J Endocrinol Invest (2014) 37:195–202 DOI 10.1007/s40618-013-0039-4
ORIGINAL ARTICLE
Phenotypical and genotypical expression of Wolfram syndrome in 12 patients from a Sicilian district where this syndrome might not be so infrequent as generally expected F. Lombardo • G. Salzano • C. Di Bella • T. Aversa • F. Pugliatti • S. Cara • M. Valenzise F. De Luca • L. Rigoli
•
Received: 13 June 2013 / Accepted: 6 December 2013 / Published online: 9 January 2014 Ó Italian Society of Endocrinology (SIE) 2014
Abstract Background Since the original description, there have been only few epidemiological studies of Wolfram syndrome (WS). Aim Aims of the present paper are to ascertain WS prevalence and expression in a district of North-eastern Sicily, i.e. a geographic area where consanguineous unions are not very unusual. Materials and methods Prevalence rates of WS in the Messina district were calculated by taking into consideration both the total population (653,737) and the populations included within the 0–30 year age range (202,681). We estimated the relative prevalence of WS among patients with youth-onset insulin-dependent diabetes mellitus (DM) who are currently aged under 30 years (256). Results Global WS prevalence in our district is 1:54,478, whereas prevalence among individuals under 30 is 1:16,890 and relative prevalence among patients with juvenile-onset insulin-dependent DM is 1:22.3. When compared with the patients with insulin-dependent DM of Messina district, WS patients did not exhibit significant differences in terms of biochemical features at DM onset, whereas age at DM diagnosis was significantly earlier in WS group. Conclusions (a) WS prevalence is not so infrequent as generally expected; (b) in our series, DM presented before 10 years in 11/12 patients and ten cases have already developed all the four peculiar manifestations of WS by
F. Lombardo G. Salzano C. Di Bella T. Aversa F. Pugliatti S. Cara M. Valenzise F. De Luca (&) L. Rigoli Department of Pediatrics, University of Messina, Via Consolare Valeria 1, 98125 Messina, Italy e-mail: [email protected]
26 years; (c) 9/12 patients exhibited a homozygous frameshift/truncation mutation (Y454_L459del_fsX454), which is the one most frequently found also in patients from other Italian regions; (d) age at DM diagnosis was significantly earlier in WS group than in the patients with insulin-dependent DM of Messina district. Keywords DIDMOAD association Epidemiological evaluation Phenotypical expression Abbreviations WS Wolfram syndrome DIDMOAD Diabetes insipidus, diabetes mellitus, optic atrophy, deafness DM Juvenile-onset diabetes mellitus OA Optic atrophy DI Diabetes insipidus D Deafness DKA Diabetic ketoacidosis
Introduction Wolfram first reported the combination of juvenile-onset diabetes mellitus (DM) and optic atrophy (OA) in 1938. With the discovery of two other essential components, diabetes insipidus (DI) and deafness (D), Wolfram syndrome (WS) is now also called the DI, DM, OA and D (DIDMOAD) syndrome. Other, less common, manifestations include renal tract abnormalities, neuropsychiatric disorders and hypogonadism [1]. WS is inherited as an autosomal recessive trait and is caused by a mutation in the gene on chromosome 4p16.1 encoding a transmembrane protein called wolframin, which
123
196
J Endocrinol Invest (2014) 37:195–202
is normally present in several tissues. This disorder has a variable clinical presentation and, due to progressive neurodegeneration, usually results in death under the 5th decade [2, 3]. Prevalence has been estimated as 1:770,000 in the UK, with a carrier frequency of 1 in 354 [4]. Such prevalence is significantly lower than what had been previously estimated in a North-American population (1:100,000 with a carrier frequency of 1 in 100), based on the 1:148 frequency of OA in a juvenile—DM Clinic [5]. According to the most recent epidemiological study, WS prevalence in North India has been estimated as 1:805,000 [6], which is not very different from the UK [4]. However, it has to be considered that, since the original description of Wolfram and Wagener [7], there have been hundreds of case reports but only few epidemiological studies [4–6], at least to the best of our knowledge. Therefore, it is not possible, on the basis of available data, to establish whether WS prevalence is the same globally or might significantly change in the different ethnic groups. Here, we describe clinical presentation and genetic analysis in a large cohort of 12 WS patients diagnosed over the last 15 years in the Pediatric Diabetes Clinic of a tertiary care Centre in North-eastern Sicily.
Materials and methods Patients Our study population consisted of 12 WS Caucasian patients (six males) from seven couples and five unrelated
families, who were diagnosed during the last 15 years in the Unit of Pediatric Endocrinology of a 660-bed tertiary care Hospital in North-eastern Sicily, which caters for the overall population of the Messina district (653,737 inhabitants). This Unit is the only recognized reference Centre of the Messina district for diagnosis, treatment and follow-up of youth-onset DM. In addition, nearly all diabetic patients between 0 and 30 years are recorded and followed in our Unit. However, to definitely confirm that the 12 patients included in the present study are the only WS individuals aged under 30 who are currently living in the Messina district, we have contacted all the other medical Units of our region. All patients had been born and are living in the Messina district. At present, they are aged between 9 and 29 years and have been followed in our Pediatric Endocrinology Unit since time of diagnosis of juvenile-onset, insulindependent, non-autoimmune DM. Sex, present age, age at diagnosis of WS and at time of detection of each WS cardinal feature are listed in Table 1, together with other anomalies that have been observed in these 12 patients up to now. The first seven patients of Table 1 descend from a fivegeneration WS pedigree had been previously reported elsewhere [8] and that was reconstructed through a proband, who exhibited all four cardinal manifestations of DIDMOAD association (Fig. 1). This proband (No. 1 of Table 1 and IV-9 of Fig. 1) was born to healthy consanguineous parents (III-5 and III-6) and has two affected sisters (Nos. 2 and 3 of Table 1 and IV-8 and 10 of Fig. 1) and one affected brother (No. 4 of Table 1 and IV-11 of
Table 1 Sex, correspondence numbers for Fig. 1, present age (years), age at diagnosis of Wolfram syndrome (WS) and at the time of the detection of the major WS clinical features, and other anomalies in the 12 patients of the present series Patients
