Utero-placental cell interactions revealed by single cell transcriptomics

Utero-placental cell interactions revealed by single cell transcriptomics

64 Abstracts / Placenta 45 (2016) 63e133 NI1.01 TRANSCRIPTOMIC PROFILING OF SNCRNA IN PLACENTA BY RNASEQ E. Magda Price 1,2, Victor Martinez 1, 3, D...

46KB Sizes 0 Downloads 55 Views

64

Abstracts / Placenta 45 (2016) 63e133

NI1.01 TRANSCRIPTOMIC PROFILING OF SNCRNA IN PLACENTA BY RNASEQ E. Magda Price 1,2, Victor Martinez 1, 3, Daiana Becker-Santos 1, 3, Irina Manokhina 1, 2, Wan L. Lam 1, 3, Wendy P. Robinson 1, 2. 1 University of British Columbia, Vancouver, Canada; 2 Child and Family Research Institute, Vancouver, Canada; 3 BC Cancer Agency, Vancouver, Canada Objectives: Small non-coding RNAs (sncRNA), such as miRNAs (21-24 nucleotides long) and piRNAs (26-31 nucleotides long), are increasingly recognized as important modifiers of gene expression. sncRNAs with tissue or condition-specific expression, that can also be detected in blood, are attractive non-invasive biomarkers of health and disease. During pregnancy, secretion and shedding of trophoblast results in release of placental sncRNAs into maternal circulation. Though several studies have identified differential sncRNA expression in complicated pregnancies, rarely are the same candidates reported. Understanding biological and technical factors associated with sncRNA expression is fundamental to developing robust, reproducible clinical tools. Methods: In this study, the sncRNA transcriptome of 26 healthy placentas, 6 first trimester (6-11 weeks gestation), 10 second trimester (14-24 weeks gestation) and 10 term (38-40 weeks gestation), was profiled using highthroughput small RNA sequencing. This approach allowed for detection of novel transcripts and sequencing of both piRNAs and miRNAs by applying population-specific read-length filters. Expression levels were normalized by sequencing depth and transcript length (RPKM) and log transformed. Results of 6,029 annotated piRNAs, 916 (15%) were expressed in >1 sample and 187 (3%) were expressed in >75% of samples of at least one trimester. Principal component analysis suggested major sources of variation due to sample processing time and library construction, thus these factors were adjusted for in further analyses. Linear modelling identified 19 piRNAs differentially expressed (FDR <0.05) between trimesters with a range of fold changes (0.30-6.28), while 61 piRNAs exhibited stable expression across trimesters (nominal p-value >0.1). Top stable and gestational ageassociated sncRNA candidates will be validated by qPCR and tested in an extended set of samples. Conclusion: Our study suggests that processing time, library batch and gestational age are important factors in the study of sncRNAs and demonstrates for the first time, that piRNAs can be detected in placental tissue. NI1.02 IMPAIRED GENE-EXPRESSION AND EPIGENETIC REGULATION OF RETINOIC ACID RECEPTOR RESPONDER 1 IN PREECLAMPSIA AND CHORIOCARCINOMA Hanna Huebner 1, Fabian B. Fahlbusch 2, Pamela L. Strissel 1, Reiner Strick 1, David Wachter 3, Matthias W. Beckmann 1, Matthias Ruebner 1. 1 Department of Gynaecology and Obstetrics, University Hospital Erlangen, Friedrich-Alexander University Erlangen-Nuremberg, Comprehensive Cancer Center Erlangen-EMN, Erlangen, Germany; 2 Department of Pediatrics and Adolescent Medicine, University Hospital Erlangen, Friedrich-Alexander University Erlangen-Nuremberg, Comprehensive Cancer Center Erlangen-EMN, Erlangen, Germany; 3 Department of Pathology, University Hospital Erlangen, FriedrichAlexander University Erlangen-Nuremberg, Comprehensive Cancer Center Erlangen-EMN, Erlangen, Germany Objectives: Human placental development resembles tumorigenesis in its invasive and proliferative capacity. In contrast to cancer, these features are tightly regulated. Disturbances within this regulation are thought to contribute to gestational diseases, like choriocarcinoma, preeclampsia (PE) and intrauterine growth restriction (IUGR). Retinoic acid receptor responder 1 (RARRES1) is a tumor-suppressor known to be epigenetically silenced in many cancers. The aim of our study was to investigate the expression and epigenetic regulation of RARRES1 in PE, IUGR and choriocarcinoma. Methods: Immunhistochemical staining of RARRES1 on healthy and pathological (choriocarcinoma, PE, IUGR) placental sections was performed. Gene-expression was analyzed by qRT-PCR of RNA derived from total placental tissue, isolated primary trophoblasts, the first-trimester cell-line Swan71 and choriocarcinoma cell-lines (Jeg-3, JAR, BeWo). The

