Skip to main content
NCBI home page
eLife logoLink to eLife
. 2015 Mar 10;4:e04883. doi: 10.7554/eLife.04883

DNA damage shifts circadian clock time via Hausp-dependent Cry1 stabilization

Stephanie J Papp 1,†, Anne-Laure Huber 1,†, Sabine D Jordan 1, Anna Kriebs 1, Madelena Nguyen 1, James J Moresco 1, John R Yates III 1, Katja A Lamia 1,*
Editor: Joseph S Takahashi2
PMCID: PMC4352707  PMID: 25756610

Abstract

The circadian transcriptional repressors cryptochrome 1 (Cry1) and 2 (Cry2) evolved from photolyases, bacterial light-activated DNA repair enzymes. In this study, we report that while they have lost DNA repair activity, Cry1/2 adapted to protect genomic integrity by responding to DNA damage through posttranslational modification and coordinating the downstream transcriptional response. We demonstrate that genotoxic stress stimulates Cry1 phosphorylation and its deubiquitination by Herpes virus associated ubiquitin-specific protease (Hausp, a.k.a Usp7), stabilizing Cry1 and shifting circadian clock time. DNA damage also increases Cry2 interaction with Fbxl3, destabilizing Cry2. Thus, genotoxic stress increases the Cry1/Cry2 ratio, suggesting distinct functions for Cry1 and Cry2 following DNA damage. Indeed, the transcriptional response to genotoxic stress is enhanced in Cry1−/− and blunted in Cry2−/− cells. Furthermore, Cry2−/− cells accumulate damaged DNA. These results suggest that Cry1 and Cry2, which evolved from DNA repair enzymes, protect genomic integrity via coordinated transcriptional regulation.

DOI: http://dx.doi.org/10.7554/eLife.04883.001

Research organism: mouse

eLife digest

Many aspects of our physiology and behavior, most notably our patterns of sleep and wakefulness, are synchronized with the day–night cycle. These circadian rhythms are generated and maintained by the circadian clock, which consists of positive and negative feedback loops formed by a large number of genes and proteins. The end result is that the rates at which thousands of proteins are produced varies rhythmically over the course of the day–night cycle.

It has long been suspected that one of the functions of this circadian clock is to control the timing of cell division. Moreover, since UV radiation can give rise to genetic mutations when cells divide, it is thought that the circadian clock limits the amount of DNA damage that occurs during daytime. Papp, Huber et al. have now confirmed that the circadian clock does indeed participate in the DNA damage response and have revealed that two proteins known to be involved in the circadian clock—Cryptochrome 1 and 2—have a central role in protecting the integrity of the genetic information in the cell. These proteins evolved from light-activated enzymes that repair DNA in bacteria.

While mammalian cryptochromes have lost their ability to repair DNA, they still prefer to bind to genetic material that has been damaged by UV radiation. Papp, Huber et al. show that DNA damage triggers cryptochrome 1 to bind to a protein called Hausp, which stabilizes the cryptochrome and prevents it from being broken down. By contrast, DNA damage triggers cryptochrome 2 to bind to a protein called Fbxl3, which has a destabilizing effect on the cryptochrome and promotes its degradation. Since the cryptochromes regulate the activity of BMAL1 and CLOCK, the proteins associated with the two master clock genes, these changes can have a significant effect on the circadian clock of an organism.

Further experiments are needed to work out how these proteins influence the activity of BMAL1 and CLOCK, and to investigate the seemingly conflicting roles of the two cryptochromes and the interactions between them.

DOI: http://dx.doi.org/10.7554/eLife.04883.002

Introduction

Mammalian circadian clocks involve transcriptional feedback loops (Green et al., 2008): Brain and muscle ARNT-like protein 1 (BMAL1) and ‘circadian locomotor output cycles kaput’ (CLOCK) activate expression of many transcripts including the period (Per1, Per2, and Per3) and cryptochrome (Cry1 and Cry2) genes, whose protein products (PERs and CRYs) inhibit CLOCK and BMAL1, resulting in rhythmic expression. Posttranslational modifications reset the clock (Green et al., 2008), including ubiquitination and subsequent degradation of CRYs by Skp-Cullin-Fbox (SCF) E3 ligases in which substrates are recruited by F-box and leucine-rich repeat proteins 3 (FBXL3) (Busino et al., 2007; Siepka et al., 2007) and 21 (FBXL21) (Dardente et al., 2008; Hirano et al., 2013; Yoo et al., 2013). Phosphorylation of CRY1 by AMP-activated protein kinase (AMPK) increases its association with FBXL3 (Lamia et al., 2009) by disrupting interaction with PER (Xing et al., 2013). CRY stability seems to be a key factor in circadian period determination: several mutants identified in forward genetic screens selected by robust changes in circadian period have been alleles of FBXL3 or FBXL21 (Godinho et al., 2007; Siepka et al., 2007; Yoo et al., 2013).

In addition to their roles in circadian clock negative feedback, Cry1 and Cry2 are key effectors of a variety of physiological pathways. In mammals, Cry1 and Cry2 modulate glucose homeostasis by repressing the transcriptional activity of the glucocorticoid receptor (Lamia et al., 2011) and the CRE-responsive element binding protein (CREB) (Zhang et al., 2010). Consistent with these results, small molecules that stabilize Cry1/2 depress glucose production in hepatocytes and may be useful in the treatment of hyperglycemia (Hirota et al., 2012). Genetic disruption of both Cry1 and Cry2 also alters the expression of proinflammatory cytokines (Narasimamurthy et al., 2012), the severity of arthritis (Hashiramoto et al., 2010), and salt-induced blood pressure elevation (Doi et al., 2010). Genetic inactivation of Cry1 and/or Cry2 has also been reported to alter rates of tumor formation (Ozturk et al., 2009), though the reported effects have varied (Fu and Kettner, 2013). In addition, Cry-deficient mice are resistant to genotoxic stress in the context of cyclophosphamide treatment (Gorbacheva et al., 2005). Consistent with the idea that Cry1/2 may be promiscuous transcriptional repressors involved in a wide variety of physiological pathways, a recent study found that Cry1 and Cry2 each bound thousands of chromatin sites independently of other clock transcription factors in mouse liver (Koike et al., 2012).

Though Cry1 and Cry2 are mostly believed to associate with chromatin via binding a variety of transcription factors, they can also interact directly with DNA. Cry1 and Cry2 evolved from prokaryotic light-activated DNA repair enzymes, known as photolyases. While they seem to have lost the [6-4]photolyase catalytic activity characteristic of their ancestral homologs (Ozturk et al., 2007), they retain the ability to bind preferentially to UV-damaged DNA containing a [6-4]photoproduct (Ozgur and Sancar, 2003). The three-dimensional structures of Cry1 and Cry2 resemble those of photolyases, including the DNA binding surfaces (Maul et al., 2008; Czarna et al., 2013; Xing et al., 2013). Together, these properties suggest that Cry1 and Cry2 could retain a residual role in sensing or responding to damaged DNA. Such conservation of function by divergent molecular mechanisms has been seen previously between cryptochromes derived from different species (Yuan et al., 2007; Lamia et al., 2009; Kim et al., 2014).

Ubiquitination of substrate proteins by E3 ligases, like SCFFbxl3 and SCFFbxl21, is reversed by ubiquitin-specific proteases (USPs) (Eletr and Wilkinson, 2014). Herpes virus associated ubiquitin-specific protease (Hausp; a.k.a. Usp7) was first identified as the cellular partner of the herpes virus protein Vmw110 (Everett et al., 1997). Hausp modulates proliferation by catalyzing the removal of polyubiquitin chains from the tumor suppressor p53 and from the p53-destabilizing E3 ligases Mdm2 and MdmX (Li et al., 2002, 2004). The affinity of Hausp for p53 is increased and for Mdm2/MdmX is decreased in response to DNA damage (Khoronenkova et al., 2012), contributing to stabilization of p53. Knockout of Hausp in mice is lethal (Kon et al., 2010), probably due to disrupted cell proliferation. A growing list of Hausp substrates has been identified recently, including several components of DNA damage response and DNA repair pathways (Nicholson and Suresh Kumar, 2011; Schwertman et al., 2012; Jacq et al., 2013; Eletr and Wilkinson, 2014). In this study, we demonstrate that Hausp participates in DNA damage-induced resetting of circadian clock time by stabilizing Cry1.

Results

Identification of Hausp as a novel regulator of Cry1 stability

In an ongoing effort to understand the molecular determinants of cryptochrome stability, we used mass spectrometry to identify novel protein partners of mammalian CRYs and found Hausp to be the most highly enriched protein in Cry1-containing complexes (Figure 1A–B, Supplementary file 1). Co-immunoprecipitation of endogenous (Figure 1C) and overexpressed (Figure 1D–E) Cry1 and Hausp confirmed the specificity of this interaction. Interestingly, Hausp interacts much more strongly with Cry1 than with the closely related Cry2 (Figure 1D). Indeed, the divergent Cry1 C-terminus is necessary and sufficient for strong interaction with Hausp (Figure 1D–E).

Figure 1. Hausp interacts with Cry1.

(A and B) Lysates from 293T cells expressing pcDNA3-2xFLAG with no insert (−), Cry1, or Cry2 after the FLAG tag with (+) or without (−) co-expression of Per2 were used to purify control, Cry1, or Cry2-containing complexes by immunoprecipitation (IP) of the FLAG tag. 5% of each purification was analyzed by SDS-PAGE and silver stain (A) and components of the resulting complexes were identified by mass spectrometry performed on the remaining 95% of the sample. The experiment was performed in triplicate and Pattern Lab for Proteomics (Carvalho et al., 2012) was used to identify statistically enriched partners. In (B) Enrichment is the ratio of spectral counts in Cry1 vs control samples for all statistically enriched partners over three experiments (e.g., lane 1 vs lane 5 from [A]). Arrows depict several established partners for Cry1 as well as the observed 37-fold enrichment for Hausp in Cry1-containing samples. (C) Endogenous Hausp bound to endogenous Cry1 was detected by immunoblot (IB) following IP from nuclear and cytoplasmic fractions of mouse embryonic fibroblasts (MEFs) harvested at the indicated times (CT, hours) following circadian synchronization by dexamethasone. (D) Top: Hausp-V5 bound to FLAG-Cry1/2 hybrids was detected by IB following IP from 293T cells. Bottom: schematic diagram showing the composition of the Cry1/2 hybrids and domains used in D and E. (E) Hausp-V5 bound to FLAG-Cry1 full length or isolated domains was detected by IB following IP.

DOI: http://dx.doi.org/10.7554/eLife.04883.003

Figure 1.

Figure 1—figure supplement 1. Circadian measurement of Hausp mRNA expression in mouse tissues.

Figure 1—figure supplement 1.

Hausp expression was measured by quantitative RT-PCR in RNA prepared from mouse liver, quadriceps, spleen, and kidneys harvested at the indicated zeitgeber times (ZT, hours after lights on) from wild-type mice housed under normal 12:12 light:dark conditions. Data represent the mean ± s.d. for three samples at each ZT.
DOI: http://dx.doi.org/10.7554/eLife.04883.004
Figure 1—figure supplement 2. Circadian measurement of Hausp protein expression in mouse tissues.

Figure 1—figure supplement 2.

Hausp, Per2, Cry1, Cry2, Tubulin, and Lamin were measured by IB in whole cell lysates or nuclei prepared from mouse liver or quadriceps harvested at the indicated ZTs. Each lane on the gel represents a sample collected from a unique animal.
DOI: http://dx.doi.org/10.7554/eLife.04883.005
Open in a new tab

The regulation of Cry1/2 protein stability is complex and involves differential expression and localization of the E3 ligase subunits Fbxl3 and Fbxl21 that compete for Cry binding and have different rates of ubiquitin conjugation (Hirano et al., 2013; Yoo et al., 2013). Similar to what has been described for Fbxl3 and Fbxl21, we found no significant tissue specificity or circadian rhythm of expression or localization for Hausp (Figure 1—figure supplements 1, 2). However, while both Hausp and Cry1 are more abundant in the cytoplasm, their interaction is stronger in the nucleus, regardless of circadian phase (Figure 1C, Circadian Time, CT, denotes hours after dexamethasone-induced synchronization of circadian cycles).

Because Hausp is an ubiquitin-specific protease, its interaction with Cry1 in the nucleus seemed likely to stabilize nuclear Cry1 by removing polyubiquitin chains. We used small hairpin RNA (shRNA)-expressing viruses to demonstrate that Hausp depletion led to decreased Cry1 protein primarily in the nuclear compartment in mouse embryonic fibroblasts (MEFs) independent of circadian phase, as expected from the ubiquity of Hausp expression (Figure 2A, Figure 2—figure supplement 1). Treatment of cells with pharmacological inhibitors of Hausp (Nicholson and Suresh Kumar, 2011; Weinstock et al., 2012) also decreases Cry1 protein, especially in the nucleus (Figure 2B), consistent with the hypothesis that Hausp stabilizes nuclear Cry1 in vivo. (Note that compound 7 also inhibits Usp47.)

Figure 2. Hausp stabilizes Cry1 via deubiquitination and alters circadian rhythms.

(A) Wild-type or Cry1−/−;Cry2−/− (Cry−/−) MEFs stably expressing a control sequence (−) or shRNA targeting Hausp (#1) were subjected to nuclear and cytoplasmic fractionation. Cry1, Hausp, Lamin, and Tubulin were analyzed by IB from fractions harvested at the indicated times following circadian synchronization with dexamethasone (CT, hours). (B) Cry1, Hausp, Lamin and Tubulin were detected by IB in nuclear and cytoplasmic fractions from MEFs treated with vehicle (DMSO, −) or Compound 7 (+). (C) Wild-type MEFs stably expressing control or Hausp-targeting shRNA or Cry−/− MEFs were treated with vehicle (DMSO, −) or MG132 (+) for 6 hr, and lysed in RIPA buffer containing iodoacetamide. 6 mg of RIPA lysates from each condition was subjected to IP with 5 μg of anti-Cry1 antibody. Ubiquitinated Cry1 (Cry1− (Ub)N), Cry1, and Hausp were detected by IB in IPs and whole cell lysates (WCL). (D, F, H) Typical results of continuous monitoring of luciferase activity from MEFs expressing Per2-luciferase fusion protein from a knock-in allele (D and F) or from U2OS cells stably expressing luciferase under the control of the Bmal1 promoter (H) with stable expression of control or either of two shRNA sequences targeting Hausp (D) or treated with Compound 7 and/or AICAR (F and H). (E, G, I) Quantitation of the circadian period of luciferase activity from experiments performed as described in (D, F, H). Data represent the mean ± s.d. for 4–8 samples per condition. **p < 0.01, ***p < 0.001 vs control samples (control shRNA for E or DMSO-treated cells for G and I).