Sex
Figure 1
Present age
Age at diagnosis
DM
DI
OA
D
Other anomalies
1
M
IV-9
20
10
4
11
10
9
Psychiatric disorders; Hashimoto’s thyroiditis
2
F
IV-8
26
14
5
18
14
13
Psychiatric disorders; Bilateral cataract
3
F
IV-10
23
18
5
11
18
7
Bilateral cataract; Psychiatric disorders
4
M
IV-11
20
15
3
10
15
6
Psychiatric disorders; Renal out-flow abnormalities
5
F
V-3
24
12
6
10
12
10
Renal out-flow abnormalities
6
M
V-4
20
10
4
17
10
13
–
7
F
V-5
20
11
3
14
11
10
Hashimoto’s thyroiditis
8
F
–
18
14
9
15
14
–
–
9
F
–
9
6
5
–
6
–
–
10
M
–
29
16
13
25
16
26
Hashimoto’s thyroiditis
11 12
M M
– –
29 27
14 10
1 6
15 18
14 10
16 24
Hypergonadotropic hypogonadism –
Mean
22.1
12.5
5.33
14.9
12.5
13.4
SD
5.6
3.3
3.11
4.5
3.3
6.8
Median
21.5
13
5
15
13
11.5
Range
9–29
6–18
1–13
10–25
6–18
6–26
DM juvenile-onset diabetes mellitus, DI diabetes insipidus, OA optic atrophy, D deafness
123
J Endocrinol Invest (2014) 37:195–202 Fig. 1 Family tree of the patients Nos. 1–7 of Table 1; generation are marked on the left with Roman numerals, while individuals are identified with Arabic number; the seven patients are indicated with black symbols; individuals whose DNA was tested are indicated by a dash above their symbol and the proband (IV-9) is indicated by an arrow
197
I 1
2
II 1
2
3
4
5
III
IV
V
1
2
1
2
1
2
3
3
3
4
Fig. 1). The other three patients of this pedigree (Nos. 5, 6 and 7 of Table 1 and V-3, 4 and 5 of Fig. 1) share two common ancestors (I-1 and I-2) and were born to two other first cousin couples (IV-1 and IV-2, IV-3 and IV-4, respectively). This very extended family comes from a small, isolated town in the Nebrodi Mountains. Patients Nos. 8 and 9 of Table 1 are two sisters belonging to another, unrelated family living on the Northeast Coast, with non-consanguineous parents. The remaining three patients (Nos. 10, 11 and 12) come from three other unrelated families living in different areas of the Messina district and were born to healthy non-consanguineous parents. Study design and methods Data concerning clinical evolution of WS from time of diagnosis onwards in these 12 patients were retrospectively gathered from clinical records, according to a standardized form that had been prepared to document the main clinical details, including age at onset of DM and other components of the syndrome, duration of the disease and familial aggregation. Among first-degree relatives, those who gave consent underwent a detailed clinical assessment, including relevant biochemistry, to rule out any features of WS. In all patients, diagnosis of WS was confirmed by genetic analysis, which was performed by studying eight exons and the exon–intron sequences of WFS1 gene. The same genetic analyses were performed also in the 14 parents of our patients. Clinical and biochemical findings at DM onset in the WS cohort (N 12) were compared with those collected in a population of children and adolescents with juvenile-onset, insulin-dependent DM (N 256) aged under 30 who are currently living in Messina district. The analyzed data were: age, HbA1c concentrations and diabetic ketoacidosis
4
4
5
6
5
5
6
7
6
8
9
10
11
7
(DKA) prevalences at DM diagnosis; honeymoon duration and daily insulin requirements at DM onset were also taken into consideration. DKA was defined when blood glucose levels were [250 mg/dl, pH was \7.3 and serum bicarbonate levels were\15 mEq/l. Honeymoon duration was assessed on the basis of the criteria that are generally adopted for definition of partial remission in T1DM: HbA1c concentrations \7.0 % and daily insulin requirement \0.5 IU/kg/day [9]. HbA1c was measured by an immunoassay technique with a monoclonal antibody directed against a sequence of the HbA1c molecule. For statistical purposes data are expressed as mean ± SD or median and range values as appropriate. Comparisons between groups were performed by the Student’s t test (normally distributed data) or the Mann– Whitney U test (non parametric data) as appropriate. Frequency rates were compared by v2-test. The level of significance was set at 0.05. Study design was approved by the Ethics Committee of our Hospital and patients and/or their parents gave informed consent. Appropriate consents for this study were also obtained from the Study Group for Diabetes Mellitus of the Italian Society for Pediatric Endocrinology and Diabetology. Our study was conducted according to the Declaration of Helsinki. DNA extraction and PCR amplification Genomic DNA was extracted from peripheral blood leukocytes by standard phenol chloroform procedure. Fourteen primer pairs were used to amplify the entire coding region of WFS1 gene, including the non-coding exon 1 and the exon–intron boundaries. Exon 8 was subdivided into seven regions (from 8a to 8g), as previously described [10]. PCR amplifications were performed in a
123
198
J Endocrinol Invest (2014) 37:195–202
Gene Amp PCR System 9700 (Perkin Elmer, Foster City, CA) with an initial denaturation step at 94 °C for 15 min and then 30 cycles at 94 °C for 30 s, 60 °C for 30 s, 72 °C for 45 s, followed by 7 min of final extension at 72 °C. WFS5 and WFS8c amplifications were performed at annealing temperature of 62 °C. The specificity of the amplified samples was checked out by agarose gel electrophoresis before further analysis.
Results Main findings at DM diagnosis All these data are summarized in Table 2. The only significant difference between WS series and the overall population with insulin-dependent DM regarded age at DM diagnosis, that was significantly earlier in WS cohort (Table 2). All the remaining data did not significantly differ in the two groups (Table 2).
Sequencing analysis Clinical findings DNA PCR samples underwent direct sequencing using a 3130 Genetic Analyzer (Applied Biosystem, Foster City, CA), according to PRISM Dye Terminator and Dye primer cycle sequencing chemistries. Sequences were compared with reference to human genomic and cDNA WSF1 sequences (GenBank accession No. AC004689 and Y18064). To avoid potential artefacts due to AmpliTaq GoldTM DNA Polymerase (PE Applied Biosystem, Foster City, CA), each sequence alteration was confirmed by sequencing both DNA strands of three independent PCR products. The sequence variants were considered mutations when they: (a) caused a nonconservative amino acid change; (b) were absent in 300 ethnically matched control chromosomes; (c) affected phylogenetically conserved residues. Other DNA variations that did not fulfil these criteria were considered polymorphisms.
