Genomic Profiling of Lower-Grade Gliomas Subtype with Distinct Molecular and Clinicopathologic Characteristics via Altered DNA-Damage Repair Features

Lower WHO grade II and III gliomas (LGGs) exhibit significant genetic and transcriptional heterogeneity, and the heterogeneity of DNA damage repair (DDR) and its relationship to tumor biology, transcriptome, and tumor microenvironment (TME) remains poorly understood. In this study, we conducted multi-omics data integration to investigate DDR alterations in LGG. Based on clinical parameters and molecular characteristics, LGG patients were categorized into distinct DDR subtypes, namely, DDR-activated and DDR-suppressed subtypes. We compared gene mutation, immune spectrum, and immune cell infiltration between the two subtypes. DDR scores were generated to classify LGG patients based on DDR subtype features, and the results were validated using a multi-layer data cohort. We found that DDR activation was associated with poorer overall survival and that clinicopathological features of advanced age and higher grade were more common in the DDR-activated subtype. DDR-suppressed subtypes exhibited more frequent mutations in IDH1. In addition, we observed significant upregulation of activated immune cells in the DDR-activated subgroup, which suggests that immune cell infiltration significantly influences tumor progression and immunotherapeutic responses. Furthermore, we constructed a DDR signature for LGG using six DDR genes, which allowed for the division of patients into low- and high-risk groups. Quantitative real-time PCR results showed that CDK1, CDK2, TYMS, SMC4, and WEE1 were significantly upregulated in LGG samples compared to normal brain tissue samples. Overall, our study sheds light on DDR heterogeneity in LGG and provides insight into the molecular pathways of DDR involved in LGG development.


Introduction
Lower-grade glioma (LGG) is one of the most common primary malignancies of the central nervous system, characterized by inevitable recurrence, high mortality, and substantial heterogeneity (van den Bent 2010). Due to the high heterogeneity of LGG, patients often have significantly different prognoses, even with the same histologic diagnosis. For stratifying patients, conventional histological classifications of gliomas are not sufficient in this era of precise medication. Currently, several potential molecular biomarkers are being identified. The IDH mutations, chromosome 1p/19q symbiosis, chromosome 7loss/10gain, EGFR amplification, and TERT promoter mutations along with mutations of ATRX, TP53, CIC, TTN, and FUBP1 have also been included in the 2021 WHO stratification classification of gliomas (Louis et al. 2021). As per a previous report, there is a classification of lower-grade gliomas as grade II and III (Gittleman et al. 2020) where 70% of patients had IDH mutations along with a combined deletion/absence of chromosome 1p19q (Molloy et al. 2020). These patients had the best prognosis with 8 years of median OS. Conversely, patients with IDH mutation and no deletion of 1p/19q (astrocytoma) displayed 6.4 years of the median OS, whereas those LGG patients with IDH wild-type mutation had 1.7 years being close to the OS of glioblastoma with IDH wild type (Paľa et al. 2019;Pallud et al. 2013;Pallud et al. 2012).
In spite of a relatively better prognosis compared to high-grade gliomas (HGGs), most LGG patients are susceptible to transformation into HGGs, recurrence, or death during the development and progression of LGG and exhibit more malignancy and aggressiveness (Binder et al. 2019;Duffau 2021). Surgical excision, radiation, chemotherapy, targeted therapy, and immunotherapy are now the predominant clinical therapies for LGG (Cai et al. 2020). However, existing therapeutic approaches are not significantly successful in improving patients' survival. Therefore, there is an urgent need to find new biomarkers to help improve the early clinical prognosis of patients with lower-grade gliomas and explore potential mechanisms of LGG progression and additional avenues of treatment.
DDR response is involved in maintaining genomic stability as well as protecting cells from endogenous and exogenous DNA damage (Li and Heyer 2008). Alterations in DDR responses leading to genomic instability and subsequently increased frequency of mutations have already been reported, which in turn further triggers the production of tumor-specific neoantigens (Chae et al. 2019;Gavande et al. 2016;Mouw et al. 2017). DDR has been reported to contain the following eight pathways that is: mismatch repair (MMR), base excision repair (BER), nucleotide excision repair (NER), homologous recombination repair (HRR), nonhomologous end-joining (NHEJ), checkpoint factors (CPF), Fanconi anemia (FA), and translesion DNA synthesis (TLS) (Scarbrough et al. 2016). There is an accurate and prompt DNA damage repair via interaction of these pathways, thus preventing the onset of gene distortion while ensuring the integrity of the cellular genome (Song et al. 2020). High mutational load is closely associated with increased neoantigen load (NAL) as well as tumor-infiltrating lymphocytes (TILs) . With respect to biomarkers, mismatch repair defects (MMR-D), HR-gene mutations, and POLE mutations (affecting the DDR signaling pathway) play an essential role while influencing the effectiveness of immune checkpoint inhibitors (ICIs). Preliminarily, mutations in repair genes are associated with increased levels of NAL, CD4 + , and CD8 + TIL and expression of cytotoxicity-related genes PD-1 and PD-L1 (Teo et al. 2018). The estimation of gene dysregulation and heterogeneity in LGG is limited by transcriptomic and proteomic analysis, particularly for LGG. Some researchers found that altered DNA damage repair regulates microglia M2 polarization through p53-mediated MDK expression and reshapes the tumor GBM microenvironment (Meng et al. 2019). In addition, it has been shown that TGF-β/Smads signaling affects radiation response and prolongs survival in malignant gliomas by regulating DNA repair genes (Tao et al. 2018). A study by Shahmoradi Ghahe Somayeh et al. finds that enhanced DNA repair capacity enhances the resistance of Glioblastoma Cells to photodynamic therapy (Shahmoradi Ghahe et al. 2021). However, studies elucidating the DDR gene's involvement in the immune profile of LGG remain scarce.
Our aim herein is to study the altered transcriptional profile of DDR genes among LGG in a comprehensive manner. We were successful in identifying two DDR gene-based isoforms based on 276 DDR genes, and these two isoforms possess distinct clinical outcomes and molecular profiles. We have also identified six DDR-related signature genes from the genomic and transcriptional levels and confirmed their expression in tumor samples and normal brain tissue by real-time quantitative polymerase chain reaction (qPCR). In addition, we explored their prognostic significance for LGG patients and the mechanism of DNA damage repair genes mediating tumor progression and their role in immune cell infiltration as well as the tumor microenvironment, which may reveal their influence on the prognosis of LGG.

