High expression of Fibroblast activation protein (FAP) predicts poor outcome in high-grade serous ovarian cancer

doi:10.21203/rs.3.rs-28242/v2

Download PDF

Research article

High expression of Fibroblast activation protein (FAP) predicts poor outcome in high-grade serous ovarian cancer

https://doi.org/10.21203/rs.3.rs-28242/v2

This work is licensed under a CC BY 4.0 License

Journal Publication

published 27 Oct, 2020

Read the published version in BMC Cancer →

You are reading this older preprint version

Read the latest preprint version →

Background: High-grade serous ovarian cancer (HSOC) is a fatal form of ovarian cancer. Previous studies indicated some potential biomarkers for clinical evaluation of HSOC prognosis. However, there is a lack of systematic analysis of different expression genes (DEGs) to screen and detect significant biomarkers of HSOC.

Method and materials: TCGA database was conducted to analyze relevant genes expression in HSOC. Outcomes of candidate genes expression, including overall survival (OS) and progression-free survival (PFS), were calculated by Cox regression analysis for hazard rates (HR). Histopathological investigation of the identified genes was carried out in 151 Chinese HSOC patients to validate gene expression in different stages of HSOC.

Results: Of all 57,331 genes that were analyzed, FAP was identified as the only novel gene that significantly contributed to both OS and PFS of HSOC. In addition, FAP had a consistent expression profile between carcinoma-paracarcinoma and early-advanced stages of HSOC. Immunological tests in paraffin section also confirmed that up-regulation of FAP was present in advanced stage HSOC patients. Prediction of FAP network association suggested that FN1 could be a potential downstream gene which further influenced HSOC survival.

Conclusions: High-level expression of FAP was associated with poor prognosis of HSOC via FN1 pathway.

Cancer Biology

Oncology

Fibroblast activation protein (FAP)

The Cancer Genome Atlas Program (TCGA)

high-grade serous ovarian cancer (HSOC)

survival

Ovarian cancer is one of the major causes of death in females globally. According to 2018 global cancer statistics, 295,414 new cases and 184,799 deaths were reported (1). In gynecological oncology, ovarian cancer is less prevalent than breast cancer and cervix cancer, however the death rate of OC is the highest (1). The most recent 2020 cancer statistics in United States also confirmed that ovarian cancer is the fifth cause of deaths of females (13,940 patients, 5% of total cancer-related death), only trailing by lung & bronchus cancer, breast cancer, colon & rectum cancer, and pancreas cancer (2). According to the NIH Surveillance, Epidemiology, and End Results program (SEER) survival statistics (2009-2015), 5-year survivorship of ovarian cancer is only 47.6% (3), which remained virtually unchanged since the last decade (4).

Based on the immunohistological variation, serous ovarian cancer is the most common subtype of ovarian cancer, which could be further categorized as high-grade and low-grade neoplasm according to tumor Federation International of Gynecology and Obstetrics (FIGO) grade (5, 6). High-grade serous ovarian cancer (HSOC) is the most common, aggressive, and fatal type of ovarian cancer. Almost 30% of patients died within five years of diagnosis (7), mainly because of lack of disease-specific symptoms, prominent biomarkers, and effective therapy or targeted drugs (8-10). Despite sharing some similar histological characters and terminology, high- and low- grade SOCs are now acknowledged as two different neoplasms (11). In 2011, the Cancer Genome Atlas (TCGA) program published the genomic and transcriptomic data of ovarian serous carcinoma, which summarized specific features of HSOC such as TP53 mutation, extensive DNA copy variation, BRCA1/BRCA2 inactive mutation, CCNE1 aberrations, and other survival-related preliminary transcriptional signatures (12-15).

Fibroblast activation protein (FAP), a cell-surface serine protease, emerges as an imperative factor in cancer-associated fibroblasts (CAFs), especially relevant to tumor occurrence and progression. Structurally, FAP consists of a cytoplasmic tail, a single transmembrane domain, and an extracellular domain (16). FAP is rarely expressed in healthy adult tissues. However, FAP is usually highly upregulated during tissue remodeling events, including cancers or cancer-associated fibroblasts (CAFs) (17-20). In addition, FAP is considered as a potential biomarker in certain tumor diagnosis and progression due to its protumorigenic specificity in both enzymatic and non-enzymatic manners (21-24).

In this study, we aim to identify potential biomarkers of HSOC survival from TCGA ovarian cancer cohort bioinformatics data. We analyzed gene expression, clinical and/or demographic information, and targeted strategies for potential HSOC biomarkers. In addition, we validated our findings by immunohistological investigation of tissues in a group of Chinese HSOC patients. Results suggest that FAP expression could be an effective biomarker for HSOC survival, which warrants further investigation as potential intervention of HSOC.

