Current Topics in Developmental Biology, Volume 55 (Current Topics in Developmental Biology)

Current Topics in Developmental Biology

Volume 55

Series Editor Gerald P. Schatten Director, PITTSBURGH DEVELOPMENTAL CENTER Deputy Director, Magee-Women’s Research Institute Professor and Vice-Chair of Ob-Gyn-Reproductive Sci. & Cell Biol.-Physiology University of Pittsburgh School of Medicine Pittsburgh, PA 15213

Editorial Board Peter Gru¨ss Max-Planck-Institute of Biophysical Chemistry Go¨ttingen, Germany

Philip Ingham University of Sheffield, United Kingdom

Mary Lou King University of Miami, Florida

Story C. Landis National Institutes of Health National Institute of Neurological Disorders and Stroke Bethesda, Maryland

David R. McClay Duke University, Durham, North Carolina

Yoshitaka Nagahama National Institute for Basic Biology, Okazaki, Japan

Susan Strome Indiana University, Bloomington, Indiana

Virginia Walbot Stanford University, Palo Alto, California

Founding Editors A. A. Moscona Alberto Monroy

Current Topics in Developmental Biology Volume 55 Edited by

Gerald P. Schatten Director, PITTSBURGH DEVELOPMENTAL CENTER Deputy Director, Magee-Women’s Research Institute Professor and Vice-Chair of Ob-Gyn-Reproductive Sci. & Cell Biol.-Physiology University of Pittsburgh School of Medicine Pittsburgh, PA 15213

Amsterdam Boston Heidelberg London New York Oxford Paris San Diego San Francisco Singapore Sydney Tokyo

Cover Photo Credit: Zebrafish Nodal signaling mutant. Frontal view of Nodal signaling mutant at approximately 1 day postfertilization. See chapter 3 figure 1 WT for futher details.

This book is printed on acid-free paper. Copyright ß 2003, Elsevier (USA).

All Rights Reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the Publisher. The appearance of the code at the bottom of the first page of a chapter in this book indicates the Publisher’s consent that copies of the chapter may be made for personal or internal use of specific clients. This consent is given on the condition, however, that the copier pay the stated per copy fee through the Copyright Clearance Center, Inc. (222 Rosewood Drive, Danvers, Massachusetts 01923), for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective works, or for resale. Copy fees for pre-2003 chapters are as shown on the title pages. If no fee code appears on the title page, the copy fee is the same as for current chapters. 0070-2153/2003 $35.00 Permissionions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone: (+44) 1865 843830, fax: (+44) 1865 853333, e-mail: [email protected]. You may also complete your request on-line via the Elsevier homepage (http://elsevier.com), by selecting ‘‘Customer Support’’ and then ‘‘Obtaining Permissions.’’

Academic Press An Elsevier Imprint. 525 B Street, Suite 1900, San Diego, California 92101-4495, USA http://www.academicpress.com

Academic Press 84 Theobald’s Road, London WC1X 8RR, UK http://www.academicpress.com International Standard Book Number: 0-12-153155-4 PRINTED IN THE UNITED STATES OF AMERICA 03 04 05 06 07 08 9 8 7 6 5 4 3 2 1

Contents

Contributors

vii

1 The Dynamics of Chromosome Replication in Yeast Isabelle A. Lucas and M. K. Raghuraman I. II. III. IV. V. VI. VII.

Introduction 1 Assays for Origin Activity 2 Organization of the Genome for Replication 11 Assembly and Activation of the Initiation Complex The Temporal Program of Origin Activation 38 Monitoring the Replication Program 49 Concluding Thoughts 59 References 59

17

2 Micromechanical Studies of Mitotic Chromosomes M. G. Poirier and John F. Marko I. II. III. IV. V. VI.

Introduction 76 Architecture and Components of Eukaryote Chromosomes Stretching Elasticity of Chromosomes 94 Bending Elasticity of Chromosomes 104 Viscoelasticity of Chromosomes 112 Combined Biochemical–Micromechanical Study of Mitotic Chromosomes 116 VII. Conclusion 125 References 133

79

v

vi

Contents

3 Patterning of the Zebrafish Embyro by Nodal Signals Jennifer O. Liang and Amy L. Rubinstein I. II. III. IV. V. VI. VII.

Introduction 143 Zebrafish Nodal Signals 144 Nodal Signaling Pathway 146 Patterning the Mesoderm and Endoderm 153 Role of Nodal in Patterning the Ventral Nervous System Patterning the Left–Right Axis 162 Future Directions 165 References 165

158

4 Folding Chromosomes in Bacteria: Examining the Role of Csp Proteins and Other Small Nucleic Acid-Binding Proteins Nancy Trun and Danielle Johnston I. Introduction 173 II. Small Nucleic Acid-Binding Proteins Implicated in Chromosome Folding in Escherichia coli 175 III. Csp Proteins in Escherichia coli 180 IV. Relationships among Escherichia coli Csp Proteins 189 V. Csp Proteins of Bacillus Subtilis 190 VI. The Crystal Structures of CspA and Related Proteins 191 VII. Distribution of Small DNA-Binding Proteins in Archaea and Bacteria 197 VIII. Conclusions 198 References 200

Index 203 Contents of Previous Volumes

215

Contributors Numbers in parentheses indicate the pages on which the authors’ contributions begin.

Danielle Johnston (173), Department of Biological Sciences, Duquesne University, Pittsburgh, Pennsylvania 15219 Jennifer O. Liang (143), Biology Department, Case Western Reserve University, Cleveland, Ohio 44106 Isabelle A. Lucas (1), Department of Genome Sciences, University of Washington, Seattle, Washington 98195 John F. Marko (75), Department of Physics and Bioengineering, University of Illinois at Chicago, Chicago, Illinois 60607 M. G. Poirier (75), Department of Physics, University of Illinois and Chicago, Chicago, Illinois 60607 Mokur K. Raghuraman (1), Department of Genome Sciences, University of Washington, Seattle, Washington 98195 Amy L. Rubinstein (143), Zygogen, Atlanta, Georgia 30303 Nancy Trun (173), Department of Biological University, Pittsburgh, Pennsylvania 15219

Sciences,

Duquesne

vii

ThisPageIntentionallyLeftBlank

1

____________________________________________________________________________

The Dynamics of Chromosome Replication in Yeast Isabelle A. Lucas and M. K. Raghuraman Department of Genome Sciences University of Washington Seattle, Washington 98195

I. Introduction II. Assays for Origin Activity A. The Autonomously Replicating Sequence Assay B. Two-Dimensional Agarose Gel Electrophoresis C. Alkaline Gel Electrophoresis Assay for Nascent DNA Strands D. Density Transfer Method to Determine the Time of Replication E. Single Molecule Analysis by Molecular Combing F. Assaying Replication on a Genomic Scale G. Strengths and Limitations of the Assays III. Organization of the Genome for Replication A. Origin Structure B. Origin Spacing and Location IV. Assembly and Activation of the Initiation Complex A. Formation of the Prereplicative Complex B. Origin Firing C. Preventing Reinitiation D. What Determines Origin Choice and EYciency? E. Fork Migration and Termination V. The Temporal Program of Origin Activation A. Setting Up and Reading the Temporal Program B. What Is the Physiological Relevance of the Temporal Program? VI. Monitoring the Replication Program A. The S Phase Response to DNA Damage and Hydroxyurea B. Detecting Ongoing, Uninterrupted Replication VII. Concluding Thoughts References

I. Introduction Chromosomal DNA replication is a task of immense complexity. No sooner has a cell divided than it begins preparation for the next round of DNA replication. Origins of replication, sites that will direct initiation of DNA synthesis at roughly 30- to 100-kb intervals along the chromosome, Current Topics in Developmental Biology, Vol. 55 Copyright 2003, Elsevier (USA). All rights reserved. 0070-2153/03 $35.00

1

2

Lucas and Raghuraman

begin to recruit the proteins that will be required for the initial steps in DNA synthesis. By the time the cells reach START, a critical decision point late in G1 phase of the cell cycle, origins will have completed the first phase of this assembly process and become licensed for initiation of DNA synthesis. DNA synthesis begins in S phase with the unwinding of DNA by replication initiation complexes and the assembly of replication forks. As replication proceeds, the forks traverse the genome, unraveling the structure of the chromosomes as they advance and leaving repackaged and duplicated DNA in their wake. Every base pair must be faithfully copied, yet no portion of the genome may be copied more than once per cell cycle. This task is made more diYcult by the temporal program of origin activation that eukaryotes follow. Origins are activated in a reproducible, sequential pattern through S phase and then must be shut oV for the remainder of the S phase even as other origins become active. How cells solve this problem of preventing reinitiation at origins that have already initiated has become clear in the last few years, but how the temporal program is laid down in the first place remains unanswered. Even more obscure is the question of why cells bother with this temporal program at all. Work on the budding yeast Saccharomyces cerevisiae has led the way in our exploration of how replication is choreographed. As with many other aspects of biology, the basic framework underlying the process of replication is conserved between yeast and higher eukaryotes, including humans. From the recent explosion of work on yeast and other model systems, we have begun to have a deeper appreciation of the dynamic aspects of replication—how the control of replication initiation is tied to phases of the cell cycle, how the activation of diVerent origins within the genome are coordinated, and how checkpoint surveillance systems monitor and ensure the orderly progress and completion of S phase. This chapter reviews our current understanding of how the program of replication is laid down and carried out in S. cerevisiae, with particular emphasis on the dynamic aspects of replication and the open questions still remaining.

II. Assays for Origin Activity With any endeavor in science, the sophistication of the questions that can be asked is limited by the techniques available, and replication is no exception. The interpretation of experimental results also hinges on a clear understanding of the methods used and their limitations. We shall begin, therefore, with a brief overview of the methods that are used most commonly in studying replication.

1. The Dynamics of Chromosome Replication in Yeast

3

A. The Autonomously Replicating Sequence Assay Replication origins had been observed previously in yeast by methods such as electron microscopy and fiber autoradiography (e.g., [1,2]), but identification of specific sequences that can function as replication origins first came from a genetic assay that required stable maintenance of a plasmid [3–5]. In this assay, DNA fragments are tested for their ability to confer autonomous replication on a plasmid that otherwise cannot be maintained as an extrachromosomal element in yeast. Sequences that support maintenance of the plasmid presumably do so by acting as origins of replication. Potential origins of replication can thus be identified in a very simple transformation assay. The autonomously replicating sequences (ARS elements) so identified are small, on the order of 100–200 bp [6]. They have been shown to function as origins of replication on plasmids in vivo (see later; [7,8]) and to various extents as origins on chromosomes. For example, ARS301 is capable of supporting maintenance of a plasmid but is considered to be a ‘‘silent’’ origin in its native context on chromosome III, i.e., it shows no detectable origin activity in the chromosome [6,9,10]. Thus, the ARS assay allows identification of potential origins of replication; other means (see later) must be used to test for origin activity on the chromosome.

B. Two-Dimensional Agarose Gel Electrophoresis Developed in 1987, the two-dimensional (2-D) agarose gel electrophoresis technique has dramatically changed how we look at replication [7]. Distinct patterns of gel migration are observed in this technique for diVerent forms of replication intermediates (Fig. 1a). These patterns allow recognition of a restriction fragment that contains an active origin (bubble-shaped replication intermediates), a fork passing through (Y-shaped intermediates), or two forks converging (double Y intermediates) (Fig. 1a). The eYciency of an origin is defined as the percentage of cell cycles in which it initiates replication or ‘‘fires.’’ The classical 2-D gel method described earlier only permits qualitative estimates of origin eYciency. A modification of this technique, fork direction analysis [11,12], allows the quantitation of origin eYciency by revealing the proportion of leftward versus rightward moving forks in a restriction fragment. By assaying fork movement on either side of an origin, one can measure the fraction of cells in which forks are moving outward from the origin (initiation event) compared to those in which forks enter the region (no initiation). Using this method, origin firing can be ranked from highly eYcient (e.g., ARS607 fires in 85% of cells) to ineYcient (e.g., ARS605 fires in <27% of cells) to dormant or inactive (e.g. ARS301) [9]; also see Newlon and Theis [6] and Friedman [13].

4

Lucas and Raghuraman i. Bubble arc 1st dimension

ii. Y arc

iii. Double Y arc

iv. Bubble arc to Y arc transition

2nd dimension

(a)

Initiation

Passive replication

Termination

Initiation— off-center origin

(b) Origin High Mr DNA (Parental strands)

Alkaline gel electrophoresis

Figure 1

(Continued).

Replication Intermediates (~1−10 kb)

1. The Dynamics of Chromosome Replication in Yeast

Figure 1

(Continued).

5

6

Lucas and Raghuraman

In a diVerent form of 2-D gel assay [8], the second dimension of gel electrophoresis is performed under alkaline, denaturing conditions to detect nascent DNA strands of increasing lengths, emanating from a single site. This technique allows the identification of origins by mapping regions containing diverging replication forks [8,14].

C. Alkaline Gel Electrophoresis Assay for Nascent DNA Strands Another technique used to detect origin activity has become popular. This method is used not to map origins but to assay origin activity. Here, the

Figure 1 Assays for replication. (a) The 2-D gel electrophoresis method. Genomic DNA cut with a restriction enzyme is separated mainly by mass in the first dimension of electrophoresis. In the second dimension, the shape of the molecule strongly affects the migration rate: branched molecules migrate more slowly than linear molecules of the same mass. The gel is then blotted and probed for the fragment of interest. Patterns of hybridization are seen for (i) bubble forms of replication intermediates; (ii) simple Y molecules; (iii) double Y molecules; and (iv) bubble-to-y transition from molecules containing an asymmetrically placed origin of replication. The replication intermediates giving rise to the 2-D gel patterns are depicted below each gel pattern. See Friedman and Brewer [11] for additional details. (b) Detection of nascent DNA strands by alkaline gel electrophoresis. Genomic DNA collected from cells released into S phase in the presence of hydroxyurea is subjected to electrophoresis under alkaline conditions. Parental, high-molecular weight strands (red) are separated from the lower molecular weight replication intermediates (black). Replication intermediates at specific sequences can then be detected by Southern blotting. (See Color Insert.) (c) Density transfer analysis of replication time. Cells grown in isotopically dense culture medium are arrested in late G1 and released into S phase in the presence of isotopically light culture medium. Samples are collected through S phase. Replicated (HL) DNA from each sample (digested with a restriction enzyme) can be separated from unreplicated (HH) DNA by centrifugation in cesium chloride solution. The location of any particular fragment within the cesium chloride density gradient is determined by slot blot analysis of gradient fractions. The plot below shows replication curves for three hypothetical fragments as determined by density transfer analysis. Replication curves are shown for fragments replicating early in S phase, in mid-S, and in late S phase. The time of half-maximum replication (trep) is marked for the late-replicating fragment. See van Brabant and Raghuraman [12] for additional details. (d) Replication profile for a hypothetical chromosome. Peaks indicate the locations of origins, and valleys represent termination zones. The shorter a peak, the later in S phase the corresponding origin fires. Peaks marked Early, Mid, and Late represent origins firing in early, mid, and late S phase, respectively. (e) A scheme for genome-wide analysis of replication by molecular combing. Chromosomal DNA molecules (black) from S phase cells labeled with BrdU are stretched and aligned on a microscope slide ([17]; see text for details). Sites of BrdU incorporation (red) would reveal sites of DNA synthesis in relation to marker sites detected by in situ hybridization with known probes (green and blue). Efficient (filled arrowhead) and inefficient origins (open arrowhead) could be detected from examination of a population of molecules (shown as being in register for convenience). The pattern of labeling that might be seen for the hypothetical chromosome is shown in with a mid- to late-S-phase sample. (See Color Insert.)

1. The Dynamics of Chromosome Replication in Yeast

7

presence or absence of nascent DNA strands arising from DNA synthesis is detected by electrophoresis in alkaline gels [10,15]. To enrich for replication intermediates, the DNA is collected from cells that have been released into S phase in the presence of the drug hydroxyurea, which blocks ribonucleotide reductase [16]. Under these conditions, DNA synthesis is slowed dramatically. Nascent strands persist longer, and origins that normally fire late in S phase are thought to be inhibited strongly [15]. The presence of a smear of lower molecular weight DNA in the gel (1–10 kb; Fig. 1b) is interpreted as evidence in favor of DNA synthesis at the probed location [10,15]. This method is relatively quick and easy, and can be informative when the appropriate controls are done. However, some caution is needed when utilizing this assay, in part because the eVects of hydroxyurea are not fully understood.

D. Density Transfer Method to Determine the Time of Replication The time of replication of specific restriction fragments can be determined by a variant of the Meselson–Stahl density transfer experiment [12] (Fig. 1c). Cells grown in isotopically dense (13C, 15N) culture medium are allowed to enter a synchronous S phase in the presence of isotopically light (12C, 14 N) culture medium. The kinetics of conversion of a restriction fragment from fully dense (heavy–heavy, HH) to hybrid density (heavy–light, HL) reflects the time of its replication (Fig. 1c). Comparison of the replication kinetics of diVerent fragments is facilitated by taking the time of half-maximal replication (trep) for each fragment.

E. Single Molecule Analysis by Molecular Combing A relatively new method that yields information at the single molecule level is called molecular combing. Here, a slide is dipped in a solution of chromosomal DNA; the free ends of the DNA molecules bind to the slide surface, and as the slide is withdrawn from the solution, the molecules are stretched and aligned along the surface of the slide. Individual molecules can then be visualized by various fluorescence microscopy techniques [17]. For example, replicating DNA can be identified by pulse labeling the cells with bromodeoxyuridine (Brdu), which can be detected on the slide by immunochemical means. Alternatively, or in addition, specific chromosomal sequences can be identified by hybridization with fluorescently labeled probes [17,18]. The advantage of this method is that visualization of single DNA molecules allows examination of events in single cells, whereas the other methods described provide information about the average events in a

8

Lucas and Raghuraman

population of cells. For example, one can ask if initiation at an origin is always associated with initiation at a diVerent origin on the same molecule. One could also assay the eYciency of particular origins. Because hundreds of molecules can be examined on a single slide, this method can give statistically meaningful results.

F. Assaying Replication on a Genomic Scale The availability of the yeast genome sequence is revolutionizing how we study replication. Many of the techniques described earlier have been modified and extended for whole genome studies. As with any young field, the techniques are still in flux and will undoubtedly be revised and refined in the years ahead. Nevertheless, it is worth describing briefly some strategies that have been used successfully thus far. As discussed later, the diVerent assays have yielded largely similar results, but with some diVerences. 1. Chromosomal Replication Profiles Based on Density Transfer Analysis An extension of the Meselson–Stahl density transfer experiment described earlier can be used to determine replication times on a genome-wide scale. Cells are transferred from isotopically dense to light culture medium as in Section IID. However, instead of assaying for replication of specific fragments, the unreplicated (HH) and replicated (HL) DNA pools obtained from these cells are separately labeled and hybridized to genomic DNA microarrays. In essence, the question being asked is how much DNA is present in the replicated vs unreplicated pool for each locus in the genome. By compiling this information over several times in S phase, one can deduce the dynamics of replication for each locus that is represented on the microarray, and thereby derive a replication profile of the genome—the order and time of replication for each segment along each chromosome (Fig. 1d; [19]). The replication profile reveals the locations of origins and termini, and the order of firing of diVerent origins. Finally, these profiles yield information about the direction and rate of replication fork migration. 2. Chromosomal Replication Profiles Based on Copy Number In principle, the progressive increase in the copy number of various loci as they are replicated in S phase (one copy at the start of S phase, two after it has replicated) can be measured relative to a nonreplicating standard. The method hinges on the ability to detect small changes in copy number—the copy number for any specific segment will be between one

1. The Dynamics of Chromosome Replication in Yeast

9

and two at intermediate times in S phase. This approach has the advantage of being less expensive and applicable to a wider range of experimental conditions, as labeling with dense isotopes is not a consideration. Originally used with Escherichia coli [20], it has been applied to budding yeast [21], yielding replication profiles that generally match those published previously using the density transfer method [19]. Other microarray-based methods (such as hybridization of nascent DNA strands obtained from timed S phase samples) can be envisioned as well and will undoubtedly be tried in the years ahead. 4. Genomic Locations of Replication Initiation Complexes A completely diVerent strategy, which complements those outlined earlier, is to identify locations of presumptive origins by looking for DNA sequences that are bound by the proteins that can form replication initiation complexes [22,23]. In this method, protein–DNA complexes are fixed by cross-linking with formaldehyde. After shearing of the chromatin, the proteins of interest are immunoprecipitated, and the bound DNA is recovered by reversing the cross-links. This DNA pool represents fragments enriched for binding sites for the proteins that were immunoprecipitated and can be identified by labeling and hybridization to DNA microarrays. For this reason, this method is often called a ‘‘ChIP-to-chip’’ method (chromatin immunoprecipitation followed by DNA chip or microarray hybridization). Sequences identified by this method will include active as well as silent origins. In addition, sites where replication proteins are bound for other possible functions (e.g., transcriptional silencing [24–26]) would also be identified.

G. Strengths and Limitations of the Assays All of these methods have their unique advantages and limitations. The ARS assay, based on genetic selection, is immensely powerful and has the additional benefit of allowing some quantitation of ARS strength. Transformation eYciency and the more rigorous plasmid loss assays give an indication of how well an ARS functions in any given yeast strain (high loss rates are indicative of poor ARS function). Such assays not only measure the intrinsic ability of the ARS to support plasmid maintenance (e.g., [27]), they also allow judgment of the eVect of mutations that might aVect origin initiation (e.g., [28]). However, this assay does not give any information on how the potential origin actually functions in the chromosome. The 2-D gel technique does allow visualization of origin activity on the chromosome, but is limited to restriction fragments in the 3- to 5-kb range,

10

Lucas and Raghuraman

although it has been applied to fragments as small as 1 kb and as large as 13–15 kb. The density transfer method has been a mainstay of replication studies since the mid-1980s. When used in combination with the 2-D gel technique, it reveals the time of firing of origins. However, one factor to keep in mind is that these experiments show the average replication time for a fragment in the population of cells. If an ineYcient origin fires early in S phase in a subpopulation of cells and is replicated passively later in S phase by the remaining cells, the density transfer analysis will give an intermediate value for the time of replication. The genomic replication profiles give a wonderful overview of replication. However, because a mathematical smoothing algorithm is applied to aid in the detection of peaks and valleys, the resolution of this technique is on the order of a few kilobases. Furthermore, the use of cesium chloride gradient centrifugation to resolve HH and HL DNA can result in under- or overrepresentation of fragments with extremes of G+C richness or poorness in HL compared to HH DNA fractions. The ChIP-to-chip method has the advantage that potential origins can be identified at much higher resolutions than the other genomic methods. No information is obtained about whether the sequences actually function as origins. For example, as discussed later, some sequences appear to form the necessary prereplicative complexes but fail to initiate DNA synthesis. Such silent or ineYcient origins or even nonorigin sequences would be detected by this method but not by the previous two methods. Molecular combing has not been used thus far in genome-wide replication analyses. However, it could in principle be used to identify origins on a genomic scale. For example, one could combine BrdU labeling of nascent DNA with in situ hybridization of marker probes to the combed DNA molecules (Fig. 1e). The marker probes could be color coded or give predictable patterns of hybridization and would provide landmarks for the identification of specific chromosomal locations. The sites of DNA synthesis (BrdU labeled) could be determined in relation to these landmarks. This method has been used to detect the pattern of origin initiations in the ribosomal RNA gene locus (rDNA) [29]. Although laborious to apply on a genomic scale, the power of this approach is that it would allow direct examination of a set of individual molecules in contrast to other methods that build aggregate profiles based on a population of cells. Methods such as fork direction gel analysis can give estimates of origin eYciency but molecular combing could allow direct measurement of the proportion of molecules in which an origin fires or does not fire (Fig. 1e). Molecular combing could also address questions not readily amenable to analysis by other techniques. For example, combing of timed S phase samples could reveal whether there is cell-to-cell variation in the order of origin initiations.

1. The Dynamics of Chromosome Replication in Yeast

11

III. Organization of the Genome for Replication A. Origin Structure The replicon model of prokaryotic DNA replication [30] proposed that two elements are necessary for regulating DNA replication. One element, the replicator, corresponds to a specific DNA sequence that is acted upon to cause initiation. The second element, the initiator, interacts in trans with the replicator to trigger the onset of DNA synthesis. The basic features of the model still hold. In bacteria, the replicator corresponds to the origin of replication, and the initiator is the set of proteins needed to act on the origin and initiate DNA synthesis. With some modifications, this model can be extended to eukaryotes as well. While defined replicators (i.e., specific sequences) have been hard to identify in higher eukaryotes, replication in budding yeast more closely resembles that seen in prokaryotes. From the limited information available [31], it appears that the site of initiation is coincident with the genetic replicator element, although in principle initiation need not occur at the replicator itself. The ARS assay has facilitated analysis of the sequence organization of replication origins in yeast. Early comparisons of the first identified ARSs revealed a short, <20-bp stretch of similarity that was essential for ARS function, flanked by short sequences important for ARS activity [32–35]. The first detailed and systematic mutagenesis of an ARS was performed on ARS1 [36]. This study created a set of linker-substitution mutations that each altered a 5- to 8-bp block of sequence all the way across the 190-bp ARS1 fragment. This analysis, in combination with similar mutational analyses of a few other ARSs, has revealed the following picture of a multipartite ARS structure (Fig. 2a). All ARSs studied to date have a core, essential element called the ‘‘A’’ element, the primary constituent of which is the ARS consensus sequence (ACS). Two or more flanking ‘‘B’’ elements lie 30 to the T-rich strand of the ACS. In addition, some ARSs have a ‘‘C’’ element that is believed to improve ARS function [37]. 1. The ACS From a comparison of several ARS elements, the 11-bp sequence was identified as being the ACS [6]. Initially, it appeared that all ARSs must have at least a 10/11 match to this consensus, but subsequently, examples of ARSs with a 9/11 match were found [38,39]. The conclusion that a perfect match to the ACS is not needed is further supported by mutagenesis—substitutions within the ACS are indeed tolerated, although some positions within the ACS seem to be invariant [6,40]. As described A TTTAC AA TTTA T T G T

12

Lucas and Raghuraman

Figure 2 Sequence organization of ARS elements. (a) ARS structure. Sequence elements contributing to ARS function are shown for ARS1 [36], ARS305 [56], ARS307 [269,270], ARS121 [38,271], and ARS1501 [58]. ARSs are aligned by the ACS (vertical tick mark below the box marked ‘‘A’’) within the A element. B1* indicates B1-like elements as determined by location; ORC footprinting has not been done at these ARSs. REN is a replication enhancer that strongly stimulates ARS function. (b) Sequence logo showing information content around the ACS (gray-shaded area). Twenty ARSs listed by Theis and Newlon [42] and aligned at the ACS were used in this analysis. The total height at each position represents the degree of sequence conservation at that position; the height of an individual letter represents the frequency of that base at that location. The sequence logo was generated using WebLogo on the internet at . For further information on sequence logos, see Schneider and Stephens [43] and on the internet at .

1. The Dynamics of Chromosome Replication in Yeast

13

later, the sequence requirement shows a fair degree of flexibility and is influenced by the context within which the ACS occurs. For example, a mutation in one of the invariant positions in the ARS307 ACS eliminates origin activity at its native location on chromosome III, but the same 200-bp fragment containing the mutated ACS functions as an ARS when moved to a diVerent location [41]. Although the 11-bp sequence is still considered to be the core, essential sequence, with more ARSs being identified and sequenced over the years, a more extended 17-bp consensus has emerged [42]. Nevertheless, the 11-bp sequence continues to be used as a hallmark of ARSs. One reason the 11-bp ACS still has such widespread currency is that while it is fairly safe to use a 9/11 match as a requirement, it is not so clear how much variation is allowed with the expanded 17-bp consensus. Even the expanded ACS does not accurately predict ARS locations by itself, as there are >2000 15/17 or better matches to that sequence in the genome. However, as computational methods improve, the additional information aVorded by the expanded consensus will likely prove useful in predicting ARS locations. For example, there is a trend underway to represent protein-binding sites not as consensus sequences but as ‘‘sequence logos’’ that capture the information content of a set of binding sites [43,44] (Fig. 2b). This approach allows scanning of genomic sequences for matches to the sequence logo and assigns probabilities of matching rather than an all-or-nothing score for matches to a consensus sequence [45]. It will be interesting to see if such probabilities of matches to sequence elements can be correlated with ARS eYciency. Increased scrutiny uncovers diversity, and so it is with ARSs. We are encountering instances of ARSs that do not fit the simple picture painted earlier. One such noncanonical type of ARS is being called a compound ARS, of which ARS101 and ARS310 are examples. Mutational analysis of these ARSs has shown that they contain two and three matches, respectively, to the 11-bp ACS, which must all be mutated to abolish ARS activity in the respective fragments [46]. Upon reexamination of 22 previously described ‘‘standard’’ origins in light of these findings, it appears that 6 of those 22 also fit the description of compound origins [46]. A complication of a diVerent sort is that an imperfect ACS match may be used by the cell as the essential element even if a perfect match is available nearby [47]. A possible explanation for this preferred use of a poorer match is that such choices are made because the imperfect match is in a better location vis a vis the other ARS elements (B1, B2, etc.) than is the perfect match. What does the ACS do? A variety of in vitro and in vivo analyses (see Section IVA) have led to the conclusion that the ACS, along with the nearest B element, nucleates assembly of the replication initiation complex by serving as a binding site for a six subunit protein complex called the origin recognition complex (ORC). This complex, once bound to origin DNA, then

14

Lucas and Raghuraman

serves as a landing pad for the stepwise assembly of other proteins required for origin firing. 2. The B Elements In addition to the A element, linker mutagenesis of ARS1 uncovered specific sequence elements termed B elements, 30 to the T-rich strand of the ACS, that contribute to ARS function—mutagenesis of those positions reduced or eliminated ARS activity (monitored as plasmid transformation eYciency) [36]. Three such elements were identified. Mutation of any single B element reduced ARS activity, whereas pairwise mutation of any two B elements almost eliminated ARS activity. The diVerent B elements could not functionally substitute for each other, e.g., replacing B2 and B3 with a copy of B1 at each of those locations did not give a functional ARS. These four elements (A and B1–B3) together functioned as an ARS when embedded with the native spacing relative to each other but in unrelated DNA, indicating that these elements together are suYcient to create a functional ARS. Once ORC subunits were purified and an in vitro ORC-binding assay was developed [48], it became possible to determine the binding site for ORC by looking for protection of the DNA from nuclease digestion, or for interference of binding by chemical modification of specific bases. Such assays showed that ORC has a bipartite binding site, with protein–DNA contacts made in the ACS and the B1 element [49,50]. The B2 element is somewhat nebulous in function. It is almost always associated with a duplex unwinding element (DUE), a region of reduced helical stability that is presumed to promote unwinding of the DNA for the initiation of DNA synthesis [51]. Mutations in this region that impair ARS function can be rescued by elevating the temperature [52], supporting this view. Moreover, the polymerization start site has been mapped to a location between B1 and B2 [53], consistent with the idea that the initial unwinding of DNA occurs at the B2 element. However, duplex unwinding is not the sole function of the B2 element, as mutations in B2 that change the sequence without changing the energy of unwinding nevertheless impair ARS activity [54]. Furthermore, changes in the B2 region appear to impair binding of a diVerent six subunit protein complex, the Mcm (minichromosome maintenance) complex, required for origin activity [54,55]. The Mcm complex may be the replicative helicase (see Section IV,A), so the two apparent functions of B2 may be related—recruiting the Mcm complex and then facilitating unwinding of the DNA perhaps mediated by the complex. The B3 element in ARS1 is a binding site for the transcription factor Abf1p [36]. While this element does contribute to ARS1 activity, it can be

1. The Dynamics of Chromosome Replication in Yeast

15

replaced by binding sites for other transcriptional activators such as Gal4p or Rap1p. 3. Deciphering the ARS Code Despite this detailed dissection of ARS structure and the apparently simple requirements for ARS function, predicting ARS locations based on sequence remains an elusive goal. In part, this diYculty arises from nonconservation of the nucleotide sequence between B elements from diVerent ARSs; i.e., mutational analysis may reveal B elements at similar locations in diVerent ARSs, but little or no similarity can be detected at the sequence level. Despite this lack of sequence similarity (with the exception of very minimal conservation, such as the presence of a conserved TT dinucleotide in B1 [56]), B elements from diVerent ARSs can functionally substitute for each other to an extent [57]. However, this interoperability is not universal, e.g., the B3 element of ARS1 may inhibit rather than promote activity at ARS305 [57]. Furthermore, not all ARSs have three B elements: ARS307 just has two and ARS121 has one B1-like and two B3-like elements (Abf1p-binding sites, originally called OBF sites; [38]) that are 50 to the T-rich strand of the ACS (Fig. 2a). The recently discovered ARS1501 has a ‘‘replication enhancer’’ 50 to the T-rich strand that stimulates ARS function [58]. Deciphering the ARS code will clearly be of considerable interest in the future; with the use of genomic approaches mentioned previously, the database of ARSs is seeing a sharp increase and should help discern subtle patterns that may have eluded detection when working with a small set of ARSs. What is clear is that the bipartite ORC-binding site (ACS and B1) is the most important, common component; the other elements contribute to various extents and may make quantitatively diVerent contributions to diVerent ARSs [56]. These contributions may include easing of the initial melting of the DNA at the origin (DUE) or facilitating a favorable chromatin structure (e.g., one that allows access to replication factors) either directly or via recruitment of other factors (Abf1p, Rap1p, etc.). Origins in higher eukaryotes are not as small and well defined as in budding yeast. Even in the fission yeast Schizosaccharomyces pombe, origins are larger. They have multiple AT-rich sequences that are recognized by SpOrc4p, but no defined consensus or essential sequences have been identified so far [59,60]. The picture becomes even murkier in higher eukaryotes. For instance, Xenopus egg extract systems, which are used frequently for the in vitro assembly of initiation complexes, will replicate any DNA that is introduced without specificity as to the sites of initiation [61,62]. Likewise, mammalian cells seem capable of replicating any introduced DNA eYciently provided it is at least 12 kb in size [63]. Nevertheless, there are several

16

Lucas and Raghuraman

reports of initiation from specific regions of a few kilobases in diVerentiated cells in culture (see Zannis-Hadjopoulos and Price [64] for a review).

B. Origin Spacing and Location Early studies with fiber autoradiography and electron microscopy led to a rough estimate of 200–400 origins in the 14 Mb yeast genome [1,65]. This estimate has been borne out by recent, genome-wide analyses of replication [21,23,66]. However, even the modern analyses have not provided a definitive answer to the question of how many origins there are. The three diVerent genomic approaches have predicted slightly diVerent numbers of origins, ranging from 260 [21] to 332 [19] to 429 [23] (Sections IIF1, IIF2, and IIF4, respectively). The explanation for these diVerences probably lies in some combination of the following factors.  The diVerent techniques used have inherent strengths and weaknesses (see Section II), resulting in false positives and negatives in the scoring of putative origins. For example, the ChIP-to-chip method detects potential origins (Section II,F,4) and does not assay origin activity directly.  All three methods require some mathematical manipulation of data, including some decision making as to what to score as an origin. For example, with the density transfer-based genomic analysis of replication [19], slope changes from positive to negative in the replication profiles (see Fig. 1d) were scored as peaks and therefore as being putative origins. This approach is sensible, but for some features in the profiles, small (possible nonreproducible) diVerences in replication times could make a diVerence in origin predictions.  There may not be one universally correct value for the number of origins. The number of origins may vary with the strain or the growth conditions used. IneYcient origins further complicate the prediction of origin numbers. Molecular combing (Section II,E, Fig. 1e) could be particularly useful in addressing the question of origin eYciency. Because statistically large numbers of molecules can be examined for origin activity at specific locations, this method has the potential to give detailed information on origin eYciency, especially if applied to the whole genome. Several observations suggest that replication origins are excluded from transcribed regions.  Initiations in the early Xenopus embryo can occur anywhere in the rDNA, but once transcription begins at the midblastula transition, initiation is restricted to the nontranscribed regions [67].

1. The Dynamics of Chromosome Replication in Yeast

17

 In budding yeast, ARSs largely map to intergenic regions [23,68]. ARSs that do fall in genic regions may be ineYcient or inactive as origins in the chromosome [46].  Transcription across an origin can erase origin activity [69].  Even in E. coli, genes are organized so that the direction of transcription and that of replication are colinear [70], consistent with the idea that transcription can interfere with replication. As proposed previously [68], perhaps the streamlining of the budding yeast genome to eliminate intergenic sequences has driven the evolution of sequence-specific origins. Higher eukaryotes, in contrast, have large intergenic regions and could therefore aVord to relax the sequence specificity of binding of the replication initiation factors. Even with the relatively tight constraints on origin locations in yeast, the requirements to establish ARS activity are not diYcult to meet [71,72], suggesting that the strategy may be to create enough potential origins such that even if each one by itself does not have a high probability of firing, the overall probability of getting no firing within any substantial stretch of the chromosome is infinitesimally low.

IV. Assembly and Activation of the Initiation Complex Much of what we know about the assembly of initiation complexes comes from a combination of nuclease protection assays (footprinting) to reveal protein binding, chromatin immunoprecipation and coimmunoprecipitation to reveal the identity of proteins bound to the DNA and associated with each other, respectively, and genetics to determine the gene requirements for the establishment of origin activity. These assays, when performed on cells growing synchronously, also help define the order of events in relation to the cell cycle. One caution to keep in mind while interpreting these results is that protein–protein or protein–DNA associations detected by immunochemical means may represent direct or indirect interactions. Despite the progress that has been made in understanding the sequence of events in initiation, that we are still far from a complete picture is underscored by the regularity with which newly discovered proteins are found to be essential for the proper completion of the process. Footprinting studies have shown that origins cycle between two chromatin states in vivo [73,74]. One of these patterns of nuclease protection resembles that seen with ORC–origin complexes assembled in vitro, whereas the other is a more extended pattern. The smaller pattern therefore is believed to result from ORC bound to origins—the postreplicative complex (postRC). As the cells complete mitosis and enter G1, the post-RC recruits

18

Lucas and Raghuraman

the Mcm complex and becomes the prereplicative complex (pre-RC), which is licensed for replication in that cell cycle. Formation of the pre-RC is manifested as an expanded protection pattern. As the cells go through S phase, the Mcm complex (but not ORC) is released from each origin, presumably as it replicates, and the protection pattern reverts to the post-RC pattern [73,74]. The molecular events underlying these chromatin states have been analyzed by a combination of genetics and biochemistry; the major steps are outlined here and discussed in greater detail later. Origins are bound constitutively by ORC [75]. These post-RCs are converted to pre-RCs in G1 upon the addition of Mcm complexes. The next step is the maturation of the preRC to the preinitiation complex (pre-IC), which has all the components needed to open the DNA in preparation for the initiation of DNA synthesis. The pre-RC to pre-IC transition begins at the G1/S boundary and seems to parallel the order of origin firing, occurring early in this window at origins that fire early in S phase and later at origins that fire later in S phase. Finally, unwinding of the DNA and the addition of components of the future replication fork mark the transition of the pre-IC to the initiation complex, which can begin polymerization of nucleotides. Note, however, that although it is convenient to visualize these events as a linear progression, the precise order is uncertain for some of the events.

A. Formation of the Prereplicative Complex Assembly of the pre-RC may occur at all ARSs regardless of whether they are actually used in the ensuing S phase. However, the generality of this conclusion has not been tested extensively. It is based primarily on the footprinting analysis of one particular silent origin, ARS301 [10,76]. It could well be that ARS301 is something of an exception. ORC is known to have a role in silencing [24–26], and ARS301 is located in one of the most strongly silenced loci in the yeast genome, raising the possibility that assembly of a pre-RC at that site is a side eVect of the role of ORC in silencing. The reduced eYciency of multiple ARSs that are located close together may appear at first glance to support the idea that even ineYcient origins form pre-RCs. In such circumstances, the ARSs interfere with each other: rarely does more than one copy within the set fire in any given S phase [77–79], possibly as a consequence of asynchrony in the time of initiation. When such tandem ARS copies were examined by footprinting, pre-RCs were found to assemble on most or all copies [10,79], although presumably only one copy in the tandem array actually fired. Nevertheless, because these ARSs are each capable of firing at high eYciency in the absence of interference from neighboring copies, this system of tandem interfering ARSs may not be directly comparable to dormant

1. The Dynamics of Chromosome Replication in Yeast

19

origins. Therefore, it is still not certain if pre-RC formation occurs uniformly at all ARS elements in the genome.

1. Components of the Pre-RC and Their Roles Formation of the pre-RC requires the six ORC subunits (Orc1p–Orc6p), the Mcm complex subunits Mcm2p–Mcm7p, and additional proteins— Noc3p, Cdc6p, Cdt1p, and Mcm10p. The ORC and Mcm proteins seem to be the components that actually participate in initiation; the function of the other proteins is to load the Mcm complex onto the origin. The order of events in assembly of the pre-RC seems to be as follows (Fig. 3a). ORC and possibly Mcm10p are associated with origins throughout the cell cycle. In G1 phase, Cdc6p interacts with ORC and perhaps with Mcm10p; the interaction with ORC probably requires Noc3p. The Mcm complex is brought into the nucleus by the action of Cdt1p; the combined action of Cdc6p and Cdt1p then results in loading of the Mcm complex onto the ORC–DNA complex. At the end of G1 phase, the presence of ORC, Mcm10p, Cdc6p, Cdt1p, and Mcm2-Mcm7p at the origin of replication marks the completion of the pre-RC formation or licensing of the origin. As discussed previously, it is still not certain if the pre-RC forms at every ARS in the genome, and whether the eYciency of pre-RC formation correlates with either the eYciency or the timing of origin activation. a. Origin Recognition Complex and Its Functions. Mutational analysis of ARS1 and of ORC genes has shown that the ORC complex binds specifically to the ACS element and that this interaction is essential for DNA replication (see Dutta and Bell [80] and Bell and Dutta [81] for reviews). ORC binding in vitro is an ATP-dependent process [48]. Two aspects of ATP involvement, binding and hydrolysis, appear to play important roles. Using a nonhydrolyzable analogue of ATP, Klemm et al. [82] found that ATP binding, but not hydrolysis, promotes origin-specific interaction of ORC with origin DNA. A subsequent study found that a mutant orc1 allele defective in ATP hydrolysis blocks assembly of the pre-RC [83]. Overexpression of CDC6 relieved this defect, suggesting that the defective Orc1p titrates Cdc6p away from the origin-associated ORC. One implication of this result is that ATP hydrolysis promotes the release of Cdc6p from the complex [83], perhaps coupled to the transition to the pre-IC. Two subunits of ORC, Orc1p and Orc5p, possess ATP-related motifs: Walker A, predicted to have a role in ATP binding and hydrolysis, and Walker B, with a role in ATP hydrolysis only [84,85]. Both Orc1p and Orc5p bind ATP, and for both genes, mutant alleles defective in ATP binding cause growth defects [82]. This phenotype is much stronger with orc1

20

Lucas and Raghuraman

Figure 3

(Continued)

1. The Dynamics of Chromosome Replication in Yeast

21

Figure 3 Assembly of pre-RC and initiation. (a) Assembly of pre-RC and transition to pre-IC. Sequential association of diVerent proteins with the origin is shown. Question marks indicate unknowns—the number of Noc3p, Cdc6p, and Cdt1p molecules and Mcm complexes, and whether there is nucleation of additional Mcm complexes flanking the origin. (See Color Insert.) (b)Transition to the initiation complex and initiation. The sequence of events and the participants are shown; the details of copy number and precise associations are not known. For example, it is not known when ORC associates with the DNA after the origin region is duplicated, nor is it known how many molecules of RPA associate with the origin. (See Color Insert.)

than with orc5. Furthermore, the ATP-dependent specificity of origin DNA binding and the subsequent hydrolysis of ATP were both found to be mediated by Orc1p [82,83]. Therefore, the principal ATP hydrolysis activity apparently resides in Orc1p. The role of the ATP-binding activity of Orc5p remains unknown. b. Noc3p. A screen for multicopy suppressors of the DNA replication defect of cdc46/mcm5 has turned up NOC3, which codes for a basic helix– loop–helix protein previously known to be involved in preribosomal RNA processing [86]. Immunoprecipitation experiments show that Noc3p is

22

Lucas and Raghuraman

associated with ORC and Mcm proteins and with origin but not nonorigin sequences. This association with origin sequences is seen throughout the cell cycle and requires ORC—loss of chromatin-bound ORC is accompanied by loss of Noc3p after a shift to restrictive temperature of a temperaturesensitive orc2-x1 strain. Furthermore, using a thermolabile allele of NOC3 to inactivate the protein in synchronized cultures, Zhang et al. [86] demonstrated that Noc3p is required for the stable maintenance of Cdc6p and Mcms at the origins. Thus, it appears that Noc3p may mediate the first steps in the assembly of the pre-RC [86]. c. Cdt1p and Cdc6p. Cdc6p was isolated in a screen for mutants in cell cycle progression [87]. Upon shift to restrictive temperature, strains carrying a temperature-sensitive allele of cdc6 fail to enter S phase, arresting in G1. Cdc6p interacts with ORC [88], and immunoprecipitation experiments with conditional alleles of CDC6 have shown that Cdc6p is necessary for the recruitment of Mcm2p–Mcm7p to origin-bound ORC [74,89,90]. Cdc6p, like Orc1p and Orc5p, is a member of the AAA+ family (ATPase associated with a variety of cellular activities) and possesses the Walker A and Walker B motifs [91,92]. These two motifs seem to be essential for Cdc6p function. Mutations in the Walker A motif (K114E; [92]) or the Walker B motif (E224G; [93]) allow the mutant Cdc6p to interact with origins but prevents loading of the Mcm complex and blocks DNA replication. Based on the analysis of cdc6 mutations in two regions thought to be important for ATPase activity in AAA+ family members (sensors 1 and 2, distinct from the Walker motifs), Takahashi et al. [94] proposed that Cdc6p may hydrolyze ATP that is bound to ORC. This hydrolysis may induce a conformational change in ORC, thereby promoting loading of the Mcm complex [94,95]. There is some uncertainty in whether Cdc6p is required for maintenance of the pre-RC. One study found that Cdc6p could be eluted from chromatin preparations without loss of Mcm5p [74]. This result agrees with a similar finding with XCdc6p in Xenopus egg extracts [96]. However, a diVerent study found a slow decay of the pre-RC footprint at the yeast 2-mm plasmid origin of replication when temperature-sensitive cdc6-1 cells were held at the restrictive temperature in late G1 [89]. A possible resolution of this conflict comes from the work of Lengronne and Schwob [97]. They proposed that pre-RC complexes formed early in G1 can undergo disassembly prior to the onset of S phase and that a second burst of Cdc6p synthesis close to the G1/S phase transition is required to restore pre-RCs at origins that have lost the complex. This idea would suggest that Cdc6p is needed just for the assembly of Mcm proteins at the origin; the apparent requirement for maintenance of the complex could be simply a manifestation of the need for an additional round of assembly.

1. The Dynamics of Chromosome Replication in Yeast

23

It has been shown in human cells [98], Xenopus [99], S. pombe [100], and S. cerevisiae [101,102] that an additional protein, Cdt1p, is also necessary for the loading of the Mcm2p–Mcm7p complex. Cdt1p levels are cell cycle regulated. Using Myc- or GFP-tagged Cdt1p in budding yeast, Tanaka and DiZey [101] demonstrated that Cdt1p is excluded from the nucleus at the onset of S phase, before any DNA replication, and reaccumulates in the nucleus at the end of mitosis. The Mcm complex (at least Mcm2p and Mcm6p) and Cdt1p interact physically, as demonstrated by coimmunoprecipitation, and appear to be interdependent for accumulation in the nucleus in the late mitosis–G1 phase. However, artificial targeting of the Mcm complex to the nucleus in a cdt1 mutant strain is not suYcient to allow DNA replication. Therefore, Cdt1p is required not merely for accumulation of Mcm2– Mcm7p in the nucleus but is also necessary for some additional role in initiation [101]. Neither Cdt1p nor Cdc6p is needed for elongation, supporting the idea that this additional role of Cdt1p is in assembling the pre-RC in combination with Cdc6p [90,101]. d. Mcm2p–Mcm7p. MCM genes, which are necessary for the initiation of DNA replication, were identified in genetic screens looking for genes implicated in plasmid maintenance or cell cycle progression ([28,103]; for a review, see Lei and Tye [104]). Mcm2p–Mcm7p are a set of highly conserved proteins. They assemble into a complex that is imported into the nucleus; the subunits of the complex are interdependent for this import [74]. Once in the nucleus, as described earlier, they are loaded onto origins by Cdc6p and Cdt1p, probably via Noc3p. As the cells progress through S phase, the Mcm2p–Mcm7p proteins are reexported into the cytoplasm [74,90,105,106]. What is the function of the Mcm complex at the origin? Human Mcm4p, Mcm6p, and Mcm7p form a complex in vitro that has weak DNA helicase activity [107], but the popular consensus seems to be that the in vivo complex is probably a hexamer of Mcm2p–Mcm7p [104]. Despite the lack of demonstrable helicase activity of the hexamer and the poor processivity of the Mcm4–6–7p subcomplex, the Mcm complex remains a prime candidate for the job of the replicative helicase [108]. Two additional observations lend support to this idea. First, morphological features of the subcomplex support the helicase model. In E. coli and in SV40 replication, a replicative helicase forms a double hexameric ring (dnaB in E. coli, T antigen in SV40) that encircles one of the strands of the template DNA and leads the fork by unwinding the template DNA [109]. Electron microscopy of the Mcm subcomplex has shown that these proteins also form double rings with six-fold symmetry that resemble the bona fide helicase complexes described earlier [110]. Furthermore, an Mcm-related protein from the archaean Methanobacterium thermoautotrophicum has been found to assemble into a homomultimer that has helicase activity in vitro and forms double hexameric rings as

24

Lucas and Raghuraman

visualized by electron microscopy [111–113]—perhaps the strongest evidence that Mcms can indeed have the properties expected of a replicative helicase. However, more recent molecular modeling of the M. thermoautotrophicum Mcm complex has led to the suggestion that this protein can form single heptameric rings [114]. It is not clear whether this protein can form double hexameric rings as well as single heptameric rings or whether the diVerence is one of image analysis. Furthermore, these studies were done in the absence of substrate DNA; the structure of the complex may change if DNA is present. The second observation supporting the helicase model comes from chromatin immunoprecipitation experiments indicating that members of the Mcm complex migrate away from origins during S phase and track along with proteins found in the replication fork [115]. In these experiments, origin-containing fragments were found to coprecipitate with Mcm proteins before the time of firing of those origins, but not at later times in S phase. Instead, fragments that were progressively further from the origin were precipitated at progressively later S phase times. The observed kinetics of redistribution of the Mcm subunits coincided with the kinetics of redistribution of DNA polymerase , consistent with the idea that Mcm proteins travel along with the replication forks, away from origins. This interpretation would predict that in the simplest scenario, there must be at least two sets of Mcm proteins at each origin, one for each fork. It is not known if the same complex travels with the fork from initiation to termination. Presumably, the Mcm complexes are released upon fork termination and exit the nucleus. Many questions about Mcm function remain unanswered. Mcm complex members are highly abundant proteins, with estimates of copy numbers ranging from 15 to 100 the number of origins [74,105,116]. If the function of Mcm proteins is in origin initiation, why do Mcm complexes so outnumber origins? How many complexes are present at each origin, and are there additional complexes flanking the core pre-RC? There is evidence in Xenopus that there are secondary complexes flanking the preRC [117]. One can speculate that Mcms run a sort of a relay race—with Mcm complexes bound weakly along the chromatin, but getting loaded tightly (and replacing the old complexes) when a fork runs into them. Another possibility is that the high concentration is needed to drive complex formation. Loading of the complex to form the pre-RC requires Cdc6p and so on, but the half-life of the complex is not known, nor is it known whether the pre-RC, once formed, can still exchange Mcm subunits with Mcms in solution. This possibility seems intuitively to be unlikely, but it has been shown that some members (e.g., Rad52p) of large recombination foci can undergo dynamic exchange with subunits that are in solution [118]; perhaps such a dynamic exchange can also occur in replication complexes.

1. The Dynamics of Chromosome Replication in Yeast

25

Despite the apparently vast excess of Mcm proteins over origins, a 50% decrease in the cellular concentration of Mcm2p but not Mcm3p increases the plasmid loss rate [105]. However, this loss rate varies with the identity of the ARS on the plasmid: no increase in plasmid loss with ARS-H2B, a 2-fold increase with ARS121, a 4-fold increase with ARS-HO, and a dramatic 44-fold increase with ARS1 [105]. Mutations in MCM3 also cause a diVerential plasmid loss. In a mutant mcm3 strain, a plasmid containing ARS1 is very unstable (0.1% of the cells retaining the plasmid after eight generations in nonselective culture medium), whereas a plasmid containing ARS121 is less unstable (10% retention) [119]. These results suggest that although intracellular Mcm protein concentrations are far higher than those of origins, Mcm proteins are functionally at limiting concentrations within the cell. Additionally, the critical concentration required for Mcm deposition or function may diVer between origins or chromosomal contexts. In fact, it is not clear how much of the Mcm complexes is associated with chromatin. While indirect immunofluorescence and chromatin fractionation studies suggest that the bulk of Mcms become chromatin associated in G1 [74,120], immunoblot analysis of fractionated cell extracts seems to indicate that a substantial portion of Mcms is cytoplasmic at all times; even in the nuclear fraction, only a portion is associated with chromatin tightly, the remainder being nuclear but readily eluted [116]. e. Mcm10p. MCM10 (identical to DNA43) was identified in the same screen as some of the MCM2–MCM7 genes (MCM2, MCM3, and MCM5; [28]; for a review, see Lei and Tye [104]). No obvious similarity has been detected between Mcm10p and Mcm2p–Mcm7p. MCM10 is an essential gene, and two-hybrid assays have shown interaction between Mcm10p and Mcm2p–Mcm7p [121,122]. Furthermore, a mutant allele mcm10-1 suppresses the defects seen in mcm7-1 and in a double mutant, mcm7-1 mcm5-461 [122]. These observations indicate that Mcm10p, along with the Mcm2p–Mcm7p complex, is needed for origin initiation. Because no direct interaction has been seen between the Mcm complex and the ORC complex, it has been suggested that Mcm10p binds to the origin and acts as a docking site for the Mcm complex [104]. This hypothesis would predict that association of Mcm10p with the origin should be independent of the other Mcms. In fact, in budding yeast, Mcm10p binding was found not to require Orc2p or Mcm2p–Mcm7p, whereas binding of the Mcm complex did require Mcm10p [122]. However, a study using immunodepletion experiments in Xenopus egg extracts has come to the opposite conclusion: that the interaction of XMcm10p with the chromatin requires prior binding of Mcm7p [123]. These experiments also showed that XMcm10p is necessary for XCdc45p loading, which marks the transition

26

Lucas and Raghuraman

to the pre-IC. XMcm10p would then not be considered as a part of the preRC but is implicated in the initiation process. It is possible that the timing and perhaps the role of Mcm2–Mcm7p and Mcm10p during the pre-RC formation diVer between the Xenopus embryonic system and budding yeast. For example, unlike ScMcm10p, XMcm10p is not bound constitutively at the chromatin in Xenopus egg extracts. Furthermore, while the budding yeast Mcm complex shows tight interaction with chromatin at discrete sites only (the origins), the Xenopus Mcm complex seems to spread out from the ORC-binding site over a larger region of DNA [117]. These diVerences between yeast and Xenopus, as well as the absence of specificity of initiation in the egg extracts, are consistent with the hypothesis of a divergent function of Mcm10p between these two model systems. Clearly, further investigation is needed. In any event, Mcm10p is important for the initiation step in Xenopus as well as in budding yeast [121,123]. Furthermore, the fact that Mcm10p is needed for the completion of DNA replication after the release of cells from a hydroxyurea block and that MCM10 interacts genetically with several genes coding for elongation factors indicates that Mcm1p could also play a role during the elongation step [124]. One suggestion is that Mcm10p could be important for the passage of replication forks across origins that have not fired [122]. In a mutant mcm10-1 strain, forks appear to pause at origins that have not been fired. If an origin is mutationally inactivated (A or B2-B3 elements mutated), then neither bubble arcs nor fork pauses are observed by 2-D gel electrophoresis at this origin. If the ARS mutation only reduces its eYciency, then some pauses are seen and, as expected, fewer bubble intermediates. Perhaps Mcm10p facilitates passage of the replication fork through the pre-RCs, which would explain why the inactive mutant ARS did not exhibit fork pauses—presumably, it did not assemble a pre-RC and therefore did not present a problem. However, it would be interesting to confirm this hypothesis by looking at replication intermediates in a mutant mcm10 strain at silent origins. By the end of G1, pre-RCs (ORC and Mcms) are assembled at all ARSs destined to be origins in that cell cycle, as judged by ChIP and footprinting assays. Activation of S-CDK at this point results in the elimination of Cdc6p and Cdt1p from the complex, their job done [90,101]. Elimination of Cdc6p and Cdt1p is not merely an incidental side eVect but an essential step in preventing reformation of pre-RCs during S phase (see Section IV,C). Subsequent events appear to parallel the time of origin firing as detected by the assays outlined earlier, i.e., they occur early in S phase at origins that fire early and later in S phase at origins firing later. These events include recruitment and assembly of proteins needed at the replication fork, some of which may need to occur simultaneously.

1. The Dynamics of Chromosome Replication in Yeast

27

B. Origin Firing Origin firing or activation is generally seen as consisting of the following set of events:  unwinding of the DNA at the origin  loading of polymerases and synthesis of the primer RNA  activation of replicative helicase At the operational level, assays for activation are less clearly defined. The presence of specific proteins (such as polymerases) at the origin and the fork are inferred from chromatin immunoprecipitation experiments. Unwound DNA can be detected by assaying for sensitivity of the DNA to potassium permanganate, which modifies unstacked Ts [125]. This assay is rarely used, however; the simpler assay of 2-D gel electrophoresis is generally preferred, even though it only detects replication bubbles that have already matured by at least a few hundred base pairs. The time of origin activation is monitored either by 2-D gel electrophoresis with DNA from timed S phase samples or by density transfer experiments. An alternative, high-resolution method used to detect origin firing is replication initiation point mapping (RIP), which detects the position of the initial primer [31]. A possible drawback of this technique is that quantitation of origin firing eYciency may be diYcult; indeed, it is generally used to map the location of the initiation point and not the eYciency or time of firing. Alkaline gel electrophoresis detection of replication intermediates at origins has gained popularity as a quick assay for successful origin firing (see Section II,C). Much of our understanding of the mechanism of origin firing comes from work on SV40 [126–128]. In this system, origin firing begins when the SV40 T-antigen (T-Ag) binds to the origin and causes unwinding of origin DNA. This unwinding is a two-step process: T-Ag first opens the DNA and the open DNA then becomes a substrate for the binding of replication protein A (RPA), the eukaryotic equivalent of E. coli single-stranded DNA–binding protein (ssb). RPA binding results in a larger region of stably unwound DNA. The T-antigen then recruits DNA polymerase -primase, which synthesizes short RNA primers. Replication factor C (RFC, the clamp loader) binds to the primer, displacing Pol and recruiting PCNA (the sliding clamp, which acts as a processivity factor for DNA polymerase) and DNA polymerase . The stage is now set for fork elongation. The T-Ag functions as a replicative helicase leading the fork; RPA, RFC, Pol-primase, and pol  follow the unwinding DNA. Some equivalent pattern of events is expected to be true for yeast also. The sequence of events as best as we know is outlined here. Several of these steps require cyclin-dependent kinase (CDK, a heterodimer of the kinase Cdc28p with cyclin, its regulatory subunit) and Dbf4-dependent kinase (DDK, a

28

Lucas and Raghuraman

heterodimer of the kinase Cdc7p with its regulatory subunit Dbf4p), which will be pointed out in the relevant places.

1. Transition to the Preinitiation Complex The earliest detected event at a licensed origin is the association of Cdc45p with the pre-RC. CDC45 was first detected as a cold-sensitive mutant that failed to undergo DNA replication at the restrictive temperature [103,129]. Interaction between Cdc45p and Mcm complex members has been found by two-hybrid analysis [130] and by coimmunoprecipitation [55,131,132]. Furthermore, synthetic lethality has been observed between cdc45 and orc2-1 [133]. These observations all pointed to an integral role of Cdc45p in origin initiation. The association of Cdc45p with origins requires the activity of CDK and DDK [55,132]. There is some confusion as to when exactly Cdc45p becomes associated with origins. In one study, cells were arrested late in G1 with  factor, and ChIP was used to monitor Cdc45p binding at the early firing origins ARS305 and ARS607 after release from the block [55]. These authors did not detect Cdc45p at the origin until 10 min after release from the block. In contrast, two other studies detected Cdc45p at ARS305 even in cells arrested with  factor [134,135]. A possible resolution to this apparent conflict is that Cdc45p may bind in two steps—an initial loose association followed by tighter binding. In any event, the apparent discrepancy may not be of particular consequence, as Cdc45p can be dispensed with until START [136]. Both sets of studies do come to the same conclusion with respect to Cdc45p binding to origins that fire later in S phase—that Cdc45p binding at late-firing origins does not occur until later in S phase, at times that approximately match the time of firing of those origins as detected by density transfer and 2-D gel analysis. Because the resolution of the studies performed to date does not allow fine discrimination of origins that fire at intermediate times in S phase, we cannot as yet tell if the program of Cdc45p binding exactly matches that of origin firing, or whether there are, for example, two or more distinct waves of Cdc45p binding. Cdc45p binding requires the presence of two other factors: Sld3p and RPA. SLD3 was originally identified as part of a screen to find genes that, when mutated, are synthetically lethal with dpb11-1 [135]. Because Dpb11p interacts with DNA polymerase  [137], the genetic interaction between SLD3 and DPB11 suggested a role for SLD3 in DNA replication. More recently, an extensive series of experiments has led to the conclusion that Sld3p and Cdc45p interact and are mutually dependent for binding to the origin [138]. This conclusion was based on several lines of evidence.

