GJ

Ep

KSa

b

c

d

a

ARR1AA

KPCGP

1

oanril

of

0h

ARTICLE IN PRESS ModelRI-2169; No. of Pages 6

Journal of Reproductive Immunology xxx (2013) xxx– xxx

Contents lists available at SciVerse ScienceDirect

Journal of Reproductive Immunology

j o ur nal homep age : w w w.elsev ier .com/ locate / j repr imm

levated complement factor C5a in maternal and umbilical cordlasma in preeclampsia

erina J. Dennya,b, Liam G. Coultharda,b, Richard H. Finnell c, Leonie K. Callawayb,d,tephen M. Taylora, Trent M. Woodruffa,∗

School of Biomedical Sciences, The University of Queensland, Brisbane, AustraliaSchool of Medicine, The University of Queensland, Brisbane, AustraliaDell Pediatric Research Unit, University of Texas, Austin, TX, United StatesRoyal Brisbane and Women’s Hospital, Brisbane, Australia

r t i c l e i n f o

rticle history:eceived 27 July 2012eceived in revised form5 November 2012ccepted 27 November 2012vailable online xxx

eywords:reeclampsiaomplementestational hypertension

a b s t r a c t

Preeclampsia is a leading cause of morbidity and mortality worldwide, encompassingsignificant short- and long-term health sequelae. Recently, there has been accumulatingevidence for a role of the complement system in the pathogenesis of numerous complica-tions of pregnancy, including preeclampsia. The present cross-sectional study comparedthe plasma concentrations of complement factors C3a and C5a between normotensivepregnancies and pregnancies complicated with either preeclampsia or gestational hyper-tension alone. We found that maternal plasma C5a concentration was significantly higherin preeclamptic pregnancy than in pregnancy affected by gestational hypertension aloneor normotensive pregnancy. Umbilical cord plasma C5a concentrations were also higherin pregnancies complicated by preeclampsia compared to gestational hypertension or nor-

regnancy motensive pregnancy. Maternal and cord plasma C5a concentrations were significantlycorrelated, suggesting that C5a can freely diffuse between maternal and fetal circulation.There were no significant differences in C3a concentrations in maternal or cord plasmabetween any groups. These results support the hypothesis that C5a may play a role inpreeclampsia, but not in gestational hypertension.

© 2013 Elsevier Ireland Ltd. All rights reserved.

. Introduction

Preeclampsia, defined clinically by the de novo devel-pment of hypertension and proteinuria in pregnancyfter 20 week’s gestation, complicates 2–8% of all preg-ancies. The clinical manifestations of preeclampsia

Please cite this article in press as: Denny, K.J., et al., Elevated

plasma in preeclampsia. J. Reprod. Immunol. (2013), http://dx.d

epresent widespread endothelial cell dysfunction andnclude: proteinuria, hypertension, hemolysis, elevatediver enzymes, and low platelet counts (HELLP syndrome);

∗ Corresponding author at: School of Biomedical Sciences, Universityf Queensland, St. Lucia, QLD 4072, Australia. Tel.: +61 7 3365 2924;ax: +61 7 3365 1766.

E-mail address: t.woodruff@uq.edu.au (T.M. Woodruff).

165-0378/$ – see front matter © 2013 Elsevier Ireland Ltd. All rights reserved.ttp://dx.doi.org/10.1016/j.jri.2012.11.006

and/or seizures (Powe et al., 2011). Preeclampsia remainsone of the leading causes of maternal and neonatal mor-bidity and mortality worldwide, with the only effectivetreatment being parturition or termination of the preg-nancy (Davey and MacGillivray, 1988; Steegers et al.,2010).

The immune system is central to hypotheses thatattempt to explain the etiopathogenesis of preeclamp-sia. An overzealous immune response may account forthe increased incidence of preeclampsia in women whohave had inadequate time to develop immune tolerance to

complement factor C5a in maternal and umbilical cordoi.org/10.1016/j.jri.2012.11.006

paternally-derived antigens, such as nulliparous women,women who have recently changed sexual partners, orwomen whose pregnancy resulted from donor insem-ination (Einarsson et al., 2003; Kyrou et al., 2010). In

ING Model

ductive

ARTICLEJRI-2169; No. of Pages 6

2 K.J. Denny et al. / Journal of Repro

contrast, prior exposure to paternal antigens in semi-nal fluid or by means of a prior pregnancy with thesame partner is protective against the development ofpreeclampsia (Dekker and Robillard, 2007). Furthermore,women who have a pre-existing autoimmune disease,such as systemic lupus erythematosus or antiphospho-lipid syndrome, are at a substantially increased risk ofdeveloping preeclampsia during their pregnancy relativeto that of the general population (Chakravarty et al.,2006).

