(Iso-)form Matters: Differential Implication of Vav3 Variants in Ovarian Cancer

A genome-wide association study identifies novel gene associations with facial skin wrinkling and mole count in Latin Americans

BACKGROUND

Genome-wide association studies (GWASs) have identified genes influencing skin ageing and mole count in Europeans, but little is known about the relevance of these (or other genes) in non-Europeans.

OBJECTIVES

To conduct a GWAS for facial skin ageing and mole count in adults < 40 years old, of mixed European, Native American and African ancestry, recruited in Latin America.

METHODS

Skin ageing and mole count scores were obtained from facial photographs of over 6000 individuals. After quality control checks, three wrinkling traits and mole count were retained for genetic analyses. DNA samples were genotyped with Illumina's HumanOmniExpress chip. Association testing was performed on around 8 703 729 single-nucleotide polymorphisms (SNPs) across the autosomal genome.

RESULTS

Genome-wide significant association was observed at four genome regions: two were associated with wrinkling (in 1p13·3 and 21q21·2), one with mole count (in 1q32·3) and one with both wrinkling and mole count (in 5p13·2). Associated SNPs in 5p13·2 and in 1p13·3 are intronic within SLC45A2 and VAV3, respectively, while SNPs in 1q32·3 are near the SLC30A1 gene, and those in 21q21·2 occur in a gene desert. Analyses of SNPs in IRF4 and MC1R are consistent with a role of these genes in skin ageing.

CONCLUSIONS

We replicate the association of wrinkling with variants in SLC45A2, IRF4 and MC1R reported in Europeans. We identify VAV3 and SLC30A1 as two novel candidate genes impacting on wrinkling and mole count, respectively. We provide the first evidence that SLC45A2 influences mole count, in addition to variants in this gene affecting melanoma risk in Europeans.

Ovarian Cancer Stem Cell Heterogeneity

Ovarian carcinoma features pronounced clinical, histopathological, and molecular heterogeneity. There is good reason to believe that parts of this heterogeneity can be explained by differences in the respective cell of origin, with a self-renewing fallopian tube secretory cell being likely responsible for initiation of an overwhelming majority of high-grade serous ovarian carcinomas (i.e., type II tumors according to the recent dualistic classification), whereas there are several mutually non-exclusive possibilities for the initiation of type I tumors, including ovarian surface epithelium stem cells, endometrial cells, or even cells of extra-Müllerian origin. Interestingly, both fallopian tube self-renewing secretory cells and ovarian surface epithelium stem cells seem to be characterized by an overlapping array of stemness signaling pathways, especially Wnt/β-catenin. Apart from this variability in the respective cell of origin, the particular clinical behavior of ovarian carcinoma strongly suggests an underlying stem cell component with a crucial impact. This becomes especially evident in high-grade serous ovarian carcinomas treated with classical chemotherapy, which entails a gradual evolution of chemoresistant disease without any apparent selection of clones carrying obvious chemoresistance-associated mutations. Several cell surface markers (e.g., CD24, CD44, CD117, CD133, and ROR1) as well as functional approaches (ALDEFLUOR™ and side population assays) have been used to identify and characterize putative ovarian carcinoma stem cells. We have recently shown that side population cells exhibit marked heterogeneity on their own, which can hamper their straightforward therapeutic targeting. An alternative strategy for stemness-depleting interventions is to target the stem cell niche, i.e., the specific microanatomical structure that secures stem cell maintenance and survival through provision of a set of stem cell-promoting and differentiation-antagonizing factors. Besides identifying direct or indirect therapeutic targets, profiling of side population cells and other ovarian carcinoma stem cell subpopulations can reveal relevant prognostic markers, as exemplified by our recent discovery of the Vav3.1 transcript variant, which filters out a fraction of prognostically unfavorable ovarian carcinoma cases.

Evaluation of Vav3.1 as prognostic marker in endometrial cancer

PURPOSE

Vav3 is a guanine nucleotide exchange factor that regulates the activity of Rho/Rac family GTPases. In a study on ovarian cancer, we recently demonstrated pronounced prognostic and predictive value of Vav3.1, a specific truncation variant of the parental Vav3 gene. Here, we sought to investigate the role of Vav3.1 in the most prevalent gynecological tumor entity, endometrial cancer.

METHODS

Vav3.1 transcript levels were determined in a large cohort of endometrial cancer patients using variant-specific PCR (n = 239), and non-malignant endometrial tissue served as control (n = 26). Expression levels of Vav3.1 were stratified according to established clinicopathological characteristics and correlated to long-term patient survival (average follow-up of > 7.5 years). Type 1 and type 2 cancers were separately investigated.

RESULTS

While Vav3.1 was markedly overexpressed in endometrial cancer tissue, we could not detect associations with clinical parameters related to prognosis, such as FIGO stage and tumor grade. Kaplan-Meier estimators of different measures of survival failed to show prognostic significance of Vav3.1 in endometrial cancer. Lack of prognostic value was observed for both type 1 and type 2 cancers.

CONCLUSIONS

Our study shows that Vav3.1 is not suited as a marker of cancer progression and/or treatment response in endometrial cancer. Feasibility and potential benefit of targeting Vav3.1 in endometrial cancer needs to be evaluated in future studies, proceeding from its clear, roughly ten-fold, induction in the malignant endometrium.