Novel mutations and the ophthalmologic characters in Chinese patients with Wolfram Syndrome

Wolfram syndrome (WFS; MIM #222300), first described in 1938 by Wolfram and Wagener, is a rare hereditary autosomal recessive disease. The prevalence of WFS was estimated to be 1 in 770,000 in UK [1] and 1 in 710,000 in the Japanese population [2]. As a progressive neurodegenerative disorder, WFS has a wide spectrum of clinical manifestations. The main phenotypes of WFS are diabetes insipidus (DI), diabetes mellitus (DM), optic atrophy (OA) and deafness (D) [1, 3]. Around 50% patients harbor all these manifestations, so WFS was also referred to as the acronym DIDMOAD syndrome [1, 3, 4]. Other common manifestations include neurologic and psychiatric disorders, renal tract abnormalities, endocrine disorders, as well as many others [5]. The major diagnosis criterion of WFS is the coincidence of early onset type 1 DM and bilateral OA before the second decade [1, 5‐7]. Because of the multi-system neurodegeneration, the prognosis of WFS is very poor and the patients’ median life expectancy is about 30 years (range 25–49 years) [1].

WFS1 on chromosome 4 is the causative gene of Wolfram Syndrome type 1 (WFS1) [8], and the loss-of-function mutations of WFS1 have been identified in most of patients with WFS [5, 9]. WFS1 encodes wolframin, an endoplasmic reticulum (ER) transmembrane protein [10]. Wolframin is widely expressed in neurons, pancreas, heart, muscle, liver, spleen and kidney [11]. It has also been detected in optic nerve glial cells and retinal ganglion cells [12, 13]. The main function of wolframin are reducing the ER stress, maintaining the Ca²⁺ homeostasis and regulating the biosynthesis and secretion of insulin [14‐16].

In addition, mutations of CISD2 are responsible for Wolfram Syndrome type 2 (WFS2; MIM #604928), which has variant features including gastrointestinal ulceration and bleeding tendency without diabetes insipidus [17‐19]. CISD2, CDGSH iron-sulfur domain-containing protein 2, located on chromosome 4q22–24, encodes endoplasmic reticulum intermembrane small protein (ERISP) [19]. Although the biological functions of CISD2 still remain incompletely defined, some studies show that it has a similar role with WFS1 in maintaining the homeostasis of Ca²⁺ and ER and the cross-talk between ER and mitochondria [20, 21].

In this study, we performed a clinical and genetic investigation on 4 unrelated Chinese patients with WFS. We systematically reviewed their clinical ophthalmologic features and identified 3 novel mutations in WFS1 and CISD2 gene. And we reported the first Chinese patient with WFS2 carried a homozygous mutation in CISD2.

In this study, we evaluated four Chinese WFS patients and descried their ophthalmologic characteristics, as well as reported three novel WFS1 and CISD2 mutations. Most patients presented at least three clinical manifestations and developed at least one in their first decade, which was consistent with the systematic review of WFS [5]. A wide range of ophthalmological findings were detected including severe vision acuity lost, declined color vision, constriction of visual fields and abnormal VEP, which were consistent with previous studies [31‐33]. Notably, the presenting ages of impaired vision of some patients were early than OA diagnosis age, which suggested the insidiousness of vision loss in WFS. This indicates that ophthalmologist should be aware of the possibility of WFS in young patients with severe bilateral optic atrophy. Detailed medical history inquiry and appropriate genetic testing are highly recommended for these patients.

There are two genes, WFS1 and CISD2, were proven to cause WFS. CISD2 is a rare causative gene and autosomal-recessive mutations in CISD2 is the pathogeny of WFS2. So far, very limited mutations have been reported in this gene (Table 3) [17‐20]. In our study, patient 1 was homozygous for the frame-shift mutation c.272_273del in CISD2, due to the parental consanguinity. This mutation was not detected in our Chinese control population. Patient 1 presented the most severe phenotype with rapid progression of disease and multisystem manifestations. The mutant CISD2 protein exerts a deleterious influence on ER-mitochondrial structure and function and ultimately participate in multisystem neurodegeneration [20]. WFS2 firstly was regarded as a subtype which has various unique features such as peptic ulcer and bleeding tendency [17‐19]. In contrast, our patient presented classical features of WFS1, including early-onset DM, progressive OA, DI and neurodegenerative features. Haematological abnormalities and peptic ulcer has not been detected so far. Our study may support the point of view that WFS1 and WFS2, caused by different genes, has a continuous clinical spectrum [20]. Since this patient was still young, with the progression of WFS2, he may develop other signs of WFS2 in the future, so long term follow-up is needed.

Mutations in WFS1 gene are responsible for most WFS patients. Since the discovery of WFS1 in 1998, more than 300 different mutations have been identified in this gene [34] and majority of them located in the exon 8 encoding the nine transmembrane segments and the C-terminal tail of wolframin [33]. In this study, we found four missense mutations located in exon 8 of WFS1, two of them were first reported including c.1618 T > G (p.Trp540Gly) and c.1048 T > A (p.Phe350Ile). The Sanger Sequence results in control population showed that these mutations are less likely to be polymorphisms. These two novel missense mutations are located in transmembrane domain. Multiple sequence alignment showed that they were positioned within evolutionary conserved regions of wolframin. And they were predicted to be deleterious by different tools (Table 2). Notably, the mutation c.2020G > A was found in two unrelated patients in our study. This mutation was previously reported in 4 patients with DM and OA without DI and deafness [26, 27]. The allele frequency of A is < 0.0001 in Han Chinese by the 1000 Genomes Project [35]. Our result indicates that this mutation is probably a hotspot in Chinese WFS patients, which needs to be verified by more cases. Only one heterozygous mutation (c.937C > T, p.His313Tyr) was found in patient 4, which was previously detected in three patients with OA, very early DM diagnosis and profound hearing loss [28‐30]. Coincidentally, patient 4 was diagnosed hearing loss much earlier than OA, which might provide an evidence that this mutation cause more hearing impairment than visual disability.

Our study showed a group of Chinese patients with WFS who had various clinical features. Genetic analysis detected three novel mutations in WFS1 and CISD2. This is the first report of Chinese patient with WFS2. Our study also illustrates the complexity and heterogeneity of WFS. So genetic testing is recommended for clinical optic nerve atrophy patients with highly suspected WFS, especially when diabetes mellitus is concomitant.