DNA Damage Repair Gene Collection and Clustering Aanalysis of Lower-Grade Glioma
On the basis of a combination of MSigDB v5.0 and knowledgebased DDR pathways, we obtained 276 DDR genes from the literature (Lin et al. 2021) (see Supplementary File 1). It is very likely that the aforementioned genes may be associated with several DDR pathways or coordinate cellular, along with molecular responses to DNA damage. First, we assessed DDR alterations and established subtypes of the DDR gene based on three LGG cohorts. Samples with repeat sequencing, no survival status, overall survival time < 1 day, no clear WHO classification, and non-primary LGG were excluded for subsequent modeling. Based on the above nadir criteria, we downloaded RNA-seq data from the TCGA database for the TCGA-LGG cohort, including 481 samples; we downloaded RNA-seq data from the Chinese Glioma Genome Atlas (CGGA) database for the CGGA-693 project (332 patients) and the CGGA325 project (162 patients), for a total of 494 samples as the CGGA-LGG cohort, and downloaded the microarray data of GSE16011 cohort from the GEO database, totaling 80 samples. For the RNA-seq data, after log(x + 1) normalization and TPM transformation, they were background corrected, normalized, and expression calculated using the combat function in the sva package and merged with the GEO cohort to remove batch effects (Fig. S1). Consensus clustering is a technique used to identify robust and stable subgroups within a dataset. It involves generating multiple random partitions of the data and calculating a consensus matrix, which represents the proportion of times two samples that are assigned to the same cluster across all partitions. The resulting consensus matrix is then used to identify subgroups based on a clustering algorithm. We performed consensus clustering of transcriptome profiles of 276 DDR (DNA damage response) genes using the CancerSubtypes package in R. The package uses a consensus clustering algorithm and provides several parameters to control the clustering process, including reps = 50, pItem = 0.8, and maxK = 9. Based on the clustering results as well as clinical ease of use, an optimum number of clusters were determined. Both training and validation cohorts underwent a similar clustering process. We further applied the Kaplan-Meier (KM) analysis using the log-rank test for comparing the difference in overall survival (OS) across two subgroups.

Identification of Clinical and Molecular Features Specific to DDR Subtypes
We observed the clinicopathological as well as molecular characteristics of various DDR subtypes and also compared these characteristics for two subgroups. The chi-square (X 2 ) test explored the distribution of clinicopathological features among various DDR subtypes. By further utilizing the TCGA database, we separately downloaded somatic mutation profiles of LGG patients and analyzed them via the "maftools" R package. We also utilized Gene Set Enrichment Analysis (GSEA), a program based on the Cluster-Profiler package of R software, to compare transcriptomic modifications across DDR-activated and DDR-suppressed subtypes.

