Original Research

ajog.org

OBSTETRICS

Decorin over-expression by decidual cells in preeclampsia: a potential blood biomarker Mohammad F. Siddiqui, MSc1; Pinki Nandi, PhD1; Gannareddy V. Girish, PhD; Karen Nygard, AHT; Genevieve Eastabrook, MD; Barbra de Vrijer, MD; Victor K. M. Han, MD; Peeyush K. Lala, MD, PhD

BACKGROUND: Decorin, a leucine-rich proteoglycan that is produced

by decidual cells, limits invasion and endovascular differentiation of extravillous trophoblast cells during early placentation by binding to multiple tyrosine kinase receptors, in particular, vascular endothelial growth factor receptor-2. OBJECTIVE: Because many studies have reported an association between poor trophoblast invasion and endovascular differentiation with preeclampsia, the studies reported here tested (1) whether decorin overexpression in the chorionic villi and/or basal decidua is associated with preeclampsia and, if so, (2) whether this association results in a hypoinvasive placenta, and (3) whether elevated plasma decorin concentration in the second trimester is a predictive biomarker for preeclampsia. STUDY DESIGN: Decorin messenger RNA expression was measured with quantitative polymerase chain reaction at the tissue level and with in situ hybridization at the cellular level using 35S-labeled antisense complimentary RNA probe in placentas from healthy control subjects and subjects with preeclampsia (14 each, 23-40 weeks of gestation). Tissue sections of the same placentas were also immunostained for decorin protein. A decorin over-expressing human endometrial stromal cell line was tested for invasion-regulatory effects on an invasive first-trimester extravillous trophoblast cell line HTR-8/SVneo plated in cocultures that were separated by a semipermeable membrane. Furthermore, we conducted retrospective measurements of plasma decorin levels during the second trimester (15-18 weeks of gestation) in a cohort of 28 body mass indexematched pairs of control subjects and subjects with preeclampsia before the onset of clinical disease. RESULTS: First, decorin messenger RNA expression at the cellular level measured with in situ hybridization exhibited profoundly higher expression levels in basal plate decidual cells within the placentas from preeclamptic

P

reeclampsia is a serious pregnancyassociated ailment that afflicts approximately 8e9% of pregnancies in North America and 3e5% pregnancies globally. It is a leading contributor to maternal morbidity, death, and premature birth.1-3 It generally is accepted

Cite this article as: Siddiqui MF, Nandi P, Girish GV, et al. Decorin over-expression by decidual cells in preeclampsia: a potential blood biomarker. Am J Obstet Gynecol 2016;215:361.e1-15. 0002-9378/$36.00 ª 2016 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ajog.2016.03.020

subjects than those from control subjects at all gestational ages, whereas no difference between the 2 subject groups was noted in villus mesenchymal cells. Similarly decorin messenger RNA expression at the tissue level in chorionic villi (primarily resulting from fetally derived mesenchymal cells) did not differ significantly between control and preeclampsia placentas. These findings were validated with immunostaining for decorin protein. Second, knocking down decorin gene in a decorin overexpressing endometrial cell line (used as an in vitro surrogate of decorin over-expressing decidual cells) in cocultures with extravillous trophoblast cells abrogated its invasion-restraining actions on trophoblast cells, which indicated paracrine contribution of decorin over-expressing decidua to the poor trophoblast invasiveness in situ. Finally, retrospective measurement of plasma decorin levels during the second trimester in 28 body mass indexematched pairs of control subjects and subjects with preeclampsia revealed elevated plasma decorin levels in all subjects with preeclampsia in all body mass index groups. A receiver operating characteristic curve analysis revealed strong diagnostic performance of plasma decorin in the prediction of preeclampsia status. Although there was no significant gestational age-related change in decorin levels during the second trimester in control or subjects with preeclampsia, we found that plasma decorin had a significant inverse relationship with body mass index or bodyweight. CONCLUSION: We conclude that decorin over-expression by basal decidual cells is associated with hypoinvasive phenotype and poor endovascular differentiation of trophoblast cells in preeclampsia and that elevated plasma decorin concentration is a potential predictive biomarker for preeclampsia before the onset of clinical signs. Key words: biomarker, decidual cell, decorin, extravillous trophoblast,

preeclampsia

that preeclampsia is a multifactorial syndrome, with many of the causative factors being associated with a hypoinvasive placenta.4,5 The clinical signs typically are manifested as hypertension after the 20th week of pregnancy, in an otherwise normotensive mother, in association with proteinuria. Although conservative treatments can reduce the severity of disease, it cannot be cured without the delivery of the placenta, which indicates a placental origin of the disease.1,2,4 The disease typically is divided into 2 stages: an early stage (stage 1) of compromised trophoblast invasion and a later stage (stage 2) of

maternal endothelial dysfunction, most likely the result of circulating toxic molecules that are released from a poorly perfused placenta.5 Stage 1 is insufficient to cause the maternal syndrome, but it interacts with maternal constitutional factors (genetic or environmental) to result in stage 2 disease. Molecules related to oxidative stress, inflammatory cytokines, and angiogenesis inhibitors have been identified in the maternal circulation, some of which have been related to maternal endothelial dysfunction and some of which have been proposed as biomarkers.5-7 A wellstudied angiogenesis inhibitor, sFlt-1,

SEPTEMBER 2016 American Journal of Obstetrics & Gynecology

361.e1

Original Research

ajog.org

OBSTETRICS

TABLE 1

Clinical characteristics of control subjects and subjects with preeclampsia included for decorin quantitative real-time polymerase chain reaction and in situ hybridizationa P value Control group Preeclampsia (normal vs (n ¼ 14) group (n ¼ 14) preeclampsia)b

Variable Mean maternal age at delivery, y  SD

27.8  7.0

27.5  6.8

>.05

Mean gestational age at delivery, wk  SD

30.6  5.3

31.3  5.3

>.05

Systolic blood pressure, mm Hg  SD

124  12

174  17

<.0000

Diastolic blood pressure, mm Hg  SD

76  8

108  12

<.0000

Proteinuria (n ‘þ’)

c

Mode of birth: vaginal birth:low segment cesarean section Mean birthweight, g  SD

0.0  0.0

3.3  0.61

9:5

7:7

1707  1028 1536  984

Intrauterine growth restriction, n (%) Fetal gender: male, n (%)

0

<.0000

>.05

3 (21.4)

10 (71.4)

9 (64.3)

a

Additional subjects (2 control and 2 preeclamptic) included for quantitative real-time polymerase chain reaction alone had the following features, respectively: maternal age: 31  1 years old, 30  2 years old; gestational age: 28  1 weeks, 27  1 weeks; mode of birth: 0:2, 0:2; intrauterine growth restriction: 0,1; b Obtained by unpaired t-test; c Proteinuria as measured by number of ‘þ’ on dipstick; e, þ, þþ, þþþ, þþþþ. Siddiqui et al. Decorin, a potential biomarker in preeclampsia. Am J Obstet Gynecol 2016.

has been identified as a predictive biomarker that is implicated in the pathogenesis of preeclampsia.7,8 However, sFlt-1 has been proposed recently instead as a protective mechanism in preeclampsia as a trophoblast-secreted response to vascular endothelial growth factor (VEGF) over-expression in the preeclampsia decidua.9 Further studies of molecules that are involved in the pathogenesis of preeclampsia are necessary to identify novel biomarkers for this multifactorial disease. The hemochorial placenta, including the human, is a highly invasive structure, in which a trophoblast subpopulation, the extravillous trophoblast (EVT) invades the decidua and uteroplacental (spiral) arteries that feed the placenta.10-12 During arterial invasion, EVT cells adopt an endovascular phenotype to replace the endothelial lining of the arteries and modify them from narrow, high-resistance contractile vessels to wide, low-resistance and flaccid tubes that allow unhindered flow of maternal blood to the placenta. During normal placentation, the spiral arteries lose their smooth muscle coat

within the endometrium and a portion of the myometrium. Poor EVT invasion and endovascular differentiation has been linked with an inadequate arterial modification that results in poor perfusion of the placenta with maternal blood. This, in turn, is believed to result in intrauterine growth restriction of the fetus and/or preeclampsia in the mother.4,13,14 We and others have shown that, unlike tumor invasion, EVT invasion of the uterus is exquisitely regulated in situ, both positively and negatively, by a large number of locally produced molecules that include growth factors, growth factor binding proteins, proteoglycans, and lipid derivatives to maintain a healthy utero-lacental homeostasis.4,5,15 EVT cell proliferation, migration, and invasiveness were shown to be regulated negatively by 2 decidua-derived factors: transforming growth factor beta (TGFb)16 and decorin, a secreted proteoglycan colocalized with TGFb in the decidual extracellular matrix.17,18 The negative regulatory actions of decorin on EVT cells were shown to be mediated differentially by multiple tyrosine kinase receptors epidermal

361.e2 American Journal of Obstetrics & Gynecology SEPTEMBER 2016

growth factor receptor, insulin-like growth factor receptor-1, and vascular endothelial growth factor receptor-2 (VEGFR)-2.19,20 We discovered that decorin is an antagonistic ligand for VEGFR-2 and that a 12 amino acid peptide within the leucine-rich-repeat 5 domain of decorin protein was responsible for the most avid VEGFR-2 binding, antagonizing VEGF actions on the EVT.20 We further showed that VEGF-A promoted endovascular differentiation of the EVT by binding to VEGFR-2, whereas decorin blocked the VEGFR-2e dependent endovascular differentiation of the EVT by blocking both ERK1/ERK2 and P38 MAPK pathways.21 We proposed that decorin over-expression or overactivity in the decidua may contribute to preeclampsia in the mother by a dual mechanism: decorin action on the EVT that leads to a hypoinvasive phenotype and decorin action on endothelial cells that inhibit angiogenesis.20-22 We explored the following objectives: (1) whether decorin over-expression in the decidua in situ is associated with preeclampsia pregnancies; (2) whether this was causal in EVT hypoinvasiveness that was reported for preeclampsia placentas in situ, and (3) whether elevated decorin levels in the maternal plasma that was collected during the second trimester potentially could predict preeclampsia.

