Skip to main content
Download PDF

Integrative analysis regarding the correlation between collagen-related genes and prostate cancer

BMC Cancer volume 24, Article number: 1038 (2024) Cite this article

Abstract

Purpose

Prostate cancer (PCa) is a common malignancy in men, with an escalating mortality rate attributed to Recurrence and metastasis. Recent studies have illuminated collagen’s critical regulatory role within the tumor microenvironment, significantly influencing tumor progression. Accordingly, this investigation is dedicated to examining the relationship between genes linked to collagen and the prognosis of PCa, with the objective of uncovering any possible associations between them.

Methods

Gene expression data for individuals with prostate cancer were obtained from the TCGA repository. Collagen-related genes were identified, leading to the development of a risk score model associated with biochemical recurrence-free survival (BRFS). A prognostic nomogram integrating the risk score with essential clinical factors was crafted and evaluated for efficacy. The influence of key collagen-related genes on cellular behavior was confirmed through various assays, including CCK8, invasion, migration, cell cloning, and wound healing. Immunohistochemical detection was used to evaluate PLOD3 expression in prostate cancer tissue samples.

Results

Our study identified four key collagen-associated genes (PLOD3, COL1A1, MMP11, FMOD) as significant. Survival analysis revealed that low-risk groups, based on the risk scoring model, had significantly improved prognoses. The risk score was strongly associated with prostate cancer prognosis. Researchers then created a nomogram, which demonstrated robust predictive efficacy and substantial clinical applicability.Remarkably, the suppression of PLOD3 expression notably impeded the proliferation, invasion, migration, and colony formation capabilities of PCa cells.

Conclusion

The risk score, derived from four collagen-associated genes, could potentially act as a precise prognostic indicator for BRFS of patients. Simultaneously, our research has identified potential therapeutic targets related to collagen. Notably, PLOD3 was differentially expressed in cancer and para-cancer tissues in clinical specimens and it also was validated through in vitro studies and shown to suppress PCa tumorigenesis following its silencing.

Peer Review reports

Introduction

Prostate cancer, a malignancy originating in prostate gland epithelial cells, predominantly affects older males. [1].In 2020, the estimated global death toll exceeded 375,000 individuals, with a total mortality rate of 7.7 per 100,000 [2, 3]. Therapeutic approaches for prostate cancer encompass active surveillance, prostatectomy (surgical removal of the prostate), and radiation therapy [4]. Following radical prostatectomy, patients experiencing disease recurrence are treated with salvage radiotherapy and/or androgen deprivation therapy (ADT) for localized recurrence or a combination of chemotherapy and innovative androgen-targeted therapies for systemic relapse. However, despite castration therapy, most patients eventually progress to castration-resistant prostate cancer (characterized by serum testosterone levels below 50 ng/dl or 1.7 nmol/l, coupled with biochemical recurrence or new lesions) within 18–24 months [5]. Consequently, there is a significant urgency to establish prognostic biomarkers and identify novel therapeutic avenues for this lethal ailment.

Collagen, a vital element of the extracellular matrix (ECM), holds significant sway in the advancement of cancer [6, 7]. Recent research indicates that ECM proteins are vitally important in anti-tumor immunity, with collagen assembly mediated by discoidin domain receptor 1(DDR1) effectively excluding T cells from the tumor microenvironment [8]. ECM has two main functions: firstly, by modulating collagen formation, it can serve as a dense physical barrier, resulting in the immune microenvironment becoming an “immune desert”, in other words, the microenvironment lacks T cells and NK cells; secondly, ECM proteins modulate the functionality of immune cells through direct interaction with ECM receptors on these cells. This is highlighted in laboratory experiments employing co-culture assays with breast tumor cells and murine T lymphocytes on type IV collagen-coated surfaces, where type IV collagen is identified as a robust suppressor of T cell reactions [9]. Prostate cancer is characterized by its reactive stroma, characterized by a dense population of fibroblasts and myoblasts [10]. Nevertheless, the role of collagen-related genes in prostate cancer remains insufficiently investigated.

Recent advancements in oncology research have seen the development of numerous prognostic models, risk scoring systems, and models predicting drug resistance. Song et al. constructed a telomere-associated gene risk model to forecast outcomes in kidney cancer patients, potentially aiding in the selection of appropriate treatment agents [11]; Huang et al. also developed a prognostic model utilizing five genes linked to collagen, demonstrating the capacity to effectively anticipate the prognosis of patients with endometrial carcinoma, with robust performance across the training, testing, and entire datasets [12]; Utilizing a sophisticated methodology, the model developed by Sui et al. has a robust effectiveness in forecasting the treatment reaction and overall prognosis of patients undergoing platinum-based chemotherapy treatments [13]. The aforementioned predictive models have demonstrated considerable predictive efficacy and hold promising prospects for clinical application.

