Development and Implementation of an Aluminum-Promoted Phosphorylation in the Uprifosbuvir Manufacturing Route

Article abstract of DOI:10.1021/acs.oprd.0c00487

A novel application of the synthesis of pronucleotide (ProTide) 5′-phosphoramidate monoesters promoted by aluminum-based Lewis acids is described. In the multikilogram synthesis of uprifosbuvir (MK-3682, 1), a clinical candidate for the treatment of hepatitis C, this methodology provided >100:1 diastereoselectivity at the phosphorus stereocenter and >100:1 selectivity for the 5′-mono phosphorylation over undesired bisphosphorylation side products. The high diastereoselectivity and mono/bis ratio achieved enabled elimination of the tedious workup associated with the tert-butyl magnesium chloride protocol commonly used to install this functionality in similar nucleotide prodrugs, achieving a near doubling of the isolated yield from 45% to 81%. The process development and purity control strategy of MK-3682, as well as handling of the pyrophoric reagent on scale, will also be discussed.

Full text of DOI:10.1021/acs.oprd.0c00487

pubs.acs.org/OPRD
Communication
Development and Implementation of an Aluminum-Promoted
Phosphorylation in the Uprifosbuvir Manufacturing Route
Zhuqing Liu,* Artis Klapars, Bryon Simmons, Ana Bellomo, Alexei Kalinin, Mark Weisel, Jerry Hill,
and Steven M. Silverman*
Cite This: https://dx.doi.org/10.1021/acs.oprd.0c00487
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: A novel application of the synthesis of pronucleotide (ProTide) 5′-phosphoramidate monoesters promoted by
aluminum-based Lewis acids is described. In the multikilogram synthesis of uprifosbuvir (MK-3682, 1), a clinical candidate for the
treatment of hepatitis C, this methodology provided >100:1 diastereoselectivity at the phosphorus stereocenter and >100:1
selectivity for the 5′-mono phosphorylation over undesired bisphosphorylation side products. The high diastereoselectivity and
mono/bis ratio achieved enabled elimination of the tedious workup associated with the tert-butyl magnesium chloride protocol
commonly used to install this functionality in similar nucleotide prodrugs, achieving a near doubling of the isolated yield from 45%
to 81%. The process development and purity control strategy of MK-3682, as well as handling of the pyrophoric reagent on scale,
will also be discussed.
KEYWORDS: ProTide synthesis, Lewis acid-mediated phosphorylation, dimethylaluminum chloride, uprifosbuvir, MK-3682,
pyrophoric reagents
INTRODUCTION
RESULTS AND DISSCUSSION
■
■
Hepatitis C virus (HCV) infection is a disease of global impact,
with more than 170 million infected and over 70 million
presenting chronic hepatitis C and requiring treatment.1
Chronic hepatitis C can lead to a spectrum of conditions
affecting the liver, ranging from inflammation to cirrhosis and
cancer.2 Uprifosbuvir (MK-3682) is a nucleoside-based
prodrug for the treatment of HCV, which, along with other
clinical candidates and approved drugs,3 utilizes the 5′-
aryloxyphosphoramidate or “ProTide” moiety introduced by
McGuigan et al. in the early 1990s as a strategy to enable
dramatic enhancement of cellular permeability and phosphor-
ylation rates.4 The original phosphorylation of uprifosbuvir
penultimate was performed under basic conditions using tert-
butylmagnesium chloride following the conditions for the
preparation of sofosbuvir (Scheme 1).5 While this process was
sufficient to provide material for early clinical studies, there
were several challenges to address to develop a process suitable
to provide material for late stage clinical studies and
commercial supplies, including the following: (1) The reaction
was stopped at 85% conversion to avoid increased levels of bis-
phosphorylation, wasting 15% of valuable penultimate and
increasing the purification burden. (2) The selectivity of
mono- vs bisphosphorylation was low (∼6:1). (3) A tedious
workup procedure was applied to remove a series of impurities.
And, (4) even after this workup, multiple crystallizations were
required to remove the bis-phosphorylation byproduct and
obtain the desired form. These complications led to a 45%
overall yield and long cycle times. Here, we present the
invention and process development of a Lewis-acid-promoted
phosphorylation approach to the preparation of uprifosbuvir.
