Preserving Salivary Gland Physiology against Genotoxic Damage - The Tousled Way

Abstract

Tousled and its homologs are evolutionarily conserved serine/threonine kinases present in plants and animals. Human Tousled-like kinases, TLK1 and TLK2, are implicated in chromatin assembly during DNA replication, chromosome segregation during mitosis, as well as in DNA damage response and repair. They share a high degree of sequence similarity, but have few non-redundant functions. Our lab has studied TLK1, and found that it increases the resistance of cells to ionizing radiation (IR) damage through expedited double strand break (DSB) repair. DSBs are life-threatening lesions which when repaired restore DNA integrity and promote cell survival. A major focus in our lab is to dissect TLK1’s role in DSB response and repair and study its usefulness in averting salivary gland hypofunction, a condition that invariably afflicts patients undergoing regional radiotherapy. The identification of anti-silencing factor 1 (ASF1), histone H3, and Rad9 as substrates of TLK1 links the protein to chromatin organization and DNA damage response and repair. However, recent findings of new interacting partners that include NEK1 suggest that TLK1 may play a broader role in DSB repair. This review provides a brief overview of the DNA damage response and DSB repair, and it highlights our current understanding of TLK1 in the process.

Introduction

Radiotherapy is a cost-effective treatment for a majority of cancer patients. Regional radiation of head and neck cancer and cervical lymphomas, or total body radiation before bone marrow transplantation inadvertently exposes normal salivary glands to radiation. Genotoxic radiation damage is a major threat to salivary gland function, and despite a slow turnover rate of cells, salivary glands are unusually sensitive to radiation. A functional decline begins within the first weeks of radiotherapy and continues well beyond its completion. Chronic irreversible loss of saliva is one of the most common untoward effects of radiation, and it harbingers oral morbidity that significantly impairs the quality of life of the patient (Lalla et al., 2017). Saliva is a biological fluid that plays a vital role in oral health. It cleanses the oral cavity and lubricates the oral mucosa to facilitate swallowing, speech and taste. Its loss affects the clearance of food debris and bacteria from the oral cavity and increases mucosal friability with resultant mucosal tears, dental decay, and opportunistic oral infections. Conventional treatment of salivary hypofunction focuses on assuaging dry mouth using salivary substitutes or prosecretory, parasympathomimetic sialagogues such as pilocarpine or cevimeline. These measures provide intermittent relief, but are largely inadequate in quenching the persistent feeling of oral dryness.

Early drop in salivary flow following radiation is not accompanied by cell death and has been attributed to a secretory malfunction of acinar cells. However, ensuing death of irreparably damaged cells and the inability to regenerate functioning secretory cells has been linked to the chronic irreversible nature of the disorder. Submandibular glands irradiated below a radiation dose threshold (<39 Gy) have been shown to recovery and overcome loss of function (Murdoch-Kinch et al., 2008). However, invariably, radiation treatment with intent to eradicate cancer far exceeds the threshold dose, and irreversible loss of salivary function commonly plagues most head and neck cancer patients (Jensen et al., 2010). To provide some resemblance of normal salivary function in patients undergoing radiotherapy, a number of preemptive and corrective strategies are being investigated to preserve, repair, or restore gland function. These include 1) the sparing of radiation to stem-cell rich regions of the salivary glands or transplantation of autologous stem cells to promote gland regeneration (Lombaert et al., 2008a, van Luijk et al., 2015), 2) pharmacological and biological interventions to increase the number of surviving cells through either improving cellular repair, stimulating cell proliferation, or bypassing cell death (Lombaert et al., 2008b, Zheng et al., 2011, Timiri Shanmugam et al., 2016), and 3) altering water permeability of surviving salivary cells to increase fluid output (Baum et al., 2012, Delporte et al., 1997). Each advance though promising has its limitations. Measures to conservatively spare regions of the salivary glands are limited by tumor location, whereas the challenge to obtain sufficient number of stem cells from biopsies constrains the translation of autologous stem cell transplantation to patient care. Alternatively, systemic administration of pharmaceuticals can be fraught with the risk of tumor protection, promoting cell proliferation or circumventing apoptosis in damage encumbered salivary cells can lead to carcinogenesis. Finally, increasing water flow from surviving salivary ductal cells is a safe alternative treatment, but it cannot replenish salivary immunoglobulins, anti-microbial proteins, mucins, or enzymes. Therefore, local therapies that promote repair and protect salivary gland function are vital to the faithful restoration of salivary flow and the lasting alleviation of xerostomia. In this regard, we have shown that a normal cellular isoform of Tousled-like kinase 1 (TLK1; abbreviations are also defined in Table 1) alleviates radiotoxicity and preserves salivary gland function in animals (Palaniyandi et al., 2011, Sunavala-Dossabhoy et al., 2012, Timiri Shanmugam et al., 2013) potentially, by facilitating DNA repair. In this review, we discuss DNA damage response to and repair of radiation-induced DSBs highlighting the role of TLK1 in the process and expounding its function in attenuating radiation-induced salivary dysfunction.

