Glucocorticoids Enhance CD163 Expression in Placental Hofbauer Cells

Abstract

Periplacental levels of glucocorticoid (GC) peak at parturition, and synthetic GC is administered to women at risk for preterm delivery. However, little is known concerning cell-type-specific effects of GC in placenta. Hofbauer cells (HBCs) are fetal macrophages that are located adjacent to fetal capillaries in placenta. The goal of the current study was to determine whether GC treatment altered HBC gene expression and function. Western blotting and flow cytometry revealed CD163 and folate receptor-β (FR-β), markers of antiinflammatory M2 macrophages, were specifically expressed by primary cultures of HBCs immunopurified from human term placentas. GC receptor mRNA and protein levels were higher in HBCs compared with placental fibroblasts. Treatment of HBCs with cortisol or dexamethasone (DEX) markedly and specifically enhanced CD163 protein and mRNA levels, whereas expression of FR-β and CD68 were largely unresponsive to GC treatment. DEX treatment also increased hemoglobin uptake by HBCs, evidence of enhanced HBC function. The level of CD163 mRNA, but not FR-β or CD68 mRNA, was stimulated in placental explant cultures by DEX treatment, and increased CD163/FR-β and CD163/CD68 mRNA ratios sensitively reflected the response to GC. Maternal GC administration was associated with increased CD163/FR-β and CD163/CD68 mRNA ratios in placentas from women with spontaneous preterm birth. In conclusion, in vitro studies indicated that GC treatment specifically up-regulated CD163 expression in HBCs and enhanced HBC function. In addition, the observed alterations in patterns of expression of macrophage marker genes associated with maternal GC administration suggest that HBCs are in vivo targets of GC action.

Despite the widespread use of antenatal glucocorticoid (GC) in women at risk for preterm delivery (1, 2), the effects of GC administration on the function of specific placental cell types remains largely unelucidated. The original study by Liggins and Howie (3) showed that antepartum betamethasone (βM) prevented respiratory distress syndrome in premature infants. From later meta-analysis, it is clear that maternal administration of βM or dexamethasone (DEX) significantly reduced the risk of respiratory distress syndrome and intraventricular hemorrhage associated with preterm delivery (4). Synthetic GC is used clinically to obviate significant metabolism and inactivation by 11β-hydroxysteroid dehydrogenase type 2, which oxidizes cortisol to receptor-inactive cortisone (5, 6). Midgestation maternal cortisol concentrations (∼300 nm) (7) are 10 times that found in fetal plasma, suggesting that the placenta actively metabolizes physiological GC (8). Studies have localized 11β-hydroxysteroid dehydrogenase type 2 to syncytiotrophoblast (SCT), the outer layer of the placental villus, which is in direct contact with maternal blood (9, 10).

Previous reports from our laboratory and others have documented that GC is a key regulator of placental gene expression (11–17), and DEX and βM are approximately 25-fold more potent than cortisol in the regulation of GC-responsive genes (5, 18). Effects of GC are mediated through GC receptor (GR), a transcription factor, which after binding of GC, translocates to the nucleus and binds to its cognate response element (i.e. a GRE) in the promoter region of GC-responsive genes (19). Human GRα and GRβ isoforms arise due to alternative splicing of the 3′ terminus of the GR gene (20). The GRβ isoform was reported to be transcriptionally inactive and inhibited GRα-mediated effects (21). Our previous study indicated that placental fibroblasts, components of the placental mesenchyme (22, 23), expressed high levels of GR, and GC treatment promoted transactivation-associated GR phosphorylation as well as homologous down-regulation of GR in this cell type (14).

Hofbauer cells (HBCs) (fetal macrophages) were first described in placental villi as large (10-30 μm), pleiomorphic, and highly vacuolated cells with a granular cytoplasm (23, 24). They are located in the placental stroma, which is below the SCT and above fetal capillaries. HBCs appear on the d 18 of gestation and are found until term (22). Because HBCs are observed before the appearance of a fetal circulation, they are likely derived from villous stem cells early in pregnancy, whereas later in pregnancy, they are suggested to arise from recruited fetal monocytes (24–26). General functions of tissue macrophages include phagocytosis of apoptotic bodies and cellular debris as well as antigen presentation in response to inflammation and infectious agents (27), although these functions remain largely unexamined for HBCs. Recent studies suggest that HBCs play an important role in early placental development by enhancing angiogenesis, vasculogenesis, and development of the placental mesenchyme (28–31). HBCs, like other tissue macrophages, can be classified as M1 (proinflammatory), characterized by increased expression of IL-1, TNF-α, CD11b, and CD40, or M2 (antiinflammatory/proangiogenic), characterized by increased expression of IL-10, TGF-β, CD163 and folate receptor (FR)-β (32, 33). Results from immunohistochemical analysis suggests that HBCs are M2 macrophages (34), which supports their putative role in early placental angiogenesis and development. Increased numbers of HBCs have been noted in placentas from pregnancies with villitis of unknown etiology (25, 26), a relatively rare noninfectious entity manifesting a placental inflammatory response akin to allograft rejection or graft-vs.-host disease (35). Our group observed focal increases in the number of HBCs in placentas from pregnancies with chorioamnionitis (36), which is characterized by microbial-driven inflammation at the maternal/fetal interface (37).

