Cytomegalovirus Infection Triggers the Secretion of the PPARγ Agonists 15-Hydroxyeicosatetraenoic Acid (15-HETE) and 13-Hydroxyoctadecadienoic Acid (13-HODE) in Human Cytotrophoblasts and Placental Cultures

Abstract

Introduction: Congenital infection by human cytomegalovirus (HCMV) is a leading cause of congenital abnormalities of the central nervous system. Placenta infection by HCMV allows for viral spread to fetus and may result in intrauterine growth restriction, preeclampsia-like symptoms, or miscarriages. We previously reported that HCMV activates peroxisome proliferator-activated receptor gamma (PPARγ) for its own replication in cytotrophoblasts. Here, we investigated the molecular bases of PPARγ activation in infected cytotrophoblasts.

Results: We show that onboarded cPLA2 carried by HCMV particles is required for effective PPARγ activation in infected HIPEC cytotrophoblasts, and for the resulting inhibition of cell migration. Natural PPARγ agonists are generated by PLA2 driven oxidization of linoleic and arachidonic acids. Therefore, using HPLC coupled with mass spectrometry, we disclosed that cellular and secreted levels of 13-hydroxyoctadecadienoic acid (13-HODE) and 15-hydroxyeicosatetraenoic acid (15-HETE) were significantly increased in and from HIPEC cytotrophoblasts at soon as 6 hours post infection. 13-HODE treatment of uninfected HIPEC recapitulated the effect of infection (PPARγ activation, migration impairment). We found that infection of histocultures of normal, first-term, human placental explants resulted in significantly increased levels of secreted 15-HETE and 13-HODE.

Conclusion: Our findings reveal that 15-HETE and 13-HODE could be new pathogenic effectors of HCMV congenital infection They provide a new insight about the pathogenesis of congenital infection by HCMV.

Fig 1. Onboarded cPLA₂ is required for PPARγ activation and inhibition of cytotrophoblast migration.

(A) TLC analysis of bodipy-phosphatidylcholine (B-PC) incubated in the presence of buffer (mock), purified PLA₂ (PLA2), native HCMV particles (HCMV), HCMV particles pre-treated by MAFP (HCMV+MAFP) or UV-irradiated (HCMV-UV). (B) PPARγ activity luciferase assay performed with HIPEC in various conditions, using a reporter plasmid responsive to PPARγ (pGL4-PPRE) or a control plasmid (PGL4), 48 h pi or post treatment onset. HCMV: HIPEC infected by live HCMV particles; vi: the viral inoculum was treated beforehand by MAFP; cm: control with 50 nM MAFP in the culture medium; GW: PPARγ inhibitor GW9662; Rosi: PPARγ agonist rosiglitazone. RLU: relative luciferase units; **: p < 0.01 (Kruskal-Wallis test). The assay was repeated twice.(C) Wound-healing assays. HCMV+MAFP: the viral inoculum was treated beforehand by MAFP; HCMV+GW: the HIPEC were infected in the presence of GW9662; HCMV-UV: the viral inoculum was UV-irradiated. Results are expressed as the percent variation relative to the control. The figure shows results of a representative experiment, out of two independent experiments, each comprising triplicate measures. NI: non-infected.

Fig 2. 15-HETE and 13-HODE are released by HIPEC infected by live HCMV.

(A) LC-MS/MS analysis of the PUFA-derived lipids in HIPEC uninfected (NI), infected with live HCMV (HCMV) or UV-irradiated HCMV (UV-HCMV) (expressed as pg/mg of protein) and from the corresponding culture supernatants (expressed as pg/ml). (B) LC-MS/MS analysis of 13-HODE and 15-HETE in the culture medium of HIPEC uninfected (NI), infected by live HCMV (HCMV) or by HCMV treated by MAFP (HCMV-MAFP). (C) Red Oil O staining of uninfected HIPECs stimulated by 13-HODE or treated by the vehicle (DMSO). (D) Wound healing assay using uninfected HIPEC treated by different doses of 13-HODE, rosiglitazone (Rosi), or the vehicle (DMSO), or untreated (NT). The results shown are from one representative experiment, out of two independent experiments, each comprising triplicate measures. (E) Transwel migration assay using uninfected HIPEC or HIPEC infected by live HCMV (HCMV), UV-irradiated HCMV (HCMV-UV) or HCMV treated by MAFP (HCMV-MAFP) (MOI = 3), in the presence or the absence of 1μM rosiglitazone, 2 μM GW9662 (GW), 50 nM MAFP (MAFP), or 13-HODE (13-HODE) or 15-HETE (15-HETE) at various concentrations (nM). Shown are the results of two independent experiments. *: p < 0.05; **: p < 0.01; ***: p < 0.001.

Fig 3. Increased amounts of 15-HETE and 13-HODE secreted by from early placenta explants infected by HCMV.

(A) LC-MS/MS analysis of the amounts of PUFA-derived lipids secreted from histocultures from 12 first trimester placentas either infected by HCMV (CMV) or uninfected (NI). *: p < 0.05; **: p < 0.01 (Wilcoxon test). (B) Immunostaining analysis of HCMV antigen IE in infected (HCMV) or uninfected (NI) placental explants cultured ex vivo. Representative views are shown. Note the positive, nuclear IE staining in the cytotrophoblast layer (ST) surrounding the villus stroma (S).

Fig 4. Proposed model of PPARγ activation in HCMV infection of placental cells.

HCMV particles (HCMV) carry onboarded cPLA2 (oPLA2) which catalyses linoleic acid (LA) and arachidonic acid (AA) release from host membrane phospholipids (PL). AA and LA undergo oxidization driven by 15-lipoxygenase (15-LOX), which generates 15HETE and 9-HODE, respectively. 15HETE and 9-HODE are activating ligands of PPARγ, which dimerizes with RXR to regulate the expression of the host and virus genomes, resulting in impaired migration abilities in vitro, and enhanced IE transcription and viral replication. M: cell membrane, C: cytoplasm, N: nucleus.