Skip to main content
NCBI home page
Journal of Cell Communication and Signaling logoLink to Journal of Cell Communication and Signaling
. 2020 Sep 4;15(1):57–70. doi: 10.1007/s12079-020-00582-3

GAB2 and GAB3 are expressed in a tumor stage-, grade- and histotype-dependent manner and are associated with shorter progression-free survival in ovarian cancer

Caglar Berkel 1,✉, Ercan Cacan 1
PMCID: PMC7904992  PMID: 32888136

Abstract

Ovarian cancer is the most lethal gynecological malignancy and molecular mechanisms of its progression and metastasis are not completely understood. Some members of GAB (GRB2-associated binding) protein family have been reported to be involved in tumor cell proliferation and metastasis in various cancer types. In the present study, we analyzed the expression of GAB proteins (GAB1, GAB2 and GAB3) in ovarian cancer compared to normal ovarian tissue, in terms of tumor stage, tumor grade and histological type. Differential expression analyses performed in R programming environment using multiple transcriptome datasets (n = 1449) showed that GAB1 expression is decreased in ovarian cancer independently of tumor stage, grade and histotype. Unlike GAB1, expression of GAB2 and GAB3 are increased from early stage to late stage and from low grade to high grade in epithelial ovarian cancer. GAB2 and GAB3 also showed histotype-dependent expression. GAB3 was computed as a top gene whose expression most significantly changed between tumor cells from primary tumor, metastases and ascites. High expression of GAB2 and GAB3 was shown to be associated with shorter progression-free survival in ovarian cancer. This study shows that GAB2 and GAB3 can be important regulators of tumor progression and metastasis in ovarian cancer.

Electronic supplementary material

The online version of this article (10.1007/s12079-020-00582-3) contains supplementary material, which is available to authorized users.

Keywords: Cancer progression; GAB2; GAB3; Metastasis; Ovarian cancer, Transcriptomics

Introduction

Ovarian cancer (OC) is the most lethal gynecological malignancy and more than 125.000 women per year die from this disease worldwide (Ferlay et al. 2015). Epithelial ovarian cancer (EOC) comprises of different histological subtypes/histotypes: clear cell, endometrium, mucinous and serous (high or low grade) (Berkel and Cacan 2020). High grade serous ovarian cancer (HGSOC) is the most common and deadly histotype. Due to the lack of characteristic symptoms and effective screening tools to detect ovarian cancer, most patients are diagnosed at an advanced stage and this leads to high mortality rates in this cancer type (Cacan 2016). Also, high recurrence and metastasis rates following a clinical response with platinum- and/or paclitaxel-based chemotherapy contribute to increased mortality (Cacan et al. 2014). A better understanding of molecular mechanisms leading to ovarian cancer progression and metastasis is urgently needed to develop novel strategies in order to increase survival rate of ovarian cancer patients.

Growth factor receptor bound protein 2 (GRB2)-associated binding (GAB) family proteins including GAB1, GAB2 and GAB3 are large scaffold/docking proteins which integrate and amplify molecular signals from various signaling molecules including growth factors, cytokines, antigen receptors and cell adhesion molecules to coordinate many signaling pathways (Wöhrle et al. 2009). Recent studies have shown the contributions of GAB proteins to tumor progression and metastasis in several cancer types including breast cancer, melanoma, head and neck squamous cell carcinoma and colorectal cancer (Bentires-Alj et al. 2006; Horst et al. 2009; Hoeben et al. 2013; Seiden-Long et al. 2008).

Amplification of GAB2 (a paralog of GAB1 and GAB3) was shown in ovarian cancer and this amplification was associated with serous histotype of ovarian cancer (Brown et al. 2008). In a study by Wang et al., GAB2 were able to promote epithelial-to-mesenchymal transition (EMT) characteristics in ovarian cancer cell lines via PI3K-Zeb1 pathway by inhibiting E-cadherin expression and enhancing Zeb1 expression (a TF involved in EMT) (Wang et al. 2012). GAB2 was also shown to transform immortalized ovarian cell lines and this GAB2-induced transformation required PI3K pathway activation (Dunn et al. 2014). This study also reported that GAB2-overexpressing ovarian cancer cell lines require GAB2 for survival. Another study showed that GAB2 overexpression leads to tumor growth and angiogenesis in ovarian cancer cells by the upregulation of several chemokine genes (Duckworth et al. 2016).

It was also shown that TGF-beta-dependent repression of miRNA-125b in ascites of ovarian cancer patients resulted in an increased expression of GAB2 (a target gene of miR-125b) and this upregulation of GAB2 promoted increased tumor cell migration (Yang et al. 2017). Huang et al. showed that lncRNA snaR and GAB2 are increased in the serum of OC patients and that lncRNA snaR overexpression resulted in increased GAB2 expression and cancer cell proliferation (Huang et al. 2018). Another study reported that GAB2 can promote cell migration/invasion in ovarian cancer by the downregulation of miR-200c (Fang et al. 2019).

Unlike GAB2, studies for its paralogs GAB1 and GAB3 in ovarian cancer are highly limited. GAB1 was reported to be involved in a non receptor tyrosine kinase FER-mediated metastasis mechanism in ovarian cancer cells (Fan et al. 2016). This is the only study reporting the involvement of GAB1 in ovarian cancer. GAB3, a paralog of both GAB1 and GAB2, has not been studied previously in the context of ovarian cancer.

In the current study, we analyzed the expression levels of GAB1, GAB2 and GAB3 in ovarian cancer tissues compared to normal ovarian tissue, in terms of tumor stage, tumor grade and histological subtypes. We saw that GAB1 levels are decreased in ovarian cancer compared to normal ovary, in all datasets, independently of tumor stage, grade and histotype. However, GAB2 levels are decreased mostly in early stage or low grade ovarian cancer, and not in serous histotype, compared to normal ovary. GAB3 levels are decreased in serous ovarian cancer and this decrease is the most significant in early stage. GAB3 levels are not significantly different in late-stage, high-grade serous ovarian cancer compared to normal ovary; and GAB3 levels are lower in endometrioid histotype compared to serous histotype. During cancer progression (from early to late stage and from low to high grade), we observed that GAB2 and GAB3 levels are mostly increased in ovarian cancer in a histotype-dependent manner.

We also computed top genes differentially expressed between cancer cells from primary tumors, metastases and ascites. GAB3 was identified to be a top hit (primary vs ascites) and a third hit (primary vs metastases) in these analyses. We also showed that high GAB2 or GAB3 expressions (but not GAB1) are associated with decreased progression-free survival (PFS) in ovarian cancer. Lastly, copy number gains for all three GAB genes were shown to be most frequent in ovarian cancer among all cancer types.

Materials and methods

Datasets

In this study, following gene expression datasets from Gene Expression Omnibus (GEO) were used: GSE6008 (n = 103), GSE51088 (n = 172), GSE12470 (n = 53), GSE18520 (n = 63), GSE26712 (n = 195), GSE9891 (n = 285) (Hendrix et al. 2006; Karlan et al. 2014; Yoshihara et al. 2009; Mok et al. 2009; Bonome et al. 2008; Tothill et al. 2008). These GEO datasets and The Cancer Genome Atlas (TCGA) (n = 578) data were loaded into R statistical computing language/environment using curatedOvarianData R / Bioconductor package which includes many clinically annotated ovarian cancer transcriptome data (Cancer Genome Atlas Research Network 2011; Ganzfried et al. 2013). Ovarian cancer gene expression datasets were selected based on the criteria that they contain data for both ‘healthy’ and ‘tumor’ sample type. GSE9891 was also included in the analysis since it contains data for 2 histological subtypes of epithelial ovarian cancer (endometrioid and serous) (Tothill et al. 2008). Some of the datasets used in current study, which contain expression data for GAB1 and GAB2 genes, do not contain data for GAB3 gene; therefore, less figures for GAB3 were shown.

