Wolfram Syndrome 1

A number sign (#) is used with this entry because Wolfram syndrome-1 (WFS1) is caused by homozygous or compound heterozygous mutation in the gene encoding wolframin (WFS1; 606201) on chromosome 4p16.

Description

Wolfram syndrome-1 is a rare and severe autosomal recessive neurodegenerative disease characterized by diabetes mellitus, optic atrophy, diabetes insipidus, and deafness (DIDMOAD). Additional clinical features may include renal abnormalities, ataxia, dementia or mental retardation, and diverse psychiatric illnesses. The minimal diagnostic criteria for Wolfram syndrome are optic atrophy and diabetes mellitus of juvenile onset. Hearing impairment in Wolfram syndrome is typically progressive and mainly affects the higher frequencies, but a small fraction of affected individuals have congenital deafness (summary by Rendtorff et al., 2011).

Autosomal dominant mutations in the WFS1 gene have been found to cause low-frequency nonsyndromic deafness (600965) as well as a Wolfram syndrome-like phenotype (614296) in which affected individuals have hearing impairment with diabetes mellitus and/or optic atrophy.

Genetic Heterogeneity of Wolfram Syndrome

Wolfram syndrome-2 (WFS2; 604928) is caused by mutation in the CISD2 gene (611507) on chromosome 4q24.

Clinical Features

Wolfram syndrome is sometimes referred to as DIDMOAD (diabetes insipidus, diabetes mellitus, optic atrophy, and deafness). Wolfram and Wagener (1938) found juvenile diabetes mellitus and optic atrophy in 4 of 8 sibs. Tyrer (1943) observed 3 of 8 sibs affected as well as 3 affected out of 4 offspring of a first-cousin marriage. Rose et al. (1966) reviewed the literature and described several cases including 2 unrelated patients, each the son of a consanguineous mating. They suggested that homozygosity for a gene with pleiotropic effects may be involved and that because of clinical heterogeneity more than one locus may be involved. All 7 patients described by Rose et al. (1966) were male. Affected females were described by others, e.g., Wolfram and Wagener (1938) and Tyrer (1943). Rorsman and Soderstrom (1967) described a family in which 3 sisters and a brother developed diabetes mellitus and optic atrophy in their teens. In one the optic atrophy appeared before the diabetes mellitus. Raiti et al. (1963) reported 2 sisters with both diabetes mellitus and diabetes insipidus. Diabetes mellitus developed at ages 9 and 5 years. Autosomal recessive inheritance was suggested. Histiocytosis X is an 'acquired' cause of double diabetes. Nevin (1974) reported a sibship of 10 of whom 2 girls, aged 14 and 11 years, had juvenile diabetes mellitus and optic atrophy. The younger girl also had diabetes insipidus.

Shaw and Duncan (1958) described 2 sisters and a niece with optic atrophy, nerve deafness, and diabetes mellitus. All 3 features had their onset in the first year of life. Ikkos et al. (1970) described first-cousin parents. Page et al. (1976) described 2 families; in both, the parents were first cousins, and 1 family had 4 affected sibs. Friedman et al. (1986) reported the birth of a healthy child from an affected woman.

Salih and Tuvemo (1991) described 2 Sudanese families with 2 affected boys in one and an affected boy and girl in the other. Diabetes mellitus was the first manifestation (at 3 to 8 years), followed by deafness and visual failure. The disease ended fatally in 1 patient at the age of 20 years. In the other 3, diabetes insipidus was confirmed using water deprivation tests for 8 hours. All 3 had severe bilateral hydronephrosis with dilated ureters and distended bladder without vesicoureteral reflux.

Wit et al. (1986) documented vasopressin deficiency in a child with Wolfram syndrome, thus confirming the central origin of the diabetes insipidus in this disorder.

The highly variable clinical picture of the Wolfram syndrome may include neurologic abnormalities such as nystagmus, mental retardation, and seizures. Rando et al. (1992) presented the cases of 2 unrelated patients who in addition to the 4 cardinal features had several other neurologic abnormalities and in whom MRIs showed widespread atrophic changes throughout the brain.