CD163 and FR-β, in addition to being known as markers of the M2 macrophages, play important functional roles in scavenging of hemoglobin (Hb) and folate transport, respectively (38, 39). CD163 expression has been shown to be GC responsive in blood-derived monocytes and in macrophages derived from the in vitro differentiation of this cell type (40–42). In the current study, we tested the hypothesis that GCs specifically regulate CD163 expression in HBCs. The rationale for this study is that our knowledge of CD163 regulation in tissue macrophages is limited, and no specific studies in placenta or HBCs have been reported. In addition, CD163 plays a key role in macrophage function, and investigation of GC effects in HBCs is warranted in light of the observation of increasing levels of cortisol across gestation and at parturition and the use of synthetic GCs in women at risk for preterm delivery (1, 2, 7). We employed cultures of HBCs isolated from term placentas to test the effects of GC on CD163 expression and HBC function. The effect of GC on expression of CD163 was also examined in placental tissue using explant cultures. In addition, we will test whether administration of GC to women at risk for preterm delivery was associated with alterations in placental expression of CD163 and other HBC markers.

Materials and Methods

Placental tissue collection and patient groups

Placentas obtained from uncomplicated cesarean delivery at term (n = 12) were used to isolate placental cells and tissue explant cultures. We also studied placental samples from women with spontaneous preterm birth (sPTB) due to spontaneous preterm labor and/or preterm premature rupture of membranes without evidence of clinical or histological chorioamnionitis (n = 19). Clinical chorioamnionitis was diagnosed by the presence of maternal fever (more than 37.8 C), uterine tenderness, foul-smelling amniotic fluid, or visualization of pus at the time of the speculum examination, and maternal (>100 beats/min) or fetal (>160 beats/min) tachycardia (43, 44). Histological examination of placentas was performed by a pathologist unaware of the clinical presentation or outcome. From each placenta, sections of chorionic plate, fetal membranes, and umbilical cord were examined for inflammation as previously described (45). GC was administered to 10 patients as two 12-mg injections of βM 24 h apart, which is in compliance with previous clinical recommendations (1, 2). Nine patients did not receive GC treatment due to emergent preterm delivery. Placentas were collected under protocols approved by the Human Investigation Committee of Yale University. Written informed consent was obtained from all participants before enrollment. Gestational age was established based on menstrual date confirmed by sonographic examination before 20 wk gestation. Characteristics of the study population are shown in Table 1.

Table 1.

Characteristics of the subjects who provided placental samples for this study

Variables	sPTB (−GC), n = 9	sPTB (+GC), n = 10	P value
Age (yr)a	30 (24-34)	23 (20-27)	0.141
Nulliparityb	4 (44)	7 (70)	0.37
Non-Caucasian raceb	4 (44)	7 (70)	0.37
Gestational age (wk)a	29 (22-34)	30 (28-33)	0.683
PPROMb	4 (44)	5 (50)	1
Birth weight (g)a	1280 (435-2062)	1465 (1115-2028)	0.391
Apgar 1 mina	6 (1-8)	6 (4-8)	0.51
Apgar 5 mina	8 (1-9)	9 (7-9)	0.436
Cesarean deliveryb	2 (22)	3 (30)	1

PPROM, Preterm premature rupture of the membranes.

^aData are presented as median (interquartile range) and analyzed by Mann-Whitney U test.

^bData presented as n (%) and Fisher's exact test.

Placentas were brought to the laboratory within 15 min of delivery, and a full thickness specimen was obtained from a central cotyledon (inner third of the placenta) noted to be free of macroscopic pathology (i.e. fibrin deposition or infarcts). The decidua basalis layer was then dissected and discarded, and approximately 1 g of underlying villous tissue was collected, frozen in liquid nitrogen, and maintained at −80 C for use in quantitative PCR (qPCR) studies. Using this system of sampling, we were able to first confirm patterns of placental genes known to be dysregulated in preeclampsia and then to identify novel dysregulated genes including ceruloplasmin, an iron-transport protein located in the syncytium (46).

