Abstract
Free full text
Chronic Chorioamnionitis Displays Distinct Alterations of the Amniotic Fluid Proteome
Abstract
Acute chorioamnionitis of infectious origin and chronic chorioamnionitis of immunologic origin are two major placental lesions of spontaneous preterm birth with elevated amniotic fluid interleukin-6 and CXCL10 concentrations, respectively. The changes in the amniotic fluid proteome associated with intra-amniotic infection and acute chorioamnionitis are well-defined, yet alterations unique to chronic chorioamnionitis remain to be elucidated. This study was conducted to determine those amniotic fluid proteins changing specifically in the presence of chronic chorioamnionitis. Amniotic fluid obtained from acute chorioamnionitis, chronic chorioamnionitis, and gestational age-matched controls were analyzed by 2-dimensional difference in gel electrophoresis and MALDI-TOF analyses. The type of histologic inflammation was used to define each condition in preterm labor cases (n=125) and term not in labor cases (n=22); and the amniotic fluid concentrations of interleukin-6, CXCL8, CXCL10, and prostaglandin F2α were also measured by specific immunoassays. Among preterm labor cases, thirty-one differentially expressed proteins were identified in chronic chorioamnionitis cases, as compared to both acute chorioamnionitis and control cases. Importantly, glycodelin-A, which maintains maternal tolerance against an allogeneic fetus, was decreased in chronic chorioamnionitis, while haptoglobin was increased. We report the amniotic fluid proteome of chronic chorioamnionitis for the first time, and the findings herein strongly suggest that there is a pathophysiological association between the changes of immunomodulatory proteins in the amniotic fluid and chronic chorioamnionitis, a histologic manifestation of maternal anti-fetal allograft rejection.
Introduction
Human parturition involves the activation of pro-inflammatory cascades in gestational tissues, such as the chorioamniotic membranes, in cases of spontaneous labor at term and preterm birth due to intra-amniotic infection. Not surprisingly, in both conditions there is an increased expression of prostaglandin-endoperoxide synthase 2 (PTGS2) in the chorioamniotic membranes [1–3], and the amniotic fluid prostaglandin concentrations are increased [4–7]. Acute chorioamnionitis following microbial invasion of the amniotic cavity (MIAC) is a major pathological lesion found in spontaneous preterm births (preterm labor and preterm prelabor rupture of membranes) [8–10]. However, our recent analysis demonstrated that chronic chorioamnionitis is more common in spontaneous preterm birth cases, found in 34% of preterm labor with intact membranes cases and 39% of preterm prelabor rupture of membranes cases. This lesion was also observed in 23% of term preeclampsia and 16% of preterm preeclampsia cases suggesting its link with the immunopathology of preeclampsia and indicated preterm birth as well [11,12].
Chronic chorioamnionitis, defined as the amniotropic infiltration of maternal T cells, is considered a histologic manifestation of maternal anti-fetal allograft rejection and graft-versus-host disease in the placenta [11,13,14]. Therefore, a major proportion of preterm births seems to be due to either intra-amniotic infection or allograft rejection. While amniotic fluid interleukin-6 (IL-6) is a sensitive marker of intra-amniotic infection and acute chorioamnionitis [15–17], chronic chorioamnionitis is associated with elevated intra-amniotic T cell chemokine CXC-motif ligand-10 (CXCL10) concentration and increased CXCL9, CXCL10, and CXCL11 mRNA expressions in the chorioamniotic membranes [11]. Intra-amniotic infection- and acute chorioamnionitis-associated changes in the amniotic fluid proteome have been elegantly addressed in previous studies [18–20], yet those associated with chronic chorioamnionitis remain to be elucidated. The assessment of changes in the amniotic fluid proteome could unveil key alterations in the intra-amniotic environment leading to the development of chronic chorioamnionitis and preterm birth. This study was conducted to determine specific changes in the abundance of amniotic fluid proteins in the presence of chronic chorioamnionitis.
Materials and methods
Patient materials
This cross-sectional study included 125 women with a singleton pregnancy who presented with preterm labor with intact membranes and 22 women with normal pregnancy at term not in labor. The study was approved by the Institutional Review Boards of participating institutions, and all participants provided written informed consent for the collection of biological materials and clinical data. Amniotic fluid samples were obtained by transabdominal amniocentesis performed to assess the microbial status of the amniotic cavity or fetal lung maturity. Histologic diagnosis of placental lesions was done by examination of hematoxylin and eosin stained sections of the chorioamniotic membranes (n=1), placental disc (n=2), and umbilical cord (n=1) in each case. The diagnosis of acute chorioamnionitis was made based on the criteria previously described [21]. The diagnosis of chronic chorioamnionitis was made when lymphocytic infiltration into the chorionic trophoblast layer or chorioamniotic connective tissue was observed. The extent of chronic chorioamnionitis was graded 1 when there were more than two foci or patchy inflammation, and 2 when diffuse inflammation was present. The stage of inflammation was scored as 1 if amniotropic lymphocytic infiltration was limited to the chorionic trophoblast layer, and 2 if lymphocytic infiltration extended into the chorioamniotic connective tissue [11].
