Natural selection of human embryos: impaired decidualization of endometrium disables embryo-maternal interactions and causes recurrent pregnancy loss

Abstract

Background: Recurrent pregnancy loss (RPL), defined as 3 or more consecutive miscarriages, is widely attributed either to repeated chromosomal instability in the conceptus or to uterine factors that are poorly defined. We tested the hypothesis that abnormal cyclic differentiation of endometrial stromal cells (ESCs) into specialized decidual cells predisposes to RPL, based on the observation that this process may not only be indispensable for placenta formation in pregnancy but also for embryo recognition and selection at time of implantation.

Methodology/principal findings: Analysis of mid-secretory endometrial biopsies demonstrated that RPL is associated with decreased expression of the decidual marker prolactin (PRL) but increased levels of prokineticin-1 (PROK1), a cytokine that promotes implantation. These in vivo findings were entirely recapitulated when ESCs were purified from patients with and without a history of RPL and decidualized in culture. In addition to attenuated PRL production and prolonged and enhanced PROK1 expression, RPL was further associated with a complete dysregulation of both markers upon treatment of ESC cultures with human chorionic gonadotropin, a glycoprotein hormone abundantly expressed by the implanting embryo. We postulated that impaired embryo recognition and selection would clinically be associated with increased fecundity, defined by short time-to-pregnancy (TTP) intervals. Woman-based analysis of the mean and mode TTP in a cohort of 560 RPL patients showed that 40% can be considered "superfertile", defined by a mean TTP of 3 months or less.

Conclusions: Impaired cyclic decidualization of the endometrium facilitates implantation yet predisposes to subsequent pregnancy failure by disabling natural embryo selection and by disrupting the maternal responses to embryonic signals. These findings suggest a novel pathological pathway that unifies maternal and embryonic causes of RPL.

Figure 1. PROK1 and PRL transcript levels in timed endometrial samples of 10 RPL patients and 20 control (CON) subjects, consisting of 10 fertile (FERT) volunteers and 10 infertile (INF) patients without a history of RPL.

PROK1 (A) and PRL (B) mRNA levels, normalized to L19 transcript levels, are expressed in arbitrary units (a.u.). Horizontal bars indicate the median expression in each group. PROK1 and PRL mRNA levels, respectively (C & D), in RPL patients with recurrent biochemical (biochem; n = 10) or fetal (n = 10) pregnancy failure. A ‘biochemical’ loss was defined as a miscarriage at 4–6 weeks gestation with ultrasound evidence of either an intrauterine pregnancy sac with no fetus or retained products of conception. A ‘fetal’ loss was defined as a pregnancy failure between 6–13 weeks gestation with prior ultrasound evidence of fetal development. Note the logarithmic y-axes.

Figure 2. RPL is associated with aberrant expression of PROK1 and PRL in decidualizing primary ESC cultures.

PROK1 (A) and PRL (B) mRNA levels, normalized to L19 transcript levels and expressed in arbitrary units (a.u.), were determined in undifferentiated ESCs (day 0) or cultures decidualized with 8-Br-cAMP and MPA (cAMP/MPA) for 2, 4 or 8 days. A total of 9 primary cultures were established from RPL patients and 12 from control (CON) subjects. Horizontal bars indicate the median expression in each group. Note the logarithmic y-axes.

Figure 3. RPL is characterised by aberrant hCG responses in decidualizing ESCs.

Primary cultures from RPL patients (n = 10) and control women (n = 10) were decidualized with 8-Br-cAMP and MPA for 72 hours in the presence or hCG or vehicle. The data show the percentage change (± SEM) in PRL (A) and PROK1 (B) mRNA levels upon hCG treatment of control and RPL cultures, relative to the expression levels in cultures treated with vehicle (dotted lines); P<0.005.

Figure 4. Woman-based analysis of time-to-pregnancy (TTP) in 560 women with a history of ≥3 consecutive first trimester miscarriages.

Histogram representing woman-based analysis of the mean (A) and mode (B) TTP in patients with RPL.