Fucosidosis is a rare autosomal recessive lysosomal storage disease caused by α-L-fucosidase deficiency due to FUCA1 gene mutations. Its estimated frequency is below 1 in 200,000 live births, depending on the country [16].
This study was the continuation of the largest Tunisian survey of fucosidosis patients diagnosed during 1987–2007. In Tunisia, the real frequency of this disease is underestimated, considering the significant number of suspected cases through pedigree analysis of affected families [17].
According to the age of onset and the degree of severity, the phenotypic expression of fucosidosis includes two phenotypes: severe (type I or form I) and milder (type II or form II) [1]. The diagnosis of fucosidosis patients was based on the characteristic pattern of urinary oligosaccharides and an enzymatic assay in leukocytes in the studied patients. Based on clinical examination, two patients (P1 and P2) were classified as type I with a severe disease, while one patient (P3) was classified as type II with a milder disease.
According to the clinical data, the confirmation of the diagnosis in all patients with fucosidosis was performed at the mean age of two years, similar to that previously described in the literature [18].
All patients presented with an early onset of psychomotor retardation within the first year of life and developed severe spastic quadriplegia. Severe growth retardation was noted in all patients, and all cases presented variable degrees of dysostosis multiplex on radiological investigations.
In this study, angiokeratoma was observed in the patients (P3 and P1) with type I and type II fucosidosis, respectively. The clinical profile of the patients (P3 and P1)was in agreement with several studies described in the literature [18] . Of note, angiokeratomas do not represent a pathognomonic criterion since this phenotypic description is present in other pathologies, such as Fabry disease and sialidosis [19].The presence of angiokeratoma has been detected in patients with type I who develop faster neurological deterioration leading to early death [20].
Cases of both patients (P1 and P2) were classified as type I disease. Nevertheless, only patient P1developed a faster neurological deterioration that led to an earlier death compared to patientP2 (and P3, who developed type II of the disease).
To the best of our knowledge, we have described the first molecular analysis of FUCA1 in three unrelated patients with fucosidosis. The genotypes of the patients were p.F77Sfs*55/ p.F77Sfs*55, p.K57Sfs*75/p.K57Sfs*75, and p.G332E/ c.662+5g>c.
With regard to the pathogenicity of the novel mutations, the frameshift mutations caused by a single base deletion (p.F77Sfs*55 and p.K57Sfs*75) are located in the glycoside hydrolase catalytic domain of the FUCA1 protein and are both predicted to introduce premature termination of glycopeptides in which the amino acids of the downstream sequence are completely altered. Although the functional test was not further characterized in this study, the two frameshifts, p.F77Sfs*55 and p.K57Sfs*75, were identified in patients P1 and P2, respectively, who did not have detectable FUCA1 activity, which was consistent with the severe observed phenotype. Furthermore, clinical variability was observed in the two patients (P1 and P2) with the two frameshift mutations. The phenotypic heterogeneity seemed to be secondary to unknown factors [20,21].
The third novel alteration, p.G332E, was a nonsense mutation, probably involving damage to protein function, based on the PolyPhen-2prediction algorithm. Additionally, we found that the missense mutation occurred in the conserved domain among human lysosomal sulfatases, and the conserved domains among sulfatases have been known to be essential for the catalytic activity [22]. The p.G332E mutation associated with the novel splice site mutation (c.662+5g>c) has been identified in patientP1.The combination of the c.662+5g>c variant of the missense mutation p.G332Eallows the patient (P1) to present a milder phenotype.
The structure of human FUCA1 was modeled by homology with the crystal structure of the bacterium TM aFuc [13]. Of interest, glycine-332 is a buried residue, and the larger charged glutamate could cause misfolding of the protein. This would be consistent with the marked loss of enzymatic activity and the severe phenotype of the homoallelic p.G332E.
The Gly332 residue is involved in the formation of an extremely structured loop that serves to reverse the direction of the seventh β-strand polypeptide to the eleventh α-helix. Additionally, the p.G332Emutation is located close to the secondary structure of elements carrying the catalytic residues buried at the end of the fourth and the sixth β-strands. Thus, this mutation could prevent the normal folding of the protein as well as its function. In the literature, only one missense p.N329Y mutation has been identified in this conserved loopin the homozygous form [23]. The p.N329Y genetic lesion has already been identified in an Australian patient presenting with a severe phenotype. Gly332 is located near the active site of FUCA1 in a conservative region, suggesting the severity of the mutation. The combination of c.662+5g>c and the missense mutation p.G332E provided patientP1 with a milder phenotype. Consequently, the fourth novel mutation, c.662+5g>c, may provide enough residual activity to avoid a severe phenotype. Of note, only one donor splice site mutation c.954+1G>A identified in intron 5 was detected in a homozygous status in an East Indian-Zambian patient who developed a severe form of fucosidosis[24]. Furthermore, several studies have shown that donor splice site mutations are generally more prevalent than the acceptor splice site variants [25].
Interestingly, the 3D structure analyses have demonstrated that the novel identified mutations (p.F57fs, p.K77fs, and p.G332E) and most of the reported mutations (p.G65D, p.S68L, p.Q82X, p.146fs, p.K156fs, p.E118fs, p.W188X, p.N334, p.E258fs, p.S270fs, p.Y335fs, and p.Y216X) are located in the catalytic domain of the FUCA1 protein (Fig.2D). These are mainly frameshift variations, which affect the helices surrounding the central axis of this catalytic domain. Moreover, among the observed missense mutations, four were close to the catalytic sites, and three nonsense mutations were located on the sides. However, only four nonsense mutations (p.E380X, p.387X, p.G402X, and p.G427X) have been identified in the C-terminal domain of the FUCA1 protein [26] (Fig. 2D).
In addition to these mutations, a large number of FUCA1 sequence variations were identified in the Tunisian fucosidosis alleles (Table 2). The noncoding variations (rs180788085, rs907245739, rs1329117558, rs965877153T, and rs1344267327) and coding variations (p.P10R, p.L172L, p.L194N, p.P213P, and p.Y216F) polymorphisms/sequence variants do not change the disease phenotype because the same polymorphisms cause severe (P1 andP2) and milder patient (P3) phenotypes. However, several findings support the notion that polymorphisms of many genes may play a role in the pathophysiology of a major disease, such as infectious diseases in diabetics, tuberculosis, and leishmaniasis [26]. A large variability in clinical responses is observed, which justifies the crucial role of SNPs in the pathophysiology of these diseases, especially when the polymorphisms are located in the promoter regions [27]. We hypothesize their participation in the regulation of molecular mechanisms.
The characterization of these mutations aims to elucidate the allelic heterogeneity of fucosidosis phenotypic aspects, thereby providing more information on the impact of the mutant residues on the FUCA1 structure. These findings will be of importance in the development of new approaches for therapies in patients with fucosidosis.