Human Cell Atlas study defines cells and their interactions at the maternal-fetal interface of early pregnancy.
A cell atlas study of early pregnancy has been published in Nature as part of the ambitious Human Cell Atlas initiative founded in 2016 by Sarah Teichmann.(1,2) Roser Vento-Tormo and Mirjana Efremova, the lead co-authors, and their colleagues in the UK analysed all the cells found in the area where the fetus anchors to the uterine lining, the so-called 'maternal-fetal' interface. With advanced genomics and refined in silico approaches, they defined key cell populations and their interactions to explain how fetal cells masterfully establish themselves in the uterus to initiate pregnancy.
On its arrival in the uterine cavity, the implantation-ready embryo encounters the primed uterine lining (decidua). The outer layer of embryonic cells, the trophoblasts, penetrate the decidua and interact with numerous cells to build the placenta. These primary interactions aim to nourish and protect the allogenic embryo from maternal immune system attack. Abnormal interactions can manifest in failed implantation or miscarriage, generally common during the first trimester.
The authors profiled the transcriptome (RNA content) of about 70,000 single cells found in the placenta, decidua and blood collected from healthy pregnancies between 6 and 14 weeks of gestation. Grouping cells with identical transcriptomes revealed the cellular subtype heterogeneity at the site of placentation. Collectively, 29 cell populations in the decidua and placenta, mostly unprecedented, were identfied. The transcriptome was then scrutinised using a repository of in-silico predicted cell-to-cell interactions (CellPhoneDB), whereby a cell enriched with a known ligand was predicted to communicate with another cell enriched for that ligand’s receptor.
First revelation was the molecular behaviour of three trophoblast populations. Villous cytotrophoblasts and syncytiotrophoblasts, both lining the placenta, were predicted to interact only with other placental cells to further placenta development and nutrient exchange. Extravillous trophoblasts (EVT), which were identified in placenta and decidua, were predicted to interact with various decidua subtypes to promote invasion, accommodation and immunological acceptance of the embryo.
Further, the study defines three stromal and two perivascular cell populations in the decidua and notes the significance of their distribution. Microscopic experiments showed the outer layer of decidua (compacta) populated by two stromal cell types producing classical PRL and IGFBP1. These were predicted to interact with invading EVTs in the pursuit of immunomodulation. The remaining stromal population expressed perivascular markers and was routed to the inner layer (spongiosa) between the glands where it may support the developing vasculature.
The hallmark of the study is the definition of three new states for decidual natural killer (dNK) cells. Since the first characterisation of dNKs in 1991, scientists have been puzzled by their roles. Now, the authors predict the diverse interactions involving the three dNKs and other decidual or trophoblast cells for facilitation of invasion, immunotolerance and vascular remodeling. Standing out is the highly secretory CD39+ dNK population that monopolises the recognition of HLA-G and the immunosuppressant galectin 9. This specific dNK population, according to the authors, interacts with EVT and stromal cells in the compacta to cease immune activation, hence sorely influencing implantation outcome.
When Vento-Tormo was asked if ablation of this subtype in animals is an obvious step forward to model implantation failure she replied that ‘first we need to ensure that dNK populations and their interactions are conserved in other species’. In response to whether variability owing to the equal treatment of samples of different gestational age troubled interpretation, Vento-Tormo commented that ‘the transcriptomic signature was conserved in each cell population throughout early gestation – only the percentage of cells within each cluster would differ’.
The reference map of maternal-fetal interface is now publicly deposited to help scientists hunt for the treasure trove establishing the fertile ground for implantation.
1. Vento-Tormo R, Efremova M, Botting RA, Turco, et al. Single-cell reconstruction of the early maternal–fetal interface in humans. Nature 2018; 563: 347-353.