Results

CLINICAL DATA

The pedigree of family 74 segregating CN is shown in Figure 1.

PATIENT 1 (IV.1)

The index patient was diagnosed with OA1 at age 3 months because of typical clinical features: CN, iris translucency, macular hypoplasia, fundus hypopigmentation, normal pigmentation of skin and hair (Fig 2A), and typical carrier signs of OA1 in his mother and maternal grandmother (see below). At age 2.5 years VA was 0.12 bin (Lea test, sc). At age 4 years his visual acuity had improved to 0.4 Lea test bin (RE plano/−2.5 cyl A 170°, LE plano/−2.5 cyl A 10°). Fundus and iris pigmentation had also increased (Fig 3A and D). Flash albino VEP showed clear asymmetry—that is, much smaller amplitudes on the left occipital hemisphere when stimulating the right eye.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

External aspect of (A) patient IV.1 at age 2.5 years and (B) patient III.3 at age 54 years. Tanning is obvious in both patients. Patient III.3 additionally has very dark hair colour.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Testing of iris translucency in patient (A) IV.1 and (B) III.3, and iris pigmentation in patient III.3 (C). Patchy hypopigmentation can be seen in the iris stroma of patient III.3 but no obvious translucency can be detected compared with patient IV.1. Fundus appearance showing macular hypoplasia and varying degrees of hypopigmentation in patients IV.1 (D), III.3 (E), II.3 (F), and the carrier III.2 (G) (mother of patient IV.1).

PATIENT 2 (III.3)

In the maternal uncle of patient IV.1 aged 51 years there were clear macular hypoplasia (Fig 3E), patches of hypopigmentation of the iris stroma (Fig 3C) but no iris translucency (Fig 3B) or fundus hypopigmentation (Fig 3E). It is interesting to note that both hair and iris stroma (outside the hypopigmented patches) were heavily pigmented (Fig 2B, Fig 3C). His VA was 20/200 in both eyes (RE −2.75 sph/−4.75 cyl A 35°, LE −3.0 sph/−4.25 cyl A 170°). Nystagmus had been present since infancy.

Histological examination of a naevus biopsy revealed macromelanosomes.

PATIENT 3 (II.3)

The maternal uncle of patient III.3 reported to have CN since early childhood. He was originally highly myopic (−18 dioptres in both eyes). At age 78 years, he had bilateral posterior chamber IOL implantation. At presentation he was 79 years old. His VA was then 20/200 in both eyes (RE +2.0 sph, LE 2.0 sph). The iris did not show any translucency. The fundus appearance was typical for excessive myopia including macular changes (Fig 3F). Nystagmus was still present at examination.

CARRIER III.2

Her visual acuity was normal (RE −3.5 sph = 0.9, LE −3.5 sph = 1.0) at age 37 years. There was no iris translucency but hypopigmented patches of the iris stroma in both eyes similar to those in her brother III.3. On funduscopy, there was a typical mosaic pattern of hypopigmented RPE intermixed with normally pigmented RPE (Fig 3G).

CARRIER II.2

At examination she was 76 years old. She was bilaterally aphakic. Her VA was 0.8 in both eyes (RE +12.5 sph, LE + 13.25 sph/−4.0 cyl A 75°). She had clear iris translucency, and on funduscopy there was a typical mosaic pattern of hypopigmented RPE intermixed with normal pigmented RPE similar to that in her daughter (III.2).

MOLECULAR GENETIC ANALYSIS

SSCP analysis showed an aberrant banding pattern for the exon 6 amplimer of the OA1 gene. Direct sequencing of the PCR product revealed a 14 bp deletion starting at nt816 and covering the splice donor signal (gt) (Fig 4C). (Nucleotide numbering based on the cDNA reported by Bassi et al,2 Genbank: NM_000273.) The mutation abolishes a restriction enzyme recognition site for Tsp509I. A loss of the restriction site could be shown in all patients. The heterozygous carrier status could be shown in the mother of patient IV.1 (Fig4A).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4

(A) Tsp509I restriction enzyme digest. The wild type sequence is cut once to produce fragments of 157 bp and 69 bp. This site is lost by the 816del14bp mutation. Lanes: (1) control, (2) patient II.3, (3) patient IV.1, (4) carrier III.2, (5) patient III.3, (6) control. (M) Marker: 100 bp ladder. (B) RT-PCR using RNA from (1) control melanocytes, (2) patient III.3 melanocytes, (3–5) preparations from a control donor eye (3) retina, (4) RPE and choroid, (5) RPE with choroid contamination. (M) Marker: 100 bp ladder. (C) Schematic showing the effect of the 816del14bp mutation on RNA splicing. The mutant gene is shown in grey.

FUNCTIONAL PROOF

Since the mutation covers the splice donor of intron 6, we assumed an effect on splicing, as the flanking sequence provides enough information to produce an alternative splice site (Fig 4C).

To evaluate this splice site RT-PCR was performed to amplify a product of 349 bp (Fig 4B). A product of the expected size could be detected in a cDNA preparation of RNA from melanocytes (lane 1), retina, RPE, and choroid + RPE (lanes 3–5) from control individuals. RNA of melanocytes from a naevus of patient III.3 showed a PCR product of reduced size (lane 2). Direct sequencing of the RT-PCR product of patient III.3 and a control sample revealed a mRNA which used nucleotides 814–815GT as splice donor signal (Fig 4C) in the RNA of patient III.3. Therefore, the open reading frame (ORF) of the resulting mRNA comes into a frame shift predicting the termination of the resulting gene product at amino acid 259 (Fig 4C).

Discussion

We report on a novel mutation (816del14bp) in theOA1 gene.

