Synthesis and evaluation of 2′-dihalo ribonucleotide prodrugs with activity against hepatitis C virus

Article abstract of DOI:10.1016/j.bmc.2019.115208

[Figure presented] Hepatitis C virus (HCV) nucleoside inhibitors have been a key focus of nearly 2 decades of HCV drug research due to a high barrier to drug resistance and pan-genotypic activity profile provided by molecules in this drug class. Our investigations focused on several potent 2′-halogenated uridine-based HCV polymerase inhibitors, resulting in the discovery of novel 2′-deoxy-2′-dihalo-uridine analogs that are potent inhibitors in replicon assays for all genotypes. Further studies to improve in vivo performance of these nucleoside inhibitors identified aminoisobutyric acid ethyl ester (AIBEE) phosphoramidate prodrugs 18a and 18c, which provide high levels of the active triphosphate in dog liver. AIBEE prodrug 18c was compared with sofosbuvir (1) by co-dosing both compounds by oral administration in dog (5 mg/kg each) and measuring liver concentrations of the active triphosphate metabolite at both 4 and 24 h post dosing. In this study, 18c provided liver triphosphate concentrations that were 6-fold higher than sofosbuvir (1) at both biopsy time points, suggesting that 18c could be a highly effective agent for treating HCV infected patients in the clinic.

Full text of DOI:10.1016/j.bmc.2019.115208

Bioorganic&MedicinalChemistryxxx(xxxx)xxxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Synthesis and evaluation of 2′-dihalo ribonucleotide prodrugs with activity
against hepatitis C virus
A. Chris Krueger⁎, Hui-Ju Chen⁎, John T. Randolph, Brian S. Brown, Geoff T. Halvorsen,
Howard R. Heyman, Tongmei Li, Christopher C. Marvin, Jason P. Shanley, Eric A. Voight,
Daniel A.J. Bow, Cecilia Van Handel, Vincent Peterkin, Robert A. Carr, DeAnne Stolarik,
Tatyana Dekhtyar, Michelle L. Irvin, Preethi Krishnan, Rodger F. Henry, Rolf Wagner,
David A. DeGoey
AbbVie Incorporated, Global Pharmaceutical Research and Development, 1 North Waukegan Road, North Chicago, IL 60064, United States
A R T I C L E I N F O
A B S T R A C T
Keywords:
Hepatitis C
HCV
Antiviral
NS5B
Hepatitis C virus (HCV) nucleoside inhibitors have been a key focus of nearly 2 decades of HCV drug research
due to a high barrier to drug resistance and pan-genotypic activity profile provided by molecules in this drug
class. Our investigations focused on several potent 2′-halogenated uridine-based HCV polymerase inhibitors,
resulting in the discovery of novel 2′-deoxy-2′-dihalo-uridine analogs that are potent inhibitors in replicon assays
for all genotypes. Further studies to improve in vivo performance of these nucleoside inhibitors identified
aminoisobutyric acid ethyl ester (AIBEE) phosphoramidate prodrugs 18a and 18c, which provide high levels of
the active triphosphate in dog liver. AIBEE prodrug 18c was compared with sofosbuvir (1) by co-dosing both
compounds by oral administration in dog (5 mg/kg each) and measuring liver concentrations of the active
triphosphate metabolite at both 4 and 24 h post dosing. In this study, 18c provided liver triphosphate con-
centrations that were 6-fold higher than sofosbuvir (1) at both biopsy time points, suggesting that 18c could be a
highly effective agent for treating HCV infected patients in the clinic.
⁎ Corresponding authors.
E-mail addresses: a.chris.krueger@abbvie.com (A. Chris Krueger), hui-ju.chen@abbvie.com (H.-J. Chen).
https://doi.org/10.1016/j.bmc.2019.115208
Received 1 October 2019; Received in revised form 31 October 2019; Accepted 4 November 2019
0968-0896/©2019ElsevierLtd.Allrightsreserved.
Pleasecitethisarticleas:A.ChrisKrueger,etal.,Bioorganic&MedicinalChemistry,https://doi.org/10.1016/j.bmc.2019.115208

A. Chris Krueger, et al.
Bioorganic&MedicinalChemistryxxx(xxxx)xxxx
1. Introduction
prodrug strategy for nucleoside drug delivery.7 Early clinical studies
with nucleoside inhibitors focused on cytidine and guanosine analogs,
such as NM283 (2), the first HCV nucleoside to enter the clinic, and
BMS-986094 (3), one of the more recent compounds to be studied in
clinical trials. These and several other structurally related analogs are
potent inhibitors of viral replication, but clinical studies were dis-
continued due to poor tolerability and/or toxicity.8
Hepatitis C virus (HCV) infection continues to be a global health
problem, with current estimates ranging from 71 to 80 million
chronically infected individuals at risk of serious and potentially life-
threatening liver disease.1 Of these chronically infected people, 10–20%
will develop cirrhosis, a very serious condition that greatly increases
the odds of developing hepatocellular carcinoma. In the United States
the most recent statistics show that HCV related mortality is on the rise
with close to 20,000 deaths listing HCV as the cause in 2014.2 HCV also
exhibits an extraordinarily high degree of genetic diversity, sub-
stantially greater than that of the HIV-1 pandemic. Currently there exist
six different major HCV genotypes (1–6) found throughout the world.
HCV genotype 1, predominant in North America, Europe, and Japan,
accounts for over half of the global infections. Genotype 2 infections are
most prevalent in North America, Europe, and Japan, while genotypes 3
and 6 infections are predominant within various parts of Southeast
Asia. In Egypt, HCV infections are almost exclusively genotype 4, while
genotype 5 is common in South Africa.3
2. Theory
Antiviral activity requires formation of the nucleoside 5′-tripho-
sphate metabolite, which is required for RNA incorporation by poly-
merase. Because uridine nucleosides are not efficiently phosphorylated
in hepatocytes, uridine-based nucleosides must be delivered to the he-
patocyte with a 5′-phosphate group in place. A 5′-phosphoramidate is
an effective prodrug approach for delivering the 5′-monophosphate to
the cell, which is then converted to the active 5′-triphosphate by cel-
lular kinases. Thus, the phosphoramidate group plays a critical role in
the potency of HCV nucleoside inhibitors, particularly in uridine ana-
logs like sofosbuvir. The 2′-beta methyl substituent on nucleoside HCV
inhibitors plays a critical role in the ability of these molecules to block
viral RNA replication. Following incorporation of the HCV nucleoside
into the RNA chain, the 2′-methyl group prevents RNA chain elongation
by steric interference with the subsequent incoming nucleotide.9
Our research program has explored 2′-beta halogenated uridine
analogs as a novel class of HCV polymerase inhibitors. Inhibitors in this
structural class had not been described at the time our investigations
began, although several reports have recently appeared.10 During this
research, attention was focused on 2′-dihalogenated analogs, which
proved to be challenging compounds to assemble. Our initial designs
included both chlorinated and brominated analogs, as both groups were
expected to provide the necessary steric bulk required for activity. We
report herein the synthesis and characterization of the potency, cellular
toxicity and pharmacokinetics for phosphoramidate prodrugs of several
new 2′-dihalogenated analogs and provide a comparison to additional
2′-substituted analogs, including those previously reported. Bioactiva-
tion of 2′-halogenated uridine analogs was found to differ from so-
fosbuvir (1) in the observed triphosphate concentrations and half-lives
in human hepatocytes in vitro and dog liver biopsy studies in vivo. Thus,
we explored a variety of 5′-phosphoramidate prodrugs to identify
analogs that efficiently deliver the active triphosphate.
In the last decade treatment of HCV has been revolutionized with
the approval of direct acting antivirals (DAAs), which provide high
functional cure rates. This functional cure is considered to be achieved
when virus RNA levels in blood remain undetectable for 12 weeks after
completion of treatment, commonly noted as a sustained virologic re-
sponse at 12 weeks (SVR12). Advances in clinical research have pro-
vided three different drug combination therapies that have been ap-
proved for HCV infections encompassing genotypes 1–6 with very high
functional cure rates.4,5
Clinical studies have investigated over 3 dozen HCV drug candidates
that act on multiple viral drug target proteins. One of these proteins is
the HCV RNA-dependent RNA polymerase, NS5B. Nucleoside poly-
merase inhibitors have been of particular interest, with over a dozen
different compounds entering clinical studies. Nucleoside drug candi-
dates are attractive NS5B inhibitors because their polymerase binding
site is highly conserved across the HCV genotypes. Consequently, they
are effective against all HCV genotypes, and present a high barrier
against the development of viral resistance. In spite of the particular
attention that nucleoside analogs have received in HCV drug research,
sofosbuvir (1) is the only nucleoside HCV polymerase inhibitor that has
been approved for HCV therapy.6 Sofosbuvir (1) (Fig. 1) is a uridine
analog that is dosed as the 5′-phosphoramidate, using the ProTide
Fig. 1. Examples of HCV drug candidates that contain different nucleoside bases.
2

