesterase to produce mononucleotide products. Although Protocols 9 and 10 specifically describe the analysis of splice-site junctions, these same procedures apply to the detection of modified bases. Thu s methods and conditions for sample preparation, nuclease digestion, thin-layer chromatography, and quantification can be directly applied to analyses of ba se modifications in pre-tRNA and tRNA substrates .
and isomerization reactions, as well as more complex side-chain addition and
transglycosylation reactions . These reactions are catalysed by an extensive set of site-specific modifying enzymes which are thought ta act at discrete stag es in a tRNA processing pathway (15). The functions of these speci fic modifications, the substrate recognition properties and catalytic mechanisms
of the modifying enzymes, and the ordering of the processing events ar e important areas of study in tRNA biosynthesis. Most current biochemical methods for analysing specific tRNA modification reactions involve the following steps: (a) Incubation of pre-tRNA substrates with processing extracts ; (b) Isolation of the reaction products;
4,3.2 Identification and quantification of splice sites
(c) Digestion of the RNA product to its constituent oligonucleotides, nuc1eotides, or nucleosides; and
Protocol 9 describes a method for identifying the splice-site junction ba sed on sequential digestion of spliced tRNA with nuclease P I followed by separation of the digestion products by TLC on PEl -cellulose. The nuclea se-resistant dinucleotide characteristic of the splice junction is then eluted and subje cted to digestion with snake venom phosphodiestera se to identify its component mononucleotides, again after TLC on PEl-cellulose. Important features of this procedure include:
(d) Separation of the modified constituents from their unmodified counterparts on the basis of distinct chemical or physical properties. The requirements for the first step have been discussed in Sections 2 and 3. Isolation of the reaction products may simply consist of phenol extraction followed by ethanol precipitation (see Protocol 3) or more extensive purification by gel electrophoresis (see Protocol 8) or chromatography (22, 23). Isol ated reaction products are then digested to their constituent oligonucleotides. nucleotides, or nucleosides by treatment with the appropriate nucleases or chemical cleavage reagents. In this step, the effects of modifications on the cleavage reaction must be considered and examples of the use of nuclease PI and snake venom phosphodiesterase are described in Section 4.3 .2. In the final step, digestion products are resolved by TLC (8, 47), HPLC (48, 49), or combined liquid chromatography/mass spectroscopy (LC /MS; 50, 51). The digestion products may be detected and quantified by UV absorbance when unlabelled substrates have been used or by radiometry or autoradiography for the products of labelled substrates. Labelling with [J2p I orthophosphate in vivo (Protocol 3) and post -labelling procedures in which the RNA is labelled after synthesis (47, 52, 53) result in uniform labelling and potentially allow the detection of all the modified nucleotides present in a single digesti on. Where the labelled RNA substrate has been synthesized in vitro, the abilit y to detect a specific modified nucleotide depends both on the choice of the labelled nucleotide and the specificity of the cleavage reagent used in the subsequent digestion (i.e, whether phosphates are retained in the 5 ' or 3 ' position in cleavage products). These two factors must be considered in designing procedures which include in vitro transcription . For example. transcription with 200
(a) The use of PEl-cellulose as an effective and general means for the separation of individual digestion products; (b) Washing with ethanol to remove the solvent components from the developed PEl-cellulose plates. Thi s is essential prior to the elution of the dinucleotide and its subsequent digestion and chromatography of the products. Washing with ethanol is effective only for LiCI-based TLC solvents. (c) The use of triethylammonium bicarbonate (TEAB) for elution of the dinucleotide from the PEl-cellulose plates . This volatile solvent can be removed by drying samples under vacuum, thus facilitating subsequent digestions and analyses.
Protoco l 9 . Identification of splice junctions using digestion with nuclease P 1 and snake venom phosphodiesterase
Equipment and reagents • PEl-cellulose thin-layer plates (20 x 20 em
eel 300 PEl plates; Macherev Nagel or Brinkman Instruments 801 503). Wash the plates in methanol then develop once
using water and finally dry them . Using a soft pencil, mark sample origins 1.5 cm from bottom edge of each plate with at least 1 cm separating individual samples
Continued 201
Chris L. Greer
6: Processing of tronsfer RNA precursors Protocol 9 . Continued
Protocol 9. Continued. e TLC tank (e.g. VWR Sci enti fic,
KT416 180 -0000 1 e u V ligh t sou rce (2 54 om, e.g . Fisher Scient ific 1 1-9 84- 201
e Celtutose powder (Sigma . C6 6 6 3l
e Glass mlcroce pilla rv tubes (e.q. 1 -5~' disp osable pi pettes: Beck t on D ickinson ,
46141 e Seelinq com pound (Seal-Ease. Beckto n Dick inson , 1015) • C apill ary t ub ing (1 x 10 0mm. Kimble
Products , 34502) • Diamond -tip scrib ing pen lVW R Products.
52865·0051 _ C arrier RNA pr epared as in Protoco l 1 • Un labe lled nu cleotide marker m ixture (prepare by addi ng 10 ~g RNas e A t o 250 ~g c arr ier RNA in 250 IJI of 10 mM Tris-Hel pH 7.4 .1 mM EDTA ; in cubate f or 2 h at 37 °C; st ore m ixture at
- 20 ' CI
• Tyg on tubing
• 32P_l abelled sp liced t RNA (labell ed as in Prot ocols 3 o r 4 ; spliced and pu rified as in Protocol 8 1 • Snake (Crotalus aunssus l venom ph os phod iesteras e !Boeh ringe r Mannhe im 108 260; 2 m g /m l in 50 % glycerol) ; st ore at 4 ° C • Nu clease P1 from Penicillium citrinum (Calbiochem 493866 ; 5 mg /ml in 20 mM sod iu m acetate buffer . pH 5.01; store at - 20°C . 20 mM sodiu m acet ate, pH 5 .0
. SVP buffer (2 5 mM Tri s-HCL pH 9 .5 , 5 mM MgCI 2 1 • 1.0 M LiCI • TEAS buffer (30% tr iethylam mon ium bicarbonate in H 20) . Adjust to pH 9 .5 w ith CO 2 (add dry ice shavin.gs or bubble CO 2 through from a CO 2 cylinder). Stor e at 4 ° C f or up to 2 weeks
tubing (include an approp riate trap to prevent fine particulates from entering the pump). Plug the tip of the capillary by drawing up a small amount of cellulose powder. With a clean rnicrospatula , scrape the area of the PEl plate co ntaining the desir ed digestion produc t . using th e vac uum to s uck the PEl- cellulose flakes into the ca pillary tube . 6. Elute the dinucleotide by torcing a small volume 120-100 ~I) of TEAS buffer through the tub e (from the wide end of the fine end) using a syringe attac hed via fine plast ic tubin g . Collect t he eluate in a microcentr ifuge tube and dry it unde r vacu um. 7 . Resuspend the residue in 1 0 -2 0 ~ 1 of water and again dry it under vacuum . Rep eat the resuspen sion and drying t w ice more. 8 . Resus pend the dried nucleotid e in 4 ~ I of SVP buffe r and add 10 jJg of carrier RNA. 9 . Add 1 jJg of snake venom phosphodieste rase and incubate t he mixture in a seale d microcap illary tu be (ste p 2) for 1 h at 37 °C. ·10 . Break the ends off the tu be and s pot th e s ample on to a PEl- cellulose plate las in step 31. Develop the plates with 1.0 M LiCI and visualize markers and labelled products as described in steps 3 and 4 .
Method Precip itate (32p] tRNA. labelled as in Proto?ol 3 or 4 and spliced an d 1. purified as in Protocol 8 . with 20 J.lg of carrier RNA (Protoc~1 1).. 2. Add O.5 ~l of nuclease P1 and transfer the mixture to a gdlaSsd~lcrocbaPt III:~~ tube . Seal the ends of the tube with sealing compoun an InCU a e tube for 2 h at 37 °C. 3. Scr ibe the ends of each microcapillary tube just abfofve the II se~lpi:~ compound using a diamond-tip pen and break the en dso manua y. the sample at the marked origin on a PEI-~ellulose plate . Also spot 5 J.lg unlabelled nucleotide marker mixture alongside the exper~mental sample ts l. Allow the plate to dry at room temperature for 5 -10 min. Place the ~I~te with origin spots at the bottom in the chromatography ~ank comamms 1 M LiCI to a depth of 0.5-1 cm . Develop t he plate until the solvent 's 1-2 cm from the top edge of the plate. 4 . Wash the plate by soaking it twice (for 10 min each ti.mel in 75 % eth anol. Allow the plate to air dry. Visualize the nucle?tlde markers . by UV illumination (at 260 nrn) and outline each spot With a soft ~encil. Mark the corners of the plate with radioac tive or fluorescent Ink f~r la ~e r alignment of film and plate . Wrap the pla~e in plastic cling filmand visua li ze labelled digestion products by autoradiography . . 5 . Elute the nuclease-resistant digestion product corresponbdin g to ;.h~ Spp~\~~ junction as follows. Draw out t he tip of a capillary tu e .to a !n and connect the drawn-out end to a vacuum pump usmg thin Tygo n
Continued 202
An example of the application of the method described in Pro tocol 9 is show n in Figure 4. Here pre -tR NAPho was labelled by in vit ro transcr iption (Protocol 4) using [a- J2P I GTP, -ATP , o r -CTP _Spli cing react ions wer e carried o ut using yeast endonuclease and DHFR-ligase and the spliced products were then separated by gel electrophoresis a nd eluted (Protocol 8). Eluted tR NA was p recipitated with 20 pg of ca rrie r RNA (P rotocol I) . Digestion with nu clea se PI was followed by T LC on PEl - cellulose (Figure 4A) . RNA subs trates labelled with ["PjGT P or [J2PjATP gave rise to the radioactive splice site dinucleotide pG ~A , whereas [32P J CTP did not. The dinucleot ides were treated with snake ven om phosphodiesterase and their monon uc leo tide p ro d uct s were separated b y TLC on PE l-cellulose (Figure 48) . An alternative procedure for investigation o f tRNA splice junctions is descri bed in Protocol 10. This makes use of combined RNase A /RNase TI digestio n of the spliced t RNA pr ior to nuclease PI digestion. The prod ucts of PI digestion are then separated by two-dimensional chromatography on cellulose thin-layer plates. The two-dimensional system described in Protocol 10 has been exceptio nally well characterized for the separation of a wide range of modified nucleoti des. Ref. 23 should be consulted for an extensive catalogue of the relative mobilities of modified nucleotides in thi s system. Identification of labelled digestion products is facilitated by inclusion of appropriate unlabelled markers which can be visualized under UV light. 203
6: Processing of tran sfer RN A p rec ur sor s A.
B,
C, - - NpG:CpN--
:"'i
..
, "
"
pC-
Chris L. Gree r
.--
,'';
..
I
,- ,
_pu-
-, »
_pC-
' pA -
pA- -
' p G ~A -
t'
t •
•
2
3
(/
- Ori 4
-On
7
PO
8
9
•
• Pi
pG
+ pC
- pGP
' p G ~ A .r-
6
Pc · pt
'I
•
• ' pG P~
pG'
'I
• t
..-:
.
lsvp
"
........ pG /
'pG -
.
PI
p G ~ C + pN
.... _
,',
• Ori
-,-
Fig ur~v~ ' Ide ntification of splice -site dinculeotide by 2 -D c h ro m ato grap hy. Sp lic ed pre tR.NA ( u n ~ f o r m l y labelled) w as dige sted w ith RNase A and RNase T 1 and the oligonu cleot id e produ ct ( AA~Al{pl c on t ai nin g the spliced jun c ti on wa s iso late d by chromatograph y on PEl -cellulose th in -layer plat es (Pro tocol 9 1. Th is elu t ed ol ig onucleot"d e w as t h en d ig ested wi th nuclease P 1 an d prod ucts were resolved by two -djme nSjo~al chromato.g~aph Y ~n a. c.ellulose plate (Protocol 10). For de ta ils , see legend t o Figure 4 . 't pseudourldlne ; on. onqm . '
5
Figu re 4 . Identification of sp lice site . (A ) Pre -tRNA """' substrate was labelled by in vitro transcr ipt ion with (o:_37 p ] GT P (lane 11. [a _J 2 P j A TP (lanes 2 , 3 1, or [ o:-32 P ] CT P (lane 51. Su bstrates were incubat ed in sp licing reactions along w ith carrier RNA , and the splic ed tRNA pro du cts were then pur ified b y gel electrophoresis. Samples were th en d igest ed w ith nucl ease Pl and the digestion products were resolved by chr omatog rap h y on PEl - cel lu lose th in-la yer plates and v isua lized by autorad iography (Proro co l 9 1. L ane 4 contain s a sample of the c arrier RNA alone processed in t he same way. lS) Spots corresponding to the nuclease P 1-resistant din ucleotide pG ~A were elu ted f rom th e pla te described in panel A and rechromatog raphed in this same system w ithout further incu bation (lan e 61 or after secondary dige st ion wi t h snake veno m phosphodiesterase (lan es 7 - 9 ; c orrespond ing to lanes 1-3 , respe ct ivel y , in panel A I. In panels A and B, the identi ti es of labe lled products are ind icated with asterisks. Un labe lled ma rkers , visualized under UV light , are outlined w ith dots and iden t if ied at left or righ t by symbols lackin g ast erisk s. Samp le orig ins are indicated by Ori. IC ) A summary of the sequential digestion pro cedure; Pl and SVP refer to nu cleas e Pl digestion and snake venom phosphod iesterase dig est ion. resp ect ivel y.
Protocol 10 , Ana lysis of tRNA splice junctions using nuclease digest ion a nd t w o -diman s lnn a] c hro matog raphy Equipment and reag ents
e Ceuurcse
thin -layer plates 110 x 10 cm cut from 20 )( 20 cm sheets ; Eastma ~ Kodak 13255); mark samp le origins with a soft pen cil at the bottom left c orner , 1 cm in fr om the adj acen t edges
- First -dim ensional solv ent ( 100 ml o f iso butyric acid , 60 ml of 0 .5 M NH 40H) ; store in a fume hood • Second -d imension al so lvent 11 12 m l of pr opan- g -ol , 46 ml of 6.0 M HCIl
- J2P·labelle d splic ed tRNA , nuclease Pl , unlabelled marker nucl eotide m ixture as in Proroco l 9 . • RNase A {bovine pancreat ic ribonuclea se (Catbioch ern, 55674 )[, 5 m g fml stock in 10 mM Tr is-HC I. pH 7.4 , 1 mM EDTA • RNase T 1 /Aspergillus oryzae ribo nuclease (Calbio chern. 556785 ) l. 5 Uni ts /ml stoc k in 10 mM Tris -H CI, pH 7.4
Meth od Figure 5 provides an illustration of this procedure, Here the pre-tRNATY' subs trate was uni formly lab elled in vivo with [J2p 1ortho phosphate (Protocol 3) and then subjected to splicing as in Protocol 8, After digestion of the spliced RNA with a mixture of RNase A an d RNase T1 the oligonucleotide AA gA,"P containing the spliced junct ion was isolated by chromatography on PE l-cellulose thin-layer plates, The eluted oligonucleoti de was digested with nuclease P I and the products were resolved by two-dimensional chro matography on cellulose plates (Protocol 10 ),
~ B Th~ iso but yric acid used in the chromatography first-dimensiona l s ol ven t IS noxIOUS and a strong irritant' carry out step 5 in a fume hood. 1.
~re pare? sp liced tA NA as described in Protocol 8 . except resuspend the 10 1in 8 ~I ot 10 mMTris-Hel, pH 7.4 , 1 mM
~~~:p" ced RNA pellet (step
2, Add 1 ~ I, carrier RNA, 0, 5 ~ I of 5 mg/ml RNase A, o.s ~ I ot 5 U/ml RNase Tl and Incubate f or 2 h at 37 0C ,
Con tinued
204 2 05
6 : Proc es sing of tran sfer RNA pre c ursors Protocol 10. Continued 3 . Spot the sample on a PEl-cellulose plate, deve lop with 1 M Lie l and recover the jun ct ion oligonucleotide (this ol igonucleo tide is the slowest m igr at ing o f th e digest ion products) as described in Protocol 9 . st eps 3 - 7.
4 . Resuspend the dried oligonucleotide in 4 ~ I of P1 buffer , add 0 .5 1-11nuclease P' and transfer to a glass microcapillary tube for digestion as des cribed in Protoco l 9 , step 2. 5 . Break the tips of the microcapilla ry tube (Protoco l 9 , st ep 3 ) and app ly 3 -51-11 of t he samp le (in 1 IJI increments with intermittent drying) at the origin of the cel lular plate. Develop th e plate w ith th e fir st -dime n sional sol vent until t he front is 0 .5-1 em from the top edge of the plate . Allow the plate to dr y in a f um e hood overnight . 6. Turn the plate through 90 0 and develop it in the sec ond dimensio n with the second-dimensional solvent un til the front is 0 .5 -1 cm from t he top . Dr y the plat e and vis ualize the marker nucleotides and labelled p rod uct s as desc ribed in Protocol 9 (step 41. Note that guanosin e and man y of it s derivatives fluoresce under UV light in propan -2 ·ol:HCI. Also , th e pre sence of excess RNA (:;::. 20 I-Ig ) causes guanosine to streak in the first dimen sion. 7 . Align the plate and autoradiograph on a light box, marking t he positions of radiolabelled spots with a soft penc il . Spots c an then be cut out, w it h sharp scissors or a razor blade , to measu re radioactivity . Y ields of individual din ucleotide s or modif ied bases ca n be cal culated relative to t he y ield of one or more ot her bases of known yi eld.
Chri s L. Greer 7. Bjork. G. R., Ericson, J . U Gustafsson C Y. H ., and Wikstrom P ~i' (1987) A ' . E. D .~ Hagervall , T. G ., Jonsson, . nnu . ne« BlOchem., 56. 263. 8. Nishimura S (1979) I' T" . • . . n ransfer RNA' structu . D . Soli. J. Abelson and P R S h' " re, aroperues and recognition (ed. Press, Cold Spring' Harbo~ , NY~ rmmet}, p. 547 . Cold Spring Harbor Laboratory 9. Rich, A. and RajBhandary U L (197 10 Kim S -H (1979) I T ' " 6). Annll. Rev. Bioehem., 45 805. . ' " . n ransfer RNA' strucu . • D. Soil J Abelson and P R S hi . I ire. properties and recognition (ed . ~ rrnmet), p. 83 . Co ld Spring Harbor Laboratory Pr ess, Coid Spring 'Harbo; ,
NY
: ~ . ~ormanlY, J . .
13 14 15: 16 17' 8' I .
and Abelson, J . (1989). Annu. Rei'. BlOchem ., 58, 1029 . eurscher. ~f. P . (1984). CRC Crit , Rev. Biochem 17 45 ?,euI " Cher, M. P. (1990). Progr. Nucleic ACids Res:' M;1 BIOI J9 '09 " e ton D A D R b . . "', . . Nishik~ra .K '~ndeD~Rer~ts, E . EM' : tand Cortese, R. (1980). Nature, 284, 143. ,. 0 ertis, . " . (1981). J. M ol BIO I 145 405 O'C Ho onno r, J ., P . and Peeb les, C. I.. (1991). M ol. Ceil. Bt~i., 11', 425'. _ pper, A. K., Banks, F., and hangelides, V. (1978). Cell, 14 211 Knapp, G.. Beckmann J S Jo hn P F ' . ( 1978). Cell, 14, 221. ' " , son,. ., Fuhrman, S. A ., and Abelson, J .
19. Wallac~,. D. M. (1987). In M ethods in enzymologv, Vol. 152 led S L B . . . erger and A. R. Kimmel), p. 33. Academic Press, San Diego. 20. Ogden, R. C. and Ad ams D A (1987) S L Ber er an ;. " . n ,¥ ethods In enzymology, Vol. 152 (ed, II: ' g d A. R. Kimmel), p. 61. Academic Press San Diego 21. .emura. T. (1989). In M ethods in enzymology, Vo l. 180 'led J E D hlb J. N. Abe lson), p. 14. . . . a erg and
"f
.