Another dataset (GSE73168) was used to compare the transcriptomes of tumor cells from primary tumors, metastases and ascites of HGSOC patients (Gao et al. 2019). All GEO datasets used in this study can be accessed using their GEO accession IDs in https://www.ncbi.nlm.nih.gov/geo/. First 6 GEO datasets and TCGA data can also be loaded via curatedOvarianData package in R (Ganzfried et al. 2013, R Core Team 2020).

Transcriptome analysis

Differential expression (DE) analyses of GAB genes in ovarian cancer were performed in terms of sample type (levels: healthy, tumor (and benign, borderline and metastatic tumors, where available)), tumor stage (levels: early stage and late stage), tumor grade (levels: low grade and high grade) and epithelial ovarian cancer histological type (histotype) (levels: clear cell, endometrium, mucinous and serous). Expression data and clinical metadata were retrieved from large Expression Set objects using functions in Biobase R package (Huber et al. 2015). A total of 1449 clinical samples were analyzed in this study.

Top hits of differentially expressed (DE) genes between tumor cells from primary tumor, metastases and ascites in HGSOC were computed using Biobase, GEOquery and limma packages (https://www.ncbi.nlm.nih.gov/geo/geo2r/) (Huber et al. 2015; Davis and Meltzer 2007; Ritchie et al. 2015). DE genes between cases were ordered based on their significance (p value). GAB3 expression values for each patient were used in data visualization.

Statistical analyses were performed between cases with unpaired Student’s t Test using ggpubr R package, and significance levels or formatted p values were given in figures and in text (Kassambara 2020). Following convention for star symbols indicating statistical significance was used: ns (non-significant): p > 0.05; *: p < = 0.05; **: p < = 0.01, ***: p < = 0.001; ****: p < = 0.0001.

Protein expression and proteome data

Protein data for GAB proteins in ovarian cancer was retrieved from the following datasets: The MaxQuant Database (http://maxqb.biochem.mpg.de/mxdb/, project P019, Eckert et al. 2019), TCGA-OV reverse phase protein array (RPPA) dataset (Cancer Genome Atlas Research Network 2011), The MD Anderson Cell Lines Project (MCLP) datasets (https://tcpaportal.org/mclp, Li et al. 2017), The CCLE (Cancer Cell Line Encyclopedia) datasets (https://portals.broadinstitute.org/ccle, Barretina et al. 2012) and The Human Protein Atlas datasets (http://www.proteinatlas.org, Uhlén et al. 2015; Uhlén et al. 2017).

Protein interaction and gene ontology analysis

GAB1, GAB2 and GAB3 protein interaction networks were identified using BioGRID protein interaction repository (https://thebiogrid.org/, Oughtred et al. 2019). For GAB1 and GAB2, only proteins which were shown to interact with GAB1 or GAB2 by at least two experiments (evidences) were included for further analyses. Molecular function (MF) ontology terms associated with GAB proteins and their protein interactors were investigated using DAVID tool (The Database for Annotation, Visualization and Integrated Discovery, https://david.ncifcrf.gov/home.jsp; Huang et al. 2009a; Huang et al. 2009b).

Survival analysis

Kaplan Meier plots were drawn using Kaplan-Meier Plotter tool for ovarian cancer (n = 1435) using the JetSet best probe set for each gene (https://kmplot.com/analysis/index.php?p=service) (GAB1: 214987_at, GAB2: 203853_s_at, GAB3: 228410_at) (Gyorffy et al. 2012). Median survival (in months) of ovarian cancer patients with low or high GAB gene (GAB1, GAB2 and GAB3) expression was computed based on progression free survival (PFS), and logrank p values were calculated. We did not restrict survival analysis to any ovarian cancer subtypes and treatment groups.

Copy number variation (CNV) analysis

Percentage of copy number variation (CNV) events in GAB1, GAB2 and GAB3 genes in TCGA (The Cancer Genome Atlas) (Cancer Genome Atlas Research Network 2011) projects across all cancer types were accessed through Genomic Data Commons Data Portal (https://portal.gdc.cancer.gov/). Exported tsv files were analyzed in R following some data wrangling, to visualize CNV gain percentages in 3 GAB genes among all cancer types. TCGA projects in y-axis were ordered in decreasing order of total CNV gain event percentages in three genes.

Data analysis and visualization

All data analyses and visualization in this study were performed in R statistical computing and graphics environment (R Core Team 2020). In addition to above mentioned R / Bioconductor packages, following R packages were used in different parts of analysis: tidyverse (an opinionated collection of R packages designed for data science), readxl, rmarkdown, knitr, gridExtra, magick and pdftools (Wickham et al. 2019; Wickham and Bryan 2019; Allaire et al. 2020; Xie 2020; Auguie 2017; Ooms 2020a, 2020b).

Corresponding datasets were indicated in figure captions (at the bottom right of each panel) and same panel background color was used for the datasets shared between figures for the ease of reading. R code written to analyze and visualize the data used in the current study is available in Supplementary files as RMarkdown and pdf documents to make it possible for other researchers to replicate/reproduce the results of our study.

Results

GAB1 expression is decreased in ovarian cancer independently of tumor stage, grade and histotype

We first analyzed GAB1 expression levels in ovarian tissue samples from healthy individuals and ovarian cancer patients, using 5 GEO (Gene Expression Omnibus) datasets (GSE6008, GSE51088, GSE12470, GSE18520 and GSE26712) and The Cancer Genome Atlas (TCGA) data. In epithelial ovarian cancer (EOC) samples, GAB1 levels are decreased compared to healthy ovarian samples (Fig. 1a, b). GAB1 levels are also decreased in borderline (ovarian tumors of low malignant potential (LMP)) and metastatic tumors, but not in benign tumors, compared to normal ovarian tissue (Fig. 1b). We then analyzed two different gene expression datasets in which data only for serous ovarian cancer (most common form of epithelial ovarian cancer) are present (Fig. 1c, d). We again observed that GAB1 expression levels in this histotype are lower than that of normal ovarian tissue (Fig. 1c, d). GSE18520 and GSE26712 datasets contain expression data only for late-stage, high-grade ovarian cancer. In these datasets, GAB1 levels are also decreased in tumor samples compared to normal ovarian tissue samples (Fig. 1e, f).

Fig. 1.

Fig. 1

Open in a new tab

GAB1 expression is decreased in ovarian cancer compared to normal ovarian tissue. GAB1 expression is decreased in epithelial ovarian cancer (EOC) (a-b), in serous histotype (a histological subtype of EOC) (c-d) and in late-stage, high-grade ovarian cancer (e-f), compared to healthy ovarian tissue. Types of analyzed ovarian cancer samples were given in figure titles. Transcriptome dataset IDs were given in figure captions at the bottom right. Same panel background color was used for the datasets shared between figures. ns (non-significant): p > 0.05; *: p < = 0.05; **: p < = 0.01, ***: p < = 0.001; ****: p < = 0.0001

In order to understand if the decreased expression of GAB1 in ovarian cancer is tumor stage- or grade-dependent, we analyzed the previous datasets in terms of tumor stage (early and late stage) and grade (low and high grade) (Fig. 2a-2g). We saw that GAB1 levels are decreased in ovarian cancer independently of tumor stage and grade (Fig. 2a-2g). In other words, both early and late stage, and both low and high grade ovarian tumors have lower GAB1 expression relative to matched healthy control samples.

Fig. 2.

