RNA polymerase II pausing regulates a quiescence-dependent transcriptional program, priming cells for cell cycle reentry

Adult stem cells persist in mammalian tissues by entering a state of reversible arrest or quiescence associated with low transcription. Using cultured myoblasts and primary muscle stem cells, we show that RNA synthesis is strongly repressed in G0, returning within minutes of activation. We investigate the underlying mechanism and reveal a role for promoter-proximal RNAPol II pausing: by mapping global Pol II occupancy using ChIP-seq, in conjunction with RNA-seq to identify repressed transcriptional networks unique to G0. Strikingly, Pol II pausing is enhanced in G0 on genes encoding regulators of RNA biogenesis (Ncl, Rps24, Ctdp1), and release of pausing is critical for cell cycle re-entry. Finally, we uncover a novel, unexpected repressive role of the super-elongation complex component Aff4 in G0-specific stalling. We propose a model wherein Pol II pausing restrains transcription to maintain G0, preconfigures gene networks required for the G0-G1 transition, and sets the timing of their transcriptional activation.


Introduction
In mammalian tissues, adult stem cells exist in a state of reversible arrest or quiescence.
This "out of cycle" or G 0 phase is characterized by the absence of DNA synthesis, highly condensed chromatin, and reduced metabolic, transcriptional and translational activity. By contrast to their terminally differentiated counterparts, quiescent cells retain the ability to reenter cell cycle. Indeed, it is this property of reversible arrest that is crucial for stem cell functions such as self-renewal and regeneration. Mounting evidence indicates that the transition into G 0 includes not only the induction of a specific quiescence program (Coller et al., 2006;Fukada et al., 2007;Liu et al., 2013;Subramaniam et al., 2013), but also suppression of alternate arrest programs such as differentiation, senescence and death (García-Prat et al., 2016;Sousa-Victor et al., 2014). Recent studies highlight the role of metabolic adaptations, epigenetic regulation and signaling mechanisms that enable cells to establish and/or maintain quiescence (reviewed in (Cheung and Rando, 2013)). Collectively, these insights have promoted the view that rather than a passive decline due to the absence of nutrients, mitogens or signaling cues, the quiescent state is actively maintained. Parallels with G 0 in yeasts (Martinez et al., 2004;Sajiki et al., 2009;Yanagida, 2009) have shown that the quiescence program is evolutionarily ancient (reviewed in (Dhawan and Laxman, 2015)).
Among the adult stem cell populations that are amenable to comprehensive analysis, the skeletal muscle stem cell or satellite cell (Mauro, 1961) holds a special position as it is readily isolated, easily imaged and can be cultured along with its myofiber niche (Bischoff, 1990;Beauchamp et al., 2000;Siegel et al., 2009). However, there is a growing appreciation that perturbing the stem cell niche leads to activation and loss of hallmarks of quiescence, and that freshly isolated stem cells while not yet replicating, have triggered the G 0 -G 1 transition (Kami et al., 1995;Fukada et al., 2007;Pallafacchina et al., 2010;Zhang and Anderson, 2014;Fu et al., 2015). Thus, culture models employing C2C12 myoblasts that recapitulate the reversibly arrested stem cell state (Milasincic et al., 1996;Sachidanandan et al., 2002;Yoshida et al., 2013) are useful adjuncts for genome-wide analysis of G 0 as distinct from early G 1 (Sebastian et al., 2009;Subramaniam et al., 2013;Cheedipudi et al., 2015) that is inaccessible in vivo. Importantly, genes isolated on the basis of induction in G 0 myoblasts in culture mark muscle satellite cells in vivo (Sachidanandan et al., 2002;Doles and Olwin, 2015), strengthening their utility as a model of quiescent muscle stem cells. Gala et al,Pol II pausing in G 0 4 Several lines of evidence suggest that quiescent cells are held in readiness for cell cycle re-entry by mechanisms that keep the genome poised for action. For example there is an increase in epigenetic modifications (H4K20me3) that promote the formation of facultative heterochromatin and chromatin condensation, and trigger transcriptional repression (Evertts et al., 2013;Boonsanay et al., 2016), while other histone modifications help to maintain the quiescence program (Srivastava et al., 2010;Juan et al., 2011;Mousavi et al., 2012;Woodhouse et al., 2013;Liu et al., 2013). Chromatin regulators induced specifically in G 0 hold genes encoding key cell cycle regulators in a poised state (Cheedipudi et al., 2015) by preventing silencing. Importantly, epigenetic regulation in G 0 ensures that the repression of key lineage determinants such as MyoD is reversible, helping to sustain lineage memory (Sebastian et al., 2009).
There is little information on mechanisms that regulate the expression of quiescencespecific genes in the context of global transcriptional repression. The regulation of RNA polymerase II (Pol II) is an integral part of the transcription cycle, with cell type and cell statespecific control (reviewed in (Adelman and Lis, 2012;Puri et al., 2015)). Pol II in quiescent adult stem cells exhibits reduced activation (Freter et al., 2010), Mediator (MED1) is associated with maintenance of quiescence (Nakajima et al., 2013) and transcription factor complex II (TFII) proteins have been implicated in differentiation (Deato and Tjian, 2007;Malecova et al., 2016). However, while components of the Pol II complex contribute to cell state changes, Pol II regulation in reversible quiescence is poorly explored.
In recent years, the view that global regulation of Pol II occurs largely at the level of transcription initiation has undergone a major revision. Elongation control at the level of promoter-proximal Pol II pausing has been implicated in developmental, stress-induced genes, neuronal immediate early genes (IEG) and mitogen-induced activation of groups of genes (Adelman and Lis, 2012;Saha et al., 2011;Levine, 2011;Gaertner et al., 2012). Further, Pol II pausing allows genes to be poised for future expression (Kouzine et al., 2013), and appears to mark regulatory nodes prior to a change in developmental state, allowing coordinated changes in gene networks. The temporal dynamics of Pol II pausing during development have been investigated (Kouzine et al., 2013), but the role of Pol II pausing in regulating cell cycle state is not well understood. We hypothesized that withdrawal of cells into alternate arrested states might preconfigure different subsets of genes by Pol II pausing, and that these genes may play critical Gala et al,Pol II pausing in G 0 5 roles in the subsequent cell fates.
Here, we investigated the role of promoter proximal RNA Pol II pausing in regulating the quiescent state and maintaining G 0 cells in a primed condition that might enable cell fate transitions. Using ChIP-seq and RNA-seq, we elucidated Pol II occupancy and the extent of transcriptional repression during different types of cell cycle arrest (reversible vs. irreversible), thereby identifying poised transcriptional networks characteristic of G 0 . We used knockdown analysis of G 0 -specific stalled genes identified by our analysis, as well as known regulators of Pol II activity, to delineate the role of pausing in quiescence-dependent functions. Our results suggest a model wherein promoter-proximal Pol II pausing regulates a specific class of repressed genes in G 0 and creates a state that is primed for cell cycle re-activation. We propose that Pol II stalling may preconfigure gene networks whose repression aids entry into and maintenance of G 0 , and confers appropriate timing to their transcriptional activation during the G 0 -G 1 transition and subsequent G 1 -S progression. Gala et al,Pol II pausing in G 0 6

Altered RNA metabolism in quiescent cells: reduced RNA content and low RNA synthesis
Global transcription mediated by all three RNA polymerases is dynamic across the cell cycle (Yonaha et al., 1995;White et al., 1995;Russell and Zomerdijk, 2005), and dampened during cell cycle exit (Hannan et al., 2000;Scott et al., 2001;Russell and Zomerdijk, 2005). To determine if this suppression is common to cells entering alternate states of arrest, we measured total cellular RNA content of C2C12 skeletal myoblasts in different states: proliferating undifferentiated myoblasts (MB), reversibly arrested undifferentiated myoblasts (quiescent/G 0 ) and permanently arrested differentiated myotubes (MT). First, using quantitative in situ imaging of cells stained with SYTO RNA-select we found that total RNA content per cell was lower in G 0 than in MB (Figure 1a,b). Interestingly, despite exiting the cell cycle, MT showed no significant reduction in total RNA content, providing evidence that global transcriptional activity distinguishes quiescence from differentiation. A time course of quiescence reversal (30'-18 hr reactivation out of G 0 ) showed that a rapid, robust increase in RNA content occurred within 30' of cell cycle reentry (Figure 1b), and was sustained as cells enter S phase. Flow cytometry of cells stained simultaneously for DNA and RNA confirmed that cells with 2N DNA content in suspension-arrested cultures showed less than half the RNA content per cell than 2N cells in proliferating cultures, validating the results of quantitative imaging and distinguishing between G 0 and G 1 (Figure 1c). Flow cytometric analysis of DNA synthesis during reentry is shown for comparison ( Figure 1d): cells begin to enter S phase between 12-16 hr after replating. Taken together, the results show that a rapid rise in cellular RNA content occurs minutes after exit from quiescence (marking the G 0 -G 1 transition) and greatly precedes the G 1 -S transition.
To determine whether reduced steady state levels of RNA in G 0 result from reduced transcription, we measured active RNA synthesis using pulse-labeling with 5-ethynyl uridine (EU) (Jao and Salic, 2008) (Figure 1e,f). EU incorporation into newly synthesized RNA was strongly suppressed in G 0 compared to MB and sharply increased within 30' of activation ( Figure 1f) correlating with the rise in RNA content (Figure 1b), and indicating a rapid restoration of the transcriptional machinery concomitant with cell cycle reentry. To evaluate the transcriptional output of muscle stem cells in situ, we isolated single myofibers with associated satellite cells and determined their EU incorporation. We found that reversibly arrested satellite cells (SC) ex vivo already show robust RNA synthesis within 30' of isolation, while Gala et al,Pol II pausing in G 0 7 differentiated myofiber nuclei (MN) in the same sample showed low incorporation of EU ( Figure   1g, Figure 1-figure supplement 1a). This observation is consistent with evidence that SC on freshly isolated fibers have already exited from quiescence during myofiber isolation (Kami et al., 1995;Fukada et al., 2007;Pallafacchina et al., 2010;Zhang and Anderson, 2014;Fu et al., 2015;Zammit et al., 2006).
Taken together, the comparison of steady state RNA levels and active RNA synthesis per cell confirm that in G 0 , low global RNA levels result directly from depressed transcriptional activity, which is rapidly reversed in a transcriptional burst within minutes of cell cycle reactivation. Further, since total RNA levels are maintained in irreversibly arrested myotubes, strong repression of RNA biogenesis is not a general function of cell cycle cessation, but rather, a feature of a specific quiescence program.

