Conserved function of medaka pink-eyed dilution in melanin synthesis and its divergent transcriptional regulation in gonads among vertebrates

Abstract

Medaka is emerging as a model organism for the study of vertebrate development and genetics, and its effectiveness in forward genetics should prove equal to that of zebrafish. Here, we identify by positional cloning a gene responsible for the medaka i-3 albino mutant. i-3 larvae have weakly tyrosinase-positive cells but lack strongly positive and dendritic cells, suggesting loss of fully differentiated melanophores. The region surrounding the i-3 locus is syntenic to human 19p13, but a BAC clone covering the i-3 locus contained orthologs located at 15q11-13, including OCA2 (P). Medaka P consists of 842 amino acids and shares approximately 65% identity with mammalian P proteins. The i-3 mutation is a four-base deletion in exon 13, which causes a frameshift and truncation of the protein. We detected medaka P transcripts in melanin-producing eyeballs and (putative) skin melanophores on embryos and an alternatively spliced form in the non-melanin-producing ovary or oocytes. The mouse p is similarly expressed in gonads, but not alternatively spliced. This is the first isolation of nonmammalian P, the functional mechanism of action of which has not yet been elucidated, even in mammals. Further investigation of the functions of P proteins and the regulation of their expression will provide new insight into body color determination and gene evolution.

Figure 1.—

Phenotype of the medaka i-3 mutant. (A) Melanin deposition in the eyes and skin is strongly suppressed in i-3 (bottom) compared with the wild type (top). (B) Phenotype of i-3 fry (middle) is almost indistinguishable from that of the tyrosinase (OCA1) mutant, i-1 (left). Slight deposition of melanin is detectable in the eyes of the MATP/AIM-1/Uw (OCA4) mutant, b^g8 (right), but not in i-3 or i-1. Pink-colored pigment cells are leucophores. (C) Tyrosinase reactions using i-1 (left), i-3 (middle), or b^g8 (right) fry were completely negative in i-1 and differentially positive in i-3 and b^g8. (D and E) Higher magnification of the dorsal parts of b^g8 (D) and i-3 (E) after the tyrosinase reaction. Note that strongly positive cells are distributed in a line in b^g8 but are absent from i-3. Distribution of these strongly tyrosinase-positive cells in b^g8 corresponds to dorsal, lateral, and ventral stripes of fully differentiated melanophores in wild-type fry (F and other data not shown; see Kelsh et al. 2004). All the fry in B–F are 0–1 day after hatching.

Figure 2.—

Positional cloning of the medaka i-3 gene. (A) Syntenic relationship between medaka and human. Sequences of ESTs on medaka LG4 were obtained from the M base, and their human orthologs were defined as the genes with the highest E-values on BlastX, except for OLc49.01g and OLc07.12d. The human orthologs of these two ESTs belong to gene families and we detected multiple human genes with high E-values. The three genes with the highest E-values are listed as putative orthologs of them. No orthologous genes were detected for OLe09.07h or OLe17.08f. (B) Chromosome walking from OLe17.08f to the i-3 locus. The numbers between neighboring loci denote the number of recombinants detected among 843 intercross siblings (1686 meioses). BAC ends mapped are indicated by solid circles. BAC 129C11 contains the entire i-3 mutation candidate region. The insert lengths of the BACs are not reflected in the lengths of the bars. (C) The i-3 mutation detected in the 13th exon. Top, Southern inbred Hd-rR is the wild type; bottom, i-3. Four-base deletion and its position in the i-3 allele are indicated by the red box and arrowhead, respectively. The black box shows a novel stop codon created by the frameshift.

Figure 3.—

(A) Nucleotide sequence of the medaka P cDNA of the HNI strain. Vertical lines show positions of introns in the genomic DNA. Black boxes show translation initiation and stop codons. Red shows deleted nucleotides in i-3. Nucleotides indicated in light blue are polymorphic in HNI, Hd-rR, and i-3 (gray residues in the 5′-most and 3′-most regions were not compared). Putative poly(A) signal is indicated in green. (B) Alignment of amino acid sequences of P proteins so far isolated, from mouse, human, pig (Sus scrofa), and medaka (Hd-rR). Red letters indicate the residues most commonly conserved at that site in the four species. Boxed residues contain exon-intron boundaries within their codons or are the first amino acids of exons. Light blue underlining shows the untranslated residues from P-o (Figure 4D and see text). Residues with black underlining are substituted with a novel seven amino acids (CSSLQSQ) in i-3. Black circles indicate positions of mutated residues reported in human OCA2 patients. Gray circles and open circles show the positions of polymorphic residues detected in human and medaka, respectively. Three of four residues with open circles (T, V, and D) are substituted into A, L, and E in HNI, respectively, while the third one (S) is substituted into G in i-3. Human mutation and polymorphism information was obtained from the Albinism Database (http://albinismdb.med.umn.edu/) and Kato et al. (2003) and Suzuki et al. (2003).

Figure 4.—

Expression of medaka P and mouse p mRNA. (A) Whole-mount in situ hybridization using medaka 6-day b^g8 embryos using P riboprobes (top, antisense; bottom, sense). Positive cells are obvious in the eye and dorsal part of the head. (B) Higher magnification of the trunk. P-positive cells are distributed on the dorsal-most part (red arrowhead) and along the lateral line (yellow arrowhead) and the ventral side (blue arrowhead), which corresponds well to positions of three melanophore stripes in wild-type embryos (Kelsh et al. 2004). (C) Expression of medaka P in adult organs. The broad horizontal line and vertical lines in the diagram of P-m mRNA show the translated region and boundaries between exons, respectively. Black, dark gray, and light gray arrowheads indicate the approximate positions of the primers used for RT-PCR. (D) An ovary- (or oocyte-) specifically spliced form of P mRNA (P-o) occurs in medaka. (Top) Diagram shows the genomic structure of the 5′ region of the medaka P locus. Triangles show the positions of exons 1–9 and an ovary-specific exon (o). Lengths of exons and introns are indicated by black and gray, respectively. RT-PCR using a primer complementary to the ovary-specific exon (open arrowhead in the middle diagram) revealed its exclusive expression in the ovary. (E) Expression of medaka P-m and P-o during embryonic development. cDNAs were synthesized from embryos collected 0–5 days after fertilization and RT-PCR was performed using the primers shown in C and D. (F) Expression of mouse p in adult organs. RT-PCR was performed using two sets of primers that amplify the whole cDNA (black and gray arrowheads) or only the 3′ region (gray arrowheads). Strong expression was observed in the eyeball, ovary, and testis. Identical results were obtained with both sets of primers, except that the relative band intensity of the whole cDNA in testis was weak, suggesting the possible loss of the 5′ region of p mRNA in the testis by alternative splicing (see text).