SRI Supplement to Reproductive Sciences - Volume 25 Number 1 - March 2018 - 237A

Scientific Abstracts

F-167

F-168
Alterations in the First and Second Trimester Maternal Urinary
Metabolome Associated with Subsequent Early-Onset Preeclampsia.
Kathryn J Gray,1 Julian Avila-Pacheco,2 David E Cantonwine,1 Thomas F
McElrath,1 Clary B Clish,2 Richa Saxena.3 1Brigham & Women's Hospital,
Boston, MA, United States; 2Broad Institute, Cambridge, MA, United
States; 3Massachusetts General Hospital, Boston, MA, United States.
INTRODUCTION: Improved mechanistic understanding of preeclampsia
(PE) is needed to facilitate novel predictive tools and treatments. As
placental dysfunction and maternal endothelial damage precede clinicallyapparent preeclampsia, we hypothesized that associated alterations in the
maternal metabolome would be detectable in early pregnancy and provide
insight into causal biologic pathways.
METHODS: Maternal 1st (T1; mean 12 wks gestation) and 2nd (T2, mean
26 wks gestation) trimester urine samples from 60 women who developed
early-onset PE were matched with 60 healthy, normotensive controls from
the LIFECODES cohort at Brigham & Women's Hospital. Matching was
based on maternal age, race, BMI, smoking status, and gestational age at
sample collection (T1, T2). Global metabolic profiling was performed on
urine samples using three well-characterized liquid-chromatography mass
spectroscopy (LC-MS) metabolic profiling methods (Broad Institute),

237A

allowing for the identification of 351 known and 1000's of unknown
metabolites. Data was normalized (by median metabolite level), log2
transformed, and features with >90% missing values removed. Paired
t-tests were performed to assess for significant metabolite differences at
each time point, and across gestation (T2/T1), and corrected for multiple
testing. Conditional logistic regression was performed to control for
confounders.
RESULTS: Significant differences in case-control metabolites (p<0.01)
were noted in the 1st trimester (10 known, 326 unknown metabolites),
2nd trimester (3 known, 480 unknown metabolites), and across trimesters
(T2/T1; 3 known, 213 unknown metabolites) (Table 1). After false
discovery rate (FDR) correction for 351 known metabolites, differences
in 1st trimester case-control carnitine levels remained significant (p=0.05).
This result remained significant after adjustment for age, race, and BMI.
CONCLUSION: Here we report the first longitudinal maternal global
urinary metabolic profiling study in the 1st and 2nd trimesters of pregnancy
in women who developed early-onset PE vs. controls. This analysis
highlights early alterations in carnitine metabolism as a pathway for further
investigation. Future work aims to extend this global metabolic profiling
to plasma samples from the same subjects, as well as to a replication PE
case-control cohort.
*Figure(s) will be available online.

F-169
sFlt-1 Transcription and Secretion is Regulated through the
Mitochondrial Electron Transport Chain - A New Mechanism to
Target for Treating Preeclampsia. Tu'uhevaha J Kaitu'u-Lino*, Fiona
C Brownfoot†, Natalie J Hannan, Roxanne Hastie†, Ping Cannon, TuongVi Nguyen, Stephen Tong. University of Melbourne, Mercy Hospital for
Women, Heidelberg, Australia.
INTRODUCTION: Preeclampsia is characterised by placental
oxidative stress and elevated release of anti-angiogenic factor sFlt-1.
Mitochondria are the energy factories of cells and are dysfunctional in
preeclamptic placentas. Inhibiting the mitochondrial electron transport
chain (ETC) decreases ATP production and upregulates mitochondrial
biogenesis molecules AMPK, SIRT1 and PGC1α. We have demonstrated
that mitochondrial electron transport chain (ETC) complex I inhibitor
metformin inhibits sFlt-1 release from primary trophoblast cells.
In this study, we aimed to test 1) whether inhibiting other mitochondrial
ETC complexes would inhibit sFlt-1 2) whether activating mitochondrial
biogenesis molecules AMPK, SIRT1 and PGC1α alters sFlt-1 release 3)
whether potential therapeutics esomeprazole and metformin that reduce
sFlt-1 release, inhibit mitochondrial respiration and alter AMPK, SIRT1
and PGC1α expression.
METHODS: For all functional studies, primary trophoblast isolated
from term placentas were used. Primary trophoblast were treated with
mitochondrial ETC complex II, III and V inhibitors rotenone, antimycin
and oligyomycin and sFlt-1 transcription measured by qRT-PCR and
secretion measured via ELISA. To activate biogenesis molecules, SIRT1
activator resveratrol and AMPK mimetic AICAR were used. Activation
of biogenesis molecules was confirmed via qRT-PCR and sFlt-1
transcription and secretion measured. To assess the effect of esomeprazole
and metformin on mitochondrial respiration, primary trophoblast were
assessed using a Seahorse Flux analyser. The effect of esomeprazole and
metformin on AMPK, SIRT1 and PGC1α was determined via qRT-PCR.
RESULTS: Inhibiting complexes II, III and V of the mitochondrial ETC
in primary trophoblast significantly decreased sFlt-1 transcription and
secretion. Similarly, activating down-stream mitochondrial biogenesis
molecules SIRT1, AMPK and PGC1α with either resveratrol or AICAR
significantly inhibited sFlt-1 transcription and secretion. Treating primary
trophoblast with esomeprazole or metformin significantly decreased
sFlt-1 secretion, coincident with decreased mitochondrial respiration
and increased expression of mitochondrial biogenesis molecules AMPK,
SIRT1 and PGC1α.
CONCLUSION: Our data suggests that inhibiting the mitochondrial
ETC or activating mitochondrial biogenesis decreases sFlt-1 secretion. In
addition, we also show that potential preeclampsia therapeutics metformin
and esomeprazole may be decreasing sFlt-1 via this mechanism.

