Biochimie 89 (2007) 516e527 www.elsevier.com/locate/biochi

Transcription elongation through a chromatin template Christophe Lavelle* Laboratoire de Microscopie Mole´culaire et Cellulaire, UMR 8126, Institut Gustave Roussy, 39 rue Camille Desmoulins, 94805 Villejuif, France Received 2 August 2006; accepted 26 September 2006 Available online 17 October 2006

Abstract DNA transaction events occurring during cell life (replication, transcription, recombination, repair, cell division) are always linked to severe changes in the topological state of the double helix. However, since naked DNA almost does not exist in eukaryote nucleus but rather interacts with various proteins, including ubiquitous histones, these topological changes happen in a chromatin context. This review focuses on the role of chromatin fiber structure and dynamics in the regulation of transcription, with an almost exclusive emphasis on the elongation step. Beside a brief overview of our knowledge about transcribed chromatin, we will see how recent mechanistic and biochemical studies give us new insights into the way cell could modulate DNA supercoiling and chromatin conformational dynamics. The participation of topoisomerases in this complex ballet is discussed, since recent data suggest that their role could be closely related to the precise chromatin structure. Lastly, some future prospects to carry on are proposed, hoping this review will help in stimulating discussions and further investigations in the field. Ó 2006 Elsevier Masson SAS. All rights reserved. Keywords: DNA topology; Nucleosome; Chromatin; Topoisomerase; Transcription

1. Introduction 1.1. Chromatin: a compact ‘‘but’’ dynamic structure. In vivo, DNA is not naked but complexed with histones, which lead naturally to the ‘‘nucleosome barrier’’ concept where nucleosomes are supposed to act as both transcription initiation (by simply hiding promoter) and elongation (by stopping polymerase progression) repressors. A clear evidence that nucleosomes are indeed not ‘‘transparent’’ (as sometimes wrongly claimed) to the transcription machinery came from genetic studies which showed that histone depletion derepressed certain important genes [1]. At the same time, many DNase I or micrococcal nuclease digestion assays gave evidences that genes are still packaged within the nucleus as chromatin, even during transcription [2e4], although chromatin structure may be altered during gene activity [2,5e8]. In fact, some studies suggest that canonical nucleosomes are lost

* Tel.: þ33 6 24 71 44 03; fax: þ33 1 42 11 54 94. E-mail address: [email protected] 0300-9084/$ - see front matter Ó 2006 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.biochi.2006.09.019

during transcription (as found for the RNA polymerase I transcribed ribosomal genes [9] and for RNA polymerase II heavily transcribed genes [10]) while others showed they are still present (as found on a number of inducible and housekeeping genes (see chapter 8 in [11] and references therein)). Interestingly, transcribed yeast chromatin and total yeast chromatin are equally sensitive to DNase I digestion, which could mean that the entire yeast genome exists in a state that represents a restricted proportion of total chromatin in higher eukaryotes [12]. In the same way, the control for ribosome gene activity has recently been shown by microscopy studies not to be mediated by changes in chromatin structure [13], a result which confirms former biochemical studies of Sogo and coworkers [9,14e16] and could be interpreted as follow: genes transcribed at high rates (i.e. densely loaded with RNA polymerases) have a disrupted chromatin and lack canonical nucleosomes. However, it remains elusive whether nucleosomes dissociate or simply unfold when polymerase get through (see further discussion below). Hence, if chromatin is sometimes merely seen as a way to pack two meters of DNA into the nucleus volume, the regulatory role of this polymorphic and highly dynamic structure is

C. Lavelle / Biochimie 89 (2007) 516e527

now largely acknowledged. This means that a nucleosome must act both as a compaction and regulation tool, or, in other words, be reasonably stable while keeping some dynamic properties to allow chromatin ‘‘opening’’ for promoter access and subsequent transcription elongation. 1.2. .polymerases have to go through Now, if transcription initiation mainly relies on promoter access (potentially regulated through nucleosome positioning and remodeling; see [17] for a review), do nucleosomes also present a significant obstacle for RNA polymerase once it has left the initiation complex and begun moving along the DNA? Since RNA polymerase needs to pass through nucleosomes, transcription elongation should be repressed by chromatin. Indeed, transcription on chromatin complex is slower in vitro than in vivo, and also slower than on naked DNA templates in vitro, probably because enhancement of pausing [18e20]. Elongation raises in fact many questions, some of the most puzzling being the way RNA polymerase progresses along chromatin template (reviewed in [21]), the subsequent fate of nucleosomes [22,23] and the need for numerous elongation factors (including ATP-dependent remodeling factors, histone chaperones and histone modification enzymes; see [24e26] for reviews and Fig. 2D) as well as topoisomerase activity [27e34] (see Fig. 2F). Also poorly documented is the question whether polymerase moves/rotates along DNA or DNA moves/rotates along a fixed polymerase [35]. All of this will be discussed below. 2. When polymerase meets nucleosome 2.1. A thirty-year-old question Given the fact that even a ‘‘simple’’ prokaryotic polymerase is a molecular motor closely associated with about 40 bp of DNA during transcription elongation, it seems inconceivable that it simply progresses around nucleosome, causing neither displacement nor structural rearrangements. At least, histones must be displaced from the transcribed strand to allow passage of the polymerase. Indeed, in vivo studies designed to evaluate the nucleosomal state of active genes have generally shown that these structures are in more ‘‘open’’ or even disrupted state [36e38]. But at the same time, in vitro studies have showed limited disruption, however depending on the experimental conditions. So, what happens to nucleosomes during transcription (or, equivalently, what happens to polymerase as it encounters nucleosomes) is a question that, although regularly reviewed [22,23,39e47], has not lost much of its mystery. As we will see, difficulty to clearly answer these questions emerges mainly from the different experimental approaches used (in vitro versus in vivo, linear mononucleosome versus reconstituted circular plasmid, prokaryotic versus eukaryotic system) and the difficulty to obtain non ambiguous results. In particular, it is often hard to make the difference between direct (mechanical, through contact) or distal

517

(topological, through DNA elastic constraints propagation) effects of polymerase tracking. 2.2. Overcoming the nucleosome barrier A simple energetic calculation shows that to transcribe the 150 or so base pairs of DNA associated with a nucleosome will require about 1500 kcal mol1 (only taking into account the 10 kcal mol1 required for each phosphate bond of the nascent RNA transcript) while complete displacement of the histone octamer requires 10e20 kcal mol1 [48]. Thus, one can reasonably ask whether the histone octamer is irrelevant to the transcribing polymerase [49]. One obvious way to test this hypothesis is to use reconstituted in vitro system. However, complexity of eukaryotic polymerases (and its necessary initiation factors) led most researchers to utilize simpler prokaryotic polymerases [23]. Experiments of this kind have been carried very soon after the nucleosome ‘‘discovery’’ [50,51] but led quickly to controversial interpretation [52,53]. At the same time, De Bernardin et al. were the first to show that nucleosomes (perhaps in an altered form) on SV40 minichromosomes do not prevent transcription elongation by RNA polymerase II in vivo [54]. However, far from being ‘‘transparent’’ to the transcription machinery (as already discussed above), nucleosomes interact with polymerase in various ways which make interpretation of the numerous experimental studies rather puzzling. 2.3. Many transcription assays. Several in vitro studies showed that transcription elongation by RNA polymerase II (Pol II) is stopped by nucleosomes in physiological salt conditions [55]. However, elongation can be recovered in high salt (>300 mM) [50,56] or with acetylation of the histone tails [57]. The common reasons could be the neutralization of charges, suggesting either that electrostatic interactions between histone octamer and DNA have to be broken for RNA polymerase to transcribe DNA organized into nucleosomes [50], or that these histone tails interact directly with components of the transcriptional machinery [57]. Another usually proposed explanation is H2AeH2B dimers destabilization [56,58,59], consistent with the rapid exchange of these dimers, compare to (H3eH4)2 tetramer, upon transcription [47,60e62]. Furthermore, some additional factors that facilitate these exchanges have been characterized as essential elongation factors, including ATP-dependent remodeling factors such as SWI/SNF [63,64] and histone chaperones such as FACT (FAcilitates Chromatin Transcription) [58,65,66] or NAP1 (Nucleosome Assembly Protein 1 [67e 69] (see also Figs. 1B and 2D). Transcription by RNA polymerase III (Pol III) is also much reduced on nucleosomal template with a tendency for the polymerase to terminate at natural pause sites in the DNA sequence that is wrapped around the nucleosome and furthermore depends on chromatin folding mediated by internucleosomal interactions [70e72]. However, as pointed out by Jackson [22], these observations obtained on tandem arrays

