Cold‐induced chromatin compaction and nuclear retention of clock mRNAs resets the circadian rhythm

Abstract Cooling patients to sub‐physiological temperatures is an integral part of modern medicine. We show that cold exposure induces temperature‐specific changes to the higher‐order chromatin and gene expression profiles of human cells. These changes are particularly dramatic at 18°C, a temperature synonymous with that experienced by patients undergoing controlled deep hypothermia during surgery. Cells exposed to 18°C exhibit largely nuclear‐restricted transcriptome changes. These include the nuclear accumulation of mRNAs encoding components of the negative limbs of the core circadian clock, most notably REV‐ERBα. This response is accompanied by compaction of higher‐order chromatin and hindrance of mRNPs from engaging nuclear pores. Rewarming reverses chromatin compaction and releases the transcripts into the cytoplasm, triggering a pulse of negative limb gene proteins that reset the circadian clock. We show that cold‐induced upregulation of REV‐ERBα is sufficient to trigger this reset. Our findings uncover principles of the cellular cold response that must be considered for current and future applications involving therapeutic deep hypothermia.

The effect of temperat ure on circadian clocks is highly complex. While these timekeepers are temperat ure-compensat ed even in homeot hermic animals, they can be synchroniz ed even by minut e oscillat ions in temperat ure wit hin the physiological temperat ure range (Saini et al., 2012). These temperat ure oscillat ions are accompanied by the rhyt hmic shut tling of Heat Shock Fact or (HSF1) bet ween nucleus and cyt oplasm (Reinke et al., 2008) and the cyclic expression of the coldinducible RNA-binding prot ein CIRP (and to a lesser ext ent RBM3) (Morf et al., 2012), which have been shown to part icipat e in the temperat ure-dependent phase reset ting.
Temperat ures below the physiological range are used in the clinic during cert ain surgical int ervent ions and for the preservat ion of organs dest ined for transplant at ion. In this int erest ing st udy the aut hors examined the effect of sub-physiological temperat ures on chromat in compact ion and circadian clock gene expression in two human cell lines (AC16,U2OS). Surprisingly, they find that chromat in assumes a more condensed st at e when lowering the temperat ure from 37{degree sign} C to 18{degree sign} C (somet imes used during cont rolled deep hypot hermia in the clinic), but shows lit tle changes at 8{degree sign} C (used for the st orage of organs) when compared to 37{degree sign} C. These changes are reversible, as chromat in t opology rapidly assumes t he normal st at e upon reset ting t he t emperat ure t o 37{degree sign} C. The t emperat ure-dependent transit ions in chromat in st ruct ure do not appear t o involve post translat ional modificat ions of hist ones or RNA polymerase II. Using cell fract ionat ion and FISH experiment s t he aut hors obt ained compelling evidence t hat cooling cells t o 18{degree sign} C engenders a ret ent ion of polyadenylat ed t ranscript s in t he nucleus, probably owing t o different configurat ions of perinuclear chromat in and nuclear pores. Again, t his nuclear t ranscript ret ent ion is reversible, in t hat it is rapidly alleviat ed upon readjust ing t he t emperat ure t o 37{degree sign} C. The aut hors t hen went on t o st udy t he impact of t ranscript release int o t he cyt oplasm for circadian gene expression upon rewarming t he cells from 18{degree sign} C t o 37{degree sign} C. Surprisingly, t hey observed t hat mRNAs encoding core clock prot eins ent er t he cyt oplasm wit h different kinet ics during t his process. Among t he examined t ranscript s, REV-ERBα mRNA is export ed most rapidly int o t he cyt osol, and this elicit s a surge of REV-ERBα prot ein accumulat ion. The aut hors t hen compared the phases of circadian PER2::luc expression in U2OS cells wit h or wit hout funct ional REV-ERBα alleles in a clever set of synchroniz at ion experiment s ( Figure 5). Briefly, PER2::luc expression was synchroniz ed t o different phases by dexamet hasone in parallel cult ures. T hese were t hen exposed to t he cooling-rewarming prot ocol, and t he phase-reset ting was examined. In experiment s in which the cooling t o 18{degree sign} C was rest rict ed t o 5 hours (i.e. a t ime period during which newly synt hesized REV-ERBα t ranscript s remained largely in t he nucleus), only wild-t ype cells could efficient ly set all different ly pre-synchroniz ed phases t o t he same phase. Hence, under t hese condit ions t he surge of REV-ERBα expression upon re-warming was required t o set all phases t o a nearly ident ical one.