1. The Dynamics of Chromosome Replication in Yeast

29

CDC45 and SLD3 on multicopy plasmids can suppress the replication defects seen in sld3 and cdc45 mutant strains, respectively. Interaction between Cdc45p and Sld3p was seen both in a two-hybrid screen and by coimmunoprecipitation of the two proteins. Sld3p associates with origins as demonstrated by ChIP assays, and this association depends on a functional ACS, at least in the case of ARS1. Also, as with Cdc45p, the time of association of Sld3p with the early- and late-firing origins ARS305 and ARS501 roughly matches the time of firing of those origins. The two proteins can be coimmunoprecipitated throughout the cell cycle, indicating that they function as a heterodimer. Consistent with this idea, the two proteins were found to be mutually dependent for binding to the origin as judged by ChIP, and the association of Cdc45p with Mcm2p requires Sld3p. However, successful coimmunoprecipitation of the two proteins required prior in vivo cross-linking with a strong cross-linking agent, suggesting that the Cdc45p–Sld3p complex is a fragile one. Finally, Sld3p is needed for the resumption of S phase upon release from a hydroxyurea block [138]. This result is interpreted as evidence that Sld3p is needed for ongoing fork elongation, but could potentially be related to recovery from aberrant, hydroxyurea-induced structures. Binding of Cdc45p and of the single-stranded DNA-binding protein RPA are also mutually dependent [55]. This observation would imply that Cdc45p, Sld3p, and RPA all bind simultaneously. However, because the Rfa1p subunit of RPA was not coimmunoprecipitated with Cdc45p and Sld3p with or without in vivo cross-linking, it appears that RPA is not an integral part of the Cdc45p–Sld3p complex [138]. 2. Initiation Transition to the initiation complex is followed by (or is concomitant with) binding of Sld2p and Dpb11p at the origin. SLD2 (also known as DRC1) was discovered along with SLD3 as being synthetically lethal with dpb11-1 when mutated [135]. In addition to this synthetic lethality, twohybrid, as well as direct physical interaction (coimmunoprecipitation), have been detected between these two proteins [135,139]. Mutant strains temperature sensitive for either gene show reduced origin activity at ARS306 as judged by the intensity of bubble arcs in 2-D gels [135], confirming the role of these proteins in replication. Several observations together suggest the outline of an attractive model for the role of Sld2p and Dpb11p in origin firing. The observations are as follows. Coimmunoprecipitation and ChIP experiments have shown that Dpb11p interacts with DNA polymerase  and that this interaction is needed for the association of both proteins with origins [140]. The interaction between these two proteins may be regulated by the S-CDK-dependent

30

Lucas and Raghuraman

phosphorylation of Sld2p. There are six preferred S-CDK phosphorylation sites on Sld2p; mutation of all six sites (the ‘‘All-A’’ allele) to prevent their phosphorylation also blocks the interaction of Sld2p with Dpb11p [141]. Furthermore, phosphorylation of Sld2p, as well as formation of the complex between Sld2p and Dpb11p, was found to peak during S phase, and it was predominantly the phosphorylated form of Sld2p that was coimmunoprecipitated with Dpb11p. Very similar results have also been seen in S. pombe [142]. ChIP experiments using temperature-sensitive dpb11-9 and cdc17-1 (Pol) strains show that while association of Pol with origins ARS305 and ARS1 requires Dpb11p, the converse is not true—association of Dpb11p with these origins occurs at the restrictive temperature in a cdc17-1 strain [140]. One limitation of this study is that these origin associations were assayed in cells released into S phase in the presence of hydroxyurea, conditions in which checkpoint responses are evoked (Section VI). This limitation is particularly relevant in light of the finding that Dpb11p and Sld2p are involved in the Rad53p-mediated S phase checkpoint [139,140]. Nevertheless, combined with the observation that Dpb11p and Pol are mutually dependent for association with origins [140], these results imply that although Pol-primase is the first polymerase activity needed at the fork, recruitment of this polymerase requires prior recruitment of Pol. The essential role of Pol seems to be its checkpoint function—indeed, its N-terminal catalytic domain is dispensable in both budding and fission yeasts, while the C-terminal domain is required [143–145]—and the requirement for sequential binding of the Pol and Pol primase may enforce the presence of this checkpoint component before the onset of DNA synthesis [81]. These observations suggest a unified model for the process of initiation (Fig. 3b). S-CDK action converts Sld2p to a phosphorylated form, which can bind to Dpb11p and recruit Pol to the origin. Successful recruitment of Pol allows recruitment of Pol-primase, at which point primer synthesis can occur. Synthesis of a short RNA primer is followed by brief extension by the DNA polymerase activity of Pol (for reviews of the organization of the replication fork, see Waga and Stillman [109] and MacNeill [146]). This enzyme is then displaced by the homotrimeric processivity factor PCNA (¼ Pol30p), loaded by the five subunit RFC in an ATP-dependent process [147]. DNA polymerase  (or ) is loaded, and DNA synthesis continues. The precise roles of Pol and Pol in leading and lagging strands are not understood, but current opinion seems to be that Pol and Pol are required for lagging strand synthesis while Pol is used for leading strand synthesis, although it can be substituted by other polymerases [109]. There are several gaps in this model with respect to the roles of Dpb11p, Sld2p, and Pol. The S-CDK-dependent phosphorylation of Sld2p is the first

1. The Dynamics of Chromosome Replication in Yeast

31

demonstration of a positive regulatory substrate for S-CDK, and the idea that phosphorylation of Sld2p is required for recruitment of Pol (via Dpb11p) is an attractive one. However, this requirement has not been demonstrated directly, although it will surely be an active area of investigation in the near future. Likewise, the order of assembly of Pol and Pol-primase is inferred from the observation that Pol binding requires Dpb11p; direct requirement of Pol has not been demonstrated. Furthermore, as mentioned earlier, those experiments were done in hydroxyurea and need to be confirmed in the absence of added drug. These limitations notwithstanding, the overall model is attractive for a couple of reasons. First, it provides at least one substrate for S-CDK that is needed to promote initiation rather than to block reinitiation (see Section IVC). Second, it suggests a role for Sld2p and Dpb11p—that of a pivotal control point, both by depending on S-CDK (which is in turn tied to Cln-CDK and therefore to cellular physiology; [148]) and by checkpoint systems [139,140]. 3. Postinitiation ChIP experiments looking for protein–DNA interactions in DNA fragments at and near origins of replication have shown that several of the proteins at the initiation complex are incorporated into the elongating replication fork, including the Mcm complex, Cdc45p, and probably Sld3p [115,138]. If the Mcm complex is indeed the replicative helicase, then finding it at the fork is no surprise. The role of Cdc45p at the fork is unknown, although experiments using thermolabile alleles of Mcms and Cdc45p may provide a possible explanation. When Mcm complex components are depleted (using alleles coding for thermolabile proteins) in cells released into S phase in hydroxyurea, replication is blocked irreversibly—no resumption of fork elongation is seen by density transfer analysis when hydroxyurea is removed whether or not new Mcm synthesis is allowed [149]. Considering how well the system is tuned to prevent the reassembly of Mcm complexes at origins (see Section IVC) the failure to reform Mcm complexes at forks should perhaps not be surprising. Alternatively, loss of Mcm complex members may damage forks in some fundamental and irreversible way. Depletion of Cdc45p results in a similar failure in the resumption of DNA synthesis after the removal of hydroxyurea. Unlike with Mcm proteins, however, new synthesis of Cdc45p allows resumption of DNA replication [150]. Perhaps it is just the polymerases and later stage components of the fork that are lost upon removal of Cdc45p, and perhaps reconstructing the fork requires rebinding of Cdc45p, Sld3p, etc. Suggestive evidence along these lines comes from fission yeast, where it appears that SpSld3p is needed to maintain the association of SpCdc45p with chromatin in hydroxyurea-treated cells

32

Lucas and Raghuraman

[151]. The role of Cdc45p at the fork could be to maintain the association of polymerases with the fork and to allow rebuilding of the fork should the polymerases disassemble. In this scenario, the fork resembles an ordered stack of components; loss of components in the stack causes loss of components added later but not those added before. A couple of tests of this idea are obvious. It predicts that reversibility of the replication block following Cdc45p depletion should be dependent on Sld2p and Dpb11p (those components being added later in the ‘‘stack’’). It also predicts that Pol and Pol should be lost from forks when Cdc45p is depleted. 4. The Requirement for CDK and DDK Tight binding of Cdc45p (and presumably Sld3p) requires the action of S-CDK and DDK [55,132]. How these kinases bring about this tight association is not known, and the order of action of the two kinases remains confusing. A series of reciprocal inactivation experiments using a galactose-regulated version of the S-CDK inhibitor SIC1 and a temperature-sensitive cdc7-1 allele suggests that DDK function requires the prior activity of S-CDK [152]. However, these interpretations contradict the interpretations drawn from experiments done in Xenopus egg extracts [153,154], which concluded that DDK must act before S-CDK. An obvious explanation is that yeast and Xenopus simply diVer in how they assemble the pre-IC. For example, the Mcm10 and Mcm complex proteins show a diVerent order of addition in yeast compared to Xenopus. An alternative possibility is that confusion arises because one or both kinases function at more than one step along the pathway. For example, S-CDK is needed not only for transition of the pre-RC to the pre-IC, but is also apparently needed for the recruitment of Pol to the initiation complex [141]. Additionally, it may well be that some proteins need to be present physically at certain steps in the assembly pathway even though their enzymatic activities are not needed until later [104]. This strategy would ensure that critical steps in the assembly pathway are initiated only when components to be used later are already present and accounted for. Such seems to be the case with Pol primase and Pol: although the catalytic activity of Pol comes into play only after that of Pol-primase, recruitment of the latter seems to require prior recruitment of the former [140]. S-CDK is known to phosphorylate Cdc6p [155,156]; perhaps the SCDK-mediated removal of Cdc6p promotes the subsequent interaction of Cdc45p with the pre-RC. DDK phosphorylates Mcm proteins (particularly Mcm2p) and Cdc45p in vitro [92,152,157], which may mean that the phosphorylation of these two substrates promotes their interaction. If so, however, the DDK-dependent phosphorylation of Cdc45p cannot be

1. The Dynamics of Chromosome Replication in Yeast

33

necessary for origin firing, as the mcm5-bob1 allele allows bypass of the requirement for CDC7 and DBF4 [158]. The bypass of the DDK requirement in mcm5-bob1, taken together with the observation that RPA binding requires DDK activity [159], suggests the following admittedly speculative sequence of events. S-CDK brings about the transition from pre-RC to pre-IC by promoting the initial association of Cdc45pSld3p with the preRC. Phosphorylation of Mcm complex subunits by DDK then allows melting of origin DNA and promotes tight association of Cdc45p with the origin and binding of RPA to the unwound DNA, forming the initiation complex (IC). DNA synthesis begins upon the subsequent binding of polymerase  and synthesis of RNA primers. This model predicts that the DDK-dependent change in the state of the Mcm complex is a crucial step in the transition from pre-IC to IC; in fact, it has been observed that origin DNA undergoes premature melting in the mcm5-bob1 strain [125]. However, Cdc45p shows precocious and elevated levels of association with the early origins ARS305 and ARS1 but not with the late origin ARS501 in mcm5-bob1 cells [160]. Therefore, the proposed DDK-dependent unwinding of DNA is not suYcient to promote binding of Cdc45p. Other questions are yet to be addressed. What is the role of Mcm10p in this transition? What exactly does Cdc45p do? Does it actively promote the disengaging of the Mcm complex from Mcm10p [104] or is it merely an intermediary for the subsequent loading of polymerases?

C. Preventing Reinitiation The classic cell fusion experiments of Rao and Johnson [161] highlighted what has come to be known as the ‘‘once and only once’’ problem—every portion of the genome must be replicated once in every cell cycle, but no more than once. Solutions to the problem might be trivial if all origins fired synchronously at the start of S phase. In reality, however, origin initiations occur throughout S phase, raising the question of how initiation at early origins (which had already fired early in S phase) is suppressed while late origins fire. This question remained one of the big puzzles in the field of DNA replication until a few years ago, when a solution to the puzzle emerged. While all the details of the solution have yet to be elucidated, a fairly robust framework has been worked out and has been reviewed extensively [162–164]. The essence of the solution is simple and can be summarized as two key points.  Initiation occurs in two steps: assembly of the complex that is needed for initiation (whereby the origin becomes ‘‘licensed’’ for firing) and activation of that complex to achieve initiation.

34

Lucas and Raghuraman

 Conditions needed for activation of the complex are incompatible with those needed for its assembly. Although some details may diVer, the same general scheme applies to all eukaryotes studied thus far. The switch between these two states is governed by S-CDK. B-type cyclins (Clb1p–Clb6p) are synthesized in a staggered series beginning in late G1— first Clb5p and Clb6p, then Clb3p and Clb4p, and finally Clb1p and Clb2p [165]. CDK thus remains active between the end of G1 and mitosis, at which time the Clbs are targeted for destruction. The rapidity of the change in ClbCDK levels in the cell is further ensured by the activity of Sic1p, a specific and stoichiometric inhibitor of Clb-CDK. Sic1p is present in G1 and is destroyed rapidly at the G1/S phase transition by CDC4-dependent ubiquitination, thereby releasing the inhibition of S-CDK activity. Thus, two cell cycle-dependent states of Clb-CDK activity are established: low activity from mitosis to late G1 and high activity through S and G2 phases [165]. The importance of these two states is that the recruitment of Mcm proteins to form the pre-RC can occur only in the absence of CDK activity, so formation of the pre-RC is restricted to a window of opportunity lasting from mitosis to late G1 (Fig. 4). In contrast, the transition of the pre-RC to the initiation complex and the initiation of DNA synthesis require CDK activity, and therefore can only occur after S-CDK levels build up after G1. Thus, an origin can have one of three fates. It can fail to form a pre-RC during G1, in

Figure 4 Preventing reinitiation. Clb-CDKs begin accumulating in late G1, peak during S and G2 phases, and drop precipitously at mitosis [165]. Assembly of pre-RCs can occur only in the absence of Clb-CDK activity and initiation requires Clb-CDK activity [162–164]. Thus, preRCs cannot form once the cells have entered S phase and reinitiations at origins are thereby prevented.

1. The Dynamics of Chromosome Replication in Yeast

35

which case it cannot initiate replication in the following S phase; it can form a pre-RC and therefore be licensed for firing, but be replicated passively (which presumably causes loss of the pre-RC); or it can be licensed and successfully initiate replication, which also causes loss of pre-RC. Whichever of the three fates it experiences, therefore, the final outcome is that the origin lacks a pre-RC by the end of S phase and cannot reacquire a new pre-RC until the cell passes through mitosis and reenters G1. Thus, by making the conditions for formation of the initiation complex and for the use of the complex mutually incompatible, the cell very elegantly gets around the bookkeeping problem of which origins have fired and which have not. Several lines of evidence led to this model of CDK-dependent regulation of pre-RC assembly. The time of activation of S-CDK in late G1 coincides with the time beyond which pre-RC formation cannot occur [89,90]; delaying the time of expression of Clb-CDKs extends this window of time for pre-RC assembly [166]. Furthermore, ectopic expression of the ClbCDK inhibitor Sic1p in G2 phase cells allows the formation of pre-RCs and an extra round of DNA synthesis [167]. Conversely, ectopic expression of Clb2p in G1 phase cells (when Clbs are not normally present) before the expression of Cdc6p prevents pre-RC formation and blocks replication [168]. CDK-dependent inhibition of pre-RC assembly occurs through multiply redundant paths involving Cdc6p, Mcms, ORC, and Cdt1p. CDK-dependent phosphorylation of Cdc6p in late G1 targets it for destruction, thereby eliminating one of the components needed for the assembly of pre-RCs [155,156,169]. Regulation of Cdc6p takes two additional forms. First, CDK activity blocks the nuclear import of Swi5p, which is needed for the transcription of CDC6 [170,171]. Second, a mutant version of Cdc6p that cannot be phosphorylated shows six- to sevenfold greater aYnity for Mcm2p, indicating that the normal, late-G1 phosphorylation of Cdc6p reduces its aYnity for Mcm2p in addition to targeting it for destruction [172]. A second mechanism preventing reformation of pre-RCs in S phase is the export of Mcm proteins from the nucleus. Visualization of Mcm4p-GFP in live yeast cells showed that it is nuclear in G1 cells but becomes predominantly cytoplasmic in S phase [106]. Export of the protein from the nucleus was Clb-CDK dependent: Mcm4-GFP maintained its nuclear localization in cells expressing a stabilized, nondegradable form of the Clb-CDK inhibitor Sic1p; conversely, ectopic expression of Clb2p resulted in premature export from the nucleus [106]. Details of a third mechanism are not known, but the phosphorylation of ORC subunits is involved. Orc2p and Orc6p contain consensus CDK phosphorylation sites, and the proteins show S-phase-specific phosphorylation. Mutation of the phosphoacceptor residues to alanine did not result in any reduction in the eYciency of origin firing but did allow rereplication within one cell cycle, resulting in accumulation of >2N DNA content.

36

Lucas and Raghuraman

However, this phenotype was seen only when the ORC mutations were combined with disruptions in the normal regulation of Mcm and Cdc6p (Mcm kept in the nucleus by fusion of a nuclear localization signal to Mcm7p, and Cdc6p blocked from degradation using a partially stabilized allele) [173]. No pairwise combination of disruptions was suYcient to cause overreplication, and even the triply disrupted strain did not undergo a complete extra round of replication; 2-D gel analysis detected reinitiation at some origins but not at others [173]. These results highlight the elaborately redundant mechanisms that cells have evolved to prevent rereplication. More recently, yet another mechanism has been uncovered—one involving the depletion of Cdt1p from the nucleus. As described earlier, Cdt1p is needed along with Cdc6p to recruit Mcms to the ORC complex at origins [101,102]. Like Mcms, Cdt1p is also exported from the nucleus during S phase, and this export is blocked by inhibition of Clb-CDK by the ectopic expression of Sic1p [101]. This newly discovered mode of CDK-dependent control may explain the failure to get a complete extra round of replication in the experiments described earlier [173]. Preventing reinitiation must be a particularly high priority in higher eukaryotes, which have evolved a completely separate, CDK-independent mechanism to regulate Cdt1p activity—one mediated by geminin, a specific inhibitor of Cdt1p that is destroyed during G1 to allow activation of Cdt1p [164].

D. What Determines Origin Choice and Efficiency? Origin eYciency on the chromosome can be high or low—some origins function in almost every cell cycle, whereas others are used less frequently. For example, of the nine ARSs mapped on chromosome VI [13,174], only four are used in at least 50% of the cell cycles; the others are used variously at eYciencies of 10–40% [13]. Likewise, 11 of the 19 potential origins on chromosome III are used in <50% of the cell cycles [66]. Why origins show such a range in eYciency is not well understood. One possibility is that an origin is ineYcient if it has low competence to assemble a pre-RC (e.g., a suboptimal set of A and B elements), leading to <100% occupancy by initiation proteins in the population. An example of such an origin is the multiple copy rDNA ARS—this ARS seems to have low firing eYciency because it is intrinsically weak [175,176]; perhaps only a fraction of the 200 copies of the rDNA ARS become licensed in any given cell cycle. As mentioned earlier, this origin uses an imperfect match to the ACS even though a perfect match is available nearby [47]. One might predict that mutation of this imperfect match to a perfect match should improve the eYciency of this ARS, and indeed, that is exactly what happens [177]. In wild-type rDNA, only one in 5–7 copies of the rDNA ARS is actually used

1. The Dynamics of Chromosome Replication in Yeast

37

on average. Ideally, one would like to ask if that value increases with the mutated, ‘‘improved’’ rDNA ARS, but of course replacing all 200 copies of the ARS in vivo is not a straightforward experiment. An additional factor contributing to the weakness of this ARS may be that its DUE is not particularly amenable to unwinding [177]. Thus, quantitative diVerences in origin activity and ARS eYciency may be a reflection of the overall cumulative (and possibly synergistic) contributions made by the diVerent ‘‘accessory’’ elements. Indeed, although ORC and the Mcm protein complexes are believed to be required for origin activity (see later), the genome-wide ChIP-to-chip experiment mentioned earlier found many sites where some but not all ORC or Mcm proteins were detected [23]. Even if this result is not a consequence of masked epitopes, it probably does not mean that some ARSs function without Mcms. Rather, a more plausible interpretation is that as a consequence of thermodynamic limitations (perhaps arising from suboptimal arrangement of the A and B elements), some ARSs have a lower probability of capturing ORC or Mcm proteins or of maintaining the complex once formed. These ARSs would presumably show a correspondingly lower frequency of initiation. One obvious test of this hypothesis (if the reduced binding of Mcm or ORC proteins is confirmed) would be to test ARS activity (by plasmid loss assays) and origin eYciency on the chromosome (using 2-D gel electrophoresis) of these predicted ARSs. Alternatively, an origin could have high competence, but low eYciency, i.e., it becomes licensed at a high frequency, but nevertheless fails to fire in some fraction of the population. That such situations exist is implied by two observations.  ARSs that seem quite capable of driving plasmid replication nevertheless can be completely or almost completely silent in their native chromosomal locations. For example, an ARS301 plasmid is lost at only a slightly higher frequency than an ARS1 plasmid, but ARS301 is normally completely silent in its native location on chromosome III [9,27,66,76]. Likewise, origins that are fully functional at one chromosomal location can be silent at a diVerent location [178].  Silent origins may form prereplicative complexes that look fully functional by the blunt instrument of nuclease footprinting [76]. As discussed earlier, this example (ARS301) may turn out to be unique by virtue of its location in a silenced region; the generality of this phenomenon will be known only upon examination of more silent origins. Origin interference is one obvious explanation for why these presumably licensed origins fail to fire, i.e., two adjacent origins may both be competent to fire, but their proximity dictates that unless they both happen to fire within moments of each other, the fork from one usually will passively replicate the other [77,79]. If the local context influences one origin so as to

38

Lucas and Raghuraman

advance or retard its firing time, the interference will have a bias imposed on it, with one origin always firing before the other and therefore showing a consistently higher eYciency [78]. In an extreme version of this phenomenon, an origin may become, in eVect, infinitely late and therefore always replicated passively even though it is fully competent to fire. DiZey and coworkers [10,76] argued that ARS301 is an example of such an origin. This origin appears to be capable of firing in its native location when the nearest eYcient origins ARS305 and ARS306 are eliminated [178] or when incoming forks are impeded by hydroxyurea treatment and the checkpoint inhibition of origin firing is abrogated by a rad53 mutation [10,76].

E. Fork Migration and Termination Very little is known about the regulation of fork migration in normal S phases. Fork migration rates deduced from a genome-wide analysis of replication times showed a broad range of <1 to >5 kb/min, with a mean of 2.9 kb/min [19]. The mechanism behind this variation in fork migration rate is not known. One possibility is that the fork rate is determined by the nucleotide sequence being replicated by the fork. Alternatively, the origin giving rise to a fork could have some influence on the properties of the fork, including its migration rate. Phosphorylation of the Mcm complex is believed to be a key event in origin firing. If the Mcm complex is indeed the replicative helicase, phosphorylation or another modification of Mcm subunits could be instrumental in regulating the progress of the fork. Fork termination is another little-understood event. With the exception of replication fork barriers, such as that found in the rDNA [179], replication is not believed to terminate at specific locations. Instead, it appears that forks terminate wherever they happen to be when they meet another oncoming fork, with the result that termination usually occurs within a zone rather than at a discrete site [180]. Variation in the precise site of termination could arise from cell-to-cell variation in the firing time of adjacent origins or in fork migration rates. How exactly forks are resolved when they meet and terminate is not understood.

V. The Temporal Program of Origin Activation Active origins of replication are not fired at the same time but follow a staggered program of activation, which is conserved from one cell cycle to the next (for a review, see Fangman and Brewer [181]). For example, the chromosome IV origin ARS1 is activated early in S phase, whereas ARS501 on chromosome V is activated late [182].

1. The Dynamics of Chromosome Replication in Yeast

39

Although origins are often referred to as ‘‘early’’ or ‘‘late,’’ it has become apparent from two recent genome-wide analyses of replication that origin activations occur throughout S phase [19,21]. One diVerence between these two studies is that while the first study, using a density transfer approach, concluded that most initiations occur around mid-S phase [19], the second study, using a marker frequency analysis, came to the conclusion that there are two broad, overlapping waves of origin activations [21]. Clearly, further experiments will be needed to resolve the diVerences between the two experiments, but one possible explanation lies in the details of the data analysis. The marker frequency analysis set criteria for the identification of origins such that many fewer locations were scored as being putative origins. Additionally, while this approach could distinguish twofold changes in copy number quite readily, it is not clear if smaller changes in copy number at intermediate times in S phase could be detected as reliably. However, S phase samples were collected at shorter intervals in the marker frequency analysis than in the density transfer experiment, and so the marker frequency experiment could, in principle, give a more accurate representation of the progression of S phase events than the density transfer experiment. Nevertheless, the two studies agree extremely well on the overall dynamics of replication for the genome, and such studies will surely be invaluable in the analysis of replication timing, e.g., in mutant strains showing defects in the S phase program.

A. Setting Up and Reading the Temporal Program 1. The Determinant(s) for Firing Time and Determinants for Origin Activity Are Separable A comparison of origin firing eYciency and firing time reveals that there is no clear correlation between the two properties. While late firing may contribute to poor eYciency of an origin because it is sometimes replicated passively before it fires, as has been suggested for ARS301 [10,76], it does not follow that late-firing origins are all ineYcient. For example, ARS501, the first late-firing origin to be discovered, shows high eYciency of firing as judged by 2-D gel analysis (e.g., see the 2-D gels in Ferguson et al. [182] and Raghuraman et al. [183]). Likewise, origins that fire early in S phase are not necessarily eYcient; the close proximity of other origins can result in interference between them such that each origin is used in only a fraction of cell cycles [77–79]. Finally, some origins show a strain-dependent variation in eYciency without a change in firing time (I.Lucas, unpublished results). These results imply that the determinants for activity of an origin are separable from those determining the time of firing of the origin. Direct testing

40

Lucas and Raghuraman

of this idea using mutations or engineered chromosomes is diYcult because changes in origin firing time or in eYciency will have indirect consequences on origins in surrounding regions by virtue of passive replication. Indirect support for the distinction between the determinants of the firing time and the determinant(s) for origin activity comes from measurements of the eYciency of an origin that has changed its replication timing. The subtelomeric late Y’ origin fires earlier in a sir3 mutant than in wild type, but the origin is very eYcient at both firing times as measured by fork direction analysis by 2-D gel electrophoresis [184]. 2. Sequence Elements That Affect Origin Activation Time a. Telomeres. It has been shown by density transfer experiments that loci immediately adjacent to several telomeres replicate late [182,185,186]. These observations have been confirmed and extended by the microarray analysis of DNA replication [19]. In addition, some origin translocation experiments have demonstrated that the telomeric regions seem to cause late activation of origins (Fig. 5a).  A copy of ARS1 inserted close to ARS501, near the right telomere of chromosome V, replicates late in S phase while the copy in the native location remains early firing [187]. Likewise, insertion of ARS1 in a late-replication region on chromosome XIV causes it to become later firing, while the endogenous copy maintains its normal early S firing time [188].  ARS501 is activated early in S phase on a circular plasmid, but reverts to being late firing when the circular plasmid is converted to a linear minichromosome by linearization and addition of telomere sequences [187]. b. Late Determinants on Chromosome XIV. A contig of five late origins has been identified on chromosome XIV (ARS1410–ARS1414; [188]); late activation of at least two of these origins is imposed on them by cis-acting flanking sequences (Fig. 5b).  When cloned into circular plasmids as small DNA inserts, ARS1412 (2.9-kb fragment) or ARS1413 (1.4-kb fragment) fires early in S phase.  However, if larger inserts containing more of the surrounding regions on either side of the origin are used, ARS1412 (17.4 kb) and ARS1413 (13.6 kb) are both activated late in S phase, at approximately the times the native chromosomal versions of the origins are activated. Deletion analysis showed that at least four separate elements contribute additively to delaying the time of activation of ARS1413 [188].

41

1. The Dynamics of Chromosome Replication in Yeast (a)

ARS1 Chr IV Early ARS1

ARS501

Tel

Chr V Late

Late

Early

Late ARS1413

(b) Chr XIV

Late

Late

Early

(c)

ARS501

Y' ARS

Late

Late

Late

Early

Early

Early

Wild type

Tel

sir3

ku70

Figure 5 Position eVects on origin activation time. (a) Origin transplantation experiments demonstrating the telomere position eVect on activation of ARS501 [182,187]. (b) Origin transplantation experiments demonstrating late determinants on chromosome XIV [188]. (c) Distinct contributions of SIR-dependent and Ku-dependent mechanism mechanisms to the telomere position eVect. Sir3p aVects the Y’ ARS but not ARS501, whereas Ku aVects both [184,193].

42

Lucas and Raghuraman

c. Sequences That Appear to Advance Origin Activation Time. Early observations in yeast describing cohesion of sister centromeres until anaphase led to the hypothesis that centromeric regions must be late replicating. Density transfer analysis of replication timing of centromere-containing fragments found, instead, that centromeric are actually replicated early in S phase [185]. As with the telomeric regions, this observation has since been confirmed and extended to all the centromeres by a genome-wide approach [19]. The 32 origins flanking centromeres that have been predicted by this genomic approach were found to replicate significantly earlier than the genomic average for all origins. This eVect seems to be correlated with proximity to the centromere, fading away with the increasing distance from the centromere. These observations raise the exciting possibility that centromeres, like telomeres, can exert an influence on the time of origin activation, although in the opposite direction (causing early instead of late activation). An alternative view is that early replication is important for some aspect of centromere function, leading to strong selective pressure for centromeres to be located in early replicating regions. Distinguishing between these possibilities will have to await further experiments to see if centromeres can advance the time of origin activation. 3. How Do These Sequences Exert Their Effect on Origin Activation Time? a. Chromatin Structure and Its Modulation. The telomeric regions in yeast are characterized by their heterochromatin-like structure: the chromatin is condensed as detected by accessibility to probes [189,190], the lysine residues of the histone tails are hypoacetylated, and transcription is repressed (reviewed in Grunstein [191]). Transcriptional silencing at telomeres is a consequence of an altered chromatin structure established by the action of the Sir complex Sir2p-Sir4p [191]. Is some aspect of chromatin structure responsible for the late origin firing in the subtelomeric regions? Some elegant experiments performed by Stevenson and Gottschling [184] have confirmed this hypothesis. Using a mutant strain, these authors demonstrated that Sir3p, a protein known to be a key player in the silenced chromatin structure found at telomeres [191], is responsible for delaying the firing of a subtelomeric origin, Y ’ [184]. However, the advancement of origin firing in the absence of Sir3p does not spread very far away from the telomere, as ARS501, which is located 25 kb away from the telomere, remains late activated [184] (Fig. 5c). This finding is consistent with the observation that the SIR3-dependent silencing of transcription extends only a few kilobases from the chromosome end, presumably because that is as far as the SIR-mediated alteration in chromatin structure extends [192]. The observations just described imply that there are at least two mechanisms by which telomeres aVect origin activation time: one acting via Sir3p

1. The Dynamics of Chromosome Replication in Yeast

43

and one by some other means. Consistent with the idea of a Sir-independent telomere eVect on origin activation time, it has been found that the telomeric Ku complex is involved in the telomeric position eVect on origins. Unlike the sir3 mutant, in which the firing time of ARS501 was unaVected, a ku70 mutant strain showed advanced firing of ARS501 [193] (Fig. 5c). Neither mutation causes a change in the activation time of origins ARS1411–ARS1414 [184,193]. Therefore, the SIR-and Ku-dependent eVect on origin firing time does not extend to all late-firing origins. Indeed, inspection of predicted Sir protein-binding sites in the genome shows no universal correlation between these sites and late-firing origins [19]. Another example linking gene silencing and late origin firing comes from work in which a minimal version of the HMR-E silencer was introduced next to the early firing origin, ARS305 [194]. These authors used the alkaline gel electrophoresis assay described earlier, testing for the appearance of replication intermediates in the presence of hydroxyurea. Origins that normally fire very early in S phase do fire in cells released into S phase in the presence of hydroxyurea, but origins that normally fire near the end of S phase show little or no evidence of firing, as judged by the appearance of replication intermediates in alkaline gels [15]. This inhibition of late origins is not seen in mutant rad53 strains. Zappulla et al. [194] found that no replication intermediates were detectable from the ARS305-(HMR-E) construct in the presence of hydroxyurea, although intermediates were seen in rad53 background. The same result has been found by targeting one of the Sir proteins, Sir4p, to ARS305. The conclusion made from these observations was that targeting of the Sir complex either via the HMR-E silencer or directly by Sir4p is suYcient to reset the firing time of ARS305 from early to late. Both of these targeting methods also led to the transcriptional silencing of APA1, the adjacent gene to ARS305. The more general conclusion is that ARS305 can be made to fire late because of the probable modification of the surrounding chromatin structure induced by the targeting of the Sir proteins [194]. Although this hypothesis is intuitively attractive and obvious, no further investigations about the chromatin structure—condensation level, acetylation level, etc.—were performed in this study. These results also leave open the question of whether there is a causal link between late firing of ARS305 and transcriptional silencing of the adjacent gene. Furthermore, activation of ARS305 in this strain has not been demonstrated directly. It is conceivable that the origin within HMR-E or the early replicated eYcient origin ARS306 located 35 kb away from ARS305 could have contributed to the nascent strands detected. More direct tests, such as 2-D gels or density transfer assays, performed in the absence of hydroxyurea would provide stronger evidence for the resetting of replication firing from early to late by chromatin silencing. Keeping these limitations in mind, these experiments are nevertheless

44

Lucas and Raghuraman

consistent with the idea that chromatin compaction can delay origin activation time. Several studies in mammalian cells seem to confirm this link between chromatin structure and the determination of replication timing. Fluorescent in situ hybridization (FISH) analysis of the replication timing of regions located on chromosome 11 in human cells has shown that treatment with either sodium butyrate or trichostatin A, two histone deacetylase inhibitors, induces an advancement of replication timing of imprinted late regions [195]. Because histone acetylation is associated with more open chromatin, blocking deacetylases should result in more acetylated and therefore more open chromatin; the observation that such treatment advances replication time is therefore again consistent with the idea that chromatin compaction leads to delayed origin firing. More recently, chromatin acetylation has been shown to be one of the determinants of replication timing in yeast [196]. In a mutant rpd3 strain deficient for a histone deacetylase, regions of chromatin showing elevated levels of acetylation also showed earlier firing of origins compared to wild type. The magnitude of the change in replication time showed a rough correlation with the magnitude of the change in the acetylation state. Furthermore, targeting of an acetylase to the late-firing origin ARS1412 specifically advanced its time of firing without aVecting control, nontargeted late origins. b. Is Localization within the Nucleus a Factor? Technological progress in real-time fluorescence microscopy has made possible two studies on chromosome dynamics and, more particularly, on origin localization within the nucleus during the cell cycle [197,198] (for reviews, see Gasser [199] and Cimbora and Groudine [200]). These studies used FISH as well as lacIGFP staining [201] of particular loci to monitor the subnuclear location of various origin and nonorigin sequences at diVerent cell cycle times. They concluded that early firing origins are not constrained to any particular location within the nucleus. In contrast, nontelomeric late-firing origins are localized more preferentially in the periphery of the nucleus. These observations strongly suggested a link between the peripheral nuclear localization of an origin and its late firing. However, even late-firing origins were not exclusively peripheral in their localization; it appears that the dwell time at the nuclear periphery is just greater for late origins than for early firing ones. One way of thinking about these results is that there is a general correlation between the probability of finding an origin sequence at the nuclear periphery and the probability of that origin firing late in S phase. Interestingly, ARS501, which remains late firing in the ensuing S phase following excision as a circle by site-specific recombination in late G1 [183], did not retain its peripheral localization following excision [197]. This observation reinforces

1. The Dynamics of Chromosome Replication in Yeast

45

the idea that although peripheral localization may be important for establishing late origin activation, continued association with the periphery is not necessary for late activation. Along the same lines, another study has shown that localization to the nuclear periphery by itself seems not to be suYcient to cause late firing. To ask if localization to the nuclear periphery per se is suYcient to cause late activation, Sterglanz’ laboratory developed an elegant system that allows tethering of a specific chromatin locus to the nuclear membrane [202]. In this system, a chimeric protein consisting of the endoplasmic reticulum (ER) membrane Yif1p fused to the DNA-binding domain of Gal4p (Gal4-DBD) is overproduced. Because the ER is contiguous with the nuclear membrane, overproduction of the fusion protein results in its diVusing along the membrane and into the inner surface of the nucleus. Consequently, the Gal4DBD protein is exposed at the inner surface of the nucleus, where it can, in principle, bind to target sites in chromosomal DNA and tether those sites to the nuclear periphery. Using this method, it has been shown previously that tethering of a mutationally compromised HMR-E silencer at the nuclear periphery can lead to silencing of a reporter gene inserted next to the crippled silencer [202]. However, in a more recent study, Sternglanz’ laboratory reported that merely tethering the normally early firing origin ARS305 to the nuclear periphery does not lead to its late activation (replication intermediates were detected in the presence of hydroxyurea), nor does it cause silencing of an adjacent expressed gene [194]. It is possible that the contribution of the nuclear periphery in late activation is that it is a zone enriched in factors such as deacetylases and chromatin compaction or remodeling factors so that an origin that is localized to that zone has a greater probability of being compacted or otherwise modified into a ‘‘late’’ chromatin structure. However, one could imagine that additional sequences might be needed for eYcient interaction between the DNA and the chromatin factors; if so, merely tethering an origin to the nuclear periphery may not be suYcient to cause late activation if the origin and its surrounding regions lack the cis-acting elements necessary to interact with the chromatin factors. 4. When Is Origin Firing Time Established? In vivo excision of the normally late-firing origin ARS501 demonstrated that establishment of late firing at ARS501 occurs between mitosis and late G1, coincident with the time of licensing of origins for firing [183]. In these experiments, ARS501 was excised as a circle either in late G1 (-factor block) or in G2/M (nocodazole block), and density transfer timing analysis was done on the excised ARS501 in the ensuing S phase. ARS501 remained late firing if it had been excised in late G1, but became early firing if excised

46

Lucas and Raghuraman

in G2/M, indicating that establishment of the late activation program at ARS501 must occur between G2/M and late G1. A study of the distribution of early and late-replicating regions of Chinese hamster ovary (CHO) nuclei isolated at diVerent stages of the cell cycle and then incubated in Xenopus egg extracts has led to a similar conclusion [203]. 5. Executing the Temporal Program What are the factors implicated in reading and executing the temporal program? The Clb-CDK and Dbf4pCdc7p kinases have been shown to be necessary for the firing of replication and to do it in the order imposed by the temporal program, i.e., they are required for proper execution of the temporal program, but are not themselves responsible for establishing that program (reviewed in Donaldson and Blow [204]). Another kinase—Rad53p, better known for its checkpoint functions—may be implicated in the execution of the temporal program in absence of any triggered checkpoint, as rad53 mutant cells have been reported to show some advancement in the time of origin firing [205]. Consistent with the idea that Rad53 has a role in normal S phase, a genetic interaction has been seen between RAD53 and both CDC7 and DBF4 [206]. a. B-Type Cyclins and Execution of the Temporal Program. The S phase Clb-CDKs, Clb5p and Clb6p in association with Cdc28p, play a role in the initiation of DNA replication. Mutant cells lacking Clb5p have defects in the activation of late origins, whereas early origins show no apparent defect [207]. It appears from these results that Clb6p is capable of activating earlybut not late-firing origins. The fate of origins that normally fire at intermediate times (mid-S) is not known; genome-wide analyses using microarrays should be able to address that question in the near future. When both Clb5p and Clb6p are missing, the origins of replication apparently are still fired in the normal order of the temporal program, but origin firing starts later in the cell cycle than in the wild-type cells [207]. It has been suggested that M phase B-type cyclins, Clb1p, Clb2p, Clb3p, and Clb4p, substitute for the missing Clb5p and Clb6p. However, precocious expression of Clb2p during S phase cannot activate late origins in a clb5 mutant, nor does it activate early origins at wild-type times in a clb5 clb6 mutant [208]. The functional diVerence between Clb5p and Clb6p (i.e., why Clb5p can apparently activate early and late origins and Clb6p can activate only early origins) is not understood. One possibility is that the two proteins have the same targets in the cell but merely diVer in concentration or catalytic activity. Alternatively, they may have diVerent substrate specificities. In any event, the interpretation that Clb5p can activate all origins and that Clb6p can only activate early origins should be treated with caution: only a

1. The Dynamics of Chromosome Replication in Yeast

47

handful of origins have been tested and it is possible that there are regions of the genome where these results do not hold up. b. Cdc7p/Dbf4p. Although Cdc7p was initially thought to be a master switch for the onset of S phase and dispensable thereafter, more recent experiments have shown that Cdc7p is needed not only at the onset of S phase but in fact thoughout S phase for the activation of early and late origins [209,210]. In light of the observation that phosphorylation of targets in the pre-IC is necessary for conversion of the pre-IC to an initiation complex, this finding is now not surprising. However, the finding raises a new question: What prevents Cdc7-Dbf4p from acting prematurely on late origins? Is the recruitment of Cdc45p and its attendant proteins the limiting step? Dissection of the replication program in a mutant mcm5-bob1 strain, in which the need for Cdc7p is bypassed, will surely be informative regarding the role of Cdc7p in the execution of the temporal program.

B. What Is the Physiological Relevance of the Temporal Program? One of the strongest arguments for a biological role of the temporal program of origin activation is that it appears to be conserved through evolution. As described earlier, the existence of such a program is well established in budding yeast. Similar studies are in their infancy in S. pombe [211], and comparisons between budding and fission yeasts in the years ahead should be informative. That Physarum has a program of origin activation has been well documented [212]. Cytological techniques, such as staining with bromodeoxyuridine, have long revealed that higher eukaryotes, including mammals, have a temporal program of replication, and more recently, genomic approaches have begun to yield more detailed views of early- and late-replicating zones in human cells [213]. Despite this wealth of descriptive information, we still have no clear answers as to the function—if any—of the temporal program. Some of the current ideas are outlined here. 1. Regulation of Transcription A strong correlation between transcriptional activity and early replication has been observed in mammalian cells and in Drosophila: genes that are transcriptionally inactive are replicated late in S phase while transcriptionally active regions are early replicating [214,215]. Genes can apparently switch from one state to the other. For example, the FMR1 gene, which is normally transcriptionally active and is replicated at mid-S phase, is silent and replicates late in individuals with the mutation that causes fragile X chromosome syndrome [216]. Similarly, the -globin locus, replicated early in erythroid

48

Lucas and Raghuraman

cells, is late replicating in nonerythroid cells [217]. If the 50 region including the locus control region is deleted (hispanic -thalassemia), this locus is no longer expressed and becomes late replicating [218]. Chromatin studies provide a second link between transcription and replication. As mentioned earlier, chromatin modifications such as acetylation and methylation have been shown to play a role in the chromatin compaction (see Wu and Grunstein [219] for a review). Acetylation is believed to destabilize the condensed chromatin fiber [220], thereby regulating the binding of nonhistone proteins to the DNA. In support of this hypothesis, a histone acetylase activity has been found associated with many transcriptional regulatory proteins in yeast, Drosophila, and humans [221–223]. Work in yeast on the specific acetylation of promoter regions has confirmed the role of this modulation of chromatin structure in the regulation of transcription [224], and chromatin acetylation has been shown to aVect replication timing also [196]. Correlations between transcriptional activity and early replication have led to the popular idea that early replication may be a way of establishing a transcriptionally active state. However, almost no such correlation is seen in yeast [19]. This diVerence between yeast and higher eukaryotes could reflect diVerences in the biology and the evolutionary histories of the diVerent organisms. It is also possible that such a correlation does exist in yeast, but not under the conditions in which almost all experiments to date have been performed. For technical reasons, almost all timing experiments have been done in haploid MATa cells; it is conceivable that replication timing has a much greater significance in other cell types, such as meiotic cells. Even in complex eukaryotes the replication timing of a given origin may be correlated with the structure of the chromatin without implicating transcriptional activity within the region. For example, the removal of just the locus control region of the human -globin locus causes a decrease in the transcription of the locus without any modification of the replication timing or the open chromatin structure [225]. Also, it has been shown in the ICF (immunodeficiency) syndrome [226] and the Roberts syndrome [227] that modifications of the chromatin structure are correlated with an advance in the time of replication without a systematic change in gene expression. Thus, while replication time may show a general correlation with transcriptional activity, the correlation is not absolute. Multiple processes could coexist, some linking both replication and transcription to chromatin structures and others that only link one or the other of these two processes to chromatin. 2. Monitoring the Completion of S Phase It has been proposed that pre-RCs could act as an S phase checkpoint signal that acts as a marker for ongoing replication [228,229]. In this view, every unused pre-RC has some probability of blocking progress into

1. The Dynamics of Chromosome Replication in Yeast

49

anaphase; the presence of pre-RCs at late origins would block premature entry into S phase until those pre-RCs had been converted to post-RCs either by passive replication or by initiation. The observation that cells that fail to assemble pre-RCs can bypass S phase altogether and directly enter mitosis is consistent with this hypothesis [171,230]. DiZey’s group argue against this idea [230]. In this experiment, a thermolabile version of Mcm4p was used to deplete cells of the Mcm complex after they had entered S phase in the presence of hydroxyurea. Under these conditions, the cells did not undergo premature mitosis. If pre-RCs were a checkpoint signal, one might have expected the cells to enter mitosis after the depletion of Mcms (and therefore of pre-RCs). However, this argument does not take into consideration the eVect of hydroxyurea in this experiment. Hydroxyurea treatment induces a checkpoint response, possibly as a consequence of exposed single strands (see Section VIA). Analysis of the experiment is thus complicated by the diYculty in separating the possible role of pre-RCs in a checkpoint that monitors ongoing replication with that of the hydroxyurea-induced checkpoint.

VI. Monitoring the Replication Program The discovery of checkpoints was a major breakthrough in our understanding of the cell cycle [231,232]. Originally considered to be cellular surveillance systems that monitor DNA damage, we now understand that they are part of a more general strategy used by cells to monitor normal cellular events and structures as well (reviewed in [163,233,234]). Importantly for human health, checkpoints have been found to promote genome stability by enforcing the ordered execution of cell cycle events and by monitoring the integrity of intracellular structures such as the spindle (reviewed in [233,235,236]). Genetic and biochemical analyses have revealed several checkpoint pathways, i.e., signals that are recognized at diVerent times in the cell cycle and are transduced in diVerent ways to achieve diVerent eVects. Many of the genes involved are shared between the pathways, and the pathways are often at least partially redundant. The general pattern of checkpoint responses is the same. Sensors such as Rad9p and Pol detect the signal (e.g., DNA damage or stalled replication forks) and transducers such as Mec1p and Rad53p transmit this signal to targets such as RPA and Pds1p to execute the proper response, such as halting the cell cycle. In addition to stopping or slowing the cell cycle, the checkpoint response may include the regulation of gene transcription (e.g., genes coding for ribonucleotide reductase [237]). Many of the genes central to checkpoint systems are highly conserved across evolution. For example, MEC1 in budding yeast is a member of the

50

Lucas and Raghuraman

phosphotidyl inositol3-kinase family and has orthologs in S. pombe (rad3) and humans (ATR). Likewise, budding yeast Rad53p is also a protein kinase with orthologs Cds1p in fission yeast and hChk2p/hCds1p in humans (see [238–240] for reviews). However, downstream targets of these genes may vary depending on the biology of the organism. For example, cell cycle arrest is eVected in S. pombe by regulating the phosphorylation state of Cdc2p (CDK). However, in S. cerevisiae, there is no prolonged G2 phase and events such as spindle formation are already under way during S phase. Therefore, targeting CDK in S. cerevisiae would not be an eVective strategy; instead, the metaphase-to-anaphase transition is blocked via PDS1, which regulates sister chromatid separation at anaphase [241]. Much of what we know about checkpoint signaling cascades comes from the analysis of protein phosphorylation patterns combined with mutational analysis. That Mec1p is a kinase, for example, was inferred from sequence analysis and the presence or absence of phosphorylation on putative targets in mec1 mutant strains. Replication-related checkpoints have been classified into three groups [242]. The G1/S phase checkpoint delays entry of cells into S phase when cells undergo DNA damage in G1 phase. Within S phase, the ‘‘intra-S’’ checkpoint responds to continuous damage to the DNA in S phase by slowing down S phase progression as judged by flow cytometry, but the cells eventually complete S phase and enter mitosis. The ‘‘S/M’’ or ‘‘S phase’’ checkpoint responds to challenges to fork progression by halting S phase and blocking mitosis (reviewed in Kolodner et al. [240]). The intra-S and S/M checkpoints are similar in that they both cause repression of origins that normally fire late in S phase. This response makes intuitive sense: in the face of continuous exposure to DNA-damaging agents or circumstances that prevent fork progression, it does not make sense for the cell to initiate new replication forks. One diVerence between intra-S and S/M checkpoints as noted earlier is that the S/M checkpoint causes a mitosis block whereas the intra-S checkpoint does not. Because checkpoints have been reviewed extensively (see references cited earlier), we focus here on the intra-S and S/M checkpoints, paying particular attention to the question of what the checkpoint signal is and what is regulated as a response.

A. The S Phase Response to DNA Damage and Hydroxyurea 1. The Effect of Hydroxyurea and Methyl Methanesulfonate In the past few years, several studies have been published on S phase progression in the presence of methyl methanesulfonate (MMS; a DNAdamaging agent) or hydroxyurea (inhibitor of ribonucleotide reductase, RNR). Historically, there were distinct motivations behind the use of these

1. The Dynamics of Chromosome Replication in Yeast

51

two types of drugs. MMS is an alkylating agent that modifies DNA; the use of this agent in analyzing how cells cope with lesions arising from such modifications is obvious. Hydroxyurea, by quenching the tyrosyl free radical in the active site of RNR, inhibits the enzyme and thereby depletes the intracellular pool of deoxyribonucleoside triphosphates (dNTPs) [16,243]. The assumption in the past had been that treatment of cells with hydroxyurea causes cells to stall in S phase because replication forks are, in eVect, starved of dNTPs. Accordingly, hydroxyurea has been used as a chemical means of inducing a cell cycle arrest, e.g., to explore whether and how the cell detects ongoing DNA synthesis. Despite the superficial diVerence between these two types of agents, it has become apparent that the cellular responses to them have many points of congruence (see later). One explanation for this similarity is that the MMS-induced lesions are processed to a form that is recognized by a checkpoint system that has elements in common with the S/M system. Additionally, as discussed later, it is also becoming apparent that hydroxyurea does more than just freeze replication forks in place. Therefore, other strategies have to be developed if one is interested in asking how cells monitor ongoing, uninterrupted DNA synthesis. When S. cerevisiae cells are allowed to enter S phase in the presence of MMS, replication is slowed but not stopped [244,245]. This response is manifested as a faster S phase progression in the presence of MMS in mutants defective in various checkpoint genes when compared to wild-type cells [244]. The response to low concentrations of hydroxyurea is similar. However, when cells are released from an -factor arrest in the presence of higher (200 mM) concentrations of hydroxyurea they fail to complete S phase for prolonged periods as judged by flow cytometry and by cell morphology. Both drugs cause loss of viability in rad53 or mec1 mutants but not in wild-type cells [246]. How replication proceeds in the presence of these drugs has been investigated by several diVerent laboratories using diVerent techniques; representative results obtained by three diVerent research groups are summarized. The activation of origins that fire either early or late in S phase has been examined by alkaline gel analysis of nascent DNA strands at or near those origins [10,15]. When wild-type cells were released from an -factor arrest into S phase in the presence of hydroxyurea, nascent strands (manifested as a smear of lower molecular weight DNA in alkaline gels) were seen at the early firing origin ARS305, but were dramatically lower in abundance at the late-firing origin ARS501 [15]. Strikingly, nascent strands were detected at both origins in rad53 and mec1 cells. The conclusion that hydroxyurea blocks late-origin firing in wild-type cells was supported by footprinting analysis of ARS1 at its native location, where it fires early in S phase, compared with its footprint on a linear minichromosome, where it is

52

Lucas and Raghuraman

expected to fire late. The footprint at the early firing ARS1 location changed from a pre-RC to a post-RC pattern in cells released into hydroxyurea, indicative of origin firing. However, ARS1 on the minichromosome retained its pre-RC footprint pattern in cells released into hydroxyurea, indicating that this copy of ARS1 had not fired [15]. Following up on the work described earlier, Tercero and DiZey [247] examined the eVects of MMS on replication. For this study, they used density transfer analysis of specific fragments along the right end of chromosome VI. In the strain used, this region is normally replicated by a fork from the eYcient, early firing origin ARS607 [13]. Replication of restriction fragments in this region was detected as conversion of the fragments from fully dense to hybrid density. The fork migration rate could be deduced from the time of replication of fragments at various distances from the origin. MMS-treated cells of both wild-type and mutant rad53 or mec1 strains showed a 5- to 10-fold reduction in the fork migration rate compared to untreated controls, indicating that the slow S phase in response to MMS treatment is not mediated by the checkpoint genes RAD53 or MEC1. The authors noted two diVerences between wild-type and mutant strains. First, the kinetics of replication of the diVerent restriction fragments indicated that there was increased activation of a subtelomeric origin in mutant compared to wild-type cells in the presence of MMS. This observation is reminiscent of the late-origin firing detected previously in the presence of hydroxyurea in checkpoint-deficient cells (see earlier discussion). A parallel study using 2-D gel electrophoresis to monitor origin firing found largely similar results [205]. Second, despite the slowdown in their migration rate, forks in most wild-type cells continued to progress and eventually replicated the region being examined. In contrast, a fraction of the mutant cell population failed to complete replication of the region, and the amount of unreplicated DNA increased with distance from the origins. An attractive interpretation of these observations that also explains the sensitivity of rad53 and mec1 cells to MMS is that slowed replication forks have some intrinsic probability of collapsing or forming dead end structures unless protected by the action of checkpoints. If two converging forks collapse and there is no licensed origin between them, the intervening DNA will remain unreplicated and would likely lead to chromosome breakage if the cells entered mitosis. These results also argue that the faster S phase progression in MMS-treated rad53 and mec1 cells compared to wild-type cells must be a consequence of the aberrant activation of late origins in the mutant strains. Although the same analysis has not been done with hydroxyurea, two other sets of experiments suggest that hydroxyurea causes a similar fork collapse and could further shed light on the eVect of hydroxyurea. The first series of experiments was based on a 2-D gel analysis of replication of the

1. The Dynamics of Chromosome Replication in Yeast

53

early activated origin ARS305 and its surrounding regions in the presence of hydroxyurea [248]. Replication bubbles were reduced in abundance in rad53 compared to RAD53 cells in hydroxyurea. Furthermore, upon release from the hydroxyurea treatment, wild-type cells completed replication rapidly as judged by the disappearance of replication intermediates and by accumulation of cells with a G2 DNA content. In contrast, replication intermediates persisted in rad53 cells, which that failed to complete replication. The 2-D gels of rad53 cell DNA showed hints of unexpected patterns of hybridization suggestive of aberrant DNA structures, consistent with there being collapsed forks. A more direct examination of replication forks by electron microscopy has proved to be extremely revealing [249]. This study uncovered evidence for the presence of a single-stranded region of 220 nucleotides on one branch at each fork in DNA from untreated wild-type cells. The size of this singlestranded gap increased to 320 nucleotides in the presence of hydroxyurea, perhaps because one nascent strand is preferentially elongated over the other. The increased exposure of single strands could represent some or all of the checkpoint signal that is detected by Mec1p and Rad53p [249]. Indeed, single-stranded regions are known to trigger the G2/M checkpoint response [250]. The size of the bubbles increased with time of incubation of the cells in hydroxyurea, suggesting that the forks could still synthesize DNA at a slowed rate (50 bp/min). The single-stranded region at each fork stayed relatively constant in size while the bubble grew in the presence of hydroxyurea. Replication intermediates from rad53 cells were markedly diVerent. Most of the forks showed extensive regions of single strandedness in the presence of hydroxyurea, with close to half of the bubbles giving the appearance of being hemireplicated, i.e., one branch of the bubble being double stranded and the other appearing single stranded. Unlike with wild-type cells, there was no evidence for growth of the bubbles; instead, the bubbles remained relatively constant in size while the region of single strandedness increased. Additional aberrations in the form of X-shaped structures characteristic of reversed forks were seen in DNA from rad53 cells but not wild-type cells, consistent with the idea that forks collapse and undergo partial reversal when Rad53p is not present to stabilize forks that are impeded. What can we learn from these experiments? First, it appears that at the concentration of hydroxyurea used in these experiments (0.2 M), replication forks are slowed drastically but not arrested [248,249]. Additional evidence to that eVect comes from an experiment that used molecular combing to examine the incorporation of bromodeoxyuridine into newly synthesized DNA [18]. This experiment also showed a drastic reduction in DNA synthesis in the presence of hydroxyurea; nevertheless, there was a small but

54

Lucas and Raghuraman

measurable increase in the amount of label incorporated over time in hydroxyurea, indicating that replication forks had not stalled altogether. Second, the fork slow down seen in the presence of MMS is not mediated by MEC1 or RAD53. Other checkpoints may be involved, or the slowdown is a result of a physical impediment to fork progression brought about by the lesions. Third, Mec1p and Rad53p respond to stalled or slowed forks both by inhibiting late-firing origins and by protecting the forks through some unknown mechanism, preventing their collapse. Forks slowed in the presence of wild-type Mec1p and Rad53p can recover from the drug treatment and continue DNA synthesis. In the absence of these guardian proteins, forks stall irreversibly and form aberrant, dead-end structures, perhaps through a recombination pathway. It is estimated that 16% of all the converging pairs of forks will undergo irreversible collapse after hydroxyurea or MMS treatment unless protected, an amount that is large enough to account for the sensitivity of rad53 and mec1 cells to hydroxyurea and MMS [247]. Informative as these observations are, they still leave several questions unanswered. For practical reasons, the aforementioned experiments have all concentrated on a very small handful of origins, some that are activated near the beginning of S phase and some that are activated close to the end of S phase. While it is clear that the late-firing representatives behave diVerently than the early firing ones in hydroxyurea and MMS, this distinction ignores the reality that many if not most origins are intermediate in their activation time [19]. How do origins that fire around the middle of S phase behave? Are there two qualitatively diVerent classes of origins, those that are repressed by checkpoints and those that are not? The results of two experiments suggest otherwise. One experiment was the footprinting analysis of an early firing copy of ARS1 compared to that of a presumed late-firing copy [15] (see Section VIA,1). In a diVerent experiment, a transcriptional silencer was inserted close to the early firing origin ARS305. This modification changed the origin, converting it from one that fires in hydroxyurea to one that does not, as judged by alkaline gel analysis of nascent DNA [194]. The conclusion from both experiments is that the response to the drug cannot be an intrinsic property of the origin. The proviso with both experiments is that neither the activity nor the firing time of the repressed origin was ascertained directly. Nevertheless, the results as interpreted are inconsistent with the idea that there is an intrinsic, qualitative diVerence between origins that are repressed by checkpoints and origins that are not. A perhaps more plausible model is that there is a stochastic element in whether an origin fires in the presence of hydroxyurea or MMS. In this scenario, the probability that the checkpoint will successfully repress an origin is related to the normal time of activation of that origin: the later the origin normally fires, the greater the probability that it will be repressed. Consequently, the

1. The Dynamics of Chromosome Replication in Yeast

55

outcome for a population of cells would be that early firing origins would be scored as being active whereas late-firing ones would not. The recent genome-wide marker frequency analysis of the replication program is particularly relevant in this regard [21]. This study concluded that there are two overlapping waves of origin initiations, and that origins that are repressed by hydroxyurea treatment correspond to the second wave of initiations. However, the study did not quantify the level of activity at each origin; future studies will certainly attempt to address that question. 2. The Response Pathway Early steps in the detection and response to DNA damage involve the RFC-like complex Rad24p-Rfc2p-5p and the PCNA-like complex Rad17p-Ddc1p-Mec3p [238,239]. Mec1p is also recruited to the lesion by its pairing partner Lcd1p [251]. Because of the putative similarity between the Rad24p complex and the Rad17p complex to RFC and PCNA, respectively, it is attractive to think that they are the primary sensors of lesions. However, some doubt has been cast on this idea by the observation that certain Mec1p targets, such as Lcd1p, are phosphorylated in response to DNA damage even in the absence of these complexes. The sensors for response to hydroxyurea include Pol and Rad24p [252,253]. Again, the contributions of these two proteins is not clear-cut. Pol apparently participates in the response when the cells are treated with hydroxyurea at 0.15 M [252] but not at 0.2 M [253]. It has been suggested that replication generates some component of the signal and that not enough replication occurs at the higher hydroxyurea concentration for Pol to be a factor [253]. In any event, without a systematic analysis of responses to diVerent concentrations of hydroxyurea, it is diYcult to make meaningful conclusions about the diVerent experimental regimes. One of the central downstream targets of Mec1p is Rad53p (Fig. 6). Rad53p is capable of autophosphorylation. However, it is phosphorylated even in mutants making only a defective, kinase-null version of Rad53p, but it is not phosphorylated in mec1 mutant strains [253,254]. Mec1-dependent phosphorylation of Rad53p is mediated by Mrc1p [255]. In S. pombe, Mrc1p is synthesized only during S phase [256]; it remains to be seen if S. cerevisiae Mrc1p shows a similar pattern of synthesis. Immunofluorescence studies find that Rad53p colocalizes with ORC during S phase and that the RecQ-like helicase Sgs1p plays a role in this localization [257]. This association occurs even in normal cells not subjected to replicational stress and may represent some function of the checkpoint system in monitoring or regulating ongoing replication. Activated Rad53p has a diverse set of downstream targets representing a variety of response pathways. Members of the replication initiation complex

56

Lucas and Raghuraman

Figure 6 Checkpoints monitoring replication. Impediment to fork progression (either from DNA damage or from hydroxyurea) is indicated by an asterisk. A fork that is impeded triggers a checkpoint response that results in phosphorylation of Rad53p.

show Rad53p-dependent phosphorylation in response to MMS or hydroxyurea. These targets include RPA [258] and Pol-primase B subunit [259], which may be phosphorylated indirectly via CDK [253]. Because phosphorylation of Dpb11p by CDK is a regulatory step in origin firing, mediation of the checkpoint response by CDK may play a role in the repression of late origins. Phosphorylated Rad53p also evokes a transcriptional response, upregulating the expression of genes coding for ribonucleotide reductase (RNR). Transcription of RNR genes is repressed by SML1; in response to the checkpoint signal Rad53p activates Dun1p, which blocks Sml1p and thereby activates the transcription of RNR [243,260]. It has been suggested that Cdc7pDbf4p is also a target of the checkpoint pathway, in part because of the observed sensitivity of cdc7 mutant cells to

1. The Dynamics of Chromosome Replication in Yeast

57

DNA-damaging agents [261]. However, the mcm5-bob1 mutant (which bypasses the need for Cdc7p) is nevertheless checkpoint proficient [210], indicating that Cdc7p plays at best a nonessential role in checkpoint function. A fourth response that may be triggered is to block the metaphase-toanaphase transition. The primary target (either directly or indirectly) is likely to be Pds1p; it is not clear how this response is engaged when the signal is hydroxyurea but not when it is MMS. Also, activation of this branch may bypass Rad53p altogether and be carried out by Mec1p [262]. In addition, complete arrest also requires functioning of an independent pathway dependent on Chk1p [263]. It has been discovered that Rad53p associates with the chromatin assembly factor Asf1p [264,265]; part of the checkpoint response therefore may be to regulate the assembly of chromatin in newly replicated DNA. Although we often treat checkpoint responses as being all or none, we should keep in mind the probable stochastic nature of the response, i.e., that a lesion has a certain probability of activating a checkpoint, which in turn has some probability of achieving a response. This idea is reinforced by the observation that some pathways show partial redundancy. For example, both Rad53p- and Chk1p-mediated pathways have to be eliminated in order to get complete elimination of the G2/M delay after DNA damage in S phase [263]. Failure of one of these pathways gives the result that some members of a genetically homogeneous population execute the arrest while others do not. A simple interpretation of such a result is that the target event has some probability of escaping each checkpoint pathway; when both pathways are functioning, the probability of escaping both is very low.

B. Detecting Ongoing, Uninterrupted Replication What keeps a cell from entering mitosis before S phase is complete? Does the cell monitor the presence of replication forks or initiation complexes? The S phase checkpoint described earlier was elucidated using hydroxyurea to prolong S phase. However, it is clear that drugs such as hydroxyurea do not simply freeze a normal S phase; replication forks are structurally diVerent in hydroxyurea-treated cells compared to untreated cells [249], and therefore may be eliciting diVerent or additional checkpoints. One experiment designed to ask if progress through a natural S phase is monitored by cells made use of a yeast artificial chromosome (YAC) containing a 170-kb region within which eYcient origins had been eliminated [266,267]. Replication of this chromosome therefore depended predominantly on a single replication fork traversing the 170-kb origin-free region. If cells cannot monitor ongoing replication, one might expect the cells to

58

Lucas and Raghuraman

enter mitosis before replication of the YAC was complete. Density transfer analysis of replication time showed that replication of this YAC was not completed until well after a normal S phase would end. However, the YAC was maintained as eYciently as a wild-type version that had all its origins intact, but the cell population doubling time in the presence of the origin-deficient YAC was increased [266,267]. Maintenance of the YAC was dependent on RAD9: in mutant rad9 cells, the YAC was either lost (in the absence of selection) or rearranged (when selection for both YAC ends was imposed) such that the size of the origin-free region was reduced substantially [267]. This result indicates that the cell does indeed monitor an ongoing ‘‘normal’’ S phase. Furthermore, because RAD9 was not known to be a major component of any of the S phase checkpoints described previously, this experiment may have uncovered a hitherto unknown checkpoint pathway. What aspect of a normal S phase might a cell monitor? Several ideas present themselves.  The cell could be detecting the presence of unreplicated chromatin. As mentioned earlier, the amount of chromatin-associated Mcm proteins is in vast excess over the number of predicted origins in the yeast genome. Perhaps the excess Mcm complexes mark unreplicated DNA and are displaced from chromatin as a replication fork passes through. The observation that elimination of Mcms in mid-S does not result in premature entry into mitosis argues against this idea [149,230].  The signal for incomplete S phase could be the presence of intact preRCs [228,229]. Consistent with this idea, in the absence of Cdc6p, cells can bypass S phase altogether and directly enter a catastrophic mitosis [171]. Again, however, the observation that the elimination of Mcm proteins during S phase does not lead to premature mitosis argues against this idea [149,230]—elimination of Mcms should presumably also destroy pre-RCs that have formed already, and yet the checkpoint appears to be intact.  Some aspect of an active replication fork could be monitored. There is some evidence from a Xenopus in vitro system that the synthesis of RNA primers by Pol-primase contributes to the checkpoint [268]. One potential complication is that there may be multiple, redundant checkpoints, making clean tests of hypotheses diYcult. For example, the Mcm destruction experiments [149,230] used hydroxyurea to block the cells in S phase prior to destruction of the Mcm proteins, raising the possibility that the checkpoint that was seen to function was not the normal S phase checkpoint but the hydroxyurea response checkpoint. Furthermore, the Mcm complex is needed for normal fork progression [149]; if the Mcm complex at a fork is disassembled, the fork may collapse or form aberrant structures, which may also trigger a checkpoint response. Because of

1. The Dynamics of Chromosome Replication in Yeast

59

such complications, none of the aforementioned possibilities for the normal S-phase checkpoint signal can be eliminated unambiguously as yet.

VII. Concluding Thoughts The past decade has been something of a golden age for the study of yeast chromosome replication. The advances to date have mainly explored S phase in haploid yeast cells of the a mating type. We know little or nothing about S phases in other cell types (such as diploids) or germinating spores. Nevertheless, what we have learned from yeast has had a tremendous impact on our understanding of the dynamics of replication in higher eukaryotes as well. The recent advent of whole-genome technologies such as microarrays and molecular combing will help prolong this golden age and should be invaluable in addressing several remaining puzzles. A big advantage of these approaches is that they allow unbiased surveys of the whole genome, as opposed to picking one or a few origins to examine under some particular condition (e.g., in a mutant strain) and then generalizing from these few examples. Furthermore, with databases of origins and their activation properties (eYciency, firing time) being compiled, we are closer to being able to apply bioinformatics tools to understanding the sequence contributions behind origin function. At the same time, single cell techniques such as molecular combing hold the prospect of asking questions that would otherwise not be feasible, e.g., whether there is an obligate order of firing of origins. When it comes to the question of the spatial organization of the genome within the nucleus, we are just beginning to scratch the surface. A challenge for the future will be to explore to what extent such organization exists and how it is integrated with the various functions of the chromosome: DNA replication, segregation, transcription, recombination, and repair.

Acknowledgments We thank Walt Fangman, Bonny Brewer, Margaret Hoang, and Gina Alvino for their very helpful comments. Work in the Fangman-Brewer laboratory is supported by NIH Grants GM18926 and CA77852.

References 1. Newlon, C. S., and Burke, W. (1980). Replication of small chromosomal DNAs in yeast In ‘‘ICN-UCLA Symposium on Molecular and Cellular Biology,’’ pp. 399–409. 2. Rivin, C. J., and Fangman, W. L. (1980). Replication fork rate and origin activation during the S phase of Saccharomyces cerevisiae J. Cell Biol. 85, 108–115.