Cell isolation and culture

Isolation of HBCs, cytotrophoblasts (CTs), and fibroblasts (FIBs) from the same placenta was carried out as we have recently described (47). In this protocol, villous tissue is initially digested with trypsin/deoxyribonuclease I digestion, and then the digestate is centrifuged on a discontinuous Percoll gradient (50/45/35/30%). Cells migrating to the 35/45% Percoll interface were recovered by centrifugation and were immunopurified by negative selection by simultaneous treatment with mouse antihuman CD9 antibody, mouse antihuman CD45 antibody, and then goat antimouse IgG -onjugated DynaBeads, yielding CTs with more than 98% purity based on immunocytochemical and flow cytometric analysis (47). Syncytiotrophoblasts were obtained by spontaneous differentiation of CTs after 72 h of culture (14, 47).

HBCs were isolated from trypsin-treated tissue fragments that were further digested with collagenase A and were loaded onto a discontinuous Percoll gradient (40/20%) and then centrifuged. Cells from the 40/20% Percoll interface were pelleted, resuspended, and then loaded onto a second discontinuous Percoll gradient (35/30/25/20%). Cells from 20/25% to 30/35% interfaces were combined and were immunopurified by negative selection using sequential treatment with anti-epidermal growth factor receptor and then anti-CD10 antibodies conjugated to magnetic beads. Cells were plated, and after 1 h, floating and weakly adherent cells were removed and discarded. This procedure yielded HBCs with purity of at least 98% based on flow cytometry and morphological criteria (47).

Cultures of FIBs were obtained from cells attached to magnetic beads containing CD10 antibody from HBC preparations. The bead-cell mixtures were washed and cultured using a 1:1 mixture of DMEM/F12 with 10% fetal bovine serum (FBS) and 1% antibiotic-antimycotic (i.e. FBS medium). Fresh FBS medium was added every 2-3 d until confluency was reached after approximately 2-3 wk. After trypsinization of first-passage cells, magnetic beads with attached cells, comprising approximately 10% of the cell population, were removed with a magnet and discarded. FIBs between passage three and eight were used in experiments.

For studies of GC effects on gene expression, HBCs were plated at a density of 3 × 10⁵ cells/cm² in RPMI 1640 containing 25 mm HEPES, 5% FBS, and 1% antibiotic-antimycotic in 12-well dishes for Western blot analysis, and six-well dishes for qPCR studies. After 1 h, HBCs were washed twice with PBS and were then incubated overnight in DMEM/F12 containing 10% FBS treated with dextran-coated charcoal. After 24 h, cells were stimulated with steroids and RU486 (catalog item M8046; Sigma Chemical Co., St. Louis, MO) in serum-free medium (SFM) consisting of DMEM/F12 supplemented with 50 μg/ml ascorbic acid and ITS+ premix (BD Biosciences, San Jose, CA) (yielding a final concentration of insulin of 6.25 μg/ml; transferrin, 6.25 μg/ml; selenous acid, 6.25 ng/ml; BSA, 1.25 mg/ml; and linoleic acid, 5.35 μg/ml).

For placental explant cultures, approximately 60 mg of minced placental tissue was rinsed in PBS and placed in a 24-well plate fitted with Transwell 0.8-μm permeable inserts (catalog item 3422; Costar, Corning Life Sciences, Pittston, PA). Explants were maintained for 18-21 h in medium containing DMEM/F12 with 10% charcoal-stripped stripped serum with and without GC as indicated. Explants were then snap-frozen and stored at −80 C before RNA extraction and qPCR.

