Identification of Estrogen-Regulated Proteins in Zebrafish Embryos by Quantitative Proteomics

Short description: In this presentation, we introduce quantitative proteomics using zebrafish embryos as a paradigm for testing the potential impact of endocrine disrupting chemicals that represent a class of exogenous compounds interfering with the biological actions of endogenous hormones. Specifically, we aimed at evaluating the effect of deyolking (which is a generally employed procedure to better identify low abundance proteins in embryos) on the results of label-free protein quantification, as well as understanding protein networks affected by estrogens by exposing this model organism to 17β-estradiol (E2). Purpose: To identify estrogen-induced differential protein expression impacted by deyolking in zebrafish embryos using a label-free quantitative proteomics approach. Methods: Along with non-treated controls, zebrafish embryos were treated short-term with 1 ppm of E2. Half of the embryos were subjected to a deyolking procedure. Embryo protein extracts were processed using a bottom-up shotgun proteomics approach. The samples were analyzed using data-dependent LC-ESI-MS/MS on an LTQ-Orbitrap Velos Pro (Thermo) connected to a nano-ACQUITY UPLC system (Waters). MS/MS spectra were searched against a composite UniProt zebrafish protein database using the Mascot and SEQUEST search algorithms within Proteome Discoverer (Thermo Scientific). Label-free quantitation was performed by an MS/MS total ion current approach using Scaffold (Proteome Software), as well as calculating spectral counts. Additionally, the differentially expressed proteins were mapped to networks and biological processes through Ingenuity Pathway Analysis (IPA, QIAGEN). Results: The increase in the zebrafish yolk protein vitellogenin is a well-known marker of estrogen exposure. With the observation of a significant increase in vitellogenin in the treated embryos compared to non-treated control embryos, we confirmed E2’s action for our subsequent proteomics study. Estrogen-regulated proteins were identified using both spectral counting and MS/MS-based total ion current method as label-free quantitative approaches. With p2-fold change as threshold, 74 proteins were differentially regulated by the hormone by combining data from both yolk-intact and deyolked samples. Of these significantly differentially regulated proteins, 3 were found to be unique to spectral counting, while 7 were unique to yolk-intact samples. Three proteins were represented in both yolk-intact and deyolked specimen but to differing degrees of up- and down-regulation. When the differentially regulated 74 proteins were submitted for pathway analysis, 53 proteins were mapped into 1 network that merged into an E2-regulated pathway. We saw repression of several proteins such as ATP synthase alpha- and beta-subunits and eukaryotic elongation factor 2 in E2-treated embryos. However, in other models, these proteins were established previously to be activated by estrogen. Therefore, the deyolking procedure significantly alters the state of the proteome in such a way that it potentially invalidates the results of quantitative proteomics studies. Conclusions: The deyolking of zebrafish embryos to increase protein coverage alters expression data obtained by quantitative proteomics. (Supported by the Robert A. Welch Foundation, BK-0031)