A DNA damage repair associated signature for HNSCC

doi:10.21203/rs.3.rs-3595930/v1

Download PDF

Research Article

A DNA damage repair associated signature for HNSCC

https://doi.org/10.21203/rs.3.rs-3595930/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background:

DNA damage repair (DDR) genes play a vital role in tumorigenesis and the tumor microenvironment. However, the specific role of DDR genes in head and neck squamous cell carcinoma (HNSCC) is not fully understood.

Method:

Gene expression profiles from the TCGA HNSCC cohort were used to identify DNA damage repair genes with prognostic value. LASSO regression was used to construct a prognostic model for HNSCC. Survival analysis and receiver operating characteristic (ROC) curve analysis were used to validate the model. Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis were performed to identify potential pathways associated with the signature model. Gene set enrichment analysis (GSEA) and the Tumor Immune Dysfunction and Exclusion (TIDE) algorithm were used to investigate the tumor microenvironment and immunotherapy response of HNSCC patients, respectively. Finally, a nomogram was constructed to validate the clinical usefulness of the signature model.

Result:

We constructed a six-gene DNA damage repair-related signature consisting of DCLRE1C, TOP3B, POLE2, POLD2, ERCC2, and RAD23B. The results of survival analysis and receiver operating characteristic (ROC) curve analysis suggested that the signature had good prognostic value. The risk score based on the signature model was significantly correlated with immune cell infiltration and expression levels of immune markers in HNSCC, and could predict the immunotherapy response for HNSCC patients. Finally, a nomogram was constructed and showed good consistent fitness between the predicted and observed values for 1-, 3-, and 5-year overall survival (OS).

Conclusion:

We developed a DNA damage repair signature for HNSCC that is associated with survival prognosis and the tumor microenvironment, suggesting its potential for clinical management.

DNA repair

tumor microenvironment

The Cancer Genome Atlas

prognosis

Head and neck squamous cell carcinoma (HNSCC) is one of the most lethal kind of cancers[1]. Emerging research had indicated that HNSCC may benefit from immunotherapy and improve patient’s survival prognosis[2]. However, biomarkers to identify patients who benefit from immunotherapy are limited.

Living organisms experience DNA damage when exposed to DNA-damaging agents. The accumulation of DNA mutations can lead to tumor development and progression.[3]. However, the inherited DNA repair mechanism would detect the DNA damage and activate the DNA repair programmes and thus ensure the integrity of DNA and the progress of DNA synthesis and division[4]. The role of DNA damage repair genes in tumorigenesis and progression has been widely investigated. DNA damage repair pathways, such as mismatch repair, base excision repair, nucleotide excision repair, and homology-directed repair, are essential for genetic stability. Defective DNA repair mechanisms can lead to accumulated genome instability and tumorigenesis. Target therapy against DNA damage repair genes is an emerging treatment for cancer therapy. For example, poly ADP-ribose polymerase (PARP) inhibitors have shown efficacy in BRCA1/2 mutant cells and tumors[5], suggesting that DNA damage repair pathways are promising target for tumor therapy. Furthermore, immunotherapy such as PD-1 and PD-L1 have improved survival in a subset of cancer patients[6], and the genomic signature correlates with the response to immune checkpoint therapies. However, the role of DNA damage repair genes in maintaining genome stability and predicting immunotherapeutic response is rarely reported in HNSCC.

In this study, we analyzed the RNA-seq data and clinical characteristics attained from In this study, we analyzed RNA-seq data and clinical characteristics from The Cancer Genome Atlas (TCGA) to explore the survival prognostic role of DNA damage repair genes in HNSCC. We constructed a DNA damage repair genes-related signature that predicts the survival prognosis and immunotherapy response for HNSCC. We also constructed a nomogram for HNSCC based on the prognostic signature and prognostic clinical characteristics to predict the 1-, 3-, and 5-year overall survival, indicating its usefulness in clinical management. Overall, this study provides a signature for predicting survival prognosis and immunotherapy of HNSCC.