Materials and Methods Placenta tissue collection Placentas that were obtained from the tissue biobank of the Translational Research Centre of the Children’s Health Research Institute were collected at preterm or near-term deliveries at the Department of Obstetrics and Gynecology, University of Western Ontario. The institutional Research Ethics Board approved tissue collection and the study protocol. Table 1 presents the clinical characteristics of the control and preeclampsia placentas that were used for decorin messenger RNA (mRNA) quantification in chorionic villi at the tissue level by quantitative real-time polymerase chain reaction (qPCR), and at the cellular level by in situ hybridization (ISH). All subjects with

ajog.org preeclampsia met criteria based on International Society for the study of Hypertension in Pregnancy and The Society of Obstetricians and Gynaecologists of Canada diagnostic criteria for the disease.23,24 Subjects with any form of preexisting hypertension or proteinuria were excluded from the control group, and subject with diabetes mellitus were excluded from both groups. Preterm delivery occurred in control subjects for indications that included spontaneous preterm labor, preterm rupture of membranes, antepartum hemorrhage, or nonreassuring fetal monitoring in the absence of intrauterine growth restriction or preeclampsia. Freshly obtained placental tissues that included chorion and decidua were either frozen in liquid nitrogen (for qPCR) or immediately fixed in 10% formalin and then embedded in paraffin (for ISH). Frozen chorionic villus tissues from healthy control and preeclamptic pregnancies (n ¼ 16 each) of matching gestational ages (23e40 weeks) were used to measure decorin expression by qPCR. For analysis, samples were ungrouped or grouped into early (23e27 weeks) or late (28e40 weeks) gestation, (n ¼ 8 each). Fixed placental tissues inclusive of basal decidua from 14 control and 14 preeclamptic pregnancies were used for ISH of decorin mRNA, because 2 samples in each group mentioned earlier yielded insufficient tissue for further analysis. For ISH analysis, tissues were divided into 2 gestational age groups: 6 each in group 1 (23e27 weeks), and 8 each in group 2 (28e40 weeks). Tissues subjected to ISH were also used for immunohistochemistry to visualize decorin protein.

Decorin mRNA expression in the placenta using qPCR Liquid nitrogen-frozen chorionic villi were crushed to powder form. The total RNA was isolated by Trizol-chloroform method, and qPCR was performed on a LightCycler (Bio-Rad Laboratories Inc, Mississauga, Ontario, Canada) with the use of custom-designed primers for human decorin (forward primer GACAACAACAAGCTTACCAGAGTA, reverse primer TGAAAAGACTCACA

OBSTETRICS

CCCGAATAAG, product size: 163 base pair) and human b-actin (forward primer GAAATCGTGCGTGACATTAAG GAG, reverse primer ATGATGGAGTT GAAGGTAGTTTCG, product size: 230 base pair) with the use of the SYBR Green PCR mix (Quanta Bioscience, Gaithersburg, MD). The respective amplicon length was verified by running the PCR product on 1% agarose gel. DDCt method was used to determine decorin mRNA expression relative to b-actin mRNA for each sample.25

Decorin mRNA expression quantitated by ISH Riboprobe preparation

We used the ISH protocol as reported earlier,26 with minor modifications. Human decorin transcript variant A1 DNA (gene bank ref ID: NM_001920.3) from a parent vector was used to amplify 1080 base pair of the decorin coding region. Forward and reverse primers for PCR sequencing with appropriate flanking restriction enzymes were designed respectively as 50 AAAGGTA CCATGAAGGCCACTATCATCCTCC30 and 50 GGCGTCTAGATTACTTATAGT TTCCGAGTTGAATGG 30 (OligoPerfect designer; Life Technologies, Rockville, MD). The PCR products were gelpurified (Qiagen, Valencia, CA) and subcloned into pGEM-4Z (Promega Corp, Madison, WI) vector, that contained SP6 and T7 RNA polymerase promoters flanking the multiple cloning regions. The resulting subcloned plasmid product, pGEM4Z-decorin, was validated by running on 1% agarose gel and sequencing (London Genomic Centre, Robarts Research Institute, London, ON). Transcription and radiolabeling of riboprobes

pGEM4Z-decorin was linearized at the appropriate digestion sites to produce the desired sense and anti-sense strands. SP6/T7 Riboprobe System- reagents and protocol (Promega Corp) was used for in vitro transcription of sense and antisense probes. Transcription and radiolabeling of 35S eUTP (1200 Ci/mmol; Amersham SJ 1303; Amersham International PLC, Buckinghamshire, UK) were done in a single reaction of 10-mL

Original Research

reaction volume each for sense and antisense probes. Specific activity of probes was measured as the total incorporated 35 S counts per minute per microliter and stored at e80 C. Tissue preparation for ISH

Sections (8-mm thick) of fixed and paraffin-embedded placental tissues were deparaffinized, rehydrated, and, after prehybridization at 45 C, hybridized with 35S -labeled anti-sense and sense complementary RNA probes (to the count of 10106 cpm/mL of incorporated 35S) overnight at 60 C and washed repeatedly with standard saline citrate solution of different strengths at different temperature, the final maximum stringency being 0.1standard saline citrate at 60 C for 10 minutes. Autoradiography

Sections were then dehydrated, coated with Kodak NTB nuclear track emulsion (Eastman Kodak Co, Rochester, NY) and exposed at 4 C for 7-14 days to assess labeling intensity. They were developed with D-19 developer (Eastman Kodak Co), fixed, stained with Harris’s hematoxylin and eosin, and mounted with Permount (Fisher Scientific, Pittsburgh, PA). Definitive analysis was done with slides that had been exposed for 14 days. Image analysis

Complete image capturing and analysis was done at Integrated Microscopy (Biotron, University of Western Ontario) using Upright Zeiss AxioImager Z1 (Carl Zeiss Co, Jena, Germany). The analysts were blinded to the identification of the slides. To ensure crisp image capturing of all exposed silver grains in the emulsion above the hybridized cells, the “Stitch and Tile” and “Extended Depth of Focus” modules of the Zeiss’s Zen imaging software (Carl Zeiss Co) were used. Quantification of grains

This was done in a blinded manner with the use of Image Pro Premier software (Media Cybernetics LP, Silver Spring, MD). Five sections (at 10 mm apart from each other) for each tissue were scanned for counting silver grains on all cells (identified by nuclear stain) in randomly

SEPTEMBER 2016 American Journal of Obstetrics & Gynecology

361.e3

Original Research

OBSTETRICS

selected fields of view (with a 100100 mm grid) to derive the average number of silver grains per cell labeled with antisense (experimental) and sense (controls, providing the background labeling) probes.

Immunohistochemistry for cytokeratin7 (CK7) and decorin protein To discriminate EVT cells from decidual cells in the basal plate and trophoblast cells lining the chorionic villi from the fetal mesenchymal cells within the villus core, randomly selected sections (n ¼ 10) used for ISH were deparaffinized, subjected to antigen retrieval (by boiling for 10 minutes in citrate buffer [0.01 M, pH 6.0]) and immunostained with a CK7 monoclonal antibody (1: 250; Abcam, Toronto, Ontario, Canada). Tissue sections treated with immunoglobulin G of the same isotype in phosphate-buffered saline (PBS) in lieu of the primary antibody served as a negative control. To gather semiquantitative information on the expression of decorin protein at the cellular level, deparaffinized tissues were subjected to immunohistochemistry after antigen retrieval as mentioned earlier. The sections were washed for 10 minutes in 3% H2O2/PBS (25 min) and pretreated with avidin, followed by biotin (to block the remaining biotin binding sites on the avidin) followed by treatment with a universal blocking reagent (Background Sniper BS966; Biocare Medical, Concord, CA) for 10 minutes. Sections were then incubated with rabbit polyclonal antiedecorin antibody (ab67449; 1:250; Abcam) overnight at 4 C. Rabbit immunoglobulin G replacing primary antibody was used as negative control. PBS washes preceded all subsequent steps: incubation with biotinylated horse anti-rabbit secondary antibody (BA1100; Vector Laboratories, Burlington, ON) at 1:250 dilution for 30 minutes; streptavidin/peroxidase solution, avidin-biotin complex, in PBS (PK-4001; Vector Laboratories) for 45 minutes; DAB substrate, enhancer, and chromogen for 2 minutes, followed by counterstaining with methyl green.