Enzyme-linked immunosorbent assay (ELISA)
The amniotic fluid concentrations of IL-6 (R&D Systems, Minneapolis, MN, USA), CXCL8 (R&D Systems), CXCL10 (R&D Systems), PGF2α (Assay Designs, Inc., Ann Arbor, MI, USA), and glycodelin-A (DRG Diagnostics, Marburg, Germany) were measured by specific immunoassays according to the manufacturers’ instructions.
2-Dimensional Difference in Gel Electrophoresis (2D-DIGE)
Three gestational age-matched amniotic fluid samples of each histologic group [no inflammation-control (gestational age: 24.9, 27.4, 33.4 weeks), acute chorioamnionitis (gestational age: 25.3, 27.4, 33.4 weeks), chronic chorioamnionitis (gestational age: 25.1, 28.9, 33.9 weeks)] were selected. These three samples were then pooled to represent each group. Methanol was added to the pooled amniotic fluid samples (volume ratio 9:1), and the precipitated proteins were resolubilized in lysis buffer [30 mM Tris-HCl (pH 8.8), 7 M urea, 2 M thiourea, 4% CHAPS]. Thirty μg of proteins from each pooled sample were labeled with Cy2, Cy3, and Cy5, respectively. The labeling reaction was stopped by adding 1.0 ul of 10 mM Lysine to each sample, and incubating in the dark on ice for an additional 15 min. The labeled samples were then mixed with 2X sample buffer [8 M urea, 4% CHAPS, 20 mg/ml dithiothreitol (DTT), 2% pharmalytes] and 100 μl destreak rehydration buffer (7 M urea, 2 M thiourea, 4% CHAPS, 20 mg/ml DTT, 1% pharmalytes) to make a total volume of 250 μl. The samples were mixed, spun, and then loaded into a strip holder. After isoelectric focusing (pH 3–10), IPG strips were incubated in equilibration buffer-1 (50 mM Tris-HCl, pH 8.8, 6 M urea, 30% glycerol, 2% SDS, 10 mg/ml DTT) for 15 min with gentle shaking, and then rinsed in equilibration buffer-2 [50 mM Tris-HCl (pH 8.8), 6 M urea, 30% glycerol, 2% SDS, 45 mg/ml DTT] for 10 min with gentle shaking. Following electrophoresis in a 12% SDS-polyacrylamide gel at 15°C, the gel was scanned using Typhoon TRIO (Amersham BioSciences, Piscataway, NJ, USA). Scanned images were then analyzed by Image Quant software version 6.0 (Amersham BioSciences), followed by differential in-gel analysis using DeCyder software version 6.5 (GE Healthcare, Piscataway, NJ, USA), to obtain the fold-changes in protein expression.
Mass Spectrometry
The spots of interest were picked up by Ettan Spot Picker (Amersham BioSciences) and digested in-gel with modified porcine trypsin protease (Trypsin Gold; Promega, Madison, WI, USA). The digested tryptic peptides were desalted by Zip-tip C18 (Millipore, Billerica, MA, USA). Peptides were eluted from the Zip-tip with 0.5 μl of matrix solution (5 mg/ml of α-cyano-4-hydroxycinnamic acid in 50% acetonitrile, 0.1% trifluoroacetic acid, 25 mM ammonium bicarbonate) and spotted on the MALDI plate (model ABI 01-192-6-AB). MALDI-TOF and TOF/TOF mass spectrometry were performed on an ABI 4700 mass spectrometer (Applied Biosystems, Framingham, MA, USA). MALDI-TOF mass spectra were acquired in reflectron-positive ion mode, averaging 4000 laser shots per spectrum. TOF/TOF mass spectrometry fragmentation spectra were acquired for each sample, averaging 4000 laser shots per fragmentation spectrum on each of the 10 most abundant ions present in each sample. Both the resulting peptide mass and the associated fragmentation spectra were submitted to a GPS Explorer workstation equipped with a MASCOT search engine (Matrix Science Ltd., London, United Kingdom) to search the redundant database of the National Center for Biotechnology Information (NCBInr). Searches were performed without constraining protein molecular weight or isoelectric point, with variable carbamidomethylation of cysteine and oxidation of methionine residues, and with one missed cleavage also allowed in the search parameters. Candidates with either a protein score of C.I.% or Ion C.I.% greater than 95 were considered significant.
Immunoblotting
Immunoblotting for haptoglobin was done using 3.5 μl of amniotic fluid. Electrophoresis was done under reducing conditions in 12% SDS-PAGE, and proteins were electro-blotted onto polyvinylidene difluoride membranes. The membranes were incubated with mouse monoclonal anti-haptoglobin-beta antibody (Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA), and then with HRP-conjugated anti-mouse IgG. Chemiluminescence was detected using a ChemiGlow West kit (Alpha Innotech Corporation, San Leandro, CA, USA). β-haptoglobin density was analyzed by FluorChem™ SP densitometry using AlphaEase®FC Software (Version 4.1.0; Alpha Innotech Corporation).
