® Check for updates
`
`Review
`
`Antiviral Chemistry 8 Chemotherapy 2011;22:23—49 (doi: 10.3851/IMP1797)
`
`Nucleotide prodrugs for HCV therapy
`
`Michae/ J Sofia”
`
`1Pharmasset, Ine, Princeton, NJ, USA
`
`*Corresponding author e-mail: michae|.sofia@pharmassetcom
`
`
`HCV infection is a significant worldwide health problem
`and is a major cause of hepatocellular carcinoma. The
`current standard of care. interferon and ribavirin. is only
`effective against a proportion of the patient population
`infected with HCV. To address the shortcomings of exist—
`ing therapy. the development of direct acting antiviral
`agents is under investigation. The HCV RNA dependent
`RNA polymerase is an essential enzyme for viral replica—
`tion and is therefore a logical target against which to
`develop novel anti—HCV agents. Nucleosides have been
`shown to be effective as antiviral agents for other viral
`diseases and therefore. have been investigated as inhibi—
`tors of HCV replication. The development of prodrugs
`of nucleoside 5’—monophosphates has been pursued
`
`to address limitations associated with poor nucleoside
`phosphorylation. This is required to produce the nucle—
`oside 5'—triphosphate which is the anabolite that is the
`actual
`inhibitor of the polymerase enzyme. Prodrugs of
`nucleoside 5’—monophosphates have been developed that
`enable their delivery into cells and in vivo into the liver.
`The implementation of these prodrug strategies has ulti—
`mately led to the identification of several prodrugs of
`nucleoside 5’—monophosphates that are potent
`inhibi—
`tors of HCV replication in vitro. They have progressed into
`the clinic and the early data demonstrate greatly reduced
`viral load levels in HCV—infected patients. This review will
`survey the state of nucleotide prodrugs for the treatment
`of HCV.
`
`Introduction
`
`HCV is known to have infected approximately 180
`million individuals worldwide [1]. It is estimated that
`of those infected with HCV approximately 80% will
`develop chronic liver disease and a significant propor—
`tion of those infected will eventually develop liver cir-
`rhosis and subsequently hepatocellular carcinoma [2].
`HCV is a single-stranded, positive sense RNA virus of
`the Flaviviridae. Six major viral genotypes with over
`100 viral subtypes have been identified for HCV. Geno-
`types la and 1b are the most prevalent genotypes in
`the western world with genotypes 2 and 3 comprising
`20—30% of this population. HCV genotypes 2—6 pre-
`dominate in the developing world [3,4]. Because HCV
`replicates in the cytoplasm of infected cells by a mem-
`brane associated replication complex and the virus has
`an RNA genome with no DNA intermediate during
`replication, no genomic templates are stably integrated
`into the host genome. Therefore, virological cures are
`possible for HCV patients. This is in contrast to other
`viruses such as HIV and HBV where the viral genome is
`integrated into the host DNA and a virological cure is
`considered remote. However, for HCV-infected patients
`virological cures are made difficult due to the high rate
`of HCV viral replication and by the high spontaneous
`
`©2011 International Medical Press 1359-6535 (print) 2040—2066 (online)
`
`mutation rate of the virus. This high mutation rate is a
`result of poor replication fidelity exhibited by the HCV
`polymerase and an apparent lack of proof reading [4].
`The current therapy for treating chronic HCV infec—
`tion consists of regular injections of ot-interferon (IFN)
`with daily oral administration of ribavirin (RBV). This
`standard of care (SOC) regimen does not act by directly
`attacking the virus but functions by boosting the host
`immune response. For genotype 1 patients regular IFN/
`RBV treatments for 48 weeks result in only 40—50% of
`patients achieving a sustained virological response (SVR)
`indicative of a cure [5,6]. However, for genotype 2 and
`3 patients the SVR rates can be as high as 75%. It is also
`known that subpopulations which include individuals of
`African ancestry tend to respond less well to IFN/RBV
`treatments [7]. Recent genome-wide association studies
`have shown that a single nucleotide polymorphism 3kb
`upstream of the IL28B gene correlates with a significant
`difference in response to IFN therapy [8]. IL28B which
`encodes the type III interferon IFN-A—3, is known to be
`upregulated by IFNs and by RNA viral infections. It has
`been shown that HCV patients who harbour a TT or TC
`allele in their ILZSB gene tend to respond less well to IFN/
`RBV treatment than do those having the CC genotype.
`
`Gilead 2005
`
`23
`
`I-MAK v. Gilead
`
`lPR2018—00125
`
`