Evaluation of Tumor Microenvironment of Lower-Grade Glioma
Here, we further performed the previously proposed metagenomic approach for assessing twenty-eight immune cell subpopulations for LGG-TME. The relative infiltration scores of 28 immune cell subpopulations were estimated using the gene set variance analysis (GSVA) algorithm, whereas metagenes for these subpopulations were retrieved utilizing a previous study. Wilcoxon tests have been applied then for estimating differences in immune characteristics across subtypes.

Development and Validation of DDR Subtype Signatures
Considering that there are too many genes, detection is not easy as per clinical applications are concerned. Therefore, for identifying the DDR subtype, we have established a gene signature. We applied Wilcoxon analysis for the identification of differentially expressed genes among DDR activating as well as repressing subtypes where log2 (fold change) > 1 and p value (P < 0.05) represented DDR subtype-specific genes. Considering that DDR-related proteins are the main performers, we further compared the relationship between transcriptome and proteome levels. A total of 5413 protein expression levels were included in the CPTAC cohort, and then we performed Spearman correlation analysis for the identification of paired transcriptomic as well as proteomic data. DDR subtype signature construction was applied to genes that exhibited a considerable correlation (Spearman correlation coefficient > 0.3) between protein and mRNA levels.

Identification of the Prognostic Value for DDR Subtype Characteristics
To determine the effectiveness of DDR subtype characteristics in predicting survival, a univariate Cox analysis was performed in each cohort revealing its relationship with OS. We also computed the hazard ratios (HR) and corresponding 95% correspondence intervals (CI). The results of the survival analysis were then integrated using a meta package. Heterogeneity was analyzed using I2 and Q tests where heterogeneity was considered to exist when I 2 was greater than 50%, Q-test P was less than 0.1, and a random-effects model was selected.

Characterization of DDR Subtypes for Predicting Immunotherapy Response
We performed an analysis of the relationship among DDR characteristics and immunotherapy responses through the IMvigor210 cohort, whereas the Kruskal-Wallis test was utilized for exploring differences in DDR characteristics scores between different immunotherapy response groups for complete response (CR), partial response (PR), stable disease (SD), and disease progression (PD). DDR characteristics of immunotherapy response were estimated using the area under the curve (AUC).

Drug Sensitivity Analysis
The chemotherapeutic drugs were retrieved from the Genome of Drug Sensitivity in Cancer (GDSC) database, while the calculation of IC50 was performed utilizing the prophetic package of R software.

Molecular Docking
Virtual screening of molecular docking of DDR signature genes using AutoDock Vina 1.1.2 to predict the most likely best-fit ligands. Constituent structures were obtained from Pubchem. Molecular optimization was performed using Syby-X software with the following optimization parameters set: Tripos force field was used, Gasteiger-Hückel charge was selected, the maximum iteration factor was set to 10000, and the energy gradient was limited to 0.005 kcal/(mol-A), and all unspecified parameters were set to default values. The three-dimensional structure of the receptor was obtained using the RCSB database (http:// www. rcsb. org/). Ligands and receptors were repaired by mgltools_win32_1.5.6 software and saved as PDBQT files, and information on receptors is detailed in Supplementary File 2. AutoDock Vina 1.1.2 software (http:// vina. scrip ps. edu/) was used to test key active ingredients and target affinity for docking between the proteins. We visualized the final graphical molecular docking data via PyMOL 2.3.

RNA Extraction and qRT-PCR
From January 2022 to June 2022, 10 lower-grade glioma tissues and 10 normal brain tissues have been obtained from 20 patients who had undergone surgical dissection as well as pathological confirmation at the First Affiliated Hospital of Xinjiang Medical University. The Medical Research Ethics Committee of the First Affiliated Hospital of Xinjiang Medical University approved the study. Total RNA was extracted by RNA reagent (Servicebio), whereas the total RNA concentration was derived utilizing Nano Drop 2000 (Thermo Fisher Scientific, USA). A two-step reaction process involving reverse transcription (RT) and polymerase chain reaction (PCR) was carried out for quantifying the mRNA levels. We carried out cDNA synthesis with the aid of the Servicebio RT First Strand cDNA Synthesis Kit (Wuhan servicebio Technology CO., LTD, Wuhan, China). We applied qRT-PCR for determining the expression levels of GAPDH, TRIP13, CDK1, CDK2, TYMS, SMC4, and WEE1 while using SYBR Green qPCR Master Mix (High ROX) (Servicebio, Wuhan, China) to detect them. The results were expressed as GAPDH. The following PCR primer sequences were designed and synthesized by Servicebio (Wuhan) Co.

