Is post-transcriptional stabilization, splicing and translation of selective mRNAs a key to the DNA damage response?

Abstract

In response to DNA damage, cells activate a complex, kinase-based signaling network that consist of two components - a rapid phosphorylation-driven signaling cascade that results in immediate inhibition of Cdk/cyclin complexes to arrest the cell cycle along with recruitment of repair machinery to damaged DNA, followed by a delayed transcriptional response that promotes cell cycle arrest through the induction of Cdk inhibitors, such as p21. In recent years a third layer of complexity has emerged that involves post-transcriptional control of mRNA stability, splicing, and translation as a critical part of the DNA damage response.  Here, we describe recent work implicating DNA damage-dependent modification of RNA-binding proteins that are responsible for some of these mRNA effects, highlighting recent work on post-transcriptional regulation of the cell cycle checkpoint protein/apoptosis inducer Gadd45a by the checkpoint kinase MAPKAP Kinase-2.

Post-transcriptional Control of Gene Expression is Essential for a Functional DNA Damage Response

In response to DNA damage, cells activate a complex interconnected signaling network that alters cell metabolism, temporarily or permanently halts cell cycle progression, activates and recruits DNA repair machinery and controls the decision of whether to initiate some type of cell death program.1 Many of these changes occur as a consequence of post-translational modifications of proteins within the DNA damage signaling network through phosphorylation, ubiquitylation or sumoylation.2 In addition, the pattern of mRNA expression also undergoes significant changes after DNA damage. In a study of human lymphoblastoid cell lines from healthy adults, for example, Chu and colleagues found that thousands of mRNAs were up or downregulated after cells were exposed to ionizing radiation (IR) or ultraviolet light.3 Similarly, as much as 20% of the genes in budding yeast showed a two-fold or greater change in mRNA levels in response to MMS treatment or ionizing radiation.4 However, it would seem intuitively obvious that the last thing one would expect cells to do shortly after incurring significant DNA damage is to start transcribing new genes. First, there is the real probability that the genetic information in the DNA strand itself is damaged. Second, the transcription process is both energy intensive (synthesis of a message with n bases requires at least n NTP molecules) and slow from the point of view of creating a final protein product if it was needed for cell cycle arrest or DNA repair. In agreement with this, a number of studies have shown that DNA damage triggers a transient repression of gene transcription in eukaryotic cells.5,6 How, then, do cells accomplish DNA damage-induced changes in mRNA expression, when transcriptional activity is globally suppressed?

One possible explanation lies in manipulating the post-transcriptional regulatory circuits that control mRNA splicing, stabilization and translation. All three of these processes depend on specific RNA-binding proteins (RBPs) which modulate mRNA half-life, spliceosome targeting, nuclear export and translatability through direct interactions with their client mRNAs.7–11 Importantly, recent data from several laboratories now suggests that these RBPs are direct targets of the DNA damage response (DDR), and are likely to be critically important in controlling the phenotypic response in cells experiencing genotoxic stress. In a mass spectrometry-based proteomic screen for binding partners of the DNA damage response molecule/tumor suppressor 14-3-3σ, for example, we found that the largest classes of interacting proteins were molecules involved in mRNA splicing or translation.12 Matusoka et al. performed a phospho-proteomic mass-spectrometry screen looking for substrates of the dedicated DDR kinases ATM, ATR and DNA-PK in response to IR.13 Surprisingly, the largest subset of ATM/ATR/DNA-PKcs substrates they identified were proteins linked to nucleic acid metabolism, and in particular proteins involved in posttranscriptional mRNA regulation. Cimprich and colleagues conducted a genome-wide RNAi screen looking for spontaneous γH2AX formation following gene knock-down.14 In that screen, mRNA-binding and processing factors emerged as the largest subset of ‘hits’. Finally, in a study of alterations in gene expression that directly compared changes in mRNA abundance with changes in gene transcription, Gorospe and colleagues found that over 50% of the observed changes in steady state mRNA levels in non-small cell lung carcinoma H1299 cells exposed to UV irradiation were directly attributable to mRNA turnover.15 Taken together, these observations, from very diverse experimental approaches, highlight the emerging importance of post-transcriptional regulatory mechanisms in the context of DDR signaling. What, then, are the molecular signals emerging from damaged DNA that are responsible for controlling these post-transcriptional events?

