Generation of a white-albino phenotype from cobalt blue and yellow-albino trouts: inheritance pattern and chromatophores analysis

Albinism is the most common color variation described in fish and is characterized by a white or yellow phenotype according to the species. In rainbow trout Oncorhynchus mykiss, aside from yellow-albino phenotypes, cobalt blue variants with autosomal, recessive inheritance have also been reported. In this study, we investigated the inheritance pattern and chromatophores distribution/abundance of cobalt blue trouts obtained from a local fish farm. Based on crosses with wild-type and dominant yellow-albino lines, we could infer that cobalt blue are dominant over wild-type and co-dominant in relation to yellow-albino phenotype, resulting in a fourth phenotype: the white-albino. Analysis of chromatophores revealed that cobalt blue trouts present melanophores, as the wild-type, and a reduced number of xanthophores. As regards to the white-albino phenotype, they were not only devoid of melanophores but also presented a reduced number of xanthophores. Cobalt blue and white-albino trouts also presented a more elongated body shape and, most remarkably, a smaller pituitary gland compared to wild-type and yellow-albino, suggesting that the allele for blue color is somehow linked with this abnormal pituitary phenotype. These phenotypes represent interesting models for research on body pigmentation in salmonids and on the mechanisms behind endocrine control of color patterning.

In rainbow trout, different color variations have been described such as the albino, 57 palomino, and cobalt/metallic blue phenotypes [11,12]. Regarding albinism, genes such as tyr-1, 58 tyr-2, and slc45a2 were shown to be implicated in pigmentation [9,10]. However, the 59 mechanism or cells behind the blue phenotype is still unresolved. In majority of fish species, 60 the blue color is caused by a light scattering phenomenon in purine and pteridine platelets in 61 the iridophores. When combined with xanthophores other colors (e.g. green) can be generated 62 whereas in their absence the blue color becomes prominent [5]. Indeed, a previous study 63 reported a reduced number of xanthophores in blue trouts [13], suggesting that blue phenotype 64 might be related to iridophores and the density of xanthophores in the skin. 65 The inheritance of albino phenotypes is determined mainly by recessive genes [11], 66 but there is also a case of a dominant strain [14]. In case of blue phenotype, Blanc

99
To verify the inheritance of these phenotypes in relation to others, both females and 100 males of cobalt blue (indicated as B+) rainbow trout were crossed between themselves and with 101 the following strains: wild-type (bb) and dominant yellow-albino (AA). Gametes were 102 manually stripped by applying gentle pressure along the abdomen and inseminated using a  Olympus, Japan). Images were captured with a CCD Camera (DP73) and analyzed by CellSens 113 software (ver. 1.12; Olympus, Japan). The number of xanthophores from five to seven fields 114 were counted for each color phenotype. The area of those cells (n = 15 cells per color 115 phenotype) was measured using Image J software. Analysis of pituitary gland among four cobalt blue and wild-type phenotype 125 The pituitary gland was carefully dissected from juveniles (about 6 months) and 126 sexually mature females and males (about 2 years) from the four color phenotypes (Table S1).

127
The relative weight was calculated as pituitary weight (g) divided by total body weight (Kg). difference between groups was analyzed by One-way ANOVA followed Bonferroni's test using 133 GraphPad Prism (v.5.00; GraphPad Software, San Diego, CA, USA) or by Student's T test.

134
Differences were considered as significant for p<0.05.

137
Color inheritance in crosses among blue, wild-type, and albino trouts 138 Crosses between blue female vs. wild-type male and wild-type female vs. blue male 139 generated balanced proportions of both cobalt blue and wild-type phenotypes (Table 1). For 140 three crosses between blue individuals (crosses #4, #5, and #6), the proportion between blue 141 and wild-type was approximately 3:1, as supported by Chi-square test (Table 1). In the cross 142 between blue female and dominant yellow-albino male, about half of the fish were yellow-143 albino as the mother while another half was white-albino ( Figure 1A).  The F1 cobalt blue trout generated from crosses #4, #5, and #6 were used for progeny 163 tests. Single-pair crosses between cobalt blue females (n=10) and wild-type males, and between 164 wild-type females and cobalt blue males (n=10) generated eight all-cobalt blue progenies. The 165 remaining 12 crosses yielded mixed-populations of blue and wild-type offspring (Table 2), 166 whereby in six crosses the proportions did not deviated from 1:1 (p ≥ 0.05); in the remaining 167 six, the proportion of wild-type were higher than that of cobalt blue in four crosses. In a cross between F0 cobalt blue and F1 white-albino trouts, white-albino, cobalt blue, 176 yellow-albino, and wild-type trouts were obtained (Table 3; Figure 2A) in a proportion that did 177 not differ significantly from 3:3:1:1 (Chi-square test; p < 0.05). In the intercross between F1 178 white-albino trouts, white-albino, yellow-albino, cobalt blue, and wild-type trouts were 179 obtained (Table 3; Figure 2B) in a proportion that did not differ significantly from 9:3:3:1 (Chi-180 square test; p < 0.05), respectively, as expected from a cross between double heterozygous, 181 based on Mendelian segregation. The F1 yellow trouts were also intercrossed, yielding yellow 182 and wild-type trouts (Table 3; Figure 2B) in a proportion that did not differ significantly from   display recessive inheritance and impairments related to growth or reproduction aspects [12,14].

249
In this study, we investigated the color inheritance and analyzed the density and types of xanthophores described for fish, such as pax7a, slc2a15b, and slc2a11b [32], that could be 294 implicated in the expression of blue color in our cobalt blue strain, but it seems that this color 295 phenotype is caused by a mutation in a gene related to neuroendocrine pathway that somehow 296 affect chromatophores rather than in a gene directly involved in chromatophore specification.

297
This assumption is based on differences found in body morphology and also in the size of 298 pituitary, that was inferior not only in cobalt blue but also in white-albino trouts that also carry 299 the allele responsible for blue phenotype (B). Although a more detailed analysis is still required,  In conclusion, this study shows that the cobalt blue rainbow trout presents dominant