Only insulin-dependent diabetes mellitus and bilateral progressive optic atrophy are necessary to make the diagnosis of DIDMOAD. Both may present in childhood, adolescence, or early adult life; typically, but not invariably, diabetes mellitus is detected first. Diverse neurologic symptoms in Wolfram syndrome homozygotes include hearing loss, urinary tract atony, ataxia, peripheral neuropathy, mental retardation, dementia, and psychiatric illnesses (Swift et al. (1990, 1991)). Swift et al. (1990) found that 60% of a series of 68 Wolfram syndrome homozygotes had episodes of severe depression, psychosis, or organic brain syndrome, as well as compulsive verbal and physical aggression. Heterozygous carriers of the Wolfram syndrome, estimated by Swift et al. (1991) to represent approximately 1% of the United States population, are thought to be predisposed to psychiatric illness. Swift et al. (1991) estimated that the risk that a Wolfram syndrome heterozygote will be hospitalized for psychiatric illness or will commit suicide is approximately 8 times that of a noncarrier.

Psychiatric illness appears to occur in most cases of Wolfram syndrome (Strom et al., 1998). The psychiatric manifestations are particularly diverse, and a predisposition for psychiatric disorders was proposed for heterozygous carriers by Swift et al. (1998).

Scolding et al. (1996) described 2 pairs of affected sibs with the cardinal features of Wolfram syndrome in addition exhibiting neurogenic respiratory failure, startle myoclonus, Parinaud syndrome, and axial rigidity. MRI of the brain demonstrated marked brainstem atrophy.

Gabreels et al. (1998) reported a disturbance in vasopressin precursor processing in the supraoptic and paraventricular nuclei of patients with Wolfram syndrome. In patients with diabetes insipidus, the authors detected virtually no cellular immunoreactivity for processed vasopressin in the supraoptic and paraventricular nuclei. On the other hand, a considerable number of cells immunoreactive for the vasopressin precursor were present in the paraventricular nucleus. The proprotein convertase PC2 (162151) and the molecular chaperone 7B2 (173120) were also absent. As expression of PC2 and 7B2 was detected in the nearby nucleus basalis of Meynert of 1 patient with Wolfram syndrome and in the anterior lobe of the other patient with Wolfram syndrome, the authors concluded that the absence of the 2 proteins in the paraventricular nucleus was not caused by mutations in their genes. Gabreels et al. (1998) concluded that in Wolfram syndrome patients with diabetes insipidus, not only does vasopressin neuron loss occur in the supraoptic nucleus, but there is also a defect in vasopressin precursor processing.

The maternally inherited diabetes-deafness syndrome of Ballanger and Wallace (520000) has phenotypic overlap with Wolfram syndrome.

Medlej et al. (2004) reported 31 Lebanese WFS patients belonging to 17 families. Criteria for diagnosis of WFS were the presence of insulin-dependent diabetes mellitus and optic atrophy unexplained by any other disease. Central diabetes insipidus was found in 87% of the patients, and sensorineural deafness confirmed by audiograms was present in 64.5%. Other less frequent features included neurologic and psychiatric abnormalities, urodynamic abnormalities, limited joint motility, cardiovascular and gastrointestinal autonomic neuropathy, hypergonadotropic hypogonadism in males, and diabetic microvascular disease (see 603933). New features, including heart malformations and anterior pituitary dysfunction, were recognized in some of the patients and contributed to the morbidity and mortality of the disease.

Haghighi et al. (2013) reported 2 unrelated consanguineous Iranian families with severe Wolfram syndrome, defined as having neurodegenerative involvement. In 1 family, 2 living affected adult male sibs had childhood-onset diabetes mellitus, optic atrophy, and neurogenic bladder; only 1 had adult-onset hearing loss. Each patient had fathered a healthy son, indicating that they had normal fertility. The brothers were homozygous for a missense mutation in the WFS1 gene (asp211-to-asn, D211N). In the second family, affected individuals had diabetes mellitus, optic atrophy, hearing loss, neurogenic bladder, and psychiatric or cognitive problems. These individuals were homozygous for a truncating mutation in the WFS1 gene (gln486-to-ter, Q486X). Heterozygous family members in both families did not show any features of the disorder.

