Abstract
Prostate cancer lineage plasticity is a key driver in the transition to neuroendocrine prostate cancer (NEPC), and the RTK/RAS signaling pathway is a well-established cancer pathway. Nevertheless, the comprehensive link between the RTK/RAS signaling pathway and lineage plasticity has received limited investigation. In particular, the intricate regulatory network governing the interplay between RTK/RAS and lineage plasticity remains largely unexplored. The multi-omics data were clustered with the coefficient of argument and neighbor joining algorithm. Subsequently, the clustered results were analyzed utilizing the GSEA, gene sets related to stemness, multi-lineage state datasets, and canonical cancer pathway gene sets. Finally, a comprehensive exploration of the data based on the ssGSEA, WGCNA, GSEA, VIPER, prostate cancer scRNA-seq data, and the GPSAdb database was conducted. Among the six modules in the clustering results, there are 300 overlapping genes, including 3 previously unreported prostate cancer genes that were validated to be upregulated in prostate cancer through RT-qPCR. Function Module 6 shows a positive correlation with prostate cancer cell stemness, multi-lineage states, and the RTK/RAS signaling pathway. Additionally, the 19 leading-edge genes of the RTK/RAS signaling pathway promote prostate cancer lineage plasticity through a complex network of transcriptional regulation and copy number variations. In the transcriptional regulation network, TP63 and FOXO1 act as suppressors of prostate cancer lineage plasticity, whereas RORC exerts a promoting effect. This study provides a comprehensive perspective on the role of the RTK/RAS pathway in prostate cancer lineage plasticity and offers new clues for the treatment of NEPC.
Similar content being viewed by others
Introduction
Prostate cancer is a prevalent malignancy that predominantly affects elderly males1. In 2016, 95% of prostate cancer patients were over 60 years old, and 51.2% of those were over 80 years old. The global incidence of prostate cancer in 2018 reached 1.3 million new cases, and there were 355,000 deaths attributed to the disease. This positions prostate cancer as the second most common cancer worldwide and the fifth leading cause of cancer-related deaths among men2.
Lineage plasticity has emerged as a critical mechanism contributing to treatment resistance in prostate cancer. Androgen deprivation therapy is a standard treatment due to the dependency of prostate cancer cells on the androgen receptor (AR)3,4. In clinical practice, some castration-resistant prostate cancer patients have tumor cells that exhibit both morphological features of neuroendocrine cancer and mixed luminal and basal cell characteristics upon recurrence5. This ability to transition from epithelial cells to neuroendocrine cells is lineage plasticity6. The acquisition of prostate cancer cell lineage plasticity can lead to cells transforming into stem-like and multilineage states, differentiating into new or distinct lineages, undergoing reprogramming to transform into a novel type of cell that can survive without the need for AR, and leading to the acquisition of AR-independent characteristics through lineage transformation7. There has been recent progress in the research on prostate cancer lineage plasticity. Reports indicate that the loss of tumor suppressor functions like TP53 and RB1 promotes the transition of AR-dependent luminal epithelial cells to AR-independent basal-like cells, and EZH2 inhibitors have been found to reverse this lineage transition and to restore sensitivity to androgen deprivation therapy8. Additionally, SOX2, through the JAK-STAT signaling pathway, can facilitate the conversion of AR-dependent prostate cancer cells into a multi-lineage state7,9. FOXA2 can induce the transition of multi-lineage state cells to Neuroendocrine Prostate Cancer (NEPC) through the KIT pathway10. ASCL1, activated downstream of the ROR2/CREB signaling pathway, contributes to prostate cancer lineage plasticity11. Specific SWI/SNF complexes are associated with PCa progression and may play a role in treatment resistance12. Moreover, epigenetic factors such as chromatin modifications and DNA methylation also play a role in promoting lineage plasticity13. Indeed, lineage plasticity is not confined to prostate cancer, as this phenomenon has been noted in both lung and breast cancer14,15. However, the research on lineage plasticity remains incomplete, there are limited clinical trials addressing these challenges, and the prognosis for patients continues to be bleak.
Public databases offer rich sources of various omics data related to prostate cancer. This study leverages these to analyze and explore lineage plasticity of prostate cancer. We conducted an analysis of multi-omics data from TCGA and SRA databases, including AR ChIP-seq, transcriptomic, single-cell transcriptomic, lncRNA, SNV, CNV, DNA methylation, and clinical data of PCa patients to identify key gene sets related to prostate cancer. Then, annotation and clustering analysis were applied to these gene sets, resulting in 6 function modules. Based on the TCGA, NEPC, CRPC cohort data, and data on prostate cancer cell stemness and lineage plasticity, it was discovered that function module 6 is more closely associated with CRPC and is related to cancer stemness and lineage plasticity. Analysis of Module 6 in relation to 10 oncogenic signaling pathways (cell cycle, Hippo signaling, Myc signaling, Notch signaling, oxidative stress response/Nrf2, PI-3-Kinase signaling, receptor-tyrosine kinase (RTK)/RAS/MAP-Kinase signaling, TGFb signaling, p53, and b-catenin/Wnt signaling)16, revealed that it is positively correlated with 3 of these cancer pathways (cell-cycle, p53, and RTK/RAS). The RTK/RAS pathway plays a crucial role in the progression, extraprostatic extension, and development of castration-resistant prostate cancer (CRPC), and contributes to prostate cancer lineage plasticity11,17.
Measuring the extent of the RTK/RAS pathway's contribution to lineage plasticity in prostate cancer is challenging. Through a series of computational analyses, we explored the connections between the marker gene set of multi-lineage state cells, prostate cancer samples, and the gene set associated with the RTK/RAS signaling pathway. The findings revealed that the RTK/RAS signaling pathway promotes lineage plasticity. Transcription factors (TFs) and genomic variations are molecular driving factors of prostate cancer lineage plasticity18. This study unveiled that both copy number variations (CNVs) and TFs contribute to the RTK/RAS pathway in prostate cancer lineage plasticity. Particularly, TFs TP63 and FOXO1 suppress and RORC drives prostate cancer lineage plasticity through the RTK/RAS pathway.
The investigation into prostate cancer lineage plasticity and its driving factors is important for combating cancer resistance. This work revealed that CNVs within the RTK/RAS pathway, along with their upstream TFs, drive lineage plasticity of prostate cancer. We also identified a set of potential target gene options for prostate cancer patients who have developed treatment resistance. Moreover, this study offers novel insights and references for drug resistance research for other forms of cancer.
Results
Differential expression genes, prognostic lncRNAs, and extraprostatic extension related genes were obtained through prostate cancer expression data analysis
There are 770 DEGs found to be regulated by AR in prostate cancer through intersecting AR Chip-seq data from the SRA and TCGA databases (Fig. 1A). These DEGs play a significant role in the progression of prostate cancer, for example: TIMP319, HSD17B320, SRD5A221, FN122, CYP19A123, SGK124, SCNN1A25, RND326, FASN27, CDK1528, SOX529 (additional details can be found in the Supplementary Table S1). GO and KEGG enrichment analyses were performed on these genes (Supplementary Tables S2 and S3) and all annotated genes are provided in Supplementary Table S4.
Then, 5 lncRNAs that were associated with the prognosis of prostate cancer was identified by univariate and multivariate Cox analyses (Supplementary Fig. S1) based on lncRNA transcriptome data from the TCGA prostate cancer database (Fig. 1B), and the genes interacting with these lncRNAs were annotated (Supplementary Tables S5–S7). The predictive model was defined as the linear combination of the expression levels of the 5 lncRNAs, which were weighted as follows: survival risk score = ((− 0.56041) × AC106881.1 + (0.81588) × PRRT3-AS1 + (1.234135) × AC004148.2 + (0.511379) × AP000844.2 + (0.297335) × AL354710.2) (p = 1.11e−07; C-index = 0.95). The threshold of risk score was 0.8789252 with linear model fits. PRRT3-AS1 has the capability to effectively inhibit the proliferation of prostate cancer cells, induce cell apoptosis and autophagy30, and it is a biomarker and therapeutic target for skin cutaneous melanoma31. AP000844.2 is involved in the PCa’s early prognosis and risk stratification32, and its expression exhibits significant differences among various aging-related subtypes in prostate cancer patients33. Additionally, Huang et al. (2022) also identified the significance of AP000844.2 in their analysis of the role of lncRNAs in prostate cancer34. AC106881.1 is an antisense RNA of UNC5C, which is involved in neuronal development35. AC004148.2 is one of the lncRNAs in the immune-related lncRNA risk model for the prognosis of head and neck squamous cell carcinoma patients36. From the analysis of differential expression between extraprostatic extension and intraprostatic tumors, a total of 372 genes associated with Gleason grading were identified (Fig. 1C). More detailed gene information and annotation results can be found in Supplementary Tables S8–S10.
Acquisition and annotation of SNV, CNV, and methylation-driven genes in prostate cancer
The analysis of prostate cancer SNV data indicates that (Fig. 1D), in high mutation samples (TP53 + SPOP + TTN), samples with TP53 mutations are significantly enriched in those with Gleason grade 5 (8.28e-03) and in samples with extraprostatic extension of prostate cancer (1.756173e-34). The corresponding DEGs associated with these mutations can be found in Supplementary Table S11, and the results of GO and KEGG analyses are presented in Supplementary Tables S12 and S13.
