100 THE DYNAMICS OF CONTRACTIONS OF MOUSE BLASTOCYSTS IN VITRO IS TYPICAL FOR EMBRYO QUALITY AND MAY PREDICT HATCHING—A TIME-LAPSE-BASED ANALYSIS

Abstract

Assessing embryo quality has been relying on morphological evaluation of embryos at a few time points during their in vitro culture. Consequently, definitive events (e.g. fragmentations, blastocoel pulsation, synchrony of divisions) were often not noticed and important time points (e.g. cleavages, length of interphases) could not be detected. Thus, morphologically sound embryos selected for transfer often carry reduced competence, resulting in reduced pregnancy and increased abortion rates. Our previous study described the predictive value of the 2nd and 3rd cleavage times and the high rate of false diagnosis at routine morphological assessment (Pribenszky et al. 2010 Reprod. Biomed. Online). Previous literature and our experience reveal that blastocyst pulsation and the event of hatching both happen in vivo as well. Our present study examines the contraction dynamics of in vivo-produced blastocysts and shows its correlation to embryo quality and the probability of hatching. NMRI mouse embryos were flushed at 3.5 days post-coitus and expanded blastocysts were selected. They were distributed to 3 groups: (1) vitrified/warmed, (2) stressed with lethal hydrostatic pressure stress (HP) and (3) untreated. Embryos were transferred individually to the microwells of a well of the well (WOW) culture dish and cultured in groups, under mineral oil at 37°C, 6% CO₂ and 90% humidity in air for 2 days, on the stages of compact digital inverted microscopes designed to be used inside of the incubator, exhibiting automated time-lapse analysis (Primo Vision time-lapse embryo monitoring system, Cryo-Innovation Technologies, Budapest, Hungary). Imaging frequency was set to 10 min. Measured parameters were diameter of the embryo when it reached a local minimum or maximum during pulsation and the relevant time for the given size. The number of measurements per embryo, pulsation frequency, speed of collapses and re-expansion, the time of the hatching and hatched statuses and the beginning of necrosis were recorded. Altogether, 117 embryos were used, in 5 replicates; data were analysed by the Cox regression model. Both HP stress and vitrification reduced the probability of hatching compared with untreated control (61, 80 and 64%, respectively). However, descriptive factors that are characteristic for these statuses were the same in all of the groups. The number of pulsations, the maximal size, the extent of expansions per pulsation and the extent of collapse per pulsation were all significant predictors for the event of hatching. Relative hazard coefficients for the unit increments of these factors were 1.39, 1.03, 1.04 and 0.96, respectively. Data showed that the way a blastocyst contracts is correlated with its quality and to its probability of hatching. Moreover, similar pulsation patterns preceded hatching, independent of the treatment of the blastocyst. The time-lapse system used was capable of detecting these events routinely in our laboratory, while not affecting culture conditions.