A. Chris Krueger, et al.
Bioorganic&MedicinalChemistryxxx(xxxx)xxxx
Scheme 1. Synthesis of halogenated nucleosides 9. (A) TMSOTf, Et3N; NBS; (B) LHMDS, NFSI; (C) TMSOTf, Et3N; Palau'Chlor; (D) LHMDS, NCS; (E) LHMDS, NCS for
5b or NBS for 5c and 5d; (F) LHMDS, NCS; (G) LHMDS, (BrCl2C)2; (H) DIBAL; (I) MsCl, Et3N; or BzCl, DMAP, pyridine; (J) (i) 2,4-bis((trimethylsilyl)oxy)pyrimidine,
TMSOTf; (ii) NH4F; (K) di-TMS-Bz-cytonsine, SnCl4; (L) (i) H2O, AcOH; (ii) NH4F.
3. Chemistry
isomer could be derived as the major product from bromination of the
chlorolactone 4d. Treatment of the silyl ketene acetal of 4a with
Palau’Chlor, followed by bromination with lithium bis(trimethylsilyl)
amide and N-bromosuccinimide, gave an inseparable mixture of pro-
ducts in a ratio of 8:1. Interestingly, the major isomer of the mixture
was identical by 1H NMR to be compound 5a (Scheme 2). The synthetic
approach was further evaluated in order to produce the quantity
needed to fully characterize compound 9a. This unexpected stereo-
chemical outcome was examined under a variety of conditions and led
to an improved synthesis involving chlorination of lactone 4a, followed
by treatment with potassium bis(trimethylsilyl)amide in toluene in
The synthesis of the halogenated nucleosides was carried out by
halogenation of a 2-deoxyribonolactone protected by bulky silyl groups
to provide flexible access to a variety of halogenated nucleosides. As
shown in Scheme 1, the two key steps involved halogenation of lactone
4a11 and subsequent stereoselective glycosylation of the activated 2-
deoxy-2,2-dihalo-D-furanose 7.
Lactone 5a was prepared by bromination11 of compound 4a fol-
lowed by treatment with lithium bis(trimethylsilyl)amide and N-
chlorosuccinimide.12 The stereochemistry of the major isomer was
preliminarily assigned as 5a, based on the assumption that the chlorine
electrophile would approach anti to the bulky silyl protecting group.
Reduction using diisobutylaluminium hydride and subsequent activa-
tion with methanesulfonyl chloride provided intermediate 7a, which
glycosylates in the presence of tin tetrachloride with bis(trimethylsilyl)-
N-benzoylcytosine, generated from N,O-bis(trimethylsilyl)acetamide
and N-(2-oxo-1,2-dihydropyrimidin-4-yl)benzamide to give a 2:1 dia-
stereomeric mixture of the nucleosides favoring the desired β-anomer.
Column chromatography yielded an inseparable mixture of the desired
compound 8a, containing, by 1H NMR, 12% of the corresponding 2,2-
dichloro compound 8e. De-silylation of 8a with ammonium fluoride
(Scheme 2) gave compound 10 as a crystalline solid. X-ray crystal-
lography of 10 (Fig. 2) confirmed the C2′-stereochemistry of compound
8a and verified the stereochemistry established by chlorination of 2-
deoxy-2-bromolactone 4b. Hydrolysis of the N-benzoylcytosine and de-
silylation of 8a gave a mixture containing mostly the desired α-Br- β
-Cl-nucleoside 9a,13 which was isolated by chiral supercritical fluid
chromatography (SFC).
combination with 1,2-dibromo-1,1,2,2-tetrachloroethane in
a 2:1
ratio,14 and quenching the reaction with methanol at −78 °C before
warming. The resulting compound 5a was stable and no decomposition
or structural changes occurred after standing at room temperature for
several weeks. Activation of lactol 6a as the benzoate and carrying out
the Vorbruggen reaction at 65 °C with 3 equivalents of tin tertachloride
and 2.5 equivalents of bis(trimethylsilyl)-N-benzoylcytosine provided
the improved stereoselectivity for the glycosylation.
Fluorination of lactone 4a,11 followed by chlorination with lithium
bis(trimethylsilyl)amide and N-chlorosuccinimide, produced the de-
sired β -chloro intermediate 5b. On the other hand, generation of the
silyl ketene acetal of the fluorolactone 4c and treatment with Palau’-
Chlor resulted in the opposite stereoselectivity yielding the corre-
sponding α -Cl- β -F-lactone in 26% yield. Diisobutylaluminium hydride
reduction and subsequent acylation of lactone 5b gave compound 7b,
which was reacted with bis-trimethylsilyl protected pyrimidine-di-one
in the presence of trimethylsilyl trifluoromethanesulfonate. The re-
sulting mixture of products was de-protected to give a 1:3 mixture of β:
α anomers. Chiral SFC purification provided the desired β-anomer
9b.15,10c
Based on the stereochemical outcomes of the halogenation sequence
leading to 5a, it was envisioned that the corresponding α -Cl- β -Br
3

A. Chris Krueger, et al.
Bioorganic&MedicinalChemistryxxx(xxxx)xxxx
Scheme 2. Unexpected stereoselectivity from chlorination/bromination of lactone 4a. (A) LHMDS, NCS; (B) LHMDS, NBS; (C) NH4F.
isolated in 5% yield. Given the need for in vitro and in vivo character-
ization of compound 9c,16,10d it was necessary to optimize the current
route due to poor stereoselectivity in two key steps. Collaboration be-
tween the discovery and process chemistry teams enabled two synthetic
approaches with improved diastereoselectivity and efficiency to deliver
compound 9c in kilogram scale, which included an X-ray structure of
this intermediate compound (Scheme 3).17 Nucleosides 9e and 9f were
prepared in similar fashion from bis-chlorination or bis-bromination of
the lactone 4a followed by transformation of the corresponding glycol
donors 7e or 7f, respectively.10a,10b
Third-Generation Route
The β-bromonucleoside 15 was synthesized from lactone 11, ob-
tained as a minor product from bromination of the corresponding 2-
chlorolactone using potassium bis(trimethylsilyl)amide and 1,2-di-
bromo-1,1,2,2-tetrachloroethane (Scheme 4). Diisobutylaluminium
hydride reduction of 11, activation of the resulting lactol with para-
methoxybenzoyl (PMBz) chloride, and protecting group transformation
produced intermediate 13. The para-methoxybenzoyl protection of the
3-hydroxyl group has demonstrated superior selectivity in the glyco-
sylation of activated 2- β -bromo-D-furanose and improved β -anomer
selectivity through anchimeric assistance.17 Executing the Vorbruggen
reaction upon 13, followed by hydrolysis of the N-benzoylcytosine with
80% aqueous acetic acid, and deprotection of the para-methoxybenzoyl
groups with ammonia in methanol gave nucleoside 15.
Fig. 2. X-ray structure of compound 10.
Although chlorination of 2-deoxy-2-fluorolactone 4c with lithium
bis(trimethylsilyl)amide and N-chlorosuccinimide gave 5b as a single
isomer, bromination of compound 4c using lithium bis(trimethylsilyl)
amide and N-bromosuccinimide gave a 3:1 inseparable mixture of
diastereomers favoring the β -bromo compound 5c. A mixture of
compounds 6c and 6d was generated from diisobutylaluminium hy-
dride reduction of 5c, and could be separated by silica gel column
chromatography. The C2 stereochemistry of 6c and 6d was determined
by HeH ROESY and H-F NOE NMR techniques. Lactols 6c and 6d were
activated as the corresponding benzoates 7c and 7d, respectively. The
unfavorable diastereoselectivity observed in the conversion of 7b was
exacerbated by the bulky β-bromo of 7c, and treatment with protected
cytosine in the presence of tin tetrachloride led to a 20:1 mixture of
diastereomers favoring the α-anomer. The desired compound 8c was
Prodrugs 16a-k and 17 were prepared by treating the corresponding
nucleosides with tert-butylmagnesium chloride and phosphoramidate
1918 or 21, whereas the prodrugs 16l, 18a, 18c, and 18l started from
racemic phosphoramidates 20 or 22, followed by chiral SFC separation
to provide the diastereomerically pure prodrugs (Scheme 5).
4

A. Chris Krueger, et al.
Bioorganic&MedicinalChemistryxxx(xxxx)xxxx
Scheme 3. Alternative synthetic approaches to compound 9c second generation route.
4. Results and discussion
group, we postulated that it would provide not only sufficient steric
interference but also polarizability to terminate the RNA chain elon-
gation for viral replication. We found that the 2′-α-bromo-2′-β-chloro
analog 16a, and the 2′-α-chloro-2′-β-bromo analog 16k both had po-
tencies that were similar to 1 with very little or no measured cellular
toxicity. Finally the 2′-α-fluoro-2′-β-bromo analog 16c displayed ex-
cellent potencies of 0.056 μM and 0.087 μM which were roughly 2 to 3
fold better than 1 with no associated cellular toxicity, while analog
16d, with the opposite stereochemistry at the 2′ position, was about 30
to 50 fold less potent. Further studies focused on promising analogs 16a
and 16c, which were evaluated for their ability to inhibit replication of
a variety of HCV genotypes, using both wild-type and variant replicon
constructs containing the S282T mutation. Activities against genotypes
2a, 3a, 4a, and 6a are compared to genotype 1 in Table 2. Both 16a and
16c provided potent inhibition of replication in replicon assays for all of
these genotypes, with measured EC50 values that were significantly
better than 1. Additionally, the loss in potency for both 16a and 16c
against the S282T resistant variants was similar to 1.
Several phosphoramidate prodrug analogs were tested bearing
modifications at the 2′ position. Inhibition of HCV genotypes 1a and 1b
RNA replication in Huh-7 cells using a subgenomic HCV replicon cell
culture system are shown in Table 1.19 As a comparator, our assay
conditions provided 50% inhibition values of 0.155 μM and 0.230 μM
for sofosbuvir (1) against genotypes 1a and 1b respectively, along with
no measured cellular toxicity in Huh-7 cells. Replacement of the po-
larizable 2′-α-fluoro group in 1 with a methyl group yielded the cor-
responding 2′-dimethyl analog 16g. This led to a surprising reduction in
replicon inhibitory potencies. On the other hand, the moderate to ex-
cellent replicon potency of analogs having two polar functional groups
on the 2′ position, such as the symmetrical 2′-dihalo analogs 16h, 16e,
and 16f, demonstrated the feasibility of halogen substitiution on C2′.
The difluoro analog 16h demonstrated a significant amount of toxicity
in Huh-7 cells in contrast to 1, 16e and 16f. The 2′-dichloro analog 16e
was more potent than 1, with EC50 values 3- to 4-fold lower against
genotype 1 replicons. Compound 16e was subsequently reported else-
where.10a While the 2′-dibromo analog 16f was not cytotoxic in Huh-7
cells, it was nevertheless approximately 10-fold less active than 16e.
Our attention then turned to unsymmetrical substitutions at the 2′-po-
sition.
Pharmacokinetic studies in dog were designed to assess the ability
of promising 2′-bromo-uridine phosphoramidate prodrugs 16a and 16c
to deliver the active 5′-triphosphate (TP) in liver. Each compound was
co-dosed with sofosbuvir (1) (5 mg/kg p.o. for both test compounds)
and the concentration of the active TP for both drugs was measured by
liver biopsy at 4 and 24 h post dosing (Table 3). This method provided a
direct comparison of active TP levels for the 2′-bromo-uridine nucleo-
side drugs with those produced by dosing a clinically effective agent.
Including sofosbuvir (1) as an internal standard also helped to correct
for variability between experiments, thus providing a more reliable
comparison of different prodrugs. Phosphoramidate 16a provided liver
TP concentrations that were > 20-fold lower than the concentrations of
sofosbuvir TP (shown in brackets in the table) at both 4 and 24 h post-
dosing. This finding was disappointing given the similar replicon po-
tency of 16a in comparison to sofosbuvir (1). Results from this study
suggest that a clinically efficacious dose for 16a would be much greater
than the approved dose of sofosbuvir (1) (400 mg/day).
We found that the 2′-α-methyl-2′-β-fluoro analog 16i (the 2′ inverse
stereochemical combination in 1) was inactive as the polarizability at
this position is quite different than in 1 and the natural substrate which
has a 2′-α-hydroxy group. Likewise 16g was also sub optimal. We also
made the analog, 2′-α-chloro-2′-β-methyl (16j) for comparison.20
Under our assay conditions, analog 16j had similar genotype 1a and 1b
potency to 1 along with no associated cellular toxicity. However analog
16b displayed roughly 3 to 4 fold better potency as compared to 1, but
also displayed some level of cellular toxicity, with a TD50 of 11 μM.
Further investigations were directed towards 2′,2′-chloro,bromo and
2′,2′-fluoro,bromo analogs. Reasoning that a chlorine or a bromine
atom possesses roughly the same Van der Waal’s radius as a methyl
Scheme 4. Synthesis of compound 15. (A) (i) DIBAL, PhMe; (ii) PMBzCl, DMAP, Et3N; (B) (i) AcOH, TBAF; (ii) PMBzCl, DMAP, Et3N; (C) di-TMS-Bz-cytonsine, SnCl4;
(D) (i) 80% aq. AcOH; (ii) NH3/MeOH.
5