518

C. Lavelle / Biochimie 89 (2007) 516e527

Fig. 1. Short-range chromatin modification due to polymerase overcoming the nucleosomal barrier. (A) Nucleosome ‘‘progressive displacement’’ model. As polymerase (arrow) arrives at nucleosome, it displaces proximal H2AeH2B and proceeds to H3eH4 region through dyad axis, disrupting distal H2AeH2B contacts while DNA begins to rebind to the proximal H3eH4 allowing recapture of DNA and H2AeH2B to the proximal face of the (H3eH4)2 tetramer. (Taken from [23], with kind permission from the authors). (B) ‘‘Histone RNA-spooling’’ model (particular case of the nucleosome ‘‘dissociation and transfer’’ model). In step 1, the RNA polymerase (RNAP) is shown advancing toward a nucleosome. In step 2, the polymerase has disrupted the interaction of DNA with the entry site H2AeH2B dimer (the entry site H2AeH2B dimer now establishes interactions with the nascent RNA, interactions that are facilitated by first binding to NAP1; at this step, there is occasional premature termination as the polymerase attempts to advance through the rest of the nucleosome). In step 3, the polymerase has successfully disrupted all the histone-DNA interactions and the complex of H2A, H2B, H3, and H4 has spooled to the RNA. The polymerase is then shown to transcribe toward another H3eH4 tetramer that is bound to the template DNA. This DNA is in a topologically restrained state, and therefore, the (H3eH4)2 tetramer will not be displaced to the nascent RNA when that region of DNA is transcribed. To return to step 1, the H2AeH2B dimer must be transferred from RNA to the H3eH4 tetramer (binding to NAP1 facilitates this transfer). With the addition of the H2AeH2B dimer, the RNA polymerase can now displace the H3eH4 tetramer as shown in step 2. (Adapted from [68], by permission of the American Chemical Society). (C) Nucleosome ‘‘spooling’’ model (nucleosomal DNA is shaded). RNA polymerase (RP) initiates transcription and rapidly transcribes the first 25 base pairs of the core causing dissociation of this DNA from the octamer; the DNA behind RP binds to the exposed surface of the octamer to form a loop (further movement of RP being severely inhibited due to steric inhibition of rotation of RP within the loop, pausing is observed); the DNA behind RP dissociates from the octamer, thus breaking the loop, and RP continues into the core (the loop might be formed and broken several times); eventually RP penetrates 60 bp into the core, the final loop is formed but then broken ahead of RP as octamer transfer is completed by DNA spooling, and RP transcribes rapidly to the end of the template. (Adapted from [86], by permission of Elsevier). (D) In accordance with the previous model, Bustamante and coworkers suggested that the backward displacement of the nucleosome (N), observed by Studitsky et al. after the RNA polymerase (P) passage, was a structural consequence of DNA wrapping in the elongation complex. The DNA in contact with the histone octamer is shown in dark and the direction of transcription is from left to right (for clarity reasons, the DNA is represented with only one turn around the histone octamer): the polymerase approaches a nucleosome from the left (A), when it encounters a nucleosome, it continues its movement by taking up the downstream DNA from the nucleosome and by giving back the upstream DNA already transcribed (BeF), when it overcomes the nucleosome, the histone octamer is displaced backwards, relative to its initial position on the DNA (G). (Reproduced from [87], with permission from Elsevier). Note that these models may coexist depending on the context (short versus long transcript, highly transcribed versus low transcribed gene, acetylated versus non acetylated nucleosome, Pol II versus Pol I or Pol III, etc).

may not have the same relevance for Pol III (supposed to transcribe only short sequences in vivo) as they have for Pol II (designed to transcribe great distances). Another study held by Felsenfeld and coworkers showed that Pol III transcribes nucleosomal DNA through a direct internal nucleosome transfer in which histones never leave the DNA template, and that polymerase pauses with a pronounced periodicity of 10 to

11 base pairs, consistent with restricted rotation in the DNA loop formed during transfer [73]. As shown by in vivo psoralen cross-linking assays, transcription by RNA polymerase I (Pol I) may occur in a particular environment devoid of nucleosomes [9,14e16]. In fact, high density of elongating Pol I itself seems to be the cause of the formation of non-nucleosomal chromatin [15] (see

C. Lavelle / Biochimie 89 (2007) 516e527

519

Fig. 2. Long range chromatin modification due to topological constraints propagation during elongation. (A) Twin-supercoiled domain model: a transcription ensemble R include the polymerase, the nascent RNA, and proteins bound to the RNA (ribosome and nascent polypeptide chains in prokaryotes; possibly some histones in eukaryotes, see Fig. 1B). As it moves in the direction of the arrow along a DNA segment (anchored on a large structure represented by the solid bars) without turning around it, the DNA in front of the polymerase becomes positively supercoiled and the DNA behind the polymerase becomes negatively supercoiled. (Reproduced from [81], with kind permission from the authors). (B) Twin domains of DNA supercoiling lead to positive supercoils downstream of the transcription complex, which conformationally unfold or split nucleosomes, and negative supercoils upstream, which tightly pack the structure. (Taken from [92], with kind permission from the authors). (C) Twin-supercoiled domain model adapted within the three-state model of chromatin fiber. At the onset of transcription, the whole fiber is torsionally relaxed; as polymerase moves along the template, the right part of the fiber shifts to a positive state (DNAs at the entry/exit of nucleosome cross positively) whereas the left part shifts to a negative state (nucleosomal entry/exit DNAs cross negatively). (Adapted from [144], thanks to Maria Barbi). (D) Transcription by Pol II may result in nucleosome redistribution and partial nucleosome depletion, but also converts nucleosomes to hexasomes in the path of polymerase; facilitation of this process could be achieved with the aid of nucleosome-specific elongation factors such as FACT. (Reprinted from [56], with permission from Elsevier). (E) The passage of RNA polymerase may induce partial release of H1 upon torsional stress and unfolding of the nucleosome structure with the displacement of an H2AeH2B dimer or complete dissociation of the histone octamer. Upon removal of the acetyl groups by deacetylases and relaxation of the torsional stress in the wake of RNA polymerase, nucleosome reforms, histone H1 binds again and folds the nucleosome fiber back into its 30 nm higher-order inactive structure. (Taken from [45], with permission from the Company of Biologists). (F) The relaxation activity of the topoisomerase enzymes may be required during transcription on chromatin templates, to relax excessive supercoils. (Taken from [32], with permission from Oxford University Press). Note that these models may be complementary: positive supercoiling in the front of the polymerase (A) may induce positive crossing of nucleosomes (C) followed (or preceded) by H1 release (E) and unfolding of the octamer (B), facilitating dimer exchanges (D) while topoisomerases help relaxing excessive constraints which would make polymerase to pause and potentially abort elongation (F).

[74] for a review). Remarkably, transcription of ribosomal genes also implies preferential localization of topo I (see discussion below). On the other side, prokaryotic polymerases (SP6 or T7) are much more processive, even in physiological salt [56], although full transcripts are scarcely obtained [75]. The velocity of elongation on the nucleosomal templates is slightly slower than on naked DNA, a difference due to a small increase in pausing on the nucleosomal templates (remarkably, the sites of this increased pausing were shown to be also pause sites on the naked DNA) [20]. However, the question remains if these polymerases displace nucleosomes [18,52,76e78] or not [53,79]. It seems that if polymerase can displace histones from DNA in vitro, the fact that it really does or not depends on subtle experimental conditions such as DNA sequence, supercoiling constraints, ionic strength and the presence of a competitor DNA/RNA. Regarding supercoiling effects (further discussed below), a very decisive experiment has been carried out by Jackson

and coworkers [80] who used a negatively supercoiled plasmid to show that transcription led to loss of nucleosomal structure only in the presence of topoisomerase I (topo I), which suggests that negative supercoiling in the wake of the polymerase is essential to nucleosome regeneration. Interestingly, they also studied transcription in the presence of RNase and showed no evidence for nucleosome loss, which could be interpreted either as a lack of superhelical stress generation by the polymerase without its nascent transcript [81] or as a need for ‘‘histone chaperone’’ role of the RNA [68,82,83]. 2.4. .led to different models These studies show that transcription occurs through nucleosomes in vitro, but there are substantial repression and premature termination leading to the apparent hindrance observed in most of these assays. Furthermore, they suggest the existence of two kinds of progression of RNA polymerases through chromatin in vitro [84]: the first is characteristic of prokaryotic RNA

