Trophoblast expression of the minor histocompatibility antigen HA-1 is regulated by oxygen and is increased in placentas from preeclamptic women

Abstract

Introduction

Maternal T-cells reactive towards paternally inherited fetal minor histocompatibility antigens are expanded during pregnancy. Placental trophoblast cells express at least four fetal antigens, including human minor histocompatibility antigen 1 (HA-1). We investigated oxygen as a potential regulator of HA-1 and whether HA-1 expression is altered in preeclamptic placentas.

Methods

Expression and regulation of HA-1 mRNA and protein were examined by qRT-PCR and immunohistochemistry, using first, second, and third trimester placentas, first trimester placental explant cultures, and term purified cytotrophoblast cells. Low oxygen conditions were achieved by varying ambient oxygen, and were mimicked using cobalt chloride. HA-1 mRNA and protein expression levels were evaluated in preeclamptic and control placentas.

Results

HA-1 protein expression was higher in the syncytiotrophoblast of first trimester as compared to second trimester and term placentas (P<0.01). HA-1 mRNA was increased in cobalt chloride-treated placental explants and purified cytotrophoblast cells (P=0.04 and P<0.01, respectively) and in purified cytotrophoblast cells cultured under 2% as compared to 8% and 21% oxygen (P<0.01). HA-1 mRNA expression in preeclamptic vs. control placentas was increased 3.3-fold (P=0.015). HA-1 protein expression was increased in syncytial nuclear aggregates and the syncytiotrophoblast of preeclamptic vs. control placentas (P= 0.02 and 0.03, respectively).

Discussion

Placenta HA-1 expression is regulated by oxygen and is increased in the syncytial nuclear aggregates and syncytiotrophoblast of preeclamptic as compared to control placentas. Increased HA-1 expression, combined with increased preeclamptic syncytiotrophoblast deportation, provides a novel potential mechanism for exposure of the maternal immune system to increased fetal antigenic load during preeclampsia.

1. Introduction

Pregnancy is a unique biological process wherein the female is confronted with an antigenically foreign fetus that is permitted and encouraged to flourish, despite a surrounding immunological environment that can recognize it as foreign. Thus, the success of pregnancy depends on maternal immunological acceptance of the fetus. Disruption of the biological processes governing maternal-fetal interactions can lead to the development of serious pregnancy complications. Preeclampsia is a common and dangerous complication that is specific to pregnancy and affects 5–8% of pregnancies worldwide [1]. It is a syndrome characterized by new onset maternal hypertension after 20 weeks gestation and proteinuria, and if untreated, can lead to maternal seizures, multi-system organ failure and even death of both mother and baby. The only successful treatment of preeclampsia is delivery of the baby and the placenta; making preeclampsia one of the leading causes of iatrogenic prematurity [2]. Numerous studies have indicated a central, causative role for the placenta in the manifestation of preeclampsia; for example, preeclampsia can develop in molar pregnancies, in which no fetus is present [3], and further, delivery of the placenta, not the baby, relieves the maternal symptoms [4]. Although the pathophysiology of preeclampsia is not well understood, alterations in vascular reactivity leading to both increased placental oxidative stress and enhanced maternal immune activation are key components of this syndrome [5].