Dynamic protein analysis of TCGA database

Gene expression data (379 cases, Workflow Type: HTSeq-Counts) and clinical information were downloaded by the TCGAbiolink package in R 3.6.0 (https://www.R-project.org/) from the official TCGA website. After screening the clinical database, 320 patients were included for G3 histologic grade indicating HSOC. Then, 2 of these patients were excluded because of lack of relevant information. Finally, a total of 318 participants were recruited in our study.

For gene expression analysis, we downloaded the entire 57,331-gene data in serous cystadenocarcinoma from the TCGA RNAseq database. After data cleaning, 19 of 57,331 genes were excluded because of their insufficient expression in total gene expression (expressions of these excluded genes were < 1 copy in all participants). Then RNA expression data were transformed to z-scores in survival analysis.

Participants recruitment

This study followed the Declaration of Helsinki and was approved by the Institution of Research Ethics Committee of the First Hospital affiliated to Soochow University. All participants were fully aware of all protocols of this study, and signed up written consent forms to authorize the utilization of their tissues and relevant information.

For the validation set of the findings in HSOC TCGA database, 151 Chinese Han patients were diagnosed and recruited by the First Hospital affiliated to Soochow University, the Nanjing Maternity and Child Health Care Hospital, and the First Hospital affiliated to the Nanjing Medical University, from January 2013 to May 2019. After surgery, patients’ tumor tissues were prepared into the paraffin sections. Their demographic and clinical characteristics were collected as well.

Immunohistochemistry staining (IHC) and IHC score

IHC was performed on paraffin sections of ovarian cancer tissues to characterize target gene (FAP) expression profile. Detailed steps can be found in our previous study (25). FAP antibody (#66562) was purchased from CST company to incubate the preparing sections overnight at 4°C for further staining.

IHC of FAP was evaluated by two independent and qualified pathologists who were blinded to actual clinical outcomes. IHC scoring was then established as follows. Percentage of positive cells and intensity of staining of FAP antibody were first calculated and then divided into these three major categories: ≤3, negative or weak; >3 and ≤6, Moderate; >6, strong.

Association Network prediction of targeting gene

Prediction of functional protein association networks of candidate genes was performed by STRING version 11.0 (https://string-db.org/) and Genecard version 4.13 (www.genecards.org). In order to increase prediction accuracy, only common predictions present in both websites were included.

GO and KEGG pathway analysis of host gene

The gene ontology (GO) functional annotation and KEGG pathway analysis of host genes of polymorphisms were carried out by using the package ‘clusterProfiler’ in R (version 4.0.1).

Statistical analysis

All statistical analyses were calculated in R 3.6.0. Overall survival (OS) and progression-free survival (PFS) analyses in SOC patients were conducted by Cox regression and the Kaplan-Meier method. Multivariate Cox analysis was applied to identify potential influence of FAP expression on OS and PFS at different clinical stages of HSOC. Other relevant demographic and clinical characteristics were compared by Student’s t test or Wilcoxon test whenever appropriate between the two groups. P value < 0.05 was considered as statistical significance in this study.

Workflow of gene identification in TCGA database

As shown in figure 1, there were 304 and 237 genes with different expressions in OS and FPS analysis, respectively. By comparing early stage (Stage I+II) with advanced stage (Stage III+IV) HSOC, we identified 544 stage-related aberrantly expressed genes in HSOC patients. After cross-checking OS, FPS and stage-related genes, only FAP and SSC5D were still present. And, FAP was finally included because of its characteristics of typical enzyme-catalyzed activity in uniprot database (www.uniprot.org/) and its potential role in HSOC patient’s survival (Figure 1).

Associations of FAP expression with overall survival (OS) and progression-free survival (PFS)

Cox regression analysis based on FAP expression was performed to generate survival curves in OS and PFS. As shown in figure 2A, low FAP expression group showed a significant protective effect on HSOC prognosis in OS (P = 0.005). Longitudinally, low FAP expression group had 91.1% survival rate in a period of 12 months, compared with 84.4% in high FAP expression group. Survival rate in 50 months decreased to 31.9% in low FAP group and 21.4% in high group, respectively (Table 1). Results of PFS also showed similar patterns to OS between high and low FAP expression groups (P = 0.008, figure 2B).

Expression of FAP in HSOC patients

Sections of 151 HSOC patients’ tumor tissues in different stages were stained with FAP-antibody. We identified an increasing trend of FAP expression with respect to severity of cancer stages (figure 3). In addition, strong positive FAP-staining generally showed in membranous and cytosolic compartments in HSOC tissues. Figure 3b revealed the total difference scores between early stage HSOC patients and advanced patients (P = 0.016). Additionally, negative FAP was observed in 33 (21.85%) patients.

