■ 1 History of Virus and Antiviral Drugs

Viruses are humanity’s invisible enemy. They wreak daily havoc by causing the flu, measles, rabies, hepatitis, smallpox, polio, and even human immunodeficiency virus (HIV). Although viruses have existed on the earth much longer than humans, it was not until in 1892 when the concept of virus took root when Chamberland experimented with viruses using the Pasteur–Chamberland filter. Solid evidence emerged when tobacco mosaic virus (TMV) crystal was isolated in 1935.

However, human ingenuity afforded successful measures to combat viruses long before 1892. For instance, Jenner successfully pioneered a vaccination for preventing smallpox in 1796, nearly one hundred years before Chamberland’s exploits and before Pasteur developed the first vaccination for rabies in 1885. The scourge of polio has been nearly wiped out thanks to Salk’s inactivated polio vaccine (IPV) available since 1954 and Sabin’s oral poliovirus vaccine (OPV) popularized in 1960. The 1951 the Nobel Prize in Physiology or Medicine was awarded to Theiler for his contributions to yellow fever vaccines.

In terms of small molecule antiviral drugs, the nucleoside iododeoxyuridine (IDU, 2), a simple analog of thymidine (3), was first synthesized and used as an antiviral drug in 1959 by Prusoff.¹ Unfortunately, due to its systemic cardiotoxicities, IDU is now only used topically to treat herpes simplex keratitis. A similar antiviral nucleoside, trifluorothymidine (TFT, Viroptic, 4), is less toxic than 2, and is also primarily used topically in eyes to kill the herpes simplex virus (HSV).²

Under the leadership of future Nobel laureate Elion, Burroughs Wellcome introduced the nucleoside analog acyclovir (Zovirax, 5) in 1978 for the treatment of HSV infection.³ While not the first antiviral agent on the market, Zovirax (5) was the first small molecule drug to be widely used to control a viral infection. Introduction of valacyclovir (Valtrex, 6), a prodrug of Zovirax (5) with higher oral bioavailability, afforded the patient a more convenient regimen because it does not have to be taken as frequently as the parent drug Zovirax (5).⁴ Coincidently the prodrug strategy employed here to improve oral bioavailability is a popular strategy in the antiviral drugs arena. The title antiviral drug sofosbuvir (Sovaldi, 1) is a prodrug as well.

1.1 HIV Antiviral Drugs

Montagnier’s discovery of the human immunodeficiency virus-1 (HIV-1) in 1983 earned him a Nobel Prize in 2008. Intensive efforts to fight the virus began as soon as acquired immunodeficiency syndrome (AIDS), caused by HIV, appeared at an alarming rate in the early 1980s. The National Institutes of Health (NIH) led a screening program by collecting compounds from major pharmaceutical companies. The first fruit of the pursuit was azidothymidine (zidovudine, AZT, 7), a sample submitted by Burroughs Wellcome. The nucleoside was initially prepared as an anticancer drug, but was shown to be an HIV reverse transcriptase inhibitor (RTI). In 1987, AZT (7) became the first antiretroviral agent approved for treating AIDS.⁵ AZT (7) works by incorporation of AZT triphosphate into the growing DNA chain of DNA elongation and thus inhibits DNA synthesis.⁶

Similar to AZT (7), d4T (stavudine, 8) was also first prepared as a potential anticancer agent and its HIV-1 inhibitory activity was revealed later. A newer drug, lamivudine (3TC, Epivir, 9),⁷ was discovered at Emory University and licensed to Glaxo. Antiviral drugs 7–9 all belong to the class of nucleoside reverse transcriptase inhibitors (NRTIs). They are also sometimes known as DNA chain terminators because of their MOA.