Identifying genes regulated in quiescence using cell number based normalization methods
Previous estimates of altered gene expression in quiescence using differential display (Sachidanandan et al., 2002) in myoblasts, or comparative microarray analysis in a variety of cell types (Venezia et al., 2004;Coller et al., 2006) including myoblasts (Subramaniam et al., 2013) and satellite cells (Fukada et al., 2007;Pallafacchina et al., 2010;Liu et al., 2013) have identified quiescence-induced genes. However, all these studies were based on normalization to equivalent amounts of total RNA which can lead to misinterpretation of global gene expression data (Lovén et al., 2012), particularly when there is sharply varying cellular RNA content across compared samples (as documented above). To quantitatively re-investigate the transcriptome of quiescent myoblasts in contrast to proliferative or differentiated cells, we used RNA-seq analysis, where gene expression was measured by normalization to equal cell number and not equal RNA (enabled by ERCC spike-in analysis as described in Materials and Methods). Post cell number normalization, the similarities between replicates and dissimilarities between samples were computed using Euclidean Distance (Figure 1-figure supplement 1b), confirming that no bias was generated by data processing.
To assess the extent of effects of normalization methods (equal RNA vs. equal cell number) on biological interpretation, we compared gene lists identified as differentially expressed between MB and G 0 derived from the same RNA-seq dataset but using the two Gala et al,Pol II pausing in G 0 8 different normalization methods. The Venn diagram (Figure 1h) shows that equal RNA normalization over-represented up-regulated genes compared to the equal cell number normalization (1583 vs. 497) and greatly under-represented down-regulated gene numbers (1564 vs. 9449) (Figure 1 -source data 1). Thus, quiescence is characterized by repression of a much larger number of transcripts than previously appreciated.

Repressed RNA biogenesis pathways distinguish quiescent from differentiated cells
To identify distinguishing features between two mitotically inactive states (reversibly arrested G 0 cells and irreversibly arrested MT), we analyzed gene ontology (GO) terms of differentially regulated genes derived from equal cell number normalization, using gProfiler tools. Consistent with our earlier microarray study (Subramaniam et al., 2013), compared to MB, genes identified as up-regulated in G 0 are distinct from those in MT (Figure 1-figure supplement 3). GO terms representing response to external stimuli, stress and extracellular matrix organization are over-represented in G 0 , whereas skeletal muscle contraction and ion transport are highlighted in MT. These contrasting terms reflect the suppression of myogenic differentiation in G 0 myoblasts. Expectedly, the down-regulated genes identified from both mitotically arrested states were enriched for cell cycle processes, DNA replication and cell proliferation ( Figure 1-figure supplement 4). Commonly down-regulated gene families that were repressed in both G 0 and MT included replication-dependent histone transcripts bearing a 3' stem-loop and lacking a polyA tail (Figure 1i). Slbp which binds the 3' stem-loop in polyA histone mRNAs regulating their stability and translation during S phase (Whitfield et al., 2000), was also repressed in both arrested states ( Figure 1i).
We focused on the ontologies that were uniquely down-regulated in G 0 but not in MT, which intriguingly, involved RNA biogenesis itself (tRNA and rRNAs), mitochondria & electron transport chain, proteasome and carbon metabolism. Thus, reversible arrest is uniquely associated with suppression of RNA metabolism, protein turnover and bioenergetics (Figure 1figure supplement 4). Further, GO terms involved in stem cell proliferation, epithelial differentiation, neurogenesis and cellular growth, were exclusively down-regulated in MT, consistent with the repression of lineage-inappropriate genes in terminally differentiated myotubes (Ali et al., 2008), and retention of progenitor features and plasticity in quiescent myoblasts. Indeed, as has been previously reported, while the satellite cell specification factor Gala et al,Pol II pausing in G 0 9 Pax7 was induced in G 0 and repressed in MT, the lineage determinant MyoD was suppressed in G 0 and induced in MT.
Further, genes identified here in our cultured quiescence model as differentially expressed (quiescent vs. cycling) were corroborated by comparison to genes reported as differentially expressed in freshly isolated quiescent satellite cells (QSCs) vs. activated satellite cells (ASCs) (Liu et al., 2013) (Figure 1-figure supplement 5-a, b). More than 98% of genes expressed in ASCs at 60 hr were expressed in proliferating MB and not in G 0 , consistent with the derivation of C2C12 cells from adult SCs and confirming the global similarity in expression profiles between this cultured cell line and SCs. By contrast with near identity of the ASCspecific gene set, only 17% of the "QSC-specific" gene set of freshly isolated SC were also identified as up-regulated in G 0 in culture. This observation suggests either that G 0 in vivo employs markedly different programs than in vitro, or that the reported "QSC-specific" gene set defines cells that are no longer in G 0 . Taken together with the strong RNA synthesis seen in freshly isolated SC (Figure 1g), this comparative expression data strongly supports the view that disturbance of the SC niche during fiber isolation activates a departure from G 0 . We conclude that the reported "QSC-specific" gene set likely reflects an early-activated cell state, which while still clearly pre-replicative, does not represent undisturbed quiescence.

The RNA Pol II transcriptional machinery is specifically altered in G 0
Since we observed strong repression of total RNA levels, reduced RNA synthesis and repression of a large class of Pol II-transcribed genes in G 0 , we hypothesized that Pol II expression itself may also be repressed, to control transcription globally. Indeed, RPB1 the largest subunit of RNA Pol II (N20 pAb) showed decreased levels in both G 0 and MT (Figure 2a levels in both mitotically inactive states, MT exhibit a nuclear localized enzyme which is   transcriptionally active, whereas G 0 cells show predominantly cytoplasmically localized Pol II, and is transcriptionally less active, consistent with reduced EU incorporation.

Identification of stage-specific Pol II recruitment in myogenic cells
To identify stage-specific programs of gene activation we used Pol II enrichment profiling. We first validated the N20 antibody using targeted ChIP analysis of the Rgs2 locus (called as G 0 -induced by RNA-seq (Figure 1 -source data 1) and by microarray analysis (Subramaniam et al, 2013)), interrogating enrichment at 4 locations across the gene using qPCR  (Adelman and Lis, 2012) where initiated Pol II may pause, and represents a checkpoint for transcription, presaging stage-specific regulation of elongation.
To compare global cell state specific variations in Pol II recruitment ,we first categorized non-overlapping genes in decile groups based on enrichment score, separately for each of the three conditions, and further plotted the normalized enrichment of RNA Pol II around the TSS ( Figure 2f). In all cases, enrichment peaked within 30-100 bases downstream of the TSS, consistent with initiated but paused polymerase complexes. Strikingly, the recruitment of Pol II on promoters in G 0 cells for the >90th percentile Pol II-enriched genes was significantly higher than that of either MB or MT, which were comparable. By contrast, profiles for the 70-90 percentile sub-groups behaved similarly for all 3 cellular states. Thus, a small group of promoters Gala et al,Pol II pausing in G 0 11 in G 0 are occupied with much higher abundance of Pol II than other genes, and may represent a G 0 -specific class of genes, with elevated ongoing Pol II recruitment. Since promoter occupancy is observed not only on genes that are highly expressed ("active"), but also on genes that are not expressed and exhibit paused polymerases ("poised"), we used the extent of promoter clearance to distinguish these classes.

Calculation of polymerase stalling index (SI): more genes are stalled in G 0 and MT than MB
Promoter clearance can be quantified using the distribution of Pol II abundance across an entire locus. The ratio of Pol II enrichment at the promoter region to that across the gene body region is defined as the Pausing Index or Stalling Index (SI), where higher SI reflects a lower promoter clearance of Pol II and indicates a point of regulation (Zeitlinger et al., 2007;Min et al., 2011;Danko et al., 2013). All genes with annotated gene length >2500 bases were used for further analysis. Genes overlapping with neighboring genes were removed from the analysis. Pol II density is calculated by base pair enrichment in ChIP-seq data, at promoters (-500 to +300 relative to TSS) and gene body (+300 to +2500).
To confirm that the SI is a function of specific enrichment of Pol II, we calculated stalling indices similarly for control input samples: each IP sample exhibited significantly more enrichment/stalling than that of its respective input ( Interestingly, >50% of analyzed genes display a SI >1 in any state, confirming that clearance of engaged Pol II from the promoter is a significant common regulatory mechanism. Comparative GO analysis for cell state-specific stalled genes (SI>2.5 in MB, MT and G 0 ) shows that metabolic and stress response genes are stalled in all three states (Figure 2-figure supplement 6). Strikingly, the proportion of genes that are stalled in the two non-dividing states (G 0 and MT) is higher than that in proliferating cells, suggesting that Pol II stalling represents a point of regulation associated with mitotic inactivity. Thus, halted cell division (either due to quiescence or differentiation) is associated with higher rates of polymerase pausing.

