In vivo bovine embryos were obtained by nonsurgical flushing of uterine horns of cows submitted to superovulatory treatment, while in vitro embryos were generated from oocytes collected from slaughtered donors. Lucifer Yellow injected into single blastomeres did not diffuse into neighboring cells until the morula stage in in vivo embryos and the blastocyst stage in in vitro embryos. In both cases diffusion was limited to a few cells. In contrast, diffusion was extensive in microsurgically isolated inner cell mass (ICM) but absent in the trophectoderm (TE). At the blastocyst stage, diffusion was always more extensive in in vivo than in in vitro embryos. Ultrastructural analyses confirmed these functional observations, and gap junction-like structures were observed at the blastocyst stage. These structures were diffuse in the ICM of in vivo embryos, scarce in the ICM of in vitro embryos and in the TE of in vivo embryos, and not observed in the TE of in vitro embryos. Blastomeres at all stages of development from the 2-cell stage to the blastocyst stage in in vitro embryos and at the morula and blastocyst stage in in vivo embryos were electrically coupled, and the junctional conductance (Gj) decreased in in vitro embryos from 4.18 +/- 1.70 nS (2-cell stage) to 0.37 +/- 0.12 nS (blastocyst stage). At each developmental stage, in vivo embryos showed a significantly (P < 0. 05) higher Gj than in vitro-produced embryos. Moreover, a significantly (P < 0.01) higher Gj was found in isolated ICM than in the respective blastocyst in both in vivo- and in vitro-produced embryos (3.5 +/- 1.4 vs. 0.7 +/- 0.3 and 2.6 +/- 1.6 vs. 0.37 +/- 0. 12 nS, respectively). The electrical coupling in absence of dye coupling in the early bovine embryo agrees with observations for embryos from other phyla. The late and reduced expression of intercellular communicative devices in in vitro-produced embryos may be one of the factors explaining their developmental low efficiency.

No abstract