Table 1.

List of abbreviations.

Abbreviations
IR	ionizing radiation
DSB	double stranded break
ssDNA	single stranded DNA
HR	homologous recombination
NHEJ	non-homologous end joining
SSA	single-strand annealing
TLK1/2	Tousled-like kinase 1/2
ASF1	anti-silencing factor 1
CDK	cyclin-dependent kinase
PARP1/2	poly ADP ribose polymerase 1/2
PI3KKs	phosphatidylinositol 3-kinase-like protein kinases
ATM	ataxia telangiectasia mutated
DNA-PKcs	DNA-dependent protein kinase catalytic subunit
MRN	Mre11-Rad50-Nbs1
ATR	ataxia telangiectasia and Rad3 related
RPA	replication protein A
9-1-1	Rad9-Rad1-Hus1
53BP1	p53 binding protein 1
WRN	Werner syndrome protein
CtIP	C terminal binding protein (CtBP) interacting protein

Induction of TLK1B in Response to Genotoxic Stress

Tousled gene was identified in Arabidopsis Thaliana as being essential to the proper development of floral and meristem organs (Roe et al., 1993). Homologs of Tousled were later found to be evolutionarily conserved in plants and animals. Humans have two homologs namely, Tousled like-kinase 1 and 2 (TLK1 and TLK2), and they share nearly 84 % amino acid sequence similarity and show substantial conservation of function (Sillje et al., 1999). Tousled and TLKs are serine/threonine kinases that harbor a catalytic C-terminal kinase domain and a regulatory N-terminal domain. Multiple variants of TLKs have been identified, and one variant of TLK1 that utilizes a downstream methionine as the start was identified in our lab and is referred to as TLK1B (Li et al., 2001). Other than the absence of the N-terminal 238 amino acids, the shorter variant is identical to the full-length protein, and it has been found to phosphorylate the same substrates identified to date. TLKs and its splice variants localize to the nucleus and perinuclear region (Li et al., 2001, Sillje et al., 1999), and they dimerize or heterodimerize to attain catalytically active conformations through auto-phosphorylation. Even though TLKs are constitutively expressed throughout the cell cycle, their activity peaks in interphase and is linked to DNA replication (Sillje et al., 1999).

TLK1B possesses a longer and structured 5′UTR, and it is dependent on high levels of rate-limiting eukaryotic translation initiation factor 4E (eIF4E) for translation (Li et al., 2001). TLK1B, therefore, is not expressed at appreciable amounts in normal cells. However, in response to genotoxic stress, AKT-mTOR mediated phosphorylation of eIF4E-binding protein (eIF4E-BP) increases the pool of free eIF4E, which translationally upregulates expression of TLK1B (Sunavala-Dossabhoy et al., 2004). Since exogenous expression of TLK1B had been shown to promote cell survival against IR (Li et al., 2001, Sunavala-Dossabhoy et al., 2003), it begged the question whether upregulation of TLK1B is a normal cellular mechanism to counter radiation injury.