EVT and human endometrial stromal cell (HESC) lines To test whether decorin over-expression in the preeclampsia decidua is associated causally with hypoinvasive phenotype of EVT cells in situ, we used an in vitro system to measure invasiveness of firsttrimester EVT cells in cocultures with a decorin over-expressing HESC line. The human EVT cell line HTR-8/SVneo was derived in our laboratory by SV40Tag immortalization of a short-lived primary first-trimester cell line HTR-827 that was produced from a chorionic villus explant in culture, as reported with methodologic details and phenotypes.28 The immortalized cell line was 100% positive for CK7 and expressed HLA-G, as verified recently.20 The cells express no measurable decorin. An immortalized HESC cell line, T-HESC, was obtained from the American Type Culture Collection, Rockville, MD (CRL-4003). These cells expressed a very high level of decorin mRNA, as revealed with qPCR. We knocked down decorin mRNA expression using small interfering RNA (siRNA) transfection (Santa Cruz Biotechnology Inc, Santa Cruz, CA), resulting in 76% knockdown (KD) of decorin mRNA (measured with qPCR), as compared with scramble siRNA. The decorin down-regulated cells were named as HESC-decorin-KD.

Invasion assays Both control (scramble siRNA transfected) and HESC-decorin-KD cells were used as regulator cells in EVT (HTR-8/ SVneo) cell invasion assays conducted in 24-well transwell inserts, as reported earlier for migration assays.21 The conditions for invasion assay29 were exactly the same as migration assay, except that the membranes were coated with a thin layer of growth factor reduced Matrigel so that the cells had to degrade the Matrigel barrier before migrating through the pores of the membranes. EVT cells at 70% confluency were serum starved overnight, and 60,000 cells in serum-free medium were plated onto the top chamber of the cell culture inserts. Fetal bovine serum (5%) was added to the lower chamber as stimulant. The

361.e4 American Journal of Obstetrics & Gynecology SEPTEMBER 2016

ajog.org bottom chambers contained no cell or control (scramble siRNA transfected) or HESC-decorin-KD cells at 1:1 and 1:2 responder (EVT)/regulatory (HESC) cell ratios. T-HESC cells were allowed to settle in the bottom chamber for 2 hours before the addition of EVT cells in the inserts of upper chambers. The assays were conducted for 48 hours at 37 C in 5% CO2, when the invasion reached a peak followed by a decline. The upper surfaces of the membranes were then wiped gently with cotton swabs to remove the cells that have not yet crossed the membranes. The membranes were then fixed with methanol, stained with 0.1% eosin, and counter-stained with methylene blue; the absolute number of invasive cells was scored visually with the use of a light microscope.

Expression of sFlt-1 in HTR-8/ SVneo cells in the presence of exogenous decorin To test whether decorin can regulate sFlt-1 expression by EVT cells, HTR-8/ SVneo cells (105 cells per well in 6 wellplates), serum starved overnight were treated with bovine cartilage-derived decorin (200 nmol/L; Sigma-Aldrich, St. Louis, MO) for 6 hours and washed with PBS. RNA was isolated, and sFlt-1 mRNA levels were measured with qPCR in treated and untreated cells.

Measurement of plasma decorin levels in control subjects and subjects with preeclampsia by enzyme-linked immunosorbent assay (ELISA) Selection of subjects used for plasma decorin measurements

A retrospective measurement of plasma decorin levels during the second trimester of pregnancy (15-18 weeks of gestation) was done in a cohort of control subjects and subjects with preeclampsia who were matched for body mass index (BMI) class. BMI was calculated with the formula: BMI¼ kg/m2. BMI classes were stratified as normal weight (BMI¼18.5e24.9 kg/m2), overweight (BMI¼25.0e29.9 kg/m2), and obese (BMI30.0 kg/m2). Patient information was entered

ajog.org prospectively into a combined prenatal diagnosis, and perinatal database and second-trimester blood samples were retrieved from 28 women with preeclampsia and 28 BMI class-matched control subjects. Patients with preexisting hypertension, renal or autoimmune disease, diabetes mellitus, age <18 or >40 years, BMI<18 kg/m2, multiple pregnancies, and fetal congenital anomalies were excluded. The resulting dataset was found to be predominantly white; thus, non-white subjects were excluded to avoid any possible effect of race on circulating levels of decorin, as has been described with other placental markers.30 The study was approved by the University of Western Ontario Health Sciences Research Ethics Board (HSREB100348). At the time of sampling, none of the subjects had clinical features of preeclampsia. Plasma samples were stored at e80 C to measure a panel of putative biomarkers. Decorin ELISA

Plasma decorin levels were measured with an ELISA kit from Sigma-Aldrich (RAB0140) using their protocol. This assay used anti-decorin capture antibody coated on a 96-well plate. Plasma samples were diluted to 1:500 in the diluents provided. Standards (recombinant human decorin provided in the kit) and samples were pipetted into the wells in duplicate, and the target protein present in a sample was bound to the wells by the immobilized antibody. The wells were washed, and a biotinylated antidecorin antibody was added. After unbound biotinylated antibody was washed away, Horseradish peroxidaseconjugated streptavidin was pipetted to the wells. The wells were again washed, and 3,3’5,5’- tetramethylbenzidine substrate solution and subsequently Stop solution were added to the wells. Level of bound decorin protein, in proportion to the intensity of developed color was measured using the plate reader at 450 nm. With the optical density values obtained, the standard curve was plotted with the use of SigmaPlot software (SigmaStat, San Jose, CA), and the unknown concentrations in the samples were derived.

OBSTETRICS

Statistical analysis Data in this study are described as meanSD or SEM as specified in the results and analyzed with the use of the GraphPad Prism statistical software (version 6; GraphPad Software, Inc, San Diego, CA) for most of the results. Unpaired 2-tailed t-test (with Welch’s corrections) was used for ISH data. Silver grain count distribution per cell as a measure of decorin mRNA by ISH revealed close to normal distribution so that mean grain counts were used for the final analyses. Linear regression was used for raw decorin values in plasma drawn at different gestational ages in control subjects and subjects with preeclampsia to identify any significant change over the period of gestation at sampling (15-20 weeks of gestation). The same was used to identify any significant relationship between plasma decorin values and subject BMI or bodyweight. For comparison of plasma decorin levels between control subjects and subjects with preeclampsia, either regardless of BMI values or stratified into BMI classes (normal weight, overweight, or obese), data were analyzed with 2-tailed MannWhitney test with the use of median and quartile distributions. To determine the predictive performance of decorin on preeclampsia status, control for BMI, or bodyweight, SPSS software (version 23; IBM Corporation, Armonk, NY) was used to conduct a logistic regression. Decorin, BMI, and bodyweight were entered as the predictors, and preeclampsia status was entered as the outcome. Finally, the diagnostic performance of plasma decorin in the prediction of preeclampsia status was estimated through a receiver operating characteristic (ROC) curve analysis. Probability values at <.05 were considered significant for all data.

Results Clinical characteristics of control subjects and subjects with preeclampsia used for decorin mRNA expression in placentas Clinical characteristics are summarized in Table 1. Although 32 subjects (16 control and 16 preeclampsia) were used for qPCR quantitation of decorin

Original Research

mRNA at the tissue level in chorionic villi, with sufficient samples being unavailable from 4 subjects (2 control, 2 preeclampsia), this allowed for decorin mRNA measurements at the cellular level by ISH in 28 subjects (14 control, 14 preeclampsia).

Decorin mRNA measured with qPCR in chorionic villous samples The qPCR data for decorin mRNA were expressed relative to b-actin mRNA, which has been reported to be unchanged in preeclampsia placentas.31 Figure 1 presents the data in preeclampsia and healthy pregnancies (n¼16 each) pooled for all gestational ages in Figure 1, A, or grouped into early (<28 weeks) and late (28 weeks) gestation in Figures 1, B and C, respectively. No significant difference was noted between preeclampsia and control placentas in any of the groups. Follow-up ISH studies were done to measure decorin mRNA at the cellular level in chorionic villi and basal plate decidua.

CK7 staining to distinguish trophoblast from mesenchymal and decidual cells in sections subjected to ISH Figures 2, GeI illustrate positive CK7 staining of trophoblast cells that were lining the chorionic villi and some EVT cells in superficial decidua basalis. As expected, neither decidual cells nor fetal mesenchymal cells were stained for CK7.