A. Chris Krueger, et al.
Bioorganic&MedicinalChemistryxxx(xxxx)xxxx
Scheme 5. Synthesis of phosphoramidate prodrugs 16, 17, and 18. (A) t-BuMgCl, 19 or 20, THF/DMPU; (B) t-BuMgCl, 21, THF/DMPU; (C) PhMgCl, 22, THF/DMPU.
Phosphoramidate 16c provided TP concentrations that were similar to
the concentrations of sofosbuvir (1) TP at both biopsy time points,
evidence that clinical success with this compound might be achievable
at a similar dose.
Further prodrug modifications were focused on the amino-acid
portion of the phosphoramidate. A key compound made in this effort
was the α-amino-isobutyric acid analog 17, which provided TP con-
centrations in dog liver that were several fold higher than 16a. Further
improvements were made by replacing the isopropyl ester with an ethyl
ester group to give 18a, which provided dog liver TP concentrations
that were roughly 10-fold higher than 16a at 4 and 24 h post dosing
after adjusting for the sofosbuvir TP internal standard levels. The Rp
isomer 18l gave similar results in a single dose PK study. When 18a and
18l were co-dosed with sofosbuvir (1) (5 mg/kg for both test com-
pounds) for 4 days to achieve near steady state TP concentrations, the
Sp isomer 18a provided higher TP concentrations at 24 h following the
final dose. The amino-isobutyric acid ethyl ester (AIBEE) prodrug pro-
vided dog liver TP concentrations for 18a that were similar to so-
fosbuvir (1) in the single dose PK studies, although sofosbuvir (1) TP
concentrations were higher in the repeat dosing study.
We investigated alternative 5′-phosphoramidate prodrugs for the 2′-
α-bromo-β-chloro uridine nucleoside in an effort to identify compounds
that deliver high TP levels in comparison to 16a.21 Physical properties
for 16a that distinguish it from sofosbuvir include much higher per-
meability in the PAMPA assay, and relatively poor aqueous solubility
(Table 4). The 2′-halogens on 16a make it more liphophilic than so-
fosbuvir (1), as evidenced by the higher LogD, which likely contributes
to the improved permeability. However, the low solubility (23 μM at pH
7.4) is believed due primarily to the crystalline nature of 16a, and was
suspected to be a contributing factor in the poor in vivo performance,
possibly due to limiting effects on drug absorption. The Rp isomer 16l,
which has the opposite phosphate stereochemistry, was less crystalline
and much more soluble. However, both 16l and 16a provided similar
TP levels in dog liver, indicating that solubility was not a significant
contributor to the poor in vivo performance of these isomeric prodrugs.
Due to the dramatic effect on triphosphate levels observed in 18a,
we also prepared the amino-isobutyric acid ethyl ester (AIBEE) prodrug
18c in an effort to deliver higher level TP concentrations of the 2′-α-
Table 1
HCV genotype 1a and 1b replicon activity, and cytotoxicity, in Huh7 cells for 2′-halogenated uridine phosphoramidate prodrugs.
Compound #
1
16 g
16 h
16e
16f
16i
16j
16b
16a
16 k
16c
16d
R1
R2
=
=
Me
Me
F
Cl
Br
F
Me
Cl
Cl
Br
Br
F
F
Me
F
Cl
Br
Me
Cl
F
Br
Cl
F
Br
Assay
GT1a EC50 (µM)
GT1b EC50 (µM)
Huh-7 TD50 (µM)
0.155
0.230
> 100
5.63
4.75
31.6
0.226
0.286
0.384
0.043
0.048
> 100
0.408
0.522
> 100
> 10
> 10
> 32
0.139
0.167
> 100
0.053
0.061
11
0.114
0.147
80
0.200
0.258
> 100
0.056
0.087
> 100
3.03
2.50
26.5
6

A. Chris Krueger, et al.
Bioorganic&MedicinalChemistryxxx(xxxx)xxxx
Table 2
Activities of 16a, 16c, and 1 in transient replicon assays for various HCV genotypes, with and without the S282T mutation.a
Cpd #
EC50 μM
1b-Con1-WT
1b-S282T
2a-WT
2a-S282T
3a-WT
3a-S282T
4a-WT
6a-WT
16a
16c
1
0.0304
0.0152
0.0620
0.1606
0.1343
0.6376
0.1882
0.1888
0.6813
0.2617
0.3082
1.002
0.0314
0.0127
0.0751
0.1418
0.0753
0.3301
0.0299
0.0257
0.0785
0.0227
0.0145
0.0867
a
WT refers to Wild Type virus.
Table 3
Concentrations of the active 5′-triphosphate of 2′-bromo-uridine nucleosides.
Compound
Structure
Nucleotide triphosphate concentration (μM)
Dog liver†
Human hepatocytes‡
R1
R2
R3
*
4 h
24 h
4 day SSa
4 h
24 h
122
Sofosbuvir (1)
Me
F
Sp
[see below]b
[see below]b
[see below]b
97
16a
16c
16l
17
Cl
Br
Cl
Cl
Cl
Cl
Br
Br
F
Sp
0.088 [1.59]
11.5 [7.39]
0.143 [1.76]
0.713 [4.20]
8.83 [15.03]
7.21 [5.35]
43.3 [7.0]
0.049 [1.02]
6.61 [8.87]
0.036 [1.26]
0.244 [3.43]
2.10 [5.44]
2.64 [4.03]
42.7 [6.64]
NT
110
192
102
157
2818
NT
22
Sp
NT
91
Br
Br
Br
Br
F
Rp
NT
24
Sp/Rp
Sp
NT
40
18a
18l
18c
2.89 [11.1]
1.43 [6.05]
8.18 [10.7]c
203
NT
736
Rp
Sp
3600
†
Concentration of nucleotide triphosphate (TP) in dog liver following a 5 mg/kg p.o. dose of the prodrug.
Concentration of TP in human hepatocytes following a 4 h incubation with the prodrug (100 μM).
‡
a
b
c
TP concentrations following once daily p.o. dosing for 4 days to achieve near steady-state TP levels (liver biopsy taken 24 h following dose 4).
Bracketed values are the concentration of sofosbuvir TP in liver (co-dosed with the test compound at 5 mg/kg p.o.).
18c (1 mg/kg) and sofosbuvir (5 mg/kg) co-dosed p.o.; NT = not tested.
fluoro-β-bromo uridine nucleoside. In a single-dose dog PK study,
prodrug 18c provided TP concentrations that were over 5-fold higher
than 16c and about 6-fold higher than sofosbuvir (1) TP at the 24 h
biopsy time point. Importantly, liver TP concentrations were the same
at both 4 and 24 h following dosing with 18c, demonstrating excellent
TP persistence with this prodrug. In a repeat dosing study in dog, 18c
(1 mg/kg) was co-dosed with sofosbuvir (1) (5 mg/kg) for four days.
The liver TP concentration provided by 18c (8 μM) was similar to so-
fosbuvir (1) TP (12 μM) at 1/5 the dose.
hepatocytes at 4 h, TP concentration dropped 4-fold by 24 h. 16c pro-
vided much higher TP concentration at 4 h, and better TP persistence
through 24 h. The AIBEE prodrugs 18a and 18c both provided very
high TP concentrations in hepatocytes at 4 h, 30-fold to 40-fold higher
than sofosbuvir (1) TP, although TP concentrations at 24 h were much
lower. 18c provided better TP persistence, with a 5-fold reduction in TP
concentration from 4 h to 24 h in comparison to over 12-fold reduction
from 4 to 24 h for 18a.
The high concentrations of nucleoside TP provided by AIBEE
phosphoramidate prodrugs of HCV active nucleosides observed both in
human hepatocytes and in dog liver biopsy studies, allude to great
potential for these compounds as potent inhibitors of HCV replication in
vivo. Yet, activity studies in genotype 1 HCV replicon assays found that
AIBEE prodrug analogs are less potent than sofosbuvir (1), 16a, and
16c, all of which have the “standard” L-alanine isopropyl ester phos-
phoramidate prodrug (Table 4). These results are consistent with a
Human hepatocyte studies were used to measure TP concentrations
produced using phosphoramidate prodrugs of 2′-bromo uridine nu-
cleosides in the target cell. In this assay, sofosbuvir (1) provided higher
TP levels at 24 h (140 μM) than at 4 h (95 μM). The remarkable per-
sistence of sofosbuvir (1) TP in hepatocytes undoubtedly contributes to
its efficacy, and is an important characteristic for an HCV nucleoside
prodrug. Although 16a provided acceptable concentrations of TP in
7