520

C. Lavelle / Biochimie 89 (2007) 516e527

polymerases and eukaryotic RNA polymerase III and involves nucleosome translocation (without release of the octamer into solution); the second is characteristic of RNA polymerase II and involves displacement of H2AeH2B dimer (without octamer translocation). These considerations lead to several mechanistic models, with or without release of histones. Models relying on histone release are: (i) the ‘‘dissociation and transfer’’ model, in which the histone octamer ahead of the polymerase is simply released and rebinds to DNA after passage (histones might dissociate and reassociate as whole octamers or as dimers and tetramers, and some ‘‘transfer agents’’ such as histone chaperones or nascent RNA may facilitate the process; see [68,69,82] and Figs. 1B and 2D); (ii) the ‘‘progressive displacement’’ model in which sequential disruption of dimers H2AeH2B allows the passage of polymerase without much unwrapping of the nucleosomal DNA (see [23,56] and Fig. 1A). These models are supported by numerous in vivo measurements of histone exchanges which showed that H2AeH2B dimers are quite labile under transcription [60e62]: (H3eH4)2 tetramer may thus serve as the central complex of proteins upon which nucleosomal assembly and disassembly can be rapidly done [22]. However, the fate of ‘‘mobile’’ histones is not clear and some studies seem to reflect a transient change in DNA/histones interactions rather than a pure depletion [3,4]. Also, one should notice that the high energy of the process could explain why, even when disruption of the histone octamer is prevented by histone crosslinking [79,85], in vitro transcription still can occur, which is not in contradiction with the above models. Namely, although histone partial release may facilitate polymerase progression, its requirement may be optional. Models without histone release are: (i) the ‘‘partial release’’ model (better referred to as ‘‘spooling’’ model by Felsenfeld and coworkers) in which successive partial unbinding of nucleosomal DNA sequence enables the passage of polymerase without disruption of the octamer (see [18,73,76,78,86,87] and Fig. 1C and D); (ii) the ‘‘unfolding’’ model in which nucleosome may be altered in a way that would permit transcription while still maintaining a critical number of histone-DNA contacts (see [36,37,43,88] and Fig. 2B). Different mechanisms have been proposed for this unfolding. Weintraub et al. first proposed a separation of the nucleosome into two symmetrically half-nucleosomes [88]. However, this model was soon discredited [43,85] and appears nowadays very unlikely since it would imply breaking the (H3eH4)2 tetramer hold by the very strong four-helix bundle motif [89]. Another structural interpretation of nucleosome unfolding is the lexosome model proposed by Allfrey and coworkers [37]. This model, based on the observation that H3 sulfhydryls showed an enhanced accessibility in transcriptionally active sequences, led to further studies of the same group who exploited this result to isolate transcriptionally active nucleosomes on mercury-agarose columns (see [90] for an example). However, despite many attempts to better characterize this structure [91], its relevance is still controversial [20]. Another model to explain this enhanced H3 accessibility at the dyad has been suggested by Lee and Garrard who proposed

a ‘‘split’’ of the nucleosome upon positive supercoiling [36,92] (see Fig. 2B). Again, the detailed structure of such a mechanism is not clear and need further investigation. Furthermore, one should note that these unfolding models mainly arose from the observation of ‘‘half-nucleosome’’ pattern upon nuclease digestion: however, this pattern can also be explained by transient H2AeH2B depletion in active genes (see ‘‘dissociation and transfer’’ and ‘‘progressive displacement’’ models), without necessity of unfolding [23]. 3. Action at a distance 3.1. Topological domains The different histone-DNA destabilization models proposed above could result from direct interaction with the polymerase but also (at least partially) from the production of topological stress. Namely, DNA is thought to be organized into discrete topological domains as a result of its attachment to cellular structure such as membranes in prokaryotes or nuclear matrix in eukaryotes [93e95], potentially mediated in the last case by some protein complexes including topoisomerase II (topo II) [96e98]. These topologically constrained looped domains are often characterized by a relatively high or relatively low level of nuclease sensitivity and may facilitate the developmental regulation of linked genes. Intriguingly, DNase I sensitivity of the active b-globin genes is reversed in vivo after treatment with novobiocin, a topo II inhibitor, which seems to indicate that torsional stress and nuclease sensitivity of active genes are linked [99]. However, since novobiocin has further been shown to precipitate histones, it raises concern about the interpretation of such studies and points out the need for using such drugs only with caution in experiments designed to implicate topoisomerase activity in chromatin dynamics [100]. Cook and coworkers have also proposed that the transcription complex itself may be part of large structures (potentially linked with nuclear skeleton) called ‘‘transcription factories’’ that participate into the functional partitioning of the genome [35,101,102]. Indeed, several studies showed that attached polymerase can work in vitro [103,104] and gave some evidences for transcription occurring at fixed site in vivo [105]. Furthermore, estimates of the force exerted by these enzymes (up to 14 pN [106]) show that they are amongst the most effective of the known molecular motors and that they have potentially enough power to pull the chromatin template through the viscous nucleus medium. These studies, along with other arguments discussed below, thus tend to favor the immobile polymerase model (however scarcely presented in textbook): this point is of particular relevance to set up the framework in which topological constraints and topoisomerases activity should be interpreted. 3.2. Transcription elongation in prokaryotes: genesis of the twin-supercoiled domain model As RNA polymerase tracks along the DNA template, it follows an helical path: transcription elongation thus requires

C. Lavelle / Biochimie 89 (2007) 516e527

a rotation of the transcription complex relative to the DNA template [107]. Given the high velocity of the process (roughly 23 bases per second according to [108]; see [109] for a careful biophysical study) and the fact that the transcription complex comprises RNA polymerase and the nascent RNA chain (furthermore carrying ribosomes in the case of prokaryotic cells), DNA has been proposed to turn around its axis during transcription rather than polymerase turning around the template (see [35] for discussion). The tracking would then generate positive supercoils ahead of the polymerase and negative supercoils behind it: this is the ‘‘twin supercoiled domain’’ model, first proposed by Wang and coworkers [81] (see Fig. 2A). The model predicts accumulation of negative and positive supercoils in intergenic regions of two divergent and convergent transcription units respectively. In prokaryotes, there are two major topoisomerases: DNA gyrase which can generate negative supercoils and thus relax positively supercoiled DNA, and topoisomerase I which can relax only negatively supercoiled DNA. Several studies using topoisomerase/gyrase inhibitors or mutant strains confirmed the supercoiling accumulation predicted by the twin supercoiled domain hypothesis [110,111]. Further in vitro [103,112,113] or in vivo [114e 118] studies on local topological changes induced by transcription also confirmed the model. 3.3. Transcription elongation in eukaryotes: unconstrained supercoiling Transcription in eukaryotic cells mainly implies type IB and type IIA topoisomerase subfamilies, to which belong respectively the topo I and topo II [119]. While both are efficient at relaxing positive and negative helical stress, they differ by their mechanism: topo I cleaves one strand of duplex DNA and uses DNA torque, without energy cofactor, to drive rotation of the duplex; topo II catalyses the ATP-dependent transport of one DNA duplex through a transient double-strand break in another duplex. Both topo I [120] and topo II [31] have been shown to play an important role in relaxation of chromatin upon transcription by Pol II, but their relative preponderance and precise function are still a matter of discussion [34]. To test whether the twin supercoiled domain model is applicable to eukaryotic transcription, Giaever and Wang used yeast thermosensitive topoisomerase mutant Dtop1/top2ts expressing Escherichia coli DNA topo I (which relaxes negative supercoils specifically): generation of positive supercoils was observed during transcription under restrictive temperature [121], and thus confirmed the model. Interestingly, psoralen cross-linking studies (that use the fact that psoralen photobinds to DNA at a rate nearly linearly proportional to negative superhelical density) have shown that, although the bulk of genomic DNA appears torsionally relaxed within nuclei [122], unconstrained negative DNA supercoils occur in various loci [95,123e125]. This demonstrates that transcription-coupled negative (and probably also positive, although it can not be detected by this technique) supercoils exist even in the presence

521

of active topoisomerases, which lend further support to the twin supercoiled domain model for transcription in the genome of eukaryotic cells. 3.4. Topoisomerase roles and partitioning Observations that topo I concentrates in the nucleolus [126e128] and other highly transcribed regions [129,130], whereas topo II is more evenly distributed throughout the interphase nucleus [97], has supported a general view of topo I as the principal relaxation enzyme of helical tension generated during transcription and other DNA-tracking processes, whereas topo II would be devoted to unlink replicated DNA duplexes [131] and condense chromosomes before segregation [132] (see also [96,119,133,134] for reviews). However, as pointed out by Roca and coworkers, numerous observations argue against this simple partition and indicate that topo II participates also in DNA relaxation tasks [34]. Indeed, inactivation of either topo I or topo II does not significantly affect transcription in yeast, whereas inactivation of both enzymes reduces rRNA synthesis and, to a lesser extent, mRNA synthesis [27,33,135,136]. In a recent study, Roca et al. shed a new light into the potential interpretation of these experimental results: they compared topo I- and topo II-mediated relaxation of highly supercoiled yeast minichromosomes (supercoiling density s > 0.4) and showed that, in contrast to what is observed with naked DNA, topo II relaxes nucleosomal DNA much faster (roughly 5 times as fast) than topo I. Considering that chromatin could impose barriers for DNA twist diffusion (which impair the DNA strand-rotation mechanism of topo I), whereas it favors the juxtaposition of DNA segments (which facilitates the DNA cross-inversion mechanism of topo II), the authors conclude that topo II is probably the main modulator of DNA topology in chromatin fibers, while the nonessential topo I would only assist DNA relaxation where chromatin structure impairs DNA juxtaposition but allows twist diffusion [34]. As we will see below, their relative importance may also vary depending on the organism (topoisomerase activity requirement could be different in human and in yeast, where chromatin has a different structure due for instance to its shorter linkers) and on the level of transcription (see for instance the ribosomal genes transcription). However, one should notice that topoisomerases studies are sometimes difficult to interpret since these enzymes have multiple roles in nucleus [96,134] and that reporter minichromosomes used in experimental assays (SV40 in most cases) can have complex topological response in particular due to alteration of its chromatin structure [137]. 3.5. Nucleosome under highly positive constraint In the former sections, we have shown that: (i) RNA polymerase is supposed to generate positive supercoils in the DNA template ahead and negative supercoils behind [81]; (ii) this positive supercoiling may play a major role in elongation by