`

`MJ Sofia
`
`Figure 1. Model of HCV NSSB RNA complex
`
`Fingers
`
`Ribbon structure of the HCV NSSB RNA dependent RNA polymerase showing the
`Palm. Finger and Thumb domains typical of a RNA polymerase. Also depicted is a
`bound template-primer strand of RNA and the nucleotide binding site.
`
`Patients who choose to undergo IFN/RBV therapy face
`not only the possibility of not responding to treatment
`but also must contend with the potential for multiple and
`sometimes serious side-effects that include influenza-like
`
`symptoms, fatigue, hemolytic anaemia and depression.
`The intolerable side effects can result in a high rate of
`drug discontinuations. Consequently, the modest cure
`rates and subpopulation differences combined with the
`side-effect profile for SOC have prompted an urgency to
`develop alternative novel, safe and effective therapies.
`As in the case of other viral diseases, the development
`of direct acting antivirals (DAAS) has become a focus.
`Development of small molecule agents to attack essen-
`tial viral proteins has the potential benefit of reducing
`toxicities and side effects associated with manipulating
`host functions and hopefully such DAA therapies will
`not have the intolerable side effects exhibited by cur-
`rent SOC.
`
`The push to identify small molecule DAAs has also
`prompted the discussion around the possibility of elimi-
`nating or at least reducing the use of IFN/RBV from
`treatment regimens. Although clinical development of
`first generation DAAs has focused on combinations
`with SOC in the hope of both shortening duration of
`treatment and increasing the cure rate, the long-term
`desire is to completely eliminate the use of IFN/RBV
`from treatment regimens. Clearly this is an aspirational
`goal and a goal that can only be achieved if an immune
`component of therapy is not absolutely required to
`eliminate those vestiges of undetectable virus
`[9].
`In addition, such a lofty goal can be realised if small
`
`24
`
`molecule DAAs either alone or more probably in com-
`bination can drive viral loads to undetectable limits and
`maintain those undetectable levels over the necessary
`course of therapy and after cessation of treatment with—
`out having viral breakthrough resulting from the emer-
`gence of resistant virus. As has been shown with HIV
`highly active antiretroviral therapy (HAART), the HCV
`treatment paradigm will likely require combinations of
`anti-HCV agents [7]. The desire is to suppress virus as
`rapidly and completely as possible in order to give the
`body’s natural immune system the opportunity to clear
`residual virus and to hold back emergence of resistant
`virus. However, what will comprise those ideal com-
`binations of DAAs is yet to be determined and studies
`to clarify this question are under active discussion and
`investigation.
`HCV has a 9.6 kb genome of positive-stranded RNA.
`This genome encodes a precursor polyprotein that is
`processed into 10 functional proteins: three structural
`proteins and seven non—structural proteins [10]. Sev-
`eral of the non-structural proteins have been the focus
`of intensive efforts to identify small molecule DAA
`agents as inhibitors of HCV replication. Of the seven
`non-structural proteins, molecules that inhibit the func-
`tions of the NS3/4 protease, NS4A, NS4B, NSSA and
`N55 B RNA dependent RNA polymerase (RdRp) have
`advanced to the clinic [11—15]. The most advanced
`agents are the N53/4 protease inhibitors telaprevir and
`boceprevir. Each of these compounds has completed
`Phase III clinical
`investigation and both have been
`shown to be efficacious in treating HCV infection when
`given in combination with SOC. However, each of these
`first generation protease inhibitors suffers from the lack
`of genotype coverage, undesired side effects, that may
`limit their usage, and the early emergence of resistant
`virus. It is therefore not surprising that even with the
`potential benefits of these first generation DAAs, ware—
`housing of patients by physicians occurs in order to wait
`for approval of more effective and tolerable agents.
`The HCV RNA dependent RNA polymerase is an
`~68 kd protein that has the typical palm-finger—thumb
`structural motif found in many viral polymerases (Fig-
`ure 1) [16,17]. HCV polymerase is an essential enzyme
`involved in RNA replication. Phylogenetic analysis
`shows a 65% homology of HCV RdRp across genotypes
`and an 80% homology within a particular genotype [3].
`The HCV polymerase active site is located in the palm
`domain where the conserved aspartic acid residue-con-
`taining GDD motif is located [18]. This conserved GDD
`motif is common to viral polymerases in general [18].
`Through a divalent metal ion (Mg"* or Mn“) the GDD
`motif functions to coordinate the binding of the ribonu-
`cleoside triphosphate. The HCV polymerase catalyzes
`the addition of a single ribonucleoside triphosphate
`monomer to the 3’—end of the growing RNA chain by the
`
`©2011 International Medical Press
`
`