General Comment s
This is an int erest ing st udy on t he effect s of sub-physiological t emperat ure on circadian gene expression t hat have hit hert o not yet been invest igat ed. A quest ion t hat remains open is how a surge of REV-ERBα accumulat ion can synchroniz e clock gene expression wit hout affect ing CLOCK and BMAL1 expression. REV-ERBα has been thought to modulate the expression of positive limb members (i.e.), however, the authors did not observe temperature-dependent changes in CLOCK and BMAL1 mRNAs. It is thus unclear how the release of REV-ERBα mRNA from the nucleus can quickly synchronize the different phases established by Dexamethason (see below). In the Introduction and Discussion sections the description of the Transcriptional/Translational Feedback Loops (TTFLs) underlying circadian rhythm generation is very confusing. As it's written it gives the impression that CRYs, PERs and REV-ERBs belong to the same functional class of "repressor/corepressor" genes encoding proteins of the negative limbs of the TTFLs. According to current belief there are two coupled feedback loops driving the molecular oscillator. In the primary TTFL CRYs and PERs are assembled in large co-repressosome complexes that annul the activity of CLOCK-BMAL1 complexes. In the secondary TTFL, REV-ERB repressors cyclically inhibit the transactivation of the BMAL1 and CLOCK genes through competitive binding with ROR activators to ROREs. Since the rhythmic expression of REV-ERBs is under control of the primary TTFL, the primary and secondary TTFLs are tightly coupled. As mentioned above, the results obtained by the authors with regard to the phase-setting role of REV-ERBα in the cooling-rewarming protocol is difficult to interpret on the basis of the "coupled TTFL model". Nonetheless,the data presented in Figure 5B are rather compelling, and this reviewer thus considers likely that REV-ERBα acts through a less conventional mechanism, perhaps by binding directly to PER2 (Schmutz et al., 2010). I am not asking the authors to figure out the detailed mechanism through which Rev-ERBa synchronizes circadian clocks upon rewarming, as this would take years. However, this issue should definitely be discussed in greater detail.

Minor Issues
I have never seen the term "clock suppressor" standing for members of the negative limbs of the TTFLs in the scientific literature. While REV-ERBs act as are repressors (in the classical sense) by directly binding to cis-acting regulatory elements (ROREs), CRYs and PERs can be considered as corepressors, since they bind to DNA-binding proteins (here CLOCK-BMAL1) rather than to DNA elements. It would be advisable to replace "clock suppressors" with "clock repressors/corepressors" or "negative limb members".
On page 8 (Paragraph "Cold exposure shapes...") the terms "genes" and "transcripts" are confused on multiple occasions. For example "651 and 1267 genes showing significant changes in the nucleus and the cytoplasm" should read "651 and 1267 transcripts showing significant changes in the nucleus and the cytoplasm". Please discriminate between "genes" and "transcripts" throughout the text.
On page 18 the sentence "However, both of these pathways cannot account for the observed phase changes..." should read "However, neither of these pathways can account for the observed phase changes....".
Referee #2: Paper summary Fischl et al demonstrate that therapeutically relevant cooling temperatures leads to the upregulation of the circadian clock suppressor transcripts and that chromatin compaction limits their nuclear export. Upon rewarming, the authors show that the reversal of chromatin compaction correlates with nuclear export of transcripts. The authors propose that the release of these mRNAs to the cytoplasm leads to their translation and resets the circadian clock. These are very interesting findings. However, from Figure 2 onwards the paper was rather difficult to read as the order the data were presented was not logical order. Moreover, some of the conclusions were not supported and additional data for figures 3, 4 and 5 is required as outlined below. Figure 1 · Page 6 line 17/18 states "....compaction and largest increase in IC occurs at 18{degree sign} C, not 8{degree sign} C...". The chromatin landscape at 8oc looks most similar to 37oc in Fig 1A, and not like the other cooled samples. Can the authors expand on why incubating cells at 8oc does not appear to have the same effect as cooling at 28oC or 18oC? · Page 6 line 18/19 states "...cold-induced compaction to chromatin is reversed (Fig. 1B)". However, fig 1B only shows partial reversal of compaction when incubating 18oC cells at 37oC for 2 hours. Moreover, the y-axis scale suggests that these are very moderate changes. Therefore, can the authors include statistical analysis to confirm these observations. · What is the viability of cells grown at 8oC for 24 hours? Can the authors include analysis of cell death in each of these different cooling conditions to show that cells are not dying upon cooling and rewarming. · Figure 3D: It is important to show that the tagging of REV-ERBa doesn't perturb its function or mRNA localisation upon cooling and rewarming. It would also be interesting to know if the translation of REV-REBa correlates with global protein synthesis rates. · Text on page 11 line 9/10 states "...nuclear-retained transcripts surge into the cytoplasm after rewarming to 37{degree sign} C where they are translated (Fig. 3D, S3H)". However, fig 3D and S3H only shows data for REV-ERBa and not PER1, CRY1 or CRY2. Please amend the text or include examples of other nuclear transcripts to support this text. · Text on page 11 line 18/19 states "...we employed RNA-FISH using probes targeting transcripts for the most highly upregulated and earliest responding gene, REV-ERBα, the later responding and nuclear-restricted gene, CRY2, and the unchanged control gene, TP53 (Fig. S3F)". Fig S3F shows ARNTL, CLOCK and TP53. This is very confusing -please amend. Figure 4B and 4F -The representative images in 4B (particularly 18oC 24h, 37oC 2h) do not support the quantified data in 2F. This makes it difficult to conclude: ".... after 24h, CRY2 transcripts accumulate in the nucleus and are released into the cytoplasm after rewarming". Please include the N value of replicates used to generate this quantification. · Page 13 line 11/12 states "...rewarming following exposure to 18{degree sign} C for 5h only triggers a cytoplasmic pulse of increased REV-ERBα without affecting PER1/2 or CRY1/2 levels ( Fig. 3-4)". How can the authors conclude this when the data for CRY1/2 or PER1/2 at the 5 hr time point are not included in figure 4? The authors must include the 5 hr time for at least CRY2 to make this conclusion. Moreover, figure 3B only shows the localisation of RNA for rewarming after 24 hr 18oC and not 5 hours. Please include the data for 5 hours to support this statement.