Furthermore, the visualization of prostate cancer CNVs and loops was performed, depicting their locations on the chromosomes (Fig. 1E). BCL237, ETV638, and CDKN1B39 are important genes in the occurrence and development of prostate tumors, experiencing not only copy number deletions but also appearing in loops. The detailed list of DEGs between CNV and non-CNV samples can be found in Supplementary Table S14, and the annotation analysis information for DEGs can be accessed in Supplementary Tables S4 and S15–S16.
Methylation-driven genes were also identified (Fig. 1F) and a detailed list of these can be found in Supplementary Table S17 and the corresponding annotation information can be accessed in Supplementary Tables S4 and S18–19.
Typically, a biological process is accomplished by a group of genes working together, while a biological phenotype is regulated by multiple biological processes. Therefore, mining the results of GO and KEGG enrichment analysis based on multi-omics data enabled the assembly of prostate cancer-related functions or pathways into a larger biological context, leading to a better understanding of the biological questions related to prostate cancer phenotypes. CRPC (Castration-Resistant Prostate Cancer) and NEPC (Neuroendocrine prostate cancer) are two classic phenotypes of prostate cancer. By integrating three cohort datasets from TCGA, CRPC, and NEPC, and quantifying the CRPC and NEPC scores in TCGA, it was found that samples in TCGA with T3b, T4, N1 stages, Gleason grade group 5, and those that received drug treatments also exhibited significantly higher CRPC and NEPC scores (Supplementary Fig. S2).
Six functional modules were identified through a comprehensive analysis of DEGs, prognostic lncRNAs, extraprostatic extension related genes, SNV, CNV, and methylation-driven genes in prostate cancer
Further clustering analysis was performed on the annotation results of the six datasets on prostate cancer, including the DEGs, genes regulated by prognostic lncRNAs, genes associated with extraprostatic extension, DEGs between mutation and non-mutation prostate cancer samples, DEGs between CNV and non-CNV prostate cancer samples, and methylation-driven genes. These pathways and functions were classified into six Functional Modules (Fig. 2A). After scoring TCGA, CRPC, and NEPC using six functional modules, it was discovered that functional modules 5 and 6 differ from modules 1–4 and are closely associated with the CRPC phenotype. Moreover, CRPC and NEPC characteristics exhibit a positive correlation with functional module 6 (Supplementary Fig. S3). Function Modules 1–4 encompass processes related to cancer occurrence and development, including the Wnt signaling pathway, positive regulation of epithelial cell proliferation, positive regulation of endothelial cell proliferation, endothelial cell migration, Ras signaling pathway, muscle cell development, cell adhesion molecules, and cell–cell junction. Function Module 5 includes biological processes related to the microenvironment and epithelial-mesenchymal transition (EMT), such as responses to endoplasmic reticulum stress and protein glycosylation hydrolase activity, acting on glycosyl bonds. Notably, Function Module 6 includes entries related to stemness, such as the regulation of G2/M transition of mitotic cell cycle, mitotic nuclear division, and response to ionizing radiation. It is important to highlight that Function Module 6 has a low Z-score (lighter color bar), indicating a significant upregulation of the pathway.
In addition, a protein–protein interaction network was constructed based on 300 genes obtained from the intersection of the six datasets (Supplementary Fig. S4A). A core network consisting of 19 genes was identified, and most genes in the core network have been reported to play important roles in prostate cancer, especially TPM1, ITGB3, and CALD1 (Supplementary Fig. S4B). RT-qPCR validation was performed on these three genes, which revealed their high expression in prostate cancer tissues (Supplementary Fig. S4C).
The relationship between Function Modules 1–6 and the six datasets is summarized in Fig. ure 2B. All six gene-sets are involved in Function Modules 1–4, while the gene-sets of CNV, lncRNA and Methylation are involved in Function Module 5. The gene-sets of lncRNA, SNV, and CNV are involved in Function Module 6. More detailed information about the six functional modules and their source data is presented in Fig. 2C–H. Some genes in the Functional Modules are derived from multiple datasets. A total of 42 genes in the six Functional Modules were investigated, and 26 of these genes exhibit expression in prostate cancer cells or tissues (Supplementary Fig. S5). Among these genes, SOX529, PRKCB40, VCL41, ARSB42, TF (Transferrin)43,44, APC45, PTEN46, TP5347, BRCA148, DCP249, DCUN1D250, MKLN151, LPAR152, EFNA5 and SORBS153,54 have been reported play important roles in prostate cancer. COL14A155, ANGPT156, SERPINF157, GJA158, SMC259, PCID260, FBXL861, and FBXL1762 have also been reported in other cancers. KCNS2, MYOCD63, and DCAF864 are involved in other biological processes. These results suggest that the genes identified in our study but not previously reported to be involved in prostate cancer may be candidate genes in prostate cancer.
GSEA analysis of the Function Modules suggested that the stemness of prostate cancer cells primarily originates from genetic mutations
GSEA analysis was performed on the six functional modules (Fig. 3). The results revealed that Function Module 1–5 (Fig. 3A–E) were negatively enriched with tumor stem cell characteristics (SUZ12_TARGETS_BENPORATH or PRC2_TARGETS_BENPORATH). In addition, Function Module 5 (Fig. 3E) also showed negative enrichment with epithelial, endothelial cells, and wound response functions, suggesting it represents a complex tumor microenvironment that may be associated with EMT in prostate cancer. Surprisingly, in contrast to Function Modules 1–5, Function Module 6 exhibited a significant positive correlation with stem cells, meaning that there is a close relationship between it and cell stemness (Fig. 3F). Moreover, it is noteworthy that the primary dataset for Function Module 6 is derived from CNV analysis. This result provides compelling evidence that CNV is helpful for the stemness of prostate cancer cells. This aligns with previous research that also found that high CNV contributes to the stemness of prostate cancer cells65.
Function Module 6 is positively correlated with lineage plasticity in prostate cancer as shown through integrative analysis of multiple datasets, which provides a valuable means for assessing the lineage plasticity in samples
Multi-lineage states represent important stages of cancer cell differentiation characterized by both stemness and multi-lineage potential. Evaluating lineage plasticity in prostate cancer is currently a research hotspot and major challenge. Based on GSEA analysis and datasets associated with stem cell-related genes66,67, our study revealed a positive correlation between the Function Module 6 and stemness (Fig. 4A). Leveraging GSEA analysis and multi-lineage states datasets7, our investigation unveiled a positive relationship between the Function Module 6 and multi-lineage states (Fig. 4B). Moreover, employing GSEA analysis along with datasets related to classic cancer driver pathways16, our findings demonstrated that the Function Module 6 is positively associated with three cancer pathways (cell-cycle, p53, and RTK/RAS) (Fig. 4C), thus, a substantial connection between Function 6 and these biological aspects was confirmed. This finding suggests that the Function Module 6 can serve as a functional module for assessing lineage plasticity in prostate cancer.
CNVs and TFs contribute to the involvement of RTK/RAS pathway genes in prostate cancer lineage plasticity
Recent research has highlighted the involvement of RTK in prostate cancer lineage plasticity11, and our analyses have indicated Function Module 6 is positively correlated with prostate cancer lineage plasticity and RTK/RAS (Fig. 4). Thus, our research centered on RTK/RAS and aims to systematically delve into the intricacies of the pathway in relation to prostate cancer lineage plasticity through bioinformatics analysis. In our approach, we overcame the challenge of quantifying the relationship between pathways and lineage plasticity by employing ssGSEA and WGCNA method calculations to derive correlation values between all genes within the Function Module 6 and the lineage plasticity of prostate cancer. We further subjected all genes in the RTK/RAS signaling pathway to GSEA analysis for their lineage plasticity, revealing that it indeed enhances lineage plasticity (Fig. 5A).
Then, we analyzed the 19 LEGs (BUB1B, CCNE2, CDK7, CENPF, ECT2, FOXM1, GINS2, KIF11, MYC, ORC6, PARD6A, PBK, RAD51, RAD51AP1, RFC3, RFC4, RPL12, TTK, and TUBB4B) which drive the prostate cancer lineage plasticity through the RTK/RAS signaling pathway as identified using GSEA. These LEGs exhibited significant upregulation in prostate cancer samples from the TCGA database (Fig. 5B and Supplementary Table S20), and their expression was also detected in single-cell transcriptome sequencing of the prostate cancer cell line LNCaP68 (Fig. 5C). Moreover, these 19 genes displayed varying levels of CNVs in both the TCGA and SU2C prostate cancer cohorts (Fig. 5D). Previous research indicates that some of the LEGs, including BUB1B69, CDK770, CENPF71, FOXM172, KIF1173,74, ORC675, PBK76, RAD5177, TTK78, CCNE279, and MYC80, have previously been reported to be associated with prostate cancer, particularly in relation to metastasis. Additionally, ECT281, GINS282, PARD6A83, RAD51AP184, RFC385, RFC486, and TUBB4B87 have been reported to function in other cancer occurrences. All these LEGs contribute to lineage plasticity.