Due to the mitochondrial toxicities often associated with NRTIs such as 7–9,⁸ rigorous efforts have been made to search for non-nucleoside reverse transcriptase inhibitors (NNRTIs).⁹ The first three NNRTIs on the market at the end of 1990s were Boehringer–Ingelheim’s neviripine (Viramune, 10), Merck’s efavirenz (Sustiva, 11), and Upjohn/Pfizer’s delavirdine mesylate (Rescriptor, 12). Newer NNRTIs currently on the market as of 2014 include etravirine (Intelence) and rilpivirine (Edurant).¹⁰

Perhaps the most visible achievement against AIDS was the emergence of HIV protease inhibitors (PIs) in the mid-1990s. A protease is an enzyme that breaks peptide bonds of the virus’s proteins via hydrolysis. HIV protease is the enzyme that breaks the virus’s polyprotein peptide bonds via hydrolysis, thus rendering the HIV virions noninfectious.

The first HIV protease inhibitor on the market, Roche’s saquinavir (Invirase, 13), was approved by the FDA in 1995 for the treatment of AIDS.¹¹ It was quickly followed by Abbott’s ritonavir (Norvir)¹² and Merck’s indinavir (Crixivan).¹³ Nelfinavir (Viracept), amprenavir (Agenerase), fosamprenavir (Lexiva), lopinavir (Kelatra when combined with ritonavir), atazanavir (Reyataz), and tipranavir (Aptivus) have emerged since the 1990s. One of the latest HIV protease inhibitor is Merck’s darunavir (Prezista, 14). It was approved for the treatment of HIV/AIDS patients who are harboring drug-resistant HIV that does not respond to other therapies. The drug was discovered by Ghosh, who started the effort strictly as an educational project, at Purdue University.¹⁴

In addition to HIV reverse transcriptase inhibitors (including both NRTIs and NNRTIs) and HIV protease inhibitors, several other novel MOA have afforded successful and alternative treatments in the clinics. Merck’s raltegravir (Isentress, 15) is the first FDA-approved inhibitor of HIV integrase.¹⁵ Pfizer’s maraviroc (Selzentry, 16) is the first-in-class CCR5 antagonist for the treatment of HIV.¹⁶

1.2 Influenza Antiviral Drugs

Currently, there are four drugs available for the treatment or prophylaxis of influenza infections. The first two adamantanes include amantadine (Symmetrel, 17) and rimantadine (Flumadine, 18), which act as M₂ ion channel inhibitors and interfere with viral un-coating inside the cell.¹⁷ They are effective only against influenza A and are associated with several toxic effects as well as drug resistance. They are now rarely used.

The influenza neuraminidase (also known as sialidase) is one of two major glycoproteins located on the influenza virus membrane envelope. The other glycoprotein is hemagglutinin. The combination of the variants of neuraminidase and hemagglutinin is used to name influenza strains, for example, N1H1 and N9H7. Two newer drugs to treat influenza are both neuraminidase inhibitors: oseltamivir (Tamiflu, 19) and zanamivir (Relenza, 20). They are kinder and gentler in comparison to adamantanes 17 and 18. More important, they work better for both influenza A and B.

Oseltamivir (19), the first orally active neuraminidase inhibitor, is a prodrug. It was discovered by Gilead in 1995.¹⁸ Gilead and Roche began co-developing it in 1995 and gained FDA approval in 1999. Zanamivir (20) is given via inhalation due to its poor oral bioavailability. It was discovered by Biota Holdings, a small Australian biotechnology concern.¹⁹ In the United States, Biota established an alliance with GlaxoSmithKline for development and marketing of zanamivir (20), which was also approved by the FDA in 1999.

1.3 Hepatitis Antiviral Drugs

Hepatitis B virus (HBV)²⁰ causes hepatitis in both humans and animals. Interferon and five nucleos(t)ides have been approved as treatments for chronic hepatitis B²¹ in many parts of the world. Interferon, unfortunately, is effective only in a subset of HBV patients. Furthermore, it is often poorly tolerated, requires parenteral administration, and is expensive.²² Lamivudine (3TC, Epivir, 9) was the first oral antiviral drug for treating HBV. Tenofovir disoproxil fumarate (Viread, 21) is an orally administered phosphate prodrug of tenofovir, an NRTI that shows potent in vitro activity against both HBV and HIV-1. Current treatments for HBV also include adefovir dipivoxil (Hepsera, 22), Novartis’s telbivudine (Tyzeka, 23), BMS’s entecavir (Baraclude, 24), and the ever-versatile ribavirin (Rebetol, 25), which has been used to treat hepatitis C virus (HCV) as well.

