- Email: [email protected]
Effects of placental terminal villous capillarisation on vascular wall shear stress and placental angiogenesis
76
Abstracts / Placenta 45 (2016) 63e133
P1.10 A PRIORI SPECIFIED RELATIONSHIPS AMONG ARTERIAL CHORIONIC SURFACE VESSEL NETWORK (PCSVN) AND AUTISM RISK Ruchit Shah 1, Theresa Girardi 1, Jen-mei Chang 2, Jennifer Craig Newschaffer 5, Dawn Misra 4, Philip Straughen 3, Richard Miller 6, 7, John Moye 7, 8, Carolyn Katzman 6, 7, Salafia 1, 7. 1 Placental Analytics, LLC, New Rochelle, NYU, USA; 2 California State, Long Beach, Long Beach, CA, USA; 3 Henry Ford Health Systems, Detroit, MI, USA; 4 Wayne State University, Detroit, MI, USA; 5 Drexel University, Philadelphia, PA, USA; 6 University of Rochester, Rochester, NY, USA; 7 Placental Consortium, Bethesda, MD, USA; 8 NICHD, Bethesda, MD, USA Objectives: Many features can be quantified from PCSVNs, and principal components analysis (PCA) can be used to identify such relationships. We selected variables that addressed specific PCSVN features, extracted components and determined whether the interrelationships varied between high (familial) and low ASD risk groups. Methods: Digital photomicrographs were obtained of 201 National Children Study placenta (NCS, considered “baseline” and 89 placentas from Early Autism Risk Longitudinal Investigation (EARLI) born into a family with an elder sibling with an ASD diagnosis. The tracings were extracted as per published protocols. PCSVN “size” was reflected by branch point number, mean and standard deviation (SD) vessel thickness, network arc length and mean and SD distance from vessel end points to the disk edge. Mean and SD of branch angles and segment tortuosity were selected, as branch angles reflect speed of growth at the time of branching and tortuosity implies prior altered intravascular pressure. The whole chorionic surface area was included to normalize factors for placental size. Analyses were performed in SPSS (SPSS 22.0, IBM, Chicago, IL). Results: Four components were extracted that accounted for 80% of data variance (Table 1). Factors scores were calculated for the four factors in both groups. Each factor differed between the two groups (each p<0.0001 (Fig. 1). Examples of extreme scores for the four factors are presented in Fig. 2. Conclusions: While individual vessel features differ between high and low ASD risk, “the company they keep” is also very distinctive. Our goals have been to most fully characterize normal PCSVN, and to draw distinction between this baseline and familial (presumably genetic) high ASD risk populations. This has been accomplished. Our next steps are to use these methods for classification within high genetic risk populations as to actual ASD diagnosis. (Supported in part by NIH LOI18 and Placental Analytics, LLC.)
Ă
Ă
P1.11 EFFECTS OF PLACENTAL TERMINAL VILLOUS CAPILLARISATION ON VASCULAR WALL SHEAR STRESS AND PLACENTAL ANGIOGENESIS Win Min Tun 1, 2, Joanna James 2, Alys Clark 1. 1 Auckland Bioengineering Institute, University of Auckland, Auckland, New Zealand; 2 Faculty of Medical and Health Sciences, University of Auckland, Auckland, New Zealand Introduction: Terminal capillary convolutes (CCs) in the tips of human placental villi are the principal sites of feto-maternal gas exchange. Overand under-capillarisation of villi occur in high altitude, and several pregnancy complications such as high placental blood pressure in umbilical venous abnormalities. Changes in wall shear stress (WSS) in CCs regulate capillary remodelling (cell proliferation) and angiogenesis. However, physiological WSS is difficult to measure due to inaccessible CCs. Thus in vitro studies of the influence of WSS on placental vascularisation are difficult to relate to in vivo function. Methods: Using an anatomically based multi-scale computational model of the placental vasculature we predict the relationship between placental Ă
Abstracts / Placenta 45 (2016) 63e133
vascularisation and WSS on endothelial cells within CCs. The model was solved in three cases 1) under-capillarisation, 2) normal and 3) overcapillarisation. Based on morphological studies, we assumed under- and over- capillarisation are defined as 0.5 and 1.7 times the normal number of CCs respectively. Results: For the same fetal cardiac output, flow (0.004m3/s, 0.002m3/s and 0.001m3/s for cases 1, 2 and 3 respectively) and predicted WSS (0.106Pa, 0.059Pa and 0.039Pa for cases 1, 2 and 3 respectively) in each CC decreased with the number of CCs. Increasing placental blood pressures (e.g. umbilical venous obstruction) increased WSS by 2.7 fold, which subsequently stimulate over-capillarisation to reduce WSS to 2.3 times its normal level. Conclusion: Our model predicts normal and abnormal ranges of WSS in placental CCs, including decreased WSS with increasing capillarisation. This model translates in vitro data on the effect of WSS on placental