Fig. 2

Open in a new tab

Decreased GAB1 expression in ovarian cancer is independent of tumor stage, grade and histotype. GAB1 expression levels in ovarian cancer compared to healthy ovarian tissue were given in terms of tumor stage (a, c, e, f), tumor grade (b, d, g) and histotype (h, i). GAB1 expression levels in endometrioid and serous histotypes in comparison were given (j). Types of analyzed ovarian cancer samples were given in figure titles. Transcriptome dataset IDs were given in figure captions at the bottom right. Same panel background color was used for the datasets shared between figures. ns (non-significant): p > 0.05; *: p < = 0.05; **: p < = 0.01, ***: p < = 0.001; ****: p < = 0.0001. endo: endometrioid, ser: serous

We then analyzed GAB1 expression levels in terms of epithelial ovarian cancer histotypes (clear cell, endometrioid (second most common form of EOC after serous histotype), mucinous, serous). Fig. 2h and i show that GAB1 levels are decreased in all histotypes compared to normal ovarian tissue. Most significant decrease in GAB1 expression seems to be in serous and endometrioid subtypes (Fig. 2h, i). In another dataset in which data only for endometrioid and serous histological types (bot not normal ovarian samples) are present, we also did not observe any difference between these subtypes in terms of GAB1 expression levels (Fig. 2j).

GAB2 expression levels change in ovarian cancer in a manner dependent on tumor stage, grade and histotype

We performed the same differential expression (DE) analyses for GAB2, a paralog of GAB1. We observed that GAB2 expression levels are lower in epithelial ovarian cancer compared to normal ovary (Fig. 3a, b). However, we did not see any significant decrease in GAB2 levels in tumors classified as benign, borderline or metastatic, compared to normal ovarian samples (Fig. 3b). In serous ovarian cancer (a subtype of EOC), GAB2 levels were not significantly different than in healthy ovarian tissue samples (Fig. 3c, d). Similarly, we did not observe any difference in GAB2 levels in late-stage, high-grade ovarian cancer samples, relative to healthy controls (Fig. 3e, f). However, at the protein level, tumor cells seem to have lower average GAB2 protein levels compared to surrounding stromal cells in high-grade serous ovarian cancer (Fig. 3g, p value is 0.1 due to low sample size in this proteomics study; mean GAB2 protein expression values were given in black).

Fig. 3.

Fig. 3

Open in a new tab

GAB2 expression is decreased in epithelial ovarian cancer; but not in its serous histotype and not in late stage, high grade ovarian cancer. GAB2 expression levels in epithelial ovarian cancer (EOC) (a-b), in serous histotype (a histological subtype of EOC) (c-d) and in late-stage, high-grade ovarian cancer (e-f), compared to healthy ovarian tissue. GAB2 protein levels are decreased in HGSOC cells compared to ovarian stroma (g). Types of analyzed ovarian cancer samples were given in figure titles. Transcriptome dataset IDs were given in figure captions at the bottom right. Same panel background color was used for the datasets shared between figures. ns (non-significant): p > 0.05; *: p < = 0.05; **: p < = 0.01, ***: p < = 0.001; ****: p < = 0.0001

Following the observation that GAB2 levels are decreased in epithelial ovarian cancer, but not in one of its histotypes (serous) and not in late-stage, high-grade ovarian cancer; we wanted to see if GAB2 levels change in ovarian cancer in a tumor stage- and grade-, or histological subtype-dependent manner. Indeed, we observed that, in early stage or low grade ovarian tumors, GAB2 levels are lower than healthy ovarian tissues (for stage: Fig. 4a, c, e; for grade: Fig. 4b, d). From early to late stage, and from low to high grade, GAB2 levels seem to be increasing in epithelial ovarian cancer back to the levels observed in healthy ovarian tissues, pointing the dynamic expression of GAB2 during cancer progression (Fig. 4a-e). In one of the four datasets analyzed (TCGA), we did not see a decreased GAB2 expression in early stage or low grade ovarian cancer (Fig. 4f, d; TCGA dataset only contains serous ovarian cancer samples and in the previous figure (Fig. 3c, d), we showed that GAB2 levels are not decreased in this histotype of EOC compared to healthy ovarian cells); however, we considered that this might be due to ovarian cancer histotype (serous), which is explained in the next paragraph. Also, at the protein level, we confirmed that GAB2 levels do not differ between tumor grades (Supp. Figure 1a, b) or stages (Supp. Figure 1c) in serous ovarian cancer, similar that observed at the mRNA level.

Fig. 4.

Fig. 4

Open in a new tab

GAB2 expression is decreased in ovarian cancer in a manner dependent on tumor stage, tumor grade and histotype. GAB2 expression levels in ovarian cancer compared to healthy ovarian tissue were given in terms of tumor stage (a, c, e, f), tumor grade (b, d, g) and histotype (h, i). GAB2 expression levels in endometrioid and serous histotypes in comparison were given (j). Types of analyzed ovarian cancer samples were given in figure titles. Transcriptome dataset IDs were given in figure captions at the bottom right. Same panel background color was used for the datasets shared between figures. ns (non-significant): p > 0.05; *: p < = 0.05; **: p < = 0.01, ***: p < = 0.001; ****: p < = 0.0001. endo: endometrioid, ser: serous

When the differential expression of GAB2 was analyzed in terms of epithelial ovarian cancer histotypes; unlike in GAB1, we did not see any significant decrease in GAB2 levels in serous histotype compared to normal ovarian tissue (Fig. 4h, i). It seems that most significant decrease in GAB2 levels among all histotypes compared to healthy tissue is in endometrioid histological subtype (Fig. 4h, i). In another dataset in which data only for endometrioid and serous histological types (bot not normal ovarian samples) are present, we also observed that GAB2 levels are lower in endometrioid histotype compared to serous histotype of EOC, in parallel with that observed in two previous datasets (Fig. 4j). Additionally, in ovarian cancer cells lines, GAB2 protein levels seem to be higher in serous histotype compared to the other histotypes (Supp. Figure 1d, e).

GAB3 expression levels change in serous ovarian cancer in a tumor stage- and grade-dependent manner

Next, we analyzed GAB3 expression levels, a paralog of both GAB1 and GAB2, in the previous ovarian cancer datasets when the expression data for GAB3 are available. We observed that GAB3 levels are decreased in serous ovarian cancer compared to normal ovarian tissue (Fig. 5a). We also saw that in early stage, GAB3 levels are more significantly decreased and that GAB3 expression increases from early to late stage, in serous ovarian cancer (most common form of EOC) (Fig. 5b). Similar to GAB2, GAB3 levels are not significantly different than normal ovarian tissues in late-stage, high-grade ovarian cancer (Fig. 5c) and GAB3 levels are lower in endometrioid histotype compared to serous histotype (Fig. 5d). We also showed that, at the protein level, GAB3 has a medium expression in normal ovarian tissues (Supp. Figure 2, top); however, its protein level is among the lowest within all cancer types, possibly showing that GAB3 levels decrease in ovarian cancer at the protein level, too (Supp. Figure 2, bottom, The Human Protein Atlas).

Fig. 5.

