PREFACE Since the discovery of AZT in 1985, soon after the discovery of the human immunode- ficiency virus (HIV), a number of laboratories of nucleosides and viral pharmacology intensified their efforts in order to come-up with more safe and potent anti-HIV agents. From these efforts, scientists further discovered ddl, ddC and d4T as potent and clinically useful agents, although these agents also possess different profiles of antiviral activity as well as side-effects. With these agents in hand we opened a new era of combination chemotherapy of viral diseases. However, retrospectively, the major breakthrough came when 3TC (lamivudine) was discovered as a potent anti-HIV agent. The discovery of 3TC was almost a shock to conventional nucleoside chemists as 3TC was an oxathiolane derivative, and furthermore it was an L-nucleoside. L-Nucleosides have previously never been demonstrated any significant biological activity. Of course, we know now that more than 70-80 % of AIDS patients are currently taking 3TC as part of combination therapy. Furthermore, the discovery of 3TC fueled the chemistry of L-nucleosides for the last ten years. Thanks to 3TC, we are now witnessing at least six L-nucleosides (FTC, L-FMAU, L-d4FC, L-OddC, L-dT and L-dC) currently undergoing clinical trials. This all happened in the 90s, thus nucleoside chemists call it the "L-nucleosides decade" Parts of this book reflect the development of the L-nucleosides chemistry and pharma- cology, plus other conventional antiviral nucleosides discovered during the last ten years. This book covers experimental antiviral agents discovered up to the middle of 2002. At The University of Georgia, I am using the chapters in this book as the text for our gradu- ate students in medicinal chemistry. It works out well because these chapters not only cover modem carbohydrate and nucleoside chemistry, but also deal with biology and chemotherapy of antiviral agents. Thus, these chapters encompass the so called "from bench to beside." That is what our aspiring graduate students in medicinal chemistry need to be exposed to due to the interdisciplinary nature of drug discovery and develop- ment. I hope this book will be useful to those who are already involved in the field for a quick review as well as graduate students who are entering the field of antiviral chemistry and chemotherapy. Finally, I would like to dedicate this book to scientists who have contributed to the field of nucleoside chemistry and biology. Some of them participated as part of the Gordon Conference in New Port, Rhode Island in 1997. Keep-up your good work! C. K. Chu Athens, Georgia USA August, 2002 Vll CHAPTER 1 RECENT ADVANCES IN ANTIVIRAL NUCLEOSIDES GIUSEPPE GUMINA, YONGSEOK CHOI and CHUNG K. CHU 1.1. Introduction During the last two decades, treatment of viral infections has advanced remarkably, thanks to the heroic efforts of chemists and pharmacologists, the rapid progress of molecular virology as well as the cumulative knowledge of more detailed mechanism of action of antiviral agents.^ In recent years, we are facing an outburst of new and emerging viral diseases, such as new strains of hepatitis and herpes viruses, Ebola virus, West-Nile virus, plus a number of exotic viruses which, although still isolated in small areas of the world, have the potential for pandemic outbreak. Besides, the threat that viruses and other microorganisms could be used as biological weapons in warfare or bioterrorism has become a reality. Although vaccination is a valuable tool to fight viral diseases and in some cases is available and successful, the difficulty associated with state- or world- wide vaccination programs makes antiviral chemotherapy a more practical approach in the fight to epidemic viral infections. Among the most successful antiviral agents, nucleoside analogs have been the drugs of choice in the treatment of a number of deseases caused by herpes simplex virus (HSV), human cytomegalovirus (HCMV), varicella zoster virus (VZV), human immunodeficiency virus type 1 (HIV-1) and human hepatitis B (HBV) and C (HCV) vuns. Since 1980, a variety of biologically interesting and promising nucleosides have been discovered, some of which are being used clinically or are undergoing preclinical or clinical development. Currently, eighteen nucleosides are clinically being used for the treatment of HIV-1, herpes virus, HBV, RSV and HCV infections (Table 1). Despite these achievements, continued discoveries of novel nucleoside analogs are needed in order to overcome common problems in antiviral chemotherapy, such as toxicity, metabolic instability and, above all, the emergence of resistant viral strains as well as of new and emerging viral diseases. So far, a number of reviews have been published, regarding specific nucleoside classes,^ general aspects of nucleosides^ and their chemistry^ as well as their antiviral activity spectrum and target of actions.