| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Nuffield Department of Obstetrics and Gynaecology, University of Oxford, John Radcliffe Hospital, Oxford, United Kingdom; and
Department of Maternal Fetal Medicine, Imperial College School of Medicine, Chelsea and Westminster Hospital, London, United Kingdom
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
throughout normal pregnancy, more IL-12p70 in the first and second trimesters, and more IL-18 in the first trimester only. Intracellular cytokine measurements (TNF- and IL12p70) showed little change by comparison. IFN-
production was suppressed in all three trimesters. In pre-eclampsia, IL-18 secretion was increased. Secreted but not intracellular measures of TNF- and IL-12p70 were also further enhanced compared with normal pregnancy. Inhibition of IFN- production was lost and involved both CD56+ NK and CD56– lymphocyte subsets. We determined whether circulating syncytiotrophoblast microparticles (STBM) could contribute to these inflammatory changes. Unbound STBM could be detected in normal pregnancy by the second trimester and increased significantly in the third. They were also bound in vivo to circulating monocytes. Women with pre-eclampsia had significantly more circulating free but not cell-bound STBMs. STBMs prepared by perfusion of normal placental lobules stimulated production of inflammatory cytokines (TNF-, IL12p70, and IL-18 but not IFN-) when cultured with PBMCs from healthy nonpregnant women. Inflammatory priming of PBMCs during pregnancy is confirmed and is established by the first trimester. It is associated with early inhibition of IFN- production. The inflammatory response is enhanced in pre-eclampsia with loss of the IFN- suppression. Circulating STBMs bind to monocytes and stimulate the production of inflammatory cytokines. It is concluded that they are potential contributors to altered systemic inflammatory responsiveness in pregnancy and pre-eclampsia. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Our work has suggested that in later pregnancy the dominant immune interaction at interface II involves the innate rather than the adaptive systems and is characterized by a mild systemic inflammatory state (6, 7). This is evident in activation of circulating leukocytes (6), the associated acute phase response (reviewed by Redman and Sargent in Ref. 8) and the parallel endothelial activation (9). We have shown that pre-eclampsia is associated with more exaggerated systemic inflammatory changes (6) and argued that pre-eclampsia is what happens when the response is so intense as to provoke the features of the disorder (7).
Pre-eclampsia is a specific disorder of the second half of pregnancy, comprising variable combinations of signs of which hypertension and proteinuria are emphasized for clinical diagnosis. The possible range of features is however much wider including dysfunction of clotting and of the liver. The condition can progress to crises which can be fatal. These include convulsions and the acute hemolysis, elevated liver enzymes, and low platelets (HELLP) syndrome,3 which is characterized by disseminated intravascular coagulation, acute hemolysis, and liver damage (10).
The cause of the inflammatory response is not known. Some workers have assumed that alloimmune reactivity to the fetoplacental unit (11) is possible although the concept has evolved in relation to murine rather than human pregnancy. There is little specific evidence for a human equivalent. Inflammatory responses are not Ag specific and our current hypothesis is that one or more factors derived from the placenta are the stimuli, which does not depend on, or evolve into, Ag-specific immunity. The relevant issues are the intensity and quality of the maternal inflammatory response and its relation to the maternal syndrome of pre-eclampsia.
Four cytokines are of potential importance in generating the excessive inflammatory response of pre-eclampsia: TNF-, IL-12, IL-18, and IFN-. They are involved in the Shwartzman reaction, which is a lethal response to endotoxin after a prior priming injection. Pregnant animals are uniquely sensitive to the Shwartzman reaction in that the priming administration of endotoxin is not needed (several authors, for example; see Ref. 12). A notable feature of the HELLP syndrome is its hyperacute presentation in some women (several authors, for example; see Ref. 13), which matches the acute time course of the Shwartzman response. In experimental animals the Shwartzman reaction results from a positive feedback loop in which activated monocytes secrete IL-12 and IL-18 that stimulate lymphocytes to produce IFN-, which magnifies further the degree of activation of the monocytes (14). Hepatic necrosis, which is a typical consequence of the HELLP syndrome, occurs in a murine variant of the Shwartzman reaction (15), in which it has been shown that IL-18, IL-12, IFN- and TNF- all contribute to the experimental pathology. The process can be totally prevented by prior administration of anti-IL-18 but is only partially ameliorated by anti-IFN- or anti-TNF- (16). We have shown that, in human pregnancy, circulating monocytes are primed to produce IL-12 (17) which may indicate that a human equivalent to the priming of animals may occur.