522

C. Lavelle / Biochimie 89 (2007) 516e527

destabilizing and possibly releasing H2AeH2B dimers, which subsequent reincorporation on tetramers/hexamers would be facilitated by negative supercoiling in the wake of the polymerase [60,80]; (iii) this model is consistent with the observation of dimers exchange during in vivo transcription [60e62], although octamers could also alternatively be translocated as a whole [79]. Now, in the absence of conclusive direct experimental data on the role of a positive wave, nucleosome reconstitution was studied on positive DNA substrates. The first result is that nucleosome can form on positively supercoiled plasmid without any visible alteration [138,139], although further similar study showed a possible altered form of positively constrained nucleosome [140]. Another system using ethidium bromide intercalation in the loop of mononucleosome on DNA minicircle failed in releasing dimer [141]. The reason is however obvious taking into account the ability of nucleosome entry/exit DNAs to form a positive crossing, as demonstrated by Prunell’s group [142,143]. In the study already mentioned, Roca and coworkers produced highly supercoiled chromatin substrates and characterized their conformation under topological changes in vivo [34]. Indeed, when DNA undergoes positive helical tension, the bulk of histones remain bound to DNA but a few additional micrococcal nuclease cleavage sites indicate an alteration of chromatin structure (as already observed with a similar approach by Lee and Garrard [92]), apparently reversible under DNA relaxation. As the authors noticed, intermediate states might occur, and individual nucleosomes may equilibrate differently according to their DNA sequences and histone modifications. This observation is much consistent with our own study on nucleosome polymorphism, where both histone tails and DNA sequence effects were characterized [143]. Furthermore, a recent work we carried out on single chromatin fiber under torsional stress led us to propose an original interpretation of positively supercoiled chromatin data and to draw a new mechanism that could facilitate the tracking of RNA polymerase through chromatin [144] (see Fig. 2C). This rather sophisticated experimental set up allowed us for the first time to manipulate under torsion single nucleosome arrays reconstituted on tandem repeat of 5S positioning sequence. As we will see below, this system revealed an unsuspected resilience of the chromatin fiber which can accommodate torsional stress relatively more easily than naked DNA.

3.6. Chromatin: a highly resilient structure The conventional biochemical and biophysical techniques used to study chromatin structure and dynamics have been recently complemented by an array of single-molecule approaches, in which chromatin fibers are investigated one at a time [145]. Among them, micromanipulation (either with magnetic or optical tweezers) is a powerful technique to test single chromatin fiber assembly [146] and response to various mechanical constraints, either under tension [147e150] or torsion [144]. Namely, force measurement has revealed the

existence of an internucleosomal attraction that maintains the higher-order chromatin structure under physiological condition [149] and shown the multi-step (and partially reversible) pealing of nucleosomal DNA [148,150]. Only recently has the first torsional manipulation of single chromatin fibers using magnetic tweezers been carried out [144]. This study showed that chromatin fiber can accommodate a surprisingly large amount of torsional stress, either negative or positive, without much change in its length. Experimental results were interpreted in the framework of the three-state model of chromatin fiber (natural extension of the three-state nucleosome model arisen from minicircle studies of Prunell and coworkers [142,143,151]) and provided a prediction of the torque as a function of torsion: this torque is less than 3 pN nm rad1 (compared to 6 pN nm rad1 for naked DNA), that is, substantially smaller than that exerted by advancing polymerase (>5 pN nm rad1) [107] and than the value predicted for nucleosome torsional ejection (9 pN nm rad1) [152]. Now, if we take for example a 10 kbp domain clamped at both ends making a topologically constrained fiber containing 50 nucleosomes (see Fig. 2C), a rough estimation shows that this fiber can accommodate the supercoiling generated by the transcription of about 100 bp (1% of the domain) without the help of topoisomerase and without exceeding the torque exerted by the polymerase [144]. Transposed to a whole chromatin domain of roughly 100 kbp, these figures mean polymerase could transcribe a single gene of 1000 bp (i.e. almost 100 turns of the template around the fixed polymerase) without topoisomerase activity. Consistent with this finding, Parvin et al. have shown that, in the absence of topoisomerase, the synthesis of transcripts <100 nucleotides were unaffected by chromatin reconstituted on a 3000 bp linear template (this figure could however be overestimated due to the lack of supercoiling constraints at the extremities of the fragment, although transient torques develop even in topologically open system [153]). This result give a scheme to explain how torsional resilience of chromatin could lead to a small driving torque and thus to a low activity of topo I, while at the same time chromatin might favor DNA transport activity of topo II by increasing the juxtaposition probability of DNA segments and facilitate the nucleosome entry/exit DNAs to shift from a closed negative state to closed positive one. However, in regions where the transcription rate is high, nucleosome may be depleted (see discussion above) and, as pointed out by Roca [34], DNA-pulling forces exerted by polymerases may prevent the local formation of supercoils: helical tension would then deform DNA mostly by twist, allowing topo I to be more efficient than topo II. This scheme is consistent with the preferential localization of topo I in highly transcribed regions, such as rDNA genes [27,29,33,129,130,154]. Thus, perturbation occurring throughout a transcribed gene seems to be dependent on the relative frequency of transcription [155] which in turn favor the activity of one relaxing mechanism (topo I twist relaxation) compare to another (topo II writhe relaxation).

C. Lavelle / Biochimie 89 (2007) 516e527

4. Conclusion and further discussion 4.1. Chromatin dynamics and nucleosome polymorphism act as a topological buffer during elongation Transcription is a multiscale process whose regulation takes place at multiple steps, including initiation and elongation. Initiation is not discussed here since it would deserve in itself a whole review. However, one should remind that it has also a lot to do with DNA supercoiling, but puzzlingly in an opposite way as for elongation: namely, transcription initiation can be stimulated owing to an accumulation of negative supercoils while synthesis of full-length RNA requires a DNA relaxing activity [33]. As far as elongation is concerned, some relevant features have been discussed here. Basically, we have shown that a molecular motor such as a polymerase push, pull and twist DNA, developing transient torques chromatin has to deal with. For instance, movement of the polymerase through a nucleosome creates approximately 20 positive supercoils (taking into account the linker DNA) while only one would be released by nucleosome disruption. If one obvious role for topoisomerases might be to relax the excessive supercoils in order to prevent transcription stalling, nucleosome itself plays a big part in this. Indeed, as stressed by Owen-Hughes [156], nucleosomes are far from being ‘‘tuna cans’’ but are instead highly polymorphic and dynamic entities which: (i) continuously fluctuate between different conformation at the entry/exit DNAs [151]; (ii) change structure to create subparticles (tetrasome [157], hexasome [56]) and other altered forms, to appropriately respond and adapt to surrounding events such as transcription or replication. This highly dynamical feature of chromatin and its components explains why this structure is paradoxically more flexible than a renormalized naked DNA of the same contour length [144]: in fact, because rotary friction is linearly dependent on the length, but has a higher dependence on radius, chromatin should be harder to spin than DNA [158]. However, this would be true if chromatin was merely seen as a compact rigid cylinder, which is obviously not the case. In fact, it is a highly tunable structure of which elastic constants are strongly sensitive to its architectural details (such as linker length, or acetylation status of the histone) and might locally reach very low values [159]. This tunable elasticity might be a key feature for chromatin function in the regulation of transcription. For instance, as transcription elongation is concerned, the capacity of chromatin to accommodate torsion should favor the smooth progression of tracking enzymes, maintain a low torque to enable polymerase tracking and protect nucleosomes from unsolicited destruction by positive supercoiling [144]. 4.2. Transcription and replication elongation: same problems, same solutions? Remarkably, most of the concept presented could apply for another crucial cellular process: the replication. Indeed, replication elongation shares obvious similarities with transcription