A. Chris Krueger, et al.
Bioorganic&MedicinalChemistryxxx(xxxx)xxxx
Table 4
In vitro potency and ADME properties of nucleoside prodrugs described in Table 3.
Compound id
Replicon EC50 (μM)
ADME properties
GT1a
GT1b
GT1b CES1a
LogD HT-LCMS
PAMPA
aq sol (μM) (pH 7.4)
Plasma stabilityb
fup
Hep Clint (L/h/kg)
Dog
Human
Dog
Human
Dog
Human
Sofosbuvir (1)
0.168
0.114
0.056
0.412
1.09
0.188
0.147
0.095
0.515
1.38
0.053
NT
3.02
3.51
3.07
3.21
3.30
3.07
3.11
2.86
0.36
2.22
0.97
1.68
2.86
3.46
3.16
0.71
> 200
23.4
93%
92%
86%
NT
96%
92%
100%
NT
0.323
0.213
0.367
NT
0.563
0.113
0.430
NT
2.4
6.8
NT
4.3
16a
16c
16l
17
16.4
15.4
12.1
11.5
9.3
0.056
0.218
0.162
NT
> 200
> 200
> 200
> 200
> 200
519
NT
100%
100%
86%
100%
100%
92%
100%
98%
0.130
0.187
0.133
0.317
0.193
0.310
0.240
0.473
8.0
5.9
12.0
7.4
18l
18a
18c
2.72
2.56
0.644
0.259
0.952
0.434
0.154
0.069
16.1
7.9
a
b
EC50 in GT1b replicons that overexpress CES1.
Percent remaining following a 4 h incubation at 37 °C in dog or human plasma; NT = not tested.
theory that the carboxyesterase 1 (CES1) enzyme is principally re-
sponsible for metabolism of prodrugs having an AIB ester group. The
CES1 enzyme is poorly expressed in Huh-7 cells used in standard HCV
replicon assays, which have been shown to rely on cathepsin A (CatA)
to initiate metabolic activation of sofosbuvir (1).22 We established a
modified replicon assay using genotype 1b replicon cells that tran-
siently express higher levels of CES1 that are more similar to levels in
human hepatocytes (Table 4, CES1 EC50). EC50 values in this modified
assay for AIBEE prodrugs 18a and 18c are equivalent to EC50 values for
the alanine isopropyl ester phosphoramidate prodrugs of the corre-
sponding nucleosides (16a and 16c) in the standard genotype 1b re-
plicon assay. Interestingly, 16c has better activity than 18c in the CES1
modified genotype 1b replicon assay. Since TP concentrations drive
cellular potency through inhibition of HCV NS5B polymerase, this re-
sult would indicate that AIBEE prodrugs do not provide higher TP
concentrations in replicon cells.
analogs 18. When AIBEE prodrug 18a was co-dosed with sofosbuvir (1)
in dog (5 mg/kg p.o. for both compounds), this compound provided
liver TP concentrations that were > 10-fold higher than those obtained
with 16a at the same dose, and similar to TP concentrations observed
for sofosbuvir (1). When AIBEE prodrug 18c was co-dosed with so-
fosbuvir (1) in dog (5 mg/kg p.o. for both compounds), this compound
provided liver TP concentrations that were 5-fold higher than so-
fosbuvir (1) TP concentrations. With regards to TP persistence in liver,
the 2′-bromo,fluoro nucleosides 16c and 18c are superior to the 2′-
bromo,chloro nucleosides 16a and 18a. In order to confirm the ad-
vantage provided by AIBEE prodrugs in vivo, a head to head repeat
dosing study in dogs for 4 days to assess near steady-state liver TP le-
vels, with 18c dosed at 1 mg/kg, and sofosbuvir (1) dosed at 5 mg/kg.
18c provided TP concentrations at 24 h following the final dose that
were similar to sofosbuvir (1) TP concentrations, confirming the results
from the single dose study. This data suggests that 18c could be a highly
effective agent for treating HCV infected patients in the clinic.
A comparison of ADME properties of phosphoramidate prodrugs
does not provide an explanation for the difference in TP levels provided
by AIBEE prodrugs. All of the prodrugs (with the exception of 16a) have
good aqueous solubility in phosphate buffer, and all are stable in dog
and human plasma, with little or no degradation observed following
incubation at 37 °C for 4 h. A significant difference in unbound fraction
of drugs in plasma was observed, with higher protein binding con-
tributing to a lower free fraction for the more lipophilic 2′- bromo,-
chloro uridine prodrugs. However, the plasma free fraction difference
did not contribute to an observable difference in plasma stability, and
importantly, measured unbound fractions were no different for pro-
drugs of the same nucleoside (ie, 16c and 18c). The measured intrinsic
clearance in human and dog hepatocytes offered no help to understand
the large increase in TP concentrations observed for the AIBEE phos-
phoramidate prodrugs.
6. Experimental section
6.1. Single crystal X-ray diffraction studies
A single crystal of compound 10 was mounted on loop. Intensity
data were collected on a Bruker APEX2. Data were collected at 100 K
with graphite-monochromated MoKα radiation (λ = 0.71073 Å) and
triumph optics. Data were collected in four sets using omega-phi scans
with omega steps of 0.3° and phi steps of 90°. A total of 2350 frames
were collected with 20 s frame exposures. Data processing was carried
out using the APEX software package.24 Corrections for Lorentz-polar-
ization effects and absorption were applied. All structures were solved
using direct methods that yielded the non-hydrogen atoms. All pre-
sented hydrogen atoms were located in Fourier-difference electron
density maps. All non-hydrogen atoms were refined anisotropically.
Hydrogen atoms associated with carbon atoms and the amide nitrogen
atom were refined in geometrically constrained riding positions. The
hydrogen atoms associated with alcohol groups and the water molecule
were refined isotropically. Refinement was carried out with the use of
SHELX-97.25 Absolute stereo-chemistry was determined by anomalous
dispertion.
5. Conclusions
A research program to discover novel 2′-halogenated uridine-based
nucleoside HCV polymerase inhibitors identified 16a and 16c, which
are potent inhibitors of HCV replication in replicon assays.23 These
unique bromine-containing 2′-dihalogenated nucleosides are syntheti-
cally challenging molecules due to the difficulty in controlling stereo-
selectivity during the halogenation and heteroarylation steps. While
both compounds have in vitro antiviral potency equal to or better than
sofosbuvir against multiple genotypes, oral dosing studies in dog found
that 16a provided nucleoside TP concentrations in dog liver that were
much lower than 16c and sofosbuvir (1). Given the importance of the
5′-phosphoramidate group for delivering the nucleoside to the liver in a
form that can be efficiently converted to the active TP metabolite, al-
ternative phosphoramidate prodrugs were investigated, resulting in the
discovery of aminoisobutyric acid ethyl ester (AIBEE) phosphoramidate
6.2. Replicon assays
All compounds were tested against 1a-H77 (GenBank accession
number NC004102) and genotype 1b-Con1 (GenBank accession number
AJ238799) stable subgenomic replicon cell lines containing the luci-
ferase reporter activity as described previously.19 In order to express
liver carboxylesterase 1(CES1) (NM_001266) in the replicon cell lines,
CES1 expressing plasmid was purchased from OriGene Technologies
8

A. Chris Krueger, et al.
Bioorganic&MedicinalChemistryxxx(xxxx)xxxx
(Rockville, MD). Approximately 6 million genotype 1b-Con1 replicon
cells were incubated overnight in the absence of antibiotics in Dul-
becco’s Modified Eagle Mediun (DMEM) containing 5% fetal bovine
serum (FBS). The CES1 transfection mixture was prepared by in-
cubating 800 µL DMEM, 24 µL turbofectin-TF (OriGene Technologies
Rockville, MD), and 8 µg CES1 DNA for 25 min at room temperature.
Media on cells incubated overnight was replaced with CES1 transfec-
tion mixture in 15 mL DMEM without antibiotics containing 5% FBS for
24 h resulting in transient expression of CES1. The inhibitory effect of
compounds on HCV replication in replicon cells with or without CES1
overexpression was determined in DMEM containing 5% FBS. The cells
were incubated for 3 days in the presence or absence of compounds, and
were subsequently lysed and processed according to the manufacturer’s
instructions (Promega, Madison, WI) to measure the luciferase reporter
activity using an EnVision instrument (Perkin-Elmer, Waltham, MA).
Previously described replicon shuttle vectors were used to generate
constructs encompassing NS5B derived from HCV genotypes 2a, 3a, 4a,
or 6a -infected patient samples.26 Mutagenesis was performed using the
Change-IT Multiple Mutation Site Directed Mutagenesis Kit (USB), or an
NS5B gene containing the mutation was synthesized (IDT) and inserted
into the appropriate shuttle vector plasmid. Transfection of Huh-7 cells
and measurement of the inhibition of HCV replication by compounds
with the luciferase assay were conducted as described previously.27
EC50 was calculated using nonlinear regression curve fitting to the 4-
parameter logistic equation and GraphPad Prism 5 software.
manner. MS analysis was conducted using a Finnigan SSQ7000 (ESI)
mass spectrometer. All final compounds were purified to > 95% purity
as determined by reverse phase HPLC performed on a Waters 2695
separation module/Waters 2489 UV/visible detector equipped with a
YMC ODS-A, 5.5 μm, 120 Å, 4.6 mm × 150 mm column using a solvent
gradient of 10–100% acetonitrile in water (0.1% TFA).
(3S,4R,5R)-3-Chloro-4-((triisopropylsilyl)oxy)-5-(((triisopro-
pylsilyl)oxy)methyl)dihydrofuran-2(3H)-one (4d). To a solution of
compound 4a (376.86 g, 847 mmol) and triethylamine (0.295 l,
2117 mmol) in dichloromethane (1.9 l) at −15 °C was slowly added
trimethylsilyl trifluoromethanesulfonate (0.352 l, 1948 mmol)
(∼45 min). The mixture was stirred for 10 min, and a solution of
Palau’Chlor (231.35 g, 1104 mmol) in dichloromethane (1.1 l) was
added slowly (∼1 h). The mixture was stirred for 30 min after addition,
quenched with 2 N HCl (1.8l), and stirred for 20 min. The layers were
separated, and the aqueous layer was extracted with dichloromethane
(1 l, 0.3 l). The combined organic layers were washed with 1 N HCl
(350 mL), 1 M Na2S2O3 (400 mL), water (400 mL), and brine. The light
yellow organic layer was dried over Na2SO4 and concentrated. Copious
white solids precipitated during concentration, so heptanes (780 mL)
was added, the resulting suspension was spun on the rotovap for 5 min,
and decanted carefully, rinsing the sticky white solids with heptanes
(3 × 50 mL). The mixture was concentrated to 416 g total mass and
purified by column chromatography on silica gel using a solvent gra-
dient of 0–25% dichloromethane in heptanes to give compound 4d
(224.65 g, 55.3% yield). 1H NMR (500 MHz, CDCl3) δ ppm 1.03 – 1.19
(m, 63H), 3.90 – 3.97 (m, 1.5H), 4.02 – 4.10 (m, 1.5H), 4.29 (dt,
J = 5.4, 3.2 Hz, 1H), 4.38 (d, J = 6.1 Hz, 1H), 4.45 (q, J = 2.4 Hz,
0.5H), 4.68 (d, J = 5.4 Hz, 0.5H), 4.71 (dd, J = 5.5, 2.5 Hz, 0.5H), 4.83
(dd, J = 6.2, 5.4 Hz, 1H).
6.3. Hepatocyte studies
Cryopreserved primary human hepatocytes were obtained from
BIOIVT, thawed and plated (∼55,000 viable cells per well) on collagen-
coated 96 well culture dishes. Cells were incubated at 37 °C (5% CO2
incubator), then plating medium was replaced after 4–6 h by Incubation
(3R,4R,5R)-3-Bromo-3-chloro-4-((triisopropylsilyl)oxy)-5-
(((triisopropylsilyl)oxy)methyl)dihydrofuran-2(3H)-one (5a).
A
solution of compound 4d (116.59 g, 243 mmol) in tetrahydrofuran
(800 mL) was stirred at room temperature and 1,2-dibromo-1,1,2,2-
tetrachloroethane (127.02 g, 390 mmol) was added. Cooled to −115 °C
(liquid nitrogen, slurry started to form) and added a 0.5 M toluene so-
lution of potassium bis(trimethylsilyl)amide (1070 mL, 535 mmol)
slowly maintaining the internal temperature between −100 to −85 °C
(∼1 h). The cooling bath was removed, and after 15 min (final tem-
perature of −79 °C), 2 N HCl (650 mL) was added. The mixture was
warmed to room temperature and the layers were separated. The or-
ganic layer was washed with brine (300 mL), dried over Na2SO4, and
concentrated to give 274 g of crude product as a dark oil. Another batch
of 5a was prepared using a similar condition starting from compound
4d (108.06 g, 225 mmol). The material from both reactions were
combined, diluted with heptane (1 l), and washed with acetonitrile
(100 mL, and 50 mL). The acetonitrile layers were combined and back-
extracted with heptanes (700 mL), and the new combined heptane
layers were washed with acetonitrile (50 mL). This acetonitrile wash
was then back-extracted with heptanes (500 mL), the heptanes layers
were combined and concentrated to give compound 5a (272.33 g,
quantitative) which was used without further purification. 1H NMR
(400 MHz, CDCl3) δ ppm 1.32–0.97 (m, 42Η), 3.99 (dd, J = 12.5,
1.9 Hz, 1Η), 4.24–4.10 (m, 2Η), 4.75 (d, J = 7.6 Hz, 1Η). MS (DCI) m/z
Medium (Williams
E Medium without phenol red, containing
Hepatocyte Maintenance Supplements, Gibco product CM4000) and
Geltrex™ (Life Technologies) 0.35 mg/mL. Following overnight in-
cubation, two separate plates of hepatocytes were incubated with
prodrug compounds (100 µM) for 4 h. Prodrug containing media was
removed and washed from hepatocytes after 4hrs. One plate from each
experiment was further incubated in incubation media only for a fur-
ther 20 h (washout period). Following the 4 or 24 h incubation period,
plates were washed with ice-cold HBSS and stored at −80 °C until ex-
traction for analysis. Samples were analyzed using LC-MS/MS.
6.4. Dog pharmacokinetic studies
Prodrugs (5 mg/kg or 1 mg/kg PO in EtOH: PEG-400: Phosal 53
MCT (10:30:60, w/w)) were cassette dosed with sofosbuvir (5 mg/kg)
to beagle dogs (3 dogs/per time point). Liver biopsy samples were
collected at 4 or 24 h post single dose or post final dose when com-
pounds were dosed for 4-days. Liver biopsy samples were flash frozen
and stored at −80 °C until analysis. The concentrations of triphosphates
in liver were analyzed by LC–MS/MS. All animal procedures were ap-
proved by the Institutional Animal Care and Use Committee at AbbVie
Inc.
576 (M + NH4)+
.
(3R,4R,5R)-3-Bromo-3-chloro-4-((triisopropylsilyl)oxy)-5-
6.5. Chemistry
(((triisopropylsilyl)oxy)methyl)tetrahydrofuran-2-yl
benzoate
(7a’). A solution of compound 5a (261 g, 468 mmol) in toluene
(1000 mL) was cooled to an internal temperature of −78 °C. A solution
of diisobutylaluminium hydride (1 M in toluene, 515 mL, 515 mmol)
was added over 45 min so that the internal temperature remained
below −65 °C. After stirring for 30 min at −70 °C, pyridine (114 mL,
1410 mmol), 4-dimethylaminopyridine (85.94 g, 703 mmol), and then
benzoic anhydride (212.1 g, 938 mmol) were added. The cooling bath
was removed, and the mixture was stirred for 45 min (final temperature
of −15 °C), and warmed with a water bath to room temperature, and
General Procedures. All starting materials were dried by azeotropic
distillation of toluene prior to subjecting to reaction conditions.
Reagents and solvents, including anhydrous solvents, were obtained
from commercial sources and used as supplied. Column chromato-
graphy was carried out on silica gel. 1H NMR spectra were measured
using either a Bruker 500 MHz or a Varian 400 MHz spectrometer.
Chemical shifts are reported in ppm (δ) and referenced to an internal
standard of tetramethylsilane (δ 0.00 ppm). 1H−1H couplings are as-
sumed to be first-order, and peak multiplicities are reported in the usual
9