Prevalence studies Prevalence rates of WS in the Messina district were calculated by taking into consideration both the total population of our district (653,737) and the population included within the 0–30 year age range (202,681). Furthermore, we also estimated the relative prevalence of WS among patients with youth-onset insulin-dependent DM of the Messina district who are currently aged under 30 years (256). To exclude any bias due to consanguinity we have also taken into consideration, for epidemiological purposes, only the five patients from non-consanguineous unions. Table 2 Diabetic ketoacidosis (DKA) prevalence and median values (and ranges) of age, glycosylated hemoglobin (HbA1c), daily insulin requirement and honeymoon duration at the time of diabetes mellitus Patients
DKA prevalence (%)
Group A (Nos. 12)
8.3
Group B (Nos. 256)
28.1
Age (years)
The minimum ascertainment criteria for diagnosis of WS, i.e. juvenile-onset, insulin-dependent, non-autoimmune DM and bilateral OA were fulfilled in all patients. In all cases, DM was the earliest detected abnormality and presentation occurred during the first decade in 11/12 patients (Table 1). The remaining cardinal features of DIDMOAD association were detected some years later and only the two youngest patients of the entire cohort (Nos. 8 and 9) have not so far developed the full range of DIDMOAD manifestations (Table 1). Other less common clinical features included psychiatric disorders (four cases from the same pedigree), renal outflow tract abnormalities (two cases from the same pedigree), Hashimoto’s thyroiditis (three cases), bilateral cataract (two sisters with the same features) and primary gonadal atrophy (one case) (Table 1). On overall, clinical picture was distinctly more complex and severe in the seven patients from the same pedigree (Table 1), although correlation with genotype was not absolute, as demonstrated by the milder phenotype observed in the patients Nos. 8 and 9 who exhibited the same genotype of patients from 1 to 7 (Table 3). However, it has to be considered that patient No. 9 is still very young. WFS1 mutations and genotypes Direct sequencing of WFS1 gene identified two mutated alleles in all patients. In 9/12 patients, the mutations were homozygous for the same allele, whilst the remaining three patients were compound heterozygotes (Table 3). We (DM) onset in Wolfram syndrome group (Group A) and in that with insulin-dependent DM (Group B)
HbA1c (%)
Insulin requirement (IU/kg/day)
Honeymoon (months)
5.0 (1–13)
10 (8.0–12.8)
0.9 (0.7–1.0)
6.0 (0–12)
9.2 (0.8–22.1)
10.1 (5.5–14.5)
0.7 (0.1–1.8)
7.0 (0–41)
v2
0.13
–
–
–
–
p
2.20
0.001
0.96
0.15
0.53
123
J Endocrinol Invest (2014) 37:195–202
199
Table 3 Mutations of WFS1 gene found in the 12 patients of the present Wolfram syndrome series Patients
Exon
Nucleotide change
Amino-acid change
Type of mutation
Protein domain
Zygosity
1
8b
c1362_1377del16
Y454_L459del_fsX454
Frameshift/truncaton
CD3
Homozygote
2
8b
c1362_1377del16
Y454_L459del_fsX454
Frameshift/truncaton
CD3
Homozygote
3
8b
c1362_1377del16
Y454_L459del_fsX454
Frameshift/truncaton
CD3
Homozygote
4
8b
c1362_1377del16
Y454_L459del_fsX454
Frameshift/truncaton
CD3
Homozygote
5
8b
c1362_1377del16
Y454_L459del_fsX454
Frameshift/truncaton
CD3
Homozygote
6
8b
c1362_1377del16
Y454_L459del_fsX454
Frameshift/truncaton
CD3
Homozygote
7 8
8b 8b
c1362_1377del16 c1362_1377del16
Y454_L459del_fsX454 Y454_L459del_fsX454
Frameshift/truncaton Frameshift/truncaton
CD3 CD3
Homozygote Homozygote
9
8b
c1362_1377del16
Y454_L459del_fsX454
Frameshift/truncaton
CD3
Homozygote
10
5/8c
c532_537del6; c1673G[A
K178_A179del; R558H
Deletion; missense
CD1; CD4
Compound heterozygote
11
8b/ IVS6
c.1328G[T; c.712?16G[A
S443I; –
Missense; splice
TM4; –
Compound heterozygote
12
8b/8c
c1230delCTCT; c.1246delTTC
L410del-fsX441; 516del
Frameshift/truncaton; deletion
TM3-TM4; ER3
Compound heterozygote
Fig. 2 a WFS1 gene, which comprises eight exons; exon 1 is non-coding. b Hypothetical structure of wolframin protein. The cytoplasmic regions are yellow, the endoplasmatic regions are green and transmembrane regions are blue. Mutations found in the patients are boxed
identified four different types of WFS1 gene mutations. The mutations were either missense type (n = 2), or frameshift/truncation type (n = 10), or deletion type (n = 2), or affecting the splice site (n = 1) (Table 3). In wolframin, the mutations were localized in the following regions: three in three cytoplamic domains (CD1, CD3 and CD4); two in the TM3 and TM4 transmembrane domains, and one in the ER3 endoplasmic reticulum domain. One mutation was found in a splicing site of WFS1 gene (Fig. 2). The most frequent alteration in our WS group was a Y454_L459del_fsX454 frameshift/truncation mutation, that was harbored by nine out of 12 analyzed WS subjects
(Nos. 1–9): a 16 base-pair deletion (c1362_1377del16) localized in the CD3 cytoplasmic domain and all these patients were found to be homozygotes for it. Missense mutations were detected in two patients. Patient No. 10 was a K178_A179del/R558H compound heterozygous. K178_A179del mutation was a six c532_537deletion localized in the N-terminal region. It was the only mutation found in exon 5. R558H mutation (c1673G[A), a missense mutation, was found in the CD4 cytoplasmic domain of the wolframin. Patient No. 11 was compound heterozygous for S443I mutation and c.712?16G[A. S443I missense mutation was found at np1328 (AGC[ATC) in exon 8, leading to a
123
200
substitution of serine to isoleucine. It was situated in the TM4 transmembrane region. c.712?16G[A was situated at intron 6. Patient No. 12 was compound heterozygous for a L410del-fsX441/F516del. L410del-fsX441 was a four nucleotide deletion (c.1230delCTCT) (frameshift/truncation mutation) localized in the TM3-TM4 transmembrane regions. F516del mutation (c.1246delTTC) was found in the ER3 endoplamic reticulum region of wolframin. Results of patients’ genetic analyses were supported by results of their parents’ molecular investigations. All the detected mutations had been previously described. Epidemiological data The 12 patients included in the present study are the only WS patients aged under 30 who are currently living in the Messina district, as also confirmed by the other medical Units of the entire Sicilian region, which were contacted for this purpose. Since the total population of children, adolescents and young adults under 30 completed years in this district is currently 202,681, WS prevalence among individuals under 30 of the Messina district might be estimated as one in 16,890. If the total population of the Messina district were considered (653,737), the prevalence of WS in this district would be 1:54,478, i.e. 18.3 cases per million population. If the seven patients from consanguineous unions were excluded, the prevalence of WS among individuals under 30 would be 1:40,536 (1:130,747 in the overall population). According to our findings, the relative prevalence of WS among cases with juvenile-onset, insulin-dependent DM aged under 30 who are currently living in Messina district might be estimated as 1 in 22.3 (1 in 51.2 considering only the WS patients from non-consanguineous unions). The 12 patients in the study have 26 first-degree relatives (14 parents and 12 siblings). All of the 14 parents come from Sicilian pedigrees, as confirmed by a detailed analysis of family background that was carried out on all patients and their first-degree relatives. Both parents of the patients Nos. 1–4, 5–6, 7 and 8–9 (Table 1) were found to be healthy carriers of the same mutation (Y454_L459del_fsX454) that had been detected at homozygous state in the respective offsprings. Each of the parents of patients Nos. 10, 11 and 12 was found to be healthy carrier of one of the mutations that had been detected at compound heterozygous state in the respective offsprings. No clinical and/or biochemical abnormalities that might be compatible with WS diagnosis were detected in the asymptomatic first-degree relatives who underwent a
123
J Endocrinol Invest (2014) 37:195–202
detailed assessment aiming to exclude any features of this syndrome.