The candidate stimulants for the systemic inflammatory response of pregnancy include circulating cytokines or antiangiogenic factors (18), products of oxidative stress (19), or subcellular debris shed from the syncytial surface of the placenta (20). We were particularly interested in the last, as in human hemochorial placentation, the outermost layer of the placenta, the syncytiotrophoblast, is in direct contact with maternal blood. Any microparticulate debris from the placenta will be shed into the maternal circulation and could interact with both maternal leukocytes and vascular endothelium. Furthermore, the syncytiotrophoblast is both class I and class II MHC negative and will therefore not provoke classical T cell allograft responses.
We devised a technique of flow cytometric measurement of the binding of syncytiotrophoblast membrane Ags to peripheral blood monocytes. We determined whether such binding occurred in normal pregnancy and, if so, at what stages of pregnancy it could be detected and if pre-eclampsia changed the degree of binding. In a final set of experiments, we investigated the role of syncytiotrophoblast membrane microparticles (STBM) that are shed into maternal blood from the placental surface during pregnancy (20) and whether they can stimulate the production of proinflammatory cytokines. For the latter purpose, microparticles were prepared from placentas from healthy pregnancies, both by the standard mechanical method (mSTBM) and from eluates of perfused isolated placental cotyledons (pSTBM), which may better represent the particles shed in vivo, and used in coculture experiments with PBMCs.
In this study, we sought to characterize the systemic inflammatory response of pregnancy and pre-eclampsia, the time course of its evolution during normal pregnancy, and the changes that are associated with pre-eclampsia. We studied the four inflammatory cytokines TNF-, IL-12, IL-18, and IFN-, which were selected because of their importance in the Shwartzman reaction. We determined how early during normal pregnancy circulating STBMs can be detected, whether they circulate in maternal blood bound to PBMCs, specifically monocytes, and whether they stimulate secretion of inflammatory cytokines from PBMCs when cultured ex vivo. We found that an enhanced inflammatory response can be elicited from PBMCs from the first trimester onwards, that it is characterized by suppression of lymphocyte production of IFN- and that this suppression is partially released in pre-eclampsia with enhanced IL-18 production. STBMs are proinflammatory in culture and could contribute to the systemic inflammatory response that we observe, both in normal pregnancy and in pre-eclampsia.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Penicillin, streptomycin, sodium pyruvate, 
-ME, PMA, ionomycin, brefeldin A, saponin, and rat IgG Ig isotypes were obtained from Sigma-Aldrich. RPMI 1640 with L-glutamine was obtained from Invitrogen Life Technologies. Recombinant human IFN- was supplied by Cambridge Biosciences. Monoclonal mouse anti-CD14 (conjugated with FITC), unconjugated anti-CD45, and rat anti-TNF- Abs (FITC-conjugated) were obtained from Serotec; mouse PE-conjugated anti-IL12p40/70, rat PE-conjugated anti-IL12p70, and FITC-conjugated anti-IFN- were obtained from BD Pharmingen and mouse FITC-conjugated anti-CD56 from Immunotech. NDOG2 and ED822 were in-house mAbs to human placental alkaline phosphatase (21) and an unknown Ag expressed on the apical surface of the syncytiotrophoblast (22), respectively. Mouse IgG used in control steps was obtained from Serotech (unconjugated) or BD Pharmingen (FITC-conjugated). PE-conjugated goat anti-mouse IgG was obtained from Dako.
Patients, normal subjects, and cell purification
To study normal gestational changes, healthy pregnant women were recruited in the first, second, and third trimesters and matched for age (±4 years) and parity (0, 1–3, 
4) with nonpregnant women. To study pre-eclampsia, patients were recruited and matched for age and parity with normal pregnant women and nonpregnant women. Pregnant women were also matched for gestation (±13 days). Normal pregnant women were studied in the first (12–14 wk), second (15–25 wk), and third (26–40 wk) trimesters of pregnancy. Nonpregnant women of reproductive age were recruited from staff members. None had any significant past medical history, current illness, or were taking regular medication (apart from folic acid and iron supplements). Pregnant patients were not in labor at the time of sampling.
Pre-eclampsia was diagnosed by new hypertension in the second half of pregnancy (blood pressure 140/90 mm Hg or an increase of >30/15 mm Hg from baseline at booking, on at least two occasions at least 6 h apart) and new proteinuria (2+ on dipstick testing on at least two occasions or 500 mg protein in a 24-h urine collection, in the proven absence of a urinary tract infection).