There has been accumulating evidence for a role of thecomplement system – a vital component of the innateimmune system – in the pathogenesis of numerous compli-cations of pregnancy, including preeclampsia, intrauterinefetal death, and recurrent spontaneous abortion (Lee et al.,2007; Richani et al., 2005; Soto et al., 2010; Lynch et al.,2012). Under normal physiological conditions, activationof complement triggers an inflammatory cascade result-ing in immune cell activation and the rapid opsonizationand destruction of foreign pathogens (Manthey et al.,2009). However, excessive or inappropriate activationof complement leading to the generation of the potentimmuno-inflammatory anaphylatoxins C3a and C5a hasbeen demonstrated to contribute to the progression ofnumerous disease states (Woodruff et al., 2008; Pavlovskiet al., 2012).

Normal human pregnancy is characterized by a signifi-cant increase in C3a and C5a in the maternal circulation,an increase that does not fluctuate with gestationalage (Richani et al., 2005; Derzsy et al., 2010). Thisincrease may be indicative of a more generalized upreg-ulation of the innate immune system, to compensate forthe suppression of the adaptive immune system dur-ing gestation, an occurrence that presumably allows forthe tolerance of paternal alloantigens from the fetus(Chaouat, 2007). However, pregnancies affected by recur-rent miscarriage, intrauterine growth restriction, andunexplained intrauterine fetal death have higher mater-nal plasma concentrations of complement anaphylatoxinsthan in healthy pregnancies, suggesting that unregu-lated elevations in maternal complement factors mayadversely affect pregnancy outcomes (Lynch and Salmon,2010).

Studies have shown that women with preeclamp-sia have significantly higher levels of C5a than healthypregnant women (Soto et al., 2010; Haeger et al., 1992).However, the data on C3a is less convincing, with somestudies showing elevated circulating C3a in preeclamp-sia (Boij et al., 2012; Lynch et al., 2012) and othersshowing no significant difference compared to healthypregnant women (Lynch et al., 2010; Soto et al., 2010).In contrast to maternal blood, no study to date hasreported anaphylatoxin levels in the fetal circulation.The present study therefore aimed to determine mater-nal and umbilical cord circulating concentrations ofboth C3a and C5a in healthy pregnancies, and preg-nancies affected by preeclampsia at term. In addition,

Please cite this article in press as: Denny, K.J., et al., Elevated

plasma in preeclampsia. J. Reprod. Immunol. (2013), http://dx.d

this study also compared preeclamptic samples withsamples from pregnancies with gestational hyperten-sion without proteinuria to demonstrate that changes incirculating anaphylatoxins are specific to preeclampsia,

PRESSImmunology xxx (2013) xxx– xxx

rather than a result of general gestational pathol-ogy.

2. Materials and methods

2.1. Study design and definitions

A cross-sectional study was conducted at the RoyalBrisbane and Women’s Hospital in which a total of 135women were recruited between May 2009 and July 2010and divided into 3 groups: (1) normal pregnant women,(2) women with gestational hypertension in the absenceof proteinuria, and (3) women with preeclampsia. Normalpregnant women were initially recruited from a weekly-booked caesarean section list. Indications for caesareansection included: maternal preference, repeat elective cae-sarean section and breech presentation or transverse lie.Once sample collection processes were streamlined, aconsecutive series of normal pregnant women having nor-mal vaginal delivery were recruited consecutively for aseries of several weeks from the birth suite. Overlappingwith this collection, all consecutive women present-ing with either gestational hypertension or preeclampsiawere approached regarding the study. The study wasstaffed by a research midwife, who approached all eli-gible women present in the hospital during workinghours. All women provided written informed consentprior to sample collection. The collection of sampleswas approved by both the Royal Brisbane and Women’sHospital Human Research Ethics Committee and theUniversity of Queensland Medical Research Ethics Commit-tee.

Inclusion criteria for normal pregnancy included: theabsence of medical, obstetrical, or surgical complications,and delivery of an appropriate-for-gestational-age infantat term. Twin pregnancies were excluded. Gestationalhypertension was defined as the development of de novohypertension (systolic blood pressure of >140 mmHg ordiastolic blood pressure of >90 mmHg) during the ges-tational period. Preeclampsia was defined as gestationalhypertension, as defined above, with the addition of sig-nificant proteinuria (>300 mg in a 24-h urine collection)(Davey and MacGillivray, 1988).

