• oculocutaneous albinism;
  • tyrosinase;
  • OCA2;
  • TYRP1;
  • SLC45A2

Dear Sir,

Oculocutaneous albinism (OCA) is an autosomal recessive disease of skin, hair and eye hypopigmentation caused by a deficiency in melanin biosynthesis. Four types of non-syndromic OCA are described. OCA1 (MIM203100), with OCA1A and OCA1B subtypes, is caused by mutations in the TYR gene that spans 118 kb in 11q14.3 and contains five exons. About 200 mutations of TYR have been described (Albinism database, OCA2 (MIM203200) is caused by mutations in the OCA2 gene that spans 345 kb in 15q11.2q12 and contains 25 exons. To date, about 80 mutations have been described in the OCA2 gene (Albinism Database). OCA3 (MIM203290) is caused by mutations in TYRP1 leading to rufous OCA. The TYRP1 gene spans 17 kb in 9p23 and is composed of eight exons. Only five mutations were described so far in TYRP1 (Albinism Database). OCA4 (MIM606574) is caused by mutations in SLC45A2 that spans 40 kb in 5p13.3 with seven exons. To date, 30 mutations have been identified in Turkish, German, Korean, Japanese, Brazilian, and Indian patients (Albinism Database).

In this paper, we analyze a series of 63 patients with OCA and describe novel mutations (including intragenic deletions) in the four OCA genes and propose a molecular analysis strategy based on the re-evaluation of frequency of the different OCA types. Genotype/phenotype correlations are discussed.

The search for point mutations was performed either by DHPLC analysis followed by sequencing of the variants or by direct sequencing, depending on the genes and exons. Intragenic deletions were screened by quantitative multiplex fluorescent PCR (QMF PCR) (Niel et al., 2004). Primer sequences and experimental conditions are available from the authors upon request.

Twenty nine OCA patients were found to carry mutations in TYR (46%; OCA1). We identified 31 different mutations including 13 novel ones: three splice site mutations, three nonsense mutation, three missense mutations, one frameshift, one deletion of three nucleotides, one deletion of the entire gene, and one deletion of a single exon (Table 1). Twenty-three patients harbored two mutations (79%). Six patients carried only one TYR mutation (21%), the second one remaining uncovered yet. The potential 5′ distal regulatory sequences including the Terminal Distal Element (TDE), the Locus Control Region (LCR) and a large complex GA repeat were analyzed in these patients (Ray et al., 2007). No mutation was found in the LCR or TDE, and no significant difference in the GA repeat size was observed between the six patients analyzed (size range: 297–317 bp) and 47 controls (size range: 307–317 bp). Our data therefore suggest that a variation in the length of this promoter polymorphism is not involved in OCA pathogenesis.

Table 1.   Novel mutations in the TYR, OCA2, TYRP1, SLC45A2 genes
GenePatient no.Novel mutationAssociated mutationEthnicityPhenotype
Mutation locationNomenclatureMutation locationNomenclature
  1. Novel mutations are in bold. OA: ocular albinism.

  2. aAlso based on positive TYR hair bulb testing.