Data Acquisition and Processing

We first downloaded the TCGA HNSCC cohort of RNA-seq data, clinical information and the probe annotation files from online database Santa Cruz Xena Browser (https://xenabrowser.net/). A total of 511 HNSCC samples were acquired and selected for the following analysis. Further, the probe IDs of HNSCC gene expression cohort were transformed into gene symbols using the annotation files. GEO database (https://www.ncbi.nlm.nih.gov/geo/) was used to acquire the HNSCC external validation cohorts GSE27020 using R package “GEOquery.”[7] The probe IDs of the BLCA gene expression matrix were transformed into gene symbols according to the annotation files.

Construction of a DNA damage repair related risk signature for HNSCC

We acquired a total of 105 DNA damage repair genes which consist of 5 DNA Damage Repair pathways from Msigdb (https://www.gsea-msigdb.org/gsea/msigdb/). Then the damage repair genes which shown significant survival prognosis in HNSCC identified by univariate Cox regression analysis (p < 0.05) were selected for the Construction of the signature model. Firstly, TCGA HNSCC cohort (n = 511) were divided into training (n = 306) and testing cohort (n = 205) at a 6:4 ratio. Then, to identify the genes with most significant prognostic value for HNSCC, we performed least absolute shrinkage and selection operator (LASSO) regression using R package “glmnet” in the training cohort and five genes were selected for the construction of the models. We calculated the risk score of each sample using the following formula:

Risk score =

Functional Enrichment analysis

We first determined the risk score of each HNSCC sample calculated by the risk signature model and divided the samples into high- and low- risk groups according to the median of risk score. Then we explored the differentially expressed genes (DEGs) between the two risk groups was by setting the cutoff criteria as |logFC| > 0.5 and adjusted P value < 0.05 using R package “limma” [8]. Next, to identify the potential biological pathways associated the risk signature, we performed GO and KEGG enrichment analysis for the DEGs using R package “clusterPofolier” [9]. Furthermore, gene set enrichment analysis (GSEA) was also performed to identify significant pathways between the two risk groups. Adjusted P value less than 0.05 was considered as the cutoff value to identify significant pathway. Finally, the DNA pathways activity of each TCGA HNSCC samples were estimated using the DNA damage repair genes by R package “GSVA” [10].

Estimating immune cells infiltration in HNSCC

We first identified the immune cell infiltration level of each TCGA HNSCC samples using a set of marker genes representing different immune cell types using the ssGSEA algorithm from R package “GSVA” [6].

Predicting immunotherapy response

To explore the potential clinical usefulness of the gene signature, we predicted the clinical response rate of each TCGA HNSCC sample to immunotherapy (Including PD-1 and CTLA-4 antibody) using the Tumor Immune Dysfunction and Exclusion (TIDE) algorithm (http://tide.dfci.harvard.edu) [11].

Construction and evaluation of Nomogram model

To better predict the survival prognosis of the HNSCC samples, we constructed a nomogram including clinical characteristic and the risk signature using the R package “rms”. Furthermore, we also evaluated the accuracy of the nomogram by constructing the Receiver operating curve to predicting the 1-, 3- and 5- year overall survival for HNSCC patients using R package “ROCsurvival”.

Survival analysis

We compared the survival prognosis between the high-risk and low-risk group using the Kaplan-Meier curve and log-rank test. R package “survival” was used to perform all the survival analyses and log-rank test. R package “surviminer” was used to visualize the Kaplan-Meier curve.

Statistical analysis

All of the statistical analyses in this study were performed in R version 4.0.2. Pearson’s correlation test was used to evaluate the correlation between the signature score and the immune cell infiltration level, the expression of the characteristic gene and the immune marker, respectively. Student's t tests were used to determine statistical significance among two groups. P value less than 0.05 was considered statistics significant.