Reaction Optimization. We previously identified general
conditions demonstrating the utility of dimethylaluminum
chloride as a Lewis acid promoter of highly chemo- and
diastereoselective phosphorylation reactions.6 The scope of the
transformation was demonstrated to be quite broad, and the
method could be applied to the synthesis of uprifosbuvir
(Scheme 2), but the use of pyridine as a solvent and the
required chromatography initially rendered the developed
conditions unsuitable for use in the uprifosbuvir API step,
requiring significant reaction development to render the
process suitable for the production of kilogram quantities of 1.
Our initial investigation of alternative phosphorylation
conditions in the synthesis of API 1 revealed multiple
aluminum-based Lewis acids to be effective in promoting the
reaction, including Et2AlCl, i-Bu2AlCl, AlCl3, Al(Ot-Bu)3, and
DABCO-(AlMe3)2, all affording the product in >50:1
diastereoselectivity and mono- to bis-chemoselectivity at
>95% conversion. Upon further analysis of the reaction profile
and API isolation with this series of reagents, we concluded
that Me2AlCl was the most suitable reagent for large scale
synthesis of MK-3682.7 While Al(Ot-Bu)3 and DABCO-
(AlMe3)2 were more air stable than Me2AlCl, the limited
supply could not meet the projected commercial demand.
Special Issue: Celebrating Women in Process Chem-
istry
Received: October 30, 2020
https://dx.doi.org/10.1021/acs.oprd.0c00487
Org. Process Res. Dev. XXXX, XXX, XXX−XXX
© XXXX American Chemical Society
A

Organic Process Research & Development
pubs.acs.org/OPRD
Communication
Scheme 1. Uprifosbuvir First Generation Supply Route
Scheme 2. Selective Aluminum-Promoted Phosphorylation in the Synthesis of Uprifosbuvir
a
Table 1. Optimization of Phosphorylation Conditions
entry
base
NEt3
pyridine
2,6-lutidine
NMM
DBU
Me2AlCl (equiv)
solvent
THF
THF
THF
THF
THF
THF
THF
THF
THF
THF
THF
THF
conversion (%)
dr (1:5)
ratio 1:4
1
2
3
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.2
0.3
0.4
0.5
0.7
0.9
0.5
0.5
0.5
0.5
0.5
0.5
0.5
91
59
73
87
91
97
93
84
94
95
98
98
97
1
88
90
96
96
97
93
10
>100
>100
39
11
58
>100
>100
>100
>100
67
>100
>100
121
186
460
320
192
72
b
4
5
6
b
b
NMI
7
8
9
2,4,6-collidine
2,6-lutidine
2,6-lutidine
2,6-lutidine
2,6-lutidine
2,6-lutidine
2,6-lutidine
2,6-lutidine
2,6-lutidine
2,6-lutidine
2,6-lutidine
2,6-lutidine
2,6-lutidine
2,6-lutidine
41
4.3
14.3
46
101
96
95
nd
40
45
37
63
58
31
10
11
12
13
14
15
16
17
18
19
20
THF
heptane
CPME
i-PrOAc
DME
MeCN
2-MeTHF
DMF
nd
44
145
94
49
67
232
a
Nucleoside 2 (1.0 equiv), 3 (1.2 equiv), base (1.5 equiv), Me2AlCl (0.5 equiv. unless otherwise indicated), 0.2 M, room temperature for 24 h,
b
then 50 °C for 4 h, THF as solvent unless otherwise indicated. Aged for 72 h at room temperature.
B
https://dx.doi.org/10.1021/acs.oprd.0c00487
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
pubs.acs.org/OPRD
Communication
Figure 1. Diagram of dimethylaluminum chloride transfer procedure.
Additionally, the corrosive nature of AlCl3 increased the
difficulty in implementing this reagent on scale. Of the
alkylaluminum chloride options, we observed a small amount
of alkyl group transfer to phosphorus in the case of i-Bu2AlCl
and Et2AlCl, but no methyl transfer was observed with
Me2AlCl. Therefore, we chose to optimize the reaction further
with this reagent knowing that implementation of appropriate
procedural and engineering controls would be required to
allow safe handling of the reagent.
additional Lewis acid (entries 11−13). No benefit to reaction
performance was observed when more than 0.5 equiv was used,
where diastereoselectivity was observed to be highest, leading
to its selection for further optimization.