Fig. 5

Open in a new tab

GAB3 expression is decreased in ovarian cancer in a manner dependent on tumor stage, tumor grade and histotype. GAB3 expression levels in serous ovarian cancer were given compared to normal ovarian tissue (a). Tumor stage-dependent expression of GAB3 in serous ovarian cancer was shown (b). Expression of GAB3 in late stage, high grade ovarian cancer was compared to that of healthy ovarian tissue (c). GAB3 expression levels in endometrioid and serous histotypes in comparison were given (d). Types of analyzed ovarian cancer samples were given in figure titles. Transcriptome dataset IDs were given in figure captions at the bottom right. Same panel background color was used for the datasets shared between figures. ns (non-significant): p > 0.05; *: p < = 0.05; **: p < = 0.01, ***: p < = 0.001; ****: p < = 0.0001. endo: endometrioid, ser: serous

Ovarian cancer cells in metastases and ascites have higher GAB3 levels compared to cancer cells in primary tissue

Based on the observation that GAB2 and GAB3 levels are increasing from early to late stage, and from low to high grade, we wanted to see if tumor cells from metastases and ascites have higher GAB2 and GAB3 levels than primary ovarian tumors in the same patients. We compared transcriptomes of tumor cells from primary tumor, metastases and ascites in HGSOC (most common form, accounting for approximately 68% of ovarian carcinoma) using the gene expression dataset GSE73168. We first computed top genes which are differentially expressed (DE) between primary tumor and tumors cells from metastases. We identified GAB3 as a third top hit in this comparison (p = 0.00006385, first two gene were NR5A2 and FOXD4L1///FOXD4). We then compared the expression profiles of primary ovarian tumors and tumor cells from ascites. In this comparison, GAB3 was identified as the top hit which is the gene whose expression is most significantly changed between these two cases (p = 7.44e-07). GAB3 expression values based on tumor cell location (primary tumor, metastases and ascites) were shown in Fig. 6.

Fig. 6.

Fig. 6

Open in a new tab

Ovarian cancer cells in metastases and ascites have higher GAB3 levels compared to cancer cells in primary tissue. GAB3 expression levels were compared between tumor cells from primary tumor, metastases and ascites in HGSOC (High grade serous ovarian cancer) (a). Samples for each patient were given a unique color. Transcriptome dataset ID was given in figure caption at the bottom right. GAB2 protein expression levels in ovarian cancer cell lines in terms of the location from which cell lines were isolated (b-d). Molecular function (MF) gene ontology (GO) terms associated with proteins in the GAB3 protein interaction network (e). Count represents total number of genes out of 6 genes for interactor proteins in that term. P values were given at the top of bar plots. ns (non-significant): p > 0.05; *: p < = 0.05; **: p < = 0.01, ***: p < = 0.001; ****: p < = 0.0001. MCLP: The MD Anderson Cell Lines Project; CCLE: Cancer Cell Line Encyclopedia

Next, using proteomics data of ovarian cancer cell lines in two different datasets (MCLP and CCLE), we found that, similar to GAB3, mean GAB2 protein level is higher in cell lines from metastases compared to cell lines from primary ovarian tissue (Fig. 6b, c). In parallel with GAB3, GAB2 protein expression is also increased in cell lines from ascites compared to cell lines from primary ovarian tissue (Fig. 6d). Although p values are higher than 0.05 in these cell line comparisons due to high variance and presence of many confounding factors in cell lines, trend in GAB2 protein levels are clear (mean GAB2 protein levels were given as yellow points and in numbers). Detailed data including cell line names on ovarian cancer cell lines present in these datasets can be found in the supplementary files.

In order to have an insight on the possible molecular functions of GAB3 in ovarian cancer, we studied gene ontology (GO) terms associated with GAB3 and its protein interactors. We first identified 6 unique physical protein interactors of GAB3 using BioGRID protein interaction repository (LYN, PTPN11, SRC, FYN, GRB2 and ABL1; GAB3 protein network were given in Supp. Figure 3). We saw that 4 of these 6 interactors are proto-oncogenes (namely LYN, SRC, FYN and ABL1; and PTPN11 was implicated in oncogenic transformation, more detail at Discussion). Next, we identified molecular function (MF) ontology terms associated with proteins in GAB3 interaction network using DAVID functional annotation tool and results were given in Fig. 6e. We also found that top interactor of GAB3, LYN, has a lower protein expression in high-grade serous ovarian cancer cells compared to ovarian stromal cells (Supp. Figure 4) (Results of GO analyses for GAB1 and GAB2 protein interaction networks were provided in the supplementary files).

High expression of GAB2 and GAB3, but not GAB1, is associated with shorter progression-free survival in ovarian cancer

Since we saw that GAB2 and GAB3 levels are increased in late stage or high grade ovarian tumors compared to early stage or low grade ovarian tumors, respectively; we wanted to see the effect of their increased expression on progression free survival (PFS) of ovarian cancer patients. We performed survival analyses using Kaplan-Meier plots, and survival plots for GAB1, GAB2 and GAB3 were given in Fig. 7. Here, we saw that high expression of GAB2 or GAB3 result in shorter median survival (high GAB2: 19 months, low GAB2: 23.56 months (p = 8.6e-06); high GAB3: 13.7 months, low GAB3: 20 months (p = 0.00046)). However, high expression of GAB1 does not result in shorter median survival (high GAB1: 21.59 months, low GAB1: 18 months (p = 0.019)) (Fig. 7).

Fig. 7.

Fig. 7

Open in a new tab

Progression-free survival (PFS) of ovarian cancer patients with low and high expression of GAB1, GAB2 and GAB3. Kaplan Meier plots showing PFS data for ovarian cancer patients with low (black) and high (red) expression of GAB1, GAB2 and GAB3. Logrank p values were given for each gene. Median survival periods were computed as follows: low GAB1: 18 months, high GAB1: 21.59 months (p = 0.019); low GAB2: 23.56 months, high GAB2: 19 months (p = 8.6e-06); low GAB3: 20 months, high GAB3: 13.7 months (p = 0.00046). HR: Hazard ratio

Copy number gains of GAB genes is the highest in ovarian cancer among all cancer types

We next analyzed The Cancer Genome Atlas data for Copy Number Variation (CNV) events for GAB1, GAB2 and GAB3 genes across all cancer types available. For all three genes, we observed that most frequent copy number gain events (CNV gain) were in ovarian cancer compared to other cancer types (Fig. 8). Percentage of CNV gain events in TCGA-OV project were as following: GAB1 (13.85%), GAB2 (31.62%) and GAB3 (20.68%) (Fig. 8). Most frequent CNV gain events in ovarian cancer was seen in GAB2, followed by GAB3 and GAB1.

Fig. 8.

Fig. 8

Open in a new tab

Copy number gain events of GAB genes is the highest in ovarian cancer among all cancer types. Percentage of copy number gain (CNV gain) events in GAB1, GAB2 and GAB3 genes in TCGA projects across all cancer types. Data from Genomic Data Commons Data Portal (https://portal.gdc.cancer.gov/). TCGA project names in y-axis were ordered in decreasing order of total CNV gain event percentages in three genes. TCGA-OV: The Cancer Genome Atlas Ovarian Cancer

Discussion

In the present study, we studied differential expression of GAB1, GAB2 and GAB3 in ovarian cancer compared to normal ovarian tissue, in terms of tumor stage, tumor grade and histotype, using publicly available transcriptome datasets including TCGA. We showed that GAB1 expression is decreased in ovarian cancer compared to normal ovary, independently of tumor stage, grade and histotype. This might be an early event in tumor development and GAB1 expression levels remain lower during cancer progression from early stage and low grade into late stage or high grade, respectively. Although increased expression of GAB1 was reported to be present and to promote metastasis in certain subtypes of breast cancer (Wang et al. 2019); its expression seems to be lower in ovarian cancer subtypes (in terms of tumor stage, grade and histotype) compared to healthy ovarian cells. Survival analysis also supports this observation, since increased expression of GAB1 does not result in shorter progression-free survival in ovarian cancer patients (high GAB1: 21.59 months, low GAB1: 18 months (p = 0.019)).