22. Tanner, N.K.(1989).l nMethods in enzy m ology Vo1. 180(e d JE D hlb J . N. Abelson), p.25. ' . . . a erg and
Acknowledge me nts H ea ther Belford and Shawn West away are gratefully a cknowledged for help ful
comm ent s, assista nce with experimental procedures, and communication of result s prio r to publication. T h is work was su ppo rt ed by grants from th e NIH (GM-35955) a nd NSF (DMB-86 14092) . c.L.G. is an Est ablished Investigato r o f t he American Heart A ssociatio n .
References I. Nichols, M., Willis, I. , and Soli, D. (1990). In M ethods in enzymology, Vol. 181
(ed . J . E. Dahl berg and J . N. Abelson), p. 377. Academic Press, San Diego. 2. Hopper, A. K. (1990). In M ethods in enzymology, Vol. 181 (ed. J. E. Dahlberg and J . N. Abelson ), p.400. Acad emic Press, San Diego . 3. Schimmel, P. R. , Soli, D.. and Abelson , J. (ed.) (1979). Transf er RNA: structure, properties and recognition. Co ld Spring Harbor Laboratory Press , Cold Spring Ha rbor, NY. 4. Soli, D., Abelson, J . N., and Schimmel. P. R. (1979). Transf er RN A: biological aspects. Cold Spring Harbo r Laboratory Press, Cold Spring Harbor, NY. 5. Spinzl, M., Dank , N. , Nock, S.. and Schon, A. (1991). Nucleic Acids Res. , 19 (suppl. ), 2127. 6. Cedergren , R. J .. Sankoff, D.. LaRue, B.. and Grosjean, H. (1981). CRCCrit. Rev . Biochem . , II , 35. 206
23. Nishimu ra, S.• Shindo·Okada, N., and Crai n P . " . F. (1987). In M ethods m enzymology , Vo l. 155 (ed R Wu) 373 . Academic Pres s, San Diego . 24. Sherman F. Fink G R ' ci H' ,p. an icks, J . B. (1986). Methods in yeast genetics . Co ld Spring H'a rb' or a oratory Press, Cold Spri ng Harbor NY G " . ' 25. Ogden, R. c.. Lee .\ 1 -C and K 26 K ' ' " napp, . (1984). Nucleic A cids Res 12 9367 . napp, G . (1989). In M ethods in enzymology, Vol. 180 (ed J ED hlbe: . J . N. Abelson), p. 110. Academic Press, San Diego. '" a erg and 27. Greer, C. 1.., Javor, B., and Abelson, J . (1983). Cell 33 899 9 . ". 28. Shapero, M H and G e L 29. E Ik 0 reer, . . (I 91). BIOchemistry . 30 6465 30. R~~~m:~ j G~~e;heimer, P" ,and Abelson, J . (1985). J. BioI. Chem. , 260, 1271. A cad. sci. us~', 83, ; ;.;;. P . , \ ails, I.. A. , and Stevens, T. H . (1986). Proc. Natl
Lb ' .,
R' H
31. Evans, C. F. and Engelke D R (1990) I . (ed JED hlb '" n M ethods In en::ymology Vol 181 . . . a erg and J N Abelson) p 439 Ad ' , . 32. Phizicky EM S h . 'R ' ' " ca ermc Press, San Diego . , artRz, . C ., and Abelson, J . (1986). J. BioI. Chem 261 2978 33 Rauhut ' R ' G" c 'p · , ., reen,. and Abelson J (1990) J B ' I '". 34 Green P Rand Ab I ., J ' .. " 10. Chem., 265, 18 180 · J . E. Dahlb~r an e s?n, . (1990). In M eth ods in enzymology , Vo l. t8 1 (ed: 35 Peebles C g d J. N. A belso n), p. 471. Academic Press, San Diego . · DR b' .: 1.., Gegenhe imer, P., and Abelson , J . (1983) Cell 32 525 36. e 0 erus E M Black P d N' . " , . 37 ' ' ., , . , an ishikura, K. (1981) Cell '3 89 · Aposto l, B. L Westaway S K Ab I " . , . Chem ., 266, 7445. , .., e son, J ., and Greer, C. I.. (1991). J. BioI. 207
6: Processing of tronsfer RN A precursors
Ch ris L. Greer
38 . de Soer , H . A . , Comstock, L. K., and Vassar , :\1. (1983 ). Proc. Na tl A cad. Sci.
69. Darr , S. C.; Pace, B.. and Pa ce, N. R. (1990). J. Bioi. Chem. , 265 , 12 927. 70. Manam , S. and Van Tu yle, G . C. (1987). J. Bioi. Chem., 262, 10 272. 71. Doer sen , C .-J. , Gu errier-Takada, c., Altman, S., and Attard i, G. (1985). J. Bioi. Chem .. 260, 5942.
39. 40 . 4 1. 42. 4 3.
USA, 80 , 21. Baccanari, D. P ., Stone, D., and Kuyper, L. (198!) . J. Bioi. Chem. , 256, 1738. Brad for d , M . M. (1976). Anal. Biochem .. 118, 197. G reer , C. L., Peebles, C. L. , Gegenheimer, P. , an d Abelson, J . (1983). Cell, 32, 537. Nagai, K. and Thoger sen, H . C. (1983). In Methods in enzymology, Vol. 153 (ed . R. Wu and L. Gro ssman ), p. 461. Academi c Press, San Diego. Pe rlman , P . S. , Peebles, C. L., and Dan iels, C . (1990) . In Intervening sequences in evolution and develop ment (ed . E. M . Stone and R. J. Sch wartz) , pp. 112, O x ford University Press, Oxford.
44. Greer, C. L., Soli, D., an d Willis, I. (1987). Mol. Cell. Bioi.. 7, 76. 45. Leirm o , S.. Harrison, C .; Ca yley, D. S., Burgess, R. R.. and Record, M. T . (1987). Biochemistry , 26, 2095. 46. Apostol, B. L. and G reer, C. L. (1991). Nucleic A cids Res.. 19, 1853. 47 . Kuchino , Y. , Hanyu. N., and Ni shimura. S. (1987). In Me thods in enzymology , Vol. 155 {ed. R. Wu), p. 379. Academic Pr ess, San Diego . 48. Buck, M. , Connick , M ., and Am es, B. N. (1983). Anal. Biochem.. 129, I. 49. Gehrke, C. W. and Kuo . K. C.; and Zumwalt, R. W . (1983). In The modified nucleoside of transfer RNA, II (ed . P . F. Agris, A . A. Kopper , and G . Ko pper), pp . 59. Allan R. Liss Inc., NY. 50. Edmonds, C. G., Vestal, ~I. , and McC loskey, J. A . (1985). Nucleic Acids Res.. 13, 8197. 51. Pom eran z. S. C. and McCloskey, J . A . (1990). In Methods in enzy mology, Vol. 193 (ed. J. A. ~IcCloskey), p. 796. Academic P ress, San Diego. 52 . Randerath. K., Gupta , R. C., and Randerath, E. (1980) . In Ate/hods in enzvmotogv, Vol. 65 {ed. l. Gro ssman and K. Mo ldave). p. 638 . Academic Pre ss, New Yo rk. 53. Silberklang, M., Gillum, A. M .. an d RajBh an dary, U. L. (1979). In Methods in enzymology, Vol. 59 (ed , K . Molda ve and L. Gross ma n) , p. 58. Academic Press, New York . 54. Chen, J.-Y. and Martin, N. C. (1988). J . Bioi. Chem., 263, 13 677. 55. Krupp, G ., Cherayil, B., Frendeway, D., Nishkikawa, S.. and Soli, D. (1986). £ M BO J., 5, 1697. 56. Leo nd s, N., Dal.io . A . , Stro be l, M . , and Engelk e. D . (1988) . N ucleic Acids Res.. 16 , 2537. 57. ~l c Ci a i n , IV. H ., Gue rrie r-Takada , c., an d Altman, S. (1987). Science, 238. 527.
72 . Gruissem, W. , Green berg, B. ~ 1. , Zuraws ki, G ., Presco tt, D . M., and Hallick R.B.
(1983). Cell, 35 , 815.
'
73. G reen berg, B. M . and Hallick, R. B. (1986). Plant M ol. Bioi. , 6, 89. 74. Wang, I\1. J ., Da vis, N. W., and Gegenheimer , P . (1988). £ M BOJ. , 7,1 567. 75. Kikuchi, Y ., Tyc. K., Filipowicz, IV., Sanger, H. L., and Gross, H . J . (1982). Nucleic A Cids Res. , 10, 7521. 76. Gegenhei mer, p .. Gabius, H .-J ., Peebles, C. L. , and Ab elson , J . (1983). J. Bioi. Chem., 258, 8365. 77. Pi ck, L. and H urwitz, J . (1986). J. Bioi. Chem., 261, 6684. 78. Sta nge, N . and Beier , H . (1987). £ M BO J. , 6, 2811. 79. Frendeway, D. , Dingerman, T ., Cool ey, L. an d Soli, D. (1985). J. Bioi Chem 260,449. . ., 80. Ga rber, R. L. and Altman , S. (1979). Cell, 17, 389. 81. Birkenmeier, E . H ., Brown, D. D., and Jordan , E. (1978). Cell, IS , 1077. 82. Ogde n, R. c ., Beck mann , J . S. , Abe lson , J . N., Kang, H. S., Soli, D., and Schmid t, O. (1979). Cell, 17, 399. 83. Gan dini-A tta rdi, D ., Marg ar h , I., and Tocchini-Valennn f G . P . (19 85) . EAfBO J. , ~ , 3289. 84. Gruissem , W ., P rescott , D. M. , Greenberg, B. M., and Hallick R. B. (1982) ell 30, 81. ' . en, 85. Laski , F. A ., Fire, A. Z ., RajBhandar y, U. L, and Sharp , P. A . (1983). Proc. Natl A cad. SCI. USA, 79, 5813. 86. Dignam , J . D., Lebo witz, R. M ., an d Roeder , R. G . (1983). Nu cleic A cids Res II , 1475. ., 87. Zas lof f, M ., Santos, T. , Romeo , P ., and Rosenb erg , .\ 1. (1982). J. Bioi Chem 257, 7857. . ., 88. Herrenknecht, K., van Tol, H ., and Gross , H . J . (1988). Nucleic A cids Res. , 16, 7713. 89. Laemmli, U. K. (1970). Natu re, 227, 680.
58 . Shapero , M . H . (1990) . Ph D Th esis, U niversity of Cali fo rnia, Irvine .
59. Reyes, V. M. and Abelson , J . (1987). Anal. Biochem.. 166, 90. 60. Hall, K. and Sampson, J . R. (1990). Nucleic Acids Res.. 18, 704 1. 6 1. Samuelsson, T. , Boren. T. , Johansen, T. -I. . and Lustig , F. ( 1988) . J. Bioi. Che rn.. 263, 13 692. 62. Kru pp, G.. Kahle, D., Vogt , T ., and Char, S. (199 1). J . Mol. Bioi.. 217, 637. 63. Rob ert son, H.. Alt man, S.. and Smith , J . (1972) . J . Bioi. Chem .. 247, 5243. 64. Fourn ier. M. J ., Webb, E., and Tan g, S. (1977). Biochem istry , 16, 3608. 65. Chudney. H. K. and Deutscher , M . P . (1988) . J. Bioi. Chem. , 263, 1518. 66 . Gardiner. K. and Pace, N. R. (1980). J. Bioi. Chern.. 255, 7507. 67. Har tmann, R. K.. Tos chka , H . Y., and Erdmann, V. A . (1991). J. Bacteriol. , 173, 268 1. 68. Th ompson, L. D. and Daniel s, C. J. (1988). J. Bioi. Chem. , 263, 17 951.
208
209
Ribozymes DAVID A. SHUB, CRAIG L. PEEBLES, and ARNOLD HAMPEL
1. Introduction The discovery of ribozymes (catalytic RNA) by Cech (I) and Altman et al. (2) led to exciting advances in nucleic acid chemistry and biology. In addition to a key role in the processing of certain RNA transcripts, evidence supports the broader involvement of catalytic RNA in the origin of life and evolutionary adaptation. This chapter focuses on the catalytic properties of group I introns, group II introns, hammerhead ribozymes and hairpin ribozymes .
2. Group I intron rib ozym es 2.1 The self-splicing activit y of group I introns The term ribozyme was coined to describe the catalytic properties of the intron in the large subunit ribosomal RNA of Tetrahymena thermophila (3). It was soon noticed that this self-splicing intron also satisfied the sequence and structural criteria for a set of introns in yeast mitochondria, designated group I by Michel et al. (4). Well over 100 group I introns have now been described (see ref. 5 for a recent compilation), most of which are in mitochondrial and chloroplast genes. However, in addition to their presence in nuclear rRNA genes of eukaryotic protists, introns of this group have also been found in genes of bacteriophage (5) and bacteria (6). Group I introns occur in genes of every kind, i.e. encoding rRNA, tRNA and mRNA . Although many of these introns (like their Tetrahymena exemplar) self-splice efficiently in vitro, others require protein co-factors (7). However. since they all share a set of structural features necessary for self-splicing (5), it is likely that all group I introns also share a common RNA-based catalytic mechanism, protein co-factors merely acting to help the RNA achieve the correct conformation . In common with the splicing reactions of group II introns in organelles (see Section 3) and snRNP-assisted pre-mRNA splicing in the nucleus (Volume I, Chapter 3), group I cleavage-ligation is accomplished through a series of trans-esterifications (Figure I), Unlike the other splicing pathways, the
David A . Shub , Craig L. Peebl es, and Arnold Hamp el
7: Ribozymes
even under co nd itio ns that are optimal fo r spli cing in vitro, The products of this hydrolysis do not go on 10 splice, and so th eir pre sence can complicate the analysis of in vitro splicing reactions.
r>
I El
·G,/ EI
-,
E2
2, 2 Detection of self- splicing introns in cellular RNA
OH -
l + 'G~ OH
EI
~~ .
~
.S~G
QOH
EI
~
+I
E2
OH
<, -G -
OH
OH
+V
------':>
QG OH
The addition of GTP to the 5 ' -end of excised introns (Figure J) can so metimes be used to demonstrate the presence of self-splicing group I introns in an RNA preparation. The advantage of this reaction is that it does not require any previous knowledge of intron or exon sequences or the ident ity of the interrupted gene. In practice, detection of group I self-splicing introns involves incubating deproteinized RNA, extracted from the organism under study, with [a- J2P IGTP and analysing the products for the presence of radioactively-labelled RNA s by denaturing polyacrylamide gel electrophoresis. The procedure is described in Protocol J. Protocol 1. De te c t io n of group I self-splicing introns' Equip m ent and reagents
Figure 1. Splicing path way and related react ions of group I int rons. The splicing path w ay is initiated by labelled GTP (-GTPl binding on a self- splicing RNA . The labelled GTP att acks the 5 ' splice junction and att aches to the S'<end of th e intran ( -G- thin line ) in th e first step of splicing . The 3 '-OH on th e liberated 5 ' exon (open box- E1QHI atta cks the 3 ' splice junction to join to the 3 ' exon (open box -E2 1and for m th e spliced exon s (ope n box -E1-open box -E 2 1. The linear lnt ron is released with the labelled GTP at the 5 '·end and the conserved G at the 3 '.end COG- thin line-G )' C ircularization by attack of the 3 ' GO+-! at an internal site rele ases oH the labelled GT? as part of a small oligonucleotide ( - G-short thin line - oHI . After prolonged incubation or at somew hat elevated pH , the circle is atta cked by hydroxide (O H - I to yield a shorte r linear intr on fragmen t (t hin line -G Oti I . Likewise for t he tran script , at high pH or in the absence of GTP , th e 3 ' splice junct ion is subject to atta ck by hydro xide ion tOH - 1. a non -product ive side react ion that yields the 5' exon attached to the intron (E1 -thin line - oH l and the 3 ' exo n (open box-E2 1.
. 5 x GTP -Iabelling buffer lO.25 M NH 4C l, 0 .'5 M MgC I, . 0 .'5 M Tris-HCI . pH 7.5 . 5 mM spermidine)
• (a _3 2P] GTP 13000Ci/m mol, 10mC i/ml) • Stop solution (2. 5 M ammonium acetate, 25 mM sodium EDTA , pH 8.0 ,300 IJg /1J1 yeast RNA )
e O.2 M OTT • RNas in tProm ega, 3 0 units/ mil
e Total RNA (5 mg /mll. isolated from the organism under study as described in Vo lume I, Ch apter 1. Prot ocols 9-11 • Ma terials and equipmen t for denatu ring polyacr ylamide gel electrophoresis and aut oradiography (Volume I, Chapter 1, Protocol 8 or V olume I, Chapter 4 , Pro tocot t t v; for RNA samples 100-600 nt in length, a 5 % polyacrylamide gel with an acrylamide:bisacrylamide ratio of 3 0: 0. 8 is suitable
Method 1. In a rnicrocentr ifuqe tube on ice. mix the follow ing:
attacking nucleophile in the first step of the group I mechanism is the 3 ' -hydroxyl group of guanosine or GTP. Consequently, the excised linear intron has an extra , uncoded G residue at its 5' -end. Thi s reaction can be used to identify and characterize the self-splicing property of group I introns. Some excised introns can cyclize; the 3 ' -hydroxyl group attacks a phospho diester bond near the 5 ' -end of the intron, with elimination of as ' oligonucleotide (3) . The circular intron can then reopen, leading to additional rounds of circularization and reduction in length. This can result in a mixture of linear and circular intron products , of various sizes, on prolonged incubation . Finally, the phosphodiester bond between the final nucleotide of th e intron and the first residue of the downstream cxon is especiall y sensitive to hydrolysis. For so me introns this reaction (Figure I) can be significant, 212
2
~I
e O.2 M OTT
0 .25
~I
e RNasin
0 .33
~I
e 5 x GTP-Iabelling buffer
e [ a - 32 Pj GTP
e RNA substrate
e H20
2~1
2
~I
to 10 ~I final volume
2. In c~bate the mixture at 37 °C for 1 h and then stop the reaction b y addinq 150 IJI of stop solution chilled on ice and 320 IJ\ of ethano l cooled to - 20 cC. 3 . Leave the mixture for 30 min at - 70° C.
Continued
213
7: Ribo zymes
DO\'id A. Shub. Craig L. Peebl es. ond Arn old Hampel
Protocol 1 . Continued
• the requirement for cellu lar components to facilitate splicing in vitro
4 . Recover the precipitated RNA by centr ifugation for 15 min at 4 ° C in a microcentrifuge.
• the low abundance of unspl iced precursors in the cells because the intr on s are loca ted in rare transcripts
5 . Wash the RNA pellet by resuspending it in 1 ml of 70% ethanol and centrifuging as in step 4.
• the RNA precu rsor s splice so rea dily in vivo that the co ncentration of self-splicing subst rates in the RNA isolated from the orga nism is too low for
6 . Repeat step 5 and vacuum dry the washed RNA pellet. 7. Analyse the RNA products by denatur ing polyacrylam ide g~l electro phoresis. To do this, dissolve the RNA in 8 ~I of H 20. Immediatel y a~d 8 ~I of formam ide-dyes loading solution and denature the RNA for 1 5 mm at 65 vC. Load the samples onto a denatur ing 5 % polyacrylam ide gel. fun the gel at 200 V overnight and autorad iograph it as described in Volume I, Chapter 1. Protocol 8 or Volume I, Chapter 4 . Protocol 11 . Autorad iographic exposures of up to 1-2 weeks might be necessary to detect weak signals. ~
Based on ref. 8.
