Overexpression of COL10A1 in CAFs predicts poor outcome and promotes ovarian cancer progression

Yaping Wang 1,2

Min Tian 1

Hongjian Zhang 1,2

Hai Zhu 1

Caixia Ma 1,3

Xiabing Li 1,2

Luyao Kang 4

Qiaohong Qin 1,2

Yiran Wang 1

Hongyu Li 1,2

Qing Liu 1

Shujun Zhao 1

Dr.

Gaili Ji 1,3,5✉ Phone+86-13939077898 Email745160500@qq.com

1 Gynecologic Oncology the Third Affiliated Hospital of Zhengzhou University 450052 Zhengzhou Henan China

2 Zhengzhou Key Laboratory of Gynecological Oncology 450052 Zhengzhou Henan China

3 Department of Gynecology and Obstetrics, the Third Affiliated Hospital Zhengzhou University 450052 Zhengzhou China

4 Shanghai Key Laboratory of Maternal Fetal Medicine, Shanghai Institute of Maternal Medicine. Fetal Medicine and Gynecologic Oncology, Shanghai First Maternity and Infant Hospital, School of Medicine Tongji University 200092 Shanghai China

5 Department of Gynecology and Obstetrics the Third Affiliated Hospital of Zhengzhou University 450000 Zhengzhou Henan P. R. China

Yaping Wang ^1,2*, Min Tian^1*, Hongjian Zhang^1,2, Hai Zhu^1*, Caixia Ma^1,3, Xiabing Li^1,2, Luyao Kang⁴, Qiaohong Qin^1,2, Yiran Wang¹, Hongyu Li^1,2, Qing Liu^1#, Shujun Zhao^1#, Gaili Ji^1,3#

¹Gynecologic Oncology, the Third Affiliated Hospital of Zhengzhou University, 450052, Zhengzhou, Henan, China;

²Zhengzhou Key Laboratory of Gynecological Oncology, 450052, Zhengzhou, Henan, China;

³Department of Gynecology and Obstetrics, the Third Affiliated Hospital, Zhengzhou University, Zhengzhou, 450052, China;

⁴Shanghai Key Laboratory of Maternal Fetal Medicine, Shanghai Institute of Maternal Medicine. Fetal Medicine and Gynecologic Oncology, Shanghai First Maternity and Infant Hospital, School of Medicine, Tongji University, Shanghai 200092 China;

^#Correspondence: Dr. Gaili Ji, Department of Gynecology and Obstetrics, the Third Affiliated Hospital of Zhengzhou University, Zhengzhou, Henan 450000, P. R. China. E-mail: 745160500@qq.com. Tel: +86-13939077898, Fax: +86-371-66992000

Yaping Wang, Min Tian and Hai Zhu contributed equally to this work.

Abstract

Background

Ovarian cancer (OC) represents one of the most lethal gynecological malignancies. Cancer-associated fibroblasts (CAFs) are present in both primary and metastatic tumors and exhibit significant functional heterogeneity, adaptability and resilience. These cells are crucial for cancer progression because of their complex signaling interactions with different cell types in the tumor microenvironment. The collagen type X alpha 1 chain (COL10A1) is notably overexpressed in CAFs and is closely associated with the initiation and progression of the disease. However, the role of the COL10A1 gene in OC-associated CAFs (OC-CAFs) remains unexplored.

Methods

Here, we identified the differentially expressed gene COL10A1 via the integration of multiple databases, including the GEO database and the TCGA-OV dataset. We subsequently performed further bioinformatics analyses concerning COL10A1. The external datasets were ultimately analyzed, and clinical samples were collected and examined via immunohistochemistry and quantitative reverse transcription polymerase chain reaction (qRT‒PCR).

Results

Our findings revealed that COL10A1 expression was markedly elevated in OC-CAFs and correlated with adverse clinicopathological characteristics and poorer patient prognosis. Functional enrichment analyses indicated that COL10A1 may facilitate tumorigenesis and progression by modulating several pathways associated with cellular growth, metabolism, proliferation, and survival, particularly the phosphoinositide 3-kinase (PI3K)/protein kinase B (AKT) signaling pathway and extracellular matrix (ECM)-receptor interactions. Furthermore, we observed that COL10A1 was associated with the infiltration of various immune cells and immune checkpoints. Importantly, in clinical samples, COL10A1 expression was significantly increased in OC-CAFs, which was related to unfavorable clinicopathological features and poorer patient prognosis.

Conclusions

Our research indicates that CAFs with elevated COL10A1 expression may enhance OC progression through the modulation of macrophage polarization. Consequently, COL10A1 serves as a novel biomarker for predicting OC prognosis and provides a promising avenue for developing therapeutic strategies targeting the tumor microenvironment.

Keywords:

COL10A1

ovarian cancer

immune infiltration

tumor microenvironment

prognosis

Introduction

Ovarian cancer (OC) ranks among the most prevalent malignant tumors and affects the female reproductive system. It is recognized as one of the gynecological cancers with the highest mortality rates. Globally, the five-year survival rate for individuals diagnosed with OC is less than 50%(1). First-line treatment strategies, including surgical intervention followed by platinum-based chemotherapy, are available. However, the challenges of chemotherapy resistance and resistance to targeted therapies continue to significantly impede the enhancement of prognostic outcomes for patients with OC(2). Therefore, there is a pressing need to investigate prognostic and drug-responsive biomarkers that could improve survival rates for patients with metastatic ovarian cancer. The process of identifying novel biomarkers could be combined with emerging technologies, thereby paving the way for innovative research avenues focused on clinical applications in the realm of tumor diagnosis and treatment(3).

The tumor microenvironment (TME) comprises tumor cells, the vascular network, the extracellular matrix (ECM), and various immune cells(4). The ECM is integral to tumorigenesis, disease progression, and the modulation of therapeutic responses. Genes associated with the ECM have the potential to serve as prognostic indicators of recurrence and overall prognosis in OC. Collagen is the main component of the ECM, participates in cancer fibrosis and affects the behavior of cancer cells. Cancer cells can reverse and reshape collagen, promoting the progression of cancer. Collagen, the major structural protein of the ECM, is involved in cancer-related fibrosis and influences cancer cell behavior. Cancer cells can remodel collagen, thereby facilitating cancer progression. Additionally, collagen interacts with macrophages, mast cells, lymphocytes, and fibroblasts, which collectively modulate cancer immunity and progression(5). Numerous clinical investigations have identified collagen as a significant prognostic factor(4). Furthermore, the relationship of collagen with chemotherapy and targeted drug resistance in cancer patients has been established(6). Owing to its evident genetic and epigenetic stability, along with its expression across nearly all cancer types, collagen holds promise as a therapeutic target or agent.

The gene encoding type X collagen, COL10A1, produces a secreted ECM protein that contains a triple helix domain, a noncollagenous carboxyl-terminal domain, and a short N-terminal domain(7). COL10A1 plays a critical role in maintaining the homeostasis of the tissue microenvironment by influencing the assembly of the ECM and its mechanical properties, thereby affecting cellular signaling and tissue mechanics(8). Recent research has indicated that the upregulated expression of COL10A1 significantly contributes to tumor progression through the activation of specific signaling pathways. This abnormal expression of COL10A1 facilitates tumor adhesion, migration, and angiogenesis while also accelerating ECM remodeling and tumor invasion(9). Several studies have demonstrated that the malignant progression of diverse solid tumors, including breast, bladder, and prostate cancers, is reliant on the regulation of the epithelial‒mesenchymal transition (EMT) process(10) and the PI3K‒AKT signaling pathway(11–13) by COL10A1. Moreover, elevated levels of COL10A1 expression are closely associated with lymph node metastasis, chemotherapy resistance, and the formation of an immunosuppressive microenvironment(14). Furthermore, COL10A1 facilitates immune evasion through the recruitment of M2-type tumor-associated macrophages (TAMs) and regulatory T cells (Tregs), which collectively suppress antitumor immune responses. Although the mechanism of action of COL10A1 in other malignancies, such as bladder cancer, has gradually been elucidated(15, 16), its expression pattern in OC, its role in OC progression, and its prognostic value have not been systematically investigated.