Discussion The present investigation covers one of the largest Italian WS populations that have been described until now [10– 12]. One of the purposes of our study was to investigate WS prevalence in a geographic area where consanguineous unions may not be very unusual, particularly in the mountain villages of the hinterland. According to our data, the frequency of WS in the total population of a Northeastern Sicilian district might be estimated as one in around 55,000 inhabitants, which is distinctly higher if compared with figures reported in both UK [4] and North India [6], i.e. 1:770,000 and 1:805,000, respectively, but not too far from that reported in North America, i.e. 1 in 100,000 inhabitants [5]. However, it should be considered that the prevalence estimates in these four epidemiological studies were based on different designs and methods. Moreover, whilst the previous investigations covered very large populations from extended regions, the present one was based on the total population of a limited district area including \700,000 inhabitants. However, the relatively elevated frequency of WS in our district was confirmed by its high prevalence rate in the context of the cases with youthonset, insulin-dependent DM aged under 30 and living in the same area: 1 in 22.3, vs 1 in 150 in the study by Gunn et al. [13] on USA populations. It should be considered that in the present study our population consisted of 12 patients from seven couples and five unrelated families, which makes it difficult to extrapolate unbiased prevalence data. Nevertheless, even though the seven patients coming from the same pedigree were excluded from this epidemiological evaluation and only the five patients from non-consanguineous unions were considered for this purpose, the frequency of WS in the total population of our district might be estimated at around 1:130,000, i.e. very close to that observed in North America and very far from UK and North India figures. Therefore, although we are aware of the risk that referral centres might sometimes overestimate prevalence of rare diseases, we could postulate, on the basis of our data and North-American ones, that WS is not necessarily so infrequent as generally expected according to other epidemiological data [4, 6]. The proportion of patients with the four components of DIDMOAD association in our series was distinctly higher than that previously reported by Barrett et al. [4]: 83.3 vs 53 %. Moreover, it should be underlined that the only two patients of our cohort who have not so far developed the
J Endocrinol Invest (2014) 37:195–202
full range of DIDMOAD manifestations are the youngest ones of the overall series. This finding might suggest that most patients presenting with DM and OA in childhood will afterwards develop the other two components of DIDMOAD association within some years, especially if glycemic control is very poor [14]. In the present study, we have also demonstrated that WS-related DM presents earlier in life than type 1 DM, thus confirming another previous report [14]. This finding might be probably due to the fact that the process of endoplasmic reticulum stress-related deterioration of betacell function in WS patients already starts at the beginning of life, whilst beta-cell destruction in type 1 DM is triggered later [14]. In line with previous reports [2, 4], also in the present series WS-related DM was autoantibody-negative in all cases and not commonly complicated by DKA. All the other investigated parameters (HbA1c concentrations, honeymoon duration and insulin requirements at DM onset) did not achieve a statistical relevance between WSrelated DM and type 1 DM groups, as well as in the report by Rohayem et al. [14]. Apart from the manifestations of DIDMOAD association, some patients of this study exhibited other less common complications, that are generally regarded as an integral part of WS: renal tract abnormalities, psychiatric disorders, primary hypogonadism [1, 4, 6] and even bilateral cataract, which should be considered as a regular component of WS, according to recent reports [15, 16]. These less typical features were observed almost exclusively in the seven patients from the same pedigree. The association with an autoimmune thyroid disease in three patients of our cohort might suggest an autoimmune origin of DM, at least in those cases. Nevertheless, all of our patients had a non-autoimmune form of DM, as also confirmed by the absence of anti-glutamic acid decarboxylase autoantibodies in each of them. In the present series, we have detected four different mutations of WFS1 gene, which were mainly concentrated in exon 8, as in most previous reports [1]. The most frequent mutation was a 16 bp deletion (Y454_L459del_fsX454), that resulted in a frameshift and premature stop codon. Its causal role in the pathogenesis of WS was confirmed by the finding that it was absent in 300 ethnically matched control chromosomes, whereas in all the affected patients with this mutation it was present in a homozygous state. This mutation had been previously reported also in many WS patients from other Italian regions [1] and therefore it cannot be considered as pathognomonic of Sicilian patients. Also the other three mutations encountered in the present series had been previously described in other study populations [10, 11]. However, it has to be considered that novel gene mutations in WS patients are very infrequently encountered and described [17].
201
We conclude that: (a) WS might not be so infrequent as generally expected; (b) in our series, DM presented before 10 years in 11/12 patients and ten cases have already developed all four manifestations of DIDMOAD association by 26 years; (c) 9/12 patients exhibited a homozygous frameshift/truncation mutation (Y454_L459del_fsX454), which is the one most frequently found also in patients from other Italian regions; (d) age at DM diagnosis was significantly earlier in WS group than in the patients with insulin-dependent DM of Messina district. Conflict of interest The authors F. Lombardo, G. Salzano, C. Di Bella, T. Aversa, F. Pugliatti, S. Cara, M. Valenzise, F. De Luca and L. Rigoli declare no conflict of interest.
References 1. Rigoli L, Lombardo F, Di Bella C (2011) Wolfram syndrome and WFS1 gene. Clin Genet 9:103–117 2. Kinsley BT, Swift M, Dumont RH, Swift RG (1995) Morbidity and mortality in the Wolfram syndrome. Diabetes Care 18:1566–1570 3. Cremers CW, Wijdeveld PG, Pinckers AJ (1977) Juvenile diabetes mellitus, optic atrophy, hearing loss, diabetes insipidus, atonia of the urinary tract and bladder, and other abnormalities (Wolfram syndrome). A review of 88 cases from the literature with personal observations on 3 new patients. Acta Paediatr Scand Suppl 264:1–16 4. Barrett TG, Bundey SE, Macleod AF (1995) Neurodegeneration and diabetes: UK nationwide study of Wolfram (DIDMOAD) syndrome. Lancet 346:1458–1463 5. Fraser FC, Gunn T (1977) Diabetes mellitus, diabetes insipidus, and optic atrophy. An autosomal recessive syndrome? J Med Genet 14:190–193 6. Ganie MA, Laway BA, Nisar S et al (2011) Presentation and clinical course of Wolfram (DIDMOAD) syndrome from North India. Diabet Med 28:1337–1342 7. Wolfram DJ, Wagener HP (1938) Diabetes mellitus and simple optic atrophy among siblings: report of four cases. Mayo Clinic Proc 13:715–718 8. Lombardo F, Chiurazzi P, Ho¨rtnagel K et al (2005) Clinical picture, evolution and peculiar molecular findings in a very large pedigree with Wolfram syndrome. J Pediatr Endocrinol Metab 18:1391–1397 9. Lombardo F, Valenzise M, Wasniewska M et al (2002) Two-year prospective evaluation of the factors affecting honeymoon frequency and duration in children with insulin dependent diabetes mellitus: the key-role of age at diagnosis. Diabetes Nutr Metab 15:246–251 10. Colosimo A, Guida V, Rigoli L et al (2003) Molecular detection of novel WFS1 mutations in patients with Wolfram syndrome by a DHPLC-based assay. Hum Mutat 21:622–629 11. Tessa A, Carbone I, Matteoli MC et al (2001) Identification of novel WFS1 mutations in Italian children with Wolfram syndrome. Hum Mutat 17:348–349 12. d’Annunzio G, Minuto N, D’Amato E et al (2008) Wolfram syndrome (diabetes insipidus, diabetes, optic atrophy, and deafness): clinical and genetic study. Diabetes Care 31:1743–1745 13. Gunn T, Bortolussi R, Little JM, Andermann F, Fraser FC, Belmonte MM (1976) Juvenile diabetes mellitus, optic atrophy, sensory nerve deafness, and diabetes insipidus—a syndrome. J Pediatr 89:565–570