Estimation of DDR Heterogeneity via Single-Cell Analysis
Based on the TISCH database, single-cell transcriptome data of a total of 3533 cells were used for the analysis of the study after filtering out low-quality cells from the GSE84456 dataset. To explore the heterogeneity of DDR features among various cell types, we evaluated the DDR features for each cell and also compared the differences between them.

Statistical Analysis
All statistical data were analyzed by R software and Perl. The chi-square test or Fisher test was used for categorical variables, and the Wilcoxon test was used for continuous data. We utilized K-M and log-rank tests for assessing survival differences. Further, the analysis of differential expression was carried out for LGG samples from TCGA (N = 518), along with normal samples from matched TCGA normal and genotype-tissue expression (GTEx) data (N = 207) via utilizing Gene Expression Profile Interaction Analysis (GEPIA) (http:// gepia. cancer-pku. cn/). A p < 0.05 was statistically significant.

Identification of DDR Gene Clustering Subtypes in Lower-Grade Glioma
On the basis of 276 DDR gene expression profiles, all LGG patients were classified into heterogeneous subtypes for revealing the DDR gene heterogeneity among LGG. Two DDR subtypes were identified following consideration of consensus clustering results as well as clinical significance. In the TCGA cohort (RNA-seq), subtype B (n = 150) was assigned as the DDR-activated subtype due to the upregulation of most DDR-related genes in subtype B; subtype A (n = 331) was assigned as the DDR-suppressed subtype (Fig. 1A). KM plots showed that patients classified as DDR-activated subgroups had poorer OS (Fig. 1D). In the validation CGGA cohort (RNA-seq), all 494 LGG patients were classified into varied subtypes based on 276 DDR gene expression, where K-means clustering showed that patients being classified into two subgroups also possess similar DDR pathway alterations as to the training cohort ( Fig. 1B), with similarly poor survival in the DDR activation subgroup (Fig. 1E). In the validation GEO cohort (Microarray), 80 LGG patients were similarly divided into different subtypes where K-means clustering showed that patients being classified into two subgroups also possess similar DDR pathway alterations as of the training cohort (Fig. 1C), with survival differences consistent with the two data sets described above (Fig. 1F).

Distinct Clinical and Molecular Profile Revealed via DDR Gene-Based Subtypes
A comparison of clinical parameters among the two groups revealed that advanced age and high grade were more common in the subgroup with DDR activation (Fig. 2A). The difference in gene mutations among the two DDR subtypes was also taken into account when comparing genomic differences (variations). Patients in the TCGA cohort with different subtypes showed significant differences in the frequency of mutations in IDH1, CIC, EGFR, PTEN, and LRP2 genes (Fig. 2B). IDH1 and TP53 were the most frequently mutated genes in patients from the TCGA cohort (Fig. 2C). Considering their significance, we have compared them and found that IDH mutations were more common in DDR suppressor subtypes (83.69% vs 60.67%) (Fig. 2D). At the same time, mutations in TP53 were not significantly different between the two groups (Fig. 2E). GSEA analysis revealed that the top 5 most activated GO terms in the DDR activation subgroup were adaptive immune response, blood vessel morphogenesis, cell-cycle checkpoints, cellcycle DNA replication, and cell cycle G1/S phase transition (Fig. 3A). In the DDR-activated subgroup, the top 5 most active Kyoto genes and genome encyclopedia terms were cell cycle, ECM-receptor interaction, focal adhesion, p53 signaling pathway, and pathways in cancer (Fig. 3B).

Identification of Varied Immune Profiles Among DDR Subtypes
As the role of immune cell infiltration in tumor progression and immunotherapeutic response is significant, this study aimed to explore the variation of this event between the two DDR subtypes. Specifically, the authors examined the activation of CD4 + T cells, B cells, CD8 + T cells, DC cells, and natural killer cells and observed significant upregulation of these cells in the DDR activation subtype, as depicted in Fig. 4A, which was consistent with the findings of the CGGA training set, as presented in Fig. 4C. The differential analysis of immune-related pathways revealed that multiple pathways were considerably activated in the DDR activation subtype, as indicated in Fig. 4B and D. Additionally, the microarray cohort showed similar modified immune profiles among different subtypes, as demonstrated in Fig. S2.