Quantification of silver grains in radio-autographs of sections counterstained with hematoxylin and eosin Cell-specific expression of decorin mRNA was assessed by ISH with the use of 35S-labeled antisense riboprobe; the background signal was provided by the sense riboprobe. Quantification of mRNA expression, given by the density of silver grains, was assessed by multiple fields of view of magnified images from both the groups in a blinded manner. Number of silver grains per cell above background provided the relative level of decorin mRNA expression. In the

SEPTEMBER 2016 American Journal of Obstetrics & Gynecology

361.e5

Original Research

ajog.org

OBSTETRICS

FIGURE 1

Decorin messenger RNA in chorionic villi measured with quantitative real-time polymerase chain reaction

Comparison of decorin messenger RNA expression measured with quantitative real-time polymerase chain reaction in composite cell populations of human chorionic villus samples of preeclampsia and control placentas of matched gestational ages: A, all ages; B, <28 weeks gestation; C, ‡28 weeks gestation. No significant difference (P > .2) was noted in any of these groups. mRNA, messenger RNA; PE, preeclampsia. Siddiqui et al. Decorin, a potential biomarker in preeclampsia. Am J Obstet Gynecol 2016.

chorionic villi, decorin mRNA was expressed in chorionic mesenchymal cells and not in the villous trophoblast. In basal plate decidua, decorin mRNA was expressed in the decidual and not in the EVT cells. Decorin mRNA was observed in much greater abundance, as was revealed by the density of silver grains, in decidual cells of the basal plate decidua of preeclampsia placentas compared with controls (Figure 2, AeF). The mean grain counts per cell (SE) presented either as gestation age groups (<28 or 28 weeks gestation) or ungrouped (Figure 2, J) demonstrate significant up-regulation of decorin mRNA in decidual cells of the basal plate decidua from preeclampsia patients compared with those in the same region of placentas from gestational ageematched control subjects. In contrast to decidual cells, there was no visible difference in decorin mRNA abundance (noted as low silver grain density above background) between the control and preeclampsia groups in corresponding chorionic villus mesenchymal cells (Figures 3,AeD). Quantitation of silver grains in fetal mesenchymal cells did not show any differential expression of decorin mRNA between preeclampsia and control (Figure 3, E). This finding is consistent with the qPCR results presented earlier.

Comparison of decorin protein expression by immunohistochemistry Immunohistochemically identified decorin protein was localized to the chorionic villus mesenchymal cells and decidual cells in placentas at all gestational ages in control subjects, as reported earlier.16,18 Consistent with the data on ISH mRNA expression presented earlier, decidual cells in the basal plate examined in 10 randomly selected sections, displayed much stronger immunostaining for decorin protein in preeclampsia placentas, whereas villus mesenchymal cells displayed equally weak immunostaining in both control and preeclampsia placentas. Representative pictures of control and preeclampsia placentas, both at 31 weeks gestation, are shown in Figure 4.

Knocking down decorin gene in a high decorin-expressing HESC partially restores invasive functions of EVT cells in coculture To investigate whether decorin overexpression in the preeclampsia decidua is associated causally with an hypoinvasive phenotype of EVT cells in situ, we used an in vitro system of measuring invasiveness of EVT cells in the presence of a high decorineexpressing ESC line T-HESC and whether knocking down decorin expression in the stromal cells

361.e6 American Journal of Obstetrics & Gynecology SEPTEMBER 2016

could modulate EVT cell invasive capacity. Figure 5, A, demonstrates that T-HESC-decorin-KD cells had 76% down-regulation of decorin mRNA relative to control scramble siRNA transfected cells. Figure 5, B, presents the immunoblot for decorin protein (54% down-regulation). Figure 5, C, shows that presence of the high decorine expressing control T-HESC at 2 different cell concentrations in the bottom chamber of transwells reduced the native invasive capacity of EVT cells in a T-HESC cell concentration-dependent manner and that knocking down decorin significantly restored the invasive capacity of EVT cells.

Decorin treatment up-regulates sFlt-1 mRNA in EVT cells Because placentas from mothers with preeclampsia have been reported to exhibit up-regulated sFlt-1 mRNA expression in EVT cells in situ,9 we tested whether exogenous decorin treatment in vitro resulted in an increased sFlt-1 mRNA level in HTR-8/SVneo cells. As shown in Figure 6, treatment with decorin (200 nmol/L) resulted in a nearly 13-fold increase in sFlt-1 mRNA expression. Treating with VEGF-A, another molecule reported to be unregulated in preeclampsia decidua,9 also caused a minor up-regulation of sFlt-1 mRNA by 1.5-fold.

ajog.org

OBSTETRICS

Original Research

FIGURE 2

Decorin messenger RNA in basal plate decidua measured with in situ hybridization A

B

C

D

E

F

G

H

I

J

A, B, D, E, Brightfield and C, F, darkfield photomicrographs of A, B, C, normal and D, E, F, preeclamptic human placenta sections (all 23 weeks of gestation) hybridized in situ with 35S-labeled decorin complementary RNA probe. Magnification bars: red¼20 mm; black¼160 mm. The decidual basalis region in preeclampsia placental sections is shown expressing abundant decorin messenger RNA (reflected by grain density) as compared with the decidua of normal placenta. G, H, I, Cytokeratin-7 staining reveals that trophoblast cells that line chorionic villi (and invade extravillous trophoblast cells) in superficial decidua basalis show immunoreactivity to cytokeratin-7, whereas decidual cells are not stained. Quantification of silver grains per decidual cell in the basal plate of preeclamptic placentas relative to J, controls shows significant over-expression of decorin in the preeclampsia deciduas in group 1 (n ¼ 6; 23e27 wks; P ¼ .034), group 2 (n ¼ 8; 28e40 weeks of gestation; P ¼ .049) and in the combined group (n ¼ 14; 23e40 wks; P ¼ .009) compared with their respective controls of matched gestational ages. Data represent mean  standard deviation. A single asterisk indicates P < 0.05; the double asterisk indicates P < 0.0.01. CF, chorionic villi; db, decidual basalis region; EVT, extravillous trophoblast. Siddiqui et al. Decorin, a potential biomarker in preeclampsia. Am J Obstet Gynecol 2016.

SEPTEMBER 2016 American Journal of Obstetrics & Gynecology

361.e7

Original Research

OBSTETRICS

FIGURE 3

Decorin messenger RNA in villus mesenchymal cells measured with in situ hybridization

B

A

D

C

the appearance of clinical signs and whether there was a relationship between maternal BMI class and plasma decorin.

Clinical characteristics of control subjects and subjects with preeclampsia used for plasma decorin measurement We performed a retrospective measurement of plasma decorin during the second trimester (15-18 weeks of gestation) in 28 subjects with preeclampsia and 28 control subjects who were matched for BMI class. Preeclampsia was diagnosed using the ISSHP and SOCG guidelines,23,24 as previously described. Clinical characteristics of these subjects are presented in Table 2. As expected, the subjects with preeclampsia were delivered at a significantly lower gestational age and had lower birthweights.

Elevated plasma decorin level in mothers during the second trimester is predictive of preeclampsia

E

A, B, Brightfield and C, D, darkfield photomicrographs of A, C, control and B, D, preeclamptic human placenta sections hybridized in situ with 35S-labeled decorin complementary RNA probe. Magnification bar: yellow¼20 mm. Chorionic villus mesenchymal cells (black arrowheads) in both control and preeclamptic placental sections show equally sparse distribution of silver grains (yellow circle). E, The grain count on quantification shows no significant difference (P > .2) between the preeclampsia and control groups. Data represent mean  standard deviation. PE, preeclampsia. Siddiqui et al. Decorin, a potential biomarker in preeclampsia. Am J Obstet Gynecol 2016.

Our in situ data that showed decorin over-expression in basal plate decidual cells at all gestational ages in clinically identified preeclampsia and in vitro data of high decorin-expressing uterine ESCs down-regulating EVT cell invasion suggested that decorin overexpression in situ was the cause of hypo-invasive phenotype of EVT that is associated with preeclampsia. Decorin has also been shown to be

ajog.org

antiangiogenic.21,22 Furthermore, as shown earlier, sFlt-1 up-regulation in decorin-treated EVT cells in vitro simulated preeclampsia-associated EVT cells in situ.9 Decorin was also reported to be elevated moderately in the blood plasma of obese subjects and patients with type-2 diabetes mellitus.32 These findings prompted us to explore whether decorin is an additional plasma biomarker for preeclampsia before

361.e8 American Journal of Obstetrics & Gynecology SEPTEMBER 2016

Because decorin can bind to fibrinogen, serum decorin measurements reveal only the free form of decorin. For this reason, we measured plasma decorin levels that included both free and fibrinogen-bound decorin. We tested both gestation and BMI-related changes in decorin. Figure 7, A, shows the scatter plots of all decorin values during the second trimester in these control subjects and subjects with preeclampsia, including the median and quartile distributions, irrespective of gestational ages (15-18 weeks) or the BMI values (20e41 kg/m2). Although there is an overlap in the raw values between the 2 groups, median plasma decorin levels were significantly elevated (P ¼.0016) in subjects with preeclampsia (median, 8.48 pg/mL) as compared with control subjects (median, 5.44 pg/mL) during the second trimester, regardless of BMI. The ratios of preeclampsia:control values of plasma decorin in 28 BMI-matched pairs were distributed normally, with a mean of 1.9  0.3 (SD), which was significantly higher than 1 (P < .0001), which indicates that BMI

ajog.org

OBSTETRICS

FIGURE 4

Decorin immunostaining in chorionic villus and decidua

Representative immunostaining of decorin protein in chorionic villus mesenchymal cells and decidual cells of the basal plate at 31 weeks of gestation in control and preeclamptic placentas. Magnification bar: black¼20 mm. Although there was no difference in the very weak immunopositivity of fetal mesenchymal cells in the chorionic villi in the 2 groups, decidual cells in the basal plate of the preeclamptic placenta show much stronger immunostaining. Siddiqui et al. Decorin, a potential biomarker in preeclampsia. Am J Obstet Gynecol 2016.