2.2. Blood collection and human C3a and C5aimmunoassays

Samples of maternal peripheral blood were collectedjust prior to delivery, and fetal blood obtained from theumbilical cord vein immediately following delivery. Bloodwas drawn directly into EDTA glass Vacutainer tubes (Bec-ton Dickinson) and the complement convertase and serineprotease inhibitor FUT-175 (Sigma) added to a final con-centration of 0.1 mg/ml to prevent in vitro complementactivation (Pfeifer et al., 1999). Samples were then cen-

complement factor C5a in maternal and umbilical cordoi.org/10.1016/j.jri.2012.11.006

trifuged at 4 ◦C and plasma aliquoted and stored at −80 ◦C.C3a and C5a enzyme-linked immunoassays (ELISAs) wereperformed according to manufacturer’s instructions (USCNLife Science Inc., Wuhan).

ARTICLE IN PRESSG Model

JRI-2169; No. of Pages 6

K.J. Denny et al. / Journal of Reproductive Immunology xxx (2013) xxx– xxx 3

Table 1Demographic and clinical characteristics of healthy normotensive pregnancies (Normal), pregnancies affected by gestational hypertension (GH), andpregnancies complicated by preeclampsia (PET).

Normal GH PET P value

Maternal characteristicsN 72 15 43Age 32 ± 0.62a 28 ± 1.80 31 ± 1.04 0.0724Nulliparity (%) 27.8 70.6b 58.1b 0.0003Body mass index (kg/m2) 29 ± 0.74 36.8 ± 2.27b 34.6 ± 1.30b < 0.0001Caesarean section (%) 14.8 41.7 29.6 0.1912Neonate characteristicsN 63 15 41Birth weight (g) 2848 ± 114.7 2371 ± 277.1 2304 ± 186.1b 0.0229

36

2

tawiotbaptpSmsa

3

wnmptwPco(bseTnansnp

iT

parametric correlation analysis was performed, revealing apositive correlation between maternal and umbilical cordplasma C5a levels (Spearman’s r = 0.28; P < 0.05; Fig. 3).

Fig. 1. Mean plasma C3a and C5a concentration of normal pregnantwomen (normal; n = 32), women with gestational hypertension (GH;n = 6), and women with preeclampsia (PET; n = 35). (A) There wereno significant differences in maternal plasma C3a concentrations

Gestational age at delivery 36 ± 0.5

a Values represent mean ± SEM.b Significant difference compared with normal pregnancy.

.3. Statistical analysis

Differences in maternal and neonatal clinical charac-eristics between the groups were compared by means of

Kruskal–Wallis one-way analysis of variance (ANOVA)ith a post hoc Dunn–Bonnferroni correction. Nullipar-

ty and caesarean section rates were analyzed by meansf chi-square analysis with Fischer’s exact test. Bartlett’sest was used to test equality of variances. Correlationsetween different parameters were examined by bivari-te analysis using Spearman correlation coefficients forarameters with skewed distributions. Multivariable logis-ic regression analysis was also used to further investigateossible confounders using the statistical software packageTATA v11.0 (Statacorp, USA). All results are expressed asean ± standard error of the mean (SEM), with significance

et at P < 0.05. Results are unadjusted for confounding vari-bles when not explicitly stated otherwise.

. Results

Demographic and clinical characteristics of patientsith preeclampsia, gestational hypertension alone, andormotensive pregnancies are presented in Table 1. Bodyass index (BMI) at delivery was significantly higher in

atients with preeclampsia (34.6 kg/m2 ± 1.30) or gesta-ional hypertension (36.8 kg/m2 ± 2.27) than in patientsith uncomplicated pregnancies (29 kg/m2 ± 0.74;

< 0.0001). Further, nulliparity was significantly moreommon in women with gestational hypertension (70.6%)r preeclampsia (58.1%) than in normotensive patients27.8%; P = 0.0003). The mean birth weight of neonatesorn to patients with preeclampsia (2304 g ± 186) was alsoignificantly lower than neonates born to patients whoxperienced a normal pregnancy (2848 g ± 115; P < 0.05).here was a trend towards lower birth weights amongsteonates born to patients with gestational hypertensionlone, but this was not statistically significant. There waso statistically significant difference in rates of caesareanection between normotensive pregnancy (14.8%), preg-ancy affected by gestational hypertension (41.7%), and

Please cite this article in press as: Denny, K.J., et al., Elevated

plasma in preeclampsia. J. Reprod. Immunol. (2013), http://dx.d

reeclamptic pregnancy (29.6%; P = 0.1912).Levels of circulating C3a and C5a were measured

n both maternal plasma and umbilical cord plasma.here was no significant difference in maternal plasma