`

`formation of a 3’,5’—phosphodiester linkage. To accom-
`plish this process, the polymerase must simultaneously
`bind a template RNA strand, a primer RNA strand and
`a ribonucleoside triphosphate monomer [19,20]. There—
`fore, investigation of nucleoside analogues is a rational
`choice for the development of inhibitors of the HCV
`NSSB polymerase.
`Of all the DAAs under clinical investigation, nucleo-
`side/nucleotide NSSB polymerase inhibitors hold the
`promise of pan-genotype coverage and a high barrier
`to development of resistant virus. As in the case of HIV
`infection where nucleosides have become the backbone
`
`of therapy (for example, TRUVADA® and Combivir®),
`HCV nucleosides/nucleotides are positioned to assume
`a similar role. To date, only nucleosides/nucleotides
`have demonstrated broad genotype coverage both in
`the laboratory and in human clinical studies [21]. In
`addition, to date, no pre-emergent resistant virus has
`been detected in clinical studies [22]. It is for these
`reasons that nucleosides/nucleotides are positioned to
`play a prominent role in developing HCV treatment
`paradigms.
`The HCV polymerase has been shown to be a
`uniquely selective polymerase as it relates to the devel-
`opment of nucleoside/nucleotide inhibitors. Over
`the last 10 years only two broad classes of nucleo—
`sides have emerged as inhibitors of this polymerase
`[23—25]. These include the 2’-methyl and the 4’-azido
`classes (Figure 2). However,
`these classes and sub-
`groups within these classes have clearly differentiated
`themselves in preclinical and clinical studies. This
`differentiation is exemplified in their viral selectivity,
`viral resistance, overall safety and clinical efficacy pro—
`files. Resistance associated with the 2’-methyl class of
`nucleosides is associated with the 8282T amino acid
`
`alteration located in the finger domain of the HCV
`polymerase [26—29]. This mutation has been shown to
`be difficult to raise in vitro and has not been detected
`
`isolates [22].
`as a pre—existing mutation in clinical
`Similarly, for the 4'-azido class, the S96T amino acid
`alteration has been identified in vitro but has not been
`
`observed in the clinic [27].
`Implementation of a prodrug strategy has played
`a prominent role in the development of a number of
`nucleosides and nucleotides for the treatment of HCV
`
`infection [23,30]. These prodrugs have been developed
`to overcome, not only, bioavailability and stability
`issues but also to address key anabolism limitations
`important to nucleoside activation. Simple ester prod—
`rugs of both the 2’-methylcytidine, 2’-0t-fluoro-2’—B-C-
`methylcytidine and 4’-azidocytidine nucleosides were
`developed to overcome both bioavailability issues and
`to curb undesirable metabolism [21,3 1,32]. Prodrugs of
`the phosphate group of nucleoside 5’-monophosphates
`were developed to address not only bioavailability
`
`Antiviral Chemistry Er Chemotherapy 22.1
`
`Nucleotide prodrugs for HCV therapy
`
`Figure 2. Two major classes of nucleosides that are known to
`be inhibitors of HCV polymerase.
`
`2’-Methyl class
`
`4’-Azido class
`
`BASE
`
`BASE
`
`HO
`
`O
`
`Ho"
`
`’x
`
`X = OH, F
`
`HO
`
`N3“
`
`_. 0
`
`Ho‘
`
`’x
`
`X = OH, F
`
`These classes include the 2’-methy| and the 4’—azido ribosides.
`
`issues but also poor in vitro and in vivo conver—
`sion of the parent nucleoside to the active nucleoside
`5’-triphosphate [33]. Because nucleosides must be con-
`verted to their 5’—triphospates to be active as inhibitors
`of the HCV polymerase, they need to undergo a series
`of phosphorylation steps catalyzed by three separate
`kinases (Figure 3). These kinases convert the nucleoside
`first to the monophosphate, then to the diphosphate
`and finally to the active triphosphate. However, it is
`not uncommon that in the phosphorylation cascade, a
`nucleoside or its corresponding mono- or diphosphate
`is a poor substrate for one of the kinases. In particular,
`it is the first kinase in the phosphorylation cascade that
`is generally the most substrate selective. Therefore, it
`is not unusual that bypassing the first kinase results in
`achieving high levels of the active triphosphate. Because
`nucleoside monophosphates are enzymatically dephos-
`phorylated and negatively charged, they do not readily
`enter cells and therefore are not desirable as drug can—
`didates. To overcome the limitations of administering
`a nucleoside monophosphate-containing agent, pro-
`drugs of the 5'—monophosphate nucleoside have been
`employed. Prodrugs of nucleoside monophosphates
`have been known for many years and a number of
`phosphate prodrug strategies have been developed to
`address the need to deliver a 5’-monophosphate nucleo-
`side into the cell [33—35]. However, there have been few
`examples where a nucleotide prodrug has been shown
`to deliver the corresponding 5’—monophosphate in vivo
`to the desired site of action [33]. Often the prodrug
`moiety decomposes prior to achieving its objective
`because of either chemical or enzymatic instability in
`the gastrointestinal tract and/or plasma.
`The development of a nucleoside phosphate prod-
`rug useful for the treatment of HCV faces several chal-
`lenges. The nucleoside phosphate prodrug must have
`sufficient chemical stability to be formulated for oral
`
`25
`
`

