Introduction
Prostate cancer ranks as the second most diagnosed cancer among men worldwide, with approximately 1.4 million new cases in 2020, accounting for 7.3% of all cancer diagnoses globally. It also has a significant mortality impact, with around 375,000 deaths reported in the same year, representing 3.8% of cancer-related fatalities.1,2 Despite advancements in diagnostic and therapeutic strategies, the molecular heterogeneity of the disease remains a major challenge in identifying reliable biomarkers and effective therapeutic targets.3
Proteogenomic analysis offers a vital link between genomic mutations and their effects on protein function, playing a crucial role in cancer research.4 Integrating proteomic and genomic data enables the identification of protein signatures associated with specific mutations or genetic alterations, which can serve as biomarkers for disease diagnosis and prognosis.5 Many studies primarily rely on transcriptomic data to infer protein expression; however, this approach is limited, as only ∼10% of the variability in protein abundance can be explained by messenger RNA levels.6 This highlights the need for methodologies that directly assess protein levels and correlate these changes with genetic mutations to ensure greater accuracy in identifying functional biomarkers and potential therapeutic targets.
Among these alterations, missense mutations hold particular significance, as they can compromise protein stability, functionality, and abundance—often impacting molecular interactions and expression levels in cancer-related genes.7,8 Recent studies indicate that missense mutations are the most frequent type of genetic alteration in prostate cancer, with highly mutated genes such as TP53, TTN, and SPOP potentially influencing tumor progression and immune infiltration.9 Mutations in GGAP2 have also been detected in approximately 50% of prostate cancer tissues analyzed, further supporting their role in tumorigenesis.10 Additionally, germline missense variants in BTNL2 have been associated with hereditary prostate cancer, reinforcing the genetic predisposition aspect of the disease.11 Given their functional impact, further investigation is needed to elucidate how missense mutations contribute to proteomic alterations in prostate cancer.
This study investigated the impact of missense mutations on protein abundance in healthy and cancerous prostate tissues. Using stable isotope labeling by amino acids in cell culture (SILAC), we aimed to quantify and compare protein expression levels across these conditions. Our objective was to identify differentially expressed proteins affected by missense mutations and distinguish them from alterations resulting from secondary tumor-related processes. Additionally, we explored how these changes might influence tumor progression, metastatic potential, and treatment resistance, providing insights into the molecular mechanisms driving prostate cancer development. By addressing these aspects, this study contributes to a more refined understanding of genome-proteome interactions, offering new perspectives for precision oncology and the development of more effective therapeutic strategies.
Materials and methods
Databases
The RefSeq database from the National Center for Biotechnology Information (NCBI) was employed to obtain reference (non-mutated) proteins due to its rigorous curation, extensive coverage, and high reliability, making it a fundamental resource for high-precision comparative analyses.12 To investigate missense mutations associated with prostate cancer, the dbPepVar web portal was utilized.13 This platform hosts a proteogenomic database that integrates genetic variation data from dbSNP with protein sequences from NCBI’s RefSeq. Furthermore, it provides comprehensive information on various mutation types identified in prostate cancer, including SNP IDs, protein alterations, and corresponding gene annotations. This integration enabled precise mapping of mutations to affected proteins, facilitating the distinction between tumor-associated variants and reference proteins, thereby supporting an in-depth proteogenomic analysis.
Sample collection and preprocessing
Mass spectrometry data for prostate cancer samples were obtained from the study by Iglesias-Gato et al.,14 which performed quantitative proteomic profiling using an SILAC-based approach. The study analyzed a total of 28 prostate tumor samples (Gleason score 6–9) and eight adjacent nonmalignant tissue samples, all derived from formalin-fixed paraffin-embedded radical prostatectomy specimens. The SILAC technique was employed to quantify proteins in both healthy and tumor tissue samples. Proteins extracted from cancerous tissues were labeled with the “L” isotope, while proteins from healthy tissues were labeled with the “H” isotope. This method allowed for direct comparisons between conditions, highlighting differences in protein abundance between healthy and tumor tissues. To analyze mass spectra and identify proteins in healthy and cancerous tissues, two distinct datasets were used. The first dataset corresponds to the process of searching mass spectra against a reference database from NCBI, while the second dataset was obtained from the dbPepVar portal and contains information on identified mutated peptides categorized by mutation type.