Case characteristics are detailed in Tables I and II. These studies were approved by the Oxfordshire Clinical Research Ethics Committee.
|
|
Peripheral venous blood (40 ml) was taken into sodium heparin (10 IU/ml), and the mononuclear cells (PBMCs) were separated by density gradient centrifugation using Ficoll-Paque PLUS (endotoxin tested). The final pellet was resuspended in RPMI 1640 with L-glutamine containing 1 mM sodium pyruvate, 50 U/ml penicillin, 50 µg/ml streptomycin, 0.02 mM 2-ME, and 10% FBS, at 2 x 106 cells/ml, and 1-ml aliquots were placed in a 24-well Sarstedt plate.
To stimulate TNF-, IL-18, and IL-12 production, PBMCs were preincubated for 2 h with 20 ng/ml recombinant human IFN-, followed by 22 h of stimulation with 1 µg/ml LPS plus 20 ng/ml IFN-. This protocol was used after preliminary comparisons of different protocols (data not shown). The final conditions were selected to optimize detection of all three cytokines for both intracellular and culture supernatant measurements. To stimulate IFN- production, PBMCs were incubated with 10 ng/ml PMA and 0.75 µg/ml (1 µM) ionomycin for 4 h. Brefeldin A (10 µg/ml) was used to inhibit PBMC cytokine secretion, for the measurements of intracellular cytokine. The cells were incubated at 37°C in 5% carbon dioxide in air. After the indicated time, they were resuspended and microfuged at 13,000 x g. The supernatants were stored at –80°C until use.
Cytokines were measured in culture supernatants using commercial ELISA kits (from PeproTech for TNF-, and from BioSource International for the remaining assays) following the manufacturers’ instructions. The plates were read on an MRX Microplate Reader using Revelation software. Samples were analyzed in duplicate together with appropriate standards and quality controls, and average values were determined.
To measure intracellular cytokines, each PBMC pellet was resuspended in 100 µl of staining buffer (47.5 ml of PBS-E, 2.5 ml of 10% sodium azide, and 250 µl of FBS). Cells were labeled in 1.5-ml Eppendorf tubes, on ice, in the dark. Abs directly conjugated to CD14 (monocytes; final dilution, 1/100) and to CD56 (NK cells; 1/50) were used to identify relevant cell types. Then, 25 µl of the Ab diluted in staining buffer were added to 0.5 x 106 PBMCs. Controls samples comprised staining buffer alone, and mouse IgG isotype Ig (MsIgG). The suspensions were incubated for 30 min, washed with 400 µl of staining buffer, and microfuged at 13,000 x g for 7 s; the pellet was resuspended in 50 µl of fixation buffer (4% paraformaldehyde in PBS-E), and left for 20 min. Samples were again washed, spun, and resuspended in 25 µl of permeabilization buffer (47.5 ml of PBS-E, 2.5 ml of 10% sodium azide, 250 µl of FBS, and 50 mg of saponin) containing the appropriate Ab and incubated for a further 30 min. Directly conjugated Abs were used to detect TNF- (final dilution, 1/5), IL-12p70 (2/5) and IL-12p40/70 (2/5) in monocytes, and IFN- (1/50) in NK cells. Control samples comprised cells in permeabilization buffer or incubated with MsIgG and rat IgG of the appropriate isotype.
Ten thousand events were collected by flow cytometry using a Beckman Coulter EPICS Altra, at 520 nm or 578 nm for analysis off-line. Monocytes and lymphocytes were identified by their optical characteristics. Gates set on single parameter histograms, using the isotype-negative controls to include 
1% of events, were used to determine the positive populations at both wavelengths. A quadrant gate was applied to a two-parameter plot corresponding to these negative control gates. Similar gates were set for the lymphocyte population, to measure the production of IFN- by both CD56+and CD56– lymphocytes. The proportions of cytokine positive monocytes or lymphocytes were measured. We were unable to develop a similar method for detecting intracellular IL-18 by monocytes using the commercially available mAb; therefore, only secreted IL-18 levels in culture supernatants could be measured.
Measurement of free STBMs in peripheral blood
Free STBM in plasma samples were measured using an in-house ELISA (23). In brief, peripheral blood samples were collected into sodium heparin and centrifuged at 2000 x g for 15 min at room temperature. The resultant plasma supernatant was diluted with PBS-E (at least 1/2) and ultracentrifuged (150,000 x g for 45 min at 4°C). The supernatant was discarded, and the final pellet was resuspended in 350 µl of PBS-BSA and stored at –80°C until use.