^ In view of these reviews, the purpose of this chapter is to give a brief overview on the most recent advances in antiviral nucleosides focusing on the structure-activity relationships with particular regard to the biochemical mode of action of the most promising nucleosides. Antiviral Nucleosides: Chiral Synthesis and Chemotherapy, Ed. by C.K. Chu. 1 — 76 © 2003 Elsevier B.V. All rights reserved. G. Gumina, Y. Choi and C. K. Chu Table 1. Antiviral nucleosides used in clinics Generic name Anti-HIV agents Zidovudine Didanosine Zalcitabine Stavudine Lamivudine Abacavir Tenofovir disoproxil Anti-HBV agent Lamivudine Anti-Herpetic agents Idoxuridine Trifluridine Acedurid Vidarabine Acyclovir Valaciclovir Penciclovir Famciclovir Ganciclovir Cidofovir Virazole Acronyms AZT ddl ddC d4T 3TC 1596U89 PMPA 3TC IdU TFT EdU araA ACV val-ACV PCV FCV DHPG (S)-HPMPC Ribavirin Target viruses HIV-1 HIV-1 HIV-1 HIV-1 HIV-1 HIV-1 HIV-1 HBV HSV-1/2 HSV-1/2 HSV-1/2 HSV-1/2 HSV-1/2, HSV-1/2, HSV-1/2, HSV-1/2, HCMV HCMV • ~ \ J >- " , \ > ) vzv vzv vzv vzv } RSV, HCV Mode of action Reverse transcriptase inhibitor/ chain terminator Reverse transcriptase inhibitor/ chain terminator DNA polymerase inhibitor; topical use Selective viral DNA polymerase inhibitor Valine ester prodrug of acyclovir Selective viral DNA polymerase inhibitor; topical use Oral prodrug of PCV Selective viral DNA polymerase inhibitor Viral RNA polymerase inhibitor Recent Advances in Antiviral Nucleosides 1.2. Structural features of nucleosides as antiviral agents Nucleoside analogs as inhibitors of viral replications usually act by interaction of their triphosphates with viral polymerases. As structural units of nucleic acids, the nucleoside triphosphates (NTPs) are the substrates for polymerase enzymes, which catalyze the polymerization of the NTPs. The biosynthesis of the NTPs is controlled by nucleoside kinases. The structural requirements of nucleosides to interact with kinases and poly- merases have important implications in the design of potential antiviral nucleosides (Figure 1). The 5'-hydroxymethyl group and the base moiety of nucleosides interact with kinases and their complementary nucleotides on the DNA template. The sugar moiety of the nucleoside can be considered as a spacer to connect the hydroxymethyl group and the base moiety.^ Therefore, modification of the sugar moiety has provided opportunities in the design of biologically active nucleosides. Some viruses, such as herpes viruses, encode their own nucleoside-phosphorylat- ing enzymes, which offers the potential for a therapeutic target.^ Nucleosides, which are preferably phosphorylated by viral enzymes rather than by the cellular homologue, are only activated in infected cells and can have high selectivity against these viruses. This is, for example, the main factor in the success of acyclovir (ACV). However, other viruses, such as HIV and HBV, do not encode nucleoside kinases. In order to be active against these viruses, nucleoside analogs have to be phosphorylated by cellular kinases. Thus, the selectivity between antiviral activity and cellular toxicity depends on the sub- strate specificity of the NTPs for viral and host polymerases, and often the therapeutic exploitation of active nucleosides is compromised by the toxicity resulting from inhibi- tion of the host enzymes or incorporation in the host nucleic acids. In general, enzymes act on one enantiomer of a chiral substrate, the specificity of which is related to the unique structure of the enzymes.^ However, recent findings have indicated that there are some exceptions to this rule among enzymes involved in the phosphory- lation of nucleosides.