Methods

Acquisition and analysis of data

Data from TCGA (https://portal.gdc.cancer.gov) containing RNA-Seq data and clinical data pertaining to prostate cancer were gathered. The GSEA database (https://www.gsea-msigdb.org/gsea/index.jsp) can provides a dataset of collagen-related genes. Using thresholds of |log2FC| > 0.5 and P < 0.05, we employed the “edgeR” package to generate a volcano plot and identify differentially expressed genes linked with collagen between tumor and adjacent noncancerous tissues. Assessing the correlation between genes associated with collagen and differential genes is accomplished through the application of the Pearson correlation coefficient. Genes linked to collagen and those displaying a correlation coefficient |R2| > 0.5 alongside a P-value < 0.05 are r recognized as substantial and selected for further scrutiny.

Construct and validate risk scores based on prognostic related genes

Participants who had a minimum follow-up period of 30 days were split into two groups, with two-thirds in the training group and one-third in the validation group. Clinical survival data and Genes underwent analysis through univariate Cox regression. These genes associated with prognosis were further scrutinized via LASSO regression analysis. Following multivariate Cox regression analysis of the targeted genes, a risk scoring model was established based on the coefficients obtained. Assessment of the scoring system’s effectiveness was carried out by analyzing the area under the curve (AUC). With the scoring model, the risk score for each patient can be ascertained. Utilizing the median risk score, researchers categorized patients into high and low-risk groups, after which a Kaplan-Meier survival plot was generated to illustrate the variance in outcomes between the two cohorts. The researchers presented the risk score and survival status of each sample through the use of risk curves and scatter plots. Furthermore, heat maps and volcanoes were employed to depict the expression states of chosen genes.

Nomogram

Researchers analyzed risk scores, age, gender, and TNM using both univariate and multivariate Cox methods to discern variables linked to prognosis. Subsequently, a nomogram was crafted, informed by the findings of multivariate Cox analysis. Assessment of the nomogram involved computing the C-index and creating the AUC, calibration plot, and DCA.

Cell culture

The prostate cancer cell lines researchers use are all of human origin, including 22Rv1, as well as a brain metastasis cell line DU145 and a bone metastasis cell line PC-3. All cell lines were obtained from the cell bank of the Chinese Academy of Sciences. The PC-3 and 22Rv1 cells were consistently cultured in RPMI-1640 medium (Gibco), while the DU145 cells were maintained in DMEM (Gibco). Both media were supplemented with 10% fetal bovine serum. The incubation process took place in a CO2 incubator with humidity control at a temperature of 37 °C.

SiRNA transfection

The three cell lines mentioned above were placed in six-well dishes with a concentration of 1 × 105 cells each, then separated into the control and siRNA groups. They were grown in culture until they reached a density of 50-70%. Before being administered to DU145, PC-3, and 22Rv1 cells for siRNA transfection, Control siRNA, PLOD3 siRNA1, and PLOD3 siRNA2 were combined with Lipofectamine RNAi Max (Invitrogen) in Opti-MEM (Sigma) for 20 min. The siRNA primer sequences are available in Supplementary Material 2.

Reverse transcription of complementary DNA followed by qRT-PCR

In order to study the gene expression, the PrimeScript RT Reagent Kit from Takara (Shiga, Japan) was utilized to synthesize complementary DNA (cDNA). We utilized the TB Green Premix Ex Taq II kit (Takara, Shiga, Japan) for RT-qPCR analysis. The cDNA samples were served as the template for this procedure. The 2^-ΔΔCT technique was useful in assessing the comparative levels of expression. GAPDH functioned as a reference for mRNA levels. The primer sequences for PLOD3 and GAPDH are provided in Supplementary Material 2.Cell death/Viability/Proliferation assays.

The viability of DU145 and 22Rv1 cell lines, in both the control and siRNA groups, was evaluated using the CCK-8 assay post transfection. These cell lines were uniformly inoculated in 96-well plates with 2000 cells/Wells, 100ul of complete medium was added to each well, and then transfected for 48 h. Following a 24-hour period after transfection, the CCK-8 assay was performed by configured CCK-8 working fluid. The CCK8 working solution is prepared by mixing 10% CCK-8 solution with 90% complete medium, and finally adding 100ul working solution to each hole. Following a 2-hour incubation, cell growth and health were evaluated by measuring the optical density (OD) at 450 nm using the Tecan Spark 10 M multimode microplate reader (Tecan, Austria). The procedure was conducted thrice with triplicate repetitions. Cell viability percentages were determined using the following formula: (difference in cell number between the experimental and blank group) divided by (difference in cell number between the control and blank group), multiplied by 100.