Filtering reference peptides
The first step in the methodology involved filtering the file containing peptides identified in the reference protein database (evidence_refseq.txt). Only records with a non-zero Intensity H value were retained, indicating the identification of a protein in healthy tissue.
Filtering of mutated peptides
Similarly, filtering was performed on the file containing mutated peptides from dbPepVar (evidence.dbPepVar.PrCa.txt). Only records with a nonzero Intensity L value were retained, indicating the presence of the protein in cancerous tissue. Additionally, only missense mutations were considered for subsequent analysis.
Data integration
After filtering, a new file was generated containing combined information from both mutated and reference peptides. The new file included the following fields: Sequence_dbsnp, Sequence_refSeq, SNP ID, Variation Type, Intensity L (extracted from evidence.dbPepVar.PrCa.txt), Intensity H (extracted from evidence_refseq.txt), Raw File, Experiment (See Supplementary File).
Validation of Peptide Correspondence
To ensure correct pairing of mutated and non-mutated peptides, the presence of both reference and mutated sequences was verified in evidence_refseq.txt and evidence.dbPepVar.PrCa.txt, respectively. Correspondence was established using Raw File and Leading Razor Protein as keys. This step ensured that comparisons were conducted for the same protein and sample, maintaining the consistency of the analyzed data.
This methodological approach enables a robust analysis of protein presence in healthy and cancerous tissues, contributing to the identification of potential clinically relevant biomarkers.
Protein abundance distribution
For gene set selection in enrichment analysis, protein abundance was determined using the Intensity H/Intensity L ratio. The objective was to assess whether non-mutated genes exhibit higher abundance in healthy tissue samples (Intensity H) compared to mutated genes in cancerous samples (Intensity L). Values > 1 were classified as RefSeq Abundant. Values < 1 were classified as Variant Abundant. Only non-zero values for Intensity H and Intensity L were considered.
Correlation analysis between missense mutations and protein abundance alterations
To assess the impact of missense mutations on protein expression, a correlation analysis was conducted to compare protein abundance in paired healthy and tumor tissue samples. Proteomic data obtained through mass spectrometry were used to quantify protein intensities in healthy (Intensity H) and tumor (Intensity L) tissues, enabling the calculation of key metrics for evaluating mutation-associated alterations.
Calculation of variables
Each identified protein was quantified in both tumor and healthy tissues from the same patient, allowing for direct comparison between its mutated and non-mutated forms. The following metrics were employed:
Protein intensity difference:
Δ=IL−IH where IL represents protein intensity in tumor tissue and IH represents intensity in healthy tissue. Negative values indicate a reduction in protein abundance associated with the missense mutation in the tumor.
Protein intensity ratio:
R=IHILValues R > 1 indicate higher protein abundance in healthy tissue, whereas R < 1 indicates higher abundance in tumor tissue.
Statistical analysis
To determine the relationship between missense mutations and changes in protein expression, we applied Spearman’s rank correlation between ΔI and R. The analysis was conducted separately for two groups based on protein abundance classification:
Refseq abundant: Proteins predominantly expressed in healthy tissue.
Variant abundant: Proteins predominantly expressed in tumor tissue.
The Spearman correlation coefficient (Rs) and the corresponding p-value (p) were calculated for each group. Statistical significance was considered at p < 0.05.
Functional enrichment analysis
Functional enrichment analysis was conducted using Gene Ontology (GO) terms to identify biological processes associated with RefSeq Abundant and Variant Abundant proteins. This aimed to elucidate the functional implications of differentially abundant proteins in healthy and cancerous tissue samples. The analysis was performed using the gseapy library with the Enrichr method, employing the GO_Biological_Process_2018 gene set.