Figure 4 ·
· The authors show that the localisation of REV-ERBa in the cytoplasm following 18oC 24 hr and rewarming 2 hr differs in AC16 and U2OS cells (Fig 4E and 4H -purple bar in cytoplasmic area), suggesting that REV-ERBa is retained/maintained in the cytoplasm following rewarming in U2OS cells but not AC16 cells. Could the authors provide and explanation as to why this may be the case? Minor Comments · Add N values to all figure legends. · Figure 1B -Add missing value to axis (0.60). · Refer to the figures in order and provide the figure reference at the appropriate place in the text. · In figures where multiple cell lines are used, please state the cell line used in the figure legend. · Axis titles missing in figures 3B, 3D · Please add protein size marker label to ponceau stain in figure 3D and S3H · Figure 3D: It is not clear if these are cytoplasmic cell extracts. Please add more detail to the figure legend. · @ 37{degree sign} C is missing from figure S3A · End of figure legend 4 states "All error bars (D-H) show S.E.M." - Figure 4D shows microscopy images and doesn't include error bars -Please check that all in-text reference to figures are correct. · Figure 5A and 5B: It would be clearer to include temperature labels (@ 18oC or @ 37oC) below the time point label to be consistent with figure 3A and 3B.
Referee #3: The manuscript by Dr. Fischl and colleagues addresses the impact of induced cold stress on the chromatin organization in human cardiomyocytes. Furthermore, the study explores the connection of this effect to the regulation of the cell-autonomous molecular clocks. The data suggest that cold exposure leads to the nuclear accumulation of several clock components, with Reverba being the most affected. Of note, exposure of the AC16 cardiomyocytes to the cold stress at 28C, 18C and 8C led to the temperature specific transcription modifications in the cellular nuclei and cytoplasm. These alterations were non-linear, with the nuclear signatures at 8C being surprisingly the mildest, i e closest to those observed at 37C controls. Interestingly, CIRBP transcript that was upregulated at 28C as one would expect, did not exhibit changes at more extreme lower temperatures. Strikingly, the transcripts upregulated at 18C were retained in the nucleus and were released to cytosol upon rewarming back to 37C, including the core-clock components of the negative limb (CRYs and Pers) and Reverba. The latter exhibited the most dramatic changes of its levels. Subsequent experiments with U2OS cells expression Per2::luciferase reporter and expressing or lacking Reverba highlighted the role of the core-clock repressors in the cell response to the extreme cold stress and in resetting the clockwork upon rewarming. Moreover, these experiments revealed that cold induced activation of Reverba was sufficient to reset the phase of cellular oscillators.
In summary, this is a well-designed and rigorously conducted study, the data is presented clearly and concisely, and followed by a meaningful discussion. The reported data lead to important conclusions both scientifically and potentially clinically.
I have only few minor suggestions for this already refined work: 1. In the Abstract, in case the space allows, I would specify what kind of "human cells" the authors are working with.
2. For the experiments with U2OS/Per2::luc ReverbaKO cells -the authors thoroughly explain the deletion protocol that leads to the functional disruption of the gene. Can the authors please clarify how did they confirm that Reverba function has been disrupted?
3. The authors report an intriguing finding that most of the transcriptional changes are observed at 18C, whereas nuclear transcriptome exhibited relatively mild alterations at extreme temperature of 8C. How do the authors explain this interesting observation?
4. Discussion p. 15 line 6: there are unnecessary parentheses appearing between the references by Ou et al and Jackson and Kochanek.

Author's response to the comments received from the reviewers
Author's response: We thank all the reviewers for their constructive assessment and insightful comments on our paper. We have addressed all the points raised in a revised manuscript and in the point by point responses below. We have added information regarding the statistical analyses and provided additional data including: a new experiment addressing the survival rate of cells exposed to 8°C; data from an additional sequencing time point to analyse the changes in core clock gene transcript levels in the nucleus and cytoplasm after rewarming following a 5h exposure to 18°C; data that confirm disruption of REV-ERBα expression in the REV-ERBα knockout cell line. We have addressed all other points raised by the referees by changing the text in the manuscript as requested. We feel that together these changes and the additional information we have included have greatly improved our manuscript and strengthened the conclusions made and we are grateful to the referees for that.