60

Lucas and Raghuraman

3. Stinchcomb, D. T., Struhl, K., and Davis, R. W. (1979). Isolation and characterization of a yeast chromosomal replicator Nature 282, 39–43. 4. Struhl, K., Stinchcomb, D. T., Scherer, S., and Davis, R. W. (1979). High-frequency transformation of yeast: Autonomous replication of hybrid DNA molecules Proc. Natl. Acad. Sci. USA 76, 1035–1039. 5. Stinchcomb, D. T., Thomas, M., Kelly, J., Selker, E., and Davis, R. W. (1980). Eukaryotic DNA segments capable of autonomous replication in yeast Proc. Natl. Acad. Sci. USA 77, 4559–4563. 6. Newlon, C. S., and Theis, J. F. (1993). The structure and function of yeast ARS elements Curr. Opin. Genet. Dev. 3, 752–758. 7. Brewer, B. J., and Fangman, W. L. (1987). The localization of replication origins on ARS plasmids in S. cerevisiae Cell 51, 463–471. 8. Huberman, J. A., Spotila, L. D., Nawotka, K. A., el-Assouli, S. M., and Davis, L. R. (1987). The in vivo replication origin of the yeast 2 micron plasmid Cell 51, 473–481. 9. Newlon, C. S., Collins, I., Dershowitz, A., Deshpande, A. M., Greenfeder, S. A., Ong, L. Y., and Theis, J. F. (1993). Analysis of replication origin function on chromosome III of Saccharomyces cerevisiae Cold Spring Harb. Symp. Quant. Biol. 58, 415–423. 10. Santocanale, C., Sharma, K., and Diffley, J. F. (1999). Activation of dormant origins of DNA replication in budding yeast Genes Dev. 13, 2360–2364. 11. Friedman, K. L., and Brewer, B. J. (1995). Analysis of replication intermediates by twodimensional agarose gel electrophoresis Methods Enzymol. 262, 613–627. 12. van Brabant, A. J., and Raghuraman, M. K. (2002). Assaying replication fork direction and migration rates Methods Enzymol. 351, 539–568. 13. Friedman, K. L., Brewer, B. J., and Fangman, W. L. (1997). Replication profile of Saccharomyces cerevisiae chromosome VI Genes Cells 2, 667–678. 14. Nawotka, K. A., and Huberman, J. A. (1988). Two-dimensional gel electrophoretic method for mapping DNA replicons Mol. Cell. Biol. 8, 1408–1413. 15. Santocanale, C., and Diffley, J. F. (1998). A Mec1- and Rad53-dependent checkpoint controls late-firing origins of DNA replication Nature 395, 615–618. 16. Yarbro, J. W. (1992). Mechanism of action of hydroxyurea Semin. Oncol. 19, 1–10. 17. Herrick, J., and Bensimon, A. (1999). Single molecule analysis of DNA replication Biochimie. 81, 859–871. 18. Lengronne, A., Pasero, P., Bensimon, A., and Schwob, E. (2001). Monitoring S phase progression globally and locally using BrdU incorporation in TK(+) yeast strains Nucleic Acids Res. 29, 1433–1442. 19. Raghuraman, M. K., Winzeler, E. A., Collingwood, D., Hunt, S., Wodicka, L., Conway, A., Lockhart, D. J., Davis, R. W., Brewer, B. J., and Fangman, W. L. (2001). Replication dynamics of the yeast genome Science 294, 115–121. 20. Khodursky, A. B., Peter, B. J., Schmid, M. B., DeRisi, J., Botstein, D., Brown, P. O., and Cozzarelli, N. R. (2000). Analysis of topoisomerase function in bacterial replication fork movement: Use of DNA microarrays Proc. Natl. Acad. Sci. USA 97, 9419–9424. 21. Yabuki, N., Terashima, H., and Kitada, K. (2002). Mapping of early firing origins on a replication profile of budding yeast Genes Cells 7, 781–789. 22. Ren, B., Robert, F., Wyrick, J. J., Aparicio, O., Jennings, E. G., Simon, I., Zeitlinger, J., Schreiber, J., Hannett, N., Kanin, E., Volkert, T. L., Wilson, C. J., Bell, S. P., and Young, R. A. (2000). Genome-wide location and function of DNA binding proteins Science 290, 2306–2309. 23. Wyrick, J. J., Aparicio, J. G., Chen, T., Barnett, J. D., Jennings, E. G., Young, R. A., Bell, S. P., and Aparicio, O. M. (2001). Genome-wide distribution of ORC and MCM proteins in S. cerevisiae: High-resolution mapping of replication origins Science 294, 2357–2360. 24. Bell, S. P., Kobayashi, R., and Stillman, B. (1993). Yeast origin recognition complex functions in transcription silencing and DNA replication Science 262, 1844–1849.

1. The Dynamics of Chromosome Replication in Yeast

61

25. Micklem, G., Rowley, A., Harwood, J., Nasmyth, K., and Diffley, J. F. (1993). Yeast origin recognition complex is involved in DNA replication and transcriptional silencing Nature 366, 87–89. 26. Foss, M., McNally, F. J., Laurenson, P., and Rine, J. (1993). Origin recognition complex (ORC) in transcriptional silencing and DNA replication in S. cerevisiae Science 262, 1838–1844. 27. Newlon, C. S., Lipchitz, L. R., Collins, I., Deshpande, A., Devenish, R. J., Green, R. P., Klein, H. L., Palzkill, T. G., Ren, R. B., Synn, S., et al. (1991). Analysis of a circular derivative of Saccharomyces cerevisiae chromosome III: A physical map and identification and location of ARS elements [published erratum appears in Genetics 130 (1), 235 (1992)] Genetics 129, 343–357. 28. Maine, G. T., Sinha, P., and Tye, B. K. (1984). Mutants of S. cerevisiae defective in the maintenance of minichromosomes Genetics 106, 365–385. 29. Pasero, P., Bensimon, A., and Schwob, E. (2002). Single-molecule analysis reveals clustering and epigenetic regulation of replication origins at the yeast rDNA locus Genes Dev. 16, 2479–2484. 30. Jacob, F., Brenner, S., and Cuzin, F. (1963). On the regulation of DNA replication in bacteria Cold Spring Harb. Symp. Quant. Biol. 28, 329–348. 31. Gerbi, S. A., and Bielinsky, A. K. (1997). Replication initiation point mapping Methods 13, 271–280. 32. Kearsey, S. (1983). Analysis of sequences conferring autonomous replication in baker’s yeast EMBO J. 2, 1571–1575. 33. Kearsey, S. (1984). Structural requirements for the function of a yeast chromosomal replicator Cell 37, 299–307. 34. Broach, J. R., Li, Y. Y., Feldman, J., Jayaram, M., Abraham, J., Nasmyth, K. A., and Hicks, J. B. (1983). Localization and sequence analysis of yeast origins of DNA replication Cold Spring Harb. Symp. Quant. Biol. 2, 1165–1173. 35. Bouton, A. H., and Smith, M. M. (1986). Fine-structure analysis of the DNA sequence requirements for autonomous replication of Saccharomyces cerevisiae plasmids Mol. Cell. Biol. 6, 2354–2363. 36. Marahrens, Y., and Stillman, B. (1992). A yeast chromosomal origin of DNA replication defined by multiple functional elements Science 255, 817–823. 37. Celniker, S. E., Sweder, K., Srienc, F., Bailey, J. E., and Campbell, J. L. (1984). Deletion mutations affecting autonomously replicating sequence ARS1 of Saccharomyces cerevisiae Mol. Cell. Biol. 4, 2455–2466. 38. Walker, S. S., Francesconi, S. C., and Eisenberg, S. (1990). A DNA replication enhancer in Saccharomyces cerevisiae Proc. Natl. Acad. Sci. USA 87, 4665–4669. 39. Shirahige, K., Iwasaki, T., Rashid, M. B., Ogasawara, N., and Yoshikawa, H. (1993). Location and characterization of autonomously replicating sequences from chromosome VI of Saccharomyces cerevisiae Mol. Cell. Biol. 13, 5043–5056. 40. Van Houten, J. V., and Newlon, C. S. (1990). Mutational analysis of the consensus sequence of a replication origin from yeast chromosome III Mol. Cell. Biol. 10, 3917–3925. 41. Rattray, A. J., and Symington, L. S. (1993). Stimulation of meiotic recombination in yeast by an ARS element Genetics 134, 175–188. 42. Theis, J. F., and Newlon, C. S. (1997). The ARS309 chromosomal replicator of Saccharomyces cerevisiae depends on an exceptional ARS consensus sequence Proc. Natl. Acad. Sci. USA 94, 10786–10791. 43. Schneider, T. D., and Stephens, R. M. (1990). Sequence logos: A new way to display consensus sequences Nucleic Acids Res. 18, 6097–6100. 44. Schneider, T. D. (1997). Information content of individual genetic sequences J. Theor. Biol. 189, 427–441.

62

Lucas and Raghuraman

45. Schneider, T. D. (1997). Sequence walkers: A graphical method to display how binding proteins interact with DNA or RNA sequences Nucleic Acids Res. 25, 4408–4415. 46. Theis, J. F., and Newlon, C. S. (2001). Two compound replication origins in Saccharomyces cerevisiae contain redundant origin recognition complex binding sites Mol. Cell. Biol. 21, 2790–2801. 47. Miller, C. A., and Kowalski, D. (1993). Cis-acting components in the replication origin from ribosomal DNA of Saccharomyces cerevisiae Mol. Cell. Biol. 13, 5360–5369. 48. Bell, S. P., and Stillman, B. (1992). ATP-dependent recognition of eukaryotic origins of DNA replication by a multiprotein complex Nature 357, 128–134. 49. Rao, H., and Stillman, B. (1995). The origin recognition complex interacts with a bipartite DNA binding site within yeast replicators Proc. Natl. Acad. Sci. USA 92, 2224–2228. 50. Lee, D. G., and Bell, S. P. (1997). Architecture of the yeast origin recognition complex bound to origins of DNA replication Mol. Cell. Biol. 17, 7159–7168. 51. Huang, R. Y., and Kowalski, D. (1993). A DNA unwinding element and an ARS consensus comprise a replication origin within a yeast chromosome EMBO J. 12, 4521–4531. 52. Umek, R. M., and Kowalski, D. (1990). Thermal energy suppresses mutational defects in DNA unwinding at a yeast replication origin Proc. Natl. Acad. Sci. USA 87, 2486–2490. 53. Bielinsky, A. K., and Gerbi, S. A. (1999). Chromosomal ARS1 has a single leading strand start site Mol. Cell 3, 477–486. 54. Wilmes, G. M., and Bell, S. P. (2002). The B2 element of the Saccharomyces cerevisiae ARS1 origin of replication requires specific sequences to facilitate pre-RC formation Proc. Natl. Acad. Sci. USA 99, 101–106. 55. Zou, L., and Stillman, B. (2000). Assembly of a complex containing Cdc45p, replication protein A, and Mcm2p at replication origins controlled by S-phase cyclin-dependent kinases and Cdc7p-Dbf4p kinase Mol. Cell. Biol. 20, 3086–3096. 56. Huang, R. Y., and Kowalski, D. (1996). Multiple DNA elements in ARS305 determine replication origin activity in a yeast chromosome Nucleic Acids Res. 24, 816–823. 57. Lin, S., and Kowalski, D. (1997). Functional equivalency and diversity of cis-acting elements among yeast replication origins Mol. Cell. Biol. 17, 5473–5484. 58. Raychaudhuri, S., Byers, R., Upton, T., and Eisenberg, S. (1997). Functional analysis of a replication origin from Saccharomyces cerevisiae: identification of a new replication enhancer Nucleic Acids Res. 25, 5057–5064. 59. Okuno, Y., Satoh, H., Sekiguchi, M., and Masukata, H. (1999). Clustered adenine/thymine stretches are essential for function of a fission yeast replication origin Mol. Cell. Biol. 19, 6699–6709. 60. Kong, D., and DePamphilis, M. L. (2001). Site-specific DNA binding of the Schizosaccharomyces pombe origin recognition complex is determined by the Orc4 subunit Mol. Cell. Biol. 21, 8095–8103. 61. Mahbubani, H. M., Paull, T., Elder, J. K., and Blow, J. J. (1992). DNA replication initiates at multiple sites on plasmid DNA in Xenopus egg extracts Nucleic Acids Res. 20, 1457–1462. 62. Hyrien, O., and Mechali, M. (1992). Plasmid replication in Xenopus eggs and egg extracts: A 2D gel electrophoretic analysis Nucleic Acids Res. 20, 1463–1469. 63. Caddle, M. S., and Calos, M. P. (1992). Analysis of the autonomous replication behavior in human cells of the dihydrofolate reductase putative chromosomal origin of replication Nucleic Acids Res. 20, 5971–5978. 64. Zannis-Hadjopoulos, M., and Price, G. B. (1999). Eukaryotic DNA replication J. Cell. Biochem.Suppl. 1–14. 65. Petes, T. D., and Williamson, D. H. (1975). Fiber autoradiography of replicating yeast DNA Exp. Cell Res. 95, 103–110.

1. The Dynamics of Chromosome Replication in Yeast

63

66. Poloumienko, A., Dershowitz, A., De, J., and Newlon, C. S. (2001). Completion of replication map of Saccharomyces cerevisiae chromosome III Mol. Biol. Cell 12, 3317–3327. 67. Hyrien, O., Maric, C., and Mechali, M. (1995). Transition in specification of embryonic metazoan DNA replication origins Science 270, 994–997. 68. Brewer, B. J. (1994). Intergenic DNA and the sequence requirements for replication initiation in eukaryotes Curr. Opin. Genet. Dev. 4, 196–202. 69. Snyder, M., Sapolsky, R. J., and Davis, R. W. (1988). Transcription interferes with elements important for chromosome maintenance in Saccharomyces cerevisiae Mol. Cell. Biol. 8, 2184–2194. 70. Brewer, B. J. (1988). When polymerases collide: Replication and the transcriptional organization of the E. coli chromosome Cell 53, 679–686. 71. Kipling, D., and Kearsey, S. E. (1990). Reversion of autonomously replicating sequence mutations in Saccharomyces cerevisiae: Creation of a eucaryotic replication origin within procaryotic vector DNA Mol. Cell. Biol. 10, 265–272. 72. Zweifel, S. G., and Fangman, W. L. (1990). Creation of ARS activity in yeast through iteration of non-functional sequences Yeast 6, 179–186. 73. Diffley, J. F., Cocker, J. H., Dowell, S. J., and Rowley, A. (1994). Two steps in the assembly of complexes at yeast replication origins in vivo Cell 78, 303–316. 74. Donovan, S., Harwood, J., Drury, L. S., and Diffley, J. F. (1997). Cdc6p-dependent loading of Mcm proteins onto pre-replicative chromatin in budding yeast Proc. Natl. Acad. Sci. USA 94, 5611–5616. 75. Diffley, J. F., and Cocker, J. H. (1992). Protein-DNA interactions at a yeast replication origin Nature 357, 169–172. 76. Santocanale, C., and Diffley, J. F. (1996). ORC- and Cdc6-dependent complexes at active and inactive chromosomal replication origins in Saccharomyces cerevisiae EMBO J. 15, 6671–6679. 77. Brewer, B. J., and Fangman, W. L. (1993). Initiation at closely spaced replication origins in a yeast chromosome Science 262, 1728–1731. 78. Brewer, B. J., and Fangman, W. L. (1994). Initiation preference at a yeast origin of replication Proc. Natl. Acad. Sci. USA 91, 3418–3422. 79. Marahrens, Y., and Stillman, B. (1994). Replicator dominance in a eukaryotic chromosome EMBO J. 13, 3395–3400. 80. Dutta, A., and Bell, S. P. (1997). Initiation of DNA replication in eukaryotic cells Annu. Rev. Cell Dev. Biol. 13, 293–332. 81. Bell, S. P., and Dutta, A. (2002). DNA replication in eukaryotic cells Annu. Rev. Biochem. 71, 333–374. 82. Klemm, R. D., Austin, R. J., and Bell, S. P. (1997). Coordinate binding of ATP and origin DNA regulates the ATPase activity of the origin recognition complex Cell 88, 493–502. 83. Klemm, R. D., and Bell, S. P. (2001). ATP bound to the origin recognition complex is important for preRC formation Proc. Natl. Acad. Sci. USA 98, 8361–8367. 84. Walker, J. E., Saraste, M., Runswick, M. J., and Gay, N. J. (1982). Distantly related sequences in the alpha- and beta-subunits of ATP synthase, myosin, kinases and other ATP-requiring enzymes and a common nucleotide binding fold EMBO J. 1, 945–951. 85. Koonin, E. V. (1993). A common set of conserved motifs in a vast variety of putative nucleic acid-dependent ATPases including MCM proteins involved in the initiation of eukaryotic DNA replication Nucleic Acids Res. 21, 2541–2547. 86. Zhang, Y., Yu, Z., Fu, X., and Liang, C. (2002). Noc3p, a bHLH protein, plays an integral role in the initiation of DNA replication in budding yeast Cell 109, 849–860. 87. Hartwell, L. H. (1973). Three additional genes required for deoxyribonucleic acid synthesis in Saccharomyces cerevisiae J. Bacteriol. 115, 966–974.

64

Lucas and Raghuraman

88. Liang, C., Weinreich, M., and Stillman, B. (1995). ORC and Cdc6p interact and determine the frequency of initiation of DNA replication in the genome Cell 81, 667–676. 89. Detweiler, C. S., and Li, J. J. (1997). Cdc6p establishes and maintains a state of replication competence during G1 phase J. Cell Sci. 110, 753–763. 90. Tanaka, T., Knapp, D., and Nasmyth, K. (1997). Loading of an Mcm protein onto DNA replication origins is regulated by Cdc6p and CDKs Cell 90, 649–660. 91. Zwerschke, W., Rottjakob, H. W., and Kuntzel, H. (1994). The Saccharomyces cerevisiae CDC6 gene is transcribed at late mitosis and encodes a ATP/GTPase controlling S phase initiation J. Biol. Chem. 269, 23351–23356. 92. Weinreich, M., Liang, C., and Stillman, B. (1999). The Cdc6p nucleotide-binding motif is required for loading mcm proteins onto chromatin Proc. Natl. Acad. Sci. USA 96, 441–446. 93. Perkins, G., and Diffley, J. F. (1998). Nucleotide-dependent prereplicative complex assembly by Cdc6p, a homolog of eukaryotic and prokaryotic clamp-loaders Mol. Cell 2, 23–32. 94. Takahashi, N., Tsutsumi, S., Tsuchiya, T., Stillman, B., and Mizushima, T. (2002). Functions of sensor 1 and sensor 2 regions of Saccharomyces cerevisiae Cdc6p in vivo and in vitro J. Biol. Chem. 277, 16033–16040. 95. Mizushima, T., Takahashi, N., and Stillman, B. (2000). Cdc6p modulates the structure and DNA binding activity of the origin recognition complex in vitro Genes Dev. 14, 1631–1641. 96. Hua, X. H., and Newport, J. (1998). Identification of a preinitiation step in DNA replication that is independent of origin recognition complex and cdc6, but dependent on cdk2 J. Cell Biol. 140, 271–281. 97. Lengronne, A., and Schwob, E. (2002). The yeast CDK inhibitor Sic1 prevents genomic instability by promoting replication origin licensing in late G(1) Mol. Cell 9, 1067–1078. 98. Rialland, M., Sola, F., and Santocanale, C. (2002). Essential role of human CDT1 in DNA replication and chromatin licensing J. Cell Sci. 115, 1435–1440. 99. Maiorano, D., Moreau, J., and Mechali, M. (2000). XCDT1 is required for the assembly of pre-replicative complexes in Xenopus laevis Nature 404, 622–625. 100. Nishitani, H., Lygerou, Z., Nishimoto, T., and Nurse, P. (2000). The Cdt1 protein is required to license DNA for replication in fission yeast Nature 404, 625–628. 101. Tanaka, S., and Diffley, J. F. (2002). Interdependent nuclear accumulation of budding yeast Cdt1 and Mcm2–7 during G1 phase Nature Cell Biol. 4, 198–207. 102. Devault, A., Vallen, E. A., Yuan, T., Green, S., Bensimon, A., and Schwob, E. (2002). Identification of Tah11/Sid2 as the ortholog of the replication licensing factor Cdt1 in Saccharomyces cerevisiae Curr. Biol. 12, 689–694. 103. Moir, D., Stewart, S. E., Osmond, B. C., and Botstein, D. (1982). Cold-sensitive celldivision-cycle mutants of yeast: Isolation, properties, and pseudoreversion studies Genetics 100, 547–563. 104. Lei, M., and Tye, B. K. (2001). Initiating DNA synthesis: From recruiting to activating the MCM complex J. Cell Sci. 114, 1447–1454. 105. Lei, M., Kawasaki, Y., and Tye, B. K. (1996). Physical interactions among Mcm proteins and effects of Mcm dosage on DNA replication in Saccharomyces cerevisiae Mol. Cell. Biol. 16, 5081–5090. 106. Labib, K., Diffley, J. F., and Kearsey, S. E. (1999). G1-phase and B-type cyclins exclude the DNA-replication factor Mcm4 from the nucleus Nature Cell Biol. 1, 415–422. 107. Ishimi, Y. (1997). A DNA helicase activity is associated with an MCM4, -6, and -7 protein complex J. Biol. Chem. 272, 24508–24513. 108. Labib, K., and Diffley, J. F. (2001). Is the MCM2-7 complex the eukaryotic DNA replication fork helicase? Curr. Opin. Genet. Dev. 11, 64–70.

1. The Dynamics of Chromosome Replication in Yeast

65

109. Waga, S., and Stillman, B. (1998). The DNA replication fork in eukaryotic cells Annu. Rev. Biochem. 67, 721–751. 110. Sato, M., Gotow, T., You, Z., Komamura-Kohno, Y., Uchiyama, Y., Yabuta, N., Nojima, H., and Ishimi, Y. (2000). Electron microscopic observation and single-stranded DNA binding activity of the Mcm4,6,7 complex J. Mol. Biol. 300, 421–431. 111. Kelman, Z., Lee, J. K., and Hurwitz, J. (1999). The single minichromosome maintenance protein of Methanobacterium thermoautotrophicum DeltaH contains DNA helicase activity Proc. Natl. Acad. Sci. USA 96, 14783–14788. 112. Chong, J. P., Hayashi, M. K., Simon, M. N., Xu, R. M., and Stillman, B. (2000). A double-hexamer archaeal minichromosome maintenance protein is an ATP-dependent DNA helicase Proc. Natl. Acad. Sci. USA 97, 1530–1535. 113. Shechter, D. F., Ying, C. Y., and Gautier, J. (2000). The intrinsic DNA helicase activity of Methanobacterium thermoautotrophicum delta H minichromosome maintenance protein J. Biol. Chem. 275, 15049–15059. 114. Yu, X., VanLoock, M. S., Poplawski, A., Kelman, Z., Xiang, T., Tye, B. K., and Egelman, E. H. (2002). The Methanobacterium thermoautotrophicum MCM protein can form heptameric rings EMBO Rep. 3, 792–797. 115. Aparicio, O. M., Weinstein, D. M., and Bell, S. P. (1997). Components and dynamics of DNA replication complexes in S. cerevisiae: Redistribution of MCM proteins and Cdc45p during S phase Cell 91, 59–69. 116. Young, M. R., and Tye, B. K. (1997). Mcm2 and Mcm3 are constitutive nuclear proteins that exhibit distinct isoforms and bind chromatin during specific cell cycle stages of Saccharomyces cerevisiae Mol. Biol. Cell 8, 1587–1601. 117. Edwards, M. C., Tutter, A. V., Cvetic, C., Gilbert, C. H., Prokhorova, T. A., and Walter, J. C. (2002). MCM2-7 complexes bind chromatin in a distributed pattern surrounding ORC in Xenopus egg extracts J. Biol. Chem. 26, 26. 118. Essers, J., Houtsmuller, A. B., van Veelen, L., Paulusma, C., Nigg, A. L., Pastink, A., Vermeulen, W., Hoeijmakers, J. H., and Kanaar, R. (2002). Nuclear dynamics of RAD52 group homologous recombination proteins in response to DNA damage EMBO J. 21, 2030–2037. 119. Gibson, S. I., Surosky, R. T., and Tye, B. K. (1990). The phenotype of the minichromosome maintenance mutant mcm3 is characteristic of mutants defective in DNA replication Mol. Cell. Biol. 10, 5707–5720. 120. Yan, H., Merchant, A. M., and Tye, B. K. (1993). Cell cycle-regulated nuclear localization of MCM2 and MCM3, which are required for the initiation of DNA synthesis at chromosomal replication origins in yeast Genes Dev. 7, 2149–2160. 121. Merchant, A. M., Kawasaki, Y., Chen, Y., Lei, M., and Tye, B. K. (1997). A lesion in the DNA replication initiation factor Mcm10 induces pausing of elongation forks through chromosomal replication origins in Saccharomyces cerevisiae Mol. Cell. Biol. 17, 3261–3271. 122. Homesley, L., Lei, M., Kawasaki, Y., Sawyer, S., Christensen, T., and Tye, B. K. (2000). Mcm10 and the MCM2-7 complex interact to initiate DNA synthesis and to release replication factors from origins Genes Dev. 14, 913–926. 123. Wohlschlegel, J. A., Dhar, S. K., Prokhorova, T. A., Dutta, A., and Walter, J. C. (2002). Xenopus Mcm10 binds to origins of DNA replication after Mcm2-7 and stimulates origin binding of Cdc45 Mol. Cell 9, 233–240. 124. Kawasaki, Y., Hiraga, S., and Sugino, A. (2000). Interactions between Mcm10p and other replication factors are required for proper initiation and elongation of chromosomal DNA replication in Saccharomyces cerevisiae Genes Cells 5, 975–989.

66

Lucas and Raghuraman

125. Geraghty, D. S., Ding, M., Heintz, N. H., and Pederson, D. S. (2000). Premature structural changes at replication origins in a yeast minichromosome maintenance (MCM) mutant J. Biol. Chem. 275, 18011–18021. 126. Herendeen, D. R., and Kelly, T. J. (1996). DNA polymerase III: Running rings around the fork Cell 84, 5–8. 127. Waga, S., Bauer, G., and Stillman, B. (1994). Reconstitution of complete SV40 DNA replication with purified replication factors J. Biol. Chem. 269, 10923–10934. 128. Waga, S., and Stillman, B. (1994). Anatomy of a DNA replication fork revealed by reconstitution of SV40 DNA replication in vitro Nature 369, 207–212. 129. Moir, D., and Botstein, D. (1982). Determination of the order of gene function in the yeast nuclear division pathway using cs and ts mutants Genetics 100, 565–577. 130. Dalton, S., and Hopwood, B. (1997). Characterization of Cdc47p-minichromosome maintenance complexes in Saccharomyces cerevisiae: Identification of Cdc45p as a subunit Mol. Cell. Biol. 17, 5867–5875. 131. Hopwood, B., and Dalton, S. (1996). Cdc45p assembles into a complex with Cdc46p/ Mcm5p, is required for minichromosome maintenance, and is essential for chromosomal DNA replication Proc. Natl. Acad. Sci. USA 93, 12309–12314. 132. Zou, L., and Stillman, B. (1998). Formation of a preinitiation complex by S phase cyclin CDK-dependent loading of Cdc45p onto chromatin Science 280, 593–596. 133. Zou, L., Mitchell, J., and Stillman, B. (1997). CDC45, a novel yeast gene that functions with the origin recognition complex and Mcm proteins in initiation of DNA replication Mol. Cell. Biol. 17, 553–563. 134. Aparicio, O. M., Stout, A. M., and Bell, S. P. (1999). Differential assembly of Cdc45p and DNA polymerases at early and late origins of DNA replication Proc. Natl. Acad. Sci. USA 96, 9130–9135. 135. Kamimura, Y., Masumoto, H., Sugino, A., and Araki, H. (1998). Sld2, which interacts with Dpb11 in Saccharomyces cerevisiae, is required for chromosomal DNA replication Mol. Cell. Biol. 18, 6102–6109. 136. Owens, J. C., Detweiler, C. S., and Li, J. J. (1997). CDC45 is required in conjunction with CDC7/DBF4 to trigger the initiation of DNA replication Proc. Natl. Acad. Sci. USA 94, 12521–12526. 137. Araki, H., Leem, S. H., Phongdara, A., and Sugino, A. (1995). Dpb11, which interacts with DNA polymerase II(epsilon) in Saccharomyces cerevisiae, has a dual role in S phase progression and at a cell cycle checkpoint Proc. Natl. Acad. Sci. USA 92, 11791–11795. 138. Kamimura, Y., Tak, Y. S., Sugino, A., and Araki, H. (2001). Sld3, which interacts with Cdc45 (Sld4), functions for chromosomal DNA replication in Saccharomyces cerevisiae EMBO J. 20, 2097–2107. 139. Wang, H., and Elledge, S. J. (1999). DRC1, DNA replication and checkpoint protein 1, functions with DPB11 to control DNA replication and the S- phase checkpoint in Saccharomyces cerevisiae Proc. Natl. Acad. Sci. USA 96, 3824–3829. 140. Masumoto, H., Sugino, A., and Araki, H. (2000). Dpb11 controls the association between DNA polymerases alpha and epsilon and the autonomously replicating sequence region of budding yeast Mol. Cell. Biol. 20, 2809–2817. 141. Masumoto, H., Muramatsu, S., Kamimura, Y., and Araki, H. (2002). S-Cdk-dependent phosphorylation of Sld2 essential for chromosomal DNA replication in budding yeast Nature 415, 651–655. 142. Noguchi, E., Shanahan, P., Noguchi, C., and Russell, P. (2002). CDK phosphorylation of Drc1 regulates DNA replication in fission yeast Curr. Biol. 12, 599–605. 143. Kesti, T., Flick, K., Keranen, S., Syvaoja, J. E., and Wittenberg, C. (1999). DNA polymerase epsilon catalytic domains are dispensable for DNA replication, DNA repair, and cell viability Mol. Cell 3, 679–685.

1. The Dynamics of Chromosome Replication in Yeast

67

144. Dua, R., Levy, D. L., and Campbell, J. L. (1999). Analysis of the essential functions of the C-terminal protein/protein interaction domain of Saccharomyces cerevisiae pol epsilon and its unexpected ability to support growth in the absence of the DNA polymerase domain J. Biol. Chem. 274, 22283–22288. 145. Feng, W., and D’Urso, G. (2001). Schizosaccharomyces pombe cells lacking the aminoterminal catalytic domains of DNA polymerase epsilon are viable but require the DNA damage checkpoint control Mol. Cell. Biol. 21, 4495–4504. 146. MacNeill, S. A. (2001). DNA replication: Partners in the Okazaki two-step Curr. Biol. 11, R842–R844. 147. Gomes, X. V., Schmidt, S. L., and Burgers, P. M. (2001). ATP utilization by yeast replication factor C. II. Multiple stepwise ATP binding events are required to load proliferating cell nuclear antigen onto primed DNA J. Biol. Chem. 276, 34776–34783. 148. Dirick, L., Bohm, T., and Nasmyth, K. (1995). Roles and regulation of Cln-Cdc28 kinases at the start of the cell cycle of Saccharomyces cerevisiae EMBO J. 14, 4803–4813. 149. Labib, K., Tercero, J. A., and Diffley, J. F. (2000). Uninterrupted MCM2-7 function required for DNA replication fork progression Science 288, 1643–1647. 150. Tercero, J. A., Labib, K., and Diffley, J. F. (2000). DNA synthesis at individual replication forks requires the essential initiation factor Cdc45p EMBO J. 19, 2082–2093. 151. Nakajima, R., and Masukata, H. (2002). SpSld3 is required for loading and maintenance of SpCdc45 on chromatin in DNA replication in fission yeast Mol. Biol. Cell 13, 1462–1472. 152. Nougarede, R., Della Seta, F., Zarzov, P., and Schwob, E. (2000). Hierarchy of S-phasepromoting factors: Yeast Dbf4-Cdc7 kinase requires prior S-phase cyclin-dependent kinase activation Mol. Cell. Biol. 20, 3795–3806. 153. Jares, P., and Blow, J. J. (2000). Xenopus cdc7 function is dependent on licensing but not on XORC, XCdc6, or CDK activity and is required for XCdc45 loading Genes Dev. 14, 1528–1540. 154. Walter, J. C. (2000). Evidence for sequential action of cdc7 and cdk2 protein kinases during initiation of DNA replication in Xenopus egg extracts J. Biol. Chem. 275, 39773–39778. 155. Drury, L. S., Perkins, G., and Diffley, J. F. (2000). The cyclin-dependent kinase Cdc28p regulates distinct modes of Cdc6p proteolysis during the budding yeast cell cycle Curr. Biol. 10, 231–240. 156. Elsasser, S., Chi, Y., Yang, P., and Campbell, J. L. (1999). Phosphorylation controls timing of Cdc6p destruction: A biochemical analysis Mol. Biol. Cell 10, 3263–3277. 157. Lei, M., Kawasaki, Y., Young, M. R., Kihara, M., Sugino, A., and Tye, B. K. (1997). Mcm2 is a target of regulation by Cdc7-Dbf4 during the initiation of DNA synthesis Genes Dev. 11, 3365–3374. 158. Hardy, C. F., Dryga, O., Seematter, S., Pahl, P. M., and Sclafani, R. A. (1997). mcm5/ cdc46-bob1 bypasses the requirement for the S phase activator Cdc7p Proc. Natl. Acad. Sci. USA 94, 3151–3155. 159. Tanaka, T., and Nasmyth, K. (1998). Association of RPA with chromosomal replication origins requires an Mcm protein, and is regulated by Rad53, and cyclin- and Dbf4dependent kinases EMBO J. 17, 5182–5191. 160. Sclafani, R. A., Tecklenburg, M., and Pierce, A. (2002). The mcm5-bob1 bypass of Cdc7p/ Dbf4p in DNA replication depends on both Cdk1-independent and Cdk1-dependent steps in Saccharomyces cerevisiae Genetics 161, 47–57. 161. Rao, P. N., and Johnson, R. T. (1970). Mammalian cell fusion: Studies on the regulation of DNA synthesis Nature 225, 159–164. 162. Diffley, J. F. (1996). Once and only once upon a time: Specifying and regulating origins of DNA replication in eukaryotic cells Genes Dev. 10, 2819–2830.

68

Lucas and Raghuraman

163. Kelly, T. J., and Brown, G. W. (2000). Regulation of chromosome replication Annu. Rev. Biochem. 69, 829–880. 164. Diffley, J. F. (2001). DNA replication: Building the perfect switch Curr. Biol. 11, R367–R370. 165. Nasmyth, K. (1996). At the heart of the budding yeast cell cycle Trends Genet. 12, 405–412. 166. Piatti, S., Bohm, T., Cocker, J. H., Diffley, J. F., and Nasmyth, K. (1996). Activation of Sphase-promoting CDKs in late G1 defines a ‘‘point of no return’’ after which Cdc6 synthesis cannot promote DNA replication in yeast Genes Dev. 10, 1516–1531. 167. Dahmann, C., Diffley, J. F., and Nasmyth, K. A. (1995). S-phase-promoting cyclindependent kinases prevent re-replication by inhibiting the transition of replication origins to a pre-replicative state Curr. Biol. 5, 1257–1269. 168. Detweiler, C. S., and Li, J. J. (1998). Ectopic induction of Clb2 in early G1 phase is sufficient to block prereplicative complex formation in Saccharomyces cerevisiae Proc. Natl. Acad. Sci. USA 95, 2384–2389. 169. Drury, L. S., Perkins, G., and Diffley, J. F. (1997). The Cdc4/34/53 pathway targets Cdc6p for proteolysis in budding yeast EMBO J. 16, 5966–5976. 170. Moll, T., Tebb, G., Surana, U., Robitsch, H., and Nasmyth, K. (1991). The role of phosphorylation and the CDC28 protein kinase in cell cycle-regulated nuclear import of the S. cerevisiae transcription factor SWI5 Cell 66, 743–758. 171. Piatti, S., Lengauer, C., and Nasmyth, K. (1995). Cdc6 is an unstable protein whose de novo synthesis in G1 is important for the onset of S phase and for preventing a ‘reductional’ anaphase in the budding yeast Saccharomyces cerevisiae EMBO J. 14, 3788–3799. 172. Jang, S. W., Elsasser, S., Campbell, J. L., and Kim, J. (2001). Identification of Cdc6 protein domains involved in interaction with Mcm2 protein and Cdc4 protein in budding yeast cells Biochem. J. 354, 655–661. 173. Nguyen, V. Q., Co, C., and Li, J. J. (2001). Cyclin-dependent kinases prevent DNA re-replication through multiple mechanisms Nature 411, 1068–1073. 174. Yamashita, M., Hori, Y., Shinomiya, T., Obuse, C., Tsurimoto, T., Yoshikawa, H., and Shirahige, K. (1997). The efficiency and timing of initiation of replication of multiple replicons of Saccharomyces cerevisiae chromosome VI Genes Cells 2, 655–665. 175. Kouprina, N. Y., and Larionov, V. L. (1983). The study of a rDNA replicator in Saccharomyces Curr. Genet. 7, 433–438. 176. Larionov, V., Kouprina, N., and Karpova, T. (1984). Stability of recombinant plasmids containing the ars sequence of yeast extrachromosomal rDNA in several strains of Saccharomyces cerevisiae Gene. 28, 229–295. 177. Miller, C. A., Umek, R. M., and Kowalski, D. (1999). The inefficient replication origin from yeast ribosomal DNA is naturally impaired in the ARS consensus sequence and in DNA unwinding Nucleic Acids Res. 27, 3921–3930. 178. Vujcic, M., Miller, C. A., and Kowalski, D. (1999). Activation of silent replication origins at autonomously replicating sequence elements near the HML locus in budding yeast Mol. Cell. Biol. 19, 6098–6109. 179. Brewer, B. J., and Fangman, W. L. (1988). A replication fork barrier at the 30 end of yeast ribosomal RNA genes Cell 55, 637–643. 180. Brewer, B. J., Diller, J. D., Friedman, K. L., Kolor, K. M., Raghuraman, M. K., and Fangman, W. L. (1993). The topography of chromosome replication in yeast Cold Spring Harb. Symp. Quant. Biol. 58, 425–434. 181. Fangman, W. L., and Brewer, B. J. (1992). A question of time: Replication origins of eukaryotic chromosomes Cell 71, 363–366.

1. The Dynamics of Chromosome Replication in Yeast

69

182. Ferguson, B. M., Brewer, B. J., Reynolds, A. E., and Fangman, W. L. (1991). A yeast origin of replication is activated late in S phase Cell 65, 507–515. 183. Raghuraman, M. K., Brewer, B. J., and Fangman, W. L. (1997). Cell cycle-dependent establishment of a late replication program Science 276, 806–809. 184. Stevenson, J. B., and Gottschling, D. E. (1999). Telomeric chromatin modulates replication timing near chromosome ends Genes Dev. 13, 146–151. 185. McCarroll, R. M., and Fangman, W. L. (1988). Time of replication of yeast centromeres and telomeres Cell 54, 505–513. 186. Wellinger, R. J., Wolf, A. J., and Zakian, V. A. (1993). Saccharomyces telomeres acquire single-strand TG1-3 tails late in S phase Cell 72, 51–60. 187. Ferguson, B. M., and Fangman, W. L. (1992). A position effect on the time of replication origin activation in yeast Cell 68, 333–339. 188. Friedman, K. L., Diller, J. D., Ferguson, B. M., Nyland, S. V., Brewer, B. J., and Fangman, W. L. (1996). Multiple determinants controlling activation of yeast replication origins late in S phase Genes Dev. 10, 1595–1607. 189. Gottschling, D. E. (1992). Telomere-proximal DNA in Saccharomyces cerevisiae is refractory to methyltransferase activity in vivo Proc. Natl. Acad. Sci. USA 89, 4062–4065. 190. Wright, J. H., Gottschling, D. E., and Zakian, V. A. (1992). Saccharomyces telomeres assume a non-nucleosomal chromatin structure Genes Dev. 6, 197–210. 191. Grunstein, M. (1998). Yeast heterochromatin: Regulation of its assembly and inheritance by histones Cell 93, 325–328. 192. Renauld, H., Aparicio, O. M., Zierath, P. D., Billington, B. L., Chhablani, S. K., and Gottschling, D. E. (1993). Silent domains are assembled continuously from the telomere and are defined by promoter distance and strength, and by SIR3 dosage Genes Dev. 7, 1133–1145. 193. Cosgrove, A. J., Nieduszynski, C. A., and Donaldson, A. D. (2002). Ku complex controls the replication time of DNA in telomere regions Genes Dev. 16, 2485–2490. 194. Zappulla, D. C., Sternglanz, R., and Leatherwood, J. (2002). Control of replication timing by a transcriptional silencer Curr. Biol. 12, 869–875. 195. Bickmore, W. A., and Carothers, A. D. (1995). Factors affecting the timing and imprinting of replication on a mammalian chromosome J. Cell Sci. 108, 2801–2809. 196. Vogelauer, M., Rubbi, L., Lucas, I., Brewer, B. J., and Grunstein, M. (2002). Histone acetylation regulates the time of replication origin firing Mol. Cell 10, 1223–1233. 197. Heun, P., Laroche, T., Raghuraman, M. K., and Gasser, S. M. (2001). The positioning and dynamics of origins of replication in the budding yeast nucleus J. Cell Biol. 152, 385–400. 198. Heun, P., Laroche, T., Shimada, K., Furrer, P., and Gasser, S. M. (2001). Chromosome dynamics in the yeast interphase nucleus Science 294, 2181–2186. 199. Gasser, S. M. (2002). Visualizing chromatin dynamics in interphase nuclei Science 296, 1412–1416. 200. Cimbora, D. M., and Groudine, M. (2001). The control of mammalian DNA replication: A brief history of space and timing Cell 104, 643–646. 201. Straight, A. F., Belmont, A. S., Robinett, C. C., and Murray, A. W. (1996). GFP tagging of budding yeast chromosomes reveals that protein-protein interactions can mediate sister chromatid cohesion Curr. Biol. 6, 1599–1608. 202. Andrulis, E. D., Neiman, A. M., Zappulla, D. C., and Sternglanz, R. (1998). Perinuclear localization of chromatin facilitates transcriptional silencing Nature 394, 592–595. 203. Dimitrova, D. S., and Gilbert, D. M. (1999). The spatial position and replication timing of chromosomal domains are both established in early G1 phase Mol. Cell 4, 983–993. 204. Donaldson, A. D., and Blow, J. J. (1999). The regulation of replication origin activation Curr. Opin. Genet. Dev. 9, 62–68.

70

Lucas and Raghuraman

205. Shirahige, K., Hori, Y., Shiraishi, K., Yamashita, M., Takahashi, K., Obuse, C., Tsurimoto, T., and Yoshikawa, H. (1998). Regulation of DNA-replication origins during cell-cycle progression Nature 395, 618–621. 206. Dohrmann, P. R., Oshiro, G., Tecklenburg, M., and Sclafani, R. A. (1999). RAD53 regulates DBF4 independently of checkpoint function in Saccharomyces cerevisiae Genetics 151, 965–977. 207. Donaldson, A. D., Raghuraman, M. K., Friedman, K. L., Cross, F. R., Brewer, B. J., and Fangman, W. L. (1998). CLB5-dependent activation of late replication origins in S. cerevisiae Mol. Cell 2, 173–182. 208. Donaldson, A. D. (2000). The yeast mitotic cyclin Clb2 cannot substitute for S phase cyclins in replication origin firing EMBO Rep. 1, 507–512. 209. Donaldson, A. D., Fangman, W. L., and Brewer, B. J. (1998). Cdc7 is required throughout the yeast S phase to activate replication origins Genes Dev. 12, 491–501. 210. Bousset, K., and Diffley, J. F. (1998). The Cdc7 protein kinase is required for origin firing during S phase Genes Dev. 12, 480–490. 211. Kim, S. M., and Huberman, J. A. (2001). Regulation of replication timing in fission yeast EMBO J. 20, 6115–6126. 212. Pierron, G., Benard, M., Puvion, E., Flanagan, R., Sauer, H. W., and Pallotta, D. (1989). Replication timing of 10 developmentally regulated genes in Physarum polycephalum Nucleic Acids Res. 17, 553–566. 213. Watanabe, Y., Fujiyama, A., Ichiba, Y., Hattori, M., Yada, T., Sakaki, Y., and Ikemura, T. (2002). Chromosome-wide assessment of replication timing for human chromosomes 11q and 21q: Disease-related genes in timing-switch regions Hum. Mol. Genet. 11, 13–21. 214. DePamphilis, M. L. (1999). Replication origins in metazoan chromosomes: Fact or fiction? Bioessays 21, 5–16. 215. Schubeler, D., Scalzo, D., Kooperberg, C., Van Steensel, B., Delrow, J., and Groudine, M. (2002). Genome-wide DNA replication profile for Drosophila melanogaster: A link between transcription and replication timing Nature Genet. 30, 30. 216. Hansen, R. S., Canfield, T. K., Fjeld, A. D., Mumm, S., Laird, C. D., and Gartler, S. M. (1997). A variable domain of delayed replication in FRAXA fragile X chromosomes: X inactivation-like spread of late replication Proc. Natl. Acad. Sci. USA 94, 4587–4592. 217. Kitsberg, D., Selig, S., Keshet, I., and Cedar, H. (1993). Replication structure of the human beta-globin gene domain Nature 366, 588–590. 218. Aladjem, M. I., Groudine, M., Brody, L. L., Dieken, E. S., Fournier, R. E., Wahl, G. M., and Epner, E. M. (1995). Participation of the human beta-globin locus control region in initiation of DNA replication Science 270, 815–819. 219. Wu, J., and Grunstein, M. (2000). 25 years after the nucleosome model: Chromatin modifications Trends Biochem. Sci. 25, 619–623. 220. Tse, C., Sera, T., Wolffe, A. P., and Hansen, J. C. (1998). Disruption of higher-order folding by core histone acetylation dramatically enhances transcription of nucleosomal arrays by RNA polymerase III Mol. Cell. Biol. 18, 4629–4638. 221. Grant, P. A., and Berger, S. L. (1999). Histone acetyltransferase complexes Semin. Cell Dev. Biol. 10, 169–177. 222. Kornberg, R. D. (1999). Eukaryotic transcriptional control Trends Cell Biol. 9, M46–M49. 223. Struhl, K. (1998). Histone acetylation and transcriptional regulatory mechanisms Genes Dev. 12, 599–606. 224. Krebs, J. E., Kuo, M. H., Allis, C. D., and Peterson, C. L. (1999). Cell cycle-regulated histone acetylation required for expression of the yeast HO gene Genes Dev. 13, 1412–1421. 225. Cimbora, D. M., Schubeler, D., Reik, A., Hamilton, J., Francastel, C., Epner, E. M., and Groudine, M. (2000). Long-distance control of origin choice and replication timing in the

1. The Dynamics of Chromosome Replication in Yeast

226.

227.

228.

229. 230.

231. 232. 233. 234. 235. 236. 237. 238.

239. 240. 241.

242. 243. 244. 245.

71

human beta-globin locus are independent of the locus control region Mol. Cell. Biol. 20, 5581–5591. Hansen, R. S., Stoger, R., Wijmenga, C., Stanek, A. M., Canfield, T. K., Luo, P., Matarazzo, M. R., D’Esposito, M., Feil, R., Gimelli, G., Weemaes, C. M., Laird, C. D., and Gartler, S. M. (2000). Escape from gene silencing in ICF syndrome: Evidence for advanced replication time as a major determinant Hum. Mol. Genet. 9, 2575–2587. Barbosa, A. C., Otto, P. A., and Vianna-Morgante, A. M. (2000). Replication timing of homologous alpha-satellite DNA in Roberts syndrome Chromosome Res. 8, 645–650. Kelly, T. J., Martin, G. S., Forsburg, S. L., Stephen, R. J., Russo, A., and Nurse, P. (1993). The fission yeast cdc18 gene product couples S phase to START and mitosis Cell 74, 371–382. Diller, J. D., and Raghuraman, M. K. (1994). Eukaryotic replication origins: Control in space and time Trends Biochem. Sci. 19, 320–325. Labib, K., Kearsey, S. E., and Diffley, J. F. (2001). MCM2-7 proteins are essential components of prereplicative complexes that accumulate cooperatively in the nucleus during G1-phase and are required to establish, but not maintain, the S-phase checkpoint Mol. Biol. Cell 12, 3658–3667. Weinert, T. A., and Hartwell, L. H. (1988). The RAD9 gene controls the cell cycle response to DNA damage in Science 241, 317–322. Weinert, T., and Hartwell, L. (1989). Control of G2 delay by the rad9 gene of Saccharomyces cerevisiae J. Cell Sci. Suppl. 12, 145–148. Clarke, D. J., and Gimenez-Abian, J. F. (2000). Checkpoints controlling mitosis Bioessays 22, 351–363. Wahl, G. M., and Carr, A. M. (2001). The evolution of diverse biological responses to DNA damage: Insights from yeast and p53 Nature Cell Biol. 3, E277–E286. Weinert, T. (1997). Yeast checkpoint controls and relevance to cancer Cancer Surv. 29, 109–132. Wassmann, K., and Benezra, R. (2001). Mitotic checkpoints: From yeast to cancer Curr. Opin. Genet. Dev. 11, 83–90. Elledge, S. J., Zhou, Z., Allen, J. B., and Navas, T. A. (1993). DNA damage and cell cycle regulation of ribonucleotide reductase Bioessays 15, 333–339. Foiani, M., Pellicioli, A., Lopes, M., Lucca, C., Ferrari, M., Liberi, G., Muzi Falconi, M., and Plevani, P. (2000). DNA damage checkpoints and DNA replication controls in Saccharomyces cerevisiae Mutat. Res. 451, 187–196. Rouse, J., and Jackson, S. P. (2002). Interfaces between the detection, signaling, and repair of DNA damage Science 297, 547–551. Kolodner, R. D., Putnam, C. D., and Myung, K. (2002). Maintenance of genome stability in Saccharomyces cerevisiae Science 297, 552–557. Yamamoto, A., Guacci, V., and Koshland, D. (1996). Pds1p, an inhibitor of anaphase in budding yeast, plays a critical role in the APC and checkpoint pathway(s) J. Cell Biol. 133, 99–110. Boddy, M. N., and Russell, P. (2001). DNA replication checkpoint Curr. Biol. 11, R953–R956. Zhao, X., Muller, E. G., and Rothstein, R. (1998). A suppressor of two essential checkpoint genes identifies a novel protein that negatively affects dNTP pools Mol. Cell 2, 329–340. Paulovich, A. G., and Hartwell, L. H. (1995). A checkpoint regulates the rate of progression through S phase in S. cerevisiae in response to DNA damage Cell 82, 841–847. Paulovich, A. G., Margulies, R. U., Garvik, B. M., and Hartwell, L. H. (1997). RAD9, RAD17, and RAD24 are required for S phase regulation in Saccharomyces cerevisiae in response to DNA damage Genetics 145, 45–62.

72

Lucas and Raghuraman

246. Desany, B. A., Alcasabas, A. A., Bachant, J. B., and Elledge, S. J. (1998). Recovery from DNA replicational stress is the essential function of the S phase checkpoint pathway Genes Dev. 12, 2956–2970. 247. Tercero, J. A., and Diffley, J. F. (2001). Regulation of DNA replication fork progression through damaged DNA by the Mec1/Rad53 checkpoint Nature 412, 553–557. 248. Lopes, M., Cotta-Ramusino, C., Pellicioli, A., Liberi, G., Plevani, P., Muzi-Falconi, M., Newlon, C. S., and Foiani, M. (2001). The DNA replication checkpoint response stabilizes stalled replication forks Nature 412, 557–561. 249. Sogo, J. M., Lopes, M., and Foiani, M. (2002). Fork reversal and ssDNA accumulation at stalled replication forks owing to checkpoint defects Science 297, 599–602. 250. Garvik, B., Carson, M., and Hartwell, L. (1995). Single-stranded DNA arising at telomeres in cdc 13 mutants may constitute a specific signal for the RAD9 checkpoint Mol. Cell. Biol. 15, 6128–6138. 251. Rouse, J., and Jackson, S. P. (2002). Lcd1p recruits Mec1p to DNA lesions in vitro and in vivo Mol. Cell 9, 857–869. 252. Navas, T. A., Zhou, Z., and Elledge, S. J. (1995). DNA polymerase epsilon links the DNA replication machinery to the S phase checkpoint Cell 80, 29–39. 253. Pellicioli, A., Lucca, C., Liberi, G., Marini, F., Lopes, M., Plevani, P., Romano, A., Di Fiore, P. P., and Foiani, M. (1999). Activation of Rad53 kinase in response to DNA damage and its effect in modulating phosphorylation of the lagging strand DNA polymerase EMBO J. 18, 6561–6572. 254. Sanchez, Y., Desany, B. A., Jones, W. J., Liu, Q., Wang, B., and Elledge, S. J. (1996). Regulation of RAD53 by the ATM-like kinases MEC1 and TEL1 in yeast cell cycle checkpoint pathways Science 271, 357–360. 255. Alcasabas, A. A., Osborn, A. J., Bachant, J., Hu, F., Werler, P. J., Bousset, K., Furuya, K., Diffley, J. F., Carr, A. M., and Elledge, S. J. (2001). Mrc1 transduces signals of DNA replication stress to activate Rad53 Nature Cell Biol. 3, 958–965. 256. Tanaka, K., and Russell, P. (2001). Mrc1 channels the DNA replication arrest signal to checkpoint kinase Cds1 Nature Cell Biol. 3, 966–972. 257. Frei, C., and Gasser, S. M. (2000). The yeast Sgs1p helicase acts upstream of Rad53p in the DNA replication checkpoint and colocalizes with Rad53p in S-phase-specific foci Genes Dev. 14, 81–96. 258. Brush, G. S., and Kelly, T. J. (2000). Phosphorylation of the replication protein A large subunit in the Saccharomyces cerevisiae checkpoint response Nucleic Acids Res. 28, 3725–3732. 259. Marini, F., Pellicioli, A., Paciotti, V., Lucchini, G., Plevani, P., Stern, D. F., and Foiani, M. (1997). A role for DNA primase in coupling DNA replication to DNA damage response EMBO J. 16, 639–650. 260. Chabes, A., Domkin, V., and Thelander, L. (1999). Yeast Sml1, a protein inhibitor of ribonucleotide reductase J. Biol. Chem. 274, 36679–36683. 261. Jares, P., Donaldson, A., and Blow, J. J. (2000). The Cdc7/Dbf4 protein kinase: Target of the S phase checkpoint? EMBO Rep. 1, 319–322. 262. Clarke, D. J., Segal, M., Jensen, S., and Reed, S. I. (2001). Mec1p regulates Pds 1p levels in S phase: Complex coordination of DNA replication and mitosis Nature Cell Biol. 3, 619–627. 263. Rhind, N., and Russell, P. (2000). Chk1 and Cds1: Linchpins of the DNA damage and replication checkpoint pathways J. Cell Sci. 113, 3889–3896. 264. Emili, A., Schieltz, D. M., Yates, J. R., III, and Hartwell, L. H. (2001). Dynamic interaction of DNA damage checkpoint protein Rad53 with chromatin assembly factor Asf1 Mol. Cell 7, 13–20.

1. The Dynamics of Chromosome Replication in Yeast

73

265. Hu, F., Alcasabas, A. A., and Elledge, S. J. (2001). Asf1 links Rad53 to control of chromatin assembly Genes Dev. 15, 1061–1066. 266. van Brabant, A. J., Fangman, W. L., and Brewer, B. J. (1999). Active role of a human genomic insert in replication of a yeast artificial chromosome Mol. Cell. Biol. 19, 4231–4440. 267. van Brabant, A. J., Buchanan, C. D., Charboneau, E., Fangman, W. L., and Brewer, B. J. (2001). An origin-deficient yeast artificial chromosome triggers a cell cycle checkpoint Mol. Cell 7, 705–713. 268. Michael, W. M., Ott, R., Fanning, E., and Newport, J. (2000). Activation of the DNA replication checkpoint through RNA synthesis by primase Science 289, 2133–2137. 269. Rao, H., Marahrens, Y., and Stillman, B. (1994). Functional conservation of multiple elements in yeast chromosomal replicators Mol. Cell. Biol. 14, 7643–7651. 270. Theis, J. F., and Newlon, C. S. (1994). Domain B of ARS307 contains two functional elements and contributes to chromosomal replication origin function Mol. Cell. Biol. 14, 7652–7659. 271. Walker, S. S., Malik, A. K., and Eisenberg, S. (1991). Analysis of the interactions of functional domains of a nuclear origin of replication from Saccharomyces cerevisiae Nucleic Acids Res. 19, 6255–6262.

ThisPageIntentionallyLeftBlank

2

____________________________________________________________________________

Micromechanical Studies of Mitotic Chromosomes M. G. Poirier* and John F. Marko*y

Departments of *Physics and yBioengineering University of Illinois at Chicago, Chicago, Illinois 60607

I. Introduction II. Architecture and Components of Eukaryote Chromosomes A. Eukaryote Chromosomes Are Made of Chromatin Fiber B. Micromechanics of Double-Stranded DNA C. Micromechanics of Chromatin Fibers D. Chromosome Structure at Scales Larger Than the Chromatin Fiber E. Why Study Mitotic Chromosomes Micromechanically? III. Stretching Elasticity of Chromosomes A. Lampbrush Chromosomes B. Mitotic Chromosome Extensibility and Elasticity C. In Vitro-Assembled Chromosomes D. Summary IV. Bending Elasticity of Chromosomes A. Expected Bending Flexibility and Fluctuations of Mitotic Newt Chromosomes B. Bending Fluctuations of Chromosomes Extracted from Cells C. Bending Fluctuations of Chromosomes in Vivo D. Bending Fluctuations of in Vitro-Assembled Xenopus Chromatids E. Bending of Chromosomes during Mitosis F. Summary V. Viscoelasticity of Chromosomes A. Observations of Slow Stress Relaxation B. Dynamics of Bending of Mitotic Newt Chromosomes C. Summary VI. Combined Biochemical–Micromechanical Study of Mitotic Chromosomes A. Whole Genome Extraction Experiments B. Combined Micromechanical–Chemical Experiments C. Shifts in Ionic Conditions Can Decondense or Hypercondense Mitotic Chromosomes D. Micrococcal Nuclease Completely Disintegrates Mitotic Chromosomes E. Restriction Enzymes with Four-Base Specificity Can Disintegrate Mitotic Chromosomes F. Summary VII. Conclusion A. Summary of Physical Properties of Mitotic Chromosomes B. Elasticity of Mitotic Chromosomes Versus Elasticity of Chromatin Fiber C. Ionic Condition Shift Experiments Current Topics in Developmental Biology, Vol. 55 Copyright 2003, Elsevier (USA). All rights reserved. 0070-2153/03 $35.00

75

76

Poirier and Marko D. DNA-Cutting Experiments E. Implications for Structure of the Mitotic Chromosome F. Future Experiments References

We review micromechanical experiments studying mechanoelastic properties of mitotic chromosomes. We discuss the history of this field, starting from the classic in vivo experiments of Nicklas (1983). We then focus on experiments where chromosomes were extracted from prometaphase cells and then studied by micromanipulation and microfluidic biochemical techniques. These experiments reveal that chromosomes have a wellbehaved elastic response over a fivefold range of stretching, with an elastic modulus similar to that of a loosely tethered polymer network. Perturbation by microfluidic ‘‘spraying’’ of various ions reveals that the mitotic chromosome can be rapidly and reversibly decondensed or overcondensed, i.e., that the native state is not maximally compacted. We compare our results for chromosomes from cells to results of experiments by Houchmandzadeh and Dimitrov (1999) on chromatids reconstituted using Xenopus egg extracts. Remarkably, while the stretching elastic response of reconstituted chromosomes is similar to that observed for chromosomes from cells, reconstituted chromosomes are far more easily bent. This result suggests that reconstituted chromatids have a large-scale structure that is quite diVerent from chromosomes in somatic cells. Finally, we discuss microspraying experiments of DNA-cutting enzymes, which reveal that the element that gives mitotic chromosomes their mechanical integrity is DNA itself. These experiments indicate that chromatin-condensing proteins are not organized into a mechanically contiguous ‘‘scaVold,’’ but instead that the mitotic chromosome is best thought of as a cross-linked network of chromatin. Preliminary results from restriction enzyme digestion experiments indicate a spacing between chromatin ‘‘cross-links’’ of roughly 15 kb, a size similar to that inferred from classical chromatin loop isolation studies. These results suggest a general strategy for the use of micromanipulation methods for the study of chromosome structure. ß 2003 Elsevier (USA).

I. Introduction The question of how double-stranded DNAs (dsDNA) that encode the genomes of cells are physically organized, or ‘‘folded,’’ is a fundamental yet unresolved problem of cell biology. This is remarkable given the large amount of eVort that has been devoted to the traditional microscopy of higher order chromatin structures. The fact that new models for large-scale

2. Micromechanical Studies of Mitotic Chromosomes

77

chromosome structure (Kimura et al., 1999; Machado and Andrew, 2000a; Dietzel and Belmont, 2001; Losada and Hirano, 2001, Stack and Anderson, 2001) continue to be proposed indicates that this question remains open. During interphase, gene expression is closely related to chromatin organization. It is inevitable that the physical layout of genes in the nucleus aVects their expression, e.g., by aVecting the transport of regulatory factors to and mRNAs from transcription loci. However, very little is known about the structures that organize interphase chromosomes. During mitosis, gene expression stops and chromosomes undergo a gross reorganization, or ‘‘condensation,’’ into segregated, cigar-shaped mitotic chromatids. Again, very little is known about how the chromatin is folded up at this stage of the cell cycle. There are many reasons why determination of the chromosome structure in any cell is challenging. However, one of the main problems is certainly that chromosomes have a dynamic structure, which changes drastically during the cell cycle (Fig. 1). Studies of chromosome structure make sense only in the context of particular points of the cell cycle in defined cell types. This chapter focuses on the folding of the chromosome in amphibian cells during mitosis, specifically at the stage between prophase and metaphase when chromosomes are completely condensed and the nuclear envelope has been disassembled, but where the chromosomes are not yet attached to the mitotic spindle. We will mainly discuss the structure of prometaphase chromosomes, specifically in epithelial cells from newt (Notophthalmus viridescens) and frog (Xenopus laevis). These are model organisms for the study of mitotic

Figure 1 Cell cycle in a newt cell. During mitosis, chromosomes condense inside the nucleus; during prophase, the nuclear envelope disassembles and chromosomes float loose in the cytoplasm; and during prometaphase, they are captured and aligned by the spindle at metaphase. The two duplicate chromatids of each chromosome are pulled apart at anaphase. Bar: 10 mm. Image is a phase-contrast, 60 oil objective.

78

Poirier and Marko

Figure 2 (Top) Prometaphase chromosome attached at its ends to pipettes outside a cell. Bar: 10 mm. Image is a DIC, 60 oil objective. (Bottom) Three possible models of how chromatin is arranged within a mitotic chromosome.

chromosome structure for the simple reason that their chromosomes are large (Fig. 2). A second problem that chromosome researchers must confront is that chromosomes are soft physical objects, with elastic stiVness far less than that of DNAs and proteins from which they are composed. This means that the structures of chromosomes can be destroyed—or changed—by preparations that leave protein and DNA secondary structures intact. This chapter is concerned with reviewing recent studies of mechanical properties of mitotic chromosomes that quantify their softness. Emphasis will be placed on the idea that mechanical measurements can be used to assay structural changes introduced biochemically. We will show how such studies can provide information about higher order chromosome structure. Section II provides a brief review of previous biophysical studies of chromosome structure and the force response of single DNA molecules and chromatin fibers, the basic constituents of chromosomes. Section II concludes with a summary that seeks to convince the reader that micromanipulation experiments are a useful tool for answering the many questions about the mitotic chromosome structure that are either contentious or unanswered. Section III reviews experiments studying the stretching elastic response of whole mitotic chromosomes. Section IV discusses the bending elasticity of mitotic chromosomes, emphasizing the connection expected between bending and stretching elasticity. The dynamics of stress relaxation in chromosomes are then discussed briefly in Section V.

2. Micromechanical Studies of Mitotic Chromosomes

79

Section VI then discusses experiments that modify chromosome structure chemically and biochemically, while monitoring the changes in chromosome mechanical properties. This includes discussion of the eVects of shifts in salt concentration and DNA-cutting enzymes. The experiments discussed in this section have clear implications for mitotic chromosome structure, and in particular rule out the ‘‘contiguous protein scaVold’’ model, which posits that chromatin fibers are organized as loop domains tethered to an internal and physically connected protein skeleton. Finally, Section VII presents a preliminary model of mitotic chromosome structure based on these results and then discusses some of the many open questions, including the topic of DNA connections between mitotic chromosomes. Work of Poirier et al. (2000, 2001, 2002; Poirier and Marko, 2002a,b) is described in more detail in Poirier (2001). Web materials, including images and movies of experiments, are available at http://www.uic.edu/jmarko.