DOI: http://dx.doi.org/10.7554/eLife.04883.006

Figure 2.

Figure 2—figure supplement 1. Validation of shRNA targeting Hausp.

Figure 2—figure supplement 1.

Hausp expression was measured by quantitative RT-PCR in RNA prepared from MEFs stably expressing the indicated shRNA. Data represent the mean ± s.d. for three samples per cell line measured in triplicate.
DOI: http://dx.doi.org/10.7554/eLife.04883.007
Figure 2—figure supplement 2. In vitro deubiquitination of Cry1 by recombinant Hausp.

Figure 2—figure supplement 2.

Full-length and ubiquitylated Cry1 were measured by IB following in vitro exposure of purified ubiquitylated Cry1 to the indicated amounts of recombinant USP7 (Hausp) or USP8. Right: quantitation of the western blots shown at left. Data represent a typical result of three independent experiments.
DOI: http://dx.doi.org/10.7554/eLife.04883.008
Figure 2—figure supplement 3. Quantitation of in vivo Cry1 ubiquitination.

Figure 2—figure supplement 3.

Quantitation of ubiquitinated Cry1 immunoprecipitated from MEFs expressing shRNA targeting Hausp (blue) or a control sequence (black). Left, western blot from Figure 2C with boxes used for quantitation. The average signal detected in the first two lanes (background nonspecific signal from Cry1−/−;Cry2−/− cells) was subtracted from each of the other lanes to generate the data show on the right.
DOI: http://dx.doi.org/10.7554/eLife.04883.009
Open in a new tab

Recombinant Hausp can deubiquitinate Cry1 in vitro (Figure 2—figure supplement 2). To examine Cry1 ubiquitination in vivo, we measured ubiquitinated Cry1 in MEFs expressing control or Hausp-targeting shRNA in the presence or absence of the proteasome inhibitor MG132 to stabilize ubiquitinated proteins. Cry1 from Hausp-depleted cells was much more highly ubiquitinated than Cry1 from control cells (Figure 2C, Figure 2—figure supplement 3) as expected if Hausp catalyzes the removal of polyubiquitin chains from Cry1 in vivo. (Note that while Cry1 is decreased in Hausp-depleted cells, we used a limiting amount of anti-Cry1 antibody for immunoprecipitation to normalize the amount of Cry1 and enable comparison between samples. Cry−/− cells were used as a control to subtract the non-specific ubiquitin signal and enable quantitative comparison of control and Hausp-depleted cells.)

Hausp activity alters circadian period length

Cryptochrome stability is a critical determinant of circadian period length, though the direction and magnitude of the period change associated with altered expression or stability of Cry1 and/or Cry2 seems to depend on the mechanism and context of altered stability (Vitaterna et al., 1999; Hirota et al., 2012; St John et al., 2014). Nonetheless, if Hausp stabilizes Cry1 by removing ubiquitin chains, reducing Hausp expression or activity is expected to alter circadian rhythms. In immortalized fibroblasts expressing a Per2-Luciferase fusion from the endogenous Per2 locus (Per2::Luc [Yoo et al., 2004]), shRNA-mediated depletion of Hausp increased circadian period (Figure 2D,E). We also observed period lengthening in immortalized Per2::Luc MEFs when Hausp activity was inhibited pharmacologically (Figure 2F,G). Because our data suggest that Hausp inhibition and AMPK activation each destabilizes nuclear Cry1, we examined whether they could synergistically increase circadian period. Using cells stably expressing luciferase under a circadian promoter (U2OS-B6 [Vollmers et al., 2008]), we observed that activation of AMPK increased the circadian period as expected (Lamia et al., 2009), inhibition of Hausp also increased period, and combined activation of AMPK and inhibition of Hausp led to a dramatic increase in period, perhaps reflecting synergistic destabilization of nuclear Cry1 (Figure 2H,I).

DNA damage increases the Cry1/Cry2 ratio

Given that the Cry1–Hausp interaction occurs primarily in the nucleus and that Hausp interaction with other partners is regulated by DNA damage, we examined the impact of DNA damage on the Hausp–Cry1 association and found that it increases the interaction (Figure 3A, Figure 3—figure supplements 1, 2). Because Hausp catalyzes the removal of polyubiquitin chains from Cry1 thereby decreasing its proteasomal degradation (Figure 2), increased Cry1–Hausp association in response to genotoxic stress leads to a prediction that DNA damage should increase Cry1 protein levels. Consistent with this hypothesis, we found that exposure to DNA damage transiently stabilized endogenous Cry1 in primary MEFs (Figure 3A–C). Intriguingly, Cry2 was destabilized following exposure to DNA damage, demonstrating that the increase in Cry1 does not merely reflect a change or synchronization of the circadian rhythm and suggesting differential regulation of these highly homologous family members, consistent with our observation that Hausp preferentially interacts with Cry1. Because Cry1 and Cry2 each can repress the other's expression, Cry2 protein could decrease in response to damage secondary to stabilization of Cry1. However, Cry2 protein decreases and Cry1 protein increases in response to DNA damage in MEFs expressing only a single Cry paralog (i.e., Cry2 in Cry1−/− MEFs and vice versa; Figure 3B). Thus, DNA damage acutely regulates Cry1 and Cry2 protein levels independently.

Figure 3. DNA damage resets the clock via Hausp-dependent stabilization of nuclear Cry1.

(A) Endogenous Hausp, Cry1, Cry2, phospho-P53 (Ser15), P53, Lamin, and Tubulin were detected by IB in Cry1 immunoprecipitates or input samples from nuclear and cytoplasmic fractions of primary MEFs treated with vehicle (−) or doxorubicin (+). (B) Cry1, Cry2, phospho-P53 (Ser15), and Actin were detected by IB in lysates from wildtype (WT), Cry1−/− or Cry2−/− MEFs treated with doxorubicin for the indicated times. (C) Cry1, Cry2, phospho-P53 (Ser15), Hausp, Lamin, and Tubulin were detected by IB in nuclear and cytoplasmic fractions from MEFs expressing control or Hausp-targeting shRNA and treated with doxorubicin for the indicated times. (D and E) Typical results of continuous monitoring of luciferase activity from primary adult ear fibroblasts expressing Bmal1-luciferase and control or Hausp-targeting shRNA and treated with 0 (black) or 10 Gy (red) irradiation 3 hr after circadian synchronization with dexamethasone. Data represent the mean luciferase counts of eight samples per condition from one of four independent experiments. (F) Quantitation of the differences in initial circadian phase of luciferase activity caused by irradiation calculated from experiments performed as described in (D and E). Data in (D–F) represent the mean ± s.d. of phase shifts observed in four independent experiments. **p < 0.01 vs control samples.

DOI: http://dx.doi.org/10.7554/eLife.04883.010

Figure 3.

Figure 3—figure supplement 1. Effect of DNA damage on Cry1-Hausp interaction in transfected 293T cells.

Figure 3—figure supplement 1.

HAUSP-V5, FLAG-Cry1, and Actin were detected by IB in IPs or whole cell lysates (WCL) from 293T cells transfected with the indicated plasmids (by calcium phosphate method) and treated with doxorubicin.
DOI: http://dx.doi.org/10.7554/eLife.04883.011
Figure 3—figure supplement 2. Proteostasis and/or membrane stress increase the Cry1-Hausp interaction.

Figure 3—figure supplement 2.

HAUSP-V5, FLAG-Cry1, and Actin were detected by IB in IPs or whole cell lysates (WCL) from 293T cells transfected with the indicated plasmids (by calcium phosphate CaP, or PEI method) and treated with (A) Mg132 or (B) 10 Gy ionizing radiation (IR). Note that the Cry1-Hausp interaction, which can be increased by DNA damage stimuli, is independently induced by other stress signals including proteostasis stress and membrane stress. The degree to which Cry1 and/or Hausp are overexpressed in transfected cells influences their affinity of interaction. The method of transfection also altered their interaction. We consistently observed induced interaction between Cry1 and Hausp after DNA damage in cells transfected with low concentrations of plasmids using standard calcium phosphate transfection protocols. However, cells transfected with PEI or otherwise expressing excessive amounts of exogenous proteins or treated with the proteasome inhibitor Mg132 displayed high basal interaction between Cry1 and Hausp that was not further increased upon treatment with DNA damaging agents.
DOI: http://dx.doi.org/10.7554/eLife.04883.012
Figure 3—figure supplement 3. Circadian time of exposure determines phase shift in response to DNA damage.

Figure 3—figure supplement 3.

Typical results of continuous monitoring of luciferase activity from mouse embryonic fibroblasts expressing Per2::Luciferase treated with 0 (black curves) or 10 Gy (red curves) ionizing radiation 4 (CT4) or 18 (CT18) hours after circadian synchronization with 1 μM dexamethasone. Data represent the mean luciferase counts of six samples per condition from one of two independent experiments. Right: quantitation of the differences in initial circadian phase of luciferase activity caused by irradiation. Data represent the mean ± propagated s.d. difference between initial phase in Mock vs irradiated samples for six samples per condition. ***p < 0.001 for a significant interaction between CT and irradiation by 2-way ANOVA.
DOI: http://dx.doi.org/10.7554/eLife.04883.013
Open in a new tab

To determine the contribution of Hausp to the damage-induced stabilization of nuclear Cry1, we examined nuclear Cry1 protein levels following DNA damage in MEFs expressing either control sequences or Hausp-targeting shRNA. Indeed, depletion of Hausp prevents damage-induced stabilization of nuclear Cry1, similar to the effect of Hausp depletion on p53 accumulation (Figure 3C). Note that the weaker interactions that we observed between Hausp and Cry2 or hybrid constructs containing the C-terminus of Cry2 compared to those containing the Cry1 C-terminus (Figure 1D) are likely artefacts of overexpression in 293T cells since we did not observe destabilization of Cry2 upon Hausp depletion (Figure 3C).

Hausp is required for clock resetting in response to DNA damage

It has been reported that DNA damage causes phase shifts of circadian rhythms (Oklejewicz et al., 2008; Engelen et al., 2013). Consistently, we observed phase shifts in primary MEFs with a peak shift following irradiation at CT2-4, (Figure 3—figure supplement 3). The requirement for Hausp in stabilization of nuclear Cry1 after DNA damage suggested Hausp could contribute to phase shifts in response to DNA damage. By examining the circadian phase of control and Hausp-depleted fibroblasts after exposure to irradiation at CT3, we found that although the circadian phase of the non-irradiated cells is similar (Figure 2D), DNA damage-induced phase shifts were greatly diminished in Hausp-deficient fibroblasts (Figure 3D–F).

Phosphorylation of both partners influences the association between Cry1 and Hausp

ATM- and PPM1G-dependent dephosphorylation of serine 18 in the N-terminus of Hausp has been reported to drive the DNA damage dependent disruption of Hausp interaction with Mdm2 and MdmX (Khoronenkova et al., 2012). Conversely, S18 de-phosphorylation may increase Hausp–Cry1 association because mutation of S18 to the non-phosphorylatable amino acid alanine (S18A) increases interaction and mutation to aspartic acid, which is chemically similar to phospho-serine, decreases the interaction (Figure 4A). However, S18 dephosphorylation cannot fully explain DNA damage induction of Cry1–Hausp interaction as evidenced by persistent stimulated association between Cry1 and Hausp S18A after DNA damage. Intriguingly, we (Figure 4—figure supplement 1, Supplementary file 2) and others (Gao et al., 2013) find that Cry1 and Cry2 interact with kinases that are activated by DNA damage and phosphorylate serine or threonine followed by glutamine, (S/T)-Q (Kim et al., 1999; O'Neill et al., 2000). Cry1 and Cry2 contain several such sequences (Figure 4—figure supplement 2), including three serines in the Cry1 C-terminal tail that are not conserved in Cry2.

Figure 4. DNA damage induced signaling modulates interactions of Cry1/2, Hausp, and Fbxl3.

Hausp-V5, FLAG-Cry1/2, phospho-P53 (Ser15), Phospho-SQ/TQ, Phospho-Cry1S588 (P-S588), Fbxl3-V5, and Actin were detected by IB in IPs and lysates (WCL) from 293T cells transfected with the indicated plasmids and lysed at the indicated times following treatment with doxorubicin (doxo) or irradiation (IR).

DOI: http://dx.doi.org/10.7554/eLife.04883.014

Figure 4.

Figure 4—figure supplement 1. Composition of Cry1- and Cry2-associated protein complexes.

Figure 4—figure supplement 1.

(A and B) Lysates from 293T cells expressing pcDNA3-2xFLAG with no insert, Cry1, or Cry2 were used to purify control, Cry1, or Cry2-containing complexes by immunoprecipitation (IP) of the FLAG tag. Components of the resulting complexes were identified by mass spectrometry. The experiment was performed in triplicate and PatternLab for Proteomics (Carvalho et al.) was used to identify statistically enriched partners. Enrichment is the ratio of spectral counts in Cry1 or Cry2 vs control samples for the top 150 statistically enriched partners over three experiments. Arrows depict several established partners for Cry1 and Cry2 as well as the observed 18-fold and 27-fold enrichment for DNA-PKcs in Cry1- and Cry2-containing samples, respectively. (C and D) Biological processes identified by Gene Set Enrichment Analysis to be statistically enriched in Cry1-associated complexes (C) or Cry2-associated complexes (D).
DOI: http://dx.doi.org/10.7554/eLife.04883.015
Figure 4—figure supplement 2. Conserved SQ/TQ motifs present in Cry1 and/or Cry2.