Responses to specific points raised by individual referees:
Referee #1

-> General Comments
This is an interesting study on the effects of sub-physiological temperature on circadian gene expression that have hitherto not yet been investigated. A question that remains open is how a surge of REV-ERBα accumulation can synchronize clock gene expression without affecting CLOCK and BMAL1 expression. REV-ERBα has been thought to modulate the expression of positive limb members (i.e.), however, the authors did not observe temperature-dependent changes in CLOCK and BMAL1 mRNAs. It is thus unclear how the release of REV-ERBα mRNA from the nucleus can quickly synchronize the different phases established by Dexamethason (see below).

3rd Aug 2020 1st Authors' Response to Reviewers
In the Introduction and Discussion sections the description of the Transcriptional/Translational Feedback Loops (TTFLs) underlying circadian rhythm generation is very confusing. As it's written it gives the impression that CRYs, PERs and REV-ERBs belong to the same functional class of "repressor/corepressor" genes encoding proteins of the negative limbs of the TTFLs. According to current belief there are two coupled feedback loops driving the molecular oscillator. In the primary TTFL CRYs and PERs are assembled in large co-repressosome complexes that annul the activity of CLOCK-BMAL1 complexes. In the secondary TTFL, REV-ERB repressors cyclically inhibit the transactivation of the BMAL1 and CLOCK genes through competitive binding with ROR activators to ROREs. Since the rhythmic expression of REV-ERBs is under control of the primary TTFL, the primary and secondary TTFLs are tightly coupled. As mentioned above, the results obtained by the authors with regard to the phase-setting role of REV-ERBα in the cooling-rewarming protocol is difficult to interpret on the basis of the "coupled TTFL model". Nonetheless,the data presented in Figure 5B are rather compelling, and this reviewer thus considers likely that REV-ERBα acts through a less conventional mechanism, perhaps by binding directly to PER2 (Schmutz et al., 2010). I am not asking the authors to figure out the detailed mechanism through which Rev-ERBa synchronizes circadian clocks upon rewarming, as this would take years. However, this issue should definitely be discussed in greater detail.
We apologise for not clearly distinguishing between the primary and secondary TTFLs. We have revised the TTFL description accordingly in the introduction and discussion in the revised manuscript.
In regard to the potential mechanisms by which the accumulation of REV-ERBα triggers phase synchronisation, we have now included a new paragraph in the discussion on page 19 in the revised manuscript.

Minor Issues
I have never seen the term "clock suppressor" standing for members of the negative limbs of the TTFLs in the scientific literature. While REV-ERBs act as are repressors (in the classical sense) by directly binding to cis-acting regulatory elements (ROREs), CRYs and PERs can be considered as corepressors, since they bind to DNA-binding proteins (here CLOCK-BMAL1) rather than to DNA elements. It would be advisable to replace "clock suppressors" with "clock repressors/corepressors" or "negative limb members".
We have now explained the distinction between the roles of clock co-repressor genes (PER1/2 and CRY1/2) and clock repressor genes (REVERBα) as part of two separate negative limbs of two TTFLs more clearly in the introduction and then throughout the rest of the text referred to them together as "negative limb genes", as advised.
On page 8 (Paragraph "Cold exposure shapes...") the terms "genes" and "transcripts" are confused on multiple occasions. For example "651 and 1267 genes showing significant changes in the nucleus and the cytoplasm" should read "651 and 1267 transcripts showing 3rd Aug 2020 1st Authors' Response to Reviewers significant changes in the nucleus and the cytoplasm". Please discriminate between "genes" and "transcripts" throughout the text.
We agree that more precision is needed and in response we have now changed all the relevant examples and clearly and consistently state that the transcript levels of genes are changing.
At this point it is worth noting that the numbers of genes showing differential transcript levels upon exposure to different cold conditions has now changed slightly from the initial submission. This is because we realised that there was an error in the original table that we were using for annotations in which certain gene names had been accidently converted to dates (e.g. SEPT1 to 1-Sep) and then excluded from further analysis as they were not recognised as genes upon intersection with genes within the RefSeq database. This error has now been corrected and has changed the significance values calculated by DESeq for a small number of genes, resulting in different numbers of genes showing significantly different transcript levels. This has not affected any of the other analyses or conclusions in any other way.
On page 18 the sentence "However, both of these pathways cannot account for the observed phase changes..." should read "However, neither of these pathways can account for the observed phase changes....". We have now changed this.

