PUNISHER rs12318065 C>A transversion: a putative somatic driver mutation for poor prognosis in colon cancer

Abstract Objective: Colon cancer (CC) remains one of the leading causes of cancer death worldwide. Several mutations/polymorphisms have been implicated in CC development and/or progression. The role of the recently identified variants related to the long non-coding RNAs (lncRNAs) family has not yet been fully uncovered. In this sense, we aimed to explore the association between the lncRNA PUNISHER rs12318065 variant and the CC risk and/or prognosis. Methods: A total of 408 CC (paired 204 cancer/non-cancer) tissues were genotyped using the TaqMan allelic discrimination assay. Results: “A” variant was associated with higher susceptibility to develop CC under heterozygote (A/C vs. C/C: OR = 1.39, 95%CI = 1.09–2.17, P=0.002), homozygote (A/A vs. C/C: OR = 2.63, 95%CI = 1.51–4.58, P=0.001), dominant (A/C-A/A vs. C/C: OR = 1.72, 95%CI = 1.15–02.57, P=0.008), and recessive (A/A vs. C/C-A/C: OR = 2.23, 95%CI = 1.34–3.72, P=0.001) models. Patients with metastasis were more likely to harbor A/A and A/C genotypes (16.7% and 14.1%) than 11% with the C/C genotype (P=0.027). Patients harboring C>A somatic mutation were more likely to develop relapse (52.6% vs. 26.5%, P=0.003), have poor survival (57.9% vs. 27.7%, P=0.001), and have shorter disease-free survival (43.2 ± 2.6 months vs. 56.8 ± 1.29 months, P<0.001) and overall survival (49.6 ± 2.4 months vs. 56.6 ± 0.99 months, P<0.001). Multivariate Cox regression analysis showed that patients with distal metastasis and C>A somatic mutation were three times more likely to die. Conclusions: To our knowledge, the present study is the first to identify that the PUNISHER rs12318065 variant could be a novel putative driver of colon cancer and is associated with poor prognosis.


Introduction
Colon cancer (CC) is a heterogeneous disease that represents one of the most common (1.9 million incidence rate) and leading causes of cancer-related mortality (0.9 million death in 2020) worldwide [1]. In addition to increased exposure to environmental risk factors due to shifting diet and lifestyle toward westernization, genetic elements have also been reported to contribute to increasing CC incidence [2,3]. Despite current advances in surgical therapy, chemotherapy, and molecular targeting therapy for CC, the overall 5-year survival rate for patients remains as low as 12% for metastatic cases of CC [4]. Therefore,

Allelic discrimination analysis
Tissue DNA was isolated from samples via "QIAamp DNA FFPE Tissue Kit" (Catalog No. 56404, Qiagen, Hilden, Germany). After dissolving and removing the paraffin by xylene, the samples were lysed under denaturing conditions with proteinase K, and the genomic DNA was obtained following the manufacturer's protocol. NanoDrop ND-1000 (NanoDrop Technologies, Inc. Wilmington, DE, U.S.A.) was used to assess the concentration/purity of the extracted DNA. A specific TaqMan probe-fluorescence assay (C 30952613 10) with VIC and FAM dyes "GAGTGGGTGCGTCTGTCCAGCGGTC[A/C]GCCCGGTGTGGTCGTGCCCGGCCCG" for each allele, respectively, was run for allelic discrimination. Real-Time polymerase chain reaction (PCR) was carried out by two coauthors independently blinded to cancer/non-cancer sample status in a StepOne™ Real-Time PCR System (Applied Biosystems, Foster City, CA, U.S.A.) as previously described [27]. Loading of DNase-free water instead of unknown DNA in each run as a negative control was applied. The PCR reaction conditions were run as follows: "40 cycles at 95 • C for 10 min, 95 • C for 15 s, annealing at 60 • C for 1 min, and final step at 60 • C for 30 s" [28]. Five percent of the samples were randomly selected and run in duplicates to ensure results reliability with a 100% concordance rate for genotype calls. The genotyping results were retrieved by the related SDS software version 1.3.1.

Statistical analysis
Data analysis was performed using the "Statistical Package for the Social Sciences (SPSS) for Windows" software (version 27.0). Genotype/allele frequencies and Hardy-Weinberg equilibrium analysis were performed using the online SNPStats software (https://www.snpstats.net/). Binary logistic regression was performed, and adjusted odds ratio (OR) and 95% confidence interval (CI) by age and sex were estimated for five genetic association models (homozygote and heterozygote comparison, dominant, recessive, and over-dominant models). McNemar's test was used to calculate the somatic mutation rate [29]. Categorical variables were presented as frequencies and percentages and compared using the chi-square (χ2) or Fisher's exact tests when appropriate. Continuous variables were shown as mean + − standard deviation and compared using the Student's t-test. A two-tailed P-value of <0.05 was considered statistically significant. Cox Proportional Hazards regression analysis was applied to detect predictors of poor survival. Kaplan-Meier survival curves were generated to compare patients with and without C to A somatic mutation.