MT
The histone genes transcribed by Pol II were excluded from stalling index calculation, as they do not meet the length criteria set for the analysis. This important gene family comprises ~80 genes that map to chr13 and chr3 in clusters, with a few genes singly located on other chromosomes. To evaluate the relative transcriptional engagement of histone loci, the read density of Pol II across the entire gene region was used. Notably, strong enrichment of Pol II was observed in both MB and G 0 but not in MT (Figure 2h, Figure 2-figure supplement 7), indicating that the histone loci are transcriptionally silent in terminal differentiation, but not in reversible arrest. Thus, RNA Pol II occupancy on histone gene loci is indicative of transcriptional activity but the low mRNA levels in G 0 (Figure 1i) likely reflect rapid post-transcriptional turnover, and the sustained Pol II engagement across the entire histone gene family represents a primed gene network preconfigured for the return to proliferation.

G 0 -stalled genes are specifically repressed in quiescence
Since promoter proximal stalling influences the transcriptional output from a given locus, the repertoire of genes regulated by this mechanism in a given state would yield insights into global Pol II-mediated regulation of that state. Our hypothesis was that control of quiescence might include uniquely stalled genes. Therefore, we analyzed the group of 727 genes that show strong stalling (SI > 2.5) in G 0 (hereafter referred to as 'G 0 -stalled genes' (Figure 2 -source data 1)). The cumulative distribution of SI for G 0 -stalled genes was compared across the three states (SI>2.5 in G 0 is shown in Figure 3a). Genes that were strongly stalled in G 0 (high SI) display much lower SI in MB and MT states, indicating that they experience enhanced stalling specifically in quiescence ( Expression levels enumerated by RNA-seq for G 0 -stalled genes showed that ~95% of these genes were specifically repressed in G 0 (Figure 3b, Figure 3-figure supplement 2). We conclude that most of the 727 genes show very low transcriptional activity in G 0 but continue to display high promoter proximal Pol II occupancy, indicative of a poised or primed state. Most of these genes are also repressed in MT but do not exhibit stalled Pol II, indicating that other mechanisms are responsible for their reduced expression in differentiated cells. Polymerase pausing has previously been associated with upstream divergent transcription (Flynn et al., 2011). Therefore, we evaluated the level of antisense RNA associated with promoters of genes categorized as G 0 -stalled, G 0 up-regulated (compared to MB) or a randomly selected group containing an equal number of genes. Interestingly, upstream antisense transcription was Gala et al,Pol II pausing in G 0 13 observed only for G0-stalled genes ( Figure 3c). Gene ontology-based clustering using DAVID revealed distinct terms encompassing G 0 -stalled genes ( Figure 3d). These include Pol IItranscribed genes involved in ribosomal machinery, mRNA biogenesis & RNA processing, as well as mitochondria-related genes. We conclude that RNA-pol II pausing is associated with high rates of antisense transcription, and distinguishes specific gene groups in G0.
Genes that display stalled polymerases in G 0 are poised for reactivation during G 1 Considering that quiescent cells display reduced levels of total RNA, the enrichment of GO terms relating to ribosomal machinery and RNA processing genes ( Fig.1) led us to hypothesize that the genes regulated by RNA Pol II pausing might be key nodes for global control of RNA metabolism. To validate this promoter proximal stalling, we chose genes involved in RNA regulation which includes genes involved in rRNA biogenesis (nucleolin -Ncl) and ribosomal protein S24 -Rps24), Pol II regulation (Pol II recycling enzyme CTD phosphatase -Ctdp1), and post-transcriptional regulators (stem loop binding protein -Slbp, that controls histone mRNA stability, and Zinc Finger CCCH-Type Containing 12A -Zc3h12A), an RNase that modulates miRNA levels). Enhanced stalling is confirmed as depicted on genome browser snapshots of Pol II occupancy on these genes ( Since these genes display active histone marks but are not actively expressed in G0 (Figure 3-figure supplement 7), we hypothesized that RNA Pol II stalling in G 0 may poise genes for subsequent activation. Comparable RNA-seq data was not available for reactivation, so we used q-RTPCR to evaluate induction of specific genes during cell cycle reentry. Within the first 30' of reactivation, we observed a sharp increase in normalized mRNA levels (referenced to G 0 ) of the immediate early gene Fos, a hallmark of the G 0 -G 1 transition (Kami et al., 1995). Id2 (stalled but active in G 0 ) was further induced during re-entry. By contrast, the G 0 -stalled active gene Rgs2 showed a rapid decrease in expression during re-entry. Interestingly, only the G 0stalled, repressed genes (Ncl, Rps24, Ctdp1, Zc3h12A, Slbp) showed enhanced expression (Figure 3e) whereas genes repressed but not stalled (MyoD, L7, and YPLE5) failed to show rapid activation upon cell cycle re-entry (Figure 3-figure supplement 8). Further, the increased level of transcripts associated with exit from quiescence reflects in increased protein levels (Ncl Gala et al,Pol II pausing in G 0 14 and Rps24; Figure 3-figure supplement 9 and 10) in both C2C12 model and primary satellite cells. These data confirm that G 0 -stalled, repressed genes are functionally activated during the G 0 -G 1 transition. Together, these patterns of expression suggest that while these G 0 -stalled genes are differentially regulated at later stages of the cell cycle, their common early reactivation is consistent with the hypothesis that reversal of stalling contributes to their coordinate restoration to transcriptional competence.

Transcriptional induction during G 0 -G 1 is associated with reduced stalling
Since expression of most G 0 -stalled genes was low in quiescence and induced during cell cycle reentry, one might expect a corresponding decrease in the proportion of promoter-proximal stalled Pol II, or an increase in the proportion of elongating Pol II (gene body). We investigated potential changes of Pol II occupancy from G 0 to early G 1 (30'-2 hr, the window of induced expression), by computing SI for each candidate gene, using targeted ChIP-qPCR. Consistent with their identification by ChIP-seq as G 0 -stalled, the SI calculated by this targeted analysis was also highest in G 0 , and decreased within 2 hr of exit from quiescence, indicative of increased promoter clearance as cells enter G 1 (Figure 3f). The IEG Fos, that exhibits strong stalling in G 0 , registered a rapid drop in SI in the first 30' after activation, at a time when its mRNA expression is high (Figure 3e), consistent with a rapid increase in promoter clearance. All the selected G 0stalled repressed genes showed similar trends. Notably, the Rgs2 and Id2 genes (G 0 stalled but active) showed the opposite trend: the SI increased as cells exited quiescence, suggesting altered regulation at the level of pausing for repressed vs. active genes. Collectively, this analysis confirms that activation of G 0 stalled genes is accompanied by reduced stalling during the early G 1 transcriptional burst.

Alterations in RNA Pol II phosphorylation during the G 0 -G 1 transition
To analyze whether changes in global transcriptional activity of Pol II across cell states are associated with global alteration in its phosphorylation status, we used western blot analysis for total Pol II and its actively transcribing phosphorylated forms (Ser5-p and Ser2-p; Figure 3g), and calculated the ratio of phospho-Pol II to total Pol II in each state (Figure 3g). At R30' we observed a sharp increase in the proportion of Ser5-p-and Ser2-p-modified polymerase, indicating a rapid rise in transcriptional competence compared to G 0 & correlating well with the Gala et al,Pol II pausing in G 0 15 rapid rise of total RNA and active synthesis shown in Figure 1b and Figure 1f, which stabilizes in 2 hr.
Taken together, we conclude that these repressed G 0 -stalled genes undergo release of Pol II stalling upon exit from quiescence that correlates with their rapid induction during this transition. Further, the data so far suggest that the release of Pol II stalling is associated with exit from quiescence, and this mode of regulation facilitates cell cycle re-entry by controlling G 0stalled genes encoding regulators of the protein synthesis machinery, mRNA biogenesis and transcriptional control.

RNA Pol II stalling mediates repression that controls the quiescent state
At the level of gene expression, the functional consequences of Pol II stalling in G 0 are repression of transcription (resulting in depressed cellular activity), while stalled loci are poised for activation in response to appropriate signals. Our observations thus far identified several G 0specific stalled genes with the appropriate functions to participate in the biology of quiescence and activation, but whether Pol II stalling is directly responsible for establishment or maintenance of the quiescent state has not been demonstrated. Since Pol II pausing on G 0 -stalled genes results in lower mRNA production from the stalled transcription unit, we directly decreased the expression level of individual G 0 -stalled genes in proliferating MB and analyzed the functional consequences for entry into quiescence. We used siRNA-mediated knockdown of the selected G 0 -stalled genes (Ctdp1, Ncl, Rps24, Slbp, Zc3h12A) in cycling MB ( We directly tested the onset of quiescence in cells knocked down for G 0 -stalled genes, by evaluating expression of quiescence specific markers p27 (cyclin-dependent kinase inhibitor 1B) (Oki et al., 2014 ) and p130 (Rb family tumor suppressor) (Carnac et al., 2000;Litovchick et al., 2004) ( observations suggest that repression of these G 0 -stalled genes is essential for quiescence, since forced reduction of their expression in proliferating cells leads to induction of markers of G0, indicative of a quiescence-like state.