Referee #2
Major comments Figure 1 · Page 6 line 17/18 states "....compaction and largest increase in IC occurs at 18{degree sign}C, not 8{degree sign}C...". The chromatin landscape at 8oc looks most similar to 37oc in Fig 1A, and not like the other cooled samples. Can the authors expand on why incubating cells at 8oc does not appear to have the same effect as cooling at 28oC or 18oC?
We agree that the image that we had used to represent the compaction observed for cells exposed to 8°C gave the impression of a similar level of compaction to that seen at 37°C. We have now changed that image to that from a different z-stack of the same nucleus, which shows a similar level of compaction to that observed at 28°C. All stacks can be viewed in the movie files. The boxplots in Fig. 1B and EV1E show the distribution of ratios of chromatin volume to nuclear volume for all cells imaged for each condition. These show that, while there is some variation between cells, in the 28°C and 8°C condition cells show on average significantly more compaction than at 37°C but significantly less than at 18°C.
The effects on chromatin compaction could be different because the level of calcium influx, that we propose as a possible mechanism for chromatin compaction, may differ between the temperatures. It can be expected that both membrane fluidity and channel responses are different between the different temperatures. We have added a sentence explaining this possibility in the discussion on page 16 of the revised manuscript: 3rd Aug 2020 1st Authors' Response to Reviewers "As both the calcium channel activity and membrane fluidity (Al-Fageeh & Smales, 2006) vary at different temperatures, the calcium release is likely to differ at the three temperatures, resulting in the observed differences in the degree of compaction." · Page 6 line 18/19 states "...cold-induced compaction to chromatin is reversed (Fig. 1B)". However, fig 1B only shows partial reversal of compaction when incubating 18oC cells at 37oC for 2 hours. Moreover, the y-axis scale suggests that these are very moderate changes. Therefore, can the authors include statistical analysis to confirm these observations.
We have now included statistics from carrying out Mann-Whitney tests showing that upon rewarming cells for 2h there is significantly less compaction than at 18°C. There is however still significantly more compaction than at 37°C. We have therefore amended the text to clarify that the compaction to chromatin upon rewarming shows reversion back towards the levels seen in cells before cooling. See page 6: "Upon rewarming back to 37°C, this cold-induced compaction to chromatin reverses back towards levels seen in cells before cooling (Fig. 1B)" · What is the viability of cells grown at 8oC for 24 hours? Can the authors include analysis of cell death in each of these different cooling conditions to show that cells are not dying upon cooling and rewarming.
We have carried out viability assays for AC16 cells exposed to 28°C, 18°C and 8°C for 24h and then back up from each of these temperatures to 37°C for 24h. Whilst at both 18°C and 8°C temperatures some cells are no longer adherent and are lost upon aspiration of the growth media (~60 % of cells remain adherent after 24h at 8°C and ~80 % of cells remain adherent after 24h at 18°C). Those that remain are largely viable as shown by the number of cells approximately doubling after 24h back at 37°C, which is similar to the doubling time of AC16 cells kept at 37°C. It is important to note that only these adherent cells are included in sequencing and imaging experiments. Cells at 28°C continue to proliferate at the lower temperature but at a slower rate.