The conditions in Protocol 1 are given as a guide and are not necessarily optimal. Many self-splicing introns requite incubation at higher temperatures (50 ' C or more) and a buffer whose ionic strength is optimal for splicing in vitro . Also , since circularization eventually remove s the labelled guanylate residue from the 5' -end of the intron (Figure I), th e time course of the splicing reaction should be investigated . The concentration of Mg?- ion s should also be optimized; circle formation may require a higher concentration than splicing. Since RNA is readily degraded, precautions must be taken against contaminating RNases during all steps. Volume I, Chapter 1 (Section 1.1) and ref. 9 list general precautions that should be taken when working with RNA . The choice of RNA isola tion method (Volume I, Chapter I, Section 3) depends on the organism of interest. Naturally, the gentlest method consi stent with good yield shoul d be used . The pre sence of a labelled band on the auto radiograph is a good indicat ion of a self-splicing intr on. Some of the other possible interp reta tions can be eliminated by treating part of the sample with alkaline phosphatase before electrophoresis. Since group I splicing leaves the " P-Iabelled phosphate exposed on the 5 ' -end of the excised intron, it should be removed by the phosphatase. In addition to the excised intron, the intermediate splicing product , consisting of the intron plus the 3 ' -exon (Figure!), can somet imes be seen as a second labelled band of lower electrophoretic mobili ty. More than two prominent bands may indicate multiple introns (see ref. 10 for an exampl e). Thu s, a positive result can reveal the presence of a gro up 1 int ron and allow its size, and that of the down stream exon , to be calculated. However, failure to detect end -lab elled RNA in this assa y cannot be considered proof of the absence of group 1introns, since the assay can fail to detect group I introns for various reason s includin g: 214
detection
2 .3 Confirmation of self-splicing by in vitro transcription of
cloned DNA containing putative group I introns Analysis of the products of in vitro transcription can provi de eviden ce for self-splicing of cloned putative intro ns. If self-splicing occ urs efficient lv in vitro, spliced products corres po nding in size to the ligat ed exons and to the excised intron should be detectabl e in addition to the pr imar y run -of f tr anscript. For this assay, the DNA fra gment containing the put ative gro up I intron has to be cloned into a vector which cont ain s pro mot ers for RNA polymera se adj acent to the cloning site. Th ere are man y commercially-available plasmids which can be used that cont ain multi ple cloning sites flank ed on eithe r side by promoter s fo r SP6. T3, or T 7 bacteriophage RNA polymera ses. Use of both the 5 ' and 3 ' flank ing promoters allows tra nscript ion of the cloned insert in either directio n. Since prematurely-terminat ed or partiallydegrade d tran script s can co mplicate the analysis of RNA produ cts aris ing fro m splicing, the in vitro transcription mixtu re should be run on a denaturing po lyacrylamide gel and the full-length tr an script isolated by elut ion (Volum e I, Chapter I, Proto col 8 or Volume I, Cha pter 4, Protocol Il) . In addition , the authors recommend th at the templ at e should not exceed 1-2 kbp for efficient transcripti on into a full-length product, best resnlts being ob tained with the sho rtest tra nscripts possible. Th erefor e, run -off transcripts shou ld nat extend into th e vector seque nces. This can be ensured by truncatio n of the DNA templat e at the multiple cloni ng site or within the cloned insert by digestio n with suitable restriction enzymes. Volum e I, Chapter 1 (Section 2.2) a nd Chapter 2, this volume (Section 2.6) give addi tional information abo ut in vitro tran scription meth od s. Proto col 2 describe s the in vitro tr anscription and an alysis o f sequences containing group 1 self-splicing introns. Templates based on pBS-M 13 + (Stratagene) and isolated by th e alkaline lysis minipreparation meth od ( 11) without RNase treatment are usually adequate for in vitro transcripti on. The DNA samples should be linearized by restricti on digestion and then extracted with phenol and ethanol precipitated . Co nta minat ion with RNases must be avo ided du ring the in vitro transcription a nd subsequent ste ps; the reaction mixture described in Proto col 2 contains RNasin but other precaut ion s shou ld also be taken as advised in Volume I, Cha pter I, Section 1.1. The GT P concentrat ion present in the transcr iption reaction mixture is sufficient to allow self-splicing of group I intro ns . 215
7: Ribozymes
Protocol 2 . In vitro tra ns c ript ion of cloned group I introns Equipment and reagents • (a_3 2 P j U TP l20 m C i/m l, 800 C i/mmol)
e RNA poly m erase (50 units /ill) f rom T3 or T 7 phag e. depend ing on the nature of
. 5 x transcription buff er
(0 .2 M T ris-Hel , pH 8 .0 , 40 mM MgCI 1 . 10 mM spermidine . 0 . 2 5 M NaC IJ
e O.75 M DTT • RNas in (Prom ega. 30 un its/m il
the vecto r used • D NA tem plate (O. 2 - 0. 5 IJg /1J1) con taining the cloned sequence of intere st down st ream of a T 3 or T7 promoter • Material s and equipment tor denaturing
e O. 5 mM UTP
polyacrylamide gel electr ophoresis and aut oradiography (see Prot ocol 1 )
• 10 mM stock solutions of ATP , CTP . and
GTP
1. Make up the follow ing reaction mixtu re (suff icient for 8 transc ription reactions) at room tempera ture to avoid forming a precipitate:
e RNasin
8 IJI 8~1
1 ~I
e lO m M ATP
1.6 ~1
e l 0 m M CTP
1.6~1
e 10 mM GTP
1.6
e O.5 mM UTP
1 .6~1
e [ a·" PjU TP
1.6 ~ I
e H~
concentration s may be requi red to achieve the best compromise between high rates of tran scription and low levels of splicing. Since all of the molecular species illustrated in Figure I ma y be a mong the products of a single transcription reaction , electrophoretograms can be difficult to interpret. If the size of the linear intron is already known from the GTp· labelling assay (see Protocol I), a band of corresponding mobil ity should certainly be present o n the autoradiograph , If a physical map of the insert is available, the insert can be truncated to different length s using appropri at e restriction enzymes. Analy sis of the in vitro transcripts and spliced products fro m these modified temp lates can be used to deduce the position of the intron within the original insert. No spliced product s should be detected from trans cripts ending within the intron. Similarly, deletions between the promoter and the intron can reveal whether any of the bands are due to premature termination
Method
• 5 x transcription buff er e O.7 5 M OTT
Dovid A . Shub , Craig L. Peebl es . and Arn old Hampel
~I
of transcription . Ultimately, the nucleotide sequence of putative ligated exon species must be deter mined and compared to the DNA seq uence in order to characterize fully the splicing event . Full cha racterization of the self-splicing reaction requires purification of the precursor transcript, generally accomplished by denaturing polyacrylamide gel electrophoresis and elution (Volume I, Chapter 1, Protocol 8 or Volume 1, Chapter 4, Protocol II) . Precursors may be uniformly labelled, in which case non-radioactive GTP is substit uted to start the reaction (see Protocol I) . Thi s discussion has been limited to the group I ribo zyme reaction involved in RNA processing, namel y splicing. Th ere are, however, a large number o f other reactions that have been demonstrated on cleverly-engineered intron ribozymes. T he interested reader should consult Cech (12) for a review of several of them .
13~
2 . Add O.41J1 of 13 or T7 RNA polymerase and then divide the reaction mixture into 41J1 aliquots. 3 . Start each individual reaction by adding 11J1 of DNA template .
4. Incubate the tubes at 37 °C for 1 h. 5 . Stop the reactions by adding 81J1 of formam ide -d yes loading buffe r to each tube. 6 . Denature the samples by heating them at 65 °C for 15 min before analysing them on a denaturing polyacrylam ide gel as described in
Volume I. Chapter 1. Protocol 8 or Volume I, Chapter 4, Protocol 11. If introns are very actively spliced du ring in vitro transcription , precu rsor molecule s may not accumulate to high levels. In such cases, the authors carry out the transcription reaction (Protocol 2) in the presen ce of very high concentrations of nucleoside triphosphates (typicall y 2 mM o f each) leaving the concentration of MgC1 2 unchanged. Thi s reduces the effective Mgh ion concentration below the level needed for splicing, while still allowing efficient transcription. Naturally. some adjustments of the NTP and Mg" ion 21 6
3, Group II intron ribozymes 3. 1 Identification of group II introns Group II introns were first recognized as a specific class of intervening sequen ces in yeast mitochondrial genes that ar e distinct from the gro up I introns (ref. 4 and reviewed in refs 12 and 13). Each group of mitochondrial introns is different from nuclear pre-mRNA introns, based on a comparison of the boundary sequences and internal structures. Group I or group II introns may be distinguished by their characteristic secondary structures and con served sequences (13, 14). Standardized secondary structure dia grams for group [ and group II int rons have been reviewed recently (13, 15). The se generalized structures can accommodate the many newly-identified members of each group. Furthermore, each group ca n be divided into subgro ups according to specific structural features that are not general for the entire group. Thus these intr on s repre sent a diverse and expand ing fam ily of molecules. A detailed description of group II intron structure is given in Section 3.2 . 21 7
7: Rib ozym es
A. TRANS-ESTERIF ICAT ION
~ 'A~criPt
G, (GCGA
~
1
B. HVDROLVSIS ~ ~ ' A transcript G, (GCGA K
(Z1,.
exon 1
~
(Z1,.
2/3's lari at
GCGAGp 1:-
G/"A intermediate (GCGA , L-.+-j
S
1
~
STEP 2
transester-ificeticn
G/"A (GCGA, ~+ V//J I, lariat i ntron s pliced exons
'l
O
STEP]
tr ansester-ific.aticn
exonl
David A . Shub. Craig L. Peebles. and Arnold Hampel
":> STEP 1
hydrolysis
'A~2/3'S
,
~ermedlate
1
STEP 2
tr-enses terificatton
'A GCGAGp 1:,")+ V//J I I Ii near i ntron spliced exone
Figure 2 . Self -splici ng pathways of group II int ro n s. The sequence used throughout is t hat of the group II self-splicing intron a5) o f yeast m ito chondria . lAl Th e standa rd self -spl icing pat hway is init iate d by trans-esterif icat ion . The first step is attack by the branchpoint adenosine l A ) on t he 5 ' splice junction to release t h e 5 ' exon (h at c hed box) from the 2/3'5 lar iat intermed iat e co m posed of th e int ron lGA GCG-t hin line) and t he 3 ' exon (open bo x). The second ste p is attack by t he 3 ' -OH o f the 5 ' exon on t h e 3 ' spli ce jun ct ion to release the splic ed exo ns an d a lariat intron. (Bl A n alte rnative self-splicinq pathway is init iat ed by hydrolysis and is t hus un c oupled fr om branch ing. Th e fir st step is attack by wa te r (HOH) on the 5 ' splice junction to release t he 5 ' ex on (h atch ed box ] fr om the linear 2 /3 's intermed iate composed o f the int ro n (pGA GCG - th in lin e) and the 3 ' exon (op en box ). The seco nd step is att ack by th e 3 ' -OH of t he 5 ' ex on on the 3 ' spli ce jun ctio n to release the sp lice d exerts and a linear int ron. Th e ar rowheads indic ate th e 3 ' direc tion or a 3 ' -OH end .
Figure 3. Schematic diagram of gr oup II Int ro n sec ondary st ru ct ure. A stylized drawing after the model of ref . 13 show ing the secondary structure of a group II in t ron as the thin lin e; t.he c~rv~d s~gmen~s are identified as unpa ired in the secondary structure with paralle l stra ight I~nes Jn~ lca t.Jng helic~r seg~ents . The 5 ' exon and 3 ' exon are indicated as h eav ier straight Hnes. Sites In vo lv ed In t ernary base pairing int erac tio ns are labelled and in dic at ed w ith the gr~y ~ ighligh t bars IESS1 , EBS2, 18Sl, 1852 , a, a' , 13', c, t ' ). Each dom ain and subdom ain (Wit hin domain 1) is in dic at ed. an d the ot he r f eatures are dis cu ssed in the text (Sect ion 3 .2 1. No attempt has been made to diagram the secondary stru ctures o f doma in s 2 an d 4.
a,
splicing-related reactions have been described (see Section 3.5). This section will focus on the various condit ions that have been used to analyse splicing and other reactions of group Il introns .
Group Il intr on s have been identified in th e mit ochondrial or chloroplast genomes of fungi, plants and protists (13). Like many group 1 introns (reviewed in ref. 12), some group Il introris are ca pable of self-s p licing in vitro (16- 21). For group Il introns (Figure 2A), the first step is an attack on the pho sphate group at the 5 ' splice junction by the 2 ' -hydrox yl gro up of an unpaired aden osine near the 3 splice junction. Thi s trans-esterification reactio n releases the 5 ' exon with a free 3 ' -hydroxyl group and creat es a ' lariat ' interm ediate. In the second step, the 3 ' -hydroxyl gro up of exon 1 att acks the phosphate group at the 3 splice junction to create the spliced exons a nd r elease an excised intron lari at. By an alogy with group I intron ribo zymes (Sect ion 2) and other transesteri fication reacti on s, group 11 ribozymes are pres u med to positio n a Mg2 -;ion as an activatin g electrophile and to fo ld specifically for alignmen t of the reactan ts to promote a n in-line attack on the ph o sphorus atom . While the prim ar y products of the group Il self-splicing reaction are linear spliced exons and excised lariat intro ns (see Sectio n 3.4 and F igure 2), severa l other
IS bounded by ' pairing segments' th at are repre sented as helical stems. Each domain is separat ed fro~ its neighbou rs by short 'jo ining segment s'; joining segments are con served In length and partly conserved in seq uence (13). The 5 ' splice junction is apparentl y selected by tertiary base-pairing interacnons bet ween exon and intr on sequences. There are two exon-bindin g sequ ences (EBS I and EBS2) in subdomain 0 of domain I ' EBSI and EBS2 a I ' , re comp ementary to '?t:on-binding sequences (18Sl and 18S2) in th e 5 ' exon (exon I) . These pairmgs a re designated 18Sl :EBS I an d 18S2:EBS2 (22).
218
219
I
I
3,2 Group II intron seco ndary structure and fun ctional anatomy The distin ctive seconda ry structure of group Il introns is usually displayed as
~ 'cloverleaf' consisting of six 'domains' numbered 1-6 (Figure 3). Each domain
7: Riboz ymes
David A . Shub, Craig L. Peebles, and Arnold Hampel
While the exon sequences adj acent to the splice sites are highly variable, ther e are recognizable consensus seque nces define d for both 5 ' and 3' termini of the int ron. The 5 ' co nsensus is -j GUGCG- and the 3' consens us is -AY j-, where j designates the splice site (13). A base-pairing contact designated e.e' involves nucleotides near the 5 ' intron boundary pairing with an internal loop o f subdomain C I of domain 1 (23). Based on phylogenetic co-variation, two additional long-range base-pairing contacts are predicted to occur within domain I, designated c:c ' and ~:~' (22, 24). . Because of its complex internal structure and demonstrated role in the specific recognition of exon I (22), domain I is generally believed to contribute much to the catalytic core of group II introns. Domain 2 is rather variable in size and sequence and has no known specific role in group II activity. Domain 3 is more uniform in size, but , like domain 2, has not yet been assigned an y
and recombination. At present, there are no published reports for applications of group II introns as ribozyme agents for down regulation of gene expression .
specific function . Domain 4 is extremely variable in size and sequence, contai ns
a large open reading frame in some group II introns, and may be entirely deleted with out severe consequences fo r self-splicing activity (25). Domain 5 is the most highly conserved in bot h size and sequence of any domain; the presence of domain 5 may be considered diagnostic for group II introns (13). Dom ain 5 is critica l for function in vitro, playing an essential role in the first transesterification step of splicing (25), although the details of domain 5 inter action with other parts of the intron are not yet characterized. Domain 6 contai ns the unpai red adenosine that is the site of bran ch format ion (17) and consequent ly may be considered one of the substrates of the ribozyme. Deletion or alterati on of domain 6 can prevent branch formation, but splicing still proceeds in vitro (26-28). So far , there are no extensive base-pairi ng interactions defined between the intron and the 3' exon (exon 2). However, the nucleotide on the 5' side and adjacent to EBSI pairs with the first nucleotide of exon 2 (28). Th is interaction has been designated the 'in ternal guide' (ig:ig'). T he last base of the intr on participates in a solitary basepair with a base in 13; this contact is termed ..,:.., ' and affects the rate of the second Irons-esterification step of splicing (23). In addi tio n to ..,:.., ', domain 6 is needed to promote accurate and effici ent reaction at the 3 ' splice junct ion (28).
3.3 Potential of group II intron ribozymes for practical use
3.4 Self-splicing of group II introns 3 .4.1 Introduction Some group II introns are capable of self-splicing in vitro (16-21). The earliest examples were a5')' and b l , both of which are derived from the mito chondrial genome of the yeast Saccharomyces cerevisiae (16- 18) and are members of subgroup liB (13). More recently, examples of subgroup IIA intro ns, including al and a2 from yeast, have been shown to be self-splicing under some conditions in vitro (19-21 and P. S. Perlman, personal communication). Most studies of mutants and alternative reactio ns have focused on a5.., and b l. All group II introns are candida tes fo r self-splicing, but it should be point ed out that selfsplicing has not been demonstrated for every group II intron tested. Several protocols ar e currently in use by various la boratorie s studyi ng group II introns. Th e original reaction condition s were devised for the prototype of self-splicing grou p II intron s, a5.., of yeast (16). The splicing co nditions were similar to those fo r in vitro tran scriptio n of the RNA substrate, since spliced products were detected among the transcr iption products. Subsequently, modifications were made to improve the rate and yield of the ribozyme reactions . The following sections (Sectio ns 3.4.2-3 .4.4) descri be th e synth esis of ribozyme substra tes, the basic pro cedure for studying self-splicing o f group II int rons, and variations on this basic metho d. Later sections (3.5. 1- 3.5.3) discuss other reactions of gro up II int rons .
3.4 .2 Preparation of transcripts for se lf-s plicing In order to detect self-splicing of a known or suspected group II intron , it is advantageous to prepare reasonable amounts of homogeneous transcripts spanning the intron and including both exon I and exon 2 segments . This may be co nveniently accompl ished by cloning a restriction fragment or an a ppropriate PCR-amplified DNA segment containing the intron into the mu ltiple clon ing site of a plasmid with a promoter derived from SP6 , T3, or T7 bacteriophage. At least 50 nt of the natu ral exon I sequence should be included to ensure that the entire length of IBSI and IBS2 are included. No more than a few bases of exon 2 are needed. Additio nal sequences derived from the plasmid can be retained at either end, usually without posing problems in the subsequent transcri ption or self-splicing reaction. It is frequently convenient to have the total length of the exons similar in size to the intron; then the spliced exons, separate exons, and intron products are easily displayed on a single gel. To prepare the template for in vitro transcription, purified plasmid DNA must
Group II introns are capable of a variety of potentially useful reactions. However, this potential is largely unrealized because of the limited availability, large size and slow reaction rates of these ribozymes. The splicing pathw ay resembles group I intron ribozymes but is less convenient to use since group II introns do not depend on a small molecule substrate to initiate the reactio n. Reactions of group I introns have the distinct advantage that their cleavage products can be labelled specifically by the addition of GTP (see Protocol I) . Nevertheless, group II introns offer high specificity , and research is continuing for improvement of group II introns as ribozyme reagents for RNA cleavage
the 3 ' -end of the transcript. Restriction enzymes that leave blunt ends or 5 ' extensions are preferred to those leaving 3 ' extensions. Cleavage within
220
221
be cut at a restriction site downstream from the inserted fragment to define
7: Rib o zymes
Do vid A . Shub . Cra ig L. Peebles. ond Arn old Hompel
the intron will pre vent splici ng, but some intron fra gments retain ribozyme activity. Care must be taken to eliminate traces of RNase typically used in plasmid preparations . An alternative method of preparing transcrip tio n templates is to use PCR to amplify the exon-intron sequence with an appropriate promoter sequence built into the upstream primer. In vitro transcription is carried ou t as described in Volume I, Chapter I , Protocol 4. A proced ure is also given in Protocol 2 of thi s chapter. Because yeast group II introns contain 40% uridylate resid ues , it is mo st co nvenient to labe l transcripts using [a -J2P j UTP . H o wever, any suitab ly labe lled nu cleo tide co ntai ning 3H , I4C a-" S, o r a _J2 p sho uld be successful. Since group II int ron t ra nsc ripts are typically ove r I kb in length, transcr iption conditions mu st be o ptimized with respect to the yield of ho mogeneous full-le ngth p ro ducts. Good quality RNA po lymerase (U nit ed States Biochemicals) and ribonucleoside tri phos phates (P harmacia) seem to be important in this respect. Otherwise, full-length transcripts must be purified b y gel electropho resis. This may be done using 4% polyacrylamide gels with 0.1 "7. SDS an d 8 ~I urea run with TBE bu ffer as de scribed in Volume I, Chapter I, Protocol 8 or Volume I, Chapter 4, Protocol II . A fter the product bands ha ve been identified by autoradiography or UV shado wing, the RNA can be eluted from the gel by soaking overnight in 0.5 M ammonium acetate, 5 mM EDTA. Ho wever , better yields and mo re rapid recovery are obtained by electroelution in either Tris-acetate-EDTA or TBE buffers with a Schleicher and Schu ell Elutrap device . After transcription, either with or without gel purificatio n, the RNA is recovered by ethanol precipitation and dissolved in water at a concentration of 103-10" c.p.m .zul . The transcripts should be stored at -20 ' c.