OA1 codes for a seven helix transmembrane G protein coupled receptor (GPCR) of melanosomal membranes. GPCRs transduce signals over membranes and activate second messenger signalling cascades by coupling to GTP binding proteins. The function of OA1 is not defined currently. It may represent a receptor for a G protein which is involved in regulating the biogenesis and maturation of melanosomes (Fig 5(1) and (2)).3

OA1 patients present macromelanosomes as a typical feature. Garneret al10 hypothesised macromelanosomes to result from promelanosomes which do not separate correctly from the endoplasmatic reticulum and Golgi apparatus. This notion is supported by a recent animal model of Oa1knockout mice.11 Ultrastructural analysis of the RPE of these mice showed normal melanosomes from E12.5 until birth. Macromelanosomes started to form at P1. Individual macromelanosomes were absent in adult skin, hair and uveal melanocytes11which showed darker pigmentation than the RPE.

We conclude that mutated OA1 causes late budding and may lead to increased melanosome size—that is, macromelanosomes which accumulate at the site of their origin owing to impaired budding signals. The same defect may impair active fusion of melanosomes and the plasma membrane (Fig 5(3)) resulting in a transient hypopigmentation as seen in patient IV.1.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5

Schematic indicating maturation steps of melanosomes possibly impaired by mutated OA1. Budding of vesicles from ER and Golgi membranes is controlled by small G proteins (Ras or Rab) which may represent ligands to the GPCR OA1. Mutations in OA1 would cause a delay in budding and, therefore, vesicles with increased size. (2) The same group of G proteins controls targeting of vesicles (melanosomes) to intracellular compartments and the plasma membrane. Mutations in OA1 would cause intracellular misrouting of vesicles. (3) As a third function of a transmembrane GPCR, interaction of OA1 with an acceptor molecule on the plasma membrane can be proposed. Absent or altered OA1 would cause impaired DOPA excretion by impaired membrane fusion. An alternative way of excretion would be passive lipid mediated membrane fusion which rarely occurs spontaneously at membrane surfaces but requires more time. Any of the steps mentioned above cause impaired excretion of DOPA and even if passive membrane fusion or vesicle targeting occurs, it will not be possible to excrete sufficient amounts of DOPA in time.

The mutation in the OA1 gene reported here affects the splice donor site of intron 6 including the complete splice signal. An influence on splicing could be shown in patient III.3 (who showed the less severe phenotype as to ocular hypopigmentation). In a preparation of cDNA from his dermal melanocytes a PCR product of the OA1 mRNA was detectable but showed an aberrant size. Since the splice donor signal was deleted nucleotides 814/816 were used as a new splice donor signal. The newly created splice signal has a quality equal to the optimal splice signal as reported by Shapiro et al.12The resulting mRNA showed a functional frame shift abolishing the correct translation of the mRNA after Leu251 and predicting the termination at codon 259. The resulting gene product lacks about one third of the amino acids including the last transmembrane domain from the extracellular loop 3 up to the C terminus.

Further examples of GPCRs relevant in the eye are rod and cone visual pigments which perceive light and activate the G protein transducin to start a second messenger cascade. Mutations in these are involved in retinal disorders. Rhodopsin interacts with the α subunit of transducin at its fourth cytoplasmic loop.13 Deletions downstream of the third extracellular loop of rhodopsin cause ADRP.1415 We therefore assume that this novel hypomorphic OA1 gene mutation impairs the interaction of OA1 with the associated G protein, thus abolishing vesicle budding (Fig 5(1)), vesicle targeting (Fig 5(2)), or active membrane fusion of melanosomes and the plasma membrane (Fig 5(3)).14

Melanosomes generate melanin from tyrosine via 3,4-dihydroxyphenylalanine (DOPA). DOPA concentration is reduced when tyrosinase activity (oculocutaneous albinism, OCA1) is reduced, or when reduced P-gene activity abolishes supply of tyrosine (OCA2). DOPA is an antimitogenic factor involved in cell cycle regulation and crucial for the maturation of the retina and the optic nerve.16 The neuroretinal pathology of OCA is caused by the lack of DOPA, thus giving an overdevelopment of retinal cells and cells of the optic nerve.16 This results in a developmental injury during embryogenesis causing a neuroretinal phenotype which cannot be reversed during postnatal development. In OA1, the studies summarised above imply that melanosomes do not actively excrete the formed DOPA, thus not reaching a DOPA concentration critical for the development of the neuroretina in due course—that is, during embryogenesis. After birth neuroretinal cells will cease to proliferate. Therefore, the neuroretinal phenotype is permanent in all three forms of albinism even if there is increasing pigmentation postnatally. In OA1 melanocytes form melanin throughout life in contrast with OCA. This will increase pigmentation over time in the iridal pigment epithelium (IPE), RPE, and skin, and also in the uveal melanocytes.11 Marked postnatal activity could be shown in the iridal stroma of cynomolgus monkeys after induction by latanoprost,17 thus explaining the decrease of iris translucency in our youngest patient. The variable degree of hypopigmentation in OA1 patients may be explained by the complex interaction of the genes involved in melanin synthesis.

Our hypothesis on the pathogenesis of the phenotype associated withOA1 mutations is additionally supported by other occasional reports of macular hypoplasia without clinically detectable hypopigmentation as the only sign of OA1 in African-American, Japanese, and also one white patient.18-22 It shows that visual acuity may be greatly impaired even in the absence of macular hypopigmentation although from literature data it is evident that there is a trend towards better visual acuity with increasing pigmentation.20 The lack of hypopigmentation may impede the recognition of OA1 in some older patients, and can only be overcome by carefully searching for additional signs such as macular hypoplasia and albino misrouting.