DNA Damage Response to IR

IR causes a plethora of DNA lesions that include oxidation of DNA bases and sugars, and the induction of single- and double-stranded DNA breaks. Reactive oxygen species generated along radiation tracks result in cluster damage with simple or complex DNA breaks that have altered 3′ ends with a 3′-phosphate or 3′-phosphoglycolate moiety, and, or oxidized base modifications or abasic sites next to the break. Double-strand breaks (DBSs) that result in two free DNA ends are the most lethal types of lesions since non-repaired breaks can induce genetic instability or mitotic catastrophe. To efficiently preserve DNA integrity for faithful transmission of genetic material from one generation to the next, cells have developed distinct mechanisms to sense DNA damage and orchestrate repair (Ciccia & Elledge, 2010). Checkpoints are introduced through activation of protein kinase cascades that ultimately inhibit cell cycle phosphatase CDC25 and suppress CDK activation until repair is complete. Not surprisingly, mutations that abrogate the checkpoint mechanism are associated with tumorigenesis (Harper & Elledge, 2007).

The binding of early sensor proteins to DSBs triggers an elaborate cellular response that implements cell arrest and recruits DNA repair machinery. The binding of poly ADP ribose polymerase 1 and 2 (PARP1/2) catalyze ADP-ribosylation of proteins around the break to signal damage and recruit repair factors. DNA ends are sensed by Ku heterodimers (Ku70/80) or the Mre11-Rad50-Nbs1 (MRN) complex, and they activate phosphatidylinositol 3-kinase-like protein kinases (PI3KKs), ataxia telangiectasia mutated (ATM) and DNA-dependent protein kinase catalytic subunit (DNA-PKcs), respectively. Ku70/80 avidly binds most DNA ends, and recruits and activates DNA-PKcs that negatively regulates ATM activity (Zhou et al., 2017). Alternately, complex DNA ends with modification or adducts are bound by the MRN complex, which recruits ATM. Activation of ATM, through dimerization and autophosphorylation, initiates the phosphorylation of many of its substrates including the MRN complex. Another member of the PI3KK family, ataxia telangiectasia and Rad3 related (ATR), is also rapidly activated after IR; however, it occurs downstream of ATM (Myers & Cortez, 2006). Similar to activation of ATR by single-stranded DNA (ssDNA) formed at replication blocks, the generation of ssDNA as DSB repair intermediates activates ATR too (Cimprich & Cortez, 2008). Newly formed ssDNA is rapidly coated by RPA, which then recruits ATR through its interacting protein, ATRIP, and stimulates Rad17-dependent loading of Rad9-Rad1-Hus1 (9-1-1) complex onto the DNA (Yang & Zou, 2006). Bridging of 9-1-1 and ATR-ATRIP complexes by TopBP1 activates ATR resulting in downstream checkpoint signaling through Chk1 (Delacroix et al., 2007). Although the bonafide downstream signal transducer of ATM is Chk2 and that of ATR is Chk1, cross-activation of checkpoint kinases has also been observed. Nonetheless, through inhibitory phosphorylation of CDC25 phosphatase, the master regulatory kinase-initiated cascades delay cell progression to provide the necessary interval for decision making, repair, and recovery.