TYR050813Exon 1c.56A > G/p.His19ArgExon 2c.823G > T/p.Val275PheFranceOCA1A
051409Exon 1c. 98A > C/p.Lys33ThrExon 1c.573delA/p.Gly191GlyfsX35TurkeyOCA1A
070788Exon 1c.488C > G/p.Ser163XExon 3c.1118C > A/p.Thr373LysFranceOCA1A
050180Exon 1c.505_507delCAG/p.Asp169delExon 5c.1467_1468insT/p.ALa489CysfsX19EnglandOCA1A
050333Exon 1c.655G > T/p.Glu219XExon 4c.1205G > A/p.Arg402GlnFranceOCA1Ba
070918Exon 2c.820-1_820GA > TG/p.Ile274_Ser277 > ProfsX23Exon 1c.242C > T/p.Pro81LeuFranceOCA1A
070996Exon 2c.824-?_1007 + ?delExon 4c.1205G > A/p.Arg402GlnFranceOCA1A
040267Exon 4c.1366 + 1 G > A/Not foundFranceNot documented
060058Exon 4c.1185-1 G > TExon 4c.1185-1 G > TMoroccoOCA1Ba
070965Exon 4c.1200G > T/p.Trp400CysExon 4c.1217C > T p.Pro406Leu FranceOCA1B/OA
040742Exon 5c.1366 + 3 A > T/Not foundAsiaOCA1/OCA3
071079Exon 5c.1393_1394insT/p.Lys465XExon 4c.1205G > A/p.Arg402GlnFranceOCA1A
97066Entire gene11q14.3 deletionExon 1c.731_732delGTFranceOCA1A (Coupry et al.,2001)
OCA2030879Exon 2c.182G > A/p.Trp61XIntron 5c.574-19A > GNetherlandsModerate hypopigmentation of skin and iris, nystagmus, photophobia
030857Exons 3–20c.228-?_2007 + ?delExon 9c.1001C > T/p.Ala334ValSpainPale skin, light yellow hair, pendular nystagmus, hyperplastic papillae
070606Exons 3–20 c.228-?_2007 + ?delExons 3–20 c.228-?_2007 + ?delPolandBlond hair, light skin, nystagmus, photophobia, visual acuity 1/10
070122Exon 7c.745C > G/p.His249AspExon 7c.647_807del/p.Ser216_Gln269 > CysfsX23CamerounSkin and hair light brown, nystagmus, depigmented retina
060714Exon 10c.898insG/p.Val300GlyfsX49Exon 22c.2195C > T/p.Ser732LeuTurkeyNot documented
061544Exon 10c.1116 + 6T > CExon 13c.1327G > A/p.Val443IleFrancePale skin, light yellow hair, nystagmus, photophobia, visual acuity 15/100
050372Exon 13c.1349C > T/p.Thr450MetExon 8c.819-822CTGG > GGTCAfricaNot documented
050088Exon 14c.1454G > T/p.Gly485ValExon 7c.647_807del/p.Ser216_Gln269 > CysfsX23FranceCreamy skin, light brown hair, visual acuity from 2/10 to 3/10
030529Exon 17c.1672G > C/p.Ala558ProExon 17c.1672G > C/p.Ala558ProFranceYellow skin and hair, nystagmus, intense photophobia
070764Exon 21c.2106delinsTTC/p.Glu702AspfsX11Exon 14c.1441 G > A/p.Ala481ThrFranceWhite skin, yellow hair, nystagmus, photophobia, visual acuity 2/10
040068Exon 22c.2228 C > G/p.Pro743ArgExon 7c.647_807del/p.Ser216_Gln269 > CysfsX23SpainPale skin, light yellow hair, blue eyes, horizontal nystagmus
030888Exon 25c.2433G > T/p.Arg811Ser Exon 19c.1954T > A /p.Trp652Arg FrancePale skin, yellow hair, blue eyes, nystagmus
TYRP1020513Exon 2c.106delT/p.Leu36XExon 5c.1067G > A/p.Arg356GluGermanyLight yellow skin, yellow gold hair with orange highlights (Rooryck et al., 2006)
040742Exon 4c.780_791del/p.Cys261_Asp264delExon 5c.1057_1060del/p.Ser354ValfsX31AsiaBlond to red hair, creamy skin, nystagmus and albinoid retina
SLC45A2030878Promoterc.-1077_1078delGAExon 7c.1502C > A/p.Ala501AspNetherlandsMild skin and hair phenotype, nystagmus, photophobia
050179Exon 1c.277G > A/p.Asp93AsnExon 1c.277G > A/p.Asp93AsnTurkeyYellow hair and eyelashes and light skin
01045Exon 3c.739A > T/p.Lys247XExon 3c.739A > T/p.Lys247XMoroccoWhite skin, yellow hair, nystagmus, visual acuity < 1/10
050588Exon 4c.889-?_1032 + ?delExon 7c.1532C > A/p.Ala511GluFranceLight yellow hair, nystagmus, photophobia
061105Exon 6c.1166_1167del/p.Lys389SerfsX55Exon 6c.1166_1167del/p.Lys389SerfsX55BelgiumNot documented

A deletion of the entire TYR gene was found in a patient (97066) presenting with a typical OCA1A form and leukodystrophy (Coupry et al., 2001; Goizet et al., 2004). A new heterozygous deletion of a large part of exon 2 was identified in patient 070996.