Based on the transcriptome data of different TF knockdowns from the GPSAdb database88, a total of 194 TFs were enriched (Supplementary Table S21) and hence impact the expression of LEGs. The TFs that for the top enrichment LEGs were identified (Fig. 5E), and among these the 10 with transcription active in prostate cancer were selected (Fig. 5F). These indicate a closer correlation between the 10 TFs and the lineage plasticity of prostate cancer. Moreover, all 10 TFs, including RORC89, AR90, LIN28B91, IRF992, NKX2-593, IRF394,95, RUNX196, TRPS197, FOXO198, and TP638, have been reported to play roles in prostate cancer, especially in castration-resistant prostate cancer. This finding further confirms the reliability of our study. It is worth noting that TP63, FOXO1, and RORC exhibit significant transcription activity (p-value < 0.05), with the transcription activity of TP63 and FOXO1 being reduced and RORC increased (Fig. 5F). It is established that TP63 regulates the expression of p53, which can inhibit prostate cancer lineage plasticity, suggesting that TP63 could potentially contribute to the inhibition of prostate cancer lineage plasticity8. FOXO1 has been shown to restrain the migration and invasion of prostate cancer cells98. On the other hand, RORC drives the expression of AR in prostate cancer89. Based on all the information presented above, TP63 and FOXO1 suppress prostate cancer lineage plasticity through the RTK/RAS pathway, whereas RORC drives prostate cancer lineage plasticity through RTK/RAS pathway.
In summary, the research indicates that both CNVs and TFs contribute to the involvement of the RTK/RAS pathway in prostate cancer lineage plasticity.
Discussion
Treatment for advanced prostate cancer patients remains undiscovered. After acquiring lineage plasticity, prostate cancer cells progress into neuroendocrine prostate cancer. Moreover, lineage plasticity also occurs in other malignant tumors. The study of prostate cancer lineage plasticity and its driving factors is a focus point in prostate cancer research, which has significant importance in advancing our understanding of cancer drug resistance and related research. Our analyses have revealed that the RTK/RAS signaling pathway promotes prostate cancer lineage plasticity. A total of 19 LEGs with high numbers of CNVs significantly contribute to this process, and the TFs that influence the expression of these genes were identified. TFs regulating the same gene exhibited contrasting effects: some suppressed prostate cancer lineage plasticity and others promoted it (Fig. 6). This provides a range of potential targets for treatment and novel insights for investigating prostate cancer lineage plasticity.
Here, we aimed to explore the critical factors in the development of prostate cancer from a more comprehensive and systematic perspective. The key gene sets associated with prostate cancer across various omics profiles were obtained from the TCGA and SRA databases, including data from AR ChIP-seq, transcriptomic, single-cell transcriptomic, lncRNA, SNV, CNV, DNA methylation, and clinical data for prostate cancer patients (Fig. 1). Network analysis was performed on the intersection of 300 genes from these six datasets, and qRT-PCR validation confirmed the 3 top hub genes (TPM1, ITGB3, and CALD1) in the network are expression in prostate cancer tissues (Supplementary Fig. S4). The 3 top hub genes represent significant alterations at every level in prostate cancer and have been reported to play important roles in the cancer65,99,100. Then, six function modules were obtained through a neighbor joining algorithm based on the GO and KEGG enrichment analysis of prostate cancer multi-omics data (Fig. 2A). Cellular proteomics analysis was performed on the intersecting genes of the 6 function modules, and the results indicated that most of these exhibited protein expression in prostate cancer (Fig. S5). Notably, APC, PTEN, TP53, and BRCA1, all tumor suppressor genes, were intersecting genes in the SNV and CNV datasets, and these are also in Function Module 6. APC acts as an antagonist of the Wnt signaling pathway, and there is a strong correlation between APC methylation and prostate cancer45. Loss of PTEN function is one of the most common driver events in the development of prostate cancer46. TP53 mutations are closely associated with tumor progression in prostate cancer, and mutations are widespread in various cancer types47. Pathogenic variants of BRCA1 have been extensively studied for their association with the risk of prostate cancer48. This suggests Function Module 6 and CNVs play crucial roles in the development of prostate cancer. Our previous work has also indicated the involvement of CNV in prostate cancer development67.
Further analysis of the six functional modules revealed that the Function Module 6 is the only one with significant positive correlation with stem cell characteristics, while the Function Module 1–5 are negatively enriched with tumor stem cell (Fig. 3). Moreover, Function Module 6 is also positively correlated with the prostate cancer multi-lineage state (Fig. 4AB), suggesting a positive correlation between it and prostate cancer lineage plasticity as the coexistence of multi-lineage states and stemness is a characteristic of a lineage plasticity cell. Notably, our analysis indicates a positive correlation between Function Module 6 and the RTK/RAS signaling pathway (Fig. 4C), which is one of the classical cancer pathways and has recently been reported to play a role in prostate cancer lineage plasticity11.
By integrating multiple computational methods, it has been confirmed that the RTK/RAS signaling pathway is significantly enriched in samples with high scores for prostate cancer lineage plasticity (Fig. 5A). All LEGs driving the RTK/RAS signaling pathway to promote prostate cancer lineage plasticity exhibit significantly higher expression in prostate cancer samples compared to normal samples (Fig. 5B), and they display high expression in scRNA-seq data for prostate cancer (Fig. 5C). Most of the LEGs, including BUB1B, CDK7, CENPF, FOXM1, KIF11, ORC6, PBK, RAD51, TTK, CCNE2, and MYC, have previously been reported to be associated with prostate cancer, particularly in the context of metastasis, and other LEGs, including ECT2, GINS2, PARD6A, RAD51AP1, RFC3, RFC4, and TUBB4B, have also been reported to function in other cancer occurrences. CENPF and FOXM1 drives the malignancy of prostate cancer101, CDK7 is a crucial gene in castration-resistant prostate cancer cells70, KIF11 is an indicator of the invasiveness of prostate cancer73,74, RAD51 is overexpressed in invasive prostate cancer77, and MYC can promotes prostate cancer development80. CCNE2 and BUB1B promotes the proliferation and migration of prostate cancer cells69,79. Additionally, ORC6 shows a positive correlation with T-regulated cell immune infiltration in prostate cancer tissues75, and PBK also drives prostate cancer76. These genes all hold potential as targets for diagnosis, prognosis, and treatment in prostate cancer. For instance, silencing TTK can inhibit the proliferation and progression of prostate cancer78, and KIF11 inhibitors have been developed as chemotherapy drugs for treating various cancers73,74. Moreover, ECT2 and GINS2 have roles in breast cancer81,82, RAD51AP1 plays a role in ovarian cancer84, RFC3 is involved in gastric and colorectal cancer85, and RFC4 is associated with oral tongue squamous cell carcinoma86, Notably, GINS2 and TUBB4B are involved in regulating cancer stem cells, PARD6A plays a role in epithelial-mesenchymal transition87, and RAD51AP1 promotes cancer development84. The roles of these genes in prostate cancer lineage plasticity should be further investigated.
CNV data was the predominant source of genes in Function Module 6 (Fig. 2H), suggesting the importance of CNV in prostate cancer lineage plasticity. Therefore, a CNV analysis of LEGs in prostate cancer was conducted, and significant CNV alterations in LEGs were identified (Fig. 5D). These provide further evidence that CNVs contribute to the driving force of prostate cancer lineage plasticity. Considering that gene expression is regulated by TFs, we identified the TFs with transcriptional activity in prostate cancer samples that can influence the expression of multiple LEGs (> 14). There are 10 such TFs, including RORC, AR, LIN28B, IRF9, NKX2-5, IRF3, RUNX1, TRPS1, FOXO1, and TP63 (Fig. 5E,F). These TFs have all been reported to play roles in prostate cancer. RORC, AR, LIN28B, IRF9, and IRF3 exert promotive roles in prostate cancer. RORC drives the expression of AR in prostate cancer89. AR is a very important gene in castration-resistant prostate cancer (CRPC), with 80% of CRPCs carrying elevated AR gene copy numbers and approximately 30% of CRPCs having high-level AR gene amplification90. LIN28B is a potential target for the treatment of neuroendocrine prostate cancer (t-NEPC)102, as it promotes the development of the neuroendocrine prostate cancer91. IRF9 and IRF3 play roles in prostate cancer progression92,94,95. Additionally, the transcriptional activity of RORC, AR, LIN28B, IRF9, and IRF3 is positively correlated with LEGs (Fig. 5F). Combining existing reports with the analysis presented in this study, it can be inferred that RORC, AR, LIN28B, IRF9, and IRF3 promoted prostate cancer lineage plasticity through the RTK/RAS pathway. NKX2-5, RUNX1, TRPS1, FOXO1, and TP63 play a inhibitory roles in the development of prostate cancer. NKX2-5 exhibits anti-cancer activity in the process of prostate cancer development93, and elevated RUNX1 expression in prostate cancer patients tends to be associated with a more favorable clinical prognosis96. TRPS1 may participate in oxidative stress-induced apoptosis in androgen-independent DU145 prostate cancer cells103. TP63 regulates the expression of p53, which can inhibit prostate cancer lineage plasticity, meaning it could potentially contribute to the inhibition of prostate cancer lineage plasticity8. FOXO1can also restrain the migration and invasion of prostate cancer cells98. The transcriptional activities of NKX2-5, RUNX1, TRPS1, FOXO1, and TP63 are negatively correlated with LEGs (Fig. 5F), suggesting these suppress prostate cancer lineage plasticity through the RTK/RAS pathway. The process of the RTK/RAS pathway driving prostate cancer lineage plasticity is a complex interplay involving multiple TFs, with TP63, FOXO1, and RORC potentially playing more crucial roles due to their relatively high activity levels. TP63 and FOXO1 inhibit prostate cancer lineage plasticity and RORC promotes it.