In colony formation experiments, 2000 cells were distributed into a 6-well plate and cultured. After being cultured for 2–3 weeks, the medium was changed to paraformaldehyde (Servicebio, Wuhan, China) and incubated at ambient temp for thirty minutes. Cells were stained with 0.1% crystal violet at ambient temperature for 15 min and then rinsed twice with PBS. After drying naturally, clones were photographed and analyzed with AID v Spot Spectrum (AID, Germany).

Invasion and migration assay

Evaluation of cell motility and invasiveness involved transwell and wound healing assays. In brief, the processed cells were readied following the instructions provided earlier, utilizing chambers (Falcon, USA) with an 8 μm pore size and Matrigel (BD Science, USA). Migration and invasion studies involved resuspending 5 × 104 cells in 200 µl serum-lacking medium and placing them in the upper chambers, with Matrigel application for invasion assays and exclusion for migration assays. The lower chambers contained 700 ul of medium with 10% serum, and the cells were incubated for the specified duration. After being washed twice with PBS, the cells were treated with paraformaldehyde (Servicebio, Wuhan, China) for 30 min. At ambient temperature, the cells were subjected to staining with 0.1% crystal violet for ten minutes, then swabbed to remove them from the upper chambers. The Leica DM2000 microscope (Leica Camera AG, Wetzlar, Germany) facilitated imaging.

To simulate wound healing, a 200 ul pipette was utilized to create a wound by lining the plate vertically, followed by gently washing the cells with PBS. Initially, pictures were taken using an Olympus IX2 inverted microscope (Olympus, Tokyo, Japan). Cultivation of cells in a serum-deprived medium was carried out for the recommended timeframe and images were taken at the end of time with Olympus IX2 inverted microscope (Olympus, Tokyo, Japan). Analysis was conducted on the movement of cells and the healing of wounds.

Immunohistochemical detection

We performed Immunohistochemical detection (IHC) using commercially available antibodies. Our pathological specimens of four groups of prostate cancer patients were obtained from the Department of Pathology, Sun Yat-sen Memorial Hospital, Sun Yat-sen University. Ethical approval was obtained for the study. IHC was performed using conventional methods. The primary and secondary antibodies we used were from SAB and Abcam. Dilution ratio and reaction times are carried out according to the protocols. We used an upright microscope from Nikon to photograph the stained area, and the software Image J was used for semi-quantitative analysis.

Statistical analysis

All statistical analyses were performed using the R software, version 4.0.2 (The R Foundation for Statistical Computing, Vienna, Austria; http://www.R-project.org). A P-value less than 0.05 was deemed to indicate statistical significance.

Results

The preliminary screening of collagen- associated genes

Information from transcriptome sequencing and thorough clinical monitoring of 498 prostate cancer patients was acquired from the TCGA dataset. Employing the ‘edgeR’ R package, a differential gene expression profile was created, implementing screening criteria of |log2(fold change)|>0.5 and FDR < 0.05. Next, GSEA identified five gene sets related to collagen. The DGRGs expression patterns were displayed by volcanoes and heatmaps (Fig. 1A, B).

Fig. 1
figure 1

Discovery of Collagen-Associated Differential Genes in a Prostate Cancer Cohort. (A) Heatmap: Illustrates the gene expression profile of patients with PCa, highlighting the patterns of upregulation and downregulation among the group. (B) Volcano Plot: Illustrates the differentially expressed genes, highlighting significant genes with substantial fold changes and statistical significance

Full size image

Identification of collagen-related genes associated with prognosis

Through Pearson correlation analysis, we discovered 64 genes associated with collagen, which are detailed in Supplementary Material 1. After conducting univariate Cox regression analysis, we identified 6 collagen-related genes that are linked to prognosis. To control overfitting, LASSO regression analysis was used, and finally, 4 collagen-related genes associated with prognosis were obtained. These selected genes were PLOD3, COL1A1, MMP11 and FMOD. (Fig. 2)

Fig. 2
figure 2

Identification of Collagen-Associated Genes Related to BRFS in PCa Patients. (A) Utilizes univariate Cox regression analysis to identify prognostic collagen-associated genes relevant to the cohort. (B, C) Demonstrates the inclusion of six prognostic collagen-associated genes in a Lasso regression model, employing penalties to mitigate the model’s overfitting effects. (D) illustrates the differences in survival between groups with high or low expression of the four collagen-related genes in the TCGA cohort

Full size image

Creation and validation of risk score model

A total of 393 patients were divided into two groups, with 262 individuals allocated to the training cohort and 131 to the validation cohort. Within the training set, multivariate Cox regression analysis was employed to develop a risk score model, elucidating the coefficients associated with specific genes.

$${\rm{Risk}}\>{\rm{score = }}\Sigma _{i\, = \,1}^4{\rm{ = coefficien}}{{\rm{t}}_{\rm{i}}}\,{\rm{ \times }}\,{\rm{Exp(gene}}{{\rm{)}}_{\rm{i}}}$$

Individuals were categorized into high and low-risk groups based on the median risk score computed using this risk assessment tool. The risk scores, gene expressions, and survival predictions for both high and low-risk groups are delineated in both the training and validation datasets (Fig. 3). The high-risk group had a worse outcome compared to the low-risk group (P < 0.05).