II. Architecture and Components of Eukaryote Chromosomes This section reviews current understanding of the components of chromosomes and overall chromosome structure. It also discusses physical properties of the components of chromosomes, essentially DNA and chromatin fiber, with emphasis on recent micromanipulation experiments. This section is not a complete review of the large literature on the chromosome structure (see Koshland and Strunnikov, 1996; Hirano, 2000), but is meant to brief the reader on some basic structural and biophysical facts about eukaryote chromosomes important to understanding the later sections. The plan of this section is to start from what is best known—the structure of the nucleosome—and then work up to gradually larger chromatin structures, which are less well understood. A. Eukaryote Chromosomes Are Made of Chromatin Fiber Eukaryote chromosomes contain similar amounts of genomic dsDNA and protein. Chromosomes of animals contain on the order of 100 Mb of dsDNA (note the useful dsDNA relation 1 Gb = 1 pg). Paradoxically, size and complexity of genomes are not obviously related (Gall, 1981): the largest human chromosomes contain about 300 Mb, whereas some amphibian chromosomes contain more than 1 Gb. At all stages of the cell cycle, this large amount of DNA is organized into nucleosomes (Kornberg, 1974), octamers of histone proteins around which dsDNA is wrapped. Each nucleosome is about 10 nm in diameter and involves about 200 bp of dsDNA (146 bp wrapped, with the balance as internucleosomal ‘‘linker’’ DNA). The structure of the nucleosome has

80

Poirier and Marko

been determined precisely using X-ray crystallography (Klug, 1984; Arents et al., 1991; Luger et al., 1997). Remarkable progress has been made in the understanding of the remodeling of nucleosome structure and chemical modification of histones themselves during gene expression (WolVe and Guschin, 2000). It is clear that there are many structural states of chromatin to understand. The molecular mass of 200 bp of dsDNA is about 120 kDa, and the molecular mass of the histone octamer plus one ‘‘linker’’ histone (which sits on the linker DNA) is about 125 kDa. Thus the relative weight of dsDNA and histones in chromosomes is roughly equal; histones are a major protein component of chromosomes. It is known that DNA bound to nucleosomes is able to unbind transiently. Quantitative experiments (Widom, 1997; Polach and Widom, 1995; Anderson and Widom, 2000) show that restriction enzyme access to DNA is attenuated exponentially as one moves into nucleosome-bound DNA. This raises the interesting question of on what time scale, and for what factors, transient access to DNA may occur via conformational fluctuation of the nucleosome itself. The clarity of understanding of nucleosome structure contrasts with the confusion about how the 10-nm-diameter nucleosomes are organized into larger scale (‘‘higher order’’) chromatin structures. Electron microscope (Thoma et al., 1979) and X-ray diVraction (Widom and Klug, 1985) studies suggest that the nucleosomes fold into a 30-nm-diameter chromatin fiber, possibly with a helical structure. However, little else about supranucleosomal organization (‘‘higher order chromatin structure’’) is solidly understood. This is a result of the relative softness of chromatin fiber, which leads to the apparent flexible polymer properties of chromatin (Cui and Bustamante, 2000; Marko and Siggia, 1997a; Sec.II.C), plus the inhomogeneity inherent to chromatin. Polymer-like flexibility may also account for observations of nonhelical chromatin fiber structures (Horowitz et al., 1994; Woodcock and Horowitz, 1995). Chromatin fiber structure is sensitive to ionic conditions. When chromatin fibers are extracted into solution at subphysiological 10 mM univalent ionic strength, they are observed in the electron microscope as 10-nm-thick ‘‘beads on a string.’’ At the more physiological ionic strength of 100 to 150 mM univalent ions, nucleosomes stack into a more condensed, and thicker, 30-nm-thick fiber (Fig. 3). At physiological ionic strength, lateral internucleosomal attractions tend to lead to aggregation of isolated fibers (Van Holde, 1989). The sensitivity of chromatin fiber to ionic strength is connected to two key concepts. First, nucleosome–nucleosome interactions have a strong electrostatic component. The eVect of altering univalent ionic strength is to change the strength and range of electrostatic interactions. At low ionic strength,

2. Micromechanical Studies of Mitotic Chromosomes

81

Figure 3 dsDNA, histones, nucleosome, 10-nm chromatin fiber, and 30-nm chromatin fiber. Structural–biological studies of chromatin have focused on the ultrastructure of isolated nucleosomes and on studying the conformation of nucleosomes in the 10- and 30-nm fiber.

electrostatic interactions are strong and have a long range, causing the likecharged nucleosomes (chromatin fiber has a net negative charge, similar to dsDNA) to repel suYciently to open chromatin fiber up. At higher ionic strength, the reduced strength and range of electrostatic repulsion are overcome by attractive nucleosome–nucleosome interactions mediated by histone tails and histone H1, and the fiber folds up. The second key concept, which is perhaps less familiar, is that chromatin fiber is relatively soft, or equivalently that internucleosomal interactions

82

Poirier and Marko

are relatively weak. The change in ionic strength from 10 to 100 mM has a drastic eVect on chromatin fiber, yet the nucleosomes themselves do not undergo major conformational changes. The strong electrostatic histone– dsDNA interactions are relatively unperturbed until much higher ionic strengths (0.8 M Na+) are reached. Similarly, the dsDNA structure is essentially insensitive to this change in ionic strength; over the range 0.01 to 0.1 M Na+ the main eVect on the double helix is an increase in the melting temperature of about 10 . The softness of chromatin fiber, relative to the relative ‘‘stiVness’’ of dsDNA and nucleosome structure, is important in understanding chromosome physical properties. The physiological 30-nm chromatin fiber is thought to be anywhere from 10- to 50-fold shorter in contour length than the underlying dsDNA. A widely used estimate results from the compaction of the 1200-bp associated with six nucleosomes into one 10-nm-thick turn of helical chromatin fiber: the resulting 120 bp/nm for chromatin is about 40 times less than the 3 bp/nm for dsDNA. In fact, this 40-fold compaction factor has not been convincingly given by experiments. Given that it is known that some nucleosomes are positioned, some are mobile, and that there are a wide range of histone modifications and variants, it seems unlikely that there is a universal chromatin fiber structure or length compaction factor.

B. Micromechanics of Double-Stranded DNA A new approach to biophysical characterization of DNA is mechanical manipulation of single molecules, with molecular tension as an experimentally controllable and measurable quantity. The general idea is to quantify intermolecular interactions by direct force measurement and to observe self-organization processes of single or small numbers of molecules. Methods used to study single dsDNAs are all based on attaching the ends of the molecule to large objects, which act as ‘‘handles’’ (Bustamante et al., 2000). The handles are used to apply controllable forces and to provide an optical marker for the molecule ends and therefore end-to-end extension. Although these techniques usually are restricted in application to molecules of at least a few kilobases in length, ingenious techniques (Bustamante et al., 2000) have been developed to measure conformational changes of just a few nanometers (Liphardt et al., 2001). This section focuses mainly on what has been learned about dsDNA mechanical properties using these techniques in preparation for discussing similar force–distance experiments on whole chromosomes. We also use the example of dsDNA to introduce some of the basic ideas of polymer elasticity used to discuss chromosome extensibility. dsDNA itself has mechanical properties that are well characterized and understood. dsDNA has a persistence length of about A=50 nm (150 bp in

2. Micromechanical Studies of Mitotic Chromosomes

83

B form) (Hagerman, 1988). The persistence length is the contour length over which thermal (Brownian) fluctuations can dynamically bend the double  helix appreciably (e.g., through a 60 bend). Thus, over dsDNA lengths of less than 150 bp, the contour is of fixed shape (the double helix is in general roughly straight, but some sequences are intrinsically rather severly bent). Over distances longer than 150 bp, a dsDNA undergoes appreciable dynamic bending. If one stretches a long dsDNA out, thermally excited bends will require a certain tension to be straightened (Fig. 4a). This tension is about kT/A, where kT ¼ 4.1  1021 J is the energy associated with a single thermal fluctuation at room temperature T  300 K (note kT ¼ RT/NA, where R is the familiar gas constant and NA is Avogadro’s number; RT is simply the thermal energy of 1 mol of thermal fluctuation, i.e., about the heat inside 1 mol of a simple gas or liquid). This characteristic tension is about 0.1 1012 Newtons (N; note 1 J/m ¼ 1 N) or about 0.1 piconewtons (pN). Below 0.1 pN, one can think of a dsDNA as being a spring, with extension proportional to applied tension; at 0.1 pN a dsDNA is extended to slightly greater than half its total contour length. At higher forces (0.1 to 10 pN) dsDNA elasticity is highly nonlinear, with tension increasing quickly as the length approaches that of the B form (3 bp/nm) (Smith et al., 1992; Bustamante et al., 1994). The characteristic tension to begin to extend a dsDNA (0.1 pN) is a small force, even by single molecule standards. Cellular motor proteins generate forces ranging from a few piconewtons (myosin: 5 pN, kinesin: 8 pN) to tens of pN (RNA polymerase: 40 pN; Yin et al., 1995), roughly because they convert chemical energy at the rate of a few kT per nanometer of motion (note that 1 kT/nm ¼ 4 pN). Another source of tension on dsDNA in vivo is DNA–protein interaction; e.g., it has been demonstrated that polymerization of RecA onto dsDNA generates forces in excess of 50 pN (Leger et al., 1998). In the cell, dsDNA thus can be stretched out and modified structurally by forces generated by the machinery that transcribes (Yin et al., 1995), replicates (Wuite et al., 2000), and repairs it. From forces of 0.1 to 10 pN, the dsDNA elastic response is well expressed by the empirical force law (Bustamante et al., 1994): " # kT x 1 1 þ  ð1Þ f ¼ A L 4ð1  x Þ2 4 L

where A is the persistence length of 50 nm and where x is the molecule endto-end extension and L its total B-form contour length. Equation (1) captures the weak initial elastic response where force increases from 0 to about kT/A as x/L increases from 0 to about 0.5, and the strong nonlinear force increase as x approaches L. These two features are generic for all flexible

84

Poirier and Marko

2. Micromechanical Studies of Mitotic Chromosomes

85

polymers that undergo random walk-like bending fluctuations when unstretched. For even larger forces (10 to 100 pN), the dsDNA secondary structure starts first to stretch (10–50 pN) and then the double helix is disrupted and stretches to an extended form at 65pN (Cluzel et al., 1996; Smith et al., 1996). This disruption has a strong DNA twisting dependence (Allemand et al., 1998; Leger et al., 1999). Measurements of the stretching of the double helix structure have shown that dsDNA can be thought of as an elastic rod, of elastic Young modulus Y  300 MPa. The meaning of Y comes from the force needed to stretch an elastic rod of uniform and circular cross section and equilibrium (unstretched) length L so as to increase its length by
L: f ¼ r2 Y


L L

ð2Þ

Here, r is the cross-sectional radius of the rod (for dsDNA, r ¼ 1 nm; note that for the general cross-section shape, r2 can be replaced by the rod crosssectional area). The Young modulus is thus the stress (force per cross-sectional area) at which an elastic rod would be doubled in length if its initial linear elasticity could be extrapolated: Y characterizes the stretching elasticity of a material in a shape-independent way. Similarly, f0 ¼ r2Y is the force at which a rod would double in length, based on extrapolation of its linear elasticity. For dsDNA, f0  1000 pN, and like most solid materials, Eq. (2) applies only for
L/L much less than unity [for dsDNA, the regime where Eq. (2) applies is from
L/L ¼ 0.0 to about 0.05, where 0.0 refers to the B-DNA length). The bending flexibility of an elastic rod is also related to Y. The bending modulus of an elastic rod, again assuming linear elasticity and circular uniform cross section, is  ð3Þ B ¼ r4 Y 4 This quantity has dimensions of energy times length. If our elastic rod is bent into a circular arc of bending radius R, the torque that must be applied is B/R, and the force that must be applied is B/R2. For dsDNA, Y ¼ 300 MPa gives B ¼ 21028 Jm. Figure 4 Comparison of elastic response of (a) single dsDNAs and (b and c) chromatin fibers. dsDNA and the chromatin fiber both display an initial low-force (sub-pN) elastic regime, followed by a higher force (few pN) regime. However, dsDNA shows a very stiV and nonlinear response (A), whereas the chromatin fiber shows a more gradual elastic response (B) (Cui and Bustamante, 2000) as it is extended. This is believed to be due to driving the chromatin fiber opening transition (10- to 30-nm fiber transition of Fig. 3) by force. (C) Low-force elasticity is also seen (Bennink et al., 2001); at higher forces, force jumps corresponding to nucleosome removal events are observed. (See Color Insert.)

86

Poirier and Marko

For rods that are thin enough to be bent by thermal fluctuation (e.g., the double helix), it is useful to relate B to the bending persistence length A: A¼

B  r4 Y ¼ kT 4 kT

ð4Þ

For dsDNA, we therefore see that Y=300 MPa gives an estimate of A ¼ 50 nm, essentially the observed value. The connection between the value of Y obtained from stretching the double helix with that obtained from separate measurement of the persistence length A shows that elementary concepts of elasticity at least roughly apply at the nanometer scale of the interior of the double helix. Micromechanical studies of DNA have allowed detailed studies of DNA stretching (Smith et al., 1992; Bustamante et al., 1994), DNA twisting and supercoiling (Strick et al., 1996), stress-driven DNA structural transitions (Cluzel et al., 1996; Smith et al., 1996; Allemand et al., 1999; Leger et al., 1999), and DNA strand separation by force (Essevaz-Roulet et al., 1997). Application of these techniques to the study of nucleic acid processing and reorganization by proteins is a direction of intensive current research. Protein–DNA interactions, which have been studied micromechanically, include force generation during transcription (Yin et al., 1995), force generated by DNA polymerase (Wuite et al., 2000), direct study of a single DNA loop formed by lac repressors (Finzi et al., 1995), dynamics of stretching of DNA by RecA (Leger et al., 1998), and observation of DNA strand exchange by topoisomerase II (Strick et al., 2000).

C. Micromechanics of Chromatin Fibers The force–extension properties of chromatin fiber extracted from chicken erythrocytes (Cui and Bustamante, 2000) have been measured. Because chromatin fibers are far more complex than single dsDNAs, their mechanical response is complicated. Three diVerent force regimes have been reported. First, a very low-force ‘‘entropic elasticity’’ regime is observed, similar to that seen for dsDNA. This initial low-force (below 0.1 pN) force response is thought to be due to the polymer flexibility of chromatin and allows an estimate of chromatin persistence length of about 30 nm, slightly shorter than dsDNA itself. This low persistence length is possible due to the zig-zag path of the linker DNA: a spring (a ‘‘Slinky’’ toy is a good example) can be bent more easily than the wire from which it is formed. However, quantitatively useful data for chromatin low-force (<0.1 pN) ‘‘polymer’’ elasticity under physiological conditions have not yet been published. At higher forces (0.1 to 5 pN), what is observed depends strongly on ionic conditions, as one would expect based on the 10- to 30-nm fiber transition

87

2. Micromechanical Studies of Mitotic Chromosomes +

observed with increasing ionic strength. At relatively low (10 mM Na ) ionic strength, a strongly nonlinear elastic response similar to that of dsDNA is observed. However, at closer to physiological ionic strength (40 mM Na+), a more gradual, nearly linear elastic response is observed for forces between 0.1 and 5 pN (visible at the left of Fig. 4b, data from Bennink et al., 2001). This can be explained in terms of the unstacking of adjacent nucleosomes, i.e., by the idea that force can be used to drive the 30- to 10-nm fiber transition. This transition is observed to be reversible and is characterized by a force constant f0  5 pN and a high degree of smooth extensibility (compare with ‘‘bare’’ dsDNA, which has a stretching force constant of 1000 pN and can be stretched by only about 5% before transforming to a new stretched form). The doubling in length of the chromatin fiber over a 5 pN increase in force observed by Cui and Bustamante (2000) can be combined with the native fiber 30-nm diameter to estimate an eVective Young modulus, Y100 kPa, far below the eVective modulus of straight DNA 300 MPa. As DNA is folded up into chromatin, its eVective modulus is reduced. At higher forces (20 pN), irreversible extension of chromatin fiber occurs (Cui and Bustamante, 2000). This has been observed to be in the form of a series of jumps of quantized length (Fig. 4b). These jumps are thought to be associated with the removal of single nucleosomes. Brower-Toland et al. (2002) showed that half-nucleosome (80 bp) winds of DNA can also be released using similar tensions. It is likely that this threshold for nucleosome removal is highly extension rate dependent, as the known binding free energy 20 to 30 kT/nucleosome indicates that one should expect equilibrium between bound and free nucleosomes for forces near 2 to 3 pN (Marko and Siggia, 1997b). Observation of this equilibrium for pure chromatin fiber would require long experimental time scales, as the barrier associated with nucleosome removal or rebinding is likely close to the 20-kT binding energy. However, use of nucleosome assembly factors such as NAP-1, which act in thermal equilibrium, may make it practical to observe chemical equilibrium between octamer on- vs oV-states (S. Leuba, private communication). The experiments of Cui and Bustamante (2000) used chromatin fibers isolated from cells, but more recent experiments have assembled chromatin fibers in vitro onto initially bare molecules of dsDNA. One way to proceed is to use salt dialysis assembly (Brower-Toland et al., 2002), which titrates histone–DNA interaction strength allowing nucleosomes to form along dsDNA. Another strategy is to use cell extract-derived chromatin assembly systems (Ladoux et al., 2000; Bennink et al., 2001), which have allowed measurement of the 10-pN forces applied during chromatin assembly. Purified chromatin assembly enzymes such as NAP-1 promise to provide a biochemically defined assembly system. These strategies promise to allow

88

Poirier and Marko

assembly of nucleosome arrays where the underlying DNA is known and which fold into 30-nm-like fibers. A key result of the chromatin fiber studies to date is that the elastic response of chromatin fiber is very diVerent from that of the underlying dsDNA. The presence of nucleosomes masks the divergent force response of dsDNA elasticity. Weak initial entropic elasticity is followed by reversible unfolding at the few piconewton force scale, whereas at larger forces 10 pN, nucleosomes are irreversibly popped oV. Many questions remain, just as one example the degree of variation of chromatin physical properties as a function of histone modifications that occur in vivo.

D. Chromosome Structure at Scales Larger Than the Chromatin Fiber Beyond the chromatin fiber, it is thought that nonnucleosome proteins act to define chromosome structure. During interphase, this includes the machinery of gene regulation and expression, centers of DNA replication (Cook, 1991), and the nuclear matrix (WolVe, 1995, Section II,D,2), all of which are beyond the scope of this chapter. This section focuses on what is known about the large-scale chromosome structure, with emphasis on the mitotic chromosome structure. We describe conclusions of classical microscopy-based studies, plus newer insights obtained from threedimensional studies of chromosome structure and dynamics. Finally, we discuss recent work on biochemical characterization of proteins from cell-free chromosome assembly systems. 1. Structural–Biological Studies of Mitotic Chromosome Structure Much of our understanding of the mitotic chromosome structure at larger scales is mainly based on relatively invasive electron microscopy (EM) studies and on optical microscopy. Based on EM visualization of DNA loops extending from an apparent protein-rich chromosome body after histone depletion (Paulson and Laemmli, 1977; Paulson, 1988), and to some extent on direct visualization of these chromatin loops in fixed cells, one commonly discussed model for mitotic chromosome structure is based on labile chromatin loops interconnected by a protein-rich ‘‘scaVold’’ (Marsden and Laemmli, 1979; Fig. 2). Other studies suggest that the scaVolding is coiled (Boy de la Tour and Laemmli, 1988). These experiments are often taken to imply the existence of a connected protein ‘‘skeleton’’ inside the mitotic chromosome (see Lewin, 2000; Lodish et al., 1995; WolVe, 1995). Paulson and Lamelli (1977) concluded that the scaVold was a fibrous network of nonhistone proteins and was responsible for the basic shape of metaphase chromosomes, and Lamelli

2. Micromechanical Studies of Mitotic Chromosomes

89

et al. (1978) emphasized that the scaVold could be isolated as a structurally independent stable entity. However, slightly later discussions (Marsden and Lamelli, 1979) suggested that the question of whether the native scaVold is stabilized through protein–protein interactions is unresolved. Intriguingly, an old literature of whole chromosome DNaase digestion experiments (Cole, 1967) suggests that if DNA is cut suYciently often, the chromosome disintegrates. Laemmli (2002) has emphasized to one of us that the conclusion that the internal protein skeleton is mechanically contiguous does not follow from his results. Therefore, the question of connectivity and mechanical integrity of the DNA and non-DNA components of the mitotic chromosome remains open; this is a primary focus of Section VI. Other microscopy studies suggest a hierarchical structure formed from a succession of coils at larger length scales (Belmont et al., 1987, 1989; Fig. 2). Proposals have since been made for a mitotic chromosome structure that combines loop and helix-folding motifs (Saitoh and Laemmli, 1993). Existing microscopy studies do not give a clear and consistent idea of chromatin structure in mitotic chromosomes, in part because of the invasive preparations necessary for EM visualization and the inability of light microscopy to observe structures smaller than 200 nm in vivo. The folding scheme of interphase chromatin inside the nucleus pre1990 was highly unsettled. With no techniques to diVerentiate diVerent chromosomes or chromosomal regions, light microscopy by itself reveals little, and electron microscopy again leads to conflicting views of chromatin structure at length scales from 10 to 100 nm. Biochemical analysis of chromatin domains (Jackson et al., 1990) suggests that the interphase chromatin is organized into 50-kb domains. 2. Three-Dimensional Microscopy Study of Chromosome Structure and Dynamics An increasing use of fluorescent labeling and optical sectioning microscopy techniques in the 1990s allowed many features of chromosome structure to be determined by mapping the physical position of specific DNA sequences with 300 nm precision. Fluorescent in situ hybridization (FISH) and other techniques applied to whole chromosomes show that diVerent chromosomes occupy diVerent regions or ‘‘territories’’ of the interphase nucleus (Cremer et al., 1993; Zink et al., 1998) and has also shown the existence of interchromosomal regions. Similar studies where specific chromosome loci were tagged have been used to measure the real-space distance between genetic markers as a function of the chromatin length between the markers. Remarkably, these studies show interphase chromosomes to have a random walk-like organization at

90

Poirier and Marko

<1-Mb scales and a ‘‘loop’’ organization at 1- to 100-Mb scales (Yokota et al., 1995). Similar studies have been used to study attachments of chromosomes to the nuclear envelope (Marshall et al., 1996). The structure of the bulk of the interphase nucleus remains uncertain, with the role of a nucleoskeleton (‘‘nuclear matrix’’) in chromosome organization still unclear (Pederson, 2000). A FISH study of loci along metaphase chromosomes has also been done to verify that genes are in linear order at >1-Mb scales. However, markers spaced by less than 1 Mb are often seen in random order, indicating that at the corresponding <1-mm scale, metaphase chromatin is not rigidly ordered (Trask et al., 1993). This lack of determined structure is consistent with the flexible loop domain picture of the metaphase chromosome structure (Fig. 2), although one might argue that the fixation used somehow distorted structures at these scales. Structural studies have also been done in vivo by the use of live cell dyes for specific structures, by the incorporation of fluorescent nucleotides (Manders et al., 1999), and by the expression of fusions of chromosome-specific proteins with green fluorescent protein (GFP) (Tsukamoto et al., 2000; Belmont, 2001). One study used both techniques to show that there are 1-mm position fluctuations of interphase chromosome loci from a range of species (Marshall et al., 1997). These fluctuations persisted even in poisoned cells, suggesting that Mb chromosome segments are free to undergo thermal fluctuation, in the manner of flexible polymers. This result is at odds with the idea of a dense, rigid nucleoskeleton and suggests instead that chromosomes have intermittent attachments, with Mb regions of chromatin free to move on micrometer-length scales. A study of the yeast (Saccharomyces cerevisiae) interphase chromosome structure by Dekker et al. (2002) is unique in its methodology and results. This study used cross-linking of isolated nuclei, followed by restriction enzyme digestion. The fragments were self-ligated, and the resulting fragments were polymerase chain reaction amplified and analyzed. The result was a statistical ‘‘map’’ of in vivo chromatin contacts, giving a statistical three-dimensional chromosome model. This technique may provide a way to map chromosome structure and dynamics in unprecedented detail. 3. Chromosome-Folding Proteins Identified Using Cell-Free Chromosome Assembly Systems An alternative to the deconstruction of chromosomes from live cells is to study chromosomes assembled in vitro using cell-derived factors. Xenopus egg extracts provide an excellent system for doing this, allowing the conversion of Xenopus sperm chromatin into either interphase nuclei or metaphase-like chromatids (Smythe and Newport, 1991). This system has

2. Micromechanical Studies of Mitotic Chromosomes

91

permitted the identification of proteins thought to be critical to organizing mitotic chromosomes, most notably the SMC protein family (Hirano and Mitchison, 1994; Strunnikov et al., 1993, 1995; Strunnikov, 1998). Hirano and Mitchison (1994) showed that if the XCAP-C/E proteins (two of the SMC proteins in Xenopus) were removed from the in vitro mitotic chromosome system, then only a cloud of tangled chromatin fibers would result instead of mitotic chromatids. Furthermore, anti-XCAPs were found to destabilize assembled mitotic chromatids, indicating that XCAP-C/Es were needed both for assembly and for maintenance of the mitotic chromosome structure. Hirano and Mitchison (1994) also found that the XCAPs were localized inside the mitotic chromatids, possibly on a helical or lattice structure. Further work of Hirano and co-workers established that XCAPs in ‘‘condensin’’ complexes (Hirano, 1997) show an ATP-dependent DNA supercoiling capability that was interpreted in terms of a DNA coiling function (Kimura et al., 1997, 1999). Other SMC-type proteins have other roles in modulating the chromosome structure (Strunnikov and Jessberger, 1999), notably holding sister mitotic chromatids together during prophase (‘‘cohesins,’’ see Michaelis et al., 1997; Guacci et al., 1997; Losada et al., 1998). Losada and Hirano (2001) suggested that the balance between condensin and cohesin SMCs determines large-scale metaphase chromosome morphology. Many questions remain about the SMC proteins, which have a remarkable structure of 100-nm coiled-coils with a central hinge (Melby et al., 1998) and ATP-binding and -hydrolyzing end domains (Fig. 5). Their distribution inside mitotic chromatids, clear revelation of their function in chromosome condensation, and whether they are the major proteins of the ‘‘mitotic protein scaVold’’ all remain unknown. Thanks to the biochemical characterizations described earlier, these questions may be answered through gradual ‘‘biochemical dissection’’ (Hirano, 1995, 1998, 1999). Transcriptionally functional interphase nuclei can also be assembled readily from Xenopus egg extracts (Smythe and Newport, 1991). This system has been used to analyze nuclear assembly, transcription, and nuclear import. To date, there has been only limited progress in identifying chromosome-organizing nuclear proteins based on in vitro-assembled nuclei, presumably due to the increased complexity of the interphase nucleus relative to the metaphase case. 4. Topoisomerase II One of the most common proteins found in mitotic chromosomes is topo II (Gasser et al., 1986), the enzyme that passes dsDNA through dsDNA and which is assumed to be the enzyme primarily responsible for removing entanglements of chromatin fiber during chromosome condensation and segregation. This idea is strongly supported by experiments using

92

Poirier and Marko

Figure 5 SMC proteins play a role in the higher order mitotic chromosome structure. Condensin SMC complexes (a) include a 100-nm-long hinged dimer with ATP- and DNAbinding domains at each end. These complexes are thought to bind together (b), or perhaps coil (c), chromatin fibers.

mitotic Xenopus egg extracts: when topo II is depleted, sperm chromatin just forms a cloud of apparently entangled chromatin fibers, which never form condensed and segregated chromatids (Adachi et al., 1991). However, a second hypothesis that topo II also plays a structural role in mitotic chromosomes is contentious (Warburton and Earnshaw, 1997). Immunofluorescence studies show that topo II is localized into helical tracks inside chromatids (Boy de la Tour and Lamelli, 1988; Sumner, 1996). Combined with the fact that topo II interacts with two strands of dsDNA, this result suggests that topo II might be part of an internal protein structure in the mitotic chromosome. However, other experiments use salt treatment to deplete topo II from mitotic chromosomes, with no apparent deleterious eVect on their structure (Hirano and Mitchison, 1993). It has been reported that the axial distribution of topo II may be triggered by cell lysis while in vivo it is mobile (Christensen et al., 2002). These experiments can be reconciled by supposing that topo II is critical for establishment of the mitotic chromosome structure by allowing dsDNA disentanglement, that it is present in a high copy number on the assembled mitotic chromosome, but

2. Micromechanical Studies of Mitotic Chromosomes

93

that it does not play a crucial role in holding the mitotic chromosome together. 5. Chromosomal Titin Titin is a huge protein of filamentous structure and is the elastic restoring element of sarcomeres (Trinick, 1996). The mechanical response of isolated titin molecules has been measured precisely using single molecule manipulation (Kellermayer et al., 1997; Reif et al., 1997; Tskhovrebova et al., 1997). Because of its structure, a long series of independently folded globular domains, titin displays initial linear elasticity followed by a series of irreversible force jumps associated with successive domain unfolding events. Remarkably, it was found that muscle titin antibodies localize onto mitotic chromosomes (Machado et al., 1998; Machado and Andrew, 2000a,b). It has been therefore speculated that a putative chromosomal titin might play a role in chromosome condensation and might be a contributor to the chromosome elastic response (Houchmandzdeh and Dimitrov, 1999).

E. Why Study Mitotic Chromosomes Micromechanically? The structure of chromosomes beyond the nucleosomal scale is poorly understood, partly because chromosomes are dynamic, having quite diVerent structures at diVerent points of the cell cycle, and partly because chromatin is inhomogeneous and soft. In particular, mitotic chromosomes are soft, without a regular structure that can be studied by X-ray crystallography. The mitotic chromosome is a logical starting point for the study of chromosome structure, as in this stage of the cell cycle the chromosome is packaged (condensed), the chromosomes are segregated from one another, and gene expression is halted, all of which appear to be simplifying factors. In addition, study of the mitotic chromosome structure will presumably shed light into the mechanism of chromosome disentanglement and condensation (Hirano, 2000), and lessons learned from study of the mitotic chromosome may be applicable to the presumably more diYcult problem of understanding the interphase nucleus. Basic questions about the mitotic chromosome of interest to us include the following: What is the physical arrangement of chromatin fiber (randomly or regularly coiled or folded?)? What are the molecules (proteins?) that accomplish this folding? What molecules are necessary to keep the mitotic chromosome folded up? How are the processes of chromosome condensation and disentanglement coordinated? All of these questions have a mechanistic as well as a biochemical character and might be attacked using a combination of biochemical and micromechanical experimental methods.

94

Poirier and Marko

In addition to studying chromosome structure, biophysical chromosome experiments provide information relevant to understanding a range of in vivo chromosome biology questions. For example, stresses applied to chromosomes are known to play a role in chromosome alignment and segregation during mitosis (Alut and Nicklas, 1989; Li and Nicklas, 1995, 1997; Nicklas, 1997, 1998; Nicklas et al., 1994, 1995, 2001; King et al., 2000). Kinetochore chromatin elasticity is central to a recent model for the capture of mitotic chromosomes on the mitotic spindle (Joglekar and Hunt, 2002), and chromosome stretching has been used to study the roles of specific proteins in chromatin compaction (Thrower and Bloom, 2001). Chromosome stiVness has also been proposed to play a role in the mechanism of meiotic synapsis (Kleckner, 1996; Zickler and Kleckner, 1999). The next sections focus on our experiments that seek to study mitotic chromosome structure using their elastic response. Elegant and pioneering experiments of Nicklas (1983) showed that meiotic metaphase I chromosomes have well-defined elastic properties. We have carried out studies reaching the same general conclusion for mitotic chromosomes removed from amphibian cells. Combining micromanipulation and the in situ reaction techniques of Maniotis et al. (1997), we are able to monitor the elastic response of whole chromosomes while biochemical reactions are being carried out on the underlying chromatin. The goal of these studies is to diagnose changes in chromosome structure made biochemically via observation of changes in chromosome elasticity.

III. Stretching Elasticity of Chromosomes The extensibility of chromosomes has been studied by a number of researchers and for many years (Callan, 1954; Bak et al., 1977, 1979; Nicklas, 1983; Claussen, 1994). Mitotic chromosomes can be observed to occasionally be stretched out by spindle forces (Fig. 6), and their extensibility is possibly a result of the fact that they contain up to meter-long DNA molecules. However, when studied in more detail, mitotic chromosomes display remarkable elasticity, with stretching properties reminiscent of a network of highly elastic filaments. This section presents some history of chromosome extensibility studies and then describes experiments studying chromosome-stretching elasticity. A. Lampbrush Chromosomes One of the earliest discussions of extensibility of chromosomes was by Callan (1954), who carried out manipulation experiments on amphibian lampbrush chromosomes using glass microneedles. The lampbrush phase

2. Micromechanical Studies of Mitotic Chromosomes

95

Figure 6 A newt (N. viridescens) tissure culture cell showing a chromosome being stretched to about twice its native length by the mitotic spindle during anaphase. Spindle forces are known to be on the order of 1 nN, indicating that the force constant of a whole chromosome is on a similar scale. Bar: 20 mm. Photo courtesy of Professor J. Tang.

occurs during female meiotic prophase in birds and amphibians and has played a special role in cell biology for three reasons. First, lampbrush chromosomes are huge, even by amphibian standards, up to 1 mm long. Second, they display large flexible loop domains tethered to a central axis (Gall, 1956). The basic idea of chromatin loops tethered to a central chromosome axis, clearly the case for lampbrushes, has been used as a basic model for chromosome structure at other cell stages, notably mitosis. Third, the large lampbrush loops are ‘‘puVed up’’ by RNA transcripts coming oV tandem polymerases. Electron microscope observation of the tandem transcription units along lampbrush loops provided early and convincing evidence of the processive nature of transcription (Miller and Beatty, 1969; Miller and Hamkalo, 1972). Particularly this aspect of lampbrush structure continues to be an area of active research (Morgan, 2002). Callan (1954) carried out lampbrush chromosome-stretching experiments using glass microneedles and observed that they could be stretched to centimeter lengths. Observations of DNAase breakage of lampbrush

96

Poirier and Marko

chromosomes by Callan and Macgregor (1958) and Gall (1963) were used to support the hypothesis that each chromatid contains a single linear DNA. This hypothesis, which has been proven to be true for small eukaryote chromosomes (e.g., for yeast, by isolation and analysis of whole genomic DNAs), is assumed to be true for all eukaryotes. In addition, these experiments made clear that lampbrush chromosomes are held together by nucleic acid and not by an internal non-DNA structure. The highly quantitative work of Gall (1963) further established that the large lampbrush loops are extended regions of individual chromatids. Marvelous pictures of lampbrush chromosomes can be found in Callan (1986); it should be noted that the large loops are apparently not in sharp focus, despite the use of flash photography. This is because the loops are in motion, i.e., undergoing thermal conformational fluctuation (Callan, 1986). This feature of lampbrush chromosomes is an example of the flexible polymer behavior of chromatin on a huge and directly observable scale (Marko and Siggia, 1997b).

B. Mitotic Chromosome Extensibility and Elasticity As mentioned earlier, direct observation of stretching of chromatids by the mitotic spindle, plus the huge length of DNA per chromatid, leads naturally to the notion that mitotic chromosomes should be extensible. This expectation was verified in work by Nicklas and Staehly (1967), who used microneedles to hook chromosomes inside grasshopper spermatocytes and demonstrated that meitoic chromosomes (metaphase I through anaphase I) were extensible and elastic, i.e., would return to native length after being stretched by up to eight times. 1. Nicklas’ Study of Chromosome Elasticity in Grasshopper Spermatocytes The first experiment to quantify the elastic response of a chromosome in vivo was carried out by Nicklas (1983) using microneedles to carry out experiments inside living cells. The cells used were grasshopper (Melanoplus sanguinipes) spermatocyte cells, which have a soft cell cortex that allows needles to grab chromosomes without breaking the cell membrane (a few additional experiments on cricket spermatocytes are also reported). Then, forces were measured by observing the bending of the microneedle and then calibrating the force needed to cause such bending. Microneedles were used that required between 0.076 and 0.25 nN/mm of deflection (1 nN ¼ 109 Newton; recall 1 Newton ¼ 1 kg m/s2). Using a film analysis technique, the minimum resolvable deflection of about 0.25 mm gave a force resolution of roughly 0.05 nN.

2. Micromechanical Studies of Mitotic Chromosomes

97

The spermatocytes go through two divisions, which reduce the original four sets of homologous chromatids to the single chromatids of sperm. Nicklas studied the first meitoic division, focusing on anaphase I. Although the main focus of this remarkable paper is on spindle force generation (the maximum force that could be applied to a chromosome by the spindle was found to be nearly 1 nN), Nicklas also reported a complete series of measurements of chromosome elasticity. Nicklas (1983) noted that during anaphase I it was possible to measure the elastic response of one and two chromatids independently by carrying out experiments on chromosomes either before or after their chromatid separation (during anaphase I, the chromatids ‘‘unpeel’’ except for the kinetochore). Using a statistical analysis of data on a number of chromosomes, he showed that attached pairs of chromatids required twice as much force to be doubled in length as did single chromatids. The elasticity observed was linear [force proportional to change in length and to cross-sectional area, see Eq. (2)]. The force needed to double a grasshopper meiotic anaphase I chromosome (two chromatids) was determined to be f0 ¼ 0.75 nN; single chromatids were found to have f0 ¼ 0.32 nN (when reading Nicklas’ paper, keep in mind 1 nN ¼ 104 dyne). This result was used to infer that the (average) Young-stretching modulus of an anaphase I chromosome was 430 Pa (again, note 1 Pa ¼ 1 N/m2 ¼ 10 dyne/cm2). The range of linear elastic response was reported to be at least up to
L/L ¼ 2 (threefold extension). The experiments of Nicklas (1983) are superb in being in vivo measurements, which are suYciently quantitative that it is completely convincing that the elastic response of the chromosomes, and not some aspect of the cell membrane or cytoskeleton, is being measured. However, this depended on the very fluid cell surface of insect spermatocytes (Nicklas, 1983; Zhang and Nicklas, 1995, 1999), a feature not shared by mammalian somatic cells. This is emphasized by Skibben and Salmon (1997), who were able to do elegant chromosome manipulations inside cultured newt lung cells during mitosis only using very stiV microneedles, with consequently no possibility to use their bending to measure force. Nevertheless, micromanipulation techniques based on microneedles have been used to carry out remarkable studies of spindle mechanics and regulation in somatic vertebrate cells. 2. Stretching Mitotic Chromosomes after Their Removal from Cells Given that stretching chromosomes inside mitotic vertebrate cells is not possible, the next best approach to study chromosome stretching is to remove chromosomes from cells into the cell buVer. This approach will always be subject to the criticism that chemical conditions outside the cell will alter chromosome structure, but using comparisons with available

98

Poirier and Marko

in vivo information, the relationship between in vivo and ex vivo chromosome structures can be understood. As described later; our own experiments, combined with those of others, convinced us that there is little or no change in chromosome structure, at least initially after removal from a mitotic cell. Classen and co-workers (1994) noted that chromosomes prepared for metaphase spread karyotyping could be highly extended. That group has used chromosome stretching to develop high-resolution chromosomebanding techniques (Hliscs et al., 1997a,b). However, the first quantitative measure of the elastic response of a mitotic chromosome extracted from a cell was carried out by Houchmandzadeh et al. (1997) using a technique reminiscent of that of Nicklas (1983). The experiments of Houchmandzadeh et al. (1997) were done on mitotic cells in primary cultures of newt lung epithelia (Notophthalmus viridescens). This organism is attractive for chromosome research because it is a vertebrate with relatively few (haploid n = 11), large (haploid genome  35 pg of dsDNA) chromosomes (Gregory, 2001). Each N. viridescens chromatid therefore contains about 3 pg = 3 Gbp, or about 1 m, of dsDNA. At metaphase, the chromosomes are between 10 and 20 mm long and have a diameter of about 2 mm. Newt epithelia cells are cultured easily as a monolayer on dishes built on cover glasses, which are open to room atmosphere, making them excellent for micromanipulation experiments (Reider and Hard, 1990). Houchmandzadeh et al. (1997) used glass micropipettes (inside diameter 2 mm; Brown and Flaming, 1986) to puncture mitotic cells and then to grab onto the chromosomes. The micropipettes were introduced into the open culture dish from above using an inverted microscope. Chromosomes were grabbed by aspirating the chromosome end into the pipette opening, with the other chromosome end anchored in the cell. The main method used by Houchmandzadeh et al. (1997) to apply controlled stretching forces to chromosomes was to use aspiration into a pipette that had been treated with bovine serum albumin so that the chromosome could slide freely while in contact with the bore of the pipette. The chromosome acted as a piston, and by controlling the aspiration pressure, it could be stretched. This technique allows sensitive measurements, but has the defect that the chromosome–pipette seal is not perfect, and the ‘‘piston’’ will be leaky. This results in an overestimation of the modulus, as part of the pressure applied to the pipette drives flow. The results were essentially that mitotic chromosomes are elastic, with a Young modulus estimated to be approximately 1000 Pa at prometaphase (i.e., chromosomes condensed, but not yet attached to spindle), compatible with the results of Nicklas after taking into account the flow eVect mentioned earlier. Over a range of twofold extension, the elasticity was remarkably linear (see Fig. 8; Houchmandzadeh et al., 1997). Experiments

2. Micromechanical Studies of Mitotic Chromosomes

99

were also carried out just after nuclear envelope breakdown (end of prophase), and it was found that chromosomes had a higher elastic modulus Y = 5000 Pa. In addition, Houchmandzadeh et al. (1997) discussed the result of severe deformation of chromosomes using untreated pipettes to which chromosomes adhere permanently. It was found that prometaphase chromosomes could be extended to as large as 100 times their native length without breaking. For extensions beyond 10 times length, the chromosomes did not return to their native length, plus a number of eVects due to ‘‘plastic’’ deformation of chromosomes were observed. When rapid extensions were made, the native chromosome could be converted to a thin fiber that was much stiVer than the native chromosome. Using calibrated pipette bending, the thin fiber was found to have Y  10,000 Pa. Finally, the thinned chromosome was observed to coil helically after stress was released. Following the study of Houchmandzadeh et al. (1997), we further developed the micropipette-based manipulation technique in order to more quantitatively measure newt chromosome mechanical properties (Poirier et al., 2000). Primary cultures of newt lung epithelia were used using the pipettes to tear holes in the cells. Calibrated micropipette bending was used as the force measurement scheme for chromosomes removed from cells and suspended between two pipettes. This allows both ends of the chromosome to be monitored, and therefore chromosome extension can be controlled precisely. Digital image acquisition and analysis were used to measure pipette bending. Measurement of the correlation between pipette images allows pipette shifts (and therefore deflections) to be determined to about 10-nm accuracy. Typically pipettes were used with bending moments 1 nN/mm of deflection, setting a limit on our force resolution of 0.01 nN = 10 pN. In practice, force resolution is usually limited by slow mechanical drifts of the pipettes. Figure 7 shows the experimental setup as viewed in the microscope. Measurements of Poirier et al. (2000) of the force–extension response of single mitotic (prometaphase) chromosomes are shown in Fig. 8. A completely reversible elastic force response was observed for extensions up to about five times native extension, with a force constant f0  1 nN. Given the 1.6 mm diameter of the chromosomes, this corresponds to a Young modulus near 500 Pa, near to the value obtained by Nicklas (1983). [The 300 Pa quoted in Poirier et al. (2000) is based on a slight overestimate of the chromosome thickness; our current best estimate is a prometaphase chromosome diameter of about 1.6 mm.] Although on the same order of magnitude as the modulus measured by Houchmandzadeh et al. (1997), the lower modulus observed by Poirier et al. (2000) indicates that the aspiration technique overestimates chromosome elasticity.

100

Poirier and Marko

Figure 7 A two-pipette chromosome experiment as carried out in Poirier et al. (2000). (a) Schematic diagram of the experimental setup. Two pipettes are used to hold a mitotic chromosome, with one pipette fabricated with a deflection force constant 1 nN/mm to allow chromosome tension to be measured. A third pipette can be moved near to chromosome to microspray reagents for combined chemical–micromechanical experiments (see Section VI). (b) Example images collected during the force–extension experiment. As the right pipette is moved, the left pipette is observed to deflect. Digital image analysis allows pipette deflections to be measured to an accuracy of about 10 nm. Bar: 10 mm.

To date we have carried out about 100 chromosome stretching experiments on newt mitotic chromosomes, and in accord with Nicklas (1983), we find appreciable variation in the force constant, roughly from f0 = 0.5 to 2 nN (see histograms of Fig. 8). Unfortunately, there are no apparent cytological markers on newt chromosomes (for a karyotype, see Hutchison and Pardue, 1975) so we are unable at present to determine whether particular newt chromosomes have consistently higher or lower force constants. It might be possible to correlate the chromosome elastic response with the chromosome number by in situ karyotyping, e.g., using sequence-sensitive DNA dyes, following force measurement. A feature of chromosome stretching that is quite obvious in all the aforementioned studies is that mitotic chromosomes do not become thin as they are stretched in the reversible elastic regime. Our measurements (Poirier et al., 2000) indicate that the fractional decrease in chromosome width is less than 0.1 times the fractional chromosome length increase. For a solidly bonded elastic medium, this ratio (called the Poisson ratio, see Fig. 8a, inset) is usually close to 0.5, corresponding to volume conservation. In contrast, the volume of a mitotic chromosome actually increases as it is being

2. Micromechanical Studies of Mitotic Chromosomes

101

Figure 8 Force–extension data for newt chromosomes. (a) Data from Poirier et al. (2000) for primary cultures of newt lung cells. The diVerent curves show successive extension–retraction cycles; their coincidence indicates that the chromosome has reversible elasticity over the fourfold range of extension shown. The elastic response is nearly linear, and the initial force increase shows that the chromosome force constant is about 1 nN. (Inset) The fractional change in chromosome width as a function of extension, and the chromosome Poisson ratio is less than 0.1. (b) Data for the newt TVI cell line for small extensions (up to two times native length) after chromosome extraction using dilute Triton X-100 (see text). In this range, the chromosome elastic response is strikingly linear, again with a force constant near 1 nN. (c) Histogram of force constants of 84 extracted newt prometaphase chromosomes, plus histograms of in vivo force constants of grasshopper spermatocyte metaphase I chromosomes (two chromatids) and single chromatids (Nicklas, 1983). Single chromatid grasshopper data had forces doubled for direct comparison with the two chromatid data sets. Distributions of force constants are similar in the newt and grasshopper systems.

102

Poirier and Marko

stretched. This can only occur if the fluid medium surrounding the chromosome flows into it as it is stretched, which in turn indicates that the chromatin fibers inside a mitotic chromosome do not adhere to one another. In Poirier et al. (2000), the behavior of chromosomes under high extensions (5 to 50 times native length) is also reported. In contrast to Houchmandzadeh et al. (1997), high extensions were studied using very slow rates of strain, typically 0.01/s (i.e., chromosome length is doubled in 100s). At these slow strain rates, it is found that permanent plastic deformation of chromosomes occurs beyond about fivefold extensions. Remarkably, after extensions beyond about 30 times native length, when stress is released, the chromosome relaxes to an elongated and swollen structure with a very low elastic modulus. This ‘‘ghost’’ chromosome appears to have the same histone content as the native chromosome, as assayed using fluorescent-labeled antihistone introduced in situ. This irreversible swelling behavior is in accord with the Poisson ratio result discussed earlier, again suggesting that the chromatin fibers inside the mitotic chromosome are not adhering to each other strongly. In the interpretation of Poirier et al. (2000), the ‘‘ghost’’ chromosome results from mechanical rupturing of the cross-linking elements that hold the chromosome in its condensed mitotic form. Following the work of Poirier et al. (2000), we further developed a number of aspects of the newt chromosome experiment. First, we obtained a newt eye lens epithelial tissue culture line (TVI line; Reese, 1976), which provides many more metaphase cells per experiment dish and avoids all the troubles of working with animals and primary cell cultures. Next, we developed a technique of using a micropipette loaded with a 0.05% solution of Triton X-100 in 60% phosphate-buVered saline, (PBS), which we spray onto the surface of a mitotic cell to produce a hole through which the mitotic chromosomes are disgorged. Finally, we now generally anchor the force-measuring pipette to the sample slide rather than placing it on a micromanipulation to reduce its mechanical drift. None of these changes in system or technique resulted in any changes to the results reported in Poirier (2000). The result of a recent experiment is shown in Fig. 8b and indicates that the initial elastic response of a mitotic chromosome is essentially linear, with a force constant near to 1 nN. We have also carried out experiments on Xenopus A6 tissue culture cells. These amphibian cells are very similar to newt cells, but have smaller chromosomes (n = 18, haploid DNA content 3 pg = 3 Gb, or about 150 Mb/chromosome). These chromosomes can be isolated and manipulated at prometaphase; they show the same general elastic properties as newt chromosomes, with a force constant of about 1 nN. This force constant indicates a Young modulus Y1000 Pa (Poirier et al., 2002b).

2. Micromechanical Studies of Mitotic Chromosomes

103

C. In Vitro-Assembled Chromosomes Dimitrov and Houchmandzadeh (1999) carried out an important study of the mechanical properties of mitotic chromatids assembled in vitro using Xenopus egg extracts. It is important to note that the system studied is assembled from sperm DNA and, as a result, isolated chromatids are assembled. Also, the chromosomes assembled in the usual egg extract ‘‘mitotic’’ reaction may have a structure unique to the first division of a fertilized egg, i.e., not precisely the structure of a somatic mitotic chromosome. Micropipettes were used to grab, manipulate, and stretch the chromatids; a force measurement was done via observation and calibration of micropipette bending using the same general scheme as shown in Fig. 7. Micropipette bending stiVnesses were roughly 1 nN/mm. The stretching experiments were carried out in buVer, following chromosome assembly. The in vitro-assembled chromatids display stretching elasticity similar to that of chromosomes isolated from cells. For small extensions, linear elasticity was observed, with a force constant 1 nN, and Young modulus Y1000 Pa. However, for extensions beyond about two times native length, the force observed during retraction is significantly less than the force observed during extension, indicating that irreversible changes have occurred. Finally, for extensions about 10 times native length, the in vitro-assembled chromatids show a force ‘‘plateau’’ and fail mechanically. Houchmandzadeh and Dimitrov (1999) also presented an explanation for the mitotic chromatid elastic response in terms of a titin-like elastic ‘‘core.’’ Roughly, the in vitro chromatids have stretching elasticity rather similar to chromosomes from cells, but are somewhat more fragile at higher extensions. It would be of great interest to have stretching data on replicated in vitro-assembled chromosomes, which would have two duplicate chromatids; replicated chromosomes can be assembled using ‘‘cycling’’ egg extracts (Smythe and Newport, 1991).

D. Summary All the available experimental data on the low-extension stretching elasticity of mitotic chromosomes are in excellent agreement; grasshopper, newt, and frog chromosomes from cells, plus egg extract chromatids, all require roughly 1 nN of force to be doubled in length. This level of force constant corresponds to Young moduli of roughly 500 Pa. Chromosomes from cells display a reversible force–extension response when stretched five times their native length; irreversible stretching occurs for larger extensions, with chromosome breakage occurring for large extensions >20 times native

104

Poirier and Marko

length. In vitro-assembled chromatids are mechanically less robust than in vivo chromosomes. An intriguing feature of the mitotic chromosome elastic response is that chromosome volume increases with initial extension. This indicates that mitotic chromosomes are not bonded solidly together and instead behave rather like a polymer network containing an appreciable amount of cytoplasm (‘‘solvent’’). The initial elastic response of mitotic chromosomes is not due to gross alteration of the chromatin fiber structure, as can be seen from comparisons of chromosome and chromatin (Cui and Bustamante, 2000) elastic responses. Chromatin fibers extracted from chicken eurythrocytes display reversible elasticity with a force constant of roughly 10 pN (pN = 1012 N). Because there will be on the order of a few thousand chromatin fibers in a chromosome cross section (the chromosomes discussed earlier are roughly a micrometer in cross section, and each fiber is roughly 30 nm thick), the 1 nN force at which chromosome length is doubled corresponds to a maximum force per chromatin fiber of a fraction of a piconewton. Therefore, the chromatin fiber structure is not being altered appreciably when chromosomes are being stretched by a factor of two; the initial elastic response of chromosomes must be due to modification of a larger scale condensed chromatin structure. Furthermore, the relatively low modulus of the chromosome indicates that the large-scale chromatin structure is remarkably soft, yet elastic.

IV. Bending Elasticity of Chromosomes During mitosis, chromosomes are observed to be bent by spindle-generated forces; knowing the bending stiVness of a mitotic chromosome therefore gives additional information about forces generated by the mitotic apparatus. However, an additional motivation for measurement of the bending stiVness of a mitotic chromosome is to study the homogeneity of the elastic response, and therefore of structure. The previous section, following the approach of Nicklas (1983), implicitly assumed that chromosomes are uniform elastic media; definition of a Young modulus really makes sense only for homogeneous materials. This assumption can be checked easily: because bending of the rod is just stretching that varies across the rod cross section (the inside of the bend is compressed, while the outside is elongated), if a chromosome has uniform elongational properties, its bending and elongational stiVnesses will be related by Eq. (3). The main result of experiments that compare the elongational and bending stiVness of chromosomes is that in vivo and for chromosomes extracted from cells, bending and stretching properties are related in the way that we

2. Micromechanical Studies of Mitotic Chromosomes

105

expect for uniform elastic media (Poirier et al., 2002b). This indicates that chromosome elasticity is due to the bulk of the cross section of the chromosome and is not mainly due to a thin, stiV, central structure. In contrast, the in vitro-assembled Xenopus chromatids studied by Houchmandzadeh and Dimitrov (1999) are far more flexible than one would expect from their Young modulus of about 1000 Pa. This is a strong indication that in vitroassembled mitotic chromatids have an internal structure distinct from that of in vivo mitotic chromosomes. The method that has been used to measure bending moduli of chromosomes is to measure spontaneous thermal-bending fluctuations. The idea is that a small rod in solution will be hit continually by the molecules of the solvent, and as a result its shape will undergo conformational fluctuations. This is the mechanism by which flexible polymers undergo conformational diVusion. Chromosomes turn out to be flexible enough that their bending fluctuations can be observed. Essentially, the bending modulus B is inversely proportional to the amplitude squared of bending fluctuations, and so by measurement of fluctuation amplitude, bending stiVness can be inferred. This technique has been used to measure the bending elasticity of a number of filamentous cell structures. Elegant experiments of this type by Gittes et al. (1993) were used to measure the bending rigidity of actin filaments and microtubules. Perhaps the simplest experiment to envision is to anchor one end of the filament being studied to a solid object (e.g., a very stiV micropipette) and then observe the fluctuations along the rod (Fig. 9a). As one moves down the rod from the anchor point, the amplitude of fluctuation perpendicular

Figure 9 Experimental arrangements for the observation of thermally excited rod bending. (a) Rod clamped at one end, and the transverse fluctuation amplitude is measured as a function of distance from the pipette along the rod. (b) Short rod measured at three nearby points equidistant along a segment of length L so as to measure the fluctuation angle.

106

Poirier and Marko

to the rod will increase. The precise relation expected for a straight, uniform rod is u2 ¼

32kTL3 4 B

ð5Þ

where the bar indicates the average of the fluctuation–amplitude squared. As one would expect, a higher temperature gives larger fluctuation, a higher bending stiVness B gives lower fluctuation, and fluctuation amplitude increases with distance from the anchor point. Equation (5) applies only near enough to the anchor point so that the typical value of the amplitude u is small compared to the distance to the anchor L. Chromosomes from amphibian cells will turn out to be stiV enough that this is always true. However, in vitro-assembled chromatids are so flexible that they undergo random walk fluctuations along their length, and to analyze their fluctuations, the generalization of Eq. (5) is needed (see Houchmandzadeh and Dimitrov, 1999). A second type of experiment can be done by the observation of bending angles along a short segment of an untethered elastic rod (Fig. 9b). If the positions of three points at the ends and midpoint of a segment of the rod of length L are measured, then a bending angle can be determined. Fluctuations of the bending angle are related to the bending modulus by a relation similar to Eq. (5): 2


 ¼

kTL3 B

ð6Þ

The probability distribution of the angle fluctuations is simply Gaussian: " # ð
Þ2 ð7Þ pð
Þexp  2 2
 and, if it can be measured, provides a check that the fluctuations are thermal and not mechanical noise. This type of measurement can be done in situations where a small rod cannot be manipulated, e.g., for a chromosome inside a cell.

A. Expected Bending Flexibility and Fluctuations of Mitotic Newt Chromosomes The previous section showed that the Young modulus of a mitotic newt chromosome was roughly Y = 500 Pa. Using Eq. (3), a chromosome crosssection radius of r ¼ 0.8 mm, we obtain an expected B = 1.6  1022 N m2, based on the assumption that the chromosome behaves as a uniform elastic

2. Micromechanical Studies of Mitotic Chromosomes

107

medium. Plugging this value of B into Eq. (5), and the maximum chromosome p ffiffiffiffi length L ¼ 20 mm, we find the root mean square fluctuation u2 ¼ 0:3 mm. Thus, the tip of a 20-mm-long newt chromosome should thermally wobble over by roughly half a micrometer, an easily observable fluctuation amplitude.

B. Bending Fluctuations of Chromosomes Extracted from Cells We have measured bending fluctuations for chromosomes extracted from newt cells and immobilized at one end in a pipette (Fig. 10). The main

Figure 10 (continues)

108

Poirier and Marko

Figure 10 Measurement of bending fluctuations for newt (N. viridescens) prometaphase chromosomes. (a) A chromosome is anchored at one end in a pipette, and bending fluctuations are observed. Bar: 4 mm. (b) Amplitude time series as a function of time for the three positions indicated by arrows in part a; the distance from the anchored end is indicated in each segment. With increasing distance from the anchor, the fluctuations increase. (c) Mean square fluctuations versus distance from the anchor point; on a log–log plot data fall on the cubic power law given by Eq. (5). The fit shown allows the bending modulus to be extracted from data.

diYculty in this experiment was finding chromosomes that had no attachments to any cell structures so that a completely free end was obtained. The fluctuations have amplitudes satisfying Eq. (5) and lead to bending moduli between 1 and 31022 N m2 (Poirier and Marko, 2002a), in good accord with the aforementioned estimate. Newt chromosomes show no sign of ‘‘hinges’’ or other easily bent regions along their length. In particular, there is no sign of hinging behavior at the kinetochore. Each newt chromosome that had its bending modulus measured subsequently had its Young modulus measured by attachment of a force-measuring pipette to the free end. The actual Young modulus agreed well will the Young modulus inferred from the bending fluctuations. In the same study (Poirier and Marko, 2002a), chromosomes extracted from Xenopus A6 cells were found to be somewhat more flexible, with bending moduli between 5 and 20  1024 N m2. This flexibility is due to the smaller cross section of the frog chromosomes (r  0.5 mm). The bending

2. Micromechanical Studies of Mitotic Chromosomes

109

modulus value is consistent with measured A6 Young moduli (200 to 800 Pa in our experiments) via Eq. (3).

C. Bending Fluctuations of Chromosomes in Vivo To check the relationship between bending moduli of newt chromosomes extracted from cells and in vivo, it would be useful to have data for mitotic chromosomes in live newt cells. Marshall et al. (2001) first did this using observations of bending fluctuations to measure the bending rigidity of mitotic chromosomes in Drosophila embryo cells. During mitosis, the mitotic spindle induces large bending fluctuations. Marshall et al. (2001) therefore compared native cells (large nonthermal bending fluctuations) with colchicine-treated cells (no microtubules, and therefore much smaller bending fluctuations). The small fluctuations of the Drosophila chromosomes in the colchicinetreated cells led to an estimate of B = 6  1024 N m2 and a Young modulus estimate of 40 Pa. The much larger fluctuations in the native cells were then used to quantify the forces being applied to the chromosomes by the mitotic apparatus. Unfortunately, no stretching data are available for Drosophila embryo mitotic chromosomes; the small size of the chromosomes makes their micromanipulation extremely challenging. We used the basic method of Marshall et al. (2001), colchicine treating newt cells, to obtain in vivo thermal bending data for mitotic chromosomes (Poirier and Marko, 2002a; note that cochicine treatment leads to essentially metaphase chromosomes). Angle fluctuations were measured in vivo over short distances along a number of chromosomes, and the probability distribution was the expected Gaussian form [Eq. (7)]. Chromosome-bending stifnesses were from 0.2 to 0.5  1022 N m2, about a factor of four smaller than obtained for isolated chromosomes. The somewhat smaller values of B obtained in vivo may reflect a change in physical properties due to the chemical diVerences between the cytoplasm and the extracellular medium. Alternately, there may be sources of nonthermal fluctuation that are weak and not disrupted by colchicine treatment. Just as one example, SMC ‘‘condensin’’ proteins have a possible motor function and could result in forces on top of thermal forces, which tend to move chromosomes around. Also, the live cells continue to crawl on their substrate, and it may be that cytoplasmic flows driven by cell crawling cause some nonthermal fluctuations. Because nonthermal forces will generally increase bending fluctuations, we can expect the in vivo measurements to provide lower bounds on B. Thus we conclude that newt chromosomes have B in vivo comparable to that measured in the extracellular medium, roughly 1022 N m2.

110

Poirier and Marko

D. Bending Fluctuations of in Vitro-Assembled Xenopus Chromatids Houchmandzadeh and Dimitrov (1999) measured the bending stiVness of mitotic chromatids assembled in Xenopus egg extracts. They observed that the roughly 20mm-long chromatids were very flexible, finding B = 1.2  1026 N m2. This is about 1000 times smaller than the value of B that we have obtained for chromosomes from Xenopus A6 cells. Thus, in vitro-assembled chromosomes are far more flexible to bend than somatic cell chromosomes from the same species. The bending modulus of in vitro-assembled chromatids is so low that they undergo flexible polymer-like bending fluctuations. The thermal persistence length of in vitro-assembled chromatids is A = B/(kT ) = 2.5 mm, meaning that they should have many thermally fluctuating bends along their length. Indeed, movies of in vitro-assembled chromatids display observable dynamical bending on a few-micrometer length scale (S. Dimitrov, private communication). In collaboration with Professor R. Heald (U. C. Berkeley), we have observed in vitro-assembled chromatids and have verified that they are extremely flexible. To the eye, they behave entirely diVerently from mitotic chromosomes isolated from cells. Paradoxically, in vitro-assembled chromosomes are extremely flexible to bend, yet have a Young (stretching) modulus Y1000 Pa similar to that of mitotic chromosomes from cells. In starker terms, Eq. (3) fails for in vitro-assembled chromosomes by a factor of 1000. This fact led Houchmandzadeh and Dimitrov (1997) to suggest that the in vitro-assembled chromosomes should be organized around a thin core, which would provide stretching elasticity, but with very little bending rigidity. Those authors present a quantitative analysis indicating that one or a few titin molecules, suspected to be a chromosomal component (Machado et al., 1998; Machado and Andrew, 2000a,b), could produce the observed elasticity while being bent easily. In comparison with the 1000-fold larger value of B obtained for chromosomes from cells (recall that for both newt and frog chromosomes from cells, the bending behavior was consistent with the Young modulus), it seems inescapable that in vitro-assembled chromosomes have a diVerent internal structure from chromosomes in somatic cells. Given that in vitro-assembled chromatids are a widely used model system for studying mitosis, this observation may be fairly important. A further interesting question is whether in vitro-assembled chromosomes, which are run through a round of DNA replication so that they are chromatid pairs, have a larger bending rigidity consistent with their Young modulus. One possibility is that cycle 1 of the fertilized frog egg may simply assemble a distinct chromosome structure; another possibility is that the reaction on unreplicated DNA may not be able to fully condense the chromatids.

2. Micromechanical Studies of Mitotic Chromosomes

111

E. Bending of Chromosomes during Mitosis If one observes cells in culture going through mitosis, chromosomes can be observed to be bent during prometaphase as they are being aligned, and then during anaphase as the chromatids are being pulled toward the spindle poles. During anaphase, the chromosomes can be bent quite severely, and to the eye it appears that the chromosome arms are being pulled back by some retarding force. Roughly, the retarding force needed to bend a chromosome into an anaphase ‘‘U’’ shape is the bending modulus divided by the square of the width of the ‘‘U’’ (Houchmandzadeh et al., 1997). For a newt chromosome with B = 1022 N m2 and a U width of a few micrometers, this retarding force is roughly 1011 N. A basic question is whether this force is possibly due to viscous drag. The drag force on the chromosome will be roughly its length times viscosity times its velocity; for newt chromosomes (L ¼ 10 mm, cytoplasm viscosity = 0.01 Pa s, velocity = 0.01 mm/s), we obtain a drag force of about 1015 N. Drag cannot generate the relatively large force needed to bend an anaphase chromosome (Nicklas, 1983). Based on this estimate, it is clear that the force that bends chromosomes and chromatids during mitosis is not simple viscous drag due to the cytoplasm. One explanation suggested by Houchmandzadeh et al. (1997) is that there is a larger amount of friction due to motion of the chromosomes through the cell than estimated on the basis of estimates of cytoplasm viscosity, perhaps because of cytosketetal filaments that oppose the motion of large objects. However, it is worth considering the alternative possibility that opposing forces are applied to the kinetochore and chromosome arms (or telomeres), which are large compared to viscous drag forces.

F. Summary Mitotic chromosomes have well-defined bending properties, and mitotic amphibian cell chromosomes have a bending modulus consistent with their stretching moduli via Eq. (3), implying that their cross sections are relatively homogeneous. Measurements of bending fluctuations inside and outside cells indicate that there is not a large diVerence in bending elasticity caused by isolation of a chromosome. Remarkably, in vitroassembled chromatids are bent far more easily than would be expected on the basis of their stretching modulus. This suggests that in vitro-assembled chromatids have an internal structure qualitatively diVerent from that of metaphase chromosomes in somatic cells.

112

Poirier and Marko

V. Viscoelasticity of Chromosomes The previous three sections emphasized two main points: that chromosomes contain a huge length of dsDNA, organized into chromatin fiber, which itself is a bulky but flexible polymer that is highly elastic, and that mitotic chromosomes have a well-defined stretching elasticity (Y500 Pa), allowing them to recoil into their native form after being stretched up to five times. This stretching elasticity is also responsible for the well-defined bending elasticity of chromosomes in vivo, with a bending modulus in accord with the Young modulus via Eq. (3). Our main focus has been on the reversible, equilibrium stress response, i.e., elastic-restoring forces in the regime where extensions are small enough not to damage chromosome internal structure and where suYcient time has elapsed that the internal stress in the chromosome is in equilibrium with the force applied by the measurement apparatus (e.g., a microneedle or micropipette). If one stretches a chromosome rapidly enough, a stretching force in excess of the equilibrium force will be required, as the stress in the chromosome will be partly due to the intrinsic elasticity, plus additional viscous stress associated with the fact that the chromosome internal structure is not able to reach its equilibrium at each moment in time. The viscous stress can also be thought of arising from the sliding friction of adjacent chromatin domains. This section is concerned with quantifying this eVect, and its conclusion will be that a mitotic chromosome has an immense ‘‘internal viscosity’’ in quantitative terms 105 times the viscosity of water.