DSB Repair

There are two predominant pathways for repair of DSBs, non-homologous end joining (NHEJ) and homology-directed recombination (HR; Figure 1). Cells deficient in either pathway display sensitivity to IR (Rothkamm et al., 2003, Hinz et al., 2005). A majority of DSBs are repaired through NHEJ that operates throughout the cell cycle, but is most required for repair of DSBs that arise in the G1 phase. HR, on the other hand, is operational only in the S/G2 phase of replicating cells i.e., when a sister chromatid is available as a template (Rothkamm et al., 2003, Hinz et al., 2005). The repair pathway of choice is mainly governed by cell cycle phase and the balance of repair factors that dictate the progression of events towards a chosen path. Recent studies that demonstrate the negative regulation of ATM by DNA-PKcs (Zhou et al., 2017), or blockage of MRN by p53 binding protein 1 (53BP1) or Werner syndrome protein (WRN) at DNA ends to prevent excessive processing of DNA ends, which promote repair by the classical NHEJ pathway (Bunting et al., 2010, Shamanna et al., 2016). On the other hand, exclusion of 53BP1 by ATM phosphorylated MOF (males absent on the first) and the entry of BRCA1 is a determinant of repair by HR (Chapman et al., 2012, Gupta et al., 2014). Recent results indicate that p53BP1binding prevents long-range resection and allows NHEJ to be the first choice pathway in S/G2 cells. In the absence of timely repair; however, BRCA1 promotes dephosphorylation of p53BP1 and favors recruitment of EXO1and repair by HR (Isono et al., 2017). However, in the absence of 53BP1 and BRCA1, repair occurs through a non-canonical NHEJ mechanism (Ochs et al., 2016). It is well-accepted that the extent of DNA end resection foretells the repair pathway choice. The initial step in 5′ strand resection by MRN endonuclease is critical to HR, and its inhibition sanctions repair through NHEJ (Shibata et al., 2014). Interaction of MRE11 with CDK phosphorylated CtBP interacting protein (CtIP; T847) in S/G2 phase stimulates MRN endonuclease (Anand et al., 2016), whereas lack of phospho-CtIP impairs resection and generation of ssDNA intermediates (Huertas & Jackson, 2009). Likewise, phosphorylation of EXO1 by CDKs heightens EXO1 recruitment to DSBs and augments 5′end resection to promote repair via HR. EXO1 phosphorylation mutants that fail to resect the 5′strand conversely, increase NHEJ (Tomimatsu et al., 2014). Collectively, the observations suggest that resection of 5′ terminated DNA ends in S/G2 phase licenses HR, whereas deficiency in HR pathway components or absence of a homologous sequence redirects repair to the NHEJ pathway.

Figure 1. Schematic of the major DNA DSB repair pathways, NHEJ and HR.

Generation of IR-induced DSBs activate DNA sensor proteins, Ku heterodimers and MRN complex, which promote recruitment of enzymes that facilitate non-homology directed end joining or homology dependent repair. Refractory DNA ends that cannot be processed by Ku-DNA-PKcs are routed for repair through the non-canonical, alternate NHEJ pathway.

NHEJ

NHEJ repair can occur through classical NHEJ or alternative NHEJ depending on the initial sensor proteins. Rapid recognition of DSBs by the Ku70/80 heterodimer initiates repair via the classical NHEJ pathway, which catalyzes the rejoining of DNA ends through minimal end-processing. Ku recruits DNA-PKcs, and through dimerization, DNA-PKcs immobilize the DNA termini and limit end resection. Activation of DNA-PKcs then facilitate entry of end-processing enzymes such as Artemis that process the ends to ligatable 5′ phosphate and 3′hydroxyl ends (Mahaney et al., 2009). Artemis is essential to removal of IR-induced 3′ phosphoglycolated DNA ends and predictably, Artemis mutants are hypersensitive to radiation (Yannone et al., 2008, Povirk et al., 2007). DNA-PKcs regulate WRN, which inhibits recruitment of MRN complex and excessive processing of DNA ends to promote repair by classical NHEJ pathway (Shamanna et al., 2016). WRN depletion results in large deletions at breaks (Oshima et al., 2002); however, cells display merely an incremental increase in IR sensitivity (Brosh & Bohr, 2007) suggesting compensatory repair through non-canonical, alternate NHEJ. Alternative NHEJ is a backup pathway that becomes apparent when classical NHEJ is compromised. PARP1, which predominantly recognizes single strand breaks, can compete with Ku for binding to DSBs (Wang et al., 2006). PARP1 promotes association of the MRN complex and thereby, end resection. MRN endonuclease is dispensable for NHEJ, but inhibition of its exonuclease results in defective repair (Shibata et al., 2014). Regions of microhomology uncovered during strand resection are annealed to restore DNA continuity - characteristics of the repair pathway. The single-strand annealing (SSA) repair pathway is similar to the alternate NHEJ; however, MRN mediated end resection is more extensive in SSA. In the absence of a sister chromatid, a required feature for HR, Rad52 pairs and anneals uncovered repetitive complementary sequences with repair resulting in sequence deletions of intervening non-complementary overhangs.