± 1.0 34 ± 0.8 0.0630

C3a concentrations between normotensive pregnancy(54.9 ng/ml ± 1.62), pregnancies affected by gestationalhypertension (57.1 ng/ml ± 1.66), or preeclamptic preg-nancies (63.8 ng/ml ± 4.37; P = 0.132; Fig. 1 A). By contrast,maternal plasma C5a concentrations were significantlyhigher in women with preeclampsia (8.2 ng/ml ± 1.30)than in normotensive pregnant women (4.5 ng/ml ± 0.50;P < 0.05; Fig. 1B). This difference remained when resultswere corrected for BMI using multilinear logistic regres-sion analysis (P < 0.001). There was no significant differencein the median plasma concentrations of C5a betweennormotensive pregnant women and women with gesta-tional hypertension (4.5 ng/ml ± 0.51). Cord plasma C5aconcentrations were significantly higher in pregnan-cies affected by preeclampsia (8.3 ng/ml ± 1.71), but notgestational hypertension (4.9 ng/ml ± 0.93), than in nor-motensive pregnancies (3.2 ng/ml ± 0.35; P < 0.01; Fig. 2B).Conversely, there were no significant differences inumbilical cord plasma C3a between any of the groups(normotensive: 61.2 ng/ml ± 8.24; gestational hyperten-sion: 34.0 ng/ml ± 7.07; preeclampsia: 79.9 ng/ml ± 10.70;Fig. 2A). To assess the relationship between plasma C5afrom maternal and umbilical C5a cord plasma, a non-

complement factor C5a in maternal and umbilical cordoi.org/10.1016/j.jri.2012.11.006

between women with GH (57.1 ng/ml ± 1.66), PET (63.8 ng/ml ± 4.37),and normotensive women (54.9 ng/ml ± 1.62; P = 0.132). (B) The meanmaternal plasma C5a concentration was higher in patients with PET(8.2 ng/ml ± 1.30) than normal pregnancies (4.5 ng/ml ± 0.50) or womenwith pregnancies affected by GH (4.5 ng/ml ± 0.51; P < 0.05).

ARTICLE ING Model

JRI-2169; No. of Pages 6

4 K.J. Denny et al. / Journal of Reproductive

Fig. 2. Mean cord plasma C3a and C5a concentration in normal preg-nancies (normal; n = 35), pregnancies with gestational hypertension (GH;n = 14), and pregnancies affected by preeclampsia (PET; n = 29). (A)There were no significant differences in cord plasma C3a concentrationsbetween women with GH (34.0 ng/ml ± 7.07), PET (79.9 ng/ml ± 10.70),and normotensive women (61.2 ng/ml ± 8.24). (B) The mean cord plasma

C5a concentration was higher in patients with PET (8.3 ng/ml ± 1.71) thannormal women (3.2 ng/ml ± 0.35) or women with GH (4.9 ng/ml ± 0.93;P < 0.01).

4. Discussion

The finding that maternal circulating C5a is elevated inpreeclampsia is in concordance with previous research insimilar clinical populations demonstrating elevated plasmaC5a throughout the gestational period (Soto et al., 2010).Evidence for how excessive circulating C5a may contributeto the clinical manifestations of preeclampsia comes fromrecent studies in murine models of pregnancy, in whichC5a has been demonstrated to be a mediator of angio-genic dysregulation, a key pathophysiological feature ofpreeclampsia (Qing et al., 2011). Specifically, a study byGirardi et al. (2006) demonstrated a causal relationshipbetween C5a and the release of the anti-angiogenic fac-tor, soluble vascular endothelial growth factor receptor-1(sVEGFR-1). It is theorized that the combination of inflam-matory cytokine release and disruption of angiogenesiscontributes to the placental insufficiency and mater-nal endothelial dysfunction observed in the preeclampticpatient (Lynch and Salmon, 2010).

The present study has additionally demonstrated, forthe first time, that maternal plasma C5a levels correlatestrongly with umbilical cord plasma C5a concentration.In pregnancies affected by preeclampsia, both the motherand fetus are exposed to significantly greater levels of this

Please cite this article in press as: Denny, K.J., et al., Elevated

plasma in preeclampsia. J. Reprod. Immunol. (2013), http://dx.d

pro-inflammatory anaphylatoxin than in uncomplicatedpregnancies. Although the precise effect of elevated circu-lating levels of C5a on the human fetus remains unknown,

Fig. 3. Bivariate correlation between maternal and cord plasma C5a con-centration. Data was pooled from all groups. There was a significantpositive correlation between maternal and cord plasma C5a concentration(Spearman’s r = 0.28; P < 0.05; n = 73).

PRESSImmunology xxx (2013) xxx– xxx

C5a has been shown to be a key effector of fetal injury innumerous antibody-dependent and antibody-independentanimal models of intrauterine growth restriction, fetaldeath, and spontaneous miscarriage (Girardi et al., 2003,2006; Denny et al., 2012). In addition, C5a has been shownto upregulate pro-labor mediators in the human placenta(Lappas et al., 2012). Taken collectively, these data sug-gest that high levels of circulating C5a may be injurious tomother and fetus, although the mechanism(s) are currentlyill-defined.