The placenta plays an important role in modulating the response of the maternal immune system to the baby during pregnancy. While the placenta tightly restricts its expression of the highly polymorphic human leukocyte antigens [6] [7], which govern adaptive immune recognition, it does express polymorphic fetal proteins encoded outside the HLA complex, known as minor histocompatibility antigens (mHAg) [8], which can also be recognized by maternal immune cells. Minor antigens are important in the success or failure of hematopoietic stem cell and solid organ transplants [9–11]. Exposure of an untolerized individual to these antigens, either as the result of pregnancy or allogeneic transplantation, can lead to the expansion of minor antigen-specific T cells [12, 13]. Further, the presence of minor antigen-specific T cells in allogeneic transplants has been linked to increased rates of graft rejection, graft-versus-host disease, and graft-versus-leukemia effects, indicating that, at least in the context of transplantation, these T cells are functional and pathophysiologically relevant [9, 11, 14, 15]. Recent work by our lab and others has demonstrated that the placenta is a source of fetal minor antigens during pregnancy, and thus contributes to maternal immunological recognition of the fetus [8]. T cells with specificity for paternally-inherited fetal antigens expand during pregnancy and can persist for up to 22 years following delivery of the baby [13, 16, 17]. HA-1 (also called HMHA1 for human minor histocompatibility antigen 1) functions as a Rho GTPase activating protein [23], is expressed in the human placenta [8] and is important for graft outcomes in HLA-matched hematopoietic stem cell transplantations. A study of 148 HLA-matched bone marrow transplant pairs noted that a mismatch of HA-1 alone was significantly correlated with increased graft-versus-host disease [11]. HA-1 is antigenic due to a single nucleotide polymorphism in which the arginine in the peptide sequence KECVLRDDLLEA is changed to a histidine, giving rise to the immunogenic peptide KECVLHDDLLEA. This single nucleotide polymorphism results in increased binding affinity of the peptide to the appropriate major histocompatibility complex (MHC) for immune recognition, in this case HLA-A*0201, A*0206, B*60 or B*40012 [24–28]. HA-1-specific CD8+ T cells have been identified in human pregnancies in which an HLA-A2-positive mother expresses the non-immunogenic form of the peptide and the fetus expresses the immunogenic form [13]. Given the expansion of specific T cells during pregnancy and the importance of HA-1 in transplantation immunology, we investigated the expression and regulation of HA-1 in the human placenta during pregnancy, when the placenta and fetus together act as an allograft that can be recognized by the maternal immune system.

2. Materials and Methods

2.1 Sample Collection

Samples were collected in accordance with the Helsinki Declaration using protocols approved by the Institutional Review Boards of the University of Kansas Medical Center and Saint Luke’s Hospital. First trimester placentas were obtained from women undergoing elective pregnancy terminations at the Center for Women’s Health in Overland Park, Kansas. Normal term placentas, defined as placentas from pregnancies at 37 gestational weeks or later with no noted complications other than fetal malpresentation (including intrauterine growth restriction and diabetes), were obtained following non-laboring Caesarian deliveries at the University of Kansas Medical Center and Saint Luke’s Hospital. Fixed and frozen tissues from preeclamptic placentas, together with gestational age- and mode of delivery-matched control placentas were obtained from the Research Centre for Women’s and Infants’ Health Biobank at Mount Sinai Hospital, Toronto, Canada. Control placentas were selected primarily based on the mode of delivery and gestational age, and excluded pregnancies complicated by gestational hypertension or intrauterine growth restriction. We compared the clinical parameters outlined in Table 1 and the only significant difference between the control and preeclampsia groups was that infants in the preeclampsia group weighed significantly less than those in the control group (p<0.01). This is not unexpected given the high concordance of preeclampsia and intrauterine growth restriction [29].

Table 1.

Clinical Characteristics of Preeclamptic and Matched Control Placentas

Sample	Clinical Information	GA (weeks/ days)	Ethnicity	Age	Parity (Gravidity/ Parity)	Antenatal glucocorticoids? (Y/N)	Maternal cigarette use? (Y/N)	BMI	Mode of Delivery	Fetal Birth Weight	Fetal Sex (M/F)
1 - PE	PE	34/6	East Indian	37	3/1	N	N	19.83	Vaginal	2140g	F
1 - C	PTL	34/5	Caucasian	31	1/0	N	N	25.91	Vaginal	2710g	M
2 - PE	PE	37/2	Caucasian	37	3/0	N	N	19.47	Vaginal	2710g	M
2 - C	Labor	37/0	Asian	30	1/0	N	N	21.51	Vaginal	2870g	M
3 - PE	PE/HELLP	36/5	Caucasian	31	1/0	N/A	N	24.34	C-section	2640g	M
3 - C	Placenta praevia	36/3	Black	35	2/1	N	N	26.78	C-section	2665g	M
4 - PE	PE/HELLP/IUGR	32/4	Caucasian	39	2/0	Y	N	33.96	C-section	1220g	F
4 - C	PTL	32/4	Caucasian	33	1/0	Y	N	21.1	C-section	2080g	M
5 - PE	PE	36/2	Asian	38	1/0	N	N	21.64	Vaginal	2265g	M
5 - C	PTL	35/6	Asian	32	1/0	Y	N	19.88	Vaginal	2920g	F
6 - PE	PE/IUGR	34/1	Black	24	1/0	N	N	23.62	C-section + labor	1550g	F
6 - C	PTL	34/1	Caucasian	27	4/1	Y	N	26.18	C-section + labor	2360g	M
7 - PE	PE	36/0	Caucasian	37	1/0	N	N	26.4	C-section + labor	2280g	M
7 - C	PTL	35/4	Other	36	1/0	N	N	21.19	C-section + labor	3030g	F
8 - PE	PE/IUGR	31/5	Black	38	4/1	Y	N	26.22	Vaginal	890g	F
8 - C	PTL	32/1	Caucasian	29	2/1	Y	N	17.99	Vaginal	1810g	M
9 - PE	PE/IUGR	34/4	Native Canadian	37	4/0	Y	N	25.15	C-section + labor	1673g	F
9 - C	PTL	34/6	Caucasian	36	2/1	N	N	35.16	C-section + labor	2180g	M

