The presence of macrophages in the stromal core of the placental villous suggests that the placenta has the ability either to initiate an in situ inflammatory response or to react to an exogenous inflammatory stimulus. Decidual macrophages and their cross-talks with the neighboring placenta cells are well-characterized, particularly as they relate to placental invasion and the tolerance of the fetus by the maternal organism at early development stages. By contrast, the role and function of placental resident macrophages at later stages of pregnancy is not as well-understood. Inflammation at the maternal–fetal interface is a common feature of gestational diseases in humans that are associated with either fetal or maternal death. The increased accumulation of macrophages in the placenta during pregnancy with preeclampsia, obesity, or diabetes mellitus suggests that disorders of maternal homeostasis have the potential to impact the placental immune function.

The knowledge on effectors and cellular pathways that control the engagement of macrophages into the inflammatory response, their migration, and maturation is rapidly extending in the context of low-grade inflammation. Peripheral blood monocytes can emigrate through the endothelial barrier into various tissues under both inflammatory and noninflammatory conditions. For instance, adipose tissue macrophages of obese individuals are derived primarily from activated systemic monocytes that extrude through vascular walls into the stroma vascular compartment. On infiltration, the monocytes differentiate into activated specialized cells with phenotypic and functional traits that are dictated by their cellular environment. It is not known currently whether this infiltrating process applies to the placenta in late gestation. Because the placenta is of fetal lineage and harbors a population of resident macrophages, the Hofbauer cells that originate from fetal myeloid cells early in development.

To gain insight into the pathways that are associated with chronic placental inflammation, this study was aimed at investigating the characteristics of the monocyte macrophage populations that are present at the maternal–fetal interface in late gestation. Based on our previous observations that, in human obesity, the placental macrophages and maternal blood monocytes are in a proinflammatory state, we sought to characterize monocytes that are derived from both the placenta and systemic circulation of obese women. Using a gene-set approach, we show that placental- and maternal-activated monocytes exhibit common genotypic and phenotypic characteristics that suggest that they both contribute to immune crosstalk at the maternal–fetal interface.

Materials and Methods

Subjects

The study cohort included 18 overweight/obese women (age, 27 ± 2 years; pregravid body mass index, 35.2 ± 1.7 kg/m ² ) with a term singleton pregnancy (38.7 ± 0.2 weeks) of 10 male and 8 female fetuses. To rule out a potential inflammatory component that would be related to parturition, our subjects were recruited at elective cesarean delivery. There were no clinical or laboratory signs of infection or history of autoimmune disorders. Maternal blood was drawn on admission to labor and delivery before placement of a normal saline solution intravenous line for hydration. Umbilical venous blood was drawn with a syringe from the double-clamped cord immediately after delivery of the placenta. The protocol was approved by the MetroHealth Medical Center Institutional Review Board and Clinical Research Unit Scientific Review Committee. Volunteers gave their written informed consent in accordance with the MetroHealth Medical Center guidelines for the protection of human subjects. Maternal body mass index was assessed with prepregnancy weight by history and measured height of the women at her first antenatal visit. Birthweight, placental weight, and neonatal percent body fat that were recorded as described were 3627 ± 122 g, 734 ± 43 g, and 16.1 ± 0.4 g, respectively.

Cell isolation and immunoselection

Placental tissue was digested with trypsin and DNase; cells were purified by density gradient centrifugation. The average yield was 10 ⁶ cells/g of tissue with >80% viability and >85% of trophoblast cells that were positive for CD133. For mononuclear cells isolation, venous maternal and umbilical cord blood was separated by Ficoll-paque plus (GE Healthcare, formerly Amersham Biosciences, Piscataway, NJ) and directly processed for CD14 immunoselection with CD14-coupled magnetic beads (Dynalab, Reynoldsburg, OH).

Flow cytometry

For fluorescence-activated cell-sorting cells were incubated with fluorescent primary antibodies (CD14-PE, CD68-PE; Abcam Inc, Cambridge, MA) and CD133-FITC (Milteny Biotech, Auburn, CA) or control immunoglobulin G. Propidium iodide was used to control for cell viability. Multicolor population analysis was performed with paint-a-gate software (version 3.0.0; PPC; Becton Dickinson Labware, Franklin Lakes, NJ). All gating and data analysis were performed with CELLQuest software (version 3.2.1f1; Becton Dickinson Labware).

RNA extraction and real time reverse-transcriptase polymerase chain reaction (RT-PCR)

Total RNA was prepared from frozen placental cells with Trizol (Invitrogen, Carlsbad, CA) and from mononuclear cells with an RNeasy kit (Qiagen, Valencia, CA). Integrity of purified RNA samples was assessed by spectrometry (Agilent Technologies Inc, Santa Clara, CA). For real-time RT-PCR, complementary DNA was analyzed in duplicate with SYBR green dye (Lightcycler; Roche Molecular Diagnosis, Indianapolis, IN). The copy number of Y chromosome was assessed from the absolute amount of DYS14 gene based on a genomic DNA standard curve 0.5-500 ng/mL (Promega Corp, Madison, WI).

Transcriptional profile analysis

Hybridization of complementary RNA to HG-U133 plus 2.0 GeneChip pangenomic oligonucleotide arrays (Affymetrix, Santa Clara, CA) that contains 47,000 transcripts was performed as previously described. The raw data files from the Affymetrix arrays were imported into Gene Spring software (Agilent Technologies Inc) and normalized before further analysis. Genes with a fold change of >2 were considered differentially expressed between groups. The statistical difference of differential gene expression was estimated by the Student t test and significance analysis of microarray. The 16,739 genes that were common to the placenta and the maternal blood gene sets represented 70% and 73% of their respective transcriptome; 1800 genes (10%) were expressed differentially ( Figure 1 ). The microarray results were validated by RT-PCR analysis of 18 selected genes with total RNA from 8 individual blood and placental monocyte preparations ( Table 2 ). Clustering algorithm s were used to evaluate how the various sample groups were derived from principle component analysis (GeneGO, St. Joseph, MI). Over-represented biologic entities were identified within each cluster and annotated with Onto-Express software (Wayne State University, Detroit, MI) and a gene ontology database. Hierarchic clustering was performed with genesis on up-regulated genes to identify the most representative group identities.

Global pattern of gene expression of monocytes from maternal blood and the placenta

The whole transcriptome of CD14 monocytes from placenta ( PMNC ; 22,794 genes) and maternal blood ( MMNC ; 23,750 genes) was obtained by screening oligonucleotide arrays that contained a total of 47,000 transcripts. Filtering of raw data identified 16,739 genes that were common to the 2 datasets; 1800 genes that had a false discovery rate of <0.05 and a fold change of >2 were considered differentially expressed between groups.

Basu. Molecular phenotype of monocytes at maternal–fetal interface. Am J Obstet Gynecol 2011.

Statistical analysis

All values are presented as means ± SE. Differences among variables were examined with 1-way analysis of variance. Statistical significant differences between variables were tested with a Fisher’s protected least significant difference post-hoc test; the nonparametric McNemar’s test was used for the comparison of the distribution between gene clusters. The data were analyzed with the Statview II statistical package (Abacus Concepts, Berkeley, CA). Statistical significance was set at .05, unless otherwise indicated.