Prognostic value of the DNA damage repair pathways in HNSCC

We first estimated the DNA damage pathways activity in HNSCC samples using single cell Gene Set Enrichment Analysis (GSEA). The results showed that high scores of base excision repair (BER), homologous recombination (HR), and mismatch repair (MMR) were associated with better survival prognosis in HNSCC (Fig. 1A, B, C). However, there was no significant prognostic difference between the high and low score groups for nucleotide excision repair (NER) and non-homologous end joining (NHEJ) (Fig. 1D, E).

These findings suggest that BER, HR, and MMR play important roles in the survival prognosis of HNSCC. Further studies are needed to investigate the underlying mechanisms and to develop targeted therapies that can improve the prognosis of patients with HNSCC.

Construction and Evaluation of a DNA repair related signature model for HNSCC

First, we identified DNA damage repair genes with prognostic value in HNSCC using univariate Cox regression. A total of 10 genes were identified with a P value less than 0.05 and were selected as the candidate genes for the construction of the signature model (Fig. 2A).

The HNSCC cohort was then randomly divided into the training (n = 306) and testing (n = 205) cohort at a ratio of 6:4. LASSO regression was then performed using the training cohort to identify the DNA damage genes that had the greatest importance on the survival outcomes in HNSCC (Fig. 2B-2C). Finally, we established a risk signature model consisting of six genes (Fig. 2D): DCLRE1C, TOP3B, POLE2, POLD2, ERCC2, and RAD23B.

The survival prognostic value of the risk signature consisting of six DNA damage repair genes was evaluated in the training and testing cohort. The risk score of each HNSCC sample was calculated according to the signature model and ordered based on the risk score in the training cohort. The scatter plot representing the overall survival status of HNSCC patients showed that the samples in the high-risk group were associated with a higher mortality rate than those in the low-risk group (Fig. 3B). We also constructed a heatmap presenting the signature gene expression profiles, which showed that HNSCC samples with higher signature scores tended to exhibit higher expression levels of POLD2, ERCC2, and RAD23B, while those with lower risk scores tended to exhibit elevated levels of TOP3B, POLE2, and DCLRE1C (Fig. 3C).

The above analysis was also performed in the testing cohort and the distribution plot representing the survival status (Fig. 3D), scatter plot indicating the risk score (Fig. 3E), heatmap presenting the gene expression (Fig. 3F) were consistent to that in the training cohort, respectively.

Finally, the survival prognostic value of the risk signature was evaluated in the training and the testing cohort. The results indicated that the higher risk score was significantly associated with poor survival prognosis in the training cohort (Fig. 4A, P < 0.001) and the testing cohort (Fig. 4B, P = 0.0072). These findings indicate the high predictive capacity of the risk signature in both the training and testing cohorts.

To further validate the prognostic value of the model, we tested the model in the external GEO cohort GSE27020. Survival analysis was performed and the Kaplan-Meier curve indicated that higher signature score was associated with poor survival prognosis in GSE27020 (Fig. 4C, P = 0.031). In summary, the DNA damage risk signature showed good performance in predicting the survival prognosis in HNSCC.

Identification of Potential Pathways Associated with the DNA Damage Repair Signature

We first explored the distribution of the Gene Set Enrichment Analysis (GSEA) score of each DNA damage repair pathway between the high- and low-risk groups. We found that all of the pathways were significantly higher expressed in the low-risk group than in the high-risk group, including base excision repair (Fig. 5A, P = 5.1e-09), homologous recombination (Fig. 5B, P = 2e-07), mismatch repair (Fig. 5C, P = 4.2e-11), non-homologous end joining (Fig. 5D, P = 0.00016), and nucleotide excision repair (Fig. 5E, P = 4.3e-05).

These findings suggest that the DNA damage repair signature is associated with the expression of these pathways. This is consistent with our previous analysis, which showed that a low GSVA score of the DNA damage repair signature was associated with poor survival prognosis in HNSCC samples.