Therefore, this study aimed to analyze COL10A1 expression in OC-CAFs via bioinformatics and clinical specimen evaluations. We explored the correlations between COL10A1 expression and tumor stage, vascular invasion, tumor grade, lymph node invasion, and poor prognosis. The objective of this study was to uncover the oncogenic properties of COL10A1 in OC-CAFs and assess its potential as a therapeutic target, thereby providing a theoretical framework for improving patient prognosis. The flowchart illustrating the study design is presented in Fig. 1.

Fig. 1

Flow chart of this study.

Materials and methods

Data collection and analysis

Normal and tumor samples from the TCGA database of mRNA sequencing data (FPKM format) (https://portal.gdc.cancer.gov/), which included 427 samples and 88 matched normal ovarian tissue samples, were downloaded. Additionally, we retrieved datasets from the Gene Expression Omnibus (GEO) database (available at https://www.ncbi.nlm.nih.gov/), namely, GSE40595, GSE193875, GSE26712, GSE66957, GSE17260, GSE140082, GSE9891, and GSE26193. Concurrently, clinical data for OC patients from both the TCGA and GEO databases were obtained. Following data acquisition, we categorized patients into high- and low-expression groups on the basis of the median expression level of COL10A1 mRNA. In total, the study included 1,558 samples (TCGA: 427; GSE40595: 31; GSE193875: 6; GSE26712: 184; GSE66957: 57; GSE17260: 110; GSE140082: 355; GSE9891: 281; GSE26193: 107) and 108 normal tissue samples (TCGA: 88; GSE40595: 8; GSE26712: 10; GSE66957: 12).

Screening of differentially expressed genes (DEGs)

First, CAFs from 3 cases of metastatic OC tissues (mCAFs) and 3 cases of nonmetastatic OC tissues (nmCAFs) were obtained from the GSE193875 dataset downloaded from the GEO database. The mCAFs and nmCAFs were compared to identify differentially expressed genes. Second, the CAFs in 31 cases of high-grade serous OC from the GSE40595 dataset downloaded from the GEO database and those in 8 cases of normal ovarian tissues were compared to identify DEGs. Differentially expressed genes were subsequently identified by comparing 427 high-grade serous OC tissues downloaded from the TCGA database and 88 normal ovarian tissues. All screenings were conducted with a threshold P value of 0.05 and an absolute log-fold-change (logFC) of 1. Finally, the DEGs in each group were identified via the web tool Venn diagrams (http://bioinformatics.psb.ugent.be/webtools/Venn). The mRNA expression of DEGs in OC tissues from the TCGA database was analyzed via R software in terms of risk status, tumor stage, tumor grade, tumor size, lymph node invasion, vascular invasion, and survival analysis. Therefore, it is possible to identify specific genes related to the prognosis and progression of patients with OC.

Construction and validation of the nomogram

A nomogram was developed using independent prognostic factors through the “rms” and “survival” R packages(17). A calibration plot was generated to assess the accuracy of the nomogram predictions. The concordance index (C-index) was used to evaluate the ability of the nomogram to distinguish between different patient groups.

Functional enrichment analysis

The objective of the DEG analysis was to ascertain functional enrichment through Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses. This was achieved by employing the R package clusterProfiler in conjunction with org.Hs.eg.db and enrichplot (version 1.0.2)(18). A false discovery rate (FDR) threshold below 0.25 and an adjusted p value below 0.05 were established as joint criteria for determining statistically significant enrichment in functional or pathway categories.

Analysis of immune infiltration

The relative infiltration levels of immune cells were quantitatively assessed via the single-sample gene set enrichment analysis (ssGSEA) approach implemented in the gene set variation analysis (GSVA) package (version 1.34.0). To investigate the association between COL10A1 expression and immune infiltration in OC, both Spearman correlation analysis and the Wilcoxon rank sum test were employed. Additionally, we utilized the SCNA module available on the TIMER website to examine tumor immune infiltration levels associated with various copy number variations (CNVs) of COL10A1. A comparison of COL10A1 expression across six immune subtypes revealed distinct immune characteristics that corresponded to the predominant features of tumor samples documented in the TCGA database(19). The CIBERSORT algorithm was subsequently applied to analyze the composition of 22 tumor-infiltrating immune cell types in OC samples with high and low COL10A1 expression(20). Furthermore, the GEPIA web server, which includes normal tissue datasets as controls, was used to calculate Spearman correlation coefficients; this analysis assessed the correlation between COL10A1 expression and the expression levels of immune marker genes in the TCGA database. Concurrently, the mRNA expression levels of ten immune checkpoint genes were compared between the high and low COL10A1 expression groups.

Tissue samples

Fresh OC tissues were collected from patients who underwent surgical procedures at the Third Affiliated Hospital of Zhengzhou University between 2018 and 2020. Additionally, normal ovarian tissues, which were excised due to medical conditions unrelated to ovarian cancer, were also collected. Subsequently, tumor samples fixed in formalin and embedded in paraffin were obtained from the hospital’s pathology department. Renowned pathologists verified the histopathological diagnoses and tumor staging at the time of sample preservation.

All participants provided written informed consent prior to inclusion in the study.

This research received approval from the Medical Ethics Committee of the Third Affiliated Hospital of Zhengzhou University, under Approval No. 2025-01-236.

RNA extraction and qRT‒PCR assays

Total RNA was extracted from tissue samples via TRIzol™ reagent (Invitrogen, Carlsbad, CA, USA). Subsequently, complementary DNA (cDNA) synthesis was performed with the PrimeScript™ RT Reagent Kit (Takara Bio, Tokyo, Japan). Quantitative real-time polymerase chain reaction (qRT‒PCR) was performed via TB Green™ Premix Ex Taq™ (Takara Bio, Tokyo, Japan) on a CFX96 Deep Well™ real-time system (Bio‒Rad, Hercules, CA, USA). The relative expression levels of the target mRNAs were determined via the 2^(-ΔΔCt) method, with normalization against GAPDH. The sequences of the primers used in this analysis are presented

in Supplementary Table 1.

Immunohistochemistry

The tissue samples were subjected to immunohistochemical (IHC) processing following the established protocol, which included deparaffinization, dehydration, and antigen retrieval. After antigen retrieval, the sections were treated with a blocking solution containing 5% BSA and then incubated overnight at 4°C with a primary antibody targeting COL10A1 at a dilution of 1:500 (Servicebio). The following day, the sections were exposed to a goat anti-rabbit IgG secondary antibody (Servicebio) at room temperature for 1 hour. The staining was developed using diaminobenzidine (DAB) to visualize the results. The overall staining scores were evaluated independently by two experienced pathologists.

Statistical analysis

Bioinformatics analyses were performed with R (version 4.2.2) and Perl (version 5.32.1)(21). To evaluate differences in clinicopathological factors, both the chi-square test and Fisher’s exact test were employed. For continuous variables, analysis of variance (ANOVA) was performed. Additionally, univariate logistic regression models were applied to assess the associations between the expression levels of COL10A1 and the relevant clinical features. The area under the receiver operating characteristic (ROC) curve (AUC) was determined to evaluate the diagnostic potential of COL10A1 for OC. Furthermore, Cox regression, Kaplan‒Meier (KM) survival analysis, and the log-rank test were conducted to investigate the prognostic significance of COL10A1. Statistical analyses of different clinical datasets were conducted via Prism 8 (GraphPad Software Inc., La Jolla, USA), with a significance threshold set at p < 0.05.