matching made the plasma decorin elevation in preeclampsia more significant (data not shown). Figure 7, B, shows individual plasma decorin values (picograms/milliliter) during the second trimester that were plotted against gestational ages at sampling in 28 control subjects and 28 subjects with preeclampsia and showed no significant gestation-related change in either group. However, the regression line for preeclampsia values was significantly elevated (P ¼ .009), as compared with control subjects. Next, we stratified the data into 3 BMI classes: normal weight, overweight, and obese using the criteria described earlier in the Methods section. Data in Figure 7, C, reveals that plasma decorin levels (showing median and quartile ranges) were significantly higher in subjects with preeclampsia than in control subjects across all BMI groups (P ¼ .007e.035). Seeing a trend in decreased decorin values among the

overweight and obese subjects as compared with those in the normal weight group both in control subjects and subjects with preeclampsia, we performed a regression analysis to examine whether there was a relationship between decorin and BMI. Plasma decorin level vs BMI values are shown in Figure 7, D, that shows a negative correlation in both control subjects and subjects with preeclampsia, not significantly different from each other. However, the values for preeclampsia were significantly higher than the control (slope elevation; P ¼ .0061). To test whether the major source of plasma decorin during pregnancy is the placenta, which does not appreciably change with maternal bodyweight, we performed a regression analysis of the relationship between decorin and maternal bodyweight. Plasma decorin level as function of bodyweight (Figure 7, E) revealed a negative correlation in both

Original Research

control subjects and subjects with preeclampsia that was not significantly different from each other. Again, the values for preeclampsia remained significantly higher than the control (slope elevation, P ¼ .0064). A logistic regression was conducted to determine the predictive performance of decorin on preeclampsia status when controlled for BMI or bodyweight. Decorin, BMI, and bodyweight were entered as predictors, and preeclampsia status was entered as the outcome. The full model significantly predicted preeclampsia status (omnibus chisquare ¼ 9.87; degrees of freedom ¼ 2; P ¼ .007) and accounted for 16e22% of the variance in preeclampsia status. In addition, an increase in decorin concentration of 1 pg/mL was associated with an increase in the odds of preeclampsia by a factor of 1.25 (95% confidence interval, 1.05e1.49; Wald ¼ 6.45; degrees of freedom ¼ 1; P ¼ .011), even after accounting for BMI, or bodyweight, which showed no significant effect. Finally, the diagnostic performance of plasma decorin in the prediction of preeclampsia status was estimated through an ROC curve analysis (Figure 7, F). The cutoff value of decorin that best predicted preeclampsia status by optimizing sensitivity and specificity was 5.96 pg/mL (area under the curve ¼ 0.75; SE ¼ 0.07; P ¼ .002; sensitivity, 89%; specificity, 61%). This resulted in a positive likelihood ratio of 2.27 and a negative likelihood ratio of 0.18.

Comment Primary findings We had earlier shown that exogenous decorin treatment blocked proliferative, migratory, and invasive abilities of EVT cells independent of TGFb by binding to multiple tyrosine kinase receptors that include VEGFR-219-21 and that decorin compromised VEGFR-2edependent endovascular differentiation of EVT cells.21 The present study revealed that (1) decorin over-expression in the basal decidua is associated with preeclampsia and is linked causally with the hypoinvasive phenotype of EVT cells and (2) elevated maternal plasma decorin level during the second trimester is

SEPTEMBER 2016 American Journal of Obstetrics & Gynecology

361.e9

Original Research

ajog.org

OBSTETRICS

FIGURE 5

FIGURE 6

Regulation of extravillous trophoblast cell invasion by the decorin level in human endometrial stromal cells

Regulation of sFLT1 messenger RNA in extravillous trophoblast cells by decorin and VEGF

Regulation of sFLT1 messenger RNA in extravillous trophoblast (HTR-8/SVneo) cells (measured with quantitative real-time polymerase chain reaction) by treatments with decorin (200 nmol/L) or VEGF-A (VEGF-121; 10 ng/mL). Significant up-regulation was noted with both treatments: A, 13-fold with decorin and B, 1.5-fold with VEGF-A. Data represent mean of triplicate measurements  standard error. DCN, decorin; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; HTR, human trophoblast; mRNA, messenger RNA; sFLT1, soluble fms-like tyrosine kinase-1; VEGF, vascular endothelial growth factor. Siddiqui et al. Decorin, a potential biomarker in preeclampsia. Am J Obstet Gynecol 2016.

A, Regulation of extravillous trophoblast cell invasiveness at 48 hours by decorin over-expressing human endometrial stromal cells. The asterisk indicates P < 0.05. B, Human endometrial stromal cellsedecorin knockdown cells show 76% knockdown of decorin messenger RNA relative to control siRNA-treated cells, resulting a down-regulation decorin protein shown in immunoblot. C, HTR-8/SVneo cells were plated on the top chamber on Matrigel-coated membrane inserts, and control (scramble siRNA transfected) or decorin knocked down human endometrial stromal cellsedecorin knockdown cells were placed at the bottom chamber at 1:1 or 1:2 extravillous trophoblast/human endometrial stromal cells ratios, in the presence of 5% fetal bovine serum. Presence of human endometrial stromal cells at latter ratio significantly (P < .05) restored the invasive capacity of extravillous trophoblast cells. The asterisk indicates P < 0.05. Data represent the mean of triplicate measurements  standard error. DCN, decorin; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; HESC, human endometrial stromal cells; mRNA, messenger RNA; Scr, scrambled; siRNA, small interfering RNA. Siddiqui et al. Decorin, a potential biomarker in preeclampsia. Am J Obstet Gynecol 2016.

potentially predictive of preeclampsia, irrespective of maternal BMI or bodyweight.

Decorin over-expression in the basal decidua is associated with preeclampsia and is linked causally with the hypo-invasive phenotype of EVT cells We found that decorin mRNA expression by composite cell populations of

the chorionic villi, as determined by qPCR, exhibited no significant difference between healthy and preeclamptic placentas, irrespective of gestational age. This was confirmed by quantitative ISH that showed equally low levels of decorin expression by villus mesenchymal cells in both groups. In contrast, decorin mRNA expression by decidual cells of the basal plate measured by ISH exhibited significantly higher

361.e10 American Journal of Obstetrics & Gynecology SEPTEMBER 2016

expression in placentas from preeclampsia compared with controls at all gestational ages that were examined. Immunohistochemical examination of decorin protein followed a similar pattern. These findings support the notion that decorin over-expression is 1 of the early pathogenic events at the fetal-maternal interface that lead to preeclampsia and that the disease is maternal and not fetal in origin. In the past, we have shown that exogenous decorin blocks EVT cell outgrowth and migration in first-trimester villus explants that are cultured on Matrigel.19 A cause-effect relationship between decorin over-expression in decidual cells and compromised trophoblast invasion in situ is strengthened further by our in vitro assay that shows that the presence of a high decorin-expressing ESC line THESC (used as a decidual cell surrogate) led to reduced invasive capacity of EVT cells and that EVT cell invasiveness was restored partially by knocking down decorin in the T-HESC cells used as regulators of invasiveness.

ajog.org Regulators of trophoblast invasion may appear as potential biomarkers for preeclampsia There is strong biochemical evidence of impaired EVT invasion of the endometrial stroma during early pregnancy in preeclampsia.33 Among a large number of locally derived molecules that positively or negatively regulate EVT invasion in situ,4 some have been suggested as biomarkers of preeclampsia. For example, IGFBP1 produced by decidual cells26 was shown to promote EVT cell invasion by binding to a5b1 integrin.34 Longitudinal studies suggested that a low plasma IGFBP1 level in the mother during early to mid pregnancy was predictive of preeclampsia, whereas the levels in near-term preeclampsia surpassed normal levels.35 Of the negative regulators, decidua-derived TGFb2 was shown to be a key inhibitor of EVT cell invasion,16,36 and high TGFb2 levels in the maternal serum was also suggested to be a biomarker of preeclampsia.37 Decorin is another negative regulator of EVT cell invasion and endovascular differentiation, which are compromised in preeclampsia. Present results that show that decorin expression was markedly high in preeclampsia decidua at all gestational ages led to the hypothesis that decorin is another potential predictive biomarker of preeclampsia. This hypothesis was fully validated by a retrospective analysis of plasma decorin levels during the second trimester in BMI-matched control subjects and subjects with preeclampsia, which shows significant elevation in subjects with preeclampsia, irrespective of gestational age, maternal BMI, or bodyweight. ROC curve analysis revealed elevated plasma decorin as a significant predictor of preeclampsia.