The results of this study provide further evidence that the DNA damage repair signature is a potential biomarker for predicting the survival prognosis of HNSCC patients. The identification of the potential pathways associated with the signature could help to elucidate the underlying mechanisms of HNSCC and to develop targeted therapies for this disease.

Exploration of the potential Biological Pathways Associated with the Signature Model

Next, to explore the biological pathways associated with the signature model, we first identified the differentially expressed genes (DEGs) between the two risk groups. A total of 2347 DEGs were identified (Fig. 6A).

We then performed Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis to explore the potential biological pathways associated with the DEGs. The GO analysis indicated that the risk signature was mainly involved in the extracellular matrix organization, extracellular structure organization, and T cell activation (Fig. 6B). The KEGG enrichment analysis indicated that the DEGs were mainly enriched in cell adhesion molecules, ECM-receptor interaction, and focal adhesion (Fig. 6C).

The GO and KEGG analysis together suggest that the risk signature may be associated with the extracellular matrix. We also explored the potential pathways using Gene Set Enrichment Analysis (GSEA) and the results indicated that the signature correlated with typical tumor progression pathways including apical junction, E2F targets, and epithelial-mesenchymal transition (EMT) (Fig. 6D).

The results of this study suggest that the DNA damage repair signature is associated with a number of biological pathways, including those involved in the extracellular matrix, T cell activation, and tumor progression. These findings provide further evidence that the signature is a potential biomarker for predicting the survival prognosis of HNSCC patients. The identification of the potential pathways associated with the signature could help to elucidate the underlying mechanisms of HNSCC and to develop targeted therapies for this disease.

The signature model correlated with tumor microenvironment and immunotherapy response

The previous analysis suggested that the risk signature was associated with the epithelial-mesenchymal transition (EMT) related pathways. We next analyzed the correlation between the risk signature and the immune microenvironment. We first estimated the infiltration levels of different immune cell types in TCGA HNSCC cohort and the distribution of the enrichment score of immune cells between the high and low risk groups were compared.

As shown in Fig. 7A, many types of immune cells had a higher infiltration level in the low-risk group than the high-risk group, including activated B cells, activated CD4 T cells, and activated CD8 T cells. This suggests that the risk signature may be associated with a favorable immune microenvironment.

Next, we performed correlation analysis between the risk score and expression of four immune markers including CTLA-4, PD-1, PD-L1, and PD-L2. We found that CTLA-4 significantly negatively correlated with the signature score (Fig. 7B). This suggests that the risk signature may be inversely correlated with the expression of CTLA-4, which is a checkpoint protein that can inhibit T cell activation.

PD-1, PD-L1, and PD-L2 did not show significant correlation with the risk score (Fig. 7C-E). This suggests that the risk signature may not be directly correlated with the expression of these checkpoint proteins.

These results suggest that the risk signature may predict the effectiveness of immunotherapy. Further studies are needed to confirm this finding.

Construction of A Nomogram Predicting Prognosis of HNSCC patient

To better predict the survival prognosis for HNSCC samples, we further tried to establish a nomogram for HNSCC. First, both univariate and multivariate Cox regression analyses were performed to evaluate whether the clinicopathological characteristics and risk score can serve as independent prognostic indicators for HNSCC patients (Fig. 8A).

We found that the risk score, sex, and TNM stage were independent prognostic factors for HNSCC patients (Fig. 8B). Next, a nomogram consisting of the five factors was constructed to predict the 1-, 3-, and 5-year prognosis of the HNSCC (Fig. 8C). We also constructed the calibration curves to evaluate the accuracy of the nomogram, and the results showed a consistent fitness between the predicted and observed values for 1-, 3-, and 5-year OS of HNSCC (Fig. 8D-8F).

The nomogram showed good performance in predicting the survival prognosis of HNSCC patients. It can be used as a clinical tool to help doctors assess the prognosis of HNSCC patients and make treatment decisions.