A. Chris Krueger, et al.
Bioorganic&MedicinalChemistryxxx(xxxx)xxxx
stirred for 30 min. The reaction was quenched by addition of mL 2 N
HCl (150 mL, exothermic to 65 °C), more 2 N HCl was added (1.8 l,
temperature to 50 °C), and stirred vigorously for 5 min becoming a
mostly biphasic solution. The mixture was diluted with heptane (1 l),
washed with 1 N HCl (350 mL), saturated aqueous NaHCO3
(3 × 400 mL), brine (350 mL), dried over Na2SO4, and concentrated.
This material was dissolved in heptane (2 l), and washed with acet-
onitrile (5 × 400 mL). The combined acetonitrile layers were back-ex-
tracted with heptanes (400 mL), which were extracted with acetonitrile
(3 × 75 mL). The combined heptanes layers were concentrated to give
compound 7a’ (276.61 g, 89% yield).
fluorobenzenesulfonimide (5.0562 g, 16.03 mmol), and the resulting
mixture was cooled to −78 °C. A solution of lithium bis(trimethylsilyl)
amide in toluene (1 M, 13.5 mL, 13.50 mmol) was added dropwise over
20 min. After an additional 20 min the reaction mixture was quenched
by the dropwise addition of an aqueous solution of HCl (1 Ν, 20 mL) at
−78 °C. The cooling bath was removed, and heptanes (40 mL) and 1 N
HC1 (50 mL) were added. The mixture was stirred and warmed to room
temperature. The layers were separated. The organic layer was washed
with 1 N HC1 (40 mL), water (20 mL) and brine (20 mL). The combined
aqueous layers were extracted with heptanes (2 × 30 mL). The com-
bined organic layers were dried over MgSO4 and concentrated. The
residue was purified by flash chromatography on silica eluting with a
solvent gradient of 0–60% dichloromethane in heptane to provide
compound 4c (4.31 g, 71.6% yield). 1H NMR (400 MHz, CDCl3) δ ppm
1.04–1.12 (m, 42Η), 3.94 (dd, J = 12.2, 2.4 Hz, 2Η), 4.11 (dt, J = 12.2,
2.1 Hz, 2Η), 4.19 (dt, 35 J = 7.0, 2.3 Hz, 2Η), 4.91 (t, J = 7.2 Hz, 1Η),
4.96 (t, J = 7.2 Hz, 1Η), 5.07 (d, J = 7.4 Hz, 1Η), 5.20 (d, J = 7.4 Hz,
1Η).
N-(1-((2R,3R,4R,5R)-3-Bromo-3-chloro-4-((triisopropylsilyl)
oxy)-5-(((triisopropylsilyl)oxy)methyl)tetrahydrofuran-2-yl)-2-
oxo-1,2-dihydropyrimidin-4-yl)benzamide (8a). A suspension of N-
(2-oxo-1,2-dihydropyrimidin-4-yl)benzamide (224 g, 1041 mmol) and
trimethylsilyl N-(trimethylsilyl)acetimidate (305 mL, 1247 mmol) in
chlorobenzene (1250 mL) was heated to 80 °C for 1 h (becoming an
orange solution). The heating mantle was removed, and a solution of
compound 7a’ (276.6 g, 416 mmol) in chlorobenzene (1250 mL) was
added at 45 °C, then tin tetrachloride (142 mL, 1210 mmol) was added
slowly (∼20 min, exothermic to 60 °C). The orange-red solution was
heated to 65–70 °C. After 13 h, the brown reaction mixture was con-
centrated to 872 g total mass (of a thick brown solution), then diluted
with methyl tert-butyl ether (2.8 l) and water (2.8 l). The resulting
suspension was filtered through Celite®, and the solid washed with
additional methyl tert-butyl ether and heptanes. The filtrate was washed
with saturated aqueous NaHCO3, and brine, dried over Na2SO4, and
concentrated to give 300 g of crude product, which was purified by
column chromatography on silica gel using a solvent gradient of 0–20%
ethyl acetate in heptane to give the product (128 g, 41% yield). The
product was further purified by addition of heptanes (384 mL) and
heating to 90 °C, aging 10 min until the thin white slurry was finely
dispersed. The suspension was cooled to room temperature over 90 min,
and the solid was collected by filtration, washed with heptanes (60 mL),
then dried in a vacuum oven at 50 °C to give compound 8a (81.6 g, 26%
yield). 1HNMR (400 MHz, DMSO‑d6) δ 0.98–1.25 (m, 42Η), 3.98 (dd,
J = 20.0, 10.4 Hz, 2Η), 4.20 (d, J = 12.4 Hz, 1Η), 4.33 (d, 1Η), 6.84 (s,
1Η), 7.39 (d, J = 7.6 Hz, 1Η), 7.50 (t, J = 7.7 Hz, 2Η), 7.62 (t,
J = 7.3 Hz, 1Η), 7.98 (d, 2Η), 8.24 (d, J = 7.7 Hz, 1Η), 11.52–11.26
(3S,4R,5R)-3-Bromo-3-fluoro-4-((triisopropylsilyl)oxy)-5-
(((triisopropylsilyl)oxy)methyl)dihydrofuran-2(3Η)-one (5c) and
(3R,4R,5R)-3-Bromo-3-fluoro-4-((triisopropylsilyl)oxy)-5-(((triiso-
propylsilyl)oxy)methyl)dihydrofuran-2(3H)-one (5d). To a solution
of compound 4c (5.8357 g, 12.61 mmol) in tetrahydrofuran (55 mL)
was added N-bromosuccinimide (2.69028 g, 15.12 mmol), and the re-
sulting mixture was cooled to −78 °C. A toluene solution of lithium bis
(trimethylsilyl)amide (1 M, 15.5 mL, 15.50 mmol) was added dropwise
over 15 min. After 10 min, thin layer chromatography (TLC) (50%
methylene chloride in heptane) showed the reaction was incomplete.
Additional lithium bis(trimethylsilyl)amide (12 mL, 12.00 mmol) was
added dropwise over 10 min as the light yellow suspension turned light
brown–red. After 5 min TLC showed completion and the reaction was
quenched with 1 N HCl (30 mL) at −78 °C. The cooling bath was re-
moved, heptanes (50 mL) and additional 1 N HCl (100 mL) was added.
After warming to room temperature the layers were separated. The
organic layer was washed with water (30 mL) and brine (25 mL). The
combined aqueous layers were extracted with heptanes (50 mL × 2).
The combined organic layers were dried over MgSO4 and concentrated
to give a light yellow oil (7.9 g). Flash chromatography using silica gel
eluted with a solvent gradient of 0–5% ethyl acetate in heptanes gave,
by 1HNMR, a 3.4:1 mixture of compounds 5c and 5d (5.9467 g, 78%
yield) as a colorless oil. 1H NMR (400 MHz, CDCl3) δ ppm 1.04–1.21
(m, 58H), 3.94–4.12 (m, 3H), 4.13–4.18 (m, 0.4H), 4.46 (ddd, J = 6.4,
4.9, 2.5 Hz, 1H), 4.79 (dd, J = 14.6, 7.7 Hz, 0.4H), 4.93 (dd, J = 4.5,
(m, 1Η). MS (ESI) m/z 758 (Μ + H)+
.
1-((2R,3R,4R,5R)-3-Bromo-3-chloro-4-hydroxy-5-(hydro-
xymethyl)tetrahydrofuran-2-yl)pyrimidine-2,4(1H,3H)-dione (9a).
A mixture of compound 8a (81.4 g, 107 mmol), acetic acid (651 mL),
and water (163 mL) was heated to 100 °C. After 4 h, the mixture was
concentrated (70 °C rotovap temp), then flushed with toluene
(2 × 250 mL) to 89 g of total mass (light yellow semi-solid), then me-
thanol (407 mL) and ammonium fluoride (39.8 g, 1075 mmol) were
added and the resulting mixture was heated to 60–65 °C. After stirring
for 16 h a slurry was observed, so acetonitrile (400 mL) and methanol
(400 mL) were added, keeping the temperature greater than 60 °C and
the material went back into solution. The solution was cooled to room
temperature and filtered, then the solid was washed with 2:1 methanol/
acetonitrile (2 × 120 mL), Silica gel (150 g) was added and the mixture
was concentrated carefully then dry-loaded material and purified by
column chromatography using a solvent gradient of 0–10% methanol in
dichloromethane to give compound 9a (27.75 g, 76% yield) con-
taining ∼ 5% a diastereomeric impurity by HPLC. 1H NMR (400 MHz,
DMSO-d6) δ ppm 3.65 (ddd, J = 13.2, 5.1, 2.9 Hz, 1H), 3.79–3.87 (m,
2H), 4.03 (dd, J = 8.8, 5.8 Hz, 1H), 5.49 (t, J = 4.7 Hz, 1H), 5.71 (dd,
J = 8.2, 0.5 Hz, 1H), 6.62 (s, 1H), 6.83 (d, J = 5.9 Hz, 1H), 8.09 (d,
2.8 Hz, 1H); MS (ESI) m/z 541.0 (M + H)+
.
(3S,4R,5R)-3-Bromo-3-fluoro-4-((tri-isopropylsilyl)oxy)-5-
(((triisopropylsilyl)oxy)methyl)tetrahydrofuran-2-ol
(6c) and
(3R,4R,5R)-3-Bromo-3-fluoro-4-((tri-isopropylsilyl)oxy)-5-(((triiso-
propylsilyl)oxy)methyl)tetrahydrofuran-2-ol (6d). To a solution of
the above mixture of compounds 5c and 5d (6.0767 g, 11.22 mmol) in
toluene (35 mL) at −78 °C was added dropwise a solution of diisobu-
tylaluminium hydride in toluene (1 Μ, 13 mL, 13.00 mmol). The solu-
tion was stirred at −78 °C for 1 h, then quenched with a 10% aqueous
solution of sodium potassium tartrate (60 mL) at −78 °C, and the
mixture was allowed to warm to room temperature, then filtered
through a pad of Celite®, and rinsed with methyl tert-butyl ether. The
layers of the filtrate were separated. The organic layer was washed with
an aqueous solution of 1 N Na2CO3 (20 mL), 1 N HCl (50 mL), water
(15 mL) and brine (15 mL), dried over MgSO4, concentrated and pur-
ified by flash chromatography eluting with a solvent gradient of 0–30%
dichloromethane in heptanes to give compound 6c (3.49 g, 57.2%
yield) and compound 6d (2.8 g, 45.9% yield). For 6c: 1H NMR
(400 MHz, CDCl3) δ ppm 1.06–1.14 (m, 42H), 3.58 (dd, J = 12.1,
1.8 Hz, 1H), 3.82–3.86 (m, 2H), 3.88–3.99 (m, 4H), 4.08 (dt, J = 5.7,
2.3 Hz, 1H), 4.22 (dtd, J = 5.9, 4.1, 1.8 Hz, 1H), 4.75 (dd, J = 6.0,
J = 8.2 Hz, 1H), 11.55 (s, 1H). MS (ESI) m/z 341 (Μ-Η)−
.
(4R,5R)-3-Fluoro-4-((triisopropylsilyl)oxy)-5-(((triisopro-
pylsilyl)oxy)methyl)dihydrofuran-2(3H)-one (4c). To a solution of
4a (5.466 g, 12.29 mmol) in tetrahydrofuran (40 mL) was added N-
10