cells into in vivo system. Our model shows that increased placental blood pressure, which causes high WSS and subsequent over-capillarisation, are unlikely to be completely reversed by this over-capillarisation, which is consistent with evidence of structural capillary damage in cases of elevated placental blood pressure. P1.12 TROPHOBLASTS STIMULATE THE RELEASE OF MMP10 BY ENDOTHELIAL CELLS Nora Lagzouli, Ben Sayer, Sandra Ashton, Judith Cartwright, Guy Whitley. St George's University of London, London, UK Background: The maternal uterine spiral arteries (SpA) change significantly during the first weeks of pregnancy resulting in increased blood flow to the developing fetus. Endothelial cells (EC) and vascular smooth muscle cells (VSMC) are lost from the vessel wall and are replaced by extravillous trophoblasts. Using a 3D model of the vessel wall we have shown that the expression of a number of genes alters following stimulation with trophoblast conditioned media (TCM), including MMP10, an enzyme involved in the breakdown of the extracellular matrix. Objective: To investigate the regulation of EC derived MMP10 by trophoblast. Methods: The human endothelial line SGHEC-7 were incubated with TCM, TGFb and IL1b for up to 72h and the release of MMP10 determined by ELISA. To characterise the response in more detail these stimuli were also used in the presence of pharmacological inhibitors including the STAT3 inhibitor, STAT3 Inhibitor VIII. First trimester decidual tissue from normal pregnancies were fixed and stained for CK7 (trophoblasts) and MMP10. Results: EC but not VSMC secreted MMP10. MMP10 expression in decidual tissue seemed to be associated with the presence of trophoblast. TCM stimulates MMP10 secretion from EC in a dose dependent manner. Trophoblast secrete a number of growth factors and cytokines including IL1b, IL6 and TGFb-1. The secretion of MMP10 was significantly stimulated by IL1b (p<0.05; n¼3) and TGFb-1 (p<0.05; n¼3) but not IL6. The secretion of MMP10 in response to TCM was significantly inhibited by the STAT3 inhibitor VIII. Conclusion: The expression and secretion of MMP10 in response to factors released by trophoblasts is restricted to EC. It would therefore appear that trophoblast may mediate SA remodelling in part via the release of MMP10 from the EC. P1.13 THREE DIMENSIONAL MODELLING OF HAEMODYNAMICS IN A RAT FETO-PLACENTAL ARTERIAL NETWORK Tim Crough, Lachlan Kelsey, Barry Doyle, Caitlin Wyrwoll. The University of Western Australia, Perth, Australia Optimal arrangement of placental vasculature and adequate blood flow is central to both placental function and fetal growth. Yet despite this, how haemodynamic forces relate to placental vascular form and the consequent ramifications for placental function is predominantly unclear. Here, we apply computational fluid dynamics to model blood flow in a three-dimensional (3D) manner in a sub-volume of a rat feto-placental arterial cast. A microfil-perfused feto-placental arterial vascular cast obtained from a healthy pregnant Wister rat (E22; near the end of gestation) underwent
77
high-resolution m-CT scans using wide field mode. Two-dimensional datasets were segmented and reconstructed into a 3D vascular network by threshold segmentation (Mimics Innovation Suite v19, Materialise NV). A sub-volume consisting of umbilical artery, large bifurcating vessels and two tortuous branches were extracted and transformed into a volume mesh for subsequent haemodynamic modelling using computational fluid dynamics (STAR-CCM+, Cd-Adapco Group). Doppler ultrasound velocity data of rat umbilical artery was measured and used as input to the pulsatile computational model. Functional measures including blood flow, velocity, pressure and wall shear stress (WSS) were measured throughout the subvolume of the feto-placental network. Biological integrity of the sub-volume was high as not only was the geometry anatomically correct, but the use of pulsatile velocity input enables functional measures of flow, velocity, pressure and WSS over the cardiac cycle. We observed major variation in WSS throughout the cardiac cycle, which cannot be captures using previous steady-state simulations. This current study is the first to generate a 3D haemodynamic model of the feto-placental vasculature in a rodent. We have overcome several existing limitations of current placental vascular models and further refinement of this technique will serve as a critical tool to further understanding of mechanisms controlling development and function of feto-placental vascularity. P1.14 CHARACTERIZATION OF GLOBAL HUMAN MICRORNA EXPRESSION PATTERNS ACROSS FETAL PLASMA, PLACENTA, AND MATERNAL PLASMA Alison Paquette 1, Tianjiao Chu 2, Xiagong Wu 1, Kai Wang 1, Cory Funk 1, Yoel Sadovsky 2, Nathan Price 1. 