Gene expression profile of human endometrial receptivity: comparison between natural and stimulated cycles for the same patients

Abstract

Background: The adjunction of exogenous hormones for controlled ovarian stimulation (COS) may alter endometrial receptiveness. In order to identify the genes misregulated under COS, we compared the endometrium gene expression profiles, from the same patients, in a natural cycle and in a subsequent COS cycle.

Methods: For the same normal-responder patients (n = 21), endometrial biopsies (n = 84) were collected during the pre-receptive (LH + 2) and receptive stages (LH + 7) of a natural cycle and, subsequently, on oocyte retrieval day (hCG + 2) and on transfer day (hCG + 5) of a stimulated cycle. Samples were analyzed using DNA microarrays. Gene expression profiles and biological pathways involved in endometrial receptivity were analyzed.

Results: Although endometrium transition profiles from pre-receptive to receptive phases are similar between patients, COS regimens alter endometrial receptivity in comparison with natural cycle. Under COS conditions, two endometrial profiles were identified and were associated either with a moderately altered receptivity profile for the majority of the patients or a strongly altered profile for a sub-category of patients. The receptive endometrium transcription profile under COS was defective for biological functions such as TGFbeta signaling, leukocyte transendothelial migration and the cell cycle.

Conclusions: Gonadotrophin treatments in COS cycles led to disruptions of the transcriptional activation of genes involved in normal endometrial receptivity. We propose that when the receptiveness of the endometrium is seriously compromised by the COS protocol, fresh embryo replacement should be cancelled, the embryo frozen and thawed embryo replacement should be performed under natural cycles.

Figures

Figure 1. Unsupervised classification with both principal component analysis (PCA) and hierarchical clustering of 84 endometrium samples during the pre-receptive and receptive stages from natural and stimulated cycles

(A) Unsupervised hierarchical clustering was performed, allowing a separation between the pre-receptive (LH+2 and hCG+2) and receptive samples (LH+7 and hCG+5). (B) PCA using the three dimensions was performed representing 43% of the data information, and confirmed the unsupervised cluster analysis. Pink circle, LH+2 and hCG+2 samples; grey circle, LH+7 samples; green circle, hCG+5 samples.

Figure 2. Supervised hierarchical clustering of hCG+2 and hCG+5 samples from stimulated endometrium cycles

(A) The majority of the genes specifically modulated during endometrial receptivity in stimulated cycles (998 genes) were up-regulated (777 up-regulated genes, 221 down-regulated genes). Red, up-regulated genes; green, down-regulated genes. (B) A hierarchical clustering was performed with this gene list, allowing the separation of the two sample groups. Rhomb black outline: “Atypical” samples.

Figure 3. The two endometrium profiles during the pre-receptive and the receptive stages in stimulated cycles

The majority of patients (18/21 patients) were characterized by a common profile, called “standard profile”. Other patients (3/21) presented an “atypical profile”. PR, pre-receptive; R, receptive; red, up-regulated genes; green, down-regulated genes.

Figure 4. Venn diagram of transcripts up-regulated and down regulated during endometrial receptivity in the natural cycle compared to the stimulated cycle

Each list of genes was determined with the SAM software with a FDR 2.

Figure 5. Down-regulated genes related to the cell cycle function in the receptive endometrium of stimulated cycles

Genes involved in (A) the cell cycle, (B) G1/S checkpoint regulation, and (C) the G2/M DNA damage checkpoint. All of these genes were down-regulated in stimulated cycles in comparison with natural cycles. Green: down-regulated genes.

Figure 6. Major differences in main systems involved in implantation between natural and stimulated cycles

To identify differential distributions of KEGG pathways, we performed a FatiGO+ analysis with the two lists of genes up-regulated gene during endometrial receptivity of the natural and stimulated cycles. There were numerous differences, between the natural and stimulated cycles, in main systems involved in the implantation process, such as the TGFβ signaling pathway, the complement and coagulation cascades, and leukocyte transendothelial migration. Red, genes up-regulated in natural cycles; pink, genes up-regulated in stimulated cycles.

Figure 7. Unsupervised classification with principal component analysis (PCA) of the hCG+2 and hCG+5 samples with the predictor list

Our predictor list of the endometrial receptivity allows the separation of 90% of hCG+5 samples from the hCG+2 samples in this PCA using the first two dimensions, representing 39% of the data information.