Impact of new cyclooxygenase 2 inhibitors on human cytomegalovirus replication in vitro.

BACKGROUND: Human cytomegalovirus (HCMV) is involved in complications on immunocompromised patients. Current therapeutics are associated with several drawbacks, such as nephrotoxicity. PURPOSE: As HCMV infection affects inflammation pathways, especially prostaglandin E2 (PGE2) production via cyclooxygenase 2 enzyme (COX-2), we designed 2'-hydroxychalcone compounds to inhibit human cytomegalovirus. STUDY DESIGN: We first selected the most efficient new synthetic chalcones for their effect against COX-2-catalyzed PGE2. STUDY SAMPLE: Among the selected compounds, we assessed the antiviral efficacy against different HCMV strains, such as the laboratory strain AD169 and clinical strains (naïve or multi-resistant to conventional drugs) and toxicity on human cells. RESULTS: The most efficient and less toxic compound (chalcone 7) was tested against HCMV in combination with other antiviral molecules: artesunate (ART), baicalein (BAI), maribavir (MBV), ganciclovir (GCV), and quercetin (QUER) using Compusyn software. Association of chalcone 7 with MBV and BAI is synergistic, antagonistic with QUER, and additive with GCV and ART. CONCLUSION: These results provide a promising search path for potential bitherapies against HCMV.

Contrasting effect of new HCMV pUL54 mutations on antiviral drug susceptibility: Benefits and limits of 3D analysis.

Human cytomegalovirus (HCMV) resistance to antiviral drugs is a major drawback of repeated or long-duration treatment in immunocompromised patients. Resistance testing is usually performed by genotypic assays. For accurate interpretation of these assays, the role of new mutations in HCMV resistance has to be assessed. Two previously unknown UL54 single point mutations (D515Y and V787A) were characterized for phenotypic drug-resistance by marker transfer analysis using bacterial artificial chromosome (BAC) mutagenesis. Increases in 50% inhibitory concentrations of ganciclovir and foscarnet were found for both mutated recombinant strains showing that mutations D515Y and V787A induce resistance to both antivirals. Importantly, none of those impacted the viral growth kinetics. For a better understanding of their molecular resistance mechanisms, a 3D homology model was used to localize the mutated amino-acids in functional domains of UL54 and predict their impact on UL54 function and resistance. However, 3D homology model analysis has limits and phenotypic characterization using BAC-HCMV is still essential to measure the role of unknown mutations.

Ex vivo model of congenital cytomegalovirus infection and new combination therapies.

INTRODUCTION: Congenital human cytomegalovirus (HCMV) infection is a major public health problem due to severe sequelae in the fetus and newborns. Currently, due to their toxicity anti-CMV treatments cannot be administered to pregnant women. We thus developed an ex vivo model of 1(st) trimester placental CMV infection to observe the route of infection across the placenta and to test the efficacy of various new drugs targeting different stages of viral cycle. METHODS: After validation of the viability of floating villi explants by ELISA ß-HCG, the kinetics of placental infection were determined by immunochemistry and qPCR in this ex vivo model. Antiviral susceptibility was determined in vitro using focus reduction assay and by qPCR in the ex vivo model. RESULTS: The ex vivo model showed viral infection in trophoblasts and mesenchymal space of floating villi. In vitro, antiviral combinations of maribavir with baïcalein or artesunate inhibited viral infection by more than 90%. On the other hand, in ex vivo model, infection was reduced by 40% in presence of maribavir and artesunate. The synergistic effect observed in vitro was not observed ex vivo. DISCUSSION: This model allowed us to understand the CMV spread in 1(st) trimester floating villi better and to analyze the anti-CMV efficacy and toxicity of new drugs that could be administered to pregnant women, either alone or in combination. CONCLUSIONS: Such an ex vivo model could be applied to other viruses such as rubella or parvovirus B19 and in new drug development.