Irrespective of repair by classical or non-canonical pathways, NHEJ, in essence, is error prone. Repair occurs at the expense of loss of genetic information and mutagenic alterations at restored junctions. Nonetheless, it is the arbitrary joining of DNA ends in response to genotoxic breaks that plays a decisive role in preventing chromosomal translocations and maintaining genome stability.

HR

PARP1 promotes binding of the MRN and repair by HR in S/G2 phase. Considering the robust 3′-5′ exonuclease activity of MRN, the presence of 5′-strand resected intermediates in HR remained in disagreement with exonuclease polarity until evidence of bidirectional 5′strand resection was first demonstrated in budding yeast (Garcia et al., 2011). Preferential endonucleolytic clipping of 5′ strand by MRN is the primary event in HR that engages EXO1 for 5′-3′ resection away from the DSB and MRN exonuclease resection in 3′-5′ direction towards the DSB. IR generated DNA breaks are often blocked by base modifications or protein adducts that make end recognition and processing a challenge. Since MRN endonucleolytic cleavage is enhanced by blocked DNA ends, this suggests its preferred involvement in the processing of ends refractory to exonucleolytic resection (Anand et al., 2016). The 3′ single stranded DNA tails generated following complementary strand resection are rapidly bound and stabilized by RPA, and in the presence of a donor template, Rad51 substitutes RPA to initiate nucleoprotein filament-mediated strand displacement and homology search and pairing. Rad51 is then displaced by Rad54, which increases access of the 3′-OH end to DNA polymerase for DNA synthesis (Spies et al., 2016). Newly formed strands either disengage to rejoin the same strand or alternatively form Holliday junctions that when resolved result in repair with or without crossover. Although HR predictably reestablishes the original sequence at repaired junctions and averts the risk of mutagenesis, it is important to recognize that crossover events can lead to loss of heterozygosity. HR is the major repair pathway of replication-associated DSBs (Saleh-Gohari et al., 2005), which include secondary DSBs that arise when replication forks encounter non-repaired nicks or breaks (Groth et al., 2012).

TLK1 in DNA Damage Response

An emerging picture of TLK1 in DNA damage response suggests that it is intimately involved in checkpoint arrest (Figure 2). Previous studies demonstrated an early, but transitory, inactivation of TLK1 through Chk1-dependent phosphorylation at serine 695 following DNA damage or replication stress, and suppression of TLK1 activity being necessary to instituting the S/G2 checkpoint (Groth et al., 2003). Conversely, restoration of TLK1 activity was found to be associated with exit from G2 arrest (Kelly & Davey, 2013). With the recent identification of NIMA (never in mitosis A)-related kinase 1 (NEK1) as a substrate of TLK1, a higher order function of TLK1 in cell cycle regulation was implicated (Singh et al., 2017). NEK1 stabilizes the ATR-ATRIP complex and primes ATR for rapid activation following DNA damage (Liu et al., 2013). The observation that TLK1 phosphorylated NEK1 T141 is required for ATR activity suggests that in undamaged cells TLK1-NEK1 constitutively primes ATR to ensure the robust activation of ATR-Chk1 signaling in response to DNA damage. Subsequent activation of Chk1 and the inactivation TLK1 halts S/G2 phase progression to prevent degeneration of replication forks and accumulation of widespread damage. After completion of repair, cell re-entry into the cell cycle is vital to avoid a state of permanent arrest, and recovery of TLK1 activity is linked to release of cells from S/G2 (Kelly & Davey, 2013). Although more work has focused on TLK1, a recent study suggests that TLK2 is also involved in release of cells from G2 arrest (Bruinsma et al., 2016) suggesting a likely redundant role of the kinase in the DNA damage response.