The capture Ab was NDOG2 (see above) which has been found to be optimal for detecting STBM in this ELISA (24). The reporter system detected endogenous alkaline phosphatase activity on the surface of the microparticles by a colorimetric reaction (Invitrogen Life Technologies). Plates containing triplicate samples were read on an MRX Microplate reader (Dynex Technologies) at 490 nm, at 5 min. The sensitivity of the assay is <10 pg/ml (25).
A pooled preparation of mSTBM from five placentas was used to prepare the standards for the STBM ELISA (see below). All the placentas had been obtained at elective cesarean section from healthy pregnant women with no history of pre-eclampsia or intrauterine growth retardation. A standard curve was determined to measure the STBM concentration in each 350-µl sample in nanograms per milliliter. The readings were adjusted by a concentration factor to calculate the STBM level in the original plasma sample.
Immune cell response to mSTBM and pSTBM
mSTBMs were prepared from normal placentae by a modification of the method of Smith et al. (26), and the protein content was determined using a BCA protein assay kit. pSTBM were prepared using a dual placental perfusion system as described by Eaton and Oakey (27). Placentas were obtained at cesarean section, without labor, from healthy pregnant women and were processed immediately. An individual lobule was isolated and perfused with modified M-199 tissue culture medium (Medium 199 with L-glutamine and Earle’s salts without NaHCO3, containing 0.8% Dextran 20, 0.5% BSA, 5000 U/L sodium heparin, and 2.75 g/L sodium bicarbonate, pH 7.4) at a controlled rate of 20 ml/min. The perfusion medium was warmed in a 37°C water bath and oxygenated with 95% O2, 5% CO2. Every 2.5 min after perfusion started, a 50-ml fraction of effluent from the maternal circulation was collected, labeled, and kept on ice. The volume of fetal effluent was measured simultaneously, and the oxygen concentration of the maternal side perfusate monitored to ensure the stability of the preparation. Pressure monitors were used to ensure no significant deviations from baseline during the experimental period. The 50-ml fractions of maternal side perfusate were centrifuged in a Beckman J6-M centrifuge at 2000 x g for 15 min at 4°C. Aliquots (12 ml) of the supernatants were spun at 150,000 x g for 45 min at 4°C in a Beckman L8-80M ultracentrifuge. Each pellet was resuspended in 350 µl of PBS-BSA and stored at –80°C until use. The protein content of each suspension was measured (BCA protein assay kit) and adjusted to 1.9 mg/ml. pSTBM in each aliquot were measured using the ELISA described previously.
To study stimulation of cytokine responses, PBMCs were isolated from nonpregnant individuals (n = 3) and cultured for 4 and 24 h, as described earlier. Cells were stimulated with 150 ng/ml mSTBM or pSTBM (prepared as described previously) or 1 µg/ml LPS and 20 ng/ml IFN- (for TNF-, IL-18, and IL-12) or 10 ng/ml PMA and 1 µM ionomycin (for IFN-). For the 24-h time point, the cells were preincubated with IFN- for 2 h. Secreted and intracellular levels of TNF-, IL-18, and IL-12p70 and IFN- were measured as described previously. In addition, the ability of pSTBM to stimulate TNF- secretion was titrated in a dose-response curve (0–300 µg/ml).
In vivo binding of STBM to circulating monocytes
PBMCs were prepared as described, and 20 x 106 cells/ml were resuspended in PBS-BSA. They were double-labeled with Abs against CD14 and ED822 that binds to a specific syncytiotrophoblast marker and has been found to be optimal for flow cytometry (24). mAb to CD45 (Serotec) was used as a positive control. Previous work (24) suggested that there is insignificant binding of STBM to other leukocyte subtypes. Aliquots of 0.5 x 106 cells in 1.5-ml Eppendorf tubes were labeled in 5 ml of PBS-E, 20 mM glucose, and 250 µl of normal human serum, while on ice and in the dark. Primary labeling was with 25 µl of unconjugated mouse monoclonal ED822 Ab (1/200) incubated for 30 min. After a washing (1 ml of PBS-BSA, microcentrifuged at 13,000 x g for 7 s), the sample was incubated for 30 min with secondary Ab (goat anti-mouse conjugated to PE at 1/100). After further washing, unoccupied binding sites were blocked using 90 µl of a 1/6 dilution of MsIgG. Monocytes were then labeled with FITC-conjugated anti-CD14. Cell pellets were resuspended in 25 µl of the Ab diluted 1/100 and incubated for 30 min. After a washing, each sample was resuspended in 1 ml of PBS/BSA, transferred to flow cytometry tubes, and kept on ice until analysis. Appropriate negative and single and double positive controls were also prepared. Samples were analyzed as described previously. The percentages of monocytes positive for both CD14 and a trophoblast marker (ED822) were determined.