To further optimize the reaction, we performed a broad
solvent screen. Low conversion was observed in heptane, likely
due to the exceedingly low solubility of nucleoside 2 under
these conditions (entry 14). In many other polar and nonpolar
solvents, thick slurries were observed (entries 15−18) due to
the low solubility of the reactants and product, which, despite
the good selectivity achieved, was deemed problematic from a
processing perspective. Nearly homogeneous solutions were
observed in solvents such as THF, 2-MeTHF, and DMF
(entries 11, 19−20), with THF (entry 11) chosen for further
development as the reaction showed the cleanest overall
profile.
Me2AlCl Workup Process. Aluminum salts are prone to
form precipitates or emulsions under basic conditions,
rendering standard aqueous workup very challenging. Indeed,
our initial isolation attempts on the lab scale resulted in
substantial product loss attributable to entrainment of the API.
Aqueous sodium potassium tartrate solutions have been used
for aluminum workups due to the strong chelating effect on the
Lewis acid.8 For this process, we observed similar results using
the more cost-effective tartaric acid, with a rapid phase split
observed when an aqueous solution of the acid was added to
the THF reaction stream. Accordingly, at the end of reaction, a
25−30 wt % tartaric acid aqueous solution (approximately 30%
by volume relative to the total amount of solution) was added
to the reaction mixture, leading to rapid phase separation and
no emulsion. Unfortunately, we observed approximately 5%
product loss in a subsequent aqueous wash to remove residual
tartaric acid remaining in the organic layer, and it took several
hours to achieve complete phase separation; neither of these
observations is surprising given the aqueous miscibility of
THF. The addition of isopropyl acetate at this stage facilitated
phase separation of the aqueous wash and minimized the
overall API loss to approximately 1%. After crystallization,
In optimizing the aluminum-promoted phosphorylation for
production of API 1, we examined several factors key to
obtaining a suitable reaction profile on scale (Table 1). Trends
around diastereoselectivity and mono/bis chemoselectivity
were observed to correlate with base strength. Lower
diastereoselectivity was observed with stronger bases such as
NEt3 and DBU (entries 1, 4, 5), likely due to the presence of
additional pentafluorophenoxide, which was observed in
control experiments to mediate epimerization of unreacted 3.
In contrast, higher diastereoselectivity was obtained using
weaker bases such as pyridine and lutidine (entries 2−3). The
reaction rate was also found to correlate with basicity, with
faster rates observed with a stronger base. Both NMI and 2,6-
lutidine offered an appropriate balance of rate and reaction
profile, with the latter being selected for further development
due to superior reaction properties as NMI produced a gum-
like precipitate during lab runs, which we anticipated would
cause challenges on scale. A brief survey of lutidine
stoichiometry indicated that the presence of at least 0.75
equiv was necessary for the reaction to achieve full conversion.
We selected 1.5 equiv as the preferred charge since this
amount safely provided full conversion of starting material with
only a slight decrease in diastereoselectivity relative to the
starting phosphoramidate.
The effect of equivalents of Me2AlCl added on the reaction
was next studied (entries 8−13). We observed an increase in
the reaction rate and diastereoselectivity with a concomitant
increase in the quantity of Me2AlCl employed up to 0.5 equiv.
The highest chemoselectivity was achieved at 0.4 equiv.
Me2AlCl, with a decrease observed in the presence of
C
https://dx.doi.org/10.1021/acs.oprd.0c00487
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
pubs.acs.org/OPRD
Communication
uprifosbuvir could be obtained in high purity with less than 10
ppm aluminum in the isolated solid.
