High expression of PGE2 enzymatic pathways in cervical (pre)neoplastic lesions and functional consequences for antigen-presenting cells

Abstract

Although human papillomavirus (HPV) DNA is detected in the majority of squamous intraepithelial lesions (SIL) and carcinoma (SCC) of the uterine cervix, the persistence or progression of cervical lesions suggest that viral antigens are not adequately presented to the immune system. This hypothesis is reinforced by the observation that most SIL show quantitative and functional alterations of Langerhans cells (LC). The aim of this study was to determine whether prostaglandins (PG) may affect LC density in the cervical (pre)neoplastic epithelium. We first demonstrated that the epithelial expression of PGE(2) enzymatic pathways, including cyclooxygenase-2 (COX-2) and microsomal prostaglandin E synthase 1 (mPGES-1), is higher in SIL and SCC compared to the normal exocervical epithelium and inversely correlated to the density of CD1a-positive LC. By using cell migration assays, we next showed that the motility of immature dendritic cells (DC) and DC partially differentiated in vitro in the presence of PGE(2) are differentially affected by PGE(2). Immature DC had a lower ability to migrate in the presence of PGE(2) compared to DC generated in vitro in the presence of PGE(2). Finally, we showed that PGE(2) induced a cytokine production profile and phenotypical features of tolerogenic DC, suggesting that the altered expression of PGE(2) enzymatic pathways may promote the cervical carcinogenesis by favouring (pre)cancer immunotolerance.

Fig. 1

CD1a, COX-2 and mPGES-1 immunostaining in cervical biopsy specimens. The normal exocervical epithelium shows a high density of CD1a⁺ cells in basal and supra-basal cell layers (a). EpM (d), SIL (g) and SCC (j) demonstrate a low density of CD1a⁺ cells. The normal squamous epithelium shows a lack of COX-2 (b) and mPGES-1 (c) immunoreactivity. EpM (e, f), SIL (h, i) and SCC (k, l) demonstrate a high but heterogeneous staining for COX-2 and mPGES-1. The negative control for CD1a, COX-2 and mPGES-1 immunostaining are respectively an exocervical epithelium (m), a SCC (n) and a SIL (o)

Fig. 2

Semi-quantitative evaluation of CD1a, COX-2 and mPGES-1 expression in normal exocervical epithelium (n = 38), EpM (n = 12), SIL (n = 18) and SCC (n = 15). Asterisks indicate statistically significant differences (*P < 0.05, **P < 0.01, ***P < 0.001)

Fig. 3

Comparison of COX-2 (a, b, e) and mPGES-1 (b, d, f) immunostaining in paired biopsy specimens (exocervix vs. EpM, SIL or SCC) of individual patients. The horizontal bars indicate the mean score obtained for each diagnostic category. The bold lines correspond to biopsies from several patients

Fig. 4

In vitro activity of the enzymatic pathways responsible for the PGE₂ production in cervical cancer cell lines. a Western blot demonstrates higher levels of COX-2 and mPGES-1 proteins in HPV-positive cell lines (CaSki, CK2, C4-II) compared with an uninfected cervical cancer cell line (C33-A). b Cells were grown in culture medium for 72 h. The supernatants were then collected to determine by radioimmunoassay the levels of PGE₂ production. Results are representative of three independent experiments performed in duplicates. The means ± standard deviation is shown. Asterisks indicate statistically significant differences compared with C33-A cells (***P < 0.001)

Fig. 5

Influence of PGE₂ on DC migration in a Boyden Chamber assay. NC non-conditioned medium; HFM conditioned medium of human fibroblasts (positive control); HFM + PGE₂: conditioned medium of human fibroblasts supplemented with PGE₂ (1 mM). Control DC generated in the absence of PGE₂ and DC differentiated in the presence of PGE₂ were tested. Results are the means ± standard deviation of ten wells from three different experiments. Asterisks indicate statistically significant differences (*P < 0.05, **P < 0.01, ***P < 0.001)

Fig. 6

Phenotype analysis of DC differentiated in the presence or not of different concentrations of PGE₂. a Analysis of cell surface expression of CD1a, CD86, HLA-DR, HLA-ABC by different DC populations using flow cytometry. One of five experiments performed for each condition is shown. b Mean percentages of positive cells ± standard deviation obtained in the different experiments. The expression of surface markers CD11c, CD1a, CD83, CD86, HLA-ABC, HLA-DR and CCR7 were investigated. Asterisks indicate statistically significant differences compared with control DC (*P < 0.05, **P < 0.01, ***P < 0.001)

Fig. 7

DC differentiated in the presence of PGE₂ attenuate allogeneic PBMC proliferation in Mixed Leucocyte Reaction. PBMC were cultured with DC generated in the presence or not of different concentrations of PGE₂ for 6 days. Proliferation of PBMC was determined by assessing the uptake of [3H]thymidine for the last 18 h of the culture. For each condition, five independent experiments were performed in sexplicates. Data are presented as means ± standard deviation of these five experiments. Asterisks indicate statistically significant differences compared with control DC (**P < 0.01, ***P < 0.001)

Fig. 8

Secretion levels of cytokines (IL-10 and IL-12p70) in culture supernatants of DC, generated in the presence or not of PGE₂, measured by ELISA after 24 h of incubation with LPS. Data are presented as means ± standard deviation of four independent experiments performed in duplicate. Asterisks indicate statistically significant differences (*P < 0.05)