The levels of plasma C5a in both the maternal andfetal circulation observed in the present study approx-imated to circulating levels of between 0.1 and 4 nM.Studies using human recombinant C5a have demonstratedthat the binding affinity (IC50) of C5a to inflammatorycells such as neutrophils is ∼0.1–0.3 nM (Woodruff et al.,2001), suggesting that the plasma levels of C5a in preg-nancy may be biologically active. Notably, C5a is elevatedeven in healthy pregnancies compared to the non-pregnantstate (Richani et al., 2005); however, its potentially inju-rious effects are presumably prevented by the regulatoryproteins decay-accelerating factor (DAF), CD46 and CD59(Tedesco et al., 1993). These regulatory proteins, local-ized to villous trophoblasts on the placenta, are capable ofpreventing inappropriate complement activation and theconsequent generation of C5a (Lynch and Salmon, 2010).In preeclampsia, however, we propose that these regula-tors are overwhelmed, leading to excessive complementactivation with the generation of excess C5a and a resul-tant pro-inflammatory maternal-foetal environment. Thisis supported by recent reports of severe preeclampsia inwomen with genetic defects in complement regulatoryproteins (Fang et al., 2008; Salmon et al., 2011).

One interesting finding from our present study wasthe lack of a significant change in plasma C3a betweennormotensive and preeclamptic pregnancies. This is in con-trast to other recent studies demonstrating elevated C3ain preeclampsia (Derzsy et al., 2010; Lynch et al., 2012,2011; Boij et al., 2012). The concentrations of C3a deter-mined in our study were in the range of 50–70 ng/ml,in line with published levels for non-pregnant subjects(Derzsy et al., 2010). This compares with mean plasmaC3a levels of between 800 ng/ml to 75,000 ng/ml in otherstudies in pregnancy, which equates to a 10–1000-foldincrease over our current findings (Derzsy et al., 2010;Soto et al., 2010; Lynch et al., 2011, 2012; Lappas, 2011;Boij et al., 2012). One difference between these studiesand the present study is the use of FUT-175, in additionto EDTA, in blood collected in our study. FUT-175 is abroad-spectrum serine protease inhibitor, which has beenpreviously demonstrated to prevent exogenous comple-ment activation and C3a generation in plasma subsequentto blood collection (Pfeifer et al., 1999). The presence ofFUT-175 in our collected samples may therefore have pre-vented ex vivo activation of complement, thus preventingany inappropriate and misleading increases in C3a in thepreeclamptic samples. Exogenous activation is particularly

complement factor C5a in maternal and umbilical cordoi.org/10.1016/j.jri.2012.11.006

evident in blood with higher levels of extrinsic complementproteases (e.g. thrombin or tissue factor) with the abilityto cleave C3 and C5 ex vivo (Amara et al., 2008). Throm-bin is elevated during pregnancy (Brenner, 2004), and to a

ING Model

J

ductive

gmcimcos

gbthpaCchneop

oiatdptiaasoppemdi

toOiiopowacoct

lpai

ARTICLERI-2169; No. of Pages 6

K.J. Denny et al. / Journal of Repro

reater extent in preeclampsia (de Boer et al., 1989), whichay account for the discrepancies between the levels of

omplement split products between our and other stud-es. We therefore recommend that future studies aiming to

easure the precise in vivo circulating concentrations ofomplement factors such as C3a and C5a consider the usef FUT-175 to prevent erroneous elevations in complementplit products in collected samples.

The present study further aimed to determine whetherestational hypertension reflected an intermediate pointetween normal pregnancy and preeclampsia with regardso plasma C3a and C5a levels. The notion of gestationalypertension being a precursor to or milder form ofreeclampsia, however, was not reflected by any observ-ble trend in circulating maternal or umbilical plasma5a levels in the present study. This finding is in con-ordance with previous studies reporting that gestationalypertension alone, in contrast to preeclampsia, was foundot to be associated with markers of inflammation (Wolft al., 2002). However, given the relatively small numberf women with gestational hypertension alone used in thisresent study, the significance of these findings is unclear.

Illuminating the distinct pathophysiological signaturef preeclampsia is of utmost importance in understand-ng the natural history of the disease, which consequentlyllows for appropriate and targeted management of hyper-ension in pregnancy. An understanding of the immuneysregulation that underlies the clinical manifestations ofreeclampsia is also key to developing novel therapeuticso delay, or ideally, prevent the need for immediate deliveryn women affected by preeclampsia prior to term. Further,lthough plasma C5a alone is unlikely to serve as a reli-ble biological marker to predict pregnancy outcomes, weuggest that it could be used in the future, together withther markers of inflammation, to differentiate betweenreeclampsia and more benign hypertensive disorders ofregnancy. However, although studies in mice indicate thatlevated C5a may be a precipitating factor in the develop-ent of preeclampsia (Girardi et al., 2006), it remains to be

etermined whether C5a concentrations are elevated earlyn preeclamptic pregnancies.