We observed that GAB2 levels are decreased in early stage or low grade ovarian cancer compared to normal ovarian cells; however, its levels are increased in late stage or high grade epithelial ovarian cancer, compared to early stage and low grade epithelial ovarian cancer, respectively. This indicates the dynamic expression of GAB2 during cancer progression, firstly decreasing in the transition from healthy ovaries to early stage / low grade ovarian cancer, and then again increasing in late stage / high grade ovarian cancer. This change in GAB2 expression was more pronounced in epithelial ovarian cancer histotypes other than serous histotype. In serous histotype, we found that GAB2 levels remain high during cancer progression and transformation (from early to late stage, and from low to high grade) at both mRNA and protein level. GAB2 was previously shown to promote EMT (epithelial-to-mesenchymal transition) characteristics and lead to tumor cell migration / invasion in ovarian cancer (Wang et al. 2012; Yang et al. 2017; Fang et al. 2019). Therefore, it can be proposed that increased expression of GAB2 in late stage or high grade ovarian cancer might contribute to higher migration and invasion observed in late stage or high grade ovarian cancer cells. In other words, ovarian cancer cells might gain more metastatic potential by the upregulation of GAB2 during tumor progression. We also saw that GAB2 levels are higher in serous histotype compared to endometrioid histotype (similar trend was also observed at the protein level) and this is also in line with the fact that high grade serous ovarian cancer is the most lethal histotype of epithelial ovarian cancer (Wu et al. 2019). This is possibly in part due to increased tumor cell migration / invasiveness promoted by higher GAB2 levels seen in serous histotype. We further analyzed progression-free survival (PFS) periods of patients with low and high expression of GAB2, and reported that high GAB2 expression is associated with decreased survival in ovarian cancer. Consequently, it can be inferred from these data that increased expression of GAB2 in late stage or high grade ovarian cancer might contribute to decreased progression-free survival of ovarian cancer patients with high GAB2 expression.

Similar to GAB2, GAB3 levels are more significantly decreased in early stage ovarian cancer compared to healthy ovaries. We observed higher GAB3 expression in late stage serous ovarian cancer compared to early stage, pointing dynamic expression of GAB3 during tumor progression. In parallel with GAB2, GAB3 levels seem to be decreasing in the transition from healthy ovaries to early stage ovarian cancer, and then again increasing from early stage to late stage OC. In late stage, high grade ovarian cancer, GAB3 levels are not statistically different from normal ovarian cells. Alike GAB2, GAB3 levels were higher in serous histotype compared to endometrioid histotype. Thus, it can be speculated that GAB3, like its paralog GAB2, might be contributing to ovarian cancer progression from early to late stage or from low to high grade, by the upregulation of its expression. Supporting this, we observed shorter progression-free survival in ovarian cancer patients with high GAB3 expression. To our knowledge, this is the first study which shows tumor stage-dependent expression of GAB3 in ovarian cancer and which reports the association of high GAB3 expression with shorter PFS in ovarian cancer patients.

Two studies reported that GAB3 can mediate cancer cell proliferation by the activation of Akt signaling, in colorectal cancer and glioma (Xiang et al. 2017; Jia et al. 2017). Although GAB3 levels were found to be decreased in ovarian cancer compared to healthy ovarian cells; increased expression of GAB3 at late stage ovarian cancer compared to early stage might explain the shorter median survival associated with higher GAB3 expression in ovarian cancer. Also, higher expression of GAB3 in serous histotype compared to endometrioid histotype might be contributing to the worse outcome seen in high grade serous ovarian cancer.

We also computed top genes differentially expressed between cancer cells from primary tumor, metastases and ascites in HGSOC. GAB3 was identified as a top hit (primary vs ascites) and a third hit (primary vs metastases) in this analysis. These data point to the fact that primary ovarian cancer cells might exploit and regulate GAB3 expression to gain more metastatic potential. It is also possible that tumor microenvironment in metastases and ascites might be contributing to the increased GAB3 expression. Yang et al. showed that TGF-beta-dependent repression of miR-125b in ascites can promote cancer cell migration by the upregulation of GAB2 (a target gene of miR-125b) (Yang et al. 2017). A similar mechanism might be also responsible for increased expression of GAB3 in metastases and ascites observed in the current study.

Using ovarian cancer cell line protein data, we also found that cell lines from metastases and ascites have higher mean GAB2 protein levels compared to cell lines from primary ovarian tissue. This is in line with the observation that GAB3 mRNA levels are higher in metastases and ascites than primary ovarian cells. Together, these data show that GAB2 and GAB3 might be contributing factors to metastatic transformation in ovarian cancer.

We identified molecular function ontology terms associated with GAB3 and its protein interactors to have an insight on the potential roles of GAB3 in ovarian cancer. 4 of its 6 protein interactors (LYN, FYN, SRC, PTPN11, ABL1 and GRB2) were found to be proto-oncogenes and one was implicated in oncogenic transformation. One GAB3 interactor, FYN, was reported to enhance tumor cell invasive activity in ovarian cancer (Zhao et al. 2011). Another protein interactor, SRC, has been shown to be involved in ovarian cancer cell proliferation and metastasis (Chen et al. 2019). Hu et al. (2017) showed that PTPN11 overexpression enhances ovarian cancer cell invasion and metastasis. In addition, inhibition of another GAB3 interactor, GRB2, was shown to result in decreased tumor growth and metastasis in preclinical models of ovarian cancer (Lara et al. 2020). Based on the fact that most of the protein interactors of GAB3 were shown to be involved in ovarian cancer progression and metastasis; and also based on other analyses performed in the current study, it is highly plausible to speculate its potential role in ovarian cancer.

Top two molecular function terms associated with proteins in GAB3 protein network were found to be ‘non-membrane spanning protein tyrosine kinase activity’ and ‘protein tyrosine kinase activity’ in this study. Fan et al. (2016) showed that a nonreceptor tyrosine kinase (FER) might potentiate metastasis in ovarian cancer. Kanlikilicer et al. (2017) reported that targeting a receptor tyrosine kinase (AXL) inhibits tumor growth and intraperitoneal metastasis in ovarian cancer. In another study, it was shown that a tyrosine kinase (DDR2) promotes metastatic transformation in ovarian cancer (Grither et al. 2018). Therefore, it can be suggested that increased GAB3 expression might be promoting metastasis in ovarian cancer with the involvement of protein tyrosine kinases (possibly as downstream targets); however, this requires further mechanistic study.

Finally, we observed that copy number gains of GAB1, GAB2 and GAB3 was the most frequent in ovarian cancer among all cancer types in The Cancer Genome Atlas. This result highlights the importance of GAB proteins in ovarian cancer. Most frequent copy number gain events was observed in GAB2, followed by GAB3 and then GAB1. Considering the observation that high expression of GAB2 and GAB3, but not GAB1, is associated with shorter survival; ovarian cancer cells might exploit amplifications in GAB2 and GAB3 genes to promote cell proliferation or metastasis, ultimately resulting in worse outcome. Comprehensive mechanistic analysis of GAB proteins in tumor development, progression and metastasis will be highly valuable in ovarian cancer research.

Conclusion

The current study identifies the tumor stage-, tumor grade- and histotype-dependent expression of GAB2 and GAB3 genes in ovarian cancer. Increased expression of GAB2 and GAB3 in late stage or high grade ovarian cancer might a contributing factor to increased metastatic potential and worse prognosis in late stage or high grade ovarian cancer. Increased expression of GAB3 in ascites and metastases, compared to primary tumor, might also point to the role of GAB3 in ovarian cancer cell migration/invasiveness and metastases. In addition, high expression of GAB2 or GAB3 results in shorter progression-free survival (PFS) in ovarian cancer patients. This study identifies GAB2 and GAB3 as possible important players in ovarian cancer progression and metastasis.