A. Chris Krueger, et al.
Bioorganic&MedicinalChemistryxxx(xxxx)xxxx
3.5 Hz, 1H), 4.87 (dd, J = 16.1, 5.7 Hz, 1H), 5.15 (dd, J = 9.7, 5.5 Hz,
1H), 5.39 (dd, J = 12.0, 2.1 Hz, 1H). For 6d: 1H NMR (500 MHz, CDCl3)
δ ppm 1.04–1.21 (m, 59Η), 3.48 (d, J = 12.7 Hz, 0.4Η), 3.77–3.84 (m,
2Η), 3.86–3.91 (m, 2Η), 3.94 (dt, J = 6.6, 2.0 Hz, 1Η), 4.08 (tdd,
J = 4.6, 3.5, 1.0 Hz, 0.4Η), 4.67 (ddd, J = 11.6, 4.6, 0.9 Hz, 0.4Η), 4.72
(dd, J = 12.8, 6.6 Hz, 1Η), 5.17 (ddd, J = 12.7, 5.9, 0.9 Hz, OH), 5.34
(dd, J = 9.2, 1.2 Hz, 1Η).
80% pure product which was further purified by flash chromatography
on silica eluting with a solvent gradient of 0–40% of (10% methanol in
acetonitrile) in dichloromethane to give pure product (45 mg) and
mixed fractions (30 mg), which were further purified by preparative
thin layer chromatography in 50% of (10% methanol in acetonitrile) in
dichloromethane. The pure product was combined to provide com-
pound 9c (59 mg, 0.172 mmol, 55.4% yield). 1H NMR (500 MHz,
DMSO‑d6) δ ppm 3.64 (ddd, J = 12.6, 5.2, 3.0 Hz, 1H), 3.79–3.87 (m,
2H), 4.24–4.33 (m, 1H), 5.35 (t, J = 5.1 Hz, 1H), 5.73–5.76 (m, 1H),
6.24 (d, J = 16.9 Hz, 1H), 6.50 (d, J = 7.1 Hz, 1H), 7.84 (d, J = 8.1 Hz,
(3S,4R,5R)-3-Bromo-3-fluoro-4-((triisopropylsilyl)oxy)-5-
(((triisopropylsilyl)oxy)methyl)tetrahydrofuran-2-yl
benzoate
(7c). To a solution of compound 6c (3.67 g, 6.74 mmol) in pyridine
(18 mL) was added 4-dimethylaminopyridine (80.6 mg, 0.660 mmol)
followed by slow addition of benzoyl chloride (1.1 mL, 9.48 mmol). A
white solid appeared after a few minutes and the suspension was stirred
at room temperature for 1 h. The suspension was concentrated at re-
duced pressure, diluted with heptanes, and filtered through a pad of
Celite® and rinsed with heptanes. The filtrate was concentrated to give
an oil which was purified by flash chromatography eluting with a sol-
vent gradient of 0–35% dichloromethane in heptane and gave com-
pound 7c (4.07 g, 93% yield) as a colorless oil. 1H NMR (500 MHz,
CDCl3) δ ppm 1.06–1.31 (m, 67Η), 3.86 (dd, J = 11.8, 2.4 Hz, 0.6Η),
3.95–4.07 (m, 4Η), 4.32 (qd, J = 4.3, 1.0.6 Hz, 1.2Η), 4.87 (dd,
J = 4.2, 2.0 Hz, 1.2Η), 4.96 (dd, J = 19.6, 7.4 Hz, 0.6Η), 6.42 (d,
J = 9.5 Hz, 0.6Η), 6.73 (s, 1Η), 7.41–7.47 (m, 4Η), 7.56–7.61 (m, 2Η),
8.02–8.05 (m, 1.2Η), 8.08–8.11 (m, 2Η).
1H), 11.57 (s, 1H); MS (DCI) m/z 342.1 (M + NH4)+
.
(S)-Isopropyl 2-(((S)-(((2R,3R,4R,5R)-4-bromo-4-chloro-5-(2,4-
dioxo-3,4-dihydropyrimidin-1(2H)-yl)-3-hydroxytetrahydrofuran-
2-yl)methoxy)(phenoxy)phosphoryl)amino)propanoate (16a). To a
solution of compound 9a (1.5 g, 4.39 mmol) in tetrahydrofuran
(12.0 mL) and 1,3-dimethyltetrahydropyrimidin-2(1H)-one (1.5 mL) at
0 °C, was added a tetrahydrofuran solution of tert-butylmagnesium
chloride (1 M, 4.39 mL, 4.39 mmol). The resulting suspension was
stirred at 0 °C for 30 min and (S)-isopropyl 2-(((S)-(perfluorophenoxy)
(phenoxy)phosphoryl)amino)propanoate (3.98 g, 8.78 mmol) was
added. The resultant mixture was allowed to warm to room tempera-
ture. After 1.5 h an aqueous solution of half saturated NaHCO3 (75 mL)
and then extracted with ethyl acetate (3 × 50 mL). The combined or-
ganic layers were dried over MgSO4 and concentrated. The crude pro-
duct was purified by flash chromatography on silica eluting with a
solvent gradient of 0–5% methanol in dichloromethane to provide
compound 16a (2.17 g, 81% yield) containing 6 wt% of 1,3-di-
methyltetrahydropyrimidin-2(1H)-one by NMR. 1H NMR (400 MHz,
DMSO‑d6) δ ppm 1.14 (d, J = 6.3 Hz, 6Η), 1.21 (d, J = 7.1 Hz, 3Η),
3.79 (m, 1Η), 4.00 (m, 1Η), 4.08 (m, 1Η), 4.30 (m, 2Η), 4.84 (m, 1Η),
5.56 (d, J = 8.2 Hz, 1Η), 6.08 (dd, J = 13.1, 10.1 Hz, 1Η), 6.62 (s, 1Η),
6.99 (d, J = 5.5 Hz, 1Η), 7.20 (m, 3Η), 7.36 (m, 2Η), 7.63 (d,
J = 8.1 Hz, 1Η), 11.58 (s, 1Η); 13C NMR (101 MHz, DMSO‑d6) δ 20.26
(d, J = 6.5 Hz), 21.84, 21.89, 50.25, 64.58, 68.52, 77.00, 80.23, 84.59,
102.55, 120.51 (d, J = 4.9 Hz), 125.14, 130.17, 150.72, 151.09 (d,
J = 6.5 Hz), 163.09, 173.05; 31P NMR (162 MHz, DMSO‑d6) δ 3.28. MS
(Z)-N-(1-((2R,3S,4R,5R)-3-Bromo-3-fluoro-4-((triisopropylsilyl)
oxy)-5-(((triisopropylsilyl)oxy)methyl)tetrahydrofuran-2-yl)-2-
oxo-1,2-dihydropyrimidin-4-yl)benzimidic acid (8c). To a suspen-
sion of N-(2-oxo-2,3-hydroxypyrimidin-4-yl)benzamide (4.66 g,
21.66 mmol) in chlorobenzene (27 mL) was added trimethylsilyl N-
(trimethylsilyl)acetimidate (5.5 mL, 22.41 mmol). The suspension was
stirred at 80 °C for 45 min then cooled to room temperature. The so-
lution was then added to compound 7c (4.07 g, 6.29 mmol) via cannula
followed by addition of freshly distilled (114–116 °C, 760 mmHg) tin
tetrachloride (3.5 mL, 29.8 mmol) at room temperature, and then the
solution was heated to 106–114 °C. After 16 h the suspension was
slowly poured into an aqueous solution of 1 M NaHCO3 (200 mL) with
vigorous stirring at room temperature, rinsing the reaction flask with
methyl tert-butyl ether (50 mL). The mixture was stirred at room tem-
perature for 1 h, filtered through Celite®, and the solid was rinsed with
methyl tert-butyl ether (100 mL). The layers of the filtrate were sepa-
rated, and the organic layer was washed with water (20 mL), 1 N HCl
(15 mL) and brine (15 mL). The filter cake and the reaction flasks were
rinsed with dichloromethane. The organic layers were combined, dried
over MgSO4 and concentrated to give a yellow oil (5.85 g), which was
purified by flash chromatography on silica eluting with a solvent gra-
dient of 0–25% ethyl acetate in heptanes to give impure product. The
resulting material was re-purified by flash chromatography on silica
(ESI) m/z 612.0 (Μ + Η)+
.
(S)-Isopropyl 2-(((S)-(((2R,3R,4S,5R)-4-bromo-5-(2,4-dioxo-3,4-
dihydropyrimidin-1(2H)-yl)-4-fluoro-3-hydroxytetrahydrofuran-2-
yl)methoxy)(phenoxy)phosphoryl)amino)propanoate
(16c).
Compound 16c was prepared using a method similar to that described
for the synthesis of compound 16a. Using compound 9c (26 mg,
0.080 mmol),
tert-butylmagnesium
chloride
(1 M,
0.080 mL,
0.080 mmol), and (S)-isopropyl 2-(((R)-(perfluorophenoxy)(phenoxy)
phosphoryl)amino)propanoate (72.5 mg, 0.160 mmol) provided com-
pound 16c (31 mg, 62.6% yield) as a light yellow solid. 1H NMR
(400 MHz, DMSO‑d6) δ ppm 1.14 (d, J = 2.2 Hz, 6H), 1.23 (dd, J = 7.2,
1.0 Hz, 3H), 3.74–3.86 (m, 1H), 4.00–4.05 (m, 1H), 4.22–4.40 (m, 2H),
4.86 (hept, J = 6.3 Hz, 1H), 5.63 (d, J = 8.2 Hz, 1H), 6.06 (dd,
J = 13.0, 10.0 Hz, 1H), 6.23 (d, J = 17.7 Hz, 1H), 6.65 (d, J = 7.1 Hz,
1H), 7.16–7.24 (m, 3H), 7.35–7.40 (m, 2H), 7.56 (d, J = 8.1 Hz, 1H),
11.58 (s, 1H); 13C NMR (126 MHz, DMSO‑d6) δ 19.68 (d, J = 6.7 Hz),
21.30, 21.34, 49.69, 64.35, 67.95, 75.60 (d, J = 16.9 Hz), 78.96,
102.38, 109.11 (d, J = 260.9 Hz), 120.01 (d, J = 4.8 Hz), 124.58,
129.62, 150.04, 150.55 (d, J = 6.4 Hz), 162.60, 172.50, 172.54; 19F
NMR (376 MHz, DMSO‑d6) δ −118.29; 31P NMR (162 MHz, DMSO‑d6)
eluting with
a solvent gradient of 0–25% ethyl acetate in di-
chloromethane to give compound 8c (230 mg, 0.311 mmol, 4.94%
yield) as an off-white solid. 1Η NMR (500 MHz, DMSO‑d6) δ ppm
0.99–1.26 (m, 42Η), 3.98 (dd, J = 12.2, 2.9 Hz, 1Η), 4.05 (d,
J = 8.5 Hz, 1Η), 4.17 (dd, J = 12.1, 2.4 Hz, 1Η), 4.70 (dd, J = 15.5,
8.2 Hz, 1Η), 6.45 (d, J = 16.0 Hz, 1Η), 7.41 (s, 1Η), 7.52 (t, J = 7.7 Hz,
4Η), 7.60–7.66 (m, 2Η), 7.88–7.93 (m, 2Η), 7.97–8.07 (m, 3Η), 11.40
(s, 2Η); MS (APCI) m/z 741.36 (Μ + Η)+
1-((2R,3S,4R,5R)-3-Bromo-3-fluoro-4-hydroxy-5-(hydro-
xymethyl)tetrahydrofuran-2-yl)pyrimidine-2,4(1H,3H)-dione (9c).
Compound 9c was prepared using a method similar to that described
for the synthesis of compound 9a. Using compound 8c (230 mg,
0.311 mmol), the reaction mixture, upon completion, was diluted with
dichloromethane, filtered through Celite® and rinsed with 10% me-
thanol in dichloromethane. The filtrate was concentrated to a solid,
which was purified by flash chromatography on silica eluting with a
solvent gradient of 0–10% methanol in dichloromethane and gave an
δ 3.24; MS (ESI) m/z 593.9 (M + H)+
.
Ethyl 2-(((S)-(((2R,3R,4R,5R)-4-bromo-4-chloro-5-(2,4-dioxo-
3,4-dihydropyrimidin-1(2H)-yl)-3-hydroxytetrahydrofuran-2-yl)
methoxy)(phenoxy)phosphoryl)amino)-2-methylpropanoate (18a)
and Ethyl 2-(((R)-(((2R,3R,4R,5R)-4-bromo-4-chloro-5-(2,4-dioxo-
3,4-dihydropyrimidin-1(2H)-yl)-3-hydroxytetrahydrofuran-2-yl)
methoxy)(phenoxy)phosphoryl)amino)-2-methylpropanoate
(18 l). Compounds 18a and 18l were prepared using a method similar
11