Immunohistochemistry
Formalin-fixed, paraffin-embedded 5-μm-thick tissue sections of the chorioamniotic membranes obtained from preterm labor cases without inflammation (n=5), with acute chorioamnionitis (n=5), and chronic chorioamnionitis (n=5) were used for immunostaining using a Ventana Discovery automatic staining system (Ventana Medical Systems, Tucson, AZ, USA). Immunostaining was performed using a mouse monoclonal anti-CD8 antibody (1:100 diluted, Dako, Carpinteria, CA, USA), a rabbit monoclonal anti-pregnancy-associated endometrial alpha-2 globulin (PAEP, glycodelin-A) antibody (1:100 diluted, Life Span Biosciences, Inc., Seattle, WA, USA), and a mouse monoclonal anti-haptoglobin beta antibody (1:100 diluted, Santa Cruz Biotechnology Inc.). Chromogen reaction was done using the DiscoveryR DAB Map Kit (Ventana Medical Systems).
Statistical analysis
Normality of the data distribution was tested using the Kolmogorov-Smirnov test. Comparisons among the study groups were performed by the unpaired t-test and post hoc analysis or Kruskall-Wallis test and Mann-Whitney U-test for continuous variables, as well as by the χ2-test and Fisher’s exact test for categorical variables as appropriate. Spearman rank correlation was used to examine the relationship of PGF2α concentration with IL-6 and with CXCL10 concentrations in amniotic fluid and between glycodelin-A concentration and gestational age. A p value of < 0.05 was considered significant.
Results
Amniotic fluid IL-6, CXCL8, CXCL10, and PGF2α
The demographics and clinical characteristics of the study population are shown in Table 1, and histological and immunohistochemical features of each group are shown in Figures 1A–1F. Amniotic fluid cultures for aerobes, anaerobes and Mycoplasmas were available in all preterm labor cases. Among 71 cases without inflammation (control), one case had a positive culture. Amniotic fluid cultures were positive in four of the 14 cases of acute chorioamnionitis (28.6%). Cultures were negative in all chronic chorioamnionitis cases (n=40). To further characterize acute chorioamnionitis and chronic chorioamnionitis cases, amniotic fluid IL-6, CXCL8, CXCL10, and PGF2α concentrations were measured by immunoassay.
Pathologic characteristics of acute and chronic chorioamnionitis. (A, B) Histological findings of a preterm labor case without inflammation, showing no evidence of extrinsic infiltrates (A). Few CD8+ T cells (arrows) are found in the decidua (B). (C, D) A case of acute chorioamnionitis showing characteristic amniotropic neutrophil infiltration into the chorionic connective tissue layer (C). CD8+ T cells (arrows) are restricted to the decidual layer (D). (E, F) A case of chronic chorioamnionitis with infiltration of lymphohistiocytes into the chorionic trophoblast layer (E), which are predominantly of CD8+ T cells (arrows) (F). Original magnification (A–F), X200.
Table 1
Demographic and clinical characteristics of the study population
| ACA−/CCA− (n=71) | ACA+/CCA− (n=14) | ACA−/CCA+ (n=40) | |
|---|---|---|---|
| Maternal age (years) | 23.5 (±5.7) | 24.3 (±8.9) | 24.7 (±6.6) |
| Nulliparous | 31 (43.7%) | 8 (57.1%) | 15 (37.5%) |
| GA at amniotic fluid sampling (weeks) | 31.0 (±3.0) | 29.8 (±3.5) | 31.4 (±2.8) |
| GA at delivery (weeks)† | 36.5 (±2.9) | 32.9 (±5.5)* | 35.4 (±3.7) |
| Interval amniocentesis-delivery (days) † | 38.2 (±28.5) | 22.0 (±24.9) | 28.3 (±23.6) |
| Birth weight (grams) † | 2753.7 (±712.7) | 2155.0 (±1125.2)** | 2548.3 (±844.1) |
| Birth weight <10th percentile | 11 (15.5%) | 1 (7.1%) | 7 (17.5%) |
Data are presented as mean (±SD) or number (percentage)
ACA: acute chorioamnionitis, CCA: chronic chorioamnionitis, GA: gestational age.
The median IL-6 concentration was higher in acute chorioamnionitis cases (median: 7.2 ng/ml, range: 0.4–388.9 ng/ml) than in chronic chorioamnionitis cases (median 0.6 ng/ml, range 0.1–15.2 ng/ml; p=0.002) or in control cases (median: 0.6 ng/ml, range: 0.0–14.8 ng/ml; p=0.001) (Figure 2A). Among acute chorioamnionitis cases, IL-6 concentration was significantly higher in those with positive amniotic fluid culture compared to those with negative culture (median 217.3, range 18.5–388.9 ng/ml vs. median 1.0, range 0.4–241.1 ng/ml; p=0.014). Using a concentration cut-off of 12.4 ng/ml identified with a receiver-operating characteristic (ROC) curve (area under the curve (AUC): 0.979; 95% CI: 0.931–1.027; p<0.001), IL-6 had sensitivity, specificity, positive predictive value, and negative predictive value of 100%, 87.5%, 66.7% and 100%, respectively, for acute chorioamnionitis in patients who delivered within seven days of amniocentesis.