1 Institute for Systems Biology, Seattle, Washington, USA; 2 Magee Womens Research Institute, Pittsburgh, PA, USA Objective: The placenta acts as a key endocrine organ during pregnancy to regulate signals between the mother and developing fetus, and releases specific microRNAs in a coordinated fashion. There is a diagnostic and research need to understand patterns of microRNA expression during pregnancy. The objective of this analysis was to characterize differences in expression patterns amongst pregnancy compartments and quantify information conveyed across compartments. Methods: microRNA sequencing was performed on 25 non-pathological term pregnancies in fetal plasma, placenta, and maternal plasma. MicroRNA expression patterns were defined using consensus clustering. Relationships between compartments amongst the same individual were defined through mutual information, and correlations of microRNAs between compartments were examined through Pearson correlations. mirSystem was used to infer relationships between microRNA target genes and pathways. Results: MicroRNAs were variably expressed across pregnancy related compartments, with overall more diversity and increased expression in the placenta, particularly of the C19MC microRNA cluster. MicroRNAs clustered into 5 distinct groups, which showed tissue specificity in their expression patterns. We identified significant mutual information shared across fetal and maternal plasma of the same individual compared to others (P <0.05). Finally, we identified 35 microRNAs that were correlated across compartments (P <0.05), and found that most of these microRNAs exhibited decreased expression in placenta compared to fetal plasma (N¼15) or maternal plasma (N¼9). Conclusion: In this first global analysis of microRNA expression across pregnancy compartments, we were able to identify specific microRNAs that were expressed in a coordinated fashion across fetal plasma, placenta, and maternal blood. We have provided a global characterization of relationships between microRNAs and individuals in the placenta, which will be valuable for understanding how these microRNAs are expressed and their function during pregnancy. P1.15 UNRAVELLING IGF-I SIGNALLING IN VILLOUS TROPHOBLAST Magda Karolczak-Bayatti, James Horn, Lynda Harris, Melissa Westwood, John Aplin. University of Manchester, Manchester, UK
Abstracts / Placenta 45 (2016) 63e133
P1.10 A PRIORI SPECIFIED RELATIONSHIPS AMONG ARTERIAL CHORIONIC SURFACE VESSEL NETWORK (PCSVN) AND AUTISM RISK Ruchit Shah 1, Theresa Girardi 1, Jen-mei Chang 2, Jennifer Craig Newschaffer 5, Dawn Misra 4, Philip Straughen 3, Richard Miller 6, 7, John Moye 7, 8, Carolyn Katzman 6, 7, Salafia 1, 7. 1 Placental Analytics, LLC, New Rochelle, NYU, USA; 2 California State, Long Beach, Long Beach, CA, USA; 3 Henry Ford Health Systems, Detroit, MI, USA; 4 Wayne State University, Detroit, MI, USA; 5 Drexel University, Philadelphia, PA, USA; 6 University of Rochester, Rochester, NY, USA; 7 Placental Consortium, Bethesda, MD, USA; 8 NICHD, Bethesda, MD, USA Objectives: Many features can be quantified from PCSVNs, and principal components analysis (PCA) can be used to identify such relationships. We selected variables that addressed specific PCSVN features, extracted components and determined whether the interrelationships varied between high (familial) and low ASD risk groups. Methods: Digital photomicrographs were obtained of 201 National Children Study placenta (NCS, considered “baseline” and 89 placentas from Early Autism Risk Longitudinal Investigation (EARLI) born into a family with an elder sibling with an ASD diagnosis. The tracings were extracted as per published protocols. PCSVN “size” was reflected by branch point number, mean and standard deviation (SD) vessel thickness, network arc length and mean and SD distance from vessel end points to the disk edge. Mean and SD of branch angles and segment tortuosity were selected, as branch angles reflect speed of growth at the time of branching and tortuosity implies prior altered intravascular pressure. The whole chorionic surface area was included to normalize factors for placental size. Analyses were performed in SPSS (SPSS 22.0, IBM, Chicago, IL). Results: Four components were extracted that accounted for 80% of data variance (Table 1). Factors scores were calculated for the four factors in both groups. Each factor differed between the two groups (each p<0.0001 (Fig. 1). Examples of extreme scores for the four factors are presented in Fig. 2. Conclusions: While individual vessel features differ between high and low ASD risk, “the company they keep” is also very distinctive. Our goals have been to most fully characterize normal PCSVN, and to draw distinction between this baseline and familial (presumably genetic) high ASD risk populations. This has been accomplished. Our next steps are to use these methods for classification within high genetic risk populations as to actual ASD diagnosis. (Supported in part by NIH LOI18 and Placental Analytics, LLC.)