These results also highlight the well-described rela-ionship between obesity and the hypertensive disordersf pregnancy (Callaway et al., 2009; Lynch et al., 2012).besity is associated with a wide range of alterations in

mmune function, and it is possible that the alterationsn complement factors seen here reflect the underlyingbesity of women with preeclampsia, rather than thereeclampsia itself. However, this work allows for a degreef differentiation. The finding that the almost equally obeseomen with gestational hypertension did not have alter-

tions in factors studied provides good evidence that thehanges we have noted here are related to the syndromef preeclampsia. Indeed, recent evidence demonstrates aombined impact of elevated complement and obesity onhe development of preeclampsia (Lynch et al., 2012).

Once thought to be a self-limiting entity, recent

Please cite this article in press as: Denny, K.J., et al., Elevated

plasma in preeclampsia. J. Reprod. Immunol. (2013), http://dx.d

arge-scale epidemiologic studies have revealed thatreeclampsia now foreshadows significant long-term dam-ge to the cardiovascular system for both mother andnfant (Romundstad et al., 2010). Relative to women

PRESSImmunology xxx (2013) xxx– xxx 5

with uncomplicated pregnancies, women with a his-tory of preeclampsia have an increased risk of an acutecoronary syndrome, cerebrovascular disease, peripheralarterial disease, and cardiovascular mortality (McDonaldet al., 2008). Furthermore, the adolescent offspring from apreeclamptic pregnancy are heavier and have an increasedblood pressure than their counterparts – significant riskfactors for future cardiovascular morbidity and mortal-ity (Vatten et al., 2003). Thus there is a pressing needto continue research into the mechanisms underlyingpreeclampsia in the hope of developing new therapiesto reduce the physical and economic burden of bothpreeclampsia and its sequelae. Our results lend weight tothe notion that C5a may be one such potential therapeutictarget (Woodruff et al., 2011) worthy of further investiga-tion.

5. Conclusions

Our results confirm previous reports of elevated mater-nal plasma levels of C5a in preeclampsia and, in addition,show that these elevated levels are reflected in the fetalcirculation. In contrast, we failed to detect any significantalterations in plasma C3a between normal and preeclamp-tic pregnancies. These findings provoke questions as toboth the origin of C5a in gestation and the significanceof selective elevation of C5a in the etiopathogenesis ofpreeclampsia.

Conflicts of interest

The authors declare no conflicts of interest

Declaration of funding source

This work was supported by the National Healthand Medical Research Council of Australia (Project Grant569693 to SMT, TMW, LKC and RHF), and the Aus-tralian Research Council (Future Fellowship FT110100332to TMW).

References

Amara, U., Rittirsch, D., Flierl, M., Bruckner, U., Klos, A., Gebhard, F., Lam-bris, J.D., Huber-Lang, M., 2008. Interaction between the coagulationand complement system. Adv. Exp. Med. Biol. 632, 71–79.

Boij, R., Svensson, J., Nilsson-Ekdahl, K., Sandholm, K., Lindahl, T.L.,Palonek, E., Garle, M., Berg, G., Ernerudh, J., Jenmalm, M., Matthiesen,L., 2012. Biomarkers of coagulation, inflammation, and angiogene-sis are independently associated with preeclampsia. Am. J. Reprod.Immunol.

Brenner, B., 2004. Haemostatic changes in pregnancy. Thromb. Res. 114,409–414.

Callaway, L.K., O’Callaghan, M., Mcintyre, H.D., 2009. Obesity andthe hypertensive disorders of pregnancy. Hypertens Pregnancy 28,473–493.

Chakravarty, E.F., Nelson, L., Krishnan, E., 2006. Obstetric hospitalizationsin the United States for women with systemic lupus erythematosusand rheumatoid arthritis. Arthritis Rheum. 54, 899–907.

complement factor C5a in maternal and umbilical cordoi.org/10.1016/j.jri.2012.11.006

Chaouat, G., 2007. The Th1/Th2 paradigm: still important in pregnancy?Semin Immunopathol. 29, 95–113.

Davey, D.A., MacGillivray, I., 1988. The classification and definition ofthe hypertensive disorders of pregnancy. Am. J. Obstet. Gynecol. 158,892–898.

ING Model

ductive

ARTICLEJRI-2169; No. of Pages 6

6 K.J. Denny et al. / Journal of Repro

de Boer, K., Ten Cate, J.W., Sturk, A., Borm, J.J., Treffers, P.E., 1989. Enhancedthrombin generation in normal and hypertensive pregnancy. Am. J.Obstet. Gynecol. 160, 95–100.