`

`MJ Sofia
`
`Figure 3. Nucleoside kinase activation pathway
`
`
`
`
`
`HO
`
`BABE:
`0
`V Y
`'P" x
`_
`'Y
`
`HO‘
`
`—-—b-
`
`Kinase 1
`
`ONO BASE
`\?
`"i
`f... x
`HO:
`;Y
`
`Kinase 2
`
`BASE
`
`Kinase 3
`
`BASE
`
`=Phosphate
`
`Nucleoside kinase activation pathway resulting in the nucleoside triphosphate which is the active substrate for a polymerase allowing incorporation of the nucleoside
`or nucleoside analogue into the growing RNA chain and thus resulting in inhibition of virus replication. Examples of kinases 1, 2. and 3 include deoxycytidine kinase
`(de). nucleoside monophosphate kinase (YMPK) and nucleoside disphosphate kinase [NDPK], respectively.
`
`administration. It must be stable to conditions of the
`
`gastrointestinal tract such that the prodrug reaches the
`site of absorption intact. The prodrug must have good
`absorption properties and must not undergo apprecia—
`ble enzymatic degradation during the absorption phase.
`Once absorbed, the prodrug needs to have sufficient sta-
`bility in the blood in order to reach the target organ: for
`example the liver in the case of HCV. The prodrug must
`then be transported into hepatocytes and release the free
`5’-monophosphate nucleoside which can subsequently
`be converted to the active triphosphate derivative. Since
`HCV is a disease of the liver, and the liver is the first
`organ the prodrug encounters after absorption, HCV is
`an ideal disease for which to develop a targeted nucle-
`otide prodrug strategy. Consequently, several of these
`nucleoside monophosphate prodrugs have advanced
`to the clinic and have demonstrated proof of concept
`for treating HCV. Here, the application of phosphate
`prodrugs in the development of nucleotide inhibitors
`of HCV and the status of nucleotide prodrugs under
`investigation for the treatment of HCV infection will
`be reviewed.
`
`Nucleotide phosphoramidates
`Nucleotide phosphoramidates were first disclosed
`by McGuigan et al. [36] as a prodrug strategy to
`deliver a nucleoside 5’—monophosphate for the treat-
`ment of HIV and cancer. The structure of a nucle-
`
`otide phosphoramidate typically consists of a nucleo-
`side 5’-monophosphate where the phosphate group
`is masked by appending an aryloxy group (usually a
`phenol) and an (IL-amino acid ester (Figure 4); how-
`ever, other related constructs have also appeared. The
`
`26
`
`phosphate group is ultimately revealed by a sequence
`of enzymatic and chemical steps that requires either
`carboxyesterase or cathepsin A to cleave the terminal
`amino acid ester, intramolecular displacement of the
`phosphate phenol and then enyzmatic cleavage of the
`amino acid moiety by a phosphoramidase or histidine
`triad nucleotide-binding protein 1 (HINT 1) [37—39].
`It
`is believed that
`the phosphoramidate prodrug
`construct
`increases lipophilicity of the nucleoside
`5’-monophosphate and therefore increases cellular
`permeability and ultimately intracellular nucleotide
`concentrations. Since the phosphoramidate prodrug
`moiety contains a chiral phosphorus centre,
`issues
`arise with regard to development of a compound that
`consists of a mixture of isomers with implications
`arising from differential activity of each of the iso-
`mers, pharmacokinetics and manufacturing optimi-
`zation, etc. In addition, the typical phosphoramidate
`contains a phenolic substituent that is released during
`metabolism to the free monophosphate. Successful
`development also considers the metabolic release of
`this phenolic substituent. Selective examples employ-
`ing the phosphoramidate strategy to achieve kinase
`bypass for nucleoside inhibitors of HIV reverse tran—
`scriptase (RT) inhibitors, for example, ddA, d4T and
`d4A, showed that in vitro whole cell enhancement
`in potency could be achieved [40—42]. Although the
`phosphoramidate strategy was explored extensively to
`deliver nucleotides for the treatment of HIV and colon
`
`cancer [33,36], proof of concept in the clinic has yet
`to be reported. However, the phosphoramidate prod-
`rug approach has proven to be a valuable strategy in
`the development of HCV nucleotide therapy.
`
`©2011 International Medical Press
`
`