Two patients had such a mild skin involvement that a pure ocular albinism could be discussed (patient 070965 in Table 1 and patient 060611 with two heterozygous known mutations: p.His367Tyr and p.Arg402Gln). Patient 060058 from consanguineous Moroccan parents, with positive tyrosinase testing on hair bulb, was homozygous for a new splice site mutation (c.1185-1 G > T). This variant might represent a new OCA1B mutation. On the contrary, patient 050813 had a severe phenotype, although he carried one OCA1B mutation (p.Val275Phe) and a novel probably highly pathogenic missense mutation (p.His19Arg) occurring next to the signal peptide, early in the protein.

OCA2 mutations were found in 18 patients (29%). Seventeen patients carried two OCA2 mutations or deletions while one patient carried only one mutation. We identified 13 novel mutations including seven missense, two frameshift, one splicing, one nonsense mutation, and two deletions spanning multiple exons (Table 1). Most of the missense mutations occur in the loops between the transmembrane domains as already described (Spritz et al., 1997). The recurrent ‘African’ deletion including exon 7 was found in six patients. Four patients were from Africa (Angola, Cameroon, Guinea, Zaire), one from Guadeloupe (with African ancestors), and one from Spain (without information about ancestors). All had an OCA2 phenotype with light skin, yellow to light brown hair, and blue to light brown eyes. Several new intragenic deletions were detected in OCA2. Two patients (030857, 070606) carried intragenic deletions including exons 3–20. In the current state of analysis, we cannot assure that the two deletions are formally identical. Patient 070606 was homozygous for the deletion, although no consanguinity was known in her family. She had blond hair and light skin, compatible with an OCA2 skin phenotype, with a quite severe ocular phenotype. It is notable that several deletions of the 15q11.2 region have been reported. Altogether, these data indicate the frequent occurrence of recombination mechanisms in this gene. In a Turkish patient (040578) with OCA and a clinical phenotype of Angelman Syndrome, we identified a heterozygous OCA2 missense mutation (p.Pro198Leu), and a 6 Mb deletion comprising the OCA2 and UBE3A genes (using 15q11.2q12 microsatellites, QMF-PCR, and array-CGH 44K from Agilent Technologies; data not shown) The breakpoints were refined to regions of 600 and 800 kb in the centromeric ‘BP2’, and in the telomeric ‘BP3’ hotspots, respectively (Christian et al., 1999). One patient (030791) may represent a case of digenism. He carried known missense mutations in TYR (p.Thr373Lys) and OCA2 (p.Arg290Gly), both in the heterozygous state. His mother did not carry any mutation whereas his father carried the two mutations. We therefore assume that these two events were not sufficient to lead to the OCA phenotype in this patient and that a third event has occurred and needs to be identified. No other mutation was found so far in the explored regions of each of the four OCA genes in this patient.

Two patients were found to be mutated in TYRP1 (3%) (OCA3), and the four mutations identified were novel ones (Table 1). Patient 02513 was already described by us (Rooryck et al., 2006). Patient 020513 harbored two heterozygous mutations in TYRP1 and one splicing mutation in TYR that abolishes the donor site of intron 4. Her mother carried the TYRP1 mutation in exon 5, and her father carried the TYRP1 mutation in exon 4 as well as the TYR mutation. Both parents were healthy, thus excluding the possibility of a digenism, but not of a triallelism. This patient had a quite classical OCA3 phenotype. Finally, one African patient (050372) had two mutations in OCA2 and one nonsense variant in TYRP1, thus raising the possibility of a triallelism. This variant has been described as a polymorphism (rs41302073, NCBI), which is doubtful for a nonsense mutation. The patient’s mother carried only one OCA2 mutation and the father could not be analyzed. Further investigation of these two cases of possible triallelism would require testing more family members.

Eleven patients were identified with mutations in the SLC45A2 gene (17%) (OCA4). Three patients had a known recurrent mutation (c.986delC) in the homozygous state and five patients carried new mutations (Table 1). The c.986delC mutation is not restricted to the German population as suggested before (Rundshagen et al., 2004) as we found it in two Spanish and one French patients. Patient 050588 carried a novel missense mutation and a deletion of exon 4. This is the first SLC45A2 intragenic deletion to be reported. The patient’s phenotype was quite severe at birth but became milder with age. Finally, patient 030878 carried a missense mutation, and a new variant in the promoter of SLC45A2, a region that was functionally characterized by (Graf et al., 2007). This variant was found in the heterozygous state in two of 100 controls. Functional analyses on RNA extracted from skin melanocytes or hair bulbs are needed in order to confirm the role of this variant. The phenotype in our OCA4 patients is variable and seems to depend upon the mutation type: it ranges from severe (nonsense or frameshift mutations) to milder forms of OCA (missense mutations, mutations in the promoter).

