53 TARGETED SCREEN FOR AMINO ACIDS THAT REGULATE BOVINE INNER CELL MASS DEVELOPMENT

Abstract

Pluripotency relies on species-specific amino acid (AA) metabolism. In the mouse, inner cell mass (ICM) and ICM-derived pluripotent stem cells (PSCs) need threonine, which is catabolized by threonine dehydrogenase (TDH) into acetyl–CoA and glycine. Depleting (Δ) the culture medium of threonine (ΔT) or blocking TDH activity induces PSC death. By contrast, human PSCs do not survive without lysine (ΔK), leucine (ΔL), or methionine (ΔM). Since isolated bovine PSCs cannot be propagated in vitro, we screened for AAs that selectively support pluripotent ICM cells in intact bovine embryos. Five days (D5) post-IVF, embryos were transferred into glutamine-free synthetic oviduct fluid (gSOF) with Eagle’s nonessential (NE) and essential (E) AAs (gSOF_AA) plus BSA. Embryos were individually cultured until D8 under different conditions. Statistical significance was determined using Fisher’s exact test for blastocyst development (morphological grading to IETS standard) and t-tests for cell numbers (differential stain) and gene expression (quantitative or qPCR). Removal of BSA reduced grade 1–3 blastocyst (B^1–3) development (37% v. 25%, n = 3; P < 0.001). Depleting NEAAs from gSOF_AA did not significantly decrease B^1–3, but depleting all 12 EAAs did (25% v. 8%, n = 6; P < 0.001). Because ΔEAA was most effective, we focused on this. Experiments were conducted in gSOF+NEAA and compared with gSOF_AA as a positive control (n = 2–6 replicates). One (ΔT, ΔM), two (ΔMT, ΔCM, ΔCT; ΔIL, ΔIK, ΔKL), three (ΔCMT, ΔIKL), or six (ΔHPRVWY) EAA drop-out did not affect blastocyst formation, even when NEAAs were also removed for ΔT and ΔM groups (n = 3). However, depleting another six (ΔCIKLMT), nine (+CMT, +IKL), or eleven EAAs (+T, +M) increasingly compromised B^1–3 (P < 0.05). Because no clear EAA candidates emerged from the screen, we focused on TDH. TDH mRNA was present at similar levels in microsurgically isolated (by microblade) trophectoderm (TE) and chemically isolated (by Triton X-100) ICM, but undetectable in five adult tissues. Despite ΔT medium showing no effect, exposure to the TDH inhibitor QC1 (50 µM) reduced B^1–3 and B^1–2 compared with a dimethylsulfoxide (DMSO) solvent control (25% v. 37% and 8% v. 19%, n = 8; P < 0.005). ICM and TE cell numbers were equally reduced in QC1 v. DMSO-treated blastocysts (10 v. 19 and 37 v. 67 with N = 21 and N = 29 embryos, respectively, n = 3; P < 0.005). Yet TDH, hypoblast (PDGRFα), epiblast (NANOG, FGF4, SOX2), and trophoblast (CDX2, KRT8) markers were not consistently affected by QC1. We next applied 3-hydroxynorvaline (3-HNV), which TDH hydrolyses into glycine and propionyl-CoA instead of acetyl-CoA. Compared with solvent controls, 3-HNV (300 µM) killed all embryos and bovine fetal fibroblasts within 3 days in ΔT medium. This toxic effect was fully rescued by >10-fold T-supplementation. Thus, 3-HNV protein incorporation, rather than acetyl-CoA reduction, may nonspecifically impair cellular function. In summary, we found that bovine ICM formation did not specifically depend on metabolizing threonine or any other single EAA.

Research was supported by AgResearch Core Funding.