A. Observations of Slow Stress Relaxation During mitosis, chromosomes are often stretched out at anaphase due to interchromatid attachments (Fig. 6), and sometimes one can observe what happens when the chromatids release while a large amount of stress (e.g., a total force near to 1 nN) is acting on them. Following stress release, one observes the chromatid end to retract back until it returns to near-native length over a time of a few seconds. Nicklas and Staehly (1967) carried out micromanipulation experiments using microneedles and intact grasshopper spermatocytes and noted that if they stretched a metaphase I chromosome to about eight times its native length and then released it, it would recoil to its native length in about 4 s. This measurement amounts to a determination of the internal viscosity of a chromosome given the subsequent measurement of the grasshopper spermatocyte metaphase I force constant of about 1 nN (Nicklas, 1983).

2. Micromechanical Studies of Mitotic Chromosomes

113

In linear approximation, the situations described earlier are described by the force balance equation: Yx ¼ 0

dx dt

ð8Þ

where x =
L/L, the chromosome extension as a fraction of its native length L (Poirier et al., 2001a). The left side of Eq. (8) is equilibrium elastic stress, just the Young modulus times the strain (note that Y has dimensions of stress, or force per area), and has the linear form observed for whole chromosome elasticity (Fig. 8). The right side, which is zero when the extension is stationary, is the simplest model of viscous stress inside the chromosome and the starting point for the description of viscoelasticity (Landau and Lifshitz, 1986). The constant 0 has dimensions of force per area times time, or viscosity. When applying Eq. (8) to free relaxation of a chromosome, we should keep in mind that the external fluid will contribute to 0 (recall water has viscosity of 103 Pa s); however, this will be a tiny correction to the mainly ‘‘internal’’ viscosity. If at time t = 0 our chromosome is released with extension x0, its subsequent extension [the solution to Eq. (8)] is just a simple exponential decay: xðtÞ ¼ x0 exp½t=t0

9Þ

where the time constant for the decay is t0 = 0 /Y. The observation by Nicklas and Staehly (1967) of t0  4 s and the measurement of Y  500 Pa (Nicklas 1983) therefore give an estimate of that 0  2000 Pa s, more than 106 times the viscosity of water. This is a very rough estimate and involves extrapolation of Eq. (8) into the high-extension regime where nonlinear corrections are important. However, it is clear that the internal relaxation rate of a mitotic chromosome is extremely slow. In our first experiments measuring newt chromosome elasticity we found that in order to obtain reversible elastic equilibrium force responses using micropipettes, very slow extension and retraction rates were necessary (Poirier et al., 2000). If one elongates a newt chromosome by a factor of two over a time less than about 10 s, the extension curve is above the equilibrium force, while the retraction curve lags below it. The nicely reversible extension–retraction curves of Fig. 8 therefore take >5 min per extension– retraction cycle. This eVect is in accord with the observations of Nicklas and Staehly (1967) and with Eq. (8), which can be amended to include the eVect of the external stress applied by a pipette. We subsequently measured the internal viscosity of the prometaphase newt chromosome more carefully (Poirier et al., 2001). We did this using chromosomes attached to two pipettes, simply by moving one pipette rapidly while acquiring visual data for the deflection of the other, force-measuring pipette. The results of this experiment are shown in Fig. 11. Generally, one

114

Poirier and Marko

Figure 11 Measurement of the dynamics of stress relaxation in a single mitotic newt (N. viridescens) chromosome. A chromosome was attached between two pipettes, and then one pipette was moved rapidly while the other pipette deflection was observed, from which the force in the chromosome was inferred. The force jumps up to a peak and decays down to its final equilibrium value. For short extensions the end of decay is observed to be exponential, with a displacement-independent decay time of about 1s. (Inset) The equilibrium chromosome elastic response (line) and final equilibrium forces reached in the step experiments (points) are in accord for small extensions where Eq. (8) is valid.

observes an initial force pulse just shortly after the pipette is moved, followed by a decay to the final equilibrium force. For small step extensions to a final chromosome extension less than about three times native length, the final force decays are nearly all the same exponential shape, in accord with Eq. (8). For small extensions, the final equilibrium force is in agreement with the slow extension–retraction chromosome elastic response, which was measured at the beginning of the experiment (inset, Fig. 11). This indicates that the chromosome is reaching its elastic equilibrium without being damaged appreciably by the rapid pipette motion. Most simply, if we use Eq. (9), we can combine the linear regime exponential relaxation time trelax = 1 s with the chromosome elastic modulus Y = 500 Pa to obtain an internal viscosity 0 = 500 Pa s, similar to the aforementioned estimate based on the observations of Nicklas and Staehly. A more complete analysis takes into account the bending stiVness of the pipette, which turns out to be the dominant elastic element in the experiment (Poirier et al., 2001a); the relaxation time turns out to be 0 L/kp (pipette force constant kp  1 nN/mm, larger than the typical chromosome force constant of 0.1 nN/mm for a 10-mm chromosome segment). This gives an internal

2. Micromechanical Studies of Mitotic Chromosomes

0

115

5

viscosity = 100 Pa s, about 10 times the viscosity of water. The only reasonable way to explain this large internal viscosity is to suppose that when the chromosome is stretched rapidly, the chromatin inside it must rearrange, and the time needed for this rearrangement is on the order of 1 s. One might imagine that since a chromosome must increase in volume as it is stretched (recall the Poisson ratio of a newt chromosome is less than 0.1, see Fig. 8a), perhaps slow relaxation is due to the flow of buVer into the chromosome (Tanaka and Fillmore, 1979). However, if this were the case, one would expect to see the chromosome width undergo a sharp contraction in the step experiments, followed by a gradual decay back up to its equilibrium width. This is not observed (Poirier et al., 2001); the chromosome width jumps to its final equilibrium value much faster than the stress inside it decays. We therefore conclude that the internal viscosity is due to chromatin rearrangements, which we infer to be taking place on 1-s time scales. Our conclusion that chromatin domains take 1-s time scales to relax when the chromosome is stretched implies that there are internal fluctuations of chromatin conformation on the 1-s time scale inside a quiescent mitotic chromosome. Internal motions of chromatin domains can possibly be on this long time scale as the chromosome is one long filament. The ‘‘loop’’ domains that must result from loose condensation of a long filament can be expected to have to undergo slow sliding motions past one another (de Gennes, 1979; Poirier et al., 2001), with slow dynamics.

B. Dynamics of Bending of Mitotic Newt Chromosomes The previous section showed that mitotic chromosomes had well-defined bending elasticity that could be measured via observation of their spontaneous thermal fluctuations, i.e., the mean square fluctuation was proportional to the cube of the length from the anchor point [Eq. (5)]. Furthermore, the bending stiVness obtained was that expected from the Young modulus [Eq. (3)]. Our conclusion was that spontaneous thermal bending fluctuations of mitotic chromosomes have the amplitudes expected for a thin elastic filament, much as observed for actin filaments and microtubules (Gittes, 1993). However, the time dependence of the fluctuations (time series of Fig. 10a) turns out to be quite distinct from the fluctuations for actin and microtubules, where the dominant friction is the external hydrodynamic drag on the filament. In that case, the characteristic time of fluctuations at the free tip grows with the fourth power of the overall length of the filament (discussed, for example, in Harnau and Reineker, 1999): thydro

L4 B

ð10Þ

116

Poirier and Marko

This formula indicates that if one compares the tip fluctuations of two chromosomes that are a factor of two diVerent in length, the tip motion of the shorter one should be about 24 = 16 times faster than the long one. This is not the case (Poirier and Marko, 2002b); in experiments on mitotic newt chromosomes of diVerent lengths, the same tip fluctuation lifetime was observed, even while the time-averaged fluctuation amplitudes obey Eq. (5). The characteristic lifetime of the fluctuations was in each case about 0.7 s. An explanation follows from the fact that the conventional theory of bending fluctuations [Eq. (10)] ignores internal viscosity, which for actin and microtubules is negligible in experimental measurements done to date. When internal viscosity is taken into account, the tip fluctuation time turns out to be the larger of Eq. (10) and the internal viscous relaxation time t0 = 0 /Y. For newt mitotic chromosomes, t0 > tfluct for chromosome lengths satisfying L/r < (0 /)1/4, where the relation between Y and B [Eq. (3)] has been used. Since we know 0 /105 and r  1 mm, the tip fluctuations will not show any length dependence for L < 20 mm, as we have observed (Poirier and Marko, 2002c).

C. Summary Newt mitotic chromosomes, although well-defined elastic solid media, have an exceedingly slow stress response. Measurements of free relaxation of stretched chromosomes and dynamical stretching experiments on newt chromosomes reveal a characteristic time t0 = 0 /Y, where t0  1 s, and where 0  500 Pa s (about 105 times the viscosity of water). This characteristic relaxation time corresponds to the time of spontaneous internal chromatin rearrangements, which are so sluggish that external hydrodynamic friction is irrelevant. These spontaneous internal rearrangements limit the rate at which bending fluctuations occur to a degree that again external hydrodynamic damping, the usual friction relevant to flexible polymers, plays no role. Mitotic chromosomes thus behave like micrometer-thick filaments composed of cross-linked flexible polymers, with slow internal dynamics on the 1-s time scale.

VI. Combined Biochemical–Micromechanical Study of Mitotic Chromosomes The previous sections discussed the physical properties of mitotic chromosomes, focusing on elastic properties in vivo, and for chromosomes extracted from cells. By themselves, they indirectly reveal information about mitotic

2. Micromechanical Studies of Mitotic Chromosomes

117

chromosome structure, such as the flexibility and conformational freedom of chromatin inside the chromosome. However, a more direct and powerful method to analyze mitotic chromosome structure is to use changes in chromosome elasticity as an indicator of changes in the chromosome structure introduced chemically. This section focuses on two sets of experiments: reversible changes in chromosome structure driven by shifting ionic conditions and irreversible changes driven by DNA-cutting enzymes. For these experiments, work of the previous sections provides a baseline elastic response. The ion experiments will provide additional information about the flexibility of chromatin in mitotic chromosomes, while the enzyme experiments will reveal that mitotic chromosomes can be disassembled completely by cutting DNA alone. These experiments suggest that the chromosome elastic response is due to chromatin itself and is not due to some internal non-DNA structure, which is to some degree counter to conventional wisdom concerning mitotic chromosome structure. The strategy of carrying out real-time observation of chemical reactions on whole chromosomes is not new. For example, the drug actinomycin-D was used to release RNA transcripts from the large ‘‘puVed up’’ loops on amphibian lampbrush chromosomes; the subsequent collapse of the loops showed that their open morphology was due to active transcription (Izawa, 1963; Callan, 1982, 1986). For mitotic chromosomes, observations of morphological changes caused by salts, proteases, and DNAase on chromosomes were carried out in the early 1960s (Cole, 1967). More recently, Maniotis et al. (1997) carried out a series of experiments on clusters of metaphase chromosomes extracted from cells using microneedles. In that work the eVects of a wide range of chemicals, including salts and nucleases, were studied by observing morphological changes in the light microscope.

A. Whole Genome Extraction Experiments Maniotis et al. (1997) developed a technique for extracting whole genomes from human and bovine tissue culture cells in both interphase (i.e., from the nucleus) and during mitosis. They used microneedles to ‘‘harpoon’’ either nucleoli during metaphase or mitotic chromosomes. Chemical experiments were then done on the chromosomes while observing on the inverted microscope, using drops of enzyme introduced into the slide on which the cells are cultured and manipulated. Maniotis et al. (1997) emphasized that when these mechanical extractions are done, the whole genome (i.e., essentially all the chromatin) is obtained due to mechanical connections between the chromosomes. These interchromosome connections are invisible fibers (evidenced by their mechanical

118

Poirier and Marko

eVects), which are RNa ase and protease insensitive, but which are cut by DNa ase and micrococcal nuclease. They therefore concluded that the chromosomes of mammalian cells are connected together at the chromatin level, i.e., that the molecule that holds the genome together is DNA. A very wide range of interesting experiments have been done by Maniotis et al. (1997), with emphasis on interphase chromosome organization and chromosome–cytoskeleton interactions, topics that are slightly outside of the focus of this review. However, two other experiments are done of particular relevance here. One class of experiments involves shifts in ionic conditions. It was observed that mitotic chromosomes can be decondensed rapidly by the introduction of drops of high concentrations of ions (500 mM MgCl2, 500 mM CaCl2, 1 M CuCl2, 1 M NaCl) and that this decondensation was reversible, unless very high concentrations of ions were used. These experiments indicate that mitotic chromatin is compacted by interactions of primarily ionic character and suggest that the condensation of the mitotic chromosome is not a precise folding, as it can be cycled chemically on a short time scale. Second, Maniotis et al. (1997) used drops of proteases (5 mg/ml trypsin and 50 mg/ml proteinase K) to examine the role of proteins in mitotic chromosome organization. It was found that these enzymes cause rapid decondensation of chromosomes into ‘‘swollen clouds,’’ a result consistent with results of experiments discussed by Cole (1967). Remarkably, Maniotis et al. (1997) found that the decondensed chromosomes could be recondensed by adding linker histone H1 at 1 mg/ml. Core histones and other nonhistone proteins could not produce this eVect. Apparently the main eVect of protein digestion is to disrupt nucleosome-stacking interactions, as mitotic chromosome morphology can be ‘‘rescued’’ using H1. It is striking that H1 is suYcient for this rescue, as one might imagine that other, rarer proteins that define a higher order chromatin structure (i.e., above the level of the 30-nm fiber) would be cut by the proteases and that this would limit the degree of recondensation that H1 could aVect. Perhaps the large concentration of H1 and its accessibility (H1 is chemically exchanging on short time scales; Lever et al., 2000; Misteli et al., 2000) make it a main target in this experiment, whereas the rare and perhaps other wellburied proteins that stabilize the higher level chromosome structure remain undamaged.

B. Combined Micromechanical–Chemical Experiments Our experiments focus on combining the in situ biochemical reaction approach of Maniotis et al. (1997) with single chromosome elasticity measurements, the aim being to do real-time quantitative monitoring of chromosome

2. Micromechanical Studies of Mitotic Chromosomes

119

structure changes. Our focus is on the study of mitotic chromosome structure, with less emphasis on the interchromosome connections studied by Maniotis et al. (1997). The basic method is to extract mitotic chromosomes to set up two-pipette micromanipulation and then to measure the initial, native stretching elastic response. Then, we bring in a third pipette of inside 4 mm, larger than the chromosome-grabbing pipettes, which has been loaded with some reagent in suitable solution (typically 60% PBS or Tris buVer, pH 7.6; see Fig. 7). This third pipette is brought within 10 mm of the chromosome and then pressure is used to spray the reagent at the chromosome. Calibration experiments using fluorescent dyes show that this results in a jet of reagent exiting the pipette, with concentrations near to those in the pipette up to 20 mm away. Beyond this distance, the reagent diVuses rapidly into the large (1 ml) volume of the sample dish. In a typical experiment, volumes of a few thousand cubic micrometers are typically sprayed (1000 cubic mm is 1012 liter = 1 picoliter). Using this technique, arbitrary reagents can be introduced onto a mitotic chromosome, and the kinetics of reactions can be observed micromechanically to some extent via the force-measuring pipette. Then, when reagent flow is stopped, the chromosome is returned rapidly (<1 s) to the initial (extracellular) buVer condition, in which the eVect of the reaction on elastic properties can be measured quantitatively.

C. Shifts in Ionic Conditions Can Decondense or Hypercondense Mitotic Chromosomes Because chromatin is electrically charged, one can expect its solvency, packing, and elasticity properties to be modified by changes in ionic conditions. In single fiber experiments, Cui and Bustamante (2000) showed that unfolding of single chromatin fibers could be modulated with an univalent salt concentration, and Maniotis et al. (1997) showed that an increased ionic strength could strongly decondense isolated mitotic chromosomes. Using our microspraying techniques, we quantified the eVect of shifts in salt concentration, and we have reproduced the abrupt decondensation eVects reported by Maniotis et al. (1997) with >200 mM univalent and divalent salt concentrations (Poirier et al., 2002). In experiments where force was monitored, we found that applied tension was reduced entirely by high concentrations of Na+ and Mg2+. However, we have also found that in the 20 to 100 mM Mg2+ and Ca2+ concentration range, mitotic chromosomes go through a range of rather strong condensation, generating contractile forces of up to 0.2 nN. As the divalent cation concentration is ramped up from zero, we observe condensation near 20 mM, followed by an abrupt return to the native degree of

120

Poirier and Marko

compaction near 50 mM and then finally the strong decondensation observed by Maniotis et al. (1997) at >200 mM. Our results are in excellent accord with old observations of the morphological change of mitotic chromosomes during ionic condition shifts (Cole, 1967). We emphasize that the use of force measurement, in addition to morphological observation, provides a quantitative measure of the degree to which chromosome structure is changed reversibly, and irreversibly, in these types of experiments (Fig. 12). A similar condensation–decondenation eVect is observed with increasing concentrations of trivalent ions. No compaction occurs with any concentration of Na+ or K+. All these decondensation and condensation eVects occur istropically; under zero tension, the fractional length and fractional width changes are nearly equal. This behavior is reconciled easily with a chromosome model, which is an isotropic network of chromatin fibers and is diYcult to square with an aniostropic chromatin loop attached to scaVold model. The eVects of divalent ions are in line with similar reentrant bundling (i.e., bundling followed by dissolution as the divalent ion concentration is raised) of stiV polyelectolytes observed in DNA (Pelta et al., 1996; Saminathan et al., 1999) and actin solutions (Tang et al., 1996). These phenomena again make clear that native mitotic chromatids are not near their maximum possible compaction [we observe up to a 30% volume decrease using trivalent ions (Poirier et al., 2002)] and also that charge interactions are important in determining the precise degree of chromatin compaction. Divalent ions at 20–100 mM concentrations overcondense chromatin, whereas ionic concentrations >300 mM lead to strong but apparently reversible decondensation. The rapid decondensation and hypercondensation discussed earlier occurred during experiments where the ionic conditions were shifted for less than 1 min. In these experiments, force curves measured after the ionic exposures showed that no appreciable changes in mitotic chromosome

Figure 12 EVects of shifts in ionic conditions on the newt chromosome structure. Images of combined chemical–micromechanical experiments with 30 mM NaCl (a), 500 mM NaCl (b), 20 mM MgCl2 (c), and 300 mM MgCl2 (d). Images show the chromosome before, during, and after an exposure to the diVerent ionic conditions. Plots show the time series of the force the chromosome supports and width of the chromosome. For 30 and 500 mM NaCl, the force decreases and the width increases. However, 20 mM MgCl2 induces an increase in the force and a decrease in width, whereas 300 mM MgCl2 causes a decrease in force and an increase in width. Note that the force signal for the 300 mM spray of MgCl2 shows a brief force ‘‘spike’’ due to transient chromatin condensation induced by the ion concentration sweeping through 10 mM concentrations. The time scale of the response of the chromosome to the ionic shifts occurs on the second time scale and shows that the internal structure of a mitotic chromosome can be changed rapidly. Bars: 10 mm. From Poirier et al. (2002).

2. Micromechanical Studies of Mitotic Chromosomes

121

elasticity, and from our point of view, mitotic chromosome structure, occurred. However, suYciently long exposures (>10 min) to high ionic strength conditions result in an irreversible change in chromosome structure and elasticity (Poirier et al., 2002); high salt exposures are well known to eventually remove even core histones from chromatin. However, appreciable protein removal does not appear to be occuring during the 10-s experiments discussed previously. This result applied to 10 mM concentrations

122

Poirier and Marko 2+

of Mg is an important control for enzyme experiments where divalent ions are often present in about this concentration (see later).

D. Micrococcal Nuclease Completely Distintegrates Mitotic Chromosomes Micrococcal nuclease (MNase) nonspecifically cuts dsDNA and is widely used to cut chromatin up into nucleosomes. This enzyme was an obvious choice to use to examine the eVect of cutting away the chromatin itself, and we were originally motivated to determine whether we could reveal whether the often-discussed internal protein ‘‘scaVold’’ (Paulson and Laemmli, 1977; Marsden and Laemmli, 1979; Earnshaw and Laemmli, 1983; Boy de la Tour and Laemmli, 1988; Saitoh and Laemmli, 1994) was mechanically contiguous. A second aim of the experiments was to determine just how much of the chromosome elastic response was due to chromatin (i.e., dsDNA) itself. Finally, we were motivated by curiosity about how to reconcile the old literature suggesting that DNase could disintegrate mitotic chromosomes (Cole, 1967) with the observation that the protein-rich chromosome scaVold could survive the biochemical removal of histones (Paulson and Laemmli, 1977). We therefore sprayed isolated newt TVI mitotic chromosomes with 1 nM MNase in suitable reaction buVer (60% PBS plus 1 mM CaCl2) with a small tension (0.1 nN) applied initially (Poirier and Marko 2002c). Tension is

Figure 13 Time course of tension in a chromosome, and chromosome morphology, during digestion by 1 nM MNase with an initial tension of 0.1 nN. Spraying starts at 80 s; force decays after 30 s; and the chromosome is cut after 450 s. The spray pipette can be seen in the upper center of the t > 120-s frames. Bar: 10 mm.

2. Micromechanical Studies of Mitotic Chromosomes

123

monitored during the spray experiment (Fig. 13). When the spray starts, the tension jumps briefly due to the slight compaction induced by the divalent Ca2+, but then the tension drops below our force resolution ( 0.01 nN = 10 pN) after 30 s. During this initial period, the morphology of the chromosome is unaVected. However, between the period 100 of 200 s, the chromosome disintegrates and is eventually severed (Fig. 13). This experiment indicates that the force-bearing and structural element of the mitotic chromosome is DNA based, i.e., chromatin itself, and indicates that the chromosome is not held together by a mechanically contiguous internal protein ‘‘scaVold.’’ A second type of experiment was also done where digestion was done with zero applied force and was then stopped before the chromosome was altered morphologically (at 30 s of 1 nM exposure). The chromosome could then be extended into a string of blobs, connected by thin chromatin strands. These strands could then be severed by a brief spray of MNase, where the peak tension applied was <100 pN. This latter experiment makes clear that the disassembly eVect observed using MNase is not tension dependent. The forces applied in this experiment are below those required to break single protein or nucleic acid chains. These experiments, which are in accord with the old literature of disintegration of mitotic chromosomes by DNAase (Cole 1967), indicate that the mitotic chromosome is essentially a cross-linked network of chromatin, i.e., that the higher order chromosome structure is stabilized by non-DNA molecules (most likely proteins), which are isolated from one another. It is diYcult to reconcile our MNase results with the ‘‘textbook’’ model (Lewin, 2000; Lodish et al., 1995; WolVe, 1995) of chromatin loops hanging from an internal mechanically contiguous protein scaVold.

E. Restriction Enzymes with Four-Base Specificity Can Disintegrate Mitotic Chromosomes Following the MNase experiments we carried out experiments with bluntcutting restriction enzymes, which cut dsDNA, leaving no overhangs, at specific base pair sequences (Poirier and Marko 2002c). These enzymes are powerful tools for analyzing the network connectivity of mitotic chromosomes. We selected enzymes that were active in physiological-like buVers (i.e., pH near 7, ionic conditions near 100 mM univalent plus 10 mM divalents). Two enzymes with four-base recognition sequences, AluI (AG^CT) and HaeIII (GG^CC), which occur every 256 bases on random-sequence DNA, were used, and they cut up mitotic chromosomes in the fashion of MNase. Figure 14 shows the result for AluI, which severs the chromosome

124

Poirier and Marko

Figure 14 Time course of tension in a chromosome during digestion by blunt-cutting restriction enzymes. Initial force in all experiments was 0.6 nN; each force curve is normalized to this initial value. Enzyme exposure is from 200 to 550 s. (Bottom curves) AluI completely reducing force to zero (cutting chromosome completely). (Middle curves) Cac 8I only partially relaxing applied tension (partially cutting chromosome). (Top curves) Only small eVects of HincII, DraI, and restriction enzyme activity buVer (no enzyme). The ‘‘step’’ increases in force for the top curves reflect the slight condensation of the chromosome driven reversibly by the divalent ions of the activity buVer.

completely after <100 s (again the force increase seen at the onset of spraying is the reversible condensing eVect of the 10 mM Mg2+ in the enzyme buVer, easily understood in the light of our previous salt experiments). After factoring in the 10-fold reduction in sequence accessibility in chromatin vs bare DNA, this experiment shows that mitotic chromosomes are not cross-linked more often than once every few kilobases. Experiments with six-base recognition sequence enzymes, StuI (AGG^CCT) and DraI (TTT^AAA), show essentially a zero force eVect (Fig. 14 shows DraI), indicating that the accessible six-base sites are rarer than chromatin cross-links. To test the accessibility of six-base-wide sites further, Fig. 14 also shows results for Cac81 where four bases are recognized out of a six-base region (GCN^NGC). This enzyme shows an intermediate eVect, partially reducing applied force, but not totally severing the chromosome. Thus, the six-base site size is partially available to the restriction enzymes. Taken together, these results are all consistent with a chromatin network model with a cross-link every few tens of kilobases and are inconsistent with an internal protein scaVold model (unless the ‘‘scaVold’’ has the form of many small

2. Micromechanical Studies of Mitotic Chromosomes

125

localized protein structures, which is of course again a cross-linked network of chromatin).

F. Summary We have described how micromechanical measurements can be combined with microspraying of reactants to carry out experiments that give unique information about the mitotic chromosome structure. Experiments with varying ionic conditions show that the mitotic chromosome can be decondensed rapidly (1 s) and reversibly or, in the case of 10 mM concentrations of divalent and trivalent cations, hypercondensed. This shows that the mitotic chromosome structure is stabilized at a less than maximal level of condensation by interactions of strong electrostatic character. Experiments with DNA-cutting enzymes show that the contiguous element that defines mitotic chromosome structure is DNA itself. Chromosomes can be dissolved completely by MNase and four-site blunt-cutting restriction enzymes, a result that is hard to reconcile with a chromatin loop scaVold model unless the underlying non-DNA scaVold disassembles spontaneously as a result of cutting of DNA.

VII. Conclusion A. Summary of Physical Properties of Mitotic Chromosomes The previous sections presented data for elastomechanical properties of mitotic chromosomes. To generalize, we have found that mitotic chromosomes stretch and bend as if they are classical elastic media, but with an enormous range of extensibility. We find that mitotic newt chromosomes can be reversibly stretched fivefold and that over this range their elastic response is nearly linear, with a Young (stretch) modulus of about 500 Pa. The mitotic chromosomes of newt and Xenopus are therefore doubled in length by forces 1 nN, similar to the elastic response of grasshopper spermatocyte metaphase I chromosomes (Nicklas, 1983), and also similar to the maximum forces applied by the mitotic spindle to chromosomes during anaphase. The large range of the linear elastic response of mitotic chromosomes is distinct from well-bonded materials such as metals (which fracture when stretched by less than a percent in length) and even most polymer gels. This basic elastic response is in accord with the classic in vivo results of Nicklas (1983) for grasshopper spermatocyte (metaphase I) chromosomes and with Xenopus mitotic chromosomes. Houchmandzadeh and Dimitrov (1999)

126

Poirier and Marko

found similar elastic behavior in their study of in vitro-assembled Xenopus chromatids. Mitotic chromosomes taken out of cells have bending elasticity consistent with their Young modulus via Eq. (3), indicating that they may be considered to be roughly uniform (at least regarding their mechanical properties). In our experiments on newt and Xenopus chromosomes (Poirier and Marko, 2002a) removed from cells, we observed remarkably uniform bending properties, with no sign of ‘‘hinged’’ regions. In cochicine-treated cells we were able to observe bending fluctuations, which indicated that in vivo the bending elasticity of mitotic newt and Xenopus chromosomes is similar to those observed for extracted chromosomes. Our observation of bending stiVness in accord with the stretching elasticity is distinct from results for in vitro-assembled Xenopus chromatids (Houchmandzadeh and Dimitrov, 1999), who found bending elasticity roughly 1/1000 of what was expected on the basis of their 1-nN stretching constant. The tremendous flexibility of the in vitro-assembled chromatids relative to chromosomes in and extracted from cells in striking and suggests to us that in vitro-assembled chromatids have an internal structure quite distinct from metaphase chromosomes in vivo. An important experiment is therefore measurement of the bending flexibility of in vitro chromosomes assembled in egg extracts and cycled through one round of DNA replication. Nicklas and Staehly (1967) noted that after being stretched and released, grasshopper chromosomes slowly recoiled to their native length. We have found that this slow relaxation is an intrinsic physical property of mitotic chromosomes and that the eVective internal viscosity inside a mitotic chromosome is roughly 100,000 times that of water. The same slow relaxation is responsible for making the bending fluctuation time of a mitotic chromosome independent of the chromosome length, which from the point of conventional polymer and colloid physics is peculiar. This slow stress relaxation indicates that chromatin fiber domains inside the folded mitotic chromosome are undergoing continual slow thermal rearrangements with a time scale on the order of 1 s. These slow rearrangements are a natural outcome of the slow motion of entangled domains of chromatin fiber.

B. Elasticity of Mitotic Chromosomes versus Elasticity of Chromatin Fiber Individual chromatin fibers extend by about a factor of two relative to their native length, under tensions of about 5 pN (Cui and Bustamante, 2000; Bennink et al., 2001; Brower-Toland, 2002). There are 2000 chromatin fibers crossing a cross-section of a newt mitotic chromosome (the cross-sectional area 3 mm2, divided by the 800-nm2 area per

2. Micromechanical Studies of Mitotic Chromosomes

127

chromatin fiber, divided by two since part of the internal volume of the chromosome is cytoplasm ‘‘solvent’’). This indicates that we can expect a force of roughly 10 nN (2000 times 5 pN) to be associated with opening up a mitotic chromosome to the end of its elastic response. The paradox here is that at this point a mitotic chromosome is actually extended about 10 its native length. This indicates that the initial elastic response of a whole mitotic chromosome is due to opening of a higher order chromatin structure. Regarding mechanical properties, to date it is not clear what the relation is between extracted chromatin fibers (Cui and Bustamante, 2000) or in vitro-assembled chromatin (Bennink et al., 2001; Brower-Toland et al., 2002) and mitotic chromosomes. One possibility is that the large range of elastic response is due to the opening of SMC-compacted chromatin fiber. Kimura et al (1997, 1999) suggested that condensins coil DNA, and perhaps when applied to chromatin, a spring-like coil results. This could be tested by single molecule experiments on chromatin fiber compacted by active condensins.

C. Ionic Condition Shift Experiments Maniotis et al. (1997) showed that mitotic chromosomes may be decondensed easily and, to some degree, reversibly recondensed, in addition to their observation that chromosomes are interconnected throughout the cell cycle. Our study of eVects of shifts in ionic conditions on extracted mitotic newt chromosomes verifies this result and extends it. First, we find that either lowering or raising the univalent ionic strength relative to the 100 nM Na+ of the amphibian cell culture medium decondenses mitotic newt chromosomes. For lowered ionic strength, decondensation most likely results from opening of a 30-nm fiber into 10-nm ‘‘beads on a string.’’ For a high salt concentration, decondensation is most likely due to a reduction in the range of electrostatic interactions, with consequent opening of folded chromatin (Poirier et al., 2002). For multivalent ions added to cell buVer, condensation (even by Mg2+) occurs at low (10 mM), concentrations; at higher concentrations (>100 mM), decondensation occurs. The low-concentration condensation may be due to ‘‘bridging’’ interactions, i.e., net attractive interactions induced by localization of the multivalent ions (Ha and Liu, 1997). An alternative explanation is that condensation occurs when the charge neutral point is reached, eliminating coulomb repulsion and allowing other attractive interactions to dominate (Nguyen et al., 2000; Nguyen and Shklovskii, 2001). To us, the most significant result of ionic condition shift experiments is that the mitotic chromosome structure can be cycled through condensation

128

Poirier and Marko

or decondensation chemically on short (1 s) time scales. Moreover, when flow is stopped to return to the original buVer conditions, the chromosome immediately returns to its native state, as assayed by morphology and mechanical properties. Irreversible lengthening and softening of the chromosome, presumably due to protein loss, occur only at high (>1 M) ionic strengths after >10-min exposures (Poirier et al., 2002).

D. DNA-Cutting Experiments Cutting dsDNA inside the mitotic newt chromosome with suYcient frequency disconnects the chromosome completely (Poirier and Marko, 2002c). MNase and four-base blunt-cutting restriction enzymes dissolve the chromosome into optically invisible fragments. By far the simplest interpretation of this experiment is that the elastic response and mechanical continuity of the mitotic chromosome are due to chromatin fiber, i.e., DNA itself. A rough estimate of the genomic distance between cuts required to disconnect the chromosome is 15 kb, based on the reduction in eVect of more rarely cutting restriction enzymes (Poirier and Marko, 2002c). Six-base blunt-cutting restriction enzymes have no eVect on the mechanical properties of whole mitotic newt chromosomes.

E. Implications for Structure of the Mitotic Chromosome The experiments reviewed earlier, taken together, consistently suggest that the mitotic chromosome has a network structure, i.e., is organized by isolated chromatin–chromatin attachments (Fig. 15). The purely mechanical measurements (stretching and bending) indicate that chromosome stretching is supported by stress spread across its whole cross section, and therefore that the mitotic chromosome structure appears to be, at the scale of a

Figure 15 Network model of a mitotic chromatid. Black curve indicates the single linear chromatin fiber, and black blobs show isolated non-DNA cross-linking elements. If the chromatin is cut suYciently often, the chromosome will be severed; the non-DNA cross-linkers are not mechanically contiguous through the chromosome.

2. Micromechanical Studies of Mitotic Chromosomes

129

whole chromatid, homogeneous. This hypothesis is also supported by the homogeneous way that whole chromosomes elongate. Dynamic stretching and bending experiments both show that the interior of a mitotic chromosome relaxes extremely slowly, on a roughly 1-s time. We hypothesize that this long time scale is due to chromatin conformational fluctuation and that the long time scale has its origin in entanglements. This implies that mitotic chromatin is not heavily constrained by chromatin-folding proteins, i.e., that there are long stretches of chromatin between ‘‘cross-links.’’ These stretches of chromatin are apparently free to undergo slow conformational motions. Shifts in ionic conditions can decondense and overcondense a mitotic chromosome rapidly. These morphological changes are reversible for short (10 s) salt treatments, and at zero stress are isotropic, again suggesting a homogeneous and not terribly highly ordered mitotic chromatin organization. At least one-third of the chromosome volume is mobile cytoplasm or buVer based on condensation experiments. Finally, our DNA-cutting experiments make clear that the mechanically contiguous structural element of the mitotic chromosome is DNA (i.e., chromatin) itself. The nonchromatin fiber content of the mitotic chromosome must be disconnected. Taken together, these results indicate that the mitotic chromosome must be a network of chromatin fiber, with isolated nonchromatin crosslinks (Fig. 15). We must rule out models for mitotic chromosome structure based on mechanically well-defined non-DNA skeletons or scaVolds. It must be emphasized that the identity of these putative chromatin cross-linkers is unknown; at present the most likely suspects are the condensin-type SMC protein complexes.

F. Future Experiments The combined chemical–micromechanical method for the study of a largescale chromosome structure explored here provides information complementary to usual biochemical assay and microscopy approaches. Traditional biochemical approaches give information about local interactions and the products of chemical reactions. Traditional microscopy gives information about morphology and structure in a given cell state or in a given preparation of molecules. Our approach allows direct study of elastomechanical properties of chromosomes and to observe how those properties are modified dynamically by chemical reactions. Our conclusions about the flexibility and connectivity of chromatin fiber in the mitotic chromosome are diYcult to support by traditional biochemical and microscopy approaches, but are rather obvious results of a combined chemical–micromechanical approach. The question of mitotic chromosome organization therefore can be attacked

130

Poirier and Marko

Table I Symbols, Units and Relationships Symbol A B f0   0 k kT kp L NA r R u Y Distance: Force:

Energy:

Name

Unit

Value

Persistence length m 0.1 Bending modulus Nm2 1022 Force constant N 109 Angle rad 102 Fluid viscosity Pasec 103 Internal viscosity Pasec 100 Boltzmann constant J/K 1.381  1023 Unit of thermal energy J 4.1  1021 Pipette stiffness N/m 103 Length m 20  106 Avagadro’s number 6.022  1023 Cross-section radius m 1  106 Gas constant J/K 8.316 Transverse fluctuation m 1  107 Young modulus Pa 500 ˚ 1 m = 106 mm = 109 nm = 1010 A 1 Newton (N) = 1 kgm/s2 = 104 dyne 1 nN = 109 N 1 pN = 1012 N   1 kT/nm = 4.1 pN (at 300 K = 27 C) 2 2 7 1 Joule (J) = 1 kgm /s = 10 erg = 0.239 cal   1 kT = 0.59 kcal/mol (at 300 K = 27 C) 2 1 Pascal (Pa) = 1 N/m 1 Nm2 = 1 Jm

Comment Mitotic newt chromosome Mitotic newt chromosome Mitotic newt chromosome Chromosome segment bend Water and cell culture media Mitotic newt chromosome Equal to R/NA   At 300 K = 27 C Typical force-measuring pipette Mitotic newt chromosome Mitotic newt chromosome Equal to NAk Mitotic newt chromosome Mitotic newt chromosome

Pressure: Bending modulus: dsDNA: 1 Gbp (109 bp) of dsDNA = 1.013 p (pg = 1012 g) = 0.34 m contour length

profitably by integrating information from all these approaches. Many interesting questions remain to be answered. A very basic variation on the stretching experiments would be the study of relative elasticity of diVerent regions of the mitotic chromosome. Use of labels for centromere, telomere, and euchromatin regions of the chromosome would allow the elasticity of diVerent types of chromatin to be studied. For example, elasticity of the kinetochore is relevant to the modeling of chromosome capture by the mitotic spindle (Joglekar and Hunt, 2002). To back up the network model of the mitotic chromosome, it is be extremely important to analyze the sizes of chromatin fragments produced. This could be done via aspiration of the fragments followed by fluorescence quantification of them after dispersal onto a slide. Also, further digestion experiments using other DNA cutters, RNases, and proteases need to be done. EVects of other chemical modifications of chromatin (e.g., acetylation,

131

2. Micromechanical Studies of Mitotic Chromosomes Table II

a

Physical Properties of Mitotic Chromosomes

Chromosome type

Young’s Bending Experimental modulus, Y Rigidity, B condition (Pa) (Jm)

In vivo N.D.b Drosophila metaphase chromosome In vivo 200 to 1000 Grasshopper (avg = 430) metaphase I and anaphase I chromosome Cell culture 100 to 1000 Newt medium (N. viridescens) prometaphase chromosome In vivo N.D. Newt prometaphase chromosome Xenopus Cell culture 200 to 800 prometaphase medium chromosome Xenopus Cell culture 300 prometaphase medium chromatid Xenopus Xenopus 1000 reconstituted egg extract chromatid

Internal viscosity (Pas)

Reference

6  1024

N.D.

Marshall et al. (2001)

N.D.

100

Nicklas and Staehly (1967), Nicklas (1983)

1–3  1022

100

2–5  1023

N.D.

Houchmandzadeh et al. (1997); Poirier et al. (2000, 2001, 2002a) Poirier et al. (2002a)

0.5–2  1023

N.D.

Poirier et al. (2002a)

5  1024

N.D.

Poirier et al. (2002a)

1.2  1026

N.D.

Houchmandzadeh and Dimitrov (1999)

a Ranges for values indicate the width of distribution of measured values, not measurement errors. b Indicates quantity not measured directly.

phosphorylation) on mitotic chromosome condensation, monitored precisely via elasticity, would also be interesting. These kinds of experiments in general give information on the poorly understood question of enzyme access in dense chromatin. Development of our techniques to study the structural roles of specific proteins might be possible. We have already demonstrated antibody labeling using microspraying for antihistone (Poirier et al., 2000) and for antiXCAPs (unpublished); our method used directly fluorescent labeling and thus observation of primary antibody reactions in situ. The simplest types of experiments would be visualization of targeting in our experiment as a function of chromosome stretching. This general technique might be useful for chromosome mapping (Clausen, 1994; Hliscs, 1997a,b), especially if diVerent parts of a chromosome could be exposed to diVerent reagents using microchannel arrays.

132

Poirier and Marko

More ambitious experiments would use fluorophores, which can generate large amounts of hydroxyls, lysing the antibody targets (Beerman and Jay, 1994). This technique has been used successfully to study the disruption of cytoskeletal proteins and might be used in conjunction with chromosome elasticity measurement to study the eVect of condensin or cohesin disruption on the mitotic chromosome structure. This type of experiment could directly test models of SMC function such as that of Losada and Hirano (2002). Study of the orientational ordering of chromatin using polarization microscopy could be informative. Purified and concentrated nucleosomes have been demonstrated to form chiral liquid crystal phases (Leforestier et al., 1999; Livolant and Leforestier, 2000); optical activity has also been observed for certain chromosomes (Livolant, 1978; Livolant and Maestre, 1988). A major question is whether animal chromosomes have similar liquid crystal organization, either in native or in stretched forms. Some very preliminary experiments on newt chromosomes in our laboratory using the CRI Polscope showed undetectable birefringence for chromosomes stretched up to four times native length. This suggests that ordered domains of mitotic chromatin are smaller than the wavelength of light, i.e., <100 nm, and that appreciable stretching of chromosomes does not induce strong orientational ordering of chromatin. However, further experiments are necessary. Our results suggest that chromatin in mitotic chromosomes is to some degree flexible and so it might be possible to pull chromatin fibers out of them. This could be done using pipettes to pull on small particles coated with antihistone or other chromatin-binding factors. Comparison of mitotic chromatin fiber physical properties obtained in such an experiment, with results of single chromatin fiber mechanical experiments (Cui and Bustamante, 2000; Bennink, 2001), would be interesting. We have repeatedly observed interchromosome fibers between mitotic chromosomes as discussed by Maniotis et al. (1997), and these objects require further study. Initial experiments have verified the result of Maniotis et al. (1997) that these fibers are cut by MNase and therefore contain nucleic acid (most likely DNA). Rough stretching experiments show that these fibers are highly and reversibly extensible, with an estimated force constant in the nanonewton range. These are therefore a more folded structure than the 30-nm fiber, but because they are barely visible in the light microscope, we estimate their thickness to be less than 200 nm. DNA staining and quantification are an objective of our current studies. We also hypothesize that these fibers are telomeric structures (the interchromosome fibers at metaphase almost always come from chromosome ends), and therefore probes for telomere DNA should be tested. An interesting question is whether these fibers are intrinsic to transformed cells (most of our work is in tumor cell lines), and therefore parallel studies in primary cell cultures are of strong interest.

2. Micromechanical Studies of Mitotic Chromosomes

133

Other chromosome structures could be studied by combined chemical– micromechanical techniques. We are interested in comparing mitotic chromosomes to meiotic chromosomes. The range of physical structures occurring during meiosis provides a motivation for micromechanical experiments. Mechanical properties of meiotic chromosomes may play a crucial role in general recombination (Kleckner, 1996; Zickler and Kleckner, 1999) and may be related to polymer physics of the chromatin loops (Marko and Siggia, 1997a). Interphase chromosomes would be extremely interesting to study as isolated objects. Maniotis et al. (1997) used purely mechanical techniques to extract whole interphase genomes, an important first step. We are searching for a biochemical method to open the nuclear envelope to allow more gentle interphase genome extractions. Finally, we note that Hinnebusch and Bendich (1997) have demonstrated that bacterial chromosomes can be extracted and studied physically Cunha et al. (2000a, 2001b) have succeeded in isolating and chemically manipulating Escherichia coli nucleoids, which might also be studied using micromechanical techniques. The wide range of genetic and biochemical tools developed for E. coli, plus the many very basic and open questions regarding the bacterial chromosome structure, make it a highly attractive system for micromanipulation study.

Acknowledgments It is a pleasure to acknowledge the help and advice of Prateek Gupta, Tamar Monhait, Chee Xiong, Eric Siggia, Herbert Macgregor, Peter Moens, Chris Woodcock, Susan Gasser, Nancy Kleckner, Lynn Zecheidrich, Nick Cozzarelli, Tatsuya Hirano, Carlos Bustamante, Didier Chatenay, Bahram Houchmandzadeh, Albert Libchaber, Michael Elbaum, Deborah Fygenson, Peter Moens, Joe Gall, Andrew Maniotis, Paul Janmey, Josef Kas, Wallace Marshall, John Sedat, Rebecca Heald, Abby Dernburg, Stefan Dimitrov, Ulrich Laemmli, Ted Salmon, Lon Kaufman, and Arnold Kaplan. This research would not have been possible without the kind gift of the TVI cell line from David Reese. Experiments at UIC were supported by grants from the Whitaker Foundation, the NSF (DMR-9734178), Research Corporation, the Johnson and Johnson Focused Giving Program, the Petroleum Research Foundation of the American Chemical Society, and the University of Illinois Foundation.

References Adachi, Y., Luke, M., and Laemmi, U. K. (1991). Chromosome assembly in vitro: Topoisomerase II is required for condensation. Cell 64, 137–148. Allemand, J. F., Bensimon, D., Lavery, R., and Croquette, V. (1998). Stretched and overwound DNA forms a Pauling-like structure with exposed bases. Proc. Natl. Acad. Sci. USA 95, 14152–14157.

134

Poirier and Marko

Alut, J. G., and Nicklas, R. B. (1989). Tension, microtubule rearrangements, and the proper distribution of chromosomes in mitosis. Chromosoma 98, 33–39. Anderson, J. D., and Widom, J. (2000). Sequence and position-dependence of the equilibrium accessibility of nucleosomal DNA target sites. J. Mol. Biol. 296, 979–987. Arents, G., Burlingame, R. W., Wang, B. C., Love, W. E., and Moudrianakis, E. N. (1991). The nucleosomal core histone octamer at 3.1 A resolution: A tripartite protein assembly and a left-handed. Proc. Natl. Acad. Sci. USA 22, 10148–10152. Bak, A. L., Zeuthen, J., and Crick, F. H. (1977). Higher-order structure of human mitotic chromosomes. Proc. Natl. Acad. Sci. USA 74, 1595–1599. Bak, P., Bak, A. L., and Zeuthen, J. (1979). Characterization of human chromosomal unit fibers. Chromosoma 73, 301–315. Beerman, A. E. L., and Jay, D. G. (1994). Chromophore-assisted laser inactivation of cellular proteins. Methods Cell Biol. 44, 715–731. Belmont, A. S. (2001). Visualizing chromosome dynamics with GFP. Trends Cell Biol. 11, 250–257. Belmont, A. S., Braunfeld, M. B., Sedat, J. W., and Agard, D. A. (1989). Large-scale chromatin structural domains within mitotic and interphase chromosomes in vivo and in vitro. Chromosoma 98, 129–143. Belmont, A. S., Sedat, J. W., and Agard, D. A. (1987). A three-dimensional approach to mitotic chromosome structure: Evidence for a complex heirarchical organization. J. Cell Biol. 105, 77–92. Bennink, M. L., Leuba, S. H., Leno, G. H., Zlatanova, J., de Grooth, B. G., and Greve, J. (2001). Unfolding individual nucleosomes by stretching single chromatin fibers with optical tweezers. Nature Struct. Biol. 8, 606–610. Boy de la Tour, E., and Laemmli, U. K. (1988). The metaphase scaffold is helically folded: Sister chromatids have predominantly opposite helical handedness. Cell 55, 937–944. Brower-Toland, B. D., Smith, C. L., Yeh, R. C., Lis, J. T., Peterson, C. L., and Wang, M. D. (2002). Mechanical disruption of individual nucleosomes reveals a reversible multistage release of DNA. Proc. Nat. Acad. Sci. USA 99, 1752–1754. Brown, K. T., and Flaming, D.-C. (1986). ‘‘Advanced Micropipette Techniques for Cell Physiology,’’ pp. 139–141. Wiley, New York. Bustamante, C., Marko, J. F., Siggia, E. D., and Smith, S. (1994). Entropic elasticity of lambdaphage DNA. Science 265, 1599–1600. Bustamante, C., Smith, S. B., Liphardt, J., and Smith, D. (2000). Single- molecule studies of DNA mechanics. Curr. Opin. Struct. Biol. 10, 279–285. Callan, H. G. (1954). Recent work on the structure of cell nuclei. In ‘‘Fine Structure of Cells,’’ pp. 89–109. Symposium of the VIIIth Congress in Cell Biology, Noordhof, Groningen. Callan, H. G. (1982). The Croonian Lecture, 1981: Lampbrush chromosomes. Proc. Roy. Soc. Lond. B 214, 417–448. Callan, H. G. (1986). ‘‘Lampbrush Chromosomes.’’ Springer, New York. Callan, H. G., and MacGregor, H. C. (1958). Action of deoxyribonuclease on lampbrush chromosomes. Nature 181, 1479–1480. Christensen, M. O., Larsen, M. K., Barthelmes, H. U., Hock, R., Andersen, C. L., Kjeldsen, E., Knudsen, B. R., Westergaard, O., Boege, F., and Mielke, C. (2002). J. Cell Biol. 157, 31–44. Claussen, U., Mazur, A., and Rubstov, N. (1994). Chromosomes are highly elastic and can be stretched. Cytogenet. Cell Gen. 66, 120–125. Cluzel, P., Lebrun, A., Heller, C., Lavery, R., Viovy, J. L., Chatenay, D., and Caron, F. (1996). DNA: An extensible molecule. Science 271, 792–794. Cole, A. (1967). Chromosome structure. In ‘‘Theoretical and Experimental Biophysics,’’ Vol. 1, pp. 305–375. Dekker, New York. Cook, P. R. (1991). The nucleoskeleton and the topology of replication. Cell 66, 627–637.

2. Micromechanical Studies of Mitotic Chromosomes

135

Cremer, T., Kurz, A., Zirbel, R., Dietzel, S., Rinke, B., Schrock, E., Speicher, M. R., Mathieu, U., Jauch, A., Emmerich, P., et al. (1993). Role of chromosome territories in the functional compartmentalization of the cell nucleus. Cold Spring Harb. Symp. Quant. Biol. 58, 777–792. Cui, Y., and Bustamante, C. (2000). Pulling a single chromatin fiber reveals the forces that maintain its higher-order structure. Proc. Natl. Acad. Sci. USA 97, 127–132. Cunha, S., Odijk, T., Suleymanoglu, E., and Woldringh, C. L. (2001a). Isolation of the Escherichia coli nucleoid. Biochimie 83, 149–154. Cunha, S., Woldringh, C. L., and Odijk, T. (2001b). Polymer-mediated compaction and internal dynamics of isolated Escherichia coli nucleoids. J. Struct. Biol. 136, 53–66. de Gennes, P.-G. (1979). ‘‘Scaling Concepts in Polymer Physics.’’ Cornell University Press, Ithaca, NY. Dekker, J., Rippe, K., Dekker, M., and Kleckner, N. (2002). Capturing chromosome conformation. Science 295, 1306–1311. Dietzel, S., and Belmont, A. S. (2001). Reproducible but dynamic positioning of DNA in chromosomes during mitosis. Nature Cell Biol. 3, 767–770. Earnshaw, W. C., and Laemmli, U. K. (1983). Architecture of metaphase chromosomes and chromosome scaffolds. J. Cell Biol. 96, 84–93. Essevaz-Roulet, B., Bockelmann, U., and Heslot, F. (1997). Mechanical separation of the complementary strands of DNA. Proc. Natl. Acad. Sci. USA 94, 11935–11940. Finzi, L., and Gelles, J. (1995). Measurement of lactose repressor-mediated loop formation and breakdown in single DNA molecules. Science 267, 378–380. Gall, J. G. (1956). On the submicroscopic structure of chromosomes. Brookhaven Symp. Biol. 8, 17–32. Gall, J. G. (1963). Kinetics of deoxyribonuclesase action on chromosomes. Nature 198, 36–38. Gall, J. G. (1981). Chromosome structure and the C-value paradox. J. Cell Biol. 91, 3s–14s. Gasser, S. M., Laroche, T., Falquet, J., Boy de la Tour, E., and Laemmli, U. K. (1986). Metaphase chromosome structure: Involvement of topoisomerase II. J. Mol. Biol. 188, 613–629. Gittes, F., Mickey, B., Nettleson, J., and Howard, J. (1993). Flexural rigidity of microtubules and actin filaments measured from thermal fluctuations in shape. J. Cell Biol. 120, 923–934. Gregory, T. R. (2001). Animal genome size database. http://www.genomesize.com. Guacci, V., Koshland, D., and Strunnikov, A. (1997). A direct link between sister chromatid cohesion and chromosome condensation revealed through the analysis of MCD1 in S. cerevisiae. Cell 91, 47–57. Ha, B.-Y., and Liu, A. J. (1997). Counterion-mediated attraction between two like-charged rods. Phys. Rev. Lett. 79, 1289–1292. Hagerman, P. J. (1988). Flexibility of DNA. Annu. Rev. Biophys. Biochem. 17, 265–286. Harnau, L., and Reineker, P. (1999). Equilibrium and dynamical properties of semiflexible chain molecules with confined transverse fluctuations. Phys. Rev. E 60, 4671–4676. Hinnebusch, B. J., and Bendich, A. J. (1997). The bacterial nucleoid visualized by fluorescence microscopy of cells lysed within agarose: Comparison of Escherichia coli and spirochetes of the genus Borrelia. J. Bacteriol. 179, 2228–2237. Hirano, T. (1995). Biochemical and genetic dissection of mitotic chromosome condensation. TIBS 20, 357–361. Hirano, T. (1998). SMC protein complexes and higher-order chromosome dynamics. Curr. Opin. Cell Biol. 10, 317–322. Hirano, T. (1999). SMC-mediated chromosome mechanics: A conserved scheme from bacteria to vertebrates? Genes Dev. 13, 11–19. Hirano, T. (2000). Chromosome cohesion, condensation, and separation. Annu. Rev. Biochem. 69, 115–144.

136

Poirier and Marko

Hirano, T., Kobayashi, R., and Hirano, M. (1997). Condensins, chromosome condensation protein complexes containing XCAP-C, XCAP-E and a Xenopus homolog of the Drosophila barren protein. Cell 89, 511–521. Hirano, T., and Mitchison, J. (1993). Topoisomerase II does not play a scaffolding role in the organization of mitotic chromosomes assembled in Xenopus egg extracts. J. Cell Biol. 120, 601–612. Hirano, T., and Mitchison, J. (1994). A heterodimeric coiled-coil protein required for mitotic chromosome condensation in vitro. Cell 79, 449–458. Hliscs, R., Muhlig, P., and Claussen, U. (1997a). The nature of G-bands analyzed by chromosome stretching. Cytogenet. Cell Genet. 79, 162–166. Hliscs, R., Muhlig, P., and Claussen, U. (1997b). The spreading of metaphases is a slow process which leads to a stretching of chromosomes. Cytogenet. Cell Genet. 76, 167–171. Horowitz, R. A., Agard, D. A., Sedat, J. W., and Woodcock, C. L. (1994). The three dimensional architecture of chromatin in situ: Electron tomograph reveals fibers composed of a continuously variable zig-zag nucleosomal ribbon. J. Cell Biol. 125, 1–10. Houchmandzadeh, B., and Dimitrov, S. (1999). Elasticity measurements show the existence of thin rigid cores inside mitotic chromosomes. J. Cell. Biol. 145, 215–223. Houchmandzadeh, B., Marko, J. F., Chatenay, D., and Libchaber, A. (1997). Elasticity and structure of eukaryote chromosomes studied by micromanipulation and micropipette aspiration. J. Cell Biol. 139, 1–12. Hutchison, N., and Pardue, M. L. (1975). The mitotic chromosomes of Nothophthalamus (¼Triturus) viridescens: Localization of C banding regions and DNA sequences complementary to 18S, 28S, and 5S ribosomal DNA. Chromosoma 53, 51–69. Izawa, M., Allfrey, V. G., and Mirsky, A. L. (1963). The relationship between RNA synthesis and loop structure in lampbrush chromosomes. Proc. Natl. Acad. Sci. USA 49, 544–551. Jackson, D. A., Dickinson, P., and Cook, P. R. (1990). The size of chromatin loops in HeLa cells. EMBO J. 9, 567–571. Joglekar, A., and Hunt, A. J. (2002). A simple mechanistic model for directional instability during mitotic chromosome movement. Biophys. J. 83, 42–58. Kellermayer, M. S. Z., Smith, S. B., Granzier, H. L., and Bustamante, C. (1997). Foldingunfolding transitions in single titin molecules characterized with laser tweezers. Science 276, 1112–1116. Kimura, K., and Hirano, T. (1997). ATP-dependent positive supercoiling of DNA by 13S condensin: A biochemical implication for chromosome condensation. Cell 90, 625–634. Kimura, K., Rybenkov, V. V., Crisona, N. J., Hirano, T., and Cozzarelli, N. R. (1999). 13S condensin actively reconfigures DNA by introducing global positive writhe: Implications for chromosome condensation. Cell 98, 239–248. King, J. M., Hays, T. S., and Nicklas, R. B. (2000). Tension on chromosomes increases the number of kinetochore microtubules, but only within limits. J. Cell Sci. 113, 3815–3823. Kleckner, N. (1996). Meiosis: How could it work? Proc. Natl. Acad. Sci. USA 93, 8167–8174. Kornberg, R. D. (1974). Chromatin structure: A repeating unit of histones and DNA. Chromatin structure is based on a repeating unit of eight histone molecules and about 200 base pairs of DNA. Science 184, 868–871. Koshland, D., and Strunnikov, A. (1996). Mitotic chromosome condensation. Annu. Rev. Cell Dev. Biol. 12, 305–333. Ladoux, B., Quivy, J. P., Doyle, P., du Roure, O., Almouzni, G., and Viovy, J. L. (2000). Fast kinetics of chromatin assembly revealed by single-molecule videomicroscopy and scanning force microscopy. Proc. Natl. Acad. Sci. USA 97, 14251–14256. Laemmli, U. K. (2002). Packaging genes into chromosomes. http://www.molbio.unige.ch/ PACKGENE/PAGE1.html.

2. Micromechanical Studies of Mitotic Chromosomes

137

Laemmli, U. K., Cheng, S. M., Adolph, K. W., Paulson, J. R., Brown, J. A., and Baumback, W. R. (1978). Cold Spring Harb. Symp. Quant. Biol. 42, 351–360. Landau, L. D., and Lifshitz, I. M. (1986). ‘‘Theory of Elasticity.’’ Pergamon, New York. Leforestier, A., Fudaley, S, and Livolant, F. (1999). Spermidine-induced aggregation of nucleosome core particles: Evidence for multiple liquid crystalline phases. J. Mol. Biol. 290, 481–494. Leger, J. F., Robert, J., Bourdieu, L., Chatenay, D., and Marko, J. F. (1998). RecA binding to a single double-stranded DNA molecule: A possible role of DNA conformational fluctuations. Proc. Natl. Acad. Sci. USA 95, 12295–12299. Leger, J. F., Romano, G., Sarkar, A., Robert, J., Bourdieu, L., Chatenay, D., and Marko, J. F. (1999). Structural transitions of a twisted and stretched DNA molecule. Phys. Rev. Lett. 83, 1066–1069. Lever, M..A., Th’ng, J. P., Sun, X., and Hendzel, M. J. (2000). Rapid exchange of histone H1.1 on chromatin in living human cells. Nature 408, 873–876. Lewin, B. (2000). ‘‘Genes VII.’’ Oxford University Press, New York. Li, X., and Nicklas, R. B. (1995). Mitotic forces control a cell-cycle checkpoint. Nature 373, 630–632. Li, X., and Nicklas, R. B. (1997). Tension-sensitive kinetochore phosphorylation and the chromosome distribution checkpoint in praying mantis spermatocytes. J. Cell Sci. 110, 537–545. Liphardt, J., Onoa, B., Smith, S. B., Tinoco, I., Jr., and Bustamante, C. (2001). Science 292, 733–737. Livolant, F. (1978). Positive and negative birefringence in chromosomes. Chromosoma 21, 45–58. Livolant, F., and Leforestier, A. (2000). Chiral discotic columnar germs of nucleosome core particles. Biophys. J. 78, 2716–2729. Livolant, F., and Maestre, M. F. (1988). Circular dichroism microscopy of compact forms of DNA and chromatin in vivo and in vitro: Cholesteric Liquid-crystalline phases of DNA and single dinoflagellate nuclei. Biochemistry 27, 3056–3068. Lodish, H., Baltimore, D., Berk, A., Zipursky, S. L., Matsudaria, P., and Darnell, J. (1995). ‘‘Molecular Cell Biology.’’ Scientific American Press, New York. Losada, A., and Hirano, T. (2001). Shaping the metaphase chromosome: Coordination of cohesion and condensation. Bioessays 23, 924–935. Losada, A., Hirano, M., and Hirano, T. (1998). Identification of Xenopus SMC protein complexes required for sister chromatid cohesion. Genes Dev. 12, 1986–1997. Luger, K., Mader, A. W., Richmond, R. K., Sargent, D. F., and Richmond, T. J. (1997). Crystal structure of the nucleosome core particle at 2.8 A resolution. Nature 389, 251–260. Machado, C., and Andrew, D. J. (2000a). D-titin: A giant protein with dual roles in chromosomes and muscles. J. Cell. Biol. 151, 639–652. Machado, C., and Andrew, D. J. (2000b). Titin as a chromosomal protein. Adv. Exp. Med. Biol. 481, 221–236. Machado, C., Sunke, C. E., and Andrew, D. J. (1998). Human autoantibodies reveal titin as a chromosomal protein. J. Cell Biol. 141, 321–333. Manders, E. M. M., Kimura, H., and Cook, P. R. (1999). Direct imaging of DNA in living cells reveals the dynamics of chromosome formation. J. Cell Biol. 144, 813–821. Maniotis, A. J., Bojanowski, K., and Ingber, D. E. (1997a). Mechanical continuity and reversible chromosome disassembly within intact genomes removed from living cells. J. Cell. Biochem. 65, 114–130. Marko, J. F., and Siggia, E. D. (1997b). Polymer models of meiotic and mitotic chromosomes. Mol. Biol. Cell 8, 2217–2231.

138

Poirier and Marko

Marko, J. F., and Siggia, E. D. (1997). Driving proteins off DNA with applied tension. Biophys. J. 73, 2173–2178. Marsden, M. P., and Laemmli, U. K. (1979). Metaphase chromosome structure: Evidence for a radial loop model. Cell 17, 849–858. Marshall, W. F., Dernburg, A. F., Harmon, B., Agard, D. A., and Sedat, J. W. (1996). Specific interactions of chromatin with the nuclear envelope: Positional determination within the nucleus in Drosophila melanogaster. Mol. Biol. Cell 7, 825–842. Marshall, W. F., Marko, J. F., Agard, D. A., and Sedat, J. W. (2001). Chromosomal elasticity and mitotic polar ejection force measured in living Drosophila embryos by four-dimensional microscopy-based motion analysis. Curr. Biol. 11, 1–20. Marshall, W. F., Straight, A., Marko, J. F., Swedlow, J., Dernburg, A., Belmont, A., Murray, A. W., Agard, D. A., and Sedat, J. W. (1997). Interphase chromosomes undergo constrained diffusional motion in living cells. Curr. Biol. 1, 930–939. Melby, T., Ciampaglio, C. N., Briscoe, G., and Erickson, H. P. (1998). The symmetrical structure of structural maintenance of chromosomes (SMC) and MukB proteins: Long, antiparallel coiled coils, folded at a flexible hinge. J. Cell Biol. 142, 1595–1604. Michaelis, C., Ciock, R., and Nasmyth, K. (1997). Cohesins: Chromosomal proteins that prevent premature separation of sister chromatids. Cell 91, 35–45. Miller, O. L., and Beatty, B. R. (1969). Visualization of nucleoar genes. Science 164, 955–957. Miller, O. L., and Hamkalo, B. A. (1972). Visualization of RNA synthesis on chromosomes. Int. Rev. Cytol. 33, 1–25. Misteli, T., Gunjan, A., Hock, R., Bustin, M., and Brown, D. T. (2000). Dynamic binding of histone H1 to chromatin in living cells. Nature 408, 877–881. Morgan, G. T. (2002). Lampbrush chromosomes and associated bodies: New insights into principles of nuclear structure and function. Chromosome Res. 10, 177–200. Nguyen, T. T., Rouzina, I., and Shklovskii, B. I. (2000). Reentrant condensation of DNA induced by multivalent counterions. J. Chem. Phys. 112, 2562–2568. Nguyen, T. T., and Shklovskii, B. I. (2001). Complexation of DNA with positive spheres: Phase diagram of charge inversion and reentrant condensation. J. Chem. Phys. 115, 7298–7308. Nicklas, R. B. (1983). Measurements of the force produced by the mitotic spindle in anaphase. J. Cell Biol. 97, 542–548. Nicklas, R. B. (1988). The forces that move chromosomes in mitosis. Annu. Rev. Biophys. Biophys. Chem. 17, 431–449. Nicklas, R. B. (1997). How cells get the right chromosomes. Science 275, 632–637. Nicklas, R. B., Campbell, M. S., Ward, S. C., and Gorbsky, G. J. (1998). Tension-sensitive kinetochore phosphorylation in vivo. J. Cell Sci. 111, 3189–3196. Nicklas, R. B., and Staehly, C. A. (1967). Chromosome micromanipulation. I The mechanics of chromosome attachment to the spindle. Chromosoma 21, 1–16. Nicklas, R. B., and Ward, S. C. (1994). Elements of error correction in mitosis: Microtubule capture, release, and tension. J. Cell Biol. 126, 1241–1253. Nicklas, R. B., Ward, S. C., and Gorbsky, G. J. (1995). Kinetochore chemistry is sensitive to tension and may link mitotic forces to a cell cycle checkpoint. J. Cell Biol. 130, 929–939. Nicklas, R. B., Waters, J. C., Salmon, E. D., and Ward, S. C. (2001). Checkpoint signals in grasshopper meiosis are sensitive to microtubule attachment, but tension is still essential. J. Cell Sci. 114, 4173–4183. Paulson, J. R. (1988). Scaffolding and radial loops: The structural organization of metaphase chromosomes. In ‘‘Chromosomes and Chromatin’’ (K. W. Adolph, Ed.), Vol. III., pp. 3–30. CRC Press, Boca Raton, FL. Paulson, J. R., and Laemmli, U. K. (1977). The structure of histone-depleted metaphase chromosomes. Cell 12, 817–828. Pederson, T. (2000). Half a century of ‘‘the nuclear matrix.’’Mol. Biol. Cell 11, 799–805.