Electronic supplementary material

ESM 1 (79.7KB, pdf)

(PDF 79 kb)

ESM 2 (5.3MB, pdf)

(PDF 5411 kb)

ESM 3 (33KB, png)

(PNG 33 kb)

ESM 4 (6KB, pdf)

(PDF 5 kb)

ESM 5 (7KB, pdf)

(PDF 7 kb)

ESM 6 (1.2MB, pdf)

(PDF 1223 kb)

ESM 7 (56KB, pdf)

(PDF 56 kb)

Funding

Caglar Berkel is funded by TUBITAK (The Scientific and Technological Research Council of Turkey) 2211-E graduate program.

Availability of data and material (data transparency)

Transcriptomics datasets used in the present study are available from Gene Expression Omnibus (GEO) and TCGA-OV (The Cancer Genome Atlas – Ovarian Cancer), and their accession IDs were given in the text. Most expression data can be accessed in R programming environment using curatedOvarianData Bioconductor data package. Protein datasets used in the study and information on how to access them were given in Materials and Methods section.

Compliance with ethical standards

Conflicts of interest/competing interests

Authors declare no conflicts of interest.

Ethics approval

Not applicable.

Consent to participate

Not applicable.

Consent for publication

Not applicable.

Code availability (software application or custom code)

R code written to analyze and visualize the data used in this study was provided as a supplementary file to make this study completely reproducible and replicable by other researchers.

Footnotes

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Contributor Information

Caglar Berkel, Email: caglar.berkel@gop.edu.tr.

Ercan Cacan, Email: ercan.cacan@gop.edu.tr.