DDR Signature as a Potential Indicator of Immunotherapy for LGG
The IMvigor210 cohort was utilized for response-related gene expression analyses to investigate the DDR subtype profile for predicting immunotherapy response. The authors observed fewer complete response/partial response (CR/PR) groups in patients with high scores, as shown in Fig. 7A, but higher DDR signature scores when compared to the stable disease/progressive disease (SD/PD) group, as demonstrated in Fig. 7B. The results from the ROC curve indicated that the DDR signature could be used to predict immunotherapy response with an area under the curve (AUC) of 0.673, as presented in Fig. 7C. Additionally, patients with higher scores in the immunotherapy cohort had shorter survival times, as depicted in Fig. 7D. Despite reports in the literature that low-grade gliomas (LGG) are insensitive to chemotherapy, the authors explored whether DDR score-based grouping could be indicative of conventional cytotoxic drugs. To this end, they calculated the IC50 values for various drugs using the prophetic package and found that the DDR score grouping was sensitive to several chemotherapeutic drugs, including A-443654, A-770041, acadesine, benzamide, motesanib diphosphate, navitoclax, ponatinib, rucaparib phosphate, saracatinib, tretinoin, and veliparib dihydrochloride, as shown in Fig. 8A-K.

Screening the DDR Score Genes and Investigation of Best-Fitting Compounds on DDR Score Genes
To investigate the most suitable compounds, we performed a virtual screen for molecular docking of these six DDR genes using AutoDock Vina 1.1.2 (Fig. 9A-F). The different drugs The waterfall plot displays the information regarding gene mutations among each sample, while individual tumor mutation load has been indicated by the top panel. The TCGA data site was used to assess the data shown. D The DDR-suppressed subgroup had a greater incidence of IDH1 mutations, E while TP53 mutations were not varied considerably across DDR-activated and DDR-suppressed subgroups ◂ 1 3 Fig. 3 Pathway study of the DDR subtype utilizing GSEA. A Comparison of the top five most remarkably altered gene ontology entities between the DDR-activated and DDR-suppressed subgroups. B The top five most important KEGG pathways were altered between the DDR-activated and DDR-suppressed subgroups had different affinities for the key targets, and A-770041 and ponatinib had a strong binding ability to each target. Each drug had a stronger affinity for WEE1, TYMS, and CDK2. A-443654, saracatinib, motesanib diphosphate, ponatinib, navitoclax, tretinoin, A-770041, and rucaparib phosphate had the stronger binding ability (affinity > − 7 kcal/mol) with all 6 targets (Table 1). Navitoclax binds within the binding pocket of CDK1 and forms hydrogen bonds with ARG145, GLN31, ILE10, and LYS89. A-443654 binds within the binding pocket of CDK2 and forms hydrogen bonds with ASP145, THR14, and LYS129. A-770041 binds within the binding pocket of SMC4 within the binding pocket of SMC4, forming hydrogen bonds with ASN1509, ASN1508, and LYS779. Ponatinib is bound within the binding pocket of TRIP13, forming hydrogen bonds with ARG389. Ponatinib is bound within the binding pocket of TYMS, forming hydrogen bonds with ILE101. A-770041 is bound within the binding pocket of WEE1, forming hydrogen bonds with SER383, and LYS331 forms a hydrogen bond (Table 1).

Expression Validation of the Six DDR-Scoring Genes by qRT-PCR
Six genes were analyzed utilizing GEPIA and GTEx databases to confirm the expression profiles of DDR-related genes, whereas 207 normal brain tissue samples as well as 518 LGG samples were identified based on TCGA. The findings indicate that except for TRIP13, expression levels for the remaining five genes (CDK1, CDK2, TYMS, SMC4, and WEE1) were significantly higher in lower-grade glioma tissues than in normal brain tissues (p < 0.05) (Fig. 10A-F). Further, to determine the survival rates of TRIP13, CDK1, CDK2, TYMS, SMC4, and WEE1, we performed a survival analysis using the GEPIA database. As shown in Fig. 10G-R, DDR-related genes displayed low expression among LGG patients, which were also statistically considerably varied from OS and DFS (p < 0.05). Finally, 10 samples of normal brain tissue and 10 samples of LGG were collected to thoroughly characterize DDR-related gene expression levels among these tissue samples. As shown in Fig. 10S, the expression levels of CDK1, CDK2, TYMS, SMC4, and WEE1 were found considerably higher for LGG samples compared to normal brain tissue samples (p < 0.05), which is consistent with the results shown by GEPIA. Fig. 4 Differences in immune profile alterations and immune-related pathways among DDR-activated and DDR-suppressed subgroups. A-B The TCGA cohort; C-D CGGA cohort, where * presents P < 0.05, ** presents P < 0.01, *** presents P < 0.001, and empty for no considerable difference database ( Fig. 11A-B). The results of the single-cell analysis revealed significant differences in the distribution of DDR feature scores in different clusters (Fig. 11C). The DDR scores were found considerably upregulated among some particular clusters, especially for tumor cells (Fig. 11D).