Figure 2. TLK1 in DNA damage response.

TLK1 constitutively phosphorylates NEK1, which primes ATR-ATRIP in non-encumbered cells. In response to DNA damage, CHK1-dependent inhibitory phosphorylation of TLK1 occurs, which is necessary for the formation of the 9-1-1 complex and for checkpoint institution. Recovery of TLK1 thereafter is followed by Rad9 phosphorylation and disassembly of the 9-1-1 complex.

TLK1 in DSB repair

The evidence of a direct link between TLK1 and DNA repair came with the discovery of the DNA damage response and repair protein, Rad9 as a phosphorylation target of TLK1(Sunavala-Dossabhoy & De Benedetti, 2009). Mammalian Rad9 interacts with Rad1 and Hus1 to form a heterotrimeric complex referred to as the 9-1-1 complex. The C-terminal tail of Rad9 is dispensable for 9-1-1 complex formation, but multiple phosphorylation sites that lie within are phosphorylated either constitutively or in response to DNA damage and they influence protein stability and its interaction with other proteins (Roos-Mattjus et al., 2003). Although the 9-1-1 complex is structurally similar to the PCNA sliding clamp, it is not associated with replicating DNA, and it is selectively reserved for DNA checkpoint response (Venclovas & Thelen, 2000). Rad9 is dispensable for NHEJ, and is selectively involved in HR; it improves viability of S/G2 cells exposed to IR (Pandita, 2006). Directed by RPA, 9-1-1 complex is preferentially loaded at junctions of single-stranded DNA and duplex DNA in HR intermediates (Ellison & Stillman, 2003). TLK1 phosphorylates Rad9 at S328 and T355 (Kelly & Davey, 2013, Sunavala-Dossabhoy & De Benedetti, 2009). Lack of Rad9 S328 phosphorylation has little effect on 9-1-1 complex formation (Sunavala-Dossabhoy & De Benedetti, 2009), but phosphorylation at the site promotes the exclusion of the protein to the cytoplasm (Awate & De Benedetti, 2016). Additionally, TLK1 phosphorylation of Rad9 T352 is linked to the exit of cells from G2 checkpoint in late stages of DNA repair (Kelly & Davey, 2013). Other than the role of NEK1 in priming ATR-ATRIP for rapid DNA damage response, NEK1 plays a direct role in HR by phosphorylating Rad54-S572, and facilitating the timely eviction of Rad51 for repair synthesis (Spies et al., 2016). Mutation at the Rad54 phosphorylation site in turn leads to the persistence of Rad51 foci and defects in HR (Spies et al., 2016). Since TLK1 phosphorylation of NEK1 T141 augments NEK1 activity, it suggests a role of TLK1 in final stages of HR repair. The current model suggests that TLK1 coordinates DNA repair (Figure 3); early downregulation of TLK1 in response to DNA damage supports the formation of the 9-1-1 complex, whilst recovery of TLK1 and phosphorylation of NEK1 and Rad9 synchronizes the completion of DNA synthesis, 9-1-1 disassembly, and checkpoint recovery.

Figure 3. TLK1 in DSB repair via HR.