Separate analyses were carried out for RefSeq Abundant and Variant Abundant genes, applying an adjusted p-value threshold of <0.1 to identify significantly enriched biological processes. Given the exploratory nature of the study, this threshold was chosen to reduce the likelihood of false negatives and to capture biologically relevant pathways that might otherwise be overlooked under more stringent significance criteria.15
Data processing was performed using the pandas library, while visualization was conducted with matplotlib. To enhance interpretability, results were represented as scatter plots, with enriched biological processes plotted according to classification categories (RefSeq Abundant vs. Variant Abundant). Dot size was proportional to the number of overlapping genes in each biological process, and a color gradient was used to indicate statistical significance, with darker shades representing lower adjusted p-values.
Mutation hotspot detection and classification of cancer driver genes
Mutation hotspot detection was performed using Oncodrive, which identified genomic regions with a high concentration of mutations. These hotspots were further analyzed to assess their functional impact and correlation with significant changes in protein abundance in tumor tissues.
To detect cancer driver genes based on positional clustering, the maftools package was used, specifically its oncodrive function. This method is based on the OncodriveCLUST algorithm,16 originally implemented in Python, which identifies cancer driver genes from a given MAF (mutation annotation format) file. The approach relies on the concept that most mutations in cancer-causing genes tend to cluster at specific loci, known as hotspots. By leveraging this positional enrichment, the method enhances the identification of key oncogenic genes.
To classify mutations based on their functional impact, the Protein Variation Effect Analyzer (PROVEAN) tool was employed to evaluate missense mutations, categorizing them as either neutral or deleterious. This classification enabled the prioritization of mutations with a higher likelihood of affecting protein functionality or stability, which are critical factors in tumor development.
Results
Protein abundance distribution
Protein abundance was assessed using the Intensity H/Intensity L ratio, comparing non-mutated genes in healthy tissues (Intensity H) with mutated genes in cancerous samples (Intensity L). Proteins were classified as RefSeq Abundant (ratio > 1) or Variant Abundant (ratio < 1), considering only non-zero intensity values.
Figure 1 presents the logarithmic distributions of L and H intensities, corresponding to proteins identified from the dbPepVar database (blue) and the RefSeq database (red), respectively. The L intensity distribution exhibits a pronounced peak at mid-log intensity values, suggesting a higher concentration of proteins within this range. In contrast, the H intensity distribution appears more dispersed, indicating a broader and more uniform protein abundance.
The overlap observed between the two distributions in specific intensity ranges suggests regions of shared protein expression across datasets. However, the distinct density curve profiles indicate differential protein abundance patterns, likely influenced by missense mutations in cancerous samples. These findings provide insights into the impact of genetic variations on protein expression and contribute to a comparative understanding of protein abundance in healthy versus cancerous tissues.
Correlation results
The results revealed a strong and statistically significant negative correlation between protein intensity difference and protein intensity ratio in both groups (Fig. 2). Proteins classified as Refseq Abundant exhibited a stronger correlation (Rs = −0.673, p = 1.71 × 10−46), while Variant Abundant proteins showed a moderate correlation (Rs = −0.522, p = 3.46 × 10−18).
The correlation analysis was further refined by investigating individual genes with high occurrence in the dataset. The results demonstrated that all five most frequent genes exhibited strong negative correlations, with HSPD1 and GAPDH showing the most significant reductions in tumor protein abundance (Table 1).
Table 1Correlation between protein abundance and gene occurrence
| Gene | Sperman Rs | p-value | Interpretation |
|---|
| HSPD1 | −0.954 | 1.52 × 10−30 | Very strong negative correlation |
| ACTB | −0.891 | 1.68 × 10−12 | Strong negative correlation |
| PPIF | −0.906 | 1.03 × 10−12 | Strong negative correlation |
| CSRP1 | −0.860 | 5.64 × 10−10 | Strong negative correlation |
| GAPDH | −0.956 | 2.60 × 10−14 | Very strong negative correlation |
These findings strongly suggest that missense mutations lead to a significant reduction in protein expression in tumor tissues, potentially due to protein degradation, structural instability, or negative regulatory mechanisms affecting transcription and translation efficiency. The results emphasize the role of proteomic dysregulation in cancer progression and reinforce the importance of studying post-translational regulation in missense-mutated proteins.
