This article is mentioned in:

Introduction

Ovarian cancer is one of the three most common malignant tumors of the female reproductive system. The mortality rate of ovarian cancer ranks first among gynecological malignancies, and the 5-year survival rate is only ~40% (1,2). Epithelial ovarian cancer is the most common histological type, accounting for ~90% of all ovarian cancers (3). Epithelial ovarian cancer cells are highly invasive, and widespread metastasis to the pelvis, abdominal cavity and retroperitoneal lymph nodes can occur in the early stage. Even if patients receive standard treatment, ~70% of the advanced patients experience recurrence, metastasis and drug resistance within 2–3 years (4), which poses a serious threat to life and health.

Transmembrane 4 L6 family member 1 (TM4SF1) is a distant relative of the four-transmembrane protein superfamily. It is highly expressed in a variety of epithelial cancer tissues, and it regulates intracellular calcium levels, tyrosine phosphorylation and protein kinase C (5,6). In addition, TM4SF1 is involved in the formation of vascular endothelial pseudopodia and angiogenesis, and plays an important role in the regulation of cell development, activation, growth and movement (7). It has been reported that the upregulation of TM4SF1 in a variety of epithelial cancer tissues is associated with a poor prognosis (8–11). Discoidin domain receptor 1 (DDR1) is a member of the DDR family of receptor tyrosine kinases, and the abnormal upregulation of DDR1 is associated with the occurrence and development of tumors. The significantly upregulated expression of DDR1 has been detected in numerous types of malignant tumors (12), including breast cancer, non-small cell lung cancer, gastric cancer and colorectal cancer, and DDR1 is regarded as an adverse prognostic factor (13–16). Similar to TM4SF1, DDR1 is critical for tumor cell survival, drug resistance, self-renewal, differentiation, adhesion and migration, and participates in the regulation of tumor progression and metastasis (17).

According to a previous study, TM4SF1 and DDR1 interact with each other and activate resting cancer cells in distant organs to cause tumor recurrence and metastasis (18). However, the relationship of the expression of TM4SF1 and DDR1 with epithelial ovarian cancer prognosis remains unclear. Therefore, the present study investigated the expression of TM4SF1 and DDR1 in epithelial ovarian cancer tissues and the association of TM4SF1 and DDR1 expression levels with the prognosis of patients.

Materials and methods

Patients and tissue samples

Patients were treated at the Department of Gynecological Oncology of Guangxi Medical University Cancer Hospital (Nanning, China) between January 2013 and June 2017. Inclusion criteria were as follows: Patients with epithelial ovarian cancer, >18 years of age, who had received cytoreductive or comprehensive staging surgery and adjuvant chemotherapy in Guangxi Medical University Cancer Hospital, and whose clinicopathological data were complete. Exclusion criteria were as follows: Patients <18 years of age, non-epithelial ovarian cancer, no debulking or staging surgery, no adjuvant chemotherapy, incomplete clinicopathological data, and patients who refused to participate in the study. A total of 94 patients with epithelial ovarian cancer were included, whose age ranged from 28 to 83 years, with a median age of 51 years according to inclusion/exclusion criteria. All patients received ovarian tumor cytoreductive or comprehensive staging surgery and adjuvant chemotherapy. In addition, 56 patients received neoadjuvant chemotherapy. All cases were confirmed by pathological diagnosis, and complete clinicopathological parameters were obtained, including age, abdominal invasion/metastasis, histopathological type, histological grade, degree of differentiation, FIGO tumor stage (19), neoadjuvant chemotherapy regimen, cancer recurrence and survival after treatment. Progression-free survival (PFS) and overall survival (OS) were defined as the time from the day of initial treatment to the day of clinical recurrence by CA125 and/or imaging such as CT, MRI or PET/CT, and to the day of death, respectively. The follow-up data were obtained by telephone calls or outpatient visits. The tissue samples of epithelial ovarian cancer were collected from January 2013 to July 2017. The present study was approved by the Ethics Committee of Guangxi Medical University Cancer Hospital (approval no. LW2023118).