^'^'^'^° For instance, herpes virus thymidine kinases (TKs) phos- phorylate both D- and L-enantiomeric forms of several uracil analogs as well as acyclic nucleosides, cellular deoxycytidine (dCyd) kinase phosphorylates both enantiomeric forms of several dCyd analogs, and some viral DNA polymerases, such as herpes viruses, HIV-1 RT and HBV DNA polymerase, are inhibited by the triphosphates of a number of L-nucleosides. These findings offer new opportunities for antiviral chemotherapy, although, at the molecular level, it is not completely understood how kinases phosphory- late both D- and L-nucleosides. In recent years, a growing number of nucleoside analogs have been discovered which exert their antiviral activity by inhibiting enzymes different from polymerases, such as inosine monophosphate dehydrogenase, 5'-adenosylhomocysteine hydrolase, orotidine 5'-monophosphate decarboxylase and CTP synthetase.^^ Such compounds may prove useful because, by targeting different enzymes, they may offer synergistic action with classic polymerase inhibitors. G. Gumina, Y. Choi and C K. Chu HO-i B 1 ^ O II •Q-P-O B -Q-P-O-P-O-P-O B 0 O "O-P-O-P-O 6' 6' Interaction with viral DNA polymerases O ^ and/or Interaction with cellular DNA polynnerases a, p, y, e [ Antiviral activity J cytotoxicity, mitochondrial toxicity, antitumor activity A: virus-encoded TK, cellular TK, dCyd kinase, dPyr kinase, dGua kinase, 5'-nucleotidase, adenosine phosphotransferase, or UL97 gene product, etc. B: Nucleotide kinases or 5-phosphoribosyl-1-pyrophosphate synthetase, etc. C: NDP kinase, phosphoenolpyruvate carboxykinase, phosphoglycerate kinase or creatine kinase, etc. Figure 1. General mode of action of nucleoside analogs. 1.3. 2'-Deoxy nucleosides and related analogs 2'-Deoxy nucleoside analogs have proved effective against DNA viruses, such as HSV, VZV, EBV and HBV. Some nucleosides of this class show poor selectivity between the viral polymerases and the host polymerases due to their structural resemblance to natural substrates. However, modification of the base moiety or the sugar portion as well as the use of the unnatural L-enantiomer have been shown to reduce cellular toxicity, as in the case of 2'-fluoro-5-methyl-p-L-arabinofuranosyluracil (L-FMAU).^^ Recent Advances in Antiviral Nucleosides Since the discovery of the first antiherpes compound, 5-iodo-2'-deoxyuridine (1, IdU)/^^ a modifications of the 5-position of the pyrimidine moiety have produced a number of active antiviral compounds (Figure 2)P Several 5-substituted 2'-deoxy-uridines, such as 5-trifluoromethyl-2'-deoxyuridine (2, TFT)!^^ and 5-ethyl-2'-deoxyuridine (3, EdU)/^^ have been approved for the treatment of herpetic keratitis. IdU and TFT are phosphory- lated to their triphosphates by the virus-encoded TK. The triphosphates inhibit HSV DNA polymerase as well as, even though to a lesser extent, cellular polymerases. EdU has higher affinity for the herpesvirus-induced TK than for the cellular TK, and its tri- phosphate is incorporated to a large extent into the viral DNA.^^'^^"^ BVdU (4, brivudin) was originally synthesized by Walker and co-workers and shown to be a potent and selective anti-herpes agent. ^"^ It is specifically phosphorylated by virus- encoded TK and nucleoside diphosphate (NDP) kinase to give BVdUTP, which may act as either an inhibitor of or a substrate for viral DNA polymerase. However, BVdU is cleaved by pyrimidine nucleoside phosphorylases to (£)-5-(2-bromovinyl)uracil (BVU), which is cytotoxic.^^ The marked loss of activity of BVdU in thymidine kinase-deficient (TK) HSV-1 or VZV strains has been bypassed by its incorporation into phosphorami- date prodrugs {vide infra)}^ The inhibitory effects of several 5-alkynyl-2'-deoxyuridine analogs on virus replication, host cell metabolism and tumor cell proliferation have been investigated, among which 5-ethynyl-2'-deoxyuridine (5) is the most cytotoxic against L1210 cells.^^ 5-Heteroaryl-substituted 2'-deoxyuridines,^^'^^ i.e. 5-(3-Bromoisoxazol- 5-yl)-2'-deoxyuridine (6), 5-(5-bromothien-2-yl)-2'-deoxyuridine (7) and 5-(5-chloro- thien-2-yl)-2'-deoxyuridine (8) also share with BVdU a common antiviral spectrum against various strains of HSV-1 and VZV, but not HSV-2, HCMV or TK" HSV-l.^'^ 0 o HO—1 ' ^ N ^ O O ' ^ N " |—OH HO OH 1 (IdU, X = l) 2 (TFT,X = CF3) 3 (EdU, X = Et) 4 (BVdU, X = (E)-bromovinyl) 5 (X = CsCH) 6 (X = 3-Br-isoxazol-5-yl) 7 (X = 5-Br-thien-2-yl) 8 (X = 5-Cl-thien-2-yl) 9 (L-ldU, X = 1) 10 (L-BVdU, X = (E)-bromovinyl) 11 (L-dT,X = CH3) Figure 2. 2'-Deoxyuridine analogs. Focher et al. have demonstrated that L-IdU (9), L-BVDU (10) and L-thymidine (11, L-dT) (Figure 2) are not recognized by human cytosolic TK in vitro, but function as a substrate for HSV-1 TK and inhibit HSV-1 proliferation in infected cells.^ L-dT is selectively phosphorylated in vivo to L-dTMP by HSV-1 TK. L-dTMP is further phosphorylated to the di- and triphosphate forms by non-stereospecific cellular kinases. L-dTTP not