Functional study on new FOXL2 mutations found in Chinese patients with blepharophimosis, ptosis, epicanthus inversus syndrome

Blepharophimosis, ptosis, epicanthus inversus syndrome (BPES, OMIM # 110100) is a rare genetic disorder with an estimated incidence of 1 in 50,000 births [1]. It can occur sporadically or associate with autosomal dominant mutations. The characteristic clinical presentations of this disease include a complex eyelid/ocular malformation characterized by blepharophimosis, ptosis, epicanthus inversus and telecanthus. The horizontal shortening of the palpebral aperture can lead to amblyopia in one or both eyes [2]. Depending on the occurrence of premature ovarian failure (POF) or not, female BPES patients are classified into two different groups, with type I patients having POF while type II referring to those with normal ovarian function [3]. However, both types of BPES are widely recognized to result predominantly from mutations in the forkhead transcriptional factor-2 (FOXL2) gene that is involved in palpebral and ovarian development [4].

The FOXL2 gene in human is approximately 2.7-kb long located on chromosome 3q22.3, which encodes a protein with 376 residues. The FOXL2 protein contains a characteristic 100 amino-acid DNA-binding forkhead domain, which categorizes the FOXL2 protein into the superfamily of Forkhead transcription factors [5]. An alanine-rich domain, also known as a polyalanine (poly-Ala) tract highly conserved across various mammals, lies downstream of the forkhead domain and is responsible for negatively regulating its transcriptional activity [5]. FOXL2 localizes in the nucleus and transcriptionally modulates genetic programs required for early eyelid development and ovary differentiation and maintenance, at the same time, represses components essential for somatic testis determination [6]. FOXL2 acts as a transcriptional repressor for multiple genes, including the human steroidogenic acute regulatory (StAR) gene. StAR mediates the transports of cholesterol across mitochondrial membrane, and controls the rate-limiting step in steroidogenesis [7]. Mutations in FOXL2 may contribute to de-repression of the StAR promoter and cause increased differentiation of granulosa cells, leading to the onset of POF [8].

A collection of more than 100 genetic alternations affecting the FOXL2 locus have been identified in patients with BPES, including frameshifts, insertions, nonsense as well as missense mutations [9, 10], with intragenic mutations accounting for the majority (71%) [11]. Although there are some claims about the correlations between different mutations in FOXL2 and BPES types, direct genotype-phenotype association remains to be further demonstrated because of the lack of de novo genetic study using animal model and the clinical heterogeneity among patients with BPES [12]. Furthermore, the mechanisms underlying individual mutations causing the development of BPES are largely unknown and could be complicated. It is likely that some mutant FOXL2 might collaborate with other cellular alternations to promote disease progression, as exemplified by the coexistence of the FOXL2 deletion and BMP15 in some BPES patients [13]. Therefore, identification of novel FOXL2 mutations and further characterization of their contributions to the pathogenesis of BPES will provide not only biomarkers for early detection of BPES but also potential implications for therapeutic intervention.

Here, we report a novel FOXL2 mutation identified in a Chinese family with BPES. Functional studies of this missense mutation (c.931C > T, p.H311Y) revealed a reduction in both FOXL2 protein expression and its transcriptional repression activity on the promoter of StAR gene, underscoring the significance of this mutation in the pathogenesis of BPES.

Genetic alternation in FOXL2 locus has been long appreciated as an important causal factor for the pathogenesis of BPES. The highly conserved FOXL2 protein is a transcription regulator containing a DNA-binding forkhead domain and a poly-Ala tract, both of which host the vast majority of BPES-associated mutations identified to date [9]. The novel mutation c.931C > T causes a single amino acid substitution at the residue 311, converting a histidine residue into a tyrosine (p.H311Y). Although located outside the forkhead domain and poly-Ala tract, the histidine 311 residue of FOXL2 protein is highly conserved across species (Fig. 1c). Given that a histidine residue is nucleophilic and usually serves a role in stabilizing the folded structures of proteins while a tyrosine residue is hydrophobic and can make a protein exceedingly unstable when ionized, the replacement of histidine by tyrosine is likely to alter the functional property of FOXL2 protein. Indeed, further tests showed that although the c.931C > T (p.H311Y) mutation does not alter mRNA levels of FOXL2; it greatly decreases its protein expressions, probably by reducing the stability of FOXL2 protein.