Immunohistochemistry (IHC) staining

Paraffin-embedded epithelial ovarian cancer tissues were obtained from the Department of Pathology of Guangxi Medical University Cancer Hospital. The tissues were cut into 4-µm paraffin sections and heated at 60°C for 20 min. The sections were then soaked in xylene and different concentrations of alcohol (95, 70 and 50%) for dewaxing followed by soaking in distilled water for 5 min for hydration. The hydrated tissue sections were then immersed in a container containing 0.01 M citrate buffer (pH 6.0) and heated to 92–98°C for 10–15 min for antigen retrieval. A SP immunohistochemical kit (cat. no. SP-9000; ZSGB-BIO) was used for the immunohistochemical detection of TM4SF1 and DDR1 expression in epithelial ovarian cancer tissue according to the manufacturer's protocol. The tissue sections were incubated with 3% hydrogen peroxide for 10 min to block endogenous peroxidase activity, followed by incubation with goat serum working solution for 15 min at room temperature to block the non-specific binding sites. Samples were sequentially incubated in rabbit anti-TM4SF1 (1:1,000; cat. no. ab113504; Abcam) or rabbit anti-DDR1 (1:3,000; cat. no. 5583; Cell Signaling Technology, Inc.) overnight at 4°C, biotin-labeled goat anti-mouse/rabbit IgG for 1 h at room temperature and streptavidin conjugated with horseradish peroxidase working solution for 15 min at room temperature. Finally, the tissue sections were stained with 3,3′-diaminobenzidine solution and counterstained with hematoxylin for 2 min. The immunostained tissue sections were independently observed by two experienced pathologists under a light microscope (Olympus Corporation). The staining was scored and determined to be positive or negative according to methods previously reported in the literature (20).

Gene Expression Profiling Interactive Analysis (GEPIA)

The interactive online analysis tool GEPIA (http://gepia.cancer-pku.cn/) was used to analyze the RNA sequencing expression data of 9,736 tumor tissues from The Cancer Genome Atlas and 8,587 normal tissues from the Genotype-Tissue Expression database (21). GEPIA was used to determine the expression levels of the TM4SF1 and DDR1 genes in different cancer types, and the differential expression of the TM4SF1 and DDR1 genes in ovarian cancer tissue and normal ovarian tissue was compared. Analysis of the difference in OS and disease-free survival (DFS) between low or high TM4SF1 and DDR1 expression in ovarian cancer tissue was then conducted. The hazard ratio (HR) and P-value or log-rank P-value were calculated and displayed in the graph.

Oncomine analysis

Oncomine is a large cancer microarray database and integrated data mining platform (https://www.oncomine.org), which includes 715 gene expression datasets and 86,733 cancer tissue and normal tissue samples (22,23). First, Oncomine was searched using ‘TM4SF1’ and ‘DDR1’ as keywords, and the results were displayed and set as: ‘P-value: 0.01’, ‘fold change: 1.5’ and ‘gene rank: all’ to analyze the differential expression of TM4SF1 and DDR1 in different types of tumors and the corresponding normal tissues. Then, ‘TM4SF1 and ‘DDR1’ were used as keywords, and ‘ovarian cancer’, ‘cancer vs. normal’ and ‘mRNA’ data types were used as filter conditions, and the meta-analysis performed by the Oncomine database was used to verify the differential expression of TM4SF1 and DDR1 genes in ovarian cancer tissue.

Protein-protein interaction (PPI) network

Search Tool for the Retrieval of Interacting Genes/Proteins (STRING; version 11.0) is an online search tool for the analysis of PPIs and functional protein networks, which contains confirmed and predicted direct and indirect PPI biological data (24). In the present study, TM4SF1 and DDR1 were used as query conditions, and the minimum required interaction score was set to >0.4, which is a medium level of confidence. The top 50 proteins interacting directly with TM4SF1 and DDR1 were queried according to the interaction score.