523

elongation [35]: it also happens in fixed structures (so-called replication factories) [160] and implies nucleosome disruption at the replication fork and reincorporation in the wake of the polymerase [161,162]. Progress of the replication fork generates positive supercoils ahead of the replication machinery and removal of these supercoils seems a prerequisite to allow further progress of the replication machinery [163]. Still a more obvious need for topoisomerases occurs at the replication termination, when two replication forks converge at the end of DNA synthesis: decatenation of the two double-strand molecules by type II enzymes is required before cell division (see [119,133,134] for discussion and comparison of topoisomerases roles in these two processes). 4.3. Experimental perspectives First, biophysical studies using cryoelectron microscopy or atomic force microscopy should help us to better characterize potential transcription intermediate such as tetrasome, hexasome or lexosome, particularly in the way they fold DNA. Then, further single chromatin fiber micromanipulation may enable us to test torsional response of chromatin bearing histone variants (H2AZ, macroH2A, etc). Ultimately, some enzymology on these single fibers, with purified polymerases, topoisomerases and additional cofactors such as transcription factors or histone chaperone, should help us to better characterize the transcription mechanisms and the different roles played by the many actors at work. These are some experiments to be done within the next decade.. Acknowledgments I would like to thank Gilles Mirambeau, Eric Le Cam and Claire Heride for discussion, and Ariel Prunell for critical reading of the manuscript and fruitful comments. I would like also to take this ‘‘special issue’’ opportunity to convey the great pleasure I had to have Michel Duguet as a referee of my PhD thesis and how I miss the exciting ‘‘topological’’ debates we had at that time. References [1] M. Han, M. Grunstein, Nucleosome loss activates yeast downstream promoters in vivo, Cell 55 (1988) 1137e1145. [2] D. Lohr, The chromatin structure of an actively expressed, single copy yeast gene, Nucleic Acids Res. 11 (1983) 6755e6773. [3] G.A. Nacheva, D.Y. Guschin, O.V. Preobrazhenskaya, V.L. Karpov, K.K. Ebralidse, A.D. Mirzabekov, Change in the pattern of histone binding to DNA upon transcriptional activation, Cell 58 (1989) 27e36. [4] M.J. Solomon, P.L. Larsen, A. Varshavsky, Mapping protein-DNA interactions in vivo with formaldehyde: evidence that histone H4 is retained on a highly transcribed gene, Cell 53 (1988) 937e947. [5] D.S. Gross, W.T. Garrard, Nuclease hypersensitive sites in chromatin, Annu. Rev. Biochem. 57 (1988) 159e197. [6] R. Reeves, Transcriptionally active chromatin, Biochim. Biophys. Acta. 782 (1984) 343e393. [7] Y.L. Sun, Y.Z. Xu, M. Bellard, P. Chambon, Digestion of the chicken beta-globin gene chromatin with micrococcal nuclease reveals the

524

[8]

[9]

[10]

[11] [12] [13]

[14]

[15]

[16]

[17] [18]

[19]

[20]

[21] [22]

[23] [24]

[25]

[26] [27]

[28] [29]

[30]

C. Lavelle / Biochimie 89 (2007) 516e527 presence of an altered nucleosomal array characterized by an atypical ladder of DNA fragments, EMBO J. 5 (1986) 293e300. C. Wu, Y.C. Wong, S.C. Elgin, The chromatin structure of specific genes: II. Disruption of chromatin structure during gene activity, Cell 16 (1979) 807e814. A. Conconi, R.M. Widmer, T. Koller, J.M. Sogo, Two different chromatin structures coexist in ribosomal RNA genes throughout the cell cycle, Cell 57 (1989) 753e761. R.M. Widmer, R. Lucchini, M. Lezzi, B. Meyer, J.M. Sogo, J.E. Edstrom, T. Koller, Chromatin structure of a hyperactive secretory protein gene (in Balbiani ring 2) of Chironomus, EMBO J. 3 (1984) 1635e1641. K.E. van Holde, Chromatin, Springer, Berlin, 1989. D. Lohr, L. Hereford, Yeast chromatin is uniformly digested by DNaseI, Proc. Natl. Acad. Sci. U.S.A. 76 (1979) 4285e4288. M. Derenzini, G. Pasquinelli, M.F. O’Donohue, D. Ploton, M. Thiry, Structural and functional organization of ribosomal genes within the mammalian cell nucleolus, J. Histochem. Cytochem. 54 (2006) 131e 145. R. Dammann, R. Lucchini, T. Koller, J.M. Sogo, Chromatin structures and transcription of rDNA in yeast Saccharomyces cerevisiae, Nucleic Acids Res. 21 (1993) 2331e2338. R. Dammann, R. Lucchini, T. Koller, J.M. Sogo, Transcription in the yeast rRNA gene locus: distribution of the active gene copies and chromatin structure of their flanking regulatory sequences, Mol. Cell. Biol. 15 (1995) 5294e5303. J.M. Sogo, P.J. Ness, R.M. Widmer, R.W. Parish, T. Koller, Psoralencrosslinking of DNA as a probe for the structure of active nucleolar chromatin, J. Mol. Biol. 178 (1984) 897e919. R.G. Roeder, Transcriptional regulation and the role of diverse coactivators in animal cells, FEBS Lett. 579 (2005) 909e915. J. Bednar, V.M. Studitsky, S.A. Grigoryev, G. Felsenfeld, C.L. Woodcock, The nature of the nucleosomal barrier to transcription: direct observation of paused intermediates by electron cryomicroscopy, Mol. Cell 4 (1999) 377e386. M.G. Izban, D.S. Luse, Transcription on nucleosomal templates by RNA polymerase II in vitro: inhibition of elongation with enhancement of sequence-specific pausing, Genes Dev. 5 (1991) 683e696. R.U. Protacio, J. Widom, Nucleosome transcription studied in a realtime synchronous system: test of the lexosome model and direct measurement of effects due to histone octamer, J. Mol. Biol. 256 (1996) 458e472. G. Orphanides, D. Reinberg, RNA polymerase II elongation through chromatin, Nature 407 (2000) 471e475. V. Jackson, What happens to nucleosomes during transcription? in: J. Zlatanova, S.H. Leuba (Eds.), Chromatin Structure and Dynamics: State-of-the-Art Elsevier, Amsterdam, 2004. K.E. van Holde, D.E. Lohr, C. Robert, What happens to nucleosomes during transcription? J. Biol. Chem. 267 (1992) 2837e2840. G.J. Narlikar, H.Y. Fan, R.E. Kingston, Cooperation between complexes that regulate chromatin structure and transcription, Cell 108 (2002) 475e487. R.J. Sims 3rd, R. Belotserkovskaya, D. Reinberg, Elongation by RNA polymerase II: the short and long of it, Genes Dev. 18 (2004) 2437e 2468. G.A. Hartzog, Transcription elongation by RNA polymerase II, Curr. Opin. Genet. Dev. 13 (2003) 119e126. S.J. Brill, S. DiNardo, K. Voelkel-Meiman, R. Sternglanz, Need for DNA topoisomerase activity as a swivel for DNA replication for transcription of ribosomal RNA, Nature 326 (1987) 414e416. S.J. Brill, R. Sternglanz, Transcription-dependent DNA supercoiling in yeast DNA topoisomerase mutants, Cell 54 (1988) 403e411. G. Cavalli, D. Bachmann, F. Thoma, Inactivation of topoisomerases affects transcription-dependent chromatin transitions in rDNA but not in a gene transcribed by RNA polymerase II, EMBO J. 15 (1996) 590e 597. I. Collins, A. Weber, D. Levens, Transcriptional consequences of topoisomerase inhibition, Mol. Cell. Biol. 21 (2001) 8437e8451.