Discussion
Imbalance or defects in DNA damage repair can affect the production of new tumor antigens and alter the immune microenvironment, thereby affecting tumor immunotherapy (Abou Khouzam et al. 2020). Delayed or incorrect repair of DNA damage can lead to alterations in the tumor genome, thereby changing the immune homeostasis in the tumor microenvironment (Luo et al. 2021). The interaction between glioma and the host immune system is being studied, and therapeutic attempts to activate the host immune system to kill tumor cells have shown some clinical efficacy (Li et al. 2021b). Despite advances in therapeutic approaches for lower-grade gliomas, the prognosis of LGG remains poor, even after surgical resection, due to the susceptibility to recurrence or transformation to HGGs. There is regular inactivation of certain DDR pathways during LGG onset and advancement, whereas mutations between DDR genes are also associated with the chemoresistance of tumor cells (Rocha et al. 2018). Therefore, DDR genes are associated with prognosis and could be utilized in predicting treatment response as well as overall prognosis for cancer patients.
Identification of molecular phenotypic subtypes is critical for improving therapeutic options and survival monitoring in the heterogeneous group of lower-grade gliomas (LGG). However, the role of DNA damage response (DDR) in LGG is not yet fully understood. To better characterize DDR-based subtypes in LGG, we conducted a comprehensive analysis of multi-omics data, including genomics, transcriptomics, and proteomics. In addition, we investigated the differences in immunotherapeutic response and immune profiles among DDR-based subtypes. Our study identified two distinct DDR states in LGG: DDR-activated and DDR-suppressed subtypes. Notably, patients in the DDR-activated subgroup exhibited aggressive clinical behavior and poor prognoses. Our findings highlight the importance of DDR status as a potential therapeutic target and prognostic marker in LGG. To describe a molecular profile for the different DDR subtypes among LGG, there were considerable genomic alterations for two subtypes. IDH1 has been found to undergo mutation more often in the DDR suppressor subtype, while mutations in TP53 were not significantly different between the two groups. Mutations in the IDH1 gene identified glioma subtypes with various biological, clinical, and radiological features (Hartmann et al. 2009;Patel et al. 2017;Yan et al. 2009).
LGG develops through early mutations in IDH1, leading to the accumulation of 2-hydroxyglutaric acid and genomewide DNA damage repair, followed by the acquisition of two sets of co-occurring genetic alterations: mutations in TP53 and ATRX or 1p/19q symbiosis and mutations in TERT, CIC, and FUBP1 (Bettegowda et al. 2011;Jiao et al. 2012;Killela et al. 2014). However, determining whether the genetic alterations that drive the development and progression of IDH mutant gliomas in terms of whether they are derived from DNA damage repair is worth investigating. Due to the challenges encountered in directly targeting alterations driving IDH mutant gliomagenesis, future studies could focus on selectively targeting IDH mutant gliomas for immunotherapy and synthetic lethality in the DDR pathway.
Interestingly, among the DDR-suppressed subgroup, CIC is the second, while EGFR is the third most frequently mutated in comparison to the DDR-activated subgroup. The protein encoded by CIC is a homolog of the Drosophila melanogaster tube Capicua gene, which is a member of the high mobility group (HMG) box superfamily of transcriptional repressors. This protein consists of a conserved HMG structural domain for DNA binding, nuclear localization, and a conserved C-terminal (Bettegowda et al. 2011). Some reports suggest that CIC can widely activate gene expression through the independent EGFR pathway. CIC loss enhances tumor formation and reduces the latency of tumor development (Yang et al. 2017). This study identifies an essential role of CIC in the regulation of neuronal cell proliferation and spectrum specification. It also shows that CIC mutations influence the pathogenesis of oligodendrogliomas and strategies for associated targeted therapies.