Functional and pathway enrichment analyses

Functional and pathway enrichment analyses were performed using R 3.63 software (https://www.r-project.org/). Using the R clusterProfiler package, TM4SF1, DDR1 and the top 50 filtered interacting proteins that were predicted to directly interact with TM4SF1 and DDR1 were used to perform Gene Ontology (GO) function and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses (25,26). The R enrichplot (https://bioconductor.org/packages/enrichplot/; version 1.20.0) and ggplot2 (https://CRAN.R-project.org/package=ggplot2; version 3.3.1) packages were used to visualize the enrichment analysis results. The GO functional enrichment analysis included the following three categories: Biological process, cell composition and molecular function (MF). A corrected P-value of <0.05 was considered statistically significant.

Kaplan-Meier (KM) plotter database analysis

KM plotter is an online database that assesses the relationship of 54,000 genes with survival rates in 21 cancer types (27). The relationships of TM4SF1 expression, DDR1 expression and combined TM4SF1 and DDR1 expression with the OS and PFS of ovarian cancer were analyzed. The parameters were set as: Split patients by, ‘upper tertile’; follow-up threshold, ‘120 months’; array quality control, ‘exclude biased arrays’; and other parameters were the system defaults.

Statistical analysis

Statistical analysis was performed using R 3.63 software and GraphPad Prism 8.0 software (GraphPad Software; Dotmatics). The χ2 test was used to compare the differences between different clinicopathological features. To examine the prognostic significance of TM4SF1 and DDR1 in epithelial ovarian cancer patients, Kaplan-Meier survival analysis (P-value by log-rank test) was used to explore the association between the OS/PFS and TM4SF1/DDR1 expression in epithelial ovarian cancer tissue. The Cox regression model was used for the prognostic analysis of patients with epithelial ovarian cancer. The R survival (https://CRAN.R-project.org/package=survival; version 3.1–12) and survminer (https://CRAN.R-project.org/package=survminer; version 0.4.7) packages were used for data analysis and visualization. P<0.05 was considered to indicate a statistically significant result.

Results

Expression of TM4SF1 and DDR1 in cancers

The biological functions and roles of TM4SF1 and DDR1 in cancer are unclear. To investigate their roles in tumorigenesis and development, differentially expressed genes were identified in various types of cancer. The results of analyses performed using GEPIA and Oncomine showed that TM4SF1 and DDR1 were highly expressed in a variety of common malignant tumor types (Fig. 1). The analysis showed that TM4SF1 and DDR1 mRNA expression levels were significantly higher in ovarian cancer than in normal ovarian tissue (both P<0.05). A meta-analysis performed using the Oncomine database, which included 10 analyses of 7 ovarian cancer datasets, also showed that TM4SF1 and DDR1 were significantly more highly expressed in ovarian cancer than in normal ovarian tissue (TM4SF1, P=0.004; DDR1: P=3.22×10−4) as shown in Fig. 2.

Figure 1. High expression of TM4SF1 and DDR1 in a variety of tumor types. Differential expression of the (A) TM4SF1 and (B) DDR1 genes in the Gene Expression Profile Interactive Analysis database. Differential expression of the (C) TM4SF1 and (D) DDR1 gene in the Oncomine database. TM4SF1, transmembrane 4 L6 family member 1; DDR1, discoidin domain receptor 1; CESC, cervical squamous cell carcinoma and endocervical adenocarcinoma; CHOL, cholangiocarcinoma; COAD, colon adenocarcinoma; ESCA, esophageal carcinoma; KIRC, kidney renal clear cell carcinoma; LUSC, lung squamous cell carcinoma; OV, ovarian serous cystadenocarcinoma; PAAD, pancreatic adenocarcinoma; PRAD, prostate adenocarcinoma; READ, rectum adenocarcinoma; STAD, stomach adenocarcinoma.

