Assistant Professor Penn State University Park, Pennsylvania, United States
Abstract Authors: Camilla H. K. Hughes1 1Department of Animal Science, Penn State University, University Park, PA
Abstract Text: The corpus luteum (CL) is a transient endocrine gland that produces progesterone and maintains pregnancy in all mammals. In the absence of a pregnancy, it regresses in response to prostaglandin (PG)F2A, whereas in the presence of an establishing pregnancy, it is rescued. Although PGF2A is the primary in vivo luteolysin, it fails to kill luteal steroidogenic cells in vitro. This draws into question the mechanisms by which PGF2A elicits its luteolytic effects and suggests that cell-cell interaction is essential. To elucidate mechanisms of luteal cell-immune cell interaction during luteolysis, we evaluated changes in gene expression using 10X Genomics single cell sequencing. Eight cows were synchronized using the PG3G Presynch-Ovsynch protocol. On day 9, half of the cows were given an injection of PGF2A four hours prior to collection of CL. CL were collected by transvaginal surgery. For each replicate, two animals were pooled, and two replicates were performed for each treatment group. CL were dissociated using collagenase and total immune cells were isolated using anti-CD45 magnetic beads, then enriched into the cell suspension at a 1:1 ratio, to provide high-resolution information about luteal immune cell populations. In total, 4,645 midcycle and 4,745 regressing luteal cells were analyzed using Seurat 5.3 in RStudio. Interestingly, small luteal steroidogenic cells (markers LHCGR, STAR, HSD3B1) from midcycle CL formed a cluster that was distinct from those derived from regressing CL. This was in contrast to other cell types, which did not cluster separately in the two conditions, suggesting that steroidogenic cells are the first to undergo substantial transcriptional changes during luteolysis. Due to the size of the fluidics for the single cell assay, most large steroidogenic cells were filtered out, but a few were identified (markers PTGFR, STAR, HSD3B1). Endothelial cells (markers VWF, CD34, CDH5), stromal cells (markers COL1A1, COL11A1, MFAP5) and pericytes (markers RGS5, PDGFRB) were also identified. Immune cell clusters, ordered by relative abundance, included T cells (marker CD3), M1 proinflammatory macrophages (markers CD68, IL1B, TNF, NOS2), dendritic cells (markers XCR1, CLEC9A, CD1B), natural killer cells (NK; markers NKG2A, CD335), and M2 resolving macrophages (markers CD68, CD163, CD36, IL10). Differential expression and functional analysis was performed separately on each cluster. Functional changes associated with steroidogenic cells undergoing luteolysis included modulation of fibrosis and a dramatic upregulation in interferon response, suggesting that these cells are exposed to interferons. T cells and NK cells expressed IFNG, while IFNA was not detected. This suggest that IFNG produced by luteal T cells drives this cytokine response in steroidogenic cells and may induce apoptotic changes. Interestingly, there were greater than 3-fold more M1 proinflammatory macrophages in the CL undergoing luteolysis than in the midcycle CL. During luteolysis, these M1 macrophages also appear to be functionally changed, with increases in mRNAs encoding cytokines IL1A, IL1B, IL23A, and CCL2. In contrast, M2 resolving macrophages did not change in number during luteolysis and their functional changes were not associated with cytokine production, but instead with proteosome function and perhaps antigen presentation. In summary, these data definitively support previous indirect evidence suggesting that changes to cytokine signaling, driven by luteal resident T cells and macrophages, may be important mediators of early luteolysis. Moreover, we have identified T cells and NK cells as the source of luteal IFNG and provided evidence that an increase in proinflammatory M1 macrophages may drive luteolytic changes.