Differences in amniotic fluid IL-6, CXCL8, CXCL10, and PGF2α concentrations according to the type of inflammation in preterm labor cases (n=125). (A, B) Amniotic fluid IL-6 and CXCL8 concentrations are elevated in cases of acute but not chronic chorioamnionitis. (C) Amniotic fluid CXCL10 concentration is elevated in cases of chronic but not acute chorioamnionitis. (D) Amniotic fluid PGF2α concentration in each group, showing an elevation in acute chorioamnionitis cases compared to other groups. ACA: acute chorioamnionitis, CCA: chronic chorioamnionitis.
The median concentration of CXCL8 was significantly higher in acute chorioamnionitis cases (median 3.4 ng/ml, range: 0.1–467.0 ng/ml) than in chronic chorioamnionitis cases (median 0.5 ng/ml, range: 0.1–15.7 ng/ml; p=0.048) or in control cases (median 0.4 ng/ml, range: 0.1–14.8 ng/ml; p=0.019) (Figure 2B).
The median amniotic fluid CXCL10 concentration was higher in chronic chorioamnionitis cases (median: 2.7 ng/ml, range: 0.0–20.8 ng/ml) than in the acute chorioamnionitis cases (median: 1.1 ng/ml, range: 0.4–4.5 ng/ml; p=0.02) and the control cases (median: 1.1 ng/ml, range: 0.0–5.8 ng/ml; p<0.001) (Figure 2C). Using a concentration cut-off of 1.1 ng/ml identified with a ROC curve (AUC: 0.769; 95% CI 0.643–0.895; p=0.001), CXCL10 had sensitivity, specificity, positive predictive value, and negative predictive value of 85.7%, 61%, 52.9% and 89.3%, respectively, for chronic chorioamnionitis in patients delivering within 28 days of amniocentesis.
Preterm labor due to intra-amniotic infection is characterized by increased prostaglandin concentrations in the amniotic fluid [4–7,22]. To determine whether chronic chorioamnionitis shows comparable changes in the amniotic fluid prostaglandins, PGF2α was also measured. The median amniotic fluid concentration of PGF2α was significantly higher in cases with acute chorioamnionitis (median: 101.2 ng/ml, range: 2.4–2216.4 ng/ml) compared to both chronic chorioamnionitis cases (median: 40.2 ng/ml, range: 6.7–514.6 ng/ml; p=0.02) and control cases (median: 36.1 ng/ml, range: 1.9–1565.8 ng/ml; p=0.01), while no differences were observed between chronic chorioamnionitis and control cases (Figure 2D). Among acute chorioamnionitis cases, PGF2α concentration tended to be higher in cases with positive culture (median 704.8, range 183.4–2216.4 pg/ml vs. median 54.7, range2.4–939.9 pg/ml; p=0.05). Amniotic fluid PGF2α concentration showed a moderate correlation with that of IL-6 (Spearman rho coefficient: 0.384, p<0.001) but not with CXCL10 (Spearman rho coefficient: 0.062, p=0.5).
The amniotic fluid concentrations of IL-6 (median 0.5 ng/ml, range 0.2–1.8 ng/ml; Figure 3A), and CXCL10 (median 0.8 ngml, range 0.1–6.0 ng/ml Figure 3C) of normal pregnant women at term were not different from those of preterm labor cases without histologic inflammation. However, CXCL8 concentration (median 1.0 ng/ml, range 0.3–.3.9 ng/ml; Figure 3B) and PGF2α concentration (median 155.3 pg/ml, range 11.8–1338.5 pg/ml; Figure 3D) of term cases were significantly higher than that of preterm labor cases (p=0.003 and p<0.001 respectively).
Amniotic fluid IL-6 (A), CXCL8 (B), CXCL10 (C), and PGF2α (D) concentrations in preterm labor cases without inflammation (n=71) and term not in labor cases (n=22). Amniotic fluid CXCL8 and PGF2α concentrations are significantly higher in term not in labor cases than in preterm labor cases (p<0.01 and p<0.001, respectively), whereas there is no difference in IL-6 and CXCL10 concentrations. PTL: preterm labor, TNL: term not in labor.
2D-DIGE of Amniotic Fluid
Figure 4A shows the 2-D electrophoresis gel with 71 spots showing differentially expressed proteins amongst the three groups. Differentially expressed spots with more than 2-fold changes in chronic chorioamnionitis compared to acute chorioamnionitis or control cases were analyzed (Figure 4B). Table 2 provides the list of 31 differentially expressed proteins identified from 41 spots in 2D-DIGE analysis of pooled amniotic fluid samples from the three groups (control, acute chorioamnionitis, and chronic chorioamnionitis). Several acute phase proteins such as α-1-antichymotrypsin, α-1-antitrypsin, and haptoglobin were increased in chronic chorioamnionitis. Several plasma proteins were also differentially represented in chronic chorioamnionitis: plasminogen and transferrin were increased, while ceruloplasmin was decreased. Orosomucoid 2 and insulin-like growth factor binding protein 1 precursor were decreased in chronic chorioamnionitis compared to the controls but did not change in acute chorioamnionitis. Interestingly, glycodelin-A was 6.1-fold and 5.5-fold reduced in chronic chorioamnionitis compared to acute chorioamnionitis and controls, respectively.