The product s of the self-splicing reac tio n are anal ysed by electrophoresis o n a 4% polyacrylamide gel with 0.1 % SDS and 8 ~I urea run in I x TBE buffer. Under th ese conditions, linear RNAs of 0.1 - 2.0 kb can be resolved reas onably well . Typical running times for a 30 ern gel are 2-4 h at 700 V. For the largest group II introns, 3% or 3.5 % polyacrylamide gels may be more appropriate. Non-linear RNAs, such as the lariat excised intron, display greatly retarded mo bility co mpared to the linear fo rm of the sa me to tal len gth. Consequently, th e p rodu ct most cha racteristic and clearly visible after succ essful self-splicing is the laria t int ro n pr oduct , since it o ften mig rates more slowly than the sta rting transcript. H owe ver , the relat ive mo bility o f th e lariat varies so mewhat with th e degr ee of cross-lin king of the gel ma trix, the buffer salt co ncentration, the den at uran t concentration, a nd the gel fu nn ing tem perature.
3 .4 .3 The self-s p lici ng reaction The basic procedure for self-splicing of group II intro ns is de scribed in Protocol 3. It is convenient to set up the splicing reaction by adding one volume of labelled tran script to an equal volume o f 2 X splicing reactio n bu ffer. For intramolecular splicing as say s, th e co ncent ration of th e t ra nscript is no t cru cial; t he a mo u nt o f RNA used is det ermined b y its specific ra dioactivity an d the amo unt of radio act ivity needed to produ ce an acceptable autoradiographic exposure time. The authors routinel y use 10"-10' c.p.m . of the tr anscript in a 10 ~I rea ction vo lume. Larger or sma ller reac tion volumes may be used a' dict at ed by th e particular circums ta nces . The incubation temperature is important, since th e temper at ure profile is rather abrupt for self-splicing by all known gro up II intro ns (for exa mple, see ref. 16). The optimum temperature in vitro is typicall y 40-50 ' C , with the best temperature depending on the intron . Incubation times are typi call y 0.5-4 h at 45 ' C . When inve stigating the self-spli cin g of a group II int ron for the first tim e. the optimal incubation temperatu re and rea ction time mu st be determ ined in pilot expe riments . 222
Protocol J . Bas ic s e lf-spl ic ing reactio n for g ro u p II int ro ns Equip m en t and reagents •
32
P-labelled RNA substrate transcribed in
. 4 % polyac ryla mi de gel (30 cm lo ng)
vitro from a suit able DNA template using SP6, T3 o r T7 RNA polymerase and
co nt aini ng 0 .1 % 50S , 8 M urea and 1 x TB E buffer plus electrophoresis and aut orad iograph y equipment as desc ribed in V olum e I, Chap t er 1, Pro tocol 8 or V olum e I, Chapter 4 , Pro tocol 11
{0'. 3 2 p ] UTP
as described in Protoco l 2
(see also Sect ion 3 .4.2 and V ol ume I, Chapter 1, Protocol 4 1; dissolve th e ANA in water at 10 3 _10 4 c.p .m .zul
e u ree -cva m ixture (0. 1 % bromophenol blue , 0 .1 % xy lene cyanol , 20 % sucr ose, 8 M urea , 2 x TBE I
. 2 x sp licing buffe r (8 0 mM T ris- HCI, pH 7,5, 20 mM M gC I2 , 4 mM spermi dine ) '
e O. 2 M EDTA
• Au toclaved plast ic microcent rifuge t ubes and dis posab le plastic pip ette t ip s
Method 1. In a mi cr oc entrifuge tube on ice , m ix 5 ~I of 2 x splicing buff er and 5 ~I of 32P-labelled RNA . 2. Incubate the mixture at 40 -5 0 ° C for 0 .5-4 hb • 3 . St op the reaction by add ing 1 ~I of 0 .2 M EDTA to che late t he M g 2 . ions in t he reac ti on mixture . 4 . Heat th e m ix tu re at 10 0 ° C for 1 min , and t hen chill it in ice wate r. 5 . Add 1 vol. of urea-dye mi xt ure and load the sample on a 4 % polyacrylamidel ur ea gel. 6. Run the gel f or 2 -4 h at 700 V using 1 x TBE buffer and th en auto . radiograph t he gel as descri bed in Vo lum e I, Ch apt er 1, Protocol 8 or Volu me I, Chapt er 4 , Protocol 11. • The co m pos iti on of the splicing buffer ca n be vari ed as descr ib ed in Sect ion 3. 4 .4 . The con dition described here is t he low -salt react ion bu ffe r (see Table 1). b The precise tempe rat ure and t ime of inc ubat ion must be dete rm ined fo r each int ron base d on pi lot exp eriments .
223
7: Ribozymes
David A . Shub, Craig L. Peebles, and Arnold Hampel
. b If f group II intro n self-splicing and alternative reactions" Table 1. Reactron u ers or
cont ributi on of these altern ative reaction s to the final pattern of self-splicing pro ducts is dependent on th e ionic compositio n of the reacti on buffer and on the nucleotide sequence o f the RNA , particularly IBS I, IBS2, EBSI, and EBS2 (Section 3.2 and Figure 3 ). For instan ce, the high KCI buffer (see Table 1) activates additional reaction pathways. Th e high ammonium chloride buffer (see Table 1) yields a pattern of a5 ~ products intermediate between those due to the high KCI and high ammonium sulphate buffers. The first of these alternative reac tion s is cis splice junction hydrol ysis or SJ H (see Figure 4A). Transcript s extending from exon I through intron domain 5 will efficiently hydrolyse the 5 ' splice junction to release exon I with a 3 ' -hydroxyl gro up a nd a linear intro n fragment with a 5 ' -ph osph ate gro up (32, 33). SJH proceeds most rapidl y in the presence of high KCI buffers (see Table I) . In addition, the linear intr on fragment und ergoes hydroly sis j ust 3 ' to alternative sites that are similar to IBSI. Since this reaction represents a relaxed speci ficity cleava ge analogous to SJH, it may be termed SJH* (34, 35) (see Figure 48, C) . Intron RNAs will also car ry out trans SJH of substrate transcripts containing the 5 ' splice junction (36) (see Figure 4D). Pre sum abl y in thi s trans reaction the group II intr on is acting as a true cat alyst, althoug h this has yet to be demon st rated conclusively. The specificity of these hydrolysis reaction s is defin ed lar gely by EBS I, a sequence block o f abo ut six bases. Amazingly, intron RNAs will also ca talyse hydrolysis of the spliced exon pro duct (34). Th is reaction has been designated 'spliced exon reopen ing' or ' SER' (see Figure 4E) to distinguish it fro m the simila r SJ H reaction. Th ere are severa l indi cations that SE R and SJ H ar e diffe rent event s. For exa mple, the group IIA int ron a l of yeast ca rries out both SJ H and self-splicing, but is inactive for SER (21). Similarl y, deletion o f dom ain 3 from a 5~ results in substa ntial reduction of the SER reaction while self-splicing a nd SJH a re much less affected (31). In applic ations o f the group II ribo zyrne, the SJH acti vity acting in trans may be the most useful, since maximal cleavage specificity is highly pred ictable based on the sequence of EBSJ. Th e role o f EBS2 is not well understood , but it evidently cont ributes to efficient hydr olysis in trans (36). Th e IBSI and IBS2 ta rget sites may be directly adja cent or separated by a few nucle ot ides. However, onl y a few dif ferent sequences are avail able at present on which to bas e these prediction s.
T ris-H e!
MgCI2
Spe rmidine
(pH7 .51' (mMI
(mMI
ImMI
10 100
2
High KCI High ammonium chl oride
40 50 50 50
l Oa lOa
0 .5 M KCI
High ammonium sulphate
50
lOa
0 .5 M am monium sulpha t e
Oe signat ion b
Low salt High Mg "
~
Based on ref. 29 .
b
Fin al com pos ition in ribozym e rea ct ion. Other buffers and differe nt co nc ent rat io ns work as w ell. Gre at er th an 50 mM inhibit s sign if icantly .
C
d
Neutral salt
a
0 .5 M am monium chlorid e
3.4.4 Modification of the reaction conditions
.
The self-splicing reaction may be accelerated by increasing the conc ent rallon of l\l g' + ions in the splicing buffer (see Protocol 3) to 0.1 M as III thehIgh M;' + buffer (see Table 1). Ribozyme reaction times are then typically 10-60 min. The additional inclusion of 0.5 M KCl or 0.5 M ammomum sulphate, . the high KCl and high ammonium sulphate buffers (see Table I ), f~~ther as III . ' . f 2 20 in may be suff'icient. increases the reaction rate so mcubatlon time s 0 fi t .. Some group II introns will splice only unde r these high-salt eonditions (I 9~ 2 1 ), but one should realize that the use of these alternative buffers may also acn vate alternative reaction pathways (see Section 3.5 below) . Even hlghe; salt concentrations can be used (29) and may provide an advantage for an a ysing the activ ities of mutant introns (30). . e Although concentrations of Mgh ions and neutral salt higher th an thos listed in Table I are tolerated by a5 ~ intron of yeast, they result III no fur~her increase in the reaction rate . Some neutral salts , such as 0.5 M LICl, show rate enhancement or change in product pattern . The anion added with the bu e~ Mg' + ions ha s only a relati vely minor effect unles s extr emel y high c~n or M M z- . ith the same amon centrations are used. Some workers use 60 m: g Ions WI f as the neutral salt and also include spermidine in th eir high- salt bu ffers ( ~~ example, ref. 31). Spermidine is tol erated under these condtuons. but can OS omitt ed with any of the high- salt bu ffers (see Table I). Addition of 0.\ !1J' ~) is tolerated in all but the low-salt conditions (see Table I and Protoco 1 .
h;t
3.5.2 Splice junction hydrolysis and sp liced exon reo pening re acti ons
y In add ition to self-splicing, group 11 intron ribozymes also carry out a va r,,:t of splicing.related hydrol ysis and tra ns-esterification reacu on s. The relat ive
RNA substrates for SJ H and SER may be synthesized by in vitro transcription of an appropriate DNA temp late in the presence of [a -J2PJ UTP (see Protocol 2 and Section 3.4 .2). For exam ple, exon 1 with domains 1-5 (E l : 12345) will undergo SJ H efficiently (33). Intron lariat and spliced exon products are needed for SE R (34). For the ribozyme reaction , mix the transcript with an appropriate high-salt reaction mixture, such as high KCl (see Table 1) . For SJ H with the E l : 12345 transcript. the RNA concentration is not important. Incubate the reac tion for
ZZ4
ZZ5
3.5 Other reactions of group II intro n ri bozymes 3.5.1 Introduction
.
Da vid A , Shu b, Craig L, Peebles , and Arn ald Hampel
7: Ribozym es
A. TRANS- SPLI CING
C.SJH*
A . SJH (c is)
W
t r ansc ri pt W'it h DS G ( GCGA ,
STEP
~CGAGp ,
H·O'f!...!
:--:--:>
'A S
+
,'-3'---:=> i ntron
f--, QG 'A~
5' ; ntron fragment
'----:=>
(GCGA , Intron - exon 2 txt
exon 1 ectxt hydrolysis
1hy dro 1ysis I!LLl
'----:=>
0....
exon 2 fragment
(GCGA Int r on- exon 2 f ragment
exon 1
1~CGA
S , ~
i nt ro n + pG.. fragment
I ,+
H'O,H
S
s plic ed exons
t
(
G.rA GCGIi I
~
teriet i nt r on
1!LLl+ Cl. exon 1 exon 2 . f II ' trons The exon and intran segmentS Figure 4 . Splicing-related ribozyme rea~tlons 0 grod~n /:7g ure the hatched box representS are represented by the same ~onve~tlons as use x is ex cn 2 . and the arr owneade are
2-
~ine 1.5 th~~tsop~ic~ju~~~~:~YdrOIVSiS
ISJ H (cis)] is .out',n~dxf~~
the 3 '-OH end or t~e 3 dlrecncn . . lice 'unct ion hydro lysis occurswith a te a . t a transcript extending through domain 5 . tB~~~a' si~e (arrow ) in the 3 ' ex on usmq a tranSfr~s spec ificit y (SJH · lexa":!ple , 11 a~ a3~o~-c~7intrOn_ex on 2 txt ). tel Sp lice jun ction .hYd,'o y .....1 ex tend ing through the mtron an ... e o . m te 2) J at a non -can on ical sit e erro n also occurs w ith a re~a.xed . spec tficltv (SJH li~eex~un~tion hydrolysis l SJH ftr a,?s)] is sh.o~ in the int ron using pu rified Hnear mt ron . ~OJ SP . ~ that acts on a second transcript contalnlF~ for an lnt ron t ranscript extending th)'OI~gl S ~~ea~nexon reop ening (SER (trans)] is show n a the 5 ' splice junction (ex on 1 tx t. '. ~ an lnt ron la riat acting on the purified spl iced exons .
226
.'
1 ntron
1-
+.
- exon 2 txt
--..r:=L..
exon 1
pG.. fragment
AS
[?Z1... (
.
GCGAG.r~
_,. J
Y- i nt ro n intermediate STEP 2
1
STEP 2 t r-ense ster-ificati on
V//J
V//'l
I,
spl i ced exons
I,
s pliced exons
+
+
AS
GCGAG.r~ {
-I
-.J
Y- i nt r on
E . S ER ( t r ans)
WZI
(GCGAGD~
1
Gp 'A
hydrolysis
exon r . the GAGCG-th!"
.....c:::J....
[?Z1... 'A':::::>
STEP I
trensester-ification
transeosteorificat ion
D. SJH ( t r a ns) ~ 1- H'o ,H
H·O' f!...\
+
fragment
1
i ntro n - exon 2 txt
exon 1
(e xample 1)
~
~
QG 'A~
~..,---..... , -:-=.".,,,,"
1 hYdr 01YSis
GCGAGp
QG, ' A
11
ly S 1S
\..--
B . SJ H*
+ 'A~
.....c:::J....
':::::>'
hydro- (GCGA
Intron fragment
exon 1
(e xa mple 2 )
H'O,H
exon 1pG.. txt
H·O'f!...!
I!LLl +
+
(222l
:--:--:>
1hy dr o1y sis
B. TRANS- SPLICING
( exa mple 1)
( exa mple 2 )
Figur e 5 . Trans -sp licing reactions of group II inrons. The exon and intr on segments are repr esent ed by the same conventions as used in Figure 2 ; the hatched box rep resen ts t he 5 ' exon (exon 11. th e GAGCG-t hin line is the int rcn . t he open box is th e 3' exe rt (exon 21, and the arrowheads are the 3 '·OH end or the 3 ' direction. fAI One example of trans-splic ing requ ires an exon 1 transcr ipt and an int ron - exon 2 t rans cr ipt. (B) A second exa mple of trans-splicing requires on e t ranscr ipt exten ding t hr ough domain 3 {5 ' txt) and ano t her t ransc ript con t aining domains 5 and 6 and excn 2 (3' txt ).
10- 30 min , For efficient SER, each tra nscript sho uld be used at a co ncent ration of about 0. 1 ~M an d incubated for 30-120 min in the high KCI buffer (see Table l ). After the ribozyme reactio n, pur ify the products by electro ph oresis and auto ra diography as described in Prot ocol 3.
3.5 .3 Trans -splici ng rea ctions of group II introns Som e unusual group II introns have been identified that are split into two or more transcripts (37-40), Th ese separate RNAs pr esu mably associate to form the correctl y-folded active intron and undergo splicing in trans (see Figure 5A, 8) . In vit ro models of such trans-splicing reactions have been constructed and used to analyse the req uireme nts fo r efficient bindi ng and catalysis. The first such model system sup plied exon 1 as a transcript with a short 3 ' extension to a second 227
7: Rib ozymes transcript that contained the entire intron and exon 2 seq uences (36). Th ese transcripts associated through IBSI :EBS I and IBS2:EBS2 to allow SJH a nd step 2 of splicing . No branching was detected . The major products were spliced exon s, a short linear intron fragment and a full-length linear intron. Product ive reactions occur under several buffer conditions, but with a requirement for high RNA concentrations such as 0.1 ~M for each participating mol ecule . Thi s is reasonable fo r a second order system limited by RNA :RNA associ ation. Later, trans-splicing system s were established between transcripts that ea ch contain ed some of the intron domains and one exon (25). In ef fect , th e self-splicing transcript was divided with in a domain or between two domains. As before, the most efficient reacti on s were ob served at high con centration s of each RNA, in the ran ge of 0.1-1.0 ~I\l for each hal f-molecul e. Thi s style of trans-splicing occurs best in the high KCI or high ammonium sulphate reaction bu ffe rs (see Table 1) . The products are spliced exons and a Y-shaped excised intron; so me separate exons and linear intron halve s are also produced as byproducts . Th ese trans-splicing systems have been useful for examining the rol es o f indi vidual domains in the overall reactivity of the intron. For example, this approach was used to show that domain 4 is entirely dispensable, while domain 5 is essential for splicing (25). The trans-splicing reaction is carried o ut using a procedure similar to that for self-splicing (see Protocol 3) . First , prepare tra nscripts for the two hal ves of the intron, for example exon I with domains 1-3 (EI : 123) and domains 5-6 with exon 2 (56:E2). Use in vitro transcription of appropriate DNA templates (Section 3.4 .2). Purification of the transcripts by electrophoresis ma y not be necessary. Th e rate and extent of the trans-splicing reaction depend on the concentrations of both reactants. Each should be pre sent at a high concentration (at least 0.1 ~M). Small reaction volumes are useful to achieve these high con centrat ions; use a 4 ~I reaction volume overlaid with mineral oil to prevent evaporation. Use an appropriate high -salt buffer (see Table 1) and incubate th e reaction mixture for 30- 120 min at 45 °C. Anal yse the yield of pr oducts by electrop hor esis and auto rad iography (see Protocol 3).
3.5.4 Reverse splicing and transposition reactions The total number of phosphodiester bonds is conserved during th e gro up II self- splicing reaction. Therefore, each step and the entire reacti on sho uld be readil y reversible. Recently, rever se reactions, where the intron lari at inserts at the product splice junction, ha ve been described and anal ysed in some detail (41, 42). Figure 6 illustrates examples of reverse splicing. Imprecise 'reversal' event s have also been disco vered, where the intron lariat insert s at sites related to IBS1 fo und within an otherwise unrelated tr ansc ript (31, 42). Such inserti on pr oducts can be ' preserved' by reverse transcript ion and cDN A cloning. Transcr ipts of th e resulting clone s ar e th en ca pable of forwar d splicing (31). The reverse reaction s have been termed RNA recombinat ion or RNA transposition (31, 42). 228
Dm 'id A. Sh u b. Cra ig L. Peebles. an d A rnold Hampel
A . REVERSE SPLICING
(
S GCGAG.r~ . '.