2. Micromechanical Studies of Mitotic Chromosomes

139

Pelta, J., Livolant, F., and Sikorav, J. L. (1996). DNA aggregation induced by polyamines and cobalthexamine. J. Biol. Chem. 27, 5656–5662. Poirier, M., Eroglu, S., Chatenay, D., and Marko, J. F. (2000). Reversible and irreversible unfolding of mitotic chromosomes by applied force. Mol. Biol. Cell 11, 269–276. Poirier, M. G. (2001). ‘‘Combined Biochemical–Micromechanical Study of Mitotic chromosomes.’’ Ph.D. thesis University of Illinois at Chicago, Available at http://www.uic.edu/ jmarko. Poirier, M. G., and Marko, J. F. (2002a). Bending rigidity of mitotic chromosomes. Mol. Biol. Cell 13, 2170–2179. Poirier, M. G., and Marko, J. F. (2002b). Effect of internal viscosity on biofilament dynamics. Phys. Rev. Lett. 88, 228103. Poirier, M. G., and Marko, J. F. (2002c). Mitotic chromosomes are chromatin networks without a contiguous protein scaffold. Proc. Natl. Acad. Sci. USA 99, 15393–15397. Poirier, M. G., Monhait, T., and Marko, J. F. (2002). Condensation and decondensation of mitotic chromosomes driven by shifts in ionic conditions. J. Cell Biochem. 85, 422–434. Poirier, M. G., Nemani, A., Gupta, P., Eroglu, S., and Marko, J. F. (2001). Probing chromosome structure using dynamics of force relaxation. Phys. Rev. Lett. 86, 360–363. Polach, K. J., and Widom, J. (1995). Mechanism of protein access to specific DNA sequences in chromatin: A dynamic equilibrium model for gene regulation. J. Mol. Biol. 254, 130–149. Reese, D. H., Yamada, T., and Moret, R. (1976). An established cell line from the newt Notophthalmus viridescens. Differentiation 6, 75–81. Reif, M., Guatel, M., Oesterhelt, F., Fernandez, J. M., and Gaub, H. E. (1997). Reversible unfolding of individual titin immunoglobulin domains by AFM. Science 276, 1109–1112. Richmond, T. J., Finch, J. T., Rushton, B., Rhodes, D., and Klug, A. (1984). Structure of the nucleosome core particle at 7 A resolution. Nature 311, 532–537. Rieder, C. L., and Hard, R. (1990). Newt lung epithelial cells: Cultivation, use, and advantages for biomedical research. Int. Rev. Cytol. 122, 153–220. Saitoh, Y., and Laemmli, U. K. (1993). From the chromosomal loops and the scaffold to the classic bands of metaphase chromosomes. Cold Spring Harb. Symp. Quant. Biol. 58, 755–765. Saitoh, Y., and Laemmli, U. K. (1994). Metaphase chromosome structure: Bands arise from a differential folding path of the highly AT-rich scaffold. Cell 76, 609–622. Saminathan, M., Antony, T., Shirahata, A., Sigal, L. H., and Thomas, T. J. (1999). Ionic and structural specificity effects of natural and synthetic polyamines on the aggregation and resolubilization of single-, double-, and triple-stranded DNA. Biochemistry 38, 3821–3830. Skibbens, R. V., and Salmon, E. D. (1997). Micromanipulation of chromosomes in mitotic vertebrate tissue cells: Tension controls the state of kinetochore movement. Exp. Cell Res. 15(235), 314–324. Smith, S. B., Cui, Y., and Bustamante, C. (1996). Overstretching B-DNA: The elastic response of individual double-stranded and single-stranded DNA molecules. Science 271, 795–799. Smith, S. B., Finzi, L., and Bustamante, C. (1992). Direct mechanical measurements of the elasticity of single DNA molecules by using magnetic beads. Science 258, 1122–1126. Smythe, C., and Newport, J. W. (1991). Systems for the study of nuclear assembly, DNA replication, and nuclear breakdown in Xenopus laevis egg extracts. Methods Cell Biol. 35, 449–468. Stack, S. M., and Anderson, L. K. (2001). A model for chromosome structure during the mitotic and meiotic cell cycles. Chromosome Res. 9, 175–198. Strick, T. R., Allemand, J. F., Bensimon, D., Bensimon, A., and Croquette, V. (1996). The elasticity of a single supercoiled DNA molecule. Science 271, 1835–1837. Strick, T. R., Croquette, V., and Bensimon, D. (2000). Single-molecule analysis of DNA uncoiling by a type II topoisomerase. Nature 404, 901–904.

140

Poirier and Marko

Strunnikov, A. V. (1998). SMC proteins and chromosome structure. Trends Cell Biol. 8, 454–459. Strunnikov, A. V., Hogan, E., and Koshland, D. (1995). SMC2, a Saccharomyces cerivisiae gene essential for chromosome segregation and condensation, defines a subgroup within the SMC family. Genes Dev. 9, 587–599. Strunnikov, A. V., and Jessberger, R. (1999). Structural maintenance of chromosomes (SMC) proteins: Conserved molecular properties for multiple biological functions. Eur. J. Biochem. 263, 6–13. Strunnikov, A. V., Larionov, V. L., and Koshland, D. (1993). SMC1: An essential yeast gene encoding a putative head-rod-tail protein is required for nuclear division and defines a new ubitquitous family. J. Cell Biol. 123, 1635–1648. Sumner, A. T. (1996). The distribution of topoisomerase II on mammalian chromosomes. Chromosome Res. 4, 5–14. Tanaka, T., and Fillmore, D. J. (1979). Kinetics of swelling of gels. J. Chem. Phys. 70, 1214–1218. Tang, J. X., and Janmey, P. A. (1996). The polyelectrolyte nature of F-actin and the mechanism of actin bundle formation. J. Biol. Chem. 271, 8556–8563. Thoma, F., Koller, T., and Klug, A. (1979). Involvement of histone H1 in the organization of the nucleosome and of the salt-dependent superstructures of chromatin. J. Cell Biol. 83, 403–427. Thrower, D. A., and Bloom, K. (2001). Dicentric chromosome stretching during anaphase reveals roles of Sir2/Ku in chromatin compaction in budding yeast. Mol. Biol. Cell 12, 2800–2812. Trask, B. J., Allen, S., Massa, H., Fertitta, A., Sachs, R., van den Engh, G., and Wu, M. (1993). Studies of metaphase and interphase chromosomes using fluorescence in situ hybridization. Cold Spring Harb. Symp. Quant. Biol. 58, 767–775. Trinick, J. (1996). Titin as scaffold and spring. Curr. Biol. 6, 258–260. Tskhovrebova, L., Trinick, J., Sleep, J.-A., and Simmons, R.-M. (1997). Elasticity and unfolding of single molecules of the giant muscle protein titin. Nature 387, 308–312. Tsukamoto, T., Hashiguchi, N., Janicki, S. M., Tumbar, T., Belmont, A. S., and Spector, D. L. (2000). Visualization of gene activity in living cells. Nature Cell Biol. 2, 871–878. Van Holde, K. (1989). ‘‘Chromatin.’’ Springer, New York. Warburton, P. E., and Earnshaw, W. C. (1997). Untangling the role of DNA topoisomerase II in mitotic chromosome structure and function. Bioessays 19, 97–99. Widom, J. (1997). Chromosome structure and gene regulation. Phys. A 244, 497–509. Widom, J., and Klug, A. (1985). Structure of the 300A chromatin filament: X-ray diffraction from oriented samples. Cell 43, 207–213. Wolffe, A. (1995). ‘‘Chromatin.’’ Academic Press, San Diego. Wolffe, A. P., and Guschin, D. (2000). Chromatin structural features and targets that regulate transcription. J. Struct. Biol. 129, 102–122. Woodcock, C. L., and Horowitz, R. A. (1995). Chromatin organization re-viewed. Trends Cell Biol. 5, 272–277. Wuite, G. J., Smith, S. B., Young, M., Keller, D., and Bustamante, C. (2000). Single-molecule studies of the effect of template tension on T7 DNA polymerase activity. Nature 404, 103–106. Yin, H., Wang, M. D., Svoboda, K., Landick, R., Block, S. M., and Gelles, J. (1995). Transcription against an applied force. Science 270, 1653–1657. Yokota, H., van den Engh, G., Hearst, J. E., Sachs, R., and Trask, R. J. (1995). Evidence for the organization of chromatin in megabase pair-sized loops arranged along a random walk path in the human G0/G1 interphase nucleus. J. Cell Biol. 130, 1239–1249.

2. Micromechanical Studies of Mitotic Chromosomes

141

Zhang, D., and Nicklas, R. B. (1995). The impact of chromosomes and centrosomes on spindle assembly as observed in living cells. J. Cell Biol. 129, 1287–1300. Zhang, D., and Nicklas, R. B. (1999). Micromanipulation of chromosomes and spindles in insect spermatocytes. Methods Cell Biol. 61, 209–218. Zickler, D., and Kleckner, N. (1999). Meiotic chromosomes: Integrating structure and function. Annu. Rev. Genet. 33, 603–754. Zink, D., Cremer, T., Saffrich, R., Fischer, R., Trendelenburg, M. F., Ansorge, W., and Stelzer, E. H. (1998). Structure and dynamics of human interphase chromosome territories in vivo. Hum Genet. 102, 241–251.

ThisPageIntentionallyLeftBlank

3

____________________________________________________________________________

Patterning of the Zebrafish Embryo by Nodal Signals Jennifer O. Liang and Amy L. Rubinstein Department of Embryology Carnegie Institution of Washington Baltimore, Maryland 21210

I. Introduction II. Zebrafish Nodal Signals A. Overview of Nodal Protein Structure B. Expression Pattern of Zebrafish nodal-Related Genes C. Nodal Signaling Mutants III. Nodal Signaling Pathway A. Transforming Growth Factor Subfamilies and Receptors B. One-Eyed Pinhead C. Nodal Antagonists D. Smad and Fast Transcription Factors E. Downstream Targets of Nodal Signals IV. Patterning the Mesoderm and Endoderm A. Specification of the Mesendoderm B. Regionalization of the Mesendoderm V. Role of Nodal in Patterning the Ventral Nervous System A. Floor Plate B. Ventral Brain VI. Patterning the Left–Right Axis A. Visceral Asymmetry B. Left–Right Asymmetry in the Dorsal Diencephalon VII. Future Directions References

I. Introduction Nodal signals are essential for establishing the body plan of vertebrate embryos. The founding member of this family, Nodal, was identified in the mouse, where it is required for the formation of the node, an important signaling center that guides development of the embryonic axis (Conlon et al., 1994; Zhou et al., 1993). Subsequent studies in zebrafish and other vertebrates demonstrated that Nodal proteins are involved in multiple patterning events during development, including initial specification of the endoderm Current Topics in Developmental Biology, Vol. 55 Copyright 2003, Elsevier (USA). All rights reserved. 0070-2153/03 $35.00

143

144

Liang and Rubinstein

and mesoderm, formation of the ventral neural tube, and regulation of left– right asymmetry. This review summarizes the significant contribution research in zebrafish has made toward identifying components of the Nodal signaling pathway and defining the functions of Nodal proteins during development.

II. Zebrafish Nodal Signals A. Overview of Nodal Protein Structure Nodal proteins are secreted signaling factors in the Transforming Growth Factor (TGF ) family that have been found in every vertebrate examined to date. In zebrafish, two Nodal-related proteins, Cyclops (Cyc) and Squint (Sqt), have been characterized (Erter et al., 1998; Feldman et al., 1998; Rebagliati et al., 1998a,b; Sampath et al., 1998) and a third has been identified (M. Rebagliati, personal communication). Cyc and Sqt are synthesized as proproteins, which are cleaved to make the mature form. The mature regions of Cyc and Sqt are highly related to each other (68% identity) and to mouse Nodal (69 and 57% identity, respectively) (Rebagliati et al., 1998a). Cyc and Sqt are also more distantly related (40%) to other TGF -related family members that have roles during development, such as Activins and Bone Morphogenetic Proteins (BMPs). The mature forms of Cyc and Sqt have seven cysteine residues that are conserved in most TGF family members (Massague, 1998). Six of the cysteines interact to form a ‘‘cysteine knot,’’ which is essential for biological activity of all TGF -like proteins (McDonald and Hendrickson, 1993). The seventh cysteine forms a disulfide bond with another Nodal polypeptide, generating the dimeric form of the ligand that binds to the corresponding receptor and leads to activation of the signaling pathway (Daopin et al., 1992; Mason, 1994; McDonald and Hendrickson, 1993).

B. Expression Pattern of Zebrafish nodal-Related Genes The cyc and sqt genes have distinct, dynamic expression patterns in the embryo (Erter et al., 1998; Feldman et al., 1998; Rebagliati et al., 1998a; Sampath et al., 1998). These expression patterns reflect the complex roles of Nodal signals during development and are important for understanding when and where patterning is taking place. Maternally derived sqt mRNA is detected in the fertilized egg, whereas zygotic transcription begins at early blastula stages and persists through the end of gastrulation (Erter et al., 1998; Feldman et al., 1998; Kikuchi et al., 2001; Rebagliati et al.,

3. Zebrafish Nodal Signals

145

1998a). Tissue-specific transcription is first detected at the 512 cell stage, making sqt one of the first markers of the dorsal blastula (Erter et al., 1998). During midblastula stages, sqt expression persists in these dorsal cells. Transcripts are also found associated with dorsal nuclei in the yolk syncytial layer (YSL) (Erter et al., 1998; Feldman et al., 1998; Rebagliati et al., 1998a). The YSL functions in specifying the embryonic shield, the zebrafish equivalent of the organizer (Solnica-Krezel and Driever, 2001). During late blastula stages, sqt expression extends around the entire margin of the embryo, in cells that will give rise to the mesoderm and endoderm (called mesendoderm in fish because some of these cells give rise to both mesodermal and endodermal cell types) (Erter et al., 1998; Feldman et al., 1998; Rebagliati et al., 1998a; Warga and Nusslein-Volhard, 1999). Throughout gastrulation, sqt is expressed in the dorsal forerunner cells (Erter et al., 1998; Feldman et al., 1998; Kikuchi et al., 2001; Rebagliati et al., 1998a), a specialized group of cells that does not internalize during gastrulation and may have important functions in generating the left–right axis (Cooper and D’Amico, 1996; Essner et al., 2002; Melby et al., 1996). The cyc gene is expressed at two distinct times in early development: first during gastrulation and then during somitogenesis. cyc is coexpressed with sqt in the marginal cells of the midblastula, which are fated to become mesendoderm (Rebagliati et al., 1998a; Sampath et al., 1998). At the onset of gastrulation, cyc is expressed in the embryonic shield, and subsequently in the developing midline mesendoderm (Rebagliati et al., 1998a; Sampath et al., 1998). Transient expression is observed in the ventral neurectoderm. Midline expression is no longer detected by early somite stages (Rebagliati et al., 1998a; Sampath et al., 1998). The second wave of cyc transcription occurs from mid to late somitogenesis in the left lateral plate mesoderm and left dorsal diencephalon (Rebagliati et al., 1998a; Sampath et al., 1998). In summary, cyc and sqt transcripts are present in many tissues with important roles in patterning the embryo, including the YSL, embryonic shield, mesendoderm, lateral plate mesoderm, and dorsal forerunner cells.

C. Nodal Signaling Mutants Research in mice first demonstrated the importance of Nodal proteins in development. However, because mice nodal mutants die shortly after the completion of gastrulation, it had been diYcult to explore later roles for Nodal signals. In zebrafish, Cyc and Sqt share the functions of mouse Nodal. Because of this, cyc and sqt single mutants have less severe phenotypes and survive to later stages (Table I and Fig. 1). Further, many other Nodal pathway mutants have been identified in zebrafish, with their phenotypes ranging

146

Liang and Rubinstein

Table I Nodal Signaling Mutants Gene

Type of protein

Phenotype

cyclops (cyc)

Nodal: secreted TGF family member

Cyclopic, ventral brain and floor plate are missing, prechordal plate reduced, notochord precursors reduced, but mature notochord normal, altered brain asymmetry, visceral asymmetry normal

squint (sqt)

Nodal protein

Variable phenotype, ranging from normal to severe forebrain defects, including cyclopia. Prechordal plate reduced Organizer does not form, lacks all endoderm and almost all mesoderm

cyc;sqt one-eyed pinhead (oep)

Membrane-bound EGF-CFC protein

Maternal zygotic oep (MZoep) schmalspur (sur)

Cyclopic. Lacks all endoderm as well as floor plate and ventral brain Does not form organizer, missing all endoderm and almost all mesoderm

Fast-1 transcription factor

MZsur

Variable phenotype, synopthalmia or cyclopia, mild defects in ventral neural tube Cyclopia, reduction in prechordal plate and floor plate

from very mild to severe (Table I and Fig. 1). These mutants have made it possible to confirm and characterize the later roles for Nodal signaling in regionalization of the mesendoderm (Section IV, B), patterning the ventral neural tube (Section V), and left–right asymmetry (Section VI). Comparisons between diVerent Nodal signaling mutants have been essential for identifying new components of the signaling pathway (Section III).

III. Nodal Signaling Pathway A. Transforming Growth Factor b Subfamilies and Receptors Nodal signal transduction is believed to conform with what has been established for other members of the TGF superfamily (Fig. 2) (reviewed by Massague, 1998). Briefly, binding of the TGF dimer brings together two types of transmembrane receptor kinases. Upon complex formation, the type I receptor is phosphorylated by the type II receptor. This activates the type I receptor, which then phosphorylates transcription factors

3. Zebrafish Nodal Signals

147

Figure 1 Zebrafish Nodal signaling mutants. Frontal views of several Nodal signaling mutants at approximately 1 day postfertilization. Note the variation in the eye phenotype among diVerent mutants and between two sur mutants from diVerent parents.

known as Smads (Massague, 1998; Massague and Wotton, 2000). Phosphorylation of Smad proteins allows them to translocate from the cytoplasm to the nucleus where they can activate pathway-specific genes. Major subfamilies of the TGF superfamily have been grouped according to their degree of homology to one another. Each subfamily appears to signal through distinct receptors and through diVerent combinations of Smad proteins. From the outset, researchers found it diYcult to place Nodal into one of these subfamilies. The degree of homology to BMPs (42%) was about the

148

Liang and Rubinstein

Figure 2

Schematic diagram illustrating the Nodal signaling pathway in zebrafish.

same as to Activins (40%), and therefore Nodals were initially considered ‘‘intermediate’’ TGF proteins, not clearly belonging to either major subgroup (Kingsley, 1994; Massague, 1998). Research in zebrafish helped change this classification, providing evidence that Nodals actually belong to the Activin subfamily (Gritsman et al., 1999). For example, it was observed that defects in zebrafish Nodal signaling mutants, such as oep, could be corrected by overexpression of Activin and Activin receptors (Gritsman et al., 1999; Mathieu et al., 2002; Peyrieras et al., 1998). Several Activin-like receptors have been identified in zebrafish, including two type II receptors, ActRIIa and ActRIIb (Garg et al., 1999; Nagaso et al., 1999), and a type I receptor, TARAM-A (Renucci et al., 1996). Although no biochemical evidence yet exists to link any of these

3. Zebrafish Nodal Signals

149

receptors with zebrafish Nodal proteins, overexpression studies have shown that activated versions of TARAM-A and human ActRI can rescue mutants defective in Nodal signaling (Gritsman et al., 1999; Peyrieras et al., 1998; Thisse and Thisse, 1999). In addition, a dominant negative TARAM-A receptor produces a cyclopic phenotype similar to that of cyc and oep mutants (Aoki et al., 2002b).

B. One-Eyed Pinhead Studies in zebrafish revealed that Nodal signaling through Activin-like receptors requires an additional protein called One-eyed pinhead (Oep). The zebrafish oep mutation is phenotypically similar but more severe than the cyc mutation, with mutants displaying defects in the prechordal plate and ventral neurectoderm (Table I; Schier et al., 1996). The oep gene encodes an EGF-CFC protein similar to the mouse proteins Cripto and Cryptic and the Xenopus protein FRL-1 (Zhang et al., 1998). Gritsman and colleagues (1999) discovered that Oep is an essential cofactor for Nodal signaling when they observed that embryos lacking both maternal and zygotic oep (MZoep) bore a strong resemblance to cyc;sqt double mutants. Furthermore, they found that overexpression of Nodal, which normally causes severe dorsalization of the embryo (Rebagliati et al., 1998a; Sampath et al., 1998), has no eVect on oep mutant embryos. (Gritsman et al., 1999) Subsequently, it was found that EGF-CFC proteins are also essential for Nodal signaling in mice (Shen and Schier, 2000). Overexpression and biochemical studies suggest that Oep/EGF-CFC proteins may also be required for signaling by other TGF family members, such as Vg1 (Cheng et al., 2003). However, it remains to be determined whether Oep function in other signaling pathways is important for normal development. Oep is membrane bound. Although this property is normally necessary for function, it can be bypassed by overexpression of a secreted form of the protein. This suggests that a high local concentration of Oep, rather than membrane association, is required to allow Nodal signaling (Zhang et al., 1998). Biochemical interactions among EGF-CFC proteins, type I receptors, and Nodal signals have been shown in vitro. In one study, coprecipitation experiments demonstrated an interaction among the mouse EGF-CFC protein Cripto, mouse Nodal, and the human type I receptor ALK4 (Yeo and Whitman, 2001). Interestingly, a complex between Cripto and ALK4 was required for binding the Nodal ligand (Yeo and Whitman, 2001). In a second study, it was found that the Nodal protein Xenopus Nodal-Related 1 could bind directly to Cripto, the Xenopus type I receptor ALK7, and Activin Receptor II b (Reissmann et al., 2001). These data provide the first evidence that

150

Liang and Rubinstein

EGF-CFC proteins are essential components of the Nodal receptor complex. The role of the Oep protein appears to be permissive rather than instructive, as overexpression of moderate amounts of oep has no eVect on wild-type (WT) embryos (Liang et al., 2000; Zhang et al., 1998). There is some suggestion that when overexpressed at a high concentrations, Oep can have an instructive function and inhibit Nodal signaling when coexpressed with cyc or sqt (Kiecker et al., 2000). The significance of this finding is unclear, however, as only very high doses of Oep mRNA caused these eVects.

C. Nodal Antagonists Precise regulation of Nodal signals is also mediated by the action of antagonizing factors, such as Antivin/Lefty1 and Lefty2, which are produced in similar domains as cyc and sqt (Bisgrove et al., 1999; Thisse and Thisse, 1999). Overexpression of a graded series of Antivin/Lefty1 mRNA concentrations produces phenotypes that mimic mutations in the Nodal pathway (Thisse et al., 2000). Lefty proteins have been classified as members of the TGF superfamily on the basis of the six conserved cysteines that form the characteristic cysteine knot motif of TGF proteins (Bisgrove et al., 1999; Meno et al., 1996; Thisse and Thisse, 1999). However, the overall degree of sequence homology to other members of the TGF superfamily is low (20–25%) and Lefty proteins lack the dimerizing cysteine found in other members of the superfamily (Meno et al., 1996). In Xenopus, another secreted protein containing the TGF cysteine knot has been shown to antagonize Nodal signals. Cerberus, so named because it promotes the growth of head structures in the embryo, binds a range of important signaling proteins, including those in the Wnt, BMP, and Nodal families (Piccolo et al., 1999). A Cerberus homologue in zebrafish has not yet been described.

D. Smad and Fast Transcription Factors A considerable body of literature exists describing the role of Smad transcription factors in transducing TGF signals (Massague, 1998; Massague and Wotton, 2000). Eight smad genes have been identified to date from human, mouse, and frog. Smads 1, 5, and 8 appear to be activated by BMP receptors, whereas Smads 2 and 3 are activated by TGF and Activin receptors. Smad4 is believed to be required for both BMP and TGF /Activin signaling. Finally, Smads 6 and 7 appear to function by inhibiting the Activity of other Smad proteins. Several zebrafish smad genes have been

3. Zebrafish Nodal Signals

151

cloned (Dick et al., 2000; Muller et al., 1999), two of which (Smad2 and Smad3) are believed to function in Activin/Nodal signal transduction. For example, overexpression of an activated form of Smad2 rescued eye and floor plate defects in cyc mutant embryos (Dick et al., 2000; Muller et al., 2000). The smad3 gene is expressed in domains that overlap cyc gene expression (Dick et al., 2000). Additional transcription factors have been identified that appear to work in concert with Smad proteins to mediate Nodal signals. The Forkhead activin signal transducer-1 (Fast-1) was first identified as part of a Smad complex binding to an activin response element (ARE) in the mix.2 gene promoter (Chen et al., 1998). Research in zebrafish provided the first direct evidence that the Fast-1 transcription factor is required for the maintenance of Nodal signaling (Pogoda et al., 2000; Sirotkin et al., 2000b). The zebrafish mutant schmalspur (sur), which displays a similar phenotype to that described for cyc, was shown to have a lesion in fast-1 gene. MZsur mutants, which lack both maternal and zygotic Fast-1 activity, are phenotypically more similar to cyc and zygotic oep mutants than MZoep mutants (Table I). This comparatively mild phenotype suggests that Fast-1 is required to transduce only late Nodal signals and thus is involved in maintenance rather than initiation of Nodal signaling. Overexpression of cyc and sqt mRNA in MZsur mutants results in upregulation of downstream targets of Nodal, such as goosecoid, further suggesting that early Nodal signaling can occur in the absence of Fast-1/Sur activity (Pogoda et al., 2000). Other transcriptional coactivators or repressors may be involved in the complex regulatory events that follow the reception of a Nodal signal. Several factors that bind to Smad proteins have been described (Massague and Chen, 2000). One of these, TG-interacting factor (TGIF), has been tentatively linked to Nodal signaling because heterozygous mutations in human TGIF cause holoprosencephaly (Gripp et al., 2000).

E. Downstream Targets of Nodal Signals Although several components of the Nodal signaling pathway have been elucidated, little is known about the genes whose transcription is regulated by receptor-activated Smad and Fast-1 proteins. The identification of these genes may help uncover the mechanism by which cyc and sqt influence patterning of the mesendoderm and the ventral nervous system. Numerous studies have described genes whose expression appears to be downregulated in cyc mutant embryos and thus may represent downstream targets of Nodal signaling. These include genes expressed in tissues that are absent or reduced in mutant embryos, such as floor plate and hatching glands. For example, expression of axial, which encodes a transcription

152

Liang and Rubinstein

factor of the forkhead/HNF3 gene family that is expressed in cells of the ventral neural tube, is strongly reduced in cyc mutants (Strahle et al., 1993). Transcription of a zebrafish gene encoding the chemoattractant protein Netrin was also shown to depend on cyc signaling (Rastegar et al., 2002; Strahle et al., 1997). The gene encoding Calreticulin, a calcium-binding protein localized to the endoplasmic reticulum, was cloned in a study designed to identify genes regulated by cyc (Rubinstein et al., 2000). The calreticulin gene is expressed in the floor plate and the hatching glands, tissues that are strongly aVected in cyc mutants (Rubinstein et al., 2000). calreticulin expression was also shown to be reduced in cyc mutants during gastrulation (Rubinstein et al., 2000). In all of these examples, however, the genes aVected in cyc mutants appear to be downregulated rather than absent altogether. It is possible that early expression of sqt in cyc mutants may partially compensate for the absence of Cyc. A comprehensive analysis of genes upregulated by Nodal signals in the zebrafish has been performed using subtractive hybridization and microarray technologies (Dickmeis et al., 2001a). Gene expression was compared between embryos injected with a constitutively active version of the Nodal receptor, Taram-A, and those injected with wild-type Taram-A. A wide variety of upregulated genes were identified, including those encoding transcription factors, cell adhesion molecules, signal transduction pathway members, cytoskeletal proteins, and metabolic enzymes. However, it is not yet known whether these genes are directly induced by Nodal signals. MZoep embryos, which presumably have all Nodal signaling blocked, lack transcription of many genes normally expressed during early embroygenesis, including gsc, hgg1, axial, and calreticulin (Gritsman et al., 1999; Rubinstein et al., 2000). However, whether lack of expression indicates that these genes are directly regulated by Nodal signaling can be diYcult to determine since the tissues that normally express these genes are absent in MZoep mutants. One way to address this issue is through overexpression of the putative target gene. Overexpression of axial results in recovery of floor plate in cyc mutant embryos, suggesting that it is an essential downstream component of the Nodal pathway (Rastegar et al., 2002). Evidence that transcription of a gene is directly activated by Nodal signaling could also be provided by determining whether Fast-1 and Smad bind to their promoters. AREs, which bind Smad/Fast-1 complexes, have been found in the regulatory regions of Activin and Nodal-responsive genes, including goosecoid (Watabe et al., 1995), Brachyury (Latinkic et al., 1997), and lim1 (Rebbert and Dawid, 1997; Watanabe et al., 2002). It is likely that AREs will be found in the promoters of other genes that are activated in direct response to Nodal signaling. The ability of Smad and Fast proteins to bind these promoters directly would provide better evidence of their in vivo regulation by Cyc or

3. Zebrafish Nodal Signals

153

Sqt. It has been suggested that cyc is directly involved in the induction of shh in the ventral midline (Muller et al., 2000) because a shh enhancer responsible for this expression is responsive to activated Smad2. This eVect could still be indirect, however, as it was not shown whether Smad2 or Fast-1 was capable of binding to this enhancer. In another study, the ARE of the Xenopus lim-1 promoter was shown to bind Fast-1 and Smad4 (Watanabe et al., 2002). Further analysis of zebrafish mutants demonstrated that lim-1 expression requires Nodal signaling (Watanabe et al., 2002). Thus, lim-1 is likely to be a direct downstream target of Nodal in vivo.

IV. Patterning the Mesoderm and Endoderm A. Specification of the Mesendoderm Nodal was first identified through its essential role in gastrulation and mesoderm formation in the mouse. During gastrulation, mesodermal cells are generated at the posterior pole of the mouse embryo, which creates a morphologically distinguishable structure called the primitive streak (Camus and Tam, 1999). As gastrulation proceeds, the primitive streak elongates, and the most anterior end of the streak forms the mouse node, which is functionally equivalent to the organizer of Xenopus and the embryonic shield of zebrafish. In mouse nodal mutants, cells fail to delaminate to form the mesodermal cell layer, and the primitive streak and node do not form (Conlon et al., 1994; Zhou et al., 1993). As a result, nodal mutant embroys lack all but a small amount of posterior mesoderm and die shortly after gastrulation (Conlon et al., 1994; Zhou et al., 1993). In zebrafish, cyc;sqt double mutants display a phenotype reminiscent of mouse nodal mutants. The embryonic shield/organizer of cyc;sqt double mutants does not form and gastrulation is severely disrupted: cells at the embryonic margin fail to involute and as a result the cell layer that will give rise to mesoderm and endoderm (mesendoderm) is missing (Feldman et al., 1998, 2000). cyc;sqt mutants have no endoderm and develop only a small amount of mesoderm in the tail (Feldman et al., 1998). Abrogating Cyc and Sqt signaling by blocking other steps in the Nodal signaling pathway similarly prevents formation of the mesendoderm. For example, MZoep embryos have a similar phenotype to cyc;sqt double mutants (Gritsman et al., 1999), as do embryos injected with high concentrations of Antivin/Lefty-1 mRNA (Thisse et al., 2000). These results indicate that zebrafish Cyc and Sqt share the early function of mouse Nodal and demonstrate that Nodal signaling has a conserved role in vertebrate mesoderm formation.

154

Liang and Rubinstein

Studies in zebrafish have also provided insights into the mechanism by which Nodal signals specify mesendoderm. Growing evidence shows that Nodal signals act to induce a mesendodermal cell fate in their target cells. Overexpression of Nodal proteins, including Cyc and Sqt, induces the expression of mesodermal and endodermal markers (Erter et al., 1998; Feldman et al., 1998; Jones et al., 1995; Joseph and Melton, 1997; Rebagliati et al., 1998a; Sampath et al., 1998). At blastoderm stages, before the onset of gastrulation, cyc and sqt are expressed in cells at the margin of the embryo and yolk cell. At about the same time, the first expression of mesendodermspecific markers (e.g., no tail, goosecoid, and floating head) is detected in the marginal cells and their neighbors. In the absence of Nodal signaling, as in cyc;sqt and MZoep embryos, the earliest expression of mesendodermal genes in the blastoderm margin is severely reduced or absent (Feldman et al., 1998; Gritsman et al., 2000). Cell-labeling experiments show that the fate map is shifted in that cells that would give rise to mesendoderm in WT embryos instead contribute to the central nervous system (Feldman et al., 2000). These data suggest that Nodal signaling is necessary and suYcient to induce mesendodermal cell fate. Further support for this hypothesis comes from studies of antivin/lefty genes. During blastula stages when the mesendoderm is induced, antivin/ lefty genes are expressed together with cyc and sqt in the margin of the embryo (Bisgrove et al., 1999; Meno et al., 1999 Thisse and Thisse, 1999). Overexpression and loss of function studies suggest that a balance between the level of Nodal signals and Antivin/Lefty Nodal antagonists controls how many cells contribute to the mesoderm and endoderm. Loss of Nodal signaling through mutations in the Nodal signaling pathway or through overexpression of antivin/lefty mRNA results in loss of mesendodermal tissue (Bisgrove et al., 1999; Feldman et al., 1998; Meno et al., 1999; Thisse et al., 2000). Reduction of Antivin/Lefty function has the opposite eVect. Mouse lefty2 mutants have a greatly expanded mesoderm (Meno et al., 1999). Although mutants have not yet been identified in zebrafish, Antivin/ Lefty proteins have been depleted using morpolinos (Agathon et al., 2001; Chen and Schier, 2002; Feldman et al., 2002), modified antisense oligonucleotides that bind to the targeted mRNA and block its translation (Nasevicius and Ekker, 2000). As in mouse lefty2 mutants, loss of Antivin/Lefty1 or Lefty2 results in the expansion of mesendodermal tissue and increased cellular internalization (Agathon et al., 2001; Feldman et al., 2002). Injection of lefty morpholinos into cyc and sqt single mutants suggests that gastrulation defects are due primarily to the deregulation of Sqt signaling (Chen and Schier, 2002; Feldman et al., 2002). Together these studies have led to a model in which the mesendoderm is induced by Nodal signaling in the margin of the blastula and is limited by a feedback mechanism requiring the action of the Antivin/Lefty class of Nodal antagonists.

3. Zebrafish Nodal Signals

155

B. Regionalization of the Mesendoderm In addition to being required in the initial step that induces the mesendoderm to form, Nodal signals play a role in regionalization of the mesendoderm. Initially, it was observed that injection of diVerent doses of frog and fish nodal mRNA induced the transcription of genes characteristic of diVerent mesodermal and endodermal cell types in Xenopus animal caps (Chen and Schier, 2001; Erter et al., 1998; Jones et al., 1995; Joseph and Melton, 1997; Sampath et al., 1998). This suggested that a gradient of Nodal signaling could be responsible for patterning diVerent mesendodermal tissues. In vivo studies in WT and mutant zebrafish embryos have provided support for this hypothesis. 1. Prechordal Plate and Notochord The strongest evidence for a graded response to Nodal signals comes from studies of the zebrafish axial mesendoderm. The axial mesendoderm is derived from cells that involute from the dorsal margin of the embryo during gastrulation, analogous to the delamination of cells in the mouse primitive streak. The first cells to move away from the margin form the prechordal plate, which will give rise to anterior mesodermal and endodermal tissues, such as hatching glands, pharyngeal endoderm, and head muscle. The next set of cells to involute will give rise to the notochord. Analysis of embryos with altered Nodal signaling suggests that high levels of Nodal signaling specify prechordal plate, whereas lower levels induce notochord. For instance, the prechordal plate is aVected in all Nodal signaling mutants, including those with only a slight reduction in Nodal signaling such as Zsqt and Zsur (Brand et al., 1996; Heisenberg and Nusslein-Volhard, 1997; Schier et al., 1996). In contrast, the notochord is not aVected in weak mutants, slightly aVected in mutants with a modest reduction in Nodal signaling, such as cyc, and severely aVected in mutants that have a drastic reduction in Nodal signaling (Thisse et al., 1994; Warga and Nusslein-Volhard, 1999), such as cyc;sqt or MZoep (Feldman et al., 1998; Gritsman et al., 1999). Levels of Cyc and Sqt in marginal cells just before the onset of gastrulation could be influencing a choice between notochord and prechordal plate cell fate. Lineage tracing experiments demonstrate that cells closest to the margin, which would also presumably receive the highest levels of Nodal signaling, are fated to become the prechordal plate whereas those further from the margin are fated to become the notochord (Warga and NussleinVolhard, 1999). At this stage, expression of the transcription factor goosecoid (gsc) overlaps with the prechordal plate fate domain and expression of flh overlaps with the notochord fate domain (Gritsman et al., 2000; Melby et al., 1996). When Nodal signaling is partially reduced, as in Zoep mutants

156

Liang and Rubinstein

or upon expression of low levels of atv, gsc expression is absent in marginal cells of the blastula and later the prechordal plate does not form (Gritsman et al., 2000). When Nodal signaling is severely reduced, as in MZoep mutants or when atv is expressed at high levels, both gsc and flh transcription is lost in the marginal cells. In this case, neither the prechordal plate nor the notochord develops (Feldman et al., 1998; Gritsman et al., 1999, 2000). In the reciprocal experiment, injection of oep mRNA into MZoep mutants during early stages results in the recovery of both gsc and flh expression, whereas injection later leads to the recovery of only flh. These results are consistent with either the duration or the level of Nodal signaling being required to specify diVerent subtypes of the axial mesendoderm (Gritsman et al., 2000). 2. Nonaxial Mesendoderm Levels of Nodal signaling may also regulate diVerentiation of nonaxial mesendodermal tissues. Tissues that are derived from cells in the blastoderm margin that are very close to Nodal-expressing cells, such as the endoderm and prechordal plate, appear to be specified by very high levels of Nodal signaling. For instance, strong activation of the Nodal signaling pathway by overexpression of an activated form of the TGF type I receptor TARAM-A (TARAM-A*) drives cells to contribute to the prechordal plate and endoderm (Peyrieras et al., 1998). Conversely, when Nodal signaling is reduced only slightly, as through injecting low concentrations of antivin mRNA, endoderm and prechordal plate formation is blocked (Thisse et al., 2000). At the opposite end of the continuum, tissues derived from cells at the animal pole, such as paraxial and ventral mesoderm, which are much further away from cells expressing Nodal proteins, seem to be specified by low levels of Nodal signals. For instance, development of these cell fates is only blocked when atv is expressed at very high levels or in the most severe Nodal signaling mutants (Feldman et al., 1998; Gritsman et al., 1999; Thisse et al., 2000). These data suggest that a concentration gradient of Nodal activity from the vegetal to animal pole of the embryo influences cell fate. However, there is likely more complexity to the mechanism of Nodal action than just a gradient of activity. The gradient model predicts that the endoderm and prechordal plate should be similarly aVected in zebrafish Nodal signaling mutants. However, this is not the case. There are several mutants, such as Zsqt, Zsur, and cyc, in which the prechordal plate is severely reduced, but the endoderm appears unaVected (Brand et al., 1996; Hatta et al., 1991; Heisenberg and Nusslein-Volhard, 1997; Schier et al., 1996; Thisse et al., 1994; Warga and Nusslein-Volhard, 1999). It is likely that additional criteria, such as when and where the Nodal signaling

3. Zebrafish Nodal Signals

157

pathway is activated, influence which mesendodermal cell fate is induced. Further, Cyc and Sqt proteins do not have entirely comparable biochemical function. Overexpression studies in the blastula show that Cyc signaling aVects gene transcription only locally, up to two cell diameters away from the source of the signal, whereas Squint acts directly over a larger range, up to eight cell diameters (Chen and Schier, 2001). Lefty proteins can also function far from the cell in which they are synthesized (Chen and Schier, 2002). A complete model of Nodal action in mesendoderm formation will require a more thorough understanding of the spatiotemporal pattern of Cyc and Sqt activity. 3. Endoderm Research in zebrafish has played a leading role in defining the role of Nodal signals in specifying the endoderm. As described earlier, the absence of endodermal tissue in Nodal signaling mutants such as Zoep, MZoep, and cyc;sqt indicated an essential role for Nodal signals in endoderm induction (Feldman et al., 1998; Gritsman et al., 1999; Schier et al., 1997). Subsequent studies have defined the intracellular signaling pathway responsible for specifying the endoderm (Ober et al., 2003). As in other Nodal signaling events, Cyc and Sqt activate an Oep-containing receptor complex on the surface of the cell (Alexander and Stainier, 1999). This complex likely contains the type I receptor TARAM-A, as expression of an activated form of this receptor cell autonomously promotes an endodermal cell fate (Alexander and Stainier, 1999; David and Rosa, 2001; Peyrieras et al., 1998). Conversely, overexpression of a dominant negative form of TARAM-A inhibits endoderm formation (Aoki et al., 2002b). However, loss-of-function studies that definitively show that this receptor acts in endoderm formation have not yet been reported. The activated Nodal receptor complex induces the expression of several transcription factors: Casanova (a Sox family member), Mixer (a Mix family homeodomain protein), and Faust/Gata5 (a zinc finger-containing protein) (Alexander and Stainier, 1999; Aoki et al., 2002a; Dickmeis et al., 2001b; Kikuchi et al., 2000, 2001; Reiter et al., 1999, 2001; Sakaguchi et al., 2001). Casanova is necessary and suYcient for the first steps in endoderm specification; the endoderm is completely absent in casanova mutants, and expression of casanova can induce the expression of early (but not late) endoderm markers in MZoep mutants (Alexander et al., 1999; Aoki et al., 2002a; Dickmeis et al., 2001b; Kikuchi et al., 2001; Sakaguchi et al., 2001). Endoderm precursors are severely reduced in bonnie and clyde (bon) mutants, which lack mixer, and in faust mutants (Kikuchi et al., 2000; Reiter et al., 1999, 2001). As neither Mixer nor Faust/Gata5 can induce endoderm formation in casanova mutants, it has been suggested that Casanova acts

158

Liang and Rubinstein

downstream or parallel to these factors (Alexander and Stainier, 1999; Reiter et al., 2001). It has been found that another Mix transcription factor, Mezzo, may also play a role in endoderm formation. Expression of the mezzo gene is regulated by Nodal signaling, and depletion of Mezzo in bon mutants using morpholinos increased the severity of the endoderm phenotype (Poulain and Lepage, 2002). These studies are important not only because they give insight into how the endoderm is formed, but because they define the pathway by which Nodal signals induce a specific mesendodermal cell fate. However, why the endoderm-specific pathway is preferentially activated in certain cells remains to be elucidated.

V. Role of Nodal in Patterning the Ventral Nervous System cyc was one of the first zebrafish mutations to be identified in early screens for defects in embryonic development (Hatta et al., 1991). cyc mutants are missing all of the ventral nervous system, including the ventral brain and the floor plate in the ventral spinal cord. The finding that cyc encodes a Nodal protein provided the first evidence that Nodal signals have a role in patterning the nervous system (Rebagliati et al., 1998b; Sampath et al., 1998). Lack of a ventral neural tube is responsible for many of the defects in cyc and other Nodal signaling mutants, including cyclopia and disorganization of central nervous system axons (Hatta, 1992; Hatta et al., 1994; Masai et al., 2000). These secondary defects are not discussed in detail here.

A. Floor Plate The medial floor plate of the ventral neural tube has an important role in the patterning and guidance of ventral nervous system axons. For example, explant experiments in the rat have shown that spinal commissural neurons are attracted by Netrin-1 secreted by the floor plate, whereas motor neurons from the brain and spinal cord are repelled by it (Colamarino and TessierLavigne, 1995). For many years, it had been well accepted that formation of the vertebrate floor plate is induced in the ventral neural tube by Shh secreted by the underlying notochord (Placzek et al., 2000). This model has been called into question, in part because of analysis of zebrafish mutants that lack the notochord and floor plate (Le Douarin and Halpern, 2000). For example, no tail mutants do not have a notochord, yet actually form more floor plate tissue than WT embryos (Halpern et al., 1997). Conversely, cyc mutants lack a floor plate, despite the presence of a shh-expressing notochord (Halpern

3. Zebrafish Nodal Signals

159

et al., 1997; Hatta et al., 1991). This suggests that a diVerentiated notochord is not required for floor plate formation. The expression pattern of the cyc gene indicates that floor plate specification begins earlier than was previously thought. cyc is expressed during gastrulation, but not later in embryogenesis when the embryonic axis is elongating and the floor plate is continuing to form caudally (Rebagliati et al., 1998a; Sampath et al., 1998). This suggests that floor plate specification begins during gastrulation and that Cyc signals are involved in the first steps of floor plate patterning. Several experiments have been described that were aimed at defining which cyc-expressing cells are important for floor plate formation. Early transplantation studies indicated that WT cells could contribute to the floor plate in a cyc mutant background (Hatta et al., 1991). However, WT cells in the notochord of cyc mutants were not suYcient to direct formation of the floor plate (Hatta et al., 1991). These data supported the idea that the notochord is not required for floor plate specification. Subsequent studies demonstrated that rescue of shh-expressing floor plate cells in cyc mutants was correlated with overexpression of cyc mRNA in precursors of the prechordal plate (Sampath et al., 1998). In contrast, expression of cyc in the presumptive notochord only was not suYcient for rescue (Sampath et al., 1998). Analysis of the oep mutation has provided additional insight into how Nodal signals act in floor plate development. Transplantation experiments showed that while WT cells frequently diVerentiate as floor plate when grafted into host embryos, oep mutant cells never do (Shinya et al., 1999). Because oep is required for the reception of Nodal signals, these data provided evidence that activation of the Nodal signal transduction pathway is required in cells fated to become the floor plate (Shinya et al., 1999). The earliest studies of cyc function suggested a mechanism of floor plate specification termed ‘‘homeogenetic’’ induction in which diVerentiated floor plate cells induce adjacent cells to adopt floor plate identity (Hatta et al., 1991). This mechanism was proposed because it was observed that cyc mutant cells adjacent to WT cells were capable of diVerentiating as the floor plate. However, another simple explanation could explain this finding: mutant cells that diVerentiated as the floor plate were responding to the secreted Cyc signal from neighboring WT cells. Similar studies performed with the oep mutation demonstrated that complicated interactions exist between potential floor plate cells. In contrast to what was observed in cyc transplant experiments, transplanted oep mutant cells prevent adjacent WT cells (up to two cells away from the mutant cell) from diVerentiating as floor plate (Shinya et al., 1999). This suggests that floor plate specification may be mediated in part by negative interactions between neighboring cells. The nature of the inhibitory interaction is not

160

Liang and Rubinstein

known. One possible model is that a tight connection between floor plate precursors, mediated by Oep, is required to transduce Nodal signals between cells and allow floor plate diVerentiation. If this is the case, then transplanted oep mutant cells would be unable to make these connections with neighboring cells, and its neighbors would be unable to diVerentiate as the floor plate. Cells expressing markers of medial floor plate form belatedly (by 48 hpf) in cyc mutant embryos (Albert et al., 2003). These medial floor plate cells also express genes characteristic of lateral floor plate, suggesting that they are derived from the lateral floor plate in a Cyc-independent manner (Albert et al., 2003). In summary, injection experiments have provided evidence that cyc is required in the presumptive prechordal plate for early floor plate formation, whereas transplant experiments suggested that activation of the Cyc signaling pathway is required in presumptive floor plate cells. Further experiments will be needed to definitely determine whether Cyc is acting as a floor plate inducer or whether it has another role, such as to ensure the proper formation and migration of the prechordal plate precursors.

B. Ventral Brain Studies of the diencephalon and telencephalon in zebrafish mutants have elucidated some of the mechanisms by which Nodal signals specify the ventral brain. Gene expression studies revealed that the posterior ventral (PV) hypothalamus is reduced or absent in mutants with a modest reduction in Nodal signaling, such as sur, whereas the anterior dorsal (AD) hypothalamus is present (Mathieu et al., 2002). In more severe mutants, the hypothalamus is completely absent (Rohr et al., 2001). This suggests that higher levels of Nodal signaling are needed to retain the PV hypothalamus. Cyc appears to function by diVerent mechanisms to specify the AD and PV domains. Transplantation experiments demonstrate that oep mutant cells can contribute to the AD but not PV hypothalamus when transplanted into MZoep mutant hosts, whereas cotransplanted WT cells can contribute to both domains (Mathieu et al., 2002). Therefore, activation of the Nodal signaling pathway is required cell autonomously in the PV hypothalamus. In contrast, expression of the constitutively active type I receptor TARAM-A* in the prechordal plate of MZoep mutant embryos was necessary and suYcient to recover the AD hypothalamus (Mathieu et al., 2002). These data are consistent with early studies, which found that recovery of ventral brain in cyc mutants correlated with cyc expression in prechordal plate precursors (Sampath et al., 1998). Further, it demonstrates that Nodal signaling acts cell nonautonomously in the AD domain.

3. Zebrafish Nodal Signals

161

Shh produced by the prechordal plate is a candidate for the AD hypothalamus-inducing signal. Overexpression studies suggest that Nodal signaling can promote shh expression in neural tissue (Muller et al., 2000). In addition, the AD hypothalamus is reduced in smoothened (smu) mutants (Mathieu et al., 2002), which have severely reduced Hedgehog signaling (Barresi et al., 2000; Chen et al., 2001; Roy et al., 2001; Varga et al., 2001). However, overexpression of shh in MZoep mutants was able to restore hypothalamic gene expression only in a limited area of the middiencephalon (Rohr et al., 2001). Therefore, it is unlikely that Shh by itself is suYcient to promote hypothalamus formation. Nodal signaling also appears to be acting upstream of Shh signals to pattern the ventral telencephalon (Rohr et al., 2001). Examination of smu, cyc, oep, and MZoep mutants reveals that the ventral telencephalon is absent (Rohr et al., 2001). Formation of the ventral telencephalon in MZoep and other Nodal signaling mutants is recovered upon expression of shh (Rohr et al., 2001). Transplant experiments indicate that Shh expression in cells of the anterior forebrain is suYcient to restore ventral telencephalon-specific gene expression in neighboring cells (Rohr et al., 2001). This is consistent with Shh acting within the ventral telencephalon. An alternative role for Cyc in patterning of the anterior neurectoderm has also been proposed (Sirotkin et al., 2000a). In this model, cyc functions to repress the development of anterior structures (Sirotkin et al., 2000a). It is based on the analysis of double and triple mutants of the zebrafish genes cyc, sqt, and bozozok (boz). boz encodes a transcription factor that is believed to act in parallel with the Nodal signaling pathway to specify the dorsal mesoderm (Shimizu et al., 2000; Sirotkin et al., 2000a). Sirotkin et al. (2000a) found that while sqt;boz double mutants form no anterior structures, sqt;boz;cyc triple mutants occasionally do exhibit evidence of anterior cell fates, such as the presence of eyes. The significance of this finding is unclear. The percentage of triple mutants that form eyes was not reported, and Shimizu et al. (2000), in a similar analysis, did not report finding such anterior structures in triple mutants. Furthermore, cyc single mutants do not appear to form an excess of anterior cell fates, which might be expected if cyc acts to repress these cell fates in WT embryos. In conclusion, research in zebrafish was instrumental in establishing a role for Cyc signals in patterning the ventral brain and floor plate. Subsequent studies are finding that Cyc acts by several mechanisms to specify diVerent regions of the ventral neural tube. Most strikingly, Cyc signals appear to function cell autonomously within the neurectoderm in some regions, such as the PV hypothalamus and possibly the floor plate precursors (Mathieu et al., 2002; Shinya et al., 1999), and cell nonautonomously in others, such as the AD hypothalamus (Mathieu et al., 2002). Future studies will help

162

Liang and Rubinstein

determine how Shh and other signals cooperate with Cyc in specific regions of the brain.

VI. Patterning the Left–Right Axis In addition to its early roles in patterning the endoderm, mesoderm, and ventral neural tube, Nodal signals have a later role in establishing morphological left–right asymmetries. A Nodal-related signal is expressed in the left lateral plate mesoderm (LPM) of all vertebrates examined, where it has a role in regulating the asymmetry of visceral organs (Hamada et al., 2002; Mercola and Levin, 2001). In zebrafish, cyc is expressed transiently not only in the left LPM, but also in the left forebrain (Rebagliati et al., 1998a; Sampath et al., 1998). Research in zebrafish has been essential in uncovering mechanisms that regulate asymmetric gene expression in the brain and lateral plate and in revealing a function for Nodal signaling in left–right brain asymmetry.

A. Visceral Asymmetry A complex pathway controls asymmetry of the visceral organs. Although some of the molecules in this pathway diVer among species, many elements are conserved (Hamada et al., 2002; Mercola and Levin, 2001). The left–right axis appears to be generated during gastrulation or earlier and is first evident in the node/organizer or surrounding cells. Left positional information is then transmitted to the LMP through a series of activating and suppressive signals. This results in expression of a Nodal-related signal and other components of the Nodal signaling pathway, such as Antivin/Lefty and the transcription factor Pitx2, in the left but not right LPM. Proper expression of the Nodal pathway components on the left side is required for the normal asymmetric development of internal organs such as the heart, lung, liver, and digestive organs. When this pathway is perturbed, asymmetry is disrupted. For example, expression of Nodal in the right LMP is correlated with the reversal of asymmetry, and lack of Nodal or bilateral Nodal with randomized asymmetry. Together, these data suggest that Nodal signaling is important for establishing the left–right axis and for asymmetric morphogenesis of visceral organs. Because cyc and other zebrafish Nodal pathway mutants survive to larval stages, it has been easier to test the function of Nodal signaling in the left LPM. On the basis of the proposed role for Nodal signals in the left LPM, one would predict that cyc mutants would have serious disruptions in the

3. Zebrafish Nodal Signals

163

left–right asymmetry of their visceral organs. Surprisingly, gene expression in the left LPM is normal. In cycm294 or cyctf219 mutants, which have point mutations that produce nonfunctional proteins, cyc, pitx2, and lefty2 are all expressed correctly in the left LPM, although cyc transcripts may be reduced slightly compared with WT siblings (Bisgrove et al., 2000; Sampath et al., 1998). Asymmetry of visceral organs also appears to be normal in cyc mutants. One of the first morphological asymmetries occurs in the vertebrate heart (Goldstein et al., 1998). In zebrafish, shortly after the fused heart tube has formed, the anterior end ‘‘jogs’’ to the left. This is followed much later by a rightward bending of the heart. These asymmetries are normal in cycm294 mutants (Chin et al., 2000), which produce Cyc protein with no activity (Sampath et al., 1998). Therefore, asymmetric gene expression and heart laterality do not require Cyc function in the left LPM. In contrast to cyc mutants, visceral asymmetry is perturbed in oep mutants (Yan et al., 1999). Although MZoep embryos lack almost all mesendodermal tissue, including the LPM, they can be rescued to viability through the injection of oep RNA at the one cell stage (Gritsman et al., 1999). Typically, injected mRNA persists to early somite stages, which is long enough to rescue early Nodal signaling in the gastrula, but not the later Nodal signaling presumed to occur in the brain and LPM (Concha et al., 2000; Liang et al., 2000; Yan et al., 1999). In rescued MZoep (Roep) mutants, tissues that depend on early Nodal signaling, such as endoderm, mesoderm, and ventral neurectoderm, are completely restored. However, expression of cyc, lefty2, and pitx2 in the left LPM is absent, and the direction of heart looping and position of the pancreas is randomized (Yan et al., 1999). The demonstrates that as in other vertebrates, Nodal signaling is required for left-right asymmetry in the LPM and visceral organs. How can these disparate results in cyc and oep mutants be explained? One possibility is that the third zebrafish nodal gene (along with the common cofactor oep) is sucient for activating gene expression in the left LPM. Another possibility is that Nodal signaling during gastrulation or earlier (which is present in cyc mutants), not Nodal signaling in the left LPM, is responsible for activating the expression of pitx2 and other downstream eVectors in the left LPM. Analysis of Roep mutants supports latter hypothesis. To explain the lack of asymmetric gene expression in the LPM of Roep mutants, it was proposed that oep is required to propagate information within the node/organizer, or from the node/organizer to the lateral plate mesoderm (Yan et al., 1999). This suggests that early Nodal signaling in the node/organizer has a role in establishing left-sided gene expression in the LPM. Research in mice also supports this modal: when Nodal activity in the node is selectively removed, asymmetric gene expression in the LPM is not established (Brennan et al., 2002).

164

Liang and Rubinstein

B. Left–Right Asymmetry in the Dorsal Diencephalon The cyc, antivin/lefty1, and pitx2 genes are expressed in the left side of the zebrafish forebrain from mid to late somitogenesis (Bisgrove et al., 1999; Essner et al., 2000; Rebagliati et al., 1998a; Sampath et al., 1998; Thisse and Thisse, 1999). Although many morphological and functional asymmetries are known in vertebrate brains, this is one of the earliest reports of genes in a signaling pathway being expressed exclusively on one side of the brain during development. Consequently, studies of asymmetric Nodal signaling in zebrafish have provided some of the first insights into how morphological asymmetries in the brain arise. Comparison of cyc, atv, and pitx2 expression with the expression of other markers of the diencephalon revealed that they are coexpressed in cells that will give rise to the pineal organ (Concha et al., 2000; Liang et al., 2000). Unlike Nodal signaling in the left LPM, functions for the Nodal signaling pathway in the left side of the presumptive pineal organ have been identified only recently Roep mutants, which lack left-sided gene expression in the brain, as well as in the lateral plate, were valuable tools for exploring the function of the Nodal/Cyc signaling cassette (Concha et al., 2000; Liang et al., 2000). Most teleost fish have an asymmetry in the pineal complex, composed of a central pineal organ and a left-sided parapineal (Borg et al., 1983; Concha and Wilson, 2001). Shortly after genes in the Nodal signaling pathway are expressed in the left forebrain, the parapineal becomes apparent just to the left of the pineal (Gamse et al., 2002). In Roep mutants, which lack Nodal signaling in the left brain, positioning of the parapineal organ is randomized (Concha et al., 2000; Gamse et al., 2002). This suggests that Cyc signaling could directly control formation of the parapineal, perhaps by influencing the migration of parapineal precursors. Later, in the adult brain, left–right positioning of the pineal organ stalk is also altered (Liang et al., 2000). This demonstrates that one role for Nodal signaling in the left side of the pineal anlage is to regulate pineal complex asymmetry. The dorsal habenula nuclei, which neighbor the pineal complex, are also often asymmetric in lower vertebrates (Concha and Wilson, 2001; Harris et al., 1996). In zebrafish, two asymmetries have been noted in the dorsal habenula. Neuropil density is greater on the left side than the right (Concha et al., 2000), and expression of a recently identified habenula-specific gene is more extensive in the left habenula than in the right (Gamse et al., 2003). Habenular asymmetry is randomized in Roep mutants (Concha et al., 2000; Gamse et al., 2003) and appears to be dependent on the presence of the parapineal (Gamse et al., 2003). Thus, the Nodal signaling cascade appears to trigger or mediate a number of left–right decisions in the zebrafish brain.

3. Zebrafish Nodal Signals

165

VII. Future Directions Nodal signals are at the center of many patterning events in the developing embryo. However, there is still much to be learned. One prominent concentration of future studies will undoubtedly be the cellular changes that are induced by activation of the Nodal signaling pathway. This will include defining genes that are (1) regulated by Nodal signaling in diVerent cell lineages and (2) involved in changing the structure and physiology of Nodal-responsive cells. Molecular screens designed to identify transcripts that are aVected by loss or activation of the Nodal signaling pathway are already generating potential Nodal targets (Dickmeis et al., 2001a; Rubinstein et al., 2000). Especially interesting is the finding that genes involved in the cytoskeleton are upregulated upon Nodal pathway activation (Dickmeis et al., 2001a). This suggests that cell motility and architecture are regulated by Nodal signals and could help explain the migration defects that are often associated with loss of Nodal signaling (Carmany-Rampey and Schier, 2001; Conlon et al., 1994; Feldman et al., 1998; Thisse et al., 1994; Zhou et al., 1993). Other important topics for future work include a more detailed characterization of the biochemistry of the Nodal signaling pathway in zebrafish, temporal dissection of diVerent patterning events, and elucidation of the mechanisms by which Nodal signaling controls morphological asymmetries of the viscera and brain.

Acknowledgments We thank Marnie Halpern for her excellent mentorship. Marnie Halpern, Joshua Gamse, and Allisan Aquilina-Beck provided insightful and helpful comments on the manuscript. S.O.L. and A.L.R. were supported by National Research Service Awards.

References Agathon, A., Thisse, B., and Thisse, C. (2001). Morpholino knock-down of antivin 1 and antivin2 upregulates nodal signaling. Genesis 30, 178–182. Albert, S., Muller, F., Fischer, N., Biellmann, D., Neumann, C., Blader, P., and Strahle, U. (2003). Cyclops-independent floor plate differentiation in zebrafish embryos. Dev. Dyn. 226, 59–66. Alexander, J., Rothenberg, M., Henry, G. L., and Stainier, D. Y. (1999). casanova plays an early and essential role in endoderm formation in zebrafish. Dev. Biol. 215, 343–357. Alexander, J., and Stainier, D. Y. (1999). A molecular pathway leading to endoderm formation in zebrafish. Curr. Biol. 9, 1147–1157. Aoki, T. O., David, N. B., Minchiotti, G., Saint-Etienne, L., Dickmeis, T., Persico, G. M., Strahle, U., Mourrain, P., and Rosa, F. M. (2002a). Molecular integration of casanova in the Nodal signalling pathway controlling endoderm formation. Development 129, 275–286.

166

Liang and Rubinstein

Aoki, T. O., Mathieu, J., Saint-Etienne, L., Rebagliati, M. R., Peyrieras, N., and Rosa, F. M. (2002b). Regulation of nodal signalling and mesendoderm formation by TARAM-A, a TGFbeta-related type I receptor. Dev. Biol. 241, 273–288. Barresi, M. J., Stickney, H. L., and Devoto, S. H. (2000). The zebrafish slow-muscle-omitted gene product is required for Hedgehog signal transduction and the development of slow muscle identity. Development 127, 2189–2199. Bisgrove, B. W., Essner, J. J., and Yost, H. J. (1999). Regulation of midline development by antagonism of lefty and nodal signaling. Development 126, 3253–3262. Bisgrove, B. W., Essner, J. J., and Yost, H. J. (2000). Multiple pathways in the midline regulate concordant brain, heart and gut left-right asymmetry. Development 127, 3567–3579. Borg, B., Ekstrom, P., and Van Veen, T. (1983). The parapineal organ of teleosts. Acta Zool. 64, 211–218. Brand, M., Heisenberg, C. P., Warga, R. M., Pelegri, F., Karlstrom, R. O., Beuchle, D., Picker, A., Jiang, Y. J., Furutani-Seiki, M., van Eeden, F. J., Granato, M., Haffter, P., Hammerschmidt, M., Kane, D. A., Kelsh, R. N., Mullins, M. C., Odenthal, J., and Nusslein-Volhard, C. (1996). Mutations affecting development of the midline and general body shape during zebrafish embryogenesis. Development 123, 129–142. Brennan, J., Norris, D. P., and Robertson, E. J. (2002). Nodal activity in the node governs leftright asymmetry. Genes Dev. 16, 2339–2344. Camus, A., and Tam, P. P. (1999). The organizer of the gastrulating mouse embryo. Curr. Top. Dev. Biol. 45, 117–153. Carmany-Rampey, A., and Schier, A. F. (2001). Single-cell internalization during zebrafish gastrulation. Curr. Biol. 11, 1261–1265. Chen, W., Burgess, S., and Hopkins, N. (2001). Analysis of the zebrafish smoothened mutant reveals conserved and divergent functions of hedgehog activity. Development 128, 2385–2396. Chen, Y., and Schier, A. F. (2001). The zebrafish Nodal signal Squint functions as a morphogen. Nature 411, 607–610. Chen, Y., and Schier, A. F. (2002). Lefty proteins are long-range inhibitors of squint-mediated nodal signaling. Curr. Biol. 12, 2124–2128. Chen, Y. G., Hata, A., Lo, R. S., Wotton, D., Shi, Y., Pavletich, N., and Massague, J. (1998). Determinants of specificity in TGF-beta signal transduction. Genes Dev. 12, 2144–2152. Cheng, S. K., Olale, F., Bennett, J. T., Brivanlou, A. H., and Schier, A. F. (2003). EGF CFC proteins are essential coreceptors for the TGF-beta signals Vg1 and GDF1. Genes Dev. 17, 31–36. Chin, A. J., Tsang, M., and Weinberg, E. S. (2000). Heart and gut chiralities are controlled independently from initial heart position in the developing zebrafish. Dev. Biol. 227, 403–421. Colamarino, S. A., and Tessier-Lavigne, M. (1995). The role of the floor plate in axon guidance. Annu. Rev. Neurosci. 18, 497–529. Concha, M. L., Burdine, R. D., Russell, C., Schier, A. F., and Wilson, S. W. (2000). A nodal signaling pathway regulates the laterality of neuroanatomical asymmetries in the zebrafish forebrain. Neuron 28, 399–409. Concha, M. L., and Wilson, S. W. (2001). Asymmetry in the epithalamus of vertebrates. J. Anat. 199, 63–84. Conlon, F. L., Lyons, K. M., Takaesu, N., Barth, K. S., Kispert, A., Herrmann, B., and Robertson, E. J. (1994). A primary requirement for nodal in the formation and maintenance of the primitive streak in the mouse. Development 120, 1919–1928. Cooper, M. S., and D’Amico, L. A. (1996). A cluster of noninvoluting endocytic cells at the margin of the zebrafish blastoderm marks the site of embryonic shield formation. Dev. Biol. 180, 184–198.