Prediction of network influenced by FAP

Based on STRING and Genecard, prediction of co-influence genes with FAP and their potential regulating effects was shown in figure 4. Specifically, fibronetin-1 (FN1), collagen family genes-COL1A1, COL1A2, COL3A1, COL5A2, Thy-1 cell surface antigen (THY1), and insulin (INS) were identified as co-influence genes.

FN1 was the only significant gene based on Cox regression with its hazardous influence on HSOC (P = 0.018, as shown in table 1). FN1 is demonstrated to over-express in ovarian cancer, which could eventually influence the formation of multicellular aggregate of ovarian cancer cells, migration and invasion of cancer cells, and aggravating platinum-resistance to deactivate chemotherapy. Additionally, epithelial-mesenchymal transition of ovarian cancer is shown to relate to aberrant expression of FN1. Therefore, FN1 and its regulatory factors, including genes, non-coding RNAs and epigenetic regulations, might be valuable candidates for ovarian cancer studies.

Considering the similar function of collagen (COL) family, we extended the search of collagen encoding genes. Additional COL genes that could influence HSOC included COL16A1 (HR = 2.50, P = 0.001 for Cox regression), COL5A1 (HR = 2.43, P = 0.002), COL8A1 (HR = 2.16, P = 0.006), and COL4A1 (HR = 1.83, P = 0.035) (Table 1).

Bioinformation of FAP in GO and KEGG analysis

For gene ontology (GO) analysis of FAP, the top three enriched GO annotations were listed as “regulation of fibrinolysis”, “negative regulation of extracellular matrix organization”, and positive regulation of execution phase of apoptosis” three in Biological Process, while as “dipeptidyl-peptidase activity”, “aminopeptidase activity”, and “metalloendopeptidase activity” in Molecular Function (figure 5). On the other hand, however, for KEGG analysis, there is no related available record for FAP in corresponding database.

In this study we extensively searched the TCGA database for HSOC patients to identify significant biomarkers for HSOC survival. We constructed a comprehensive summary of differentially expressed genes (DEGs) of HSOC, especially with regard to patients’ survival. This is the first study, to our knowledge, to evaluate the effects of gene expression on HSOC survival through data mining.

FAP is a typical plasma membrane-bound serine protease, which is implicated in matrix digestion and invasion (26, 27). Overexpression of FAP has been investigated and believed to be associated with prognosis in many diseases, especially in cancers (20, 22, 28). Some studies have demonstrated that high FAP expression is a negative prognostic factor for epithelial ovarian cancer (29). Increased expression of FAP is believed to cause recurrence of epithelial ovarian cancer after chemotherapy (29, 30). Considering that HSOC is a common histological type of epithelial ovarian cancer, our findings in TCGA database on HSOC support this finding from a new perspective.

Based on the location of FAP in HSOC tissue, we suggest that FAP could be a strong positive cell-surface receptor. FAP has high expression in HSOC patients in both Chinese population (over 60% in our study) and other ethnics (over 50%) (30). In addition, performance of therapeutic targeting at FAP also suggests its effectiveness for cancers (31). Therefore, FAP could be a potential biomarker for drug delivery or even direct therapy for HSOC.

Cancer-associated fibroblasts (CAFs), including FAP, usually participate in extracellular matrix structure remodeling and tumor microarray reconstruction. In this study, we also retrospectively explored the TCGA database and investigated the effects of other CAF members on HSOC prognosis, such as ACTA2, PDGFRα/β, S100A4, and αSMA. These CAF members are widely adopted as biomarkers of HSOC. Nevertheless, none of these genes showed significant association with HSOC survival. This result may be partly due to negative expression of αSMA (30).

We identified FN1 as the only potential gene regulated by FAP in HSOC. Currently, there is no direct evidence to establish the causal relationship between FAP and FN1. Nevertheless, in this study, we demonstrated similar eventual clinical outcomes induced by FAP and FN1, positive correlation between their expression, significance of their role in overall survival of HSOC, and direct regulating relationship predicted by STRING and Genecard. All these novel findings suggested FAP as a novel trigger for FN1, at least for HSOC survival. For THY1 gene, though it was reported as a putative tumor suppressor of ovarian cancer, our study did not produce the same results as previous ones (32), despite THY1’s presence in our prediction of potential FAP association networks. It is possible that THY1 cooperates with FAP during HSOC occurrence but not in prognostic period.