[31] N. Mondal, J.D. Parvin, DNA topoisomerase IIalpha is required for RNA polymerase II transcription on chromatin templates, Nature 413 (2001) 435e438. [32] N. Mondal, Y. Zhang, Z. Jonsson, S.K. Dhar, M. Kannapiran, J.D. Parvin, Elongation by RNA polymerase II on chromatin templates requires topoisomerase activity, Nucleic Acids Res. 31 (2003) 5016e 5024. [33] M.C. Schultz, S.J. Brill, Q. Ju, R. Sternglanz, R.H. Reeder, Topoisomerases and yeast rRNA transcription: negative supercoiling stimulates initiation and topoisomerase activity is required for elongation, Genes Dev. 6 (1992) 1332e1341. [34] J. Salceda, X. Fernandez, J. Roca, Topoisomerase II, not topoisomerase I, is the proficient relaxase of nucleosomal DNA, EMBO J. 25 (2006) 2575e2583. [35] P.R. Cook, The organization of replication and transcription, Science 284 (1999) 1790e1795. [36] M.S. Lee, W.T. Garrard, Transcription-induced nucleosome ‘splitting’: an underlying structure for DNase I sensitive chromatin, EMBO J. 10 (1991) 607e615. [37] C.P. Prior, C.R. Cantor, E.M. Johnson, V.C. Littau, V.G. Allfrey, Reversible changes in nucleosome structure and histone H3 accessibility in transcriptionally active and inactive states of rDNA chromatin, Cell 34 (1983) 1033e1042. [38] H. Weintraub, M. Groudine, Chromosomal subunits in active genes have an altered conformation, Science 193 (1976) 848e856. [39] C.C. Adams, J.L. Workman, Nucleosome displacement in transcription, Cell 72 (1993) 305e308. [40] G. Felsenfeld, D. Clark, V. Studitsky, Transcription through nucleosomes, Biophys. Chem. 86 (2000) 231e237. [41] R.D. Kornberg, Y. Lorch, Irresistible force meets immovable object: transcription and the nucleosome, Cell 67 (1991) 833e836. [42] R.D. Kornberg, Y. Lorch, Chromatin structure and transcription, Annu. Rev. Cell Biol. 8 (1992) 563e587. [43] F. Thoma, Structural changes in nucleosomes during transcription: strip, split or flip? Trends Genet. 7 (1991) 175e177. [44] D.J. Steger, J.L. Workman, Remodeling chromatin structures for transcription: what happens to the histones? Bioessays 18 (1996) 875e884. [45] J. Ausio, Structure and dynamics of transcriptionally active chromatin, J. Cell Sci. 102 (Pt 1) (1992) 1e5. [46] J.L. Workman, Nucleosome displacement in transcription, Genes Dev. 20 (2006) 2009e2017. [47] C. Thiriet, J.J. Hayes, Histone dynamics during transcription: exchange of H2A/H2B dimers and H3/H4 tetramers during pol II elongation, Results Probl. Cell Differ. 41 (2006) 77e90. [48] J.M. Gottesfeld, K. Luger, Energetics and affinity of the histone octamer for defined DNA sequences, Biochemistry 40 (2001) 10927e10933. [49] B.M. Turner, Chromatin and Gene Regulation, Blackwell Science, Oxford, 2001. [50] B. Wasylyk, P. Chambon, Transcription by eukaryotic RNA polymerases A and B of chromatin assembled in vitro, Eur. J. Biochem. 98 (1979) 317e327. [51] P. Williamson, G. Felsenfeld, Transcription of histone-covered T7 DNA by Escherichia coli RNA polymerase, Biochemistry 17 (1978) 5695e 5705. [52] Y. Lorch, J.W. LaPointe, R.D. Kornberg, Nucleosomes inhibit the initiation of transcription but allow chain elongation with the displacement of histones, Cell 49 (1987) 203e210. [53] R. Losa, D.D. Brown, A bacteriophage RNA polymerase transcribes in vitro through a nucleosome core without displacing it, Cell 50 (1987) 801e808. [54] W. De Bernardin, T. Koller, J.M. Sogo, Structure of in-vivo transcribing chromatin as studied in simian virus 40 minichromosomes, J. Mol. Biol. 191 (1986) 469e482. [55] C.H. Chang, D.S. Luse, The H3/H4 tetramer blocks transcript elongation by RNA polymerase II in vitro, J. Biol. Chem. 272 (1997) 23427e23434. [56] M.L. Kireeva, W. Walter, V. Tchernajenko, V. Bondarenko, M. Kashlev, V.M. Studitsky, Nucleosome remodeling induced by RNA polymerase

C. Lavelle / Biochimie 89 (2007) 516e527

[57]

[58]

[59]

[60]

[61] [62]

[63]

[64]

[65]

[66]

[67]

[68]

[69]

[70]

[71]

[72]

[73]

[74]

[75]

[76] [77]

II: loss of the H2A/H2B dimer during transcription, Mol. Cell 9 (2002) 541e552. M. Pineiro, P.J. Gonzalez, F. Hernandez, E. Palacian, Interaction of RNA polymerase II with acetylated nucleosomal core particles, Biochem. Biophys. Res. Commun. 177 (1991) 370e376. R. Belotserkovskaya, S. Oh, V.A. Bondarenko, G. Orphanides, V.M. Studitsky, D. Reinberg, FACT facilitates transcription-dependent nucleosome alteration, Science 301 (2003) 1090e1093. P.J. Gonzalez, E. Palacian, Interaction of RNA polymerase II with structurally altered nucleosomal particles. Transcription is facilitated by loss of one H2A.H2B dimer, J. Biol. Chem. 264 (1989) 18457e18462. V. Jackson, In vivo studies on the dynamics of histone-DNA interaction: evidence for nucleosome dissolution during replication and transcription and a low level of dissolution independent of both, Biochemistry 29 (1990) 719e731. L. Louters, R. Chalkley, Exchange of histones H1, H2A, and H2B in vivo, Biochemistry 24 (1985) 3080e3085. S. Schwager, J.D. Retief, P. de Groot, C. von Holt, Rapid exchange of histone H2A and H2B in sea urchin embryo chromatin, FEBS Lett. 189 (1985) 305e309. R. Bash, H. Wang, C. Anderson, J. Yodh, G. Hager, S.M. Lindsay, D. Lohr, AFM imaging of protein movements: histone H2AeH2B release during nucleosome remodeling, FEBS Lett. 580 (2006) 4757e 4761. J.K. Davie, C.M. Kane, Genetic interactions between TFIIS and the Swi-Snf chromatin-remodeling complex, Mol. Cell. Biol. 20 (2000) 5960e5973. R. Belotserkovskaya, D. Reinberg, Facts about FACT and transcript elongation through chromatin, Curr. Opin. Genet. Dev. 14 (2004) 139e146. G. Orphanides, G. LeRoy, C.H. Chang, D.S. Luse, D. Reinberg, FACT, a factor that facilitates transcript elongation through nucleosomes, Cell 92 (1998) 105e116. Y. Ishimi, J. Hirosumi, W. Sato, K. Sugasawa, S. Yokota, F. Hanaoka, M. Yamada, Purification and initial characterization of a protein which facilitates assembly of nucleosome-like structure from mammalian cells, Eur. J. Biochem. 142 (1984) 431e439. V. Levchenko, V. Jackson, Histone release during transcription: NAP1 forms a complex with H2A and H2B and facilitates a topologically dependent release of H3 and H4 from the nucleosome, Biochemistry 43 (2004) 2359e2372. J. Mazurkiewicz, J.F. Kepert, K. Rippe, On the mechanism of nucleosome assembly by histone chaperone NAP1, J. Biol. Chem. 281 (2006) 16462e16472. S.J. Felts, P.A. Weil, R. Chalkley, Transcription factor requirements for in vitro formation of transcriptionally competent 5S rRNA gene chromatin, Mol. Cell. Biol. 10 (1990) 2390e2401. J.C. Hansen, A.P. Wolffe, Influence of chromatin folding on transcription initiation and elongation by RNA polymerase III, Biochemistry 31 (1992) 7977e7988. R.H. Morse, Nucleosomes inhibit both transcriptional initiation and elongation by RNA polymerase III in vitro, EMBO J. 8 (1989) 2343e2351. V.M. Studitsky, G.A. Kassavetis, E.P. Geiduschek, G. Felsenfeld, Mechanism of transcription through the nucleosome by eukaryotic RNA polymerase, Science 278 (1997) 1960e1963. M. Toussaint, G. Levasseur, M. Tremblay, M. Paquette, A. Conconi, Psoralen photocrosslinking, a tool to study the chromatin structure of RNA polymerase Ietranscribed ribosomal genes, Biochem. Cell Biol. 83 (2005) 449e459. T.E. O’Neill, M. Roberge, E.M. Bradbury, Nucleosome arrays inhibit both initiation and elongation of transcripts by bacteriophage T7 RNA polymerase, J. Mol. Biol. 223 (1992) 67e78. D.J. Clark, G. Felsenfeld, A nucleosome core is transferred out of the path of a transcribing polymerase, Cell 71 (1992) 11e22. M.F. O’Donohue, I. Duband-Goulet, A. Hamiche, A. Prunell, Octamer displacement and redistribution in transcription of single nucleosomes, Nucleic Acids Res. 22 (1994) 937e945.