A. Chris Krueger, et al.
Bioorganic&MedicinalChemistryxxx(xxxx)xxxx
to that described for compound 16a. Using compound 9a (554 mg,
Appendix A. Supplementary material
1.622 mmol),
tert-butylmagnesium
chloride
(1 M,
1.784 mL,
1.784 mmol), and ethyl 2-methyl-2-(((perfluorophenoxy)(phenoxy)
phosphoryl)amino)propanoate (1103 mg, 2.433 mmol) provided
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.bmc.2019.115208.
a
mixture of the products, which was purified by chiral SFC using a Regis
Whelk-O1 (S,S) column, eluting with 20% methanol in supercritical
CO2 to provide 18a (417 mg, 42.1% yield) (the second of two stereo-
isomers to elute) and 18l (176 mg, 17.8% yield) (the first of two ste-
reoisomers to elute). For 18a: 1H NMR (500 MHz, DMSO‑d6) δ ppm
1.12 (t, J = 7.1 Hz, 3Η), 1.33 (s, 3Η), 1.37 (s, 3Η), 3.97–4.14 (m, 4Η),
4.25–4.40 (m, 2Η), 5.52 (dd, J = 8.2, 2.2 Hz, 1Η), 5.96 (d, J = 9.6 Hz,
1Η), 6.62 (s, 1Η), 7.01 (s, 1H), 7.14–7.24 (m, 3Η), 7.36 (dd, J = 8.6,
7.3 Hz, 2Η), 7.66 (d, J = 8.2 Hz, 1Η), 11.58 (d, J = 2.2 Hz, 1Η); 13C
NMR (126 MHz, DMSO-d6) δ 13.85, 26.87 (d, J = 5.4 Hz), 26.96 (d,
J = 6.7 Hz), 56.02, 60.60, 63.87, 76.47, 79.64, 84.15, 101.95, 120.10
(d, J = 4.7 Hz), 124.52, 129.56, 150.19, 150.64 (d, J = 6.5 Hz),
162.55, 174.58; 31P NMR (162 MHz, DMSO‑d6) δ 1.83; MS (ESI) m/z
611.8 (M + H)+. For 18l: 1H NMR (400 MHz, DMSO‑d6) δ 1.14 (t,
J = 7.1 Hz, 3H), 1.34 (s, 3H), 1.36–1.42 (m, 3H), 3.98–4.09 (m, 4H),
4.26–4.35 (m, 1H), 4.39 (dd, J = 11.8, 5.3 Hz, 1H), 5.53 (dd, J = 8.2,
2.1 Hz, 1H), 6.02 (d, J = 9.7 Hz, 1H), 6.65 (s, 1H), 7.06 (s, 1H), 7.17 (t,
J = 7.3 Hz, 1H), 7.22 (dt, J = 8.6, 1.1 Hz, 2H), 7.37 (dd, J = 8.6,
7.3 Hz, 2H), 7.62 (d, J = 8.2 Hz, 1H), 11.59 (d, J = 2.2 Hz, 1H); MS
References
1. Hatzakis A, Chulanov V, Gadano AC, et al. The present and future disease burden of he-
patitis C virus (HCV) infections with today’s 3 treatment paradigm. - volume 2. J. Viral
Hepat. 2015;22:26–45 www.who.int/mediacentre/factsheets/fs164/en/; access date
March 19, 2018.
[2]. Centers for disease control and prevention website–www.cdc.gov/hepatitis/statistics/;
see press release https://www.cdc.gov/nchhstp/newsroom/2016/hcv-press-release.
html.
[3]. Messina JP, Humphreys I, Flaxman A, et al. Global distrubition and prevalence of he-
patitis C virus genotypes. Hepatology. 2015;61:77–87.
[4]. (a) Feld J, Jacobson IM, Hezode C, et al. Sofosbuvir and velpatasvir for HCV genotype 1,
2, 4, 5, and 6 infection. N Engl J Med. 2015;373:2599–2607
(b) Bhardwaj N, Chodavarapu K, Martin R, et al. Intrapatient viral diversity and treat-
ment outcome in patients with genotype 3a hepatitis C virus infection on sofosbuvir-
containing regimens. J Viral Hepatitis. 2018;25:344–353
(c) Struble K, Chan-Tack K, Qi K, Naeger LK, Birnkrant D. Benefit-risk assessment for
sofosbuvir/velpatasvir/voxilaprevir based on patient population and hepatitis C virus
genotype: U. S. Food and Drug Administration’s evaluation. Hepatology (Hoboken, NJ, U.
S.). 2018;67:482–491
(d) Younossi ZM, Stepanova M, Jacobson IM, et al. Sofosbuvir and velpatasvir with or
without voxilaprevir in direct-acting antiviral-naive chronic hepatitis C: Patient-reported
outcomes from POLARIS 2 and 3. Aliment Pharmacol Ther. 2018;47:259–267
(e) Boglione L, Pinna SM, Lupia T, Cariti G, Di Perri G. Retreatment with sofosbuvir/
velpatasvir in cirrhotic patients with genotype 4 who failed a previous interferon-free
regimen: A case series. Antiviral Ther. 2018. https://doi.org/10.3851/IMP3226.
[5]. (a) Zeuzem S, Foster GR, Wang S, et al. Glecaprevir-pibrentasvir for 8 or 12 weeks in
HCV genotype 1 or 3 infection. N Engl J Med. 2018;378:354–369
(ESI) m/z 611.8 (M + H)+
.
Ethyl
2-(((S)-(((2R,3R,4S,5R)-4-bromo-5-(2,4-dioxo-3,4-dihy-
dropyrimidin-1(2H)-yl)-4-fluoro-3-hydroxytetrahydrofuran-2-yl)
methoxy)(phenoxy)phosphoryl)amino)-2-methylpropanoate
(18c). To a solution of compound 9c (1.31 g, 4.04 mmol) in tetra-
hydrofuran (13 mL) and 1,3-dimethyltetrahydropyrimidin-2(1H)-one
(DMPU) (1.3 mL) at an internal temperature of 1.1 °C was added a
tetrahydrofuran solution of phenylmagnesium chloride (2 M, 2.5 mL,
5.00 mmol) dropwise over 5 min keeping the internal temp less than
3 °C. The viscous off-white slurry was stirred for 30 min, then (S)-ethyl
2-methyl-2-(((perfluorophenoxy)(phenoxy)phosphoryl)amino)pro-
panoate18 (2.01 g, 4.44 mmol) was added as a solid. The suspension was
allowed to slowly warm to room temperature while stirring for 16 h,
and then quenched with 1 N HCl (15 mL, internal temp 18 to 27 °C) and
extracted with ethyl acetate (3 × 15 mL). The combined organic layers
were washed with brine (2 × 10 mL) and 4:1 brine/1N HCl (10 mL),
dried over sodium sulfate, filtered and concentrated to give 4.2 g of a
yellow oil which was purified by flash chromatography on silica eluting
with a solvent gradient of 30–75% ethyl acetate in dichloromethane to
give a white foam (1.9563 g). The impure product was dissolved in
isopropyl acetate (20 mL), washed with water (2 × 10 mL), brine
(10 mL), then dried over sodium sulfate, filtered and concentrated to
give 18c (1.66 g, 66.9% yield) as a white solid. HPLC showed 96.9 area
% desired product at 210 nm, with 0.26 area% DMPU and 0.11 area% i-
PrOAc. nH NMR (500 MHz, DMSO‑d6) δ ppm 1.12 (t, J = 7.1 Hz, 3Η),
1.33 (s, 3Η), 1.37 (s, 3Η), 4.02 (q, J = 7.1 Hz, 3Η), 4.22–4.41 (m, 3Η),
5.58 (d, J = 8.1 Hz, 1Η), 5.92 (d, J = 9.6 Hz, 1Η), 6.22 (d, J = 17.7 Hz,
1Η), 6.66 (s, 1Η), 7.13–7.23 (m, 2Η), 7.32–7.39 (m, 2Η), 7.56 (d,
J = 8.1 Hz, 1Η), 11.39 (s, 1Η).
(b) Lin C-W, Dutta S, Zhao W, Asatryan A, Campbell A, Liu W. Pharmacokinetic inter-
actions and safety of coadministration of glecaprevir and pibrentasvir in healthy vo-
lunteers. Eur J Drug Metab Pharmacokinet. 2018;43:81–90
(c) Lin C-W, Dutta S, Asatryan A, et al. Pharmacokinetics, safety, and tolerability fol-
lowing single and multiple doses of pibrentasvir in a first-in-human study. Clin Pharmacol
Drug Dev. 2018;7:44–52
(d) Kwo PY, Poordad F, Asatryan A, et al. Glecaprevir and pibrentasvir yield high re-
sponse rates in patients with HCV genotype 1–6 without cirrhosis. J Hepatol.
2017;67:263–271.
[6]. (a) Sofia MJ. Sofosbuvir: the discovery of a curative therapy for the treatment of he-
patitis C virus. Successful Drug Discovery, 2. p. 163–88. DOI: 10.1002/9783527800315.
ch8.
(b) Gentile I, Maraolo AE, Buonomo AR, Zappulo E, Borgia G. The discovery of so-
fosbuvir: a revolution for therapy of chronic hepatitis C. Expert Opinion on Drug
Discovery, 10. p. 1363–77. DOI: 10.1517/17460441.2015.1094051.
(c) Kumar S, Jacobson IM. Antiviral therapy with nucleotide polymerase inhibitors for
chronic hepatitis C. J Hepatol. 2015;61:S91–S97
[7]. Pradere U, Garnier-Amblard EC, Coats SJ, Amblard F, Schinazi RF. Synthesis of nu-
cleoside phosphate and phosphonate prodrugs. Chem Rev. 2014;114:9154–9218.
[8]. Coats SJ, Garnier-Amblard EC, Amblard F, et al. Chutes and ladders in hepatitis C nu-
cleoside drug development. Antiviral Res. 2014;102:119–147.
[9]. (a) Appleby TC, Perry JK, Murakami E, et al. Viral replication. Structural basis for RNA
replication by the hepatitis C virus polymerase. Science. 2015;347:771–775
(b) Migliaccio G, Tomassini JE, Carroll SS, et al. Characterization of resistance to non-
obligate chain terminating ribonucleoside analogs that inhibit hepatitis C virus replica-
tion. in Vitro. J Biol Chem. 2003;278:49164–49170.
[10]. (a) Pinho P, Kalayanov G, Westerlind H, et al. Discovery of β -D-2’-deoxy-2’-di-
chlorouridine nucleotide prodrugs as potent inhibitors of hepatitis C virus replication.
Bioorg Med Chem Lett. 2017;27:3468–3471
(b) Chen Z, Cox BD, Garnier-Amblard EC, et al. Synthesis and anti-HCV activity of a
series of β -D-2’-deoxy-2’-dibromo nucleosides and their corresponding phosphoramidate
prodrugs. Bioorg Med Chem Lett. 2017;27:5296–5299
(c) Zhou S, Mahmoud S, Liu P, et al. 2’-Chloro,2’-fluoro ribonucleotide prodrugs with
potent pan-genotypic activity against hepatitis C virus replication in culture. J Med Chem.
2017;60:5424–5437
(d) Mengshetti S, Zhou L, Sari O, et al. Discovery of a series of 2’- α -fluoro,2’- β -bromo-
ribonucleosides and their phosphoramidate prodrugs as potent pan-genotypic inhibitors
of hepatitis c virus. J Med Chem. 2019;62:1859–1874.
[11]. Cen Y, Sauve AA. Diastereo-controlled electrophilic fluorinations of 2-deoxyr-
ibonolactone: syntheses of all corresponding 2-deoxy-2-fluoro-lactones and 2'-deoxy-2'-
fluoro-NAD+s. J Org Chem. 2009;74:5779–5789.
Declaration of Competing Interest
[12]. A mixture of halogenated products were generated and the ratio was determined by 1H
NMR as 5a: C2-epi:5e:5f∼17:5:2:1.
The authors declare the following competing financial interest(s):
All authors were employees of AbbVie at the time of the study. ACK,
HC, JTR, BSB, GTH, HRH, TL, CCM, JPS, EAV, DB, CVH, VP, RAC, DS,
TD, PK, RFH, RW, and DAD are current employees of AbbVie. MLI is a
former AbbVie employee. The design, study conduct, and financial
support for this research were provided by AbbVie. AbbVie participated
in the interpretation of data, review, and approval of the publication.
[13]. Chiral SFC analysis of the products determined 85% 9a and 14% 9e.
[14]. Rakhlin VI, Grigor’ev SV, Mirskov RG, et al. Preparative synthesis of N-bromohex-
amethyldisilazane. Russian J General Chem. 2003;73:1952–1953.
[15]. Kalayanov G, Torssell S, Wahling H. Preparation of nucleotides as hepatitis C virus
polymerase inhibitors US 20150175648A1.
[16]. Chau J, Chen HJ, DeGoey DA, et al. Anti-viral compounds US 10053474 Β2.
[17]. Voight EA, Brown BS, Greszler SN, et al. Synthesis of ABBV-168, a 2'-bromouridine for
the treatment of hepatitis C. J Org Chem. 2019;84:4723–4734.
12