Phenotypic Heterogeneity

Hardy et al. (1999) and Sam et al. (2001) described Wolfram syndrome with a distinctive phenotype, namely, central respiratory failure. All of the patients were homozygous for a 4-bp deletion at position 2648-2651 in exon 8 of the WFS1 gene (606201.0012). In the patient with the 4-bp deletion reported by Hardy et al. (1999), there was severe brainstem atrophy and central respiratory failure requiring tracheostomy. Her affected sister had died at age 28 from brainstem atrophy and central respiratory failure. Five patients (from 3 families) who were heterozygous for the 4-bp deletion did not have respiratory failure. The 33-year-old patient reported by Sam et al. (2001) was diagnosed as having diabetes mellitus, a neurogenic bladder, and bilateral optic atrophy at the age of 10, 13, and 15, respectively. Audiometry was normal, and there was no evidence of diabetes insipidus. After an episode of respiratory arrest at age 32, she required intubation, ventilation, and subsequently, tracheostomy. MRI scan showed marked brainstem atrophy.

The Wolfram syndrome family (family K) unlinked to 4p that was studied by Collier et al. (1996) had 2 affected sibs in whom optic atrophy was first diagnosed at the age of 6 months and 2 years, respectively, more than a decade before the onset of diabetes mellitus. No period of normal vision was recorded, and symptoms of diabetes insipidus, renal dysfunction, and neurologic abnormalities were not present. In contrast, the affected subjects in the 11 families linked to 4p developed diabetes mellitus either at the same time or before the onset of optic atrophy, with the exception of 1 family, where optic atrophy developed 2 years earlier in 1 sib. Thus, although meeting the ascertainment criteria they had set for Wolfram syndrome, Collier et al. (1996) concluded that the phenotype in the unlinked family was atypical.

El-Shanti et al. (2000) found that 3 families linked to 4q (WFS2) contained several patients with profound upper gastrointestinal ulceration and bleeding.

Mapping

In a genomewide linkage analysis using microsatellite repeat polymorphisms in 11 families segregating for Wolfram syndrome, Polymeropoulos et al. (1994) found that a Wolfram syndrome locus (WFS1) is linked to markers on the short arm of chromosome 4, with a maximum lod score of 6.46 at theta = 0.02 for marker D4S431.

In 5 families, Inoue et al. (1998) confirmed linkage of WFS to markers on 4p. On the basis of meiotic recombinants and disease-associated haplotypes, they localized the WFS1 gene to a BAC/P1 contig of less than 250 kb. They found mutations in a novel gene designated WFS1 encoding a putative transmembrane protein in all affected individuals in 6 WFS families. WFS1 appears to function in survival of islet beta-cells and neurons, thus explaining the pleiotropic features of Wolfram syndrome.

In 12 UK families with Wolfram syndrome, Collier et al. (1996) confirmed linkage to 4p, with a maximum 2-point lod score of 4.6 with the dopamine receptor D5 gene (126453), assuming homogeneity, and of 5.1, assuming heterogeneity. Overlapping multipoint analysis using 6 markers at a time produced definite evidence for locus heterogeneity: the maximum multipoint lod score under homogeneity was less than 2, whereas when heterogeneity was allowed for an admixture, a lod of 6.2 was obtained in the interval between D4S432 and D4S431, with the peak close to the marker D4S3023. One family with an atypical phenotype was definitely unlinked to the region. Haplotype inspection of the remaining 11 families, which appeared to be linked to 4p and had typical phenotypes, revealed crossover events during meiosis, which also placed the gene in the interval D4S432 and D4S431. In these families no recombinants were detected with the marker D4S3023, which maps within the same interval. Of the 12 families studied, one had more than 1 affected sib and the twelfth had consanguineous parents.