5 at levels that were easily purged during crystallization, during
our initial scaleups, we observed a structurally related impurity
in ∼1.5 LCAP. We inferred it to be cyclic phosphoramidate 6
(Figure 3) based on its mass, and upon standing for
Demonstration of Kilogram Scale Process. On the lab
scale, a 1 M solution of Me2AlCl in heptane can be handled
using typical safety precautions inherent to flammable
organometallic solutions. However, due to commercial
availability on scale, it was necessary to utilize neat Me2AlCl
for our kilo lab runs, and given its pyrophoric nature and water
reactivity, it was necessary to implement additional procedural
and engineering controls. The risks associated with this process
were minimized by transferring neat Me2AlCl from a cylinder
into a blow can containing the desired amount of heptane to
reduce its interaction with air and moisture as shown in Figure
1 (red line). This transfer allowed for an accurate mass
measurement while keeping the entire system inert. The
Me2AlCl solution could then be charged slowly to the reactor
containing a THF solution of the other reaction components
under moderate pressure at 0 °C (Figure 1, black line), and
after transfer, the line was flushed with heptane to prevent any
buildup of dimethylaluminum chloride.9 After addition of the
Lewis acid, the reaction mixture was aged at room temperature
for 12 h before the temperature was raised to 50 °C for 3 h,
which we found to substantially improve overall conversion
while maintaining high diastereoselectivity, giving 97.7%
conversion (95 LCAP 1) with a >100:1 diastereoselectivity
and mono/bis ratio.
Figure 3. Structure of observed cyclic phosphoramidate impurity 6.
approximately 3 days, we observed precipitation of the
impurity from the mother liquor that enabled us to isolate
and fully characterize it to confirm its identity. It also allowed
additional crystallization studies necessary to formulate a
control strategy. The solubility of impurity 6 in the final
crystallization solvent system was measured to be only 0.9 mg/
mL. As a result, this impurity could pose a risk to the isolated
API purity on scale due to the longer cycle times required. To
address this concern, we designed a control strategy through
deeper understanding of the impurity formation, thereby
ensuring that the level remained low during the reaction and
mitigating risk to batch quality associated with potential
crystallization of the impurity.
The improved reaction purity profile attributed to the Lewis
acid-promoted phosphorylation allowed for API to be
crystallized much more easily than the previous generation
process. After tartaric acid workup, high purity product could
be obtained from multiple crystallization solvent systems. This
flexibility was of additional benefit as over 20 forms of
uprifosbuvir have been identified, some showing dramatically
reduced bioavailability. We knew that crystallization from
isopropanol/heptane provided the desired API form, and
fortunately, a single crystallization provided MK-3682 in the
desired form and in high purity. Under these conditions, a
second crystallization was not needed for purity upgrade or
form turnover. On the kilogram scale, we observed the growth
of large crystals unsuitable for formulation studies (Figure 2a).
The slow crystallization kinetics of impurity 6 were
confirmed with seeding studies (Table 2). Elevated levels of
Table 2. LCAP of Impurity 6 Present in Isolated Product
time
seeded with pure MK-3682 seeded with MK-3682 and 1.5% 6
15 h
<0.02
0.03
0.5
<0.02
0.03
0.73
0.78
2 days
3 days
4 days
1.08
the impurity were observed in the isolated product after aging
the crystallization mixture for 3−4 days; these levels were
observed to be similar to those observed when MK-3682 was
coseeded with 1.5 wt % phosphoramidate 6. This observation
is consistent with a slow, unseeded crystallization of this
impurity from the solvent system being used to crystallize
uprifosbuvir and caused a concern on scale due to the
processing time.
Reinvestigating our optimized conditions, we observed the
level of the cyclic impurity 6 to increase linearly with the
amount of lutidine used in the reaction (Figure 4).10 The
reaction temperature, however, played a more profound role in
formation of the impurity, as only a moderate increase in the
reaction temperature led to a significant increase in impurity
formation during the reaction. For example, when 1.25 equiv
of 2,6-lutidine was used, a change in internal temperature from
35 to 45 °C increased the impurity level from 1.7% to 4.8% at
36 h under the reaction conditions. We found that the impurity
level in the reaction could be controlled at 0.5−0.6 LCAP at
temperatures not exceeding 35 °C for a period of 14−16 h.
These conditions provided formation of uprifosbuvir in
approximately 95% conversion and 92 LCAP. These finalized
conditions (Scheme 3) were employed in multiple pilot plant
campaigns and provided high quality material on the
multikilogram scale.
