Introduction

Oral cavity cancer is known to be the 6th most common malignancy worldwide1, accounting for 1.9% of cancer-related deaths2. Each year, there are ~ 354,864 new reported cases of oral cancer, representing 2% of all cancer cases2. According to the Cancer Statistics 2023, only in the United States there are 34,470 registered new cases and 7440 deaths from oral cavity cancer3. In countries such as Pakistan, Bangladesh, India, and Sri Lanka, oral cancer ranks as the second most common cancer, both in terms of its prevalence and its impact on prognosis by influencing cancer development, progression, and response to treatment4. Oral squamous cell carcinoma (OSCC), the most prevalent neoplasm of the oral cavity (including the palate, floor of mouth, and tongue), accounts for over 90% of oral cancer cases5 and can result in social isolation, emotional distress, and significant functional impairments as a consequence of both the disease and its treatments6,7,8. The global annual incidence of new OSCC cases is increasing, with the highest rates observed in Asia, followed by Western countries. This trend places OSCC among the 10th most prevalent types of cancer worldwide9.

In Pakistan, lip and oral cancers are the second most prevalent cancer site overall, accounting for 10.9% of cases across both genders, and the most common cancer among males, with 15.9% of new cases. Risk factors for head and neck cancer are multifaceted and include tobacco use, alcohol consumption10, chewing habits (areca nut, betel quid, gutka, pan masala, and naswar11) and viral infections such as human papillomavirus (HPV) and Epstein-Barr virus (EBV)10. Additionally, alterations in oncogenes (e.g., PIK3CA, RAS) and tumor suppressor genes (e.g., TP53, CDKN2A, NOTCH1) contribute to tumor development12. Other nonspecific risk factors include poor oral hygiene, cigar and pipe smoking, and occupational hazards like working in the nickel industry10. Some studies question the role of family history in oral cancers13 while several epidemiological studies suggest a possible correlation between familial cancer history and increased risk for OSCC14,15. Additionally, low socioeconomic status (SES) is a key predictor of head and neck squamous cell carcinoma (HNSCC), with a higher incidence reported in low-income populations16,17,18. In the local context, these chewing habits are more prevalent among individuals with lower socioeconomic backgrounds and educational levels, and are particularly common among males19,20. In our population, the increasing burden of oral squamous cell carcinoma can be attributed primarily to the chewing habits and the widespread impact of low socioeconomic status. These factors, deeply ingrained in our society, play a significant role in the high prevalence of this disease in Pakistan.

The development of OSCC is also influenced by intrinsic factors, including age and genetic predispositions21. Genetic predisposition can influence the functionality of the DNA damage repair (DDR) pathways in oral epithelial cells. These pathways include mismatch repair (MMR), base excision repair (BER), nucleotide excision repair (NER), homologous recombination (HR), and non-homologous end joining (NHEJ), with direct reversal repair (DR) and interstrand crosslink (ICL) repair addressing specific lesions22. These mechanisms are crucial for maintaining genomic stability23. Double-strand breaks (DSBs), the most severe DNA damage, can be repaired by HR, NHEJ, or both24. This complex network of genes, detects DNA lesions, halts the cell cycle to allow for repair, and induces programmed cell death if repair is not feasible25,26. Mutations in these DDR genes can affect the efficiency of detecting and repairing DNA lesions which leads to genomic instability27,28, a key driver of tumorigenesis29,30.

Head and neck cancer (HNC), including oral squamous cell carcinoma (OSCC), is managed by several treatment modalities like surgery, radiotherapy, immunotherapy, and chemotherapy serving as the cornerstone for locally advanced tumors31. Cisplatin, the most widely used chemotherapeutic agent induces cell death by forming DNA adducts and arresting the cell cycle24,32, and is often combined with agents like paclitaxel, docetaxel, 5-FU (TPF regimen), and immunotherapies (e.g., pembrolizumab, nivolumab) for advanced or metastatic stages33. However, only about 40% of patients with locally advanced HNC respond to treatment34, and the 5-year survival rate remains around 50%35,36. Resistance to cisplatin arises through increased DNA repair capacity, particularly via nucleotide excision repair (NER), which plays a key role in clearing cisplatin–DNA adducts24,32,37. Targeting DDR pathways, such as through the use of DDR inhibitors (e.g., Mitomycin C, Cisplatin, Etoposide), has emerged as a promising strategy to enhance the effectiveness of DNA-damaging treatments and overcome resistance38,39.

In OSCC, mutations in the TP53 gene are associated with reduced survival and resistance to chemotherapy and radiotherapy. TP53 mutations disrupt a DNA damage response network involving the kinases ATM and ATR, which activate effector kinases CHEK1 and CHEK2 to regulate cell cycle checkpoints and coordinate DNA repair. Disruption of this network impairs the cell’s ability to repair DNA and control the cell cycle, contributing to both therapeutic resistance and malignancy in OSCC40. In the early stages of oral cancer, the levels of DDR markers like γ-H2AX, RAD51, and 53BP1 are higher, showing that the tumor is actively repairing its DNA. Additionally, changes in proteins like TP53 and BRCA1, which are involved in DNA repair, suggest the tumor is using these mechanisms to grow and survive. Because these DDR markers are linked to how the cancer develops, they can be used to predict how the cancer will progress and how well a patient might do (prognosis)41,42. Thus these DDR markers may serve as independent prognostic biomarkers.