Results

Identification of specific gene signatures

To identify the differentially expressed genes (DEGs) associated with CAFs and the prognosis of OC patients, we used R software to compare the mRNA expression levels across the three datasets. Following the analysis of transcriptomic differences between mCAFs and nmCAFs, we identified the DEGs and visualized them via volcano plots (Fig. 2A). The corresponding expression levels of the identified DEGs were visualized via a heatmap (Fig. 2B). We subsequently assessed the DEGs between CAFs and high-grade serous ovarian cancer (HGSOC) tissues compared with normal ovarian tissues and visualized the DEGs via volcano plots (Fig. 2C) and heatmaps (Fig. 2D). We also determined the DEGs between normal ovarian tissue and HGSOC tissue, again visualizing the DEGs via volcano plots (Fig. 2E) and heatmaps (Fig. 2F). We integrated the differential analysis outcomes from the three datasets. By generating a Venn diagram (Fig. 2G), we identified six pivotal genes (MMP11, MAFB, MX2, MOXD1, CHI3L1, and COL10A1) that exhibited a consistent pattern of elevated expression across all three datasets.

Fig. 2

Differences in gene expression between CAFs in ovarian cancer and normal ovarian tissues. (A) Volcano plot of the mRNA levels from the GSE40595 dataset; the x-axis represents the log2-transformed fold change ratios. The y-axis is the log10-transformed p value. The red dots represent the DEGs with a fold change > 1. The blue dots represent the DEGs with a fold change <-1. In this study, a volcano plot was constructed to display the different genes associated with patients with CAFs from metastatic ovarian cancer (mCAFs) compared with patients with CAFs from nonmetastatic ovarian cancer (nmCAFs). (B) Heatmap of the candidate genes associated with CAFs in patients with metastatic ovarian cancer from GSE40595. (C) Volcano plot of the mRNA levels of different genes in high-grade serous ovarian cancer samples and normal ovary samples. (D) Heatmap of the candidate genes associated with high-grade serous ovarian cancer from the GSE193875 dataset. (E) Volcano plot of mRNA levels from the TCGA database. (F) Heatmap of the candidate genes associated with ovarian cancer from the TCGA database. (G) Venn diagram representing the distribution of DEGs in different groups.

Relationship between COL10A1 expression and the clinicopathological features of OC patients

To evaluate the clinical significance of these six DEGs, we used violin plots (Fig. 3 and Figs. 1–5) to visualize the expression levels of these genes across various clinical parameters. The findings indicate that COL10A1 is significantly associated with multiple clinical factors, including the risk level of OC (Fig. 3A, p = 5.46e-104), tumor stage (Fig. 3B, p = 1.98e-04), tumor size (Fig. 3D, p = 9.12e-03), venous invasion (Fig. 3E, p = 0.04), and lymphatic invasion (Fig. 3F, p = 0.04). Moreover, we examined the relationships between COL10A1 expression in the TCGA database and clinical parameters (Table 1), the results of which were consistent with those described above. COL10A1 expression was significantly related to tumor stage (p = 0.035), lymph node invasion (p = 0.016), and tumor progression status (p = 0.002). Furthermore, logistic regression analysis of the TCGA-OV dataset revealed that COL10A1 expression levels are significantly correlated with clinical stage, lymphatic invasion, and tumor status (Table 2). On the basis of these findings, COL10A1 in CAFs was identified as a core biomarker for ovarian cancer in this study. We place particular emphasis on analyzing the mechanisms linking its expression levels to the clinicopathological characteristics and survival outcomes of ovarian cancer patients.

Fig. 3

Expression of COL10A1 in ovarian cancer. (A) Expression data after the median high and low populations were grouped on the basis of the expression level of COL10A1. (B) COL10A1 expression in different stages, namely, early stage (I/II) and advanced stage (III/IV). (C) The expression of COL10A1 at different grades: well differentiated/moderately differentiated (G1/G2) and poorly differentiated/undifferentiated (G3/G4). (D) COL10A1 expression in tissues of different sizes. (E) The expression of COL10A1 under different lymph node infiltration states. (F) The expression of COL10A1 in different vascular metastasis states.

Fig. 5

The value of COL10A1 in the diagnosis and prognosis of ovarian cancer. (A) Forest plot based on multivariate Cox analysis for overall survival. (B) A nomogram for the prediction of the one-, three-, and five-year overall survival rates of patients with ovarian cancer. (C) ROC curves for classifying ovarian cancer versus normal ovarian tissues in the TCGA database. (D-F) Calibration curves of the nomogram for the prediction of the one-, three-, and five-year overall survival rates of patients with ovarian cancer. PD, progressive disease. SD, stable disease. PR, partial response. CR, complete response. G1/G2, well differentiated/moderately differentiated. G3/G4, poorly differentiated/undifferentiated.

Table 1

Relationship between COL10A1 expression and clinicopathological parameters
Characteristics (n, %)	TCGA cohort (n = 381)				Validation cohort (n = 62)
Characteristics (n, %)	High COL10A1 expression	Low COL10A1 expression	X²	P value	High COL10A1 expression	Low COL10A1 expression	X²	P value
n	190	191			31	31
Age			1.4134	0.234			0.065	0.799
<= 60	110 (28.9%)	99 (26%)			17 (27.4%)	16 (25.8%)
> 60	80 (21%)	92 (24.1%)			14(22.6%)	15 (24.2%)
Clinical stage			4.449	0.035			18.895	< 0.001
Stage I&Stage II	17 (4.5%)	7 (1.9%)			2(3.2%)	18 (29.0%)
Stage III&Stage IV	172 (45.5%)	182 (48.1%)			29 (46.8%)	13 (21.0%)
Venous invasion			3.283	0.070			0.793	0.069
No	26 (24.8%)	15 (14.3%)			19 (30.6%)	20 (32.3%)
Yes	29 (27.6%)	35 (33.3%)			12 (19.4%)	11 (17.7%)
Lymphatic invasion			5.819	0.016			17.404	< 0.001
No	32 (21.5%)	16 (10.7%)			27 (43.5%)	11 (17.7%)
Yes	46 (30.9%)	55 (36.9%)			4 (6.5%)	20 (32.3%)
Histologic grade			< 0.001	0.984			6.458	0.011
G1&G2	23 (6.2%)	23 (6.2%)			10 (16.1%)	20 (32.3%)
G3&G4	163 (43.9%)	162 (43.7%)			21 (33.9%)	11 (17.7%)
Tumor residual			3.508	0.061			16.672	< 0.001
No	41 (12.2%)	27 (8%)			25 (40.3%)	9 (14.5%)
Yes	128 (38%)	141 (41.8%)			6 (9.7%)	22 (35.5%)
Tumor status			9.458	0.002			12.762	< 0.001
Tumor free	48 (14.2%)	24 (7.1%)			24 (38.7%)	10 (16.1%)
With tumor	123 (36.4%)	143 (42.3%)			7 (11.3%)	21 (33.9%)
Primary therapy outcome			2.034	0.154			0.791	0.347
PD&SD	20 (6.5%)	29 (9.4%)			9 (14.5%)	6 (9.7%)
PR&CR	135 (43.7%)	125 (40.5%)			22 (35.5%)	25 (40.3%)

Table 2

COL10A1 expression correlated with clinicopathological characteristics analyzed by logistic regression
Characteristics	TCGA cohort (n = 381)		Validation cohort (n = 62)
Characteristics	OR (95% CI)	P value	OR (95% CI)	P value
Age (> 60 vs. <= 60)	1.278 (0.853–1.915)	0.235	1.241 (0.838–1.840)	0.282
Clinical stage (Stage III&IV vs. Stage I&II)	2.570 (1.040–6.349)	0.041	2.502 (1.105–5.669)	0.028
Venous invasion (Yes vs. No)	2.092 (0.936–4.673)	0.072	0.830 (0.526–1.308)	0.422
Lymphatic invasion (Yes vs. No)	2.391 (1.168–4.896)	0.017	1.565 (1.070–2.289)	0.021
Histologic grade (G3&G4 vs. G1&G2)	0.994 (0.536–1.843)	0.984	1.542 (1.064–2.235)	0.022
Tumor residual (Yes vs. No)	1.673 (0.973–2.875)	0.063	1.298 (0.888–1.897)	0.178
Tumor status (With tumor vs. Tumor free)	2.325 (1.347–4.014)	0.002	3.319 (1.670–6.594)	< 0.001
Primary therapy outcome (SD&PD vs. CR&PR)	1.566 (0.843–2.909)	0.156	0.931 (0.613–1.415)	0.739