Therefore, based on multi-omics data of prostate cancer, we have identified Function Modules associated with prostate cancer lineage plasticity, and the RTK/RAS signaling pathway has been found to promote prostate cancer lineage plasticity. Both CNV and TFs contribute to this process.
Conclusions
In this study, we identified 6 functional modules based on multiple omics datasets from prostate cancer obtained from the TCGA and SRA databases. Analysis of the functional modules revealed that Module 6 has a positive enrichment with both prostate cancer lineage plasticity and the RTK/RAS signaling pathway. The CNVs and TFs, such as TP63, FOXO1, and RORC, promote the RTK/RAS pathway-driven prostate cancer lineage plasticity by upregulating the LEGs. This research revealed the crucial role of the RTK/RAS signaling pathway in promoting prostate cancer lineage plasticity, providing a comprehensive list of potential therapeutic targets for prostate cancer lineage plasticity. Furthermore, it offers valuable insights for studying lineage plasticity in other types of malignancies.
Methods
Joint analysis of AR ChIP-seq data and transcriptome data in the TCGA database
Twenty raw AR ChIP-seq datasets were downloaded from NCBI's SRA database and named GSE700793, including 13 from cancer patients and 7 from normal samples. Adaptor sequences were removed using Trimmomatic104, followed by removal of low-quality bases using FASTX-Toolkit105, and lastly quality-control metrics were obtained using FastQC106. Resulting filtered data was aligned to a reference genome using Bowtie2107. Then, peaks were called using MACS108, and annotated with HOMER109 to obtain the AR-regulated genes. Raw count transcriptome data were obtained from the TCGA database (https://www.cancer.gov/tcga) and analyzed for differential gene expression using edgeR110. Subsequently, the AR-regulated genes were merged with differentially expressed genes (DEGs) from the transcriptome to detect common genes. Finally, clusterProfiler111 was used to perform GO and KEGG (https://www.kegg.jp/kegg/kegg1.html)112 enrichment analysis on these common genes with a significance level of p-value < 0.05 and q-value < 0.05.
Univariate and multivariate Cox analyses of LncRNA and functional predictions
LncRNA data was downloaded and extracted from the transcriptome of PCa in the TCGA database, and differential expression lncRNAs were obtained using edgeR. Fifteen lncRNAs (p < 0.01) were obtained based on the differential expression and clinical data downloaded from the TCGA database using the survrval package113 for univariate Cox analysis. 15 lncRNAs were identified through univariate Cox analysis, and subsequently, 5 lncRNAs were finalized after multivariate Cox analysis. Then, the accuracy of the model was validated through ROC curves114 based on TCGA and GEO115 test sets. After performing a correlation analysis between lncRNAs and mRNAs, two lncRNAs were ultimately obtained that exhibited correlation. These two lncRNAs were then subjected to GO and KEGG enrichment analysis using clusterProfiler111, with p-value < 0.05 and q-value < 0.05.
Analysis of extraprostatic extension and intraprostatic tumors data
Based on the definitions of the American Joint Committee on Cancer (AJCC) tumor, node, metastasis (TNM) system, prostate cancer staged T3-T4 was classified as extraprostatic extension, and stage T2 was classified as intraprostatic tumors. DEGs were identified based on these two classifications. Utilizing the Gleason primary and secondary scores in the TCGA clinical information, samples were classified according to the Gleason grading system (Grade Group 1 = Gleason 6 or less; Grade Group 2 = Gleason 3 (primary) + 4 (secondary) = 7; Grade Group 3 = Gleason 4 + 3 = 7; Grade Group 4 = Gleason 8; Grade Group 5 = Gleason 9–10). Employing the five Gleason grades, genes associated with Gleason grading were extracted from DEGs, and subsequent GO and KEGG enrichment analyses were performed.
SNV data analysis from the TCGA database
Based on clinical information of extraprostatic extension and the Gleason grade in the TCGA database, a SNV waterfall plot was generated using GenVisR package116. The formula for enrichment analysis (hypergeometric test) is as follows:
N represents the total population size, M represents the number of elements in the population with a specific attribute, n represents the number of draws, and k represents the number of hits for the element with the specific attribute. Based on Formula (1), perform enrichment analysis on TP53 mutation samples among high mutation samples. Genes related to SNVs were obtained from prostate cancer SNV and transcriptome data, and GO and KEGG enrichment analysis was performed on these.
Analysis of CNV and Hi-C data
Based on CNV data from TCGA, gene copy numbers were obtained using the formula 2^(Segment + 1). Based on the results of gene copy numbers, values greater than 3.5 were defined as 2 (amplification), values greater than 2.5 and less than or equal to 3.5 were defined as 1 (Single gain), values less than 0.5 were defined as -2 (Double deletion), values greater than or equal to 0.5 and less than 1.5 were defined as -1 (Single deletion), and the remaining values were defined as 0 (normal). This ultimately led to the construction of a CNV matrix with four types of mutations and one normal type. Genes with significant differences in CNV between tumor and normal samples were identified through chi-square tests. Based on prostate cancer transcriptomic and CNV data, genes whose expression were significantly affected by CNV were identified with the Kruskal–Wallis test. Utilizing the list of genes with significant differences in CNV, the count of samples with the most prevalent mutation type within this list was extracted from the CNV matrix, and the proportion of these samples in relation to tumor samples was computed. GO and KEGG enrichment analyses were performed on genes with significant CNV differences to obtain results related to functions and pathways. Obtain prostate cancer Hi-C data117, and after quality control118, use Hic-Pro119 and Juicer120 to identify differential loops between prostate cancer and normal samples. Ultimately, CNV and loop circos plots were generated using the RCircos121 package.
Analysis of methylation data from the TCGA database
Based on prostate cancer methylation and transcriptome data from the TCGA database, methylation-driven genes were obtained using the MethyMix package122, and GO and KEGG enrichment analysis were performed on these.
Protein–protein interaction network analysis and RT-qPCR experiments
Protein–protein interaction data for prostate cancer-related genes were obtained from the STRING database. A network was visualized using the software Cytoscape (3.9.1 version; https://cytoscape.org/index.html)123, and the core interacting protein network was extracted.
Total RNA from all samples was extracted using the TRIzol reagent (Invitrogen, USA), and genomic DNA contamination was removed using DNase I (Promega, USA). RNA yield was measured using a NanoDrop ND-2000 (Thermo Fisher Scientific, USA), and integrity was assessed on a 1% agarose gel. cDNA was synthesized using oligo-dT primers and SuperScript III (Takara, JP). Real-time Quantitative PCR (RT-qPCR) was performed using the Plus real-time PCR system (Applied Biosystems, USA). mRNA RT-qPCR was carried out using the SYBR Prime ScriptTM RT-PCR kit (Takara, JP). Data were analyzed using the relative 2−ΔΔCT method124. RT-qPCR primer sequences are listed in Supplementary Table S22.
Functional and pathway clustering of multi-omics data
Based on the results of GO and KEGG enrichment analysis from multi-omics data, a matrix of kappa values125,126,127 is constructed using gene pairs. Then, based on the kappa value matrix, a clustering tree is constructed using the bottom-up Neighbor Joining Algorithm128. The clustering results of functions and pathways were visualized using the R package ggtree129. The information for each cluster is counted using the UpSetPlot package130, and then a gene list is constructed based on the gene set of each cluster and the logFC values from the transcriptomic data analysis results. Finally, each cluster is annotated using the GSEA algorithm131, and the HALLMARK and SU2C databases9. Furthermore, protein expression identification of the prostate cancer genes in the clustering analysis results was performed using the protein data from the Human Protein Atlas database.
Transcriptome data for CRPC132 and NEPC133 were obtained from cBioPortal134 (Supplementary Fig. S6), and differentially expressed genes between CRPC, NEPC, and TCGA were identified using limma135. Based on the differentially expressed genes, CRPC and NEPC scores for TCGA were calculated using the gsva package136. Then, using the gene sets of six functional modules and the gsva package, sample scores for TCGA, CRPC, and NEPC were calculated. Finally, the relationship between CRPC, NEPC, and functional module 6 was determined using fgsea137.