Ă
Ă
P1.11 EFFECTS OF PLACENTAL TERMINAL VILLOUS CAPILLARISATION ON VASCULAR WALL SHEAR STRESS AND PLACENTAL ANGIOGENESIS Win Min Tun 1, 2, Joanna James 2, Alys Clark 1. 1 Auckland Bioengineering Institute, University of Auckland, Auckland, New Zealand; 2 Faculty of Medical and Health Sciences, University of Auckland, Auckland, New Zealand Introduction: Terminal capillary convolutes (CCs) in the tips of human placental villi are the principal sites of feto-maternal gas exchange. Overand under-capillarisation of villi occur in high altitude, and several pregnancy complications such as high placental blood pressure in umbilical venous abnormalities. Changes in wall shear stress (WSS) in CCs regulate capillary remodelling (cell proliferation) and angiogenesis. However, physiological WSS is difficult to measure due to inaccessible CCs. Thus in vitro studies of the influence of WSS on placental vascularisation are difficult to relate to in vivo function. Methods: Using an anatomically based multi-scale computational model of the placental vasculature we predict the relationship between placental Ă
Abstracts / Placenta 45 (2016) 63e133
vascularisation and WSS on endothelial cells within CCs. The model was solved in three cases 1) under-capillarisation, 2) normal and 3) overcapillarisation. Based on morphological studies, we assumed under- and over- capillarisation are defined as 0.5 and 1.7 times the normal number of CCs respectively. Results: For the same fetal cardiac output, flow (0.004m3/s, 0.002m3/s and 0.001m3/s for cases 1, 2 and 3 respectively) and predicted WSS (0.106Pa, 0.059Pa and 0.039Pa for cases 1, 2 and 3 respectively) in each CC decreased with the number of CCs. Increasing placental blood pressures (e.g. umbilical venous obstruction) increased WSS by 2.7 fold, which subsequently stimulate over-capillarisation to reduce WSS to 2.3 times its normal level. Conclusion: Our model predicts normal and abnormal ranges of WSS in placental CCs, including decreased WSS with increasing capillarisation. This model translates in vitro data on the effect of WSS on placental cells into in vivo system. Our model shows that increased placental blood pressure, which causes high WSS and subsequent over-capillarisation, are unlikely to be completely reversed by this over-capillarisation, which is consistent with evidence of structural capillary damage in cases of elevated placental blood pressure. P1.12 TROPHOBLASTS STIMULATE THE RELEASE OF MMP10 BY ENDOTHELIAL CELLS Nora Lagzouli, Ben Sayer, Sandra Ashton, Judith Cartwright, Guy Whitley. St George's University of London, London, UK Background: The maternal uterine spiral arteries (SpA) change significantly during the first weeks of pregnancy resulting in increased blood flow to the developing fetus. Endothelial cells (EC) and vascular smooth muscle cells (VSMC) are lost from the vessel wall and are replaced by extravillous trophoblasts. Using a 3D model of the vessel wall we have shown that the expression of a number of genes alters following stimulation with trophoblast conditioned media (TCM), including MMP10, an enzyme involved in the breakdown of the extracellular matrix. Objective: To investigate the regulation of EC derived MMP10 by trophoblast. Methods: The human endothelial line SGHEC-7 were incubated with TCM, TGFb and IL1b for up to 72h and the release of MMP10 determined by ELISA. To characterise the response in more detail these stimuli were also used in the presence of pharmacological inhibitors including the STAT3 inhibitor, STAT3 Inhibitor VIII. First trimester decidual tissue from normal pregnancies were fixed and stained for CK7 (trophoblasts) and MMP10. Results: EC but not VSMC secreted MMP10. MMP10 expression in decidual tissue seemed to be associated with the presence of trophoblast. TCM stimulates MMP10 secretion from EC in a dose dependent manner. Trophoblast secrete a number of growth factors and cytokines including IL1b, IL6 and TGFb-1. The secretion of MMP10 was significantly stimulated by IL1b (p<0.05; n¼3) and TGFb-1 (p<0.05; n¼3) but not IL6. The secretion of MMP10 in response to TCM was significantly inhibited by the STAT3 inhibitor VIII. Conclusion: The expression and secretion of MMP10 in response to factors released by trophoblasts is restricted to EC. It would therefore appear that trophoblast may mediate SA remodelling in part via the release of MMP10 from the EC. P1.13 THREE DIMENSIONAL MODELLING OF HAEMODYNAMICS IN A RAT FETO-PLACENTAL ARTERIAL NETWORK Tim Crough, Lachlan Kelsey, Barry Doyle, Caitlin Wyrwoll. The University of Western Australia, Perth, Australia Optimal arrangement of placental vasculature and adequate blood flow is central to both placental function and fetal growth. Yet despite this, how haemodynamic forces relate to placental vascular form and the consequent ramifications for placental function is predominantly unclear. Here, we apply computational fluid dynamics to model blood flow in a three-dimensional (3D) manner in a sub-volume of a rat feto-placental arterial cast. A microfil-perfused feto-placental arterial vascular cast obtained from a healthy pregnant Wister rat (E22; near the end of gestation) underwent