G 0 -stalled genes contribute to self-renewal
To investigate the role of these G 0 stalled genes in self-renewal, we evaluated the effect of knockdown on colony formation after exit from quiescence. G 0 cells display enhanced selfrenewal compared to cycling cells, supporting the view that the quiescence program includes a self-renewal module (Collins et al., 2007;Subramaniam et al., 2013). Knockdown of the same three genes that affected RNA synthesis and G 1 blockade in proliferating cells (Ctdp1, Ncl, Rps24) also showed a decrease in colony forming ability compared with control cells (Figure 4f, . Since equal numbers of viable cells were plated for the CFU assay, the effect of knockdown on genes that are normally induced very rapidly (30'-2 hr) would be to dampen their induction upon G 1 entry, accounting for the sharp decrease in colony formation. This finding further demonstrates the importance of timely reactivation of G 0 -stalled genes during the G 0 -G 1 transition, failure of which compromises self-renewal.
We conclude that repression of key cellular processes identified by analysis of G 0specific Pol II stalling is sufficient to cause cycling cells to acquire quiescence-like features.
Consistent with the pre-existing transcriptional repression of G 0 -stalled genes, further repression by siRNA knockdown does not appear to affect the quiescent state itself, but failure to de-repress their expression upon exit (normally mediated by release of RNA Pol II stalling) results in attenuated cell cycle re-entry and diminished self-renewal. Collectively, our results demonstrate that Pol II undergoes a stage-specific regulation to stall on key genes that are required to poise quiescent cells for efficient cell cycle re-entry.

Aff4, a component of the Pol II super-elongation complex regulates G 0 -stalled genes
Release of paused Pol II is mediated by promoter engagement of P-TEFb, a complex comprised of CDK9 and cyclin T. P-TEFb availability and recruitment is regulated by Hexim1, Aff4, and Brd4, with direct consequences on Pol II elongation (Adelman and Lis, 2012;Jonkers and Lis, 2015). To determine if these key regulators of elongation affected expression of G 0stalled genes, we used siRNA treatment (in G 0 ) ( Since G 0 -stalled genes were induced by release of pausing in G 1 , we hypothesized that the up-regulation of these genes in Aff4 knockdown cells might prime cells for accelerated cell cycle progression. Indeed, EdU incorporation during cell cycle reentry (6 hr and 12 hr after G 0 exit) showed that only Aff4 depletion led to accelerated S phase kinetics (~22% EdU + at 12hr To test if Aff4 acts directly on promoters of G 0 stalled genes to regulate their timely expression during exit from quiescence, we examined the occupancy of Aff4 on the promoters of G 0 -stalled genes using ChIP ( Figure 5 f). During quiescence, Aff4 occupancy was indeed located at promoters of G 0 -stalled genes (Ncl and Rps24) but not on a G 0 expressed genes (Rgs2), which is consistent with a direct role for Aff4. Further, these promoters continue to be bound by Aff4 as cells exit quiescence. Taken together, these results indicate that in quiescence, the SEC component Aff4 unexpectedly acts as repressor of G 0 -stalled genes, thereby promoting stalling and restraining exit from quiescence, in spite of activating chromatin marks.
Collectively, our results demonstrate that in reversibly arrested myoblasts, Pol II stalling contributes to repression of a set of genes that are not thus marked in permanently arrested myotubes, and that primed reactivation of these genes is critical for reversal of quiescence and for the timing of G 1 -S progression ( Figure 6).

Discussion
In this study, we highlight the role of Pol II pausing in the maintenance and exit from the quiescent self-renewing state. While cell cycle exit is known to be associated with reduced transcriptional activity, we show here that this is not uniformly true for all mitotically inactive states: reversible arrest, which is typical of adult stem cells, elaborates global transcriptional control mechanisms distinct from terminally arrested differentiated cells. Consequently, Pol II localization and expression levels, as well as the extent of repression of global RNA content and synthesis differ quantitatively between reversibly and irreversibly arrested cells. In particular, genes regulating RNA biogenesis are among the most strongly repressed in quiescent but not differentiated cells. Strikingly, promoter-proximal polymerase pausing revealed by ChIP-seq is enhanced in G 0 , genes encoding regulators of RNA biogenesis experience prominent Pol II pausing, and release of pausing on these genes is critical for cell cycle re-entry and self-renewal.
Finally, we identify the SEC component Aff4 as a key regulator of quiescence-specific Pol II stalled genes, that also unexpectedly restrains cell cycle re-activation. We conclude that Pol II stalling mediates transcriptional repression that controls quiescence, and that G 0 -stalled genes contribute to competence for G 1 entry and self-renewal ( Figure 6). Given the similarities of gene expression between myoblasts in culture and satellite cells ex vivo, we suggest that similar regulatory mechanisms may also function in muscle stem cells.

Regulation of the RNA Pol II transcriptional machinery distinguishes reversible and irreversible arrest
Reduced transcriptional output is typical of cells whose proliferative activity has slowed, and has been documented in many systems (Bertoli et al., 2013) including quiescent adult stem cells (Freter et al., 2010). Here, we explored the differences in Pol II activity as myoblasts withdrew into alternate arrested states -G 0 or differentiation-and found several notable differences. First, while the expression of total Pol II was strongly inhibited in both MT and in G 0 , the cytoplasmic localization of Pol II was increased in quiescent cells rather than predominantly nuclear as in MT. Second, the levels of active phospho-Pol II (both Ser5p indicative of initiated Pol II and Ser2p indicative of elongating Pol II) were lower in G 0 than in MT, suggesting quiescence-dependent regulation at the level of signaling to the CTD kinases/phosphatases. Finally, consistent with the severely depressed total RNA levels in G 0 Gala et al, Pol II pausing in G 0 20 compared to either MB or MT, the extent of active RNA synthesis (EU incorporation) was much lower in G 0 than in MB, but rapidly reversed in response to cell cycle reactivation. Importantly, active RNA synthesis differed in muscle cells ex vivo-analysis of freshly isolated skeletal myofibers showed enhanced EU incorporation in activated SC nuclei compared to the differentiated myonuclei. Together, these findings suggest that not only are quiescent cells in culture marked by specific mechanisms that control RNA biogenesis, but also that analysis of RNA synthesis may be useful in distinguishing global cellular states in vivo.

Revisiting the quiescent transcriptome using RNA-seq reveals major differences in RNA biogenesis compared to permanent arrest
Several studies have reported global gene expression profiles of quiescent cells and revealed a core quiescence signature as well as cell type-specific programs (Fukada et al., 2007;Liu et al., 2013;Hausburg et al., 2015). All studies along these lines so far have been comparative analysis using microarrays, where the experimental design forces the use of cRNA derived from equal total RNA. However, our study clearly shows that the altered RNA metabolism in G 0 myoblasts leads not only to lower RNA synthesis, but also strongly reduced total RNA content. The marked reduction in total RNA content led us to re-evaluate the quiescent transcriptome using RNA-seq: by normalizing transcript levels to cell number and not equal RNA, we reveal a revised picture of the quiescent transcriptome. A major new insight offered by this analysis is the strong repression of a much larger number of genes than previously appreciated, bringing the repression of RNA biogenesis into sharp focus. Interestingly, these pathways are not repressed in differentiated cells, consistent with the observations that Pol II regulation distinguishes two mitotically inactive states, leading to the elaboration of distinct global controls.
The transcriptome studies carried out here also show a distinct quiescence signature in

Poising of the replication-dependent histone gene network in reversible arrest
A key new finding from our study concerns the differential transcriptional competence of the histone genes in arrested cells. Although replication-dependent histone mRNA expression is tightly S-phase linked, histone genes are transcribed constitutively, while mRNA processing and degradation is regulated by Slbp (Kaygun and Marzluff, 2005). We now show that while Slbp expression is repressed in both arrested states, Pol II is retained on histone loci only in G 0 and not MT. Transcriptional control of Slbp has not been previously reported, and Pol II pausing on this regulator might be the pivot for priming the entire histone gene family. Since the replicationdependent histone loci while stalled and repressed, remain primed during quiescence, and stalling of Pol II on the Slbp gene is specific to G 0 , our finding provides a new node for reversible control of replication in G 0 that is silenced in MT.
Further, the expression of histone variants implicated in marking transcriptionally active or poised regions in the genome (H3f3a and H3f3b (Tang et al., 2013)) is unchanged in quiescence but repressed in MT. These variants (Yuen and Knoepfler, 2013) are translated from polyA + transcripts and are required for normal development and growth control (Tang et al., 2013). It is tempting to speculate that expression of specific replication-independent histones is critical for quiescence, and that the condensed chromatin structure (Evertts et al., 2013) and reduced nuclear size peculiar to G 0 cells (Figure 1e, 4c) may require input of specific histone isoforms.