Fig. 3
figure 3

Evaluation of the Risk Score Model Based on 4 Collagen-Linked Genes in Training and Validation Sets. (A, C) for the training dataset and (B, D) for the validation dataset illustrate the model’s efficacy. Low-risk patients had better biochemical recurrence-free survival than high-risk patients, as shown by Kaplan-Meier curves. (A, B) highlight that patients with low-risk scores typically experienced extended survival years and reduced recurrence rates compared to their high-risk counterparts. (C) in the training set and (D) in the validation set further reinforce these findings. The AUC of the model is greater than 0.7 in both the training set and the validation set (E, F)

Full size image

Construction and assessment nomogram

In light of the Univariate Cox analysis results, patient age, tumor and nodal stage, Gleason score, and risk score were singled out as significant determinants impacting BRFS among individuals with prostate cancer (Fig. 4A). Furthermore, a multivariate Cox regression analysis revealed that age, T stage, and risk score were identified as independent prognostic factors for the BRFS of patients with prostate cancer (Fig. 4B). Consequently, a prognostic nomogram was created utilizing age, T stage and risk score (Fig. 4C). Additionally, we performed individual evaluations on the effectiveness of the nomogram prediction model on the training dataset. (Fig. 4D shows the training dataset and Fig. 4E shows the validation dataset, Fig. 4F, G are two external validation sets). Over the timeframes of 1, 3, and 5 years, the calibration graph showcased the precision of the predictive model nomogram. Significant clinical utility for PCa patients at 1, 3, and 5 years was indicated by DCA through the nomogram’s analysis. The ROC curve confirmed the strong performance of the predictive model.

Fig. 4
figure 4

Establishment of Prognostic Nomogram. Univariate (A) and multivariate (B) Cox regression were conducted on the risk score and clinical data to determine the outcome. Demonstrating predictive strength and reliability, the nomogram (C) confirms the utility of the risk score model in both the training dataset (D) and validation dataset (E). In addition, we further verified the predictive performance of our model in two external validation sets (F, G)

Full size image

The change of PLOD3 expression can affect the functional phenotype of prostate Cancer cell

Using RT-qPCR, we verified a notable reduction in the PLOD3 expression level In PCa cell lines in comparison with normal prostate epithelial cells (RWPE-1) (Fig. 5A). In order to study how PLOD3 affects cellular activities, we decreased its levels in DU145, PC-3, and 22Rv1 cancer cells (Fig. 5B, C, D). The CCK-8 assay showed a decrease in cellular viability within the DU145 and PC3 cells treated with si-PLOD3 in comparison to the negative control condition (Fig. 5E, F). Treatment with si-PLOD3 resulted in significant reductions in the proliferation, metastasis and invasion capabilities of cells (Figs. 5G, H and I and 6J and K).

Fig. 5
figure 5

Validation of PLOD3’s cellular functions. (A) Demonstrates the increased expression of PLOD3 in PCa cells compared to normal prostate epithelial cells, emphasizing its possible involvement in the advancement of cancer. (B, C, D) Evaluate the transfection efficiency in both the control and siRNA groups across three distinct cell types, providing insight into the experimental setup’s effectiveness. (E, F) show decreased cellular viability within the siRNA group in comparison to the blank group for all cells tested, with emphasis on the significance (p-value) seen on day five. (G, H, I) Show a notable decrease in colony-forming capacity in the siRNA group relative to the control group, as evidenced by cell cloning experiments, suggesting an impact on cell proliferation

Full size image
Fig. 6
figure 6

Effect of PLOD3 knockdown on migration and invasion ability. (A, B) Invasion and migration assays revealed a substantial decrease in invasion and migration abilities within the siRNA group relative to the control group, highlighting the gene’s impact on the aggressiveness of cancer cells

Full size image

Expression of PLOD3 in malignant tissues and non-malignant regions

We performed PLOD3 IHC on pathological specimen sections from four groups of prostate cancer patients, which increases the clinical relevance of our findings. As shown in our results, PLOD3 was expressed differently in patients’ cancerous and para-cancerous tissues, with high PLOD3 expression in cancer tissues (Fig. 7A, B).