`

`Figure 4. The phosphoramidate prodrug decomposition pathway that results in the release of the nucleoside 5'-monophosphate
`
`Nucleotide prodrugs for HCV therapy
`
`on
`>\._-\’
`o
`HN—P—O
`I
`I
`
`R1
`
`(i)
`Ar
`
`Nucleotide phosphoramidate
`
` Carboxyesterase
`or
`
`CathepsinA
`
`9H
`
`(I)
`Ar
`
`Spontaneous
`
`R2,
`
`if
`H
`N—IT—O
`yoI
`
`O
`
`
`H20
`
`HINT-1
`
`
`(Phosphoramidase)
`
`fl
`HO-r-OOH
`
`R2
`
`0I
`
`I
`HN—P—o
`|
`9H
`
`0
`
`o
`l
`”
`
`2’—C—Methyl ribonucleotide phosphoramidates
`The 2’-C—methylcytidine nucleoside, NM107 (1; Fig-
`ure 5), was shown to be an inhibitor of HCV in cell
`
`culture (50% effective concentration [ECso] =1.23 M)
`and its triphosphate (3) was demonstrated to be a
`potent inhibitor of the HCV polymerase enzyme (50%
`inhibitory concentration [IC50] =0.09—0.18 HM) acting
`as a nonobligate chain terminator [43]. NM107 also
`showed broad antiviral activity against not only HCV
`but also bovine virus diarrhoea virus (BVDV), yellow
`fever virus, dengue virus and West Nile virus [31]. To
`overcome bioavailability issues the 3’-valinate ester
`prodrug, NM283 (valopicitabine; 2) [31,44], was taken
`into clinical development. In a Phase I monotherapy
`study, NM283 demonstrated proof of concept deliver-
`ing an ~1.2 log10 IU/ml reduction in viral load at a dose
`of 800 mg twice daily given over 14 days. Unfortu-
`nately, NM283 was discontinued because of significant
`gastrointestinal toxicity in Phase II studies [24,30].
`Subsequent work on the development of the 2’-C-
`methyl class of nucleosides focused on 5’-phosphate
`nucleotide prodrugs. It was observed that the 2'-C-
`methylcytidine triphosphate (3) was highly active as
`an inhibitor of the NSSB polymerase, yet the parent
`
`Antiviral Chemistry 8t Chemotherapy 22.1
`
`nucleoside NM107 (1) was only modestly active in the
`whole cell based replicon assay. Studies had shown that
`NM107-triphosphate (3) formation was inefficient,
`particularly because of poor conversion of the nucleo-
`side to its monophosphate by 2'-deoxycytidine kinase
`[45]. Consequently, to circumvent this phosphoryla-
`tion problem and potentially improve the therapeutic
`index by increasing nucleoside triphosphate levels in
`the liver, a phosphoramidate prodrug approach was
`investigated [45]. This effort lead to the identification
`of phosphoramidate derivative 4 (Figure 5; ECSOSOJ
`uM) showing substantial increases in potency rela—
`tive to NM283 [45]. The activity of compound 4 cor-
`related with the levels of triphosphate produced in
`human hepatocytes and these levels were shown to be
`much higher than that seen with NM283 (2). In vivo
`studies assessing liver nucleoside triphosphate levels
`after oral administration in hamsters showed low tri-
`
`phosphate concentrations only twofold higher than
`obtained with NM283. Since substantial
`liver tri-
`
`phosphate levels were seen after subcutaneous admin-
`istration in vivo and the compounds were shown to
`be stable in simulated gastric fluid, it was concluded
`that low oral bioavailability or metabolic degradation
`
`27
`
`