77
high-resolution m-CT scans using wide field mode. Two-dimensional datasets were segmented and reconstructed into a 3D vascular network by threshold segmentation (Mimics Innovation Suite v19, Materialise NV). A sub-volume consisting of umbilical artery, large bifurcating vessels and two tortuous branches were extracted and transformed into a volume mesh for subsequent haemodynamic modelling using computational fluid dynamics (STAR-CCM+, Cd-Adapco Group). Doppler ultrasound velocity data of rat umbilical artery was measured and used as input to the pulsatile computational model. Functional measures including blood flow, velocity, pressure and wall shear stress (WSS) were measured throughout the subvolume of the feto-placental network. Biological integrity of the sub-volume was high as not only was the geometry anatomically correct, but the use of pulsatile velocity input enables functional measures of flow, velocity, pressure and WSS over the cardiac cycle. We observed major variation in WSS throughout the cardiac cycle, which cannot be captures using previous steady-state simulations. This current study is the first to generate a 3D haemodynamic model of the feto-placental vasculature in a rodent. We have overcome several existing limitations of current placental vascular models and further refinement of this technique will serve as a critical tool to further understanding of mechanisms controlling development and function of feto-placental vascularity. P1.14 CHARACTERIZATION OF GLOBAL HUMAN MICRORNA EXPRESSION PATTERNS ACROSS FETAL PLASMA, PLACENTA, AND MATERNAL PLASMA Alison Paquette 1, Tianjiao Chu 2, Xiagong Wu 1, Kai Wang 1, Cory Funk 1, Yoel Sadovsky 2, Nathan Price 1. 1 Institute for Systems Biology, Seattle, Washington, USA; 2 Magee Womens Research Institute, Pittsburgh, PA, USA Objective: The placenta acts as a key endocrine organ during pregnancy to regulate signals between the mother and developing fetus, and releases specific microRNAs in a coordinated fashion. There is a diagnostic and research need to understand patterns of microRNA expression during pregnancy. The objective of this analysis was to characterize differences in expression patterns amongst pregnancy compartments and quantify information conveyed across compartments. Methods: microRNA sequencing was performed on 25 non-pathological term pregnancies in fetal plasma, placenta, and maternal plasma. MicroRNA expression patterns were defined using consensus clustering. Relationships between compartments amongst the same individual were defined through mutual information, and correlations of microRNAs between compartments were examined through Pearson correlations. mirSystem was used to infer relationships between microRNA target genes and pathways. Results: MicroRNAs were variably expressed across pregnancy related compartments, with overall more diversity and increased expression in the placenta, particularly of the C19MC microRNA cluster. MicroRNAs clustered into 5 distinct groups, which showed tissue specificity in their expression patterns. We identified significant mutual information shared across fetal and maternal plasma of the same individual compared to others (P <0.05). Finally, we identified 35 microRNAs that were correlated across compartments (P <0.05), and found that most of these microRNAs exhibited decreased expression in placenta compared to fetal plasma (N¼15) or maternal plasma (N¼9). Conclusion: In this first global analysis of microRNA expression across pregnancy compartments, we were able to identify specific microRNAs that were expressed in a coordinated fashion across fetal plasma, placenta, and maternal blood. We have provided a global characterization of relationships between microRNAs and individuals in the placenta, which will be valuable for understanding how these microRNAs are expressed and their function during pregnancy. P1.15 UNRAVELLING IGF-I SIGNALLING IN VILLOUS TROPHOBLAST Magda Karolczak-Bayatti, James Horn, Lynda Harris, Melissa Westwood, John Aplin. University of Manchester, Manchester, UK