Figure 2. Oncomine analysis shows the high expression of TM4SF1 and DDR1 in ovarian cancer tissue. Expression of (A) TM4SF1 and (B) DDR1 in ovarian cancer compared with normal ovarian tissue. The mRNA expression of (C) TM4SF1 and (D) DDR1 in ovarian cancer and normal tissue was compared in 10 analyses. Red indicates upregulation. *P<0.05. TM4SF1, transmembrane 4 L6 family member 1; DDR1, discoidin domain receptor 1; TPM, transcripts per million.

Enrichment analysis of TM4SF1- and DDR1-interacting proteins

The top 50 proteins that directly interact with TM4SF1 and DDR1 were queried using the STRING online tool, and 28 and 25 proteins were found to interact with TM4SF1 and DDR1, respectively. Among these interacting proteins, both CYSTM1 and NT5M interacted with TM4SF1 and DDR1. In addition, TM4SF1 interacted with DDR1 (Fig. 3).

Figure 3. Protein-protein interaction network of TM4SF1, DDR1 and their interacting proteins. TM4SF1, transmembrane 4 L6 family member 1; DDR1, discoidin domain receptor 1.

GO enrichment analysis of the interacting proteins showed that they were mainly enriched in terms associated with the extracellular matrix (ECM), collagen, integrin and their complexes as well as adhesion sites, the endoplasmic reticulum and other intracellular and extracellular components. GO enrichment analysis also indicated that these interacting proteins had the MFs of binding with integrin, cell adhesion molecules, growth factors, proteoglycans, neurotrophic factor receptors and γ-catenin. The GO analysis also indicated that the interacting proteins were involved in cell-matrix adhesion and the response of cells to stimulation by amino acids, which suggests that their biological functions are mediated via the activation of integrin and collagen-mediated signaling pathways (Fig. 4A; Table SI). KEGG enrichment analysis showed that these interacting proteins were mainly enriched in the terms ‘focal adhesion’, ‘ECM-receptor interaction’, ‘regulation of actin cytoskeleton’, ‘microRNAs in cancer’, ‘proteoglycans in cancer’, ‘PI3K/Akt signaling pathway’, ‘neurotrophin signaling pathway’ and a variety of tumor-related pathways, such as ‘small cell lung cancer’, ‘chronic myeloid leukemia’, ‘thyroid cancer’ and ‘bladder cancer’ (Fig. 4B; Table SII).

Figure 4. GO and KEGG enrichment analyses of proteins interacting with TM4SF1 and DDR1. (A) GO and (B) KEGG pathways found to be enriched with TM4SF1- and DDR1-interacting proteins. GO, Gene Ontology; KEGG, Kyoto Encyclopaedia of Genes and Genomes; TM4SF1, transmembrane 4 L6 family member 1; DDR1, discoidin domain receptor 1; BP, biological process; CC, cellular constituent; MF, molecular function.

Expression of TM4SF1 and DDR1 proteins in epithelial ovarian cancer and their association with clinicopathologic features

The IHC results of the 94 patients with epithelial ovarian cancer showed that 46 patients (48.94%) were positive for TM4SF1 protein, and 52 patients (55.32%) were positive for DDR1 protein. In addition, 26 patients (27.66%) were positive for both TM4SF1 and DDR1 proteins. Representative staining images of TM4SF1 and DDR1 proteins in epithelial ovarian cancer tissues are shown in Fig. 5. Analysis using χ2 tests showed that only the positive expression of DDR1 protein differed significantly according to the histological grade, FIGO stage and presence or absence of intraperitoneal metastases (all P<0.05; Table I).