The prognostic value of COL10A1 in OC

The associations between DEG expression levels and the prognosis of OC patients were assessed via the Kaplan–Meier (KM) method. Using the median risk score as a classification threshold, patients were categorized into high-risk and low-risk groups. Comparative analysis revealed that OS for the COL10A1 high-risk group was significantly lower than that for the low-risk group (p = 0.022, Fig. 4). To identify prognostic factors, we performed both univariate and multivariate Cox regression analyses, as shown in Figs. 4G and 5A. Univariate analysis revealed that several factors were closely linked to OS: clinical stage (III/IV vs I/II: HR = 3.135, 95% CI = 1.947–6.811, p = 0.037); tumor status (tumor vs. tumor-free: HR = 9.598, 95% CI = 4.487–20.532, p < 0.001); primary therapy outcome—stable disease (SD) vs. progressive disease (PD): HR = 0.441, 95% CI = 0.217–0.896, p = 0.024; complete response (CR) vs. PD: HR = 0.154, 95% CI = 0.095–0.250, p < 0.001; age (> 60 vs. ≤60: HR = 1.352, 95% CI = 1.045–1.749, p = 0.022); lymph node invasion (yes vs. no: HR = 2.613, 95% CI = 1.854–3.387, p = 0.02); and residual tumor status (yes vs. no: HR = 2.223, 95% CI = 1.441–3.430, p < 0.

Fig. 4

Prognostic value of DEGs in patients with ovarian cancer evaluated via the Kaplan‒Meier method. (A) OS of ovarian cancer patients with high versus low CHI3L1 expression. (B) OS of ovarian cancer patients with high versus low COL10A1 expression. (C) OS of ovarian cancer patients with high versus low MAFB. (D) OS of ovarian cancer patients with high versus low MMP11. (E) OS of ovarian cancer patients with high versus low MOXD1 expression. (F) OS of ovarian cancer patients with high versus low MX2. (G) Forest map based on univariate Cox analysis for overall survival. HR, hazard ratio; CI, confidence interval. PD, progressive disease. SD, stable disease. PR, partial response. CR, complete response. G1/G2, well differentiated/moderately differentiated. G3/G4, poorly differentiated/undifferentiated.

The multivariate analysis results revealed that several clinical factors serve as independent prognostic indicators for OS in patients diagnosed with OC. Advanced clinical stage (III/IV vs I/II: HR = 2.707, 95% CI = 1.410–7.101, p = 0.042) and tumor presence (tumor-positive vs. tumor-negative: HR = 17.169, 95% CI = 4.192–70.317, p < 0.001) significantly influenced OS. Compared with progressive disease, stable disease was associated with poorer survival (HR = 1.520, 95% CI = 1.096–1.246, p = 0.036), whereas complete response versus progressive disease was associated with better survival (HR = 0.194, 95% CI = 0.114–0.329, p < 0.001). Age older than 60 years versus 60 years or younger was also a significant factor (HR = 2.324, 95% CI = 1.816–1.965, p = 0.042), as was lymph node involvement (yes vs no: HR = 1.368, 95% CI = 1.060–2.857, p = 0.036). Conversely, residual tumor status (yes vs no: hazard ratio (HR) = 1.116, 95% CI = 0.189–0.660, p = 0.682) was not statistically significant in the multivariate model, as detailed in Table 3.

Table 3

COL10A1 expression correlated with clinicopathological characteristics analyzed by COX regression
Characteristics	TCGA cohort (n = 381)				Validation cohort (n = 62)
	Univariate analysis		Multivariate analysis		Univariate analysis			Multivariate analysis
	Hazard ratio (95% CI)	P value	Hazard ratio (95% CI)	P value		Hazard ratio (95% CI)	P value		Hazard ratio (95% CI)	P value
Clinical stage
Stage I&Stage II	Reference		Reference			Reference			Reference
Stage III&Stage IV	3.135 (1.947–6.811)	0.037	2.707 (1.410–7.101)	0.042		4.310 (1.301–14.279)	0.017		1.571 (1.031–2.393)	0.035
Tumor status
Tumor free	Reference		Reference			Reference			Reference
With tumor	9.598 (4.487–20.532)	< 0.001	17.169 (4.192–70.317)	< 0.001		1.970 (1.339–2.899)	< 0.001		2.012 (1.191–3.399)	0.009
Primary therapy outcome
PD	Reference		Reference			Reference			Reference
SD	0.441 (0.217–0.896)	0.024	1.520 (1.246–1.096)	0.036		1.844 (1.189–2.857)	0.006		1.715 (1.100–2.674)	0.017
PR	0.652 (0.384–1.108)	0.114	0.571 (0.324–1.008)	0.053		1.140 (0.557–2.333)	0.72
CR	0.154 (0.095–0.250)	< 0.001	0.194 (0.114–0.329)	< 0.001		2.987 (1.860–4.797)	< 0.001		2.055 (1.418–2.978)	< 0.001
Age
<= 60	Reference		Reference			Reference			Reference
> 60	1.352 (1.045–1.749)	0.022	2.324 (1.965–1.816)	0.042		2.236 (1.291–3.875)	0.004		0.828 (0.521–1.317)	0.425
Histologic grade
G1&G2	Reference					Reference			Reference
G3&G4	1.239 (0.838–1.833)	0.283				3.747 (1.404–10.004)	0.008		2.844 (1.446–5.593)	0.002
Venous invasion
No	Reference					Reference
Yes	0.896 (0.487–1.649)	0.723				0.744 (0.465–1.191)	0.218
Lymphatic invasion
No	Reference		Reference			Reference			Reference
Yes	1.613 (1.854–3.387)	0.020	1.368 (1.060–2.875)	0.036		1.931 (1.208–3.088)	0.006		1.779 (1.192–2.653)	0.005
Tumor residual
No	Reference		Reference			Reference			Reference
Yes	2.223 (1.441–3.430)	< 0.001	1.116 (0.660–1.890)	0.682		2.959 (1.224–7.157)	0.016		2.122 (0.634–7.106)	0.222

Construction and validation of a nomogram

To predict the prognosis of OC patients, nomogram plots were created using independent factors related to OS (Fig. 5B). The results from the nomogram suggested that a higher cumulative score was correlated with a worse prognosis. To evaluate the nomogram's efficacy in distinguishing patients with different clinical outcomes, we constructed an ROC curve for COL10A1 (Fig. 5C), revealing its good diagnostic utility for OC (AUC = 0.867). Furthermore, calibration curves generated via the bootstrap method were employed to assess the predictive performance of the nomogram, resulting in a concordance index (C-index) of 0.867 (95% CI = 0.833–0.901), indicating good predictive accuracy for OS in OC patients. The calibration curves also demonstrated that the predicted survival rates at 1, 3, and 5 years (Figs. 5D-F) were in close agreement with the observed survival rates, underscoring the applicability of the nomogram in a clinical context. Together, the ROC analysis and calibration curves validated the nomogram as a reliable tool for assessing survival outcomes in OC patients.

Mechanisms of COL10A1 in OC

We subsequently evaluated the relationships between the expression levels of the DEGs and COL10A1 through Spearman’s correlation analysis (Fig. 6A-B). To further elucidate the mechanistic role of COL10A1 in OC-CAFs, we performed GO and KEGG pathway analyses. Our findings revealed that the biological processes (BPs) significantly influenced by COL10A1 include external encapsulating structure organization, extracellular structure organization, and extracellular matrix organization. Additionally, the identified CCs include the collagen-containing extracellular matrix and components of the basal cell layer. The molecular functions (MFs) associated with OC, such as extracellular matrix structural constituents and cell adhesion mediator activity, were also notably regulated by COL10A1 (Fig. 6C). To link these analyses, both the GO and KEGG results highlight the involvement of COL10A1 in pathways critical to OC progression. Furthermore, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis revealed that COL10A1 is linked primarily to the PI3K‒AKT signaling pathway, whereas genes related to cell adhesion molecules were strongly enriched in ECM‒receptor interaction pathways (Fig. 6C). These findings suggest that COL10A1 in CAFs likely facilitates the progression of ovarian cancer through these pathways, thereby reaffirming the critical role of COL10A1 in the onset and progression of ovarian cancer, as well as its potential utility in prognostic assessments.