Characteristics of the study population
The study included paired samples of 204 colon cancer patients. The mean age was 58.3 years + − 12.3, and 60.8% were men. Of these, 68 patients (33.3%) died during the follow-up period of over five years. Those who expired were more likely to be men (70.6% vs. 55.9%, P=0.049), have lesions in the transverse or descending colon (57.4% vs. 45  Data are presented as frequency (percentage). Two sided-Chi-square test was used. The bold values indicate statistical significance at a P-value below 0.05.

Single-nucleotide polymorphism (SNP) analysis of PUNISHER variant
Genotype frequency of rs12318065 agreed with HWE in the control group (P=0.06). MAF (A allele) accounted for 0.33 in controls. According to the 1000 Genome Project, the same allele frequencies were 0.02 in Africans, 0.10 in Asians, 0.25 in Americans, and 0.13 in Europeans. In comparison between malignant and adjacent colon tissues, a higher frequency of the A allele was more representative in cancer tissues compared with control tissues (45% vs. 33%,  P=0.002). Correspondingly, A/A and C/A genotypes were more prevalent in cancer specimens (26.0% and 38.2%) compared with counterpart non-cancer control tissues (13.7% and 37.7%, P=0.003) (Figure 3).

Impact of genotypes on cancer risk
As depicted in Figure 4

Somatic mutation burden analysis
Tumor-normal paired analysis revealed genotype concordance in 166 out of 204 tissue samples, accounting for 81.4% of patients. In contrast, 38 samples showed the allelic difference between paired samples with a higher representation of the A allele in tumor samples. The C/C and A/C genotypes in 13 and 12 patients, respectively, were mutated to the A/A genotype in paired (cancer and non-cancer) tissues. In addition, 13 other cases with C/C genotype showed a change in one gene locus to A/C ( Table 2).

Association of PUNISHER genotypes with clinical and pathological features
PUNISHER rs12318065 genotypes were associated with distal metastasis; patients with metastasis were more likely to harbor A/A and A/C genotypes (16.7% and 14.1%) compared with 11% with C/C genotype (P=0.027). In addition, a higher frequency of mortality was reported in A/A (33.3%) and A/C (32.1%) groups compared with the C/C genotype (23.3%) (

Survival analysis
In comparison between patients who had C-to-A somatic mutation and non-cancer counterparts, Kaplan-Meier curves showed patients harboring conversion had shorter disease-free survival ( Figure 5).

