The reproductive strategies of vertebrates are diverse. Seahorses (Pisces: Syngnathidae) possess the unique characteristic of male pregnancy; i.e., males, not females, incubate embryos in a specialized structure called a ‘brood pouch’. The brood pouch is formed along the ventral midline of the tail. The lumen of the brood pouch is surrounded by loose connective tissue, called pseudoplacenta, and dermis. We visualized and evaluated the morphology of brood pouch formation in Hippocampus abdominalis to gain generalizable insights into this process in seahorses. First, we employed several staining methods to characterize the pseudoplacenta and dermis of the brood pouch of mature male seahorses. The pseudoplacenta is composed mainly of reticular fibers, while the dermis is composed mainly of collagenous fibers. Further observations showed that pouch formation is initiated by linear projections of epithelia on both ventrolateral sides of the body. These projections elongated toward the ventral midline, eventually fused together, and then formed a baggy structure composed of a single dermis layer with neither smooth muscle nor pseudoplacenta. Finally, the pseudoplacenta was formed, together with two layers of dermis and smooth muscle. Thus, a fully developed brood pouch was established. The morphology of the luminal epithelium also changed during pouch formation. We analyzed the localization of C-type lectins as markers; haCTL II was localized in both the outer and luminal epithelia of the brood pouch throughout development in the male seahorse, whereas haCTL IV, which was not detected in the early stage of seahorse development, became localized only in the luminal epithelium as development proceeded. We categorized the processes of brood pouch formation during male seahorse development into three stages: (1) the early stage, characterized by formation of a baggy structure from the primordium; (2) the middle stage, characterized by the differentiation and establishment of brood pouch-specific tissues; and (3) the late stage, characterized by a fully formed pouch with developing blood vessels and a pouch fold ultimately capable of carrying and incubating embryos.

Duplication and subsequent neofunctionalization of the teleostean hatching enzyme gene occurred in the common ancestor of Euteleostei and Otocephala, producing two genes belonging to different phylogenetic clades (clade I and II). In euteleosts, the clade I enzyme inherited the activity of the ancestral enzyme of swelling the egg envelope by cleavage of the N-terminal region of egg envelope proteins. The clade II enzyme gained two specific cleavage sites, N-ZPd and mid-ZPd but lost the ancestral activity. Thus, euteleostean clade II enzymes assumed a new function; solubilization of the egg envelope by the cooperative action with clade I enzyme. However, in Otocephala, the clade II gene was lost during evolution. Consequently, in a late group of Otocephala, only the clade I enzyme is present to swell the egg envelope. We evaluated the egg envelope digestion properties of clade I and II enzymes in Gonorynchiformes, an early diverging group of Otocephala, using milkfish, and compared their digestion with those of other fishes. Finally, we propose a hypothesis of the neofunctionalization process. The milkfish clade II enzyme cleaved N-ZPd but not mid-ZPd, and did not cause solubilization of the egg envelope. We conclude that neofunctionalization is incomplete in the otocephalan clade II enzymes. Comparison of clade I and clade II enzyme characteristics implies that the specificity of the clade II enzymes gradually changed during evolution after the duplication event, and that a change in substrate was required for the addition of the mid-ZPd site and loss of activity at the N-terminal region. We infer the process of neofunctionalization of the clade II enzyme after duplication of the gene. The ancestral clade II gene gained N-ZPd cleavage activity in the common ancestral lineage of the Euteleostei and Otocephala. Subsequently, acquisition of cleavage activity at the mid-ZPd site and loss of cleavage activity in the N-terminal region occurred during the evolution of Euteleostei, but not of Otocephala. The clade II enzyme provides an example of the development of a neofunctional gene for which the substrate, the egg envelope protein, has adapted to a gradual change in the specificity of the corresponding enzyme.

Background: Hatching enzyme is a protease that digests the egg envelope, enabling hatching of the embryo. We have comprehensively studied the molecular mechanisms of the enzyme action to its substrate egg envelope, and determined the gene/protein structure and phylogenetic relationships. Because the hatching enzyme must have evolved while maintaining its ability to digest the egg envelope, the hatching enzyme-egg envelope protein pair is a good model for studying molecular co-evolution of a protease and its substrate. Results: Hatching enzymes from medaka (Oryzias latipes) and killifish (Fundulus heteroclitus) showed species-specific egg envelope digestion. We found that by introducing four medaka-type residue amino acid substitutions into recombinant killifish hatching enzyme, the mutant killifish hatching enzyme could digest medaka egg envelope. Further, we studied the participation of the cleavage site of the substrate in the species-specificity of hatching enzyme. A P2-site single amino acid substitution was responsible for the species-specificity. Estimation of the activity of the predicted ancestral enzymes towards various types of cleavage sites along with prediction of the evolutionary timing of substitutions allowed prediction of a possible evolutionary pathway, as follows: ancestral hatching enzyme, which had relatively strict substrate specificity, developed broader specificity as a result of four amino acid substitutions in the active site cleft of the enzyme. Subsequently, a single substitution occurred within the cleavage site of the substrate, and the recent feature of species-specificity was established in the hatching enzyme-egg envelope system. Conclusions: The present study clearly provides an ideal model for protease-substrate co-evolution. The evolutionary process giving rise to species-specific egg envelope digestion of hatching enzyme was initiated by amino acid substitutions in the enzyme, resulting in altered substrate specificity, which later allowed an amino acid substitution in the substrate.