Fig. 6

Functional enrichment analysis and immune infiltration level of COL10A1 in ovarian cancer. (A) Correlation chord plot of the correlation between COL10A1 expression and the DEGs. (B) Heatmap of the correlation between COL10A1 expression and DEGs. (C) GO and KEGG analyses of the DEGs. (D) Correlations between COL10A1 expression and the relative abundances of 24 types of immune cells. The size of the dots corresponds to the absolute Spearman’s correlation coefficient values. (E) Comparison of the immune infiltration levels of 24 types of immune cells between the high- and low-COL10A1 expression groups. (F) Correlations between COL10A1 expression and immune subtypes in ovarian cancer. (G) COL10A1 expression varied among different immune subtypes. (H) Differences in immune checkpoint gene expression between patients with high and low COL10A1 levels. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

Correlation between COL10A1 expression and immune infiltration

This investigation further used the CIBERSORT algorithm to assess the infiltration levels of 24 distinct immune cell types within OC tumor tissues. The analysis revealed a significant positive correlation between the expression of COL10A1 and the infiltration of most immune cell types (Fig. 6D, p < 0.05). These include various immune cells, such as macrophages; immature dendritic cells (iDCs); mast cells; Th1 cells; effector memory T (Tem) cells; CD8 + T cells; CD56dim natural killer (NK) cells; neutrophils; T cells; regulatory T (Treg) cells; B cells; and dendritic cells (DCs). Stratifying patients into high and low COL10A1 expression cohorts on the basis of median levels revealed that the high-expression cohort exhibited a significant increase in immune cell populations, including cytotoxic cells, DCs, iDCs, macrophages, mast cells, T cells, and Tregs, compared with the low-expression cohort (Fig. 6E, p < 0.05). The expression levels of COL10A1 varied across the four molecular subtypes (C1–C4), with the highest expression observed in the wound healing subtype and the lowest in the lymphocyte depletion subtype (Fig. 6F). Notably, tumor infiltration levels were influenced by copy number variation (CNV) in COL10A1, with macrophage infiltration levels significantly reduced in patients with chromosome arm-level deletion of COL10A1 (Fig. 6G, p < 0.05).

In addition, normal tissues were used as controls to determine the relationship between COL10A1 expression and gene markers of immune cells in OC. This analysis included B cells; general T cells; CD4 + and CD8 + T cells; monocytes; mast cells; TAMs; M1 and M2 macrophages; neutrophils; NK cells; dendritic cells; Th1, Th2, TFH, Th17, and Tregs; and various immune checkpoints (Supplementary Table 2). Finally, we assessed the expression of immune checkpoint genes in patients with high and low COL10A1 levels via data from the TCGA database. The findings indicated that the high-expression group of COL10A1 presented elevated expression levels of nine genes (PD-L1 [CD274], CTLA4, PD-L2, LAG3, CD276, LMTK3, TIGIT, TIM3), suggesting a potential association between COL10A1 and the immune response in OC (Fig. 6H) (22).

Clinical features and laboratory data

The clinical characteristics of the participants involved in this investigation are summarized in Table 1. During the detection and assessment phases, the mean age of individuals with normal ovarian function was 51.1 years (SD = 11.8), whereas patients diagnosed with OC had a mean age of 50.9 years (SD = 12.1). Within the cohort, 6 patients were classified as stage I, and 24 patients were classified as stage II. The distribution further included 30 patients at stage III and 12 patients at stage IV. Notably, 33 patients exhibited venous infiltration, whereas 23 did not. Additionally, lymph node metastasis was present in 24 patients and absent in 38 patients. Tumor grading revealed that 30 cases were Grade 1 or Grade 2, whereas 32 cases were Grade 3. Furthermore, there were 28 patients with residual tumors and 34 patients without residual tumors. The cohort comprises 28 patients with active tumors and 34 patients in remission.

Validation of the elevated expression of COL10A1 in external datasets and clinical patients

To assess the expression levels of COL10A1 in patients diagnosed with OC, we conducted a comparative analysis of COL10A1 expression across various clinicopathological parameters via the datasets GSE26712, GSE66957, GSE17260, GSE140082, GSE9891, and GSE26193 (Fig. 7A-H). Our findings revealed that COL10A1 was markedly overexpressed in OC tissues and was significantly associated with clinical stage, tumor grade, and differentiation status. Additionally, we collected tissue samples from 62 newly diagnosed OC patients and 30 control patients (normal ovarian tissues). Compared with those in the control cohort, the COL10A1 expression levels in OC patient tissues were significantly elevated (Fig. 7I, p < 0.0001). Notably, patients with advanced OC presented with higher COL10A1 levels (Fig. 7J, p < 0.0001). Furthermore, COL10A1 expression in OC tissues with lymph node metastasis was significantly greater than that in tissues from OC patients without lymph node metastasis (Fig. 7K, p < 0.0001). Interestingly, COL10A1 levels were higher in low-grade OC tissues than in high-grade counterparts (Fig. 7L, p < 0.05), suggesting a complex relationship between COL10A1 expression and tumor differentiation that warrants further investigation. Importantly, the mRNA expression levels of COL10A1 were significantly correlated with key clinical parameters, including clinical stage, tumor grade, lymph node metastasis, and overall tumor status, which encompasses tumor size, metastasis, and patient prognosis (Table 1 and Table 2).

Fig. 7

Associations between COL10A1 mRNA expression levels and the clinical characteristics of OC patients. (A) COL10A1 expression in ovarian cancer and normal ovaries from patients in GSE26712. (B) COL10A1 expression in ovarian cancer and normal ovaries from patients in GSE66957. (C) The expression levels of COL10A1 in different groups according to tumor grade in GSE17260. (D) The expression levels of COL10A1 in different groups according to tumor grade in GSE140082. (E) The expression levels of COL10A1 in different groups at different clinical stages in GSE9891. (F) The expression levels of COL10A1 in different groups at different clinical stages in the GSE26193 dataset. (G) The expression levels of COL10A1 in different groups at different clinical stages in the GSE17260 dataset. (H) The expression levels of COL10A1 in different groups at different clinical stages in GSE140082. (I) The mRNA expression of COL10A1 in ovarian cancer and normal ovaries from patients was measured via qRT‒PCR. (J) The mRNA expression levels of COL10A1 in different groups at different clinical stages were measured via qRT‒PCR. (K) The mRNA expression levels of COL10A1 in different groups of patients with lymph node metastasis were measured via qRT‒PCR. (L) mRNA expression levels of COL10A1 in different groups according to tumor grade, as determined via qRT‒PCR. N, normal ovary. OC, ovarian cancer. G1/G2, well differentiated/moderately differentiated. G3/G4, poorly differentiated/undifferentiated. N0, without lymph node metastases; N1, with lymph node metastases. *P < 0.05; ****P < 0.0001.