Heterogeneity

El-Shanti et al. (2000) provided conclusive evidence of the existence of a second autosomal recessive form of Wolfram syndrome (WFS2; 604928). In 3 of 4 families, linkage to 4p16.1 was excluded and linkage to 4q22-q24 was established.

Rotig et al. (1993) found that some cases of Wolfram syndrome may have a mitochondrial basis (598500). They reported the case of a girl who presented in early infancy with insulin-dependent diabetes mellitus. She gradually developed hearing loss and optic atrophy. The progressive organ involvement and the observation of a mild hyperlactatemia pointed to a possible disorder of the mitochondrial energy supply and led Rotig et al. (1993) to identify a generalized deficiency of the respiratory chain and a 7.6-kb heteroplasmic deletion of the mtDNA. The deletion was not present in either parent.

Barrientos et al. (1996) reported studies in 2 kindreds of Spanish Caucasian origin which combined the observations of Polymeropoulos et al. (1994) and Rotig et al. (1993). The families harbored multiple deletions of mitochondrial DNA. The deletions reached percentages as high as 85 to 90% in affected tissues such as the central nervous system of 1 patient, whereas in other tissues from the same patient and from other members of the family the percentages of deleted mitochondrial DNA genomes were only 1 to 10%. In both families, Barrientos et al. (1996) demonstrated linkage to markers on 4p16; maximum multipoint lod score = 3.79 at theta = 0 (P less than 0.03). This was stated to be the first evidence of implication of both genomes in a recessive disorder (see 157640 for a description of a dominant disorder, progressive external ophthalmoplegia, caused by a mutation on 10q and multiple deletions in the mitochondrial genome). In the first family, Wolfram syndrome had been diagnosed in 4 sisters whose parents were first cousins. All the sisters first presented with insulin-dependent diabetes mellitus and dyschromatopsia followed by severe optic atrophy in their 30s. Later they developed psychiatric abnormalities in the form of anxiety, abnormal behavior, anterograde amnesia, sphincter disturbances, anosmia, walking instability, tremor, dysphagia, and swallowing difficulties. The second family was nonconsanguineous. A 27-year-old male developed IDDM at 8 years of age. When he was 16, bilateral atrophy of the optic nerves and neurosensory deafness for high frequencies were detected. Two years later, visual loss was almost complete and he developed diabetes insipidus when he was 23 years old. Barrientos et al. (1996) proposed a semidominant mode of inheritance of the predisposition to multiple mitochondrial DNA deletions: both parents of the affected individuals showed multiple mitochondrial deletions interpreted as representing the heterozygous state which in some individuals was manifested by deafness, diabetes mellitus, or psychiatric illness. See also 598500.

Now that a specific gene on 4p16.1 that is mutated in Wolfram syndrome has been identified and its characteristics initially determined, it may be possible to determine the relationship between the autosomal mutations and the multiple deletions in mitochondrial DNA observed by Rotig et al. (1993), Barrientos et al. (1996), and others.

Among 31 Lebanese WFS patients from 17 families, Medlej et al. (2004) found WFS1 gene mutations in 3 families (23.5%); no abnormalities were detected in mitochondrial DNA.

Molecular Genetics

Strom et al. (1998) identified loss-of-function mutations in both alleles of the WFS1 gene in patients with Wolfram syndrome. Homozygous mutations were found in 5 families; compound heterozygosity was found in 3 other families. In a ninth family, only a heterozygous stop mutation was found. No mutations in either allele were detected in 3 other families.

Hardy et al. (1999) performed direct DNA sequencing to screen the entire coding region of the WFS1 gene in 30 patients from 19 British kindreds with Wolfram syndrome. DNA was also screened for structural rearrangements (deletions and duplications) and point mutations in mtDNA. No pathogenic mtDNA mutations were found in this cohort. The authors identified 24 mutations in the WFS1 gene: 8 nonsense mutations, 8 missense mutations, 3 in-frame deletions, 1 in-frame insertion, and 4 frameshift mutations. Of these, 23 were novel mutations, and most occurred in exon 8. Most patients were compound heterozygotes for 2 mutations, and there was no common founder mutation. No clear-cut correlations between any of the observed mutations and disease severity were found. There were no obvious mutation hotspots or clusters.