Analysis of the relationship between prostate cancer lineage plasticity and RTK/RAS signaling pathways
Based on the GSEA results, clusters associated with stem cell function were selected. Stemness-related genes66,67, multi-lineage state-related genes7, 10 classic cancer pathways16, and the GSEA algorithm were used to analyze the relationship between stem cell function-related clusters and both prostate cancer lineage plasticity and classic cancer pathways.
Firstly, the ssGSEA algorithm131 was used to quantify the plasticity phenotype scores of prostate cancer samples from the TCGA database based on the prostate cancer lineage plasticity gene set7,65,66. Then, we employed WGCNA138 to compute the relationship between the multi-lineage state phenotype of these samples and the gene set of Function 6 module (stemness module). We retained the positive correlation values between the two. And finally, the relationship between RTK/RAS signaling pathways16,139,140 and prostate cancer lineage plasticity was analyzed employing GSEA.
Analysis of the transcriptional expression, copy number variation, and transcription factor activity of leading-edge genes
Based on the GSEA analysis results for stemness, shTP53/RB1 + Enz, and RTK/RAS, 19 leading-edge genes (LEGs) associated with prostate cancer lineage plasticity and the RTK/RAS signaling pathway were obtained. Single-cell transcriptomic data were retrieved from the SRA database, and the single-cell expression matrix was produced using cutadapt118, hisat2141, and featureCounts142. Cell population analysis was subsequently conducted using Seurat143. Based on the analysis results of prostate cancer transcriptomic data from the TCGA database and single-cell transcriptomic data, the expression profiles of the LEGs were displayed. The expression profile of LEGs in prostate cancer cells was investigated based on the transcriptome data from TCGA prostate cancer database and single-cell transcriptome data68. The copy number variation of LEGs in prostate cancer is explored based on the copy number variation data in the TCGA and SU2C databases. TFs that can influence the expression of LEGs were obtained from the GPSAdb database88. The activity of these TFs was inferred using the reconstruction of accurate cellular network algorithm based on the ARACNe-AP adaptive partitioning strategy144 and the analytic rank-based enrichment analysis algorithm based on VIPER scores145.
Ethics approval and consent to participate
This study was approved by the Ethics Committee of the Wuxi No.2 People's Hospital (NO: 2023Y-27).
Data availability
The original contributions presented in the study are included in the article/supplementary data, further inquiries can be directed to the corresponding author.
Abbreviations
- CNVs:
-
Copy number variations
- AR:
-
Androgen receptor
- NEPC:
-
Neuroendocrine Prostate Cancer
- CRPC:
-
Castration-resistant prostate cancer
- MSPC:
-
Mesenchymal and stem-like prostate cancer cells
- EMT:
-
Epithelial-mesenchymal transition
- DEGs:
-
Differentially expressed genes
- LEGs:
-
Leading-edge genes
References
Siegel, R. L., Miller, K. D. & Jemal, A. Cancer statistics, 2019. CA Cancer J. Clin. 69, 7–34. https://doi.org/10.3322/caac.21551 (2019).
Bray, F. et al. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J. Clin. 68, 394–424. https://doi.org/10.3322/caac.21492 (2018).
Pomerantz, M. M. et al. The androgen receptor cistrome is extensively reprogrammed in human prostate tumorigenesis. Nat. Genet. 47, 1346–1351. https://doi.org/10.1038/ng.3419 (2015).
Tan, M. H. et al. Androgen receptor: Structure, role in prostate cancer and drug discovery. Acta Pharmacol. Sin. 36, 3–23. https://doi.org/10.1038/aps.2014.18 (2015).
Beltran, H. et al. Molecular characterization of neuroendocrine prostate cancer and identification of new drug targets. Cancer Discov. 1, 487–495. https://doi.org/10.1158/2159-8290.CD-11-0130 (2011).
Mu, P. et al. SOX2 promotes lineage plasticity and antiandrogen resistance in TP53- and RB1-deficient prostate cancer. Science 355, 84–88. https://doi.org/10.1126/science.aah4307 (2017).
Deng, S. et al. Ectopic JAK-STAT activation enables the transition to a stem-like and multilineage state conferring AR-targeted therapy resistance. Nat. Cancer 3, 1071–1087. https://doi.org/10.1038/s43018-022-00431-9 (2022).
Ku, S. Y. et al. Rb1 and Trp53 cooperate to suppress prostate cancer lineage plasticity, metastasis, and antiandrogen resistance. Science 355, 78–83. https://doi.org/10.1126/science.aah4199 (2017).
Chan, J. M. et al. Lineage plasticity in prostate cancer depends on JAK/STAT inflammatory signaling. Science 377, 1180–1191. https://doi.org/10.1126/science.abn0478 (2022).
Han, M. et al. FOXA2 drives lineage plasticity and KIT pathway activation in neuroendocrine prostate cancer. Cancer Cell 40, 1306–1323. https://doi.org/10.1016/j.ccell.2022.10.011 (2022).
Tabrizian, N. et al. ASCL1 is activated downstream of the ROR2/CREB signaling pathway to support lineage plasticity in prostate cancer. Cell Rep. 42, 112937. https://doi.org/10.1016/j.celrep.2023.112937 (2023).
Cyrta, J. et al. Role of specialized composition of SWI/SNF complexes in prostate cancer lineage plasticity. Nat. Commun. 11, 5549. https://doi.org/10.1038/s41467-020-19328-1 (2020).
Storck, W. K. et al. The role of epigenetic change in therapy-induced neuroendocrine prostate cancer lineage plasticity. Front. Endocrinol. (Lausanne) 13, 926585. https://doi.org/10.3389/fendo.2022.926585 (2022).
Shaurova, T., Zhang, L., Goodrich, D. W. & Hershberger, P. A. Understanding lineage plasticity as a path to targeted therapy failure in EGFR-mutant non-small cell lung cancer. Front. Genet. 11, 281. https://doi.org/10.3389/fgene.2020.00281 (2020).
Christin, J. R. et al. Stem cell determinant SOX9 promotes lineage plasticity and progression in basal-like breast cancer. Cell Rep. 31, 107742. https://doi.org/10.1016/j.celrep.2020.107742 (2020).
Sanchez-Vega, F. et al. Oncogenic signaling pathways in the cancer genome Atlas. Cell 173, 321–337. https://doi.org/10.1016/j.cell.2018.03.035 (2018).
Mulholland, D. J. et al. Pten loss and RAS/MAPK activation cooperate to promote EMT and metastasis initiated from prostate cancer stem/progenitor cells. Cancer Res. 72, 1878–1889. https://doi.org/10.1158/0008-5472.CAN-11-3132 (2012).
Cheng, C. et al. Gremlin1 is a therapeutically targetable FGFR1 ligand that regulates lineage plasticity and castration resistance in prostate cancer. Nat. Cancer 3, 565–580. https://doi.org/10.1038/s43018-022-00380-3 (2022).
Adissu, H. A. et al. Timp3 loss accelerates tumour invasion and increases prostate inflammation in a mouse model of prostate cancer. Prostate 75, 1831–1843. https://doi.org/10.1002/pros.23056 (2015).
Mostaghel, E. A. Steroid hormone synthetic pathways in prostate cancer. Transl. Androl. Urol. 2, 212–227. https://doi.org/10.3978/j.issn.2223-4683.2013.09.16 (2013).
Aggarwal, S., Singh, M., Kumar, A. & Mukhopadhyay, T. SRD5A2 gene expression inhibits cell migration and invasion in prostate cancer cell line via F-actin reorganization. Mol. Cell Biochem. 408, 15–23. https://doi.org/10.1007/s11010-015-2478-z (2015).
Das, D. K. et al. miR-1207-3p regulates the androgen receptor in prostate cancer via FNDC1/fibronectin. Exp. Cell Res. 348, 190–200. https://doi.org/10.1016/j.yexcr.2016.09.021 (2016).
Massillo, C. et al. CTBP1/CYP19A1/estradiol axis together with adipose tissue impacts over prostate cancer growth associated to metabolic syndrome. Int. J. Cancer 144, 1115–1127. https://doi.org/10.1002/ijc.31773 (2019).
Liu, W. et al. SGK1 inhibition-induced autophagy impairs prostate cancer metastasis by reversing EMT. J. Exp. Clin. Cancer Res. 37, 73. https://doi.org/10.1186/s13046-018-0743-1 (2018).
Wu, L. et al. Upregulation of SCNN1A promotes cell proliferation, migration, and predicts poor prognosis in ovarian cancer through regulating epithelial-mesenchymal transformation. Cancer Biother. Radiopharm. 34, 642–649. https://doi.org/10.1089/cbr.2019.2824 (2019).
Paysan, L., Piquet, L., Saltel, F. & Moreau, V. Rnd3 in cancer: A review of the evidence for tumor promoter or suppressor. Mol. Cancer Res. 14, 1033–1044. https://doi.org/10.1158/1541-7786.MCR-16-0164 (2016).
Migita, T. et al. Fatty acid synthase: A metabolic enzyme and candidate oncogene in prostate cancer. J. Natl. Cancer Inst. 101, 519–532. https://doi.org/10.1093/jnci/djp030 (2009).