Identification of differentially abundant proteins in amniotic fluid samples from chronic chorioamnionitis cases. (A) Two-dimensional difference in gel electrophoresis (2D-DIGE) images showing differentially expressed proteins between acute chorioamnionitis cases (Cy3: green) and chronic chorioamnionitis cases (Cy5: red). The spots are labeled by number. (B) Three-dimensional images from DeCyder software analysis for spot 42 marked in the 2D-DIGE image (A), which is increased in chronic chorioamnionitis. The spot was identified as β-haptoglobin by MALDI-TOF-MS/MS. ACA: acute chorioamnionitis, CCA: chronic chorioamnionitis.
Table 2
Amniotic fluid peptides showing significant changes with chronic chorioamnionitis cases compared to acute chorioamnionitis cases or cases without chorioamnionitis
| Spot number | Protein name | Accession number | MW (Da) | Protein PI |
|---|---|---|---|---|
| 4 | Proteoglycan 2 preproprotein | gi|46276889 | 25189.3 | 6.3 |
| 7 | Ceruloplasmin | gi|1620909 | 115398.4 | 5.4 |
| 8 | Albumin preproprotein | gi|4502027 | 69321.5 | 5.9 |
| 10 | Serum albumin precursor | gi|6013427 | 69180.4 | 5.9 |
| 13 | Complement component 6 | gi|38488662 | 104739.1 | 6.4 |
| 15 | Complement factor B | gi|13278732 | 85508.5 | 6.6 |
| 19 | Plasminogen | gi|4505881 | 90510.2 | 7.0 |
| 22 | Transferrin | gi|115394517 | 76909.6 | 7.0 |
| 25 | Alpha-2-HS-glycoprotein | gi|156523970 | 39315.7 | 5.4 |
| 27 | Alpha-1-antichymotrypsin precursor | gi|177933 | 45453.4 | 5.3 |
| 28 | Kininogen 1 variant | gi|62898910 | 47822.6 | 6.3 |
| 29 | Alpha-1 antitrypsin variant | gi|110350939 | 46577.0 | 5.4 |
| 30 | Chain A, the intact and cleaved human antithrombin III complex | gi|999513 | 49008.0 | 6.0 |
| 32 | Chain A, structure of human serum albumin with S-Naproxen and the Ga module | gi|168988718 | 65777.6 | 5.6 |
| 36 | Vitamin D-binding protein/group specific component | gi|455970 | 52916.0 | 5.3 |
| 40 | Chain A, crystal structure of a Serpin:protease complex | gi|11514321 | 37624.4 | 5.3 |
| 42 | Haptoglobin | gi|3337390 | 38209.2 | 6.1 |
| 45 | PRO2619 | gi|11493459 | 56745.2 | 6.0 |
| 47 | Hemopexin, isoform CRA_d | gi|119589127 | 43201.2 | 6.2 |
| 48 | Phosphogluconate dehydrogenase | gi|40068518 | 53105.9 | 6.8 |
| 49 | Enolase 1 variant | gi|62897945 | 47167.3 | 7.0 |
| 52 | Albumin-like | gi|763431 | 52047.8 | 5.7 |
| 55 | Leucine-rich alpha-2-glycoprotein 1 | gi|16418467 | 38154.1 | 6.5 |
| 61 | Progestagen-associated endometrial protein | gi|55958594 | 14792.5 | 5.2 |
| 62 | Chain A, crystal structure of lipid-free human apolipoprotein A-I | gi|90108664 | 28061.5 | 5.3 |
| 63 | Orosomucoid 2 | gi|48145977 | 23599.7 | 5.0 |
| 64 | Chain A, crystal structure of the trigonal form of human plasma retinol-binding protein | gi|157830446 | 20945.1 | 5.3 |
| 65 | Chain E, crystal structure of the transthyretin-retinol binding protein-Fab complex | gi|212374946 | 20189.7 | 5.1 |
| 67 | Insulin-like growth factor binding protein 1 precursor | gi|4504615 | 27885.3 | 5.1 |
| 68 | Chain B, structural basis for the inhibition of insulin-like growth factors by Igf binding proteins | gi|114793741 | 9808.5 | 5.3 |
| 71 | Albumin, isoform CRA_p | gi|119626079 | 22498.5 | 8.2 |
Spot numbers shown in 2D-DIGE image (Figure 2A).
To confirm the results of 2D-DIGE, amniotic fluid glycodelin-A concentration was measured by immunoassay. When all results from the cases were analyzed, there was no difference among the groups (p = 0.2). When examining cases ≤ 32 weeks of gestation (n = 68), however, based on the fact that amniotic fluid glycodelin-A concentration rapidly decreases by the end of the second trimester [23], and that two out of three amniotic fluid samples used for 2D-DIGE were obtained before 29 weeks of gestation, median amniotic fluid glycodelin-A concentration was significantly lower in chronic chorioamnionitis cases (median: 1639.9 ng/ml, range: 45.2–12718.0 ng/ml) compared to control cases (median: 3184.3 ng/ml, range: 28.3–97991.2 ng/ml; p < 0.05) (Figure 5A). Glycodelin-A concentration in acute chorioamnionitis cases (median: 3166.8 ng/ml, range: 278.9–55608.1 ng/ml) also tended to be higher than that found in chronic chorioamnionitis cases (p = 0.091). Such differences were not present in samples collected after 32 weeks (p = 0.9). This seems to be associated with a marked physiological decrease in amniotic fluid glycodelin-A concentration. Indeed, there was a significant negative correlation between glycodelin-A concentration and gestational age at amniocentesis (Spearman rho coefficient = −0.479; p < 0.001). Even when the controls were divided into high and low amniotic fluid glycodelin-A using median concentration as the cut-off value, the gestational age of low glycodelin-A cases (median: 30.4 weeks, range: 26.6 – 32.0 weeks) was significantly higher than that of high glycodelin-A cases (median: 27.4 weeks, range: 22.7 – 31.7 weeks). However, there was no difference in gestational age between control and chronic chorioamnionitis cases, indicating that decreased amniotic fluid glycodelin-A concentration is a feature of chronic chorioamnionitis.