GA = Gestational Age, PE = Preeclampsia, C = Control, PTL = Pre-term Labor, HELLP = Hemolysis, Elevated Liver enzymes, Low Platelets, IUGR = Intrauterine Growth Restriction, N/A = Not Available, BMI = Body Mas Index, M = Male, F = Female

2.2 Placental explant preparation

First trimester (n=6) villi were dissected and explants of approximately 500mg were cultured in 12-well plates fitted with Netwell inserts (Corning, Tewksbury, MA). Explants were cultured in DMEM-F12 media (Corning Cellgro, Herndon, VA) containing 10% fetal bovine serum (Atlanta Biologicals, Lawrenceville, GA), 5ng/ml epidermal growth factor (PeproTech, Rocky Hill, NJ), 5μg/ml insulin (Sigma-Aldrich, St. Louis, MO), 10μg/ml transferrin (Sigma-Aldrich), 100μg/ml L-glutamine (Corning Cellgro), 20nM sodium selenite (Sigma-Aldrich), 400 IU hCG (Sigma-Aldrich), 100μg/ml streptomycin (Corning Cellgro) and 100 IU penicillin (Corning Cellgro).

2.3 Cytotrophoblast isolation

Cytotrophoblast cells were isolated as previously described [30, 31]. Briefly, term placentas were collected following normal, non-laboring, Caesarian deliveries and approximately 40g of villous material was dissected. The tissue was enzymatically digested using Trypsin (Gibco, Grand Island, NY) and DNase (Sigma-Aldrich), washed, and layered over a Percoll gradient (Sigma-Aldrich). Cells were further washed and subjected to negative immuno-magnetic selection using an anti-class I (W6/32) MHC antibody. Only cell preparations resulting in greater than 90% purity as assessed by cytokeratin-7 positivity by flow cytometry were used. Spontaneous differentiation of cytotrophoblast cells cultured for 24 hours at 2%, 8% and ambient oxygen concentrations is minimal, as assessed by quantification of human chorionic gonadotropin [31].

2.4 Cobalt chloride treatment

First trimester placental explants (n=6) and purified term cytotrophoblast cells (n=5) were plated overnight at 37°C with 5% carbon dioxide, 95% normal air. The following day, cells were treated with either medium alone or media containing 100μM cobalt chloride. After 24 hours, cytotrophoblast cells were lysed in TRI Reagent (Sigma-Aldrich) and processed for RNA isolation as described below; placental explants were frozen in liquid nitrogen prior to mRNA analysis.

2.5 Oxygen cultures

After overnight plating, cytotrophoblast cells (n=6) were washed with pre-equilibrated medium and were then cultured under 2% or 8% oxygen or ambient air (21% oxygen) for 24 hours using an InVivo 300 (Ruskinn/Baker, Sanford, ME) or an H35 (Hypoxygen, Frederick, MD) hypoxia chamber, or in a standard incubator for the ambient air condition. After 24 hours, cells were collected in TRI Reagent and processed as described below.

