The role of double‐strand break repair, translesion synthesis, and interstrand crosslinks in colorectal cancer progression—clinicopathological data and survival

DNA repair is a new and important pathway that explains colorectal carcinogenesis. This study will evaluate the prognostic value of molecular modulation of double‐strand break repair (XRCC2 and XRCC5); DNA damage tolerance/translesion synthesis (POLH, POLK, and POLQ), and interstrand crosslink repair (DCLRE1A) in sporadic colorectal cancer (CRC).

In the pursuit of eliminating TNM inconsistencies, CRC molecular complexity and its heterogeneous clinical presentations have been leading to the research of novel prognostic and predictive biomarkers, including DNA repair components. For example, 15% of sporadic CRC patients who harbor DNA mismatch repair (MMR) system defects and, consequently, microsatellite instability (MSI), 6 have better stage-adjusted survival and reduced likelihood of metastasis when compared with microsatellite stable tumors. 7,8 However, MSI has several limitations that restrict its use as a practical prognostic factor across all stages of CRC, as its clinical value is restricted to stage II CRC, where adjuvant chemotherapy is not recommended. 9 Nevertheless, associations of DNA damage and imbalances in other pathways engaged in their repair with CRC risk, progression, response to therapy and prognosis have been widely reported. We and others recently reported that disturbances in gene and/or protein expression of DNA damage response sensors and effectors-including double-strand break repair (DSBR), DNA damage tolerance/translesion synthesis (DDT/TLS) and interstrand crosslink repair (ICLR) pathways-have minimal association with clinicopathological features and response to therapy in CRC. 10,11 Despite the lack of definitive evidence so far, a plethora of reports have been suggesting an intersection between CRC and DNA repair systems, which may be mediated by MMR defects (by inducing other somatic mutations that disrupt DNA repair mechanisms) or not. 12 Double-strand breaks (DSBs) are the most critical type of genotoxic stress and their repair is a central cellular mechanism to preserve genomic stability. 13 DSBs are processed by homologous recombination or classical nonhomologous end-joining DNA repair pathways, and disruptions of these pathways favor the accumulation of damage in rapidly dividing cells, leading to mutagenesis or apoptosis. 14 Since DSBs result in the loss of integrity of both complementary strands, proficiency of error-prone repair is required.
However, loss of genetic information and genomic instability is an immediate consequence to guarantee cell survival. DDT mechanisms are mediated by Y-family translesion DNA polymerases (such as pol κ, pol η, and pol θ), which bypass DNA adducts, imbalanced dNTP pools, and unusual template structures. As a consequence, to impede fork collapse and apoptosis due to unrepaired DSB, translesion DNA polymerases induce mutation. 15

| Immunohistochemistry
Immunohistochemistry for MLH1, MSH2, XRCC5 (Ku80), Polκ, p53, and ki67 was carried out according to MacDonald et al. 22 The sections were incubated with the following primary antibodies, all

| Statistical analysis
Gene expression means between normal and neoplastic tissue were compared using independent sample t Student or Mann-Whitney   (Table 2).
Low XRCC5 gene expression was associated with promoter methylation (P = .015) and low XRCC5 (Ku80) protein expression (P = .0001). POLK overexpression was associated with high correspondending protein contents (P = .0001), but not with the absence of promoter methylation (P = .581) (Table S1). Promoter methylation and gene expression of POLH and DCLRE1A were not associated (data not shown).

| Associations of DSBR, ICLR, and DDT/TLS key components with clinicopathological and molecular features of CRC patients
Tumors with low expression of POLH exhibited mucinous histology (P = .05), but smaller invasive depth (P = 0.038). Low tumor expression of POLK was associated with the presence of distant metastases (P = 0.042). Promoter methylation of POLK was associated with smaller invasive depth (P = .011) and methylation of POLH to welldifferentiated tumors (.023). In addition, POLK promoter methylation was associated with tumors with high Ki67 contents (P = .036) and low expression of DCLRE1A was associated with tumors with low Ki67 contents (P = .042) (

| Prognostic value of DNA repair component modulation in patients with CRC
Kaplan-Meier's survival analyses indicated that patients whose  POLK and POLH encode members of DNA polymerase type-Yfamily of proteins, Pol κ, and Pol η, respectively. Variations in expression or activity of Y-family DNA polymerases could possibly produce TLS pathway imbalance and, therefore, mutagenesis. 33 However, the magnitude to which these alterations are oncogenic drivers or whether it impacts clinical outcomes is still unknown.
In our study, we found upregulation of POLK and downregulation of POLH in neoplastic tissues in comparison to paired normal tissues. The oncological relevance of pol κ and pol η in cancer is most firmly established concerning response to treatment. Upregulation of pol κ confers resistance to temozolomide in glioblastoma, 34,35 and upregulation of pol η to platinum drugs in HNSCC, lung, gastric adenocarcinomas, and ovarian cancers. [36][37][38] Contrary to our results, low levels of POLK were previously observed in CRC. 39,40 Conversely, others reported an increase of pol κ expression in brain and lung cancers. 41,42 Low expression of POLH and POLK were found in tumors with mucinous histology and vascular metastasis, although in the early stages of development. POLK promoter methylation was strongly associated with better DFS. Conversely, unmethylated POLH and POLK promoters were associated with more advanced and poorly differentiated tumors.
Despite finding more aggressive colorectal tumors harboring high POLK levels, this fact was not a predictor of DFS and OS. On the other hand, POLK promoter methylation was associated with better Finally, despite its sample size limitation, to the best of our knowledge, our study is one of the few to report associations between POLK, POLH modulation and clinical features and prognosis of CRC patients. Furthermore, we believe that this is the first study to evaluate DCLRE1A gene expression and promoter methylation in colorectal tumors.

| CONCLUSION
Components of the pathways involved in DSBR, DDT/TLS, and ICLR are a new horizon in the DNA repair pathway discussion.
There are few reports about these and the influence on clinicopathological features and survival is still a big question.
This study revealed that low expression or unmethylated POLH and POLK were related to worse tumors. In this context, POLK