Dekker, G., Robillard, P.Y., 2007. Pre-eclampsia: Is the immune maladap-tation hypothesis still standing? An epidemiological update. J. Reprod.Immunol. 76, 8–16.

Denny, K.J., Woodruff, T.M., Taylor, S.M., Callaway, L.K., 2012. Complementin Pregnancy: a delicate balance. Am. J. Reprod. Immunol.

Derzsy, Z., Prohaszka, Z., Rigo Jr., J., Fust, G., Molvarec, A., 2010. Activationof the complement system in normal pregnancy and preeclampsia.Mol. Immunol. 47, 1500–1506.

Einarsson, J.I., Sangi-Haghpeykar, H., Gardner, M.O., 2003. Sperm expo-sure and development of preeclampsia. Am. J. Obstet. Gynecol. 188,1241–1243.

Fang, C.J., Fremeaux-Bacchi, V., Liszewski, M.K., Pianetti, G., Noris, M.,Goodship, T.H., Atkinson, J.P., 2008. Membrane cofactor protein muta-tions in atypical hemolytic uremic syndrome (aHUS), fatal Stx-HUS, C3glomerulonephritis, and the HELLP syndrome. Blood 111, 624–632.

Girardi, G., Berman, J., Redecha, P., Spruce, L., Thurman, J.M., Kraus, D.,Hollmann, T.J., Casali, P., Caroll, M.C., Wetsel, R.A., Lambris, J.D., Holers,V.M., Salmon, J.E., 2003. Complement C5a receptors and neutrophilsmediate fetal injury in the antiphospholipid syndrome. J. Clin. Invest.112, 1644–1654.

Girardi, G., Yarilin, D., Thurman, J.M., Holers, V.M., SalmoN, J.E., 2006.Complement activation induces dysregulation of angiogenic factorsand causes fetal rejection and growth restriction. J. Exp. Med. 203,2165–2175.

Haeger, M., Unander, M., Norder-Hansson, B., Tylman, M., Bengtsson, A.,1992. Complement, neutrophil, and macrophage activation in womenwith severe preeclampsia and the syndrome of hemolysis, elevatedliver enzymes, and low platelet count. Obstet. Gynecol. 79, 19–26.

Kyrou, D., Kolibianakis, E.M., Devroey, P., Fatemi, H.M., 2010. Is the useof donor sperm associated with a higher incidence of preeclampsiain women who achieve pregnancy after intrauterine insemination?Fertil. Steril. 93, 1124–1127.

Lappas, M., 2011. Lower circulating levels of complement split proteinsC3a and C4a in maternal plasma of women with gestational diabetesmellitus. Diabet. Med. 28, 906–911.

Lappas, M., Woodruff, T., Taylor, S., Permezel, M., 2012. Complement C5aregulates pro-labor mediators in human placenta. Biol. Reprod.

Lee, J., Oh, J., Choi, E., Park, I., Han, C., Kim Do, H., Choi, B.C., Kim, J.W., Cho, C.,2007. Differentially expressed genes implicated in unexplained recur-rent spontaneous abortion. Int. J. Biochem. Cell Biol. 39, 2265–2277.

Lynch, A.M., Murphy, J.R., Gibbs, R.S., Levine, R.J., Giclas, P.C., Salmon, J.E.,Holers, V.M., 2010. The interrelationship of complement-activationfragments and angiogenesis-related factors in early pregnancy andtheir association with pre-eclampsia. BJOG 117, 456–463.

Lynch, A.M., Eckel, R.H., Murphy, J.R., Gibbs, R.S., West, N.A., Giclas,P.C., Salmon, J.E., Holers, V.M., 2012. Prepregnancy obesity and com-plement system activation in early pregnancy and the subsequentdevelopment of preeclampsia. Am. J. Obstet. Gynecol. 206, 428 e1-8.

Lynch, A.M., Gibbs, R.S., Murphy, J.R., Giclas, P.C., Salmon, J.E., Holers, V.M.,2011. Early elevations of the complement activation fragment C3a andadverse pregnancy outcomes. Obstet. Gynecol. 117, 75–83.

Please cite this article in press as: Denny, K.J., et al., Elevated

plasma in preeclampsia. J. Reprod. Immunol. (2013), http://dx.d

Lynch, A.M., Salmon, J.E., 2010. Dysregulated complement activation asa common pathway of injury in preeclampsia and other pregnancycomplications. Placenta 31, 561–567.

Manthey, H.D., Woodruff, T.M., Taylor, S.M., Monk, P.N., 2009. Comple-ment component 5a (C5a). Int. J. Biochem. Cell Biol. 41, 2114–2117.