Western blotting

Protein was extracted from cells using ProteoJet cell lysis buffer (Fermentas Life Sciences, Hanover, MD) supplemented with Complete protease inhibitor cocktail (Roche Diagnostics, Mannheim, Germany) and quantitated by DC Protein Assay (Bio-Rad Laboratories, Hercules, CA). Protein was run on a Tris-glycine 4-15% SDS-PAGE gel (Bio-Rad) and then transferred to a nitrocellulose membrane. After transfer, the membrane was incubated for 1 h at room temperature in Odyssey blocking buffer and then overnight at 4 C with primary antibody in PBS containing 0.1% Tween with 5% powdered milk. Mouse anti-CD163 (catalog item MCA1853; AbD Serotec, Raleigh NC), sheep anti-FR-β (catalog item AF5697; R&D Systems, Minneapolis, MN), rabbit anti-GR, which detects GR-α and GR-β (catalog item sc-8992; Santa Cruz Biotechnology, Santa Cruz, CA), and mouse anti-heat-shock protein 90 (housekeeping protein, catalog item MAB3286; R&D Systems) primary antibodies were all used at 1:2000 dilution. Alexa Fluor680-conjugated donkey antimouse (catalog item A10038;Molecular Probes, Invitrogen, Carlsbad, CA), IR-Dye800CW-conjugated donkey antisheep (catalog item 613731168; Rockland, Gilbertsville, PA), and donkey antirabbit (catalog item 611731127; Rockland) secondary antibodies were all used at a 1:10,000 dilution. Target and housekeeping proteins were detected simultaneously on the same membrane. The membrane was washed in PBS containing 0.1% Tween, and secondary antibodies were added for 1 h at room temperature. After washing, the red signal from the 680-nm fluorophore and the green one from the 800-nm fluorophore were visualized and quantitated with an Odyssey infrared imager (LI-COR, Lincoln, NE).

Flow cytometry

CTs were fixed before plating by incubation with 4% paraformaldehyde/PBS for 10 min at room temperature and were stored in PBS at 4 C. For HBCs, after 1 h of culture, cells were detached from the substratum using 5 mm EDTA and 4 mg/ml lidocaine-HCl in Hanks' balanced salt solution (48, 49) and fixed and stored as above. FIBs were analyzed after trypsinization of passage-three cultures. To detect surface antigen, 2 × 10⁵ paraformaldehyde-fixed cells were washed once using staining buffer (0.5% BSA/PBS). For analysis of intracellular and surface antigen (i.e. total) expression, staining buffer containing 0.1% saponin was used to permeabilize cells. Blocking of Fcγ receptor was then carried out by incubating fixed cells with human IgG (100 μg/ml) in staining buffer for 15 min at room temperature. Allophycocyanin-conjugated mouse anti-CD163 (catalog item 333609; BioLegend, San Diego, CA), unconjugated sheep anti-FR-β (see above), allophycocyanin-conjugated mouse anti-CD90, (fibroblast marker, catalog item 559869; BD Biosciences), and unconjugated mouse anti-cytokeratin-7 (epithelial cell/CT marker, catalog item M7018; Dako) primary antibodies and isotype-matched controls were then added and incubated for 45 min at 4 C as described (47). Cells were then washed twice with staining buffer. For conjugated antibodies, cells were then resuspended in 300 μl PBS for analysis. For unconjugated antibodies, cells were resuspended in staining buffer and incubated with goat antimouse antibody-DyLight649 (catalog item 115496146; Jackson ImmunoResearch, West Grove, PA) or donkey antisheep antibody-fluorescein isothiocyanate (catalog item 713096147; Jackson ImmunoResearch) secondary antibody conjugates for 45 min at 4 C. Cells then were washed and resuspended in 300 μl PBS for analysis (47). The forward-scatter threshold was set to exclude cell debris, and then 10,000 events were collected using FACSCalibur and CellQuest software (BD Biosciences). The results were analyzed using FlowJo software (Tree Star, Ashland, OR). The percentage of positive cells was based on comparison with the isotype-matched control antibody, for which gating was set at 1%.

Quantitative PCR

RNA was extracted from cells and placental tissue using Trizol reagent (Life Technologies, Grand Island, NY). cDNA was synthesized from 3 μg RNA using oligo-deoxythymidine 12-18 (catalog item 18418-012; Invitrogen) and SuperScript II reverse transcriptase (catalog item 18064-014; Invitrogen) in a 20-μl reaction volume according to the manufacturer's instructions. Levels of specific cDNAs were determined with an ABI 7500 RealTime PCR system (Applied Biosystems, Foster City, CA) using TaqMan gene expression assays for GR-α (catalog item 01005211), GR-β (catalog item 00354508), FR-β (catalog item 00265255), CD163 (catalog item 01016661), CD68 (catalog item 00154355), and 18S RNA (catalog item 99999901). qPCRs were performed in duplicate in a 20-μl volume of TaqMan Universal PCR Master Mix containing 1 μl reverse transcription cDNA and 1 μl of assay primer-probe mix. Mean cycle threshold (CT) values were analyzed using the 7500 System SDS software version 1.4. Gene expression was normalized to the housekeeping gene 18S using the formula 2^−ΔΔCT. Results are expressed as relative expression compared with an endogenous control in each experiment as we have previously described (13).