Figure 4—figure supplement 2.

Sequence alignment of mouse and human Cry1 and Cry2 indicating the positions and conservation of several SQ/TQ motifs. (Numbers correspond to the amino acid positions in mouse Cry1.)
DOI: http://dx.doi.org/10.7554/eLife.04883.016
Open in a new tab

Using an antibody that recognizes phospho-(S/T)Q, we determined that Cry1 is rapidly phosphorylated in response to either chemically or radiation-induced DNA damage, while damage-induced phosphorylation of Cry2 on (S/T)Q was reduced and delayed compared to that of Cry1 (Figure 4B and not shown), indicating that the rapid phosphorylation of Cry1 in response to DNA damage likely occurs on the non-conserved C-terminal tail. Mutating each or all of the Cry1 C-terminal SQ serines to alanine decreased or abolished, respectively, the phospho-(S/T)Q signal after DNA damage, indicating that these are the sites in Cry1 that are rapidly phosphorylated in response to DNA damage (Figure 4C, and not shown). Notably, S588 is the only one of these sites on which phosphorylation has been directly detected in vivo (Lamia et al., 2009; Hegemann et al., 2011). We generated an antibody that specifically recognizes Cry1 phosphorylated on S588 and measured a rapid increase in the presence of this phosphorylated species after exposure to DNA damage (Figure 4D). Consistent with the reported stabilization of Cry1 by mimicking phosphorylation at S588, mutation of this site to aspartic acid increased its association with Hausp (Figure 4E, left).

DNA damage stimulates the association of Cry1/2 with Fbxl3

Because the effects of DNA damage on Cry1 and Cry2 stability are not fully explained by regulated interaction with Hausp, we examined the effect of DNA damage and subsequent phosphorylation events on the interactions between Cry1/2 and Fbxl3. To our surprise, prolonged exposure to DNA damage increases the interactions of Cry1 and especially Cry2 with Fbxl3 (Figure 4F), probably contributing to the transient nature of the Cry1 stabilization and to Cry2 destabilization following damage. Notably, mutation of Cry1 S588 to aspartic acid, which increases Cry1 interaction with Hausp, decreases the association of Cry1 with Fbxl3 (Figure 4E, right) suggesting that phosphorylation of the unique Cry1 C-terminus may oppose the increase in Fbxl3 binding to Cry1, possibly explaining the preferential induction of Fbxl3 binding to Cry2 compared to Cry1.

Genetic disruption of Cry1/2 alters the transcriptional response to DNA damage

Given that Cry1 and Cry2 are transcriptional repressors and that we found a robust regulation of their stability by DNA damage, we asked whether the transcriptional response to DNA damage is altered by genetic disruption of Cry1 or Cry2. By measuring the induction of transcripts activated by DNA damage in fibroblasts (Kenzelmann Broz et al., 2013), we found that genetic loss of Cry1 or Cry2 enhances or suppresses, respectively, the induction of Cdkn1a (p21) by genotoxic stress and alters the dynamic response of other established damage responsive transcripts as well (Figure 5A–E and Figure 5—figure supplement 1). Although the chromatin association of cryptochromes may be different in different cell types, both Cry1 and Cry2 bind some of these loci in mouse liver (Koike et al., 2012) (Figure 5—figure supplement 2). In addition, Cry1 and Cry2 bind to chromatin regions near several genes encoding proteins that participate in DNA repair (Supplementary file 3). Interestingly, the expression of several of those genes in response to DNA damage is also altered by genetic loss of Cry1 and/or Cry2 (Figure 5F–I), suggesting that cryptochromes may modulate the activation of DNA repair in response to damage. The regulation of some transcripts in Cry1−/−;Cry2−/− cells resembles that in Cry1−/− cells (e.g., Rrm2b, Gadd45a, p16ink4a), suggesting that Cry1 is more relevant to their regulation than is Cry2. For other transcripts (e.g., p21, Puma, Xrcc1), the response to DNA damage in Cry1−/−;Cry2−/− cells is closer to the response in Cry2−/− cells suggesting that Cry2 is more important for regulation of those targets. A full understanding of how Cry1 and Cry2 influence gene expression following DNA damage will require further study.

Figure 5. Cry1/2 deficiency alters transcriptional response to DNA damage.

Expression of the indicated transcripts was measured by quantitative PCR (qPCR) in cDNA from wildtype (black), Cry1−/− (blue), Cry2−/− (red), and Cry1−/−;Cry2−/− (gray) fibroblasts treated with doxorubicin for the indicated times. *p < 0.05, **p < 0.01, ***p < 0.001 for effect of genotype by repeated measures ANOVA analysis (blue—WT vs Cry1−/−; red—WT vs Cry2−/−; gray—WT vs Cry1−/−;Cry2−/−).

DOI: http://dx.doi.org/10.7554/eLife.04883.017

Figure 5.

Figure 5—figure supplement 1. Transcriptional response to irradiation-induced DNA damage.

Figure 5—figure supplement 1.

Expression of the indicated transcripts was measured by quantitative PCR (qPCR) in cDNA prepared from wild-type (black), Cry1−/− (blue), and Cry2−/− (red) fibroblasts at the indicated times following exposure to 5 Gy irradiation.
DOI: http://dx.doi.org/10.7554/eLife.04883.018
Figure 5—figure supplement 2. Circadian pattern of Cry1 and Cry2 binding to selected chromatin sites.

Figure 5—figure supplement 2.

Association of Cry1 (blue) or Cry2 (red) with chromatin at the indicated locations in ChIP sequencing data set published by Koike et al. (2012). Data represent the reported Cry1 or Cry2 signal normalized to the reported ‘input’ signal for each locus.
DOI: http://dx.doi.org/10.7554/eLife.04883.019
Open in a new tab

Cry2−/− cells accumulate DNA damage

We next asked whether disruption of Cry1/2-dependent transcriptional regulation causes a functional deficit in the cellular response to DNA damage in cryptochrome-deficient cells. Indeed, Cry2−/− and Cry1−/−;Cry2−/− fibroblasts accumulate damaged DNA, reflected by an increased percentage of non-dividing cells containing multiple 53BP1-positive foci (Figure 6A,B). Accumulation of damaged DNA in cells lacking Cry2 was surprising given that Cry2−/− mice are viable and fertile and that genetic disruption of Cry1 and Cry2 decreases tumor formation in p53-deficient animals (Ozturk et al., 2009). To determine whether the increased accumulation of DNA damage that we observed in cells could possibly be relevant in vivo, we analyzed breeding records of a large number of progeny from Cry1+/−;Cry2+/− mice over several years: while Cry1 genotypes segregate in the expected Mendelian ratios, the Cry2−/− genotype is significantly underrepresented (Figure 6C). This is similar to reduced survival observed in mice with genetic defects in established components of the DNA damage response or DNA repair pathways (Tsai et al., 2005; Mukherjee et al., 2010; Crossan et al., 2011) and is consistent with the possibility that animals lacking Cry2 are prone to genetic instability, though we cannot exclude other possible explanations for the reduced viability of Cry2−/− mice.

Figure 6. Cry2−/− cells accumulate damaged DNA.

Figure 6.

Open in a new tab

(A) Representative early passage (P3–4) primary wildtype (WT), Cry1−/−, Cry2−/−, and Cry1−/−;Cry2−/− adult ear fibroblasts stained with anti-53BP1 antibody (green) and DAPI (blue). Insets show enlarged view of indicated cells. (B) Quantitation of 53BP1-positive cells prepared as described in (A). Nuclei containing more than five 53BP1 puncta and negative for BrdU labeling were considered positive for DNA damage. Data represent the mean ± s.d. for at least 200 cells per genotype. (C) Chi-squared analysis of the distributions of Cry1 and Cry2 wildtype (black), heterozygous (gray), and homozygous knockout (red) genotypes establishes a significantly reduced survival of Cry2−/− mice. **p < 0.01 by chi-squared analysis with 2 degrees of freedom (χ2 = 10.39).

DOI: http://dx.doi.org/10.7554/eLife.04883.020

Discussion

Diverse organisms use circadian clocks to optimize the timing of physiological processes in relation to predictable diurnal changes in the external environment (Dodd et al., 2005; Lamia et al., 2008; Marcheva et al., 2010; Sadacca et al., 2011). It has long been suspected that clocks influence the timing of cell division to temporally separate DNA replication from predictably recurring exposure to DNA damage (Sahar and Sassone-Corsi, 2009; Sancar et al., 2010). This hypothesis is supported by the non-random distribution of cell division within circadian cycles (Nagoshi et al., 2004). Theoretically, in order for clocks to enable such a separation, their timing must be responsive to genotoxic stress, analogous to entrainment by metabolic signals, which enables clocks to optimally coordinate cellular metabolism with externally fluctuating metabolic demands (Ramsey and Bass, 2011; Jordan and Lamia, 2013). Indeed, others have shown that DNA damage shifts circadian clock time (Oklejewicz et al., 2008). In this study, we confirm that phenomenon and describe a molecular mechanism by which Hausp-mediated deubiquitination and stabilization of Cry1 contributes to it.

Furthermore, we provide evidence for specific and divergent roles of the circadian transcriptional repressors Cry1 and Cry2 in modulating the transcriptional response to DNA damage, thus addressing the longstanding question of whether circadian rhythmicity per se is sufficient to minimize the occurrence or accumulation of DNA damage. Indeed, our results suggest that circadian rhythmicity as such is not protective because Cry2−/− cells maintain robust circadian rhythms (Khan et al., 2012), but they accumulate DNA damage at rates comparable to arrhythmic Cry1−/−;Cry2−/− cells. Therefore, it appears that the genome-protective function of normal circadian rhythms is dependent on the expression of specific clock components, including Cry2. This distinction may underlie some of the controversy over the importance of circadian clocks for maintaining genomic integrity (Fu and Kettner, 2013).

Increased accumulation of DNA damage in Cry2−/− cells would be expected to lead to an increased mutation rate; consistent with that hypothesis, we observed sub-Mendelian inheritance of the Cry2−/− genotype. Though the potential for circadian clocks to influence the cellular response to DNA damage has been controversial (Gaddameedhi et al., 2012), our results are also consistent with several lines of evidence supporting a conserved role for clocks in modulating the DNA damage response and/or DNA repair (Kang et al., 2009, 2010; Cotta-Ramusino et al., 2011; Gaddameedhi et al., 2011). Interestingly, we identified several proteins that specifically bind damaged DNA or participate in DNA repair in Cry1- and/or Cry2-associated complexes (Supplementary files 1, 2), suggesting that cryptochromes may influence DNA repair by non-transcriptional mechanisms as well.

While Cry1 and Cry2 have long been established as repressors of Clock:Bmal1-driven gene expression, and we observe altered expression of several transcripts in response to DNA damage in Cry1/2-deficient cells, it remains unclear how Cry1/2 regulate gene expression. Though it is not the focus of this study, the composition of the Cry1- and Cry2-associated protein complexes suggests that regulation of mRNA processing may be an important function for Cry1 and Cry2: a large number of RNA binding and RNA processing factors were found associated with Cry1 and Cry2 (Figure 4—figure supplement 2; Supplementary files 1, 2). This is consistent with other recent literature describing the association of RNA processing machinery in complex with Per proteins (Padmanabhan et al., 2012) and the importance of post-transcriptional regulation in circadian clock function generally (Kojima et al., 2011).

It has long been hypothesized that the C-termini of cryptochromes are important for distinguishing their species-specific functions (Green, 2004) and for enabling regulated interactions with protein and possible nucleic acid partners (Czarna et al., 2011; Zoltowski et al., 2011; Engelen et al., 2013). Here, we identify a Cry1-specific partner (Hausp) that interacts with the C-terminus in isolation. Furthermore, we describe phosphorylation events in Cry1, Cry2, and Hausp that are influenced by DNA damage and contribute to their regulated interactions. It appears that DNA damage initiates a complex cascade of signal transduction that alters circadian clock dynamics in a complicated manner. The large number of phosphorylation events in Cry1, Cry2, and Hausp induced by genotoxic stress supports an important role for these proteins in sensing or responding to such stress. Understanding the specific functions of each of these modifications will require further study.

These results suggest that mammalian cryptochromes, Cry1 and Cry2, represent a node of interaction between circadian clocks and the cellular response to genotoxic stress. Cry1 and Cry2 play non-redundant roles in this pathway, highlighting the importance of analyzing their roles independently rather than relying on the exclusive use of doubly deficient cells and animals to understand their functions. Further study will determine whether Cry1 and Cry2 modulate the transcriptional response to DNA damage via sequence-specific DNA binding transcription factors or through direct binding of damaged DNA as has been observed in vitro (Ozgur and Sancar, 2003). Finally, accumulation of damaged DNA in Cry2−/− cells suggests that Cry2 is a particularly important integrator of circadian rhythms and genomic integrity. Taking all of these data together, we conclude that Cry1 and Cry2 cooperatively regulate the transcriptional response to genotoxic stress and the inverse relationship of their stability in response to DNA damage contributes to transient activation of gene networks that protect genome integrity (Figure 7).

Figure 7. Model depicting a novel mechanism by which the regulation of Cry1 and Cry2 enables coordination of the transcriptional response to genotoxic stress.

Figure 7.

Open in a new tab

In quiescent cells, Cry1 and Cry2 repress transcription of target genes. Upon DNA damage, Cry2 is degraded, relieving repression. As Cry1 accumulates, it replaces Cry2 and returns gene expression to normal levels resulting in transient activation. In Cry1−/− cells, gene expression is enhanced, while in Cry2−/− cells, damage-induced transcription is suppressed. Note that this model does not explain the dynamics of altered response observed for all transcripts but may apply to the average change in the transcriptional response to DNA damage in Cry1/2-deficient cells.