The identification of distinct mutations in individual cancer samples underscores the importance of characterizing the molecular changes specific to each cancer patient to advance personalized therapeutic approaches43. For instance, next-generation sequencing (NGS) technologies have significantly transformed cancer genomics research by offering an extensive approach to detecting somatic alterations in cancer genomes44,45. This includes the identification of point mutations, insertions, deletions, and other genetic modifications46.

A comprehensive analysis at the exome level was conducted to understand the occurrence of mutations in key DNA damage repair genes (TP53, ATM, ATR, CHEK1 and CHEK2) and their correlation with clinicopathological factors. We examined whole-exome sequencing (WES) data from 27 Oral squamous cell carcinoma patients and 7 paired adjacent normal tissues. The novelty of the manuscript lies in the 1st time report of the different mutations through whole exome sequencing from the unexplored Pashtun population belonging to Khyber Pakhtunkhwa. In addition, we have used diverse tools to characterize these mutations and elaborate their structural or functional significance.

Material and methods

Sample selection

Inclusion criteria

Both males and females of all ages which were clinically and histopathologically confirmed for OSCC (Stage I-IV) were included in this study.

Exclusion criteria

Patients with tumor recurrence, those treated with alternative therapies (Chemotherapy, Radiotherapy), or those with other cancers were excluded from the study.

Data collection procedure

Tissue biopsies were collected from patients meeting the inclusion criteria from Hayatabad Medical Complex (HMC) and Khyber College of Dentistry (KCD), Peshawar. Written informed consent (English) was obtained from patients or their guardians and the nature of the study, aims and objectives were explained in local languages (Urdu/Pashto). After obtaining the informed consent, their details were recorded on a structured proforma which included their enrollment ID’s, demographic details, clinical investigations and history.

OSCC tissue collection and processing

Tissue samples were collected in 10% formalin under sterile conditions and transported to Khyber Medical University (KMU). Following a 20-h fixation period, the samples were subjected to gross examination for documenting parameters such as size and color. Tumor regions were carefully embedded in cassettes and processed for 16 h in a tissue processor to produce formalin-fixed paraffin-embedded (FFPE) blocks. For slide preparation, 5 µm sections were taken on a microtome and transferred to glass slides, deparaffinized in xylene, and stained with hematoxylin and eosin (H&E) using standard protocols. The H&E-stained slides were examined by an expert pathologist to generate a detailed histopathology report for further analysis. A total of 34 FFPE blocks (27 Tumor, 7 Paired Normal) were selected for Whole Exome sequencing.

DNA extraction and whole exome sequencing

Formalin-fixed paraffin-embedded (FFPE) tumor blocks with high tumor cellularity (> 50%) were selected for DNA extraction. To ensure adequate DNA recovery, a core with a minimum diameter of 2.5 mm, or two cores each with a diameter of 1 mm, were obtained from each block for extraction. The genomic DNA was extracted using the manufacture protocol as per QIAamp DNA FFPE Tissue kit (Catalog No. 56404). The quality and quantity of the extracted DNA was evaluated using 1% agarose gel electrophoresis and Qubit® dsDNA HS Assay Kit (Thermo Fisher) on a Qubit® 2.0 Fluorometer to confirm the presence of an adequate amount of DNA. Successfully passed QC DNA samples were selected for whole exome sequencing.

Following the isolation of DNA from the tumor-rich cores of the FFPE blocks, the quality of DNA was assessed using agarose gel electrophoresis. Only DNA samples with an average fragment size greater than 200 bp were selected for library preparation. A total of 200–300 ng of high-quality DNA was used for library preparation.

The size distribution of the prepared libraries was evaluated using 2% agarose gel electrophoresis. Libraries with a concentration exceeding 10 ng/μL were processed for high-throughput paired-end sequencing on the Illumina platform.

DNA libraries were prepared by fragmenting the DNA and ligating paired-end adapters to the fragments. Exonic regions of the genome were captured by hybridizing the prepared libraries with the Illumina DNA Prep with Exome 2.5 Enrichment Kit, in accordance with the manufacturer’s protocol. The captured exonic regions were subsequently amplified to ensure adequate material for sequencing.

The enriched libraries were then sequenced using the Illumina NextSeq 500 platform, employing a 150-bp read length and achieving a mean sequencing depth of 100 × , facilitating high-throughput sequencing of the targeted exome regions. The resulting libraries underwent cluster amplification after dilution to a final concentration of 2 nM in 10 µL. Following cluster generation, the flow cells were loaded onto the sequencer. Raw sequencing data were provided in FASTQ format for subsequent analysis. Per-exon coverage for all aligned samples was calculated using Bedtools coverage and the GENCODE (version 47) GTF file. The average exon coverage was 67.25x.