Statistics
Data were compared using the Wilcoxon matched pairs test for studies with equal numbers of matched subjects in each group, the Mann-Whitney U test for studies with unequal numbers in the groups and the sign test for aggregated data.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
There were different groups of subjects for the studies of cytokine production by PBMCs (group A), of free syncytiotrophoblast microparticles (group B), and of leukocyte-bound STBM (group C). In each category, there were two subgroups for changes through normal pregnancy and changes in pre-eclampsia. Matching ensured that the relevant groups were comparable except for pregnancy outcome which was associated with substantially earlier gestational ages at delivery in pre-eclampsia. The patient details are shown in Tables I–III.
|
Short term cytokine production by unstimulated monocytes or lymphocytes for 24 h or less was low, without significant differences between any of the groups (data not shown). Therefore, only data for stimulated monocytes and lymphocytes are presented.
The overall response in terms of TNF-, IL-12p70, and IL-18 was a significant increase in secretion relative to the nonpregnant baseline in the first trimester. The patterns for each cytokine differed in the remaining two trimesters. IL-18 and IL-12p70 production showed a progressive decline after the first trimester, almost returning to the baseline level by the third in the case of IL-18. TNF- secretion, however, reached a plateau in the second trimester that was sustained in the last trimester (Fig. 1A). The changes in intracellular cytokine expression (TNF- and IL-12p70 only) were less variable. There were small first trimester increases, which were significant for TNF- only (Fig. 1B) and a nonsignificant decline in IL-12p70 which followed the trend for secreted cytokine.
|
had a reciprocal pattern. Pregnancy samples were characterized by less production in all three trimesters which was, however, significant only in the third trimester. When intracellular production was analyzed, the pattern was similar, with a more sustained inhibition in the CD56– lymphocytes than in the CD56+ NK (Fig. 1C). Cytokine production by monocytes and lymphocytes in pre-eclampsia
This comparison was confined to third trimester pregnancies and nonpregnant women. In general the trends documented in Fig. 1A for late normal pregnancy were confirmed for secreted TNF-, IL-12p70 and Il-18. In pre-eclampsia secreted TNF- was slightly diminished and no different from nonpregnancy samples; secreted IL-12p70 was unchanged from normal pregnancy and secreted Il-18 increased above normal pregnancy, to a significant extent relative to the nonpregnant baseline (Fig. 2A). Intracellular TNF- and IL-12p70 were once again no different between normal pregnancy and nonpregnancy but the % positive PBMCs were significantly increased in pre-eclampsia for TNF- only (Fig. 2B). The reciprocal inhibition of IFN-, both secreted and intracellular was again observed in normal pregnancy. The striking change was loss of this inhibition in pre-eclampsia (Fig. 2A and 2C).
|
Free STBM were not detected in first trimester samples but were present at significant levels in the second trimester and increased further in the third (Fig. 3A). In the first trimester, there was no significant binding of STBM compared with the nonpregnant controls (Fig. 3B); but significant binding was detected in the second trimester with a further increase in the third trimester.
|
0.001) but no further increase was apparent in pre-eclampsia (Fig. 3B). Measurement of cytokine production from PBMCs after stimulation by pSTBMs
mSTBM did not stimulate TNF- (Fig. 4A), IL-12, IL-18, or IFN- (data not shown) production by PBM from nonpregnant women. In contrast, pSTBMs were able to stimulate moderate levels of TNF- (Fig. 4A) with a significant increase in a dose-response curve (Fig. 4B). Not only TNF- but also IL-18 and low levels of IL-12p70 and IFN- production could be detected after only 4 h of culture of pSTBM with PBMCs (Fig. 5A; p < 0.05). These findings were confirmed by measuring intracellular production in PBMCs (Fig. 5B). pSTBM also stimulated low intralymphocytic levels of IFN- production from both CD56+ and CD56– lymphocytes.
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
was suppressed, although this was not statistically significant in the second trimester (p = 0.10, n = 10).
The response in pre-eclampsia was also altered in important ways, relative to matched normal pregnancy samples. IL-18 production was significantly increased and production of IFN- was no longer suppressed. These patterns are consistent with the consensus that there is a type 2 bias during normal human pregnancy (28, 29) and extend our earlier findings in relation to normal pregnancy (6, 17).