Both WT and mutant FOXL2 localized exclusively in the nucleus, suggesting that the p.H311Y mutation does not affect the protein’s nuclear localization signal or disturb interactions with nuclear transporters. A previous study has reported that mutations in the forkhead domain of FOXL2 are more likely to cause cytoplasmic mislocalization and aggregation of the protein [22]. The c.931C > T (p.H311Y) mutation occurs outside the forkhead and the poly-Ala domains, and does not seem to alter the subcellular localization and aggregation pattern of FOXL2. This is confirmed by another report by Beysen et al. that a missense mutation located outside the forkhead domain (p.S217F) had no effect on intracellular protein distribution [23]. Interestingly, luciferase assay of the StAR gene showed loss of FOXL2 repressor function upon p.H311Y mutation as compared to the WT FOXL2. This could be due to a decrease of total available FOXL2 in the nucleus. It is also possible that the mutation may decrease intranuclear mobility, as well as binding affinity of FOXL2 to the StAR gene. Further studies are required to corroborate these assumptions.

Given the critical roles of FOXL2 mutations in the development and progression of BPES, enormous efforts have been made in the correlations between genotypes and phenotypes. Initial studies have been focused more on the FOXL2 structural alternations and ended with a preliminary genotype-phenotype correlation, that is, mutations resulting in truncated FOXL2 proteins without the poly-Ala tract are likely associated with BPES type I, whereas poly-Ala expansions might preferentially give rise to BPES type II [12]. However, clear correlation seems difficult for the mutations contributing to mutant proteins but with an intact forkhead domain and poly-Ala tract. Therefore, the functional properties of the affected FOXL2 proteins, including their transactivation capacity, subcellular localization, aggregation as well as protein 3D structure, are taken into account when classifying pathogenic FOXL2 mutations [23‐25]. For example, using luciferase reporter systems, Dipietromaria et al. were able to demonstrate that loss-of-function FOXL2 mutants are likely BPES type I mutations [24]. Based on this theory, it is speculated that mutations outside the forkhead domain without affecting the FOXL2 transactivation are possibly associated with BEPS type II. Indeed, this seems true when analyzing all the known non-forkhead missense mutations. Including this novel one reported by our study, there are six different non-forkhead missense FOXL2 mutations: p.K193C, p.Y215C, p.S217C, p.S217F, p.Y258N and p.H311Y. Among these mutations, p.K193C was identified in a type II family [9]; p.Y215C, p.S217C and p.S217F have been experimentally validated using the dual 4xFLRE-luc and SIRT1-luc reporter systems and classified as BPES type II [24]. Although the individual carrying the p.Y258N mutation displayed POF, she was not a BPES patient [26]. As for our novel p.H311Y mutation, we are not able to tell its phenotype association at this point. On one hand, the single female patient in this affected family is still too young to pathologically classify the BPES subtype; on the other hand, the experimental approach we employed in this study was completely different than those used by Dipietromaria et al. (StAR-luciferase construct rather than the dual 4xFLRE-luc/SIRT1-luc reporters) [24], making it impossible to compare the transactivation capacity of p.H311Y. Therefore, whether or not our novel p.H311Y belongs to the BEPS type II still deserves further experimental investigation and clinical follow-up.

Furthermore, the severe phenotype associated with p.H311Y mutation is somehow consistent with other published missense mutations located outside the forkhead domain. For example, a heterozygous FOXL2 missense mutation c.C650G (p.S217C) identified in an Iranian family with BPES gives rise to a striking phenotype with bilateral amblyopia [27]. Interestingly, the same mutation (p.S217C) has also been reported in an Indian family with mild eyelid phenotype [22], implying that additional genetic and/or epigenetic variations between these two p.S217C families might underlie the discrepancy in disease presentation. Despite that FOXL2 located on 3q23 is the only gene known to cause BPES, considering the previous reports that the whole segment of 3q21–24 contributes to BPES [28], and that FOXL2 deletion coexists with BMP15 variants in a BPES patient [13], it is reasonable to speculate that other genetic and/or epigenetic factors synergize mutant FOXL2, especially those with intact transactivation capacity, to cause and/or exacerbate BPES. Further investigation on this aspect will be of interest, as it might reveal some novel components that are therapeutically targetable.

In summary, we not only identified a novel mutation, c.931C > T (p.H311Y), in the FOXL2 gene in a Chinese family with BPES, but also confirmed that this missense mutation causes a reduction in both the expression and the activity of FOXL2 protein. The novel mutation reported here further expands the mutation spectrum of the FOXL2 gene and contributes to the understanding of the molecular pathogenesis of BPES.