Variant annotation

The FASTQC reports were generated to ensure sequencing quality. FASTQ files were aligned to the human reference genome (hg 38) using the Burrows-Wheeler Aligner (BWA), Following alignment, variant calling and post-alignment processing were conducted using best practices from the Genome Analysis Toolkit (GATK) and Picard-tools-1.109. To minimize false positive detection rates, variants with a depth of coverage (DP) less than 20, a genotype quality (GQ) less than 20, and a quality value (QV) less than 50 were excluded.

For variant annotation the VCF files was then uploaded to ANNOVAR resulting in CSV files with complete information of the variants. R Studio was used for filtering of silent single nucleotide variants (SNVs).

Bioinformatics analysis

Pathogenicity of the mutations, was identified using various in silico prediction tools such as SIFT, PolyPhen-2, Mutation Taster, FATHMM and PROVEAN. The impact of SNVs on the stability of the protein was determined using SAAFEQ-SEQ prediction tool. ISPRED-SEQ prediction tool was used to analyze the interaction site mutations. ConSurf tool was used to evaluate evolutionary conservation and determine the other attributes of the mutated residues such as their nature i.e. functional, structural, buried and exposed.

The mutations were mapped using cBioportal (https://www.cbioportal.org/mutation_mapper). The wild and mutated proteins with mutations on interaction site residues were superimposed and visualized in PyMOL after modelling in the Swiss-Model. Ramachandran plot was generated to compare the wild and mutant type proteins using PROCHECK server.

Molecular dynamic simulations

GROMACS 5.1 was used to perform molecular dynamics simulations. The input file includes a pdb format file having 3D structure of protein generated by Swiss Model. A topology file, which contains the force field parameters and molecule types was generated. Solvation was performed using a solvent model like spc216.gro and ions are added to neutralize the system. Energy minimization step was performed to ensure the system starts from a stable state. NVT (constant number of particles, volume, and temperature) and NPT (constant number of particles, pressure, and temperature) simulations was performed. After the Equilibration of system, the production MD simulation is performed to collect data over the desired time period. After the NVT and NPT simulation, Root Mean Square Deviation (RMSD), radius of gyration (Rg), temperature, pressure, density, and potential energy parameters were calculated. The GROMACS results were visualized using ORIGIN PRO 2024b.

Associations study

The genes mutations were associated with demographic and clinicopathological data of OSCC patients. Various pathological parameter such as (Tumor grade, Tumor site) were included. In addition, we also studied the potential association of the mutations with other social and risk factors such as naswar, tobacco etc. Kaplan–Meier curves were generated to represent the survival probability over time for different mutation categories. This analysis was conducted using R Studio, leveraging the survival and survminer packages to facilitate the evaluation. The log-rank test was employed to compare survival distributions between groups, allowing us to determine whether the presence of specific mutations significantly impacted overall survival (Fig. 1).

Fig. 1
figure 1

An Overview of the Mutational Landscape of DNA Damage Repair Genes in Patients with Oral Squamous Cell Carcinoma.

Full size image

MuTarget based analysis

MuTarget is an open-access platform that enables researchers to link genetic mutations with gene expression changes across various human cancer types. In this study, we utilized MuTarget analysis to identify DDR mutant genes associated with the altered expression of other genes in oral squamous cell carcinoma. A P-value threshold of < 0.05 was set to define statistical significance.

An overview of the scheme of study is indicated in Fig. 2.

Fig. 2
figure 2

Scheme of study.

Full size image

Results

Demographic details

Twenty-seven (27) patients fulfilling the inclusion criteria were enrolled in the study. The study sample included 19 males and 8 females. The cohort included 70.4% of female participants (Fig. 3A) while 59% of the participants aged > 56 years (Fig. 3B). The histopathological Grading revealed that 14/27 (51.8%) were classified as well-differentiated and 13/27 (48.14%) as moderately differentiated. The distribution of OSCC by anatomical site was as follows: 10/27 cases in the tongue, 5/27 in the lip, 4/27 in the buccal mucosa, and 8/27 in other locations such as (Mandible, Oral cavity, floor of mouth, palate of mouth) (Fig. 3C). Concerning tobacco use, 10/27 (37%) patients were non-tobacco users, 15/27 (55.5%) used naswar users, and 2/27 (7.4%) were smokers. Regarding family 12/27 (%) patients revealed positive family history for cancer. Additionally, 8/27 (%) patients had a history of dental problems like infections and mouth swelling etc., whereas 19/27 (70.37%) participants apparently didn’t show any historical complications (Fig. 3D). The details are summarized in inset Fig. 3A–D.