Decorin plays a direct role in compromising uterine angiogenesis, trophoblast invasiveness, and endovascular differentiation in preeclampsia and appears as a potential blood biomarker In recent years, many studies have identified high levels of anti-angiogenic molecules in the sera of mothers with

OBSTETRICS

Original Research

TABLE 2

Clinical characteristics of control subjects and subjects with preeclampsia were included for plasma decorin enzyme-linked immunosorbent assay during the second trimester (15e20 weeks of gestation) P value Control Preeclampsia (normal vs group (n ¼ 28) group (n ¼ 28) preeclampsia)a

Variable Mean maternal age at birth, y  SD

29.9  0.86

30.1  0.96

>.05 (.8773)

Prepregnancy BMI (kg/m )

28.0  1.74

27.9  1.43

>.05 (.9635)

Mean gestational age at birth, wk  SD

39.0  0.52

35.5  0.66

<.05 (.0001)

Mean gestational age at sampling, wk  SD

16.6  0.801

16.8  0.802 >.05 (.354)

Fetal gender: male:female

13:15

18:10

Mode of birth: vaginal birth:lower segment cesarean section

19:9

21:7

2

Mean birthweight, g  SD

3156  112

Intrauterine growth restriction, n (%)

3 (10.7)

2472  164

<.05 (.0011)

5 (17.8)

a

Obtained by unpaired t-test. Siddiqui et al. Decorin, a potential biomarker in preeclampsia. Am J Obstet Gynecol 2016.

preeclampsia.38 Some key examples are soluble endoglin39,40 and sFlt-1,41 the latter being a strong ligand for placenta growth factor (PLGF) and a weaker ligand for VEGF. It is a splice variant of Flt-1 (or VEGFR-1) protein, that can block angiogenesis by sequestering PLGF and VEGF. Low PLGF in the maternal serum has been linked with preeclampsia, and the sFLt-1/PLGF ratio is considered to be a reliable preeclampsia biomarker in maternal serum.41 Although compromised angiogenesis in the pregnant uterus has been suggested as a mechanism in the pathogenesis of preeclampsia, none of the antiangiogenic molecules reported so far in preeclampsia have been linked directly with poor invasion and endovascular differentiation of the EVT. In contrast, this mechanistic relationship has been noted with decorin.19-22 Soluble endoglin may compromise trophoblast differentiation in the EVT pathway by acting as a coreceptor for TGFb.42 Similarly, sFlt-1 may compromise trophoblast invasion indirectly by sequestering invasion-promoting PLGF or VEGF that is produced locally. SFlt-1, a trophoblast-secreted product, is believed to cause maternal endothelial dysfunction that leads to many of the

symptoms, such as hypertension and proteinuria.8 Reduction of circulating sFlt-1 alleviated preeclampsia-like symptoms in a mouse model,41 and a pilot study in human subjects that used extracorporeal apheresis to reduce circulating sFlt-1 in early-onset preeclampsia demonstrated some clinical benefit.43 However, a recent study9 raises caution against the rationale for removing circulating sFlt-1. These authors proposed that sFlt-1 is produced by EVT in response to excessive VEGF-A production by ESCs in preeclampsia pregnancy and that it circulates in the maternal blood as a defense mechanism. The authors created a mouse model of decidual VEGF-A overproduction to create a preeclampsia-like syndrome in which sFlt-1 was up-regulated in the trophoblast and knocking down sFlt-1 in this model ameliorated preeclampsia symptoms.9 SFlt-1 up-regulation in the trophoblast induced by VEGF-A results from modulation of alternate splicing of Flt-1.44 We have shown that decorin plays a dual role in the decidua by controlling VEGFR-2edependent endovascular differentiation of the EVT and also by inhibiting angiogenesis by blocking VEGF action on endothelial cells.20-22

SEPTEMBER 2016 American Journal of Obstetrics & Gynecology

361.e11

Original Research

OBSTETRICS

ajog.org

FIGURE 7

Plasma decorin levels in body mass indexematched control subjects and subjects with preeclampsia during the second trimester

Plasma DCN levels in BMI-matched control and PE subjects during the second trimester. A, Scatter plot of plasma DCN concentrations (ng/ml) during the second trimester in 28 pairs of control and PE subjects are plotted irrespective of gestational age (including median  quartile ranges), the median values in PE being significantly higher (P ¼ 0.0015, Mann-Whitney test) than control subjects. B, Plasma DCN concentrations (ng/ml) in BMI-matched control and PE subjects as a function of gestational age (16-18 weeks). The data were plotted irrespective of their BMI values. Neither control nor PE subjects revealed any significant gestational age-related change in DCN levels (regression slopes not different from zero). However the slope of elevation of the PE data is highly significant (P ¼ 0.009). C, Values of plasma DCN concentrations stratified according to BMI: normal weight (NW), overweight (OW), and obese (OB). Lines represent median (quartile ranges). PE values are significantly higher (P ¼ 0.007 to 0.035) in different BMI groups. D, Plasma DCN

361.e12 American Journal of Obstetrics & Gynecology SEPTEMBER 2016

ajog.org The present study further shows that exogenous decorin treatment resulted in a marked up-regulation of sFlt-1 in EVT cells. We suggest that decorin over-expression in the decidua is 1 mechanism that contributes to sFlt-1 up-regulation in the EVT cells that were noted in situ in preeclampsia placentas.9 Whether the overproduction of decorin in the decidua is a response to VEGF overproduction or vice versa or whether the 2 phenomena are completely unrelated remains to be investigated. Identification of 1 early biomarkers in maternal blood that can predict the development of preeclampsia is a goal that has been achieved only partially. For example, sFlt-1, sFlt-1/PlGF ratio,41 and soluble endoglin39 are now used as predictive biomarkers in some clinical settings. Present results reveal that decorin may be an additional biomarker that is elevated during the second trimester, before the clinical onset of preeclampsia. Although plasma decorin levels in subjects with preeclampsia that was measured during the second trimester were higher than control subjects, regardless of their BMI values or gestational ages, BMI-matched pairs revealed a more significant elevation in subjects with preeclampsia. These data suggest that the elevation of decorin may be used as a predictive biomarker for preeclampsia, when the data are controlled for BMI. Even with our limited data of plasma decorin values, the predictive performance of decorin on the development of preeclampsia, analyzed with logistic regression or ROC curve, was significant, irrespective of the BMI status or weight of subjects. However, this must be validated further in a prospective study with a larger subject population.

OBSTETRICS

Relationship of decorin to maternal obesity or bodyweight in preeclampsia: possible tissue source of plasma decorin Obesity is a major risk factor for preeclampsia, and decorin levels are reported to be moderately elevated in obese subjects.33 Thus, it is possible that obesity is an independent up-regulator of systemic decorin production, irrespective of preeclampsia. Our limited data in pregnant subjects could not confirm this possibility. In contrast, we found a negative correlation of plasma decorin level with BMI and bodyweight. This may support the hypothesis of the decidua being a primary source of plasma decorin in pregnant subjects, because the size of the placenta is independent of maternal BMI or bodyweight. A similar analysis for serum hCG levels in pregnancy was rationalized by Krantz et al.45 Our findings of a significant elevation of plasma decorin in subjects with preeclampsia in all BMI and bodyweight groups is consistent with our findings of overproduction by the decidua. With the present study being limited to the second trimester, further studies are needed to define the BMI and bodyweight-related changes in plasma decorin levels in control subjects and subjects with preeclampsia at different gestational ages, including first and third trimesters, and the temporal kinetics of normal pregnancy-related changes in plasma decorin levels in a larger control population. The cellular sources of secreted decorin entering circulation during pregnancy may be multiple and include decidual cells and other stromal cells (eg, in adipose tissues). Visceral fat was shown to be higher in decorin mRNA expression than subcutaneous fat, and a

Original Research

minor subpopulation of stromal cells adjacent to blood vessels revealed strong decorin immune-staining.33 Molecular regulation in decorin over-production by preeclampsia-associated decidual cells currently remains unknown. Because this over-production is highly selective for decidual cells, this regulation must be local rather than systemic. A metastasis suppressor cell surface molecule KAI1 (CD82) was found to up-regulate decorin expression during decidualization of ESCs in vitro, whereas interleukin-1b induced KAI1 expression.46 Thus, it is likely that decorin induction in the decidua is mediated by interleukin-1beKAI1 axis. Indeed, secretion of interleukin-1b was reported to be increased in preeclamptic placentas,47 which indicates that this inflammatory cytokine may play a role in decorin over-expression in preeclampsia decidua, which is a possibility that is being tested currently. Finally, a cause-effect relationship of decorin over-expression in the decidua and development of preeclampsia remains to be tested with a mouse model that was made to over-express decorin in the decidua alone.