A new quiescence-specific regulatory node: increased global Pol II pausing
Accumulating evidence points to quiescence as a poised state where diverse regulatory mechanism at the level of chromatin (Liu 2013; Cheedipudi, 2015), mRNA mobilization (Crist, 2013), and signaling (Rodgers, 2014) work to prime G 0 cells for cell cycle reactivation.
Extending this concept, our study addresses the hypothesis that genes important for cycling cells are kept poised by Pol II pausing, permissive for future expression. Indeed, we identify a distinct class of genes that are regulated by pausing, and allow quiescent cells to boost the activity of the core cellular machinery rapidly after activation. It is quite likely that Pol II pausing in G0 may enable the retention of open chromatin around key (inactive) promoters during quiescence, permitting their activation during cell cycle reentry. Genes exhibiting quiescence-specific stalling are also transcriptionally repressed specifically in G 0 , suggesting a functional relationship between promoter-proximal pausing and low transcriptional output in this state. As this repressed gene set includes RNA metabolism and regulatory proteins, high in the hierarchy of cellular control, Pol II mediated repression of these genes would lead to a reduction in overall cellular throughput, contributing to quiescence. Further, ontologies that include Ca + channels and neuro-muscular signaling were found to be specific for myotubes, suggesting that Pol II stalling in differentiated cells may control responses to neuronal stimulation. Overall, our finding of state-specific paused networks in myoblasts supports the emerging idea of Pol II elongation control as a major regulatory step in eukaryotic gene expression.

Quiescence-repressed genes are poised for activation in G 1
An important hallmark of quiescent cells concerns their extended kinetics of S-phase entry compared to that of a continually cycling population. This additional phase -the G 0 -G 1 transition-has long been appreciated to integrate extracellular signaling (Pledger et al., 1978), but the regulation of its duration is still incompletely understood (Coller, 2007). Our study suggests that Pol II pausing and its reversal may play a role in determining these kinetics. Unlike genes such as Rgs2 and Id2 that are up-regulated in quiescent cells, 95% of G 0 -stalled genes, are repressed. The activation of some G 0 -stalled genes preconfigures changes in cellular processes that are essential for cell cycle reentry. Transcriptional activation of this gene set is accompanied by loss of Pol II stalling, increased active phosphorylation marks (Ser2-p and Ser5-p), and a rapid burst in active RNA synthesis very early in G 1 . The timing of these events suggests that the increase in active Pol II marks is linked to the exit from quiescence. Taken together, we suggest that release of polymerase stalling and the burst of active transcriptional output during reversal of quiescence promotes cell cycle entry and fine-tunes the transcriptional cascade required for derepressing a variety of cellular processes essential for progression from G 0 to G 1 to S.

Repression of G 0 -stalled genes contributes to blocked proliferation and sustained cell potency
Our studies support the hypothesis that transcriptional repression of G 0 -stalled genes contributes to quiescence entry since direct knockdown of these genes in proliferating cells leads to arrest and compromised self-renewal. Perturbation of ribosomal biogenesis genes (Ncl and Rps24) and the Pol II recycling phosphatase (Ctdp1) led to cell cycle arrest of cycling myoblasts, consistent with the findings that these genes are anti-tumorgenic in other systems (Ugrinova et al., 2007;Badhai et al., 2009;Zhong et al., 2016). Interestingly, Ctdp1/FCP1, which controls the restoration of active Pol II during the transcription cycle, was earlier identified in a genetic screen for quiescence-regulatory genes in yeast (Sajiki et al., 2009). Further, in mammalian cells, Ctdp1 has been implicated in controlling global levels of polyA transcripts as well as in rapid induction of heat shock genes (Kobor et al., 1999;Cho et al., 2001;Mandal et al., 2002;Fuda et al., 2012), both functions ascribed to regulatory functions of Pol II mediated transcription. Thus, stalling control of the Ctdp1 gene itself and its reduced expression in G 0 , may suggest a feedforward mechanism by which RNA Pol II activity is further depressed in G 0 .
The ribosomal machinery genes, Ncl and Rps24 are involved in the very first processing step that generates pre-rRNA (Ginisty et al., 1998;Choesmel et al., 2008). Rps24 mutations result in Diamond-Blackfan Anemia characterized by ribosome biogenesis defects in HSCs and erythroid progenitors (Choesmel et al., 2008;Song et al., 2014). Stalling of genes involved in ribosomal RNA processing in G 0 suggests an upstream blockade on ribosome biogenesis, and by extension, the translational machinery. Thus, G 0 -specific Pol II pausing participates in keeping the major biosynthetic pathways in check to maintain the quiescent state.

The SEC component Aff4 regulates G 0 -stalled genes to control the timing of S-phase entry
Quiescence-specific regulation of Pol II has not been previously reported. Regulators of stalling such as DSIF and NELF are widely involved and likely to participate in all states.
Therefore, we investigated specific components of the P-TEFb regulatory system complex as good candidates for the release of stalling of specific target genes during G 1 (Lu et al., 2016).
Release of stalled polymerase is brought about by coordinated regulation of P-TEFb mediated by SEC, Brd4, Hexim1 (Lu et al., 2016;Puri et al., 2015). Hexim1 haplo-deficiency in mice Gala et al,Pol II pausing in G 0 24 increases satellite cell activity and muscle regeneration, highlighting a role in self-renewal (Galatioto et al., 2010;Hong et al., 2012). While Hexim1 knockdown did not affect the timing of cell cycle re-entry, an effect in quiescence cannot be ruled out and the repression of some G 0stalled genes warrants further investigation. The pause-release regulator that had the most interesting phenotype in our study was Aff4. As a SEC component, Aff4 is known to display target gene specificity (Luo et al., 2012), promoting release of paused polymerase on heat shock genes and MYC (Kühl and Rensing, 2000;Luo et al., 2012;Schnerch et al., 2012). By contrast, we identified a new repressive role for Aff4 in G 0, such that knockdown led to enhanced expression of G 0 -stalled genes, and accelerated exit from quiescence. Therefore, it is tempting to speculate that Aff4 mediates differential regulation (activation and repression) of stalling on different target genes, thereby regulating cell cycle progression or arrest. Moreover, the accelerated S-phase entry in knockdown quiescent cells indicates that Aff4 guards against premature activation of quiescent cells, thereby regulating the timing of cell state transitions.
In conclusion, by combining a genome-wide analysis of Pol II occupancy with its functional transcriptional output in distinct mitotically arrested states, we uncover a repressive role for Pol II pausing in the control of RNA metabolic genes that is central to the entry into quiescence. We also define a new regulatory node where the SEC component Aff4 acts to restrain expression of these genes in G 0 to set the timing of their activation during G 1 entry and affects S-phase progression. Overall, this study extends the role of Pol II pausing in regulating not only developmental lineage transitions (Scheidegger and Nechaev, 2016), but also distinct cell cycle states and in particular, sheds light on the type of arrest typical of adult stem cells. Burgering and support from Utrecht for exchange visits between JD and BB labs.

Author contributions
HG and JD designed research, interpreted data and wrote the paper; HG performed research, acquired and analysed data, and wrote custom scripts for NGS analysis and quantitative image analysis; DS, NV and AA performed research. The authors declare no conflict of interest.

Materials and Methods
Cell Culture. C2C12 myoblasts (MB) were obtained from H. Blau (Stanford University) and a subclone A2 (Sachidanandan et al., 2002) were used in all experiments. Myoblasts were maintained in growth medium (GM; DMEM supplemented with 20% FBS and antibiotics) and passaged at 70-80% confluency. Differentiation. Myotube (MT) differentiation was induced in proliferating cultures at 80% confluence after washing with PBS and incubation in differentiation medium (DM: DMEM with 2% horse serum), replaced daily for 5 days. Multinucleated myotubes appear after 24 h in DM. Quiescence induction: G 0 synchronization by suspension culture of myoblasts was as described (Sachidanandan et al., 2002). Briefly, subconfluent proliferating MB cultures were trypsinized and cultured as a single-cell suspension at a density of 10 5 cells/mL in semisolid media (DMEM containing 1.3% methyl cellulose, 20% FBS, 10 mM HEPES, and antibiotics). After 48 h, when ~98% of cells have entered G 0 , arrested cells were harvested by dilution of methyl cellulose media with PBS and centrifugation. G 0 cells were reactivated into the cell cycle by replating at a sub-confluent density in GM and harvested at defined times (30` -24hr) after activation. Upon replating, G 0 cells undergo a G 0 -G 1 transition to enter G 1 ~6 h after reattachment; S-phase peaks at 20-24 h.

Muscle fiber isolation and culture
Animal work was conducted in the NCBS/inStem Animal Care and Resource Center and in the CCMB Animal Facility. All procedures were approved by the inStem and CCMB Institutional Animal Ethics Committees following norms specified by the Committee for the Purpose of Control and Supervision of Experiments on Animals, Govt. of India.
For quantification of active RNA synthesis, 10 µM EU was added to media. Freshly isolated or cultured fibers were analyzed by immuno-staining. Single fibers were fixed in 4% PFA for 5min, washed 3X with PBS, picked and placed on charged slides (Thermo-Fisher) for immuno-staining.

Knockdown of target genes using RNAi
C2C12 cells were cultured in growth medium until 80% confluent. The cells were then trypsinized and ~0.3 X 10 6 cells were plated on 100 mm tissue culture dishes. Approximately 12-14hr post plating, the cells were transfected with siRNA (400 picomoles of siRNA with 40 µl lipid) using Lipofectamine RNAiMAX (Invitrogen) as per manufacturer's instructions. The cells were incubated with the RNA-Lipid complex for at least 18-24hr following which they were tested to evaluate the extent of knockdown by qRTPCR and western blot and then used for further experimental analysis. Typically, siRNA-mediated knockdown was in the range of 50-90% at RNA level. siRNA used in this study are detailed in Table S3.

Immunofluorescence and microscopy
Cells plated on cover slips or harvested from suspension cultures were washed with PBS, fixed in 2% paraformaldehyde at room temperature, and permeabilized in PBS + 0.2% TritonX100. Primary antibodies (Table S4) (Carpenter et al., 2006;Kamentsky et al., 2011).