We have also explored the role of DDR in the LGG immune microenvironment, where different immunological profiles exist for DDR subtypes. Activated (CD4 T cells, B cells, CD8 T cells, and DC cells) and natural killer cells were observed considerably upregulated in the DDR activation subgroup. Researchers demonstrated that KML001 (a telomere-targeting drug) may inhibit cell proliferation and cytokine production and promote apoptosis via disrupting telomere integrity and DNA repair machinery. As a result of treatment with KML001, dysfunctional telomere-induced Fig. 7 A LGG patient's proportion for CR/PR and SD/PD across high-and low-risk DDR scores. B DDR score comparison among patients making CR/PR with those keeping an SD/PD. C A ROC curve was utilized for measuring DDR subtype signature performance in predicting immunotherapy outcomes. D Analysis of K-M curve across high-and low-risk DDR scores in the IMvigor210 cohort foci (TIF), DNA damage marker H2AX, and topoisomerase cleavage complex (TOPcc) accumulation increased, leading to the attrition of telomeres (Cao et al. 2019). According to Zhao et al., during HCV infection, inadequate ATM (DNA repair enzyme) results in increased damage to DNA. Consequently, HCV T cells are rendered prone to apoptosis which also contributes to the loss or dysregulation of naive T cells (Zhao et al. 2018). Moreover, selective induction of DNA repair pathways in human B cells activated by CD4 + T cells is also reported (Wu et al. 2010). The study by Galgano Alessia et al. shows that CD8 T cells engage unique DDRs during exponential division that is not observed in other exponentially dividing cells, in T lymphocytes after UV or X irradiation, or in non-metastatic tumor cells. When CD8 T cells divide, all DDR pathways as well as cell cycle checkpoints got affected, whereas only one DDR pathway is affected for other cell types (Galgano et al. 2015). These results illustrate the role of CD8 T cells in maintaining genome integrity despite their extensive division which highlights DDR's fundamental role in the efficiency of CD8 immune responses. Karo Jenny et al. reported that NK cells incapable of expressing either RAGs or RAG endonuclease activity during ontogeny display cell-intrinsic hyperresponsiveness but are less capable of surviving after virus-driven proliferation, reduced DNA damage response mediators expression, and defects in repairing DNA breaks (Karo et al. 2014). These findings suggest that DDR subtypes possess unique differences in terms of the infiltration of immune cells, implying varied immunotherapeutic responses among subtypes.
We adjusted mRNA levels according to DDR subtypes, classified them into high and low expression groups based on median mRNA expression, and constructed DDR Score. We explored the gene expression analysis from the IMvigor210 cohort for response-related genes and found that patients with high DDR score scores in the CR (complete response)/PR (partial response) group were fewer and the CR/PR group had higher DDR signature scores compared to the SD (stable disease)/PD (progressive disease) group. We combined the six DDR genomes as a DDR subtype identification signature that showed excellent performance in training and validating cohorts to classify patients into different DDR subtypes. Further, we were able to identify 11 drugs possessing considerable sensitivity in the prognostic model Numerous studies have identified that DDR hub genes (TRIP13, CDK1, CDK2, TYMS, SMC4, and WEE1) play an essential role in tumor progression and metastasis. Thyroid hormone receptor-interacting factor 13 (TRIP13) is an important regulator of spindle assembly checkpoint and double-strand break repair. As per previous reports, TRIP13 is identified as aberrantly expressed among several human cancers, whereas TRIP13 knockdown inhibits cell proliferation, induces cell cycle arrest and promotes apoptosis, impairs cell viability, and finally interferes with the growth of tumor xenografts. In addition, TRIP13 can directly bind to the epithelial growth factor receptor (EGFR) and regulate the EGFR signaling pathway (Gao et al. 2019). Recent studies have indicated that high expression of TRIP13 promoted the proliferation and migration of ESCC cells and induced NDP resistance via enhancing the repair of DNA damage and inhibiting apoptosis (Zhang et al. 2022). Additionally, TRIP13 overexpression restored the impacts of miR-129-5p overexpression on malignant cell phenotypes and cell cycle. MiR-129-5p down-regulated TRIP13 expression, thereby restraining the malignant progression of CRC cells (Cao et al. 2022). Cycle protein-dependent kinase (CDK) is an essential member of the protein kinase family, and among them, CDK1 and CDK2 play critical roles in cell cycle regulation, checkpoint activation, and DNA damage repair (Li et al. 2021a). In human cells, there are mainly the following CDKs that regulate the cell cycle by binding to the