Differential abundance of glycodelin-A and β-haptoglobin in chronic chorioamnionitis cases with gestational ages of 32 weeks or less. (A) Amniotic fluid glycodelin-A concentration is significantly lower in chronic chorioamnionitis cases (n=21) compared to that of control cases (n=37) and acute chorioamnionitis cases (n=10). (B) Immunoblotting results showing increased abundance of β-haptoglobin subunit in chronic chorioamnionitis cases. Amniotic fluid (3.5 μl) was electrophoresed in 12% SDS-PAGE under reducing conditions. P: a pooled amniotic fluid sample as a normalizer in densitometric analysis. (C) The results of densitometric analysis of β-haptoglobin immunoblotting. Haptoglobin is significantly increased in the amniotic fluid of chronic chorioamnionitis cases (n=8) compared to control cases (n=8) and to acute chorioamnionitis cases (n=7). The graph depicts the relative density of each sample using the density of a pooled amniotic fluid sample as a normalizer. ACA: acute chorioamnionitis, CCA: chronic chorioamnionitis.
The differences in the amniotic fluid concentration of haptoglobin observed by 2D-DIGE were further assessed by immunoblotting in a subgroup of samples associated with isolated acute chorioamnionitis (n = 11), chronic chorioamnionitis (n = 16), and control cases (n = 16). Densitometric analysis of the immunoblotting results was performed. The abundance of the β-subunit of haptoglobin in the amniotic fluid of cases collected at ≤ 32 weeks of gestation (n = 23) was higher in chronic chorioamnionitis cases (median sample haptoglobin/pooled sample haptoglobin ratio: 2.9, range: 0.4 – 4.1) compared to both acute chorioamnionitis cases (median sample haptoglobin/pooled sample haptoglobin ratio: 0.6, range: 0.4 – 2.3, p<0.05) and control cases (median sample haptoglobin/pooled sample haptoglobin ratio: 0.7, range: 0.3–2.3, p<0.05) (Figure 5B, 5C), while the difference was not observed in cases > 32 weeks of gestation (n = 20), as was the case with glycodelin-A.
Glycodelin-A and Haptoglobin Immunoreactivity in the Chorioamniotic Membranes
For the localization of glycodelin-A and haptoglobin expression, sections of the chorioamniotic membranes were immunostained for glycodelin-A and haptoglobin. Intense cytoplasmic glycodelin-A immunoreactivity was found in the endometrial epithelial cells. On the other hand, decidual stromal cells showed a predominantly membranous staining pattern (Figure 6A and 6B). Haptoglobin immunoreactivity was localized in the cytoplasm of decidual leukocytes, and also in the cytoplasm of myofibroblasts and macrophages in the chorioamniotic connective tissue (Figure 6C and 6D). Neutrophil infiltrates in acute chorioamnionitis were also weakly immunoreactive.
Immunoreactivity of glycodelin-A and haptoglobin in the chorioamniotic membranes. (A) Distinct glycodelin-A immunoreactivity in the endometrial epithelial cytoplasm (arrows) in a preterm labor case without inflammation. Membranous immunoreactivity is far more prominent in the decidual stromal cells. (B) A chronic chorioamnionitis case showing similar immunoreactive patterns in the endometrial epithelial cytoplasm (arrows). (C–F) Coarsely granular haptoglobin immunoreactivity is mainly found in decidual leukocytes and in myofibroblasts (C) and macrophages of the chorionic connective tissue from a preterm labor case without inflammation (E). Similar immunoreactive patterns also found in a chronic chorioamnionitis case. Decidual leukocytes (D) and myofibroblasts and macrophages in the chorionic connective tissue (F). Original magnification (A, B), X200; (C–F), X400.
Discussion
The primary findings of this study are: 1) There are changes in the amniotic fluid proteome unique to chronic chorioamnionitis compared to acute chorioamnionitis or no inflammation cases; and, 2) Acute chorioamnionitis and chronic chorioamnionitis are different in terms of the changes in amniotic fluid IL-6, CXCL8, CXCL10, and PGF2α concentrations. The changes in amniotic fluid chemokine concentrations (CXCL8, CXCL10) were consistent with amniotropic infiltration patterns of maternal/fetal neutrophils and maternal T cells in acute chorioamnionitis and chronic chorioamnionitis, respectively. These observations further support the fact that acute chorioamnionitis and chronic chorioamnionitis are manifestations of different pathological processes, particularly in the mechanisms leading to preterm labor.