Figure 5. Staining of TM4SF1 and DDR1 protein in ovarian cancer tissue. (A) TM4SF1 protein is negative and DDR1 protein is positive. (B) TM4SF1 protein is positive, and DDR1 protein is negative. (C) TM4SF1 and DDR1 protein are both positive. TM4SF1, transmembrane 4 L6 family member 1; DDR1, discoidin domain receptor 1.

Table I. Expression of TM4SF1 and DDR1 in patients with ovarian cancer and different clinicopathologic features.

Table I.

Expression of TM4SF1 and DDR1 in patients with ovarian cancer and different clinicopathologic features.

Parameter	N	TM4SF1(+), n (%)	P-value	DDR1(+), n (%)	P-value	TM4SF1(+) + DDR1(+), n (%)	P-value
Age, years
≤50	46	27 (58.70)	0.064	26 (56.52)	0.818	16 (34.78)	0.131
>50	48	19 (39.58)		26 (50.00)		10 (20.83)
Histological grade
G1-2	20	11 (55.00)	0.541	6 (30.00)	0.010	5 (25.00)	0.764
G3	74	35 (47.30)		46 (62.16)		21 (28.66)
FIGO stage
I–II	12	5 (41.67)	0.590	2 (16.67)	0.004	2 (16.67)	0.571a
III–IV	82	41 (50.00)		50 (60.98)		24 (29.37)
Histopathological type
Serous carcinoma	82	37 (45.12)	0.053	47 (57.32)	0.308	22 (26.83)	0.901a
Non-serous carcinoma	12	9 (75.00)		5 (41.67)		4 (33.33)
Abdominal invasion
Yes	63	33 (52.38)	0.341	42 (66.67)	0.002	20 (31.75)	0.207
No	31	13 (41.94)		10 (32.26)		6 (19.35)
Neoadjuvant chemotherapy
Yes	56	25 (44.62)	0.312	32 (57.14)	0.666	14 (25.00)	0.484
No	38	21 (55.26)		20 (52.63)		12 (31.58)
Lymphatic metastasisb
Yes	25	13 (52.00)	0.453	15 (60.00)	0.113	7 (28.00)	0.282
No	31	13 (52.00)		12 (38.70)		5 (16.13)

a Yates's correction for the χ² test applied due to low expected counts.

b Comparisons were made in the 56 patients with resected lymph nodes. TM4SF1, transmembrane 4 L6 family member 1; DDR1, discoidin domain receptor 1.

Relationship of TM4SF1 and DDR1 expression with the prognosis of ovarian cancer

Bioinformatics database analysis showed that patients with ovarian cancer and high expression of TM4SF1 had significantly lower DFS or PFS than those with low TM4SF1 expression (DFS: HR=1.3, P=0.047, n=424, Fig. 6A; PFS: HR=1.17, P=0.019, n=1,435, Fig. 6E). However, the expression of DDR1 was not found have a significant association with patient DFS or PFS, and neither TM4SF1 nor DDR1 expression had a significant association with patient OS (Fig. 6B-D and F-H). However, multigene analysis performed using KM plotter showed that higher expression of TM4SF1 and DDR1 coexpression was significantly associated with a shorter PFS in patients with ovarian cancer (HR=1.15, P=0.039, n=1,435, Fig. 6I) but not with OS (Fig. 6J).

Figure 6. Relationship of TM4SF1 and DDR1 expression with OS, DFS and PFS in patients with ovarian cancer determined using GEPIA and KM plotter analyses. GEPIA results of (A) DFS and (B) OS according to TM4SF1 expression and (C) DFS and (D) OS according to DDR1 expression. KM plotter results of (E) PFS and (F) OS according to TM4SF1 expression; (G) PFS and (H) OS according to DDR1 expression; and (I) PFS and (J) OS according to coexpression of TM4SF1 and DDR1. TM4SF1, transmembrane 4 L6 family member 1; DDR1, discoidin domain receptor 1; OS, overall survival; DFS, disease-free survival; PFS, progression-free survival; GEPIA, Gene Expression Profile Interactive Analysis; KM, Kaplan-Meier.