Janet lntron
1
+021
B. BRANCH ADDiTiON
I,
3pliced exona
AS (GCGr'-:J
Jerist' i ntron
1
REVERSE STEP 2
t,..ansntedfication
~
(
~
2/3', lertat GCGAG.r~ermedi.te
(GCGA~
,
1
STEP I
exon
' AS '-:J
1-
intron
REVERSE STEP I
C_ PARTIAL REVERSAL
tranSf'stE'rjfication
(
REVERSE
tranush,..lfie.ahon
~
exonl
+~ exon 1
~ , ~n9th GCGAG L-1;ecuroor
'AS GCGAGp " ) t , +{V2I I, 1ntron 3pliced exons
,
1
REVERSE STEP 2
tr anspst.,..ification
e~'
+
GCGAGp t
I
S~
Ii neer 213',
Figure 6. Group " int ro n reve rse splicing reaction Th . represented by the same conventions as u . . s. e exon and mttcn segments are . h . sed In FIgure 2 ; the hatched box represents exon 1 the G AGCG~th ' Ii In me IS t e mtrc n the o p box i 2 ' end Orthe 3 ' direction . IAl l ariat i~tron an~ns ~~~~ ::on , an.d the arrowheads are the 3 '·OH reverse splicing. (BI A lariat intron reacts w ithPthe exono~s~r~ ln~Ubated togethe r for complete only. lei Partial reversal with a linear Int ron releases a f 0 ecu e t 0 revers~ step .' of splicing 2 (linear 2/3 's). ree exon 1 and a uneer Intron- exon
~g~in , reverse splicing and transposi tion reactions are followed using van atlons on the basic splicing procedure described in Secti on 3 4 3 and P rotocol 3 To hR ' . prep~re l eNA transcripts, carry out a self-splicing rea ction an . ' d puri fy the la~tat mtron and target tra nscript (spliced exons) by preparative gel electro pho resIS a nd electr oelu tion as described in Prot ocol3 p ifi . of target transcript b . un rcanc n b ," s ma y not e necessar y. Th e substrates ma y a lso be prepared > In vuro transcnptlon of appropriate cDNA clones . . . As for the ather t r a n ' require high RNA S reac~lOns, reverse splicing or transpo sition reactions concentratIon s, up to 1 jJM for each RNA reactant. For an 229
7: Ribozym es
R NA of about 300 nt, I ~M is equivalent to I ug in 10 ~1. After preparatio n, determine the yield and concentration of each RNA by UV absorption or by calculation based on the specific radioactivity (see Volume I , Chapter I, Section 2.3.4) . Set up the reaction mixture using the high ammo nium sulphate buffer (see Table I). Incubate the reacti on mixture either at 45 · C for 30-60 min or at 30 ·C for 1-4 h, depending on the intron in use. There is evidence th at intron bl accumulates reversal products to a greater extent at 30 ·C (42) . After incubation, analyse the products by gel electrophoresis and autoradiograp hy (see Protocol 3) . Reversal and transposition products are readily detected due to their reduced mobility compared to the input R NA s, despite the typically lo w efficiency (0.1-2.0'70) of these reactions . There is no requirement for either spermidine or elevated concentrations of Mg 2 + ions to detect reverse splicing reactions. although these additions may improve efficiency. Intron bl reversal products appear to accumulate more readily than a5)' reversal products . However, for transposition it may be necessary to use elevated concentrations of up to 0.24 M MgSO, .
4 . Hammerhead and ha irpin ribozymes
Dovid A . Sh u b, Craig L. Peebl es. and Arnald Hampel Hammerhead 3'
N • N
III
Ribozyme ..•......
',""
A
230
[]
....
'NNNN
X NNNNN 3'
~~~~~
····7N~~NNN~
····•··...
Hairpin
4 .1 Introduction Ribozymes of the hammerhead type (43, 44) and hairpin type (45 -47) are the catalytic centres of certain small plant viroid and satellite RNAs (48). The full-length RNAs which contain the catalytic centre are typically 350-400 nt in length, and the RNA has no coding function (49). It has been suggested that the inherent catalytic activity in these RNAs is required as part o f their replication cycle (50) . The catalytic cent re itself is the su bj ect o f th is section. The hammerhead and hairpin ribozymes form characteristic but different two -d imensiona l st ruc tu res with t heir RNA substr ates . Figure 7 shows t he mod els for t he trans cleavage o f target substr ates by these tw o classe s of ribozyrne . The hammer head st ruct ure co nt ai ns three helices (I , II, and III ). T wo o f these (I a nd Ill) are o f variab le length (five base pairs a re sh own in Figure 7) and invo lve two region s of the R NA subs trate o n eit her side of the event ua l clea va ge point . Th e third helix (II ) is for med internally an d delimits the ribozyme catalytic d omain o f 32 nt . Thirteen of the nu cleot id es involved in the structure are conser ved as sho wn in Figure 7. The hairpin ribozyme ha s fo ur separate helices; helices I a nd 2 in vol ve the R NA su bst rate while helices 3 and 4 are internal. Wh en helix I is 6 bp lo ng. the minim um seq uence of th e cat alytic component is 50 nt long . Both of these ribozymes cleav e t he RNA substra te to give a 5 ' cleavage frag ment with a 2 ' , 3' cyclic ph o sphate at th e 3 ' terminus and a 3 ' cleavage fra gment wit h a 5 ' -hyd rox yl gro u p at th e 5 ' te rminus (48).
Substrate
N • N
@~: G
5'
Ribozyme
U GUG
u
•••
G
3
GUAUAUUA ~ CCUGG
'2.,, ,.I. . ,.
.... ....
GU
NNNSN
S'
SUbstrate
\""
......
CNNNNNN ••• N-->
CAc-=---~~~~-;';~~ CAC A A GACC A NNNS NNNNNN •• •N " CAAAG
Hell1l4
AAGA
Hell. 3
Helix 2
Hell. 1
Figure 7 . Ham merhe ad and ha irpin ribozvmes. T he diagram s show cons ens us se uences of
:~~~::~::eSad anhd hairPint ribozym~s . In each case, cle av age of the subst rate R~A OCCurs .
ee t e text or a deta iled
descnpnon.
Su bstr at e requirement s d iffer for the two types of ribozyme. Th e ha mmerhead has a GU X requ irem ent (helix III) wit h cleav age occu rr ing most e lClent y a fter G UC, . G U U , o r GUA (51). While it ap pea rs at fir st tha t the ~a~rpIn subst rate .r eq uIrement is similar , in that clea vage is mo st effi cient fo r ~ st r~tes contarnmg a GU C seq uence, clea vage occu rs bef ore the G rather than a ter t e three base sequence as with the hammerhead. In selecting the cleavage ~nes forthe hairpin ribozyme, in addition to the GUC requirement, the adjacent asde pair ~n helix 2 must be either a G or a C. In Figure 7 the sym bo l S is .. Th ' use for either G or C at thi .h IS posmon. us th e target seq uence is SNGUC wit cleavage occurring between the Nand G (52) . Mechanistic stud ies o~
s~~strat~
231
7: Ribozy m es hai
.
Dm 'id A. Shu b, Craig L. Peebles, and Arnold Hom pel
riboz "mes represe nt a new and exciting aspect of
~~:C;:;'%~s~~:. ~~~ile~~f~ted a~ount of information has be~n obtained to date
t:
no~ (53 54) the specific details of the catalytic mechanisms for n bozymes 'b ' I cida ted However because of their small sizes, the hammer ea an yet een e U I . .. ' f d over other la rger and more hairpin systems have distinct advan tages or stu y 1f T I x ribo meso Ho pefull y, in the very near future, the ro es 0 spec~ IC and other features of the ribozymes in the catalytic mechanism will be known.
~~:~~na l gro~ps
4,2 Ap pli cations o,f hammerhead and hairpin ribozymes in molecular biology , , hai d hammerhead ribozymes ha ve two major applicatio ns. To date airpm an d fi d 5 I 3 I termi ni on These are gene ta rgeting and the generation of e me or RNA transcripts . i Gene targeting ,, ' , , : h t RNA or at least the sub strate recogrunon sequence, IS very hairpin ribozymes have the potential to be as antisense molecules for the tar geted down-regul ati on of specific gene expression (46, 55-58; see Section 4 ,3),
Sm~e \:~%~~head a~d ~~~a~;tic
~sed
ii Generation of defined transcript termini . ,h ' b d to generate RNA transcripts wit Cis-cleaving autocatalytic cassettes can e use , . f h' bility d r ed 5 ' and 3 ' termini (59-62; Section 4.4) . A pp lications 0 t I.S cap a) d e 10 en to ro d uce plant viral transcripts with defined 3.' terrrnm (59 an a met hod to pr oduce fully catalytically both a sho rt defi ned 5 ' an d 3 ' term inus (62) . Th IS method was then a P , to de liver an HIV -I spec ific hairpin ribozyme in vivo by usmg a ha::~l:
~a::e~:centl/as
actlv~ nbOzym~Sa Wt'~~
autocatalytic cassette (58). T he cis-cleaving autocatal}1'C casse~~d~~~~r: fu lly short defined 3 ' termin us on the HI V- \ specific nbozyme and p ivit . catalytically functional ribozyme in vivo with excellent ann H IV- I act }.
4,3 Use of hammerhead and hairpin ribozymes to target specific RNA transcripts 4.3 .1 Design of the ribozyme
"
.
_ ecific
Certain rules need to be fo llowed in orde: to design ribozy me.s ~~rs~~~~r:;e and ta rge ting. It is a simple matte r first to l~e.ntlfY a~ a~~~p~~iCh base pairs to iven in Figure 7. then to design either a hammerhead o r airp m n oz the ' ar ms' of th e substrate according to th e struct uresI' gi I d 11l which ib bstrate ' arms' he lees an , The hammerh ead has two n oz yrne-su h hai , ibo zyme has tWO ' I th sam e len gth T e airp m r , . ' Helix 2 is fixed at 4 op can be of approx imat e y e ribozyme-sub strat e helices with very different properties. , ith lengt hs of while helix 1 is va riable with op timal clea vage rates occurn ng WI 232
6- 12 bp dep ending o n the specific nu cleotide seq uence. Thus fo r bo th system s various lengths of these helices must be tes ted for o pti mal cat a lytic efficiency (k", / K m ) with a ny pa rti cu la r RNA SUbstrate. It is imp ortant to select tar get sequences which give reason able tu rnover (k,,,) va lues to ob viate the dan ger of generat ing a ribozyme with very low rates of reaction which would give , for all practical purposes, a non-catal ytic ant isense reactio n. For exa mple, th e indi scrim inate use of unnecessari ly long helices between the ribozyme and substrate would lead to such low off-ra tes for th e product that the ra tes of reac tion wo uld be neg ligible for any type of gene targ eting effect. In the basic ribozyme-catalysed rea ction, both the ribo zyme seque nce and the RNA substrate are prepared by in vitro transcription. Using a DNA synthesizer a nd the procedure recommended by th e manu facturer, synthesize DNA templates corresponding to the complements of the desired ribozyme and of the RNA substrate. Also include the 17 nt prom ot er of T7 bacteriophage (sequence 3 ' -ATTATGCTGAGTGATAT-5 ' ) upst ream of the seq uence to be transcribed . Although transcr iption will stan imm ediately fo llo wing nucleotide 17, severa l addi tion al nucl eotides downstream ar e also pa rt of the prom oter. It is suggested th at nucleotides 18, 19 and 20 be chosen as CCC for the ribozyme template and CGC for the substrate temp late. Th is will sta rt th e ribo zyrne transcript with a 5 'GGG and th e subst ra te with 5 'GCG . When in vitro transcription is ca rried o ut using [ a _J2p JCTP sub str ate (63), this will ensure at least o ne lab elled nucleotide in the substra te for detection pu rposes. Followin g in vitro transcription of these DNA templ at es using T7 RNA polymera se (see Protocol 2 and also Volume I, Chapter I , Protocol 4 and Chapter 2, th is volume, Protocol 4) purify the J2P -labelled substrate and ribozy me on 20% pol yacrylamidel7 M urea gels (9) using appropriate size ma rker sta ndar ds . Locate the appropriate radioa ctive RNA transc ript s by autoradiograph y and elut e the RNA (Volume I, C ha pter I , Protocol 8, Volume I, Chapter 4, Protocol I I ). To ensure the identity o f the RNA, it can be seq uenced directly wit h RNA base-spec ific en zymes. Finally carry out th e ribozyme-catal ysed cleavage reacti on and anal yse the products by denaturing polyacr ylamide gel electro pho resis. Details of this seq uence of investigation are given in Protocol 4. Pro tocol 4 . Carrying out a ribozyme catalyt ic reaction wi t h hammerhead or hairpin ribozymes Equip m ent and reag ents e DNA syn t hesi zer and appropriat e n uc leo tide mo no me rs f or th e templa te sequ ences to be sy nt hesized (ribozvme and RNA sub st rat e); see th e m an u f ac t urer ' s recommendation s
e 1 7 RNA po lym erase and othe r reagent s f or in vitro tran scr ipt ion as d escr ibe d in Protocol 2 o r as in Vol ume I. Ch apte r 1, Prot ocol 4 and Ch ap t er 2, thi s vol u me,
• [ Q'~32p JCTP (spec if ic act ivit y greater tha n 3000 Ci /m mo lJ
Protocol 4
Continued 23 3
Dcvid A . Shub, Cra ig L. Peebles. and Arn old Hampel
7: Ribozym es
Protocol 4 . Continued
Protocol 4 . Continued . 20 % polyacr ylam idel7 .0 M urea gel and other materials fo r denat uring electro ph or esis . ;;I s desc ribed in V olum e I. Chapt er 1. prot ocol 8 and V olume I, Chapter 4, Prot ocol 11
• Ext ract ion buffer 10. 5 M ammonium
a ce t a te . p H 7 .0. 2 mM Na · EDTA. 0 .5 m g lml 50S )
. 5 mg I ml tRNA (veast!
. 0 .5 ml m icr ocentrifuge tubes ~nd pest les designed to f it int o micr ocentrltuge tubes
13 . For the ribo zy me clea vag e reaction , add the foll ow ing reagents, in order, t o a 0 .5 ml m ic roc entrif uge tub e on ic e and m ix :
e H20
. 70 % ethanol . 2 mM Na-EDTA erE buffer 110mM Tris-He l, pH 8.0. 1 mM EDTA I . 4 x cleavage bu ff er (0 . 16 M T ris-H CI•
pH 7.5 . 48 m M MgCI 2 • 8mM spe rm id ine )
• ribo zyme
14. 15 .
e Speedva c conce ntra tor
Method . h moter-ribozyme and 17 phag e 1. Synthesize the appropriate T7 P a~e pr~ DNA synthesizer and th e promoter-substrate templates uSing t e manufact urer 's protocol . NA I
Its using purified T7 R
po y-
2. Separately tra nscr ibe the te~p a e hods de scribed in ProtoC Ol 2 3 2P merase and [ a · lCTP follow ing the ~e~ Chapt er 2 this v olume , or as in Volume I. Chapt er 1 , Protoco or ' . . . Protocol 4 . h d ts of eac h transcription reaction 3 . To purify the transcripts, run t e pro rel b .d app ropriate radioactiv e on a 20 % po lyacr ylamide l7 M urea ge eSI e t 018 or Volume \ size markers as descr ibed in Volume l . Chapter 1, Pro oc , '
Chapter 4. Protocol 11.
2 ul
16.
17.
2 1-11 e RNA substrate 2 ul For the zero time control. remove 31-11 into a separ ate tube c ont aining 3 J.l1 of formam ide-dyes mi xture. Heat the remain ing 3 J.ll of ribozyme reaction mixture to 80 °C and then cool it to 25 ° C. Add 1 IJI of 4 x cleavage buffer and mix . Inc ubat e the sample at either 37 °C or 25 ° C for a time per iod which w ill give signi f icant cleavage , i.e. at least 10% . Normally this will be 10-30 min . The choice of incubation temperature is dependent on t he intended use of the ribozyme . If it is a mechanistic stud y or if the int ended use is for gen e targeting in plants, 25 °C is used. If the ribozyme is to be devel oped fo r mammalian gene targeting expe riments , then 37 °C is used. Add 41J1 of formam ide-dyes mixture to the rib ozyme reaction (st ep 15 1 and separate the products of the ribozvrne reaction and t he zero time contro l (step 141 on a 20 % polyacrylamid el7 M urea gel (step 3 ). Auto rad iograph the gel and cut out th e appropriate bands (steps 4 and 51. Quantify the RNA products by Cerenkov counting (step 12) and determine the percentage cleavage of the substrate. after taking account of the zero time co nt rol va lue (st ep 14 ).
32p I b lied
1 . t detect the -a e 4. Autoradiograph the gel for less than .min 0 transcripts (e.g. using Kodak xomat film ). h i n d pla ce each gel fragment 5 . Excise the appropriate bands from t e ge a in a 0 .5 ml microcentrifuge tube. 6 . Add 0 .35 ml of extraction buffer to each tube and crush the gel fragment us ing a pestle . . 7 . Shake the samples for 20 min and then centrifuge the tubes In a mi crocentrifuge for 5 min. . b . t 1 5 ml microcentnfuge tu es 8 . Carefully remove the supernatants In o . h I M ix well and and add 1 ul of 5 mglml tRNA and 2.5 vol. et7~n~C for 20 min to leave the tubes at - 20 °C overn ight or at precipitate the RNA . . f or 30 m in 9 . Recover each RNA by centr if ugation in a mi crocentflfuge
at 10 OOOg at 4 cC. OS hi h will otherwise 10. Wash each RNA pellet to.remove.ua~es.of .S the~u~ace of th e pellet inhibit the ribozyme reaction , by simp y nnstnq 2 M EDTA and bri efl y and the in side of the tube w ith 70% ethanol , m centrifuging as in step 9. t tor 1\ . a Speed va c conc en ra . 't t , Repeat step 10 and dry the RNA pe ets In I . described in Volume , 12 . Quantif y the yields by Cerenkov co~~tl~g as ntrations to 40 fmol/~1 Chapter 1, Section 2 .3.4. Adjust the Ina ~on~.e I "ng the RNA pellets (ribozyme) and 400 fmol/~1 (RNA substrateI Y lSS0 VI in th e appropriate v olumes of TE buffer. Continued 234
4 .3. 2
Detennination of the kinetic parameters for the ribozyme reacti on
Once the reactivity of the basic ribozyme/ sub strate reaction has been established, mor e detailed kineti c a na lyses can be done. The Michaelis constant (Km ) and the turnover number (k" ,) are determined by carrying out the ba sic ribo zyme reaction (see Protocol s i over a range o f substrate concentrations both ab ove and below the K m and for times suitable to give initial velocities (64). At least two lime points ar e needed, and no more than 10"70 of the substrate should be cleaved to ensure that initi al rates can be calculated . These initial velocities (Vi) are then plotted as a function of v;l [substrate] . The K m is calculated from the negati ve slope, and k" , is calculat ed by dividing th e y intercept by th e ribo zyme concent ratio n. Suitable ribo zymes for gene targeting sho uld have catalytic effici encies (k" ,/ K m ) in the range of 5-30 min - , ~l\I - ' . 4 .3 .3
Use of selected suitable ribozymes for gene targeti ng
Upon completion of the studies to identi fy a useful ribo zyme, that is, one which has a suitable catalytic effi ciency, the ribo zyme can th en be used for gene targeting. The speci fic method of delivery and expre ssion o f th e ribozyme in vivo will be dependent on the gene being targeted an d the system being studied . For certain applications. the use of autocatal ytic cassettes to terminate the ribozymes in vivo following expressio n are useful (Secti on 4.4) . 235
David A . Shub, Craig L. Peeble s. and Arnold Hampel 7: Ribozymes
transcription of the cassette (see Protocol 4) , it will autocatalytically cleave to generate a defined terminus . It is important to note that the termini generat ed
S' Cassette (a)
s'
{
13'
Y
s:
short ribozyme-derived
sequence hamm erh ead
transcription would have termination cassettes at either or both of the 5 '
and 3 ' termini. A cassette on the 3 ' terminus of the ribozyme is especially useful, because then the newly-transcribed ribo zyme can be processed with the elimination of the requisite 3 ' poly(A) tail and normal transcriptional terminat ion sequences (58) . The resulting riboz yme is then very small and perhaps more likely to behave as an optimally functioning catalyst in vivo .