Functional enrichment analysis
Figure 3 presents the functional enrichment analysis aimed at identifying biological processes and molecular functions overrepresented in two distinct protein groups. The RefSeq Abundant group was primarily associated with metabolic and structural processes characteristic of healthy tissues, whereas the Variant Abundant group was enriched for biological pathways typically linked to oncogenic processes in cancerous tissues. This analysis provides insights into the functional distinctions between these protein groups, contributing to the understanding of their roles in physiological and pathological conditions.
The functional enrichment analysis compared two protein groups: RefSeq Abundant, predominantly composed of proteins commonly found in healthy tissues, and Variant Abundant, where shifts in protein abundance are associated with oncogenesis-related processes. Importantly, all data used in this analysis originate from proteomic mass spectrometry, meaning that the identified proteins were experimentally detected in tumor samples rather than being merely theoretical predictions.
In the RefSeq Abundant group, essential metabolic processes such as glucose metabolism (GO:0006006), hexose biosynthetic process (GO:0019319), and aerobic cellular respiration (GO:0009060) were identified. Proteins such as PGK1 and MDH2 play critical roles in these pathways, promoting cellular energy production and maintaining homeostasis. For instance, PGK1 is involved in mitochondrial metabolism, DNA replication, and repair, in addition to its central role in glucose metabolism. Furthermore, protein stabilization processes (GO:0050821), such as the response to unfolded proteins (GO:0006986), highlight the importance of heat shock proteins (HSPA9, HSPA4, and HSPB1) in protecting and maintaining protein integrity, particularly under cellular stress conditions.17 Other identified processes include homotypic cell-cell adhesion (GO:0034109) and platelet aggregation (GO:0070527), which are essential for maintaining tissue integrity and regulating the cellular microenvironment.18
Conversely, the Variant Abundant group exhibited a predominance of cancer-associated processes, suggesting that missense mutations contribute to changes in protein abundance patterns that favor oncogenic pathways. Notably, genes such as HSPA9, PPIF, and CFL1 were significantly enriched in processes like the negative regulation of apoptosis (GO:0043066), neutrophil activation in the immune response (GO:0002283), and neutrophil degranulation (GO:0043312). Additionally, genes such as PNP, PYGB, and CTSD emphasize the crucial role of neutrophils in the innate immune response, which may become dysregulated in the tumor microenvironment.19 Additionally, metabolic processes like gluconeogenesis (GO:0006094) and glucose metabolism remain prominent, indicating that while metabolic pathways are essential in both groups, they may be functionally reprogrammed in cancerous tissues to support tumor growth and survival under adverse conditions.
The analysis of proteomic data demonstrated that the RefSeq Abundant group is enriched in proteins associated with normal metabolic and structural functions, characteristic of a healthy cellular environment. In contrast, the Variant Abundant group exhibits enrichment in pathways related to immune modulation, apoptosis evasion, and metabolic reprogramming—hallmarks of oncogenesis. The fact that these findings are derived from proteomic mass spectrometry data strengthens the evidence that missense mutations lead to detectable changes in protein abundance, rather than being mere genomic alterations without functional consequences.
While this analysis provides strong evidence linking missense mutations to tumor-associated proteomic changes, further studies—such as functional assays and structural protein analyses—are needed to establish causation and elucidate the specific mechanistic consequences of these mutations on protein function.
Frequently mutated genes
The mutational analysis presented in Figure 4 is based on proteomic data obtained through mass spectrometry, meaning the identified mutations correspond to proteins that are not only genetically altered but also expressed and detected at the proteomic level. This distinction is crucial, as it confirms that these missense mutations are not merely theoretical genomic variants but are actively present in the tumor proteome, potentially influencing cellular processes.
Among the most frequently mutated proteins, several were previously identified in the functional enrichment analysis, including PGK1, MDH2, and HSPA9. These proteins play key roles in metabolic processes (GO:0006006, GO:0006094) and protein stabilization (GO:0050821), pathways enriched in the Abundant Variant group. The presence of these mutations at the proteomic level reinforces the hypothesis that missense mutations contribute to alterations in protein abundance and functional activity, impacting metabolic adaptations and cellular stress responses in tumor cells.