525

[78] V.M. Studitsky, D.J. Clark, G. Felsenfeld, A histone octamer can step around a transcribing polymerase without leaving the template, Cell 76 (1994) 371e382. [79] T.E. O’Neill, J.G. Smith, E.M. Bradbury, Histone octamer dissociation is not required for transcript elongation through arrays of nucleosome cores by phage T7 RNA polymerase in vitro, Proc. Natl. Acad. Sci. U.S.A. 90 (1993) 6203e6207. [80] P. Pfaffle, V. Gerlach, L. Bunzel, V. Jackson, In vitro evidence that transcription-induced stress causes nucleosome dissolution and regeneration, J. Biol. Chem. 265 (1990) 16830e16840. [81] L.F. Liu, J.C. Wang, Supercoiling of the DNA template during transcription, Proc. Natl. Acad. Sci. U.S.A. 84 (1987) 7024e7027. [82] H.F. Peng, V. Jackson, Measurement of the frequency of histone displacement during the in vitro transcription of nucleosomes: RNA is a competitor for these histones, Biochemistry 36 (1997) 12371e 12382. [83] B. ten Heggeler-Bordier, S. Muller, M. Monestier, W. Wahli, An immuno-electron microscopical analysis of transcribing multinucleosomal templates: what happens to the histones? J. Mol. Biol. 299 (2000) 853e 858. [84] V.M. Studitsky, W. Walter, M. Kireeva, M. Kashlev, G. Felsenfeld, Chromatin remodeling by RNA polymerases, Trends Biochem. Sci. 29 (2004) 127e135. [85] H.J. Gould, G.J. Cowling, N.R. Harborne, J. Allan, An examination of models for chromatin transcription, Nucleic Acids Res. 8 (1980) 5255e 5266. [86] V.M. Studitsky, D.J. Clark, G. Felsenfeld, Overcoming a nucleosomal barrier to transcription, Cell 83 (1995) 19e27. [87] C. Rivetti, S. Codeluppi, G. Dieci, C. Bustamante, Visualizing RNA extrusion and DNA wrapping in transcription elongation complexes of bacterial and eukaryotic RNA polymerases, J. Mol. Biol. 326 (2003) 1413e1426. [88] H. Weintraub, A. Worcel, B. Alberts, A model for chromatin based upon two symmetrically paired half-nucleosomes, Cell 9 (1976) 409e 417. [89] K. Luger, A.W. Mader, R.K. Richmond, D.F. Sargent, T.J. Richmond, Crystal structure of the nucleosome core particle at 2.8 A resolution, Nature 389 (1997) 251e260. [90] J. Walker, T.A. Chen, R. Sterner, M. Berger, F. Winston, V.G. Allfrey, Affinity chromatography of mammalian and yeast nucleosomes. Two modes of binding of transcriptionally active mammalian nucleosomes to organomercurial-agarose columns, and contrasting behavior of the active nucleosomes of yeast, J. Biol. Chem. 265 (1990) 5736e5746. [91] D.P. Bazett-Jones, E. Mendez, G.J. Czarnota, F.P. Ottensmeyer, V.G. Allfrey, Visualization and analysis of unfolded nucleosomes associated with transcribing chromatin, Nucleic Acids Res. 24 (1996) 321e 329. [92] M.S. Lee, W.T. Garrard, Positive DNA supercoiling generates a chromatin conformation characteristic of highly active genes, Proc. Natl. Acad. Sci. U.S.A. 88 (1991) 9675e9679. [93] D.A. Jackson, P. Dickinson, P.R. Cook, Attachment of DNA to the nucleoskeleton of HeLa cells examined using physiological conditions, Nucleic Acids Res. 18 (1990) 4385e4393. [94] D.A. Jackson, P. Dickinson, P.R. Cook, The size of chromatin loops in HeLa cells, EMBO J. 9 (1990) 567e571. [95] P.R. Kramer, R.R. Sinden, Measurement of unrestrained negative supercoiling and topological domain size in living human cells, Biochemistry 36 (1997) 3151e3158. [96] V. Borde, M. Duguet, The mapping of DNA topoisomerase sites in vivo: a tool to enlight the functions of topoisomerases, Biochimie 80 (1998) 223e233. [97] M.O. Christensen, M.K. Larsen, H.U. Barthelmes, R. Hock, C.L. Andersen, E. Kjeldsen, B.R. Knudsen, O. Westergaard, F. Boege, C. Mielke, Dynamics of human DNA topoisomerases IIalpha and IIbeta in living cells, J. Cell Biol. 157 (2002) 31e44. [98] P.N. Cockerill, W.T. Garrard, Chromosomal loop anchorage of the kappa immunoglobulin gene occurs next to the enhancer in a region containing topoisomerase II sites, Cell 44 (1986) 273e282.

526

C. Lavelle / Biochimie 89 (2007) 516e527

[99] B. Villeponteau, M. Lundell, H. Martinson, Torsional stress promotes the DNAase I sensitivity of active genes, Cell 39 (1984) 469e478. [100] M. Cotten, D. Bresnahan, S. Thompson, L. Sealy, R. Chalkley, Novobiocin precipitates histones at concentrations normally used to inhibit eukaryotic type II topoisomerase, Nucleic Acids Res. 14 (1986) 3671e 3686. [101] P.R. Cook, The nucleoskeleton and the topology of transcription, Eur. J. Biochem. 185 (1989) 487e501. [102] F.J. Iborra, A. Pombo, J. McManus, D.A. Jackson, P.R. Cook, The topology of transcription by immobilized polymerases, Exp. Cell Res. 229 (1996) 167e173. [103] P.R. Cook, F. Gove, Transcription by an immobilized RNA polymerase from bacteriophage T7 and the topology of transcription, Nucleic Acids Res. 20 (1992) 3591e3598. [104] D.A. Schafer, J. Gelles, M.P. Sheetz, R. Landick, Transcription by single molecules of RNA polymerase observed by light microscopy, Nature 352 (1991) 444e448. [105] P. Dickinson, P.R. Cook, D.A. Jackson, Active RNA polymerase I is fixed within the nucleus of HeLa cells, EMBO J. 9 (1990) 2207e2214. [106] H. Yin, M.D. Wang, K. Svoboda, R. Landick, S.M. Block, J. Gelles, Transcription against an applied force, Science 270 (1995) 1653e 1657. [107] Y. Harada, O. Ohara, A. Takatsuki, H. Itoh, N. Shimamoto, K. Kinosita Jr., Direct observation of DNA rotation during transcription by Escherichia coli RNA polymerase, Nature 409 (2001) 113e115. [108] A.W. Shermoen, P.H. O’Farrell, Progression of the cell cycle through mitosis leads to abortion of nascent transcripts, Cell 67 (1991) 303e 310. [109] S.F. Tolic-Norrelykke, A.M. Engh, R. Landick, J. Gelles, Diversity in the rates of transcript elongation by single RNA polymerase molecules, J. Biol. Chem. 279 (2004) 3292e3299. [110] D. Lockshon, D.R. Morris, Positively supercoiled plasmid DNA is produced by treatment of Escherichia coli with DNA gyrase inhibitors, Nucleic Acids Res. 11 (1983) 2999e3017. [111] G.J. Pruss, K. Drlica, Topoisomerase I mutants: the gene on pBR322 that encodes resistance to tetracycline affects plasmid DNA supercoiling, Proc. Natl. Acad. Sci. U.S.A. 83 (1986) 8952e8956. [112] P. Droge, A. Nordheim, Transcription-induced conformational change in a topologically closed DNA domain, Nucleic Acids Res. 19 (1991) 2941e2946. [113] Y.P. Tsao, H.Y. Wu, L.F. Liu, Transcription-driven supercoiling of DNA: direct biochemical evidence from in vitro studies, Cell 56 (1989) 111e118. [114] J.K. Lodge, T. Kazic, D.E. Berg, Formation of supercoiling domains in plasmid pBR322, J. Bacteriol. 171 (1989) 2181e2187. [115] E.A. Ostrander, P. Benedetti, J.C. Wang, Template supercoiling by a chimera of yeast GAL4 protein and phage T7 RNA polymerase, Science 249 (1990) 1261e1265. [116] A.R. Rahmouni, R.D. Wells, Direct evidence for the effect of transcription on local DNA supercoiling in vivo, J. Mol. Biol. 223 (1992) 131e 144. [117] H.Y. Wu, S.H. Shyy, J.C. Wang, L.F. Liu, Transcription generates positively and negatively supercoiled domains in the template, Cell 53 (1988) 433e440. [118] L. Moulin, A.R. Rahmouni, F. Boccard, Topological insulators inhibit diffusion of transcription-induced positive supercoils in the chromosome of Escherichia coli, Mol. Microbiol. 55 (2005) 601e610. [119] J.C. Wang, Cellular roles of DNA topoisomerases: a molecular perspective, Nat. Rev. Mol. Cell. Biol. 3 (2002) 430e440. [120] A. Khobta, F. Ferri, L. Lotito, A. Montecucco, R. Rossi, G. Capranico, Early effects of topoisomerase I inhibition on RNA polymerase II along transcribed genes in human cells, J. Mol. Biol. 357 (2006) 127e138. [121] G.N. Giaever, J.C. Wang, Supercoiling of intracellular DNA can occur in eukaryotic cells, Cell 55 (1988) 849e856. [122] R.R. Sinden, J.O. Carlson, D.E. Pettijohn, Torsional tension in the DNA double helix measured with trimethylpsoralen in living E. coli cells: analogous measurements in insect and human cells, Cell 21 (1980) 773e783.

