0933: Spatial Transcriptomics Reveal Altered Immune Dynamics Regulating Placental Development In a Humanized-TLR8 Mouse Model of Spontaneous Anti-Phospholipid Antibody Induced Pregnancy Loss

Background/Purpose: TLR8 has been implicated in adverse outcomes of human pregnancy. We previously reported a mouse model of spontaneous aPL-induced pregnancy loss in Sle1 mice expressing a human TLR8 transgene (Sle1.hTLR8tg). These mice experience dystocia, fetal resorptions, and placental developmental abnormalities, with both the Sle1 and hTLR8 loci required for the phenotype. Growth defects occur as early as embryonic day (e)8.5; bulk RNA-seq and flow cytometry revealed immune dysregulation, including decreased expression of IL15, decreased uterine (u)NK cells and increased myeloid cells as well as junctional zone thinning and placental ischemia. It is unclear how hTLR8 initiates the signaling cascade that leads to these changes in immune dynamics within Sle1.hTLR8tg placentas, or how other placental cell populations interact with uterine NK cells to regulate their recruitment and function.

Methods: To better define the spatial relationships during Sle1.hTLR8tg pregnancy, placental tissue was harvested from Sle1 and Sle1.hTLR8tg mice at e9.5 (decidualization and spiral artery remodeling occurring) and e13.5 (fully formed placenta). Spatial transcriptomics (n=2-3 per group, per timepoint) was performed using the 10x Xenium platform and a predesigned Mouse Tissue Atlas panel (379 gene targets) together with a custom panel (89 immune genes, including hTLR8).

Results: K-Means clustering identified major placental regions, including decidua, junctional zone, labyrinth, and mesometrial lymphoid aggregate of pregnancy (MLAp) (Fig.1A). hTLR8 was expressed only in Sle1.hTLR8tg placentas, primarily in decidua and MLAp, and in low amounts in the junctional zone and labyrinth (Fig.1B). hTLR8 expression co-localized preferentially with Itgam (CD11b) (Fig.2A) but also with a small number of stromal cells. At e9.5, Sle1.hTLR8tg placentas displayed clusters of immune cells (Cd8a, Ifng, Ncr1, Klrd1, Klrb1c, Itgam) at the junction of the MLAp and decidua (Fig.2B) not observed in Sle1 placentas and persisting through d13.5, suggesting an inflammatory/cytotoxic phenotype. An increase in placental CD8 cells was confirmed by flow cytometry. NK cell receptors (Klrb1c, Ncr1, Klrd1) and granzymes showed variable downregulation patterns, depending on the sub-type and placental region (Fig.3A-B) Ifng and Cd8a were expressed both in the MLAp and decidua (Fig.3C).

Conclusion: Several unique uNK cell subsets populate mouse and human placentas. The different distribution and variable downregulation of NK cell receptors and granzymes in our spatial data suggests a greater degree of uNK cell heterogeneity than previously appreciated. Future studies include understanding if these NK receptor expression patterns reflect different functions (e.g. differential granzyme production) and looking for potential correlations with human uNK subsets.

Additionally, increased expression of Cd8a and Ifng in the MLAp and their close proximity to clusters of myeloid and uNK cells suggests a pro-inflammatory environment by which myeloid cells are activated via hTLR8 and activate CD8 cells to kill target cells such as stromal and uNK cells that are required for production of key cytokines and placental growth factors.