Somatolactin alpha (SLa) is the closest relative of growth hormone (GH) secreted from the pars intermedia of the pituitary of “fish” species. For unknown reasons, tetrapods have secondarily lost the SLa gene during evolution after the lungfish branched off . Although functions of SLa have been extensively studied using various species, results are often conflicting (see refs in ) and its physiological roles in vivo remain largely unclear.
Medaka (Oryzias latipes) is known as a powerful model for studying development and genetics , and there are many kinds of skin-color variants that have long been kept as pets in Japan . One of the variants, color interfere (ci), is colored pale gray (instead of wild-type brown) because of abnormal proliferation and differentiation of pigment cells (chromatophores) in the skin. We previously identified that ci has a frame-shift mutation on the SLa gene  and could rescue the ci phenotype by transgenic overexpression of SLa ; i.e., the pale gray skin of ci became dark brown in the SLa-transgenic strain (Actb-SLa:GFP). Because both ci and Actb-SLa:GFP fish normally develop, grow, and reproduce as wild-type fish under ordinary breeding conditions, SLa seemed to play an essential role only in skin-color regulation. Indeed, it is well known that medaka, and many other fish species, acclimate their skin color to their surroundings (cryptic coloration ) and SLa is actually upregulated in dark conditions [5, 7–9]. More recently, we found that the colors formed by SLa function as sexual traits in medaka (nuptial coloration ). These findings suggest that SLa plays important roles for successful survival and reproduction in nature via skin-color regulation.
A receptor for SLa (SLR) was identified in salmon  (but see ). Unexpectedly, its medaka orthologue was ubiquitously expressed in various organs with the highest level, not in the skin, but in the liver and muscles . Hence, we expected that the pale ci and the dark Actb-SLa:GFP should have additional defects most likely in these organs. From this point of view, phenotypes of the cobalt variant of rainbow trout looked suggestive, because the fish lacks most of the pars intermedia of the pituitary, decreases SLa in the plasma, has pale skin (as ci), and is obese due to excessive fat accumulation in the abdominal cavity . Furthermore, the physiological function of SLa in lipid metabolism has also been suggested in other species [11, 15–17]. As expected, we detected more triglycerides in the liver and muscles of ci than we did in those of wild-type fish . Strangely, however, this fat accumulation in ci could not be rescued in Actb-SLa:GFP, unlike the case for skin coloration . Thus, the causal relationship between SLa and lipid metabolism remains an open question that should be carefully reassessed.
Generally, traits (phenotypes) of animals can be affected by genes (genotypes) and environments. In previous experiments, we used fish that were born and bred in a laboratory, believing that the different strains were under identical conditions and could be used to evaluate genetic effects on phenotypes. Water temperature, water quality, and light cycle were stably controlled by an automated water filtration and circulating system, and we fed fish using the same diets. However, the amount of diet, the speed of water flow into tanks, algae growing on tank walls, excrement remaining at the bottom, or other conditions may not be exactly identical between tanks, and such differences may significantly affect lipid metabolism. Thus, we hypothesized that the unrescued lipid phenotype of Actb-SLa:GFP  reflects not only genetic differences between the mutant/transgenic strains, but also such (seemingly subtle) environmental artifacts, which misled our conclusion regarding SLa’s function (hepatic triglycerides were actually shown to be sensitive to diet ). In this study, we raised various medaka strains under various breeding conditions and measured hepatic triglycerides to reassess the potential role of SLa in lipid metabolism.