2.6 RNA Isolation and RT-PCR

RNA was extracted from purified cytotrophoblast cells, first trimester placental explants and frozen placental tissues using TRI Reagent according to the manufacturer’s directions. The RNA concentration was quantified using spectrophotometry and 0.5μg were reverse transcribed using Superscript III Reverse Transcriptase (Invitrogen, Carlsbad, CA) and random primers (Invitrogen, Carlsbad, CA) in a reaction volume of 20 μl. 2 μl of the reverse transcription (RT) reaction was subjected to polymerase chain reaction (PCR) using Taq DNA polymerase (Fermentas, Glen Burnie, MD). RT-PCR for HA-1 was carried out using a TaqMan Gene Expression Assay for HA-1 (Hs00299628_m1, Applied Biosystems, Foster City, CA) or beta-actin (Hs99999903). Average fold change for each sample was calculated using ddCt values and normalized to the beta-actin control.

2.7 Immunohistochemistry

Paraformaldehyde-fixed (4%), paraffin embedded placentas cut to a thickness of 5μm were collected from preeclamptic (n=8) and gestational age- and mode of delivery-matched control women (n=8) at the Research Center for Women’s and Infants’ Health Biobank at Mount Sinai Hospital, Toronto, Canada. Slides were deparaffinized in Histo-clear (Fisher Scientific, Pittsburgh, PA), rehydrated and stained as previously described [32]. Tissues were subjected to heat antigen retrieval using Reveal buffer (BioCare Medical, Walnut Creek, CA), and nonspecific antibody binding was blocked in 10% goat serum (Sigma-Aldrich). A rabbit polyclonal anti human HA-1 antibody (Sigma-Aldrich, 12 μg/ml) or rabbit IgG (Vector Labs, Burlingame, CA, 12ug/ml) was added to tissue sections overnight at 4°C. Following addition of biotinylated goat anti-rabbit secondary antibodies (Vector Labs, 10 μg/ml) and depletion of endogenous peroxidases, binding of primary antibody was detected using streptavidin-peroxidase and aminoethyl carbazole (Invitrogen, Carlsbad, CA).

2.8 H-score Quantification of HA-1 Expression

Expression of HA-1 in first trimester, second trimester, term uncomplicated and preeclamptic placental tissues was calculated using the H-score method [33]. In the preeclamptic and matched control placentas, expression of HA-1 in the syncytiotrophoblast, syncytial nuclear aggregates and Hofbauer cells was quantified. Identification of syncytiotrophoblast and Hofbauer cells was based on location and morphology. Syncytial nuclear aggregates (SNAs) were defined as 10 or more syncytial nuclei clustered together that either protruded from a single villus or were not in contact with any other villi, as previously described [34]. Briefly, 10–20 400X images were taken of each placenta using a Nikon 80i microscope. The expression of HA-1 in each placental villus, cell or SNA was scored by a blinded reviewer as 3 (strong immunoreactivity), 2 (moderate), 1 (weak), or 0 (no immunoreactivity). At least 200 villi were quantified for each placenta. The H-score was quantified using the following formula: H-score = 3*(%strong) + 2*(%moderate) + 1*(%weak).

2.9 Statistical Analyses

All statistical analyses were conducted using the SigmaStat software (Systat, Chicago, IL). The differences between HA-1 mRNA expression in cobalt chloride-treated and untreated first trimester placental explants and purified term cytotrophoblast cells were calculated using two-tailed, paired T-tests. The differences in HA-1 mRNA expression across gestation and in purified term cytotrophoblast cells cultured at 2%, 8% and 21% oxygen were calculated using one-way ANOVAs on Ranks with Student-Newman-Keuls post-tests. The differences between HA-1 mRNA expression and H-scores in preeclamptic and normal placental tissues were calculated using two-tailed, paired T-tests. All results with a P-value of ≤ 0.05 were considered significantly different.