PRESSImmunology xxx (2013) xxx– xxx

McDonald, S.D., Malinowski, A., Zhou, Q., Yusuf, S., Devereaux, P.J., 2008.Cardiovascular sequelae of preeclampsia/eclampsia: a systematicreview and meta-analyses. Am. Heart J. 156, 918–930.

Pavlovski, D., Thundyil, J., Monk, P.N., Wetsel, R.A., Taylor, S.M., Woodruff,T.M., 2012. Generation of complement component C5a by ischemicneurons promotes neuronal apoptosis. FASEB J. 26, 3680–3690.

Pfeifer, P.H., Kawahara, M.S., Hugli, T.E., 1999. Possible mechanismfor in vitro complement activation in blood and plasma samples:futhan/EDTA controls in vitro complement activation. Clin. Chem. 45,1190–1199.

Powe, C.E., Levine, R.J., Karumanchi, S.A., 2011. Preeclampsia, a dis-ease of the maternal endothelium: the role of antiangiogenic factorsand implications for later cardiovascular disease. Circulation 123,2856–2869.

Qing, X., Redecha, P.B., Burmeister, M.A., TomlinsoN, S., D’Agati, V.D.,Davisson, R.L., Salmon, J.E., 2011. Targeted inhibition of complementactivation prevents features of preeclampsia in mice. Kidney Int. 79,331–339.

Richani, K., Soto, E., Romero, R., Espinoza, J., Chaiworapongsa, T., Nien, J.K.,Edwin, S., Kim, Y.M., Hong, J.S., Mazor, M., 2005. Normal pregnancyis characterized by systemic activation of the complement system. J.Matern. Fetal Neonatal Med. 17, 239–245.

Romundstad, P.R., Magnussen, E.B., Smith, G.D., Vatten, L.J., 2010.Hypertension in pregnancy and later cardiovascular risk: commonantecedents? Circulation 122, 579–584.

Salmon, J.E., Heuser, C., Triebwasser, M., Liszewski, M.K., Kavanagh,D., Roumenina, L., Branch, D.W., Goodship, T., Fremeaux-Bacchi, V.,Atkinson, J.P., 2011. Mutations in complement regulatory proteinspredispose to preeclampsia: a genetic analysis of the PROMISSEcohort. PLoS Med. 8, e1001013.

Soto, E., Romero, R., Richani, K., Espinoza, J., Chaiworapongsa, T., Nien,J.K., Edwin, S.S., Kim, Y.M., Hong, J.S., Goncalves, L.F., Yeo, L., Mazor,M., Hassan, S.S., Kusanovic, J.P., 2010. Preeclampsia and pregnan-cies with small-for-gestational age neonates have different profilesof complement split products. J. Matern. Fetal Neonatal Med. 23,646–657.

Steegers, E.A., Von Dadelszen, P., Duvekot, J.J., Pijnenborg, R., 2010. Pre-eclampsia. Lancet 376, 631–644.

Tedesco, F., Narchi, G., Radillo, O., Meri, S., Ferrone, S., Betterle, C., 1993.Susceptibility of human trophoblast to killing by human complementand the role of the complement regulatory proteins. J. Immunol. 151,1562–1570.

Vatten, L.J., Romundstad, P.R., Holmen, T.L., Hsieh, C.C., Trichopoulos, D.,Stuver, S.O., 2003. Intrauterine exposure to preeclampsia and ado-lescent blood pressure, body size, and age at menarche in femaleoffspring. Obstet. Gynecol. 101, 529–533.

Wolf, M., Sandler, L., Jimenez-Kimble, R., Shah, A., Ecker, J.L., Thadhani,R., 2002. Insulin resistance but not inflammation is associated withgestational hypertension. Hypertension 40, 886–891.

Woodruff, T.M., Costantini, K.J., Crane, J.W., Atkin, J.D., Monk, P.N., Taylor,S.M., Noakes, P.G., 2008. The complement factor C5a contributes topathology in a rat model of amyotrophic lateral sclerosis. J. Immunol.181, 8727–8734.

Woodruff, T.M., Nandakumar, K.S., Tedesco, F., 2011. Inhibiting the C5–C5a

complement factor C5a in maternal and umbilical cordoi.org/10.1016/j.jri.2012.11.006

receptor axis. Mol. Immunol. 48, 1631–1642.Woodruff, T.M., Strachan, A.J., Sanderson, S.D., Monk, P.N., Wong, A.K., Fair-

lie, D.P., Taylor, S.M., 2001. Species dependence for binding of smallmolecule agonist and antagonists to the C5a receptor on polymor-phonuclear leukocytes. Inflammation 25, 171–177.