Fig. 7
figure 7

Expression of PLOD3 in prostate cancer patients. (A) A representative picture of differential expression of PLOD3 in cancer and adjacent tissues of prostate cancer patients was presented, with a higher magnification, (B) A semi-quantitative histogram of four sets of immunohistochemical results is presented

Full size image

Discussion

In a significant proportion of cases, individuals with clinically localized prostate cancer, undergoing local treatments such as surgery or radiation, are likely to achieve remission. However, a subset of these patients might harbor undetectable, either localized or metastatic, residual malignancy, which predisposes them to a potential biochemical recurrence of the disease. In progression, these individuals face the risk of developing clinical metastatic progression, ultimately leading to mortality from prostate cancer [14]. Further complicating treatment, genomic sequencing reveals marked individual differences in PCa, emphasizing its heterogeneity. This variability poses considerable challenges in formulating effective and personalized treatment strategies that are both efficacious and specifically tailored [15,16,17].

Within the tumor microenvironment (TME), there exists a diverse population of active stromal cells and an intricate network of stromal protein polymers [18]. These elements synergistically propel tumor proliferation, incursion, metastasis, and the therapeutic response [19, 20]. The significance of TME in the realm of cancer biology is increasingly acknowledged, one of the primary differences between normal tissue and tumor tissue is the characteristics of the stromal components within the tumor microenvironment [21], Numerous studies have demonstrated that stromal genes are significantly upregulated in prostate cancer (Pca) compared to normal prostate tissue [22,23,24,25], Additionally, the transcriptomic profile of the stroma surrounding prostate cancer tissue (distant from the cancerous tissue) shows significant differences from that of non-cancerous prostate tissue samples [26]. Collagen, a predominant structural protein within animal organisms, is distinguished by its intricately intertwined chain structure, marked by the recurring Xaa-Yaa-Gly tripeptide units (XYG repeats) [27]. Within the tumor microenvironment, collagen represents the most prevalent stromal protein polymer, known to enhance tumor rigidity, modulate tumor immunity, and facilitate the metastatic progression of cancer [28, 29]. Recent scholarly investigations have shed light on the pivotal role played by modulated collagen in the tumor microenvironment in contributing to tumor advancement. This influence manifests either as a physical barrier or via direct receptor engagement triggering signal transduction. These studies underscore the pivotal regulatory function of collagen in oncogenesis. Notably, Sun et al. discovered that collagen rearrangement, induced by DDR1, obstructs immune infiltration [8]; In their research, Peng DH et al. substantiated that increased collagen deposition within tumors can impede the migration of CD8 + T cells into the tumor. Furthermore, the interplay between collagen and the inhibitory receptor Lair1 of T cells, mediated by SHP-1 signaling, culminates in the depletion of CD8+ T cells [30].

Beyond its influence on the tumor microenvironment, collagen also significantly impacts the biological behaviors of tumor cells. An increase in collagen levels can activate relevant signaling pathways within these cells. Research by Liu et al. has established that COL1A1 plays a crucial role in the radiotherapy resistance of cervical cancer cells, significantly contributing to anti-apoptotic mechanisms via the Caspase-3/PI3K/AKT pathway [31]. Historical research indicates that collagen XXIII is upregulated in prostate cancer, identifiable in patient tissue samples and urinary excretions. This has led to its recognition as a biomarker indicative of the progression and metastatic development of prostate cancer [32, 33]. Intriguingly, in fibrosarcoma, the elevated expression of collagen paradoxically serves to inhibit tumor proliferation and metastatic spread [34]. “This observation underscores that modulating collagen expression across various cancer cell types yields divergent impacts on the progression of cancer.

In our research, to identify significant genes, initially, genes with differential expression within the prostate cancer cohort were pinpointed using the TCGA database, followed by an intersection with collagen-related genes from the GSEA database. Subsequently, employing univariate regression and lasso regression to mitigate overfitting, we developed a predictive model comprising four genes (PLOD3, COL1A1, MMP11, FMOD). The ROC curve provided validation for the model’s effectiveness, indicating that patients with increased risk scores experience a poorer prognosis. Moreover, the model’s predictive accuracy was further validated using two external datasets (GSE116918 and GSE70770), enhancing its utility. Nonetheless, the model’s practical value requires further clinical corroboration due to the absence of extensive clinical validation cohorts. In preceding literature, certain genes encompassed within our prognostic apparatus have been delineated as oncogenes or anti-oncogenes in a multitude of tumor varieties. Illustratively, In colorectal cancer patients, metastasis occurrence is facilitated by COL1A1’s modulation of the WNT/PCP pathway, as revealed by recent research [35]; In cases of head and neck squamous cell carcinoma (HNSCC), COL1A1 fosters the advancement of this carcinoma via a ligand-receptor interaction involving CD44 [36]. In a similar vein, MMP11’s stabilization of the Smad2 protein facilitates the advancement of breast cancer [37]. FMOD acts as a negative modulator in tumor progression, exhibiting an inverse correlation with the progression and metastasis of prostate cancer [38]. PLOD3 is capable of encoding lysyl hydroxylase 3, an enzyme predominantly engaged in the biosynthesis and glycosylation activities of diverse collagens [39]. It is associated with various human tumors and can serve as a prognostic and immunotherapeutic marker for colorectal cancer [40]. Knockdown of PLDO3 can inhibit the progression of non-small cell lung cancer [41]. Although its expression in the human pan-cancer system, including prostate cancer, has been characterized [42], its role in prostate cancer has not been fully reported. In this research, via in vitro functional assays on PLOD3 knockdown cells, we substantiated its role as an oncogene in prostate cancer, notably advancing tumor proliferation, migration, and especially its invasive prowess. However, we recognize that the cell lines used in this study did not cover all prostate cancer states. For example, the VCap cell line, which expresses AR highly, shows androgen dependence in both in vivo and in vitro cultures, and represents advanced prostate cancer cell lines to a certain extent, was not included. Further research with more reliable and comprehensive cell lines will be necessary in the future to provide a more complete understanding of PLOD3’s role in prostate cancer. In addition to the verification of cell function, we also conducted a preliminary evaluation of the clinical relevance of PLOD3, which is highly expressed in cancer tissues of prostate cancer patients, indicating the clinical relevance of our study. Last but not least, this research represents the inaugural demonstration of its function in prostate cancer, offering critical evidence for subsequent investigations into the ailment.