Furthermore, we conducted an immunohistochemical analysis to assess COL10A1 expression in 30 normal ovarian tissue samples and 63 tissue samples. The results revealed positive immunohistochemical staining of COL10A1, indicated by brown granules (Fig. 8A). COL10A1 expression was significantly elevated in OC tissues compared with normal tissues (Fig. 8B, p < 0.0001). The expression levels were greater in advanced-stage OC than in early-stage OC (Fig. 8C, p < 0.01). Additionally, tissues from patients with lymph node metastasis presented increased COL10A1 expression compared with those without metastatic involvement (Fig. 8D, p < 0.05). Among the 39 samples from patients with advanced-stage OC, 23 had high immunohistochemical expression of COL10A1, accounting for 59.0% of the total sample. In contrast, only 5 out of 30 early-stage OC samples presented high COL10A1 expression, representing 22.7% of the total sample. This difference was statistically significant (Fig. 8E, P < 0.05). Among the 34 samples from patients with lymph node metastasis, 19 demonstrated high COL10A1 expression, accounting for 55.9% of the samples. Moreover, 9 out of 27 early-stage OC samples presented high expression, accounting for 33.3% (Fig. 8F, P < 0.05). Among the 63 tissue samples analyzed, 54 were positive for COL10A1 expression, with an overall positive staining rate of 85.71%. In contrast, 15 out of 30 normal ovarian tissue samples were positive, resulting in a positive staining rate of 50% (Fig. 8G, P < 0.05). Among the advanced-stage OC samples, 36 out of 39 were positive for COL10A1, reflecting a positive staining rate of 92.3%. Conversely, 16 out of 30 early-stage OC samples were positive, yielding a positive staining rate of 72.7% (Fig. 8H, P < 0.05). Finally, among the 34 lymph node metastatic OC samples, 31 were positive for COL10A1 expression, corresponding to a positive staining rate of 91.18%. In comparison, 21 out of 27 early-stage OC samples were positive, representing a positive staining rate of 77.78% (Fig. 8I, P < 0.05). These findings strongly validate the elevated expression of COL10A1 in patients diagnosed with OC.

Fig. 8

Immunohistochemical staining results of normal ovary and ovarian cancer tissues. (A) Immunohistochemical staining results of ovarian cancer and normal ovaries (200× and 400×). (B) Protein expression of COL10A1 in ovarian cancer and normal ovaries from patients, as determined by IHC. (C) Protein expression levels of COL10A1 in different groups at different clinical stages determined by IHC. (D) The protein expression levels of COL10A1 in different groups of patients with lymph node metastasis were determined by IHC. (E) Proportion of samples with high expression of COL10A1 protein in different clinical stages, as determined by IHC. (F) Proportion of samples with high expression of the COL10A1 protein in different lymph node metastases, as determined by IHC. (G) Proportion of samples with positive COL10A1 protein expression in ovarian cancer and normal ovarian tissues. (H) The proportion of samples with positive COL10A1 protein expression in different clinical stages determined by IHC. (I) Proportion of samples with positive COL10A1 protein expression in different lymph node metastases determined by IHC. (J) Kaplan‒Meier curves showing the associations between the expression of COL10A1 and OS according to the immunohistochemical results. (K) Kaplan‒Meier curves showing the association between the expression of COL10A1 and DFS according to the immunohistochemical results. N, normal ovary. OC, ovarian cancer. I/II, early stage (clinical stages I and II). III/IV, advanced stage (clinical stage III and IV). N0, without lymph node metastases; N1, with lymph node metastases. IRS, immune responsive score. HR, hazard ratio. *P < 0.05; **P < 0.01; ****P < 0.0001.

Validation of prognosis evaluation

Overall survival and progression-free survival were significantly longer in OC patients with elevated COL10A1 expression in tumor tissues than in those with decreased COL10A1 expression (Fig. 8J, p = 0.0009; Fig. 8K, p = 0.0001). As detailed in Table 3, univariate Cox regression analysis revealed that several factors, including clinical stage (HR = 4.310, p = 0.017), tumor status (HR = 1.970, p < 0.001), primary treatment outcome (HR = 1.844, p = 0.006; R = 2.987, p < 0.001), age (HR = 2.236, p = 0.004), tumor grade (HR = 3.747, p = 0.008), lymph node invasion (HR = 1.931, p = 0.006), and residual tumor (HR = 2.959, p = 0.0016), were significantly correlated with poor prognosis. Notably, multivariate Cox regression analysis indicated that clinical stage (HR = 1.571, p = 0.035), tumor status (HR = 2.012, p = 0.009), primary treatment outcome (HR = 1.715, p = 0.017; R = 2.055, p < 0.001), tumor grade (HR = 2.844, p = 0.002), and lymph node invasion (HR = 1.779, p = 0.005) were independent prognostic indicators influencing overall survival in OC patients. In summary, these findings underscore the notable prognostic significance of COL10A1 expression levels in individuals diagnosed with OC.

Discussion

OC is the gynecological malignancy with the highest mortality rate and consistently has a five-year survival rate of less than 50%. Although significant advancements have been made in the diagnosis and treatment of ovarian cancer, persistent challenges in improving long-term survival rates remain, largely due to the growing issue of drug resistance, which continues to hinder effective therapy(23). Therefore, there is a pressing need to identify novel prognostic indicators for ovarian cancer, as well as to develop innovative treatment protocols.

Collagen serves as the primary constituent of the extracellular matrix. An increasing body of research has validated its role in facilitating tumor initiation and metastasis(24). Recent investigations have revealed that collagen exerts an immunomodulatory effect within the tumor microenvironment, particularly influencing tumor-associated macrophages and T cells, which in turn affect tumor progression, prognosis, and responses to immunotherapy(25). The immunomodulatory properties of tumor-associated collagen fibers may pave the way for the development of current treatment strategies and novel therapeutic interventions(26).

The COL10A1 gene, which encodes the alpha 1 chain of type X collagen, is located on human chromosome 6q22.1. This collagen type serves as a principal matrix component within the extracellular matrix and is expressed predominantly in hypertrophic chondrocytes located in the bone growth plate. Under physiological conditions, it plays a significant role in endochondral ossification, thereby regulating bone development and mineralization. Its protein structure features a highly conserved COLF1 domain (triple helix region) and an NC1 domain (C-terminal noncollagen region), with the latter being essential for collagen trimer assembly and the stability of the ECM. While earlier studies have focused primarily on its implications in bone and joint disorders (27, 28), more recent research has revealed elevated COL10A1 expression in various tumor tissues(29). This finding has garnered increasing attention regarding the mechanisms by which COL10A1 may influence cancer progression, as well as the potential diagnostic and therapeutic applications of COL10A1 in clinical oncology.

In this investigation, we conducted a comprehensive screening analysis using the TCGA and GEO databases and reported that COL10A1 is closely related to multiple clinical parameters in OC. Next, we employed the TCGA database to assess the expression levels of COL10A1, revealing that high expression is notably associated with OS in OC patients. These results were subsequently verified in the GEO database. To further substantiate our results, we measured the mRNA expression levels of COL10A1 in 62 tissue samples and 30 normal ovarian tissue samples via qRT‒PCR. Additionally, we evaluated the protein expression of COL10A1 in 63 tissues and 30 normal ovarian samples via IHC, and the results were consistent with the above bioinformatics research results. In our study, high expression of COL10A1 was associated with clinicopathological features such as tumor stage, grade and lymph node metastasis in OC patients. We divided OC patients into two groups, one with high COL10A1 protein expression and the other with low COL10A1 protein expression, through qRT‒PCR and IHC staining, respectively. The Kaplan‒Meier survival curve indicated that patients with high COL10A1 protein expression might have a poorer clinical prognosis than those with low COL10A1 protein expression. This further validated the results of the sequencing data analysis. Overall, the results of this study indicate that COL10A1 is a promising diagnostic and prognostic biomarker for patients with OC.

To elucidate the underlying mechanisms contributing to its pronounced growth behavior, we performed GO and KEGG enrichment analyses on the TCGA dataset. These enrichment analyses are associated with several well-established signaling pathways related to the ECM, including the PI3K‒Akt signaling pathway and ECM‒receptor interactions. Previous investigations have established connections between COL10A1 and various biological processes and pathways, such as the PI3K‒Akt signaling pathway, EMT, inflammatory responses, apoptosis, TGF-β signaling, and hypoxia, all of which are relevant to the onset and progression of cancer(30), which is consistent with our research results. Tumor-infiltrating immune cells constitute a crucial element of the tumor microenvironment and significantly influence tumor growth, progression, treatment efficacy, and patient prognosis(31). Research has indicated that a higher immune infiltration score is notably correlated with CR and PFS outcomes following pembrolizumab therapy(32, 33).

