First trimester infection with Listeria monocytogenes is characterized by decidual and placental inflammation in pregnant cynomolgus macaques

s / Placenta 45 (2016) 63e133 65 how heterogeneous are trophoblasts? Is there a stem cell population analogous to the ones isolated from mouse models? What signals drive differentiation into various cell lineages? I aim to identify transitory populations of cells within the human placenta that organize and direct organogenesis. Methods: Candidate trophoblast progenitor populations were identified through a high throughput flow cytometry screen of 370 CD antigens in trophoblast enriched fractions isolated from either week 6 or 10 placentas. Dynamic subpopulations were selected for further validation through immunohistochemistry and targeted flow cytometry, and global gene expression. Results: Flow cytometry data was read into R for automated and data directed gating using flowCore and OpenCyto packages. 37 dynamic subpopulations were identified between the two timepoints and 6 were selected for validation through immunohistochemistry and targeted flow cytometry. Among these, populations expressing either EpCAM (Epithelial Cell Adhesion Molecule) or CDCP1 (CUB Domain-Containing Protein 1) were investigated in detail because their decreasing expression coincided with cell differentiation into complex villus structures and they are known to mark discrete trophoblast subpopulations in the mouse. Targeted flow cytometry and immunohistochemistry analysis confirmed distinct EpCAM+ and CDCP1+ populations become restricted to discrete cellular subpopulations. The gene profiles of these subpopulations suggest EpCAM+ cells contribute to the villous cytotrophoblast lineage while CDCP1+ cells become extravillous trophoblasts (EVTs). More importantly, gene set enrichment suggests both subpopulations remain mitotically active. This is particularly interesting for the CDCP1+ population because EVTs are believed to be a terminally differentiated lineage. Conclusion: My work provides strong evidence to support the existence of transient populations of trophoblast cells in the human placenta and contributes to our understanding of how trophoblast populations communicate and organize into a mature, multicellular tissue. NI1.05 FIRST TRIMESTER INFECTION WITH LISTERIA MONOCYTOGENES IS CHARACTERIZED BY DECIDUAL AND PLACENTAL INFLAMMATION IN PREGNANT CYNOMOLGUS MACAQUES K.B. Wolfe , M.A. Hayes , G.J. Wiepz , G.I. Bondarenko , M. Durning , H.A. Simmons , N.G. Faith , C.C. Czuprynski , T.G. Golos . University of Wisconsin e Madison, Madison, USA; Wisconsin National Primate Research Center, Madison, USA Objectives: Listeria monocytogenes (LM) colonizes the reproductive tract in pregnant women and causes significant fetal morbidity and mortality, including stillbirth and neonatal infection. Maternal infection is typically subclinical, despite severe pregnancy impact. We hypothesize that the decidua carries a bacterial burden and serves as a site of immune cell activation following infection, leading to placental pathology and fetal loss. Methods: Timed-bred cynomolgus macaques received 107 LM intragastrically at 5-6 weeks gestation. Blood culture and MOX plating confirmed bacteremia and fecal shedding. Gestation age-matched control animals received mock inoculation. When fetal demise was detected, or at scheduled fetectomy, tissues were collected surgically. Paraffin-embedded decidua was stained immunohistochemically for CD3, CD68, Von Willebrand factor (VWF) and neutrophil elastase (NE). Tissue cytokines were assessed by Luminex assays. Results: Fetal demise occurred 7-13 days post-inoculation. Both reproductive tract and fetal tissues had extensive bacterial burden. Infected decidua had increased expression of CD3, CD68, VWF, and NE. Infected decidua, placenta and amniotic fluid had high levels of IFNg, IL1b, IL-2, MCP-1, MIP1-a, and TNFa. Control decidua had detectable but substantially lower levels of IL1b and TNFa. Except for MCP-1, no cytokines were detected in control amniotic fluid. Tissues collected at postinfection days 3 and 4, prior to fetal demise, had no reproductive tract or fetal colonization. By day 5, there was significant reproductive