The involvement of NEK1 in Rad54 phosphorylation directly places TLK1 in HR. It is suggested that recovered TLK1 activity after DNA damage promotes NEK1-Rad54 driven Rad51 eviction for completion of repair synthesis in HR.

TLKs and Chromatin Remodeling in Repair

Chromatin structure can pose a significant challenge to the detection and timely repair of DNA damage. Chromatin needs to remodel for repair machinery to access, repair, and restore DNA continuity for normal cellular function. Consequently, in response to DNA damage, cells have developed mechanisms to locally create an open and relax chromatin conformation for repair. Various histone-modifying enzymes and histone chaperones act in concert to create open chromatin structures for repair and thereafter, to reassemble chromatin upon completion of repair. The histone chaperone, anti-silencing factor 1 (ASF1), is involved in DNA replication and the ‘chromatinization’ of newly replicated DNA. ASF1 and histone H3 as substrates of TLK1 (Li et al., 2001, Sillje & Nigg, 2001) link the protein to chromatin assembly and structure. Multiple residues within ASF1 C-terminal tails are phosphorylated by TLKs in vitro, and unlike ASF1b, phosphorylation of ASF1a augments protein stability by suppressing proteosomal degradation (Pilyugin et al., 2009). Although experiments in vitro suggest that ASF1 phosphorylation is dispensable for nucleosome assembly (De Benedetti, 2010), the alteration at all 4 TLK1 recognition sites (S166, S175, S192 and S199) affects histone recruitment and progression of cells through S-phase in vivo (Klimovskaia et al., 2014). It is clear that phosphorylation of ASF1 is not essential for function; however, the modification increases the avidity for soluble histones and favors the formation of ASF1-histone complex (Klimovskaia et al., 2014).

Phosphorylation of histone H3 S10 is directly linked to chromatin structure. Plant Tousled and mammalian TLK1 have been shown to phosphorylate histone H3 in vitro (Li et al., 2001, Ehsan et al., 2004), and loss-of-function TLK1 mutants in mammalian cells impair proper cell division (Sunavala-Dossabhoy et al., 2003). Histone H3 (S10) is regarded as a marker of heterochromatin, and it increases as cells progress from late G2 to mitosis. Other than being a marker of repressed chromatin, H3 S10 is an indicator of active chromatin as well. A simple explanation for the apparent paradox is that the context of the ‘histone code’ dictates the type of chromatin configuration. A drop in H3 phosphorylation occurs soon after DNA damage and time-lapsed kinetics of H3 phosphorylation is inversely related to phospho-H2AX, a marker of unresolved DSBs (Sunavala-Dossabhoy et al., 2005). Indeed, the observation is consistent with the thought that the immediate decline in TLK1 activity following DNA damage promotes decondensation of chromatin for damage recognition and end resection and homology search during repair, whereas TLK1 restoration facilitates ‘rechromatinization’ of repaired DNA ends.

Conclusion

Guarding the integrity of the genome is critical to a cell’s survival and function. In a recent proteomic screen, a large set of novel TLK1 interaction partners were identified (Singh et al., 2017), and it is conceivable that the role of TLK1 in cell survival is far more comprehensive. However, the generation of viable TLK1 and conditional TLK2 knockout mice implicates the possibility of shared substrates and interchangeable support of functions during development (Segura-Bayona et al., 2017). Likewise, knockout of TLK1 and TLK2 in mouse embryonic fibroblasts significantly increased genomic instability compared to either kinase alone. Nevertheless, despite the high degree of functional redundancy, absence of TLK1 and a corresponding reliance on TLK2 in placental tissue suggests the presence of unique dependencies in certain tissues (Segura-Bayona et al., 2017). Preemptive expression of TLK1 alleviates radiation-induced rat salivary dysfunction, and our results point to a role of TLK1 in facilitating DNA damage response and repair in the process. However, whether the expression of TLK1 exclusively, or through cross-activation TLK2 confers a radiotolerant phenotype remains to be investigated.