3' Cassette
s
(b)
are a 5 ' -hydroxyl group and a 3 ' terminus with a 2 ' , 3 ' cyclic phosphate group. To use an autocatalytic cas sene, it is cloned either in front of the desired transcript (to give a defined 5 ' terminus) or following the desired transcript (to give a defined 3 ' terminus) as shown in Figure 8. Among direct applications for cis-acting autocatalytic cassettes is the generation of defined 5 ' and 3 ' termini on ribozyrnes designed to be delivered for gene targeting (58, 62). That is, the ribozyme to be delivered by in vivo
desired transc ript
5' 1 C
====~~ ' - 3'
3'
Y
short riboz yme-d e rived
Sequence
h ammerh ead
s'
(e)
de sired tran script
s: ((===:::11--- 3 '
Y
Acknowledg ements The work in Section 2 was suppor ted by grants from the NIH to D. A. S. The work in Section 3 was support ed by grants from the American Cancer Society and NSF to C . L. P., who thanks them for their support. C. L. P. also thanks P. S. Perlman for scientific collaboration throughout this project , J. S. Franzen for the original d rawings adapted into severa l of the figures, and M . Zhang for scient ific and technical assistance. The work in Section 4 was support ed by grants from the Biotechnology Research and Development Corporation a nd NIH to A. H . A. H . thanks M . Altschuler and R. Tritz for technical and scientific assi stance.
short ribo zyme-derived
3'
se que nce
hairpin
sin th e hamm erh ead and ha irpin riboz~mes .. 9 3 "9ure 8 Transcriptional termination cassett~s u ' n Ib J hammerhead ribozyme to gIve k F . . e 5' terrnlO3 t to . w m ar s la) hammerhead ribozyrn .e t~ gl V . 3 ' termin ation . In each ca se . the arro termination, and (e) ha irpin nbozvrne to qrve the site of autocatalytic cleavage.
t transcripts with
4.4 Use of autocatalytic c~s~ettes to genera e
defin ed 5 ' or 3 ' termini b d forthe generation of defin ed Hairpin and hammerhead ribozymes ma Y e us~ the use of cis-cleaving autOh 5 ' or 3 ' termini on RNA transcnpts t r~~g cassettes are created by simply catalytic cassettes (59-62). These autoca~ar;b~zyme such that they are on the forming loops between the substrate an 8 U on completion of th e in VItro same transcript as illustrated in Figure . P 236
References 1. Cech , T ., Zaug , A., and Grabowski, P. (1981) . Cell, 27, 486. 2. Guerrier-Takada, C.. Gardiner, K., Marsh. T., Pace, N. , and Altman, S. (1983). Cell, 35, 849. 3. Kruger, K.. Grabowski, P. 1., Zaug, A. 1., Sands, 1., Gott schlin g, D. E.. and Cech, T. R. (1982) . Cell, 31, 147. 4 . Michel, F., Jacquier, A., and Dujon, B. (1982) . Biochimie, 64, 867. 5. Michel, F. and Westhof, E. (1990) . J. Mol. Bioi., 216, 585. 6 . Reinhold-Hurek, B. and Shub, D. A. (1992) . Nature, 357, 173. 7. Lambowitz, A. M. and Perlman, P. S. (1990) . TlBS, 15, 440. 8. Garriga. G. and Lambowit z, A. M. (1984) . Cell, 39 , 631. 9. Sambrook, L , Fritsch, E. F. , and Maniatis, T. (1989) . Molecular cloning. A laboratory manual. Cold Spring Harbor Press, Cold Spring Harbo r N Y. 10. oou, J. M. , Shub, D. A., and Belfort , M. (1986). Cell, 47, 81. t 1. Birnboim, H. C. and Doly, 1. (1979). Nucleic A cids Res., 7, 1513 . 12. Cech, T. R. (1990) . Annu . Rev. Biochem., 59 , 543. 237
7: Ribozymes 13. Michel. F. , Umesono. K.. and Oze ki. H. (1989). Gene , 82, 5. 14. Cec h, T . R. (1988). Gene, 73, 259. 15. Burke, J . M., Belfort, M ., Cech, T . R., Davies, R. W. , Schweyen, R. J ., Shu b, D. A.. Szostak, J . W. , and Tabak , H . F. (1987). Nucleic Acids Res.. 15, 72 17. 16 . Peebles, C. L. , Perlman , P . S.. Mecklenburg, K. L., Petrillo , ~1. L. , Tabor , J . H. , Jarrell, K. A ., and Cheng, H . L. (1986). Cell, 44, 213. 17. van dec Veen, R.; Amberg, A . C.; Van der Horst, G ., Bon en, L.; Tabak, H . F. ; and Grivell . L. A . (1986). Cell, 44. 225. 18. Schme lzer, C. and Sch weyen, R. J . (1986). Cell. 46 , 557. 19. K uck , U.. Go deha rdt, I. , and Schmidt, U. (1990) . Nucle ic A cids Res., 18, 2691. 20 . Schmid t, U.• Rieder er, B.. Marl. M ., Schmelzer , C.; and Stahl , U. (1990). £MBO J.• 9. 2289. 21. Hebbar . S. K.. Belcher . S. ~I. , and Perlma n, P . S. (1992). Nucleic Acids Res.. 20, 1747. 22. Iacqui er , A . and Michel. F. (1987) . Cell, 50, 17. 23. Jacqui er , A . and Michel. F. (1990). J. Mol. Bioi., 213, 437. 24. Michel, F. an d Jacquier, A . (1987). Cold Sp ring Harb. Symp. Quant, Bioi.. 52, 201. 25. Jarrell , K. A ., Dietrich . R. A. , and Perlman, P . S. (1988). Mol. Cell. Bioi.. 8, 2361. 26. Schmelze r , C. an d Mull er. M. W. (1987). Cell, 51 , 753. 27. van der v een. R., Kwakman , J . H ., and Gr ivell, L. A. (1987). £A180J.• 6, 3827. 28. Ja cqu ier , A . and Jacquesson -Breuleux, N . (1991) . J. Mo l. Bioi. , 219, 4 15. 29. Peebles, C. L. . Benatan. E. J .• Jarr ell, K. A. and Perlman , P . S. (1987). Cold Spring H arb. Sym p. Quant. Bioi.• 52, 223. 30. Koch, J . L., Bou lan ger , S. c.. Dib-Hajj, S. D., Heb bar . S. K.. and Perlm an, P. S. (1992). Mol. Cell. Bioi., 12, 1950. 3 1. Walla sch , C. . MorlvM. ; Niemer, 1., and Schmelzer , C. ( 199 1). Nucleic A cids Res. . 19. 3307. 32. Jarrell , K. A. (1987). P h. D . Dissertatio n, Ohio Sta te University, Co lumbus, Ohio. 33. Franzen, J . S., Zha ng, M., and Peeb les, C. L. (1993). Nucleic A cids Res. , 21, 627. 34. Jarrell, K. A .. Dietri ch , R. c., an d Perlm an , P. S. (1988). Mol . Cell. Bioi.. 8, 2361. 35 . Muller, 7\1. W. , Schweye n, R. L , and Schm elzer , C. (1988) . Nucleic Acids Res. , 16, 7383. 36. Jacqu ier, A . an d Rosbash, ~1. (1986). Science. 234. 1099. 37. Choquet, Y ., Go ldschmidt-Clermo nt. M ., Girard-Bascou, J., Kuck, V" Bennoun, P.. and Rochaix, J . D. (1988) . Cell, 52 , 903. 38. Hildebrand , ~ I., Hallick, R. P .. Passavant, C. W.• and Bourque, D. P . (1988). Proc. NaIl Acad. Sci. USA , 895, 372. 39. Kch chi , T ., Ogura , Y. , Umeso no, K. , Yamada . Y., Kamano , T ., Ozeki , H ., and Ohyama, K. (1988). Curro Genet., 14, 147. ~O . Goldschmid t-Clermont , M., Choquet. Y., Gl rard-Bascou, J.. Michel , F.. Schirmer Rah ire, M., a nd Rochaix, J . D. (199 1). Cell, 65 , 135. 41. Au gustin , S.. Muller , M . W .• an d Schweyen , R. J . (1990). Nature , 343 , 383. 42. Morl , M. and Schme lzer, C. (1990). Nucleic Acids Res. , 18, 6545. 43. Forster, A . and Symons. R. (1987). Cell, 49, 211. 44. Uhlenbeck, O . (1987). Nature , 328, 596. 45. Ham pel, A . an d Trit z, R. (1989). Biochemistry , 28,4929. 238
David A. Shub. Craig L. Peebles. and Arnald Hamp el 46 . Hampel , A ., Tritz, R., Hicks, l\ I., and Cruz , P . (1990) . Nuclei c Acids Res
18, 299.
.,
47. Haselo ff, J . and Gerla ch , W . (1989) . Gene. 82, 43. 48. ~uzayan, J., Ger lach, w.. and Bruening , G. (1986) . Nature, 323 , 349 . 49 . Keese , ~. and Symons, R. (J98 7) In Viroids and viroid-lik e pathogens (ed . J. S. Seman clk), p. I. CRC Press, Bo ca Rato n, FL. g, 50 . B,.ruenin G., B uza~an , J. , ~a~ pel , A ., and Gerlach , W. ( 1988). In RNA genetics; ~ olume 1/: retrOViruses, viroids, and RNA recombinati on (ed . E. Dom ingo J . Holland. and P . L. Ahlquist). p . 127. e RC Press, Boca Rato n, Flo ' 51. ROSSI. J . and Sarver , N . (1990). TlBT£ CH,8. 179.
Sarv~r,
N ., Jo hnston , J\1. , Hamp el, A., Zaia , J., Can tin, E., Chang, P., and J. (1990) ..In Gene regulation and A IDS: transcrip/ional activation , retroviru ses and pathogenesis (ed. Ta kis Papas) . Gulf Publishing Co., Hou sto n, TX, p. 305 . ;~. Fedor, ~1. and Uhlenbec k, O. (1990). Proc. Na tl Acad. Sci. USA. 87 , 1668. . Pieken, W ., Olsen, D., Benseler, F., A u rup, H ., and Eckstein F. ( /99 1) Science 253, 314. ' . , 52.
RO SSI,
Sar\i~r. N ., Cantin, N ., Pairoj, C.. Zaia , J., Ladn e, P., Stephens, D ., and ROSSI , J. (1990). Science, 247, 1222. 56. Haselo ff , J . and Ge rlach, W. (1988). Natu re, 334 , 584. 57 . Scanlo n, K., Jiao , L., Funaro , T ., Wang, W., Tone. T ., Ro ssi , J., and KashaniSabet, ~1. (199 1). Proc. Natl A cad . Sci. USA 88 10 591 " . 58 O · . j wan g, J., Ham pel, A ., Looney, D ., Wong-Staal, F., and Rappaport , J . ( 1992). Proc. Nat! A cad. Sci. USA . 89, 10 802. 59. Dzian c tt, A . a nd Bujar ski, J . (1989). Proc. NaIl Acad. Sci. USA , 86 , 4827. 60. Grosshans, C. and Cech, T. (1991). Nucleic Acids Res.• 19, 3875. 61. Taira, K., Man san ao , 0. , Shinshi , H Maeda H and Fur ukawa K (1990) P I ' Engineering, 3, 733. " , ., " . TO em 55 .
62. Altschu ler, M. , Tritz, R., and Hampel, A. (1992). Gene, 122, 85. 63. MJlltga~, J ., Groebe, D. , Witherall, G.• and Uhlenbeck. O. (1987). Nucleic A cids Res., is , 8783. 64. Fersh t, A. (1977) . Enzyme structu re and mechanism . W. H . Freeman and Co ., San Francisco , CA.
239
Suppliers of specialist items
American Type Cu lture Cnllectio n (ATCC), 1230 1 Park Lawn Drive, Rockville, :'-10 20852, USA . Amersham Inl ernat ion al P LC , Lin co ln Place, Green End, Aylesbu ry, Bucks HP20 2T P , UK. Amersham Co rporation. 2636 South Clearbrook Drive , A rling ton Heights, IL 60005 , USA . Amicon Ltd , Upper Mill , Stoneho use, G loocester G U O 2BJ, UK. Amicon Division , WR Grace & Co ., 72 Cherry hill Drive, Beverley, MA 01915· 1065, USA . Applied Biosystems Inc" 850 Lincol n Cent er Dri ve, Foster City, CA 94404, USA . Applied Biosystem s Ltd. , Kelvin Close, Birch wood Science Park No rt h, Warrington \VA 3 7PB , UK. J T Baker Che micals Inc., 222 Red Sch ool Lane, PO Box 492, Ph illipsburg, NJ 08865·9944, USA. J T Bak er UK Ltd ., Wyvols Court, Basingsto ke Road, Swa llowfield, Nr Reading, Berk s RG7 IP Y, UK. Barnstead Instru ment Co. , 2555 Kerper Bou levard , Dubuqu e, IA 5200 1, USA . Beckman In strumen ts UK Ltd ., Progress Road , Sands Industri al Estate , H igh Wycombe, Bucks. HP12 4J L, UK. Beckman Instruments Inc. , PO Bo x 3100, 2500 Harbo r Boul evard, Fullerton, CA 92634, USA . Becton Dick inson Labware, 2 Bridgewater Lan e, Lincoln Pa rk, NJ 07035, USA . Beclon Dickin son Ltd., Between To wns Road , Cowley, Oxford OX4 3LY, UK. Belleo Glass Co. , PO Box 8, 340 Endrido Road , Vineland, NJ 08360, USA . c/o Scient ific Laboratories Su pplies, Unit 27, Nottingham So uth an d Wilfo rd Industrial Estate, Rudd ington Lane, Wilford, Notting ham NG II 7EP, UK. Bethesda Resear ch Laboratories; see Gibco BRL. Bio-Rad Laborato ries Ltd ., Maylands Avenue, Hemel Hempstead , Herts H P 2 no, UK.
Appendix 1
Appendix 1
Bio-rad Laboratori es, Division Headquarters, 3300 Regatta Boulevard , Richmond , CA 94804. USA . Bodm an, Aston, PA 19014, USA. Boehringer Man nheim GmbH Biochernica, PO Box 31 01 20, 0 -6800 Mannheim . Germany. Boehringer Man nheim UK (Diagnostics/ Biochemicals) Ltd., Bell Lane, Lewes, East Sussex BN7 ILG, UK. Boehringer Ma nnheim Co rporation, Biochemical Products, PO Box 50414, Indianapolis, IN 46250-0413, USA. Bra un (B. Braun Melsun gen AG ), PO Box 120, 0 -3508 Melsungen , Germany. Braun (B. Bra un Medical Ltd ), 13-14 Farmborough Close, Aylesbury Vale Industrial Park, Bucks HP20 IDQ, UK. Brinkma n Instrument Co" Cantigue Road , Westbury, NY 11590, USA. cia Chem lab Instruments Ltd., Ho rnminster House, 129 Upmin ster Road , Hornchu rch, Essex, UK. Calbiochem, PO Box 12087, San Diego, CA 92112-4180, USA. Calbiochern-Novabiochem (UK) Ltd. , 3 Heathcoat Building, Highfields Science Pa rk, University Boulevard, Nottingham NG7 2QJ , UK. Co rni ng Medical and Scientific Glass Co. , Med field , MA 02052, USA. Ciba Corning Diagnostics, Colchester Road , Halstead , Essex C09 2DX , UK. Difco Laboratories Ltd. , Central Avenue, East Molesey, Surrey KT8 OSE , UK. Difco Labo rato ries, PO Box 331058, Detroit , Michigan 48232-7058, USA. Du Pon t Co. (Biotechnology Systems Divisiun ), PO Box 80024, Wilmingto n, DE 19880-0024, USA. Do Pont UK Ltd ., Wedgwood Way, Stevenage, Hens SG I 4QN, UK. Eastman Kod ak Co. , PO Box 92822, LRPD-l ool Lee Road , Rochester, NY 14692-7073, USA. cia Pha se Separations Sales, Deeside Indu str ial Park, Deeside, Clwyd CH5 2NU, UK. Falcon ; contact Becto n Dickinson . Fisher Scientific Co ., 50 Fadem Road , Springfield, NJ 07081, USA. Flow (ICN Flow), Eagle House, Peregrine Business Park , Gomm Road , High Wyco mbe HPI3 7DL, UK. Flow (ICN Biomedical Inc.), 3300 Highland Avenue, Costa Mesa, CA 92626, USA. Fotodyne Inc. , 16700 West Victo r Road , New Berlin, Wisconsin, USA' . Gibco BRL (Life Techno logies Ltd) , Trident House, Renfrew Road, Paisley PA3 4EF, UK. , Gibco BRL (Life Technologies Inc.), 3175 Staler Road , Grand Island , N\ 14072-0068, USA. Gilson Fra nce SA, BP 45, 72 rue Gambetta, 95400 Villiers le Bel, France. clo Anderman Ltd ., 20 Charles Street, Luton, Beds LU2 OEB , UK. Hoefer Scientific Instroments, PO Box 77387-0387, 654 Minnesota Street. San Francisco, CA 94107, USA. Hoefer UK Ltd., Newcastle, Staffs ST5 OTW, UK.
HyCion e La~oralor.ies, 1725 State Highway 89-91, Logan , UT 84321, USA. 1B~~~lernalional BIOtechnologies Inc.), PO Box 9558, New Haven CT 06535,
242
IBI Ltd ., 36 Clifton Road , Ca mbridge CB I 4ZR , UK. ICN Biomedicals; see Flow (ICN) . ISCQ Inc., PO Box 5347, Lincoln, NE 68505, USA. cia Jon es Chromatog ra phy Ltd ., Tir-y-Berth Indu strial Estate New Road Hengoed , Mid Glam organ CF8 8A U, UK. " Isolab Inc., PO Box 4350, A kron, O H 44398-6003, USA. cia Genetic Research Instrument s ltd ., Gene Hou se, Dunmow Road , Felstead, CM6 3LD, UK. ~imble Products, 1022 Spruce Street, Vinela nd, NJ 08360, USA . Kodak ; see Eastman Kodak . Konles Glass Co., Vineland , NJ 08360, USA . ci a ~urkard Scient.ific Sales, PO Box 55, Uxbridge, Middx UB8 2RT , UK. M.A . BlOproducts (MIcrobiological Associates) , Biggs Ford Road, Building 100, Walkersville, MD 21793, USA. cia Lab Impex Ltd., Waldergrove Road , Teddington, Middx., TWII 8LL, UK. Macherey Nagel, Postfach 307, Ne umann-Neandemrasse, 0-5160 Duren, German y.
cia Ca mlab Ltd ., Nuffie ld Road , Ca mb ridge CB4 ITN , UK. Millipor e Int erte ch, PO Box 255, Bed ford , MA 01730, USA. Millipore UK Ltd ., The Boulevard , Blackm oor Lane, Watford, Herts WDI 8YW, UK. Nalge ce., PO Box 20365, Rochester, NY 14602-0365, USA. cia FSA Labo rato ry Supplies, Bishop Meadow Road , Loughb orough Leics ' . LEII ORG , UK.