Additionally, proteins involved in cell adhesion (GO:0034109), immune regulation (GO:0002283), and apoptotic evasion (GO:0043066), such as ACTN4, NCL, and HSPA9, also exhibit recurrent missense mutations. These findings align with the Abundant Variant profile, suggesting that proteins carrying these mutations may actively participate in processes linked to tumor progression, immune modulation, and resistance to cell death.
The deleterious percentage analysis (left panel) further indicates that some of these missense mutations may compromise protein function, leading to structural instability or altered interactions within key signaling pathways. The presence of multi-hit mutations in some samples (black) suggests that specific proteins may be under selective pressure in tumors, further supporting their potential role in oncogenesis.
Integrating proteomic data with functional enrichment analysis strengthens the evidence that missense mutations influence protein abundance and cellular pathways in cancer. The confirmation of these mutations at the proteomic level suggests they are not only present but also biologically relevant, potentially contributing to metabolic reprogramming, apoptotic resistance, and immune system interactions in the tumor microenvironment. While these findings indicate a strong correlation between missense mutations and oncogenic processes, further studies—including functional assays and structural analyses of mutated proteins—are needed to establish their precise mechanistic impact on tumor progression.
The mutational analysis presented in the figure is based on proteomic data obtained through mass spectrometry, meaning the identified mutations correspond to proteins that are not only genetically altered but also expressed and detected at the proteomic level. This distinction confirms that these missense mutations are not merely theoretical genomic variants but are actively present in the tumor proteome, potentially influencing cellular processes.
Among the most frequently mutated proteins, several were previously identified in the functional enrichment analysis, including PPIF (rs765825575; NP_005720.1; p.Gln153His), PGK1 (rs148399096; NP_000282.1; p.Gly254Ala), MDH2 (rs782048786; NP_005909.2; p.Gly295Ala), and HSPA9 (rs745377968; NP_004125.3; p.Asn149Asp). These proteins play key roles in metabolic processes (GO:0006006, GO:0006094) and protein stabilization (GO:0050821), pathways enriched in the Abundant Variant group. The presence of these mutations at the proteomic level reinforces the hypothesis that missense mutations contribute to alterations in protein abundance and functional activity, impacting metabolic adaptations and cellular stress responses in tumor cells.
Additionally, proteins involved in cell adhesion (GO:0034109), immune regulation (GO:0002283), and apoptotic evasion (GO:0043066), such as ACTN4 (rs143174736; NP_004915.2; p.Gln392Glu), NCL (rs766672041; NP_005372.2; p.Val311Leu|rs535468192; NP_005372.2; p.Thr305Met), and HSPA9 (rs745377968; NP_004125.3; p.Asn149Asp), also exhibit recurrent missense mutations. These findings align with the Abundant Variant profile, suggesting that proteins carrying these mutations may actively contribute to tumor progression, immune modulation, and resistance to cell death.
The deleterious percentage analysis (left panel) further indicates that some of these missense mutations may compromise protein function, leading to structural instability or altered interactions within key signaling pathways. The presence of multi-hit mutations in some samples (black) suggests that specific proteins may be under selective pressure in tumors, further supporting their potential role in oncogenesis.
Mutation hotspots
Mutational hotspots were identified using Oncodrive software, as summarized in Table 2. These hotspots represent regions with significant mutation accumulation, potentially influencing protein stability, function, and oncogenic processes.