References

  1. Allaire, JJ, Xie, Y, McPherson, J, Luraschi, J, Ushey, K, Atkins, A, Wickham, H, Cheng, J, Chang, W and Iannone, R (2020). Rmarkdown: dynamic documents for R. R package version 2.2. URL https://rmarkdown.rstudio.com
  2. Auguie, B (2017). gridExtra: miscellaneous functions for “grid” graphics. R package version 2.3. https://CRAN.R-project.org/package=gridExtra
  3. Barretina J, Caponigro G, Stransky N, Venkatesan K, Margolin AA, Kim S, Wilson CJ, Lehár J, Kryukov GV, Sonkin D, Reddy A, Liu M, Murray L, Berger MF, Monahan JE, Morais P, Meltzer J, Korejwa A, Jané-Valbuena J, Mapa FA, Thibault J, Bric-Furlong E, Raman P, Shipway A, Engels IH, Cheng J, Yu GK, Yu J, Aspesi P, de Silva M, Jagtap K, Jones MD, Wang L, Hatton C, Palescandolo E, Gupta S, Mahan S, Sougnez C, Onofrio RC, Liefeld T, MacConaill L, Winckler W, Reich M, Li N, Mesirov JP, Gabriel SB, Getz G, Ardlie K, Chan V, Myer VE, Weber BL, Porter J, Warmuth M, Finan P, Harris JL, Meyerson M, Golub TR, Morrissey MP, Sellers WR, Schlegel R, Garraway LA. The Cancer cell line encyclopedia enables predictive modelling of anticancer drug sensitivity. Nature. 2012;483(7391):603–607. doi: 10.1038/nature11003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bentires-Alj M, Gil SG, Chan R, Wang ZC, Wang Y, Imanaka N, Harris LN, Richardson A, Neel BG, Gu H. A role for the scaffolding adapter GAB2 in breast cancer. Nat Med. 2006;12(1):114–121. doi: 10.1038/nm1341. [DOI] [PubMed] [Google Scholar]
  5. Berkel C, Cacan E. In silico analysis of DYNLL1 expression in ovarian cancer chemoresistance. Cell Biol Int. 2020;44:1598–1605. doi: 10.1002/cbin.11352. [DOI] [PubMed] [Google Scholar]
  6. Bonome T, Levine DA, Shih J, Randonovich M, Pise-Masison CA, Bogomolniy F, Ozbun L, Brady J, Barrett JC, Boyd J, Birrer MJ. A gene signature predicting for survival in suboptimally debulked patients with ovarian cancer. Cancer Res. 2008;68(13):5478–5486. doi: 10.1158/0008-5472.CAN-07-6595. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Brown LA, Kalloger SE, Miller MA, Shih I, McKinney SE, Santos JL, Swenerton K, Spellman PT, Gray J, Gilks CB, Huntsman DG. Amplification of 11q13 in ovarian carcinoma. Genes Chromosom Cancer. 2008;47(6):481–489. doi: 10.1002/gcc.20549. [DOI] [PubMed] [Google Scholar]
  8. Cacan E. Histone Deacetylase-1-mediated suppression of FAS in Chemoresistant ovarian Cancer cells. Anticancer Res. 2016;36(6):2819–2826. [PubMed] [Google Scholar]
  9. Cacan E, Ali MW, Boyd NH, Hooks SB, Greer SF. Inhibition of HDAC1 and DNMT1 modulate RGS10 expression and decrease ovarian cancer chemoresistance. PLoS One. 2014;9(1):e87455. doi: 10.1371/journal.pone.0087455. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Cancer Genome Atlas Research Network Integrated genomic analyses of ovarian carcinoma. Nature. 2011;474(7353):609–615. doi: 10.1038/nature10166. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Chen D, Cao L, Wang X. MPZL1 promotes tumor cell proliferation and migration via activation of Src kinase in ovarian cancer. Oncol Rep. 2019;42(2):679–687. doi: 10.3892/or.2019.7199. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Davis S, Meltzer PS. GEOquery: a bridge between the gene expression omnibus (GEO) and BioConductor. Bioinformatics (Oxford, England) 2007;23(14):1846–1847. doi: 10.1093/bioinformatics/btm254. [DOI] [PubMed] [Google Scholar]
  13. Duckworth C, Zhang L, Carroll SL, Ethier SP, Cheung HW. Overexpression of GAB2 in ovarian cancer cells promotes tumor growth and angiogenesis by upregulating chemokine expression. Oncogene. 2016;35(31):4036–4047. doi: 10.1038/onc.2015.472. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Dunn GP, Cheung HW, Agarwalla PK, Thomas S, Zektser Y, Karst AM, Boehm JS, Weir BA, Berlin AM, Zou L, Getz G, Liu JF, Hirsch M, Vazquez F, Root DE, Beroukhim R, Drapkin R, Hahn WC. In vivo multiplexed interrogation of amplified genes identifies GAB2 as an ovarian cancer oncogene. Proc Natl Acad Sci U S A. 2014;111(3):1102–1107. doi: 10.1073/pnas.1311909111. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Eckert MA, Coscia F, Chryplewicz A, Chang JW, Hernandez KM, Pan S, Tienda SM, Nahotko DA, Li G, Blaženović I, Lastra RR, Curtis M, Yamada SD, Perets R, McGregor SM, Andrade J, Fiehn O, Moellering RE, Mann M, Lengyel E. Proteomics reveals NNMT as a master metabolic regulator of cancer-associated fibroblasts. Nature. 2019;569(7758):723–728. doi: 10.1038/s41586-019-1173-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Fan G, Zhang S, Gao Y, Greer PA, Tonks NK. HGF-independent regulation of MET and GAB1 by nonreceptor tyrosine kinase FER potentiates metastasis in ovarian cancer. Genes Dev. 2016;30(13):1542–1557. doi: 10.1101/gad.284166.116. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Fang Z, Li T, Chen W, Wu D, Qin Y, Liu M, Wu G, He L, Li H, Gu H. Gab2 promotes cancer stem cell like properties and metastatic growth of ovarian cancer via downregulation of miR-200c. Exp Cell Res. 2019;382(1):111462. doi: 10.1016/j.yexcr.2019.06.007. [DOI] [PubMed] [Google Scholar]
  18. Ferlay J, Soerjomataram I, Dikshit R, Eser S, Mathers C, Rebelo M, Parkin DM, Forman D, Bray F. Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer. 2015;136(5):E359–E386. doi: 10.1002/ijc.29210. [DOI] [PubMed] [Google Scholar]
  19. Ganzfried BF, Riester M, Haibe-Kains B, Risch T, Tyekucheva S, Jazic I, Wang XV, Ahmadifar M, Birrer MJ, Parmigiani G, Huttenhower C, Waldron L (2013) curatedOvarianData: clinically annotated data for the ovarian cancer transcriptome. Database : the journal of biological databases and curation, 2013, bat013. 10.1093/database/bat013 [DOI] [PMC free article] [PubMed]
  20. Gao Q, Yang Z, Xu S, Li X, Yang X, Jin P, Liu Y, Zhou X, Zhang T, Gong C, Wei X, Liu D, Sun C, Chen G, Hu J, Meng L, Zhou J, Sawada K, Fruscio R, Grunt TW, Wischhusen J, Vargas-Hernández VM, Pothuri B, Coleman RL. Heterotypic CAF-tumor spheroids promote early peritoneal metastatis of ovarian cancer. J Exp Med. 2019;216(3):688–703. doi: 10.1084/jem.20180765. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Grither WR & Longmore GD (2018) Inhibition of tumor-microenvironment interaction and tumor invasion by small-molecule allosteric inhibitor of DDR2 extracellular domain. PNAS 115(33), E7786–E7794. 10.1073/pnas.1805020115 [DOI] [PMC free article] [PubMed]
  22. Gyorffy B, Lánczky A, Szállási Z. Implementing an online tool for genome-wide validation of survival-associated biomarkers in ovarian-cancer using microarray data from 1287 patients. Endocr Relat Cancer. 2012;19(2):197–208. doi: 10.1530/ERC-11-0329. [DOI] [PubMed] [Google Scholar]
  23. Hendrix ND, Wu R, Kuick R, Schwartz DR, Fearon ER, Cho KR. Fibroblast growth factor 9 has oncogenic activity and is a downstream target of Wnt signaling in ovarian endometrioid adenocarcinomas. Cancer Res. 2006;66(3):1354–1362. doi: 10.1158/0008-5472.CAN-05-3694. [DOI] [PubMed] [Google Scholar]
  24. Hoeben A, Martin D, Clement PM, Cools J, Gutkind JS. Role of GRB2-associated binder 1 in epidermal growth factor receptor-induced signaling in head and neck squamous cell carcinoma. Int J Cancer. 2013;132(5):1042–1050. doi: 10.1002/ijc.27763. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Horst B, Gruvberger-Saal SK, Hopkins BD, Bordone L, Yang Y, Chernoff KA, Uzoma I, Schwipper V, Liebau J, Nowak NJ, Brunner G, Owens D, Rimm DL, Parsons R, Celebi JT. Gab2-mediated signaling promotes melanoma metastasis. Am J Pathol. 2009;174(4):1524–1533. doi: 10.2353/ajpath.2009.080543. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Hu Z, Li J, Gao Q, Wei S, Yang B. SHP2 overexpression enhances the invasion and metastasis of ovarian cancer in vitro and in vivo. OncoTargets and therapy. 2017;10:3881–3891. doi: 10.2147/OTT.S138833. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Huang D, Sherman BT, Lempicki RA. Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists. Nucleic Acids Res. 2009;37(1):1–13. doi: 10.1093/nar/gkn923. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Huang D, Sherman BT, Lempicki RA. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc. 2009;4(1):44–57. doi: 10.1038/nprot.2008.211. [DOI] [PubMed] [Google Scholar]
  29. Huang Y, Hu Y, Jin Z, Shen Z. LncRNA snaR upregulates GRB2-associated binding protein 2 and promotes proliferation of ovarian carcinoma cells. Biochem Biophys Res Commun. 2018;503(3):2028–2032. doi: 10.1016/j.bbrc.2018.07.152. [DOI] [PubMed] [Google Scholar]