Cell cycle analysis using flow cytometry
Adherent cells were trypsinized, washed in PBS and pelleted by centrifugation.
Suspension-arrested cells were recovered from methyl-cellulose by dilution with PBS followed

Analysis of p27 induction:
A C2C12 line expressing p27mVenus (a fusion protein consisting of mVenus and a defective mutant of p27-CDKI, (p27K(-) (Oki et al., 2014)) was generated by stable transfection. Adherent myoblasts expressing p27-mVenus (excitation/emission= 515/528), +/-knockdown of various G0-stalled genes were trypsinized and fixed in 4% paraformaldehyde for 10 minutes at room temperature. Suspension cultured cells were recovered as described above. Following fixation, cells were analyzed for mVenus expression using a Gallios cytometer (Beckman Coulter). Kaluza or FlowJo software was used to acquire and analyze the data.

Western blot analysis
Soluble lysates of adherent cultures or suspension cells (2x10 6 ) were obtained after quantification of absorbance at 630 nm. Proteins were resolved on 8% Acrylamide gels. and transferred to PVDF membrane. The membrane was incubated with primary antibody overnight at 4°C, then washed in 1X TBS + 0.1% Tween for 10' each followed by incubation with secondary antibody conjugated with HRP (Horse radish peroxidase) for 1 hr. After a brief wash with TBS-T for 10', ECL western blotting detection reagent for HRP was used for chemiluminescent detection using gel documentation system (Vilber Lourmat). Details of antibodies and dilutions used in this study are given in Table S4.  For Equal cell number normalization: HTcount packages (Anders et al., 2015) was used to quantify reads per gene for each of the samples using reference annotation file. This was further normalized using DEseq2 package using custom R-scripts generated specifically for this normalization (Source Code File -RNA_Seq_DEseq_processing). The sizeFactor is estimated for the Group B counts of ERCC spike-in RNA mix since the group B genes must not vary across all samples of both mixes. The calculated sizeFactor for each sample using Group B as reference gene set is further used to normalize other genes in the respective library including those in the other ERCC spike in mix groups (Groups A, C, D). This approach ensures that the RNA-seq libraries are scaled to reference spike-in controls, and not to the sequencing depth of the library. For each of the sub groups A, C and D, we found that the mix sets 1 and 2, cluster as distinct groups and the levels of intensity within each of the mix sets is comparable. Also, the fold differences between the mix sets 1 and 2 were observed to be as expected as per the

Cross-comparison between normalization methods:
Overlap of gene identity for differentially expressed genes identified from both normalization methods (Equal RNA and equal cell number) (>2 log2 fold change and p-value<0.05) is represented using size adjusted Venn diagrams separately for both up-regulated and down-regulated genes ( Figure 1h).

Correlation between replicates and samples: Dissimilarity between the sample and replicates
were computed using the normalized gene count matrix to get the sample to sample distance using Euclidean Distance Method using DEseq packages. The distance matrix is further is clustered hierarchically and heatmap is plotted (

ChIP-seq
After quality assessment, the raw ChIP-seq reads were aligned using paired option using bowtie2 to mouse genome (NCBIM37/mm9).

RNA Pol II enrichment at TSS:
Only non-overlapping gene >2500 bases were used for analysis (except for histone genes panels). Using these criteria 15170 genes were used for analysis. The read density counts around TSS (+/-300) was enumerated for all samples. Independently, for each cell state (MB, MT, G 0 ), the TSS regions were sub group into greater than 90 percentile and 70-90 percentiles based on read density score sorted lowest to highest value. Thus for each cell state (both input and IP) the average distribution of read densities were plotted for the each subgroups (Figure 2 f).

Stalling index:
Only non-overlapping gene >1500 bases were used for analysis (except for histone genes panels). Using these criteria 16095 genes were used for analysis. To calculate RNA Pol II stalling index, the number of ChIP-seq reads at promoter region (-500 to +300 bp from TSS) and gene body (+300 to +2500 for genes longer than 2.kb and +300 to +1500 for genes between 1.5kb to 2.5kb) of each gene was enumerated using Seqmonk. Stalling index for all genes were plotted in Figure 2g. The statistical significance of RNA Pol II stalling index was assess using the Benjamini-Hochberg method corrected Fisher's exact tests and to control the false discovery cut off was (p-value<0.005) and Stalling index greater than 2.5 in G 0 IP samples were used in Figure 3a.     h. RNA Pol II is enriched on histone genes in G 0 but not in MT. Violin plots of the read density across 68 histone genes show specific enrichment in all three IP samples over the respective input samples. With respect to the corresponding input sample, enrichment is highest in MB (pvalue < 10e -12 ), retained at near equivalent levels in G 0 (p-value < 10e -11 ) and significant but much lower in MT (p-value < 10e -5 ) (Mann-Whitney test).  e. Quantitative RT-PCR analysis of G 0 -stalled genes in quiescence and during cell cycle re-entry.
The expression in reactivation time points is normalized to expression in G 0 , and mean relative expression level is plotted with increasing time as cells exit quiescence (n=3). Fos, a canonical IEG is transiently induced only at 30` after reactivation and rapidly repressed. G 0 -stalled but expressed genes Rgs2 (which is repressed during entry) and Id2 (which continues to express as cells exit quiescence). All G 0 -stalled and repressed genes (Ctdp1, Ncl, Rps24, Slbp & Zc3h12A) show induction within 30' after cell cycle activation, albeit with differential kinetics at later points during G 1 -S progression.
f. Stalling index of G 0 -stalled genes decreases during cell cycle reactivation and is associated with expression dynamics. Calculation of stalling index by ChIP-qPCR was done using ratio of ChIP enrichment for primers targeting TSS vs. gene body (+500-1000 bps) in each time point: G 0 , R30' R2h. Both IEG and G 0 -stalled genes show rapid decrease in stalling index at 30`. G 0expressed genes Rgs2 and Id2 show gradual increase (Rgs2) and no change (Id2) in stalling index respectively, correlating with expression dynamics (significance calculated using Students t-test * = p-value <0.05 and $ = p-values between 0.05-0.15).
g. Western blots for total RNA Pol II and active phosphorylated forms at G 0 , and 30'-2hr after reactivation. Densitometric analysis shows an increased proportion of ser2-p/ser5-p modification at R30 and R2 compared to G 0 , correlating with activation. Values represent mean + SD, n=3, *p-value<0.05.  f. Aff4 controls the expression of G 0 -stalled genes during quiescence and reactivation. Aff4 occupancy on promoter of G 0 -stalled genes (Ncl and Rps24) is higher than that of non-stalled gene Rgs2. Aff4 remains associated with promoters during early reactivation, while stalling is reduced. Proliferating cells can exit G1 into either reversible arrest (quiescence/G 0 ) or permanent arrest (differentiation). While both states of mitotic inactivity show enhanced pausing, we identify a distinct regulation of quiescence-specifically stalled genes and its role in cellular fate. Pol II pausing appears to maintain repressed transcription, and thereby contributes to reversibly arrested cellular state and self-renewal. In quiescent cells, Aff4 restrains transcription of stalled genes and blocks cell cycle reentry. During exit from quiescence (G 0 àG 1 transition), relief of Pol II pausing leads to coordinated transcriptional activation of poised genes and cellular processes, and reconfigures cellular physiology for timely cell cycle progression.  Fig. 1 Decreased total RNA content, repressed total RNA synthesis and major down-regulation of the transcriptome in quiescent myoblasts and satellite cells.   Fig S1b Cell number normalization of the RNA-seq dataset does not affect quantification as the replicated cluster together. The levels of ERCC spike in controls was used to normalize the sequencing depth (see methods) of the two replicates of MB, G0 and MT cell states. In order to confirm that the modified normalization method employed here does not affect the replicates the Euclidean distances between the samples was calculated from the regularized log transformation (Deseq package). The distance is color coded and shown in the key scale in the heat map. The dendogram shows the distance between samples and their replicates indicating the replicates behave similarly and three states are indeed distinct to each other.

Fig. S1a
Satellite cells associated with freshly isolated myofiber are activated rapidly during the isolation protocol. Boxplot for median intensity of EU incorporation: Pax7 positive (green) satellite cells & myonuclei (red) for ex vivo cultures at various time points (30min, 3hrs and 24hrs post isolation).

Fig. S1b
Cell number normalization of the RNA-seq dataset does not affect quantification as the replicates cluster together.
In order to confirm that the modified normalization method employed in this study does not affect the similarity between samples and replicates, the Euclidean distances between the samples was calculated from the regularized log transformation (Deseq package). The distance is color coded and shown in the key scale in the heat map. The dendogram shows the distance between samples and their replicates, indicating that the replicates behave similarly, while the three states are indeed distinct from each other.

Fig. S1a
Satellite cells associated with freshly isolated myofiber are activated rapidly during the isolation protocol. Boxplot for median intensity of EU incorporation: Pax7 positive (green) satellite cells & myonuclei (red) for ex vivo cultures at various time points (30min, 3hrs and 24hrs post isolation).

Fig. S1b
Cell number normalization of the RNA-seq dataset does not affect quantification as the replicates cluster together.
In order to confirm that the modified normalization method employed in this study does not affect the similarity between samples and replicates, the Euclidean distances between the samples was calculated from the regularized log transformation (Deseq package). The distance is color coded and shown in the key scale in the heat map. The dendogram shows the distance between samples and their replicates, indicating that the replicates behave similarly, while the three states are indeed distinct from each other.

