Analysis of variants in GATA4 and FOG2/ZFPM2 demonstrates benign contribution to 46,XY disorders of sex development

Abstract

Background: GATA-binding protein 4 (GATA4) and Friend of GATA 2 protein (FOG2, also known as ZFPM2) form a heterodimer complex that has been shown to influence transcription of genes in a number of developmental systems. Recent evidence has also shown these genes play a role in gonadal sexual differentiation in humans. Previously we identified four variants in GATA4 and an unexpectedly large number of variants in ZFPM2 in a cohort of individuals with 46,XY Differences/Disorders of Sex Development (DSD) (Eggers et al, Genome Biology, 2016; 17: 243).

Method: Here, we review variant curation and test the functional activity of GATA4 and ZFPM2 variants. We assess variant transcriptional activity on gonadal specific promoters (Sox9 and AMH) and variant protein-protein interactions.

Results: Our findings support that the majority of GATA4 and ZFPM2 variants we identified are benign in their contribution to 46,XY DSD. Indeed, only one variant, in the conserved N-terminal zinc finger of GATA4, was considered pathogenic, with functional analysis confirming differences in its ability to regulate Sox9 and AMH and in protein interaction with ZFPM2.

Conclusions: Our study helps define the genetic factors contributing to 46,XY DSD and suggests that the majority of variants we identified in GATA4 and ZFPM2/FOG2 are not causative.

Keywords: FOG2; GATA4; ZFPM2; disorders of sexual development; functional analysis; mutations.

Figure 1

Protein schematic of GATA4 and ZFPM2/FOG2 showing coding variants identified in 46,XY individuals. (a) The human GATA4 protein (NP_002043) is 440 amino acids with the following functional elements: two highly conserved N‐terminal and C‐terminal zinc finger domains (N‐Zn and C‐Zn, grey filled boxes); three transactivation domains (TAD1‐3); and a nuclear localization signal (NLS, black box). Positions of four GATA4 heterozygous missense variants identified in four individuals are shown. One variant p.W228C (case1) initially identified by our study and later reported by LaPiscina et al. occurs in the N‐terminal zinc finger, along with the first variant published in association with 46,XY DSD by Lourenco et al. (used as a positive control in this study). The other three variants (p.A346V, p.P394T, p.P407Q) occur in the C‐terminal transactivation domain (TAD3) and are reported in ClinVar in association with CHD (solid black triangle). (b) The human ZFPM2 protein (NP_036214) is 1,151 amino acids long with the following functional elements: eight zinc finger domains (grey filled boxes); nuclear localization signal (NLS) (solid black box); a CTBP2 interaction domain (light grey box); residues known to undergo post‐translational modifications (grey arrow head). Position of nine ZFPM2 heterozygous missense and one in‐frame deletion variants found in ten individuals. All variants lie outside of known functional domains, except the p.M703L variant which occurs within the 7th zinc finger domain. The solid black triangles represent variants reported in ClinVar, the white triangles indicate unreported variants

Figure 2

Transactivation assays assessing GATA4, ZFPM2/FOG2 and co‐factor activity on gonadal promoter elements. (a) Transcription factors GATA4, ZFPM2 and co‐factor NR5A1 were transfected into HEK293 cells and transcriptional activity of mTesco enhancer element construct was measured with luciferase as a reporter. Wild‐type transcription factors were tested individually as well as in combination, empty vector was used as a negative control (white and black bars). Wild‐type co‐factor NR5A1 and ZFPM2 were transfected with individual GATA4 variants (dark grey bars), or wild‐type NR5A1 and GATA4 with ZFPM2 variants (light grey). Maximal transactivation of mTesco was observed with wild‐type NR5A1 alone. Wild‐type GATA4 or ZFPM2 (alone or in combination) is able repress NR5A1 activation. Most variants tested were able to maintain repression of NR5A1 activation of mTesco (compared to wild‐type NR5A1/ZFPM2/GATA4, black bar, horizontal dotted line), only GATA4 p.G221R (previously published positive control, Lourenco et al.) and p.W228C variants showed loss of NR5A1 repression. (b) Transcriptional activity of GATA4, ZFPM2 and co‐factor WT1 (−KTS) on human AMH promoter (+10–[−270]) assessed by luciferase assays. Wild‐type transcription factors were tested individually as well as in combination, empty vector was used as a negative control (white and black bars). Wild‐type co‐factor WT1 (−KTS) and ZFPM2 were transfected with individual GATA4 variants (dark grey bars), or wild‐type WT1 (−KTS) and GATA4 with ZFPM2 variants (light grey bars). Maximal transactivation of hAMH was observed with wild‐type WT1 (−KTS), ZFPM2 and GATA4 (black bar, horizontal dotted line). A similar level of transactivation compared to the wild‐type was observed for the majority of variants tested, except for GATA4 p.G221R and p.W228C variants which showed loss of activation. For all transactivation assays: Data represented as the mean and SEM of at least three independent experiments (n = 3), as a fold change relative to the empty vector control (background), each assay was run in technical triplicate. P‐values were calculated using a one‐way ANOVA multiple comparisons (Dunnett test) (compared to the wild‐type—black bar, horizontal dashed line), p‐value **** < .0001

Figure 3

Protein interaction analysis of GATA4 and ZFPM2/FOG2 complex by co‐immunoprecipitation and Western blot analysis. (a) Detection of GATA4 variant and wild‐type ZFPM2 protein complex. Co‐immunoprecipitation was performed by transiently overexpressing wild‐type ZFPM2 protein with GATA4 wild‐type or variant protein in HEK293 cells. Pull down of the GATA4/ZFPM2 complex was performed using the FLAG‐tag of the ZFPM2 protein construct. The complex was then detected by Western blot analysis using the GATA4 protein antibody. The protein input was verified by Western blotting using ZFPM2 antibody. An interaction was detected for wild‐type GATA4 and ZFPM2 proteins (WT). A decreased interaction was detected for the GATA4 variants p.G221R (positive control) and p.W228C. (b) Detection of ZFPM2 variant and wild‐type GATA4 protein complex. In this case ZFPM2 wild‐type or variant proteins were over‐expressed with wild‐type GATA4 protein in HEK293 cells. Pull down of the GATA4/ZFPM2 complex was detected using the HA‐tag of the GATA4 protein construct. The complex was detected by Western blot analysis using a ZFPM2 antibody. The protein input was verified by Western blotting using the GATA4 antibody. An interaction was detected for wild‐type GATA4 and ZFPM2 proteins (WT). All ZFPM2 variants showing the GATA4/ZFPM2 interaction were retained