Fig. 3
figure 3

Sample characteristics and details; (A) Gender wise distribution; (B) Age wise distribution; (C) Histopathological grading and locations; (D) General characteristics.

Full size image

Mutational profile

Results from the WES data were analyzed to map mutations on five genes i.e. TP53, ATR, ATM, CHEK1 and CHEK2 with potential role in the oncogenic transformation and each were characterized using different parameters. The overall summary of the mutations on these genes indicated in Table 1. Variants found exclusively in tumor tissue samples were categorized as somatic mutations, while variants present in both tumor and paired normal tissue samples were classified as germline mutations47. We identified that 16.7% (07/42) of the mutations are not reported previously (Fig. 4A). The identified novel mutations on TP53 are TP53p.C110Afs*5, TP53p.D2Ifs*2 and TP53p.W14Cfs*25. Other novel mutations identified are ATRp.E125Hfs*9, ATMp.E2932Rfs*6, CHEK1p.E76Kfs*21 and CHEK2p.P448Lfs*51. On the TP53 and ATR genes, a single germline mutation was reported i.e. TP53p.P33R and ATRp.M211T while no germline mutations were identified on ATM and CHEK2. The total germline mutations were reported to be 11.9% (5/42). In total, the somatic mutations were 88.09% (37/42) and were distributed as 92.8%, 90% and 40% on TP53, ATR and CHEK1 respectively, while no germline mutations were reported on ATM and CHEK2 (Fig. 4B). The majority of these mutations were nonsynonymous SNVs i.e. 66.6% (28/42), followed by frameshift deletions (7/42) i.e. 16.6% and stop gain mutations (5/42) representing 11.9% (Fig. 4C). Notably, the CHEK1p.I437V mutation was reported in all patients (100%) of oral cancers, whereas, TP53p.P33R and ATRp. M211T were reported in 77.8% and 74.07% patients and therefore, these mutations can be a potential candidate for biomarker applications as indicated in Fig. 4D.

Table 1 Mutational Spectrum of TP53, ATR, ATM, CHEK1 and CHEK2.
Full size table
Fig. 4
figure 4

Characterization of mutations across TP53, ATR, ATM, CHEK1 and CHEK2: (A) shows the novel mutations, (B) the distribution of germline versus somatic mutations; (C) the types of mutations observed within the cohort; and (D) mutations with potential biomarker applications.

Full size image

The inset Fig. 5A–E illustrates the lollipop plots of mutations revealing the mutations and their location on their respective genes. Figure 5A–E shows the lollipop plots for TP53, ATR, ATM, CHEK1 and CHEK2 respectively.

Fig. 5
figure 5

Lollipop plots obtained by cBioportal visualization tool Mutation Mapper showing the distribution of different mutations (A) TP53; (B) ATR; (C) ATM; (D) CHEK1 and (E) CHEK2 in enrolled cohort.

Full size image

The inset of Fig. 6A–E reveals the exon wise distribution of the mutations which was found to be highest on exon 3 for TP53 (7/14; 50%), exon 4 for ATR (5/10; 50%), exon 1 for CHEK1 (3/5; 60%).

Fig. 6
figure 6

Frequency of Mutations on Exons: (A) TP53; (B) ATR; (C) ATM; (D) CHEK1 and (E) CHEK2.

Full size image

Pathogenicity predictions

SIFT, PolyPhen-2, Mutation Taster, FATHMM and PROVEAN bioinformatics tools were used for the pathogenicity predictions as indicated n Fig. 7A–E.

Fig. 7
figure 7

Pathogenicity of TP53, ATR, ATM, CHEK1 and CHEK2 Mutations Based on SIFT, PolyPhen-2, Mutation Taster, FATHMM and PROVEAN: (A) SIFT Predictions; (B) PolyPhen-2 Predictions; (C) Mutation Taster Predictions; (D) PROVEAN Predictions and (E) FATHMM Predictions, (F) ClinVar, (G) Alpha missense.

Full size image

The predictions from the SIFT database revealed that 10/42 (23.8%) of the mutations were of deleterious nature, whereas, predictions from the PoyPhen-2 database revealed that 7/42 (16.7%) of the mutations were classified under probably damaging. Predictions from Mutation Taster revealed 10/42 (23.8%) mutations as disease causing, whereas, 12/42 (28.5%) were predicted as deleterious by PROVEAN. TP53p.R43H, TP53p.L125Q, TP53p.R116Q, TP53p.C110Y, TP53p. L62Fmutations were predicted as pathogenic across all databases. ATR mutations i.e. ATRp.H120Y, was reported to be pathogenic by SIFT, PolyPhen-2 and Mutation Taster. ATM mutations such as ATMp.P1054R and ATMp.T2934N was predicted as deleterious by SIFT, PolyPhen-2, Mutation Taster and PROVEAN. No predictions are reported for CHEK2p.P448Lfs*51. The results are summarized in Fig. 7A–E.