Discussion
Accumulating evidence recognized the impact of gene polymorphism in increased colon cancer risk through known and yet unknown cellular and molecular changes [30]. A slew of research has discovered that lncRNA SNPs are substantially associated with cancer risk [31]. The lncRNA gene-related polymorphisms may have a significant impact on lncRNA expression levels, processing, or secondary structure, culminating in cancer genesis and progression and disparities in treatment responses [32,33]. SNPs may potentially cause the lncRNA to behave abnormally, leading to dysregulation of downstream signaling cascades and target gene expression [34,35].
In the present study, we found that PUNISHER rs12318065 AA and AC genotype carriers have a considerably higher risk of CC. The A allele was more common in cancer versus non-cancerous tissues and was considered a risk allele for CC under all genetic models. Also, the A/A and A/C genotypes were associated with a greater risk of distant metastases and mortality. Furthermore, nearly 63.8% of the cancer tissues showed a tendency for C>A shift, and the frequency of C>A somatic mutation was shown to be higher in the adenocarcinoma subtype. Additionally, patients harboring C>A somatic mutation were more likely to relapse. To our knowledge, there were no studies that uncovered the relation between PUNISHER rs12318065 and CC risk and/or outcome.
Accumulating evidence in the era of lncRNAs genetic variants, notably SNPs, has recently revealed that SNPs in lncRNAs are linked to an increased risk of colon/colorectal cancers. For example, Li et al. [31] reported that the  Data are presented as frequency (percentage). N: number; DFS: disease-free survival; OS: overall survival. A two sided-Chi-square test was used. Binary logistic regression analysis was performed. Odds ratio (OR) and 95% confidence intervals (CI) are shown. The bold values indicate statistical significance at a P-value below 0.05. lncRNA "Colorectal Cancer Associated Transcript 1 (CCAT1)" rs67085638 C>T was associated with an increased risk of CC, and the rs7013433 A>T was correlated to an advanced stage of CRC in Chinese population. Also, Cao et al. [36] found that the AA genotype of the lncRNA maternally expressed gene 3 (MEG3) rs7158663 was significantly increased the CRC risk, in particular, in those over 60 years and with a positive family history of cancer. Another study has demonstrated that the HOXA transcript at the distal tip (HOTTIP) rs3807598 (GG vs. CC) and rs2067087 (CC vs. GG) increased the CRC risk by 1.57-and 1.70-fold, respectively, while the rs17501292 variant was associated with improvement in OS of CRC patients with ulcerative/invasive tumors [37]. The metastasis-associated lung adenocarcinoma transcript 1 (MALAT1) rs664589 G allele was thought to change MALAT1's binding to miR-194-5p, resulting in increased gene expression and accelerated CRC growth and metastasis [38]. Similarly, the GAS5 gene promoter rs55829688 variant was implicated in changing the gene expression via influencing the binding affinity of the transcriptional factor YY1 to the promoter region [39], and the prostate cancer-associated transcript 1 (PCAT1) rs2632159 may influence CRC risk by altering EBF, LUN-1, and TCF12 binding, thereby up-regulating PCAT1 expression and hence potentiate its carcinogenic role [40]. Currently, stratified analysis by patients' characteristics reveals an association between the PUNISHER rs12318065 variant and the rate of C>A somatic mutation with decreased OS and DFS. Our data also showed that tumors harboring this type of transversion were a significant independent predictor of overall survival with the presence of distant metastasis. In an attempt to potentially predict the impact of the studied variant on disease risk and/or outcome, we run the "HaploReg v4.1" (https://pubs.broadinstitute.org/mammals/haploreg/haploreg.php) (last accessed on February 2022), which is an updated and validated bioinformatics tool specified for exploring annotations of the non-coding genome variants, based on the 1000 Genomes Project, and predicting the effect of variants on the regulatory motifs and gene expression based on the expression quantitative trait locus (eQTL) studies [41]. Interestingly, the obtained results showed that the PUNISHER rs12318065 A variant could influence binding with the transcriptional factors: POL2, STAT1, and ZNF263, which are linked to carcinogenesis [42][43][44][45][46], also it could alter regulatory motifs, including mrg1, HOXa9 1, and Sin3Ak-20 disc6 some of which have been confirmed to be associated with colon cancer carcinogenesis and metastasis [47]. We further evaluated, in silico, the potential function of other SNPs that were in high (r 2 > 0.80) linkage disequilibrium (LD) with rs12318065 and found that some of these polymorphisms were in regulatory regions, including the promoter, the enhancer, and DNase hypersensitivity sites (Table 5). These results could support the potential association of the studied variant with dysregulated gene expression and function that warrant further future functional studies to confirm these speculations. PUNISHER dysregulation has been associated with several cancers, including the CC, through various mechanisms and molecular pathways [18,[48][49][50]. It was implicated in regulating fibroblast growth factor receptor 1 (FGFR1) expression in CRC by sponging microRNA-497, and its up-regulation was associated with poor patient survival [10]. Other studies have suggested that PUNISHER and LINC-PINT may create a negative feedback regulation loop in colon cancer [51]. Furthermore, PUNISHER could induce endothelial-mesenchymal transition (EMT) and increase CRC cell proliferation, motility, and invasion via targeting the miR-4,668-3p/SRSF1 axis [52]. It also could raise the Cofilin-1 (CFL1) expression that mediates EMT, cell migration, and invasion in CRC via competitively binding to miR-182-5p [52].
The advantage of our study lies in the relatively large sample size, and we are the first, up to our knowledge, to report a significant association between the PUNISHER rs12318065 variant and colon cancer risk and poor prognosis. However, the present study lacks the mechanistic and functional works that uncover the specific role of the studied variant in CC. Further studies are warranted to study the impact of this variant on gene expression level and to explore the potential association of this polymorphism with chemoresistance either in CC or other cancer types. More research into novel lncRNA-based genetic biomarkers for predicting CC susceptibility and/or clinical prognosis is recommended.

Conclusion
In summary, the present study found that the PUNISHER rs12318065 C>A transversion is associated with increased colon cancer risk and poor prognostic indicators in terms of short survival time and tumor relapse.

Data Availability
All supporting data are included within the main article.