Prediction model for aneuploidy in early human embryos based on the transcriptomic signature

To analyze in the same human embryo the complete chromosomal status, gene expression, and morphokinetic, in order to establish an aneuploidy prediction model for early human embryo development. Human zygotes donated to research were thawed and cultured from the zygote stage to the third mitotic division and development monitored by time-lapse imaging. Embryos were then disassembled into individual blastomeres for chromosome and gene expression analysis at the single-cell level. Morphology and kinetic parameters were evaluated using the PN disappearance (PNd) as a reference point. Half of the cells of each embryo were analyzed by array comparative genomic hybridization (aCGH) (Illumina Inc, UK). The other half were analyzed via high-throughput qPCR (Fluidigm Co, USA) by studying the expression of 86 genes selected from literature. R and Babelomics software packages were implemented for functional and expression analysis, and SPSS was used for statistical tests. From a total of 85 embryos (Female age: 33.7 ± 4.3 y.o), gene expression data was obtained in 78 embryos. Complete aCGH results were collected in 57 embryos, with an aneuploidy rate of 50.9%. Both chromosomal and expression results were obtained in 53 embryos. Interestingly, time from PNd to start of first cytokinesis was the only kinetic parameter significantly different in aneuploid versus euploid embryos (2.8 vs. 2.4 h, respectively) (p=0.02). In addition, the time frame between PNd and the subsequent 30 hours was the single period in which the genetic landscape was significantly different between euploid and aneuploid embryos. A total of twenty genes were differentially expressed in aneuploid versus euploid embryos (p<0.05). The twelve genes that showed the greatest differences were selected to define the transcriptomic signature of early human embryo aneuploidy generation. Finally, an aneuploidy predictor model was developed based on this gene expression signature with a proven accuracy of 85.2%. This is the first study to analyze chromosomal status, gene expression, and morphokinetic simultaneously in the same human embryo. The transcriptomic signature of 12 genes allowed the development of a highly accurate model for aneuploidy prediction in early human embryo development with the potential for clinical translation.