Analysis of the primary clinical data showed that the 94 patients with epithelial ovarian cancer had a median follow-up of 33 months, and detected no significant difference in the median OS between TM4SF1-positive and TM4SF1-negative patients (29 vs. 47 months). By contrast, the median OS of DDR1-positive patients was significantly shorter than that of DDR1-negative patients (31 vs. >73 months, P<0.05). The median OS of patients with TM4SF1 and DDR1 coexpression was significantly shorter than that of patients lacking TM4SF1 and DDR1 coexpression (21 vs. 49 months, P<0.05). In addition, the median PFS of TM4SF1-positive patients was significantly shorter than that of TM4SF1-negative patients (18 vs. 26 months, P<0.05). The median PFS of patients with TM4SF1 and DDR1 coexpression was significantly shorter than that of patients lacking TM4SF1 and DDR1 coexpression (12 vs. 26 months, P<0.05), while the expression of DDR1 alone was not associated with the median PFS (Fig. 7). Thus, the TM4SF1 and DDR1 coexpression indicated that patients with epithelial ovarian cancer had shorter PFS and OS.

Figure 7. Association of TM4SF1 and DDR1 protein expression with the prognosis of patients with ovarian cancer. Overall survival curves of patients with or without (A) TM4SF1, (B) DDR1 and (C) TM4SF1 + DDR1 protein expression. Progression-free survival curves of patients with or without (D) TM4SF1, (E) DDR1 and (F) TM4SF1 + DDR1 protein expression. TM4SF1, transmembrane 4 L6 family member 1; DDR1, discoidin domain receptor 1.

Analysis of prognostic factors in epithelial ovarian cancer

Regarding the clinicopathological factors of the patients, univariate analysis showed that FIGO stage, abdominal invasion, DDR1 expression and coexpression of DDR1 and TM4SF1 were factors affecting the OS of patients with epithelial ovarian cancer, while FIGO stage, TM4SF1 expression and coexpression of TM4SF1 and DDR1 were factors affecting the PFS of epithelial ovarian cancer patients (Table II).

Table II. Univariate analysis of the association of clinicopathological factors with OS and PFS in ovarian cancer.

Table II.

Univariate analysis of the association of clinicopathological factors with OS and PFS in ovarian cancer.

	OS			PFS
Parameter	HR	95% CI	P-value	HR	95% CI	P-value
Age	0.877	0.506–1.520	0.641	0.866	0.541–1.385	0.548
Histological grade	1.146	0.573–2.292	0.700	1.387	0.742–2.592	0.305
FIGO stage	6.860	1.650–28.520	0.008	2.999	1.287–6.990	0.011
Histopathological type	0.877	0.374–2.057	0.763	1.574	0.680–3.640	0.289
Abdominal invasion	2.253	1.193–4.257	0.012	1.631	0.981–2.711	0.059
Neoadjuvant chemotherapy	1.750	0.953–3.214	0.071	1.565	0.946–2.589	0.082
Lymph node excision	0.617	0.355–1.072	0.087	0.628	0.390–1.010	0.055
TM4SF1	1.454	0.838–2.525	0.183	1.646	1.026–2.642	0.039
DDR1	2.324	1.271–4.249	0.006	1.555	0.962–2.513	0.072
TM4SF1 + DDR1	2.559	1.450–4.517	0.001	2.572	1.566–4.224	<0.001

[i] OS, overall survival; PFS, progression-free survival; HR, hazard ratio; 95% CI, 95% confidence interval; TM4SF1, transmembrane 4 L6 family member 1; DDR1, discoidin domain receptor 1.

Cox multivariate analysis showed that TM4SF1 and DDR1 coexpression was the only independent risk factor affecting OS and PFS in the patients with epithelial ovarian cancer (Fig. 8), suggesting that the expression of TM4SF1 and DDR1 may be synergistically involved in the development of epithelial ovarian cancer.