Khanim et al. (2001) stated that mutation analysis of the WFS1 gene had identified mutations in 90% of patients with Wolfram syndrome.

Hansen et al. (2005) identified mutations in the WFS1 gene in 8 affected members of 7 Danish families with Wolfram syndrome. Four of the mutations were novel. Mutations were identified in 11 of 14 disease chromosomes; in 3 families, only 1 mutation was found.

Genotype/Phenotype Correlations

Cano et al. (2007) studied 12 patients from 11 families with Wolfram syndrome and identified 8 novel and 7 previously described mutations in the WFS1 gene. In a metaanalysis of 5 published clinical and molecular studies of WFS1 involving a total of 96 patients, they found that the presence of 2 inactivating mutations predisposed to an earlier age of onset of both diabetes mellitus and optic atrophy, as well as a more complete and earlier clinical expression of Wolfram syndrome.

Zalloua et al. (2008) performed family-based linkage analysis followed by systematic screening of WFS1 exons in Lebanese juvenile-onset insulin-dependent diabetes probands and found homozygous or compound heterozygous WFS1 mutations in 22 (5.5%) of the 399 probands, of whom 17 were diagnosed with WFS and 5 with nonsyndromic nonautoimmune diabetes mellitus. Overall, 38 probands and affected family members were homozygous or compound heterozygous for WFS1 mutations, 11 (29%) of whom had nonsyndromic DM; all of the latter patients carried a complex WFS1 mutation (606201.0024), which the authors designated WFS1(LIB) and which resulted in the delayed onset or absence of extrapancreatic features of WFS. The oldest nonsyndromic DM patients included a 20-year-old patient and two 23-year-old patients. In addition, there were 2 patients with an initial diagnosis of nonsyndromic DM that was revised to WFS when they developed optic atrophy during the course of the study; Zalloua et al. (2008) noted that longer follow-up or a specific study of adult patient populations would be needed to determine whether a subset of the WFS1(LIB) patients are exempted from extrapancreatic manifestations during their lifetime.

Chaussenot et al. (2011) performed a detailed clinical study of 59 patients with Wolfram syndrome from 49 families. The minimum ascertainment criteria for the diagnosis of WFS, diabetes mellitus and optic atrophy, was found in 56 (95%) of the 59 patients. Diabetes mellitus presented in 57 (97%) patients, at a median age of 6 years, followed by optic atrophy in 58 (98.5%) at 10 years of age. Neurologic complications occurred in 31 (53%) of the patients, at a median age of 15 years, which the authors noted was much younger than previously reported. Of the 31 patients with neurologic signs, 17 (55%) had dysfunction of the brainstem or cerebellum, including cerebellar ataxia often associated with dysarthria, dysphagia, or nystagmus; 12 (39%) had peripheral neuropathy; and 10 (32%) had cognitive impairment. Chaussenot et al. (2011) noted that diagnosis and follow-up of neurologic complications are important because death in WFS patients arises mainly by respiratory failure or dysphagia due to brainstem involvement. In the overall group, renal tract anomalies presented in 31 (53%) of the 59 patients, at a median age of 12 years, followed by deafness in 27 (46%) at 16 years of age; diabetes insipidus was seen in 17 (29%) of the patients, appearing at a median age of 15 years. Other complications included psychiatric symptoms in 23 (59%) of the patients, gastrointestinal anomalies in 7 (12%), and bilateral cataract in 3 (5%). Primary gonadal atrophy occurred in 8 of 32 male patients in the study. Sequencing of the WFS1 gene revealed 109 mutated alleles corresponding to 56 different WFS1 mutations, with 2 mutated alleles detected in 53 (90%) of 59 patients; only 1 heterozygous mutation was found in 3 patients, and no mutation was found in WFS1 or in the CISD2 gene (611507) in 3 patients. Chaussenot et al. (2011) observed that the age of onset of both diabetes mellitus and optic atrophy was significantly delayed when patients carried 1 or 2 missense mutations, whereas the development or age of onset of neurologic symptoms were not correlated with genotype.