`

`MJ Sofia
`
`
`Figure 5. 2’—C-Methylcytidine nucleosides and nucleotide phosphoramidate prodrug inhibitors of HCV replication
`
`NH2
`
`NH2
`
`/
`
`W
`N
`0 N\\<
`O
`
`HO
`
`i
`Hci
`
`-.
`”OH
`NM107
`
`5050:123 pM
`1
`
`/
`
`W
`N
`0 Njf
`o
`
`,’
`’OH
`
`HO
`
`NHZ
`
`o
`
`Kt
`
`0
`
`H
`
`1
`
`NM283
`
`2
`
`NHZ
`
`/ '17
`I.
`IN
`O
`O
`O
`N-_/
`_||_ _n_ _n_ ,—
`HOFI’OFI’OIIDOMOV‘III
`OH
`OH
`OH
`'I.
`.' p.
`O
`
`HO
`
`’OH
`
`|C50=O.09—0.1 8 uM
`3
`
`ECSO=0.22 pM, CC5D=7 uM
`4
`
`EC50=0.24 11M
`5
`
`0 N
`i
`OWE-i-O’xjfi
`
`O
`
`OH
`
`9 N
`OVW-i-Oyjfig
`
`OH
`
`O
`
`O
`
`Hd
`
`’OH
`
`Hd
`
`’OH
`
`NTP (human hepatocytes) AUCMm=190 pMOh
`
`7
`
`ECSO=8.2 pM, CCSO=>100 pM
`NTP (human hepatocytes) AUC(Hh)=1,720 oM-h
`6
`
`
`
`AUC, area under the curve; CC”. 50% cytotoxic concentration; ECW 50% effective concentration; Kim, 50% inhibitory concentration,
`
`28
`
`©2011 International Medical Press
`
`

`

`Nucleotide prodrugs for HCV therapy
`
`in the intestine was the reason for the lack of oral
`
`efficacy [45].
`Another series of 2’-C-methylcytidine phosphorami-
`date prodrugs having an acyloxyethylamino phospho-
`ramidate promoiety was studied (Figure 5) [46]. The
`phosphoramidate prodrug 5 provided up to a 30-fold
`improvement in HCV replicon potency over NM283
`(2), and it was also shown that this activity correlated
`to levels of nucleoside triphosphate in rat and human
`hepatocytes. However, when administered orally to rats
`these phosphoramidates demonstrated no improvement
`in production of triphosphate in rat liver relative to
`NM283 (2). These results put into question the oral
`bioavailability and conversion of these prodrugs to the
`nucleoside triphosphate in the liver.
`Phosphoramidate monoesters of 2’—C—methylcytidine
`were also explored (Figure 5) [45]. In this case the ami—
`date moiety was either an ot-amino acid (6) or an acy-
`loxyethylamino substituent (7). Although phosphorami-
`date monoester 6 was shown to have inferior replicon
`potency relative to its phenolic ester counterpart (<200-
`fold) it showed higher levels of triphosphate formation
`in human hepatocytes. Formation of nucleoside triphos-
`phate levels were observed in hepatocytes of various spe-
`cies for both 6 and 7 but the phosphoramidate monoester
`
`6 was shown to be superior to 7. In both cases, liver levels
`of nucleoside triphosphate 3 were achieved in vivo after
`subcutaneous administration but not after oral adminis-
`
`tration suggesting a lack of oral bioavailability.
`Other 2’-C-rnethyl nucleosides containing purine
`bases were also investigated as inhibitors of HCV,
`including the 7-deaza-2’-C-methyladenosine derivative
`MK0608 (8) which showed potent inhibition of HCV
`replication in vitro (EC50=0.25 HM) and demonstrated
`SVR in an HCV-infected chimpanzee animal model
`[47,48]. Although MK0608 (8) was never progressed
`into clinical development, it did demonstrate the poten-
`tial of purine nucleosides as inhibitors of HCV. Sub-
`sequently, phosphoramidate prodrugs of 2’-C—methyl
`purine analogues were studied to determine if improve-
`ment in potency could be achieved by kinase bypass.
`Although application of the phosphoramidate prodrug
`strategy was not successful for the adenosine derivative
`11 (Figure 6), its application to the guanosine analogue
`12 resulted in an 84-fold increase in activity in the rep—
`licon assay relative to the parent nucleoside (Figure 7)
`[49]. This result was rationalized by the observation
`that the guanosine triphosphate (13; IC50=0.13 uM)
`was more potent as an inhibitor of the HCV polymer—
`ase than the adenosine triphosphate (10;
`IC50=1.9
`
`
`
`Figure 6. 2’-C—Methy|adenosine nucleoside and nucleotide phosphoramidate prodrug inhibitors of HCV replication
`
`HO"
`
`"OH
`
`no“
`
`”or.
`
`ECSO=O.25 uM, CC50=>50 uM
`8
`
`EC5°=O.3 uM, ccsu=>5o uM
`9
`
`NH2
`N
`x _
`” f~£
`o
`o
`o
`M
`{if
`n
`II
`II
`Pl
`HO—Ff—O—lID—O—T—O’VDY
`N5"
`
`OH
`
`OH
`
`OH
`
`.
`
`.--
`
`Etc
`
`9H3
`
`N
`NH2
`/
`ingot oE /\N
`VH (I)
`\/ Z
`’J
`
`HCS
`
`’OH
`
`ICSD=1.9 uM, EC50=O.3 MM
`10
`
`ECSD=0.25 uM, ccso=>5o uM
`11
`
`
`
`CCSO, 50% cytotoxic concentration; EC“. 50% effective concentration; Kim, 50% inhibitory concentration.
`
`Antiviral Chemistry 8 Chemotherapy 22.1
`
`29
`
`