The TIMER methodology was used to investigate the relationship between COL10A1 expression and the degree of tumor immune infiltration. The findings indicated that samples exhibiting elevated COL10A1 expression presented notable increases in the numbers of cytotoxic T cells, DCs, iDCs, macrophages, mast cells, T cells, and Tregs. Furthermore, variations in tumor immune infiltration were observed across different COL10A1 copy number variations. Specifically, macrophage infiltration was reduced in cases where the chromosomal arm, containing COL10A1, was absent. Analysis via the GEPIA web server revealed a significant positive correlation between COL10A1 expression and gene markers associated with TAMs, as well as M1 and M2 macrophages. These results suggest that COL10A1 may modulate immune infiltration in OC by affecting macrophage abundance or activation. In summary, macrophages may play a pivotal role in shaping COL10A1 expression within the immune microenvironment of OC.

Macrophages are a universal cellular component in all tissues and body chambers(34). Macrophages play a dual role in cancer by exerting both tumor-promoting and antitumor effects(35). COL10A1 is associated with increased infiltration of M2-type macrophages in prostate cancer(12). In colorectal cancer, the expression of COL10A1 can promote the aggregation of tumor-associated fibroblasts(36), thereby promoting epithelial‒mesenchymal transition (EMT) and metastasis, which are closely related to tumor metastasis(37). M2-polarized macrophages play a significant role in tumor-promoting and anti-inflammatory activities(38), which may be the fundamental reason for the poor prognosis of OC patients with high COL10A1 expression. We speculate that COL10A1 may play a significant role in the recruitment of infiltrative immune cells and the regulation of immunity in OC, thereby influencing patient prognosis. However, more research is needed to confirm this hypothesis, especially the effect of COL10A1 on the M2 polarization of macrophages in the OC microenvironment.

Macrophages represent a fundamental cellular element present in all tissues and bodily compartments. They play a dichotomous role in cancer, functioning as both promoters and inhibitors of tumor development. COL10A1 expression is correlated with increased infiltration of M2 macrophages in prostate cancer. In colorectal cancer, COL10A1 expression facilitates the accumulation of tumor-associated fibroblasts. This accumulation enhances EMT and metastasis, processes closely associated with tumor spread. M2-polarized macrophages play a pivotal role in mediating tumor-promoting and anti-inflammatory responses; this dual function may be a critical factor contributing to the unfavorable prognosis observed in OC patients with elevated COL10A1 levels. We hypothesize that COL10A1 significantly influences the recruitment of infiltrating immune cells and modulates immune responses in OC, thereby affecting patient outcomes. Nonetheless, further investigation is warranted to validate this hypothesis, particularly regarding the impact of COL10A1 on M2 macrophage polarization within the OC microenvironment.

Conclusion

COL10A1 is overexpressed in OC-CAFs and is correlated with poor prognosis and key clinicopathological features. Functionally, it modulates OC progression via the PI3K-AKT and ECM-receptor signaling pathways and influences immune checkpoint expression and immune cell infiltration in the tumor microenvironment. These findings highlight high COL10A1 expression in CAFs as a promising prognostic biomarker and therapeutic target for OC.

Acknowledgments

Not applicable.

Funding Source

Declaration

This study was supported in part by grants from the National Natural Science Foundation of China (Nos. 82272332 and 82203652), the joint construction project of the Henan Provincial Medical Science and Technology Research Program (LHGJ20220534), and the Henan Province Key R&D and Promotion Special Project (Scientific and Technological Research Project) (242102311175).

Data Availability

The mRNA sequencing data can be gained from the TCGA and GEO database. Further inquiries can be directed to the corresponding author.

Author Contribution

Shujun Zhao, Gaili Ji and Qing Liu conceived and designed the study. Yaping Wang and Min Tian wrote the manuscript. Xiabing Li, Luyao Kang, Hongjian Zhang, Yiran Wang, Hongyu Li, Qing Liu and Hai Zhu collected relevant information. Caixia Ma revised the manuscript. All the authors read and approved the final version of the manuscript.

Ethics approval and consent to participate

This research received approval from the Medical Ethics Committee of the Third Affiliated Hospital of Zhengzhou University under Approval No.

2025-01-236 and conducted in accordance with the Declaration of Helsinki.

Patient consent for publication

Not applicable.

Declaration of competing interest

The authors declare that they have no competing interests in this section.

References

Armstrong DK, Alvarez RD, Backes FJ, Bakkum-Gamez JN, Barroilhet L, Behbakht K, et al. NCCN Guidelines® Insights: Ovarian Cancer, Version 3.2022. J Natl Compr Canc Netw. 2022;20(9):972–80.

Wang L, Wang X, Zhu X, Zhong L, Jiang Q, Wang Y, et al. Drug resistance in ovarian cancer: from mechanism to clinical trial. Mol Cancer. 2024;23(1):66.

Kuroki L, Guntupalli SR. Treatment of epithelial ovarian cancer. BMJ. 2020;371:m3773.

Sherman MH, Beatty GL. Tumor Microenvironment in Pancreatic Cancer Pathogenesis and Therapeutic Resistance. Annu Rev Pathol. 2023;18:123–48.

Vickers R, Porter W. Immune Cell Contribution to Mammary Gland Development. J Mammary Gland Biol Neoplasia. 2024;29(1):16.

Yamada KM, Doyle AD, Lu J. Cell-3D matrix interactions: recent advances and opportunities. Trends Cell Biol. 2022;32(10):883–95.

Liu J, Chai XX, Qiu XR, Sun WJ, Tian YL, Guo WH, et al. Type X collagen knockdown inactivate ITGB1/PI3K/AKT to suppress chronic unpredictable mild stress-stimulated triple-negative breast cancer progression. Int J Biol Macromol. 2024;273(Pt 1):133074.

Nagarajan A, Varadhan V, Manikandan MS, Kaliaperumal K, Palaniyandi T, Kaliamoorthy S, et al. Signature of collagen alpha-1(x) gene expression in human cancers and their therapeutic implications. Pathol Res Pract. 2025;266:155811.

Meng Z, Zhang R, Wu X, Piao Z, Zhang M, Jin T. LncRNA HAGLROS promotes breast cancer evolution through miR-135b-3p/COL10A1 axis and exosome-mediated macrophage M2 polarization. Cell Death Dis. 2024;15(8):633.

10.

Wang X, Ma S, Li S, Jia W, Zhang D. The regulatory effect of CoL10A1 to the intracranial vascular invasion and cell proliferation in breast cancer via EMT pathway. Sci Rep. 2025;15(1):11040.

11.

Wang X, Bai Y, Zhang F, Li D, Chen K, Wu R, et al. Prognostic value of COL10A1 and its correlation with tumor-infiltrating immune cells in urothelial bladder cancer: A comprehensive study based on bioinformatics and clinical analysis validation. Front Immunol. 2023;14:955949.

12.

Jiang K, Xu LZ, Cheng F, Ning JZ. COL10A1 Facilitates Prostate Cancer Progression by Interacting With INHBA to Activate the PI3K/AKT Pathway. J Cell Mol Med. 2024;28(23):e70249.

13.

Peng J, Liu X, Mao Y, Lv M, Ma T, Liu J, et al. Upregulation of collagen type X alpha 1 promotes the progress of triple-negative breast cancer via Wnt/β-catenin signaling. Mol Carcinog. 2024;63(8):1588–98.

14.

Yang W, Wu X, Zhou F, Collagen Type X. Alpha 1 (COL10A1) Contributes to Cell Proliferation, Migration, and Invasion by Targeting Prolyl 4-Hydroxylase Beta Polypeptide (P4HB) in Breast Cancer. Med Sci Monit. 2021;27:e928919.

15.

Zhou W, Li Y, Gu D, Xu J, Wang R, Wang H, et al. High expression COL10A1 promotes breast cancer progression and predicts poor prognosis. Heliyon. 2022;8(10):e11083.

16.

Zhang M, Jin M, Gao Z, Yu W, Zhang W. High COL10A1 expression potentially contributes to poor outcomes in gastric cancer with the help of LEF1 and Wnt2. J Clin Lab Anal. 2022;36(9):e24612.

17.

Zheng Y, Jiang P, Tu Y, Huang Y, Wang J, Gou S, et al. Incidence, risk factors, and a prognostic nomogram for distant metastasis in endometrial cancer: A SEER-based study. Int J Gynaecol Obstet. 2024;165(2):655–65.