After extensive data mining of TCGA database, we identified FAP as a significant biomarker for HSOC survival. FAP overexpression led to worse outcome of HSOC patients, especially in the advanced clinical stage. FN1 expression is potentially down regulated by FAP and further influences HSOC survival.

Conflict of Interest

The authors have no conflicts of interest to declare.

Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA: a cancer journal for clinicians. 2018: 68(6): 394.
Siegel RL, Miller KD, Jemal A. Cancer statistics, 2020. CA: a cancer journal for clinicians. 2020: 70(1): 7.
Surveilllance, Epidemiology, and End Result Program. Cancer Stat Facts: Ovarian Cancer. https://seer.cancer.gov/statfacts/html/ovary.html.
Vaughan S, Coward JI, Bast RC, Jr., Berchuck A, Berek JS, Brenton JD, et al. Rethinking ovarian cancer: recommendations for improving outcomes. Nature reviews Cancer. 2011: 11(10): 719.
Cho KR, Shih Ie M. Ovarian cancer. Annual review of pathology. 2009: 4: 287.
Silverberg SG. Histopathologic grading of ovarian carcinoma: a review and proposal. International journal of gynecological pathology : official journal of the International Society of Gynecological Pathologists. 2000: 19(1): 7.
Montavon C, Stricker GR, Schoetzau A, Heinzelmann-Schwarz V, Jacob F, Fedier A. Outcome in serous ovarian cancer is not associated with LATS expression. Journal of cancer research and clinical oncology. 2019: 145(11): 2737.
Cole AL, Austin AE, Hickson RP, Dixon MS, Barber EL. Review of methodological challenges in comparing the effectiveness of neoadjuvant chemotherapy versus primary debulking surgery for advanced ovarian cancer in the United States. Cancer epidemiology. 2018: 55: 8.
Ozols RF. Challenges for chemotherapy in ovarian cancer. Annals of oncology : official journal of the European Society for Medical Oncology / ESMO. 2006: 17 Suppl 5: v181.
Bowtell DD, Bohm S, Ahmed AA, Aspuria PJ, Bast RC, Jr., Beral V, et al. Rethinking ovarian cancer II: reducing mortality from high-grade serous ovarian cancer. Nature reviews Cancer. 2015: 15(11): 668.
Robert J. Kurman, Marta-Luisa Carcangiu, Simon Herrington, Young. RH. WHO Classification of Tumours of Female Reproductive Organs. Fourth Edition. IARC publication. 2014.
Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell. 2011: 144(5): 646.
Zhang H, Liu T, Zhang Z, Payne SH, Zhang B, McDermott JE, et al. Integrated Proteogenomic Characterization of Human High-Grade Serous Ovarian Cancer. Cell. 2016: 166(3): 755.
Kaldawy A, Segev Y, Lavie O, Auslender R, Sopik V, Narod SA. Low-grade serous ovarian cancer: A review. Gynecologic oncology. 2016: 143(2): 433.
Cancer Genome Atlas Research N. Comprehensive genomic characterization of squamous cell lung cancers. Nature. 2012: 489(7417): 519.
Aertgeerts K, Levin I, Shi L, Snell GP, Jennings A, Prasad GS, et al. Structural and kinetic analysis of the substrate specificity of human fibroblast activation protein alpha. The Journal of biological chemistry. 2005: 280(20): 19441.
Niedermeyer J, Garin-Chesa P, Kriz M, Hilberg F, Mueller E, Bamberger U, et al. Expression of the fibroblast activation protein during mouse embryo development. The International journal of developmental biology. 2001: 45(2): 445.
Mankaney G, Leone P, Cruise M, LaGuardia L, O'Malley M, Bhatt A, et al. Gastric cancer in FAP: a concerning rise in incidence. Familial cancer. 2017: 16(3): 371.
Yang X, Lin Y, Shi Y, Li B, Liu W, Yin W, et al. FAP Promotes Immunosuppression by Cancer-Associated Fibroblasts in the Tumor Microenvironment via STAT3-CCL2 Signaling. Cancer research. 2016: 76(14): 4124.
Chen L, Qiu X, Wang X, He J. FAP positive fibroblasts induce immune checkpoint blockade resistance in colorectal cancer via promoting immunosuppression. Biochemical and biophysical research communications. 2017: 487(1): 8.
Hamson EJ, Keane FM, Tholen S, Schilling O, Gorrell MD. Understanding fibroblast activation protein (FAP): substrates, activities, expression and targeting for cancer therapy. Proteomics Clinical applications. 2014: 8(5-6): 454.
Dong R, Guo J, Zhang Z, Zhou Y, Hua Y. Polyphyllin I inhibits gastric cancer cell proliferation by downregulating the expression of fibroblast activation protein alpha (FAP) and hepatocyte growth factor (HGF) in cancer-associated fibroblasts. Biochemical and biophysical research communications. 2018: 497(4): 1129.
Ruan P, Tao Z, Tan A. Low expression of miR-30a-5p induced the proliferation and invasion of oral cancer via promoting the expression of FAP. Bioscience reports. 2018: 38(1).
Kilvaer TK, Rakaee M, Hellevik T, Ostman A, Strell C, Bremnes RM, et al. Tissue analyses reveal a potential immune-adjuvant function of FAP-1 positive fibroblasts in non-small cell lung cancer. PloS one. 2018: 13(2): e0192157.
Gao Y, Tang X, Cao J, Rong R, Yu Z, Liu Y, et al. The Effect of HER2 Single Nucleotide Polymorphisms on Cervical Cancer Susceptibility and Survival in a Chinese Population. Journal of Cancer. 2019: 10(2): 378.
O'Brien P, O'Connor BF. Seprase: an overview of an important matrix serine protease. Biochimica et biophysica acta. 2008: 1784(9): 1130.
Yu DM, Yao TW, Chowdhury S, Nadvi NA, Osborne B, Church WB, et al. The dipeptidyl peptidase IV family in cancer and cell biology. The FEBS journal. 2010: 277(5): 1126.
Yu LN, Liu Z, Tian Y, Zhao PP, Hua X. FAP-a and GOLPH3 Are Hallmarks of DCIS Progression to Invasive Breast Cancer. Frontiers in oncology. 2019: 9: 1424.
Mhawech-Fauceglia P, Yan L, Sharifian M, Ren X, Liu S, Kim G, et al. Stromal Expression of Fibroblast Activation Protein Alpha (FAP) Predicts Platinum Resistance and Shorter Recurrence in patients with Epithelial Ovarian Cancer. Cancer microenvironment : official journal of the International Cancer Microenvironment Society. 2015: 8(1): 23.
Mhawech-Fauceglia P, Wang D, Samrao D, Kim G, Lawrenson K, Meneses T, et al. Clinical Implications of Marker Expression of Carcinoma-Associated Fibroblasts (CAFs) in Patients with Epithelial Ovarian Carcinoma After Treatment with Neoadjuvant Chemotherapy. Cancer microenvironment : official journal of the International Cancer Microenvironment Society. 2014: 7(1-2): 33.
Pure E, Blomberg R. Pro-tumorigenic roles of fibroblast activation protein in cancer: back to the basics. Oncogene. 2018: 37(32): 4343.
Abeysinghe HR, Cao Q, Xu J, Pollock S, Veyberman Y, Guckert NL, et al. THY1 expression is associated with tumor suppression of human ovarian cancer. Cancer genetics and cytogenetics. 2003: 143(2): 125.