tract but not fetal colonization, with villous necrosis, syncytiotrophoblast degeneration, and LM in intervillous spaces, decidual cells and basal plate arteries. No bacteria were noted in chorionic plate vessels or amniotic fluid. Conclusion: LM infection in pregnancy initiates decidual leukocyte expansion/trafficking and cytokine secretion which we propose leads to vascular and cellular damage detrimental to successful pregnancy. Ongoing experiments to analyze tissues prior to fetal demise will give further insight into the precipitating temporal and spatial decidual changes. NI1.06 INTEGRATING MULTI-DIMENSIONAL GENOMICS DATA TO BUILD REGULATORY MAPS OF PLACENTAL DEVELOPMENT Rebekah Starks, Geetu Tuteja. Iowa State University, Ames, IA, USA Objectives: The placenta is a complex organ, in which distinct regulatory programs must be activated at specific times and locations during development. In this study we aim to predict transcriptional regulatorymodules during placental development using a systems-level approach. Methods: Our approach integrates RNA-Seq, H3k27Ac ChIP-Seq, ATACSeq, and binding site predictions. RNA-Seq and H3k27Ac data for mouse placenta assayed at e7.5 and e9.5 are from GEO (GSE65809). A transcription factor (TF) was considered upregulated at one time point compared to the other if it had an FPKM 10, fold 2, and q-value <0.05. We used phylogenetic footprinting for binding site predictions, with a motif match threshold of 800 and motif score p-value <0.05. ATAC-Seq libraries were generated at e7.5 and e9.5 using published protocols. Results: We have previously shown that e7.5-specific enhancer peak data combined with binding site analysis can be used to predict e7.5-specific regulatory modules. Here we improve the predictability of regulatory modules by first identifying TFs that are upregulated at a time point, and then predicting binding sites for those TFs in time point specific enhancers. Using this approach we identify 8 TFs with significant motif occurrence in e9.5-specific peaks (p<0.001). We also demonstrate that analyzing pairs of TF binding sites and incorporating ATAC-Seq data can further aid in predicting regulatorymodules. In a preliminary analysis we predict a subset of e9.5-specific enhancers, with both Plagl1 and Pparg motifs, that are significantly associated with genes involved in the insulin pathway (p<0.05). Conclusions: Our initial results suggest Plagl1 and Pparg may work together to regulate metabolic genes in the placenta. Furthermore, integration of multi-dimensional data can be used to predict regulatory modules on a genome-wide scale. Following the development of a refined computational pipeline, we will test the functionality of our predictions in placental cell lines. NI2.01 GLYCOLYTIC UTILIZATION AND CAPACITY OF HUMAN CYTOTROPHOBLASTS ARE HIGHER THAN SYNCYTIOTROPHOBLASTS Amy Valent, Kevin Kolahi, Kent Thornburg. Oregon Health and Science University, Portland, OR, USA Objective: We have previously shown that cytotrophoblast (CTB) have a greater capacity to metabolize exogenous lipids using oxidative phosphorylation than do syncytiotrophoblast (SCT). Thus, SCT may have a greater capacity for glycolysis. We tested the hypothesis that the glycolytic capacity of cultured human SCT is higher than in CTB. Methods: CTB were isolated from normal term human placentas delivered via cesarean section. Growth medium was replaced with non-buffered medium to accurately measure extra cellular acidification rate (ECAR) in the Seahorse XFe96. CTB were studied at 8 hours and cultured, differentiated syncytialized cytotrophoblasts (SCT) at 72 hours. Glycolytic function of CTB and SCT was measured under basal conditions, in the presence of 5mM or 25mM glucose, a mitochondrial ATP synthase inhibitor (Oligomycin), and a non-metabolized glucose analog that inhibits glycolysis through competition for glucose hexokinase (2-deoxy-glucose). ECAR was normalized to total DNA content. Two-way ANOVA was used to compare ECAR between CTB and SCT. Results: As shown in Fig. 1, CTB had significantly higher rates of glycolysis compared to SCT in both normal glucose (5mM) and elevated (25mM) concentrations [(53±6 vs 18±3) & 63±6 vs 19±4 mpH min-1/100ng DNA),