We studied isolated and stimulated PBMCs by measuring secreted inflammatory cytokines from cultured PBMCs and also intracellular cytokine production by flow cytometry on the same samples. Hence, the measures were not of basal production, which was negligible, but reflect priming and not in vivo activity. The gestational time courses of inflammatory cytokine production differed. For TNF-, an early increase was sustained throughout pregnancy; for IL-12 it peaked in the second trimester and declined thereafter; for IL-18, the increase was only significant in the first trimester. These changes could not be simply ascribed to changes in capacity to produce IFN- which was suppressed throughout pregnancy. In normal pregnancy, there is a monocytosis from an early stage (30) which might contribute to at least part of the differences, relative to the nonpregnant control women, in the different gestational patterns for secreted and intracellular TNF- and to a lesser extent for IL12p70. In pre-eclampsia, there is no consensus about changes in monocytes counts relative to normal pregnancy (31, 32). Monocyte counts were not available to interpret these results; thus, this possibility cannot be discounted, but it is unlikely given the evidence for inflammatory priming during pregnancy in animals (12) and humans (17).
It has been previously reported that unstimulated PBMCs from pregnant women produce less IL-12 relative to nonpregnancy, which the authors suggested explains the type 2 predominance of pregnancy (33, 34). These authors also observed increased spontaneous IL-18 production in the third trimester relative to nonpregnancy and argue that IL-18 alone is predominantly a Th2 cytokine, which is converted to a Th1 cytokine by IL-12 which induces IL-18R (35).
The changes in culture were not mirrored in the measures of intracellular TNF- or IL-12, which varied little throughout pregnancy apart from a small significant increase in TNF- in the first trimester when compared with normal pregnancy. Measures of the intensity of labeling for intracellular cytokine were similarly unremarkable (data not shown).
For TNF- these results are comparable to those we previously obtained for third trimester samples but do not replicate the evidence of priming for IL-12 production that we also demonstrated (17). In the present study, more intense stimulation with LPS (1000 vs 40 ng/ml) and IFN- (20 vs 5.5 µg/ml) was used, and a 2-h prestimulation with IFN- was added. In addition, monocytes were identified differently, i.e., on optical characteristics not by CD14 expression because CD14 is shed with monocyte activation (36) and may then become an unreliable monocytes marker. The more intense stimulation used in the present study may have obliterated the differences that were previously seen.
Basal production of IL-12 appears to be suppressed during normal pregnancy but, under certain conditions of stimulation, primed for increased production (13, 33). The most likely explanation for this conclusion is that, in vivo, the production of IL-12 is suppressed by the concurrent suppression of IFN- production. In vitro stimulation included exogenous IFN-, which would be expected to reverse the suppressive effects that low availability of IFN- in vivo causes.
Our present work is, in general, consistent with data of McCracken et al. (29) who studied isolated T cells from women with normal third trimester pregnancies. They demonstrated that inhibition of intracellular production of IFN- was caused specifically by loss of NF-
B activity. This, they speculated, could result from the high levels of circulating steroids which are characteristic of all stages of pregnancy from the second half of the first trimester onwards. This study extends their findings in three ways: we show 1) that IFN- production is reduced in all three trimesters; 2) that it is not limited to CD56– cells, which are predominantly T cells, but involves circulating NK cells (CD56+); and 3) that the inhibition is reduced or lost in pre-eclampsia. The involvement of NK cells is consistent with our previous report that suggested that both NK and NKT cells play a role in determining the type 1/type 2 balance in pregnancy and pre-eclampsia (2). In this study, the phenotype of the CD56+ cells was not investigated in detail, and the CD56– cells were not characterized other than for this attribute and their presence in the optical gate characteristic of lymphocytes. It was therefore not possible to explore this issue in detail, and we cannot exclude the participation of T cells.
We investigated circulating STBMs, shed from the placenta, as a potential stimulus for systemic inflammatory changes, particularly in pre-eclampsia. Free particles were detectable in plasmas of women with third trimester pregnancies and were significantly increased when pre-eclampsia was confirmed, which confirms our previous report (24). In this study, we show that the particles cannot be detected in the first trimester but increase thereafter. In normal pregnancy, significant monocyte binding of STBMs could also be detected by the second trimester, with a further increase in the third trimester. In pre-eclamptic patients, there were also significant levels of monocytes binding STBMs, but no further increase compared with gestation-matched normal pregnant controls, unlike the changes in unbound STBM.