DOI: http://dx.doi.org/10.7554/eLife.04883.021

Materials and methods

Mass spectrometry

For mass spectrometry analysis, samples were denatured, reduced, and alkylated before an overnight digestion with trypsin. Peptide mixtures were analyzed by liquid chromatography mass spectrometry using an Accela pump and LTQ mass spectrometer (Thermo Fisher Scientific, Waltham, MA) using a four-step multidimensional protein identification technology (MudPIT) separation (MacCoss et al., 2002). Tandem mass spectrometry spectra were collected in a data-dependent fashion and resulting spectra were extracted using RawXtract. Protein identification was done with Integrated Proteomics Pipeline (IP2) by searching against the UniProt Human database and filtering to <1% false positive at the protein level using DTASelect. Statistically enriched partners for Cry1 were identified by Pattern Lab (Carvalho et al., 2012).

Cell culture and transfection

All cells were grown in complete Dulbecco's Modified Eagle Medium (DMEM) (cat #10569; Invitrogen) unless otherwise indicated. 293T and U2OSB6 cell media were supplemented with 10% fetal bovine serum, and 1% penicillin and streptomycin. MEF media were supplemented with 15% fetal bovine serum, and 1% penicillin and streptomycin. 293T cells were grown in a 37°C incubator maintained at 5% CO2 and 20% O2 (high oxygen). MEF cells were grown in high oxygen conditions as above (Figure 2C,D,F) or a 37°C incubator maintained at 5% CO2 and 3% O2 (low oxygen). Note that in most cases, the MEFs were initially cultured in high oxygen at the time of harvest even when they were later grown in low oxygen. MG-132 was used at a concentration of 10 μM for 6–8 hr or overnight as indicated. AICAR (cat #A61170010; Toronto Research Chemicals, Canada) was used at a concentration of 1–2 mM as indicated. Hausp inhibitor Compound 7 (Progenra, Malvern, PA) was used at a concentration of 10 μM for 6–8 hr prior to cell lysis or as indicated. Doxorubicin (cat #ICN15910101; Thermo Fisher Scientific) was used at a concentration of 0.5 μg/ml for 16–24 hr or as indicated. Ionizing radiation exposure was performed using a 137Cs γ-radiation source at the indicated times after dexamethasone synchronization. Transfections were carried out using calcium phosphate or polyethylenimine (cat #23966-2; PEI; Polysciences Inc, Warrington, PA) by standard protocols.

Plasmids and shRNA

pcDNA3-2xFlag-mCRY1, pcDNA3-2xFlag-mCRY2, and pcDNA3-Fbxl3-v5 are as described previously (Lamia et al., 2009). pCl-neo Flag HAUSP deposited by Dr Bert Vogelstein was purchased from Addgene (Addgene plasmid 16655) (Cummins and Vogelstein, 2004) and cloned into pcDNA 3.2/V5/GW-CAT purchased from Invitrogen (cat #K244020) using standard protocols. Lentiviruses expressing Bmal1-luciferase and Per2-luciferase were from Dr Satchidananda Panda. Five shRNAs against Hausp and one shRNA against Gapdh were purchased from Open Biosystems. pLKO.1 sh_scramble deposited by Dr David Sabatini was purchased from Addgene (Addgene plasmid 1864) (Sarbassov et al., 2005). Either sh_Scramble or sh_Gapdh was used as controls for sh_Hausp. pLenti-lox-GFP shRNA p19-2 for immortalizations deposited by Dr Tyler Jacks was purchased from Addgene (Addgene plasmid 14091) (Sage et al., 2003). psPAX plasmid (Addgene plasmid 12,260) and pMD2.G plasmid (Addgene plasmid 12259) deposited by Dr Didier Trono used for infection also purchased from Addgene. Cry hybrid constructs were a gift from Dr Andrew C Liu (Khan et al., 2012); the hybrid coding sequences were transferred to pcDNA3-2xFlag using standard protocols; several observed mutations in the hybrid coding sequences were corrected by site-directed mutagenesis. All mutations were generated using Agilent Site-Directed Mutagenesis kit and protocols (cat #200521).

Cell lines

HEK 293T cells were from the American Type Culture Collection (ATCC, Manassas, VA). U2OS-B6 cells were a gift from Dr Satchidananda Panda. MEFs were isolated from embryos of the indicated genotypes at E15.5 and were used as primary (passaged no more than 10 times and grown in 3% oxygen), immortalized with pLenti-lox-GFP shRNA p19-2, or spontaneously immortalized. Ear Fibroblasts were isolated from 3-month-old littermates. Ear punches were put in 70% ethanol for 2 min, washed in PBS, cut into small pieces using a scalpel and transferred to a 15-ml tube. 2 ml of trypsin 0.25% was added and samples were incubated for 1 hr at 37°C in a water bath, vortexing briefly every 10 min. The trypsin was inactivated with 8 ml of EF media (DMEM/15%FBS/PS1%). Cells were spun down 5 min at 1000 rpm, re-suspended in 3 ml of EF media and transferred into a well of a 6-well plate. The medium was changed the next day. Fibroblasts grew after 3–5 days.

Generation of viruses and stable cell lines

Lentiviral shRNA were transfected into HEK 293T cells using psPAX and pMD2.G packaging plasmids for virus generation. Viral supernatants were collected 48 hr after transfection, filtered through a 0.45-μm filter, supplemented with 6 μg/ml polybrene and added to parental cell lines. After 4 hr, additional media were added to dilute the polybrene to <3 μg/ml. 48 hr after viral transduction, the infected cells were split into selection media containing 5 μg/ml puromycin. Selection media were replaced every 2–3 days until selection was complete (as judged by death of mock-infected cells; typically 1–2 weeks).

Immunoprecipitation, western blotting, and immunostaining

293T whole cell extracts and mouse liver lysates were prepared using Lysis buffer containing 1%TX-100 as previously described (Lamia et al., 2004). MEF cell extracts were prepared from RIPA buffer containing 1% TX-100, 147 mM NaCl, 12 mM sodium deoxycolate, 0.1% SDS, 50 mM Tris pH 8.0, 10 mM EDTA, 50 μM PMSF, phosphatase inhibitors (cat #P5266 and cat #P0044; Sigma, St. Louis, MO) and protease inhibitor (cat #11697498001; Roche, Switzerland). For ubiquitination experiments, iodoacetamide was added to the buffer to a final concentration of 5 mM (Fisher AC122270050).

Antibodies were anti-Flag M2 agarose beads, anti-Flag polyclonal, anti-v5 polyclonal, anti-Lamin A, anti-aTubulin, and anti-βactin from Sigma, anti-Hausp and anti-V5 from Bethyl Labs (Montgomery, TX, cat #A300-033A and cat #A190-120A), anti-53BP1 from Novus Biologicals (Littleton, CO, NB100-304), Cry1-CT and Cry2-CT as described (Lamia et al., 2011), anti-p21 from Santa Cruz Biotechnologies (Dallas, TX, cat #sc-6246), anti-p53 as previously described (Pasini et al., 2004), and anti-Ubiquitin, anti-phospho-P53 (S15), and anti-phosphoATM/ATR substrate (phospho-SQ/TQ) from Cell Signaling Technology (Danvers, MA). Anti-Cry1-phosphoS588 antibody was affinity purified from rabbit antisera raised against a phospho-S588 containing peptide.

Immunofluorescence

Cells were grown on glass coverslips and pulse-labeled for 30 min by adding 10 µM of BrdU to the cell culture medium, washed three times with PBS before fixation with 4% (wt/vol) paraformaldehyde in PBS for 15 min at room temperature (RT) and permeabilized with 0.5% (vol/vol) Triton X-100 in PBS for 10 min at RT. Coverslips were blocked with 1% BSA in PBS for 30 min at RT. For BrdU co-staining, cells were subjected to a DNase I treatment (Sigma; 200 U/ml in 30 mM Tris HCl pH 8.1, 0.33 mM MgCl2, 0.5 mM Mercaptoethanol, 1% BSA, and 0.5% Glycerol) for 1 hr at 37°C in the presence of anti-BrdU 1/50 (BD Pharmingen). Then, coverslips were washed three times with PBS prior to incubation with primary antibodies (anti-53BP1; 1/3000) for 2 hr at RT in blocking buffer. Cells were washed with PBS and incubated with secondary antibodies (Alexa Fluor 488 goat anti-rabbit 1/150 and Alexa Fluor 594 goat anti-mouse 1/150) for 1 hr at RT in blocking buffer. Cells were then washed three times with PBS and stained 15 min with DAPI (0.4 μg/ml in PBS1X) to visualize DNA. The coverslips were mounted onto glass slides with Fluoromount G (Electron microscopy Science). For quantification, at least 200 cells were counted following IF analysis. Cells with at least five 53BP1 foci and negative for BrdU labeling were considered positive for DNA damage. Images were processed using Image J software.

Nuclear and cytoplasmic fractionation of cultured cells

Cells were washed once with ice cold PBS, fresh cold PBS was added and the cells were transferred to a 5-ml tube and centrifuged 5 min at 2000 rpm. The resulting pellets were washed with cold PBS and transferred to 1.5-ml eppendorf tubes and centrifuged 5 min at 2000 rpm. The resulting pellets were resuspended in Solution A (10 mM Hepes pH 8, 1.5 mM MgCl2, 10 mM KCl, plus protease and phosphatase inhibitors), and incubated for 15 min at 4°C. An equal volume of Solution B (solution A + 1% NP40) was added and the samples were further incubated for 5 min at 4°C and centrifuged 5 min at 3000 rpm. Supernatants from this step represent the cytoplasmic fraction. The remaining nuclear pellets were then washed twice with cold PBS, lysed in RIPA buffer and either used directly (nuclear lysates) or diluted sixfold into IP buffer for immunoprecipitation.

Lumicycle analysis of circadian period and phase shifts

U2OSB6 cells, MEFs, or adult ear fibroblasts were plated at 100% confluency in 35-mm dishes (cat #82050-538; VWR, Radnor, PA). The next day, cells were treated for 1–2 hr in normal growth medium containing 1 mM dexamethasone and 100 μM D-luciferin. Media were removed and replaced with media containing DMEM, 5% FBS, 1% penicillin-streptomycin, 15 mM Hepes, pH 7.6, and 100 μM D-luciferin. Plates were sealed with vacuum grease (Dow Corning high vacuum grease; cat #59344-055; VWR) and glass cover slips (cat #22038999; 40CIR-1, Fisher Scientific) and placed into the Lumicycle 32 from Actimetrics, Inc. (Wilmette, IL). Data were recorded using Actimetrics Lumicycle Data Collection software and analyzed using Actimetrics Lumicycle Analysis program. Background subtraction of the recorded data was performed with Running Average setting, and fit by least mean squares calculation to a damped sine wave to calculate the period, amplitude, and phase of the curves. Only data with a goodness of fit percentage of 80 or above was included in the analysis.

Quantitative RT-PCR

RNA was extracted from mouse tissues or cells with Qiazol reagent using standard protocols (cat #799306; Qiagen, Germany). cDNA was prepared using QScript cDNA Supermix (cat #101414-106; VWR) and analyzed for gene expression using quantitative real-time PCR with iQ SYBR Green Supermix (cat #1708885; Biorad, Hercules, CA). For analysis of transcriptional response to DNA damage (doxorubicin or irradiation), cells were used at approximately 70% confluency (Table 1).

Table 1.

Primers used for qPCR

DOI: http://dx.doi.org/10.7554/eLife.04883.022

Cdkn2a (p21): Fwd: CCAGGCCAAGATGGTGTCTT Rev: TGAGAAAGGATCAGCCATTGC
Mdm2: Fwd: CTGTGTCTACCGAGGGTGCT Rev: CGCTCCAACGGACTTTAACA
Rrm2b: Fwd: GACAGCAGAGGAGGTTGACTTG Rev: AAAACGCTCCACCAAGTTTTCA
Puma: Fwd: GTACGGGCGGCGGAGACGAG Rev: GCACCTAGTTGGGCTCCATTTCTG
Gadd45a: Fwd: AAGACCGAAAGGATGGACACG Rev: CAGGCACAGTACCACGTTATC
Rad23b: Fwd: ACCTTCAAGATCGACATCGACC Rev: ACTTCTGACCTGCTACCGGAA
Rad51l3: Fwd: GGAGCTTTGTGCCCAGTACC Rev: TCCCCAATGTCCCAATGTCTAT
Xrcc1: Fwd: AGCCAGGACTCGACCCATT Rev: CCTTCTCCAACTGTAGGACCA
p16ink4a: Fwd: GTGTGCATGACGTGCGGG Rev: GCAGTTCGAATCTGCACCGTAG
Rad51: Fwd: TCACCAGCGCCGGTCAGAGA Rev: CCGGCCTAAAGGTGCCCTCG
Open in a new tab

In vitro deubiquitination assay

293T cells transiently expressing Flag-tagged Cry1 were treated with 10 μM MG132 for 18 hr and lysed in RIPA buffer containing Roche complete protease inhibitors, 1 mg/ml iodoacetamide, and 50 μM PMSF. Flag-Cry1 was immunoprecipitated for 2 hr with M2-agarose (Sigma A2220), washed five times in RIPA buffer and three times in reaction buffer, and eluted for 1 hr in reaction buffer (60 mM Hepes pH 7.4, 5 mM MgCl2, 4% glycerol, 2 μg/ml aprotinin, 50 μM PMSF, 2 mg/ml BSA) containing 3XFLAG peptide. Equal volumes of eluted Flag-Cry1 were combined with the indicated amounts of recombinant Hausp (cat #E-519; USP7, Boston Biochem) or USP8 (cat #E-520; Boston Biochem, Cambridge, MA) and the reactions were incubated for 30 min at 30°C before adding SDS sample buffer and boiling for 5 min. The resulting samples were separated by 8% SDS-PAGE and Cry1 was detected by immunoblot.

Mice

Cry1−/−;Cry2−/− mice were from Dr Aziz Sancar (Thresher et al., 1998); Per2::Luciferase mice (Yoo et al., 2004) were purchased from Jackson laboratories (Bar Harbor, ME). All animal care and treatments were in accordance with The Scripps Research Institute guidelines for the care and use of animals under protocol #10-0019.

