Machine learning reveals sex-specific 17β-estradiol-responsive expression patterns in white perch (Morone americana) plasma proteins.

Abstract

With growing abundance and awareness of endocrine disrupting compounds (EDCs) in the environment, there is a need for accurate and reliable detection of EDC exposure. Our objective in the present study was to observe differences within and between the global plasma proteomes of sexually mature male and female white perch (Morone americana) before (Initial Control, IC) and after 17β-estradiol (E2 ) induction. Semiquantitative nanoLC-MS/MS data were analyzed by machine learning support vector machines (SVMs) and by two-way ANOVA. By ANOVA, the expression levels of 44, 77, and 57 proteins varied significantly by gender, treatment, and the interaction of gender and treatment, respectively. SVMs perfectly classified male and female perch IC and E2 -induced plasma samples using the protein expression data. E2 -induced male and female perch plasma proteomes contained significantly higher levels of the yolk precursors vitellogenin Aa and Ab (VtgAa, VtgAb), as well as latrophilin and seven transmembrane domain-containing protein 1 (Eltd1) and kininogen 1 (Kng1). This is the first report that Eltd1 and Kng1 may be E2 -responsive proteins in fishes and therefore may be useful indicators of estrogen induction.