Figure 2. Control of uprifosbuvir particle size through wet milling.
Fortunately, a wet milling process was developed to readily
decrease the particle size to meet our D50 < 50 μm
specification (Figure 2b). From this campaign, uprifosbuvir
was isolated in 82% yield and 99.7 LCAP.
Reoptimization of Reaction Conditions: In-process
Control of Impurity 6. While the typical end-of-reaction
profile was very clean, containing only residual starting
materials, bis-phosphorylated byproduct 4, and diastereomer
D
https://dx.doi.org/10.1021/acs.oprd.0c00487
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
pubs.acs.org/OPRD
Communication
Figure 4. Formation of 6 over time at 35−45 °C. Reaction conditions: nucleoside (1.0 equiv), 2 (1.2 equiv), 2,6-lutidine (as shown), Me2AlCl (0.5
equiv), 0.2 M in THF.
Scheme 3. Dimethylaluminum Chloride-Promoted Process for the Synthesis of Uprifosbuvir 1
jacket temperature was then set to −10 °C, and dimethyla-
luminum chloride (∼1 M solution in heptane premade from
neat reagent in the blow can, 1.895 kg) was charged under
nitrogen pressure over approximately 1 h. The charge
apparatus was rinsed with heptane (2 × 375 mL). The
resulting suspension was warmed to 30−35 °C and aged for
15−16 h. The batch was cooled to an internal temperature of
10 °C, and a 30 wt % tartaric acid solution (8.14 kg) was
added, followed by isopropyl acetate (4.5 L). The aqueous
phase was separated, and the organic layer was washed with a 3
wt % aqueous sodium chloride solution (2 × 7.5L). A constant
volume distillation was then performed adding isopropanol to
the reactor until the solvent composition reached <5 mol %
THF, as determined by 1H NMR spectroscopy. A final
concentration of the organic layer was performed to reach an
11.8 L total volume. The solution was filtered at 50 °C, and the
batch was heated to an internal temperature of 50−60 °C and
seeded (45 g seeds). Heptane (12 L) was charged slowly to the
batch while maintaining a constant internal temperature, at
which point the batch was slowly cooled to 20 °C and allowed
to crystallize over 14 h. The batch was heated to 60 °C and wet
milled for 3 h to meet the desired particle size specifications,
after which it was cooled to 20 °C, filtered, and washed with
isopropanol/heptane (1:1). The wet cake was dried under a
CONCLUSIONS
■
We have shown the development and implementation of an
efficient ProTide synthesis promoted by dimethylaluminum
chloride that proceeds with high chemoselectivity for mono- vs
bisphosphorylation and with high diastereoselectivity in the
uprifosbuvir manufacturing route. The formation of a cyclic
phosphoramidation impurity was identified and studied, and a
control strategy was developed to minimize any associated risk
to the process. When techniques developed to safely handle
the pyrophoric Me2AlCl were used, this process was
successfully executed on a multikilogram scale to produce
API meeting all required specifications for use in late-stage
clinical trials.
