ANTIVIRAL ACTIVITY AND MECHANISM OF ACTION OF A NOVEL URACIL ANALOG FOR VARICELLA-ZOSTER VIRUS

Abstract

The alphaherpesvirus varicella-zoster virus (VZV) is widespread in humans. VZV causes primary and recurrent diseases that are preventable with live vaccines and are treatable with antiviral drugs. New antiviral drugs for varicella-zoster virus (VZV) with increased potency are needed, especially to prevent post-herpetic neuralgia. The purpose of this project was to evaluate β-L-1-[5-(E-2-Bromovinyl)-2-(hydroxymethyl)-1,3-dioxolan-4- yl)] uracil (L-BHDU) and 5′-O-valyl-L-BHDU for efficacy, safety, resistance, and mechanism of action in three models of VZV replication: primary human foreskin fibroblasts (HFFs), skin organ culture (SOC) and in SCID-Hu mice with skin xenografts. We found that L-BHDU and valyl-L-BHDU were safe and effective against VZV in culture and in a mouse model. Herpes simplex virus Type 1 was also sensitive to LBHDU in cultured cells. The mechanism of action of L-BHDU and its effect on drugdrug interactions were not known. Given its similar structure to brivudine (BVdU), we addressed whether L-BHDU, like BVdU, inhibits 5-fluorouracil (5-FU) metabolism. LBHDU did not interfere with 5FU metabolism, indicating that L-BHDU is a safer drug than BVdU. However, L-BHDU antagonized the activity of acyclovir (ACV), BVdU and foscarnet (PFA) in cultured cells, which was due to competition for phosphorylation by VZV thymidine kinase (TK). The mechanism of action of L-BHDU was studied by evaluating its activity against related α-herpesviruses and by analyzing resistant VZV viii strains. VZV strains resistant to L-BHDU (L-BHDUR ) were cross-resistant to ACV and BVdU but not to PFA and cidofovir (CDV). Whole genome sequencing of L-BHDUR strains identified mutations in ATP-binding (G22R) and nucleoside binding (R130Q) domains of VZV TK. The purified L-BHDUR TKs were enzymatically inactive and failed to phosphorylate the drug. In wild type VZV- infected cells, L-BHDU was converted to L-BHDU mono- and diphosphate forms; cells infected with L-BHDUR virus did not phosphorylate the drug. We also investigated whether addition of nucleosides reversed LBHDU inhibition of VZV in dividing and quiescent HFFs. Excess thymidine and uridine, but not purines, in proportion to L-BHDU restored VZV replication only in dividing cells, suggesting that the active form of L-BHDU interfered with pyrimidine biosynthesis. Like other herpesviruses, VZV infection induced thymidine triphosphate (dTTP) in confluent cells while L-BHDU treatment decreased the dTTP pool. Some herpesviruses raise dNTP pools by inducing cellular enzymes. However, VZV infection did not increase cellular thymidylate synthase (TS) expression to facilitate viral replication. Furthermore, the active form of L-BHDU did not interfere with cellular metabolism, suggesting a viral target. Further studies are required to identify the target(s) of L-BHDU active form(s).