3. Results

3.1 HA-1 is expressed in placental tissues across gestation

Previous work by our lab has shown that HA-1 and other mHAgs are expressed in placental lysate, purified cytotrophoblast cells and fetal cord blood [8]. Nonquantitative analysis suggested that HA-1 is strongly expressed in the syncytiotrophoblast of first trimester, but is reduced in second and third trimester placentas. To further explore the possibility of differential regulation of HA-1 in the placenta across gestation, we used semiquantitative analysis to examine expression of HA-1 protein in the syncytiotrophoblast of first (n=5), second (n=4) and third (n=5) trimester placentas (Figure 1). The overall cellular distribution of HA-1 agreed with our previous findings: Hofbauer cells from all stages of gestation exhibited strong staining for HA-1, while immunoreactivity in trophoblast populations was most evident in the first trimester. Furthermore, H-score analysis revealed that HA-1 protein was more than 4- and nearly 40-fold higher in the syncytiotrophoblast of the first trimester placentas as compared to the second (P=0.001) and third (P<0.001) trimester placentas, respectively. In contrast, second trimester syncytiotrophoblast HA-1 expression was minimal, and remained so at term (P>0.05). Although there was a small subset of second trimester and term placentas with clear expression of HA-1 in the syncytiotrophoblast, most second trimester and term placentas showed little or no HA-1 expression in the syncytiotrophoblast, suggesting that in many cases HA-1 expression in the syncytiotrophoblast may be confined to the first trimester.

Figure 1. Expression of HA-1 in the human placenta across gestation.

HA-1 expression in the syncytiotrophoblast (small arrows) and Hoffbauer cells (large arrows) of first trimester (n=5), second trimester (n=4) and term placental sections (n=5) was examined by immunohistochemistry and quantified using the H-score method. Expression of HA-1 in the syncytiotrophoblast was significantly higher in first trimester placentas as compared to second trimester (P<0.01) and third trimester placentas (P<0.01). There was no difference in HA-1 protein expression between the second and third trimester placentas (P=0.277). * = P<0.01, ns=no significance.

3.2 Cobalt chloride treatment increases the expression of HA-1 mRNA in first trimester placental explants and purified term cytotrophoblast cells

Oxygen tension in the placenta increases dramatically from the first to the second trimester of pregnancy as a result of trophoblast remodeling of the maternal spiral arteries, suggesting that oxygen could be, at least in part, responsible for the observed alterations in HA-1 expression across gestation [35]. Therefore, to examine potential regulation of HA-1 expression by oxygen, we treated first trimester placental explants and purified term cytotrophoblast cells with 100μM of the hypoxia mimetic, cobalt chloride (CoCl₂). Cells or explants were treated in the presence or absence of 100μM CoCl₂ for twenty-four hours, and HA-1 mRNA was quantified by real time RT-PCR (Figure 2). There was a significant increase in HA-1 mRNA expression in the first trimester placental explants (n=6, p=0.04) and in the purified term cytotrophoblast cells (n=5, p<0.01) treated with cobalt chloride as compared to the untreated controls.

Figure 2. Cobalt chloride and low oxygen increase HA-1 mRNA expression in culture.

First trimester explants (n=6) and purified term trophoblast cells (n=5) were cultured in the presence of cobalt chloride, a hypoxia mimetic. Purified term trophoblast cells from normal, C-section deliveries (n=6) were cultured in 2%, 8% or 21% oxygen for twenty-four hours. HA-1 mRNA expression, determined using qRT-PCR, was normalized to beta-actin and fold change relative to the untreated controls or 21% oxygen condition, as appropriate, was calculated using the ddCt method. HA-1 mRNA expression was elevated in the first trimester placental explants and purified term cytotrophoblast cells treated with 100μM cobalt chloride as compared to the untreated controls (P=0.04 and P<0.01, respectively). HA-1 mRNA expression was significantly higher in the 8% and 2% oxygen conditions as compared to 21% oxygen (P<0.01). HA-1 mRNA expression was also significantly higher in the 2% oxygen condition as compared to the 8% oxygen condition (P<0.01). *=P<0.01.

3.3 Low oxygen increases the expression of HA-1 mRNA in purified term cytotrophoblast cells

To directly examine a potential role of oxygen in regulating HA-1 mRNA expression, purified term cytotrophoblast cells (n=6) were cultured at 2% (hypoxia for the term placenta), 8% (normoxia for the term placenta) and 21% (ambient) oxygen for twenty-four hours. The expression of HA-1 mRNA was then determined using qRT-PCR. HA-1 mRNA expression was significantly different between all three groups (p<0.01), with the highest expression seen in the cells cultured at 2% oxygen (Figure 2).