Nonetheless, an in-depth examination of how collagen alterations impact the tumor microenvironment was not undertaken in this study, marking a limitation; to enhance understanding of this model’s theoretical underpinnings, future research should elucidate the functions of these genes in living organisms and their specific molecular pathways, areas we intend to further investigate in forthcoming studies.

Conclusion

Our risk score model based on collagen-related genes can accurately predict the prognosis of prostate cancer patients. In vitro experiments, PLOD3, which is differentially expressed in cancer and adjacent cancer cells, can widely regulate the biological functions of prostate cancer cells.

Data availability

This study analyzed datasets that were made publicly accessible through The Cancer Genome Atlas (https://portal.gdc.cancer.gov/). The data sets utilized and/or examined in the present research can be obtained from the corresponding author upon request.

Abbreviations

Pca:

Prostate cancer

BRFS:

Biochemical recurrence-free survival

IHC:

Immunohistochemical

OD:

Optical density

TCGA:

The Cancer Genome Atlas

ECM:

Extracellular matrix

AUC:

Area under the curve

DCA:

Decision Curve Analysis

References

  1. Attard G, et al. Prostate cancer. Lancet. 2016;387(10013):70–82.

    Article  PubMed  Google Scholar 

  2. Gandaglia G, et al. Epidemiology and Prevention of prostate Cancer. Eur Urol Oncol. 2021;4(6):877–92.

    Article  PubMed  Google Scholar 

  3. 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. 2018;68(6):394–424.

    Article  PubMed  Google Scholar 

  4. Mottet N, et al. EAU-EANM-ESTRO-ESUR-SIOG guidelines on prostate Cancer-2020 update. Part 1: screening, diagnosis, and local treatment with curative intent. Eur Urol. 2021;79(2):243–62.

    Article  CAS  PubMed  Google Scholar 

  5. Cornford P, et al. EAU-EANM-ESTRO-ESUR-SIOG guidelines on prostate Cancer. Part II-2020 update: treatment of relapsing and metastatic prostate Cancer. Eur Urol. 2021;79(2):263–82.

    Article  CAS  PubMed  Google Scholar 

  6. Dangi-Garimella S, et al. Three-dimensional collagen I promotes gemcitabine resistance in pancreatic cancer through MT1-MMP-mediated expression of HMGA2. Cancer Res. 2011;71(3):1019–28.

    Article  CAS  PubMed  Google Scholar 

  7. Wishart AL et al. Decellularized extracellular matrix scaffolds identify full-length collagen VI as a driver of breast cancer cell invasion in obesity and metastasis. Sci Adv, 2020. 6(43).

  8. Sun X, et al. Tumour DDR1 promotes collagen fibre alignment to instigate immune exclusion. Nature. 2021;599(7886):673–8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  9. Robertson C, et al. Extracellular matrix modulates T cell clearance of malignant cells in vitro. Biomaterials. 2022;282:121378.

    Article  CAS  PubMed  Google Scholar 

  10. Penet MF, et al. Structure and function of a prostate Cancer dissemination-permissive Extracellular Matrix. Clin Cancer Res. 2017;23(9):2245–54.

    Article  CAS  PubMed  Google Scholar 

  11. Li SC, et al. Telomere-related gene risk model for prognosis and drug treatment efficiency prediction in kidney cancer. Front Immunol. 2022;13:975057.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  12. Huang H, et al. A novel five-gene metabolism-related risk signature for predicting prognosis and immune infiltration in endometrial cancer: a TCGA data mining. Comput Biol Med. 2023;155:106632.

