In this study, we identified seven different pathogenic variants, three of which are novel, as the cause of OCA1 in ten Pakistani families. Two of the novel missense mutations, p.Pro21Leu and p.Cys35Arg, replaced highly evolutionarily conserved amino acid residues. The p.Cys35Arg mutation was found in two families, and SNP analysis revealed a common haplotype harboring this allele, which indicates a founder effect. The other common alleles found in the OCA1 families were p.Gly419Arg and p.Arg278* (Table 2). Both of these variants have been previously found in other Pakistani families [23, 25, 35]. Although no genotype data are available from the previous studies, our SNP analysis revealed common haplotypes in families sharing the same mutations. Our findings, along with the results of previous studies, indicate that p.Cys35Arg, p.Arg278* and p.Gly419Arg are the three most common mutations causing OCA1 in Pakistani families [23, 25, 35].
Interestingly, two of the novel variants, p.Cys35Arg and p.Tyr411His, might have temperature-sensitive behavior that could be due to a subtle conformational defect or gross protein misfolding . A phenotypic evaluation of individuals homozygous for this allele did not reveal a temperature-sensitive phenotype. The loss of pigmentation in exposed skin areas was not strikingly different than that in less-exposed skin areas (e.g., legs, chest or abdomen); this finding could be due to the effect of the hot local climate (37°C - 52°C) on melanocyte growth and melanogenesis in the skin .
The association of the rs1042602 cSNP of TYR with squamous cell carcinoma of the skin in Caucasians and with pigmentation variation in the south Asian population has been documented [59, 60]. This cSNP results in the substitution of serine with tyrosine at position 192 (p.Ser192Tyr) within the first copper-binding site (CuA) of tyrosinase . Enzymatic analyses have revealed an approximately 40% reduction in the catalytic activity of tyrosinase, due to the p.Ser192Tyr mutation . Interestingly, the distribution of the rs1042602 cSNP alleles varies significantly among different individuals from different geographical origins within Pakistan and thus provides a useful marker for epidemiological studies (Additional file 2: Figure S1).
In the cohort of Pakistani families segregating OCA studied here, fourteen out of forty families have mutations in OCA2. These results indicate that OCA2 is more prevalent than OCA1 in Pakistan in contrast to the Indian population [36, 62]. Of the six distinct mutations in the OCA2 gene, three (p.Asp486Tyr, p.Leu527Arg, c.1045-15 T > G) have not been found in any of the various ethnic populations analyzed to date; therefore, they may be specific to Pakistani albino individuals. Of the two known mutations, p.Pro743Leu was previously identified in individuals of Caucasian, African-American and European ancestry [37, 63–65], whereas the second mutation, p.Ala787Thr, was initially identified in the Chinese population . Another missense mutation, p.Ala787Val, affecting the same codon of the OCA2 gene as p.Ala787Thr, was also reported to cause oculocutaneous albinism , further confirming the necessity of the alanine residue at this position for proper OCA2 protein function.
Affected individuals of five OCA2 families were homozygous for the c.1045-15 T > G mutation. Although the in vivo effects of c.1045-15 T > G are not known, this mutation is expected to produce only the OCA2 isoform without exon 10, which is predicted to encode a protein with no third transmembrane domain. Therefore, the level of normal OCA2 full-length protein required for the transport of tyrosinase to the plasma membrane might be affected by c.1045-15 T > G and thus cause OCA.
We found inter-familial variation in the clinical phenotype among the families segregating the same alleles of TYR or OCA2 (Table 2 and Additional file 1: Table S1). However, no obvious genotype-phenotype correlation was observed. A significant overlap in the range of phenotypes in individuals with TYR (OCA1) and OCA2 mutations was found, which makes genetic screening obligatory for the diagnosis of the type of albinism of the affected individuals. We compared the frequencies of all mutant TYR and OCA2 alleles among our forty OCA families. Mutations in exons 1, 2 and 4 combined accounted for all of the mutant alleles of the TYR gene in our cohort (Table 1). For the OCA2 gene, mutations in exons 10 and 14 collectively accounted for ~67% (10/15) of mutant alleles. Taken together, hierarchical mutation screening of these five exons of OCA genes in the nonsyndromic albino Pakistani population might reveal pathogenic alleles in approximately 43% (95% confidence interval: 28.5 - 57.9%) of cases and would be a cost-effective approach for molecular diagnosis.
Mutations in the protein coding exons or in the splice junctions of TYR
TYRP1 and SLC45A2 and SLC24A5 were not found in 16 families. There are at least three possible reasons to explain these findings. First, cryptic mutations might be present in the regulatory or splicing elements of these genes. Presently, we do not know the locations of the regulatory elements of these genes. Secondly, although family clinical histories and evaluation suggested no other clinical phenotype besides OCA, some of these families might have syndromic OCA. Currently, mutations in at least 12 loci have been causally linked with syndromic OCA [1, 2]. There may also be an additional gene in which mutant alleles cause nonsyndromic OCA. Several previous studies of the four known OCA genes revealed no mutations in some of the affected individuals screened [36, 68, 69]. A need clearly exists for further genetic examination of this disease, as does the opportunity to understand this complex disorder more clearly.