SAAFEQ-SEQ tool was used to predict the effect of mutations on the stability and structural integrity of proteins. All single nucleotide variants were determined to have a destabilizing effect on the respective proteins as indicated in Table S1 Additionally, the interaction site prediction tool (ISPRED-SEQ) was used to identify the interaction site mutations of of TP53, ATR, ATM, CHEK1 and CHEK2 as indicated in Table S2. The mutations TP53p.R43H, TP53p.R116Q, TP53p.C110Y, TP53p.E214X, TP53p.R210X, TP53p.C110Afs*5 and TP53p.S108Ffs*23 were identified as interaction site (IS) mutations on TP53, while ATRp.M1932T and CHEK1p.E76Kfs*21 were identified as IS mutations on ATR and CHEK1, respectively. No IS mutations were detected on ATM and CHEK2. Overall, 21% of the mutations were predicted as IS mutations. A summary of these ISPRED-SEQ results is provided in Table S2 and Fig. 8.

Fig. 8
figure 8

ISPRED and ConSurf Predictions of TP53, ATR, ATM, CHEK1 and CHEK2 Mutations.

Full size image

To determine the evolutionary conservation scores of the mutation residues, ConSurf tool was used and the results are indicated in Fig. 8 and Table S3. ConSurf tool scores the evolutionary conservation based on their structural and functional significance. The results indicate that all TP53 interaction site mutations (TP53p.R43H, TP53p.C110Y, TP53p.R116Q) and TP53p.L62F are situated in highly conserved regions, each receiving the highest conservation score of 9. TP53p.R43H and TP53p.R116Q were of exposed and functional nature while TP53p.C110Y and TP53p.L62F were predicted as structural residues with buried nature. For ATM mutations, ATMp.T2934N was at highest conservation score (9) and predicted as buried while ATMp.P1054R and ATMp.D1853N conservation scores were 8 and were of exposed and functional nature. For ATR, the IS mutation ATRp.M1932T was found to be in an exposed region with a conservation score of 4.

PyMOL was used for the superimposition and visualization of the IS mutations as indicated in Fig. 9.

Fig. 9
figure 9

TP53 and ATR interacting site mutation visualization and superimposition in PyMOL.

Full size image

MDS simulations

Based on our ISPRED findings, we shortlisted only the interacting site SNV mutations for molecular dynamic simulations i.e. TP53p.R43H, TP53p.R116Q, TP53p.C110Y and ATRp.M1932T using GROMACS package 5.1 and various parameters such as Rg (Radius of gyration), root-mean-square deviation (RMSD), pressure, temperature, density, and potential for the mutant and wild type proteins were studied. Ramachandran plots were also generated. The results are indicated in Figure S1A–J to Figure S4A–J for TP53p.R116Q, TP53p.R43H, TP53p.C110Y and ATRp.M1932T respectively. All the mutant version of the proteins had larger radius of gyration as indicated in Figure S5, which indicates the perturbed protein folding behavior. Generally, a larger Rg value is associated with aberrant protein folding and less stability. Among the studied mutations, the TP53p.R116Q, revealed average Rg values of 2.12 nm for wild type and 2.93 nm for the mutant protein. For ATR IS mutation i.e. ATRp.M1932T, the Rg values for mutant protein was 1.80096 which was higher as compared to the wild type.

Root mean square deviation (RMSD) measurements were used to evaluate the structural changes across the backbone of the mutant and wild type. For TP53p.R43H, deviations were detected at 5 ns (Figure S2B). TP53p.C110Y showed a substantial deviation after 4 and 7 ns as indicated in Figure S3B, while ATRp.M1932T exhibited major deviations after 4 ns and minor deviations after 9 ns (Figure S4B). Additional parameters such as temperature, pressure, and density showed both minor and major fluctuations, indicating that these mutations have affected the stability of proteins.

Ramachandran plots for the selected interaction site (IS) mutations were generated to examine changes in the favorable regions of both mutant and normal proteins, as detailed in Figures S1–S4. For the TP53p.C110Y (Figure S3I–J) and TP53p.R43H (Figure S2I–J) mutations, the percentages of residues in favorable regions were found to be 87.7% for the mutant and 88.2% for the normal for each. The TP53p.R116Q showed, with 87.7% and 84.1% of residues in favorable regions for the normal and mutant proteins, respectively (Figure S1I–J). For the ATRp.M1932T mutation, the percentage of residues in favorable regions was 93.3% and 93.2% for normal and mutant proteins (Figure S4I–J).