Both 21 and 22 are prodrugs designed to boost the bioavailability of the parent drugs. For example, 21 is initially hydrolyzed by carboxyesterase to liberate one molecule of isopropanol and carboxylate 26, as shown in Scheme 1.²³–²⁵ Carboxylate 26 then spontaneously loses a molecule of CO₂ to provide another transient molecule 27, which spontaneously loses a molecule of formaldehyde to provide phosphate 28. Repeating the same sequence on the other side chain then delivers the active API as phosphoric acid 29. Similar to AZT, 29 as a mono-phosphate is converted in vivo to 29-ATP, which is the active species serving as a DNA chain terminator.

Hepatitis C, especially chronic hepatitis C, can severely damage the liver and cause liver carcinoma and eventually liver failure. More than 130 million people worldwide are infected with HCV.

Structurally, HCV has 10 proteins: three structural proteins and seven nonstructural (NS) proteins. As shown in Fig. 9.1,²⁶ all the HCV enzymes including NS2/3 and NS3/4A proteases, NS3 helicase, NS5A, and NS5B RNA–dependent RNA polymerase (RdRp) are essential for HCV replication and have proven to be attractive targets for the development of anti-HCV therapy.

Older therapies for HCV infection include pegylated interferon-α (PEG-Intron) alone or in combination with ribavirin (Rebetol, 25).²⁷ However, the antiviral activity of interferon is indirect, has to be given via IV, and possesses toxicities, often causing flu-like symptoms. Ribavirin (25) is a nonspecific agent with inhibitory activity toward some host proteins. Specifically targeted antiviral therapy for HCV would be more efficacious and have fewer side effects.²⁷ The resulting therapies include HCV NS2 and NS3/4 protease inhibitors, NS3 helicase inhibitors, NS4B and NS5A, and NS5B replication factor inhibitors, as well as HCV NS5B polymerase inhibitors. Sovaldi (1), the focus of this chapter, is a phosphoramide prodrug of an HCV NS5B polymerase inhibitor and belongs to the class of HCV NS5B polymerase inhibitors.

Fig. 9.1. Hepatitis C virus (HCV) genome and potential drug discovery targets.²⁶

Reproduced with Permission

Excitingly, two oral anti-HCV drugs, both HCV NS3/4A serine protease inhibitors, were approved by the FDA in 2011: boceprevir (Victrelis, 30),²⁸ by Schering–Plough/Merck, and telaprevir (Incivek, 31),²⁹ by Vertex. These two innovative medicines were the first wave of safe, efficacious, and convenient treatment options for patients with HCV. Two years later, Janssen’s simeprevir (Olysio, 32), a second-generation protease inhibitor for the treatment of genotype 1 (G1) HCV infection, was approved by the FDA.³⁰ It is also an HCV NS3/4A serine protease inhibitor. Sovaldi (1) was also approved in 2013. Sovaldi (1) and the three HCV NS3/4A serine protease inhibitors 30–32 are also known as direct-acting antivirals (DAAs).

■ 2 Pharmacology

2.1 Mechanism of Action

Sovaldi (1) has a pan-genotypic antiviral effect although it might be less efficient in genotype 3 (G3). As a phosphoramide prodrug, it is not active in vitro. Instead, it undergoes extensive metabolisms in vivo under the influence of a battery of enzymes in the human body via a relatively complex activation pathway. The end product is a uridine triphosphate analog PSI-7409 (35). It is a potent HCV NS5B polymerase inhibitor in vitro. This active species is nucleotide analog (chain terminator) and is responsible for the drug’s antiviral effects.