3.4 HA-1 is increased in the syncytiotrophoblast and syncytial nuclear aggregates of preeclamptic placentas

Previously published observations of increased expression of HIF-1α in preeclamptic placentas [36], the disrupted placental blood flow characteristic of this disease [37, 38], and the observed upregulation of HA-1 mRNA by low oxygen, led us to postulate that HA-1 expression may be dysregulated in placentas from preeclamptic patients. Preeclamptic and control placentas were matched for gestational age and mode of delivery (Table 1). Using qRT-PCR, we found that HA-1 mRNA expression (n=9 matched pairs) was significantly increased in placentas from preeclamptic pregnancies as compared to control placentas (Figure 3A, p=0.015). We also used immunohistochemistry to determine the expression of HA-1 protein in the Hofbauer cells, syncytial nuclear aggregates (SNAs) and syncytiotrophoblast layers of preeclamptic and control placentas (n=8 matched pairs; pair 4 was excluded due to poor tissue fixation). Expression of HA-1 in the SNAs and syncytiotrophoblast of preeclamptic placentas was highly variable and showed an uneven distribution within tissue sections, possibly as a result of local differences in oxygen tension and/or HIF-1α expression. Nonetheless, HA-1 protein expression was significantly higher in the SNAs and syncytiotrophoblast of preeclamptic placentas when compared to the matched control placentas (Figure 3B, C, E, F, p=0.02 and 0.03, respectively). It is worth noting that the majority (approximately 80%) of the SNAs observed are likely sectioning artifacts; in the absence of analysis of serial sections, we cannot discriminate between true syncytial knots, villous tips, and bridges between adjacent villi [33]. In addition, we quantified the number of SNAs/villi and found that there was no difference between the preeclamptic and control placentas (0.39 vs. 0.28 SNAs/villi, p=0.15). Despite the observed differences of trophoblast HA-1 expression the between control and preeclamptic placentas, HA-1 expression in the Hofbauer cells (placental macrophages) was similarly high in both groups, (p=0.69), suggesting that regulation of HA-1 expression differs among different cell types.

Figure 3. HA-1 expression is increased in preeclamptic placentas as compared to controls.

Whole placental lysate and placental sections were obtained from preeclamptic and gestational-age and mode of delivery matched controls. (A) HA-1 mRNA expression was determined using qRT-PCR normalized to beta-actin. HA-1 mRNA expression (fold change) was significantly higher in the preeclamptic placentas as compared to the control placentas (n=9, P=0.015). (B–D) HA-1 protein expression in the syncytiotrophoblast and syncytial nuclear aggregates was significantly elevated in the preeclamptic placentas as compared to the controls (n=8, p=0.03 and p=0.02, respectively), whereas there was no difference in HA-1 expression in the Hofbauer cells of preeclamptic placentas as compared to controls (n=8, p=0.69). (E–G) HA-1 protein expression in the syncytiotrophoblast layer (black arrows), syncytial nuclear aggregates (red arrows) and Hofbauer cells of preeclamptic and control placentas was determined by a blinded reviewer using the H-score method. PE = Preeclampsia.

IV. Discussion and Conclusions

Herein we report the first work to investigate mechanisms that regulate the expression of the minor histocompatibility antigen HA-1. We found that expression of HA-1 is significantly higher in the syncytiotrophoblast of first as compared to second and third trimester placentas. We also found that treatment with cobalt chloride, culture in hypoxic conditions, and preeclampsia all increase trophoblast expression of HA-1. One of the major changes that occurs during the transition from the first to the second trimester of pregnancy is a dramatic increase in placental oxygen delivery as a result of maternal spiral artery remodeling by invasive trophoblast cells and the loss of cytotrophoblast cell aggregates which block arterial blood flow to the placenta during early pregnancy [35, 39]. This process may help explain why HA-1 expression in the syncytiotrophoblast decreases as normal gestation progresses but appears to be maintained in preeclampsia, where the remodeling of the spiral arteries is defective.