123
202 14. Rohayem J, Ehlers C, Wiedemann B et al (2011) Diabetes and neurodegeneration in Wolfram syndrome: a multicenter study of phenotype and genotype. Diabetes Care 34:1503–1510 15. Hansen L, Eiberg H, Barrett T et al (2005) Mutation analysis of the WFS1 gene in seven Danish Wolfram syndrome families; four new mutations identified. Eur J Hum Genet 13:1275–1284
123
J Endocrinol Invest (2014) 37:195–202 16. Titah SM, Meunier I, Blanchet C et al (2012) Cataract as a phenotypic marker for a mutation in WFS1, the Wolfram syndrome gene. Eur J Ophthalmol 22:254–258 17. Rigoli L, Lombardo F, Salzano G et al (2013) Identification of one novel causative mutation in exon 4 of WFS1 gene in two Italian siblings with classical DIDMOAD syndrome phenotype. Gene 526:487–489
ORIGINAL ARTICLE
Phenotypical and genotypical expression of Wolfram syndrome in 12 patients from a Sicilian district where this syndrome might not be so infrequent as generally expected F. Lombardo • G. Salzano • C. Di Bella • T. Aversa • F. Pugliatti • S. Cara • M. Valenzise F. De Luca • L. Rigoli
•
Received: 13 June 2013 / Accepted: 6 December 2013 / Published online: 9 January 2014 Ó Italian Society of Endocrinology (SIE) 2014
Abstract Background Since the original description, there have been only few epidemiological studies of Wolfram syndrome (WS). Aim Aims of the present paper are to ascertain WS prevalence and expression in a district of North-eastern Sicily, i.e. a geographic area where consanguineous unions are not very unusual. Materials and methods Prevalence rates of WS in the Messina district were calculated by taking into consideration both the total population (653,737) and the populations included within the 0–30 year age range (202,681). We estimated the relative prevalence of WS among patients with youth-onset insulin-dependent diabetes mellitus (DM) who are currently aged under 30 years (256). Results Global WS prevalence in our district is 1:54,478, whereas prevalence among individuals under 30 is 1:16,890 and relative prevalence among patients with juvenile-onset insulin-dependent DM is 1:22.3. When compared with the patients with insulin-dependent DM of Messina district, WS patients did not exhibit significant differences in terms of biochemical features at DM onset, whereas age at DM diagnosis was significantly earlier in WS group. Conclusions (a) WS prevalence is not so infrequent as generally expected; (b) in our series, DM presented before 10 years in 11/12 patients and ten cases have already developed all the four peculiar manifestations of WS by
F. Lombardo G. Salzano C. Di Bella T. Aversa F. Pugliatti S. Cara M. Valenzise F. De Luca (&) L. Rigoli Department of Pediatrics, University of Messina, Via Consolare Valeria 1, 98125 Messina, Italy e-mail: [email protected]
26 years; (c) 9/12 patients exhibited a homozygous frameshift/truncation mutation (Y454_L459del_fsX454), which is the one most frequently found also in patients from other Italian regions; (d) age at DM diagnosis was significantly earlier in WS group than in the patients with insulin-dependent DM of Messina district. Keywords DIDMOAD association Epidemiological evaluation Phenotypical expression Abbreviations WS Wolfram syndrome DIDMOAD Diabetes insipidus, diabetes mellitus, optic atrophy, deafness DM Juvenile-onset diabetes mellitus OA Optic atrophy DI Diabetes insipidus D Deafness DKA Diabetic ketoacidosis
Introduction Wolfram first reported the combination of juvenile-onset diabetes mellitus (DM) and optic atrophy (OA) in 1938. With the discovery of two other essential components, diabetes insipidus (DI) and deafness (D), Wolfram syndrome (WS) is now also called the DI, DM, OA and D (DIDMOAD) syndrome. Other, less common, manifestations include renal tract abnormalities, neuropsychiatric disorders and hypogonadism [1]. WS is inherited as an autosomal recessive trait and is caused by a mutation in the gene on chromosome 4p16.1 encoding a transmembrane protein called wolframin, which
123
196
J Endocrinol Invest (2014) 37:195–202
is normally present in several tissues. This disorder has a variable clinical presentation and, due to progressive neurodegeneration, usually results in death under the 5th decade [2, 3]. Prevalence has been estimated as 1:770,000 in the UK, with a carrier frequency of 1 in 354 [4]. Such prevalence is significantly lower than what had been previously estimated in a North-American population (1:100,000 with a carrier frequency of 1 in 100), based on the 1:148 frequency of OA in a juvenile—DM Clinic [5]. According to the most recent epidemiological study, WS prevalence in North India has been estimated as 1:805,000 [6], which is not very different from the UK [4]. However, it has to be considered that, since the original description of Wolfram and Wagener [7], there have been hundreds of case reports but only few epidemiological studies [4–6], at least to the best of our knowledge. Therefore, it is not possible, on the basis of available data, to establish whether WS prevalence is the same globally or might significantly change in the different ethnic groups. Here, we describe clinical presentation and genetic analysis in a large cohort of 12 WS patients diagnosed over the last 15 years in the Pediatric Diabetes Clinic of a tertiary care Centre in North-eastern Sicily.
Materials and methods Patients Our study population consisted of 12 WS Caucasian patients (six males) from seven couples and five unrelated
families, who were diagnosed during the last 15 years in the Unit of Pediatric Endocrinology of a 660-bed tertiary care Hospital in North-eastern Sicily, which caters for the overall population of the Messina district (653,737 inhabitants). This Unit is the only recognized reference Centre of the Messina district for diagnosis, treatment and follow-up of youth-onset DM. In addition, nearly all diabetic patients between 0 and 30 years are recorded and followed in our Unit. However, to definitely confirm that the 12 patients included in the present study are the only WS individuals aged under 30 who are currently living in the Messina district, we have contacted all the other medical Units of our region. All patients had been born and are living in the Messina district. At present, they are aged between 9 and 29 years and have been followed in our Pediatric Endocrinology Unit since time of diagnosis of juvenile-onset, insulindependent, non-autoimmune DM. Sex, present age, age at diagnosis of WS and at time of detection of each WS cardinal feature are listed in Table 1, together with other anomalies that have been observed in these 12 patients up to now. The first seven patients of Table 1 descend from a fivegeneration WS pedigree had been previously reported elsewhere [8] and that was reconstructed through a proband, who exhibited all four cardinal manifestations of DIDMOAD association (Fig. 1). This proband (No. 1 of Table 1 and IV-9 of Fig. 1) was born to healthy consanguineous parents (III-5 and III-6) and has two affected sisters (Nos. 2 and 3 of Table 1 and IV-8 and 10 of Fig. 1) and one affected brother (No. 4 of Table 1 and IV-11 of
