How Rap and its GEFs control liver physiology and cancer development. C3G alterations in human hepatocarcinoma

Abstract

Rap proteins regulate liver physiopathology. For example, Rap2B promotes hepatocarcinoma (HCC) growth, while Rap1 might play a dual role. The RapGEF, Epac1, activates Rap upon cAMP binding, regulating metabolism, survival, and liver regeneration. A liver specific Epac2 isoform lacking cAMP-binding domain also activates Rap1, promoting fibrosis in alcoholic liver disease. C3G (RapGEF1) is also present in the liver, but mainly as shorter isoforms. Its function in the liver remains unknown. Information from different public genetic databases revealed that C3G mRNA levels increase in HCC, although they decrease in metastatic stages. In addition, several mutations in RapGEF1 gene are present, associated with a reduced patient survival. Based on this, C3G might represent a new HCC diagnostic and prognostic marker, and a therapeutic target.

Keywords: C3G; Epac; Rap; hepatocarcinoma; liver; liver diseases.

Figure 1.. Schematic representation of Rap1 activation by GEFs.

Rap switches from an inactive form, bound to GDP, to an active conformation, bound to GTP. GEFs, such as C3G or Epac, mediate GDP dissociation, favoring GTP binding to Rap. To inactivate it, GAPs activate Rap intrinsic GTPase activity, promoting GTP hydrolysis.

Figure 2.. Tissue-specific pattern of RapGEF1 (C3G) mRNA expression.

(A) Histogram showing RapGEF1 mRNA expression in fetal liver, adult liver, lung, colon, kidney, pancreas, prostate, ovary, whole brain and whole blood, expressed as the average normalized expression. The dotted line marks the median. These data were extracted from BioGPS platform using GeneAtlas U133A datasets composed of 176 samples of different human tissues. (B) Histogram showing RapGEF1 mRNA expression in liver, lung, colon, breast, kidney, ovary, pancreas, prostate, uterus and whole blood, expressed as the median TPM. These data were extracted from GTExportal (GTEx Analysis Release V6P), with a dataset composed of 175 liver samples. TPM: Transcript per kilobase million.

Figure 3.. Comparison of RapGEF1 mRNA expression in normal versus tumoral liver tissue.

(A) Boxplot showing RapGEF1 mRNA expression in normal and tumoral liver tissue, expressed as log2 of RSEM (accurate transcript quantification for RNA-Seq data). These data were extracted from GENT platform, derived from the analysis of 50 normal liver samples and 194 liver tumor samples using the Affymetrix GeneChip Human Genome U133 Plus 2.0 Array. (B) Boxplots showing RapGEF1 expression in normal and tumoral liver tissues, expressed as log2 RSEM (right panel) or log2 RPKM (left panel). These data come from Firebrowse platform, using TCGA datasets, with 50 control liver samples and 373 hepatocellular carcinoma (HCC) tumor samples (right panel) or nine control liver samples and 17 HCC tumor samples (left panel). A t-test statistical analysis using Graphpad 6.01 software revealed a significant difference between tumor samples and controls (***p < 0.0001; left panel). (C) Boxplot showing RapGEF1 mRNA expression in control (normal) liver, HCC, HCC-PDX and HCC metastatic samples, expressed as log2 RSEM. These data were obtained from Gene Investigator software using Affymetrix Human Genome U133 Plus 2.0 Array analysis, with a dataset of 535 control liver samples, 232 HCC samples, 30 HCC-PDX samples and 15 HCC-metastatic samples. GENT: Gene expression across normal and tumoral tissue; PDX: Patient-derived xenograft; RPKM: Reads per kilobase per million; RSEM: Accurate transcript quantification from RNA-Seq data with or without a reference genome.

Figure 4.. Somatic mutations and other genetic alteration in RapGEF1 gene.

Data from cBioportal database, using three gene sets: TCGA with 366 cases, the AMC (Hepatology, 2014) with 321 cases and Inserm (Nat. Gent., 2015) with 243 cases. (A) Piled histogram showing RapGEF1 deep deletion, amplification, missense mutation, or truncating mutation, expressed as the percentage of alterations in the gene set. (B) Schematic representation of C3G domains, from N- to C-terminal: negative regulatory domain of GEF activity, E-cadherin binding domain, SH3-binding domain (rich in prolines), Y504 (site of phosphorylation) and GEF catalytic domain composed of REM and Cdc25-Homologous Ras GEF domain. Dots above C3G scheme represent nonsense mutations (dark dot) or missense mutations (grey dot) with annotations at the site of the involved aa. aa limiting each domain are displayed. aa: Amino acid; AMC: Asian Medical Centre; fs: Frame shift insertion; REM: Ras exchange motif.

Figure 5.. Survival of patients with alterations in RapGEF1 gene as compared with patients with no alterations.

Kaplan–Meier survival curves for both patients without alterations in RapGEF1 (blue curve, 354 cases) and with alteration(s) in RapGEF1 (red curve, 11 cases), expressed as the percentage of alive patients along the time (expressed in months). These data and the graphic (with modifications) were extracted from cBioportal platform using TCGA database.

Figure 6.. Functions of Rap proteins and its GEFs in hepatocellular carcinoma or other liver pathologies.

This scheme summarizes the most relevant actions of Rap1/2 and their main GEFs, Epac and C3G, in different liver pathological contexts. Upper left panel shows both the anti- and pro-tumorigenic effect of Rap1, as well as that of Rap2B promoting tumor growth and migration in human HCC cell lines. The potential implication of Ras/ERKs cascade and FAK is also indicated. Upper right panel reflects the positive or negative correlation between C3G levels in human patients’ samples and HCC initiation or progression, respectively. Rap2 could mediate C3G pro-tumorigenic effect, although other mechanisms are possible. The relationship between C3G gene mutations and a poor prognosis is also included. Lower panel shows the antagonist contribution of Epac1 and Epac2 proteins in ALF and other effects of Epac1/Rap1 in the partial hepatectomy response. ALF: Alcoholic liver fibrosis; HCC: Hepatocellular carcinoma.