Huang, Y. et al. Verification of cell cycle-associated cyclin-dependent kinases facilitated prostate cancer progression by integrated bioinformatic analysis and experimental validation. Heliyon 8, e10081. https://doi.org/10.1016/j.heliyon.2022.e10081 (2022).
Hu, J. et al. Sox5 contributes to prostate cancer metastasis and is a master regulator of TGF-beta-induced epithelial mesenchymal transition through controlling Twist1 expression. Br. J. Cancer 118, 88–97. https://doi.org/10.1038/bjc.2017.372 (2018).
Fan, L., Li, H. & Wang, W. Long non-coding RNA PRRT3-AS1 silencing inhibits prostate cancer cell proliferation and promotes apoptosis and autophagy. Exp. Physiol. 105, 793–808. https://doi.org/10.1113/EP088011 (2020).
Zhang, W. et al. The integration of single-cell sequencing, TCGA, and GEO data analysis revealed that PRRT3-AS1 is a biomarker and therapeutic target of SKCM. Front. Immunol. 13, 919145. https://doi.org/10.3389/fimmu.2022.919145 (2022).
Liu, S. et al. Identification of potential key genes for pathogenesis and prognosis in prostate cancer by integrated analysis of gene expression profiles and the cancer genome Atlas. Front. Oncol. 10, 809. https://doi.org/10.3389/fonc.2020.00809 (2020).
Huang, L. et al. Comprehensive characterization of ageing-relevant subtypes associated with different tumorigenesis and tumor microenvironment in prostate cancer. Front. Mol. Biosci. 9, 803474. https://doi.org/10.3389/fmolb.2022.803474 (2022).
Huang, H. et al. The influence of lncRNAs on the prognosis of prostate cancer based on TCGA database. Transl. Androl. Urol. 10, 1302–1313. https://doi.org/10.21037/tau-21-154 (2021).
Chen, G. et al. Netrin-1 receptor UNC5C cleavage by active delta-secretase enhances neurodegeneration, promoting Alzheimer’s disease pathologies. Sci. Adv. 7, eabe4499. https://doi.org/10.1126/sciadv.abe4499 (2021).
Jiang, X. et al. An immune-lncRNA risk model to predict prognosis for patients with head and neck squamous cell carcinoma. Preprint at https://https://doi.org/10.1101/2022.03.10.483771 (2022).
Westaby, D. et al. Abstract B020: BCL2 expression is enriched in AR-independent advanced prostate cancer. Cancer Res. 83, B020–B020. https://doi.org/10.1158/1538-7445.prca2023-b020 (2023).
Alfahed, A. et al. Prognostic values of gene copy number alterations in prostate cancer. Genes https://doi.org/10.3390/genes14050956 (2023).
Faisal, F. A. et al. CDKN1B deletions are associated with metastasis in African American men with clinically localized, surgically treated prostate cancer. Clin. Cancer Res. 26, 2595–2602. https://doi.org/10.1158/1078-0432.CCR-19-1669 (2020).
Daniunaite, K. et al. Promoter methylation of PRKCB, ADAMTS12, and NAALAD2 is specific to prostate cancer and predicts biochemical disease recurrence. Int. J. Mol. Sci. 22, 6091. https://doi.org/10.3390/ijms22116091 (2021).
Ai, J. et al. Vinculin and filamin-C are two potential prognostic biomarkers and therapeutic targets for prostate cancer cell migration. Oncotarget 8, 82430–82436. https://doi.org/10.18632/oncotarget.19397 (2017).
Feferman, L. et al. Arylsulfatase B (N-acetylgalactosamine-4-sulfatase): Potential role as a biomarker in prostate cancer. Prostate Cancer Prostatic Dis. 16, 277–284. https://doi.org/10.1038/pcan.2013.18 (2013).
Richardson, D. R. Iron and neoplasia: Serum transferrin receptor and ferritin in prostate cancer. J. Lab. Clin. Med. 144, 173–175. https://doi.org/10.1016/j.lab.2004.05.016 (2004).
Peng, J. et al. Adenoviral vector for enhanced prostate cancer specific transferrin conjugated drug targeted therapy. Nano Lett. 22, 4168–4175. https://doi.org/10.1021/acs.nanolett.2c00931 (2022).
Chen, Y. et al. APC gene hypermethylation and prostate cancer: A systematic review and meta-analysis. Eur. J. Hum. Genet. 21, 929–935. https://doi.org/10.1038/ejhg.2012.281 (2013).
Wise, H. M., Hermida, M. A. & Leslie, N. R. Prostate cancer, PI3K, PTEN and prognosis. Clin. Sci. (Lond.) 131, 197–210. https://doi.org/10.1042/CS20160026 (2017).
Ecke, T. H. et al. TP53 gene mutations in prostate cancer progression https://www.ncbi.nlm.nih.gov/pubmed/20592345 (2010).
Nyberg, T., Tischkowitz, M. & Antoniou, A. C. BRCA1 and BRCA2 pathogenic variants and prostate cancer risk: Systematic review and meta-analysis. Br. J. Cancer 126, 1067–1081. https://doi.org/10.1038/s41416-021-01675-5 (2022).
Khan, S., Ayub, H., Khan, T. & Wahid, F. MicroRNA biogenesis, gene silencing mechanisms and role in breast, ovarian and prostate cancer. Biochimie 167, 12–24. https://doi.org/10.1016/j.biochi.2019.09.001 (2019).
Vava, A. Characterising the mechanism of DCUN1D1 activity in prostate cancer and identifying DCUN1D1 inhibitors for prostate cancer treatment http://hdl.handle.net/11427/37438 (2020).
Karthikeyan, S. K. et al. Identification of androgen response-related lncRNAs in prostate cancer. Prostate 83, 590–601. https://doi.org/10.1002/pros.24494 (2023).
Shi, J. et al. LPAR1, correlated with immune infiltrates, is a potential prognostic biomarker in prostate cancer. Front. Oncol. 10, 846. https://doi.org/10.3389/fonc.2020.00846 (2020).
Li, C. et al. Integrated network analysis to determine CNN1, MYL9, TAGLN, and SORBS1 as potential key genes associated with prostate cancer. Clin. Lab. https://doi.org/10.7754/Clin.Lab.2023.220224 (2023).
Cai, C. et al. ADAM10-cleaved ephrin-A5 contributes to prostate cancer metastasis. Cell Death Dis. 13, 453. https://doi.org/10.1038/s41419-022-04893-8 (2022).
Kong, R. et al. COL14A1 promotes self-renewal of human liver cancer stem cells through activation of ERK signaling. J. Bio-X Res. 4, 10–17. https://doi.org/10.1097/jbr.0000000000000087 (2021).
Flores-Perez, A. et al. Dual targeting of ANGPT1 and TGFBR2 genes by miR-204 controls angiogenesis in breast cancer. Sci. Rep. 6, 34504. https://doi.org/10.1038/srep34504 (2016).
Xu, B. et al. TXNDC5 is a cervical tumor susceptibility gene that stimulates cell migration, vasculogenic mimicry and angiogenesis by down-regulating SERPINF1 and TRAF1 expression. Oncotarget 8, 91009–91024. https://doi.org/10.18632/oncotarget.18857 (2017).
Meng, S. et al. GJA1 expression and its prognostic value in cervical cancer. Biomed. Res. Int. 2020, 8827920. https://doi.org/10.1155/2020/8827920 (2020).
Feng, Y. et al. Potential functional variants in SMC2 and TP53 in the AURORA pathway genes and risk of pancreatic cancer. Carcinogenesis 40, 521–528. https://doi.org/10.1093/carcin/bgz029 (2019).
Zhang, J. et al. A novel amplification gene PCI domain containing 2 (PCID2) promotes colorectal cancer through directly degrading a tumor suppressor promyelocytic leukemia (PML). Oncogene 40, 6641–6652. https://doi.org/10.1038/s41388-021-01941-z (2021).
Chang, S. C. et al. Human FBXL8 Is a Novel E3 ligase which promotes BRCA metastasis by stimulating pro-tumorigenic cytokines and inhibiting tumor suppressors. Cancer 12, 2210. https://doi.org/10.3390/cancers12082210 (2020).
Raducu, M. et al. SCF (Fbxl17) ubiquitylation of Sufu regulates Hedgehog signaling and medulloblastoma development. EMBO J. 35, 1400–1416. https://doi.org/10.15252/embj.201593374 (2016).
Miano, J. M. Myocardin in biology and disease. J. Biomed. Res. 29, 3–19. https://doi.org/10.7555/JBR.29.20140151 (2015).
Nowak, M. et al. DCAF8, a novel MuRF1 interaction partner, promotes muscle atrophy. J. Cell Sci. 132, jcs233395. https://doi.org/10.1242/jcs.233395 (2019).
Dai, Y. & Gao, X. Inhibition of cancer cell-derived exosomal microRNA-183 suppresses cell growth and metastasis in prostate cancer by upregulating TPM1. Cancer Cell Int. 21, 145. https://doi.org/10.1186/s12935-020-01686-x (2021).