Table 2Summary of mutational hotspots identified using Oncodrive software
| Gene | Cluster | Mutation frequency |
|---|
| CSRP1 | 1 | 2 |
| GEMIN6 | 1 | 4 |
| HSPA9 | 1 | 2 |
| HSPD1 | 1 | 2 |
| PPIF | 1 | 2 |
| ACTB | 1 | 6 |
Among the identified genes, ACTB (rs11960; NP_004578.2; p.Ser766Leu|rs587779770; NP_001092.1; p.Gly74Ser) exhibited the highest mutation frequency (six occurrences), followed by GEMIN6 (rs767892869; NP_079051.9; p.Pro96Leu) (four occurrences). Notably, genes such as HSPA9 (rs745377968; NP_004125.3; p.Asn149Asp), HSPD1 (rs998941208; NP_955472.1; p.Val51Leu), and PPIF (rs765825575; NP_005720.1; p.Gln153His), which play crucial roles in protein stabilization (GO:0050821) and cellular stress responses, also presented recurrent mutations. These findings align with the functional enrichment analysis, which identified heat shock proteins as key players in maintaining protein homeostasis and cellular integrity in both normal and tumor environments.
Additionally, CSRP1 (rs34504522; NP_004069.1; p.Asn28Asp) and GEMIN6 (rs767892869; NP_079051.9; p.Pro96Leu), identified in the hotspot analysis, are linked to cytoskeletal organization and RNA-processing pathways—processes often dysregulated in cancer cells. Since these genes were also detected in proteomic mass spectrometry data, their mutational presence further supports the hypothesis that missense mutations contribute to altered protein abundance and functional disruption in oncogenesis.
The identification of these mutation-enriched regions underscores their potential role in tumor progression and metabolic reprogramming, suggesting that specific mutations may be under selective pressure in cancer. Further functional and structural studies are needed to determine how these mutations impact protein stability, interaction networks, and cellular signaling pathways in prostate cancer.
Discussion
This study investigated the correlation between missense mutations and protein abundance in healthy and cancerous prostate tissues, offering insights into their role in prostate cancer progression. Proteins involved in metabolic processes, cell adhesion, and immune response were identified, highlighting significant alterations that may influence genomic stability and tumor evolution. Additionally, novel findings expand our understanding of molecular mechanisms that remain underexplored in the literature, particularly those related to metabolic regulation, immune response, and cellular homeostasis.
Proteomic abundance analysis revealed distinct patterns between healthy and tumor tissues, suggesting that missense mutations contribute to proteomic imbalances. These alterations may be linked to metabolic reprogramming and apoptotic evasion in cancer cells, reinforcing the need for functional investigations to validate their biological impact.
A highly significant negative correlation observed in the analysis further indicates that missense mutations are associated with substantial changes in protein abundance. This inverse relationship suggests that affected proteins may undergo downregulation, disrupting cellular homeostasis and promoting oncogenic processes such as metabolic reprogramming and apoptosis evasion. Enrichment analysis supports these findings, showing that downregulated proteins are significantly associated with key biological pathways, including oxidative phosphorylation, glycolysis, and immune response modulation. These pathways are critical for tumor adaptation, enabling energy production under hypoxic conditions and facilitating immune evasion. Additionally, enrichment of cell adhesion and extracellular matrix organization pathways suggests that these mutations may enhance invasiveness and metastatic potential.
Unlike previous studies that primarily focused on the metabolic role of PGK1 in prostate cancer,20 our analysis integrates proteogenomic data to establish direct associations between genetic variants and proteomic alterations, providing a more comprehensive understanding of tumor biology. While PGK1 has been recognized for its involvement in glycolysis and oncogenic signaling, our approach highlights how specific genetic variants influence its expression and functional modulation, potentially leading to novel therapeutic strategies. Additionally, HSPA9 has been linked to prostate cancer progression, with studies reporting its high expression in neoplastic prostate tissues and its association with an increased risk of recurrence after salvage therapy.21 Similarly, MDH2 overexpression has been correlated with tumor progression and resistance to docetaxel treatment, indicating its potential as a prognostic biomarker and therapeutic target.22
Our study further identified genetic variants associated with these proteins, including PGK1 (rs148399096; NP_000282.1; p.Gly254Ala), MDH2 (rs782048786; NP_005909.2; p.Gly295Ala), and HSPA9 (rs745377968; NP_004125.3; p.Asn149Asp), suggesting that these alterations may have functional consequences in tumor biology. By incorporating quantitative proteomics, our study not only confirms the metabolic reprogramming of prostate cancer but also reveals how these key proteins influence immune interactions and treatment responses, offering new insights for targeted therapies.