3. Zebrafish Nodal Signals

167

Daopin, S., Piez, K. A., Ogawa, Y., and Davies, D. R. (1992). Crystal structure of transforming growth factor beta2: A usual fold for the superfamily. Science 257, 369–372. David, N. B., and Rosa, F. M. (2001). Cell autonomous commitment to an endodermal fate and behaviour by activation of Nodal signalling. Development 128, 3937–3947. Dick, A., Mayr, T., Bauer, H., Meier, A., and Hammerschmidt, M. (2000). Cloning and characterization of zebrafish smad2, smad3 and smad4. Gene 246, 69–80. Dickmeis, T., Aanstad, P., Clark, M., Fischer, N., Herwig, R., Mourrain, P., Blader, P., Rosa, F., Lehrach, H., and Strahle, U. (2001a). Identification of nodal signaling targets by array analysis of induced complex probes. Dev. Dyn. 222, 571–580. Dickmeis, T., Mourrain, P., Saint-Etienne, L., Fischer, N., Aanstad, P., Clark, M., Strahle, U., and Rosa, F. (2001b). A crucial component of the endoderm formation pathway, CASANOVA, is encoded by a novel sox-related gene. Genes Dev. 15, 1487–1492. Erter, C. E., Solnica-Krezel, L., and Wright, C. V. (1998). Zebrafish nodal-related 2 encodes an early mesendodermal inducer signaling from the extraembryonic yolk syncytial layer. Dev. Biol. 204, 361–372. Essner, J. J., Branford, W. W., Zhang, J., and Yost, H. J. (2000). Mesendoderm and left-right brain, heart and gut development are differentially regulated by pitx2 isoforms. Development 127, 1081–1093. Essner, J. J., Vogan, K. J., Wagner, M. K., Tabin, C. J., Yost, H. J., and Brueckner, M. (2002). Conserved function for embryonic nodal cilia. Nature 418, 37–38. Feldman, B., Concha, M. L., Saude, L., Parsons, M. J., Adams, R. J., Wilson, S. W., and Stemple, D. L. (2002). Lefty antagonism of squint is essential for normal gastrulation. Curr. Biol. 12, 2129–2135. Feldman, B., Dougan, S. T., Schier, A. F., and Talbot, W. S. (2000). Nodal-related signals establish mesendodermal fate and trunk neural identity in zebrafish. Curr. Biol. 10, 531–534. Feldman, B., Gates, M. A., Egan, E. S., Dougan, S. T., Rennebeck, G., Sirotkin, H. I., Schier, A. F., and Talbot, W. S. (1998). Zebrafish organizer development and germ-layer formation require nodal-related signals. Nature 395, 181–185. Gamse, J. T., Shen, Y. C., Thisse, C., Thisse, B., Raymond, P. A., Halpern, M. E., and Liang, J. O. (2002). Otx5 regulates genes that show circadian expression in the zebrafish pineal complex. Nature Genet. 30, 117–121. Gamse, J. T., Thisse, C., Thisse, B., and Halpern, M. E. (2003). The parapineal mediates leftright asymmetry in the zebrafish diencephalon. Development 130, 1059–1068. Garg, R. R., Bally-Cuif, L., Lee, S. E., Gong, Z., Ni, X., Hew, C. L., and Peng, C. (1999). Cloning of zebrafish activin type IIB receptor (ActRIIB) cDNA and mRNA expression of ActRIIB in embryos and adult tissues. Mol. Cell. Endocrinol. 153, 169–181. Goldstein, A. M., Ticho, B. S., and Fishman, M. C. (1998). Patterning the heart’s left-right axis: From zebrafish to man. Dev. Genet. 22, 278–287. Gripp, K. W., Wotton, D., Edwards, M. C., Roessler, E., Ades, L., Meinecke, P., RichieriCosta, A., Zackai, E. H., Massague, J., Muenke, M., and Elledge, S. J. (2000). Mutations in TGIF cause holoprosencephaly and link NODAL signaling to human neural axis determination. Nature Genet. 25, 205–208. Gritsman, K., Talbot, W. S., and Schier, A. F. (2000). Nodal signaling patterns the organizer. Development 127, 921–932. Gritsman, K., Zhang, J., Cheng, S., Heckscher, E., Talbot, W. S., and Schier, A. F. (1999). The EGF-CFC protein one-eyed pinhead is essential for nodal signaling. Cell 97, 121–132. Halpern, M. E., Hatta, K., Amacher, S. L., Talbot, W. S., Yan, Y. L., Thisse, B., Thisse, C., Postlethwait, J. H., and Kimmel, C. B. (1997). Genetic interactions in zebrafish midline development. Dev. Biol. 187, 154–170.

168

Liang and Rubinstein

Hamada, H., Meno, C., Watanabe, D., and Saijoh, Y. (2002). Establishment of vertebrate leftright asymmetry. Nature Rev. Genet. 3, 103–113. Harris, J. A., Guglielmotti, V., and Bentivoglio, M. (1996). Diencephalic asymmetries. Neurosci. Biobehav. Rev. 20, 637–643. Hatta, K. (1992). Role of the floor plate in axonal patterning in the zebrafish CNS. Neuron 9, 629–642. Hatta, K., Kimmel, C. B., Ho, R. K., and Walker, C. (1991). The cyclops mutation blocks specification of the floor plate of the zebrafish central nervous system. Nature 350, 339–341. Hatta, K., Puschel, A. W., and Kimmel, C. B. (1994). Midline signaling in the primordium of the zebrafish anterior central nervous system. Proc. Natl. Acad. Sci. USA 91, 2061–2065. Heisenberg, C. P., and Nusslein-Volhard, C. (1997). The function of silberblick in the positioning of the eye anlage in the zebrafish embryo. Dev. Biol. 184, 85–94. Jones, C. M., Kuehn, M. R., Hogan, B. L., Smith, J. C., and Wright, C. V. (1995). Nodalrelated signals induce axial mesoderm and dorsalize mesoderm during gastrulation. Development 121, 3651–3662. Joseph, E. M., and Melton, D. A. (1997). Xnr4: A Xenopus nodal-related gene expressed in the Spemann organizer. Dev. Biol. 184, 367–372. Kiecker, C., Muller, F., Wu, W., Glinka, A., Strahle, U., and Niehrs, C. (2000). Phenotypic effects in Xenopus and zebrafish suggest that one-eyed pinhead functions as antagonist of BMP signalling. Mech. Dev. 94, 37–46. Kikuchi, Y., Agathon, A., Alexander, J., Thisse, C., Waldron, S., Yelon, D., Thisse, B., and Stainier, D. Y. (2001). Casanova encodes a novel Sox-related protein necessary and sufficient for early endoderm formation in zebrafish. Genes Dev. 15, 1493–1505. Kikuchi, Y., Trinh, L. A., Reiter, J. F., Alexander, J., Yelon, D., and Stainier, D. Y. (2000). The zebrafish bonnie and clyde gene encodes a Mix family homeodomain protein that regulates the generation of endodermal precursors. Genes Dev. 14, 1279–1289. Kingsley, D. M. (1994). The TGF-beta superfamily: New members, new receptors, and new genetic tests of function in different organisms. Genes Dev. 8, 133–146. Latinkic, B. V., Umbhauer, M., Neal, K. A., Lerchner, W., Smith, J. C., and Cunliffe, V. (1997). The Xenopus Brachyury promoter is activated by FGF and low concentrations of activin and suppressed by high concentrations of activin and by paired-type homeodomain proteins. Genes Dev. 11, 3265–3276. Le Douarin, N. M., and Halpern, M. E. (2000). Origin and specification of the neural tube floor plate: Insights from the chick and zebrafish. Curr. Opin. Neurobiol. 10, 23–30. Liang, J. O., Etheridge, A., Hantsoo, L., Rubinstein, A. L., Nowak, S. J., Izpisua Belmonte, J. C., and Halpern, M. E. (2000). Asymmetric nodal signaling in the zebrafish diencephalon positions the pineal organ. Development 127, 5101–5112. Masai, I., Stemple, D. L., Okamoto, H., and Wilson, S. W. (2000). Midline signals regulate retinal neurogenesis in zebrafish. Neuron 27, 251–263. Mason, A. J. (1994). Functional analysis of the cysteine residues of activin A. Mol. Endocrinol 8, 325–332. Massague, J. (1998). TGF-beta signal transduction. Annu. Rev. Biochem. 67, 753–791. Massague, J., and Chen, Y. G. (2000). Controlling TGF-beta signaling. Genes Dev. 14, 627–644. Massague, J., and Wotton, D. (2000). Transcriptional control by the TGF-beta/Smad signaling system. EMBO J. 19, 1745–1754. Mathieu, J., Barth, A., Rosa, F. M., Wilson, S. W., and Peyrieras, N. (2002). Distinct and cooperative roles for Nodal and Hedgehog signals during hypothalamic development. Development 129, 3055–3065.

3. Zebrafish Nodal Signals

169

McDonald, N. Q., and Hendrickson, W. A. (1993). A structural superfamily of growth factors containing a cystine know motif. Cell 73, 421–424. Melby, A. E., Warga, R. M., and Kimmel, C. B. (1996). Specification of cell fates at the dorsal margin of the zebrafish gastrula. Development 122, 2225–2237. Meno, C., Gritsman, K., Ohishi, S., Ohfuji, Y., Heckscher, E., Mochida, K., Shimono, A., Kondoh, H., Talbot, W. S., Robertson, E. J., Schier, A. F., and Hamada, H. (1999). Mouse Lefty2 and zebrafish antivin are feedback inhibitors of nodal signaling during vertebrate gastrulation. Mol. Cell 4, 287–298. Meno, C., Saijoh, Y., Fujii, H., Ikeda, M., Yokoyama, T., Yokoyama, M., Toyoda, Y., and Hamada, H. (1996). Left-right asymmetric expression of the TGF beta-family member lefty in mouse embryos. Nature 381, 151–155. Mercola, M., and Levin, M. (2001). Left right asymmetry determination in vertebrates. Annu. Rev. Cell Dev. Biol. 17, 779–805. Muller, F., Albert, S., Blader, P., Fischer, N., Hallonet, M., and Strahle, U. (2000). Direct action of the nodal-related signal cyclops in induction of sonic hedgehog in the ventral midline of the CNS. Development 127, 3889–3897. Muller, F., Blader, P., Rastegar, S., Fischer, N., Knochel, W., and Strahle, U. (1999). Characterization of zebrafish smad1, smad2 and smad5: The amino-terminus of smad1 and smad5 is required for specific function in the embryo. Mech. Dev. 88, 73–88. Nagaso, H., Suzuki, A., Tada, M., and Ueno, N. (1999). Dual specificity of activin type II receptor ActRIIb in dorso-ventral patterning during zebrafish embryogenesis. Dev. Growth Differ. 41, 119–133. Nasevicius, A., and Ekker, S. C. (2000). Effective targeted gene ‘knockdown’ in zebrafish. Nature Genet. 26, 216–220. Ober, E. A., Field, H. A., and Stainier, D. Y. (2003). From endoderm formation to liver and pancreas development in zebrafish. Mech. Dev. 120, 5–18. Peyrieras, N., Strahle, U., and Rosa, F. (1998). Conversion of zebrafish blastomeres to an endodermal fate by TGF-beta-related signaling. Curr. Biol. 8, 783–786. Piccolo, S., Agius, E., Leyns, L., Bhattacharyya, S., Grunz, H., Bouwmeester, T., and De Robertis, E. M. (1999). The head inducer Cerberus is a multifunctional antagonist of Nodal, BMP and Wnt signals. Nature 397, 707–710. Placzek, M., Dodd, J., and Jessell, T. M. (2000). The case for floor plate induction by the notochord. Curr. Opin. Neurobiol. 10, 15–22. Pogoda, H. M., Solnica-Krezel, L., Driever, W., and Meyer, D. (2000). The zebrafish forkhead transcription factor FoxH1/Fast1 is a modulator of nodal signaling required for organizer formation. Curr. Biol. 10, 1041–1049. Poulain, M., and Lepage, T. (2002). Mezzo, a paired-like homeobox protein is an immediate target of Nodal signaling and regulates endoderm specification in zebrafish. Development 129, 4901–4914. Rastegar, S., Albert, S., Le Roux, I., Fischer, N., Blader, P., Muller, F., and Strahle, U. (2002). A floor plate enhancer of the zebrafish netrin1 gene requires cyclops (Nodal) signaling and the winged helix transcription factor FoxA2. Dev. Biol. 252, 1–14. Rebagliati, M. R., Toyama, R., Fricke, C., Haffter, P., and Dawid, I. B. (1998a). Zebrafish nodal-related genes are implicated in axial patterning and establishing left-right asymmetry. Dev. Biol. 199, 261–272. Rebagliati, M. R., Toyama, R., Haffter, P., and Dawid, I. B. (1998b). Cyclops encodes a nodalrelated factor involved in midline signaling. Proc. Natl. Acad. Sci. USA 95, 9932–9937. Rebbert, M. L., and Dawid, I. B. (1997). Transcriptional regulation of the Xlim-1 gene by activin is mediated by an element in intron I. Proc. Natl. Acad. Sci. USA 94, 9717–9722. Reissmann, E., Jornvall, H., Blokzijl, A., Andersson, O., Chang, C., Minchiotti, G., Persico, M. G., Ibanez, C. F., and Brivanlou, A. H. (2001). The orphan receptor ALK7 and the Activin

170

Liang and Rubinstein

receptor ALK4 mediate signaling by Nodal proteins during vertebrate development. Genes Dev. 15, 2010–2022. Reiter, J. F., Alexander, J., Rodaway, A., Yelon, D., Patient, R., Holder, N., and Stainier, D. Y. (1999). Gata5 is required for the development of the heart and endoderm in zebrafish. Genes Dev. 13, 2983–2995. Reiter, J. F., Kikuchi, Y., and Stainier, D. Y. (2001). Multiple roles for Gata5 in zebrafish endoderm formation. Development 128, 125–135. Renucci, A., Lemarchandel, V., and Rosa, F. (1996). An activated form of type I serine/ threonine kinase receptor TARAM-A reveals a specific signaling pathway involved in fish head organizer formation. Development 122, 3735–3743. Rohr, K. B., Barth, K. A., Varga, Z. M., and Wilson, S. W. (2001). The nodal pathway acts upstream of hedgehog signaling to specify ventral telencephalic identity. Neuron 29, 341–351. Roy, S., Qiao, T., Wolff, C., and Ingham, P. W. (2001). Hedgehog signaling pathway is essential for pancreas specification in the zebrafish embryo. Curr. Biol. 11, 1358–1363. Rubinstein, A. L., Lee, D., Luo, R., Henion, P. D., and Halpern, M. E. (2000). Genes dependent on zebrafish cyclops function identified by AFLP differential gene expression screen. Genesis 26, 86–97. Sakaguchi, T., Kuroiwa, A., and Takeda, H. (2001). A novel sox gene, 226D7, acts downstream of Nodal signaling to specify endoderm precursors in zebrafish. Mech. Dev. 107, 25–38. Sampath, K., Rubinstein, A. L., Cheng, A. M., Liang, J. O., Fekany, K., Solnica-Krezel, L., Korzh, V., Halpern, M. E., and Wright, C. V. (1998). Induction of the zebrafish ventral brain and floorplate requires cyclops/nodal signaling. Nature 395, 185–189. Schier, A. F., Neuhauss, S. C., Harvey, M., Malicki, J., Solnica-Krezel, L., Stainier, D. Y., Zwartkruis, F., Abdelilah, S., Stemple, D. L., Rangini, Z., Yang, H., and Driever, W. (1996). Mutations affecting the development of the embryonic zebrafish brain. Development 123, 165–178. Schier, A. F., Neuhauss, S. C., Helde, K. A., Talbot, W. S., and Driever, W. (1997). The oneeyed pinhead gene functions in mesoderm and endoderm formation in zebrafish and interacts with no tail. Development 124, 327–342. Shen, M. M., and Schier, A. F. (2000). The EGF-CFC gene family in vertebrate development. Trends Genet. 16, 303–309. Shimizu, T., Yamanaka, Y., Ryu, S. L., Hashimoto, H., Yabe, T., Hirata, T., Bae, Y. K., Hibi, M., and Hirano, T. (2000). Cooperative roles of Bozozok/Dharma and Nodalrelated proteins in the formation of the dorsal organizer in zebrafish. Mech. Dev. 91, 293–303. Shinya, M., Furutani-Seiki, M., Kuroiwa, A., and Takeda, H. (1999). Mosaic analysis with oep mutant reveals a repressive interaction between floor-plate and nonfloor-plate mutant cells in the zebrafish neural tube. Dev. Growth Differ. 41, 135–142. Sirotkin, H. I., Dougan, S. T., Schier, A. F., and Talbot, W. S. (2000a). Bozozok and squint act in parallel to specify dorsal mesoderm and anterior neuroectoderm in zebrafish. Development 127, 2583–2592. Sirotkin, H. I., Gates, M. A., Kelly, P. D., Schier, A. F., and Talbot, W. S. (2000b). Fast1 is required for the development of dorsal axial structures in zebrafish. Curr. Biol. 10, 1051–1054. Solnica-Krezel, L., and Driever, W. (2001). The role of the homeodomain protein Bozozok in zebrafish axis formation. Int. J. Dev. Biol. 45, 299–310. Strahle, U., Blader, P., Henrique, D., and Ingham, P. W. (1993). Axial, a zebrafish gene expressed along the developing body axis, shows altered expression in cyclops mutant embryos. Genes Dev. 7, 1436–1446. Strahle, U., Fischer, N., and Blader, P. (1997). Expression and regulation of a netrin homologue in the zebrafish embryo. Mech. Dev. 62, 147–160.

3. Zebrafish Nodal Signals

171

Thisse, B., Wright, C. V., and Thisse, C. (2000). Activin- and Nodal-related factors control antero-posterior patterning of the zebrafish embryo. Nature 403, 425–428. Thisse, C., and Thisse, B. (1999). Antivin, a novel and divergent member of the TGFbeta superfamily, negatively regulates mesoderm induction. Development 126, 229–240. Thisse, C., Thisse, B., Halpern, M. E., and Postlethwait, J. H. (1994). Goosecoid expression in neurectoderm and mesendoderm is disrupted in zebrafish cyclops gastrulas. Dev. Biol. 164, 420–429. Varga, Z. M., Amores, A., Lewis, K. E., Yan, Y. L., Postlethwait, J. H., Eisen, J. S., and Westerfield, M. (2001). Zebrafish smoothened functions in ventral neural tube specification and axon tract formation. Development 128, 3497–3509. Warga, R. M., and Nusslein-Volhard, C. (1999). Origin and development of the zebrafish endoderm. Development 126, 827–838. Watabe, T., Kim, S., Candia, A., Rothbacher, U., Hashimoto, C., Inoue, K., and Cho, K. W. (1995). Molecular mechanisms of Spemann’s organizer formation: Conserved growth factor synergy between Xenopus and mouse. Genes Dev. 9, 3038–3050. Watanabe, M., Rebbert, M. L., Andreazzoli, M., Takahashi, N., Toyama, R., Zimmerman, S., Whitman, M., and Dawid, I. B. (2002). Regulation of the Lim-1 gene is mediated through conserved FAST-1/FoxH1 sites in the first intron. Dev. Dyn. 225, 448–456. Yan, Y. T., Gritsman, K., Ding, J., Burdine, R. D., Corrales, J. D., Price, S. M., Talbot, W. S., Schier, A. F., and Shen, M. M. (1999). Conserved requirement for EGF-CFC genes in vertebrate left-right axis formation. Genes Dev. 13, 2527–2537. Yeo, C., and Whitman, M. (2001). Nodal signals to Smads through Cripto-dependent and Cripto-independent mechanisms. Mol. Cell 7, 949–957. Zhang, J., Talbot, W. S., and Schier, A. F. (1998). Positional cloning identifies zebrafish oneeyed pinhead as a permissive EGF-related ligand required during gastrulation. Cell 92, 241–251. Zhou, X., Sasaki, H., Lowe, L., Hogan, B. L., and Kuehn, M. R. (1993). Nodal is a novel TGFbeta-like gene expressed in the mouse node during gastrulation. Nature 361, 543–547.

ThisPageIntentionallyLeftBlank

4

____________________________________________________________________________

Folding Chromosomes in Bacteria: Examining the Role of Csp Proteins and Other Small Nucleic Acid-Binding Proteins Nancy Trun and Danielle Johnston Department of Biological Sciences Duquesne University Pittsburgh, Pennsylvania 15282

I. Introduction II. Small Nucleic Acid-Binding Proteins Implicated in Chromosome Folding in Escherichia coli A. HNS B. HU and IHF C. FIS D. The Csp Family III. Csp Proteins in Escherichia coli A. CspA B. CspB C. CspC D. CspD E. CspE F. CspG and CspI G. CspF and CspH IV. Relationships among Escherichia coli Csp Proteins V. Csp Proteins of Bacillus subtilis VI. The Crystal Structures of CspA and Related Proteins A. Structure of CspA B. Sequence Homologues of the Csp Family in Eukaryotes—Y-Box Factors including Human YB-1 C. Structural Homologues of the Csp Family—the Sac 7d Protein D. Dimerization of a Csp Protein—Structure of the Bacillus subtilis CspB Protein VII. Distribution of Small DNA-Binding Proteins in Archaea and Bacteria VIII. Conclusions References

I. Introduction In systems where chromosome condensation has been studied most extensively, one overriding principle that has emerged is that no one mechanism is responsible for all of the folding of the DNA that must take place. Rather, Current Topics in Developmental Biology, Vol. 55 Copyright 2003, Elsevier (USA). All rights reserved. 0070-2153/03 $35.00

173

174

Trun and Johnston

several mechanisms contribute to the final condensed state of a chromosome. Each mechanism requires specific proteins or other components to carry out its function. In bacteria, several small nucleic acid-binding proteins have been implicated in chromosome condensation. This review examines the distribution of these proteins in the three domains, what is known about them, and the roles they play in the cell. We describe the heat-unstable nucleoid protein (HU), integration host factor (IHF), histone-like nucleoid structuring (HNS), and factor for inversion stimulation (Fis) proteins and focus on the cold-shock and cold-shock-like (Csp) protein family. In eukaryotes, the most basic level of condensation is naked DNA wrapped around nucleosomes. Nucleosoming of DNA leads to approximately sixfold condensation of the chromosome. The nucleosomed DNA is condensed to a 30-nm fiber, and this fiber is folded into loops that contain approximately 50–100 kb of DNA. The looped fiber is further condensed through several steps to give the final structure of the mitotic chromosome. Two major classes of proteins used by eukaryotes in the condensation process are the histones that make up the nucleosomes and structural maintenance of chromosomes (SMC) proteins that are used for condensing chromosomes, sister-chromatid cohesion, and in other processes such as DNA repair and recombination (Hirano, 2002; Jessberger, 2002). In the archaea domain, several species have been shown to contain histone proteins and to wrap DNA around these structures (Sandman et al., 1990, 1994). DNA in the nucleosome-like structures (NLS) is constrained in positive supercoils in the archaea rather than in negative supercoils, such as eukaryotic nucleosomes. The NLS from the archaeon Methanothermus fervidus contain two proteins that are related to eukaryotic histones (Sandman et al., 1990, 1994). These archaeal histones form dimers in solution and compact DNA. Homologues of the archaeal histones have been found in several but not all sequenced archaeal genomes and in none of the sequenced bacterial genomes. Several of the sequenced archaeal genomes contain SMC homologues, although detecting SMCs in BLAST searches depends on which SMC is used as the bait. Using the Bacillus subtilis SMC as the bait, 4 out of the 10 sequenced archaeal genomes contain one homologue. Twenty-five out of the 47 sequenced bacterial genomes contain a homologue of the B. subtilis SMC. Again, each genome contains only one homologue. In Escherichia coli, the single circular chromosome of 4639 kb is arranged into 50- to 100-kb domains that form a rosette structure called the nucloid. Each domain is independently supercoiled with supercoiling condensing the DNA by 2-fold at most. This supercoiled rosette structure is 10-fold smaller than a randomly folded molecule of the same size. However, it must be condensed 2- to 3-fold more in order to fit inside the cell in the space allotted to the nucleoid. E. coli contains a single SMC-like protein called MukB. The

4. Folding Chromosomes in Bacteria

175

mukB gene can be deleted from the chromosome, but the resulting cells exhibit a 1000-fold increase in cells without chromosomes (anucleate cells), dramatically unfolded chromosomes, and will only grow at temperatures  below 20 C. Despite much eVort, no structural or functional homologues of either eukaryotic or archaeal histone proteins have been found in E. coli. Through independent studies on individual genes, several small nucleic acidbinding proteins have been implicated in condensation. These include HU, HNS, Fis, CspC, and CspE.

II. Small Nucleic Acid-Binding Proteins Implicated in Chromosome Folding in Escherichia coli Escherichia coli contains a number of small proteins between 5 and 20 kDa that bind to nucleic acids. These include HNS, HU, and Fis. Some of the proteins bind nonspecifically to nucleic acids and others exhibit sequence specificity. They have been shown to play roles in homologous recombination, transposition, site-specific recombination, DNA inversion, DNA replication, gene regulation, and chromosome condensation. When nucleoids are isolated from cells, four major proteins are found associated with it and are released upon digestion of the DNA by DNase (Murphy and Zimmerman, 1997). These proteins are RNA polymerase, HNS, HU, and Fis. In the course of studying their functions, roles for them in chromosome condensation were uncovered. None of them are dedicated to chromosome condensation, rather they all play multiple roles in the cell.

A. HNS HNS is a 15.4-kDa protein that exists as a homodimer or an oligomer in vivo and in vitro. HNS binds to DNA with no sequence specificity but with some preference for curved DNA. It contains three distinct domains: an N-terminal domain, the central region that is responsible for multimerization, and the C-terminal domain that binds DNA (Dame et al., 2000). HNS plays roles in gene regulation, DNA replication, recombination, and transpositions, as well as chromosome condensation (AZerbach et al., 1998; Atlung and Hansen, 2002; Hulton et al., 1990; O’Gara and Dorman, 2000; White-Ziegler et al., 2000). E. coli contains a second protein called StpA that is 58% identical in amino acid sequence to HNS. It can form heterodimers with HNS, and can substitute for it in some reactions (Johansson et al., 2001). Using atomic force microscopy, the structural changes to relaxed circular DNA induced by HNS were investigated. HNS binding at one dimer of HNS per 12 bp of DNA produced two diVerent classes of HNS-induced

176

Trun and Johnston

structural changes to the DNA. One class of molecules contains loops at the ends of the molecules with a ‘‘straight’’ region between the loops. This class of HNS-induced structures condenses the molecules by 3%. The other class of structural changes produces foci of HNS molecules at the base of the DNA loops. This second class of structural changes condenses the DNA by 25%. When the HNS-to-DNA ratio is increased to one HNS dimer per 6 bp, the resulting DNA molecules resemble the first class of structures; however, foci between loops on diVerent DNA molecules are more numerous and the compaction factor is increased to >50% (Dame et al., 2000). From these experiments, the following model for HNS condensation of DNA has been proposed. First, HNS binds randomly to the DNA without condensing it considerably. When DNA strands are in close contact, the bound HNS interacts with another strand of DNA or with another molecule of bound HNS. This interaction forms a bridge between DNA strands and condenses the DNA. Other HNS molecules bound to DNA strands can interact with these bridges of HNS, forming HNS-DNA foci and further compacting the DNA (Dame et al., 2000). This model requires stoiciometric amounts of HNS and DNA for condensation. HNS reaches a maximum concentration of 20,000 copies per cell in E. coli during exponential growth. The amount of HNS decreases 60% to 8000 molecules/cell by late stationary phase (Azam et al., 1999). At 20,000 copies per cell or 10,000 dimers per cell, HNS can coat between 1 and 2.5% of the chromosome, depending on if it binds every 6 or every 12 bp.

B. HU and IHF HU is a 9.5-kDa basic protein that functions as a heterodimer in E. coli (Laine et al., 1978). The heterodimer is composed of one subunit of HU (encoded by hupA) and one subunit of HU (encoded by hupB) (Laine et al., 1978). Homodimers of either subunit can exist. In a deletion of hupA, homodimers of HU exist at 10% of the concentration of heterodimers in wild-type cells. In a hupB deletion, HU homodimers exist at 50% of the concentration of heterodimers in wild-type cells (Bonnefoy et al., 1989). The heterodimer of HU binds to cruciform, nicked, and gapped DNA with a greater aYnity than for linear DNA. The HU2 homodimer behaves very similar to the heterodimer. The homodimer of HU2 and the heterodimer HU bind to linear DNA with a greater aYnity than the HU 2 homodimer (Pinson et al., 1999). Deletion of both hupA and hupB from E. coli alters a number of processes, including transposition, recombination, replication, transcription, and chromosome condensation. The crystal structure of HU has been determined and it consists of three strands and three  helices. A dimer of HU binds to the minor groove of

4. Folding Chromosomes in Bacteria

177

DNA and one of the  helices acts as an arm, extending around the DNA to the opposite face of the DNA helix. This conformation of the protein bends the DNA between the two HU subunits (Lavoie et al., 1996). HU heterodimers bind tightly to DNA and essentially coat the molecule by binding one heterodimer about every 9 bp of linear DNA (Aki and Adhya, 1997; Lavoie et al., 1996; Pinson et al., 1999). During exponential growth, there are between 30,000 and 55,000 HU molecules/cell. The amount of HU molecules decreases upon entry into the stationary phase until it reaches 10,000 and 20,000 molecule/cell during the late stationary phase (Azam et al., 1999). At 30,000–55,000 monomers per cell or 15,000–27,500 dimers per cell, HU could coat between 3 and 5% of the chromosome. For many of its functions, only a few dimers of HU bound at specific locations are required. For example, the DNA conformation created by several dimers of HU binding to Mu sequences is necessary and suYcient for transposition (Lavoie et al., 1996). HU binds to the Gal operon in a siteand orientation-specific fashion, but the binding may not necessarily be sequence specific. The binding of HU occurs after GalR binds to its two operators and the HU creates a DNA loop that brings the GalR dimers together, forming a tetramer that represses transcription (Aki and Adhya, 1997). An intermediate of DNA repair and recombination, a nicked DNA molecule, forms a complex with HU. HU can bind the 30 overhang of a dsDNA break, and the D loop, but not to dsDNA branches (Kamashev and Rouviere-Yaniv, 2000). For these processes, HU does not have to coat the DNA. However, in all cases where it has been examined, HU must bind and bend DNA in order to carry out its functions. Escherichia coli contains a second heterodimeric protein called IHF that shares 35% homology at the amino acid level with HU. In isolated nucloids, some IHF is present (Murphy and Zimmerman, 1997). IHF can also substitute for HU in a number of diVerent assays. In contrast to HU, IHF binds to DNA in a sequence-specific manner. It recognizes a 13-bp consen sus sequence (A/T ATCAANNNNTT-Pu) and induces a 140 bend in the DNA when it binds (Yang and Nash, 1989). The concentration of IHF is 12,000 molecules/cell during exponential growth, increases to 55,000 molecules/cell during the transition from exponential growth to stationary phase, and decreases to 30,000 molecules/cell when the cells enter late stationary phase (Azam et al., 1999).

C. FIS Fis is an 11.2-kDa protein that functions as a homodimer (Yuan et al., 1991). It is a nonspecific DNA-binding protein that plays roles in DNA inversion, gene regulation, and chromosome condensation. Fis binding to

178

Trun and Johnston

nicked plasmid DNA increases the number of negative supercoils trapped upon ligation of the DNA molecule. At low concentrations of Fis and plasmid DNA (1 dimer per 325 bp), the DNA molecule appeared to have more branches (72% more branches) as shown by electron microscopy. Under these conditions, Fis was located at points in the DNA where one strand of DNA crosses another and in DNA loops. As the concentration of Fis is increased, DNA molecules contain more branches. When the concentration of Fis is increased to >1 dimer per 80 bp of DNA, the DNA forms compacted structures. This implies that Fis may condense DNA by binding to and stabilizing branches in DNA molecules (Schneider et al., 2001). Fis is a very abundant protein during exponential growth (60,000 molecules/cell) but decreases rapidly to undetectable levels upon entry into stationary phase. Recovery from stationary phase creates a Fis peak prior to the first round of cell division. The expression of fis is autoregulated (Azam et al., 1999). At 60,000 molecules per cell or 30,000 dimers per cell and one Fis molecule per 80 bp of DNA, there is enough Fis to condense approximately 52% of the chromosome. Fis–DNA and HNS–DNA complexes have been studied using a restriction enzyme access assay. Fis–DNA or HNS–DNA complexes were formed in vitro followed by the addition of a restriction enzyme (FokI) that had multiple recognition sites within the DNA. In Fis–DNA complexes, some of the restriction sites were blocked whereas others were cleavable. This implies that Fis binds preferentially to certain sites in DNA before it begins to bind randomly to the DNA. The same experiment conducted using HNS–DNA complexes demonstrated that HNS could block cleavage of all restriction sites equally, supporting sequence-independent binding. Similar experiments were conducted using a diVerent restriction enzyme (SfiI) whose cleavage of DNA molecules is enhanced in supercoiled and branched DNA. Incubating DNA with Fis prior to SfiI treatment enhanced the eYciency of cleavage. Increasing the amount of Fis in the complex increased the amount of cleavage by SfiI. Similar experiments with HNS produced no increase in SfiI-dependent cleavage and gradually reduced the eYciency of the enzyme in a concentration-dependent manner (Schneider et al., 2001). While Fis and HNS both compact DNA, these experiments demonstrate that they do so by diVerent mechanisms.

D. The Csp Family The role(s) for CspC and CspE in chromosome condensation was identified by their ability to suppress the defects associated with a mutation in the E. coli SMC gene, mukB. Overexpression of cspC or cspE reverses both the temperature-sensitive phenotype and the production of anucleate cells seen

179

4. Folding Chromosomes in Bacteria

CspA

CspA

CspB

CspC

CspD

CspE

CspF

CspG

CspH

CspI

80%

69%

49%

70%

44%

73%

47%

70%

64%

50%

69%

43%

77%

47%

70%

49%

83%

40%

67%

43%

71%

51%

29%

44%

30%

47%

46%

67%

49%

73%

44%

77%

46%

47%

79%

CspB

84%

CspC

79%

74%

CspD

64%

63%

70%

CspE

81%

77%

91%

74%

CspF

61%

62%

59%

46%

59%

CspG

84%

84%

80%

66%

79%

61%

CspH

57%

63%

62%

46%

60%

87%

59%

CspI

79%

81%

83%

66%

79%

60%

90%

% Identity

in mukB mutations (Yamanaka et al., 1994). Overexpression of cspE also leads to resistance to the DNA decondensing agent, camphor. In cells treated with camphor, the size and shape of the cell do not change but the DNA decondenses to fill the entire cytoplasm. Overexpression of cspE makes cells resistant to camphor and prevents the nucleoid from decondensation by camphor both in vivo and in vitro (Hu et al., 1996). CspC and CspE are members of a larger family of proteins that share between 50 and 85% amino acid identity (Fig. 1). Prokaryotic cells frequently contain more than one csp gene and, in fact, E. coli K12 contains nine csp genes. What makes this so remarkable is that E. coli does not usually contain

47%

59%

% Similarity Figure 1 Percent identity and percent similarity of amino acid sequences of the nine Csp proteins from E. coli. Based on percent identity, the nine Csp proteins group into pairs: CspA with CspB, CspC with CspE, CspG with CspI, and CspF with CspH. CspD appears to be a single gene, consistent with the fact that it is regulated diVerently from all of the other csp genes.

180

Trun and Johnston

multiple copies of such similar genes. It contains only a handful of duplicate genes and, in fact, ribosomal RNA genes are only present in seven copies in the chromosome of E. coli K12. The remainder of this review focuses on the structure, function, and regulation of the Csp family of proteins.

III. Csp Proteins in Escherichia coli 



When E. coli cells growing exponentially at 37 C are shifted to 10 C, the cells enter a 4-h lag phase before exponential growth resumes (Jones et al., 1987). Two-dimensional gel electrophoresis revealed that the synthesis of most proteins is reduced immediately following the temperature shift; however, 2 h into the lag period 28 proteins are synthesized. Only 1 protein of the 28 is not present at 37 C (Jones et al., 1987). The rates of protein synthesis were determined for the 28 proteins made during the lag phase. The proteins were classified into three diVerent categories: synthesis of the protein (1) decreased, (2) remained constant, or (3) increased. The third category contains 13 cold shock-induced proteins (Jones et al., 1987). Pulse labeling cells after temperature shift and separating the proteins by SDS– PAGE identified a 7.4-kDa cold shock-induced protein (CS7.4). CS7.4 was sequenced and used to create a degenerate oligonucleotide probe to locate the CS7.4 gene by Southern blot analysis. The gene was identified, cloned, sequenced, and designated as cspA (Goldstein et al., 1990). CspA is the founding member of the cold shock-like family of proteins (Csp family). The large majority of the members of the Csp family are prokaryotic homologues of CspA and were identified by amino acid sequence homology (see later). In E. coli K12, nine diVerent proteins are members of the Csp family (Fig. 2). The members were identified using diVerent techniques. Cold shock induction located cspA (Goldstein et al., 1990), whereas Southern blot analysis using cspA as a probe identified cspB and cspC (Lee et al., 1994). In a screen for a multicopy suppressor of a chromosome partitioning defect in mukB, cspC and cspE (Yamanaka et al., 1994) were identified. Sequencing of the upstream region of clpA identified an open reading frame that is homologous to cspA and was named cspD (Gottesman et al., 1990). Searching for genes induced at low temperatures isolated the promoter of cspG, and a computer search identified the cspG gene (Nakashima et al., 1996). The other homologues were identified by searching the complete sequence of the E. coli genome [cspF, cspH, and cspI (Yamanaka and Inouye, 1997)]. Only four of the nine homologues are induced upon cold shock. These are CspA, CspB, CspG, and CspI (Goldstein et al., 1990; Lee et al., 1994; Nakashima et al., 1996; Wang et al., 1999).

4. Folding Chromosomes in Bacteria

181

Figure 2 Alignment of the amino acid sequences from the Csp proteins of E. coli. Residues shown in blue are identical among the different homologues and those shown in both blue and bold type are identical among all 9 homologues. Red lines above the sequences indicate the position of the strands in the crystal structure of CspA (see text and Fig. 4) CspD is the least homologous to the other Csp proteins and contains an alanine´ and valine-rich 7 amino acid tail. (See Color Insert.)

A. CspA The cspA gene maps to 80.1 min on the E. coli chromosome (Fig. 3). It contains its own promoter and is not part of an operon. There are conflicting results on the regulation of cspA. One set of studies indicated that CspA is the major cold shock protein of E. coli (Goldstein et al., 1990; Jones et al., 1987) and is induced about 18-fold upon a temperature downshift from   37 C to 15 C (Lee et al., 1994). In these studies, the regulation of CspA was tightly controlled such that the protein was virtually undetectable at  37 C (Goldstein et al., 1990; Lee et al., 1994). Regulation of cspA occurred by increasing the transcription of the cspA gene, increasing the stability of the cspA mRNA at low temperatures (Tanabe et al., 1992), and negatively regulating transcription of cspA by CspE at 37 C (Bae et al., 1999). The mRNA of cspA contains a 159 base untranslated region (UTR) at the 50 end (Goldstein et al., 1990). Overproduction of just the first 143 bases of the UTR produced a derepression of cspA at low temperatures. This part of the UTR contains a ‘‘cold box’’ sequence (UGACGUACAGA) and forms a stem–loop structure. Overproduction of the UTR increased the cellular concentrations of CspA, as well as CspB and CspG, and the proteins remained at high concentrations for an extended period of time. If both 50 UTR and CspA are overexpressed from a pBR322 derivative (30–50 copies/cell), the concentration of CspA protein returns back to wild-type levels, implying that CspA regulates its own expression (Jiang et al., 1996). Deleting cspA causes an increase in the production of CspB and CspG when the temperature is decreased, suggesting that CspA might regulate cspB and cspG (Bae et al., 1997). Growth inhibition and cell death at low temperatures occur if truncated forms of cspA are overexpressed in a cspA deletion strain (Xia et al., 2001a). A second set of studies indicated that CspA is present at high concentrations during exponential growth at 37 C and induced 2.5-fold by cold shock

182

Trun and Johnston crcA 0/100

cspE crcB 14.1

cspA

cspD

80.1 19.9

cspH 20.6 cspG

35.3

cspI cspB cspF

41.0 cspC

Figure 3 Chromosomal positions of the nine csp genes of E. coli. Arrows indicate the direction of transcription of the genes. Only the cspE gene is in an operon; there is a single predicted promoter in front of cspE and crcB and both genes are required for the phenotypes associated with them. cspG and cspH are located next to each other but are transcribed in opposite orientation. cspB, cspF, and cspI are located close together and their presence in the genome may be the result of a duplication and a transposition event. These three homologues are only present in E. coli K12.

(Brandi et al., 1999). It was shown that the cspA transcript and protein product were induced during early exponential growth of E. coli K12 and remained detectable throughout the growth curve. Under these conditions, cspA transcription was stimulated by Fis and was inhibited by HNS (Brandi et al., 1999). Further experiments are needed to clarify the discrepancies between these two studies. 1. Functions of CspA CspA binds to single-stranded RNA (ssRNA) that is longer than 75 bases. This binding has two eVects. First, with CspA bound, ssRNA is more susceptible to degradation by RNases, presumably by preventing the formation of secondary structures that stabilize the RNA. Second, CspA binding to mRNA allows eYcient translation of that mRNA at low temperatures (Jiang et al., 1997). Using stalled transcription complexes from a template

4. Folding Chromosomes in Bacteria

183

containing a rho-independent terminator, the eVect of purified CspA was tested to determine if CspA acts as a transcription antiterminator in vitro (Bae et al., 2000). Lengths of the transcripts produced demonstrated that the read-through eYciency was increased in the presence of CspA (Bae et al., 2000). Another function of CspA is to regulate other cold shock-induced proteins. It has been shown that CspA acts as a transcriptional activator of two diVerent cold shock-induced genes, hns and gyrA (La Teana et al., 1991). A 110-bp segment of the hns promoter was used to identify the regulatory region that responds to cold induction and to determine if CspA binds to this region of hns (La Teana et al., 1991). This segment of the hns promoter contains the sequence CCAAT. Eukaryotic homologues of Csp proteins (see later) bind to the CCAAT sequence. To determine if the CCAAT box is important for regulation, a promoter of hns without the CCAAT sequence, from Proteins vulgaris, was fused to a promoter-less cat gene and tested for cold shock induction in E. coli. The P. vulgaris promoter fusion was induced upon temperature shift to a similar extent as an E. coli promoter fusion, indicating that the CCAAT consensus sequence is not required for cold shock induction (Brandi et al., 1994). CspA also regulates gyrA (Jones et al., 1992). The gyrA promoter contains three ATTGG sequences (the complement of the CCAAT consensus sequence). Spacing between ATTGG sequences aligns them on one face of the helix (ATTGG – 15 bp – ATTGG – 23 bp – ATTGG – 95 bp – ATG). When at least one of these sites is present in an oligonucleotide that matches the promoter region, a CspA–DNA complex is formed as seen by an electromobility shift assay (EMSA). When there are no ATTGG sites present in the promoter region, there is no complex formed and no shift in an EMSA. These results indicate that the ATTGG sequence is necessary for CspA binding to the gyrA promoter (Jones et al., 1992).

B. CspB The cspB gene was originally identified by Southern blot analysis of the E. coli chromosome using the cspA gene as a probe (Lee et al., 1994). It is located at 35.3 min on the E. coli chromosome (Fig. 3) and, like cspA, contains its own promoter and is not part of an operon. Production of a translational lacZ fusion to cspB allowed for the determination of its expression   patterns. Upon a shift of exponentially growing cells from 37 C to 15 C, the cspB–lacZ gene fusion was induced approximately 17-fold, indicating that CspB is also a cold shock protein (Lee et al., 1994). No in vivo functions for CspB are known.

184

Trun and Johnston

Systematic evolution of ligand exponential enrichment (SELEX) was used to determine if several of the Csp proteins exhibit any sequence specificity when binding to nucleic acids. SELEX involves alternating cycles of protein binding to random pools of RNA and amplification of the bound RNA. After a number of cycles, the remaining pool of RNA contains the preferred sequences for the particular protein. The RNA can then be reverse transcribed, cloned, and sequenced. The SELEX technique showed that CspB binds preferentially to sequences that contain a stretch of U (usually five nucleotides) followed by a C (such as UUUUUC) (Phadtare and Inouye, 1999). CspB binds to ssDNA that contains this consensus sequence and to the complementary sequence with a lesser aYnity. The apparent dissociation constant (Kd ¼ 1.26  105 M vs Kd ¼ 5.5  106 M) and the percentage of maximum binding (9% vs 28%) increase when comparing pre-SELEX probes and the consensus sequence probe, respectively (Phadtare and Inouye, 1999). CspB will bind to an 88 nucleotide ssDNA molecule containing its consensus sequence and to the complementary strand of this ssDNA. The dissociation constant for these sequences was similar to results determined for RNA, with the complementary ssDNA being comparable to pre-SELEX RNA probes. CspB also binds to a 45 nucleotide ssDNA molecule that contains three repeats of the consensus sequence. The aYnity of CspB for this molecule is increased when compared to the 88 nucleotide sequence, presumably because the consensus sequence was repeated three times (Phadtare and Inouye, 1999).

C. CspC cspC was originally identified by Southern blot using cspA as a probe (Lee et al., 1994). cspC is located at 41.0 min on the chromosome (Fig. 3) and is not part of an operon. The expression of cspC was determined using a translational fusion of cspC and lacZ. The expression of this fusion was found to be moderate (424 units of -galactosidase) at 37 C and remained unchanged after cold shock. cspC is expressed under normal growth conditions, including during lag, exponential, and stationary phases (Lee et al., 1994; Phadtare and Inouye, 2001). Overexpression of cspC increases the production of at least four proteins, UspA, OsmY, Dps, and RpoS, by increasing the stability of their mRNA (Phadtare and Inouye, 2001). CspC was functionally identified as a multicopy suppressor of a mutation in the mukB gene (Yamanaka et al., 1994). Point mutations in mukB have several phenotypes, including a 1000-fold increase in anucleate cells  and the inability to grow at temperatures above 30 C. Overexpression of cspC reverses both of these phenotypes. What is remarkable about the suppression is that CspC is a 6.9-kDa protein and MukB is a 177-kDa

4. Folding Chromosomes in Bacteria

185

protein and yet CspC can partially substitute for MukB (Yamanaka et al., 1994). Using SELEX, it was determined that CspC binds to the consensus sequence AGGGAGGGA (Phadtare and Inouye, 1999). The apparent dissociation constant (Kd ¼ 3.9  106 M vs Kd ¼ 2.5  106 M) and the percentage of maximum binding (11% vs 25%) increase when comparing pre-SELEX probes and the consensus sequence. CspC binds to an 88 nucleotide ssDNA probe that contains the consensus sequence and to the complementary sequence with a lesser aYnity, similar to CspB. However, CspC will not bind to a 45 nucleotide ssDNA molecule containing three tandem repeats of the consensus sequence (Phadtare and Inouye, 1999).

D. CspD The cspD gene was initially discovered by sequencing the upstream region of clpA (Gottesman et al., 1990). CspD maps to 19.9 min on the chromosome (Fig. 3) and is not part of an operon. A translational fusion of cspD to lacZ showed that expression of cspD–lacZ was suppressed about 15-fold upon cold shock (Lee et al., 1994). Induction of the fusion occurred at the transition between exponential growth and stationary phase (Yamanaka and Inouye, 1997). The induction of cspD is not dependent on the stationary phase sigma factor (rpoS), is slightly repressed (by about 40%) by HNS, and is inversely dependent on the growth rate (CspD is present in greater amounts in slower growing cells) (Yamanaka and Inouye, 1997). CspD has no known biological functions.

E. CspE CspE maps to 14.1 min on the E. coli chromosome and is the only one of the E. coli csp genes that is part of an operon (Fig. 3). cspE is followed by the crcB gene, and one predicted promoter resides in front of cspE and crcB. The expression pattern of cspE was determined using a translational  fusion of cspE to lacZ. The expression of cspE is constitutive at 37 C, and -galactosidase activity indicates that cspE is expressed at high levels ( 1400 units of -galactosidase). There is an 50% increase in -galactosidase activity from the cspE–lacZ fusion during the lag period and early exponential growth (Bae et al., 1999). In a strain deleted for cspA, the expression of cspE is increased 1.5-fold after cold shock (Bae et al., 1999). Using two-dimensional gel electrophoresis, proteins that are regulated by cspE were identified. A strain containing a wild-type copy of cspE in the chromosome was compared to a strain carrying a cspE deletion. In the

186

Trun and Johnston

absence of cspE, four proteins are increased and seven proteins are decreased in amount (Bae et al., 1999). CspE increases the production of UspA (universal stress protein), OsmY (osmotically inducible protein), Dps (DNA protection protein during starvation), and RpoS (the stationary phase sigma factor). One of the negatively regulated proteins is CspA (Bae et al., 1999). In the cspE deletion strain, five times more cspA mRNA is produced than in wild-type cells, yet the half-life of the cspA mRNA showed no significant increases from that of the wild-type strain. This implies that CspE regulates cspA by altering transcription and not by altering the stability of cspA mRNA (Bae et al., 1999). The alteration of transcription of cspA could be dependent on the cspA promoter or be attributed to elongation or termination of the mRNA. To determine if the cspA promoter is required for CspE repression, it was replaced with a constitutive promoter from the lpp gene and placed back into the E. coli chromosome at the normal position of cspA. The amount of CspA produced from the constitutive promoter was tested by two-dimensional gel electrophoresis in the presence and absence of CspE. In a cspE deletion strain, the amount of CspA from the fusion increased fourfold over the amount produced in the presence of CspE. This indicates that CspE represses the transcription of cspA independently of the promoter (Bae et al., 1999). Further experiments have shown that CspE increases a transcriptional pause at two diVerent positions proximal to the promoter of cspA. These sites are immediately downstream of the ‘‘cold box’’ sequence in the 50 UTR of cspA (Bae et al., 1999).

1. Chromosome Condensation Phenotypes of CspE CspE, like CspC, was originally identified as a multicopy suppressor of a point mutation in the E. coli SMC protein, mukB (Yamanaka et al., 1994). The overexpression of cspE can suppress both the production of anucleate cells and the temperature-sensitive colony formation phenotypes of the mukB106 mutation. (Yamanaka et al., 1994). To examine if the overexpression of cspE increases the amount of the mutant MukB produced, a mukB– lacZ fusion was examined. The overexpression of cspE did not increase the transcription of mukB, indicating that CspE suppresses mukB mutations by a diVerent mechanism (Yamanaka et al., 1994). Camphor vapors are lethal to cells because they decondense the chromosome (Hu et al., 1996). Overexpression of a chromosomal fragment of DNA containing the cspE gene confers resistance to camphor. In addition to cspE, this fragment also carried crcB and a third gene crcA that is located upstream of cspE and is expressed from its own promoter. When all three

4. Folding Chromosomes in Bacteria

187

genes are overexpressed, cells are 10,000-fold more resistant to camphor. Overexpression of this region confers no growth defects on cells and does not alter the amount of DNA found within the cell (Hu et al., 1996). Cells overexpressing any one of the three genes alone or crcA with cspE do not confer resistance to camphor. Overexpression of diVerent combinations of crcA, crcB, and cspE were tested for their ability to suppress diVerent mukB mutations (Hu et al., 1996). When all three genes are overexpressed, they can partially suppress the temperature-sensitive phenotypes associated with a deletion of mukB and with two diVerent point mutations in mukB. The phenotypes of camphor resistance and suppression of mukB mutations cannot be separated, suggesting that all three genes are required for both phenotypes. Microscopic observation of cells indicates that overexpression of crcA, crcB, and cspE prevents the nucleoids from being decondensed by camphor or the absence of mukB (Hu et al., 1996).

2. CspE Binds to RNA Photocross-linking RNA to nearby macromolecules in vitro identified at least nine diVerent proteins that interact with nascent RNA in an active transcription complex. Nine diVerent promoters were used in these experiments, and only a small number of the proteins cross-linked to the majority of the promoters. One of these proteins was CspE, which associated with seven of the nine promoters tested (Hanna and Liu, 1998). One of the promoters that CspE interacts with is lPR. This promoter was used to determine at which stage of transcription CspE associates with the complex. CspE will associate with the nascent RNA in the transcription complex when 10 nucleotides of RNA have been synthesized. If 88 nucleotides of nascent RNA have been synthesized, then CspE can no longer associate with the transcription complex. The same RNA that CspE associates within an active transcription complex was not a suitable substrate for CspE when the RNA is released from the complex (Hanna and Liu, 1998). CspE alters the eYciency of Q-mediated antitermination from the lPR promoter in vitro. The addition of CspE decreased the amount of antitermination at a concentration of 300 nM and completely abolished antitermination at 1 mM (Hanna and Liu, 1998). Using the SELEX technique, it was determined that CspE binds to RNA-containing sequences that are AU rich, especially doublets and triplets of A and U (Phadtare and Inouye, 1999). Single-stranded DNA containing AT consensus sequences was synthesized and used to determine that CspE will also bind to ssDNA containing AT sequences.

188

Trun and Johnston

3. CspE Binds to DNA Conflicting reports of CspE binding to double-stranded DNA (dsDNA) have been published (Phadtare and Inouye, 1999; Hanna and Liu, 1998). In studies where the AU sequence preference of CspE was determined, the authors also synthesized the dsDNA corresponding to the RNA bound by CspE. CspE was unable to bind to this dsDNA (Phadtare and Inouye, 1999). In contrast, in studies where CspE was found to be a part of active transcription complexes, CspE was shown to bind a 105-bp fragment from the lPR promoter (Hanna and Liu, 1998). In the presence of CspE, the dsDNA fragment from the lPR promoter was retarded on a polyacrylamide gel. There was not one finite retarded band, but a series of bands depending on the ratio of CspE to DNA. The more CspE, the more bands appeared on the gel (Hanna and Liu, 1998).

4. Antitermination by CspE After it was demonstrated that CspE interacts with nascent RNA chains in an active transcription complex (Hanna and Liu, 1998), the extent of the interaction was studied (Bae et al., 2000). These studies were carried out using purified CspE in an in vitro assay with stalled transcription complexes. The addition of CspE decreased the amount of transcription termination at a Rho-independent terminator (Bae et al., 2000). Antitermination could be the result of CspE interacting with the RNA polymerase, CspE interacting with elongation factors, or CspE interacting with terminator structures. CspE does not interact directly with RNA polymerase or with the C-terminal domain of the  subunit of RNA polymerase, a known binding site for transcription terminators (Bae et al., 2000). Rho-independent terminators contain two diVerent components: a stem–loop structure and a series of U residues (Uptain et al., 1997). Using a construct with a terminator missing the U residues, RNA polymerase paused at the stem–loop structure. In the presence of CspE, this pause was reduced significantly, indicating that CspE acts as a transcription antiterminator by preventing the formation of secondary structures in RNA (Bae et al., 2000).

F. CspG and CspI CspG maps to 20.6 min on the E. coli chromosome and is present as a single gene with its own promoter (Fig. 3). Studies looking for promoters that are induced by cold temperatures identified the promoter of the cspG gene (Nakashima et al., 1996). The expression of cspG was examined using

4. Folding Chromosomes in Bacteria

189

a cspG–lacZ gene fusion. cspG is induced seven-fold by coldshock, and the level of cspG expression is higher in cells that are grown at lower temperatures (Nakashima et al., 1996). An analysis of the cspG promoter showed that, like the promoter of cspA, the cspG promoter contains a large 50 UTR region (Nakashima et al., 1996; Tanabe et al., 1992). No biological functions for CspG have been identified. The cspI gene maps to 35.3 min on the E. coli chromosome (Fig. 3). It is not in an operon and it contains its own promoter. The cspI gene was identified by searching the complete sequence of the E. coli K12 genome for additional csp-related sequences (Yamanaka and Inouye, 1997). To determine the expression patterns of cspI, both transcriptional and translational fusions with lacZ were constructed. With both fusions, the amount of -galactosidase activity increased approximately eight-fold upon transi  tion from 37 C to 15 C. Primer extension experiments were carried out to determine the promoter region of cspI. The cspI transcript contains a long 50 UTR containing the ‘‘cold box’’ consensus sequence (Wang et al., 1999). No biological functions for CspI have been identified. G. CspF and CspH Both cspF and cspH were identified from BLAST searches of the E. coli genome (Yamanaka and Inouye, 1997). Their existence in the genome is known, but nothing about their expression or function has been determined.

IV. Relationships among Escherichia coli Csp Proteins When the amino acid sequences of the nine Csp proteins from E. coli are compared, several interesting relationships emerge (Fig. 1). CspA and CspB are 80% identical, CspC and CspE are 83% identical, CspF and CspH are 77% identical, and CspG and CspI are 79% identical. The Csp proteins seem to come in pairs. It is interesting to note that where it has been looked at, these pairs of Csp proteins are regulated similarly (A and B and G and I are induced by cold shock, C and E are constitutive) and, in the case of CspC and CspE, have similar biological functions. CspD is the least similar to any of the other CSP proteins and is regulated in a unique manner. When the number and amino acid sequences of the Csp proteins are compared among E. coli K12, two strains of E. coli O157, and Salmonella enterica serovar typhimurium LT2, the concept of pairs of Csp proteins is strengthened. Of these four very closely related bacteria, E. coli K12 contains nine homologues, whereas the other three genomes contain six homologues each. Both strains of E. coli O157 contain CspA, CspC, CspD,

190

Trun and Johnston

CspE, CspG, and CspH. S. enterica serovar typhimurium contains CspA, CspB, CspC, CspD, CspE, and CspH. The Salmonella CspB (stCspB) is 80% identical to E. coli K12 CspG (ecCspG) and 67% identical to E. coli K12 CspB (ecCspB). Because of this, we believe that Salmonella CspB is more likely to be a Salmonella CspG homologue. Only E. coli K12 contains the CspB, CspF, and CspI homologues. S. enterica serovar typhimurium and the two strains of E. coli O157 contain one Csp protein from the CspA/CspB pair, one from the CspG/CspI, one from the CspF/CspH, CspD, and both members of the CspC/CspE pair that are constitutively expressed. This diVerence in the number of homologues may be the result of a chromosomal duplication of three of the csp genes in E. coli K12 after it diverged from Salmonella and E. coli O157. When sequences of the nine csp’s of E. coli K12 and their positions in the genome are compared, an unusual relationship emerges. CspB is most identical to CspA (80%), CspF is most identical to CspH (77%), and CspI is most identical to CspG (79%). Locations on the chromosome show that cspH and cspG are near each other with opposing transcriptional orientations. CspF and CspI are also near each other with opposing transcriptional orientations; however, cspB is inserted between the two genes. Examining the sequence surrounding cspA identified remnants of a transposable element downstream of the gene. While the exact sequence of events is not known, it is reasonable to propose two events. The cspA gene was inserted between cspH and cspG. The cspH and cspG genes were duplicated and moved to the current location of cspF, cspB, and cspI.

V. Csp Proteins of Bacillus subtilis Bacillus subtilis contains three csp genes: cspB, cspC, and cspD. From BLAST searches, CspB is the most identical to the E. coli CspC (67%) and CspE (68%) proteins. bsCspD shows 64% identity to ecCspA and 60% identity to ecCspE. bsCspC is 72% identical to ecCspA. All three of the B. subtilis csp genes are induced by cold shock (Graumann et al., 1996). cspB and cspC are also induced by entry into stationary phase (Graumann and Marahiel, 1999). A deletion of any one csp increases the production of other Csp proteins (Graumann et al., 1996; 1997). In a cspB deletion, cells exposed to coldshock have a reduction in the induction of cold-induced proteins (Graumann et al., 1996). Deletion of any two of the csp results in a de creased growth rate at 15 C. The cspC/D deletion also had a slight increase (15–30 min) in the length of the lag phase. cspB/C and cspB/D double deletions also have an increase in cell lysis after entry into the stationary phase (Graumann et al., 1997). A triple deletion of chromosomal csp genes could be created only when cspB was provided on a plasmid. This constructed

4. Folding Chromosomes in Bacteria

191

strain could not be cured of the cspB carrying plasmid (Graumann et al., 1997). This suggests that removal of all three Csp proteins from B. subtilis is a lethal event. It also indicates that these Csp proteins can substitute for each other. A similar situation exists in E. coli, where only when a cell contains chromosomal deletions of cspA, cspB, cspE, and cspG are phenotypes observed. In this quadruple deletion, the cells are cold sensitive and have a defect in cell division. Overexpression of all of the E. coli csp genes except cspD will complement the defects of the quadruple deletion strain (Xia et al., 2001b).

VI. The Crystal Structures of CspA and Related Proteins A. Structure of CspA The structure of CspA has been determined by both solution nuclear magnetic resonance (NMR) (Newkirk et al., 1994) and X-ray crystallography (Schindelin et al., 1994) with both techniques generating the same general structure (Fig. 4A). CspA is a very compact molecule organized into five antiparallel strands ( 1– 5). A barrel is formed by the strands with 1– 3 found on one side of the protein and 4– 5 on the other side of the protein (Newkirk et al., 1994; Schindelin et al., 1994). CspA contains two RNA-binding sequences: RNP1 and RNP2. These sequences are consensus sequences from the ribonucleoprotein (RNP) motif class of proteins. RNP1 is an octamer with a consensus sequence of K/R GF/ YG/A FVXF/Y and RNP2 is a hexamer with a consensus sequence of L F /I / YV/I G/KM/GL (Nagai et al., 1990). Sequences of the RNP-like motifs in CspA are 75% identical to RNP1 and 34% identical to the RNP2 consensus sequences. RNP1 resides on 1 and RNP2 resides on 2. The NMR structure of CspA was compared to the structure of a CspA– oligonucleotide complex, where the oligonucleotide was designed to match the first 24 bases of the CspA transcript (Newkirk et al., 1994). Monitoring the chemical shifts of the resonance frequencies can identify the residues of the protein involved in complex formation. In well-structured proteins, only residues on the interface of the protein–DNA complex will have a resonance frequency shift. Significant shifts indicate that a particular residue has direct involvement with DNA (Dotsch, 2001). Binding of the oligonucleotide to CspA significantly alters the amide backbone resonance of a number of residues: Lys10, Ala14, Lys16, Gly17, Gly19, Asp24, Val30, Phe31, Val32, His33, Ser35, Ala36, Ile37, Gln38, Asn39, Tyr42, Lys43, Ser44, Leu45, Val51, Glu56, Ala59, Lys60, Gly61, Ala63, and Ser69 (Newkirk et al., 1994). These diVerences are seen on the face of the protein where the RNP sites are located and in the loop region between 3 and 4, indicating that

192

Trun and Johnston

Figure 4 Crystal structures of the E. coli CspA protein (A), the human YB-1 cold shock domain (B), the B. subtilis CspB protein (C), a dimer of the B. subtilis CspB (D), and the S. acidocaldarius Sac7d protein (E). All of the proteins contain a barrel composed of either five or six strands. Where it has been determined, the barrel binds to nucleic acids in a nonspecific manner. Structures shown in yellow are sheets, structures shown in pink are  helices, and structures shown in blue are turns. All of the structures are from Molecules ‘R Us (http://molbio.info.nih.Gov/cgi-bin/pdb). (See Color Insert.)

4. Folding Chromosomes in Bacteria

193

the RNP sites could be functional. This area also contains a number of aromatic amino acids that could potentially interact with ssDNA or RNA, although no diVerences were identified in the aromatic amino acid residue peaks (Newkirk et al., 1994). A number of crystal structures from diVerent proteins are superimposable on the structure of CspA from E. coli K12 (ecCspA). These include the coldshock domain (CSD) of the human YB-1 protein, Sac7D from Sulfolobus acidocaldarius, and CspB from B. subtilis (bsCspB). In the case of all of these proteins, the crystal structures can be superimposed based on the folding patterns of the five strands, but not in the connecting loops, the N terminus, or the C terminus.