The increased expression of HA-1 in purified cytotrophoblast cells following treatment with CoCl₂ and culture in a low oxygen environment suggests that HA-1 expression may be regulated, at least in part, by oxygen in vivo. This regulation likely occurs via the oxygen-sensitive transcription factor, hypoxia inducible factor (HIF) 1α. Under normoxic conditions, HIF-1α is ubiquitinated and degraded by prolyl hydroxylase, thus preventing its action, while under hypoxic or stress conditions, or in vitro in the presence of cobalt ions, the prolyl hydroxylase is impaired and HIF-1α function is maintained [40–42]. In support of the idea that HIF-1α regulates HA-1 expression in trophoblast cells, there are at least 3 sequences (5′-GCGTG-3′) in the upstream promoter region of the HA-1 gene that could serve as HIF-1α binding sites (800, 219 and 100 base pairs upstream of the HA-1 transcription start site). HIF-1α expression is also increased in preeclamptic as compared to normal placentas [36]. HIF-1α expression in the preeclamptic placenta may be due to changes in oxygen tension, but HIF-1α can also be upregulated by other processes that occur in preeclampsia, including oxidative stress [43] and heightened proinflammatory cytokines [44]. One limitation of this study as a model of pre-eclampsia is the culture of purified cytotrophoblast cells at static oxygen conditions, rather than cycling conditions of hypoxia and normoxia, which would more closely mimic pre-eclampsia in vitro [45]. We chose to examine HA-1 expression at static oxygen conditions in order to better clarify the precise role of oxygen in directly regulating HA-1 expression and remain interested in future studies examining the role of varying oxygen conditions in regulating HA-1 expression.

HA-1-specific CD8+ T cells have been found in the peripheral blood following human pregnancy [13], indicating that exposure of the maternal immune system to fetal HA-1 occurs during normal pregnancies. Several factors may alter maternal exposure to HA-1, and possibly other minor histocompatibility antigens, in preeclampsia. First, the alteration in placental blood flow at the maternal-fetal interface in preeclamptic women leads to areas of ischemia/reperfusion [46], the production of reactive oxygen species (ROS) [45, 47–49] and an increased deportation of syncytiotrophoblast fragments [50–54]. The increased release of syncytiotrophoblast microparticles, including the subset of syncytial nuclear aggregates known as syncytial knots, into the peripheral blood, together with increased HA-1 expression could mean that the maternal immune system is exposed to higher concentrations of HA-1 in preeclampsia than in normal pregnancies.

Second, the production of acute phase reactants and pro-inflammatory cytokines, specifically TNF-α and IL-6 [55], is characteristic of preeclampsia. This increase in pro-inflammatory cytokines together with increased expression of fetal antigen could combine to elicit a heightened antigen-specific maternal immune response in pregnancies complicated by preeclampsia. Although heightened fetal antigen-specific immune responses in preeclampsia have not been definitively shown to date, some authors have suggested the possibility that disruptions in maternal-fetal tolerance [4, 56, 57], a shift from Th2-mediated immunity to Th1-mediated immunity [58, 59], and a reduction in quantity or function or regulatory T cells [60] may play important roles in the pathophysiology of preeclampsia.

In conclusion, we have shown that low oxygen levels as well as preeclampsia increase HA-1 expression by trophoblast cells of the human placenta. The increased expression of HA-1 in the syncytiotrophoblast and syncytial nuclear aggregates of preeclamptic placentas combined with increased syncytiotrophoblast deportation and elevated levels of proinflammatory cytokines may have significant effects on maternal immune activation during preeclamptic as compared to normal pregnancies. The effects of this enhanced immune activation on current and subsequent pregnancies is unknown and may provide important insights into the manifestation of preeclampsia, other pregnancy complications and the success or failure of HA-1-mismatched transplantations. Further, given the function HA-1 as a Rho GTPase activating protein, and the potential for HA-1 expression to influence actin cytoskeletal remodeling [23], the regulation of HA-1 expression is likely to be important for understanding not only the maternal immune response to the fetus during uncomplicated and preeclamptic pregnancies, but other biological processes as well.

Highlights.

• Higher expression of HA-1 in 1^st as compared to 2^nd and 3^rd trimester syncytiotrophoblast.

• Expression of HA-1 mRNA is increased in trophoblast cells treated with cobalt chloride.

• HA-1 mRNA expression is increased in term trophoblast cells cultured in low oxygen (2%).

• Increased expression of HA-1 mRNA and protein in preeclamptic as compared to control placentas.