Clearly, the Matsuoka et al. data suggest that ATM and ATR are likely to play a direct role. In addition to the ATM-Chk2 and ATR-Chk1 modules, however, genotoxic stress also activates the p38^MAPK-MAPKAP2 Kinase-2 (MK2) pathway downstream of ATM and ATR,9,16–22 and the function of this pathway is particularly critical for cell cycle arrest and prevention of apoptosis in cells with defective p53 function. We recently discovered that major targets of the p38^MAPK/MK2 pathway are RNA-binding proteins involved in upregulating the cell cycle checkpoint protein Gadd45α.23

Gadd45α is Post-Transcriptionally Regulated in Response to DNA Damage

Gadd45α is a known component of the DDR that has been mechanistically linked to numerous cellular processes, including cell cycle arrest and apoptosis.24,25 Gadd45α expression is rapidly induced after genotoxic stress both in a p53-dependent and -independent fashion.23,26–33 In addition to transcriptional control, Gadd45α mRNA levels are post-transcriptionally stabilized in response to UV or MMS exposure.34 Recent work from Gorospe and co-workers identified the RNA-binding proteins (RBPs) AUF1 and TIAR as critical negative regulators of Gadd45α mRNA at the post-transcriptional level.35 In resting cells, AUF1 and TIAR are found in ribonucleo-protein (RNP) complexes with Gadd45α mRNA. AUF1 targets Gadd45α mRNA for degradation, while TIAR prevents the association of Gadd45α mRNA with translating polysomes. The combined effect of these two inhibitory mechanisms is a potent repression of Gadd45α protein biosynthesis.

Upon exposure to UV or treatment with MMS, Lal et al. showed that AUF1 and TIAR rapidly dissociate from Gadd45α mRNA, correlating with a substantial increase in Gadd45α mRNA stability and increased Gadd45α protein accumulation.35 The nature of the molecular signals responsible for the DNA damage-induced dissociation of AUF1 and TIAR remained enigmatic, as did the potential involvement of other damage-induced RBPs in this process. In an attempt to understand why p53-defective cells required p38^MAPK/MK2 activity for cell cycle checkpoint function, we recently made the surprising discovery that this pathway is a critical regulator of RBPs mediating posttranscriptional stabilization of Gadd45α mRNA in response to genotoxic stress.23 Specifically, we discovered three novel p38^MAPK/MK2-dependent posttranscriptional mechanisms of Gadd45α mRNA control. First, we found that RNAi-mediated MK2 depletion prevented the accumulation of Gadd45α mRNA and protein in response to doxorubicin and identified the known MK2 substrate hnRNP A0 as a novel Gadd45α mRNA-binding protein.23 We and others showed that MK2 directly phosphorylates hnRNP A0 on Ser-84.23,36 This specific phosphorylation event was critical for hnRNP A0:Gadd45α mRNA RNP complex formation, which in turn was required to stabilize Gadd45α mRNA in response to DNA damage. Second, we found that MK2 directly phosphorylates Poly-(A) ribonuclease (PARN) on Ser-557 in response to doxorubicin. PARN phosphorylation at this site was required for prolonged Gadd45α mRNA and protein expression after doxorubicin. The molecular details of this apparent inhibition of Gadd45α mRNA degradation remain somewhat unclear; despite our best efforts, we have not been able to observe any changes in PARN activity in vitro, following MK2-mediated phosphorylation on Ser-557. Finally, we could show that the MK2-activating kinase p38^MAPK directly phosphorylates TIAR after doxorubicin and that p38^MAPK inhibition completely prevented the doxorubicin-mediated dissociation of TIAR:Gadd45α mRNA RNP complexes.

In additional experiments, we showed that Gadd45α operates as part of a positive feedback loop that is necessary to maintain MK2 activity at late times, but not at early times, after DNA damage. RNAi-mediated depletion of Gadd45α specifically prevented the phosphorylation and activation of MK2 from ∼12 hrs onwards after DNA damage, likely through a late loss of p38 activity. These data suggest that the initial activation of MK2 after genotoxic stress does not depend on Gadd45α, but subsequent p38/MK2-dependent stabilization of Gadd45α, through phosphorylation of TIAR, PARN and hnRNP A0, becomes essential for maintaining MK2 activity at late times.23