EXPERIMENTAL SECTION
■
Synthesis of Uprifosbuvir (1). To a 100 L glass lined
vessel equipped with a nitrogen inlet, overhead stirrer, and
thermocouple was charged THF (21.5 L) followed by 1-
((2R,3R,4R,5R)-3-chloro-4-hydroxy-5-(hydroxymethyl)-3-
methyltetrahydrofuran-2-yl)pyrimidine-2,4(1H,3H)-dione
(1.500 kg, 5.42 mol) and (R)-isopropyl 2-(((R)-(perfluoro
phenoxy)(phenoxy)phosphoryl)amino)propanoate (2.95 kg,
6.51 mol). The reactor was cooled to an internal temperature
of 0 °C. 2,6-Lutidine (0.944 L, 8.13 mol) was charged slowly
to the vessel while maintaining a temperature of <5 °C. The
1
vacuum at 40 °C to give 1 (2.41 kg, 81%). H NMR (500
E
https://dx.doi.org/10.1021/acs.oprd.0c00487
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
pubs.acs.org/OPRD
Communication
MHz, DMSO-d6): δ (ppm) 11.64−11.40 (s, 1H), 7.74−7.49
(d, J = 8.1 Hz, 1H), 7.48−7.32 (t, J = 7.9 Hz, 2H), 7.32−7.22
(d, J = 8.3 Hz, 2H), 7.24−7.11 (t, J = 7.3 Hz, 1H), 6.33−6.21
(s, 1H), 6.21−6.06 (m, 2H), 5.62−5.52 (dd, J = 8.1, 1.7 Hz,
1H), 4.95−4.68 (hept, J = 6.2 Hz, 1H), 4.50−4.35 (dd, J =
11.4, 3.7 Hz, 1H), 4.35−4.20 (m, 1H), 4.17−4.01 (d, J = 9.2
Hz, 1H), 3.92−3.67 (ddt, J = 24.4, 10.0, 7.0 Hz, 2H), 3.39−
3.24 (s, 1H), 1.50−1.31 (s, 3H), 1.31−1.19 (d, J = 7.1 Hz,
3H), 1.18−1.05 (dd, J = 8.5, 6.4 Hz, 6H). 13C NMR (125
MHz, DMSO): δ 172.61, 172.57, 162.62, 150.69, 150.64,
150.50, 129.68, 124.55, 119.90, 119.86, 102.06, 80.03, 77.73,
72.34, 67.99, 63.59, 49.68, 22.37, 21.33, 21.28, 19.74, 19.69.
LC-MS calcd for C22H30ClN3O9P [M + H]+ m/z: 546.14.
Found: 546.2.
ACKNOWLEDGMENTS
■
The authors would like to thank Aaron Moment, Morgan
Crawford, and Darryl Chang for discussions and experimental
contributions. We thank Ryan Cohen and Rose Mathew for
help with impurity characterization. We thank Rebecca Ruck,
Paul Bulger, Neil Strotman, and L.-C. Campeau for proof-
reading suggestions.
ABBREVIATIONS USED
■
V = volume
LCAP = HPLC/UPLC area percentage
IY = isolated yield
AY = assay yield
REFERENCES
ASSOCIATED CONTENT
■
■
(1) Hepatitis C. https://www.who.int/news-room/fact-sheets/
detail/hepatitis-c.
(2) (a) Manns, M. P.; Buti, M.; Gane, E.; Pawlotsky, J. M.; Razavi,
H.; Terrault, N.; Younossi, Z. Hepatitis C Virus Infection. Nat. Rev.
Dis. Prim 2017, 3, 17006. (b) Hepatitis C. https://www.who.int/
news-room/fact-sheets/detail/hepatitis-c.
(3) (a) Chang, W.; Bao, D.; Chun, B. K.; Naduthambi, D.;
Nagarathnam, D.; Rachakonda, S.; Reddy, P. G.; Ross, B. S.; Zhang,
H. R.; Bansal, S.; et al. Discovery of PSI-353661, a Novel Purine
Nucleotide Prodrug for the Treatment of HCV Infection. ACS Med.
Chem. Lett. 2011, 2, 130. (b) Liao, L.; Wang, G.; Yokokawa, F. WO
sı
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.oprd.0c00487.
NMR characterization of specified impurities is provided
along with spectral data (PDF)
AUTHOR INFORMATION
■
Corresponding Authors
́
2014/082935A, Jun 5, 2014. (c) Cui, H.; Ruiz-Perez, L. M.;
Zhuqing Liu − Department of Process Research and
Development, Merck & Co., Inc., Rahway, New Jersey 07065,
United States; Email: zhuqing_liu@merck.com
Steven M. Silverman − Department of Process Research and
Development, Merck & Co., Inc., Rahway, New Jersey 07065,
United States; orcid.org/0000-0003-0792-7071;
Email: steven.silverman@merck.com
́
Gonzalez-Pacanowska, D.; Gilbert, I. H. Potential Application of
Thymidylate Kinase in Nucleoside Analogue Activation in Plasmo-
dium Falciparum. Bioorg. Med. Chem. 2010, 18, 7302. (d) Jiang, P.;
Guo, X.; Fu, H.; Jiang, Y.; Zhao, Y. One-Pot Synthesis of Aryl
Phosphoramidate Derivatives of AZT/d4T as Anti-HIV Prodrugs.