18.

Zhou Y, Zhou B, Pache L, Chang M, Khodabakhshi AH, Tanaseichuk O, et al. Metascape provides a biologist-oriented resource for the analysis of systems-level datasets. Nat Commun. 2019;10(1):1523.

19.

Thorsson V, Gibbs DL, Brown SD, Wolf D, Bortone DS, Ou Yang TH, et al. The Immune Landscape of Cancer. Immunity. 2018;48(4):812–e3014.

20.

Newman AM, Steen CB, Liu CL, Gentles AJ, Chaudhuri AA, Scherer F, et al. Determining cell type abundance and expression from bulk tissues with digital cytometry. Nat Biotechnol. 2019;37(7):773–82.

21.

Sepulveda JL. Using R and Bioconductor in Clinical Genomics and Transcriptomics. J Mol Diagn. 2020;22(1):3–20.

22.

Xu Q, Zheng J, Su Z, Chen B, Gu S. COL10A1 promotes tumorigenesis by modulating CD276 in pancreatic adenocarcinoma. BMC Gastroenterol. 2023;23(1):397.

23.

Konstantinopoulos PA, Matulonis UA. Clinical and translational advances in ovarian cancer therapy. Nat Cancer. 2023;4(9):1239–57.

24.

De Martino D, Bravo-Cordero JJ. Collagens in Cancer: Structural Regulators and Guardians of Cancer Progression. Cancer Res. 2023;83(9):1386–92.

25.

Maller O, Drain AP, Barrett AS, Borgquist S, Ruffell B, Zakharevich I, et al. Tumour-associated macrophages drive stromal cell-dependent collagen crosslinking and stiffening to promote breast cancer aggression. Nat Mater. 2021;20(4):548–59.

26.

Rømer AMA, Thorseth ML, Madsen DH. Immune Modulatory Properties of Collagen in Cancer. Front Immunol. 2021;12:791453.

27.

Wu H, Wang S, Li G, Yao Y, Wang N, Sun X, et al. Characterization of a novel COL10A1 variant associated with Schmid-type metaphyseal chondrodysplasia and a literature review. Mol Genet Genomic Med. 2021;9(5):e1668.

28.

Wei X, Qiu J, Lai R, Wei T, Lin Z, Huang S, et al. A human organoid drug screen identifies α2-adrenergic receptor signaling as a therapeutic target for cartilage regeneration. Cell Stem Cell. 2024;31(12):1813–e308.

29.

Thorlacius-Ussing J, Jensen C, Nissen NI, Cox TR, Kalluri R, Karsdal M, et al. The collagen landscape in cancer: profiling collagens in tumors and in circulation reveals novel markers of cancer-associated fibroblast subtypes. J Pathol. 2024;262(1):22–36.

30.

Zhang F, Wang X, Bai Y, Hu H, Yang Y, Wang J, et al. Development and Validation of a Hypoxia-Related Signature for Predicting Survival Outcomes in Patients With Bladder Cancer. Front Genet. 2021;12:670384.

31.

Hu S, Ding M, Lou J, Qin J, Chen Y, Liu Z, et al. COL10A1(+) fibroblasts promote colorectal cancer metastasis and M2 macrophage polarization with pan-cancer relevance. J Exp Clin Cancer Res. 2025;44(1):243.

32.

Yi Q, Zhu G, Zhu W, Wang J, Ouyang X, Yang K et al. Oncogenic mechanisms of COL10A1 in cancer and clinical challenges (Review). Oncol Rep. 2024;52(6).

33.

Necchi A, Raggi D, Gallina A, Ross JS, Farè E, Giannatempo P, et al. Impact of Molecular Subtyping and Immune Infiltration on Pathological Response and Outcome Following Neoadjuvant Pembrolizumab in Muscle-invasive Bladder Cancer. Eur Urol. 2020;77(6):701–10.

34.

Li M, Yang Y, Xiong L, Jiang P, Wang J, Li C. Metabolism, metabolites, and macrophages in cancer. J Hematol Oncol. 2023;16(1):80.

35.

Mantovani A, Allavena P, Marchesi F, Garlanda C. Macrophages as tools and targets in cancer therapy. Nat Rev Drug Discov. 2022;21(11):799–820.

36.

Esposito M, Yerly L, Shukla P, Hermes V, Sella F, Balazs Z, et al. COL10A1 expression distinguishes a subset of cancer-associated fibroblasts present in the stroma of high-risk basal cell carcinoma. Br J Dermatol. 2024;191(5):775–90.

37.

Wu J, Ouyang P, Huang R, Cui Y, Yang Z, Xu W, et al. METTL16 Promotes Stability of SYNPO2L mRNA and leading to Cancer Cell Lung Metastasis by Secretion of COL10A1 and attract the Cancer-Associated Fibroblasts. Int J Biol Sci. 2024;20(11):4128–45.

38.

Boutilier AJ, Elsawa SF. Macrophage Polarization States in the Tumor Microenvironment. Int J Mol Sci. 2021;22(13).

Legends

Supplementary Fig. 1.

Expression of CHI3L1 in ovarian cancer. (A) Expression data after the median high and low populations were grouped on the basis of the expression level of CHI3L1. (B) Expression of CHI3L1 at different stages: early stage (I/II) and advanced stage (III/IV). (C) Expression of CHI3L1 at different grades: well differentiated/moderately differentiated (G1/G2) and poorly differentiated/undifferentiated (G3/G4). (D) The expression of CHI3L1 in tissues of different sizes. (E) The expression of CHI3L1 under different lymph node infiltration states. (F) The expression of CHI3L1 in different vascular metastasis states.

Supplementary Fig. 2.

The expression of MAFB in OC. (A) Expression data after the median high and low populations were grouped on the basis of the expression level of MAFB. (B) MAFB expression in different stages, namely, the early stage (I/II) and advanced stage (III/IV). (C) The expression of MAFB at different grades: well differentiated/moderately differentiated (G1/G2) and poorly differentiated/undifferentiated (G3/G4). (D) MAFB expression in tissues of different sizes. (E) MAFB expression under different lymph node infiltration states. (F) MAFB expression in different vascular metastasis states.

Supplementary Fig. 3.

Expression of MMP11 in ovarian cancer. (A) Expression data after the median high and low populations were grouped on the basis of the expression level of MMP11. (B) The expression of MMP11 in different stages, early stage (I/II) and advanced stage (III/IV) disease. (C) The expression of MMP11 at different grades: well differentiated/moderately differentiated (G1/G2) and poorly differentiated/undifferentiated (G3/G4). (D) The expression of MMP11 in tissues of different sizes. (E) The expression of MMP11 under different lymph node infiltration states. (F) The expression of MMP11 in different vascular metastasis states.

Supplementary Fig. 4.

MOXD1 expression in ovarian cancer. (A) Expression data after the median high and low populations were grouped on the basis of the expression level of MOXD1. (B) Expression of MOXD1 at different stages: early stage (I/II) and advanced stage (III/IV). (C) Expression of MOXD1 at different grades: well differentiated/moderately differentiated (G1/G2) and poorly differentiated/undifferentiated (G3/G4). (D) MOXD1 expression in tissues of different sizes. (E) MOXD1 expression in different lymph node infiltration states. (F) Expression of MOXD1 in different vascular metastasis states.

Supplementary Fig. 5.

Expression of MX2 in ovarian cancer. (A) Expression data after the median high and low populations were grouped on the basis of the expression level of MX2. (B) MX2 expression in different stages, namely, early stage (I/II) and advanced stage (III/IV). (C) The expression of MX2 at different grades: well differentiated/moderately differentiated (G1/G2) and poorly differentiated/undifferentiated (G3/G4). (D) MX2 expression in tissues of different sizes. (E) MX2 expression in different lymph node infiltration states. (F) MX2 expression in different vascular metastasis states.

Supplementary Fig. 6.

COL10A1 expression levels are correlated with infiltrating immune cells in ovarian cancer.

Yes