Table 1 Sex, correspondence numbers for Fig. 1, present age (years), age at diagnosis of Wolfram syndrome (WS) and at the time of the detection of the major WS clinical features, and other anomalies in the 12 patients of the present series Patients
Sex
Figure 1
Present age
Age at diagnosis
DM
DI
OA
D
Other anomalies
1
M
IV-9
20
10
4
11
10
9
Psychiatric disorders; Hashimoto’s thyroiditis
2
F
IV-8
26
14
5
18
14
13
Psychiatric disorders; Bilateral cataract
3
F
IV-10
23
18
5
11
18
7
Bilateral cataract; Psychiatric disorders
4
M
IV-11
20
15
3
10
15
6
Psychiatric disorders; Renal out-flow abnormalities
5
F
V-3
24
12
6
10
12
10
Renal out-flow abnormalities
6
M
V-4
20
10
4
17
10
13
–
7
F
V-5
20
11
3
14
11
10
Hashimoto’s thyroiditis
8
F
–
18
14
9
15
14
–
–
9
F
–
9
6
5
–
6
–
–
10
M
–
29
16
13
25
16
26
Hashimoto’s thyroiditis
11 12
M M
– –
29 27
14 10
1 6
15 18
14 10
16 24
Hypergonadotropic hypogonadism –
Mean
22.1
12.5
5.33
14.9
12.5
13.4
SD
5.6
3.3
3.11
4.5
3.3
6.8
Median
21.5
13
5
15
13
11.5
Range
9–29
6–18
1–13
10–25
6–18
6–26
DM juvenile-onset diabetes mellitus, DI diabetes insipidus, OA optic atrophy, D deafness
123
J Endocrinol Invest (2014) 37:195–202 Fig. 1 Family tree of the patients Nos. 1–7 of Table 1; generation are marked on the left with Roman numerals, while individuals are identified with Arabic number; the seven patients are indicated with black symbols; individuals whose DNA was tested are indicated by a dash above their symbol and the proband (IV-9) is indicated by an arrow
197
I 1
2
II 1
2
3
4
5
III
IV
V
1
2
1
2
1
2
3
3
3
4
Fig. 1). The other three patients of this pedigree (Nos. 5, 6 and 7 of Table 1 and V-3, 4 and 5 of Fig. 1) share two common ancestors (I-1 and I-2) and were born to two other first cousin couples (IV-1 and IV-2, IV-3 and IV-4, respectively). This very extended family comes from a small, isolated town in the Nebrodi Mountains. Patients Nos. 8 and 9 of Table 1 are two sisters belonging to another, unrelated family living on the Northeast Coast, with non-consanguineous parents. The remaining three patients (Nos. 10, 11 and 12) come from three other unrelated families living in different areas of the Messina district and were born to healthy non-consanguineous parents. Study design and methods Data concerning clinical evolution of WS from time of diagnosis onwards in these 12 patients were retrospectively gathered from clinical records, according to a standardized form that had been prepared to document the main clinical details, including age at onset of DM and other components of the syndrome, duration of the disease and familial aggregation. Among first-degree relatives, those who gave consent underwent a detailed clinical assessment, including relevant biochemistry, to rule out any features of WS. In all patients, diagnosis of WS was confirmed by genetic analysis, which was performed by studying eight exons and the exon–intron sequences of WFS1 gene. The same genetic analyses were performed also in the 14 parents of our patients. Clinical and biochemical findings at DM onset in the WS cohort (N 12) were compared with those collected in a population of children and adolescents with juvenile-onset, insulin-dependent DM (N 256) aged under 30 who are currently living in Messina district. The analyzed data were: age, HbA1c concentrations and diabetic ketoacidosis
4
4
5
6
5
5
6
7
6
8
9
10
11
7
(DKA) prevalences at DM diagnosis; honeymoon duration and daily insulin requirements at DM onset were also taken into consideration. DKA was defined when blood glucose levels were [250 mg/dl, pH was \7.3 and serum bicarbonate levels were\15 mEq/l. Honeymoon duration was assessed on the basis of the criteria that are generally adopted for definition of partial remission in T1DM: HbA1c concentrations \7.0 % and daily insulin requirement \0.5 IU/kg/day [9]. HbA1c was measured by an immunoassay technique with a monoclonal antibody directed against a sequence of the HbA1c molecule. For statistical purposes data are expressed as mean ± SD or median and range values as appropriate. Comparisons between groups were performed by the Student’s t test (normally distributed data) or the Mann– Whitney U test (non parametric data) as appropriate. Frequency rates were compared by v2-test. The level of significance was set at 0.05. Study design was approved by the Ethics Committee of our Hospital and patients and/or their parents gave informed consent. Appropriate consents for this study were also obtained from the Study Group for Diabetes Mellitus of the Italian Society for Pediatric Endocrinology and Diabetology. Our study was conducted according to the Declaration of Helsinki. DNA extraction and PCR amplification Genomic DNA was extracted from peripheral blood leukocytes by standard phenol chloroform procedure. Fourteen primer pairs were used to amplify the entire coding region of WFS1 gene, including the non-coding exon 1 and the exon–intron boundaries. Exon 8 was subdivided into seven regions (from 8a to 8g), as previously described [10]. PCR amplifications were performed in a
123
198
J Endocrinol Invest (2014) 37:195–202
Gene Amp PCR System 9700 (Perkin Elmer, Foster City, CA) with an initial denaturation step at 94 °C for 15 min and then 30 cycles at 94 °C for 30 s, 60 °C for 30 s, 72 °C for 45 s, followed by 7 min of final extension at 72 °C. WFS5 and WFS8c amplifications were performed at annealing temperature of 62 °C. The specificity of the amplified samples was checked out by agarose gel electrophoresis before further analysis.
Results Main findings at DM diagnosis All these data are summarized in Table 2. The only significant difference between WS series and the overall population with insulin-dependent DM regarded age at DM diagnosis, that was significantly earlier in WS cohort (Table 2). All the remaining data did not significantly differ in the two groups (Table 2).
Sequencing analysis Clinical findings DNA PCR samples underwent direct sequencing using a 3130 Genetic Analyzer (Applied Biosystem, Foster City, CA), according to PRISM Dye Terminator and Dye primer cycle sequencing chemistries. Sequences were compared with reference to human genomic and cDNA WSF1 sequences (GenBank accession No. AC004689 and Y18064). To avoid potential artefacts due to AmpliTaq GoldTM DNA Polymerase (PE Applied Biosystem, Foster City, CA), each sequence alteration was confirmed by sequencing both DNA strands of three independent PCR products. The sequence variants were considered mutations when they: (a) caused a nonconservative amino acid change; (b) were absent in 300 ethnically matched control chromosomes; (c) affected phylogenetically conserved residues. Other DNA variations that did not fulfil these criteria were considered polymorphisms.