The inventor Pharmasset’s initial significant contribution to the discovery of Sovaldi (1) was installation of the fluorine atom, affording a series of fluorine-containing nucleosides with unique in vitro and in vivo characteristics.³¹ One of the company’s early leads was 2′-α-fluoro-2′-β-methyl nucleoside PSI-6130 (33), which is a potent and selective inhibitor of HCV NS5B polymerase, which exhibited antiviral activity in cell culture systems with an EC₉₀ of 4.6 µM in an HCV replicon assay and was efficacious in humans.³² EC₉₀ is the efficacious concentration that provokes a response that is 90% of the maximum; the smaller the EC₉₀ value, the more potent the compound is.

Although PSI-6130 (33) was the first of this class to advance to clinical trials, it suffered from poor bioavailability. PSI-6206 (34), a metabolite of 33 by deamination of the cytosine, had a better PK profile. Despite the fact that PSI-6206 (34) is inactive in vitro (EC₉₀ > 100 µM), its corresponding triphosphate form, PSI-7409 (35), is a potent inhibitor of HCV polymerase in in vitro assays.

The long and complicated metabolic pathway from Sovaldi (1) to its active form PSI-7409 (35) is shown in Scheme 2.³³–³⁵ At first, Sovaldi (1) is converted to corresponding carboxylic acid 36 under the influence of human cathepsin A (CatA) and carboxylesterase 1 (CES1). Carboxylic acid 36 then undergoes a fast, non-enzymatic intramolecular nucleophilic attack to form a cyclic alaninyl phosphate intermediate 37, which undergoes a hydrolysis of the cyclic phosphate to linear phosphate as carboxylic acid 38 with concomitant release of a molecule of phenol. Carboxylic acid 38 is further hydrolyzed by the histidine triad nucleotide-binding protein-1 (Hint1) enzyme to form monophosphate nucleotide 39. It is phosphorylated to the diphosphate nucleotide 40 under the influence of uridine monophosphate-cytidine monophosphate (UMP-CMP) kinase. Finally, 40 is further phosphorylated to the active triphosphate nucleotide PSI-7409 (35) with the aid of nucleoside diphosphate kinase (NDPK).

2.2 Structure–Activity Relationship

Extensive SAR investigations were carried out with regard to the optimal substituents for the phosporamidate.³⁶ The final fine-tuning section of the SAR is summarized below (Table 9.1).

Pharmasset assessed their prodrugs’ anti-HCV activity using the clone A replicon and a quantitative real-time PCR assay measured the EC₉₀. Each compound was also simultaneously evaluated for cytotoxicity by assessing the levels of cellular rRNA at 50 μM. Thus 0% would indicate the compound was devoid of cytotoxicity.

As shown in Table 9.1, when the ester moiety (R¹) on 41 was Me, Et, i-Pr, or Cy, compound 41 had submicromolar EC₉₀ values. Although R¹ as n-Bu, i-Bu and n-pentane (not shown) afforded even more potent compounds; they also showed cytotoxicity and were thus not further pursued. Next, the α-substitution on alanine was altered.

As far as the R² on phosphoramidate 41 was concerned, 1-naphthyl ester (not shown) was the most potent compound, but its cytotoxicity (95.4% at 50 μM) was even higher than that of 41p (91.5% at 50 μM). On the other hand, when R² was an alkyl group, the resulting compounds were not active. Therefore, the efforts of optimizing R² focused on simple phenyl or mono-halogenated phenyl substituents. Due the potential toxicities associated with poly-halogenated phenols, they were not considered.

Table 9.1. HCV Replicon Activity of Phosphoramidate Prodrugs: Simultaneous Carboxylate and Phenolic Ester Modification of the Phosphoramidate Moiety³⁶

Stay updated, free articles. Join our Telegram channel

Join