Transcriptional signature of the decidua in preeclampsia

An emerging concept is that the genesis of preeclampsia is rooted in dysregulated decidualization (1). In PNAS, Garrido-Gomez et al. (2) demonstrate a distinct transcriptional signature in the endometrial stromal cells (ESCs) of preeclampsia patients compared with control cells during in vitro decidualization (before pregnancy) and in vivo decidualization (after pregnancy). This finding has important implications for the development of therapies focused on improving decidualization to prevent severe preeclampsia.

Garrido-Gomez et al. (2) used the secretion of prolactin (PRL) and insulin-like growth factor binding protein 1 (IGFBP1) proteins as a marker of decidualization. We established an in vitro decidualization model in immortalized human ESC cells (ATCC Catalog No. CRL-4003) by treatment with cAMP and medroxyprogesterone acetate for 5 d. RNA-seq analysis revealed 1,359 differentially expressed genes, of which 584 genes were up-regulated and 775 genes were down-regulated upon in vitro decidualization [fold-change > 2 and false-discovery rate (FDR) < 0.05] (Fig. 1A). PRL and IGFBP1 were among the most significantly up-regulated genes, which was further confirmed by quantitative RT-PCR (Fig. 1B). Thus, we concluded that the mRNA level of PRL and IGFBP1 is also a good marker for decidualization. However, in the microarray data from Garrido-Gomez et al. (2), PRL and IGFBP1 were not identified as differentially expressed genes during in vitro decidualization for either ESC cells from preeclampsia patients or control cells, because variance between samples is too large and sample size is too small (n = 3∼7) (Fig. 1C). Additionally, Garrido-Gomez et al. (2) analyzed the transcriptome of decidua basalis (DB) and decidua parietalis (DP) using a laser-microdissection approach in a small number of samples (n = 4). Previously, three other independent studies had also measured transcriptome changes in the DB of preeclampsia patients (3–5) (Fig. 2A). In comparison, no up-regulated genes and only two down-regulated genes were consistent in at least one of these three studies (Fig. 2 B and C). Although there was a difference in sample collection, analysis platform, and data-processing method (2–5), it has been suggested that small sample size is largely responsible for inconsistent findings between microarray studies (6). Due to insufficient statistical power in small sample size studies, true differential genes are often missed and many of the detected ones are false positive. To support this conclusion, we compared in vitro decidualization data (table S4 in ref. 2, 129 differentially expressed genes) and in vivo decidualization data (table S6 in ref. 2, 79 differentially expressed genes in DB; table S7 in ref. 2, 227 differentially expressed genes in DP) from Garrido-Gomez et al. (2). As expected, no obvious overlap was observed (Fig. 2 D and E).

Fig. 1.

(A) Volcano plot for the comparison of RNA-seq data between control and in vitro decidualization (IVD). Nonchanged genes are shown in blue, while differently expressed genes (fold-change > 2 and FDR < 0.05) are denoted in red or green. (B) Validation of PRL and IGFBP1 gene expression by quantitative RT-PCR. The GAPDH gene was used as the reference gene for normalization. Data are presented as the mean ± SEM. Primer sequences: PRL-forward: aagctgtagagattgaggagcaaa; PRL-reverse: tcaggatgaacctggctgacta; IGFBP1-forward: ccaaactgcaacaagaatg; IGFBP1-reverse: gtagacgcaccagcagag; GAPDH-forward: gaaggtgaaggtcggagt; GAPDH-reverse: gatggcaacaatatccactt. (C) PRL and IGFBP1 gene expression upon IVD according to microarray data from Garrido-Gomez et al. (2).

Fig. 2.

(A) Overview of genomic studies performed on decidua basalis from preeclampsia patients. Included are Garrido-Gomez et al. (2), Lian et al. (3), Løset et al. (4), and Yong et al. (5). (B and C) Overlap of up-regulated genes and down-regulated genes among these four studies, respectively. Differentially expressed genes shared in two or more studies are labeled on the Venn diagram. (D and E) Overlap of up-regulated genes and down-regulated genes between in vitro decidualization and in vivo decidualization (DB and DP) from Garrido-Gomez et al. (2), respectively.

In conclusion, we suggest that the transcriptional signature of the decidua in preeclampsia should be tested in a larger set of samples simultaneously to raise statistical power and reliability of the findings. Only then can proper insight into the molecular mechanisms underlying preeclampsia at the decidual side be obtained.