Figure 8. Forest plot of the association of TM4SF1 and DDR1 coexpression with prognosis in ovarian cancer. Forest maps of (A) overall survival and (B) progression-free survival in patients with ovarian cancer according to TM4SF1 and DDR1 coexpression. TM4SF1, transmembrane 4 L6 family member 1; DDR1, discoidin domain receptor 1; NACT, neoadjuvant chemotherapy; AIC, Akaike information criterion.

Discussion

Ovarian cancer is a highly malignant gynecological tumor. Despite the development of treatment methods, the prognosis of patients is far from ideal. The pathogenesis of ovarian cancer is not fully understood, and few effective therapeutic targets are available. TM4SF1 and DDR1 proteins have been reported to be expressed in a variety of cancer tissues and are associated with poor prognosis in patients (8–11,17), and database analysis has yielded similar results. However, the role of TM4SF1 and DDR1 in the occurrence, development and prognosis of epithelial ovarian cancer remains unclear.

Antitumor immunity induced by TM4SF1 has been shown to inhibit the growth and migration of TM4SF1-positive cancer cells in vitro and in vivo (28). In addition, a previous study found that the positive rate of TM4SF1 protein expression in epithelial ovarian cancer tissues, particularly metastatic lymph nodes, was significantly higher than that in benign ovarian tumors and normal ovarian tissues, and the expression status of TM4SF1 in epithelial ovarian cancer was significantly associated with FIGO staging and tissue differentiation (29). Similarly, the present study also found that the positive rate of TM4SF1 protein in epithelial ovarian cancer tissues was higher than that in normal ovarian tissue, although the rate was slightly lower than that previously reported. This difference may be due to certain cases having received neoadjuvant chemotherapy and the prolonged storage of some specimens as well as different experimental conditions, sample sizes and operators. With regard to the effect of TM4SF1 on the prognosis of patients, the PFS of TM4SF1-positive patients was significantly shorter than that of TM4SF1-negative patients but TM4SF1 expression status did not affect the OS of patients. The present study found that the positive expression of TM4SF1 was not an independent risk factor for PFS and OS in epithelial ovarian cancer. In view of the difference in the positive rate of TM4SF1 in the present study compared to previous studies, the effect of neoadjuvant chemotherapy on the tissue expression of TM4SF1 requires further research.

The effects of DDR1 on the biological functions, disease progression and prognosis of different types of tumor cells vary, but there is evidence to show that DDR1 is involved in prometastatic and prosurvival signals (30). Although DDR1 was not found to be an independent risk factor affecting the prognosis of epithelial ovarian cancer in the present study, univariate analysis showed that it was significantly associated with OS. The study found that the positive rate of DDR1 in epithelial ovarian cancer tissues was 55.32%, while the positive expression rate in patients with advanced stage (stages III–IV), low differentiation (or high histological grade) and abdominal metastasis was significantly higher than that in patients with early stage (stages I–II), well-differentiated tissues and no abdominal metastasis. Advanced stage and high-grade cancer are poor prognostic factors for epithelial ovarian cancer. DDR1 was highly expressed in the sample types with these factors in the present study, suggesting that DDR1 upregulation is positively associated with the severity of the disease and that high pathological grade affects the OS of patients. These findings are consistent with previous results reported by Quan et al (31), implying that DDR1 might play a role in promoting the survival and metastasis of epithelial ovarian cancer cells.