Among the changes in amniotic fluid proteome associated with chronic chorioamnionitis, decreased glycodelin-A concentration is particularly intriguing. Glycodelin, also known as progesterone-associated endometrial protein, placental protein 14, and progestagen-dependent endometrial protein, is a dimeric glycoprotein belonging to the lipocalin family [23,24]. At least four isoforms of identical amino acid sequence but with different oligosaccharide structures and biological functions are present: glycodelin-A was isolated from the amniotic fluid [25], glycodelin-F from the follicular fluid [26], glycodelin-C from the extracellular matrix of the cumulus oophorus [27], and glycodelin-S from epithelial cells of the seminal glands [28]. The most abundant isoform in the female reproductive tract is glycodelin-A, which is produced by endometrial and decidual epithelial cells and secreted in the endometrial fluid and amniotic fluid and into the circulation [24,29]. Glycodelin-A has two main functions: contraceptive activity due to the inhibition of the binding of spermatozoa to the zona pellucida [30], and immunosuppression. The immunosuppressive activity is mediated by several different mechanisms and is effective on all predominant immune cell lineages at the maternal-fetal interface: lymphocytes, monocytes, and natural killer cells.
T cells are the most well-defined target of the immunosuppressive action of glycodelin-A. The proliferative response of T lymphocytes to allogeneic stimulation [31] and to phytohemagglutinin [32] is inhibited by glycodelin-A in a dose-dependent manner. This property is dependent on interaction with CD45, a tyrosine phosphatase with regulatory activity on the T cell receptor signaling pathway, which results in an elevation of the T cell receptor activation threshold [33]. While both Th1 and Th2 responses are inhibited by glycodelin-A, the Th1 differentiation pathway is more sensitive to the inhibition of T cell receptor signaling, thus leading an overall shift toward Th2 differentiation [34]. Moreover, glycodelin-A induces apoptosis of activated T cells [35] by an intrinsic pathway [36]. The cell receptor (or one of the receptors) responsible for its apoptotic activity on T cells was shown to be CD7, a galactose-containing surface glycoprotein expressed in activated T cells. CD7 is also recognized by galectin-1 and galectin-3 [37]. Glycodelin-A also inhibits monocyte chemotaxis [38] and induces monocyte apoptosis via the intrinsic mitochondrial stress-induced pathway [39]. Furthermore, glycodelin-A can inhibit the cytotoxic activity of circulating natural killer cells [40], possibly through induction of an apoptotic cascade which, unlike for T cells and monocytes, does not involve the mitochondrial pathway [39].
Because of its abundance at the maternal-fetal interface, it was suggested that glycodelin-A maintains maternal immune tolerance against fetal allo-antigens [39]. Indeed, a large body of evidence supports an important role for glycodelin-A in the maintenance of physiologic pregnancy: 1) women with unexplained infertility have a lower glycodelin-A concentration in uterine flushing [41]; 2) uterine fluid and serum glycodelin-A concentrations are reduced in women with a history of recurrent miscarriage [42,43]; 3) glycodelin-A expression is reduced in decidual tissues from first-trimester spontaneous abortions compared to normal pregnancies [44]; and 4) reduced decidual expression of glycodelin-A has been reported in pregnancies complicated with preeclampsia, HELLP syndrome, and fetal growth restriction, compared to normal pregnancies [45]. Our observation that amniotic fluid glycodelin-A concentration is decreased in chronic chorioamnionitis is novel. This strongly suggests that the impairment of a physiologic constraint of the maternal immune system against fetal antigens is associated with and precedes the development of chronic chorioamnionitis.
Glycodelin-A production and release are under hormonal control, as it is up-regulated by progesterone [46], with estrogen exerting a priming effect [47]; human chorionic gonadotrophin (hCG) also stimulates glycodelin-A expression in endometrial cells [48], while glycodelin-A induces hCG expression by the cytotrophoblasts [49]. The concentration of glycodelin-A in the amniotic fluid, as well as in the maternal serum, is known to vary with gestational age. Amniotic fluid concentration of glycodelin-A increases during early pregnancy to peak at 15–16 weeks of gestation, and then it rapidly drops until 19 weeks, has a temporary elevation around 23 weeks, and finally decreases until term [23,50]. A significant correlation between glycodelin-A concentration and gestational age was also confirmed in this study. Taken together, the elevation in T cell chemokine CXCL10 concentration and the decreased concentration of the T cell suppressor glycodelin-A in the amniotic fluid are consistent with the notion that a derangement of the mechanisms governing maternal immune tolerance toward the semi-allogeneic fetus is involved in the pathogenesis of chronic chorioamnionitis [11].
This study also identified a set of proteins that were differentially expressed in amniotic fluid associated with chronic chorioamnionitis cases compared to acute chorioamnionitis cases or control cases, suggesting chronic chorioamnionitis as a distinct pathological process. Many of these proteins, such as α 1-antitrypsin, α1-antichymotrypsin, haptoglobin, ceruloplasmin, and orosomucoid 2 are known as acute phase proteins [51]. These are mainly, but not exclusively, produced by hepatocytes in response to pro-inflammatory stimuli, such as infections, trauma, tissue infarction, and cancer [51]. The origin and the role of these proteins in the amniotic fluid are not yet known; however, they are likely to serve several functions besides the modulation of the acute inflammatory process. Haptoglobin, for instance, inhibits T cell response to phytohemagglutinin [52,53], B-cell response to lipopolysaccharide [54,55] and the release of T helper 2 cytokines IL-4, IL-5, IL-13, and IL-10 [53]. Moreover, haptoglobin acts as an adipokine with chemotactic action toward monocytes-macrophages in white adipose tissue [56]. It is therefore plausible that haptoglobin and other acute phase proteins present in the amniotic fluid are involved in the immunologic regulation of maternal-fetal interactions, although their roles have yet to be explored.