Figure 3
· Page 9 line 20 states "...rewarming cells back from 18{degree sign}C to 37{degree sign}C leads to a rapid readjustment of the nuclear-accumulated transcripts back to levels seen before cooling (Fig. 3A, top panel)". Can the authors include statistical analysis to support this conclusion.
We have included p-values from paired t-tests to show the significance of any difference in standardized RNA levels between key time points in the source data file for each of the standardized RNA level boxplots. We have not included these p-values in the figure as we do not think this adds value. This is because many genes are involved for each comparison, so many of the small differences observed between time points are significant. Therefore, upon rewarming from 18°C back to 37°C for 2h, nuclear-accumulated transcripts are still 3rd Aug 2020 1st Authors' Response to Reviewers significantly elevated relative to their level at 37°C, but these levels are much closer to their levels at 37°C than at 18°C for 24h.
We have amended the text to clarify this. See page 9: "Interestingly, rewarming cells back from 18°C to 37°C leads to a rapid readjustment of the nuclear-accumulated transcripts back to levels close to those seen before cooling (Fig. 3A, top panel)." · Figure 3D: It is important to show that the tagging of REV-ERBa doesn't perturb its function or mRNA localisation upon cooling and rewarming. It would also be interesting to know if the translation of REV-REBa correlates with global protein synthesis rates.
The Western blots match what we expect to see based on the RNA-seq and RNA-FISH results. When REV-ERBα RNA is elevated in the nucleus but not in the cytoplasm then there is no increase in the tagged-REV-ERBα Western blot signal (e.g. after 5h at 18°C). When REV-ERBα RNA is elevated in the cytoplasm (e.g. after 24h at 18°C or upon rewarming after 5h at 18°C) then the tagged-REV-ERBα Western blot signal is increased. This suggests that the tag does not affect the localisation of the REV-ERBα RNA. While we cannot be sure that the tag does not disrupt REV-ERBα function, the tagged strain was not used in any other experiments other than for these Western blots and so any disruption to function is of less importance. The change in the level of REV-ERBα translation after 24h at 18°C or upon rewarming after 5h at 18°C is greater than that of the rate of global protein synthesis based on the Ponceau stain and TP53 translation based on the control Western blot. · Text on page 11 line 9/10 states "...nuclear-retained transcripts surge into the cytoplasm after rewarming to 37{degree sign}C where they are translated (Fig. 3D, S3H)". However, fig 3D and S3H only shows data for REV-ERBa and not PER1, CRY1 or CRY2. Please amend the text or include examples of other nuclear transcripts to support this text.
We agree and have made this point clearer by amending the text accordingly: "The nuclearretained transcripts surge into the cytoplasm after rewarming to 37°C where, as shown for REV-ERBα, they can be translated (Fig. 3D, EV3H)." · Text on page 11 line 18/19 states "...we employed RNA-FISH using probes targeting transcripts for the most highly upregulated and earliest responding gene, REV-ERBα, the later responding and nuclear-restricted gene, CRY2, and the unchanged control gene, TP53 (Fig. S3F)". Fig S3F shows ARNTL, CLOCK and TP53. This is very confusing -please amend.
We apologise for the confusion caused and have re-written this section to clarify this issue on page 11 as follows: "To corroborate the observed RNA-Seq expression patterns of clock genes, we employed RNA-FISH using probes targeting the following transcripts: REV-ERBα, showing the highest upregulation and earliest response (Fig. 3B); CRY2, showing a later response and nuclearrestricted upregulation (Fig. 3B); TP53, showing no changes and acting as a control (Fig.  EV3F)." 3rd Aug 2020 1st Authors' Response to Reviewers  Figure 4B and 4F -The representative images in 4B (particularly 18oC 24h, 37oC 2h) do not support the quantified data in 2F. This makes it difficult to conclude: ".... after 24h, CRY2 transcripts accumulate in the nucleus and are released into the cytoplasm after rewarming". Please include the N value of replicates used to generate this quantification.
We respectfully disagree with this comment. The representative images show an increase in CRY2 transcript nuclear dots at the 18°C 24h time point and an increase in cytoplasmic dots at the 18°C 24h, 37°C 2h time point. These dots have been counted for many images of cells exposed to each condition (see source data file and Table S7). The number of images and replicates for each condition was stated in the methods. We have now included the following statement in the figure legend: "Number of images and actual p-values are presented in Table  S7 and the source data file" and included the relevant information in the source data file. The counts for each image are also included in the source data file. All the RNA FISH images have also been deposited in the Image Data Resource (https://idr.openmicroscopy.org) under accession number idr0089. The codes used to automate the counting of dots per nuclear and cytoplasmic area have also been deposited on Github (https://github.com/hjf34/Cold). · Page 13 line 11/12 states "...rewarming following exposure to 18{degree sign}C for 5h only triggers a cytoplasmic pulse of increased REV-ERBα without affecting PER1/2 or CRY1/2 levels ( Fig. 3-4)". How can the authors conclude this when the data for CRY1/2 or PER1/2 at the 5 hr time point are not included in figure 4? The authors must include the 5 hr time for at least CRY2 to make this conclusion. Moreover, figure 3B only shows the localisation of RNA for rewarming after 24 hr 18oC and not 5 hours. Please include the data for 5 hours to support this statement.
In the original submission, we showed by RNA-seq that, out of the core circadian clock genes, REV-ERBα was the only one to show elevated transcript levels in the nucleus after 5h and by RNA-FISH and western blot that rewarming following this short exposure triggered a cytoplasmic pulse of increased REV-ERBα transcripts and an increase in REV-ERBα protein levels. We assumed that, because none of the other core circadian clock gene transcripts showed elevated levels in the nucleus, we would see no change in their cytoplasmic levels upon rewarming. We agree with the reviewer that we could not be sure of this, so we have now carried out two RNA-seq replicates for nuclear and cytoplasmic fractions of AC16 cells exposed to 18°C for 5h and then returned to 37°C for 90min. These show an increase in cytoplasmic REV-ERBα transcript levels as the nuclear levels decrease, as expected; however, there is also an increase in the cytoplasmic and nuclear levels of PER2 transcripts but not CRY1/2 or PER1 transcripts (see Fig. S1). We also observe this increase in PER2 transcript levels upon rewarming following exposure to 18°C for 24h (Fig. 3B). We have now amended the text to account for this new observation. See page 13: "Rewarming following exposure to 18°C for 5h triggers a cytoplasmic pulse of increased REV-ERBα without affecting PER1 or CRY1/2 levels ( Fig. 3-4, S1). We could therefore use this to assess the effect of this cold-induced cytoplasmic surge of REV-ERBα 3rd Aug 2020 1st Authors' Response to Reviewers on the circadian rhythm in isolation from the other clock negative limb genes, except for PER2, which shows increased transcript levels after rewarming (Fig. S1)." · The authors show that the localisation of REV-ERBa in the cytoplasm following 18oC 24 hr and rewarming 2 hr differs in AC16 and U2OS cells (Fig 4E and 4H -purple bar in cytoplasmic area), suggesting that REV-ERBa is retained/maintained in the cytoplasm following rewarming in U2OS cells but not AC16 cells. Could the authors provide and explanation as to why this may be the case?
Analysis of the RNA-seq data for U2OS cells shows that the pattern of expression changes of REV-ERBα transcripts relative to cells kept at 37°C in both the nucleus and cytoplasm is very similar to that of AC16 cells for all the time points for which we have data in both cell lines (18°C for 5h, 18°C for 24h, 18°C 24h and then back at 37°C for 2h) (compare Fig. 3B to Fig.  EV3G). RNA-FISH, unlike 3' end RNA-seq, does not require transcripts to be polyadenylated in order for their detection. The RNA-FISH data are therefore likely to detect some transcripts that are in the process of being degraded, which have had their polyA tails removed or shortened. We suspect that the degradation rate for the REV-ERBα transcripts in the cytoplasm upon rewarming may be slightly slower in U2OS cells than in AC16 cells, leading to slightly longer maintenance of these transcripts in the cytoplasm, particularly of transcripts in the process of being degraded that can be detected by RNA-FISH but not RNAseq, thus giving the higher cytoplasmic RNA-FISH signal in U2OS cells after two hours back at 37°C, following 24h at 18°C. We do not believe that this elevated level of cytoplasmic transcript signal will last for long given that the levels of nuclear transcripts are reduced by this time point. Furthermore, transcripts without poly A tails will not be translated and so will not affect REV-ERBα protein levels.  figure 5 should incorporate AC16 cells to ensure that the effects observed at the later time points are not cell line specific, and this would also increase the therapeutic relevance of the study.
As explained above we do not think that there are major differences between the patterns of expression changes shown by the two cell lines after two hours back at 37°C, following 24h at 18°C. In fact, we show that the expression changes are remarkably similar between the two cell lines for all circadian clock gene transcripts across all time points when measured using RNA-seq. We have also demonstrated that after 24h at 18°C, REV-ERBα is not required for the effects observed on the phase or amplitude of the circadian rhythm (Fig. 5A). Therefore, any difference in REV-ERBα levels between the two cell lines at this time point are not of great importance.
The U2OS PER2::LUC cell line was used to carry out the investigations into the circadian rhythm as it is a well-established cell line for such studies (e.g. Jagannath et al., 2013). We believe that as we have shown that the expression patterns for the circadian clock gene transcripts are sufficiently similar between the two cell lines, it is adequate to assume that the 3rd Aug 2020 1st Authors' Response to Reviewers circadian rhythms of these two cell lines will be affected in a similar manner to exposure to 18°C.
Minor Comments · Add N values to all figure legends.
The N values for each experiment have been included in the methods or in the source data file. For cases where inclusion of the N value for each condition would lead to overly lengthy figure legends (e.g. Fig. 4) we have included a statement in the legend explaining where to find the N values. · Figure 1B -Add missing value to axis (0.60).
This axis has now been extended to include this value. · Refer to the figures in order and provide the figure reference at the appropriate place in the text. This has now been changed to read E-H. We have checked that all other in-text references to the figures are correct.
· Figure 5A and 5B: It would be clearer to include temperature labels (@ 18oC or @ 37oC) below the time point label to be consistent with figure 3A and 3B.