corresponding cyclin: (1) CDK2 promotes G1/S transition by binding to cyclinE and promotes replication initiation and S phase by binding to cyclinA, and (2) CDK1 interacts with cyclinB, whose expression in G2/M phase increases gradually and reaches the peak, to activate cells to enter M phase, maintain M phase, ensure normal mitosis, and prevent cells from entering G1 phase ahead of time (Liu et al. 2020;Tadesse et al. 2020). In vitro studies have shown that CDK2 is responsible for promoting the G1/S transition and DNA replication initiation in normal cells. When CDK2 is absent, CDK1 can completely compensate for the function of CDK2 by promoting cell entry into S-phase through binding to cyclinE and DNA replication initiation through binding to cyclinA (Roskoski 2019). In addition, CDK1 can fully compensate for the absence of CDK2 in cell cycle regulation and has many other vital functions that cannot be replaced by CDK2, such as CDK1 capacity of promoting DNA double-strand break repair by Fig. 9 A Combination pattern diagram of navitoclax and CDK1. Yellow represents hydrogen bonding; amino acid residue includes ARG45, GLN31, ILE10, and LYS89. B Combination pattern diagram of A-443654 and CDK2. Yellow represents hydrogen bonding; amino acid residue includes ASP145, LYS129, and THR14. C Combination pattern diagram of A-770041 and SMC4. Yellow represents hydrogen bonding; amino acid residue includes ASN1509, ASN1508, and LYS776. D Combination pattern diagram of ponatinib and TRIP13. Yellow represents hydrogen bonding; amino acid residue includes ARG389. E Combination pattern diagram of ponatinib and TYMS. Yellow represents hydrogen bonding; amino acid residue includes ILE101. F Combination pattern diagram of A-770041 and WEE1. Yellow represents hydrogen bonding; amino acid residue includes LYS331 and SER383  . 10 Analysis of DDR score gene expression. A-F An expression profile of six genes (TRIP13, CDK1, CDK2, TYMS, SMC4, and WEE1) between TCGA (518 LGG samples) and GTEx (207 normal brain samples) cohorts compared via GEPIA. G-R K-M survival curves evaluating OS as well as disease-free survival for six genes. S Bar plots representing the expression of six genes between LGG and normal brain samples determined via qRT-PCR (*p < 0.05) homologous recombination and cell cycle checkpoint activation through phosphorylation of BRCA1 (Kalra et al. 2017; Leal-Esteban and Fajas 2020). Thus, CDK1 and CDK2 are central to regulating many biological processes, including cell cycle regulation, DNA replication, and DNA damage repair, and closely link these biological processes to the cell cycle process. Further, the downstream target of FOXM1 is TYMS. TYMS knockdown reversed the 5-FU resistance caused by FOXM1 overexpression and re-sensitized HCC cells to 5-FU treatment. This reveals that TYMS functions as an oncogene in HCC and that inhibiting the FOXM1-TYMS axis may increase patient survival and offer novel treatment options for individuals with advanced HCC ). Li et al. proved that SMC4 which has the highest positive regression coefficient in glioma risk model is strongly linked with malignant progression and TMZ resistance of gliomas in an E2F1-dependent manner (Li et al. 2022). Lal Shruti et al. demonstrated that WEE1 inhibition in pancreatic cancer cells is dependent on DNA repair status in a context-dependent manner (Lal et al. 2016). Finally, we applied qRT-PCR for measuring the expression levels of six DDR genes. The results showed that our samples shared the same expression pattern as the ones from the public database.
While our study provides valuable insights into the identification of DDR subtypes in different cohorts of patients with lower-grade gliomas, it is important to acknowledge its limitations. Firstly, further validation of the optimal cutoff values for identifying DDR subtypes may be required using additional expression detection platforms. Secondly, our study mainly relied on multiple cohort data to provide robust survival information and molecular characterization. Future research may focus on in vivo and in vitro mechanisms to gain a better understanding of DDR subtype alterations. As such, our follow-up research will aim to confirm the clinical relevance and molecular mechanisms of our findings.

Conclusions
In summary, our research shows that DDR heterogeneity and DDR classification subtypes exist among LGG patients. The distinct DDR subtypes are prognostically relevant mechanisms in LGG and may contribute to the development of highly effective therapeutic targets and biomarkers for immunotherapy of LGG patients.