Determination of Hb uptake by HBCs

Uptake of Hb was carried out in the presence of haptoglobin (Hp) as previously described (50). Human HbA0 and Hp phenotype 2-2 were obtained from Sigma (catalog items H0267 and H9762, respectively). Human Hb was labeled with an Alexa Fluor 488 protein labeling kit (Life Technologies Corp./Molecular Probes, Grand Island, NY). Hb:haptoglobin (Hp) complexes were generated by incubating 3, 10, or 30 μg each of labeled Hb and unlabeled Hp for 10 min in SFM. Hb:Hp complexes were centrifuged at 161,000 × g at 4 C for 5 min to pellet particles before their addition to cells. For experiments, 0.6 × 10⁶ HBCs were plated in a 24-well-plate, incubated overnight in FBS medium, and then maintained for 24 h in SFM with or without 100 nm DEX. HBCs were washed twice with SFM and incubated with Hb:Hp complexes for 2 h in a cell incubator at 37 C. HBCs were then washed three times with Ca²⁺/Mg²⁺-free PBS and were detached using a solution containing Hanks' balanced salt solution/HEPES buffer (pH 7.3) supplemented with 0.05% trypsin, 5 mm EDTA, and 4 mg/ml lidocaine-HCl. Cells were then harvested and washed twice using cold Ca²⁺/Mg²⁺-free PBS containing 0.5% BSA. Fluorescence was immediately determined by FACS using a Stratedigm S1000EX flow cytometer (Stratedigm Corp., San Jose, CA) at the Yale Cell Sorter Core Facility.

Statistical analysis

Normally distributed results were analyzed by ANOVA and Student's t test and are expressed as a mean + se. Results that were not normally distributed were analyzed by Mann-Whitney Rank sum test and are presented as a median with quartiles. Proportions obtained from classification of patients were compared with Fisher's exact test. SigmaStat software (Jandel Scientific, San Rafael, CA) was used for statistical analyses. P < 0.05 was considered significant in all studies.

Results

Expression of CD163, FR-β, and GR in HBC cultures

Western blotting revealed that cultures of HBCs expressed CD163 and FR-β, but SCTs and FIBs did not (Fig. 1A). Two-color flow cytometry indicated coexpression of CD163 and FR-β in HBCs, no expression of FR-β in FIBs and CTs, and cell-specific expression of CD90 and cytokeratin in FIBs and CTs, respectively (Fig. 1B). qPCR with GRα-specific primers and Western blotting were then used to compare the levels of GR in HBCs and FIBs. We observed that levels of GR mRNA (Fig. 2A) and protein were 6- and 3-fold higher, respectively, in HBCs compared with FIBs. GR protein was detected at a molecular mass of approximately 95 kDA, which is consistent with the molecular mass reported for placental GRα (51) as well as with the results of our previous study that examined the regulation of GR expression in cultures of placental FIBs (14). We did not detect a protein species at a molecular mass of approximately 92 kDA corresponding to GR-β (51), and no appreciable levels of GR-β mRNA were detected in HBCs and FIBs by qPCR (not shown).

Fig. 1.

Assessment of CD163 and FR-β levels in placental cell cultures using Western blotting and flow cytometry. A, Protein was extracted from whole placental tissue or cultures of SCTs, FIBs, and HBCs, and levels of expression of CD163, FR-β, and heat-shock protein 90 (HSP90) were determined by Western blotting. Seventy micrograms of placental tissue protein were used for analysis, whereas 10 μg of protein from each of the three cell cultures was used. B, Dual-color flow-cytometric analysis presented as dot plots was used to compare cell-type-specific staining (black dots) with that of isotype-matched control antibody (light gray dots). In the left panel, virtually all of the black dots were localized to the upper right-hand quadrant, indicative of costaining for CD163 and FR-β proteins in HBCs. Conversely, no costaining was noted for FR-β and CD90 (fibroblast marker) in FIBs (middle panel) or for cytokeratin 7 (CK7, epithelial/trophoblast marker) in CTs (right panel). Representative results for both methodologies are shown in a single experiment representing three identically conducted ones.

Fig. 2.

