Human leucocyte antigen (HLA) expression of primary trophoblast cells and placental cell lines, determined using single antigen beads to characterize allotype specificities of anti-HLA antibodies

Abstract

Human trophoblast cells express an unusual repertoire of human leucocyte antigen (HLA) molecules which has been difficult to define. Close homology between and extreme polymorphism at the classical HLA class-I (HLA-I) loci has made it difficult to generate locus-specific monoclonal antibodies (mAbs). The problem of defining an antibody's reactivity against the thousands of existing HLA-I allotypes has often made it impossible to determine the HLA bound by a mAb in biological samples from a normal outbred population. Here we have used commercially available beads coated with individual HLA-I to characterize experimentally the reactivity of nine mAb against 96 common HLA-I allotypes. In conjunction with donor HLA-I genotyping, we could then define the specific HLA molecules bound by these antibodies in normal individuals. We used this approach to analyse the HLA expression of primary trophoblast cells from normal pregnancies; the choriocarcinoma cells JEG-3 and JAR; and the placental cell lines HTR-8/SVneo, Swan-71 and TEV-1. We confirm that primary villous trophoblast cells are HLA null whereas extravillous trophoblast cells express HLA-C, HLA-G and HLA-E, but not HLA-A, HLA-B or HLA-DR molecules in normal pregnancy. Tumour-derived JEG-3 and JAR cells reflect extravillous and villous trophoblast HLA phenotypes, respectively, but the HLA repertoire of the in vitro derived placental cell lines is not representative of either in vivo trophoblast phenotype. This study raises questions regarding the validity of using the placental cell lines that are currently available as model systems for immunological interactions between fetal trophoblast and maternal leucocytes bearing receptors for HLA molecules.

Figure 1

Characterization of antibody reactivity against human leucocyte antigen class-I (HLA-I) molecules. The reactivity of antibodies used in this study was tested experimentally against 96 of the common HLA-I allotypes, using commercially available beads coated with individual HLA-I molecules. Binding of isotype control (open points) or the indicated murine monoclonal antibodies (filled points) is shown for each of the HLA-I allotypes tested. Plotted results are the median fluorescence intensity of at least 100 normally-distributed beads, and representative of at least three independent experiments for each antibody.

Figure 2

Human leucocyte antigen (HLA) expression of primary extravillous trophoblast cells (EVT). Trophoblast cells were isolated from normal human pregnancies in the first trimester and analysed by triple-colour flow cytometry. (a) EVT were identified using the following gates. Live cells selected by scatter (R1), placental macrophages labelled with monoclonal antibody (mAb) to ILT3 and excluded from subsequent analysis (R2), and the specific marker HLA-G identifies EVT (R3). (b) HLA-G expression by EVT was confirmed by three different mAbs: MEM-G/9, G233 and MEM-G/11, with histograms shown using the gates R1 AND NOT R2. (c) Expression of classical HLA-I (Tu149, B1.23.2, MA2.1, 22E-1, Tu155) and HLA-II (L243) was investigated using the mAbs indicated. Histograms shown use the gates R1 AND R3 AND NOT R2. Staining of the indicated mAb is shown as a filled trace with isotype controls as open plots. The results shown in (a–c) are representative of the majority of trophoblast samples from over 50 independent pregnancies. (d–f) Dot plots are shown for mAb MA2.1 (d), Tu155 (e) and 22E-1 (f) binding to placental cell preparations from selected donors. In these individuals the mAb can be seen binding to HLA-G^–, but not EVT cells.

Figure 3

Human leucocyte antigen (HLA) expression of primary villous trophoblast cells. Trophoblast cells were isolated from normal human pregnancies in the first trimester and analysed by triple-colour flow cytometry. (a) Villous trophoblast cells were identified using the following gates. Live cells selected by scatter (R1), placental macrophages labelled with monoclonal antibody (mAb) to ILT3 and excluded from subsequent analysis (R2), and villous trophoblast identified by staining for epidermal growth factor receptor (EGFR) (R3). (b) HLA-I (BBM.1, W6/32, MA2.1, Tu155) and HLA-II (L243) expression was investigated using the indicated mAb. Each histogram uses the gates R1 AND R3 AND NOT R2. Staining of the indicated mAb is shown as a filled trace with isotype controls as open plots. The results are representative of staining for at least three independent pregnancies for each mAb.

Figure 4

Human leucocyte antigen (HLA) expression of choriocarcinoma cell lines. The cell lines JEG-3 and JAR were analysed by single-colour flow cytometry using mAb to classical HLA-I (W6/32, B1.23.2, 22E-1, MA2.1, Tu155), HLA-G (G233, MEM-G/9) and HLA-II (L243). Histograms use the single scatter gate, show binding of the indicated mAb (filled trace) and isotype control (open trace) and are representative of at least three independent experiments.

Figure 5

Human leucocyte antigen (HLA) expression of primary trophoblast and choriocarcinoma cell lines after exposure to interferon-γ (IFN-γ). HLA expression of the choriocarcinoma cell lines JAR and JEG-3, and primary extravillous trophoblast was measured. Histograms show staining of the indicated monoclonal antibody (mAb) after culture with IFN-γ (red traces), culture without IFN-γ (green traces) and staining of isotype control mAb (broken trace) which was the same ±IFN-γ. Results are representative of at least three independent experiments with each mAb.

Figure 6

Human leucocyte antigen (HLA) expression of placental cell lines transformed in vitro. (a) Constitutive surface HLA expression of the placental cell lines TEV-1, Swan-71 and HTR-8 was analysed by single-colour flow cytometry. Classical HLA-I (W6/32, Tu149, B1.23.2, MA2.1, 22E-1, Tu155), HLA-G (G233, MEM-G/11) and HLA-DR (L243) expression was measured using the monoclonal antibody (mAb) shown. Histograms use the single scatter gate shown for each cell line and include staining of the indicated mAb (filled trace) and isotype control (open trace). (b) HLA surface protein expression after interferon-γ (IFN-γ) stimulation of Swan-71 and HeLa cells. Filled traces show binding of the isotype control. Staining of the indicated mAb is shown for unstimulated cells (single line), cells treated with IFN-γ for 24 hr (heavy trace) and 48 hr (dashed trace). (c) Messenger RNA (mRNA) expression of HLA and related genes after 24 hr IFN-γ stimulation of HeLa, JEG-3, HTR-8 and Swan-71 cells. Each of these results are representative of at least three replicate experiments and the identity of HLA-A mRNA was confirmed by sequencing of the polymerase chain reaction product.