Our approach offers significant advantages over previous studies by identifying mutational hotspots in key genes such as ACTB, PPIF, HSPA9, and CSRP1 and directly correlating them with protein abundance and functional implications. Unlike traditional genomic analyses that identify mutations without considering their actual protein expression, our proteogenomic approach integrates both genomic and proteomic data, confirming the presence of these variants in tumor tissues. For instance, mutations in ACTB (rs11960 and rs587779770) affect essential structural domains involved in cell adhesion, potentially promoting metastasis and immune evasion.23 Similarly, the PPIF mutation (rs765825575) may disrupt apoptosis regulation, contributing to tumor resistance.24,25
Moreover, the detection of mutations in CSRP1, along with its overexpression in tumor tissues,26 highlights its potential role in platelet aggregation-mediated immune protection, an aspect that remains underexplored. These findings demonstrate that our proteogenomic approach not only enables the identification of genetic variants but also provides functional validation, improving the identification of therapeutic targets and prognostic biomarkers for prostate cancer.
Notably, the mutations identified in this study are not reported in the Catalogue of Somatic Mutations in Cancer database, underscoring their novelty and potential clinical relevance. This finding highlights the effectiveness of the proteogenomic approach in detecting previously unreported variants with functional impact, enhancing the identification of therapeutic targets and prognostic biomarkers in prostate cancer.
Using PROVEAN,27 mutations were classified as either neutral or deleterious, prioritizing those with greater functional impact. Deleterious mutations in ACTN4 (rs143174736; p.Gln392Glu) and PPIF (rs765825575; p.Gln153His) were identified as key drivers of tumor progression. ACTN4 is involved in cytoskeletal remodeling, epithelial-mesenchymal transition (EMT), and metastasis in multiple cancers, including lung and breast cancer, as reported by Tentler et al.28 Additionally, actin-binding proteins that interact with ACTN4 contribute to cytoskeletal reorganization and tumor progression in prostate cancer, according to Fu et al.29PPIF influences mitophagy, apoptosis resistance, and immune modulation, reinforcing its role in tumor cell survival and dissemination, as evidenced by Feng et al.30
PGK1, a key enzyme in glycolysis, is pivotal in metabolic reprogramming, gene expression, and tumor progression, contributing to cellular proliferation, migration, and invasion.19 Similarly, MDH2, an essential enzyme in the Krebs cycle, facilitates mitochondrial metabolism and ATP production, underscoring its role in metabolic adaptation in prostate cancer.31 Functionally significant mutations in PGK1 (rs148399096; p.Gly254Ala) and MDH2 (rs782048786; p.Gly295Ala) further highlight their involvement in metabolic alterations, a hallmark of tumorigenesis. The presence of HSPA9 (rs745377968; p.Asn149Asp) mutations, a key molecular chaperone, suggests a role in protein stability and tumor adaptation under stress conditions, consistent with findings by Murphy et al.17
Unlike traditional genomic studies that rely on sequencing data without functional validation, our proteogenomic approach integrates genomic and proteomic data, ensuring direct experimental confirmation of mutation effects. This strategy offers several advantages, including the identification of novel mutations absent in the Catalogue of Somatic Mutations in Cancer database, direct correlation with protein abundance, and a deeper understanding of tumor progression mechanisms. The classification of deleterious mutations in cytoskeletal, metabolic, and mitochondrial proteins underscores their role in metastasis, metabolic adaptation, and therapy resistance, highlighting their potential as biomarkers and therapeutic targets.
By providing a functionally validated molecular profile of prostate cancer mutations, this study advances precision oncology and supports the development of mutation-driven targeted therapies. Future research should focus on functional assays and structural modeling to further characterize these mutations and explore their therapeutic implications.