Head and neck squamous cell carcinoma (HNSCC) is a common cancer with a poor prognosis. There are limited treatment options available that can improve the overall survival of HNSCC patients. Additionally, there are few biomarkers that can predict the survival prognosis of HNSCC patients.

In this study, we explored the DNA damage repair pathways in HNSCC and developed a signature model for the survival prognosis in HNSCC. Many genes in our model had been reported to exert important role in the progression in different kind of tumor. For example, Pol δ is a DNA polymerase which participate in the chromosomal DNA synthesis and different kind of DNA damage repair pathways including base excision repair, nucleotide excision repair, and mismatch repair[12]. Recent studies had revealed the important role of POLD2 in the progression of tumor. Xu et al found that decreased expression of POLD2 improve the efficacy of DNA-damaging therapy in Glioblastoma in part by inhibit the cell proliferation and invasiveness[13]. The six signature genes in our established model had a significant prognostic value in HNSCC, indicating that the signature genes may exert complex biological and profound role in the progression of the HNSCC, and it is needed to be explored.

Emerging evidence had shown that epithelial-mesenchymal transition (EMT) related transcription factors controlling the DNA damage repair pathways[14]. For example, ZEB1, a conical EMT TF, was reported to activates the transcription of ATM by a series of reaction and contribute to the enhanced activity of homologous recombination pathway[15]. Furthermore, we also identified that the DNA damage repair related signature associated with many EMT associated biological pathways, including ECM-receptor interaction and Focal adhesion, suggesting that the interaction between the EMT and DNA damage repair pathways is one of the mechanisms of our signature model.

Finally, deficiency of the DNA damage repair pathways might lead to the accumulation of error during the DNA synthesis and production of neoantigens by tumor cell, indicating the potential efficacy of immunotherapy[16]. Our functional enrichment results indicated that the risk signature correlated with immune related pathways and further analysis showed the correlation between the risk signature and the immune cell infiltration level and immune marker, suggesting the risk signature may serve as a biomarker for predicting immune therapy response. We estimated the immunotherapy response of the two risk groups classified by the risk signature. It turned out that the lower risk group is more likely to benefit from the immune therapy, indicating that the signature may had the potential to predict the immunotherapy response and contribute to the clinical decision.

In summary, we explored the survival prognosis of DNA damage repair pathways in HNSCC and established a DNA damage repair related signature for predicting the survival and immunotherapy response for HNSCC, suggesting a potential use in clinical management.

Consent to publish

All authors consent the publish of this manuscript.

Acknowledgements

None

Authors^'contributions

Zhaoen Ma,Wenjing Liao and Saixuan Yang contributed equally to this work. Xiaowen Zhang, Wenjing Liao and Zhaoen Ma conceived and designed this study. Zhaoen Ma, Wenjing Liao,Saixuan Yang,Yali Xu and Guangui Chen analyzed the data. Zhaoen Ma,Wenjing Liao and Saixuan Yang wrote the paper. All authors revised and approved the manuscript.

Funding

No funding was used in this study.

Availability of data and materials

HNSCC Data generated and analyzed in the present study are available from the UCSC TCGA data portal (http://xena.ucsc.edu/public/).