Fig. S1a
Satellite cells associated with freshly isolated myofiber are activated rapidly during the isolation protocol. Boxplot for median intensity of EU incorporation: Pax7 positive (green) satellite cells & myonuclei (red) for ex vivo cultures at various time points (30min, 3hrs and 24hrs post isolation).

Fig. S1b
Cell number normalization of the RNA-seq dataset does not affect quantification as the replicates cluster together.
In order to confirm that the modified normalization method employed in this study does not affect the similarity between samples and replicates, the Euclidean distances between the samples was calculated from the regularized log transformation (Deseq package). The distance is color coded and shown in the key scale in the heat map. The dendogram shows the distance between samples and their replicates, indicating that the replicates behave similarly, while the three states are indeed distinct from each other.  In order to confirm that the modified normalization method employed in this study does not affect the similarity between samples and replicates, the Euclidean distances between the samples was calculated from the regularized log transformation (Deseq package). The distance is color coded and shown in the key scale in the heat map. The dendogram shows the distance between samples and their replicates, indicating that the replicates behave similarly, while the three states are indeed distinct from each other.         Left panel = Equal cell number normalization: Group B transcript used as reference for normalization and histogram shows their value is close to zero. Group A transcripts show average difference of 2 (log2 scale) between the replicate set 1 (ERCC mix1) and replicate set 2 (ERCC mix 2).

Fig S1c
Biological process comparison for two mitotically arrested states (Up regulated genes) gProfiler analysis carried out for genes down-regulated in G0 vs MB and MT vs MB comparisons. The output is graphically represented here using "moderate" filtering option for Biological process and cellular component . The enriched terms in each state is listed to the left. The square with red intensity indicates the significance of each state. The most significant among the 2 states is highlighted with bold boundaries around square and corresponding p value is shown next to it and shaded in red square..  Gene ontology analysis of genes differentially regulated between two mitotically arrested states (up-regulated genes). Biological process comparison for two mitotically arrested states (Up regulated genes) gProfiler analysis carried out for genes up-regulated in G0 vs MB and MT vs MB comparisons. The output is graphically represented here using moderate option for Biological process and cellular component. The terms enriched in each state are listed on the left. The intensity of color in red square on right indicates the significance of each state. The most significant among the 2 gene lists are highlighted with bold boundaries around square and corresponding p value is shown next to it and shaded in red square. Yellow highlight terms are enrich specifically in G0 and green are represented from MT state.  Gene ontology analysis of genes differentially regulated between two mitotically arrested states (up-regulated genes). Biological process comparison for two mitotically arrested states (Up regulated genes) gProfiler analysis carried out for genes up-regulated in G0 vs MB and MT vs MB comparisons. The output is graphically represented here using moderate option for Biological process and cellular component. The terms enriched in each state are listed on the left. The intensity of color in red square on right indicates the significance of each state. The most significant among the 2 gene lists are highlighted with bold boundaries around square and corresponding p value is shown next to it and shaded in red square. Yellow highlight terms are enrich specifically in G0 and green are represented from MT state.  Gene ontology analysis of genes differentially regulated between two mitotically arrested states (up-regulated genes). Biological process comparison for two mitotically arrested states (Up regulated genes) gProfiler analysis carried out for genes up-regulated in G0 vs MB and MT vs MB comparisons. The output is graphically represented here using moderate option for Biological process and cellular component. The terms enriched in each state are listed on the left. The intensity of color in red square on right indicates the significance of each state. The most significant among the 2 gene lists are highlighted with bold boundaries around square and corresponding p value is shown next to it and shaded in red square. Yellow highlight terms are enrich specifically in G0 and green are represented from MT state.  Gene ontology analysis of genes differentially regulated between two mitotically arrested states (up-regulated genes). Biological process comparison for two mitotically arrested states (Up regulated genes) gProfiler analysis carried out for genes up-regulated in G0 vs MB and MT vs MB comparisons. The output is graphically represented here using moderate option for Biological process and cellular component. The terms enriched in each state are listed on the left. The intensity of color in red square on right indicates the significance of each state. The most significant among the 2 gene lists are highlighted with bold boundaries around square and corresponding p value is shown next to it and shaded in red square. Yellow highlight terms are enrich specifically in G0 and green are represented from MT state.  Gene ontology analysis of genes differentially regulated between two mitotically arrested states (up-regulated genes). Biological process comparison for two mitotically arrested states (Up regulated genes) gProfiler analysis carried out for genes up-regulated in G0 vs MB and MT vs MB comparisons. The output is graphically represented here using moderate option for Biological process and cellular component. The terms enriched in each state are listed on the left. The intensity of color in red square on right indicates the significance of each state. The most significant among the 2 gene lists are highlighted with bold boundaries around square and corresponding p value is shown next to it and shaded in red square. Yellow highlight terms are enrich specifically in G0 and green are represented from MT state.  Gene ontology analysis of genes differentially regulated between two mitotically arrested states (up-regulated genes). Biological process comparison for two mitotically arrested states (Up regulated genes) gProfiler analysis carried out for genes up-regulated in G0 vs MB and MT vs MB comparisons. The output is graphically represented here using moderate option for Biological process and cellular component. The terms enriched in each state are listed on the left. The intensity of color in red square on right indicates the significance of each state. The most significant among the 2 gene lists are highlighted with bold boundaries around square and corresponding p value is shown next to it and shaded in red square. Yellow highlight terms are enrich specifically in G0 and green are represented from MT state.  Gene ontology analysis of genes differentially regulated between two mitotically arrested states (up-regulated genes). Biological process comparison for two mitotically arrested states (Up regulated genes) gProfiler analysis carried out for genes up-regulated in G0 vs MB and MT vs MB comparisons. The output is graphically represented here using moderate option for Biological process and cellular component. The terms enriched in each state are listed on the left. The intensity of color in red square on right indicates the significance of each state. The most significant among the 2 gene lists are highlighted with bold boundaries around square and corresponding p value is shown next to it and shaded in red square. Yellow highlight terms are enrich specifically in G0 and green are represented from MT state.  Gene ontology analysis of genes differentially regulated between two mitotically arrested states (up-regulated genes). Biological process comparison for two mitotically arrested states (Up regulated genes) gProfiler analysis carried out for genes up-regulated in G0 vs MB and MT vs MB comparisons. The output is graphically represented here using moderate option for Biological process and cellular component. The terms enriched in each state are listed on the left. The intensity of color in red square on right indicates the significance of each state. The most significant among the 2 gene lists are highlighted with bold boundaries around square and corresponding p value is shown next to it and shaded in red square. Yellow highlight terms are enrich specifically in G0 and green are represented from MT state.

Fig S1d
Biological process comparison for two mitotically arrested states (Down regulated genes) gProfiler analysis carried out for genes down-regulated in G0 vs MB and MT vs MB comparisons. The output is graphically represented here using "moderate" filtering option for Molecular function and KEGG pathways . The enriched terms in each state is listed to the left. The square with red intensity indicates the significance of each state. The most significant among the 2 states is highlighted with bold boundaries around square and corresponding p value is shown next to it and shaded in red square.. Gene ontology analysis of genes differentially regulated between two mitotically arrested states (Down-regulated genes). Biological process comparison for two mitotically arrested states (Down regulated genes) gProfiler analysis carried out for genes down-regulated in G0 vs MB and MT vs MB comparisons. Yellow highlight terms are enrich specifically in G0 and green are represented from Both G0 and MT states.

Fig. S1e
Gene ontology analysis of genes differentially regulated between two mitotically arrested states (Down-regulated genes). Biological process comparison for two mitotically arrested states (Down regulated genes) gProfiler analysis carried out for genes down-regulated in G0 vs MB and MT vs MB comparisons. Yellow highlight terms are enrich specifically in G0 and green are represented from Both G0 and MT states. Gene ontology analysis of genes differentially regulated between two mitotically arrested states (Down-regulated genes). Biological process comparison for two mitotically arrested states (Down regulated genes) gProfiler analysis carried out for genes down-regulated in G0 vs MB and MT vs MB comparisons. Yellow highlight terms are enrich specifically in G0 and green are represented from Both G0 and MT states.

Fig. S1e
Gene ontology analysis of genes differentially regulated between two mitotically arrested states (Down-regulated genes). Biological process comparison for two mitotically arrested states (Down regulated genes) gProfiler analysis carried out for genes down-regulated in G0 vs MB and MT vs MB comparisons. Yellow highlight terms are enrich specifically in G0 and green are represented from Both G0 and MT states. Gene ontology analysis of genes differentially regulated between two mitotically arrested states (Down-regulated genes). Biological process comparison for two mitotically arrested states (Down regulated genes) gProfiler analysis carried out for genes down-regulated in G0 vs MB and MT vs MB comparisons. Yellow highlight terms are enrich specifically in G0 and green are represented from Both G0 and MT states.

