Discussion
In the present study, we reported an X-linked myopia case and identified the variant associated with HM in a Southern Chinese family. A heterozygous/hemizygous novel nonsense ARR4 mutation, c.569C>G (p.S190*), was found to be associated with HM and co-segregated within affected family members. Furthermore, it was found that the unstable mutant ARR4 protein induced an ER stress response and was rapidly degraded via the proteasome degradation pathway. Combining with previous studies, we found that ARR4 may mediate a disease with pattern of inheritance which is contrary to the conventional genetics.
ES has been shown to be remarkable at detecting variants associated with monogenic disorders. Notably, compared to 21-25% diagnostic rate in proband-only ES, a diagnostic rate of 30-37% was observed among affected individuals when ES was applied to testing trios(Farwell et al., 2015; Valencia et al., 2015). Consequently, trio-based ES analysis was used in this study involving a HM family. Following bioinformatic analysis, the ARR4 c.569C>G (p.S190*) mutation was identified as the explanation of the proband’s phenotype as evidenced by not being present in public databases and a high amino acid conservation. The ARR4 gene, located on chromosome Xq13.1, has been linked to abundant expression in retinal tissue. To the best of our knowledge, because only three HM families have been identified harboring mutations in ARR4 worldwide(Xiao et al., 2016), it is easy to omit pathogenic ARR4 variants due to the lack of clinical associations to this gene. EXOMISER, an exome analysis toolkit, provided an effective way for prioritizing patients’ candidate genes based on modified phenotypic databases associated with zebrafish models(Smedley et al., 2015).
As shown in Figure 2B, of the three previously described pathogenicARR4 mutations, two were missense mutations, while the remaining variant was a nonsense mutation. When including c.569C>G (p.S190*), domain predictions indicated that all four mutations may affect the arrestin C-/N-terminal domains (Figure 2B). Previous studies have demonstrated that both the C- and N-terminal domains play a critical role in binding to the receptor molecule(Gurevich & Benovic, 1993; Hirsch, Schubert, Gurevich, & Sigler, 1999). Alignment of multiple ARR4 amino acid sequences have shown that the serine at codon 190 is highly conserved (Figure 2C).
Arrestin is an important family of proteins that can desensitize G-protein coupled receptors (GPCRs). Rhodopsin belongs to a class of GPCRs that can sense external light signals and transmit these to cells that produce vision. In mammals, there are two types of visual arrestins: arrestin-1 (ARR1 , also called S-antigen[SAG ] or 48K protein)(Kuhn, Hall, & Wilden, 1984; Pfister et al., 1985; Vishnivetskiy et al., 2018; Wacker, Donoso, Kalsow, Yankeelov, & Organisciak, 1977), and arrestin-4 (ARR4 , also called cone arrestin [CAR ], or X-arrestin [ARRX ])(Craft, Whitmore, & Wiechmann, 1994; Murakami, Yajima, Sakuma, McLaren, & Inana, 1993). After being activated by light, opsins are phosphorylated by rhodopsin, and arrestin-1 and arrestin-4 can bind to phosphorylated opsins to inhibit the transduction of downstream signals(Deming et al., 2015). In murine models, these two vision-inhibiting proteins are required for normal inactivation of rhodopsin in cone cells(Nikonov et al., 2008). Misfolded proteins can be recognized on the endoplasmic reticulum and transported to the proteasome for degradation(Griciuc, Aron, & Ueffing, 2011). With elevated expression of ER-stress-related markers that can be observed, the mutant ARR4 protein was found to mediate an ER stress. Consistent with this finding, it was shown that the mutant proteins were degraded via the proteasome pathway due to the proteasome inhibition effects observed on mutant APRE19 cells.
The pathogenicity of the ARR4 gene was predicted and its association with HM in female patients were described for the first time in 2016. Consequently, an X-linked, female-limited pattern of inheritance was considered for ARR4 -associated HM cases; a statement strengthened by the observation that none of the hemizygous male family members were diagnosed with HM(Xiao et al., 2016). However, in our study it was demonstrated that the proband’s father was affected with HM and possessed pathogenic ARR4 mutation. This suggested that the hereditary pattern of the ARR4 gene may not be X-linked, female-limited. HM is a complex, heterogeneous disease which is influenced by environmental and genetic factors. Occurring in females, an X chromosome is compacted during X-chromosome inactivation, which results in the random silencing of one of the X chromosomes(Galupa & Heard, 2018). X-chromosome inactivation could however not explain the phenotypes observed in this study since a nonsense mutation was identified in an affected male hemizygote, while unaffected male hemizygotes carried of a missense mutation in other families. So we speculate that the mechanism of loss of function caused by null variants may be more detrimental for the ARR4 gene. Both Arr1 andArr4 genes are expressed in mouse cone cells. If only one of the two genes is knocked out, the recovery time will slow down after being stimulated by saturated strong light, but knockout of both genes will greatly increase the delay of recovery time, which indicates that there may be functional compensation between Arr1 and Arr4 in mice(Nikonov et al., 2008). So it becomes reasonable to question that if there is also a complementary mechanism in human retina, which leads to the occurrence and absence of diseases in different hemizygous individuals? Further proof is needed in the future. In Craniofrontonasal syndrome, the female heterozygote of EFNB1 gene located on X chromosome is usually patient, but the non-mosaic hemizygous male is usually not affected or has only mild symptoms(Niethamer et al., 2020). Some studies believe that this is a kind of ”cellular interaction” caused by the mosaic state of EPHRIN-B1 protein, and the mechanism is not clear(Twigg et al., 2004; Wieacker & Wieland, 2005; Wieland et al., 2004). Polydactyly was found in female heterozygous mice (Efnb1 +/-), but not in hemizygous males (Efnb1 -) and homozygous females (Efnb1 -/-). Meanwhile, six sporadic male patients with the mosaic state were found to be severely affected. Interestingly, different proportion of mosaic mutations of EFNB1 can be detected in various tissues of them including peripheral blood(Twigg et al., 2013). However, we reviewed the exome sequencing results of high-myopia male patients in this study, and found that the sequencing depth of c.569C>G (p.S190*) site was 183X, and all of them were pathogenic variant G bases, which seemed to be inconsistent with the conclusion in the above study. However, the phenotype of high myopia caused by ARR4 mutation is limited at the present. So, it’s unknown that if ARR4 mutation is also affected by ”cellular interference” like EFNB1 , resulting in both normal and affected hemizygous males? This is the goal of our research in the future.