    Article  CAS  PubMed  Google Scholar 

  13. Sui Q, et al. Cisplatin resistance-related multi-omics differences and the establishment of machine learning models. J Transl Med. 2022;20(1):171.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  14. Zumsteg ZS, et al. The natural history and predictors of outcome following biochemical relapse in the dose escalation era for prostate cancer patients undergoing definitive external beam radiotherapy. Eur Urol. 2015;67(6):1009–16.

    Article  PubMed  Google Scholar 

  15. Zhu Y, et al. Epidemiology and genomics of prostate cancer in Asian men. Nat Rev Urol. 2021;18(5):282–301.

    Article  CAS  PubMed  Google Scholar 

  16. Oliveira-Barros EG, et al. Prostate cancer molecular profiling: the Achilles heel for the implementation of precision medicine. Cell Biol Int. 2017;41(11):1239–45.

    Article  CAS  PubMed  Google Scholar 

  17. Vidal AC, Freedland SJ. Obesity and prostate Cancer: a focused update on active surveillance, race, and Molecular Subtyping. Eur Urol. 2017;72(1):78–83.

    Article  PubMed  Google Scholar 

  18. Lambrechts D, et al. Phenotype molding of stromal cells in the lung tumor microenvironment. Nat Med. 2018;24(8):1277–89.

    Article  CAS  PubMed  Google Scholar 

  19. Balkwill FR, Capasso M, Hagemann T. The tumor microenvironment at a glance. J Cell Sci. 2012;125(Pt 23):5591–6.

    Article  CAS  PubMed  Google Scholar 

  20. Ahmed F, Haass NK. Microenvironment-driven dynamic heterogeneity and phenotypic plasticity as a mechanism of Melanoma Therapy Resistance. Front Oncol. 2018;8:173.

    Article  PubMed  PubMed Central  Google Scholar 

  21. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell. 2011;144(5):646–74.

    Article  CAS  PubMed  Google Scholar 

  22. Karkampouna S et al. Stroma Transcriptomic and Proteomic Profile of prostate Cancer metastasis xenograft models reveals Prognostic Value of Stroma signatures. Cancers (Basel), 2020. 12(12).

  23. Sun C, et al. Bioinformatics Analysis of Stromal Molecular Signatures Associated with breast and prostate Cancer. J Comput Biol. 2019;26(10):1130–9.

    Article  CAS  PubMed  Google Scholar 

  24. Planche A, et al. Identification of prognostic molecular features in the reactive stroma of human breast and prostate cancer. PLoS ONE. 2011;6(5):e18640.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  25. Heidegger I, et al. Prediction of clinically significant prostate Cancer by a specific collagen-related Transcriptome, Proteome, and Urinome signature. Eur Urol Oncol; 2024.

  26. Tyekucheva S, et al. Stromal and epithelial transcriptional map of initiation progression and metastatic potential of human prostate cancer. Nat Commun. 2017;8(1):420.

    Article  PubMed  PubMed Central  Google Scholar 

  27. Shoulders MD, Raines RT. Collagen structure and stability. Annu Rev Biochem. 2009;78:929–58.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  28. Whatcott CJ, et al. Desmoplasia in primary tumors and metastatic lesions of pancreatic Cancer. Clin Cancer Res. 2015;21(15):3561–8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  29. Yamauchi M, et al. The fibrotic tumor stroma. J Clin Invest. 2018;128(1):16–25.

    Article  PubMed  PubMed Central  Google Scholar 

  30. Peng DH, et al. Collagen promotes anti-PD-1/PD-L1 resistance in cancer through LAIR1-dependent CD8(+) T cell exhaustion. Nat Commun. 2020;11(1):4520.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  31. Liu S, Liao G, Li G. Regulatory effects of COL1A1 on apoptosis induced by radiation in cervical cancer cells. Cancer Cell Int. 2017;17:73.

    Article  PubMed  PubMed Central  Google Scholar 

  32. Banyard J, Bao L, Zetter BR. Type XXIII collagen, a new transmembrane collagen identified in metastatic tumor cells. J Biol Chem. 2003;278(23):20989–94.

    Article  CAS  PubMed  Google Scholar 

  33. Banyard J, et al. Collagen XXIII expression is associated with prostate cancer recurrence and distant metastases. Clin Cancer Res. 2007;13(9):2634–42.

    Article  CAS  PubMed  Google Scholar 

  34. Liu J, et al. Molecular mechanism of Gd@C(82)(OH)(22) increasing collagen expression: implication for encaging tumor. Biomaterials. 2018;152:24–36.

    Article  CAS  PubMed  Google Scholar 

  35. Zhang Z, et al. COL1A1 promotes metastasis in colorectal cancer by regulating the WNT/PCP pathway. Mol Med Rep. 2018;17(4):5037–42.