In summary, we report unique changes in the amniotic fluid proteome associated with chronic chorioamnionitis for the first time, and propose that acute chorioamnionitis and chronic chorioamnionitis comprise two major clusters of spontaneous preterm births having distinct perturbations in the intra-amniotic protein environment. While acute chorioamnionitis is a result of microbial invasion of the amniotic cavity and intrauterine infection [57,58], chronic chorioamnionitis could be a consequence of derangement of the hormone immune system (progesterone, glycodelin-A) affecting CD8+ T cell activity [59]. The findings herein strongly suggest that compromised tolerogenic pressure in the intrauterine environment leads to maternal anti-fetal rejection in the form of chronic chorioamnionitis in the chorioamniotic membranes.
Acknowledgments
This work was supported (in part) by the Perinatology Research Branch, Division of Intramural Research, Eunice Kennedy Shriver National Institute of Child Health and Human Development, NIH, DHHS. The authors would like to thank Dr. Hongjin Huang for her excellent technical support in proteomic analysis. The authors are grateful to the patients who agreed to participate in our studies, the nurses, laboratory staff, and clinicians who made this work possible.
Footnotes
The authors declare no conflicts of interest.
Author contribution statement
GO, DCL, JL, ZD, and CJK carried out experiments and analyzed the data. RR, FG, NGT, TC, PM, SSH, and CJK conceived experiments and analyzed data. All authors participated in writing the paper and approved the submitted manuscript.Reference List
Full text links
Read article at publisher's site: https://doi.org/10.1002/path.2825
Read article for free, from open access legal sources, via Unpaywall:
https://europepmc.org/articles/pmc3080026?pdf=render
Citations & impact
Impact metrics
Citations of article over time
Article citations
Proteomic analysis of plasma to identify novel biomarkers for intra-amniotic infection and/or inflammation in preterm premature rupture of membranes.
Sci Rep, 13(1):5658, 06 Apr 2023
Cited by: 2 articles | PMID: 37024561 | PMCID: PMC10079851
Intra-amniotic inflammation in the mid-trimester of pregnancy is a risk factor for neuropsychological disorders in childhood.
J Perinat Med, 51(3):363-378, 29 Sep 2022
Cited by: 4 articles | PMID: 36173676 | PMCID: PMC10010737
Toward a new taxonomy of obstetrical disease: improved performance of maternal blood biomarkers for the great obstetrical syndromes when classified according to placental pathology.
Am J Obstet Gynecol, 227(4):615.e1-615.e25, 03 Sep 2022
Cited by: 16 articles | PMID: 36180175 | PMCID: PMC9525890
The Cellular Transcriptome in the Maternal Circulation During Normal Pregnancy: A Longitudinal Study.
Front Immunol, 10:2863, 17 Dec 2019
Cited by: 36 articles | PMID: 31921132 | PMCID: PMC6928201
Role of galectin-glycan circuits in reproduction: from healthy pregnancy to preterm birth (PTB).
Semin Immunopathol, 42(4):469-486, 29 Jun 2020
Cited by: 9 articles | PMID: 32601855 | PMCID: PMC7508936
Review
Free full text in Europe PMC
Go to all (37) article citations
Similar Articles
To arrive at the top five similar articles we use a word-weighted algorithm to compare words from the Title and Abstract of each citation.
The frequency, clinical significance, and pathological features of chronic chorioamnionitis: a lesion associated with spontaneous preterm birth.
Mod Pathol, 23(7):1000-1011, 26 Mar 2010
Cited by: 137 articles | PMID: 20348884 | PMCID: PMC3096929
Midtrimester amniotic fluid concentrations of interleukin-6 and interferon-gamma-inducible protein-10: evidence for heterogeneity of intra-amniotic inflammation and associations with spontaneous early (<32 weeks) and late (>32 weeks) preterm delivery.
J Perinat Med, 40(4):329-343, 01 Jun 2012
Cited by: 100 articles | PMID: 22752762 | PMCID: PMC3498502
CXCL10 and IL-6: Markers of two different forms of intra-amniotic inflammation in preterm labor.
Am J Reprod Immunol, 78(1), 19 May 2017
Cited by: 39 articles | PMID: 28544362 | PMCID: PMC5488235
Interleukin 6 determination in the detection of microbial invasion of the amniotic cavity.
Ciba Found Symp, 167:205-20; discussion 220-3, 01 Jan 1992
Cited by: 66 articles | PMID: 1425014
Review
Funding
Funders who supported this work.
Eunice Kennedy Shriver National Institute of Child Health and Human Development
Intramural NIH HHS (1)
Grant ID: ZIA HD002400-19