Conclusion Our present study revealed that decorin is selectively over-expressed in maternal decidua, but not fetal mesenchymal cells, which reenforces the concept of the maternal origin of preeclampsia. Our in vitro studies that demonstrate the cause-effect relationship of decorin overexpression in decidua with compromised EVT invasiveness in preeclampsia require further validation with primary cells. Elevated plasma decorin levels that were observed during the second trimester, before clinical onset of

=

concentrations as a function of BMI show a negative correlation in both control and PE groups, the slopes not being different from each other. The negative correlation for the combined DCN values was significant (P ¼ 0.008). Values for PE remained significantly higher than the control (slope of elevation, P ¼ 0.0061). E, Plasma DCN concentrations as a function of maternal body weight show a negative correlation in both control and PE groups, the slopes not being different from each other. The negative correlation for the combined DCN values was significant (P ¼ 0.009). Values for PE remained significantly higher than the control (slope of elevation, P ¼ 0.0064). F, ROC curve analysis of plasma DCN as a predictor of PE. The cut-off value of DCN that best predicted PE status by optimizing sensitivity and specificity was >2.98 ng/ml (area under the curve ¼ 0.75, SE ¼ 0.07, P ¼ 0.002; sensitivity 89%, specificity 61%). This resulted in a positive likelihood ratio of 2.27 and a negative likelihood ratio of 0.18. BMI, body mass index; Cont, control; DCN, decorin; NW, normal weight; OB, obese; OW, overweight; PE, preeclampsia; ROC, receiver operator characteristic.

Siddiqui et al. Decorin, a potential biomarker in preeclampsia. Am J Obstet Gynecol 2016.

SEPTEMBER 2016 American Journal of Obstetrics & Gynecology

361.e13

Original Research

OBSTETRICS

preeclampsia, make decorin a novel potential biomarker. This view was supported by ROC-curve analysis of plasma decorin as a predictor of preeclampsia. However, its utility as a biomarker can be validated only by further prospective studies in a larger population. Furthermore, the sources of elevated decorin in the maternal blood require further investigation, because the placenta may not be the sole source in decorin secretion into maternal circulation. n Acknowledgment We are grateful to Dr Michael Miller (Research Associate, Children’s Health Research Institute) and Dr Wenquing He (Associate Professor, Department of Statistics and Actuarial sciences, University of Western Ontario) for assisting with the statistical analyses. We thank the mothers who donated their placentas and plasma samples.

References 1. Ilekis JV, Reddy UM, Roberts JM. Preeclampsia: a pressing problem: an executive summary of a National Institute of Child Health and Human Development workshop. Reprod Sci 2007;14:508-23. 2. Young BC, Levine RJ, Karumanchi SA. Pathogenesis of preeclampsia. Annu Rev Pathol 2010;5:173-92. 3. Lisonkova S, Joseph KS. Incidence of pre-eclampsia: risk factors and outcomes associated with early- versus late-onset disease. Am J Obstet Gynecol 2013;209:544.e1-12. 4. Lala PK, Chakraborty C. Factors regulating trophoblast migration and invasiveness: possible derangements contributing to preeclampsia and fetal injury. Placenta 2003;24: 575-87. 5. Roberts JM, Hubel CA. The two stage model of preeclampsia: variations on the theme. Placenta 2009;30:S32-7. 6. Redman CW, Sargent IL. Latest advances in understanding pre-eclampsia. Science 2005; 308:1592-4. 7. Roberts JM, Rajakumar A. Preeclampsia and soluble fms-like tyrosine kinase 1. J Clin Endocrinol Metab 2009;94:2252-4. 8. Maynard SE, Min JY, Merchan J, et al. Excess placental soluble fms-like tyrosine kinase 1 (sFlt1) may contribute to endothelial dysfunction, hypertension, and proteinuria in preeclampsia. J Clin Invest 2003;111:649-58. 9. Fan X, Rai A, Kambham N, et al. Endometrial VEGF induces placental sFLT1 and leads to pregnancy complications. J Clin Invest 2014;124:4941-52. 10. Cross JC, Werb Z, Fisher SJ. Implantation and the placenta: key pieces of the developmental puzzle. Science 1994;266: 1508-18.

11. Kaufmann P, Castelucci M. Extravillous trophoblast in the human placenta: a review. Troph Res 1997;10:21-65. 12. Lala PK, Hamilton GS. Growth factors, proteases and protease inhibitors in the maternal-fetal dialogue. Placenta 1996;17: 545-55. 13. Brosens JJ, Pijnenborg R, Brosens IA. The myometrial junctional zone spiral arteries in normal and abnormal pregnancies: a review of the literature. Am J Obstet Gynecol 2002;187: 1416-23. 14. Meekins JW, Pijnenborg R, Hanssens M, McFadyen IR, VanAsshe A. A study of placental bed spiral arteries and trophoblast invasion in normal and severe pre-eclamptic pregnancies. BJOG 1994;101:669-74. 15. Guzeloglu-Kaayisli O, GuzelogluKaayisli UA, Taylor HS. The role of growth factors and cytokines during implantation: endocrine and paracrine interactions. Semin Reprod Med 2009;27:62-79. 16. Graham CH, Lysiak JJ, McCrae KR, Lala PK. Localization of transforming growth factor-b at the human fetal-maternal interface: role in trophoblast growth and differentiation. Biol Reprod 1992;46:561-72. 17. Lysiak J, Hunt J, Pringle GA, Lala PK. Localization of transforming growth factor b and its natural inhibitor decorin in the human placenta and decidua throughout gestation. Placenta 1995;16:221-31. 18. Xu G, Guimond MJ, Chakraborty C, Lala PK. Control of proliferation, migration, and invasiveness of human extravillous trophoblast by decorin, a decidual product. Biol Reprod 2002;67:681-9. 19. Iacob D, Cai J, Tsonis M, et al. Decorinmediated inhibition of proliferation and migration of the human trophoblast via different tyrosine kinase receptors. Endocrinology 2008;149: 6187-97. 20. Khan GA, Girish GV, Lala N, DiGugliemo J, Lala PK. Decorin is a novel VEGFR-2 binding antagonist for human extravillous trophoblast. Mol Endocrinol 2011;25:1431-43. 21. Lala N, Girish GV, Bosworth CA, Lala PK. Mechanisms in decorin regulation of vascular endothelial growth factor-induced human trophoblast migration and acquisition of endothelial phenotype. Biol Reprod 2012;87: 1-14. 22. Lala PK, Nandi P. Mechanisms of trophoblast migration, endometrial angiogenesis in preeclampsia: the role of decorin. Cell Adh Migr 2016:1-15. 23. Magee LA, Pels A, Helewa M, et al. Diagnosis, evaluation, and management of the hypertensive disorders of pregnancy: executive summary. J Obstet Gynaecol Can 2014;36: 575-6. 24. Magee LA, Pels A, Helewa M, et al. The hypertensive disorders of pregnancy. Best Pract Res Clin Obstet Gynaecol 2015;29:643-57. 25. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time

361.e14 American Journal of Obstetrics & Gynecology SEPTEMBER 2016

ajog.org quantitative PCR and the 2(-Delta Delta C(T)) method. Methods 2001;25:402-8. 26. Han VKM, Bassett M, Walton J, Challis JRG. The expression of insulin-like growth factor (IGF) and IGF-binding protein (IGFBP) genes in the human placenta and membranes: evidence for IGF-IGFBP interactions at the feto-maternal interface. J Clin Endocrinol Metab 1996;81: 2680-93. 27. Graham CH, Hawley TS, Hawley RG, et al. Establishment and characterization of first trimester human trophoblast cells with extended life span. Exp Cell Res 1993;206: 204-11. 28. Irving JA, Lysiak JJ, Graham CH, Han VKM, Hearn S, Lala PK. Characteristics of trophoblast cells migrating from first trimester chorionic villus explants and propagated in culture. Placenta 1995;16:413-33. 29. Majumder M, Tutunea-Fatan E, Xin X, et al. Co-expression of a9b1 integrin and VEGF-D confers lymphatic metastatic ability to a human breast cancer cell line MDA-MB- 468LN. PLoS One 2012;7:e35094. 30. Spencer K, Ong CY, Liao AW, Nicolaides KH. The influence of ethnic origin on first trimester biochemical markers of chromosomal abnormalities. Prenat Diagn 2000;20: 491-4. 31. Aban C, Leguizamón GF, Cella M, Damiano A, Franchi AM, Farina MG. Differential expression of endocannabinoid system in normal and preeclamptic placentas: effects on nitric oxide synthesis. Placenta 2012;34: 67-74. 32. Bolton K, Segal D, McMillan J, et al. Decorin is a secreted protein associated with obesity and type-2 diabetes. Int J Obesity 2008;32: 1113-21. 33. DeGroot C, O’Brien T, Taylor RN. Biochemical evidence of impaired trophoblastic invasion of decidual stroma in women destined to have preeclampsia. Am J Obstet Gynecol 1996;175:24-9. 34. Gleeson LM, Chakraborty C, McKinnon T, Lala PK. Insulin-like growth factor-binding protein 1 stimulates human trophoblast migration by signaling through a5b1 integrin via mitogen-activated protein kinase pathway. J Clin Endocrinol Metab 2001;86:3665-74. 35. Anim-Nyame N, Hills FA, Soorana SR, Steer PJ, Johnson MR. A longitudinal study of maternal plasma insulin-like growth factor binding protein-1 concentrations during normal pregnancy and pregnancies complicated by pre-eclampsia. Hum Reprod 2000;15:2215-9. 36. Graham CH, Lala PK. Mechanisms of placental invasion of the uterus and their control. Biochem Cell Biol 1992;70:867-74. 37. Shaarawy M, El Meleigy M, Rasheed K. Maternal serum transforming growth factor beta-2 in preeclampsia and eclampsia, a potential biomarker for the assessment of disease severity and fetal outcome. J Soc Gynecol Investig 2001;8:27-31.