We used mechanically prepared STBM and compared them with preparations from perfusates of normal term placentas because the latter are considered to be more physiological and free of contamination with microparticles derived from maternal sources (37, 38). The optimal conditions required for leukocyte activation by STBM remain undetermined. STBMs alone were added, whereas in vivo the presence of one or more proinflammatory cytokines would be predicted to enhance stimulation. STBMs stimulated only low levels of IFN- production in vitro, although it was always more than the negative control for both plasma and intralymphocytic measurements. The negative control samples were limited by the possible size of the experiments. They might have included microparticles from other sources, such as mechanically processed RBC membranes, eluates from perfused nonplacental tissues, or inert synthetic particles. We did not examine microparticles prepared from pre-eclampsia placentas, although these might better reflect the in vivo situation. A further difficulty with these kinds of experiments is selecting an appropriate control for trophoblast microparticles as the syncytiotrophoblast is a unique tissue. A continuing problem at this stage of microparticle research is that the different preparations and microparticle subtypes have not been standardized (39). The importance of this is underlined by the different effects of sSTBM and pSTBM. As stated, we standardized our preparations in terms of their protein content, but it is clearly important that more work is done to address this problem. A final question is whether the level of cytokine production depends on the concentration of STBM. A pSTBM dose of 150 µg/ml was chosen for these initial experiments, which is more than one order of magnitude greater than that detected in the circulation. However, what is measured in the circulation is probably only a small part of the total of shed STBM given that most will be rapidly cleared by monocytes and the reticuloendothelial system. Thus, clearance and concentration in the spleen and liver may create different and more reactive conditions.
The release of high levels of microparticles from the placenta into the maternal circulation in normal pregnancy and pre-eclampsia constitutes a specialized challenge to the maternal immune system. However, it is not unique as apoptotic release of microparticulate membrane fragments into peripheral blood has been implicated in several pathological conditions (summarized by Ahn in Ref. 39). Circulating microparticles may be derived from leukocytes (40), endothelial cells (41), platelets (42), or RBC (43). The particles carry bioactive molecules that deliver intercellular signals (42). They are thought to have important proinflammatory and procoagulatory roles (44, 45), including promoting interaction between endothelial cells and monocytes (40), platelet activation (46), and apoptosis (47). In relation to our studies, a key question is, "Can these circulating STBMs contribute to the immunoregulation of normal pregnancy or its dysregulation in pre-eclampsia?"
STBMs, shed from perfused placentas ex vivo, induced mononuclear cell production of the proinflammatory cytokines TNF-, IL-18, and IL-12, demonstrating that they have proinflammatory potential as described for microparticles from other cellular types (44, 45). In another study, STBMs prepared by mechanical methods increased IFN- production by peripheral blood T cells purified from healthy blood donors (38). Using different techniques, we observed a small but consistent increase in IFN- production. We did not test STBM prepared from pre-eclampsia placentas, but others find that they activate inflammatory leukocytes significantly more than STBM from normal pregnancies (48). This suggests a qualitative difference, perhaps owing to oxidized lipids (49), in addition to the quantitative increases that we document here. The mechanism of stimulation is not known. It is possible that oxidized lipids or other molecules carried by STBM are detected directly as danger signals (50) to stimulate directly inflammatory responses. An alternative explanation is that dendritic cells or monocytes process the particles, which can then be presented as small molecules by classical or nonclassical routes.
The systemic inflammatory response was detected from the first trimester onwards whereas circulating STBM could be detected only during the second and third trimesters. We have previously emphasized the two interfaces of maternal-fetal interactions (see Introduction). When the uteroplacental circulation opens toward the end of the first trimester (51), interface II begins to be activated. Before then, STBM cannot be released directly into the maternal circulation, and at that time and immediately afterward what is released is limited by the small size of the placenta and may be undetectable with our current assays. We speculate that earlier systemic inflammation is secondary to the intense localized interactions between decidual immune cells (maternal) and trophoblast (fetal) at interface I. Therefore, our current data raise the possibility that shed STBM, which become detectable in the second trimester, increase in the third and are further enhanced in pre-eclampsia, contribute to a second phase of systemic inflammation by stimulating an increasing counter stimulus to the immunosuppression of normal pregnancy and have the potential to overcome it. If IFN- production were promoted, (for example in pre-eclampsia or coincident infection), the capacity to over-produce IL-12 and IL-18 could be released and generate the feed-forward loop that promotes the Shwartzman reaction. This may help to explain the unique sensitivity of pregnant animals to the Shwartzman reaction, without an additional priming dose of endotoxin. These changes probably relate to immune interface II as discussed in the Introduction. However, the inflammatory changes in the first trimester more likely reflect events secondary to the intense remodeling in the decidua, which involves inflammatory and immune processes (52). We have not studied the regression of the systemic inflammatory changes after delivery. The little that we do know is that detectable STBM are cleared almost completely from plasma samples within 24 h (A. Reddy, I. L. Sargent, and C. W. Redman, unpublished data).