A. Chris Krueger, et al.
Bioorganic&MedicinalChemistryxxx(xxxx)xxxx
[18]. Barreca G, Rasparini M, Poggiali M, et al. Nucleoside phosphoramidates useful for the
treatment of viral infections and preparation thereof PCT Int. Appl. WO2016151542A1.
[19]. Krishnan P, Beyer J, Mistry N, et al. In vitro and in vivo antiviral activity and resistance
profile of ombitasvir, an inhibitor of hepatitis C virus NS5A. Antimicrob Agents Chemother.
2015;59:979–987.
[22]. Murakami E, Tolstykh T, Bao H, et al. J Biol Chem. 2010;285:34337–34347.
[23]. Compound 16c is described in a recent publication by other researchers.10d The data we
report for 16c agrees well with data included in that publication.
[24]. Bruker AXS Inc., APEX2,version 2014.11-0, Madision Wisconsin; 2014.
[25]. Sheldrick GM. University of Göttingen, Germany; 1998.
[20]. Alexandre F-R, Badaroux E, Bilello JP, et al. The discovery of IDX21437: design,
synthesis and antiviral evalution of a 2’- α -chloro2’-β -c-methyl branched uridine pro-
nucleotides as potent liver-targeted HCV polymerase inhibitors. Bioorg Med Chem Lett.
2017;27:4323–4330.
[26]. Middleton T, He Y, Pilot-Matias T, et al. A replicon-based shuttle vector system for as-
sessing the phenotype of HCV NS5B polymerase genes isolated from patient populations.
J Virol Methods. 2007;145:137–145.
[27]. Tripathi RL, Krishnan P, He Y, et al. Replication efficiency of chimeric replicon con-
taining NS5A-5B genes derived from HCV-infected patient sera. Antiviral Res.
2007;73:40–49.
[21]. Randolph JT, Li T, Krueger AC, et al., Discovery of 2-aminoisobutyric acid ethyl ester
(AIBEE) phosphoramidate prodrugs for delivering nucleoside HCV NS5B polymerase
inhibitors, submitted to Bioorg Med Chem Lett.
13