Table 1

Significant difference of Predicted gene influenced by FAP in HSOC overall survival COX analysis
Ensembl ID	Gene	β-coefficient	Hazard Rate (HR)	Standard error of coefficient	Z-value	P-value
ENSG00000078098	FAP	0.740663	2.097325	0.271298	2.730071964	0.006332
ENSG00000115414	FN1	0.663511	1.941597	0.28008	2.369006093	0.017836
ENSG00000084636	COL16A1	0.916471	2.500451	0.279922	3.274028048	0.00106
ENSG00000130635	COL5A1	0.887919	2.430066	0.289441	3.067700678	0.002157
ENSG00000144810	COL8A1	0.768551	2.15664	0.279769	2.747088883	0.006013
ENSG00000187498	COL4A1	0.608281	1.837271	0.289627	2.100225766	0.035709

Download PDF

Journal Publication

published 27 Oct, 2020

Read the published version in BMC Cancer →

Review #2 received at journal
13 Sep, 2020
Editorial decision: Minor revision
13 Sep, 2020
Review #1 received at journal
09 Sep, 2020
Reviewer #2 agreed at journal
19 Aug, 2020
Reviewer #1 agreed at journal
17 Aug, 2020
Reviewers invited by journal
14 Aug, 2020
Editor assigned by journal
20 Jul, 2020
Editor invited by journal
19 Jul, 2020
Submission checks completed at journal
30 Jun, 2020

You are reading this older preprint version

Read the latest preprint version →

High expression of Fibroblast activation protein (FAP) predicts poor outcome in high-grade serous ovarian cancer

Status:

Journal Publication

Version 2

Abstract

Figures

Introduction

Methods And Materials

Results

Discussion

Conclusion

Declarations

References

Table

Status:

Journal Publication

Version 2