Our study also identified a PNP mutation (rs1049564; NP_000261.2; p.Gly51Ser), which has not been previously reported in prostate cancer proteogenomics. PNP is essential for purine metabolism and immune regulation, influencing nucleotide homeostasis and purinergic signaling. Its role in immune modulation affects T lymphocytes and potentially neutrophils, with implications for the tumor microenvironment.32 In prostate cancer, PNP is regulated by miR-1 and miR-133a, linking it to tumor progression.33 However, direct evidence connecting PNP mutations to neutrophil activation, degranulation, or tumor immune modulation remains limited.
Unlike previous studies focused on transcriptomic profiling, our proteogenomic approach directly links PNP mutations to neutrophil-related pathways, revealing their potential role in immune evasion and tumor progression. This highlights PNP p.Gly51Ser as a novel immune-modulatory mutation, reinforcing the value of proteogenomic integration in identifying therapeutic targets for prostate cancer.
Despite these significant findings, our study has limitations. First, reliance on available databases may underestimate the complexity of genomic variants, potentially leading to the omission of rare but functionally relevant mutations. Additionally, the identified biomarkers require further clinical validation to confirm their relevance as therapeutic targets. The lack of access to detailed clinical data also limited our ability to perform clinicopathologic analysis, restricting a deeper understanding of patient-specific disease progression, treatment response, and clinical outcomes.
These findings emphasize the importance of integrating proteogenomic analyses in cancer research. Future studies should investigate additional mutation types, including nonsense, frameshift, and structural variants, to expand our understanding of their proteomic impact. Moreover, in vitro and in vivo functional validations are necessary to confirm the biological relevance of the identified proteins and their roles in tumor progression. Another promising direction involves exploring targeted therapies focused on metabolic and cell adhesion pathways, which were identified as key regulators in prostate cancer.
Expanding proteogenomic analyses to include immunogenicity and tumor microenvironment interactions may also provide novel therapeutic insights. The results emphasize the importance of proteogenomic integration in cancer studies, advocating for future work to explore additional mutation types of mutations and perform in vitro and in vivo functional validations. Additionally, exploring therapies that target at the identified metabolic and cell adhesion pathways may provide new strategies for prostate cancer management.
Conclusions
This study provides direct experimental evidence that missense mutations influence protein abundance in prostate cancer. Using a SILAC-based quantitative proteomics approach, we identified significant proteomic differences between healthy and tumor tissues, demonstrating that missense mutations correlate with changes in protein expression. Specifically, RefSeq Abundant proteins were linked to essential metabolic and structural processes, while Variant Abundant proteins were associated with pathways involved in apoptosis inhibition, immune modulation, and metabolic reprogramming.
Spearman’s correlation analysis revealed a significant negative correlation (p < 0.05) between protein intensity difference and protein intensity ratio, indicating that missense mutations contribute to alterations in protein abundance. Additionally, Oncodrive analysis identified ACTB and PPIF as mutation hotspots. ACTB mutations may affect cell adhesion and metastasis, while PPIF variants are linked to mitophagy regulation and tumor survival. PROVEAN classification further confirmed that specific mutations in PGK1, HSPA9, and MDH2 have deleterious effects on protein stability and function.
Unlike sequencing-based studies, which infer the effects of mutations, our proteogenomic approach directly validates their impact on protein expression, reinforcing their biological and clinical relevance. This methodology enabled the identification of novel mutations in PNP, CSRP1, and GEMIN6, suggesting potential roles in immune modulation, cytoskeletal organization, and metabolic adaptation. Furthermore, our findings highlight possible interactions between neutrophil-associated proteins and tumor immune evasion, warranting further investigation into their therapeutic potential.
Despite the strengths of this study, including the integration of proteomic and genomic data, some limitations must be considered. Dependence on existing databases may introduce biases, and further experimental validation is required to confirm the functional impact of the identified mutations. Future research should focus on in vitro and in vivo functional assays to establish mechanistic links between missense mutations and tumor progression, as well as their potential applications in targeted therapy and precision oncology.
In summary, this study advances the understanding of prostate cancer biology by demonstrating the role of missense mutations in shaping the tumor proteome. These findings contribute to the identification of potential biomarkers and therapeutic targets, supporting the development of more precise diagnostic and treatment strategies.