    CAS  PubMed  PubMed Central  Google Scholar 

  36. Choi JH, et al. Single-cell transcriptome profiling of the stepwise progression of head and neck cancer. Nat Commun. 2023;14(1):1055.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  37. Zhuang Y et al. MMP11 promotes the proliferation and progression of breast cancer through stabilizing Smad2 protein. Oncol Rep, 2021. 45(4).

  38. Treacy PJ et al. Association between expression of connective tissue genes and prostate Cancer growth and progression. Int J Mol Sci, 2023. 24(8).

  39. Qi Y, Xu R. Roles of PLODs in collagen synthesis and Cancer progression. Front Cell Dev Biol. 2018;6:66.

    Article  PubMed  PubMed Central  Google Scholar 

  40. Shi J, et al. Integrated Profiling identifies PLOD3 as a potential prognostic and Immunotherapy relevant Biomarker in Colorectal Cancer. Front Immunol. 2021;12:722807.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  41. Li WH, et al. PLOD3 regulates the expression of YAP1 to affect the progression of non-small cell lung cancer via the PKCδ/CDK1/LIMD1 signaling pathway. Lab Invest. 2022;102(4):440–51.

    Article  CAS  PubMed  Google Scholar 

  42. Gong S et al. A human Pan-cancer System analysis of Procollagen-Lysine, 2-Oxoglutarate 5-Dioxygenase 3 (PLOD3). Int J Mol Sci, 2021. 22(18).

Download references

Acknowledgements

We would like to acknowledge the use of data from TCGA managed by the National Cancer Institute and the National Human Genome Research Institute.

Funding

This study was supported by the Key-Area Research and Development Program of Guangdong Province (2023B1111030006), National Natural Science Foundation of China (82372766 and 82072841), Natural Science Foundation of Guangdong Province (2021A1515010199), Guangdong Provincial Clinical Research Center for Urological Diseases (2020B1111170006) and Guangdong Science and Technology Department (2020B1212060018).

Author information

Author notes

  1. Yunfei Xiao, Cong Lai and Jintao Hu contributed equally to this work.

Authors and Affiliations

  1. Department of Urology, Sun Yat-sen Memorial Hospital, Sun Yat-sen University, No. 107 Yanjiang West Road, Guangzhou, Guangdong, 510000, China

    Yunfei Xiao, Cong Lai, Jintao Hu, Yelisudan Mulati, Xiaoting Xu, Jiawen Luo, Degeng Kong, Cheng Liu & Kewei Xu

  2. Guangdong Provincial Key Laboratory of Malignant Tumor Epigenetics and Gene Regulation, Sun Yat-sen Memorial Hospital, Sun Yat-sen University, Guangzhou, Guangdong, 510000, China

    Yunfei Xiao, Cong Lai, Jintao Hu, Yelisudan Mulati, Xiaoting Xu, Jiawen Luo & Degeng Kong

  3. Guangdong Provincial Clinical Research Center for Urological Diseases, Guangzhou, Guangdong, 510000, China

    Cheng Liu & Kewei Xu

  4. Sun Yat-sen University School of Medicine, Sun Yat-sen University, Shenzhen, Guangdong, 518000, China

    Kewei Xu

Authors
  1. Yunfei Xiao
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Cong Lai
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Jintao Hu
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Yelisudan Mulati
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Xiaoting Xu
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Jiawen Luo
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Degeng Kong
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Cheng Liu
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Kewei Xu
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

Yunfei Xiao, Cong Lai, Jintao Hu, and Kewei Xu conceived and orchestrated the study. Cong Lai and Jintao Hu, Xiaoting Xu executed data gathering and analysis from public databases. Experiments were conducted and data were analyzed by Yunfei Xiao. Yunfei Xiao also authored the manuscript. Yelisudan Mulati, Jiawen Luo, Degeng Kong, Cheng Liu, and Kewei Xu revised the manuscript. All authors read and approved the final manuscript. Yunfei Xiao, Cong Lai, and Jintao Hu contributed to this article equally and should be considered as co-first authors.

Corresponding authors

Correspondence to Cheng Liu or Kewei Xu.

Ethics declarations

Ethics approval and consent to participate

Informed consent to participate was obtained from all participants involved in the study. This study, which included human data and human tissue, was approved by the Institutional Review Board (IRB) at Sun Yat-sen Memorial Hospital under reference number [SYSKY-2023-769-01].

Consent for publication

Not applicable.

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.

Electronic supplementary material

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, 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 you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. 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-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Xiao, Y., Lai, C., Hu, J. et al. Integrative analysis regarding the correlation between collagen-related genes and prostate cancer. BMC Cancer 24, 1038 (2024). https://doi.org/10.1186/s12885-024-12783-9

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12885-024-12783-9

Keywords

BMC Cancer

ISSN: 1471-2407

Contact us