The immunohistochemical staining of epithelial ovarian cancer tissue sections showed that TM4SF1 and DDR1 proteins were mainly distributed in the cell membrane and that both proteins were expressed in some cases. STRING analysis indicated that there is an interaction between TM4SF1 and DDR. Moreover, the coexpression of M4SF1 and DDR1 was found to be significantly associated with PFS and OS in patients with epithelial ovarian cancer and also serve as an independent poor prognostic factor. Therefore, TM4SF1 and DDR1 may play important biological functions through interaction. Gao et al (18) found that TM4SF1 interacts with DDR1 in breast cancer cells and that collagen I promotes their interaction and the reactivation of dormant metastatic breast cancer cells in multiple organs, leading to breast cancer recurrence and multiple organ metastasis. Due to the limitations of immunostaining for performing colocalization analysis, whether TM4SF1 and DDR1 proteins are colocalized and interact with each other in epithelial ovarian cancer cells and whether a mutual regulatory mechanism between TM4SF1 and DDR exists requires further study.

The present study evaluated the relationship between TM4SF1 and DDR1 expression alone and in combination with the prognosis of patients with epithelial ovarian cancer. The OS and PFS results in the primary dataset are consistent with the results of online databases in general. However, it is worth noting some differences in the OS of DDR1 alone or the combined analysis of TM4SF1 and DDR1 in different datasets. Regarding the prognostic factors for OS and DFS, the results based on GEPIA and KM analysis were not as strong as some of those obtained from the primary patient dataset. We hypothesize that the possible reasons for these differences are as follows: Firstly, there may have been some bias in the statistical analysis, such as admission bias and information bias. In the primary patient dataset, the prognostic data of epithelial ovarian cancer were obtained from a single center and hospital, with a high proportion of advanced cases, a relatively high positive rate of TM4SF1 and DDR1, and a worse prognosis for patients. By contrast, the results based on GEPIA and KM data are from a wide range of sources, and the classification or inclusion criteria might not be completely consistent for different data. In addition, as aforementioned, some cases that had received neoadjuvant chemotherapy and some old specimens were included in the 94 cases of epithelial ovarian cancer. The lack of cases and inadequate follow-up time are likely to have affected the accuracy of the results. Furthermore, differences in experimental conditions may unavoidably contribute to diagnostic bias or detection bias.

The present study has mainly focused on clinical studies. In subsequent studies, an increase in the number of cases and extension of the follow-up time is necessary. In addition, the relevant regulatory mechanisms of TM4SF1 and DDR1 in epithelial ovarian cancer cells requires further research in vitro and in vivo.

In conclusion, TM4SF1 and DDR1 proteins were found to be coexpressed in some epithelial ovarian cancer tissues, and may be adverse prognostic factors for patients with epithelial ovarian cancer. Also, there may be an interaction or mutual regulatory mechanism between TM4SF1 and DDR1.

Supplementary Material

Supporting Data

Acknowledgements

Not applicable.

Funding

The present study was supported by grants from the National Natural Science Foundation of China (grant no. 82060476), the Guangxi Natural Science Foundation (grant no. 2017GXNSFDA198009), the Guangxi Zhuang Autonomous Region Key Clinical Specialty Construction Project (grant no. 2018-39), the Guangxi Medical High-level Backbone Personnel Training ‘139’ Project (grant no. 2018-22) and the Special Fund of the 17th Guangxi New Century ‘Ten, Hundred, Thousand’ Talent Project (grant no. 2014210).

Availability of data and materials

The datasets generated and/or analyzed during the current study are not publicly available due to their containing information that could compromise the privacy of research participants but are available from the corresponding author on reasonable request.

Authors' contributions

ZH and HY were responsible for study conception and carried out the literature research, clinical study, data analysis, experimental operations, manuscript writing and editing. ZY was responsible for study design, experimental guidance and manuscript modification. ZH, HY and ZY confirm the authenticity of all the raw data. All authors have read and approved the final version of the manuscript.

Ethics approval and consent to participate

All procedures were approved by the Ethics Committee of Guangxi Medical University Cancer Hospital and were performed in keeping with the standards set out in the Declaration of Helsinki and Laboratory Guidelines of Research in China. All patients provided written informed consent.

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Glossary

Abbreviations

Abbreviations:

References