Deduced primary structure of two forms of vitellogenin in Japanese common goby (Acanthogobius flavimanus)

Complete nucleotide sequences of two forms of vitellogenin (Vg) cDNA in Japanese common goby were determined from a liver cDNA library of E(2)-treated male fish. These two Vg cDNAs contained complete open reading frames encoding 1664 and 1238 amino acid residues including signal peptides, respectively. From comparison of the deduced amino acid sequences of both Vgs and the partial amino acid sequences of the yolk proteins, the longer sequence was concluded to be cDNA of the Vg-530 and the shorter one was that of the Vg-320 of the Japanese common goby which were reported in our previous paper. The deduced sequence of Vg-530 without signal peptide was arranged by lipovitellin heavy-chain (LvH), phosvitin (Pv), lipovitellin light-chain (LvL), and beta'-component beta'-c) domains from the N-terminus, and showed a range of 40-45% sequence identity to those of other fish. Furthermore, the deduced sequence of Vg-320 showed no obvious Pv domain, has a shortened C-terminal coding region after the LvH domain, and showed a close similarity to the phosvitin-less Vg of zebrafish. Moreover, biochemical analysis of the yolk proteins verified that Vg-530 cleaves into the Lv-Pv complex (molecular mass: 470 kDa) and beta'-c (33 kDa), while Vg-320 showed no change when incorporated into oocytes. The present study demonstrated the existence of the two different forms of Vgs at both the cDNA and protein level, and showed molecular alteration of the two Vgs during vitellogenesis. Two Vg sequence data will aid in designing nucleotide probes for detecting Vg gene expressions as a biomarker of environmental estrogens.