Miranda, A. et al. Cancer stemness, intratumoral heterogeneity, and immune response across cancers. Proc. Natl. Acad. Sci. 116, 9020. https://doi.org/10.1073/pnas.1818210116 (2019).
Dai, Z. & Liu, P. High copy number variations, particular transcription factors, and low immunity contribute to the stemness of prostate cancer cells. J. Transl. Med. 19, 206. https://doi.org/10.1186/s12967-021-02870-x (2021).
Horning, A. M. et al. Single-cell RNA-seq reveals a subpopulation of prostate cancer cells with enhanced cell-cycle-related transcription and attenuated androgen response. Cancer Res. 78, 853–864. https://doi.org/10.1158/0008-5472.CAN-17-1924 (2018).
Fu, X. et al. Overexpression of BUB1B contributes to progression of prostate cancer and predicts poor outcome in patients with prostate cancer. Oncol. Targets Ther. 9, 2211–2220. https://doi.org/10.2147/OTT.S101994 (2016).
Pallasaho, S. et al. Castration-resistant prostate cancer cells are dependent on the high activity of CDK7. J. Cancer Res. Clin. Oncol 149, 5255–5263. https://doi.org/10.1007/s00432-022-04475-3 (2023).
Lin, S. C. et al. Dysregulation of miRNAs-COUP-TFII-FOXM1-CENPF axis contributes to the metastasis of prostate cancer. Nat. Commun. 7, 11418. https://doi.org/10.1038/ncomms11418 (2016).
Liu, Y. et al. FOXM1 promotes the progression of prostate cancer by regulating PSA gene transcription. Oncotarget 8, 17027–17037. https://doi.org/10.18632/oncotarget.15224 (2017).
Piao, X. M. et al. Kinesin family member 11 mRNA expression predicts prostate cancer aggressiveness. Clin. Genitourin. Cancer 15, 450–454. https://doi.org/10.1016/j.clgc.2016.10.005 (2017).
Wang, H. et al. KIF11: A potential prognostic biomarker for predicting bone metastasis-free survival of prostate cancer. Oncol. Lett. 24, 312. https://doi.org/10.3892/ol.2022.13432 (2022).
Lin, Y. et al. ORC6, a novel prognostic biomarker, correlates with T regulatory cell infiltration in prostate adenocarcinoma: A pan-cancer analysis. BMC Cancer 23, 285. https://doi.org/10.1186/s12885-023-10763-z (2023).
Warren, A. Y. et al. A reciprocal feedback between the PDZ binding kinase and androgen receptor drives prostate cancer. Oncogene 38, 1136–1150. https://doi.org/10.1038/s41388-018-0501-z (2019).
Mitra, A. et al. Overexpression of RAD51 occurs in aggressive prostatic cancer. Histopathology 55, 696–704. https://doi.org/10.1111/j.1365-2559.2009.03448.x (2009).
Chen, S. et al. Silencing TTK expression inhibits the proliferation and progression of prostate cancer. Exp. Cell Res. 385, 111669. https://doi.org/10.1016/j.yexcr.2019.111669 (2019).
Liu, B. et al. A novel androgen-induced lncRNA FAM83H-AS1 promotes prostate cancer progression via the miR-15a/CCNE2 axis. Front. Oncol. 10, 620306. https://doi.org/10.3389/fonc.2020.620306 (2020).
Dong, H. et al. SOX4 is activated by C-MYC in prostate cancer. Med. Oncol. 36, 92. https://doi.org/10.1007/s12032-019-1317-6 (2019).
Yi, M. et al. Increased expression of ECT2 predicts the poor prognosis of breast cancer patients. Exp. Hematol. Oncol. 11, 107. https://doi.org/10.1186/s40164-022-00361-3 (2022).
Peng, L. et al. GINS2 regulates matrix metallopeptidase 9 expression and cancer stem cell property in human triple negative Breast cancer. Biomed. Pharmacother. 84, 1568–1574. https://doi.org/10.1016/j.biopha.2016.10.032 (2016).
Lu, Z. et al. Partitioning defective 6 homolog alpha (PARD6A) promotes epithelial-mesenchymal transition via integrin beta1-ILK-SNAIL1 pathway in ovarian cancer. Cell Death Dis. 13, 304. https://doi.org/10.1038/s41419-022-04756-2 (2022).
Pires, E., Sung, P. & Wiese, C. Role of RAD51AP1 in homologous recombination DNA repair and carcinogenesis. DNA Repair (Amst) 59, 76–81. https://doi.org/10.1016/j.dnarep.2017.09.008 (2017).
Kim, Y. R. et al. Mutational and expressional analysis of RFC3, a clamp loader in DNA replication, in gastric and colorectal cancers. Hum. Pathol. 41, 1431–1437. https://doi.org/10.1016/j.humpath.2010.03.006 (2010).
Zhang, J., Wang, L. & Xie, X. RFC4 promotes the progression and growth of Oral Tongue squamous cell carcinoma in vivo and vitro. J. Clin. Lab. Anal. 35, e23761. https://doi.org/10.1002/jcla.23761 (2021).
Dharmapal, D. et al. Beta-tubulin isotype, TUBB4B, regulates the maintenance of cancer stem cells. Front. Oncol. 11, 788024. https://doi.org/10.3389/fonc.2021.788024 (2021).
Guo, S. et al. GPSAdb: A comprehensive web resource for interactive exploration of genetic perturbation RNA-seq datasets. Nucleic Acids Res. 51, D964–D968. https://doi.org/10.1093/nar/gkac1066 (2023).
Wang, J. et al. ROR-gamma drives androgen receptor expression and represents a therapeutic target in castration-resistant prostate cancer. Nat. Med. 22, 488–496. https://doi.org/10.1038/nm.4070 (2016).
Waltering, K. K., Urbanucci, A. & Visakorpi, T. Androgen receptor (AR) aberrations in castration-resistant prostate cancer. Mol. Cell Endocrinol. 360, 38–43. https://doi.org/10.1016/j.mce.2011.12.019 (2012).
Lovnicki, J. et al. LIN28B promotes the development of neuroendocrine prostate cancer. J. Clin. Invest. 130, 5338–5348. https://doi.org/10.1172/JCI135373 (2020).
Erb, H. H. et al. IL6 sensitizes prostate cancer to the antiproliferative effect of IFNalpha2 through IRF9. Endocr. Relat. Cancer 20, 677–689. https://doi.org/10.1530/ERC-13-0222 (2013).
Daremipouran, M. et al. NKX2-5, a potential tumor suppressor gene in prostate cancer. Cancer Res. 71, 3033–3033. https://doi.org/10.1158/1538-7445.am2011-3033 (2011).
Deveci Ozkan, A. et al. Anti-inflammatory effects of nobiletin on TLR4/TRIF/IRF3 and TLR9/IRF7 signaling pathways in prostate cancer cells. Immunopharmacol. Immunotoxicol. 42, 93–100. https://doi.org/10.1080/08923973.2020.1725040 (2020).
Gambara, G. et al. TLR3 engagement induces IRF-3-dependent apoptosis in androgen-sensitive prostate cancer cells and inhibits tumour growth in vivo. J. Cell Mol. Med. 19, 327–339. https://doi.org/10.1111/jcmm.12379 (2015).
Lin, T. C. RUNX1 and cancer. Biochim. Biophys. Acta Rev. Cancer 1877, 188715. https://doi.org/10.1016/j.bbcan.2022.188715 (2022).
Chang, G. T. et al. The TRPS1 transcription factor: Androgenic regulation in prostate cancer and high expression in breast cancer. Endocr. Relat. Cancer 11, 815–822. https://doi.org/10.1677/erc.1.00853 (2004).
Zhang, H. et al. FOXO1 inhibits Runx2 transcriptional activity and prostate cancer cell migration and invasion. Cancer Res. 71, 3257–3267. https://doi.org/10.1158/0008-5472.CAN-10-2603 (2011).
Zhu, C. et al. ITGB3/CD61: A hub modulator and target in the tumor microenvironment https://www.ncbi.nlm.nih.gov/pubmed/31934272 (2019).
Zhu, H., Lin, Q., Gao, X. & Huang, X. Identification of the hub genes associated with prostate cancer tumorigenesis. Front. Oncol. 13, 1168772. https://doi.org/10.3389/fonc.2023.1168772 (2023).
Aytes, A. et al. Cross-species regulatory network analysis identifies a synergistic interaction between FOXM1 and CENPF that drives prostate cancer malignancy. Cancer Cell 25, 638–651. https://doi.org/10.1016/j.ccr.2014.03.017 (2014).
Shrivastava, G. et al. Targeting LIN28: A new hope in prostate cancer theranostics. Fut. Oncol. 17, 3873–3880. https://doi.org/10.2217/fon-2021-0247 (2021).
Chang, G. T. et al. Proteomic analysis of proteins regulated by TRPS1 transcription factor in DU145 prostate cancer cells. Biochim. Biophys. Acta 1774, 575–582. https://doi.org/10.1016/j.bbapap.2007.03.011 (2007).
Bolger, A. M., Lohse, M. & Usadel, B. Trimmomatic: A flexible trimmer for Illumina sequence data. Bioinformatics 30, 2114–2120. https://doi.org/10.1093/bioinformatics/btu170 (2014).