Prevalence studies Prevalence rates of WS in the Messina district were calculated by taking into consideration both the total population of our district (653,737) and the population included within the 0–30 year age range (202,681). Furthermore, we also estimated the relative prevalence of WS among patients with youth-onset insulin-dependent DM of the Messina district who are currently aged under 30 years (256). To exclude any bias due to consanguinity we have also taken into consideration, for epidemiological purposes, only the five patients from non-consanguineous unions. Table 2 Diabetic ketoacidosis (DKA) prevalence and median values (and ranges) of age, glycosylated hemoglobin (HbA1c), daily insulin requirement and honeymoon duration at the time of diabetes mellitus Patients
DKA prevalence (%)
Group A (Nos. 12)
8.3
Group B (Nos. 256)
28.1
Age (years)
The minimum ascertainment criteria for diagnosis of WS, i.e. juvenile-onset, insulin-dependent, non-autoimmune DM and bilateral OA were fulfilled in all patients. In all cases, DM was the earliest detected abnormality and presentation occurred during the first decade in 11/12 patients (Table 1). The remaining cardinal features of DIDMOAD association were detected some years later and only the two youngest patients of the entire cohort (Nos. 8 and 9) have not so far developed the full range of DIDMOAD manifestations (Table 1). Other less common clinical features included psychiatric disorders (four cases from the same pedigree), renal outflow tract abnormalities (two cases from the same pedigree), Hashimoto’s thyroiditis (three cases), bilateral cataract (two sisters with the same features) and primary gonadal atrophy (one case) (Table 1). On overall, clinical picture was distinctly more complex and severe in the seven patients from the same pedigree (Table 1), although correlation with genotype was not absolute, as demonstrated by the milder phenotype observed in the patients Nos. 8 and 9 who exhibited the same genotype of patients from 1 to 7 (Table 3). However, it has to be considered that patient No. 9 is still very young. WFS1 mutations and genotypes Direct sequencing of WFS1 gene identified two mutated alleles in all patients. In 9/12 patients, the mutations were homozygous for the same allele, whilst the remaining three patients were compound heterozygotes (Table 3). We (DM) onset in Wolfram syndrome group (Group A) and in that with insulin-dependent DM (Group B)
HbA1c (%)
Insulin requirement (IU/kg/day)
Honeymoon (months)
5.0 (1–13)
10 (8.0–12.8)
0.9 (0.7–1.0)
6.0 (0–12)
9.2 (0.8–22.1)
10.1 (5.5–14.5)
0.7 (0.1–1.8)
7.0 (0–41)
v2
0.13
–
–
–
–
p
2.20
0.001
0.96
0.15
0.53
123
J Endocrinol Invest (2014) 37:195–202
199
Table 3 Mutations of WFS1 gene found in the 12 patients of the present Wolfram syndrome series Patients
Exon
Nucleotide change
Amino-acid change
Type of mutation
Protein domain
Zygosity
1
8b
c1362_1377del16
Y454_L459del_fsX454
Frameshift/truncaton
CD3
Homozygote
2
8b
c1362_1377del16
Y454_L459del_fsX454
Frameshift/truncaton
CD3
Homozygote
3
8b
c1362_1377del16
Y454_L459del_fsX454
Frameshift/truncaton
CD3
Homozygote
4
8b
c1362_1377del16
Y454_L459del_fsX454
Frameshift/truncaton
CD3
Homozygote
5
8b
c1362_1377del16
Y454_L459del_fsX454
Frameshift/truncaton
CD3
Homozygote
6
8b
c1362_1377del16
Y454_L459del_fsX454
Frameshift/truncaton
CD3
Homozygote
7 8
8b 8b
c1362_1377del16 c1362_1377del16
Y454_L459del_fsX454 Y454_L459del_fsX454
Frameshift/truncaton Frameshift/truncaton
CD3 CD3
Homozygote Homozygote
9
8b
c1362_1377del16
Y454_L459del_fsX454
Frameshift/truncaton
CD3
Homozygote
10
5/8c
c532_537del6; c1673G[A
K178_A179del; R558H
Deletion; missense
CD1; CD4
Compound heterozygote
11
8b/ IVS6
c.1328G[T; c.712?16G[A
S443I; –
Missense; splice
TM4; –
Compound heterozygote
12
8b/8c
c1230delCTCT; c.1246delTTC
L410del-fsX441; 516del
Frameshift/truncaton; deletion
TM3-TM4; ER3
Compound heterozygote
Fig. 2 a WFS1 gene, which comprises eight exons; exon 1 is non-coding. b Hypothetical structure of wolframin protein. The cytoplasmic regions are yellow, the endoplasmatic regions are green and transmembrane regions are blue. Mutations found in the patients are boxed
identified four different types of WFS1 gene mutations. The mutations were either missense type (n = 2), or frameshift/truncation type (n = 10), or deletion type (n = 2), or affecting the splice site (n = 1) (Table 3). In wolframin, the mutations were localized in the following regions: three in three cytoplamic domains (CD1, CD3 and CD4); two in the TM3 and TM4 transmembrane domains, and one in the ER3 endoplasmic reticulum domain. One mutation was found in a splicing site of WFS1 gene (Fig. 2). The most frequent alteration in our WS group was a Y454_L459del_fsX454 frameshift/truncation mutation, that was harbored by nine out of 12 analyzed WS subjects
(Nos. 1–9): a 16 base-pair deletion (c1362_1377del16) localized in the CD3 cytoplasmic domain and all these patients were found to be homozygotes for it. Missense mutations were detected in two patients. Patient No. 10 was a K178_A179del/R558H compound heterozygous. K178_A179del mutation was a six c532_537deletion localized in the N-terminal region. It was the only mutation found in exon 5. R558H mutation (c1673G[A), a missense mutation, was found in the CD4 cytoplasmic domain of the wolframin. Patient No. 11 was compound heterozygous for S443I mutation and c.712?16G[A. S443I missense mutation was found at np1328 (AGC[ATC) in exon 8, leading to a
123
200
substitution of serine to isoleucine. It was situated in the TM4 transmembrane region. c.712?16G[A was situated at intron 6. Patient No. 12 was compound heterozygous for a L410del-fsX441/F516del. L410del-fsX441 was a four nucleotide deletion (c.1230delCTCT) (frameshift/truncation mutation) localized in the TM3-TM4 transmembrane regions. F516del mutation (c.1246delTTC) was found in the ER3 endoplamic reticulum region of wolframin. Results of patients’ genetic analyses were supported by results of their parents’ molecular investigations. All the detected mutations had been previously described. Epidemiological data The 12 patients included in the present study are the only WS patients aged under 30 who are currently living in the Messina district, as also confirmed by the other medical Units of the entire Sicilian region, which were contacted for this purpose. Since the total population of children, adolescents and young adults under 30 completed years in this district is currently 202,681, WS prevalence among individuals under 30 of the Messina district might be estimated as one in 16,890. If the total population of the Messina district were considered (653,737), the prevalence of WS in this district would be 1:54,478, i.e. 18.3 cases per million population. If the seven patients from consanguineous unions were excluded, the prevalence of WS among individuals under 30 would be 1:40,536 (1:130,747 in the overall population). According to our findings, the relative prevalence of WS among cases with juvenile-onset, insulin-dependent DM aged under 30 who are currently living in Messina district might be estimated as 1 in 22.3 (1 in 51.2 considering only the WS patients from non-consanguineous unions). The 12 patients in the study have 26 first-degree relatives (14 parents and 12 siblings). All of the 14 parents come from Sicilian pedigrees, as confirmed by a detailed analysis of family background that was carried out on all patients and their first-degree relatives. Both parents of the patients Nos. 1–4, 5–6, 7 and 8–9 (Table 1) were found to be healthy carriers of the same mutation (Y454_L459del_fsX454) that had been detected at homozygous state in the respective offsprings. Each of the parents of patients Nos. 10, 11 and 12 was found to be healthy carrier of one of the mutations that had been detected at compound heterozygous state in the respective offsprings. No clinical and/or biochemical abnormalities that might be compatible with WS diagnosis were detected in the asymptomatic first-degree relatives who underwent a
123
J Endocrinol Invest (2014) 37:195–202
detailed assessment aiming to exclude any features of this syndrome.