Friday Posters

A Non-Human Primate Model of Defective Uterine Spiral Artery
Remodeling That Replicates Symptoms of Preeclampsia (PE) and
Fetal Growth Restriction (FGR). ED Albrecht*,2 GJ Pepe*.1 1Eastern
Virginia Medical School, Norfolk, VA, United States; 2Univ MD, Baltimore,
MD, United States.
INTRODUCTION: During human pregnancy, uterine spiral arteries
are remodeled (UAR) into low resistance vessels to promote placental
and fetal development. A defect in UAR underlies PE and FGR, which
result in maternal and neonatal morbidity/mortality. An animal model
of impaired UAR is needed to investigate UAR regulation and develop
methods to manage PE and FGR. Rodent models replicate PE and FGR
symptoms, but not a defect in UAR, and rodents and humans differ in
uteroplacental anatomy and UAR. We have established the baboon as a
translational model and since estradiol (E2) regulates cellular remodeling
processes, we hypothesized that elevating E2 in early pregnancy would
elicit a model of defective UAR.
METHODS: Baboons (Papio anubis) were treated with E2 (25 μg/
kg BW, sc) daily on d25-59 of gestation (term=184d). Placentas were
cesarean delivered on d60 or 165 and UAR quantified by image analysis,
VEGF protein by proximity ligation assay, and sFlt-1 mRNA by RT/PCR.
Statistical analyses: ANOVA/Newman-Keuls.
RESULTS: Serum E2 on d60 was 0.15 ± 0.01 ng/ml in untreated (n=14)
and 0.75 ± 0.07 in E2-treated (n=9) baboons. The percentage of uterine
spiral arteries >25 μm dia remodeled by d60 was 5-fold lower (P<0.01)
in E2-treated (6.5 ± 1.8) vs untreated (30.6 ± 7.2) baboons. VEGF protein
in placental anchoring villi of E2-treated baboons (4.9 ± 1.1 signals x 104,
n=7) was 50% lower (P<0.05) vs untreated (n=8) baboons. In contrast,
sFlt-1 mRNA in the trophoblastic shell (55 ± 17 units/18S rRNA) and
levels in uterine vein (1,127 ± 197 pg/ml) were ~3-fold greater (P<0.05)
in E2-treated vs untreated animals on d60. Maternal vascular endothelial
function, assessed by flow-mediated dilation, was lower (P<0.05) and
arterial blood pressure 20% higher (P<0.01) with E2 treatment. Uterine
artery blood flow quantified by Doppler on d160 was lower (P<0.05) in
E2-treated (9.9 ± 1.7 ml/min/kg BW) vs untreated (13.2 ± 2.5) animals.
Fetal body weight on d165 was 12% lower (P<0.05) in E2-treated (755
± 68 gm, n=9) vs untreated (858 ± 19 gm, n=14) animals.
CONCLUSION: An innovative paradigm of elevating E2 in early baboon
pregnancy was established and characterized by defective UAR and
hallmarks of human PE and FGR, including increased placental sFlt-1 and
decreased placental VEGF expression, increased maternal blood pressure,
decreased uterine arterial flow and FGR. This nonhuman primate paradigm
provides a model to elucidate UAR regulation, establish noninvasive
imaging technologies to detect defective UAR and develop methods to
manage adverse conditions of pregnancy arising from impaired UAR.
Supported by NIH R01 HD93070.

Reproductive Sciences Vol. 25, Supplement 1, March 2018

Table of Contents for the Digital Edition of SRI Supplement to Reproductive Sciences - Volume 25 Number 1 - March 2018