[123] E.R. Jupe, R.R. Sinden, I.L. Cartwright, Stably maintained microdomain of localized unrestrained supercoiling at a Drosophila heat shock gene locus, EMBO J. 12 (1993) 1067e1075. [124] M. Ljungman, P.C. Hanawalt, Localized torsional tension in the DNA of human cells, Proc. Natl. Acad. Sci. U.S.A. 89 (1992) 6055e6059. [125] K. Matsumoto, S. Hirose, Visualization of unconstrained negative supercoils of DNA on polytene chromosomes of Drosophila, J. Cell Sci. 117 (2004) 3797e3805. [126] M.O. Christensen, R.M. Krokowski, H.U. Barthelmes, R. Hock, F. Boege, C. Mielke, Distinct effects of topoisomerase I and RNA polymerase I inhibitors suggest a dual mechanism of nucleolar/nucleoplasmic partitioning of topoisomerase I. J. Biol. Chem. 279 (2004) 21873e 21882. [127] M.T. Muller, W.P. Pfund, V.B. Mehta, D.K. Trask, Eukaryotic type I topoisomerase is enriched in the nucleolus and catalytically active on ribosomal DNA, EMBO J. 4 (1985) 1237e1243. [128] H. Zhang, J.C. Wang, L.F. Liu, Involvement of DNA topoisomerase I in transcription of human ribosomal RNA genes, Proc. Natl. Acad. Sci. U.S.A. 85 (1988) 1060e1064. [129] G. Fleischmann, G. Pflugfelder, E.K. Steiner, K. Javaherian, G.C. Howard, J.C. Wang, S.C. Elgin, Drosophila DNA topoisomerase I is associated with transcriptionally active regions of the genome, Proc. Natl. Acad. Sci. U.S.A. 81 (1984) 6958e6962. [130] Y. Mao, I.R. Mehl, M.T. Muller, Subnuclear distribution of topoisomerase I is linked to ongoing transcription and p53 status, Proc. Natl. Acad. Sci. U.S.A. 99 (2002) 1235e1240. [131] I. Lucas, T. Germe, M. Chevrier-Miller, O. Hyrien, Topoisomerase II can unlink replicating DNA by precatenane removal, EMBO J. 20 (2001) 6509e6519. [132] Y. Adachi, M. Luke, U.K. Laemmli, Chromosome assembly in vitro: topoisomerase II is required for condensation, Cell 64 (1991) 137e148. [133] J.C. Wang, DNA topoisomerases, Annu. Rev. Biochem. 65 (1996) 635e 692. [134] J.L. Nitiss, Investigating the biological functions of DNA topoisomerases in eukaryotic cells, Biochim. Biophys. Acta. 1400 (1998) 63e81. [135] M.R. Gartenberg, J.C. Wang, Positive supercoiling of DNA greatly diminishes mRNA synthesis in yeast, Proc. Natl. Acad. Sci. U.S.A. 89 (1992) 11461e11465. [136] M. Yamagishi, M. Nomura, Deficiency in both type I and type II DNA topoisomerase activities differentially affect rRNA and ribosomal protein synthesis in Schizosaccharomyces pombe, Curr. Genet. 13 (1988) 305e314. [137] R.M. Givens, R.A. Saavedra, J.A. Huberman, Topological complexity of SV40 minichromosomes, J. Mol. Biol. 257 (1996) 53e65. [138] D.J. Clark, G. Felsenfeld, Formation of nucleosomes on positively supercoiled DNA, EMBO J. 10 (1991) 387e395. [139] D.J. Clark, R. Ghirlando, G. Felsenfeld, H. Eisenberg, Effect of positive supercoiling on DNA compaction by nucleosome cores, J. Mol. Biol. 234 (1993) 297e301. [140] V. Jackson, Influence of positive stress on nucleosome assembly, Biochemistry 32 (1993) 5901e5912. [141] A. Sivolob, F. De Lucia, B. Revet, A. Prunell, Nucleosome dynamics. II. High flexibility of nucleosome entering and exiting DNAs to positive crossing. An ethidium bromide fluorescence study of mononucleosomes on DNA minicircles, J. Mol. Biol. 285 (1999) 1081e1099. [142] F. De Lucia, M. Alilat, A. Sivolob, A. Prunell, Nucleosome dynamics. III. Histone tail-dependent fluctuation of nucleosomes between open and closed DNA conformations. Implications for chromatin dynamics and the linking number paradox. A relaxation study of mononucleosomes on DNA minicircles, J. Mol. Biol. 285 (1999) 1101e1119. [143] A. Sivolob, C. Lavelle, A. Prunell, Sequence-dependent nucleosome structural and dynamic polymorphism. Potential involvement of histone H2B N-terminal tail proximal domain, J. Mol. Biol. 326 (2003) 49e63. [144] A. Bancaud, N. Conde e Silva, M. Barbi, G. Wagner, J.F. Allemand, J. Mozziconacci, C. Lavelle, V. Croquette, J.M. Victor, A. Prunell, J.L. Viovy, Structural plasticity of single chromatin fibers revealed by torsional manipulation, Nat. Struct. Mol. Biol. 13 (2006) 444e450.

C. Lavelle / Biochimie 89 (2007) 516e527 [145] J. Zlatanova, S.H. Leuba, Chromatin fibers, one-at-a-time, J. Mol. Biol. 331 (2003) 1e19. [146] S.H. Leuba, M.A. Karymov, M. Tomschik, R. Ramjit, P. Smith, J. Zlatanova, Assembly of single chromatin fibers depends on the tension in the DNA molecule: magnetic tweezers study, Proc. Natl. Acad. Sci. U.S.A. 100 (2003) 495e500. [147] M.L. Bennink, S.H. Leuba, G.H. Leno, J. Zlatanova, B.G. de Grooth, J. Greve, Unfolding individual nucleosomes by stretching single chromatin fibers with optical tweezers, Nat. Struct. Biol. 8 (2001) 606e 610. [148] B.D. Brower-Toland, C.L. Smith, R.C. Yeh, J.T. Lis, C.L. Peterson, M.D. Wang, Mechanical disruption of individual nucleosomes reveals a reversible multistage release of DNA, Proc. Natl. Acad. Sci. U.S.A. 99 (2002) 1960e1965. [149] Y. Cui, C. Bustamante, Pulling a single chromatin fiber reveals the forces that maintain its higher-order structure, Proc. Natl. Acad. Sci. U.S.A. 97 (2000) 127e132. [150] L.H. Pope, M.L. Bennink, K.A. van Leijenhorst-Groener, D. Nikova, J. Greve, J.F. Marko, Single chromatin fiber stretching reveals physically distinct populations of disassembly events, Biophys. J. 88 (2005) 3572e3583. [151] A. Sivolob, A. Prunell, Nucleosome conformational flexibility and implications for chromatin dynamics, Philos. Trans. A Math. Phys. Eng. Sci. 362 (2004) 1519e1547. [152] A. Sarkar, J.F. Marko, Removal of DNA-bound proteins by DNA twisting, Phys. Rev. E Stat. Nonlin. Soft Matter. Phys. 64 (2001) 061909. [153] P. Nelson, Transport of torsional stress in DNA, Proc. Natl. Acad. Sci. U.S.A. 96 (1999) 14342e14347.

527

[154] P.J. Ness, T. Koller, F. Thoma, Topoisomerase I cleavage sites identified and mapped in the chromatin of Dictyostelium ribosomal RNA genes, J. Mol. Biol. 200 (1988) 127e139. [155] I.L. Cartwright, S.C. Elgin, Nucleosomal instability and induction of new upstream protein-DNA associations accompany activation of four small heat shock protein genes in Drosophila melanogaster, Mol. Cell. Biol. 6 (1986) 779e791. [156] A. Flaus, T. Owen-Hughes, Mechanisms for ATP-dependent chromatin remodelling: farewell to the tuna-can octamer? Curr. Opin. Genet. Dev. 14 (2004) 165e173. [157] M. Alilat, A. Sivolob, B. Revet, A. Prunell, Nucleosome dynamics. Protein and DNA contributions in the chiral transition of the tetrasome, the histone (H3eH4)2 tetramer-DNA particle, J. Mol. Biol. 291 (1999) 815e841. [158] F. Kouzine, J. Liu, S. Sanford, H.J. Chung, D. Levens, The dynamic response of upstream DNA to transcription-generated torsional stress, Nat. Struct. Mol. Biol. 11 (2004) 1092e1100. [159] E. Ben-Haim, A. Lesne, J.M. Victor, Chromatin: a tunable spring at work inside chromosomes, Phys. Rev. E Stat. Nonlin. Soft Mater. Phys. 64 (2001) 051921. [160] P. Hozak, A.B. Hassan, D.A. Jackson, P.R. Cook, Visualization of replication factories attached to nucleoskeleton, Cell 73 (1993) 361e373. [161] A.T. Annunziato, Split decision: what happens to nucleosomes during DNA replication? J. Biol. Chem. 280 (2005) 12065e12068. [162] C. Gruss, J. Wu, T. Koller, J.M. Sogo, Disruption of the nucleosomes at the replication fork, EMBO J. 12 (1993) 4533e4545. [163] A.T. Annunziato, Inhibitors of topoisomerases I and II arrest DNA replication, but do not prevent nucleosome assembly in vivo, J. Cell Sci. 93 (Pt 4) (1989) 593e603.