ajog.org 38. Levine RJ, Maynard SE, Qian C, et al. Circulating angiogenic factors and the risk of preeclampsia. N Engl J Med 2004;350:672-83. 39. Levine RJ, Lam C, Qian C, et al. Soluble endoglin and other circulating antiangiogenic factors in preeclampsia. N Engl J Med 2006; 355:992-1005. 40. Venkatesha S, Toporsian M, Lam C, et al. Soluble endoglin contributes to the pathogenesis of preeclampsia. Nat Med 2006;12:642-9. 41. Bergmann A, Ahmad S, Cudmore M, et al. Reduction of circulating soluble Flt-1 alleviates preeclampsia-like symptoms in a mouse model. J Cell Mol Med 2010;14:1857-67. 42. Caniggia E, Taylor CV, Knox Riche JW, Lye SJ, Letarte M. Endoglin regulates trophoblast differentiation along the invasive pathway in human placental villous explants. Endocrinology 1997;138:4977-88. 43. Thadhani R, Kisner T, Hagmann H, et al. Pilot study of extracorporeal removal of soluble fms-

OBSTETRICS

like tyrosine kinase 1 in preeclampsia. Circulation 2011;124:940-50. 44. Saito T, Takeda N, Amiya E, et al. VEGF-A induces its negative regulator, soluble form of VEGFR-1, by modulating its alternative splicing. FEBS Lett 2013;587:2179-85. 45. Krantz DA, Hallahan TW, Macri VJ, Macri JN. Maternal weight and ethnic adjustment within a first-trimester Down syndrome and trisomy 18 screening program. Prenat Diagn 2005;25: 635-40. 46. Gellerson B, Briese J, Oberndorfer M, et al. Expression of the metastasis suppressor KAI1 in decidual cells at the human maternal-fetal interface. Am J Pathol 2007; 170:126-39. 47. Amash A, Holcberg G, Sapir O, Huleihel M. Placental secretion of interleukin-1 and interleukin-1 receptor antagonist in preeclampsia: effect of magnesium sulfate. J Interferon Cytokine Res 2012;32:432-41.

Original Research

Author and article information From Departments of Anatomy and Cell Biology (Drs, Nandi, Girish, and Lala and Mr Siddiqui), Oncology (Dr Lala), Biology (Ms Nygard), Pediatrics (Dr Han), Obstetrics and Gynecology (Drs Eastabrook, de Vrijer, and Han), and Children’s Health Research Institute (Drs Eastabrook, de Vrijer, Han and Lala), Schulich School of Medicine and Dentistry, University of Western Ontario, London, Canada. 1 These authors contributed equally to this article. Received Dec. 23, 2015; revised March 10, 2016; accepted March 11, 2016. Supported by operating grants from the Canadian Institutes of Health Research (P.K.L. and V.K.M. H.), a Translational Research Grant of the Children’s Health Research Institute (P.K.L. and G.E.), and a Lawson Health Research Institute Internal Research Fund grant (B.deV.). The authors report no conflict of interest. Corresponding author: Peeyush K. Lala, MD, PhD. [email protected]

SEPTEMBER 2016 American Journal of Obstetrics & Gynecology

361.e15

A4 Septiembre.pdf

A receiver operating characteristic curve. analysis revealed strong diagnostic performance of plasma decorin in the. prediction of preeclampsia status. Although ...

3MB Sizes 5 Downloads 349 Views

Recommend Documents

A4.GTAV.Xero.Market.Paper.Model.by.Papermau.2014.pdf
A4.GTAV.Xero.Market.Paper.Model.by.Papermau.2014.pdf. A4.GTAV.Xero.Market.Paper.Model.by.Papermau.2014.pdf. Open. Extract. Open with. Sign In.

a4.pdf
∆ιάλειμμα. Whoops! There was a problem loading this page. Retrying... Whoops! There was a problem loading this page. Retrying... a4.pdf. a4.pdf. Open. Extract.

A4 brochure Mphil
www.iiitmk.ac.in. Ecological Informatics (EI) is a trans-disciplinary subject that integrates ecology, information technology, computational and social sciences in.

A4 Octubre.pdf
METHODS: Data were obtained from the U.S. Zika Preg- nancy Registry, an enhanced surveillance system of pregnant. women with laboratory evidence of ...

¢2£%a4 I
Sep 24, 1979 - anamid HDS-9 Trilobe). The hydrotreating was con ducted at a reactor temperature of 650° F., hydrogen pressure of 900 psig, a liquid hourly ...

A4.Desktop.Architecture.Four.Simple.Little.Houses.by.Papermau.2015 ...
Page 1 of 4. Page 1 of 4. Page 2 of 4. Page 2 of 4. Page 3 of 4. Page 3 of 4. A4.Desktop.Architecture.Four.Simple.Little.Houses.by.Papermau.2015.pdf. A4.

A4-Eschenfelder.pdf
There was a problem previewing this document. Retrying... Download. Connect more apps... Try one of the apps below to open or edit this item.

A4 brochure Mphil
All students admitted to the program, with minimum 65% marks or 7.0 CGPA in M.Sc. will be awarded SPEED-IT. First class M.Sc. in Natural/ Physical Sciences ...

Exhibitor Directory - A4 - Austrade
Jul 28, 2014 - Fax: +61 3 6231 0604 ... 3. The rise of Chinese civilisation over the past millennia is said to be ..... Room 1215, A building, Galaxy Central.

A4-Eschenfelder.pdf
There was a problem loading more pages. Retrying... Whoops! There was a problem previewing this document. Retrying... Download. Connect more apps.

a4.pdf
Whoops! There was a problem previewing this document. Retrying... Download. Connect more apps... Try one of the apps below to open or edit this item. a4.pdf.

A4-Eschenfelder.pdf
Page 1 of 17. 1/17. Social Science Data Archives: A Historical Social Network Analysis1. Kristin R. Eschenfelder, Morgaine Gilchrist Scott, Kalpana Shankar, ...

A4.R2D2.Paper.Model.by.Disney.Family.Custom.by.Papermau.pdf
Page 3 of 3. Page 3 of 3. A4.R2D2.Paper.Model.by.Disney.Family.Custom.by.Papermau.pdf. A4.R2D2.Paper.Model.by.Disney.Family.Custom.by.Papermau.pdf. Open. Extract. Open with. Sign In. Main menu. Displaying A4.R2D2.Paper.Model.by.Disney.Family.Custom.b

Maundy Thursday Reflection A4
“One of you will betray me.” 4 ... He will be betrayed by a friend, one who sat with him as he dined that night had already put plans in .... love will triumph over all.

A4 Septiembre.pdf
There was a problem previewing this document. Retrying... Download. Connect more apps... Try one of the apps below to open or edit this item. A4 Septiembre.

A4.The.Dirty.Dog.Saloon.Vignette.Paper.Model.by.Papermau.2015.pdf
A4.The.Dirty.Dog.Saloon.Vignette.Paper.Model.by.Papermau.2015.pdf. A4.The.Dirty.Dog.Saloon.Vignette.Paper.Model.by.Papermau.2015.pdf. Open. Extract.

A4 Junio.pdf
database that was managed by the data. coordinating center, which was. responsible for data analysis. Pharmacokinetic studies. Steady-state pravastatin PK ...

A4 Julio.pdf
There was a problem previewing this document. Retrying... Download. Connect more apps... Try one of the apps below to open or edit this item. A4 Julio.pdf.

A4 Agosto.pdf
data were provided by the laboratory for. the purpose of this study because this. information is not routinely included in. the laboratory reports. Statistical analysis.

Maundy Thursday Reflection A4
He had told them that on the night before their liberation they should eat a hasty meal. • With sandals on their feet, they were to eat bread made in haste without ...

dvds-A4.pdf
Sign in. Loading… Page 1. Whoops! There was a problem loading more pages. Retrying... dvds-A4.pdf. dvds-A4.pdf. Open. Extract. Open with. Sign In.

CORE-pdf-A4-P.cdr - GitHub
Page 1. EL - TEC. - *

2017_LIF_Brochure A4.pdf
Lennon © Thomas House/Jinx Lennon. 7. Song of Love © Angelica ... Burke: No 10. The market © Liverpool ..... 2017_LIF_Brochure A4.pdf. 2017_LIF_Brochure ...

A4.Ammu.Natin.Gun.Shop.Paper.Model.V2.by.Papermau.2015.pdf
Whoops! There was a problem loading more pages. Whoops! There was a problem previewing this document. Retrying... Download. Connect more apps... Try one of the apps below to open or edit this item. A4.Ammu.Natin.Gun.Shop.Paper.Model.V2.by.Papermau.20