3rd Aug 2020 1st Authors' Response to Reviewers
We have now included labels in this figure to clearly indicate in which time period the cells are at 37°C and at 18°C.

Referee #3
I have only few minor suggestions for this already refined work: 1. In the Abstract, in case the space allows, I would specify what kind of "human cells" the authors are working with.
With all due respect we would rather not specify the cell lines in the abstract as it may imply that the findings are cell type-specific and our results suggest that this is not the case.
2. For the experiments with U2OS/Per2::luc ReverbaKO cells -the authors thoroughly explain the deletion protocol that leads to the functional disruption of the gene. Can the authors please clarify how did they confirm that Reverba function has been disrupted?
Thank you for highlighting this point. We have now included RNA-FISH data probing for REV-ERBα transcripts in the U2OS REV-ERBα KO cell line after exposing these cells to 18°C for 24h (Fig. EV4A). This shows a large significant reduction in REV-ERBα transcript levels relative to the WT cell line showing that the deletion has disrupted REV-ERBα expression. Some signal for REV-ERBα transcripts remains in the KO cells as only the promoter and first exon of REV-ERBα has been deleted. This deletion of the first exon should also cause a frame shift in the protein, which is likely to also disrupt its function. We have amended the text in the methods to clarify that we have not directly tested whether the function has been disrupted. See page 21: "This leads to a frame shift of subsequent exons, which is likely to disrupt REV-ERBα function." 3. The authors report an intriguing finding that most of the transcriptional changes are observed at 18C, whereas nuclear transcriptome exhibited relatively mild alterations at extreme temperature of 8C. How do the authors explain this interesting observation?
In the discussion section on pages 16-17 we provide a possible explanation for this observation. We suggest that a drop in temperature to 8°C may deplete ATP to levels, which will impact on RNA polymerase II processivity. We have now added that enzyme activity at 8°C and 18°C is also likely to differ which contributes to the observed difference in the transcriptional response at the two temperatures. We suggest that at 18°C enzyme RNA polymerase II enzyme activity still sufficient to enable the gradual changes to the nuclear transcriptome observed over the 24h time period.
Accordingly, we have expanded the relevant section in the discussion and added the following lines on pages 16/17 of the revised manuscript: "Furthermore, it is likely that RNA polymerase II enzyme activity is much more severely compromised at 8°C compared to 18°C where a transcriptional response is still maintained.

3rd Aug 2020 1st Authors' Response to Reviewers
However, it is worth noting that the transcriptional response at 18°C is gradual and requires exposure for 24h to affect the transcript levels of 1000 genes (EV2F)" 4. Discussion p. 15 line 6: there are unnecessary parentheses appearing between the references by Ou et al and Jackson and Kochanek.
Thank you, we have changed this mistake in the revised version. Re: EMBOJ-2020-105604R Cold-induced chromatin compaction and nuclear retention of clock mRNAs resets the circadian rhythm Dear Andre, Thank you for submitting your revised manuscript, we have now received the reports from the three initial referees (see comments below). I am pleased to say that they overall find that their comments have been satisfactorily addressed and now support publication. Referee #2 raises a minor issue regarding the figure legends, which you will see is in part also brought up by our data editors and should be resolved in the final revised version. In addition to this, I would like to ask you to also address a number of editorial issues that are listed in detail below. Please make any changes to the manuscript text in the attached document only using the "track changes" option. Once these remaining issues are resolved, we will be happy to formally accept the manuscript for publication.
Thank you again for giving us the chance to consider your manuscript for The EMBO Journal. I look forward to receiving your final revision. Please feel free to contact me if you have further questions regarding the revision or any of the specific points listed below. This in an int erest ing piece of work and will be of value to the field. The aut hors have addressed the majorit y of my concerns. However, N values to provide significance to the dat a really need to be in the figures legends and not elsewhere in the document that is hard to access, even if this makes the legends a lit tle long.
Referee #3: The aut hors have conduct ed a thorough revision of their work, and have addressed all the raised comment aries in a sat isfact ory manner. Your article will be processed for publication in The EMBO Journal by EMBO Press and Wiley, who will contact you with further information regarding production/publication procedures and license requirements.
Should you be planning a Press Release on your article, please get in contact with embojournal@wiley.com as early as possible, in order to coordinate publication and release dates.
Congratulations on your successful publication, and thank you again for this contribution to The EMBO Journal! Please continue to consider EMBO Journal for your work in the future. 3. Were any steps taken to minimize the effects of subjective bias when allocating animals/samples to treatment (e.g. randomization procedure)? If yes, please describe.
For animal studies, include a statement about randomization even if no randomization was used.
4.a. Were any steps taken to minimize the effects of subjective bias during group allocation or/and when assessing results (e.g. blinding of the investigator)? If yes please describe.
4.b. For animal studies, include a statement about blinding even if no blinding was done 5. For every figure, are statistical tests justified as appropriate?
Do the data meet the assumptions of the tests (e.g., normal distribution)? Describe any methods used to assess it.
Is there an estimate of variation within each group of data?
Is the variance similar between the groups that are being statistically compared?
6. To show that antibodies were profiled for use in the system under study (assay and species), provide a citation, catalog number and/or clone number, supplementary information or reference to an antibody validation profile. e.g., Antibodypedia (see link list at top right), 1DegreeBio (see link list at top right).

Reporting Checklist For Life Sciences Articles (Rev. June 2017)
This checklist is used to ensure good reporting standards and to improve the reproducibility of published results. These guidelines are consistent with the Principles and Guidelines for Reporting Preclinical Research issued by the NIH in 2014. Please follow the journal's authorship guidelines in preparing your manuscript.

B-Statistics and general methods
the assay(s) and method(s) used to carry out the reported observations and measurements an explicit mention of the biological and chemical entity(ies) that are being measured. an explicit mention of the biological and chemical entity(ies) that are altered/varied/perturbed in a controlled manner.
a statement of how many times the experiment shown was independently replicated in the laboratory.
Any descriptions too long for the figure legend should be included in the methods section and/or with the source data.
In the pink boxes below, please ensure that the answers to the following questions are reported in the manuscript itself. Every question should be answered. If the question is not relevant to your research, please write NA (non applicable). We encourage you to include a specific subsection in the methods section for statistics, reagents, animal models and human subjects.
definitions of statistical methods and measures: a description of the sample collection allowing the reader to understand whether the samples represent technical or biological replicates (including how many animals, litters, cultures, etc.).
The data shown in figures should satisfy the following conditions: Source Data should be included to report the data underlying graphs. Please follow the guidelines set out in the author ship guidelines on Data Presentation.
Please fill out these boxes ê (Do not worry if you cannot see all your text once you press return) a specification of the experimental system investigated (eg cell line, species name).
Standard protocols were followed. Sample numbers are detailed for each experiment.
graphs include clearly labeled error bars for independent experiments and sample sizes. Unless justified, error bars should not be shown for technical replicates. if n< 5, the individual data points from each experiment should be plotted and any statistical test employed should be justified the exact sample size (n) for each experimental group/condition, given as a number, not a range; Each figure caption should contain the following information, for each panel where they are relevant: