In vitro fertilization and embryo culture strongly impact the placental transcriptome in the mouse model

Abstract

Background: Assisted Reproductive Technologies (ART) are increasingly used in humans; however, their impact is now questioned. At blastocyst stage, the trophectoderm is directly in contact with an artificial medium environment, which can impact placental development. This study was designed to carry out an in-depth analysis of the placental transcriptome after ART in mice.

Methodology/principal findings: Blastocysts were transferred either (1) after in vivo fertilization and development (control group) or (2) after in vitro fertilization and embryo culture. Placentas were then analyzed at E10.5. Six percent of transcripts were altered at the two-fold threshold in placentas of manipulated embryos, 2/3 of transcripts being down-regulated. Strikingly, the X-chromosome harbors 11% of altered genes, 2/3 being induced. Imprinted genes were modified similarly to the X. Promoter composition analysis indicates that FOXA transcription factors may be involved in the transcriptional deregulations.

Conclusions: For the first time, our study shows that in vitro fertilization associated with embryo culture strongly modify the placental expression profile, long after embryo manipulations, meaning that the stress of artificial environment is memorized after implantation. Expression of X and imprinted genes is also greatly modulated probably to adapt to adverse conditions. Our results highlight the importance of studying human placentas from ART.

Figure 1. Experimental design.

In vivo fertilization and preimplantation development (Control group) and in vitro fertilization and preimplantation embryo development (IVF groups). Oocyte collection was performed in M2 and then in M16. Sperm capacitation and the fertilization step were conducted in M16. Embryo culture was performed in two different culture media: M16 and G1/G2. The timeline in hours (h) is relative to the hCG injection (0 h). For all groups, the blastocysts were transferred in pseudopregnant d3.5 F1 females.

Figure 2. Number of transcripts differentially regulated according to the conditions of fertilization and early embryo development.

The number of transcripts either up- (red) or down-regulated (blue) represents the number of transcripts changed two-fold (pale color) or four-fold (dark color) after comparison of the IVF groups (IVF M16 and IVF G1/G2 groups) versus the control group.

Figure 3. Hierarchical clustering analysis according to the different procedures (IVF M16 and G1/G2 groups and control group).

Figure 4. Histograms of gene clusters identified by DAVID for genes induced (A) and repressed (B) in IVF G1/G2.

Left ordinate represents the number of genes present in each cluster and the right ordinate represents the enrichment score as defined (see text).

Figure 5. Magnitude of induced (up-regulated more than two-fold), repressed (down-regulated more than two fold) and unchanged genes located on the X-chromosome (A) and imprinted genes (B) compared to the complete set of genes in IVF G1/G2 samples (P<0.0001, χ²-test).

Figure 6. Expression profile of gene expression ratios between in vitro cultured embryos and controls on the X-chromosome do not show a clusterization near the Xist gene.

A. Induction ratios (log2) of genes according to their position on the sequence of X-chromosome in the IVF samples. B. Ideogram of Mus musculus X-chromosome and localization of Xist gene.

Figure 7. Comparison of levels of induction for 15 genes between real time RT-PCR and microarrays data.

Comparison of induction ratios between qRT-PCR and microarray data. The PCR primers were chosen at positions corresponding to Affymetrix features for analysis by qRT-PCR. As clearly shown, there is a very good correlation of the results between the two approaches.

Figure 8. Projections of promoters of induced and repressed genes on axes 1 and 8 after PCPCA (Promoter Content Principal Component Analysis).

The axes were defined by PCA from a cross table representing the number of putative transcription factor binding sites in each promoter (60 promoters analyzed corresponding to the 17 most induced and 17 most repressed genes). Among the variable was also added the induction ratio, which permitted to identify the axes that were the most strongly correlated with these ratios (axes 1 and 8). These were chosen for the graphical representation. As clearly shown, there is a very clear dichotomy between induced and repressed promoters when the coordinates calculated for the two axes are used. For clarity, examples of promoters were annotated on the graph.

Figure 9. Correlation of representation of the FKHD factors in the promoters of genes highly induced or repressed in the IVF samples.

By Student t-test analysis, the FKHD binding site was shown to be strikingly overrepresented in induced genes compared to repressed genes. The graph represents promoters of repressed genes on the left side, and of induced genes on the right side. The ordinate corresponds to the number of putative FKHD binding sites found in the promoters using Genomatix™, while the abscissa classifies the promoters according to the Log2 of the induction ratio between the G1/G2 culture conditions versus controls (r = 0.55, P<0.0001, t-test).