Acknowledgements

We thank Drs Andrew Liu (University of Memphis), Satchidananda Panda (The Salk Institute), Eros Lazzerini Denchi (The Scripps Research Institute), and Ben Nicholson (Progenra, Inc.) for providing materials and reagents and Drs Eros Lazzerini Denchi, Supriya Srinivasan, Reuben Shaw, and Joseph Bass for helpful discussions and critical reading of the manuscript. KAL, SJP, ALH, AK, SDJ, and MN were supported by the Searle Scholars Fund, the Sidney Kimmel Foundation for Cancer Research, the Lung Cancer Research Foundation, the National Institute of Diabetes, Digestive and Kidney Diseases (K01DK090188-03 and R01DK097164-01), and a research fellowship from the Deutsche Forschungsgemeinschaft (DFG, to SDJ). JJM and JRY were supported by the National Institute of General Medical Sciences (8 P41 GM103533) and the National Institute on Aging (R01AG027463).

Funding Statement

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Funding Information

This paper was supported by the following grants:

  • National Institutes of Health (NIH) K01, DK090188 to Katja A Lamia.

  • Kinship Foundation Searle Scholars Award to Katja A Lamia.

  • Sidney Kimmel Foundation for Cancer Research Cancer Scholar Award to Katja A Lamia.

  • Lung Cancer Research Foundation (LCRF) research project grant to Katja A Lamia.

  • National Institutes of Health (NIH) R01, DK097164 to Katja A Lamia.

  • National Institutes of Health (NIH) R01, AG027463 to John R Yates.

  • National Institutes of Health (NIH) P41, GM103533 to John R Yates.

  • Deutsche Forschungsgemeinschaft research fellowship to Sabine D Jordan.

Additional information

Competing interests

The authors declare that no competing interests exist.

Author contributions

SJP, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article.

A-LH, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article.

SDJ, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article.

KAL, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article.

AK, Acquisition of data, Analysis and interpretation of data, Contributed unpublished essential data or reagents.

MN, Acquisition of data, Analysis and interpretation of data.

JJM, Acquisition of data, Analysis and interpretation of data.

JRY, Acquisition of data, Analysis and interpretation of data.

Ethics

Animal experimentation: This study was performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. All of the animals were handled according to approved institutional animal care and use committee (IACUC) protocols (#10-0019) of The Scripps Research Institute.

Additional files

Supplementary file 1.

Cry1-associated proteins. Lysates from 293T cells expressing pcDNA3-2xFLAG with no insert (control) or Cry1 after the FLAG tag were used to purify control or Cry1-containing complexes by immunoprecipitation (IP) of the FLAG tag. Components of the resulting complexes were identified by mass spectrometry. The experiment was performed in triplicate and PatternLab for Proteomics (Carvalho et al.) was used to identify statistically enriched partners in Cry1-associated complexes compared to the control. Enrichment (Cry1/control) is the ratio of spectral counts in Cry1 vs control samples for all statistically enriched partners over three experiments.

DOI: http://dx.doi.org/10.7554/eLife.04883.023

elife04883s001.xlsx (29.7KB, xlsx)
DOI: 10.7554/eLife.04883.023
Supplementary file 2.

Cry2-associated proteins. Lysates from 293T cells expressing pcDNA3-2xFLAG with no insert (control) or Cry2 after the FLAG tag were used to purify control or Cry1-containing complexes by immunoprecipitation (IP) of the FLAG tag. Components of the resulting complexes were identified by mass spectrometry. The experiment was performed in triplicate and PatternLab for Proteomics (Carvalho et al.) was used to identify statistically enriched partners in Cry2-associated complexes compared to the control. Enrichment (Cry2/control) is the ratio of spectral counts in Cry2 vs control samples for all statistically enriched partners over three experiments.

DOI: http://dx.doi.org/10.7554/eLife.04883.024

elife04883s002.xlsx (31.3KB, xlsx)
DOI: 10.7554/eLife.04883.024
Supplementary file 3.

Chromatin binding of circadian transcription factors to loci encoding DNA repair proteins. Published data (Koike et al., 2012, Table S2) was searched for the text string ‘repair’ to make a preliminary identification of chromatin regions near genes involved in DNA repair that were found to be associated with each of the seven circadian transcription factors Cry1, Cry2, Per1, Per2, Clock, Npas2, and Bmal1.