methylation pattern of RARRES1 was quantified by pyrosequencing. Changes of cell-cell contact were determined by Electric cell-substrate impedance sensing. Results: Choriocarcinoma cell-lines showed a hypermethylation of the RARRES1 promotor, accompanied by significant reduced gene-expression. In analogy, DNA derived from choriocarcinoma tissue showed a higher RARRES1 methylation as compared to healthy first trimester DNA. In contrast, a significant higher RARRES1 expression was observed in primary trophoblasts from PE cases compared to controls. Additionally, we found a significant induction of RARRES1 expression relative to increasing celldensity. In concordance, RARRES1 overexpression in Jeg-3 cells enhanced the measured impedance, indicating stronger cell-cell connectivity. Conclusions: Our findings strengthened the hypothesis that RARRES1 functions in a tumor-suppressive manner and is dysregulated in placental diseases. We showed that RARRES1 expression is tightly regulated by DNAmethylation. Based on our findings, we hypothesize high RARRES1 expression, as observed in PE, might increase cell-cell adhesion and thus negatively influences trophoblast invasion or proliferation. On the other hand, epigenetic silencing of RARRES1 in choriocarcinomas might reduce cell adhesion and promote the tumorigenic potential of trophoblastic cells by enhancing epithelial-to-mesenchymal transition. NI1.03 UTERO-PLACENTAL CELL INTERACTIONS REVEALED BY SINGLE CELL TRANSCRIPTOMICS Mihaela Pavlicev 1, Helen Jones 1, Kathryn Owens 1, Arun Chavan 2, Gunter Wagner 2, Louis Muglia 1. 1 Cincinnati Children's Hospital Medical Center, Cincinnati, OH, USA; 2 Yale University, New Haven, CT, USA Objectives: The high heterogeneity of placental function is reflected in correspondingly heterogeneous array of cell types and their interactions. While informative, tissue level is of limited help for understanding normal biology of placenta, as the expression profile of the tissue is inevitably biased towards the expression of most numerous cell types. Characterization of cultured cell lines resolves this problem only partially, as many of the functions ensue from cell interactions. Methods: We applied single cell transcriptomics to characterize placental expression, aiming specifically at the trophoblast cells. We captured 87 cells, characterizing populations of intravillous and extravillous trophoblast cells. In addition, we used laser micro-dissection to complement the data with syncytiotrophoblast transcriptome, and data from primary cell culture of endometrial stromal fibroblast- as well as differentiated decidual cells. We focused specifically on feto-maternal interactions between placental cells and maternal decidua. Results: Using known ligand-receptor complementarity we reconstructed the cell-cell interaction network among placental and endometrial cells. The numerous genes involved in these feto-maternal interactions include a large group of growth factors and related molecules, as well as proteins involved in immune reaction, and hormonal signaling. The data demonstrates presence of both supportive and defensive signals from maternal decidua. Interesting is also high specificity of G-protein coupled receptors, which appear as good candidate for cell type characterization. Conclusions: Transcriptomics carries a potential and also a challenge to translate the amount of data to knowledge, and integrate it with existing knowledge. A focus on cell communication and its dynamics facilitates prioritization of the ongoing processes, and this is particularly important in placenta, an epitome of interaction. Beyond characterization, our data shows many interactions that are specific to extravillous trophoblastdecidual interface and syncytial-decidual interface and are not possible without decidualization. These interactions in particular may be interesting in the context of ectopic invasive placentation. NI1.04 DYNAMICALLY REGULATED TROPHOBLAST ISOLATED FROM EARLY HUMAN PLACENTAS

SUBPOPULATIONS

Frances Wong, Brian Cox. University of Toronto, Toronto, ON, Canada Objective: The placenta arises from a network of trophoblast and mesoderm cells but many questions about early cellular development remain: