ART culture conditions change the probability of mouse embryo gestation through defined cellular and molecular responses

STUDY QUESTION Do different human ART culture protocols prepare embryos differently for post-implantation development? SUMMARY ANSWER The type of ART culture protocol results in distinct cellular and molecular phenotypes in vitro at the blastocyst stage as well as subsequently during in vivo development. WHAT IS KNOWN ALREADY It has been reported that ART culture medium affects human development as measured by gestation rates and birthweights. However, due to individual variation across ART patients, it is not possible as yet to pinpoint a cause–effect relationship between choice of culture medium and developmental outcome. STUDY DESIGN, SIZE, DURATION In a prospective study, 13 human ART culture protocols were compared two at a time against in vivo and in vitro controls. Superovulated mouse oocytes were fertilized in vivo using outbred and inbred mating schemes. Zygotes were cultured in medium or in the oviduct and scored for developmental parameters 96h later. Blastocysts were either analyzed or transferred into fosters to measure implantation rates and fetal development. In total, 5735 fertilized mouse oocytes, 1732 blastocysts, 605 fetuses and 178 newborns were examined during the course of the study (December 2010–December 2011). PARTICIPANTS/MATERIALS, SETTING, METHODS Mice of the B6C3F1, C57Bl/6 and CD1 strains were used as oocyte donors, sperm donors and recipients for embryo transfer, respectively. In vivo fertilized B6C3F1 oocytes were allowed to cleave in 13 human ART culture protocols compared with mouse oviduct and optimized mouse medium (KSOM(aa)). Cell lineage composition of resultant blastocysts was analyzed by immunostaining and confocal microscopy (trophectoderm, Cdx2; primitive ectoderm, Nanog; primitive endoderm, Sox17), global gene expression by microarray analysis, and rates of development to midgestation and to term. MAIN RESULTS AND THE ROLE OF CHANCE Mouse zygotes show profound variation in blastocyst (49.9–91.9%) and fetal (15.7–62.0%) development rates across the 13 ART culture protocols tested (R2= 0.337). Two opposite protocols, human tubal fluid/multiblast (high fetal rate) and ISM1/ISM2 (low fetal rate), were analyzed in depth using outbred and inbred fertilization schemes. Resultant blastocysts show imbalances of cell lineage composition; culture medium-specific deviation of gene expression (38 genes, ≥4-fold) compared with the in vivo pattern; and produce different litter sizes (P ≤ 0.0076) after transfer into fosters. Confounding effects of subfertility, life style and genetic heterogeneity are reduced to a minimum in the mouse model compared with ART patients. LIMITATIONS, REASONS FOR CAUTION This is an animal model study. Mouse embryo responses to human ART media are not transferable 1-to-1 to human development due to structural and physiologic differences between oocytes of the two species. WIDER IMPLICATIONS OF THE FINDINGS Our data promote awareness that human ART culture media affect embryo development. Effects reported here in the mouse may apply also in human, because no ART medium presently available on the market has been optimized for human embryo development. The mouse embryo assay (MEA), which requires ART media to support at least 80% blastocyst formation, is in need of reform and should be extended to include post-implantation development