~ew England Biolabs GIIBL), 32 Toze r Road , Beverley, MA 01915-5510, USA. New England Nuclear (NEN), Du Pont Co . , NEN Resea rch Products 549 Alban y Street, Boston, MA 02118, USA. ' Du Po.nt !! K Ltd., Wedgwood Wa y, Steve nage, Herts SGI 4Q N, UK. Pharmacls Blosystems Ltd. (Biolechnolo~y' Division), Davy Avenue, Knowlhill, MIlton Keynes, MK5 8P H. UK. Pha rmacia- LKB Biotechn ology Inc. , PO Box 1327, 800 Cen tennial Avenue, PIScata way, NJ 08855-1327, USA. Pierce, PO B? x 117, 3747 Nort h Merida n Road , Rockford , IL 61105, USA. .cIa Life SCience Lab s Ltd., Sedgewick Road , Lu ton, Beds LU4 9DT, UK. :,erce Europe BV, PO Box 1512.3260 BA Oud- Beijerland , T he Netherlands . rome ga Ltd . , Delta House, Enterprise Road , Chi lwonh Research Centre Southampton SO l 7NS, UK. ' Pr~me~a , 2800 Woods Hollow Road, Madison, WI 53711-5399, USA. Qalgen Inc. , Studio Cit y, CA 91604, USA. ~~ HYbaid Ltd., 111-113 Waldegrave Road, Teddington, l\liddx TW I I 8LL, 243
Appendix 1 Sartorius AG, P ost fach 32-43, Weender Land str asse 94· 108, 0·3400 Gotti ngen, Ge rma ny . Sartorius Ltd. , Longmead , Blenheim Road, Epsom , Surrey KT9 9QN, UK. Sartorius ;-';orth America Inc., 140 Wilbur Place, Bo hemia, Long Island, NY 11716, USA . Schleicher & Schuell , Post fach 4, 0 -3354 Oassell, Germany. c/o Ande rman & Co . Ltd ., 145 Lon don Road , Kingston-upon -Tham es, Surrey KT2 6N H , UK. Searle ; contact Amersham . Sigma Chemical Co. Ltd. , Fancy Road, Poole, Do rset , BH 17 7N H, UK . Sigma Inc. , PO Box 14508, St Lo uis, ~I O 63178, USA. Sorvall ; co ntact Du Pont . Squibb Pharmaceuticals , 1 Squibb Drive, Cra nberry, NJ 08512-9579, USA. Stratagene Inc. , 11011 Nort h To rr ey Pin es Road, La J olla , CA 92037, USA. Stratagene Ltd ., Un it 140, Cambridge Inn ovati on Cent re, Milton Road , Cambridge C B4 4FG, UK. United States Biochemical (USB) Corporation, Box 22400, Cleveland, O H 44122, USA . c/ o Ca mbridge Bioscience Ltd., 25 Signet Court, Stourbridge Co mmon Business Centre , Swans Road , Ca mbridge C B5 8LA , UK. University of Wisconsin, Genetics Computer Group , University Avenue, Madiso n, W I 53706, USA. UV Products Inc . , 5100 Wal nut Grove, San Gabriel, CA 9 1778, USA . UV Products Ltd. , Science Park , Milton Road, Ca mb ridge C B4 4B N, UK . Van Waters & Rogers, P O Box 6016, Cerr itos , CA 90702, US A . Virtis Co. Inc., Route 208, Gardiner , NY 12525, USA . c/o Damon/ IEC (U K) Ltd ., Unit 7, Lawren ce Way, Brewers Hill Road , Dunstable, Beds. LU6 lBO, UK. VWR Scientific Products, PO Box 7900, Sa n Francisco, CA 94120, USA . Wako Pure Chemicals, Dosho-M achi, Osak a , Japan . Waring Commercial, c/o Christison Scient ific Equipment Ltd., Alb an y Road , Ea st Gateshead Industrial Estate, Gateshead NE8 3AT, UK. What man Scientific Ltd ., Whatman Hou se, St Leonards Road , Maid stone, Kent MEI 6 OLS, UK. Worthington Biochemical Corporation , H alls Mill Road , Freehold, NJ 07728, USA. c/ o Ca mbridge Bioscience Ltd. , 25 Signet Co urt , Sto urbridge Common Business Cent re , Swa ns Road , Cambridge C B5 8LA, UK.
244
Contents of Volume I 1. Synthesis and purification of RNA substrates Benoit Chabot
2, Cha rac terization of RNA Paula J. Grabowski
3, Splicing of mRNA precursors in mammalian ce lls Ian C. Eperon and Adrian R. Krainer
4, Isolation and characterization of ribonucleoprotein complexes Angus I. Lamond and Brian S. Sp roat
5, Analysis of ribonucleoprotein interactions Cindy L. Will. Berthold Kaster, and Reinhard i.iihrmann
6, Analysis of pre-mRNA splicing in yeast Andrew Newman
Index .\!any of (he standard methods for preparation of RNA substrates/or processing reactions, and for anatvsis of RSA will be found in Volume I agarose (na tive) gel electrophoresis analysis of kineto pla st RNA 87-8 a utocata lytic C aS~CI [eS, ribozyme fo r generating 5' and 3 ' R;,\A term ini
236- 7 bac teriophage RNA polyrnerases. for in vitro tra nscript io n of capped mR NA 43-4 gui de R NA ., 91
pre -mRNA for 3' end-processi ng 2- 3 pre-r RNA pre-I R NA
1 ~-8
181- 5 substra tes for cryp tic RNa se 100-1 see also in vitro tran scri p tion base modificat ion s in IRNA
199-206
calf thym us. 3 ' end-processing factor" 29- 33 capping enzymes, see guanylyl transferase. RNA triphosphatase capping o f cellu lar m R:":A
by enzym ic recapping 47- 8 by in vitro transcription 43- 4 viral RiS'A using insect cyto plas mic polvbedrosis virus 39-41 using reo virus 41-2 using vaccinia virus 42- 3 cap structures in mRNA chemical diversity 35- 6 enzymic cleavage 55-6 fo rma tio n in vn-o 36-7 ra diolabelling by decapping an d reca pping 45- 8 by in vitro transcriptio n 43-4 by O-methylation 51-2 ove rview 38-9. 45 by perio date oxidation and borohydride red uction 39- 43 by viral transcript ion 39- 43 role in mRNA stab ilization 3 sepa rat ion of. by ion exchange chromatography 57-9 paper ch romatograp hy 59-61 pa per electropho resis 62- 3
T LC on PEl -cellulose 61-2 two-dimension al chro matography 64 - 5 cells, tissue cu lture. extract prepa rat ion. for mRNA turnover 114-19 . see also HeLa cell cis-splicing 225-6 cleavage and polyadenylation of mRJ"A in vitro, co upled 21- 3 cleavage an d polyadenylation specificity factor ICP SF) assay 32-3 interaction with AA UAAA sequence 6-7 .
\5 in vo lvement in 3 ' end-p rocessing of
mRN A 1-2.24 puri ficatio n from calf thymus 29-30 HeLa cells 27-8 separation from poly( A) polymerase 27-9 cleavage of m RI'A 1-2,21-7 r RNA 135-8, 149-51. 153- 5 CP SF. see cleava ge and pol yadenylation specificity facto r cross-linking RNA- prOlein 160 -4 RNA-RNA 1>2-3 see also Volume I cryp tic RNa se 96. 100-1 cytoplasmic extr acts for mR NA tu rnove r 111- 14. 119- 26 fo r rR NA processi ng 150- 5
deca ppi ng of RNA
chemic a l 45-7 enzymic 52-5 deletion a na lysis of rR NA processing 155 Dj-lf'R c- Iigase fusion protein. expression \9 \->
edit ing. RNA . in kineto pla st mitochondria assa y of enzymes involved crypt ic RNa se 100-1 gRNA:m RNA chimaera-forming activity 10\-3 overview
96- 7
Ind ex editi ng, RNA , in kin eto pla st mitoc hon d ria (conL )
assay of enzymes involved (cont.) Rl" A ligase 99 -1 00 term ina l uridvlvl transfera se 97- 9 choice o f experimental orga nism 71- .=! clea vage-ligatio n mod el 69 - 70 clo ning of L. tarentotae 7~ enz yme cascade model 69- 70 gRN As hvbrid selection 88-90 idcntification 94-6 in g RNA:mRNA chimaeras 69-70, 90,96 in vitro transc ription 91 growth and ma intenance of L. tarentotae 72-5 kinetoplast DNA (kONA) isolation 76- 8 schizode me typing of protozoa 78-9 see also kineto pla st DNA kinetoplast mitochondria l fraction 69- 71. 8 1- 5. see also mitochondria kinetopl ast RNA (kRNA) isolation 87-8 Nort hern blot ana lysis 88 occurrence 69 -7 1 p e R am plification of partia lly ed ited m RNAs and gRNA: m RNA chima eras 90, 92-3 tra ns-esterifica tio n mod el 69-70. 90, 96, 101- 2 ty pes of edi ting 69 - 71 electro n micr o scop y of r RNA syn thesis 143-4 electr o phoresis , see agar ose (native) gel electrophoresis, formalde hyde- agarose gel electrophoresis , pa per elect ro pho resis. polyacrylamide/urea (denat ur ing) gel elect rophores is, 5 0S- PAG E; see also Vo lu me I end-labelling of RNA by ligation 15- 16, l,n-8 using T-1. pclynucleori de kinase 1-1.6- 7 see also Volu me I endonuclease , yeast preparation 189- 91 in p re-t RNA pr ocessing 186 , 196-9 endonucleolyric cleavage of m RN A in abse nce of pclyadenylation 21, 23 ass ay 23 co upled to po lyadenylatio n 21-3 exam ple 24- 5 overview 1- 2. 21 p urificatio n of factors invo lved 26 - 7 3 ' end- process ing o f mR NA in vitro clea vage of RNA su bstrates 23 c1ea vage-polyadenylatio n 2 1-3
exa mples 24-5 ext ract preparation 4-6 general considerations 2-4 interpretation of data 2-1.-5 non -specific pclyadenylation 2~ o verview 2 1, 26 polyadenylation co mplex format ion analysis by electrophore sis 7-10 analysis b y modi ficatio n-interference 15-21 ana lysis by RNa se pro tection 10- 14 bac kgro und 6-7 basic procedure 7-10 time co urse 9-10 pol ya denylati o n of RNA 23-4 RNA subst ra tes 2-~ , 15- 16 3 ' end- processing of mRNA in vivo 1- 2, 33 3 ' end-p rocessing of rR NA 135- 7 , 169- 70 3 ' end -p roc essing o f tRNA 175. 183. 188- 9 5 ' end- proc essing of m RN A, see cap ping , cap structures 5 ' end-processing of r RN A 135- 7.148-51 . 153- 5 5' end-processingof t RNA 175. 183, 186. 188-9 enzvrne cascade mode l, RNA editing 69-70 e.\o~-b i ndi ng seq uences in group II ribozy mes 219-20,225 . 228 expression o f DH FR- ligase fus ion p rotein 19 1- 4 exte rna l tra nscribed space r. pre-rR NA 135- 9, 142, 144, 148, 150-1, 155, 157- 60 extract preparation , for RNA processing , see lysolecit hin ext ract s, nuclear extracts, post-nuclear superna ta nt . post-polysomal super natant, pro cessing extr acts ; see also Volume I
for maldehydc -agarose gel electrophoresis in No rth ern blotti ng of k RN A 88 of r RNA 13 9-~2
gel-shift analysis of prolein-r RNA inte ract ions 156 - 7 gene ta rget ing with ha m mer head and hairpin rlbozvmes 235- 6 glob in m RN A turn over 127- 30 gradient centr ifugatio n, see isop ycnic density gradi ents . rate zonal cent rifugat ion gRN A: m RNA chimaeras amplification by PC R 90 an alysis by RT - PC R 103 for mat ion in mitochond rial ext racts 96- 7, 101-3 ro le in RNA ed iting 70
248
Index group I intron ribozymes de tection 2 13-15 mechani sm 211- 13 occurren ce 211 group II intron ribozvm es cis-splicing 225- 6 ' disti nction from group I int ron ribo zymes 217-20 ide ntifica tion o f 217- 19 occ urr ence 2 17- 19 pOIentia I uses 219-21 rever se splicing and transpo sit ion rea ctio ns 228- 30 seco ndary structures in 219 -20 self-splicing o f ba sic proce d ure 222- 3 mechani sm 218-20 mod ificatio n of reaction conditions 2 2~ tran scri pts for 221-2 spliced exon reo pening reactions 225- 7 sp liced junction hyd ro lysis 225- 7 transcripts fo r. prepa ration 221-2 trans-splicing reactions 227- 8 gui de RNA hybrid selectio n o f 88-90 identific ation by compu ter-assis ted seq uence co mparison 94-6 in edit ing of mitochondrial RNA 69 - 71, 87 synt hesis by in vitro tran scrip tio n 91 see also gRN A: m RNA chimaeras
hammerhead and hairpin ribozym es in au to catal ytic ca ssettes 236- 7 de sign o f 232- 3 fo r genera ting transc ripts with defin ed term ini 236- 7 in gene ta rgeting 232-6 kinetics of 233. 235 mecha nisms 230-2 occurrence 230 structure 230- 3 HeL a cell extracts for 3 ' end-processing 4-6 rR NA processing 150- 5 m "Gvspecific pyrop hospha ta se 53 p urifica tion of 3' end -proce ssing factors 26-9 heparin-agar o se chromato gra p hy. puri fication of yeas t endo n uclease 189- 9 1 histone m RNA end -p roce ssing 2 turnover 129 hyb rid selectio n of gui de RNAs 88- 90 rR NA intermedia tes 139-42
int ernal transcribed spacer of r RN A 136 inr ro n-bind ing sequences, in group II ribo zym es 2 19- 21.225.228 int rons group I self-splicing ribozymes 21 I -1 7 grou p II self-splicing ribozym es 2 17- 30 in vitro t ransc ript io n to co nfirm exist ence of group I ribo zymes 215- 17 of RNA su bstrates capped 43- 4 fo r 3 ' end-pr ocessing 2- 4 for gro up II self-splicing im rons 22 1-2 for hammerh ead and hairpin rib ozyrnes 233- 5 fo r pre-m RNA turn over 108-9. 121-6. 128- 30 pre -rRN A 144-6 p re-t RN A 181- 5 viral 38-43 see also Volume I io n exc hange chromatography of ca pped oligo nucleotides 57- 9 DHFR- Iigase fusion protein 192- 5 po ly(A ) po lymerase and CPSF 27- 30 pre-tR NA substr ate 179-80 iso pycni c gradient cen trifuga tio n using CsCI for kDNA 75- 8 for maxi circle DNA 79-8 1 usi ng Reno grafin o r Pe rcoll isolation of kinetoplast mitochondrial fract ion 82-6 relati ve merits 82. 9 1
kinet op la st DNA (k DNA) a nalysis 78- 9. 82 int egrit y of 78. 80 isolat ion 76-8 ' max icircle' DNA 75.79-81 'minicircle' DNA 75 properties 75-6 schizodeme typing of kine toplastid protozoa 78-9 transcripts of 85-7 visua lizatio n 78-80 kinetoplasud p rotozoa choice of species 71-2 edi ting enzymes, assay 96-103 grow th and maint enance 72-5 guide RNAs of 88- 9 1, 94-6 k DNA pr eparation 76- 8 I mito chondria, isolation 81-5 mit och on drial tran scripts 85-8 RNA ed itin g in 69 - 71 run-on tra nsc riptio n in mi tocho ndr ia 9 1-4
249
Index
Index kinet o pla st RNA (kR NA) analysis 87-8 co mposi tio n 85- 7. 90 edit ing 69- 71 hybr id selection of g R NA~ from 88- 90 isolati on 85- 8 PCR a mplificatio n of mRNA and gRNA :mRNA chimaeras 90
labelling RNA of caps 39-43, 47-52 end-labelling by ligation 15-16,147-8 using polynucleotide kina se 146- 7 by group I intron ribo zyme s 213- 15 by in vitro tr anscription 8RNA ' 91 group I intron s 215-17 group II introns 221-3 hammerhead and hai rpin ribo zymes 233-5 mRNA 43-4 rRNA 144-6 l RNA 181-5
in vim by liga tion 15- 16. 99-1 00 r RNA 139-40 t RNA 177- 81 by ru n-on tra nscriptio n 9 1-4 see also Volume 1 lariats electrophoresi s of 221 in group II intron self-splicing 218-19 , 225-6, 228-9 purification 229 lysolecithin lysis o f cells 4, 117- 18 ext racts fo r mRNA turnover 117- 19, 123-5 messeng er RNas es 112, 119. 122-3 methotrexate-agarose chromatography 192-3 methylation of RNA in vitro 38- 42,47-8 ,51 -2 in V;\'O 35-7 micrococcal nuclea se removal of mRNA fr om poly somes 112- 13 treatment of extra cts 118-19, 125-6, 164-5 mitochondria with group 1 self-splicing intr cns 211 with group 11 sett -spliclns introns 217-1 8 kinetoplast editing enzymes 96-103 extract s for editing 97. 102-3
guide RNAs 88-9 1, 94-6 isolation 81- 5 kDl" A prepar a tion 75-81 RNA edit ing in 69- 71 run o n trans cript ion in 91- 4 tra nscri pts of 85-8 . mod ification-int erferen ce analysis of polyaden ylat ion co mplexes 15-21, see also Vol ume I mRNA analvsis 126- 30 cap ping 35-4-l , see also capping , cap str uctures destabilizing factor 129-3 0 editing , in kinetopla st mitochond ria "" 69- 71, 90-4 , 96-103 , see also editing. guide RNA end-processing 1-25 , see also 3' endproce ssing in gRNA :mRNA chimaeras 69- 70, 90, 92_3 ,96_7 ,10 1_3 polyadenylati on 21-33 . see ~/so 3' end~ processing , polya den ylat lon co mplexes, pol yadenylation of mR NA. poly(AI polym era se preparati on IlO-11 splicing gro up I introns 211 ~ 17. see also group I int ro n T1bol ymes gro up II int ro ns 217 ~ 30 . see also group II iruro n ribo zymes see also Volum e 1 stability regulat or y factor 112-30 substrates for mRNA t urnover 108-11 ., 1 turnover in vitro of endogenou s mRNA lO8-9. 119- _ , 124, 127- 8 extract for. pr epar at ion 111-19 o f in vitro transcripts 108-9 ,121-6, 128-30 2"'-3 messenger RNa!'.cs 112,11 9, 1 _ ., 6 of purified cellula r mRN A 108-9.1 . 1rat es 130- 1 regulation 129-30. 132 tro ubleshooting 131-2 see also Volume I RN .\ mRNP. end ogen ou s subst ra te for m . turnover 109 see also Volume I Northern blot analysis of kRNA 88 . RNA of snRNA!'. associated with pr e-r 157, 160 see also Volume 1 nuclear extr act!'. for 3' end-processing of mR NA 4-6