Ethics approval and consent to participate

None

Competing interests

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Siegel, R.L., et al., Cancer statistics, 2022. CA Cancer J Clin, 2022. 72(1): p. 7-33.
Yu, C., et al., Current status and perspective of tumor immunotherapy for head and neck squamous cell carcinoma. Front Cell Dev Biol, 2022. 10: p. 941750.
Chatterjee, N. and G.C. Walker, Mechanisms of DNA damage, repair, and mutagenesis. Environ Mol Mutagen, 2017. 58(5): p. 235-263.
O'Connor, M.J., Targeting the DNA Damage Response in Cancer. Mol Cell, 2015. 60(4): p. 547-60.
Li, H., et al., PARP inhibitor resistance: the underlying mechanisms and clinical implications. Mol Cancer, 2020. 19(1): p. 107.
Charoentong, P., et al., Pan-cancer Immunogenomic Analyses Reveal Genotype-Immunophenotype Relationships and Predictors of Response to Checkpoint Blockade. Cell Rep, 2017. 18(1): p. 248-262.
Davis, S. and P.S. Meltzer, GEOquery: a bridge between the Gene Expression Omnibus (GEO) and BioConductor. Bioinformatics, 2007. 23(14): p. 1846-7.
Ritchie, M.E., et al., limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res, 2015. 43(7): p. e47.
Yu, G., et al., clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS, 2012. 16(5): p. 284-7.
Hanzelmann, S., R. Castelo, and J. Guinney, GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinformatics, 2013. 14: p. 7.
Jiang, P., et al., Signatures of T cell dysfunction and exclusion predict cancer immunotherapy response. Nat Med, 2018. 24(10): p. 1550-1558.
Uchimura, A., et al., DNA polymerase delta is required for early mammalian embryogenesis. PLoS One, 2009. 4(1): p. e4184.
Xu, Q., et al., ShRNA-based POLD2 expression knockdown sensitizes glioblastoma to DNA-Damaging therapeutics. Cancer Lett, 2020. 482: p. 126-135.
Kciuk, M., et al., Cancer-associated transcription factors in DNA damage response. Biochim Biophys Acta Rev Cancer, 2022. 1877(4): p. 188757.
Zhang, X., et al., ZEB1 confers chemotherapeutic resistance to breast cancer by activating ATM. Cell Death Dis, 2018. 9(2): p. 57.
Jiang, S., et al., Prognostic value of PD-1, PD-L1 and PD-L2 deserves attention in head and neck cancer. Frontiers in Immunology, 2022. 13.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

A DNA damage repair associated signature for HNSCC

Status:

Version 1

Abstract

Background:

Method:

Result:

Conclusion:

Figures

Introduction

Materials and methods

Data Acquisition and Processing

Construction of a DNA damage repair related risk signature for HNSCC

Functional Enrichment analysis

Estimating immune cells infiltration in HNSCC

Predicting immunotherapy response

Construction and evaluation of Nomogram model

Survival analysis

Statistical analysis

Result

Prognostic value of the DNA damage repair pathways in HNSCC

Construction and Evaluation of a DNA repair related signature model for HNSCC

Identification of Potential Pathways Associated with the DNA Damage Repair Signature

Exploration of the potential Biological Pathways Associated with the Signature Model

The signature model correlated with tumor microenvironment and immunotherapy response

Construction of A Nomogram Predicting Prognosis of HNSCC patient

Discussion

Declarations

Consent to publish

Acknowledgements

Authors^'contributions

Funding

Availability of data and materials

Ethics approval and consent to participate

Competing interests

References

Additional Declarations

Status:

Version 1

A DNA damage repair associated signature for HNSCC

Status:

Version 1

Abstract

Background:

Method:

Result:

Conclusion:

Figures

Introduction

Materials and methods

Data Acquisition and Processing

Construction of a DNA damage repair related risk signature for HNSCC

Functional Enrichment analysis

Estimating immune cells infiltration in HNSCC

Predicting immunotherapy response

Construction and evaluation of Nomogram model

Survival analysis

Statistical analysis

Result

Prognostic value of the DNA damage repair pathways in HNSCC

Construction and Evaluation of a DNA repair related signature model for HNSCC

Identification of Potential Pathways Associated with the DNA Damage Repair Signature

Exploration of the potential Biological Pathways Associated with the Signature Model

The signature model correlated with tumor microenvironment and immunotherapy response

Construction of A Nomogram Predicting Prognosis of HNSCC patient

Discussion

Declarations

Consent to publish

Acknowledgements

Authors' contributions

Funding

Availability of data and materials

Ethics approval and consent to participate

Competing interests

References

Additional Declarations

Status:

Version 1

Authors^'contributions