Levels of GR expression in HBCs and FIBs. Levels of GR mRNA (A) and protein (B) in FIBs and HBCs were determined by qPCR and Western blotting, respectively. Levels of GR mRNA were normalized to 18S RNA expression, and GR protein levels were normalized to that of heat-shock protein 90 (HSP90). Cumulative results from four (A) or three (B) independent experiments using different placentas are expressed as a mean + se. *, P < 0.05 vs. FIBs; **, P < 0.01 vs. FIBs.

Regulation of CD163 levels in HBC cultures by GC

We then determined whether GC treatment affected expression of CD163 and FR-β in cultures of HBCs. We observed that treatment of HBCs with 100 nm cortisol or DEX for 24 h significantly enhanced CD163 protein levels 11- and 13-fold, respectively (Fig. 3, A and B). No effect was seen with estradiol, progesterone, and testosterone treatment. Conversely, relatively small (66%) nonsignificant increases in FR-β protein expression were noted in response to GC treatment (Fig. 3, A and C). Western blotting results also indicated DEX was 10-fold more potent than cortisol, with dose-dependent effects noted between 1 and 10 nm (Fig. 3D). Up-regulation of CD163 levels was observed between 4 and 48 h of DEX treatment (Fig. 3E). qPCR revealed that treatment of HBCs for 24 h with 100 nm cortisol or DEX significantly enhanced CD163 mRNA expression 10- and 15-fold, respectively (Fig. 4A). In addition, we observed that treatment of HBCs for 24 h with cortisol and DEX nonsignificantly enhanced FR-β mRNA levels 2.1- and 2.0-fold, respectively (Fig. 4B), and did not affect levels of CD68 mRNA (Fig. 4C). Treatment with the other steroids had no significant effect on levels of CD163, FR-β, and CD68 mRNA (Fig. 4, A–C). Of note, the presence of RU486 significantly inhibited cortisol- and DEX-mediated increases in levels of CD163 protein and mRNA (Supplemental Fig. 1, published on The Endocrine Society's Journals Online web site at http://endo.endojournals.org), indicating direct GR-dependent induction in these responses.

Fig. 3.

Regulation of CD163 and FR-β expression in HBC cultures by GC. HBCs were maintained for 24 h without [control (Ctrl)] and with 100 nm cortisol (Cort), DEX, estradiol (E2), progesterone (Prog), and testosterone (Test), and expression of CD163, FR-β, and HSP90 was determined by Western blotting. A, An experiment, representing four identically conducting ones using four different placentas; B and C, quantitation of steroid effects on CD163 and FR-β levels in four independent experiments with results expressed as a mean + se; D and E, dose dependence (D) and the time course (E) of DEX and Cort effects on CD163 and FR-β expression in an experiment representing three identically conducted ones. *, P < 0.001 vs. control.

Fig. 4.

Effect of steroid treatment on CD163, FR-β, and CD68 mRNA levels in HBCs. HBCs were incubated for 24 h in the absence [control (Ctrl)] and presence of 100 nm cortisol (Cort), DEX, estradiol (E2), progesterone (Prog), and testosterone (Test), and expression of CD163 (A), FR-β (B), and CD68 (C) mRNA was determined by qPCR and normalized to the level of 18S RNA. Cumulative results from four independent experiments using four different placentas, expressed as relative expression (mean + se), are shown. *, P < 0.001 vs. control.

Regulation of Hb uptake in HBCs by GC treatment

HBCs were maintained with and without 100 nm DEX for 24 h, and uptake of the indicated concentration of fluorescently labeled Hb (complexed to Hp) was assessed by flow cytometry (Fig. 5). We noted that GC treatment promoted a significant 4- to 6-fold increase in Hb uptake at the three concentrations of labeled Hb tested. This result indicated that the GC-mediated up-regulation in CD163 expression in HBCs was accompanied by enhanced Hb scavenging function in this cell type.

Fig. 5.

Regulation of Hb uptake in HBCs by DEX treatment. HBCs were maintained for 24 h without (unfilled solid-line histogram) or with (black filled histogram) 100 nm DEX, and uptake of a complex consisting of 3 μg/ml (A), 10 μg/ml (B), or 30 μg/ml (C) each of labeled Hb and unlabeled Hp was determined by flow cytometry. The dotted-line histogram was obtained from cells not incubated with Hb:Hp complexes. The values for mean fluorescence intensity (MFI) are indicated above the histograms in a representative experiment. D–F, Cumulative MFI results are presented below the corresponding histograms and are expressed as a mean + se from three independent experiments using three different placentas. *, P < 0.01 vs. control (Ctrl).