Clinicopathological associations

Association with demographic and histopathological data

The distribution of various mutations across demographic and clinical factors reveals distinct patterns. The TP53p.P33R mutation is most prevalent among individuals ≤ 56 (9/11; 81.80%) with higher frequencies observed in male participants (15/19; 78.90%). The distribution of TP53p.P33R across tumor site was 70% (7/10) in tongue, 75% (3/4) in buccal mucosa and 80% (4/5) in lip (80%). In addition, TP53p.P33R was found in well differentiated (10/14; 71.4%) and moderate differentiated tumors (11/13; 84.6%) (Figure S6A). The ATR mutation, i.e. ATRp.M211T was found prevalent in age group ≤ 56 (9/11; 81.82%), with higher frequencies in males (16/19; 84.21%) and is associated with tumors in the tongue and buccal mucosa, as well as well-differentiated tumors. Mutations like ATRp.V895M and ATRp.V316I were not found in patients in the age group ≤ 56 years and was not found in the female group (Figure S6B).

ATM p.D1853N and ATM p.H1380Y mutations are most common in individuals over 56 years and in females, with ATMp.D1853N showing associations with lip tumor site (3/5; 60%) and moderately differentiated tumors (2/13; 15.38%) (Figure S6C). CHEK1p.I437V mutation is universally present across all ages, genders, and tumor sites, and is associated with both well-differentiated and moderately differentiated tumors. CHEK1p.T29A, CHEK1p.P31A and CHEK1p.W79X mutations are more common in older individuals (> 56 years) and associated with moderately differentiated tumors (Figure S6D). The results are summarized in Figure S6A-D.

Association with risk factors

Strong association was observed between TP53p.P33R and naswar users (13/15; 86.60%), with potential relation to family history (11/12; 91.60%). In contrast, variants like TP53p.R43H, TP53p.C110Y, TP53p.L62F, TP53p.E214X, TP53p.C110Afs*5 and TP53p.L125Q show no association with naswar users. Variants such as TP53p.R43H, TP53p.C110Y, TP53p.R116Q, TP53p.G134X and all the 3 frameshift mutations were associated with positive family history (Figure S7A). In case of ATR mutations ATRp.M211T was found to be prevalent among naswar users (11/15; 73%). Mutations such as ATRp.R2361Q was found only in individuals with a no family history of cancers (Figure S7B). The ATM mutations like ATMp.H1380Y was only found in naswar users. Specifically, mutations such as ATMp.D1853N and ATMp.H1380Y are notably prevalent among naswar users and having dental problems, with ATMp.H1380Y also linked to individuals with a positive family history of cancer. Mutations like ATMp.H674R, ATMp.C1251F, ATMp.L1420F, ATMp.T2934N, ATMp.E2932Rfs*6 and ATMp.P1054R are primarily found in non-tobacco users (Figure S7C). CHEK1 mutations show that CHEK1p.I437V is present in all groups, while CHEK1p.T29A, CHEK1p.W79X and CHEK1p.P31A are common among naswar users and those with a positive family history except CHEK1p.W79X. All the mutations except CHEK1p.E76Kfs*21 were showed link with dental problems. CHEK1p.E76Kfs*21 appears in non-tobacco users (Figure S7D). The results are summarized in Figure S7A–D.

Association with overall survival

We conducted Kaplan–Meier survival analysis using R Studio and its associated libraries, survival and survminer, to evaluate the association between the identified mutations and overall survival in oral squamous cell carcinoma (OSCC) patients. This analysis aimed to elucidate the impact of specific genetic mutations on patient prognosis.

The results demonstrated that TP53, ATR, and ATM mutations did not exhibit a significant correlation with overall survival, with p-values of 0.64, 0.74, and 0.52, respectively. However, within the TP53 gene, a significant relationship was observed specifically for patients with the germline TP53p.P33R mutation compared to those with both germline and somatic mutations, with a p-value of 0.012. Patients harboring both germline and somatic mutations in TP53 exhibited a lower overall survival compared to those with only the germline mutation (Fig. 10A–D).

Fig. 10
figure 10

Kaplan–Meier Plot showing Survival analysis: (A) TP53 Wild Type Vs Mutated; (B) TP53 Germline Vs Germline + Somatic; (C) ATR Wild Type Vs Mutated; (D) ATM Wild Type Vs Mutated.

Full size image

MuTarget based analysis of DDR genes

The MuTarget analysis, filtered for cancer hallmark genes, revealed significant findings for TP53 and ATR mutations but showed no associated gene expression changes for ATM mutations. In TP53-mutant oral squamous cell carcinoma (OSCC), genes such as SERPINE1, CDK6, MET, MMP10, and CAV1 were upregulated, reflecting key cancer hallmarks including cell cycle dysregulation, metastasis, and immune evasion. SERPINE1 suggests enhanced metastasis, while CDK6 indicates unchecked cell cycle progression. MET promotes tumor motility, and MMP10 and CAV1 are linked to increased invasiveness (Figure S8A). In ATR-mutant cancers, FEN1 was downregulated (impairing DNA repair), while EDIL3 was upregulated, suggesting heightened invasiveness and metastatic potential (Figure S8B). However, for ATM mutations, the analysis showed no significant changes in cancer hallmark genes, indicating a lack of associated molecular alterations. Additionally, mutations in CHEK1 and CHEK2 also had limited findings due to small sample sizes.