De Heredia et al. (2013) analyzed clinical and genetic data of 412 patients with Wolfram syndrome published in the aforegoing 15 years. They found that 15% of published patients did not fulfill the ascertainment criteria of juvenline onset of diabetes mellitus and bilateral optic atrophy. They also found that genotypic prevalence differences may exist among countries; that diabetes mellitus and optic atrophy might not be the first 2 clinical features in some patients; that mutations are not uniformly distributed in WFS1; and that the age at onset of diabetes mellitus, hearing defects, and diabetes insipidus may depend on the patient's genotypic class, as does disease progression rate. Among 392 patients with age specified for any clinical symptom, the onset followed this pattern: diabetes mellitus during the first decade of life, optic atrophy during the early second decade, diabetes insipidus and hearing defects (mainly deafness) in the second decade, and urologic problems and neurologic problems during the third decade. Although the median age of onset of urologic symptoms among 76 patients was 20 years, a large proportion of patients developed symptoms at 10 to 20 years, with 3 peaks observed: 1 at 13, 1 at 21, and 1 at 33. The median age of 29 deceased patients was 27 +/- 11.4 years, with 2 peaks, 1 at 24 and the other at 45. De Heredia et al. (2013) reported 178 different and nonuniformly distributed mutations, concentrated mainly in transmembrane domains and glycosylation sites. In addition, they found a high concentration of mutations at the N-terminus between amino acids 94 and 237 and in the last 100 amino acids. Only 6 mutations were found in more than 5% of patients. De Heredia et al. (2013) classified the 178 mutations into 3 types according to predicted effects on WFS1 protein expression. The 337 patients fell into 3 main genotypic classes: class A, no WFS1 protein produced; class B, reduced expression of a defective WFS1 protein; and class C, expression of a defective WFS1 protein. The genotypic class-phenotype correlation performed indicated that the differences observed in age of onset of diabetes mellitus, diabetes insipidus, and hearing defects correlated with the genotypic classification performed. Because not all patients met diagnostic criteria despite 2 mutations in WFS1, de Heredia et al. (2013) suggested changing the diagnostic criteria to either patients with diabetes mellitus and optic atrophy at any age, which would include 96% of patients, or to patients with any 2 of the DIDMOAD symptoms, which would include almost 99% of patients. They favored the latter criterion, especially since 20% of patients with Wolfram syndrome do not present with clinical features other than diabetes mellitus.

History

Borgna-Pignatti et al. (1989) described 2 Italian children, related as first cousins, who developed megaloblastic and sideroblastic anemia, neutropenia, and borderline thrombocytopenia. The authors characterized these children as having DIDMOAD syndrome. In both children, thiamine pyrophosphate in erythrocytes and thiamine pyrophosphokinase activity were lower than the lowest values observed in control subjects. A month after institution of treatment with thiamine, the hematologic findings had returned to normal and insulin requirements had decreased. Withdrawal of thiamine repeatedly induced relapse of the anemia and increase in insulin requirements. Schwingshandl and Borkenstein (1989) found no improvement in insulin dependence after treating 2 patients with DIDMOAD syndrome with thiamine; these patients did not have anemia. In addition, Borgna-Pignatti et al. (1989) stated that they observed no response in glycemic control to thiamine treatment in 4 patients with classic DIDMOAD syndrome without anemia. Borgna-Pignatti et al. (1989) suggested that diabetes, deafness, and optic atrophy with and without anemia may actually represent different disorders. A later study by Neufeld et al. (1997) determined that the original patients reported by Borgna-Pignatti et al. (1989) in fact had thiamine-responsive megaloblastic anemia syndrome (TRMA; 249270), with linkage to chromosome 1q. These findings underscored the phenotypic overlap between DIDMOAD syndrome and TRMA, but demonstrated that they are distinct.