`

`MJ Sofia
`
`Figure 7. 2’-C—Methylguanosine nucleoside and nucleotide phosphoramidate prodrug inhibitors of HCV replication
`
`N
`f /
`N
`
`O
`NH
`N4
`
`NH2
`
`HO“
`
`0
`
`\/
`-.
`
`o
`o
`0
`HO—iDI—O—lDI—O—lDI 0*
`I
`I
`|
`OH
`OH
`OH
`
`N
`
`(/
`N
`
`/
`
`O
`NH
`N94
`
`NH2
`
`o
`
`Ho"
`
`’OH
`
`E05535 uM
`12
`
`n2
`F
`R10m ”AH
`ll/
`0
`
`0
`H
`HN—T—O
`0
`
`_
`
`/N
`( /
`N
`
`0
`
`NH
`N¢<
`NH2
`
`o
`
`/
`"H.
`
`HO
`
`’OH
`
`Hd
`
`’0H
`
`|C50=0.13 uM
`13
`
`54
`
`CH
`O
`3
`?
`O A ll
`\ll HN-P—0/\/o
`I
`0
`o
`.,
`I
`'
`HO:
`
`N
`
`(/
`N
`Z:
`‘OH
`
`OCHa
`\
`N
`,4
`
`NH2
`
`/
`
`N
`
`14 R1=CH2CH2Ph, R2=Ala EC50=0.08 uM, ccso=>1oo uM
`15 R‘:o-C|Ph, R2=Val EC50=0.43 uM, ccsu=>100 uM
`
`lNX—08189
`ECSD=O.010 uM, ccm=7 uM
`16
`
`o
`
`N
`
`IDX184
`
`E05504 IIM, CCSD>1OO 1.1M
`17
`
`
`
`CC”, 50% cytotoxic concentration; EC“. 50% effective concentration; ICED. 50% inhibitory concentration.
`
`uM), yet in the whole cell replicon assay the adenosine
`analogue (9; EC50=0.3 uM; Figure 6) was more potent
`than the guanosine analogue (12; EC50=3.5 uM; Figure
`7). Therefore, this finding coupled with the observa—
`tion that low levels of triphosphate were detected in
`cells for the guanosine nucleoside relative to that for
`the adenosine nucleoside hinted at poor phosphoryla-
`tion in the case of the guanosine nucleoside. The devel-
`opment of the 2’-C—methylguanosine phosphoramidate
`derivative was further explored by systematically eval-
`uating structural modifications to the phosphorami-
`date moiety in an effort to achieve increased HCV rep-
`licon potency, plasma stability across multiple species,
`and appropriate relative stability in intestinal and liver
`S9 preparations. Ultimately,
`these studies were able
`to identify 2’-C—methylguanosine phosphoramidate
`
`30
`
`prodrugs 14 and 15 (Figure 7) [50]. These prodrugs
`contained benzyl or alkyl L-alanine and L-valine amino
`acid ester moieties and a naphthyl phosphate ester and
`exhibited a 10—30-fold enhancement in HCV repli-
`con potency with acceptable plasma and intestinal S9
`stability suitable for progression into in vivo studies.
`However, when mice were dosed orally in order to
`assess liver levels of the nucleoside triphosphate 13, the
`2’—C—methylguanosine phosphoramidates 14 and 15
`(Figure 7) produced substantial liver levels of triphos-
`phate but these levels were not significantly improved
`over that observed when mice were dosed with the par-
`ent guanosine nucleoside 12.
`Further investigation of the 2’-C—methylguanosine
`phosphoramidate series led to the evaluation of substi—
`tution at the C-6-position of the guanosine base with the
`
`©2011 International Medical Press
`
`