Discussion

The quest for genetic biomarkers in oral squamosus cell carcinoma (OSCC) continues to be a focal point of research, particularly given the role of genetic mutations in disrupting the molecular cascades that regulate DNA damage repair mechanism. The DNA damage repair (DDR) genes and their function is crucial in various aspects of cancer. Deficiencies in DDR mechanisms are well recognized for tumor development and progression48,49. Protein interactions, crucial for maintaining normal cellular functions50, which can be significantly altered by mutations, which results in a truncated protein eventually destabilizing the protein interactome. Using the STRING server and GeneMania, we analyzed the interaction networks for TP53, ATM, ATR, CHEK1 and CHEK2 (Figures S9 and S10). STRING analysis identified several key KEGG pathways for TP53, including p53 signaling pathway, DNA damage repair (Homologous recombination) and cell cycle checkpoints. For ATM, ATR, CHEK1 and CHEK2, notable pathways include homologous recombination, fanconi anemia and p53 signaling with exception of mismatch repair pathway exclusively associated with ATM (Figure S9). GeneMania interactions further illustrated the extensive signaling networks potentially disrupted by mutations in TP53, ATM, ATR, CHEK1 and CHEK2, contributing to oral squamous cell carcinoma progression as shown in Figure S10.

For pathogenicity analysis 5 bioinformatics tools prediction were utilized. SIFT is a commonly utilized computational tool that predicts the functional effects of mutations by assessing the evolutionary conservation of amino acid residues and their likely impact on protein function51. PolyPhen-2 predicts the impact of amino acid changes on protein function by analyzing sequence features and evolutionary conservation, categorizing mutations as benign, possibly damaging, or probably damaging52. Mutation Taster predicts the potential pathogenicity of genetic variants by evaluating their likelihood to cause disease based on known mutations and functional data, categorizing them as disease-causing or benign53. FATHMM assesses the functional impact of genetic variants on protein function by integrating sequence features and evolutionary conservation, classifying mutations as either damaging or neutral54,55. PROVEAN assesses the impact of amino acid substitutions and insertions/deletions on protein function by analyzing sequence conservation and potential functional changes, categorizing mutations as either deleterious or neutral56.

Research on these key DNA damage repair genes mutations in underrepresented populations, such as the Pakistan, is sparse. TP53 is a key tumor suppressor that responds to DNA damage by halting the cell cycle at the G1/S or G2/M checkpoints, allowing time for repair. It activates genes involved in DNA repair and, if damage is irreparable, induces apoptosis to prevent tumor development, thereby maintaining genomic stability28.

Previous literature reports that TP53 is the most frequently mutated gene across various cancer tissues57. Our study also supports this finding, revealing a high frequency of TP53 mutations in oral squamous cell carcinoma patients from Pakistan. In one study TP53 mutations was observed in 80.6% of OSCC cases and our small cohort reported it in 85.2% cases. The Non synonymous SNVs was reported as the most common mutations in TP53, accounting for 64.2% of cases58. In our cohort, 6/12 of TP53 mutations were identified as non-synonymous SNVs, with TP53p.P33R being prevalent that has been reported in lung cancer59. Notably, we identified 2 novel frameshift deletion mutation (TP53p.C110Afs*5 and TP53p.D2Ifs*2) in exon 3. TP53p.C110Afs*5 was present on DNA binding domain which has an important role in gene expression. Another frameshift insertion somatic mutation, TP53p.W14Cfs*25, was found in the trans activation domain (TAD). The TAD is responsible for the activation of TP53. This domain interacts with various transcriptional coactivators and initiates the transcription of TP53 involved in cell cycle arrest, DNA repair, and apoptosis. Mutations in this domain can impair the protein’s ability to activate transcription of its target genes. Overall, the TP53 mutation sites were diverse but most of the mutations (7/12) were present in DNA binding domain and missense mutations were significantly reported. These OSCC patient findings are similar to previous literature of TP53 mutations in other cancers60,61. Previous studies62 have shown an association between TP53 mutations and overall survival (OS) (p = 0.009). However, in our study, TP53 did not demonstrate a significant correlation with OS (p = 0.64). In contrast, when comparing germline mutations to germline + somatic mutations, TP53 showed a significant relationship with OS (p = 0.012).

ATM (Ataxia Telangiectasia Mutated) is a crucial protein kinase involved in detecting and responding to DNA damage, particularly double-strand breaks, and coordinating DNA repair and cell cycle regulation63. ATM mutations are frequently reported in the DNA damage response (DDR) across various cancers64, particularly in non-small cell lung carcinoma, where ATM is identified as the most frequently mutated DDR gene65 while ATM protein expression loss in prior studies has been reported in up to 41% of tumors66. Our study supports this literature as ATM stands as the 2nd most frequent gene mutated on OSCC samples. It is mainly associated with female sex as reported by R. Biagio et al. in non-small cell lung carcinoma67. We identified a novel frameshift mutation (ATMp.E2932Rfs*6) in the PI3/PI4 kinase domain, which is critical for regulating cell growth and signaling. This is a truncating mutation and ATM truncating mutations can lead to various forms of C-terminally truncated ATM proteins. Inherited truncations are associated with ataxia-telangiectasia syndrome, which significantly elevates cancer risk, including a 20% to 30% lifetime risk of lymphoid, gastric, breast, central nervous system, skin, and other cancers68. In our cohort of OSCC no germline mutation of ATM was reported. ATM shows significant relation with OS in metastatic colorectal cancer69 (p = 0.01) while in OSCC cohort its relation with OS was not significant (p = 0.52).