rR NA proces sing 148-50, 16-l- 9 t RNA proces sing 187-9 see also Volume 1 nuclea se-free ca rriers for ethanol precipitation 176-7 nuclea se PI in analysis and prepa rat ion of RNA cap s 40, 42, 52, 54-5, 62 in identification of IR NA splice sites 201 - 3 nucleo la r endonuclease , cleavage of prer RNA ISS nucl eolar extract, fo r rRN A processi ng 150-5 nucleo tide analysis by TL C o n PEl-cellulose 201-4 by two-dim en siona l chro matography 203, 205-6 nucl eotide seq uence analysis of gR NAs 94-6 o f pa rt ially-edited mRN A 90, 92-3
oligo nucleotide-directed RNa se H digestion o f RNA in rRNA processing extracts 165- 8 in Xen opus oocytes 168-9 see also Volume I
pap er electr oph oresis of mRN A caps 62- 3 periodate oxidation in deca ppin g mR NA 45- 7 in lab elling caps 48-5 I poly(A)-binding protein, in mRN A tu rnover 128-30 polyacrylam ide (native) gel electrophor et ic analysis of poly adenyla tton co mp lexes 7-10 polyacrylamide /urea (denaturing) gel electroph ore sis assay of gRNA:m RNA chimaeras 102- 3 a ssay of mitochondrial cryptic RNase 101 for gro up II self-splicing intron s 221 of pre-tRNA 180-1,1 83- 5 in scbizodeme ana lysis of kDN A 78-9, 82 of spliced tR NA 197-9 see also Volume I pot yad enyla tio n complexes an alysis by electrophoresis 7-10 modification-interference 15-20 RNase protection 10- 14 fo rm atio n in extracts 6- 10 polyadenylation of mRNA cleavage and polyadenylaticn specificity facto r, pu rifica tion 26-30, 32- 3 in extracts coupled to endonucleolytic cleavage 2 1- 3
inte rpretation o f data 24- 5 method and analysis 23-4 over\"ie" 21 prec leaved RNA substrates 23- 4 general co nside rat ions 2- 4 in mammalian cells 1- 2 no n-specific 24 poly(A) polyme rase , pur ification 26- 32 substrates for 2-4 , 15- 16 poly(A) polymerase assa y 30-2 in 3 ' end -processi ng of mRNA 1-2,24 purificatio n 27-30 separat ion fr om CP SF 27- 9 polymerase chain reaction (pe R) amplification of pa rt ially ed ited mRNA and gRNA: mRN A chima era s 90, 92-3 see also Volume 1 polysomes as endogenous subst rate fo r mR NA tu rnover 109, 119- 21 no n-specific mR NA degradat ion in 131 preparation 110-11, 114-16 ribosomal salt wash (RSW) from 116- 17 as source o f messenger RNa se 112 turnover of m RNA in 127-8 pos t-nuclea r supe rnatant mRNA turn over in 123- 5 prepara tio n 117-18 treatment with micrococcal nuclease 1\8-19 post- polyscmal supern ata nt (5 130) messenge r RNases in 119 pr epar atio n 114-16 pre- mRNA, pre· rRNA, pr e-tR NA, see in vitro transcriptio n, lab ellin g RNA, RNA substrates process ing ext rac ts for edi ting of mRNA 97 -103 end-processing of mRNA 4-6 ,21-5 mR NA turnover 111- 26 r RNA 148- 55 tRNA 185-91,195-9 see also cyto plasmic extracts, lysolecithin extracts. nuclear extracts, nucleolar extract, post-nuclear superna ta nt , post-polysomal superna tant, and Vo lume I psoralen . cross-linking of snRNAs and rRr--;A 142-3 pyrcphcsphatases, in cap analysis 39, 52-6 rate zona l centrifugation, analysis of r RNA protein associations 157-9 retic ulocyt e lysate , mRNA turnover in 113. 125- 6
251 250
Index reverse splicing of group II imrons 226- 8 ribo so mal RNA processing in vitro a nalys is by cross -lin king 160-4 deletion a nalysis 155 electro phor esis 149- 50, 153- 5 gel-s hif t analysis 156-7 immunolo gical method s 169 micrococcal nuclease d igestion 1~- 5 oligonucleotide- RNas e H targeting
165-8 fate zona l centrif ugati o n 157- 60 cleavage of pre-r RNA by nucleolar endonucl ease 155 p re-rRNA su bstrate, preparat ion 144-8 process ing extracts prepa rat ion 148-53 proces sing of pre-rRNA in 149-50, 153-5 RNA- prot ein associations involved 155-64 ..n RNAs invo lved 164- 9 riboso ma l RNA processing in wvc ana lysis by cr oss- linking 142-3 elec tro n micro scopy 143-4 electrophoresis 137-42 hybrid selectio n 139- 4 1 micro injectio n of oocyt es 137, 168- 70 o ligo nucleot ide- RNase H tar getin g 168-9 in a ..soc ia rion with tran scription termin ation 169- 70 end- p rocessing 169- 70 labelling o f r RNA in cells 139-40 pat hways invo lved 135- 7 in X enopus oocytes 137, 168- 70 ribosomal salt wash (R5 \\.1 p reparatio n fro m polysomes 116-1 7 a s so u rce of messeng er RNa ses 119 for tu rnover o f mRNA 122- 3 ribo zyme s de finition 211 gro up I iruron ribozym es 21 1-1 7 group 1I imron ribo zyme s 217-30 hammerhea d and hairpin ri bo zyme s 230- 7 RNA an alysis by electr ophoresis 87-8 , 102 -3 , 180- 1, 183- 5. 197- 9 modification-interference 15-21 No rt hern blott ing 88, 157, 160 RNa se prot ection 10-14 see also Volum e I ca pping 39-44,47 -8 caps 35- 7,39- 43,45 -5 2 editing 69-103 , see a/so ed iting of RNA end -labelling 15-16, 146- 8, see a/so 3' end-p rocessing of mR NA
gua nylrransferase (ca pping enzyme) 37-8 , ' 5. '7-8 ligase in editing of RNA 96. 99- 100 in end-l abe lling o f RI'A 15- 16 prepa ration as fusi on protein 191-4 in pr e-tRNA splicing 186, 196-7 methylt ra nsfera ses 35-8 , 47- 8, 51- 1 po lymerase, see bacteriophage R~A polymerases, in vitro transcription self-splicing gro up I i m r c ns 211- 15 group II ir uron s 218-22 see also ribo zyme s splicing, see splicing substrates for processing reactions cap ping of mR NA 38- 44 with de fined termi ni 236- 7 edi ting mito ch ond rial mR NA 82-8 end -processing of mRNA 2-4 ,15-16, 23-5 grou p I introns 215- 17 group II in tro ns 221-2 ha mmer head a nd hairpin ribozymes 233- 5 pre- rRNA 144-8 pre-tRNA 177- 85 tu rno ver of mRNA 109-11 , 119-26 see a/so Vo lum e I targeting by ha mmerhead and hairpin ribozymes 232-6 RNa se H 165- 9 tripho sphatase 147 RNA-protein interact io ns in rRNA processing a nalysis by gel-shift 156-7 rate zonal centrifugation 157-60 UV cross -link ing 160-4 RNa se A, ident ific at ion of t RNA splice sites 203-6 RNase , cryptic, in RNA editing 96. 100-1 RNa se H , for targeted destruction of RNA 165-9, see also Vo lume I RNa se inhibitors, in m RNA turnover 110-11 .11 3-1' RNa se protectio n, analysis of polyadenvlancu complexes 10- 14 RNa se TI analysis of polvadenylation complexes 13 ident ification of IRNA splice sites 103 -6 RNase T2 analysis of mR NA caps 52, 54-5, 59-60, 63-' analysis of poly adeny lation complexes 13 RNP com plexes , a nalysis by electron microscopy 143-4 gel-shift a ssay 156-7
252
Index imm uno log ical methods 169 psoralen cross-linking 142- 3 rate zonal cent rifugarion 157-60 U \' cross-linking 160- 4 see also Volum e I run -off tran scr ipti on, see in Vitro tr amcript ion
terminal uridylyl transferase assay 97-9 in RNA editing 96-7
TLC of
capped oligon ucleo tides 61-2 rRNA nu cleotides 201- 4 Irans-acting m RNA stability regulato ry factor ~ffecl on m RN A tu rnover 129-30 " III pos~-polysomal supematanr, 112-13 schizodeme typi ng of kineroplaslid proto zoa trans-act 109 processing fac to rs 78- 9 in 3.' , end- process ing o f m RN A 1-2,6- 7 5 DS- PAGE, o f snRNP pr ot eins 162-4 , see pur ification 26- 33 a/so Volume I {rans-c1.ea ~ag~ of RNA, by ha mme rhe ad an d self-splicing imrons hai rpin ribozymes 230-2 detection in cellula r RNA 213-1 5 I rans -ester ifica t i on gro up I !n RNA editing 69 -70, 90, 96 , 101- 2 detectio n 213-15 10 self-splici ng im ro ns 211 - 13,21 8-20 2'4 mechani sm 211-13 tran sfer RNA ' Occurren ce 21 1 base modificat ion s 199 - 206 grou p II la belling in cells 178- 79 ba sic react ion 222- 3 processing distinL"tio n from group I 218 assay 195- 9 mechanism 218-19 endonuclease 189- 91 modi ficatio ns 224 ext rac t pr epa ra rion 186-9 transcrlpr, for 221-3 ligase 191-5 site-specific labe lling o f cap s overv iew of pat hway s 175-6 chem ica l 48-52 pre -rRNA su bstra tes, prepa ration 177- 9, by dec appi ng and recap ping .15- 8 181- 8 ove rvie w 38- 9, 45 p urification of factors 189- 91 by tra nsc ript ion 39-44 splice sire ide mification 199-'06 snake, veno m phosphodi ester ase, in ident ifica splicing 175,1 86, 189 , 196-9(Jon of [RNA splice sites 201-3 tran spo snlon, see re verse splicing o f gro up II snRNA m rrons a nalysis by Nort hern blot ting 157, 160 tra ns-sp licing o f gcoup 11 intr on s 227- 8 cleav age by RNa se H 165- 6 168 turnover of mRNA in ext ra cts cross-lin king to pre -r RNA 142-3 analysis 126- 30 10 pr e-rRNA processing 136, 142 choice of substrate 108- 10 see a/so Volume I dat a interpre ra no n 130 - 1 SP6 RNA pol ymerase, see bacteriop hage o f dea denylat cd mRNA 128 RNA pol ymerase, in vitro tran scription deca y ra tes 130-1 sp hero plas ts, yeas t J87- 8 deteclion of m RNA 126- 3 1 splici ng using endogenous m RN A 119- 21, 124 of group I introns 211-1 7 examples 126-30 of group II In trons 217-30 using exogeno us mR NA 121-6 reverse splicing 226 - 8 exper imental adva ntages 107- 8 spliced exon reopenin g 225 - 7 extract preparation 111-19 splice juncti on hydr ol ysis 225-7 in hibi tion of non- specific RNase 113-14 splice iu ncuo ns in IRNA 20 1-6 polY(A)-~inding prot ein , effec t of 128- 30 of ' RN A 175. 186, 189. 196-9 prepar ano n of u ndegraded m RN A 110- 11 see ~/so gro~p I intr on nbozymes, grou p 11 problem solving 131-2 .. mrron ribozym es, and Vo lume J q uanti ficat ion 126-31 sta bility fa cto r for mRNA, see trans-a ctin g regulat ion 129- 30, 132 m RNA sta bility regulato ry fa ctor two -dim ens ional chromat ography of capped oligon ucleor ide 64- 5 IRNA nucleOlides 205-6 T3, T7 RNA POlymera.~e, see bacte riophage RNA polyme rase. in vitm tra nscr ipli o n
U V cross-linking of rRN A-prOlein met hod 162-4
253
ORDER OTHER TITLES OF INTEREST TODAY
Index UV cross-linking of rRNA-prot ein (co n t . ) prep aration of substituted rRNA 162- 3 synthe sis o f 4-S-UTP 160- 2
viral trans cription of capp ed RNA 38- 43
Xen opus oocvtes, pr e-rRNA processing in 168-70 yeast, RNA processing in gro up II int ron ribozymes 217,221,225 tRNA processi ng 173- 81, 183, 186-91, 194, 196- 8, 201 see also Volume I
Price list for: UK, Europe, Rest of World (excluding US and Canada)
138.
Pla sm iu s (2.'e)
Hardy .
136.
0-1 9-963 445-9 £30. 00 Paperback 0-19- 963 444-0 £19 .50 R\:A Proce ssin g: Vol. II Higgins. Sr . &
1<.(;
(Edl
117 .
Ha mes . [J.D. (Ed s) 0 -19-96347 1-8 £3 0.00 0-19-9 63470- X £19 ,50 Higgins . 5,1 . & Hames .
Sp ira lbo und hardbac k Pape rback
135.
J ii4 133.
R:\A Pro ces sin g: Vol. B.U. (Eds ) Sp ira lbo un d hard back Pape rback
0-19 -963344 -4 0 -' 9-963343 -6
:\;MR or Mac r omolecules Gas Chromatograph y
£40 .00 £2 7.50 Stern , CIl. &
Essential De velopmental Biolo gy Ho llan d , P.\V,H. (Eds)
0-19-96 3423-8 0· 19·9 6 3422- X
Cellular Interactions in Development 0 ./" (Ed) 0-19-963391- 6 0- 19-96 3390- 8
Sp ira lbo und hard back Paperb ack
129
127 .
Behaviour-al Neuroscience: vnlume II /\, (Ed)
Hartl ey.
12 6.
Gene Targeting
0- 19-9633 58 -4 0- 19-96 3357 -6 0-19 · 96 3407 -6 0- 19-9634036 -8
125.
112.
111 .
123 . 121.
120.
254
119 . 118 .
J. (Ed)
immunocytochemist ry
Beesley.
0-1 9-96 3 270-7 0-19-963269-3
Protei n Pho sphoryl at ion Tumour lmmu nu hiulu gy R e. & others [Eds ] Spiralbou nd hard back Paper back
in Situ Hvbr-idiz afinn
£32.50 £22.50
0- 19-963 366-5 0- 19-963 3 65-7
Dafni. ,\ , £32.50 £22.50
0- 19-9 6329 9·5 0-19-963298-7
Wil kinson , I),G, (Ed )
Spir al boun d ha rdback Paper back
0-1 9-963328-2
Spiral bo und hard hac k Paper back
0-19-963 139· 5 0-19-9 63 138· 7 Stevenso n . n.R
£3 0.00
£3 5.00 £25. 00
. Unllin .
ou
105 ,
Hiomechanlcs-Matertal s
£30 .00 £19.50
Vincent . J,F.\'. lEd)
0-19 -963223-5 £35 .00 0-19 -96 3222 -7 £25.00 Anim al Cell Cu ll u r e (2ie ) FreshllPY . IU, (bl l Spir albound ha rdbac k 0 -19-9632 12-X £30.00 Pape rbac k 0-19-963213-8 £19.50 Mole cular Plant Pathology: Volume II tIurr. Spi r al bo und hardback Pap erb ack
104 .
£3 5 ,00 £2250
103
S.)., \ lcP her soll . \I.J. &: others (F.ds) Spiralbound hardback 0-19-9 63352-5 Paper ba ck 0-1 9-9 63351 -7
£32 5 0 £2 2.5 0 £40 ,00 £2 7.50
Ce ll-Cel l In ter a ct ions \\'. 1. & uthers (Eds )
Herrington. C.S . & l\!l Gee. I. jEds) Spir alboun d har d back 0-1 9-9 63237-5 Pape rbac k 0-1 9·9 63 236-7
£3 0.00 £18 .50
Lallagher, G.. Rees. 0- 19·9633 70-3 0- 19-963369- X
Pollinatinn El:nlngy
Sp ir albound hard bac k 0-19· 963 319 -3 £32 .50 Paperb ack 0- 19·9 63318 -5 £22.50 106 _ Dia gnostic Molecular Pathology: Vo lume I
Hardie. D,C. (Ed ) 0 -19 -963306-1 0·19·963 305- 3
Sp ir albou nd hard bac k Pape rback
£35 .00
£30 ,00 19.50
Fukuda. \ 1. & KOIJdtd, A. IEd s) 0- 19-96 3372-X £32.50 0-19-96 3371-1 £22 ,50 124 . Human Geneti c Dise ase Analysi s (2ie ) Va d es . K.E. (Ed)
Sp ira lbo u nd ha rd back Paperback
0-19 -96309 8 -4
0- 19· 96309 9· 2 £25.00 Testa . x.r; & Mnlmeux . C;
0-19 -963327-4 £18 .50 Prot ein .:n gineer in g Rues. A. H.. Sternberg. \'-J .E, & others (Eds )
Paperba ck
122.
£45 .00 £25 .00 £2 5.00
Hamilton . H.I. & Hamiltun .
Sp ira lbo und har dbac k Paperb ack
109 .
£32 .50 £22.50 Rickwood . D. (Ed)
0-19-9632 08-1 0-19-963211-1 0-19-9630 99-2
Sp ir albound hardback Paper back
110
£5 0.00 £27.50
107.
Glyt:nbiology
0-19·963309 -6 0· 19-963 308 -8
Haemopuiesls IEds l
£3 2.50 £22 .50
0-19-9 63326-6 0·1 9 -963325-8
Pr ep arat ive Cen tr ifuga tio n
Lipid Anal vsls
£35 ,00 £2 5.00
Bolam J.P. {Ed]
~l~ uro a n a tn m y
Shicla IEd s') Spi ral bo und ha rdbac k Paperba ck
Spira lbo u nd har d back
Sp iralbound hardback Paperback
Hiflgin s ,
£35 .00 £25 ,00
[ovner :\. 1.. (Ed )
Sp ira lb ound hard back Paperback
Experimental
Spir albo uod ha rdback Paper bac k Paperb ack
Davison , ,\ .J. & Elhotr.
H_!l.1. (Eds) Sp ira lbo und hardback Paperb ack
&
0-19 -96 3292-8 0-19 -963 291 -X Elect rop hy siology Wa llis, lJ,1. (Ed ) Spi ralbound ha rdback 0-'9-9633 48-7 Paper back 0-19-963347-9 Biological Data A n a lysi s Frv. l.C, lEd) Sp ir albound ha rdb ack 0-1 9-963 340-1 Pap erback 0-19-963339-8 Sp ir albound ha rdba ck Pape r back
113 .
£30 .00 £18 .50
Sehgal.
0- 19· 96 3458-0 £32.50 0·19-963457-2 £22.50 Behavioura l Neuroscl en ce: Vnlu r ne I Sahgal. A . (Ed) Sp ira lb ou nd ha rd back 0- 19-963 368·1 £32 .50 Pap erback 0-19- 963 367 -3 £22.50
Molecular Vir ology
11.
£30,00 £19 .50
Spira lb ound ha rdback Paperb ack
128
115 .
0-19-96 3272 -3 0-19-96 3271- 5
Sp ira lbo und hard back Paperback
13 1.
"'
Baugh. P , (Ed)
Sp ir albou nd hard back Pape rback 132 ,
£ 32,50 £22,5 0
Hanu- s. IUl.
Spiratbound ha rdbac k Paperb ack
Roberts. G.c."-. lEd)
0 -19-963225 -1 0-19-96 3224 -3
Sp iralb ound ha rd back Pape rb ack
£3 0.00 £ 19 ,50
Ge ne Tra nscript io n
5,1. (Eds )
Spiralbc u nd ha rdback
102
Signal Tra ns d uctlun Spir atboun d ha rdback Pape rb ack
101 .
0-19 -9632 9 6-0 0-19- 963 295-2
Magl~p.
£30 .00 £18.50
\\"ill~m<m .
T_{Eds) £32 .50 £22.50 100 . Dia gno st ic Molecular Pa t ho logy : Volu me II : Cell and Ti ssu e Genotyping Herr ington. c.s. & McGee, I.On IEJs)
Ht st ocomp atibthty Test io g Dver , P . & Middleton. l J. IEds) Sp iralbo und hardback 0- 19 -9633 64 _9 Paper back 0 -19 -963 363-0
99 .
Letr.hman . 1).5 . (Ed)
0·19-963342-8 £30 .00 0-19- 963 341 -X £19 .50 Growth Facln rs Mckav, I. & Leigh. l. (Eds ) Sp iralb o und hardback - 0 -19 -963 360- 6 £30 .00 Pap erback 0- 19-9633 59-2 £19 .50 £32.50 £22 .50
Spiralb ound hardback Paper back
A.I. &
£32 .50 £22.50
C . (F.d ]
Spiratbo und ha rdbac k Paperback
Transcription Fac tors
Pro lei n Ta r geting
~ tilligan.
0- 19·9 63 206-5 0-19-9632 10-3
Spi ral boun d ha rdba ck 0-19·963 239- 1 Paperba ck 0-1 9· 963238-3 Neurona l Ce ll Li nes W ood , JS. (Ed ) Spira lbo u nd hardback 0-19 -963 346 -0 Pap er back 0-1 9-9 63345- 2
£30.00 £19 .50 £32 .50 £22.50