Fig. S1e
Gene ontology analysis of genes differentially regulated between two mitotically arrested states (Down-regulated genes). Biological process comparison for two mitotically arrested states (Down regulated genes) gProfiler analysis carried out for genes down-regulated in G0 vs MB and MT vs MB comparisons. Yellow highlight terms are enrich specifically in G0 and green are represented from Both G0 and MT states. Gene ontology analysis of genes differentially regulated between two mitotically arrested states (Down-regulated genes). Biological process comparison for two mitotically arrested states (Down regulated genes) gProfiler analysis carried out for genes down-regulated in G0 vs MB and MT vs MB comparisons. Yellow highlight terms are enrich specifically in G0 and green are represented from Both G0 and MT states. Gene ontology analysis of genes differentially regulated between two mitotically arrested states (Down-regulated genes). Biological process comparison for two mitotically arrested states (Down regulated genes) gProfiler analysis carried out for genes down-regulated in G0 vs MB and MT vs MB comparisons. Yellow highlight terms are enrich specifically in G0 and green are represented from Both G0 and MT states. Fig. S1f,g Identifying similarity between G0 in cultured C2C12 myoblasts (this study) and freshly isolated satellite cells.
f. Percentage of overlapping genes between differentially expressed genes in G0 vs. MB comparison (Up/Down regulated) for each of gene sets QSCs (437) and ASC36h (355) ASC60h (1588) gene set identified in (Liu et al., 2013). Note that although Quiescent SC has genes higher fraction of genes that are Down regulated in G0 vs MB comparisons suggesting QSC gene list depicts partially activated cell state. ASC 60hrs overlap almost entirely with G0 down regulated genes ie ASC60 hrs is similar to MB.
g. Leading edge analysis using Gene set enrichment analysis (GSEA) comparing G0 vs. MB : Geneset used here is 437 gene identified as QSCs specific genes (Liu et al., 2013) and expression dataset generated in our study was used as reference. Note that these genes partition into bimodal distribution one node enriched for both G0 and MB state. Identifying similarity between G0 in cultured C2C12 myoblasts (this study) and freshly isolated satellite cells.
a. Percentage of overlapping genes between differentially expressed genes in G0 vs. MB comparison (Up/Down regulated) for each of gene sets QSCs (437) and ASC36h (355) ASC60h (1588) gene set identified in (Liu et al., 2013). Note that although Quiescent SC has genes higher fraction of genes that are Down regulated in G0 vs MB comparisons suggesting QSC gene list depicts partially activated cell state. ASC 60hrs overlap almost entirely with G0 down regulated genes ie ASC60 hrs is similar to MB.
b. Leading edge analysis using Gene set enrichment analysis (GSEA) comparing G0 vs. MB : Geneset used here is 437 gene identified as QSCs specific genes (Liu et al., 2013) and expression dataset generated in our study was used as reference. Note that these genes partition into bimodal distribution one node enriched for both G0 and MB state

G0_IP
MB_IP MT_IP -2kb +1 +2kb -2kb +1 +2kb -2kb +1 +2kb Fig S2b. RNA pol II read density across TSS Intensity of RNA pol II enrichment is represented around TSS (+/-2kb) for non overlapping genes. The dark shade represent higher intensity as observed in the ChIPseq data set for all three cellular states.  For calculation of stalling index the input is used as reference to set stringent cutoffs. The pulldown sample has rightward tail of the stalling index compared to Input and non overlapping region with >95percent confidence is used to identify the stalled gene. The vertical line shows the cut off value for G0 sample. For calculation of stalling index the input is used as reference to set stringent cutoffs. The pulldown sample has rightward tail of the stalling index compared to Input and non overlapping region with >95percent confidence is used to identify the stalled gene. The vertical line shows the cut off value for G0 sample.     RNA Pol II abundance at histone 1 cluster is higher in G0 and MB and not in MT Genome browser snapshot was generated for the Histone 1 gene cluster at coordinates (chr 13:23622400-23677600) on mm9 build. The top panel shows the gene location on chromosome and its orientation (red indicates forward and blue in reverse). The read density is normalized per million within each sample and plotted with the color code where red signifies higher density and blue indicates lower density. The intergenic regions are removed in order to represent the density at higher resolution. Strong enrichment of Pol II seen in MB persists in G0 but not MT, despite down regulation of transcripts in both arrested states.                                     Representative merge immunofluorescence images for MB, MT,G0 R30min and R2hrs stained for G0 stalled genes(red) and DAPI (Blue). NCL (left column) shows distinct nucleolar staining whereas and RPS24 (Right column) is cytoplasmic staining. For both these G0 stalled gene MB shows maximum amount of staining whereas both the arrested state (G0 and MT) shows reduced levels. Moreover G0 has lower amount of staining which induces rapidly after reactivation(R30min and R2hrs)

Fig. S3d
Restoration of expression of G0-stalled genes during cell cycle reactivation. Representative immunofluorescence images for MB, MT, G0, R30min and R2hrs stained for G0 stalled genes (red) and DAPI (Blue). Ncl (left column) shows distinct sub-nuclear puncta (nucleolar) staining whereas and Rps24 (Right column) is found in the cytoplasm. For both these G0 stalled genes MB shows maximum protein expression, whereas the two arrested states (G0 and MT) show very reduced levels. Expression is rapidly induced after reactivation (R30min and R2hrs).

Fig. S3d
Restoration of expression of G0-stalled genes during cell cycle reactivation. Representative immunofluorescence images for MB, MT, G0, R30min and R2hrs stained for G0 stalled genes (red) and DAPI (Blue). Ncl (left column) shows distinct sub-nuclear puncta (nucleolar) staining whereas and Rps24 (Right column) is found in the cytoplasm. For both these G0 stalled genes MB shows maximum protein expression, whereas the two arrested states (G0 and MT) show very reduced levels. Expression is rapidly induced after reactivation (R30min and R2hrs).

Fig. S3d
Restoration of expression of G0-stalled genes during cell cycle reactivation. Representative immunofluorescence images for MB, MT, G0, R30min and R2hrs stained for G0 stalled genes (red) and DAPI (Blue). Ncl (left column) shows distinct sub-nuclear puncta (nucleolar) staining whereas and Rps24 (Right column) is found in the cytoplasm. For both these G0 stalled genes MB shows maximum protein expression, whereas the two arrested states (G0 and MT) show very reduced levels. Expression is rapidly induced after reactivation (R30min and R2hrs).

Fig. S3d
Restoration of expression of G0-stalled genes during cell cycle reactivation. Representative immunofluorescence images for MB, MT, G0, R30min and R2hrs stained for G0 stalled genes (red) and DAPI (Blue). Ncl (left column) shows distinct sub-nuclear puncta (nucleolar) staining whereas and Rps24 (Right column) is found in the cytoplasm. For both these G0 stalled genes MB shows maximum protein expression, whereas the two arrested states (G0 and MT) show very reduced levels. Expression is rapidly induced after reactivation (R30min and R2hrs).

Fig. S3d
Restoration of expression of G0-stalled genes during cell cycle reactivation. Representative immunofluorescence images for MB, MT, G0, R30min and R2hrs stained for G0 stalled genes (red) and DAPI (Blue). Ncl (left column) shows distinct sub-nuclear puncta (nucleolar) staining whereas and Rps24 (Right column) is found in the cytoplasm. For both these G0 stalled genes MB shows maximum protein expression, whereas the two arrested states (G0 and MT) show very reduced levels. Expression is rapidly induced after reactivation (R30min and R2hrs).

Fig. S3d
Restoration of expression of G0-stalled genes during cell cycle reactivation. Representative immunofluorescence images for MB, MT, G0, R30min and R2hrs stained for G0 stalled genes (red) and DAPI (Blue). Ncl (left column) shows distinct sub-nuclear puncta (nucleolar) staining whereas and Rps24 (Right column) is found in the cytoplasm. For both these G0 stalled genes MB shows maximum protein expression, whereas the two arrested states (G0 and MT) show very reduced levels. Expression is rapidly induced after reactivation (R30min and R2hrs).

Fig. S3d
Restoration of expression of G0-stalled genes during cell cycle reactivation. Representative immunofluorescence images for MB, MT, G0, R30min and R2hrs stained for G0 stalled genes (red) and DAPI (Blue). Ncl (left column) shows distinct sub-nuclear puncta (nucleolar) staining whereas and Rps24 (Right column) is found in the cytoplasm. For both these G0 stalled genes MB shows maximum protein expression, whereas the two arrested states (G0 and MT) show very reduced levels. Expression is rapidly induced after reactivation (R30min and R2hrs). Restoration of expression of G0-stalled genes during cell cycle reactivation. Representative immunofluorescence images for MB, MT, G0, R30min and R2hrs stained for G0 stalled genes (red) and DAPI (Blue). Ncl (left column) shows distinct sub-nuclear puncta (nucleolar) staining whereas and Rps24 (Right column) is found in the cytoplasm. For both these G0 stalled genes MB shows maximum protein expression, whereas the two arrested states (G0 and MT) show very reduced levels. Expression is rapidly induced after reactivation (R30min and R2hrs). NCL staining is observed in satellite stem cells (SC) associated with single myofibers cultured ex vivo. SC marked by Pax7 (green, arrowhead) can be distinguished from differentiated myonuclei which are Pax7 negative (MN, asterisk) within the underlying myofiber. At 0hrs and 3 hrs post isolation, SC shows induced NCL levels whereas MN shows little staining.  Extent of knockdown of G0 stalled genes in proliferative MB and quiescent G0 state G0 state compared to its respective NTS control. RNA content, active RNA synthesis and nuclear area altered in proliferating MB after siRNA-mediated knockdown of candidate G0-stalled genes (NTS-non targeting control siRNA, Ctdp1, Ncl, Rps24, Slbp and Zc3h12A). After 36hr, cells were co-stained for total RNA (SYTO® RNASelectTM -green) and EU incorporation following a 30` pulse (red) (scale bar 10µm).   Colony forming assay for G0 cells treated with siRNA for G0 stalled genes,represented as % CFU. Reduced self renewal as estimated by colony forming ability for Ncl, Rps24 and Ctdp1 knockdown but not for Slbp and Zc3h12A knockdown (*p-value <0.05, n=3 student`s T-test)