Regulation of CD163, FR-β, and CD68 expression in placental explant cultures by GC

Similar to what was noted with HBC cultures, DEX treatment of placental explants for 18-21 h promoted an approximate 6-fold increase in CD163 mRNA expression (Fig. 6A), whereas neither levels of FR-β mRNA (Fig. 6B) nor levels of CD68 mRNA (Fig. 6C) were significantly affected by DEX treatment. Previous studies from our group and others demonstrated that HBCs are not equally distributed throughout placental villi (22, 36, 52), indicating that the numbers of HBCs would likely vary in randomly selected placental tissue samples. To address this issue, in the current study, we normalized expression of CD163 mRNA to that of GC-unresponsive HBC mRNAs (i.e. FR-β and CD68) to correct for potential disparities in placental distribution of HBCs. Of note, when mRNA levels in explant cultures were expressed as a ratio of CD163/FR-β mRNA (Fig. 6D) and CD163/CD68 mRNA (Fig. 6E), although the induction by DEX treatment (6- and 9-fold increases, respectively) was similar to the nonnormalized value shown above (Fig. 6A), smaller interquartile ranges and a higher level of statistical significance was obtained.

Fig. 6.

Effect of DEX treatment on CD163, FR-β, and CD68 mRNA levels in placental explant cultures. A–C, Placental explants were maintained for 18-21 h without [control (Ctrl)] and with 100 nm DEX, and levels of CD163 mRNA (A), FR-β mRNA (B), and CD68 mRNA (C) were determined by qPCR; D and E, ratios of CD163 mRNA/FR-β mRNA (D) and CD163/CD68 mRNA (E) were derived from qPCR results. Results are expressed as either a median with quartiles or a mean + se from seven independent experiments. *, P < 0.01 vs. control; **, P < 0.001 vs. control.

Maternal GC administration and placental levels of CD163, FR-β, and CD68

We then examined whether administration of maternal GC was associated with changes in placental expression of CD163, FR-β, and CD68. As shown in Table 1, the two groups of patients with sPTB without (sPTB−GC) and with (sPTB+GC) maternal GC administration did not differ in terms of gestational age at delivery, cesarean delivery percentage, or the seven other variables analyzed. Using qPCR, we observed that levels of placental CD163 mRNA expression were increased 2.4-fold in the sPTB+GC group compared with sPTB−GC group, although this effect did not reach statistical significance (Fig. 7A). In contrast, maternal GC administration was not associated with changes in levels of placental FR-β (Fig. 7B) and CD68 mRNA (Fig. 7C). Of note, maternal GC administration was associated with significant 2.2- and 3.1-fold increases in the ratio of placental CD163/FR-β mRNA (Fig. 7D) and CD163/CD68 mRNA (Fig. 7E), respectively. Taken together, these results show that GCs specifically up-regulate CD163 expression in cultures of HBCs and placental explants, and patterns of HBC marker gene expression in placental tissue are altered in association with maternal GC administration.

Fig. 7.

Levels of macrophage marker expression in placentas from preterm deliveries with and without maternal GC administration. A–C, Levels of expression of placental CD163 mRNA (A), FR-β mRNA (B), and CD68 mRNA (C) were determined by qPCR from preterm deliveries without (−GC, n = 9) and with (+GC, n = 10) maternal GC administration; D and E, ratios of CD163 mRNA/FR-β mRNA (D) and CD163 mRNA/CD68 mRNA (E) were derived from qPCR results. Results are expressed as a median with quartiles. *, P < 0.05 vs. −GC group.

Discussion

Initial studies in this paper using Western blotting and qPCR demonstrated that HBCs expressed CD163 and FR-β, markers for M2 macrophages. This pattern of expression was cell-type specific, because SCTs, CTs, and FIBs, other major cell types found in the placental villus, did not express these proteins. This result supports the previous characterization of HBCs as antiinflammatory, proangiogenic M2 macrophages, which was based on immunohistochemical analysis (34). FR-β is a marker of M2 macrophages in breast and colon tumors, where they are suggested to support tumor angiogenesis and growth and as such are a predictor of poor clinical outcome (33). In addition to being markers of M2 macrophages, CD163 and FR-β are important regulators of macrophage function (33, 38, 39, 53). CD163, a membrane protein, scavenges Hb:Hp complexes by endocytosis, thereby reducing heme-mediated oxidative damage (39, 53). Its expression is enhanced by GC treatment in macrophages derived in vitro from circulating monocytes (40–42). No information is available concerning modulation of CD163 expression in placenta or HBCs by GC. Two major proteins mediate folate uptake into cells: FR and the reduced folate carrier (38). FR-β is specifically expressed by monocytes and macrophages, and based on its extremely high affinity for folate, it plays a more dominant role when circulating folate concentrations are low (38).

