The clinical features of WS are varied and include T1DM, OA, progressive sensorineural deafness, DI, autonomic nervous system dysfunction and, ultimately, brainstem atrophy and premature death. Although the medical and family histories are vital for the diagnosis of WS, genetic testing is becoming increasingly important. The majority of WS cases are caused by WFS1 mutations. Wolframin, the protein encoded by WFS1, is predicted to have a hydrophobic central domain comprising 9 membrane-spanning segments connected to a hydrophilic N-terminal domain and a hydrophilic C-terminal tail .(2) Wolframin is a component of the ER, which carries out post-translational modification, folding and assembly of newly synthesized proteins such as insulin. WFS1 may play a role in the negative regulation of ER stress signaling, thereby inhibiting the apoptosis of cells (including pancreatic beta cells) in response to ER stress.(9)
Wolframin is abundantly expressed in the pancreas, brain, heart and muscle. N-glycosylation of wolframin is thought to affect its biogenesis and stability, and there are five predicted Asn-glycosylation sites at amino acid positions 28, 335, 500, 661 and 746 .(3) Hofmann et al. reported that the R629W mutation causes instability and rapid degradation of wolframin .(2) The proband in the present case report carried a compound heterozygous W666X and K705Ifs*7 mutation (Fig. 4), the latter of which is a frameshift mutation that completely changes the N-glycosylation site. Mutation analysis has indicated that individual amino acids in specific regions are critical for correct protein folding, with substitutions at these loci inducing major functional changes .(10) Thus, the amino acid mutations identified in the proband in this study may be critical for correct folding of wolframin.
Most mutations underlying WS are located in exon 8 of WFS1, which encodes the transmembrane region and C-terminal tail of wolframin. However, the same mutations in exon 8 can have heterogeneous clinical presentations. The patient in our study had a compound heterozygous mutation (c.2113_2114insT and c.1997G > A) and presented with three of the four components of DIDMOAD (DI, DM and OA) but not hearing impairment, which was reported previously for a patient with the c.1997G > A missense mutation.(11) Phenotype variation has also been described for another mutation (c.1346C > T, p.T4491): one affected patient showed urethral involvement and severe anorexia while her younger sister exhibited microalbuminuria without neurological or urinary tract involvement.(12) The c.1997G > A mutation identified in the proband described here causes premature termination at codon 666 and an incomplete hydrophilic C-terminal tail. Most mutations associated with deafness occur in exon 8, which contains the conserved C-terminal domain that seems vital to cochlea function.(13) The C-terminal may interact with other proteins, and its deletion may disrupt these interactions and protein function.(14) The patient in our study also had an nucleobase insertion at codon 705 that resulted in frameshift and rearrangement. This insertion mutation may be in a region that impacts on the interaction with other proteins, causing the phenotypic difference with regard to hearing impairment. However, it should be noted that WS in our patient could also have been due to an allelic variant in the promoter regulatory region in the intronic sequences of WFS1 or a mutation in a different gene that was not detected with the current technology.
The proband in this case report developed T1DM at the age of 2 years, DI in the first decade, OA, neurogenic bladder and urinary tract infections in the second decade, and neurological abnormalities later in life. Although clear genotype-phenotype correlations have yet to be characterized, it has been suggested that missense mutations have a relatively mild phenotype while inactivating mutations (deletions, insertions, nonsense mutations and splice site mutations) cause more severe disease .(15) This is supported by another study reporting that patients with WS caused by missense mutations in WFS1 had a mild phenotype.(16) Consistent with the above hypothesis, the patient in our study had a nonsense mutation and insertion in WFS1 and exhibited a severe phenotype, including the onset of DM at only 2 years of age. Inactivation of both WFS1 alleles is known to be associated with early-onset DM.(17) Interestingly, a study of juvenile-onset DM in Lebanon identified a particular WFS1 mutation encoding a protein with an extended C-terminal domain that resulted in a delayed onset or absence of extrapancreatic features.(18) In vitro and in vivo animal experiments have found that wolframin depletion causes a reduction in insulin content, impairment of glucose-stimulated insulin secretion and activation of apoptosis.(19) Interestingly, single nucleotide polymorphisms in WFS1 are strongly associated with DM risk in the general population.(20) Since the proband’s mother had a family history of T2DM, it will be interesting to determine whether the K705Ifs*7 mutation is associated with an increased risk of T2DM. Further research is needed to establish the association between WFS1 genotype and phenotype.
In summary, we have described a patient with WS likely caused by a novel compound heterozygous mutation in WFS1 (c.2113_2114insT and c.1997G > A). The patient developed DM at a very young age followed by DI and OA in the next decade and neurological abnormalities later in life. His severe phenotype may be related to the presence of inactivating (nonsense and frameshift) mutations in WFS1. Since genetic mutations tend to be population-specific, the identification of this novel mutation may help in future screening for carriers of WS.