Practically, the strategy for molecular analysis of OCA patients can be partially adapted according to the clinical form of the OCA, to tyrosinase activity measurements (on hair bulb or skin biopsy), and to the patient’s ethnicity. Tyrosinase-negative patients will be tested for TYR first, whereas tyrosinase-positive patients will have OCA2 tested first. In our experience, the tyrosinase activity test on hair bulb proved to be robust only if performed in patients older than 2 yr and if the hair bulb was analyzed within 72 h from sampling (unpublished data). Quite characteristically, one of our patients (061001) with OCA4 was tyrosinase-negative at the age of 1 yr, but turned tyrosinase-positive when she was 3 yr old. The patient’s ethnic origin comes also into consideration in defining the investigatory strategy, since the different forms of OCA are not represented equally in the various populations worldwide. Since OCA1 is more prevalent than OCA2 in Europe, a Caucasian patient from Europe will therefore have the TYR gene tested first. On the other hand, a Black African patient will have OCA2 screened in a first instance, with the prevalent exon 7 deletion being searched for first, and point mutations and other deletions being searched for next.

OCA1A and B subtypes correspond to different TYR mutations. The nonsense, frameshift, and the missense mutations that completely abolish the tyrosinase catalytic activity (probably including the novel one p.His19Arg) cause OCA1A, whereas missense mutations such as p.Val275Phe, p.Pro406Leu, p.Arg422Gln, and the novel splicing mutation c.1185-1 G > T lead to OCA1B. The p.Arg402Gln variant is classically considered as a polymorphism though it has been proved by several authors that it causes the thermo-sensitive retention of TYR in the endoplasmic reticulum of HeLa cells and melanocytes (Berson et al., 2000; Halaban et al., 2000). We found p.Arg402Gln in the heterozygous state in association with another TYR mutation in trans in several OCA patients, and we therefore suggest that this variant should now be considered as a genuine mild mutation. Depending upon the type of mutation it is associated with, p.Arg402Gln can be found in patients with either OCA1A or OCA1B.

SLC45A2 mutations were found to be more prevalent (17%) in our data set than previously described, i.e. 5–6% of OCA in Caucasians (Rundshagen et al., 2004), and 10% in Indians (Sengupta et al., 2007). Other authors reported that only one mutation was identified in one-third of patients (Sengupta et al., 2007), while we identified both mutations in our five OCA4 patient. SLC45A2 should therefore be tested before TYRP1, which represents the least frequent form of OCA worldwide.

To conclude, we have analyzed 63 patients and described 37 novel mutations in the four OCA genes. Two mutations were identified in 52 patients of the 63 from our series (83%). The thorough analysis of the four genes, including the search for intragenic rearrangements, as well as, when available, the analysis of the promoter regions, proved important in order to attain this high yield of mutation identification. It should be noted that 10% of all mutated alleles were deletions. A single mutation was identified in seven cases (six patients with a TYR mutation and one with an OCA2 mutation), thus indicating that mutations remained uncovered either in one of the four OCA1–4 genes or in an as yet unidentified gene(s). In these patients and in those in whom no mutation had been identified, we explored the DCT gene (dopachrome tautomerase), coding for TYRP2, a possible OCA candidate gene. No mutation was found in the 10 exons of the gene.

Although our series of patients is not representative of the worldwide population, it provides a comprehensive study of OCA patients from different origins (mainly Europeans) and allowed us to estimate the relative frequencies of the different forms of OCA: 46% OCA1, 29% OCA2, 3% OCA3, and 17% OCA4. Five percent of cases remain unresolved. Our data highlight the fact that OCA classification best relies on molecular analysis of the four genes as almost identical phenotypes can be observed in OCA1–4.


  1. Top of page
  2. Acknowledgements
  3. References

The authors thank the french Ministry of Research of France, the french Ministry of Health, and the Association Genespoir for their financial support. Part of the experimental work was performed on the Genotyping and Sequencing Facility of Bordeaux which was established thanks to grants from the Conseil Régional d’Aquitaine (no. 20030304002FA and no. 20040305003FA) and from the FEDER (no. 2003227).


  1. Top of page
  2. Acknowledgements
  3. References