B. Sequence Homologues of the Csp Family in Eukaryotes—Y-Box Factors including Human YB-1 Y-box factors are a family of eukaryotic proteins that have multiple proposed functions requiring interactions with DNA and/or RNA. One of the main functions of some members of this family is to act as transcription factors that bind to a consensus sequence called the Y box (CCAAT; for review see Ladomery, 1997; WolVe, 1994; WolVe et al., 1992). Y-box factors have three distinct domains: an N-terminal domain, a CSD, and a C-terminal domain. The N terminus is a variable domain whose function is unknown. The C-terminal domain contains alternating acidic and basic regions of approximately 30 residues (Murray et al., 1992). This region is responsible for protein–protein interactions and promotes RNA binding (Bouvet et al., 1995). The CSD is conserved among all Y-box factors, has 43% identity with CspA (Fig. 5), and is involved in nucleic acid binding, presumably through RNP1 and RNP2 motifs (Bouvet et al., 1995). The NMR structure from the CSD of the human Y-box protein, YB-1, was determined (Fig. 4B). The CSD is organized into a barrel formed from five antiparallel strands, similar to the structure of CspA (Kloks et al., 2002). The CSD was also used to determine the amino acids responsible for binding ssDNA. Resonance shifts were observed mainly in amino acids between the 1– 2 and the 4– 5 loops (Kloks et al., 2002). Residues implicated in DNA binding are Trp15, Phe24, Phe35, and His37. Lys14 and Tyr22 are also believed to be involved with binding DNA based on their location in the protein and their side chains (Kloks et al., 2002). It was also determined that the CSD of YB-1 binds preferentially to pyrimidine-rich sequences, but does not bind with any sequence specificity. The intact YB1 protein does bind specifically to the Y-box sequence. Because of this discrepancy, it has been proposed that the CSD of YB-1 is responsible for nonspecific binding and that other regions of the protein provide the

194

Trun and Johnston

Figure 5 Alignment of the primary amino acid sequences of the E. coli CspA protein (ECCspA), the human YB-1 cold shock domain (YB-1CSD), the B. subtilis CspB protein (BSCspB), and the S. acidocaldarius Sac7d protein (Sac7D). Residues forming sheets are shown in red and those forming  helices are in blue. The crystal structure of Sac7d indicates that this protein is circularly permuted at the primary amino acid level when compared to Csp sequences. This leads to the third strand of Sac7d corresponding to the first strand from the Csp proteins. Locations of the RNP1 and RNP2 motifs in ecCspA, YB-1, and bsCspB that interact with RNA are boxed. Consensus sequences for the RNP motifs are shown above the sequences. Either of the amino acids listed can be present at each position. (See Color Insert.)

sequence dependence interactions (Kloks et al., 2002). This interpretation is also in agreement with binding studies using the E. coli CspA protein that show it does not bind specifically to the Y-box sequence. C. Structural Homologues of the Csp Family—The Sac7d Protein Sac7d is a 7-kDa, 66 amino acid protein from Sulfolobus acidocaldarius, an archaea that has an optimal growth temperature range between 70 and  80 C. Sac7d binds to double-stranded DNA and increases the melting tem perature by over 30 C (McAfee et al., 1995). Sac7d and CspA are 40% identical at the amino acid level (Fig. 5). The crystal structures of Sac7d (Edmondson et al., 1995) and of a Sac7d–DNA complex (Robinson et al., 1998) have been determined (Fig. 4E). The solution structure of Sac7d revealed a compact protein consisting of five strands and one  helix. Two of the strands are organized into an antiparallel ribbon and the other three strands from an antiparallel sheet. The  helix is found at the C terminus (Edmondson et al., 1995). Even though there is low sequence identity at the amino acid level between Sac7d and CspA, they have very similar protein structures. The crystal structure of Sac7d complexed to dsDNA duplexes demonstrated that Sac7d binds to the minor groove of DNA with no sequence specificity (Robinson et al., 1998). The binding widens the minor groove and  introduces a kink in the DNA, forming a 72 bend in the process. The bend is mediated by the intercalation of the Val26 and Met29 amino acids of Sac7d. The intercalating amino acid residues are found near the beginning of the 4 strand (Val26 is between 3- 4 and Met29 is the second amino

4. Folding Chromosomes in Bacteria

195

acid of 4). The triple-stranded sheet of Sac7d is the region of the protein that makes contact with both sides of the DNA double helix through hydrogen bonds and salt bridges. On one side of the helix, Lys7, Y8, Lys9, Lys28, and Lys48 make strong interactions with the phosphate backbone of DNA. The strongly interacting amino acids are found in or near 1 and 4 (Lys7 and Y8 are in 1, Lys9 is between 1 and 2, Lys28 is in 4, and Lys48 is between 5 and the helix). On the other side of the helix, Lys22, Trp24, and Lys39 form weak interactions with the phosphate backbone. The weak interacting amino acids are found in 3 and 5 (Lys22 and Trp24 in 3 and Lys39 in 5) (Robinson et al., 1998).

D. Dimerization of a Csp Protein—Structure of the Bacillus subtilis CspB Protein Both the solution NMR and the crystal structure of the B. subtilis CspB (bsCspB) were determined with both techniques giving the same results (Schindelin et al., 1993; Schnuchel et al., 1993). The structure is five antiparallel strands connected by turns and loops to form a barrel (Fig. 4C). Within this structure, three of the strands ( 1– 3) form a sheet containing a DNA-binding motif. The sheet also contains an RNP1 domain (Schindelin et al., 1993; Schnuchel et al., 1993). CspB adopts an L shape  because the sheet sits at 90 from the other two strands ( 4– 5) (Schindelin et al., 1993). Determination of the crystal structure revealed that CspB forms a dimer (Fig. 4D). The crystallization of CspB was performed under two diVerent conditions to establish that dimerization was not an artifact of the crystallization process. Dimers are created by six diVerent hydrogen bonds between the 4 strands of two CspB molecules. These two molecules are orientated antiparallel to each other with their N-terminus juxtaposed to the end of the 4 strand of the opposite molecule (Schindelin et al., 1993). It is believed that this type of multimerization would be impossible for ecCspA because of the extended N-terminus (three additional amino acids in ecCspA) and sequence diVerences in the 4 strand. All four of these crystal structures contain the same basic barrel that is used for binding nucleic acids in a sequence-independent manner. Each protein decorates this barrel in diVerent fashions, presumably to add specificity to the functions that it carries out. In YB-1, the barrel is embedded in a much larger protein that must interact with a variety of other molecules in the cell. In this case, the basic structure of the cold shock domain has not changed. The specificity needed by the protein is added to this domain. The bsCspB has the unique ability to dimerize. If the N-terminal three amino

196

Trun and Johnston

g

CspE Homologues Vibrio cholerae 4 4 1 1 0 0 HU and IHF Homologues Xylella fastidiosa 2 3 1 1 0 1 Fis Homologues Pseudomonas aeruginosa 6 5 1 0 0 1 HNS Homologues Haemophilus influenzae 1 3 1 1 0 0 / 1 Sac7d Homologues Pasteurella multocida 2 3 1 1 0 0 / 1 SMC Homologues Buchnera sp 2 3 1 1 0 0 Salmonella typhimurium 6 4 1 2 0 0 / 1 Escherichia coli K12 9 4 1 2 0 0 / 1 EDL933 6 4 1 2 0 0 / 1 Escherichia coli O157:H7

b

Neisseria meningitidis

a

Rickettsia prowazekii 1 3 1 0 0 0 Mesorhizobium loti 10 3 1 0 0 1 Caulobacter crescentus 4 4 1 0 0 1

Proteobacteria

VT2-Sakai 6 4 1 2 0 0 / 1 MC58 1 3 1 0 0 1 Z2491 1 3 1 0 0 1

26695 0 1 0 0 0 0 J99 0 1 0 0 0 0 jejuni 0 1 0 0 0 0 3 16 1 0 0 0

Heliobacter pylori

e

Campylobacter Magnetococcus

Bacillales

Bacillus halodurans 1 2 0 0 0 1 Bacillus subtilis 3 2 0 0 0 1 COL 3 1 0 0 0 1 Staphylococcus aureus Mu50 3 1 0 0 0 1 N315 3 1 0 0 0 1

B a c t e r i a

Firmicutes

Lactobacillales

Mollicutes

Spirochaetales Cyanobacteria Actinobacteria Bacteroidetes Chlamydiae

Enterococcus faecalis 6 1 0 0 0 1 Streptococcus pyrogenes 1 1 0 0 0 1 TIGR4 0 1 0 0 0 1 Streptococcus pneumoniae R6 0 1 0 0 0 1 Lactococcus lactis 2 2 0 0 0 1 Mycoplasma genitalium 0 0 0 0 0 1 Mycoplasma pneumoniae 0 1 0 0 0 1 Mycoplasma pulmonis 0 1 0 0 0 1 Ureaplasma urealyticum 0 1 0 0 0 0

Borrelia burgdorferi 0 1 0 0 0 1 Treponema pallidum Nicols 0 1 0 0 0 1 Synechocystis sp 0 1 0 0 0 1 Mycobacterium leprae 2 1 0 Mycobacterium tuberculosis Porphyromonas gingivalis Chlamydia pneumoniae Chlamydia trachomatis

Deinococcus-Thermus Thermotogales Aquificae

A r c h a e a

0 0 1 CDC1551 2 1 0 0 0 1 H37Rv 2 1 0 0 0 1 0 2 0 0 0 0 AR39 0 1 0 0 0 0 CWL029 0 1 0 0 0 0 J138 0 1 0 0 0 0 0 1 0 0 0 0

Deinococcus radiodurans 1 1 0 0 0 0 Thermotoga martima 2 1 0 0 0 1 Aquifex aeolicus 1 2 0 0 0 1

Crenarchaeota

Aeropyrum pernix 0 0 0 0 0 0 Sulfolobus acidocaldarius 0 0 0 0 3 0

Euryarchaeota

Methanococcus jannaschii 0 0 0 0 0 1 Methanobacterium thermoautotrophicum 0 0 0 0 0 0 Pyrococcus abyssi 0 0 0 0 0 1 Pyrococcus horikoshii 0 0 0 0 0 1 Archaeoglobus fulgidus 0 0 0 0 0 0 Thermoplasma acidophilum 0 1 0 0 0 1 Thermoplasma volcanium 0 1 0 0 0 1 Halobacterium sp 2 0 0 0 0 0

Figure 6 Distribution of homologues of the E. coli CspE, E. coli HU/IHF, E. coli Fis, E. coli HNS, S. acidocaldarius Sac7d, and B. subtilis SMC proteins among archaea and bacteria. Numbers after each species indicate the number of homologues present in its genome, and the color indicates the identity of the protein being used as the bait. Amino acid sequences for each protein were used in BLAST searches against the complete genome of each organism

4. Folding Chromosomes in Bacteria

197

acids of ecCspA do indeed prevent it from dimerizing, then it follows that the E. coli CspB, CspF, CspG, CspH, and CspI proteins should also not dimerize. ecCspD does not have this N-terminal extension and could dimerize, and ecCspC and ecCspE have only a two amino acid extension. The primary amino acid sequence of Sac7d is circularly permuted when compared to Csp proteins and it contains a large  helix at the C terminus. The  helix does not play a role in DNA binding. Rather, strands 3, 4, and 5 in the barrel of Sac7d fit snuggly into the minor groove of DNA. The RNP1 and RNP2 consensus sequences found in the Csp proteins are in strands 2 and 3, which correspond to strands 3 and 4 in Sac7d. What is equally interesting in the Csp proteins is a cleft in the protein between strands 1–2–3–4 and 5– 6. This cleft is wide enough in the CspA crystal structure to accommodate the phosphate backbone of a single strand of nucleic acid and it is deep enough to accommodate bases from this single strand. Along one side of the cleft are five aromatic amino acids that could interact with the bases, and on the other side of the surface of the cleft is a lysine residue that could interact with the phosphate backbone. Given that this cleft seems well designed for binding single-stranded nucleic acids, it would be very surprising if it were not functional. Thus, even though Csp proteins are very small, they contain at least one and potentially two nucleic acid-binding domains.

VII. Distribution of Small DNA-Binding Proteins in Archaea and Bacteria Using the amino acid sequences of the E. coli HNS, HU, IHF, Fis, and CspE proteins and Sac7d from S. acidocaldarius, 47 bacterial genomes and 10 archaeal genomes were searched for homologues using the comprehensive microbial resource (CMR) BLAST program from The Institute for Genome Research (TIGR) (Fig. 6). If these proteins are widely used for the condensation of chromosomes lacking histones, then they should be present in many diVerent species. Searches using the amino acid sequence of the HU and IHF proteins of E. coli yielded the same homologues in all of the genomes examined. This is because these two proteins are very similar to each other. Homologues of these proteins were identified in all but 2 of the bacterial genomes searched, Mycoplasma genitalium and Mycoplasma pneumoniae, and were found in 2 of the 10 archaeal genomes, Thermoplasma acidophilum and Thermoplasma volcanium. All of the other 45 bacterial genomes contained between 1 and listed. For SMC homologues, the B. subtilis SMC was used as the bait; however, because the B. subtilis SMC does not identify the E. coli SMC MukB, we also used MukB as bait. Results from the MukB searches are listed after the slashes (in the -proteobacteria). (See Color Insert.)

198

Trun and Johnston

16 homologues of HU/IHF . The majority of species containing multiple homologues occur in the proteobacteria. Using the amino acid sequence of Fis to search the genomes revealed that Fis homologues are only found in the proteobacteria and are only in single copy in any given genome. Homologues of Fis were not observed in any of the archaeal genomes. Fis homologues split among the proteobacteria. All of the and subdivisions contain Fis homologues and all but one genome from the  subdivision contain Fis homologues, whereas the e and  subdivisions do not contain Fis homologues. Homologues of HNS were only observed in the subdivision of proteobacteria, with only one genome (Pseudomonas aeruginosa) not containing a homologue. The number of homologues varied between one and two, with closely related species containing the same number. Sequence homologues of Sac7d were only identified in one of the archaeal genomes, S. acidocaldarius, which contained three homologues. This is the bacterium in which Sac7d was originally identified. No sequence homologues were identified in bacterial genomes; however, structural homologues (Csp proteins) exist in these genomes. The majority of the bacterial genomes and one archaeal genome, Halobacterium sp., contain CspE homologues. Of the proteobacteria genomes searched, 84% of them contain a homologue to CspE, with only the –e division having no homologues. The majority of gram-positive bacteria (65%) contain a homologue of CspE.

VIII. Conclusions Chromosome folding is a complex process that requires several diVerent levels of condensation to arrive at a folded, functional DNA molecule. Eukaryotes and some archaea use nucleosomes as the first level of condensation. Other archaea and the bacteria do not contain the histone proteins that make up nucleosomes and thus they must use a diVerent mechanism(s). Several small nucleic acid-binding proteins have been shown to condense chromosomes in bacteria, and these proteins may be what the bacteria and some archaea use in place of nucleosomes. The nucleosomes condense DNA approximately six-fold, whereas bacteria need a mechanism that will condense the DNA about two- to three fold. From the distribution of the homologues of the histones and small nucleic acid-binding proteins an interesting relationship occurs. Genomes seem to contain histone protein homologues or small nucleic acid-binding protein homologues but not both. These two condensing mechanisms may be mutually exclusive. The Csp and HU/IHF proteins are the most widely distributed among the bacteria and archaea, with Fis being found only in proteobacteria

4. Folding Chromosomes in Bacteria

199

and HNS only in -proteobacteria. In contrast, SMC proteins are found in genomes from all three domains. Bacteria and archaea, when they have a homologue, only carry a single SMC homologue per genome, whereas eukaryotes usually have multiple SMC proteins. No single small nucleic acid-binding protein has been shown to be responsible for all of the condensation of the chromosome in bacteria. Rather, the strategy that seems to have been adopted is to use many proteins in diVerent combinations. Many of these small proteins can substitute for each other in functional assays and so their ability to substitute for each other in condensation is not surprising. Most of these proteins, in addition to binding nucleic acids, also bend the nucleic acid. These two features may be all that is needed to condense the chromosome the additional two- to three fold necessary for it to fit inside the cell. If this is true, then any protein that is abundant enough and binds and bends DNA could contribute to condensation. Every small nucleic acid-binding protein that has been shown to play a role in condensation is also involved in other cellular processes, including regulating other genes. This has made it diYcult to determine if their roles in condensation are direct or indirect. In vitro studies of the interactions of the proteins with DNA are usually needed to clarify the role(s) of the protein in condensation and to determine how much protein is needed to condense the DNA. In the case of Fis, one dimer binding every 80 bp is enough to condense the DNA, whereas for HU, one dimer can bind every 6–12 bp. While these numbers dramatically aVect the calculations of how much condensation each protein could carry out, there are no experiments that directly address how much condensation each protein carries out on an intact chromosome. Therefore, the calculations lead to the least amount of condensation each protein could carry out. Two members of the Csp family of proteins in E. coli, CspC and CspE, have been implicated in chromosome condensation using genetic assays. Overexpression of cspE in vivo suppresses a deletion of mukB and prevents the chromosome from being decondensed by camphor. In vitro, CspE binds to dsDNA, ssDNA, and RNA. Other members of the Csp family bend the DNA in addition to binding to it. These characteristics are reminiscent of other small nucleic acid-binding proteins involved in chromosome condensation. In order for these proteins to be eVective in condensing a chromosome, they must be very abundant. Initial reports of the concentration of CspE in cells indicate that this characteristic also applies to CspE (Bae et al., 1999). It is thought to be present at 50,000 copies per cell. The Csp family of proteins is unusual in bacteria, where duplicate genes are rare. In the 31 genomes that contain a Csp homologue, 71% contain more than one. From the two cases where it has been examined, Csp proteins have overlapping functions: Deletion of one csp gene has no detectable

200

Trun and Johnston

phenotypes but the cells die when many or all of the genes are deleted. This group of proteins was initially discovered by their induction when cells are   diluted and shifted from 37 C to 15 C. Because of this phenotype, they were named cold shock proteins. It is now clear that not all Csp proteins are induced by cold shock, rather they are always present in cells. This increases the possibilities for this unique and interesting family of proteins.

References Afflerbach, H., Schroder, O., and Wagner, R. (1998). Mol Microbiol. 28, 641–653. Aki, T., and Adhya, S. (1997). EMBO J. 16, 3666–3674. Atlung, T., and Hansen, F. G. (2002). J. Bacteriol. 184, 1843–1850. Azam, T., Iwata, A., Nishimura, A., Ueda, S., and Ishihama, A. (1999). J. Bacteriol. 181, 6361–6370. Bae, W., Jones, P. G., and Inouye, M. (1997). J. Bacteriol. 179, 7081–7088. Bae, W., Phadtare, S., Severinov, K., and Inouye, M. (1999). Mol. Microbiol. 31, 1429–1411. Bae, W., Xia, B., Inouye, M., and Severinov, K. (2000). Proc. Natl. Acad. Sci. USA 97, 7784–7789. Bonnefoy, E., Almeida, A., and Rouviere-Yaniv, J. (1989). Proc. Natl. Acad. Sci. USA 86, 7691–7695. Bouvet, P., Matsumoto, K., and Wolffe, A. P. (1995). J. Biol. Chem. 270, 28297–28303. Brandi, A., Pon, C., and Gualerzi, C. O. (1994). Biochime 76, 1090–1098. Brandi, A., Spurio, R., Gualerzi, C. O., and Pon, C. (1999). EMBO J. 18, 1653–1659. Dame, R. T., Wyman, C., and Goosen, N. (2000). Nucleic Acids Res. 28, 3504–3510. Dotsch, V. (2001). Methods Enzymol. 339, 343–357. Edmondson, S. P., Qiu, L., and Shriver, J. W. (1995). Biochemistry 34, 13289–13304. Goldstein, J., Pollitt, S., and Inouye, M. (1990). Proc. Natl. Acad. Sci. USA 87, 283–287. Gottesman, S., Clark, W. P., and Maurizi, M. R. (1990). J. Biol. Chem. 265, 7886–7893. Graumann, P. L., and Marahiel, M. A. (1999). Arch. Microbiol. 171, 135–138. Graumann, P. L., Schroder, K., Schmid, R., and Marahiel, M. A. (1996). J. Bacteriol. 178, 4611–4619. Graumann, P. L., Wendrich, T. M., Weber, M. H. W., Schroder, K., and Marahiel, M. A. (1997). Mol. Microbiol. 25, 741–756. Hanna, M. M., and Liu, K. (1998). J. Mol. Biol. 282, 227–239. Hirano, T. (2002). Genes Dev. 16, 399–414. Hu, K. H., Liu, E., Dean, K., Gingras, M., DeGraff, W., and Trun, N. J. (1996). Genetics 143, 1521–1532. Hulton, C. S. J., Seirafi, A., Hinton, J. C. D., Sidebotham, J. M., Waddell, L., Pavitt, G. D., Owen-Hughes, T., Spassky, A., Bic, H., and Higgins, C. F. (1990). Cell 63, 631–642. Jessberger, R. (2002). Nature Rev. Mol. Cell Biol. 10, 767–768. Jiang, W., Fang, L., and Inouye, M. (1996). J. Bacteriol. 178, 4919–4925. Jiang, W., Hou, Y., and Inouye, M. (1997). J. Biol. Chem. 272, 196–202. Johansson, J., Eriksson, S., Sonden, B., Wai, S. N., and Uhlin, B. E. (2001). J. Bacteriol. 183, 2343–2347. Jones, P. G., Krah, R., Tafuri, S., and Wolffe, A. P. (1992). J. Bacteriol. 174, 5798–5802. Jones, P. G., VanBogelen, R. A., and Neidhardt, F. C. (1987). J. Bacteriol. 169, 2092–2095. Kamashev, D., and Rouviere-Yaniv, J. (2000). EMBO J. 19, 6527–6535.

4. Folding Chromosomes in Bacteria

201

Kloks, C. P. A. M., Spronk, C. A. E. M., Lasonder, E., Hoffmann, A., Vuister, G. W., Grzesiek, S., and Hilbers, C. W. (2002). J. Mol. Biol. 316, 317–326. Ladomery, M. (1997). BioEssays 19, 903–909. Laine, B., Sautiere, P., Biserte, G., Cohen-Solal, M., Gros, F., and Rouviere-Yaniv, J. (1978). FEBS Lett. 89, 116–120. La Teana, A., Brandi, A., Falconi, M., Spurio, R., Pon, C., and Gualerzi, C. O. (1991). Proc. Natl. Acad. Sci. USA 88, 10907–10911. Lavoie, B. D., Shaw, G. S., Millner, A., and Chaconas, G. (1996). Cell 85, 761–771. Lee, S. J., Xie, A., Jiang, W., Etchegaray, J.-P., Jones, P. G., and Inouye, M. (1994). Mol. Microbiol. 11, 833–839. McAfee, J. G., Edmondson, S. P., Datta, P. K., Shriver, J. W., and Gupta, R. (1995). Biochemistry 34, 10063–10077. Murphy, L. D., and Zimmerman, S. B. (1997). J. Struct. Biol. 119, 321–335. Murray, M. T., Schiller, D. L., and Franke, W. W. (1992). Proc. Natl. Acad. Sci. USA 89, 11–15. Nagai, K., Oubridge, C., Jessen, T. H., Li, J., and Evans, P. R. (1990). Nature 348, 515–520. Nakashima, K., Kanamaru, K., Mizuno, T., and Horikoshi, K. (1996). J. Bacteriol. 178, 2994–2998. Newkirk, K., Feng, W., Jiang, W., Tejero, R., Emerson, S. D., Inouye, M., and Montelione, G. T. (1994). Proc. Natl. Acad. Sci. USA 91, 5114–5118. O’Gara, J., and Dorman, C. J. (2000). Mol. Microbiol. 36, 457–466. Phadtare, S., and Inouye, M. (1999). Mol. Microbiol. 33, 1004–1014. Phadtare, S., and Inouye, M. (2001). J. Bacteriol. 183, 1205–1214. Pinson, V., Takahashi, M., and Rouviere-Yaniv, J. (1999). J. Mol. Biol. 287, 485–497. Robinson, H., Gao, Y.- G., McCrary, B. S., Edmondson, S. P., Shriver, J. W., and Wang, A. H.-J. (1998). Nature 392, 202–205. Sandman, K., Grayling, R. A., Dobrinski, B., Lurz, R., and Reeve, J. N. (1994). Proc. Natl. Acad. Sci. USA 91, 12624–12628. Sandman, K., Krzycki, J. A., Dobrinski, B., Lurz, R., and Reeve, J. N. (1990). Proc. Natl. Acad. Sci. USA 87, 5788–5791. Schindelin, H., Jiang, W., Inouye, M., and Heinemann, U. (1994). Proc. Natl. Acad. Sci. USA 91, 5119–5123. Schindelin, H., Marahiel, M. A., and Heinemann, U. (1993). Nature 364, 164–168. Schneider, R., Lurz, R., Luder, G., Tolksdorf, C., Travers, A., and Muskhelishvili, G. (2001). Nucleic Acids Res. 29, 5107–5114. Schnuchel, A., Wiltscheck, R., Czisch, M., Herrler, M., Willimsky, G., Graumann, P. L., Marahiel, M. A., and Holak, T. A. (1993). Nature 364, 169–171. Tanabe, H., Goldstein, J., Yang, M., and Inouye, M. (1992). J. Bacteriol. 174, 3867–3873. Uptain, S. M., Kane, C. M., and Chamberlin, M. J. (1997). Annu Rev Biochem. 66, 117–172. Wang, N., Yamanaka, K., and Inouye, M. (1999). J. Bacteriol. 181, 1603–1609. White-Ziegler, C. A., Villapakkam, A., Ronaszeki, K., and Young, S. (2000). J. Bacteriol. 182, 6391–6400. Wolffe, A. P. (1994). BioEssays 16, 245–251. Wolffe, A. P., Tafuri, S., Ranjan, M., and Familari, M. (1992). New Biol. 4, 290–298. Xia, B., Etchegaray, J.-P., and Inouye, M. (2001a). J. Biol. Chem. 276, 35581–35588. Xia, B., Ke, H., and Inouye, M. (2001b). Mol. Microbiol. 40, 179–188. Yamanaka, K., and Inouye, M. (1997). J. Bacteriol. 179, 5126–5130. Yamanaka, K., Mitani, T., Ogura, T., Niki, H., and Hiraga, S. (1994). Mol. Microbiol. 13, 301–312. Yang, C.-C., and Nash, H. A. (1989). Cell 57, 869–880. Yuan, H. S., Finkel, S. E., Feng, J.-A., Kaczor-Grzeskowiak, M., Johnson, R. C., and Dickerson, R. E. (1991). Proc. Natl. Acad. Sci. USA 88, 9558–9562.

ThisPageIntentionallyLeftBlank

Index

A ‘‘A’’ element, 11 Activation incompatible with assembly, 34 in initiation phase, 33 of initiation complex, 17–38 staggered program of, 38 Activin response element (ARE), 151 Activins, 144 nodals in family of, 148 Alkaline gel electrophoresis, 4 detection of replication intermediates in origin firing, 27 linking gene silencing and late origin firing, 43 Amphibian cells lampbrush chromosomes, 94–96 study of chromosome folding in, 77 Anaphase, 77 bending of chromosomes during, 111 elasticity of chromosomes during, 96 U shape, 111 Antitermination, by CspE protein, 188 Antivin/lefty genes, 150, 154 acting distance from cell in which synthesized, 157 role in left-right asymmetry in dorsal diencephalon, 163–164 Archaea, distribution of small DNA-binding proteins in, 197–198 Architecture, of eukaryote chromosomes, 79–94 ARS1, 40 ARS301, 18 ARS305, 43, 53 ARS501, 40, 43, 45 ARS1413, 40 ARS consensus sequence (ACS), 11–14 Aspiration technique, overestimating chromosome elasticity, 99 Assays, for origin activation alkaline gel electrophoresis for nascent DNA strands, 6–7

autonomously replicating sequence assay, 3 copy number replication profiles, 8–9 density transfer analysis of chromosomal replication profiles, 8 density transfer method, 7 genomic locations of activation-replication complexes, 9 on genomic scale, 8–9 single molecule analysis by molecular combing, 7–8 strengths and limitations of, 9–10 two-dimensional agarose gel electrophoresis, 3–6 Assembled chromosomes, 103 Assembly incompatible with activation, 34 in initiation phase, 33 of initiation complex, 17–38 of prereplicative complex, role of Noc3p in, 21–22 Asymmetry, role of nodal signals in establishing, 162–163 Autonomously replicating sequence (ARS) assays, 3 advantage and disadvantages of, 9 identifying sequence organization of replication origins, 11 Autonomously replicating sequence (ARS) elements deciphering code of, 15–16 mutational analysis of, 13 predicting locations of, 15 sequence organization of, 12 Avogadro’s number, 83 Axial gene, in nodal signaling, 152

B ‘‘B’’ elements, 14–15 B-form contour length, 83 B-type cyclins, and execution of the temporal program, 46–47 203

204 Bacillus subtilis, 174 Csp proteins in, 190–191 CspB protein in, 195–197 Bacteria Csp family of proteins unusual in, 199 distribution of small DNA-binding proteins in, 197–198 Bacterial chromosomes, extracting and studying, 133 Bending elasticity, 78 of chromosomes extracted from cells, 107–109 of mitotic chromosomes, 104–111, 125 of newt chromosomes, 106–107 related to Young modulus, 85 thermally excited, 83 Bending fluctuations dynamics of, in mitotic newt chromosomes, 115–116 of chromosomes extracted from cells, 107–109 of chromosomes in vivo, 109 of mitotic chromosomes, 111 of newt chromosomes, 106–107 of in vitro-assembled Xenopus chromatids, 110 Binding sites, for transcription factor AbfIp, 14–15 Biochemical-micromechanical chromosome studies combined micromechanical-chemical experiments, 118–119 eVect of micrococcal nuclease on mitotic chromosomes, 122–123 eVect of restriction enzymes with four-base specificity, 123–125 shifts in ionic conditions, 119–122 whole genome extraction experiments, 117–118 Bioformatics tools, applying to genomic sequencing, 59 Biophysical studies, of chromosome structure, 78 Bipartite ORC-binding site, 15 Bone morphogenic proteins, 144 Bonnie and clyde mutants, endoderm precursors reduced in, 157 Brachyury gene, 152 BrdU labeling, of nascent DNA, 10 Bubble arc, 4

Index C Calreticulin gene, 152 Camphor vapors, decondensing chromosomes, 186 Casanova transcription factor, 157 Cdc6p, 22–23 Cdc45p, 30 Cdc7p protein, 47 CDK-dependent regulation, of prereplicative complex assembly, 35 Cdt1p, 22–23 CDK-independent mechanism for regulating, 36 Cell adhesion molecules, 152 Cell cycle dynamic structure changes during, 77 in newt cells, 77 Cell fates determined via concentration grading of nodal activity from vegetal to animal pole, 156 nodal signals inducing mesoendodermal, 153 Cell-free chromosome assembly systems, 90–91 Cell fusion experiments, 33 Cell-labeling experiments, in studies of nodal signaling, 154 Cellular concentration, inversely related to plasmid loss rate, 25 Centromeric regions, late activation of, 42 Cesium chloride gradient centrifugation, limitations of, 10 Checkpoint pathways, 49 conserved across evolution, 49–50 monitoring replication, 56 multiple and redundant, 58 Checkpoint systems, 31 discovery of, 49 mutation abrogating inhibition of origin firing by, 38 Chinese hamster ovary nuclei, factors influencing late origin firing time in, 46 ChIP-to-chip method, 9 advantages and limitations of, 10 exploring protein-DNA interactions, 31–32 role in determining origin spacing and location, 16 Chromatin, orientational ordering of, 132 Chromatin compaction delaying origin activation time, 44

Index induced by ionic interactions, 118 Chromatin domains, kilobase size of, 89 Chromatin fiber as component of chromosomes, 79–82 as telomeric structures, 132 comparative elasticity of, 126–127 disentangled by topoisomerase II, 91–93 entropic elasticity regime of, 86 force-extension properties of, 86 irreversible extension of, 87 in vitro assembly of, 87 linear elastic response of, 87 micromechanics of, 86–88 physical arrangement of, 93 Chromatin fragments, size analysis of, 130 Chromatin immunoprecipitation, 9 Chromatin loop isolation studies, 76 Chromatin organization, related to gene expression during interphase, 77 Chromatin states, cycling of origins of replication between, 17–18 Chromatin structure eVect on origin activation time, 42–44 higher-order, 80 Chromosomal DNA replication, complexity of, 1 Chromosomal replication profiles based on copy number, 8–9 based on density transfer analysis, 8 Chromosomal titin, 93 Chromosome-banding techniques, high resolution, 98 Chromosome disentanglement, 93–94 Chromosome elasticity, due to bulk of cross section of chromosome, 105 Chromosome-folding proteins, 90–91 Chromosome stiVness, 94 Chromosome structure at very large scales, 88–93 biophysical studies of, 78 irreversible changes in, 121–122 driven by DNA-cutting enzymes, 117 modifying chemically and biochemically, 79 network organization of, 128–129 of in vivo vs. in vitro-assembled chromosomes, 109–110, 126 preliminary model of mitotic, 79 real-time monitoring of changes to, 118–119 reversible changes driven by ionic conditions, 117

205 three-dimensional microscopy study of, 89 topoisomerase II critical for establishing, 92–93 Chromosome volume, increased with initial extension, 104 Chromosome XIV, 40 late determinants on, 41 Chromosomes architecture and components of, 79–94 bending elasticity of, 104–111 dynamic structure during mitosis, 77 in vitro assembly of, 90–91 physical properties of, 79 softness of, 78 Clb-CDK, cell cycle-dependent states of, 34 Clb6p protein, and execution of the temporal program, 46 Cold shock induction, 200 CspA as regulator of other induced proteins, 183 identifying Csp proteins via, 180–181 Compaction factor, of chromatin fiber, 82 Compound ARS, 13 Compound origins, 13 Condensation eVect of shifts in ionic conditions on, 119–122 during mitosis, 77 multiple factors responsible for DNA folding in, 173–174 nucleosomes as first level in eukaryotes, 198 role of Fis in, 177–178 role of nucleic acid-binding proteins in, 175 studying mechanism of, 93–94 Copy number, chromosomal replication profiles based on, 8–9 Cricket spermatocytes, chromosome studies of, 96–97 Cross-linking agents, 29 Csp protein family, 174, 178–180 appearing in pairs, 189 chromosomal positions of, 182 dimerization of, 195–197 in Bacillus subtilis, 190–191 initial identification of, 180–181 relationships among members of, 189–190 sequence homologues of, 193–194 structural homologues of, 194–195 unusual in bacteria, 199

206 CspA gene, 180–181 alignment of primary amino acid sequences in, 194 functions of, 182–183 structure of, 191–193 CspB gene, 183–184 CspC gene, 184–185 CspD gene, 185 CspE gene, 185 as transcription antiterminator, 188 binding to RNA, 187 chromosome condensation phenotypes of, 186–187 homologues in bacterial and archaeal genomes, 198 CspF gene, 189 CspG gene, 188–189 CspH gene, 189 CspI gene, 188–189 Cyclin-dependent kinase (CDK), 27–28 role in origin firing, 32–33 role in regulating assembly of prereplicative complex, 35 Cyclops (Cyc) nodal protein, 144, 145 aVecting gene transcription locally, 156 influencing choice between notochord and prechordal plate cell fate, 155 involved in preliminary floor plate patterning, 159 mutants missing ventral nervous system, 158 mutation and one-eyed pinhead, 149–150 Cytokinesis, 77 Cytoskeletal proteins, 152 studies of disruption of, 132 D Data manipulation, required to determine origin spacing and location, 16 Dbf4-dependent kinase (DDK), 27–28 role in origin firing, 23–33 Dbf4p protein, 47 Density transfer analysis chromosomal replication profiles based on, 8 of replication time, 58 of temporal program of origin activation, 39 revealing time of firing of origins, 10, 27 Density transfer method, for determining time of replication, 7

Index Direct force measurement, quanitifying intermolecular interactions by, 82 DNA biophysical characterization of, 82 CspE binding to, 188 stretching, 86 twisting and supercoiling, 86, 91 unwinding in origin activation process, 27 DNA-binding proteins, in archaea and bacteria, 197–198 DNA chip hybridization, 9 DNA connections, between mitotic chromosomes, 79 DNA cutters, future experiments in, 130 DNA-cutting experiments, 128 DNA damage, S phase response to, 50–57 DNA inversion, role of Fis in, 177–178 DNA-protein interactions, 83 DNA replication elements necessary for regulating, 11 role of nucleic acid-binding proteins in, 175 DNA synthesis, beginning in S phase, 2 Docking site, Mcm10p serving as, 25–26 Domains, of histone-like nucleoid structuring (HNS) protein, 175 Dorsal diencephalon, left-right asymmetry in, 163–164 Double-stranded DNAs (dsDNAs), 76 cross-sectional radius of, 85 elastic rod nature of, 85 micromechanics of, 82–86 persistence length of, 82 sources of tension in extension of, 83 springlike nature of, 83 structural stiVness of, 82 Double Y arc, 4 Downregulation, of genes in cyc mutants, 152 Dpb11p, role in origin firing, 29 Drosophila, 131 bending fluctuations of chromosomes in vivo, 109 regulation of transcription in origin activation of, 47–48 Duplex unwinding element (DUE), 14 E EYciency determinants of origin, 36–38 of origin firing, 27

207

Index Elasticity homogeneity of chromosomal, 104 of chromatin fiber, 126–127 of chromosomes in grasshopper spermatocytes, 96–97 of dsDNA vs. chromatin fiber, 85, 88 of mitotic chromosomes, 76, 78, 96–102, 126–127 relative among diVerent regions of chromosomes, 130 Electron microscopy in studies of mitotic chromosome structure, 88 in studies of origin spacing and location, 16 use in observing replication origins, 3 Electrostatic component, of nucleosome interactions, 80 Elongation Cdt1p and Cdc6p unneeded for, 23 in process of origin firing, 27 Empirical force law, 83 Endoderm patterning via nodal signals, 153, 157–158 specification by nodal proteins, 143–144 Equilibrium elastic stress, 113 Escherichia coli chromosome folding in, 174–175 initial identification of Csp proteins in, 180–189 relationships among Csp proteins in, 189–190 small nucleic acid-binding proteins involved in chromosome folding, 175–178 Eukaryote chromosomes, architecture and components of, 79–84 Eukaryotes, replicators identified in, 11 Expression pattern, of zebrafish nodal-related genes, 144–145 Extensibility and increased chromosome volume, 104 of mitotic chromosomes, 96–102 F Factor for inversion stimulation (Fis) protein, 174, 177–178 homologues found only in proteobacteria, 198 Fates, of origins, 34–35 Faust/Gata 5 transcription factor, 157 Fiber autoradiography studies of origin spacing and location, 16

use in observing replication origins, 3 Firing time, determinants of, 39 Flexibility, of chromatin structures, 80 Flh gene, overlap with notochord cell fate domain, 155 Floor plate Calreticulin gene expressed in, 152 role of nodal signals in patterning, 158–160 role of Smad proteins in specifying, 150–151 specification mediated by inhibitory interactions, 159 Fluorescence microscopy techniques, visualizing single molecules with, 7–8 Fluorescent in situ hybridization (FISH) analysis, 44 of loci along metaphase chromosomes, 90 of whole chromosomes, 89 Fluorophores, 132 Folding multiple factors responsible for chromosomal, 173–174 of mitotic chromosomes, 76 Footprinting, 17 Force-extension response in newt chromosomes, 101 of prometaphase chromosomes, 99 reversible, 103 Force jumps, irreversible, 93 Force regimes, of chromatin fibers, 86–87 Fork direction analysis, 3, 10 Fork elongation, 27 Fork migration and termination, 38 deducible via density transfer analysis, 52 Fork pauses, not exhibited by inactive mutant ARS, 26 Fork rate, determined by nucleotide sequence replicated, 38 Fork termination. See Fork migration and termination Forkhead activin signal transducer-1 (Fast-1), 151 Four-base specificity, of restriction enzymes, 123–125 Frog chromosome folding, 77 G Gastrulation floor plate specification beginning during, 159 role of nodal signals in, 153

208 Geminin, regulating Cdt1p activity, 36 Gene expression, related to chromatin organization during interphase, 77 Gene regulation role of Fis in, 177–178 role of nucleic acid-binding proteins in, 175 Genetic markers, real-space distances between, 89–90 Genome organization for replication, 11–17 replication profile of, 8 spatial organization of, 59 Genome-wide ChIP-to-chip experiment, 37 Genomic replication profiles, strengths and limitations of, 10 Ghost chromosomes, 102 Goosecoid gene, 152 and prechordal plate cell fate domain, 155 Grasshopper spermatocytes, 131 chromosome elasticity in, 96–97, 125 Green fluorescent protein (GFP), in chromosome structural studies, 90 GyrA promoter, regulated by CspA, 183 H Halobacterium sp., containing CspE homologues, 198 Hatching glands Calreticulin gene expressed in, 152 nodal signals specifying, 151–153, 155 Head muscle, nodal signals specifying, 155 Helical stability, reduced in duplex unwinding elements (DUE), 14 Histone-like nucleoid structuring (HNS) protein, 174, 175–176 homologues of, 198 Histone proteins, 79 archaeal, 174 role in nucleosomes and the condensation process, 174 Homologous recombination, role of nucleic acid-binding proteins in, 175 Homomultimer, Mcm-related protein assembling into, 23–24 HU protein, 174, 176–177 Human Y-box protein YB-1, 193–194 Hydroxyurea eVects of, 7, 50–55 impeding incoming forks, 38

Index S phase response to, 50–57 sensors responding to, 55 I IHF protein, 177 Immunodepletion experiments, in Xenopus egg extracts, 25 Immunoprecipiation experiments, 21 In vitro-assembled chromatids, 103, 126 Initiation two steps of, 33 unified model for, 30 Initiation complexes, 2 assembly and activation of, 17–38 initiation phase of, 29–31 origin firing in, 27–33 prereplication complex in, 17–26 transition to preinitiation complex, 28–29 Initiation event, 3 Initiation phase, of origin firing, 29–31 Initiator, role in origin structure, 11 Interchromosomal regions, 89 Internal viscosity, of mitotic chromosomes, 112 Interphase extracting whole genomes during, 117–118 gene expression related to chromatin organization during, 77 Invariant positions, mutations in, 13 Ionic conditions aVecting condensation/decondensation of mitotic chromosomes, 119–122 insensitivity of dsDNA structure to, 82 sensitivity of chromatin fiber to, 80 shift experiments in, 127–128 Isotopes, labeling with dense, 9 K Kinetochore hinging behavior at, 108 measuring elasticity of, 130 L Labeling, with dense isotopes, 9 Lagging strand synthesis, 30

Index Lampbrush chromosomes, 94–96 Left-right asymmetry in dorsal diencephalon, 163–164 role of nodal signals in patterning, 143–144, 146, 161–164 Light microscopy, limitations of chromosome structure studies, 89 Lim1 gene, 152 Loop domains, of lampbrush chromosomes, 96 M Mcm complex, 14, 30 activated by postreplicative complex, 17–18 as component of the prereplicative complex, 23–25 cellular concentration and plasmid loss rate, 25 DDK-dependent change in state of, 33 in prereplicative complex, 19 needed for fork progression, 58 Mcm10p, 25–26 Mechanical properties, of mitotic chromosomes, 78 Mec1p, 54 Meiotic chromosomes, extensibility and elasticity of, 96 Meiotic synapsis, role of chromosome stiVness in, 94 Meselson-Stahl density transfer experiment, 7 Mesendoderm regionalization via nodal signals, 154–158 specification via nodal signals, 153–154 Mesoderm patterning via nodal signals, 153–158 specification by nodal proteins, 143–144 Metabolic enzymes, 152 Metaphase chromosome folding during, 77 elasticity of chromosomes during, 96 Metaphase-to-anaphase transition, blocking of, 57 Methanothermus fervidus, 174 Methyl methanesulfonate, eVects on S phase, 50–55 Mezzo transcription factor, 157 Microarray-based methods, 59

209 of assaying chromosomal replication profiles, 9 Microarray hybridization, 9 Micrococcal nuclease, eVect on mitotic chromosomes, 122–123 Micromanipulation experiments, 79 in study of chromosome structure, 76 Micromechanical-chemical experiments, 118–119 Micromechanical studies, of mitotic chromosomes, 93–94 Microneedles in grasshopper spermatocyte experiments, 96 in lampbrush chromosome experiments, 95–96 in whole genome extraction experiments, 117–118 Mitotic chromosomes combined biochemical-micromechanical studies of, 116–125 combined micromechanical-chemical experiments in, 118–119 disassembled by cutting DNA, 117 eVect of ionic conditions on condensation/ decondensation, 119–122 eVect of micrococcal nuclease on, 122–123 eVect of restriction enzymes with four-base specificity on, 123–125 elasticity of, 126–127 expected bending flexibility and fluctuations of, 106–107 force-extension response in prometaphase, 99 physical properties of, 125–126 rationale for micromechanical studies of, 93–94 slow stress relaxation of, 112–115 stretching range of, 76 structural-biological studies of, 88–89 viscoelasticity of, 112–116 whole-genome extraction experiments in, 117–118 Mixer transcription factor, 157 MMS, eVects on replication, 51–52 Molecular combing, 59 in determining origin eYciency, 16 single molecule analysis by, 7–8 strengths and limitations of, 10 Molecular mass, of dsDNA, 80 Molecule end-to-end extension, 83

210 Mouse nodal signals, 145–146 identified via gastrulation and mesoderm formation, 153 MukB gene, 178 CspC as multicopy supressor of mutations in, 184 in Escherichia coli, 174–175 Mutagenesis, 11 Mutations, eVect on origin initiation, 9 N Nascent DNA strands alkaline gel electrophoresis assay for, 6–7 BrdU labeling of, 10 Network model, of mitotic chromosomes, 128–129 Neural tube, role of nodal proteins in specifying formation of, 143–144, 146 Newt chromosomes, 131 chromosome stretching experiments in, 100 expected bending flexibility and fluctuations of, 106–107 eye lens epithelial tissue culture line, 102 folding of, 77 force-extension data for, 101 slow stress response of, 116 stretching and elasticity of, 95 Noc3p, role and function in prereplicative complex, 21–22 Nodal antagonists, 150 Nodal proteins, 143–144 structural overview in zebrafish, 144 Nodal signaling mutants, in zebrafish, 145–146, 147 Nodal signaling pathway, 146–153 cellular changes induced by, 164–165 direction of future studies in, 164–165 one-eyed pinhead protein and, 149–150 schematic diagram of, 148 Nodal signals aVecting expression of mezzo gene, 157 antagonized by antivin/lefty genes, 154 determining cell fates, 154 downstream targets of, 151–153 in zebrafish, 144–146 patterning mesoderm and endoderm, 153–158 role in dorsal diencephalon asymmetry, 163–164

Index role in establishing body plan of vertebrate embryos, 143 role in patterning floor plate, 158–160 role in patterning left-right axis, 161–164 role in patterning ventral nervous system, 158–161 role in specifying ventral brain, 160–161 role in visceral asymmetry, 162–163 Nonaxial mesendoderm, diVerentiation via nodal signals, 156–157 Notochord formation via nodal signals, 155–156 not required for floor plate formation, 158–159 Notophthalmus viridescens, 77 chromosomal stretching and elasticity in, 95, 98 Nuclear assembly, of mitotic chromosomes, 91 Nuclear envelope, biochemical methods of opening, 133 Nuclear matrix, in chromosome organization, 90 Nuclease protection assays, 17. See also Footprinting Nucleic acid, binding lampbrush chromosomes, 96 Nucleic acid-binding proteins implicated in chromosome folding of E. coli, 175–178 role in chromosome condensation and folding, 174 Nucleoskeleton, 90 Nucleosome-like structures (NLS), 174 Nucleosome-nucleosome interactions, 80 Nucleosomes, 79–80 O OV-center origin, 4 One-eyed pinhead (Oep) protein, 149–150 Optical microscopy, in studies of mitotic chromosome structure, 88 ORC proteins, in prereplicative complex, 19 Origin activation assays for, 2–10 determinants of, 39–40 sequence elements aVecting, 40–42 temporal program of, 2, 38–49, 47–49 two waves of, 39 Origin activation time

Index aVected by localization within the nucleus, 44–45 chromatin compaction delaying, 44 eVect of chromatin structure and modulation on, 42–44 position eVects on, 41 sequences advancing, 42 Origin choice, determinants of, 36–38 Origin eYciency, determinants of, 36–38 Origin firing CDK and DDK requirements in, 32–33 factors establishing time of, 45–46 Initiation phase of, 29–31 mechanism of, 27 postinitiation phase of, 31–32 preventing reinitiation in, 33–36 role of Sld2p and Dpb11p in, 29 studies of obligate order of, 59 transition to preinitiation complex, 28–29 Origin initiations, cell-to-cell variation in order of, 10 Origin interference, 37–38 Origin recognition complex (ORC), 13–14 determining binding site for, 14 functions in prereplicative complex, 19–21 role in silencing the prereplicative complex, 18 Origin spacing and location, 16–17 Origins of replication, 1 capable of supporting maintenance of a plasmid, 3 cycling between two chromatin states in vivo, 17–18 early vs. late, 39 structure of, 11–16 three fates of, 34–35 P Passive replication, 4 Persistence, of nascent strands, 7 Persistence length, of double-stranded DNA, 82–83 Pharyngeal endoderm, nodal signals specifying, 155 Phenotypes, in nodal signaling mutants, 146 Phosphorylation, of ORC subunits, 35 Physarum, temporal program of origin activation in, 47 Plasmid loss rate, inversely related to cellular concentration, 25

211 Plasmid maintenance, ability of autonomic replication sequence to support, 9 Poisson ratio, 100, 102 Polarization microscopy, 132 Polymer-like flexibility, as property of chromatin fiber, 80 Polymerases, loading in origin activation process, 27 Posterior ventral hypothalamus, retained by nodal signaling, 160 Postinitiation phase, of origin firing, 31–32 Postreplicative complex, 17–18 Prechordal plate formation via nodal signals, 155–156 role of one-eyed pinhead in specifying, 149 Preinitiation complex, transition to, 28–29 Prereplicative complex, 18 assembly of, 21 cdt1p and Cdc6p in, 22–23 components of, 19–26 formation of, 18–26 in Xenopus, 24 Mcm genes in, 23–26 Noc3p in, 21–22 Primer RNA, synthesis in origin activation process, 27 Primitive streak, 153 Prometaphase, 77, 78 force-extension response of chromosomes during, 99 internal viscosity of newt chromosomes during, 113–114 Prophase, chromosome folding during, 77 Proproteins, in zebrafish nodal signaling, 144 Proteases, future experiments in, 130 Protein-binding sites, represented as sequence logos, 13 Protein digestion, disrupting nucleosomestacking interactions, 118 Protein-DNA complexes, 9, 17, 86 Protein-protein associations, 17 Protein scaVold model historical use of, 88 invalidity of, 79 Proteins able to form replication-initiation complexes, 9 role in mitotic chromosome organization, 118

212 Pseudomonas aeruginosa, 198 Pulse labeling, for identifying replicating DNA, 7

R Rad53p, 54, 55 association with chromatin assembly factor Asf1p, 57 Random walk fluctuations, 106 Reconstituted chromosomes, 131 elasticity of, 76 Reinitiation, preventing in origin firing, 33–36 Replication detecting uninterrupted, 57–59 determining time of, 7 dynamic aspects of, 2 transcription interfering with, 17 Replication factor C, 27 Replication forks assembly of, in S phase, 2 examination by electron microscopy, 53 monitoring of active, 58 proteins incorporated into, 31 Replication-initiation complexes, genomic locations of, 9 Replication intermediates, 53 Replication origins excluded from transcribed regions, 16–17 identification of specific sequences for, 3 Replication profiles, 8–9 Replication program, monitoring, 49–59 Replicative helicase, 23 activation in origin firing process, 27 Replicator, role in origin structure, 11 Response pathway, in face of DNA damage, 55–57 Restriction enzyme digestion, 90 ability to disintegrate mitotic chromosomes, 123–125 future experiments in, 130–131 Restriction sites, HNS blocking cleavage of, 178 Ribonucleotide reductase (RNR), genes coding for, 56 RNA, CspE binding to, 187 RNA-binding sequences, of CspA gene, 191 RNA polymerase, 175 RNases, future experiments in, 130 Rod bending, thermally excited, 105

Index S S-CDK, governing switch from assembly to activation, 34 S phase cell monitoring aspects of, 58 centromeric regions replicated in, 42 DNA synthesis beginning in, 2 monitoring completion of, 48–49 response to DNA damage and hydroxurea, 50–57 S phase checkpoint, 50 Saccharmomyces cerevisiae, 2 entering S phase in presence of MMS, 51 Sac7d protein, 194–195 Salt concentration, eVects of shifts in, 79 Salt dialysis assembly, of chromatin fiber, 87 Schizosaccharomyces pombe, 15, 30 temporal program of origin activation in, 47 Sequence elements, aVecting origin activation time, 40–42 Sequence logos, representing protein-binding sites as, 13 Shh gene, 152 expression promoted by nodal signaling in neural tissue, 160 Sic1p, 34 Signal transduction pathway members, 152 Silent origins ARSs, 37 forming apparently functional prereplicative complexes, 37 undetected by ChIP-to-chip method, 10 Single molecule analysis, via molecular combing, 7–8 Single-stranded RNA (ssRNA), binding of CspA to, 182 Sir proteins, targeting of, 43 Sir3p protein, 42 Six-base recognition sequence enzymes, 124 Sld2p, role in origin firing, 29 Sld3p, 31 Smad proteins, 146–147 and fast transcription factors, 150–151 SMC protein family, 92 critical to organizing mitotic chromosomes, 91 role in condensing chromosomes, 174 Softness, of chromatin fiber, 81 Southern blot analysis, identifying Csp proteins via, 180–181

213

Index Spatiotemporal patterns, of Cyclops and Squint activity, 157 Squint (Sqt) nodal protein, 144 aVecting gene transcription over distances, 156 gastrulation defects due to deregulation of, 154 influencing choice between notochord and prechordal plate cell fates, 155 Stationary phase, 177–178 StiVness, of chromosomes, 94 StpA protein, 175 Stress relaxation dynamics in chromosomes, 78 measurement of dynamics in news chromosomes, 114 slow in mitotic chromosomes, 112–115, 128–129 Stress response, reversible, 112 Stretching elasticity, 78 after removal of mitotic chromosomes from cells, 97–102 irreversible, 103 of lampbrush chromosomes, 94–96 of mitotic chromosomes, 96–102, 125 of in vitro assembled chromosomes, 103 Stretching range, 78 of mitotic chromosomes, 76 Structure, of origins of replication, 11–16 Supranucleosomal organization, 80 SV40, 27

Time of replication, density transfer method for determining, 7 Tip fluctuations, 116 Titin, as elastic restoring element of sarcomeres, 93 Topoisomerase II, 91–93 observation of DNA strand exchange by, 86 Transcription, 91 CspA acting as antiterminator of, 183 evidence of the processing nature of, 95 interfering with replication, 17 regulation of, in temporal program of origin activation, 47–48 zygotic in zebrafish nodal-related genes, 144–145 Transcription factors, 152 Smad proteins as, 146–147, 150–151 Transcriptional silencing, 9, 43 close to ARS305, 54 Transforming growth factor beta, 144 subfamilies and receptors, 146–149 Transmembrane receptor kinases, 146–147 Transposition, role of nucleic acid-binding proteins in, 175 Two-dimensional agarose gel electrophoresis in identifying Csp proteins in E. coli, 180 in identifying replication origins, 3–6 limitations of, 9–10 monitoring timing of origin activation, 27, 52

T

U

Teleost fish, asymmetry in pineal complex of, 164 Telomeres, aVecting origin activation time, 40 Telophase, 77 Temporal program conserved through evolution, 47 execution of, 46–47 of origin activation, 38–49 physiological relevance of, 47–49 setting up and reading, 39–47 Termination, 4 Thermal-bending fluctuations, in mitotic chromosomes, 105–106 Three-dimensional microscopy, study of chromosome structure and dynamics, 89–90

Universal stress protein (UspA), production increased by CspE, 186 Unreplicated chromatin, cell monitoring of, 58 Upregulated genes, in nodal signaling, 152 V Ventral brain, role of nodal signals in patterning, 160–161 Ventral midline, Shh gene involved in, 152 Ventral nervous system, role of nodal signals in patterning, 158 Ventral neural tube, 146 role of nodal signals in forming, 144 Ventral neurectoderm, role of one-eyed pinhead in specifying, 149

214 Vertebrate embryos, role of nodal signals in establishing body plan of, 143 Visceral organs, role of nodal signals in establishing asymmetry of, 162–163 Viscoelasticity, of mitotic chromosomes, 112–116 Viscous stress, in mitotic chromosomes, 112

W Whole-genome extraction experiments, 117–118 Whole-genome technologies, 59 Wild-type rDNA, eYciency of, 36–37

X X-ray crystallography, in study of nucleosome structure, 80 Xenopus chromatids, 131 bending fluctuations of in vitro-assembled, 110 Xenopus egg extracts, 77 and monitoring of active replication forks, 58 ARS code in, 15–16 CDK and DDK requirements in, 32 elasticity in reconstituted, 76 immunodepletion experiments in, 25–26 in vitro assembly of chromosomes, 90–91, 103 study of origin firing time in, 46 Xenopus embryo initiations in, 16

Index role of Cerberus protein in Nodal signaling, 150 Y Y arc, 4 Y-box factors, 193–194 Yeast dynamics of chromosome replication in, 1–59 genome sequence of, 8–9 replication origins excluded from transcribed regions in, 17 Saccharomyces cerevisiae, 2 Schizosaccharomyces pombe, 15 Yeast artificial chromosome (YAC), 57–58 Yolk syncytial layer (YSL), in zebrafish, 145 Young modulus, 125 defined, 85 of mitotic newt chromosomes, 109 of in vitro mitotic chromosomes, 98–99 Z Zebrafish cyc;sqt double mutants, 153 embryonic shield formation in, 153 left-right asymmetry in dorsal diencephalon of, 163–164 nodal signals in, 144–146, 148 required for left-sided gene expression, 163

Contents of Previous Volumes

Volume 47 1 Early Events of Somitogenesis in Higher Vertebrates: Allocation of Precursor Cells during Gastrulation and the Organization of a Moristic Pattern in the Paraxial Mesoderm Patrick P. L. Tam, Devorah Goldman, Anne Camus, and Gary C. Shoenwolf

2 Retrospective Tracing of the Developmental Lineage of the Mouse Myotome Sophie Eloy-Trinquet, Luc Mathis, and Jean-Franc¸ois Nicolas

3 Segmentation of the Paraxial Mesoderm and Vertebrate Somitogenesis Olivier Pourqule´

4 Segmentation: A View from the Border Claudio D. Stern and Daniel Vasiliauskas

5 Genetic Regulation of Somite Formation Alan Rawls, Jeanne Wilson-Rawls, and Eric N. Olsen

6 Hox Genes and the Global Patterning of the Somitic Mesoderm Ann Campbell Burke

7 The Origin and Morphogenesis of Amphibian Somites Ray Keller

8 Somitogenesis in Zebrafish Scott A. Halley and Christiana Nu¨sslain-Volhard

9 Rostrocaudal Differences within the Somites Confer Segmental Pattern to Trunk Neural Crest Migration Marianne Bronner-Fraser

215

216

Contents of Previous Volumes

Volume 48 1 Evolution and Development of Distinct Cell Lineages Derived from Somites Beate Brand-Saberi and Bodo Christ

2 Duality of Molecular Signaling Involved in Vertebral Chondrogenesis Anne-He´le`ne Monsoro-Burq and Nicole Le Douarin

3 Sclerotome Induction and Differentiation Jennifer L. Docker

4 Genetics of Muscle Determination and Development Hans-Henning Arnold and Thomas Braun

5 Multiple Tissue Interactions and Signal Transduction Pathways Control Somite Myogenesis Anne-Gae¨lle Borycki and Charles P. Emerson, Jr.

6 The Birth of Muscle Progenitor Cells in the Mouse: Spatiotemporal Considerations Shahragim Tajbakhsh and Margaret Buckingham

7 Mouse–Chick Chimera: An Experimental System for Study of Somite Development Josiane Fontaine-Pe´rus

8 Transcriptional Regulation during Somitogenesis Dennis Summerbell and Peter W. J. Rigby

9 Determination and Morphogenesis in Myogenic Progenitor Cells: An Experimental Embryological Approach Charles P. Ordahl, Brian A. Williams, and Wilfred Denetclaw

Volume 49 1 The Centrosome and Parthenogenesis Thomas Ku¨ntziger and Michel Bornens

2 g-Tubulin Berl R. Oakley

Contents of Previous Volumes

217

3 g-Tubulin Complexes and Their Role in Microtubule Nucleation Ruwanthi N. Gunawardane, Sofia B. Lizarraga, Christiane Wiese, Andrew Wilde, and Yixian Zheng

4 g-Tubulin of Budding Yeast Jackie Vogel and Michael Snyder

5 The Spindle Pole Body of Saccharomyces cerevisiae: Architecture and Assembly of the Core Components Susan E. Francis and Trisha N. Davis

6 The Microtubule Organizing Centers of Schizosaccharomyces pombe Iain M. Hagan and Janni Petersen

7 Comparative Structural, Molecular, and Functional Aspects of the Dictyostelium discoideum Centrosome Ralph Gra¨f, Nicole Brusis, Christine Daunderer, Ursula Euteneuer, Andrea Hestermann, Manfred Schliwa, and Masahiro Ueda

8 Are There Nucleic Acids in the Centrosome? Wallace F. Marshall and Joel L. Rosenbaum

9 Basal Bodies and Centrioles: Their Function and Structure Andrea M. Preble, Thomas M. Giddings, Jr., and Susan K. Dutcher

10 Centriole Duplication and Maturation in Animal Cells B. M. H. Lange, A. J. Faragher, P. March, and K. Gull

11 Centrosome Replication in Somatic Cells: The Significance of the G1 Phase Ron Balczon

12 The Coordination of Centrosome Reproduction with Nuclear Events during the Cell Cycle Greenfield Sluder and Edward H. Hinchcliffe

13 Regulating Centrosomes by Protein Phosphorylation Andrew M. Fry, Thibault Mayor, and Erich A. Nigg

14 The Role of the Centrosome in the Development of Malignant Tumors Wilma L. Lingle and Jeffrey L. Salisbury

15 The Centrosome-Associated Aurora/lpl-like Kinase Family T. M. Goepfert and B. R. Brinkley

218

Contents of Previous Volumes

16 Centrosome Reduction during Mammalian Spermiogenesis G. Manandhar, C. Simerly, and G. Schatten

17 The Centrosome of the Early C. elegans Embryo: Inheritance, Assembly, Replication, and Developmental Roles Kevin F. O’Connell

18 The Centrosome in Drosophila Oocyte Development Timothy L. Megraw and Thomas C. Kaufman

19 The Centrosome in Early Drosophila Embryogenesis W. F. Rothwell and W. Sullivan

20 Centrosome Maturation Robert E. Palazzo, Jacalyn M. Vogel, Bradley J. Schnackenberg, Dawn R. Hull, and Xingyong Wu

Volume 50 1 Patterning the Early Sea Urchin Embryo Charles A. Ettensohn and Hyla C. Sweet

2 Turning Mesoderm into Blood: The Formation of Hematopoietic Stem Cells during Embryogenesis Alan J. Davidson and Leonard I. Zon

3 Mechanisms of Plant Embryo Development Shunong Bai, Lingjing Chen, Mary Alice Yund, and Zinmay Rence Sung

4 Sperm-Mediated Gene Transfer Anthony W. S. Chan, C. Marc Luetjens, and Gerald P. Schatten

5 Gonocyte–Sertoli Cell Interactions during Development of the Neonatal Rodent Testis Joanne M. Orth, William F. Jester, Ling-Hong Li, and Andrew L. Laslett

6 Attributes and Dynamics of the Endoplasmic Reticulum in Mammalian Eggs Douglas Kline

7 Germ Plasm and Molecular Determinants of Germ Cell Fate Douglas W. Houston and Mary Lou King

Contents of Previous Volumes

219

Volume 51 1 Patterning and Lineage Specification in the Amphibian Embryo Agnes P. Chan and Laurence D. Etkin

2 Transcriptional Programs Regulating Vascular Smooth Muscle Cell Development and Differentiation Michael S. Parmacek

3 Myofibroblasts: Molecular Crossdressers Gennyne A. Walker, Ivan A. Guerrero, and Leslie A. Leinwand

4 Checkpoint and DNA-Repair Proteins Are Associated with the Cores of Mammalian Meiotic Chromosomes Madalena Tarsounas and Peter B. Moens

5 Cytoskeletal and Ca21 Regulation of Hyphal Tip Growth and Initiation Sara Torralba and I. Brent Heath

6 Pattern Formation during C. elegans Vulval Induction Minqin Wang and Paul W. Sternberg

7 A Molecular Clock Involved in Somite Segmentation Miguel Maroto and Olivier Pourquie´

Volume 52 1 Mechanism and Control of Meiotic Recombination Initiation Scott Keeney

2 Osmoregulation and Cell Volume Regulation in the Preimplantation Embryo Jay M. Baltz

3 Cell–Cell Interactions in Vascular Development Diane C. Darland and Patricia A. D’Amore

4 Genetic Regulation of Preimplantation Embryo Survival Carol M. Warner and Carol A. Brenner

220

Contents of Previous Volumes

Volume 53 1 Developmental Roles and Clinical Significance of Hedgehog Signaling Andrew P. McMahon, Philip W. Ingham, and Clifford J. Tabin

2 Genomic Imprinting: Could the Chromatin Structure Be the Driving Force? Andras Paldi

3 Ontogeny of Hematopoiesis: Examining the Emergence of Hematopoietic Cells in the Vertebrate Embryo Jenna L. Galloway and Leonard I. Zon

4 Patterning the Sea Urchin Embryo: Gene Regulatory Networks, Signaling Pathways, and Cellular Interactions Lynne M. Angerer and Robert C. Angerer

Volume 54 1 Membrane Type-Matrix Metalloproteinases (MT-MMP) Stanley Zucker, Duanqing Pei, Jian Cao, and Carlos Lopez-Otin

2 Surface Association of Secreted Matrix Metalloproteinases Rafael Fridman

3 Biochemical Properties and Functions of Membrane-Anchored Metalloprotease-Disintegrin Proteins (ADAMs) J. David Becherer and Carl P. Blobel

4 Shedding of Plasma Membrane Proteins Joaquı´n Arribas and Anna Merlos-Sua´rez

5 Expression of Meprins in Health and Disease Lourdes P. Norman, Gail L. Matters, Jacqueline M. Crisman, and Judith S. Bond

6 Type II Transmembrane Serine Proteases Qingyu Wu

7 DPPIV, Seprase, and Related Serine Peptidases in Multiple Cellular Functions Wen-Tien Chen, Thomas Kelly, and Giulio Ghersi

Contents of Previous Volumes

221

8 The Secretases of Alzheimer’s Disease Michael S. Wolfe

9 Plasminogen Activation at the Cell Surface Vincent Ellis

10 Cell-Surface Cathepsin B: Understanding Its Functional Significance Dora Cavallo-Medved and Bonnie F. Sloane

11 Protease-Activated Receptors Wadie F. Bahou

12 Emmprin (CD147), a Cell Surface Regulator of Matrix Metalloproteinase Production and Function Bryan P. Toole

13 The Evolving Roles of Cell Surface Proteases in Health and Disease: Implications for Developmental, Adaptive, Inflammatory, and Neoplastic Processes Joseph A. Madri

14 Shed Membrane Vesicles and Clustering of Membrane-Bound Proteolytic Enzymes M. Letizia Vittorelli

ThisPageIntentionallyLeftBlank