In the current study, we investigated the effect of steroid treatment on the expression of CD163, FR-β, and CD68, a pan-macrophage marker. Of the several steroids tested, only cortisol and DEX significantly enhanced CD163 mRNA and protein expression in cultures of HBCs. DEX was found to be 10-fold more potent than cortisol, which supports our previous results that examined fetal fibronectin regulation in CT cultures (18), and likely reflects the relative affinities of GR for these steroids in HBCs (54). Concentrations of DEX as low as 1 nm enhanced levels of CD163 in HBCs, and higher concentrations of DEX increased CD163 expression after only 4 and 8 h of treatment. This suggests that its expression is highly sensitive to regulation by this synthetic GC. Conversely, expression of FR-β and CD68 were not significantly altered by GC treatment. These results indicated that GC treatment specifically modulated CD163 expression in HBCs. Our studies also indicated that treatment of HBCs with DEX increased uptake of Hb by HBCs. In our study, the uptake of Hb:Hp complexes was assessed to mimic the process of Hb uptake in vivo (39, 55). Because CD163 is a major scavenger of Hb (39, 53), this suggests that endogenous or administered GC may reduce heme-mediated placental damage. Levels of fetal Hb were observed to be elevated in both maternal plasma and placenta in pregnancies with preeclampsia (56, 57), suggesting that placental CD163 function may be overwhelmed in these patients.

The effect of GC treatment on macrophage gene expression was then examined in cultures of placental explants, a model in which the three-dimensional structure of the placental villus is maintained (58). Similar to results obtained in HBC cultures, DEX treatment was found to markedly enhance expression of CD163 mRNA in explant cultures, whereas it had no effect on levels of FR-β mRNA and CD68 mRNA. Although DEX treatment promoted a statistically significant 7-fold induction in CD163 mRNA expression in explant cultures, a large inter-experimental variability was noted as reflected by large interquartile ranges. The observation of greater variability in GC effects on CD163 expression in placental tissue compared with that noted in HBC cultures suggested nonuniformity of HBC levels and CD163 expression in tissue samples. Our group and others have reported that HBCs are not equally distributed throughout placental tissue (22, 36, 52), indicating that HBC number would likely vary in randomly selected placental tissue samples. Unlike levels of CD163 mRNA, levels of FR-β and CD68 mRNA were not found to be significantly affected by GC treatment. Therefore, we hoped that by normalization of CD163 mRNA levels to the expression of these two GC-unresponsive genes, we would in effect provide a within-HBC control and reduce interplacental variability. This turned out to be the case because the use of CD163/FR-β mRNA and CD163/CD68 mRNA ratios reduced interplacental variability in explant culture experiments.

In the current study, the expression of macrophage markers was examined in preterm placentas delivered from pregnancies with and without maternal βM treatment. Similar to results obtained in explant studies, large interplacental variability in CD163 expression was noted in the two groups of delivered placentas. The approximate 2- to 3-fold increase in placental levels of CD163 noted in the GC-treated group reached statistical significance only when normalized to levels of FR-β mRNA and CD68 mRNA. In addition, the magnitude of enhancement of placental CD163 expression that was associated with maternal GC administration was modest compared with GC effects in cultures of HBCs and placental explants. Contributing factors likely include the variable period of time from maternal GC administration to delivery, which has been noted to influence the oxidative state of placentas in a sex-specific manner (59). Furthermore, the parturition-associated rise in cortisol levels in maternal and cord blood (60, 61) may enhance basal expression of CD163 in placenta, thereby blunting the response to subsequent βM administration. We were also limited by the relatively small size of the study. This precluded examination of the effect of duration of antenatal GC exposure and time to delivery after GC administration as well as comparison of the effects of βM and DEX treatment. Because all women at risk for preterm delivery are administered GC in compliance with recommendations by the National Institutes of Health and American College of Obstetricians and Gynecologists, it is difficult to obtain placental samples from patients who were not treated with GC before delivering preterm. The nine patients in the untreated group were from pregnancies in which GC treatment was not feasible due to emergent preterm delivery.

In conclusion, our in vitro studies using primary cultures of HBCs and placental explants indicate that GCs specifically up-regulate CD163 expression and HBC function. In addition, the observed changes in the patterns of expression of macrophage marker genes associated with maternal GC administration suggest that HBCs are targets of GC action in vivo.