DOI: http://dx.doi.org/10.7554/eLife.04883.025

elife04883s003.xlsx (46.5KB, xlsx)
DOI: 10.7554/eLife.04883.025

References

  1. Busino L, Bassermann F, Maiolica A, Lee C, Nolan PM, Godinho SI, Draetta GF, Pagano M. SCFFbxl3 controls the oscillation of the circadian clock by directing the degradation of cryptochrome proteins. Science. 2007;316:900–904. doi: 10.1126/science.1141194. [DOI] [PubMed] [Google Scholar]
  2. Carvalho PC, Yates JR, III, Barbosa VC. Improving the TFold test for differential shotgun proteomics. Bioinformatics. 2012;28:1652–1654. doi: 10.1093/bioinformatics/bts247. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Cotta-Ramusino C, McDonald ER, III, Hurov K, Sowa ME, Harper JW, Elledge SJ. A DNA damage response screen identifies RHINO, a 9-1-1 and TopBP1 interacting protein required for ATR signaling. Science. 2011;332:1313–1317. doi: 10.1126/science.1203430. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Crossan GP, van der Weyden L, Rosado IV, Langevin F, Gaillard PH, McIntyre RE, Sanger Mouse Genetics Project. Gallagher F, Kettunen MI, Lewis DY, Brindle K, Arends MJ, Adams DJ, Patel KJ. Disruption of mouse Slx4, a regulator of structure-specific nucleases, phenocopies Fanconi anemia. Nature Genetics. 2011;43:147–152. doi: 10.1038/ng.752. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Cummins JM, Vogelstein B. HAUSP is required for p53 destabilization. Cell Cycle. 2004;3:689–692. doi: 10.4161/cc.3.6.924. [DOI] [PubMed] [Google Scholar]
  6. Czarna A, Berndt A, Singh HR, Grudziecki A, Ladurner AG, Timinszky G, Kramer A, Wolf E. Structures of Drosophila cryptochrome and mouse cryptochrome1 provide insight into circadian function. Cell. 2013;153:1394–1405. doi: 10.1016/j.cell.2013.05.011. [DOI] [PubMed] [Google Scholar]
  7. Czarna A, Breitkreuz H, Mahrenholz CC, Arens J, Strauss HM, Wolf E. Quantitative analyses of cryptochrome-mBMAL1 interactions: mechanistic insights into the transcriptional regulation of the mammalian circadian clock. The Journal of Biological Chemistry. 2011;286:22414–22425. doi: 10.1074/jbc.M111.244749. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Dardente H, Mendoza J, Fustin JM, Challet E, Hazlerigg DG. Implication of the F-Box Protein FBXL21 in circadian pacemaker function in mammals. PLOS ONE. 2008;3:e3530. doi: 10.1371/journal.pone.0003530. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Dodd AN, Salathia N, Hall A, Kevei E, Tóth R, Nagy F, Hibberd JM, Millar AJ, Webb AA. Plant circadian clocks increase photosynthesis, growth, survival, and competitive advantage. Science. 2005;309:630–633. doi: 10.1126/science.1115581. [DOI] [PubMed] [Google Scholar]
  10. Doi M, Takahashi Y, Komatsu R, Yamazaki F, Yamada H, Haraguchi S, Emoto N, Okuno Y, Tsujimoto G, Kanematsu A, Ogawa O, Todo T, Tsutsui K, van der Horst GT, Okamura H. Salt-sensitive hypertension in circadian clock-deficient Cry-null mice involves dysregulated adrenal Hsd3b6. Nature Medicine. 2010;16:67–74. doi: 10.1038/nm.2061. [DOI] [PubMed] [Google Scholar]
  11. Eletr ZM, Wilkinson KD. Regulation of proteolysis by human deubiquitinating enzymes. Biochimica Et Biophysica Acta. 2014;1843:114–128. doi: 10.1016/j.bbamcr.2013.06.027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Engelen E, Janssens RC, Yagita K, Smits VA, van der Horst GT, Tamanini F. Mammalian TIMELESS is involved in period determination and DNA damage-dependent phase advancing of the circadian clock. PLOS ONE. 2013;8:e56623. doi: 10.1371/journal.pone.0056623. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Everett RD, Meredith M, Orr A, Cross A, Kathoria M, Parkinson J. A novel ubiquitin-specific protease is dynamically associated with the PML nuclear domain and binds to a herpesvirus regulatory protein. The EMBO Journal. 1997;16:1519–1530. doi: 10.1093/emboj/16.7.1519. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Fu L, Kettner NM. The circadian clock in cancer development and therapy. Progress in Molecular Biology and Translational Science. 2013;119:221–282. doi: 10.1016/B978-0-12-396971-2.00009-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Gaddameedhi S, Reardon JT, Ye R, Ozturk N, Sancar A. Effect of circadian clock mutations on DNA damage response in mammalian cells. Cell cycle. 2012;11:3481–3491. doi: 10.4161/cc.21771. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Gaddameedhi S, Selby CP, Kaufmann WK, Smart RC, Sancar A. Control of skin cancer by the circadian rhythm. Proceedings of the National Academy of Sciences of USA. 2011;108:18790–18795. doi: 10.1073/pnas.1115249108. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Gao P, Yoo SH, Lee KJ, Rosensweig C, Takahashi JS, Chen BP, Green CB. Phosphorylation of the cryptochrome 1 C-terminal tail regulates circadian period length. The Journal of Biological Chemistry. 2013;288:35277–35286. doi: 10.1074/jbc.M113.509604. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Godinho SI, Maywood ES, Shaw L, Tucci V, Barnard AR, Busino L, Pagano M, Kendall R, Quwailid MM, Romero MR, O'neill J, Chesham JE, Brooker D, Lalanne Z, Hastings MH, Nolan PM. The after-hours mutant reveals a role for Fbxl3 in determining mammalian circadian period. Science. 2007;316:897–900. doi: 10.1126/science.1141138. [DOI] [PubMed] [Google Scholar]
  19. Gorbacheva VY, Kondratov RV, Zhang R, Cherukuri S, Gudkov AV, Takahashi JS, Antoch MP. Circadian sensitivity to the chemotherapeutic agent cyclophosphamide depends on the functional status of the CLOCK/BMAL1 transactivation complex. Proceedings of the National Academy of Sciences of USA. 2005;102:3407–3412. doi: 10.1073/pnas.0409897102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Green CB. Cryptochromes: tail-ored for distinct functions. Current Biology. 2004;14:R847–R849. doi: 10.1016/j.cub.2004.09.040. [DOI] [PubMed] [Google Scholar]
  21. Green CB, Takahashi JS, Bass J. The meter of metabolism. Cell. 2008;134:728–742. doi: 10.1016/j.cell.2008.08.022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Hashiramoto A, Yamane T, Tsumiyama K, Yoshida K, Komai K, Yamada H, Yamazaki F, Doi M, Okamura H, Shiozawa S. Mammalian clock gene Cryptochrome regulates arthritis via proinflammatory cytokine TNF-alpha. Journal of immunology. 2010;184:1560–1565. doi: 10.4049/jimmunol.0903284. [DOI] [PubMed] [Google Scholar]
  23. Hegemann B, Hutchins JR, Hudecz O, Novatchkova M, Rameseder J, Sykora MM, Liu S, Mazanek M, Lénárt P, Hériché JK, Poser I, Kraut N, Hyman AA, Yaffe MB, Mechtler K, Peters JM. Systematic phosphorylation analysis of human mitotic protein complexes. Science Signaling. 2011;4:rs12. doi: 10.1126/scisignal.2001993. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Hirano A, Yumimoto K, Tsunematsu R, Matsumoto M, Oyama M, Kozuka-Hata H, Nakagawa T, Lanjakornsiripan D, Nakayama KI, Fukada Y. FBXL21 regulates oscillation of the circadian clock through ubiquitination and stabilization of cryptochromes. Cell. 2013;152:1106–1118. doi: 10.1016/j.cell.2013.01.054. [DOI] [PubMed] [Google Scholar]
  25. Hirota T, Lee JW, St John PC, Sawa M, Iwaisako K, Noguchi T, Pongsawakul PY, Sonntag T, Welsh DK, Brenner DA, Doyle FJ, III, Schultz PG, Kay SA. Identification of small molecule activators of cryptochrome. Science. 2012;337:1094–1097. doi: 10.1126/science.1223710. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Jacq X, Kemp M, Martin NM, Jackson SP. Deubiquitylating enzymes and DNA damage response pathways. Cell Biochemistry and Biophysics. 2013;67:25–43. doi: 10.1007/s12013-013-9635-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Jordan SD, Lamia KA. AMPK at the crossroads of circadian clocks and metabolism. Molecular and Cellular Endocrinology. 2013;366:163–169. doi: 10.1016/j.mce.2012.06.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Kang TH, Lindsey-Boltz LA, Reardon JT, Sancar A. Circadian control of XPA and excision repair of cisplatin-DNA damage by cryptochrome and HERC2 ubiquitin ligase. Proceedings of the National Academy of Sciences of USA. 2010;107:4890–4895. doi: 10.1073/pnas.0915085107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Kang TH, Reardon JT, Kemp M, Sancar A. Circadian oscillation of nucleotide excision repair in mammalian brain. Proceedings of the National Academy of Sciences of USA. 2009;106:2864–2867. doi: 10.1073/pnas.0812638106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Kenzelmann Broz D, Spano Mello S, Bieging KT, Jiang D, Dusek RL, Brady CA, Sidow A, Attardi LD. Global genomic profiling reveals an extensive p53-regulated autophagy program contributing to key p53 responses. Genes & Development. 2013;27:1016–1031. doi: 10.1101/gad.212282.112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Khan SK, Xu H, Ukai-Tadenuma M, Burton B, Wang Y, Ueda HR, Liu AC. Identification of a novel cryptochrome differentiating domain required for feedback repression in circadian clock function. The Journal of Biological Chemistry. 2012;287:25917–25926. doi: 10.1074/jbc.M112.368001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Khoronenkova SV, Dianova, Ternette N, Kessler BM, Parsons JL, Dianov GL. ATM-dependent downregulation of USP7/HAUSP by PPM1G activates p53 response to DNA damage. Molecular Cell. 2012;45:801–813. doi: 10.1016/j.molcel.2012.01.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Kim ST, Lim DS, Canman CE, Kastan MB. Substrate specificities and identification of putative substrates of ATM kinase family members. The Journal of Biological Chemistry. 1999;274:37538–37543. doi: 10.1074/jbc.274.53.37538. [DOI] [PubMed] [Google Scholar]
  34. Kim YM, Choi J, Lee HY, Lee GW, Lee YH, Choi D. dbCRY: a Web-based comparative and evolutionary genomics platform for blue-light receptors. Database. 2014;2014:bau037. doi: 10.1093/database/bau037. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Koike N, Yoo SH, Huang HC, Kumar V, Lee C, Kim TK, Takahashi JS. Transcriptional architecture and chromatin landscape of the core circadian clock in mammals. Science. 2012;338:349–354. doi: 10.1126/science.1226339. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Kojima S, Shingle DL, Green CB. Post-transcriptional control of circadian rhythms. Journal of Cell Science. 2011;124:311–320. doi: 10.1242/jcs.065771. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Kon N, Kobayashi Y, Li M, Brooks CL, Ludwig T, Gu W. Inactivation of HAUSP in vivo modulates p53 function. Oncogene. 2010;29:1270–1279. doi: 10.1038/onc.2009.427. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Lamia KA, Papp SJ, Yu RT, Barish GD, Uhlenhaut NH, Jonker JW, Downes M, Evans RM. Cryptochromes mediate rhythmic repression of the glucocorticoid receptor. Nature. 2011;480:552–556. doi: 10.1038/nature10700. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Lamia KA, Peroni OD, Kim YB, Rameh LE, Kahn BB, Cantley LC. Increased insulin sensitivity and reduced adiposity in phosphatidylinositol 5-phosphate 4-kinase beta-/- mice. Molecular and Cellular Biology. 2004;24:5080–5087. doi: 10.1128/MCB.24.11.5080-5087.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Lamia KA, Sachdeva UM, DiTacchio L, Williams EC, Alvarez JG, Egan DF, Vasquez DS, Juguilon H, Panda S, Shaw RJ, Thompson CB, Evans RM. AMPK regulates the circadian clock by cryptochrome phosphorylation and degradation. Science. 2009;326:437–440. doi: 10.1126/science.1172156. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Lamia KA, Storch KF, Weitz CJ. Physiological significance of a peripheral tissue circadian clock. Proceedings of the National Academy of Sciences of USA. 2008;105:15172–15177. doi: 10.1073/pnas.0806717105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Li M, Brooks CL, Kon N, Gu W. A dynamic role of HAUSP in the p53-Mdm2 pathway. Molecular Cell. 2004;13:879–886. doi: 10.1016/S1097-2765(04)00157-1. [DOI] [PubMed] [Google Scholar]
  43. Li M, Chen D, Shiloh A, Luo J, Nikolaev AY, Qin J, Gu W. Deubiquitination of p53 by HAUSP is an important pathway for p53 stabilization. Nature. 2002;416:648–653. doi: 10.1038/nature737. [DOI] [PubMed] [Google Scholar]
  44. MacCoss MJ, McDonald WH, Saraf A, Sadygov R, Clark JM, Tasto JJ, Gould KL, Wolters D, Washburn M, Weiss A, Clark JI, Yates JR., III Shotgun identification of protein modifications from protein complexes and lens tissue. Proceedings of the National Academy of Sciences of USA. 2002;99:7900–7905. doi: 10.1073/pnas.122231399. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Marcheva B, Ramsey KM, Buhr ED, Kobayashi Y, Su H, Ko CH, Ivanova G, Omura C, Mo S, Vitaterna MH, Lopez JP, Philipson LH, Bradfield CA, Crosby SD, JeBailey L, Wang X, Takahashi JS, Bass J. Disruption of the clock components CLOCK and BMAL1 leads to hypoinsulinaemia and diabetes. Nature. 2010;466:627–631. doi: 10.1038/nature09253. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Maul MJ, Barends TR, Glas AF, Cryle MJ, Domratcheva T, Schneider S, Schlichting I, Carell T. Crystal structure and mechanism of a DNA (6-4) photolyase. Angewandte Chemie. 2008;47:10076–10080. doi: 10.1002/anie.200804268. [DOI] [PubMed] [Google Scholar]
  47. Mukherjee S, Ridgeway AD, Lamb DJ. DNA mismatch repair and infertility. Current Opinion in Urology. 2010;20:525–532. doi: 10.1097/MOU.0b013e32833f1c21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Nagoshi E, Saini C, Bauer C, Laroche T, Naef F, Schibler U. Circadian gene expression in individual fibroblasts: cell-autonomous and self-sustained oscillators pass time to daughter cells. Cell. 2004;119:693–705. doi: 10.1016/j.cell.2004.11.015. [DOI] [PubMed] [Google Scholar]
  49. Narasimamurthy R, Hatori M, Nayak SK, Liu F, Panda S, Verma IM. Circadian clock protein cryptochrome regulates the expression of proinflammatory cytokines. Proceedings of the National Academy of Sciences of USA. 2012;109:12662–12667. doi: 10.1073/pnas.1209965109. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Nicholson B, Suresh Kumar KG. The multifaceted roles of USP7: new therapeutic opportunities. Cell Biochemistry and Biophysics. 2011;60:61–68. doi: 10.1007/s12013-011-9185-5. [DOI] [PubMed] [Google Scholar]
  51. O'Neill T, Dwyer AJ, Ziv Y, Chan DW, Lees-Miller SP, Abraham RH, Lai JH, Hill D, Shiloh Y, Cantley LC, Rathbun GA. Utilization of oriented peptide libraries to identify substrate motifs selected by ATM. The Journal of Biological Chemistry. 2000;275:22719–22727. doi: 10.1074/jbc.M001002200. [DOI] [PubMed] [Google Scholar]
  52. Oklejewicz M, Destici E, Tamanini F, Hut RA, Janssens R, van der Horst GT. Phase resetting of the mammalian circadian clock by DNA damage. Current Biology. 2008;18:286–291. doi: 10.1016/j.cub.2008.01.047. [DOI] [PubMed] [Google Scholar]
  53. Ozgur S, Sancar A. Purification and properties of human blue-light photoreceptor cryptochrome 2. Biochemistry. 2003;42:2926–2932. doi: 10.1021/bi026963n. [DOI] [PubMed] [Google Scholar]
  54. Ozturk N, Lee JH, Gaddameedhi S, Sancar A. Loss of cryptochrome reduces cancer risk in p53 mutant mice. Proceedings of the National Academy of Sciences of USA. 2009;106:2841–2846. doi: 10.1073/pnas.0813028106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Ozturk N, Song SH, Ozgur S, Selby CP, Morrison L, Partch C, Zhong D, Sancar A. Structure and function of animal cryptochromes. Cold Spring Harbor Symposia on Quantitative Biology. 2007;72:119–131. doi: 10.1101/sqb.2007.72.015. [DOI] [PubMed] [Google Scholar]
  56. Padmanabhan K, Robles MS, Westerling T, Weitz CJ. Feedback regulation of transcriptional termination by the mammalian circadian clock PERIOD complex. Science. 2012;337:599–602. doi: 10.1126/science.1221592. [DOI] [PubMed] [Google Scholar]
  57. Pasini D, Bracken AP, Jensen MR, Lazzerini Denchi E, Helin K. Suz12 is essential for mouse development and for EZH2 histone methyltransferase activity. The EMBO Journal. 2004;23:4061–4071. doi: 10.1038/sj.emboj.7600402. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Ramsey KM, Bass J. Circadian clocks in fuel harvesting and energy homeostasis. Cold Spring Harbor Symposia on Quantitative Biology. 2011;76:63–72. doi: 10.1101/sqb.2011.76.010546. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Sadacca LA, Lamia KA, deLemos AS, Blum B, Weitz CJ. An intrinsic circadian clock of the pancreas is required for normal insulin release and glucose homeostasis in mice. Diabetologia. 2011;54:120–124. doi: 10.1007/s00125-010-1920-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Sage J, Miller AL, Pérez-Mancera PA, Wysocki JM, Jacks T. Acute mutation of retinoblastoma gene function is sufficient for cell cycle re-entry. Nature. 2003;424:223–228. doi: 10.1038/nature01764. [DOI] [PubMed] [Google Scholar]
  61. Sahar S, Sassone-Corsi P. Metabolism and cancer: the circadian clock connection. Nature Reviews Cancer. 2009;9:886–896. doi: 10.1038/nrc2747. [DOI] [PubMed] [Google Scholar]
  62. Sancar A, Lindsey-Boltz LA, Kang TH, Reardon JT, Lee JH, Ozturk N. Circadian clock control of the cellular response to DNA damage. FEBS Letters. 2010;584:2618–2625. doi: 10.1016/j.febslet.2010.03.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Sarbassov DD, Guertin DA, Ali SM, Sabatini DM. Phosphorylation and regulation of Akt/PKB by the rictor-mTOR complex. Science. 2005;307:1098–1101. doi: 10.1126/science.1106148. [DOI] [PubMed] [Google Scholar]
  64. Schwertman P, Lagarou A, Dekkers DH, Raams A, van der Hoek AC, Laffeber C, Hoeijmakers JH, Demmers JA, Fousteri M, Vermeulen W, Marteijn JA. UV-sensitive syndrome protein UVSSA recruits USP7 to regulate transcription-coupled repair. Nature Genetics. 2012;44:598–602. doi: 10.1038/ng.2230. [DOI] [PubMed] [Google Scholar]
  65. Siepka SM, Yoo SH, Park J, Song W, Kumar V, Hu Y, Lee C, Takahashi JS. Circadian mutant Overtime reveals F-box protein FBXL3 regulation of cryptochrome and period gene expression. Cell. 2007;129:1011–1023. doi: 10.1016/j.cell.2007.04.030. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. St John PC, Hirota T, Kay SA, Doyle FJ., III Spatiotemporal separation of PER and CRY posttranslational regulation in the mammalian circadian clock. Proceedings of the National Academy of Sciences of USA. 2014;111:2040–2045. doi: 10.1073/pnas.1323618111. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Thresher RJ, Vitaterna MH, Miyamoto Y, Kazantsev A, Hsu DS, Petit C, Selby CP, Dawut L, Smithies O, Takahashi JS, Sancar A. Role of mouse cryptochrome blue-light photoreceptor in circadian photoresponses. Science. 1998;282:1490–1494. doi: 10.1126/science.282.5393.1490. [DOI] [PubMed] [Google Scholar]
  68. Tsai PS, Nielen M, van der Horst GT, Colenbrander B, Heesterbeek JA, van Vlissingen JM. The effect of DNA repair defects on reproductive performance in nucleotide excision repair (NER) mouse models: an epidemiological approach. Transgenic Research. 2005;14:845–857. doi: 10.1007/s11248-005-1772-z. [DOI] [PubMed] [Google Scholar]
  69. Vitaterna MH, Selby CP, Todo T, Niwa H, Thompson C, Fruechte EM, Hitomi K, Thresher RJ, Ishikawa T, Miyazaki J, Takahashi JS, Sancar A. Differential regulation of mammalian period genes and circadian rhythmicity by cryptochromes 1 and 2. Proceedings of the National Academy of Sciences of USA. 1999;96:12114–12119. doi: 10.1073/pnas.96.21.12114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  70. Vollmers C, Panda S, DiTacchio L. A high-throughput assay for siRNA-based circadian screens in human U2OS cells. PLOS ONE. 2008;3:e3457. doi: 10.1371/journal.pone.0003457. [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Weinstock J, Wu J, Cao P, Kingsbury WD, McDermott JL, Kodrasov MP, McKelvey DM, Suresh Kumar KG, Goldenberg SJ, Mattern MR, Nicholson B. Selective Dual inhibitors of the Cancer-Related Deubiquitylating proteases USP7 and USP47. ACS Medicinal Chemistry Letters. 2012;3:789–792. doi: 10.1021/ml200276j. [DOI] [PMC free article] [PubMed] [Google Scholar]
  72. Xing W, Busino L, Hinds TR, Marionni ST, Saifee NH, Bush MF, Pagano M, Zheng N. SCF(FBXL3) ubiquitin ligase targets cryptochromes at their cofactor pocket. Nature. 2013;496:64–68. doi: 10.1038/nature11964. [DOI] [PMC free article] [PubMed] [Google Scholar]
  73. Yoo SH, Mohawk JA, Siepka SM, Shan Y, Huh SK, Hong HK, Kornblum I, Kumar V, Koike N, Xu M, Nussbaum J, Liu X, Chen Z, Chen ZJ, Green CB, Takahashi JS. Competing E3 ubiquitin ligases govern circadian periodicity by degradation of CRY in nucleus and cytoplasm. Cell. 2013;152:1091–1105. doi: 10.1016/j.cell.2013.01.055. [DOI] [PMC free article] [PubMed] [Google Scholar]
  74. Yoo SH, Yamazaki S, Lowrey PL, Shimomura K, Ko CH, Buhr ED, Siepka SM, Hong HK, Oh WJ, Yoo OJ, Menaker M, Takahashi JS. PERIOD2::LUCIFERASE real-time reporting of circadian dynamics reveals persistent circadian oscillations in mouse peripheral tissues. Proceedings of the National Academy of Sciences of USA. 2004;101:5339–5346. doi: 10.1073/pnas.0308709101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  75. Yuan Q, Metterville D, Briscoe AD, Reppert SM. Insect cryptochromes: gene duplication and loss define diverse ways to construct insect circadian clocks. Molecular Biology and Evolution. 2007;24:948–955. doi: 10.1093/molbev/msm011. [DOI] [PubMed] [Google Scholar]
  76. Zhang EE, Liu Y, Dentin R, Pongsawakul PY, Liu AC, Hirota T, Nusinow DA, Sun X, Landais S, Kodama Y, Brenner DA, Montminy M, Kay SA. Cryptochrome mediates circadian regulation of cAMP signaling and hepatic gluconeogenesis. Nature Medicine. 2010;16:1152–1156. doi: 10.1038/nm.2214. [DOI] [PMC free article] [PubMed] [Google Scholar]
  77. Zoltowski BD, Vaidya AT, Top D, Widom J, Young MW, Crane BR. Structure of full-length Drosophila cryptochrome. Nature. 2011;480:396–399. doi: 10.1038/nature10618. [DOI] [PMC free article] [PubMed] [Google Scholar]
eLife. 2015 Mar 10;4:e04883. doi: 10.7554/eLife.04883.026

Decision letter

Editor: Joseph S Takahashi1

eLife posts the editorial decision letter and author response on a selection of the published articles (subject to the approval of the authors). An edited version of the letter sent to the authors after peer review is shown, indicating the substantive concerns or comments; minor concerns are not usually shown. Reviewers have the opportunity to discuss the decision before the letter is sent (see review process). Similarly, the author response typically shows only responses to the major concerns raised by the reviewers.