Gordon, A. & J. Hannon, G. FASTQ/A short-reads pre-processing tools http://hannonlab.cshl.edu/fastx_toolkit/index.html (2010).
Andrews, S. FASTQC. A quality control tool for high throughput sequence data https://www.bioinformatics.babraham.ac.uk/projects/fastqc/ (2010).
Langmead, B. & Salzberg, S. L. Fast gapped-read alignment with Bowtie 2. Nat. Methods 9, 357–359. https://doi.org/10.1038/nmeth.1923 (2012).
Zhang, Y. et al. Model-based Analysis of ChIP-Seq (MACS). Genome Biol. 9, R137. https://doi.org/10.1186/gb-2008-9-9-r137 (2008).
Heinz, S. et al. Simple combinations of lineage-determining transcription factors prime cis-regulatory elements required for macrophage and B cell identities. Mol. Cell 38, 576–589. https://doi.org/10.1016/j.molcel.2010.05.004 (2010).
Robinson, M. D., McCarthy, D. J. & Smyth, G. K. edgeR: A Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 26, 139–140. https://doi.org/10.1093/bioinformatics/btp616 (2010).
Yu, G., Wang, L. G., Han, Y. & He, Q. Y. clusterProfiler: An R package for comparing biological themes among gene clusters. OMICS 16, 284–287. https://doi.org/10.1089/omi.2011.0118 (2012).
Kanehisa, M. et al. KEGG for taxonomy-based analysis of pathways and genomes. Nucleic Acids Res. 51, D587–D592. https://doi.org/10.1093/nar/gkac963 (2023).
Terry M. Therneau & Patricia M. Grambsch. Modeling Survival Data: Extending the Cox Model. Vol. 350 1431–8776 (Springer, 2000).
Heagerty, P. J., Lumley, T. & Pepe, M. S. Time-dependent ROC curves for censored survival data and a diagnostic marker. Biometrics 56, 337–344. https://doi.org/10.1111/j.0006-341X.2000.00337.x (2000).
Sharma, N. V. et al. Identification of the transcription factor relationships associated with androgen deprivation therapy response and metastatic progression in prostate cancer. Cancers. https://doi.org/10.3390/cancers10100379 (2018).
Wagner, A. H. et al. GenVisR: Genomic visualizations in R. Bioinformatics 32, 3012–3014. https://doi.org/10.1093/bioinformatics/btw325 (2016).
Rhie, S. K. et al. A high-resolution 3D epigenomic map reveals insights into the creation of the prostate cancer transcriptome. Nat. Commun. 10, 4154. https://doi.org/10.1038/s41467-019-12079-8 (2019).
Martin, M. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet.journal 17, 10. https://doi.org/10.14806/ej.17.1.200 (2011).
Servant, N. et al. HiC-Pro: An optimized and flexible pipeline for Hi-C data processing. Genome Biol. 16, 259. https://doi.org/10.1186/s13059-015-0831-x (2015).
Durand, N. C. et al. Juicer provides a one-click system for analyzing loop-resolution Hi-C experiments. Cell Syst. 3, 95–98. https://doi.org/10.1016/j.cels.2016.07.002 (2016).
Zhang, H., Meltzer, P. & Davis, S. RCircos: An R package for Circos 2D track plots. BMC Bioinform. 14, 1–5. https://doi.org/10.1186/1471-2105-14-244 (2013).
Gevaert, O. MethylMix: An R package for identifying DNA methylation-driven genes. Bioinformatics 31, 1839–1841. https://doi.org/10.1093/bioinformatics/btv020 (2015).
Shannon, P. et al. Cytoscape: A software environment for integrated models of biomolecular interaction networks. Genome Res. 13, 2498–2504. https://doi.org/10.1101/gr.1239303 (2003).
Livak, K. J. & Schmittgen, T. D. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 25, 402–408v. https://doi.org/10.1006/meth.2001.1262 (2001).
Cohen, J. A coefficient of agreement for nominal scales. Educ. Psychol. Meas. 20, 37–46. https://doi.org/10.1177/001316446002000104 (2016).
Northcott, P. A. et al. The whole-genome landscape of medulloblastoma subtypes. Nature 547, 311–317. https://doi.org/10.1038/nature22973 (2017).
Huang, D. W. et al. The DAVID Gene Functional Classification Tool: A novel biological module-centric algorithm to functionally analyze large gene lists. Genome Biol. 8, R183. https://doi.org/10.1186/gb-2007-8-9-r183 (2007).
Saitou, N. & Nei, M. The neighbor-joining method: A new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4, 406–425. https://doi.org/10.1093/oxfordjournals.molbev.a040454 (1987).
Yu, G. et al. ggtree: An r package for visualization and annotation of phylogenetic trees with their covariates and other associated data. Methods Ecol. Evol. 8, 28–36. https://doi.org/10.1111/2041-210X.12628 (2017).
Lex, A. et al. UpSet: Visualization of intersecting sets. IEEE Trans. Vis. Comput. Graphics 20, 1983–1992. https://doi.org/10.1109/TVCG.2014.2346248 (2014).
Subramanian, A. et al. Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles. Proc. Natl. Acad. Sci. U S A 102, 15545–15550. https://doi.org/10.1073/pnas.0506580102 (2005).
Abida, W. et al. Genomic correlates of clinical outcome in advanced prostate cancer. Proc. Natl. Acad. Sci. U S A 116, 11428–11436. https://doi.org/10.1073/pnas.1902651116 (2019).
Beltran, H. et al. Divergent clonal evolution of castration-resistant neuroendocrine prostate cancer. Nat. Med. 22, 298–305. https://doi.org/10.1038/nm.4045 (2016).
Gao, J. et al. Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal. Sci. Signal 6, pl1. https://doi.org/10.1126/scisignal.2004088 (2013).
Ritchie, M. E. et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 43, e47. https://doi.org/10.1093/nar/gkv007 (2015).
Hanzelmann, S., Castelo, R. & Guinney, J. GSVA: Gene set variation analysis for microarray and RNA-seq data. BMC Bioinform. 14, 7. https://doi.org/10.1186/1471-2105-14-7 (2013).
Korotkevich, G. et al. Fast gene set enrichment analysis https://github.com/ctlab/fgsea (2021).
Langfelder, P. & Horvath, S. WGCNA: An R package for weighted correlation network analysis. BMC Bioinform. 9, 559. https://doi.org/10.1186/1471-2105-9-559 (2008).
Liberzon, A. et al. The molecular signatures database (MSigDB) hallmark gene set collection. Cell Syst. 1, 417–425. https://doi.org/10.1016/j.cels.2015.12.004 (2015).
Way, G. P. et al. Machine learning detects pan-cancer Ras pathway activation in the cancer genome Atlas. Cell Rep. 23, 172–180. https://doi.org/10.1016/j.celrep.2018.03.046 (2018).
Kim, D. et al. Graph-based genome alignment and genotyping with HISAT2 and HISAT-genotype. Nat. Biotechnol. 37, 907–915. https://doi.org/10.1038/s41587-019-0201-4 (2019).
Liao, Y., Smyth, G. K. & Shi, W. featureCounts: An efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics 30, 923–930. https://doi.org/10.1093/bioinformatics/btt656 (2014).
Hao, Y. et al. Integrated analysis of multimodal single-cell data. Cell 184, 3573–3587. https://doi.org/10.1016/j.cell.2021.04.048 (2021).
Lachmann, A., Giorgi, F. M., Lopez, G. & Califano, A. ARACNe-AP: Gene network reverse engineering through adaptive partitioning inference of mutual information. Bioinformatics 32, 2233–2235. https://doi.org/10.1093/bioinformatics/btw216 (2016).
Alvarez, M. J. et al. Functional characterization of somatic mutations in cancer using network-based inference of protein activity. Nat. Genet. 48, 838–847. https://doi.org/10.1038/ng.3593 (2016).
Acknowledgements
We will give our sincere gratitude to the reviewers for their valuable suggestions; and we give our great thanks to NIH and TCGA database to provide us related data. We would like to thank Dr. Daniel Petticord at the University of Cornell for his assistance with English language and grammatical editing of the manuscript.
Funding
This research was funded by the National Natural Science Foundation of China (NNSFC) (Grant No. 32200676), Project of Wuxi Health Committee (Q202229) and General Program of Maternal and Child Health Research of Wuxi Health Commission (FYKY202304).
Author information
Authors and Affiliations
Contributions
W.G., Z.X., L.S., H.W., and D.Z. conceived the study. W.G. and Z.X. collected omics data and conducted analysis; W.G., Z.X., and L.S. visualized diagrams; W.G., H.W., and D.Z. wrote and revised the draft; D.Z. supervised the project progress.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Wei, G., Zhang, X., Liu, S. et al. Comprehensive data mining reveals RTK/RAS signaling pathway as a promoter of prostate cancer lineage plasticity through transcription factors and CNV. Sci Rep 14, 11688 (2024). https://doi.org/10.1038/s41598-024-62256-z
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-024-62256-z
- Springer Nature Limited