`

`Nucleotide prodrugs for HCV therapy
`
`intention of increasing lipophilicity and thus improving
`cellular uptake relative to the natural guanosine deriva-
`tive 12. This led to the identification of the C—6-O-methyl
`derivative INX—08189 (16; Figure 7) containing both a
`neopentyl ester on the L-alanyl amino acid moiety and
`a naphthyl ester on phosphorus of the phosphorami—
`date [51]. In INX-O8189 (16), in addition to metabolic
`conversion of the phosphoramidate pro-moiety to the
`5’-monophosphate, the 6-O-methyl group of the purine
`base is metabolized to the guanine base. INX-08189
`(16) demonstrated exceptional potency in the HCV 1b
`replicon assay (EC50=0.01 HM; CC50=7 uM), was active
`against genotype 1a and 2 replicons and produced sub-
`stantial levels of the guanosine triphosphate 13 in pri—
`mary human hepatocytes over 48 h. The known 2’-C—
`methyl nucleoside SZ82T mutant replicon was shown
`to be moderately resistant (3- to 10-fold) to INK-08189
`(16) [52]. No difference in HCV replicon potency was
`observed for each of the individual diastereoisomers of
`
`INK—08189 (16), thus INX—08189 was advanced into
`clinical development as a mixture of isomers at the
`phosphorus centre of the prodrug.
`In a single-ascending-dose Phase la study in healthy
`volunteers administered doses ranging from 3 mg to
`100 mg, INX-08189 (16) was shown to be generally
`well tolerated at all closes with no drug-related serious
`adverse events and pharmacokinetics (PK) supporting
`once daily oral dosing [53,54]. INX-O8189 was pro-
`gressed into a Phase Ib study in treatment—naive geno-
`type 1 HCV patients dosed once daily at either 9 mg
`or 25 mg. Antiviral activity was observed with a mean
`HCV RNA reduction of -0.71 and -1.03 log10 IU/ml,
`respectively [54,55]. These early human data demon-
`strated clinical proof of concept for INK-08189 and
`consequently, the compound is continuing to undergo
`clinical evaluation.
`
`Another application of phosphoramidate prodrug
`technology for
`delivering a
`2’-C—methylguanosine
`5’-monophosphate is exemplified by IDX184 (17; Figure
`7) [56]. In this case the phosphoramidate prodrug moi—
`ety utilized a benzyl amine in place of an amino acid for
`the amine substituent and an S-acetyl—2-thioethyl moiety
`(SATE) as the phosphate ester substituent in place of the
`more common aryl group. The SATE group is a known
`phosphate ester prodrug construct [33,57,58]. Mecha-
`nism for release of the desired 2’-C-methylguanosine
`5’-monophosphate is believed to involve both CYP450-
`dependent and independent processes. Based on the pro-
`drug structure one would anticipate that prodrug release
`would involve cleavage of the terminal thioester fol-
`lowed by loss of the phosphate ester via intramolecular
`attack of the free thiol group releasing the phosphate and
`an equivalent of episulfide, then enzymatic cleavage of
`the benzyl phosphoramide. However, a detailed mecha-
`nism for the IDX184 (17) prodrug cleavage has not been
`
`Antiviral Chemistry Et Chemotherapy 22.1
`
`reported to date. In the whole cell HCV replicon assay
`IDX184 (17) was shown to be a potent inhibitor of HCV
`replication (EC50=0.4 uM; CC50>100 mM) and was also
`active in the genotype 2a JFH1 replicon (EC50=O.6—11
`[.LM) [59]. At a concentration of 2.5 uM it cleared HCV
`replicon RNA after 14 days of treatment. Like other
`2’-C—methyl nucleosides it was shown to select for the
`NSSB 5282T mutation in the HCV replicon. In combi-
`nation with the protease inhibitor IDX320, IFN-0t, or
`RBV, IDX184 exhibited an additive or synergistic profile
`[60]. When administered orally to cynomolgus monkeys,
`IDX184 (17) produced high live triphosphate levels
`relative to oral administration of the parent nucleoside
`with what appears to be high hepatic extraction. Sub-
`sequent studies in a HCV—l-infected chimpanzee model
`showed that oral administration of IDX184 (17) at 10
`mg/kg over 3 days produced a median viral load decline
`of approximately -2.3 log10 at day 3 and 4 [61]. Conse-
`quently, IDX184 (17) was progressed into the clinic and
`in a Phase Ia single ascending dose study was shown to
`be generally safe and well tolerated at oral doses from 5
`mg to 100 mg [56]. PK assessment supported a liver tar-
`geting mechanism for the compound. In a Phase Ib mon—
`otherapy study in HCV genotype-l-infected patients,
`IDX184 (17) was administered at doses from 25 to 100
`mg once a day for 3 days. Day 4 viral load assessment
`showed that at the highest dose of 100 mg, a -0.74 log10
`IU/ml reduction in viral load was observed [62]. Fol-
`lowing the positive Phase I clinical results, a Phase II
`study was initiated in which IDX184 (17), at doses from
`50—200 mg, was combined