Why is this late MK2 activity important? We and others had shown previously that MK2 acts as a cell cycle checkpoint effector kinase that is activated downstream of ATM and ATR18,37 much like Chk1 and Chk2. And just like Chk1 and Chk2, MK2 phosphorylates Cdc25B and C to create a 14-3-3-binding site, sequestering Cdc25B/C in the cytoplasm to prevent CyclinB/Cdk1 activation. In our recent experiments, we found that nuclear-localized Chk1 is responsible for initiating an early G₂/M checkpoint, but this arrest is lost after about 12 hours. In cells that lack p53 function, MK2 activity, primarily in the cytoplasm, is required to then maintain this arrest.18,23 Intriguingly, we could show that MK2 rapidly exits the nucleus after DNA damage and remains active in the cytoplasm for at least 30 hrs following the genotoxic insult. MK2 depletion resulted in premature nuclear re-entry of CDC25B/C followed by checkpoint collapse and a catastrophic attempt to undergo mitosis. Importantly, the time at which we observed premature CDC25B/C nuclear re-entry in MK2-depleted cells corresponded perfectly to the time when MK2 activity had dropped to baseline levels in Gadd45α-depleted cells that were treated with doxorubicin. These data strongly suggest that the positive feedback loop involving MK2-dependent stabilization of Gadd45α and Gadd45α-dependent maintenance of MK2 activity, are essential for prolonged cell cycle arrest through cytoplasmic Cdc25B/C sequestration in response to doxorubicin. Together, our data indicate that a feed forward loop consisting of p38^MAPK, MK2 and Gadd45α is critical for allowing cells time to recover from doxorubicin-induced genotoxic damage before entering the next mitotic cell division (Fig. 1).

Concluding Remarks

The DNA damage response has classically been thought of as a signaling network consisting of rapid phosphorylation-driven signal transduction on one hand and a delayed transcriptional response mediated through transcription factors, such as p53. In addition to these well-established branches of DNA damage signaling, posttranscriptional regulatory mechanisms are now beginning to emerge as critical components of the DNA damage response. Recent mechanistic studies are pinpointing at a number of RNA-binding proteins, such as HuR, TIAR, AUF1, hnRNP A0 and PARN as well as protein kinases, such as ATM, ATR, Chk2, p38^MAPK and MK2 that control the function and localization of these RNA-binding proteins.9,11,13,14,23,35,38 These reports strongly suggest that cells employ complex regulatory circuits that control mRNA transcript splicing, stability and translatability in response to genotoxic stress. The major challenges in this emerging area of research will be identifying the relevant RBPs, mapping the transcripts that are post-transcriptionally regulated by specific RBPs, and determining whether individual RBPs control a subset of transcripts with related function in the DNA damage response a so-called RNA operon.39 New technologies, such as genome-wide RNAi screening and next generation sequencing of cell lines and primary tumor samples should allow the systematic search for novel transcripts and RNA-binding proteins that are critical regulators of the DNA damage response.

Figures and Tables

Figure 1 Transcriptional and post-transcriptional regulation of Gadd45α. A simplified schematic overview of regulatory mechanisms that control the expression of Gadd45α. In resting cells, the expression of Gadd45α is transcriptionally suppressed through c-Myc and a repressive complex consisting of ZBRK1 and BRCA1. In addition, translation of existing Gadd45α transcripts is suppressed by TIAR, while AUF1 binding to Gadd45α mRNA targets it for degradation. In response to DNA damage, both Gadd45α transcription and translation are upregulated. The Gadd45α transcriptional inhibitor c-Myc is translationally repressed through miR-34c. While miR-34c can be induced by p53 following genotoxic stress, miR-34c expression in p53-deficient cells depends on the p38/MK2 signaling complex.40 Gadd45α transcription is induced by a variety of damage- and stress-activated transcription factors including p53, WT1, Oct1, NF-YA, FoxO3a, Egr-1 and C/EBPα. Furthermore, both TIAR and AUF1 dissociate from the Gadd45α mRNA after genotoxic stress. The disruption of the TIAR:Gadd45α mRNA RNP depends on p38-mediated TIAR phosphorylation. Gadd45α mRNA degradation is prevented by MK2-dependent Ser-557 phosphorylation of PARN, while Gadd45α mRNA is actively stabilized by binding to the MK2-dependent phospho-Ser-84 form of hnRNP A0. The resulting accumulation of Gadd45α protein functions within a positive feedback loop that maintains p38/MK2 activity at late times following DNA damage. Prolonged MK2 activity in turn is required to maintain Cdc25B and C in an inactive state sequestered in the cytoplasm.

Acknowledgements

We apologize to our colleagues for the omission of many important contributions to the field, and their references, due to space limitations. We thank the members of the Yaffe and Reinhardt laboratories for helpful discussions. This work was supported by the National Institutes of Health (GM68762, CA112967, ES015339 to M.B.Y.), the Deutsche Forschungsgemeinschaft (RE2246/1-1, RE2246/2-1 and SFB-832 to H.C.R.), the Deutsche Nierenstiftung (to H.C.R.), the Austrian Science Fund (FWF, Erwin-Schroedinger-Fellowship to S.M.), the Anna Fuller fund of New Haven (to I.G.C.) and the David H. Koch Fund (H.C.R. and M.B.Y.).