|
Acknowledgments |
|---|
|
Disclosures |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Footnotes |
|---|
1 This project was supported by Action Research Grant SP3441 and by the Oxford Radcliffe Hospitals Trust Research and Development Fund. 
2 Address correspondence and reprint requests to Dr. Christopher W. Redman, Nuffield Department of Obstetrics and Gynaecology, John Radcliffe Hospital, Oxford, U.K. E-mail address: christopher.redman{at}obs-gyn.ox.ac.uk
3 Abbreviations used in this paper: HELLP syndrome, acute hemolysis, elevated liver enzymes, and low platelets; STBM, general term for syncytiotrophoblast microparticles; mSTBM, STMsB prepared by a mechanical method; pSTBM, STMsB prepared by perfusion of the maternal surface of the placenta; MsIgG, mouse IgG isotype Ig.
Received for publication August 30, 2006. Accepted for publication January 29, 2007.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
, and tumor necrosis factor are the key cytokines of the generalized Shwartzman reaction. J. Exp. Med. 180: 907-915. - and Fas ligand-mediated hepatotoxic pathways in endotoxin-induced liver injury in mice. J. Immunol. 159: 3961-3967. [Abstract]B expression in T cells in humans is essential for the maintenance of the cytokine profile required for pregnancy success. J. Immunol. 172: 4583-4591. This article has been cited by other articles:
|
|
 |
|
  K. King, S. Smith, M. Chapman, and G. Sacks Detailed analysis of peripheral blood natural killer (NK) cells in women with recurrent miscarriage Hum. Reprod., October 9, 2009; (2009) dep349v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Yang Gu, Chang Liu, J. S. Alexander, L. J. Groome, and Yuping Wang Chymotrypsin-Like Protease (Chymase) Mediates Endothelial Activation by Factors Derived From Preeclamptic Placentas Reproductive Sciences, September 1, 2009; 16(9): 905 - 913. [Abstract] [PDF]
|
|
|
|
 |
|
 H. S. Nielsen, R. Steffensen, K. Varming, A. G.S. Van Halteren, E. Spierings, L. P. Ryder, E. Goulmy, and O. B. Christiansen Association of HY-restricting HLA class II alleles with pregnancy outcome in patients with recurrent miscarriage subsequent to a firstborn boy Hum. Mol. Genet., May 1, 2009; 18(9): 1684 - 1691. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. P. Murphy, C. Tayade, A. A. Ashkar, K. Hatta, J. Zhang, and B. A. Croy Interferon Gamma in Successful Pregnancies Biol Reprod, May 1, 2009; 80(5): 848 - 859. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. A. R. Lok, A. N. Boing, I. L. Sargent, S. R. Sooranna, J. A. M. van der Post, R. Nieuwland, and A. Sturk Circulating Platelet-derived and Placenta-derived Microparticles Expose Flt-1 in Preeclampsia Reproductive Sciences, December 1, 2008; 15(10): 1002 - 1010. [Abstract] [PDF]
|
|
|
|
 |
|
 A. F. Orozco, C. J. Jorgez, C. Horne, D. A. Marquez-Do, M. R. Chapman, J. R. Rodgers, F. Z. Bischoff, and D. E. Lewis Membrane Protected Apoptotic Trophoblast Microparticles Contain Nucleic Acids: Relevance to Preeclampsia Am. J. Pathol., December 1, 2008; 173(6): 1595 - 1608. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Hering, F. Herse, S. Verlohren, J.-K. Park, M. Wellner, F. Qadri, R. Pijnenborg, A. C. Staff, B. Huppertz, D. N. Muller, et al. Trophoblasts Reduce the Vascular Smooth Muscle Cell Proatherogenic Response Hypertension, February 1, 2008; 51(2): 554 - 559. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. V. Ilekis, U. M. Reddy, and J. M. Roberts Review Article: Preeclampsia A Pressing Problem: An Executive Summary of a National Institute of Child Health and Human Development Workshop Reproductive Sciences, September 1, 2007; 14(6): 508 - 523. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