Thank you for sending your work entitled “Circadian clocks protect genome integrity via Cry1/2-coordinated transcription” for consideration at eLife. Your article has been favorably evaluated by Janet Rossant (Senior editor) and 3 reviewers, one of whom, Joseph S Takahashi, served as the guest Reviewing editor.

The Reviewing editor and the other reviewers discussed their comments before we reached this decision, and the Reviewing editor has assembled the following comments to help you prepare a revised submission.

In a proteomic screen for Cry1 interactors, Papp and colleagues find Usp7 (Hausp), and show that the interaction is mediated by the c-terminus of Cry1. In the first part of the manuscript, the authors detail their discovery of a DNA damage-regulated interaction of Cry1 with the deubiquitinase, Hausp. They show that this interaction is specific for Cry1, mediated by its unique C-terminus, and that DNA damage-induced phosphorylation of SQ/TQ sites in the Cry1 C-terminus play a role in assembling the complex. Stabilization of Cry1 after deubiquitination by Hausp occurs only in the nucleus. Importantly, they demonstrate a reciprocal relationship between the damage-induced interaction with Hausp (stabilizing) and interaction with the E3 ubiquitin ligase, Fbxl3 (destabilizing). In the absence of stabilizing Hausp-interactions with Cry2, the authors show that Cry2 undergoes a net destabilization after DNA damage. The authors present data that Hausp is required for DNA damage-dependent phase shifting of the clock, presumably through regulation of Cry1 protein levels. These data are consistent with findings from Oklejewicz et al. that DNA damage induces circadian phase advances through a transcription-independent mechanism. Together, these data provide one of the first molecular descriptions of how DNA damage impacts the stability of circadian proteins.

Comments to be addressed:

1a) Depletion of Usp7 leads to decreased Cry1 expression. However, it also leads to long period length. In contrast, loss of Cry1 by RNAi or mouse knockouts leads to shortened period length, while loss of Fbxl3, which targets Crys for degradation, leads to long period. What explains this apparent discrepancy? In addition, the small molecule KL001 binds directly to Cry1/2 and blocks interaction with Fbxl3 to stabilize Cry1/2 proteins, yet also elicits a long period phenotype (Hirota, Science, 2012, St. John, PNAS, 2014). The authors should strive to be more inclusive of data that illustrate differential effects of Cry1 stabilization on period.

1b) Genotoxic stress stabilizes Cry1 and destabilizes Cry2, increasing the Cry1/Cry2 ratio. However, Cry1-/- cells (Cry1/Cry2 ratio of 0), have increased responses to genotoxic stress, while Cry2-/- cells have blunted responses to genotoxic stress. As E-box repressors, both genes have the capacity to repress each other. Is it possible that Cry2 is up-regulated in Cry1-/- cells and plays the dominant role in responses to genotoxic stress, possibly through its unique c-terminus? Along these lines, are Cry1 and Cr2 differentially recruited to the promoters of DNA damage response genes in Figure 5 in Koike et al.?

2) Figure 4E shows nicely that DNA damage induces interaction of Cry1/2 (Cry2 more noticeably than Cry1) with Fbxl3. However, the experiments probing the role of DNA-damage-dependent phosphorylation of Cry2 T363 or Fbxl3 S388 on their interaction is much less convincing than other data presented in Figure 4. In the absence of data that provide a clearer picture of how DNA damage-dependent phosphorylation regulates this interaction, we suggest removing panels F and G from Figure 4 and their discussion from the manuscript. We do not feel that removing these data detracts from the major findings of the work.

3) In the second part of the manuscript (Figure 5 and 6), the authors look to extend their description of the damage-dependent regulation of Cry1/2 stability to downstream events in cells. Here, the data for Cry1/2 regulation is less convincingly direct. Some genes, like p21, show striking Cry1 or Cry2-dependent changes from WT of opposite magnitude, while others show only a Cry1-dependent change in magnitude or kinetics of induction (Mdm2, Puma). It is unclear whether statistical significance should be determined by comparing damage-dependent transcription to the zero point (i.e. without damage) or between genotypes, if we are to assess the role of Cry1/2 in the responses. Moreover, it is not clear from the data presented in Figure 5 how direct the role of Cry1 or Cry2 is in transcriptional regulation of these genes. The authors are admittedly careful not to attribute changes in mRNA to direct regulation by Cry1/2, although it feels as though they imply this direct regulation by citing the Koike et al. findings that Cry1/2 are recruited to sites throughout chromatin. It is unclear what 'Cry1/2-coordinated responses' means in this context and should be better clarified.

4) Finally, in Figure 6, the authors make an intriguing connection between knockout of Cry2 and the accumulation of DNA damage that may play a role in altering Mendelian ratios of survival in Cry2-/- mice. While these data are interesting, we do not see a clear connection between Figures 1-4 and Figure 6. If anything, the DNA damage-dependent mRNA profiles shown in Figure 5 seem to suggest that Cry2-/- cells have a transcription response that is not altogether different from WT cells. In addition, Mendelian transmission ratio distortion cannot be directly attributed to “increased mutation rate” without supporting evidence. Many other scenarios are possible. The most obvious would be a defect due to a change in CRY2 targets or function.

5) In Figure 2C and in the fourth paragraph of the Results section, ubiquitination is measured only in cells, therefore, the ubiquitination cannot be assessed “directly” as stated since this requires in vitro ubiquitination assays using recombinant proteins as shown for CRY by Busino et al., 2007, in Science and Siepka et al., 2007, in Cell. The deubiquitination assays should be performed in vitro. It is curious that the Methods section describes a form of “in vitro deubiquitation assay” but this is not what is shown in Figure 2C and I did not find such results elsewhere. The assay describes the use of immunoprecipitated CRY1 from 293T cells which is not ideal since other interacting CRY1 protein would be co-IP'd. In vitro SCF E3 ligase assays are normally performed using baculovirus expressed CRY constructs and E3 ligase components.

6) In Figure 2C, the rightmost two lanes for Cry-/- there is a (Ub)N signal on the western blot that is as strong as the some of the wildtype lanes on the left. What accounts for this background signal? This emphasizes the need for in vitro ubiquitination assays.

7) Introduction section, first paragraph: The original references for FBXL3 should be cited instead of the Gatfield and Schibler paper (the FBXL21 original papers are cited correctly). Busino et al., 2007, Science; Siepka et al., 2007, Cell; Godinho et al., 2007, Science; Siepka et al., 2007.

8) Introduction section, second paragraph: In the list of physiological effects of Cry1/2 null alleles, the response to genotoxic stress induced by cyclophosphamide should be mentioned (Gorbacheva et al., 2005, PNAS) since, in this case, Cry1/2 knockout mice are resistant to this agent, which on the surface is the opposite of the results reported in this manuscript. In the case of cyclophosphamide, it is a CLOCK;BMAL1 target that confers circadian time-dependent resistance to cyclophosphamide. In addition, a role for CRY has been shown for p53 cancer risk by Ozturk et al., 2009, PNAS.

9) Figure 1D: What is responsible for the weak interaction signal for FLAG-N constructs 2-5?

10) Figure 2F: Why is this experiment only 2.5 days in length? Too short, should be at least 5 cycles.

eLife. 2015 Mar 10;4:e04883. doi: 10.7554/eLife.04883.027

Author response

1a) Depletion of Usp7 leads to decreased Cry1 expression. However, it also leads to long period length. In contrast, loss of Cry1 by RNAi or mouse knockouts leads to shortened period length, while loss of Fbxl3, which targets Crys for degradation, leads to long period. What explains this apparent discrepancy? In addition, the small molecule KL001 binds directly to Cry1/2 and blocks interaction with Fbxl3 to stabilize Cry1/2 proteins, yet also elicits a long period phenotype (Hirota, Science, 2012, St. John, PNAS, 2014). The authors should strive to be more inclusive of data that illustrate differential effects of Cry1 stabilization on period.

We thank the reviewers for their insight into the complexity of the relationship between Cry1/2 stability and circadian period length and we apologize for the oversimplified discussion of this subject in our original manuscript. Indeed, the mechanism by which Cry1/2 expression or stability is altered influences the ultimate effects on circadian period length. There are several examples in which a change in the average level of Cry1 or Cry2 protein expression can either increase or decrease circadian period depending on the details, some of which are cited by the reviewers. For example, KL001 increases period even in Cry1-/- cells (Hirota, Science, 2012) showing that stabilization of Cry2 leads to long period length even though knockout or knockdown of Cry2 also increases circadian period. Mathematical modeling spurred the authors of that study to hypothesize that Cry1 and Cry2 have similar roles in period determination and that nuclear stabilization of either isoform leads to period lengthening. We find that destabilization of nuclear Cry1 by shRNA-mediated knockdown of Hausp or pharmacological inhibition of Hausp increases period, which is consistent with the long period in cells treated with AMPK activators that also destabilize nuclear Cry1 but in contrast to short periods observed in Cry1-/- cells and long periods in KL001-treated cells in which nuclear Cry1 is stabilized. This discrepancy could be attributed to temporal segregation of the actions of Fbxl3 and Hausp, perhaps one of them is more important during the phase of Cry1/2 accumulation and the other critical during the destruction phase. Alternatively, while the inhibition of Fbxl3 binding and ubiquitylation of Cry1 by KL001 and the acceleration of Cry1 ubiquitylation by Hausp knockdown or inhibition have opposing effects on nuclear Cry1 stability, their similar effects on circadian period length could reflect actions of different subsets of Cry1 (e.g. Cry1 with different posttranslational modifications or in different subnuclear locations) or differential effects on Cry2: for example, in Cry1-/- cells and in KL001-treated cells, Cry2 is stabilized while in Hausp-depleted cells, Cry2 is unaffected. Updating mathematical models to incorporate new clock protein partners, posttranslational modifications and the effects on period upon interference with them may help to predict the relative importance of new pathways in various phases of circadian negative feedback (St. John, PNAS, 2014). We have updated the text to accommodate these uncertainties about the relationship between Cry1/2 stability and circadian period determination.

1b) Genotoxic stress stabilizes Cry1 and destabilizes Cry2, increasing the Cry1/Cry2 ratio. However, Cry1-/- cells (Cry1/Cry2 ratio of 0), have increased responses to genotoxic stress, while Cry2-/- cells have blunted responses to genotoxic stress. As E-box repressors, both genes have the capacity to repress each other. Is it possible that Cry2 is up-regulated in Cry1-/- cells and plays the dominant role in responses to genotoxic stress, possibly through its unique c-terminus? Along these lines, are Cry1 and Cr2 differentially recruited to the promoters of DNA damage response genes in Figure 5 in Koike et al.?

Based on the transcriptional responses to DNA damage that we observed in Cry1-/- and Cry2-/- cells, we hypothesized that Cry1 and Cry2 redundantly repress some sequence-specific DNA binding transcription factor(s) that contribute to induction of transcription in response to genotoxic stress. For example: if Cry2 has a slightly higher affinity than does Cry1 for transcription factor X, in quiescent wildtype cells, X is repressed by Cry2 and poised for activation upon Cry2 degradation following DNA damage. Accumulation of Cry1 after DNA damage would allow Cry1 to replace the degraded Cry2 and return expression of the target genes of X to basal levels resulting in transient activation in wildtype cells while expression of those genes would be sustained in Cry1-/- cells as we observed. The activation of those genes would be blunted in Cry2-/- cells due to Cry1 being bound to X in the absence of Cry2 since Cry1 would not be degraded in response to the damage stimulus (Figure 7). As the reviewers noted, this pattern of altered regulation was most clear and consistent in the case of Cdkn1a (p21) expression, but Rrm2b also exhibits increased induction in Cry1-/- cells and blunted induction in Cry2-/- cells (Figure 5) and Gadd45a and Puma are hyperactivated in Cry1-/- cells while the response in Cry2-/- cells is suppressed in response to irradiation (Figure 5–figure supplement 1). We did not detect induction of Gadd45a or Puma in either wildtype or Cry2-/- cells in response to doxorubicin treatment under the conditions used, so we can’t say whether their response to doxorubicin would be blunted in Cry2-/- cells.

The reviewers correctly point out that a model in which Cry2 positively contributes to activation of target genes by a DNA binding transcription factor X would also explain our data. We had not considered this because Cry1 and Cry2 are generally believed to act as transcriptional repressors and the observed degradation of Cry2 in response to DNA damage could explain activation of associated transcription factors but it is certainly possible that a transiently induced state of Cry2 (for example ubiquitylated Cry2) is required to activate transcription and we thank the reviewers for this creative suggestion.