ATR (Ataxia Telangiectasia and Rad3 Related) is a crucial protein kinase that plays a central role in managing replicative stress (RS) and regulating the cell cycle. Loss of G1 checkpoint control is nearly universal in cancer, leading to an increased dependence on the S and G2/M checkpoints and ATR signaling to manage DNA damage. Thus mutation in ATR gene may lead towards cancer progression70. In our study one interacting site mutation (ATRp.M1932T) of ATR was in FAT domain (FRAP, ATM, and TRRAP) that facilitates protein–protein interactions and substrate recognition, and 2 mutations (ATRp.R2361Q and ATRp.R2425Q) was present in PI3/PI4-kinase domain that is crucial for its kinase activity in phosphorylating substrates involved in DNA damage response and cell cycle regulation. A novel truncating mutation (ATRp.E125Hfs*9) was also observed in our cohort. Truncating mutations in ATR produce C-terminally truncated proteins, leading to significant genetic instability71 when combined with mismatch repair deficiencies. In mice, heterozygous ATR loss correlates with increased tumorigenesis72.

CHEK1, a serine/threonine kinase from the CHEK family, plays a crucial role in mediating cell cycle arrest in response to DNA damage. Its action involves activating ATM and ATR, which phosphorylate TP53 and other CHEKs, thereby initiating DNA repair mechanism. Its mutations is reported in endometrial, colorectal and stomach carcinomas73,74,75. CHEK1 is altered in 0.80% of all cancers and mutated in 2.62% of malignant solid tumors76. In contrast to this finding our OSCC cohort shows germline mutation of CHEK1 (CHEK1p.I437V) in 100% of samples while other 2 germline mutations (CHEK1p.T29A and CHEK1p.P31A) was present in 18.5% of cases. The somatic mutations were recorded in 3.5% of cases which supports the given literature76.

The CHEK2 gene encodes checkpoint kinase 2 (CHK2), crucial for the ATM-CHK2-p53 pathway that responds to DNA double-strand breaks and prevents early tumorigenesis77. Initially linked to moderate breast cancer risk, CHEK2 mutations are now associated with a wider range of cancers78,79, making it a key focus in genetic testing for hereditary cancer.

In our study cohort only 1 novel somatic truncating mutation (CHEK2p.P448Lfs*51) was reported in Pkinase domain which is the central domain responsible for its catalytic activity, structural and regulatory role.

The lack of significant association between these genes and overall survival (OS) may be attributed to the small cohort size. Larger sample sizes are needed to more accurately determine the relationship and potential impact of these genes on OS.

Conclusions

This study analyzed genetic mutations in sporadic oral squamous cell carcinoma (OSCC) patients, focusing on TP53, ATR, ATM, CHEK1, and CHEK2 genes. The cohort of 27 patients revealed a high frequency of somatic mutations, with TP53 showing the highest mutation frequency. A notable finding was the CHEK1p.I437V mutation, present in all patients, suggesting its potential as a biomarker, whereas, TP53p.P33R and ATRp.M211T was reported in 77.7% and 74% patients indicating their potential as biomarker. Analysis of the mutations’ pathogenicity through various bioinformatics tools highlighted the complex nature of predicting their impact, with some mutations, like TP53p.R43H, consistently predicted as pathogenic. All the 42 mutations were predicted to have a destabilizing effect on protein, which was confirmed through various bioinformatic tools and molecular dynamic simulations. The simulations also showed that the radius of gyration of mutant proteins was higher as compared to the wild type indicating their instability and perturbed folding behavior. The association results showed that the TP53p.P33R was found predominantly in the naswar users. Kaplan–Meier survival analysis indicated that while mutations in TP53, ATR, and ATM did not significantly affect overall survival, patients with both germline and somatic TP53 mutations had a significantly lower survival rate compared to those with only germline mutations. These findings underscore the importance of understanding mutation-specific effects and their potential clinical implications in OSCC.

This study has several limitations, including a small sample size and its conduction across a limited number of centers, which may have introduced bias into the final results. Additionally, due to the small sample size, it was not feasible to effectively coordinate molecular research with clinical data. We recommend further studies in larger cohorts from the same population to further characterize the penetrance of the mutations in OSCC patients. The findings shall help in developing strategies for the management of OSCC patients in local settings.