Discussion The present investigation covers one of the largest Italian WS populations that have been described until now [10– 12]. One of the purposes of our study was to investigate WS prevalence in a geographic area where consanguineous unions may not be very unusual, particularly in the mountain villages of the hinterland. According to our data, the frequency of WS in the total population of a Northeastern Sicilian district might be estimated as one in around 55,000 inhabitants, which is distinctly higher if compared with figures reported in both UK [4] and North India [6], i.e. 1:770,000 and 1:805,000, respectively, but not too far from that reported in North America, i.e. 1 in 100,000 inhabitants [5]. However, it should be considered that the prevalence estimates in these four epidemiological studies were based on different designs and methods. Moreover, whilst the previous investigations covered very large populations from extended regions, the present one was based on the total population of a limited district area including \700,000 inhabitants. However, the relatively elevated frequency of WS in our district was confirmed by its high prevalence rate in the context of the cases with youthonset, insulin-dependent DM aged under 30 and living in the same area: 1 in 22.3, vs 1 in 150 in the study by Gunn et al. [13] on USA populations. It should be considered that in the present study our population consisted of 12 patients from seven couples and five unrelated families, which makes it difficult to extrapolate unbiased prevalence data. Nevertheless, even though the seven patients coming from the same pedigree were excluded from this epidemiological evaluation and only the five patients from non-consanguineous unions were considered for this purpose, the frequency of WS in the total population of our district might be estimated at around 1:130,000, i.e. very close to that observed in North America and very far from UK and North India figures. Therefore, although we are aware of the risk that referral centres might sometimes overestimate prevalence of rare diseases, we could postulate, on the basis of our data and North-American ones, that WS is not necessarily so infrequent as generally expected according to other epidemiological data [4, 6]. The proportion of patients with the four components of DIDMOAD association in our series was distinctly higher than that previously reported by Barrett et al. [4]: 83.3 vs 53 %. Moreover, it should be underlined that the only two patients of our cohort who have not so far developed the
J Endocrinol Invest (2014) 37:195–202
full range of DIDMOAD manifestations are the youngest ones of the overall series. This finding might suggest that most patients presenting with DM and OA in childhood will afterwards develop the other two components of DIDMOAD association within some years, especially if glycemic control is very poor [14]. In the present study, we have also demonstrated that WS-related DM presents earlier in life than type 1 DM, thus confirming another previous report [14]. This finding might be probably due to the fact that the process of endoplasmic reticulum stress-related deterioration of betacell function in WS patients already starts at the beginning of life, whilst beta-cell destruction in type 1 DM is triggered later [14]. In line with previous reports [2, 4], also in the present series WS-related DM was autoantibody-negative in all cases and not commonly complicated by DKA. All the other investigated parameters (HbA1c concentrations, honeymoon duration and insulin requirements at DM onset) did not achieve a statistical relevance between WSrelated DM and type 1 DM groups, as well as in the report by Rohayem et al. [14]. Apart from the manifestations of DIDMOAD association, some patients of this study exhibited other less common complications, that are generally regarded as an integral part of WS: renal tract abnormalities, psychiatric disorders, primary hypogonadism [1, 4, 6] and even bilateral cataract, which should be considered as a regular component of WS, according to recent reports [15, 16]. These less typical features were observed almost exclusively in the seven patients from the same pedigree. The association with an autoimmune thyroid disease in three patients of our cohort might suggest an autoimmune origin of DM, at least in those cases. Nevertheless, all of our patients had a non-autoimmune form of DM, as also confirmed by the absence of anti-glutamic acid decarboxylase autoantibodies in each of them. In the present series, we have detected four different mutations of WFS1 gene, which were mainly concentrated in exon 8, as in most previous reports [1]. The most frequent mutation was a 16 bp deletion (Y454_L459del_fsX454), that resulted in a frameshift and premature stop codon. Its causal role in the pathogenesis of WS was confirmed by the finding that it was absent in 300 ethnically matched control chromosomes, whereas in all the affected patients with this mutation it was present in a homozygous state. This mutation had been previously reported also in many WS patients from other Italian regions [1] and therefore it cannot be considered as pathognomonic of Sicilian patients. Also the other three mutations encountered in the present series had been previously described in other study populations [10, 11]. However, it has to be considered that novel gene mutations in WS patients are very infrequently encountered and described [17].
201
We conclude that: (a) WS might not be so infrequent as generally expected; (b) in our series, DM presented before 10 years in 11/12 patients and ten cases have already developed all four manifestations of DIDMOAD association by 26 years; (c) 9/12 patients exhibited a homozygous frameshift/truncation mutation (Y454_L459del_fsX454), which is the one most frequently found also in patients from other Italian regions; (d) age at DM diagnosis was significantly earlier in WS group than in the patients with insulin-dependent DM of Messina district. Conflict of interest The authors F. Lombardo, G. Salzano, C. Di Bella, T. Aversa, F. Pugliatti, S. Cara, M. Valenzise, F. De Luca and L. Rigoli declare no conflict of interest.
References 1. Rigoli L, Lombardo F, Di Bella C (2011) Wolfram syndrome and WFS1 gene. Clin Genet 9:103–117 2. Kinsley BT, Swift M, Dumont RH, Swift RG (1995) Morbidity and mortality in the Wolfram syndrome. Diabetes Care 18:1566–1570 3. Cremers CW, Wijdeveld PG, Pinckers AJ (1977) Juvenile diabetes mellitus, optic atrophy, hearing loss, diabetes insipidus, atonia of the urinary tract and bladder, and other abnormalities (Wolfram syndrome). A review of 88 cases from the literature with personal observations on 3 new patients. Acta Paediatr Scand Suppl 264:1–16 4. Barrett TG, Bundey SE, Macleod AF (1995) Neurodegeneration and diabetes: UK nationwide study of Wolfram (DIDMOAD) syndrome. Lancet 346:1458–1463 5. Fraser FC, Gunn T (1977) Diabetes mellitus, diabetes insipidus, and optic atrophy. An autosomal recessive syndrome? J Med Genet 14:190–193 6. Ganie MA, Laway BA, Nisar S et al (2011) Presentation and clinical course of Wolfram (DIDMOAD) syndrome from North India. Diabet Med 28:1337–1342 7. Wolfram DJ, Wagener HP (1938) Diabetes mellitus and simple optic atrophy among siblings: report of four cases. Mayo Clinic Proc 13:715–718 8. Lombardo F, Chiurazzi P, Ho¨rtnagel K et al (2005) Clinical picture, evolution and peculiar molecular findings in a very large pedigree with Wolfram syndrome. J Pediatr Endocrinol Metab 18:1391–1397 9. Lombardo F, Valenzise M, Wasniewska M et al (2002) Two-year prospective evaluation of the factors affecting honeymoon frequency and duration in children with insulin dependent diabetes mellitus: the key-role of age at diagnosis. Diabetes Nutr Metab 15:246–251 10. Colosimo A, Guida V, Rigoli L et al (2003) Molecular detection of novel WFS1 mutations in patients with Wolfram syndrome by a DHPLC-based assay. Hum Mutat 21:622–629 11. Tessa A, Carbone I, Matteoli MC et al (2001) Identification of novel WFS1 mutations in Italian children with Wolfram syndrome. Hum Mutat 17:348–349 12. d’Annunzio G, Minuto N, D’Amato E et al (2008) Wolfram syndrome (diabetes insipidus, diabetes, optic atrophy, and deafness): clinical and genetic study. Diabetes Care 31:1743–1745 13. Gunn T, Bortolussi R, Little JM, Andermann F, Fraser FC, Belmonte MM (1976) Juvenile diabetes mellitus, optic atrophy, sensory nerve deafness, and diabetes insipidus—a syndrome. J Pediatr 89:565–570
123
202 14. Rohayem J, Ehlers C, Wiedemann B et al (2011) Diabetes and neurodegeneration in Wolfram syndrome: a multicenter study of phenotype and genotype. Diabetes Care 34:1503–1510 15. Hansen L, Eiberg H, Barrett T et al (2005) Mutation analysis of the WFS1 gene in seven Danish Wolfram syndrome families; four new mutations identified. Eur J Hum Genet 13:1275–1284
123
J Endocrinol Invest (2014) 37:195–202 16. Titah SM, Meunier I, Blanchet C et al (2012) Cataract as a phenotypic marker for a mutation in WFS1, the Wolfram syndrome gene. Eur J Ophthalmol 22:254–258 17. Rigoli L, Lombardo F, Salzano G et al (2013) Identification of one novel causative mutation in exon 4 of WFS1 gene in two Italian siblings with classical DIDMOAD syndrome phenotype. Gene 526:487–489