In order to gain additional insight into the roles of Cry1 and Cry2 in transcriptional regulation after DNA damage, and in particular to determine whether the effects observed in either Cry1-/- or Cry2-/- cells could be attributed to increased expression of the remaining cryptochrome, we examined transcription in response to DNA damage in Cry1-/-; Cry2-/- doubly deficient (DKO) fibroblasts in addition to the single knockout cells. We also measured several additional transcripts including some inspired by the reviewers’ suggestion to mine the Koike et al. ChIP data set for possible direct targets of Cry1 and/or Cry2. The results are illuminating and suggest that both Cry1 and Cry2 can participate in the regulation of transcription in response to DNA damage, though their relative importance seems to depend on the specific transcript measured and on the duration of the DNA damage response.

The expression pattern for p21 in DKO cells looks similar to that of Cry2-/- cells, though we generally observe at least some induction of p21 in DKO cells and less or none in Cry2-/- cells, suggesting that the loss of Cry2 is sufficient to explain some blunted induction of p21 but that increased Cry1 protein could contribute to the more severe loss of p21 induction in Cry2-/- cells compared to DKO cells. In contrast, the prolonged activation of p21 in response to DNA damage in Cry1-/- cells seems to require the expression of Cry2 since it does not occur in the DKO cells, suggesting that this phenomenon could involve either release of repression upon Cry2 degradation or a positive action of a subset of Cry2 (e.g. uniquitylated), or an indirect effect of the loss of Cry2 on another pathway that modulates p21 expression.

The expression patterns of Gadd45a and Rrm2b in DKO cells resemble the patterns in Cry1-/- cells, suggesting that it is the loss of Cry1 rather than accumulation of Cry2 that causes hyperactivation of those transcripts. Interestingly, the Gadd45a locus was observed bound to both Cry1 and Cry2 and the Rrm2b locus was uniquely bound by Cry1 in the Koike et al. data set, suggesting that Cry1-dependent regulation of these genes could involve direct binding (Figure 5–figure supplement 2).

While we are cautious about using ChIP-sequencing data from unstimulated liver samples to interpret the effects of genetic loss of Cry1 and Cry2 on the transcriptional response to DNA damage stimuli in fibroblasts, we examined the association of each of the circadian transcription factors (Koike et al.) not only with the established DNA damage responsive genes that we initially analyzed but also with chromatin regions that encode proteins involved in DNA repair. Interestingly, Cry1 and Cry2 were bound to several more such sites than were the other circadian factors (Supplementary File 3).

In order to assess whether these associations in unstimulated livers are relevant to DNA damage response in fibroblasts, we measured the expression of these transcripts in response to DNA damage in WT and Cry1/2-deficient cells and found that several of their expression profiles are altered by genetic disruption of Cry1 and/or Cry2. Most striking, Rad23b expression is induced only in the wildtype cells: disruption of Cry1 and/or especially Cry2 prevents its activation by DNA damage.

These analyses that were inspired by the reviewer’s suggestions indicate that Cry1 and Cry2 can both directly and indirectly modulate the transcriptional response to DNA damage. A more complete understanding of how Cry1 and Cry2 influence gene expression during DNA damage response will require analysis of genome-wide transcription and chromatin association of Cry1 and Cry2 in wildtype and Cry1/2-deficient cells in response to DNA damage, which is beyond the scope of this study. We have updated the text to clarify the state of our understanding and we changed the title of the manuscript to emphasize other aspects of this work that are more central to the novel and important findings described herein.

2) Figure 4E shows nicely that DNA damage induces interaction of Cry1/2 (Cry2 more noticeably than Cry1) with Fbxl3. However, the experiments probing the role of DNA-damage-dependent phosphorylation of Cry2 T363 or Fbxl3 S388 on their interaction is much less convincing than other data presented in Figure 4. In the absence of data that provide a clearer picture of how DNA damage-dependent phosphorylation regulates this interaction, we suggest removing panels F and G from Figure 4 and their discussion from the manuscript. We do not feel that removing these data detracts from the major findings of the work.

We agree with the reviewers’ comments and have removed these data from the manuscript.

3) In the second part of the manuscript (Figure 5 and 6), the authors look to extend their description of the damage-dependent regulation of Cry1/2 stability to downstream events in cells. Here, the data for Cry1/2 regulation is less convincingly direct. Some genes, like p21, show striking Cry1 or Cry2-dependent changes from WT of opposite magnitude, while others show only a Cry1-dependent change in magnitude or kinetics of induction (Mdm2, Puma). It is unclear whether statistical significance should be determined by comparing damage-dependent transcription to the zero point (i.e. without damage) or between genotypes, if we are to assess the role of Cry1/2 in the responses. Moreover, it is not clear from the data presented in Figure 5 how direct the role of Cry1 or Cry2 is in transcriptional regulation of these genes. The authors are admittedly careful not to attribute changes in mRNA to direct regulation by Cry1/2, although it feels as though they imply this direct regulation by citing the Koike et al. findings that Cry1/2 are recruited to sites throughout chromatin. It is unclear what 'Cry1/2-coordinated responses' means in this context and should be better clarified.

We agree that we have not thoroughly investigated the roles of Cry1 and Cry2 in the transcriptional response to DNA damage and that the relationship is likely to be complex and transcript dependent as discussed in response to point #1b above. We also agree that a more appropriate statistical analysis should be applied and have now analyzed these data by repeated measures ANOVA to ask whether the Cry1/2 genotype significantly alters the response to the stimulus (Figure 5). As noted by the reviewers we have been careful not to attribute the observed transcriptional changes to direct regulation by Cry1 and/or Cry2 because we are not able to precisely discern the mechanism(s) by which they modulate the transcriptional response to DNA damage. We have clarified the text on these points.

4) Finally, in Figure 6, the authors make an intriguing connection between knockout of Cry2 and the accumulation of DNA damage that may play a role in altering Mendelian ratios of survival in Cry2-/- mice. While these data are interesting, we do not see a clear connection between Figures 1-4 and Figure 6. If anything, the DNA damage-dependent mRNA profiles shown in Figure 5 seem to suggest that Cry2-/- cells have a transcription response that is not altogether different from WT cells. In addition, Mendelian transmission ratio distortion cannot be directly attributed to “increased mutation rate” without supporting evidence. Many other scenarios are possible. The most obvious would be a defect due to a change in CRY2 targets or function.

We agree that many other scenarios could explain the Mendelian ratio distortion in Cry2-deficient animals. We performed this analysis because we were surprised to observe striking accumulation of DNA damage in Cry2-/- cells and wondered whether this could possibly be relevant in vivo given that Cry2-/- mice are viable and fertile, while several genetic models of mice with deficiencies in DNA damage response or DNA repair have been shown to be born at sub-Mendelian ratios (Tsai et al., 2005; Mukherjee et al., 2010; Crossan et al., 2011). While we cannot determine the root cause of the Mendelian ratio distortion at this time, it is at least consistent with the possibility that Cry2 is important for maintaining genomic integrity in vivo consistent with the results that we observed in cultured cells. We have updated the text to clarify this.

In addition, we added new quantitative analysis of the proteomics data sets that identified Hausp as a strong partner for Cry1, which also identified several proteins that interact with damaged DNA and/or participate in DNA repair in complex with Cry1 and/or Cry2 (Supplementary files 1 and 2). We have updated the Discussion to include this information and to better connect the different parts of the paper.

5) In Figure 2C and in the fourth paragraph of the Results section, ubiquitination is measured only in cells, therefore, the ubiquitination cannot be assessed “directly” as stated since this requires in vitro ubiquitination assays using recombinant proteins as shown for CRY by Busino, 2007, in Science and Siepka, 2007, in Cell. The deubiquitination assays should be performed in vitro. It is curious that the Methods section describes a form of “in vitro deubiquitation assay” but this is not what is shown in Figure 2C and I did not find such results elsewhere. The assay describes the use of immunoprecipitated CRY1 from 293T cells which is not ideal since other interacting CRY1 protein would be co-IP'd. In vitro SCF E3 ligase assays are normally performed using baculovirus expressed CRY constructs and E3 ligase components.

We have added data (Figure 2–figure supplement 2) demonstrating that purified recombinant HAUSP (USP7) can remove ubiquitin chains from ubiquitinated Cry1 purified from 293T cells treated with the proteasome inhibitor MG132 while the related deubiquitinase USP8 did not remove ubiquitin from Cry1. While immunoprecipitated Cry1 may contain other interacting proteins, the dose-dependent removal of ubiquitin from this substrate by recombinant HAUSP (and not recombinant USP8) added to the reaction in vitro suggests that Hausp can directly catalyze the removal of ubiquitin from Cry1, though it is possible that it activates an associated protein that directly deubiquitinates Cry1. We were not able to use baculovirus expressed Cry1 for this assay because the C-terminal tail of Cry1 is required for its interaction with Hausp and Cry1 can only be stably expressed and purified from baculovirus-infected Sf9 cells if the C-terminal tail is excluded. (Instability of full length Cry2 was described in Xing et al., 2013, and we were similarly able to stably express and purify only Cry1 and Cry2 lacking the C termini in baculovirus expression systems).

6) In Figure 2C, the rightmost two lanes for Cry-/- there is a (Ub)N signal on the western blot that is as strong as the some of the wildtype lanes on the left. What accounts for this background signal? This emphasizes the need for in vitro ubiquitination assays.

The rightmost two lanes in Figure 2C were included to control for antibody specificity in the immunoprecipitation. The signal in the first lane (WT cells without proteasome inhibition) is similar to that in the control lanes, suggesting that under normal conditions, ubiquitinated Cry1 is rapidly degraded by the proteasome and is not detectable by this assay, at least not above this background signal, which likely represents other ubiquitinated proteins that are non-specifically immunoprecipitated in all of the samples. In the second lane, from WT cells in which proteasome activity is blocked by mg132 treatment, we clearly see an increase in the specific signal for ubiquitinated Cry1. We now include quantitation of these blots subtracting the average signal in the rightmost two lanes from the other lanes to quantitatively assess the changes in Cry1 ubiquitination resulting from Hausp depletion in the presence and absence of proteasome inhibition (Figure 2–figure supplement 3).

7) Introduction section, first paragraph: The original references for FBXL3 should be cited instead of the Gatfield and Schibler paper (the FBXL21 original papers are cited correctly). Busino et al., 2007, Science; Siepka et al., 2007, Cell; Godinho et al., 2007, Science; Siepka et al., 2007.

We have corrected these citations.

8) Introduction section, second paragraph: In the list of physiological effects of Cry1/2 null alleles, the response to genotoxic stress induced by cyclophosphamide should be mentioned (Gorbacheva et al., 2005, PNAS) since, in this case, Cry1/2 knockout mice are resistant to this agent, which on the surface is the opposite of the results reported in this manuscript. In the case of cyclophosphamide, it is a CLOCK;BMAL1 target that confers circadian time-dependent resistance to cyclophosphamide. In addition, a role for CRY has been shown for p53 cancer risk by Ozturk et al., 2009, PNAS.

We have updated the text to include these additional physiological effects of Cry1/2 genetic disruption.

9) Figure 1D: What is responsible for the weak interaction signal for FLAG-N constructs 2-5?

In cells overexpressing Hausp and Cry2, we observe an interaction between the two proteins, albeit much weaker than that between Hausp and Cry1; similarly, all of the Cry1/2 hybrids interact weakly with Hausp. Those that contain the Cry2 C-terminus (#2-5) interact weakly with Hausp; those that contain the Cry1 C-terminus (#1, #6) bind strongly. We do not believe that the Cry2-Hausp interaction is biologically relevant as we were not able to detect interaction between endogenous Cry2 and Hausp, and Cry2 protein levels were not altered by shRNA knockdown (Figure 3C) or pharmacological inhibition of Hausp (not shown).

10) Figure 2F: Why is this experiment only 2.5 days in length? Too short, should be at least 5 cycles.

We have updated Figure 2F to include 5 cycles.

Associated Data

    This section collects any data citations, data availability statements, or supplementary materials included in this article.

    Supplementary Materials

    Supplementary file 1.

    Cry1-associated proteins. Lysates from 293T cells expressing pcDNA3-2xFLAG with no insert (control) or Cry1 after the FLAG tag were used to purify control or Cry1-containing complexes by immunoprecipitation (IP) of the FLAG tag. Components of the resulting complexes were identified by mass spectrometry. The experiment was performed in triplicate and PatternLab for Proteomics (Carvalho et al.) was used to identify statistically enriched partners in Cry1-associated complexes compared to the control. Enrichment (Cry1/control) is the ratio of spectral counts in Cry1 vs control samples for all statistically enriched partners over three experiments.

    DOI: http://dx.doi.org/10.7554/eLife.04883.023

    elife04883s001.xlsx (29.7KB, xlsx)
    DOI: 10.7554/eLife.04883.023
    Supplementary file 2.

    Cry2-associated proteins. Lysates from 293T cells expressing pcDNA3-2xFLAG with no insert (control) or Cry2 after the FLAG tag were used to purify control or Cry1-containing complexes by immunoprecipitation (IP) of the FLAG tag. Components of the resulting complexes were identified by mass spectrometry. The experiment was performed in triplicate and PatternLab for Proteomics (Carvalho et al.) was used to identify statistically enriched partners in Cry2-associated complexes compared to the control. Enrichment (Cry2/control) is the ratio of spectral counts in Cry2 vs control samples for all statistically enriched partners over three experiments.

    DOI: http://dx.doi.org/10.7554/eLife.04883.024

    elife04883s002.xlsx (31.3KB, xlsx)
    DOI: 10.7554/eLife.04883.024
    Supplementary file 3.

    Chromatin binding of circadian transcription factors to loci encoding DNA repair proteins. Published data (Koike et al., 2012, Table S2) was searched for the text string ‘repair’ to make a preliminary identification of chromatin regions near genes involved in DNA repair that were found to be associated with each of the seven circadian transcription factors Cry1, Cry2, Per1, Per2, Clock, Npas2, and Bmal1.

    DOI: http://dx.doi.org/10.7554/eLife.04883.025

    elife04883s003.xlsx (46.5KB, xlsx)
    DOI: 10.7554/eLife.04883.025

    Articles from eLife are provided here courtesy of eLife Sciences Publications, Ltd

    RESOURCES