Centrosome fragmentation and malposition is common in the first cell division of equine ICSI embryos

In vitro production of embryos via Intra-Cytoplasmic-Sperm-Injection (ICSI) has become popular in the horse industry. Although improving, the blastocyst yield in clinical practice remains sub-optimal (∼20%: Stout. JEVS 2020; 89:103011). The high incidence of developmental arrest of ICSI embryos is suggested a result of mitotic errors during the first zygotic cell divisions. A better understanding of the cause of these errors could help develop strategies to prevent mitotic errors in equine in vitro produced (IVP) embryos. In somatic cells, the centrosomes, composed of two centrioles surrounded by pericentriolar material are the major microtubule-organizing centers and play essential roles in spindle assembly and chromosome segregation. Mammalian oocytes lack centrosomes and although two centrioles are re-introduced by the spermatozoon at fertilization, centrosomes make only a minor contribution to zygotic spindle assembly (Schneider et al., J.Cell. Biol. 2021; 220(11):e202010106). Although not essential for spindle assembly, the role of the centrosomes in ensuring fidelity of chromosome segregation during zygotic cell division is unclear. The aim of this study was to evaluate the position and number of centrosomes in horse ICSI zygotes during the first mitotic division. Cumulus-Oocyte-Complexes (COCs) were obtained from Warmblood mares by Ovum-Pick-Up, maintained in H-SOF at room temperature overnight and then matured in vitro, fertilized by ICSI and cultured as described previously (Ducheyne et al. Reprod. Fertil. Dev. 2019; 31(12):1830). After 24h, injected oocytes were fixedand immunostained for alpha-tubulin, acetylated tubulin, (peri-)centrosome material (NEDD1) and chromatin (DAPI). For each zygote, confocal microscopy was performed using a Leica SPE II microscope and a 0.17µm interval z stack generated. Of 326 COCs collected, 177 reached MII and were injected and cultured further. Of the 177, 136 zygotes were fixed and imaged, 13 were at the two pronucleus, 11 at prophase/prometaphase, 14 at metaphase, 11 at anaphase or telophase, 54 at the two cell stage and 33 were either degenerated or poorly stained. During the two pronucleus and prophase/prometaphase stages, centrosomes mainly localized at the interface between the two pronuclei. In 56% (n=14/25) of zygotes in pro-metaphase and metaphase, independent spindles formed around each of the two parental pronuclei. The majority of dual spindles were closely adjacent and synchronous. Moreover, in 64.2% (n=9/14) of the metaphase zygotes, the (peri-)centrosomes were either malpositioned (e.g. not at the poles of the spindle) or fragmented. Fragmented or multiple centrosomes are known to predispose to abnormal chromosome-microtubule attachment and aberrant chromosome segregation. We therefore hypothesize that malposition and fragmentation of the centrosomes contributes to mitotic errors and developmental arrest of equine ICSI embryos. Acknowledgements to CAPES, Brazil, for financial support (Code 88882.433332/2019-01 and 88887.467895/2019-00)