  30. Huber W, Carey VJ, Gentleman R, Anders S, Carlson M, Carvalho BS, Bravo HC, Davis S, Gatto L, Girke T, Gottardo R, Hahne F, Hansen KD, Irizarry RA, Lawrence M, Love MI, MacDonald J, Obenchain V, Oleś AK, Pagès H, Reyes A, Shannon P, Smyth GK, Tenenbaum D, Waldron L, Morgan M. Orchestrating high-throughput genomic analysis with Bioconductor. Nat Methods. 2015;12(2):115–121. doi: 10.1038/nmeth.3252. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Jia P, Li F, Gu W, Zhang W, Cai Y. Gab3 overexpression in human glioma mediates Akt activation and tumor cell proliferation. PLoS One. 2017;12(3):e0173473. doi: 10.1371/journal.pone.0173473. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Kanlikilicer P, Ozpolat B, Aslan B, Bayraktar R, Gurbuz N, Rodriguez-Aguayo C, Bayraktar E, Denizli M, Gonzalez-Villasana V, Ivan C, Lokesh G, Amero P, Catuogno S, Haemmerle M, Wu SY, Mitra R, Gorenstein DG, Volk DE, de Franciscis V, Sood AK, et al. Therapeutic targeting of AXL receptor tyrosine kinase inhibits tumor growth and Intraperitoneal metastasis in ovarian Cancer models. Molecular therapy Nucleic acids. 2017;9:251–262. doi: 10.1016/j.omtn.2017.06.023. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Karlan BY, Dering J, Walsh C, Orsulic S, Lester J, Anderson LA, Ginther CL, Fejzo M, Slamon D. POSTN/TGFBI-associated stromal signature predicts poor prognosis in serous epithelial ovarian cancer. Gynecol Oncol. 2014;132(2):334–342. doi: 10.1016/j.ygyno.2013.12.021. [DOI] [PubMed] [Google Scholar]
  34. Kassambara A (2020) ggpubr: 'ggplot2' Based Publication Ready Plots. R package version 0.4.0. https://CRAN.R-project.org/package=ggpubr
  35. Lara OD, Bayraktar E, Amero P, Ma S, Ivan C, Hu W, Wang Y, Mangala LS, Dutta P, Bhattacharya P, Ashizawa AT, Lopez-Berestein G, Rodriguez-Aguayo C, Sood AK. Therapeutic efficacy of liposomal Grb2 antisense oligodeoxynucleotide (L-Grb2) in preclinical models of ovarian and uterine cancer. Oncotarget. 2020;11(29):2819–2833. doi: 10.18632/oncotarget.27667. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Li J, Zhao W, Akbani R, Liu W, Ju Z, Ling S, Vellano CP, Roebuck P, Yu Q, Eterovic AK, Byers LA, Davies MA, Deng W, Gopal YN, Chen G, von Euw EM, Slamon D, Conklin D, Heymach JV, Gazdar AF, et al. Characterization of human Cancer cell lines by reverse-phase protein arrays. Cancer Cell. 2017;31(2):225–239. doi: 10.1016/j.ccell.2017.01.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Mok SC, Bonome T, Vathipadiekal V, Bell A, Johnson ME, Wong KK, Park DC, Hao K, Yip DK, Donninger H, Ozbun L, Samimi G, Brady J, Randonovich M, Pise-Masison CA, Barrett JC, Wong WH, Welch WR, Berkowitz RS, Birrer MJ. A gene signature predictive for outcome in advanced ovarian cancer identifies a survival factor: microfibril-associated glycoprotein 2. Cancer Cell. 2009;16(6):521–532. doi: 10.1016/j.ccr.2009.10.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Ooms, J (2020a). Magick: advanced graphics and image-processing in R. R package version 2.3. https://CRAN.R-project.org/package=magick
  39. Ooms, J (2020b). Pdftools: text extraction, rendering and converting of PDF documents. R package version 2.3.1. https://CRAN.R-project.org/package=pdftools
  40. Oughtred R, Stark C, Breitkreutz BJ, Rust J, Boucher L, Chang C, Kolas N, O'Donnell L, Leung G, McAdam R, Zhang F, Dolma S, Willems A, Coulombe-Huntington J, Chatr-Aryamontri A, Dolinski K, Tyers M. The BioGRID interaction database: 2019 update. Nucleic Acids Res. 2019;47(D1):D529–D541. doi: 10.1093/nar/gky1079. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. R Core Team (2020). R: A language and environment for statistical computing. R Foundation for statistical computing, Vienna, Austria. URL https://www.R-project.org/
  42. Ritchie, ME, Phipson, B, Wu, D, Hu, Y, Law, CW, Shi, W and Smyth, GK (2015). Limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic acids research, 43(7), e47. 10.1093/nar/gkv00730.Kassambara, A. (2020). Ggpubr: ‘ggplot2’ based publication ready plots. R package version 0.3.0. https://CRAN.R-project.org/package=ggpubr [DOI] [PMC free article] [PubMed]
  43. Seiden-Long I, Navab R, Shih W, Li M, Chow J, Zhu CQ, Radulovich N, Saucier C, Tsao MS. Gab1 but not Grb2 mediates tumor progression in met overexpressing colorectal cancer cells. Carcinogenesis. 2008;29(3):647–655. doi: 10.1093/carcin/bgn009. [DOI] [PubMed] [Google Scholar]
  44. Tothill RW, Tinker AV, George J, Brown R, Fox SB, Lade S, Johnson DS, Trivett MK, Etemadmoghadam D, Locandro B, Traficante N, Fereday S, Hung JA, Chiew YE, Haviv I, Australian Ovarian Cancer Study Group. Gertig D, DeFazio A, Bowtell DD. Novel molecular subtypes of serous and endometrioid ovarian cancer linked to clinical outcome. Clinical cancer research : an official journal of the American Association for Cancer Research. 2008;14(16):5198–5208. doi: 10.1158/1078-0432.CCR-08-0196. [DOI] [PubMed] [Google Scholar]
  45. Uhlén, M, Fagerberg, L, Hallström, BM, Lindskog, C, Oksvold, P, Mardinoglu, A, Sivertsson, Å, Kampf, C, Sjöstedt, E, Asplund, A, Olsson, I, Edlund, K, Lundberg, E, Navani, S, Szigyarto, CA, Odeberg, J, Djureinovic, D, Takanen, JO, Hober, S, Alm, T, … Pontén, F (2015). Proteomics. Tissue-based map of the human proteome. Science (New York, N.Y.), 347(6220), 1260419. 10.1126/science.1260419 [DOI] [PubMed]
  46. Uhlén M, Zhang C, Lee S, Sjöstedt E, Fagerberg L, Bidkhori G, Benfeitas R, Arif M, Liu Z, Edfors F, Sanli K, von Feilitzen K, Oksvold P, Lundberg E, Hober S, Nilsson P, Mattsson J, Schwenk JM, Brunnström H, Glimelius B et al (2017) A pathology atlas of the human cancer transcriptome. Science (New York, N.Y.) 357(6352):eaan2507. 10.1126/science.aan2507 [DOI] [PubMed]
  47. Wang Y, Sheng Q, Spillman MA, Behbakht K, Gu H. Gab2 regulates the migratory behaviors and E-cadherin expression via activation of the PI3K pathway in ovarian cancer cells. Oncogene. 2012;31(20):2512–2520. doi: 10.1038/onc.2011.435. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Wang X, Peng J, Yang Z, Zhou PJ, An N, Wei L, Zhu HH, Lu J, Fang YX, Gao WQ. Elevated expression of Gab1 promotes breast cancer metastasis by dissociating the PAR complex. Journal of experimental & clinical cancer research : CR. 2019;38(1):27. doi: 10.1186/s13046-019-1025-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Wickham, H and Bryan, J. (2019). Readxl: read excel files. R package version 1.3.1. https://CRAN.R-project.org/package=readxl
  50. Wickham H, Averick M, Bryan J, Chang W, D’Agostino McGowan L, François R, Grolemund G, Hayes A, Henry L, Hester J, Kuhn M, Lin Pedersen T, Miller E, Bache SM, Müller K, Ooms J, Robinson D, Seidel DP, Spinu V, Takahashi K, Vaughan D, Wilke C, Woo K, Yutani H. Welcome to the tidyverse. Journal of open source software. 2019;4(43):1686. doi: 10.21105/joss.01686. [DOI] [Google Scholar]
  51. Wöhrle FU, Daly RJ, Brummer T. Function, regulation and pathological roles of the gab/DOS docking proteins. Cell communication and signaling : CCS. 2009;7:22. doi: 10.1186/1478-811X-7-22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Wu AY, Gu LY, Cang W, Cheng MX, Wang WJ, Di W, Huang L, Qiu LH. Fn14 overcomes cisplatin resistance of high-grade serous ovarian cancer by promoting Mdm2-mediated p53-R248Q ubiquitination and degradation. Journal of experimental & clinical cancer research : CR. 2019;38(1):176. doi: 10.1186/s13046-019-1171-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Xiang S, Wang N, Hui P, Ma J. Gab3 is required for human colorectal cancer cell proliferation. Biochem Biophys Res Commun. 2017;484(4):719–725. doi: 10.1016/j.bbrc.2017.01.095. [DOI] [PubMed] [Google Scholar]
  54. Xie, Y (2020). Knitr: A general-purpose package for dynamic report generation in R. R package version 1.28
  55. Yang L, Zhang X, Ma Y, Zhao X, Li B, Wang H. Ascites promotes cell migration through the repression of miR-125b in ovarian cancer. Oncotarget. 2017;8(31):51008–51015. doi: 10.18632/oncotarget.16846. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Yoshihara K, Tajima A, Komata D, Yamamoto T, Kodama S, Fujiwara H, Suzuki M, Onishi Y, Hatae M, Sueyoshi K, Fujiwara H, Kudo Y, Inoue I, Tanaka K. Gene expression profiling of advanced-stage serous ovarian cancers distinguishes novel subclasses and implicates ZEB2 in tumor progression and prognosis. Cancer Sci. 2009;100(8):1421–1428. doi: 10.1111/j.1349-7006.2009.01204.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Zhao G, Chen J, Deng Y, Gao F, Zhu J, Feng Z, Lv X, Zhao Z. Identification of NDRG1-regulated genes associated with invasive potential in cervical and ovarian cancer cells. Biochem Biophys Res Commun. 2011;408(1):154–159. doi: 10.1016/j.bbrc.2011.03.140. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ESM 1 (79.7KB, pdf)

(PDF 79 kb)

ESM 2 (5.3MB, pdf)

(PDF 5411 kb)

ESM 3 (33KB, png)

(PNG 33 kb)

ESM 4 (6KB, pdf)

(PDF 5 kb)

ESM 5 (7KB, pdf)

(PDF 7 kb)

ESM 6 (1.2MB, pdf)

(PDF 1223 kb)

ESM 7 (56KB, pdf)

(PDF 56 kb)

Data Availability Statement

Transcriptomics datasets used in the present study are available from Gene Expression Omnibus (GEO) and TCGA-OV (The Cancer Genome Atlas – Ovarian Cancer), and their accession IDs were given in the text. Most expression data can be accessed in R programming environment using curatedOvarianData Bioconductor data package. Protein datasets used in the study and information on how to access them were given in Materials and Methods section.

Articles from Journal of Cell Communication and Signaling are provided here courtesy of The International CCN Society

RESOURCES