Synlett 2005, 16, 2537. (e) Zhou, X. J.; Pietropaolo, K.; Sullivan-
Bolyai, J.; Kuca, B.; Liu, W.; Xu, L.; Belanger, B.; Khan, S.; Mayers, D.
J. Resistance to Direct Antiviral Agents in Patients With Hepatitis C
Virus Infection. J. Hepatol. 2009, 50, S351. (f) Sommadossi, J.-P.;
Gosselin, G.; Pierra, C.; Perigaud, C.; Peyrottes, S. Compounds and
Pharmaceutical Compositions for the Treatment of Viral Infections.
US 7902202 B2, Mar 08, 2011. (g) Vernachio, J. H.; Bleiman, B.;
Bryant, K. D.; Chamberlain, S.; Hunley, D.; Hutchins, J.; Ames, B.;
Gorovits, E.; Ganguly, B.; Hall, A.; Kolykhalov, A.; Liu, Y.;
Muhammad, J.; Raja, N.; Walters, C. R.; Wang, J.; Williams, K.;
Patti, J. M.; Henson, G.; Madela, K.; Aljarah, M.; Gilles, A.;
McGuigan, C.. INX-08189, a Phosphoramidate Prodrug of 6-O-
Methyl-2′-C-Methyl Guanosine, Is a Potent Inhibitor of Hepatitis C
Virus Replication with Excellent Pharmacokinetic and Pharmacody-
namic Properties, Antimicrob. Agents Chemother.. 2011, 55, 1843.
(4) (a) McGuigan, C.; Pathirana, R. N.; Mahmood, N.; Devine, K.
G.; Hay, A. J. Aryl Phosphate Derivatives of AZT retain Activity
Against HIV1 in Cell Lines Which Are Resistant to the Action of
AZT. Antiviral Res. 1992, 17, 311. (b) McGuigan, C.; Pathirana, R.
N.; Mahmood, N.; Hay, A. J. Aryl Phosphate Derivates of AZT Inhibit
HIV Replication in Cells Where the Nucleoside is Poorly Active.
Bioorg. Med. Chem. Lett. 1992, 2, 701.
Authors
Artis Klapars − Department of Process Research and
Development, Merck & Co., Inc., Rahway, New Jersey 07065,
United States
Bryon Simmons − Department of Process Research and
Development, Merck & Co., Inc., Rahway, New Jersey 07065,
United States
Ana Bellomo − Department of Process Research and
Development, Merck & Co., Inc., Rahway, New Jersey 07065,
United States
Alexei Kalinin − Department of Process Research and
Development, Merck & Co., Inc., Rahway, New Jersey 07065,
United States
Mark Weisel − Department of Process Research and
Development, Merck & Co., Inc., Rahway, New Jersey 07065,
United States
Jerry Hill − Department of Process Research and Development,
Merck & Co., Inc., Rahway, New Jersey 07065, United States
(5) (a) Uchiyama, M.; Aso, Y.; Noyori, R.; Hayakawa, Y. J. Org.
Chem. 1993, 58, 373. (b) Ross, B. S.; Ganapati Reddy, P.; Zhang, H.
R.; Rachakonda, S.; Sofia, M. J. Synthesis of Diastereomerically Pure
Nucleotide Phosphoramidates. J. Org. Chem. 2011, 76, 8311.
(6) Simmons, B.; Liu, Z.; Klapars, A.; Bellomo, A.; Silverman, S. M.
Mechanism-Based Solution to the ProTide Synthesis Problem:
Selective Access to Sofosbuvir, Acelarin, and INX-08189. Org. Lett.
2017, 19, 2218.
(7) This decision was made after taking into account both the
reaction purity profiles observed with each reagent and the ability to
implement a process on the anticipated commercial scale. While some
reagents appeared comparable to Me2AlCl in terms of reaction profile,
Complete contact information is available at:
https://pubs.acs.org/10.1021/acs.oprd.0c00487
Author Contributions
The manuscript was written through contributions of all
authors. All authors have given approval to the final version of
the manuscript.
Notes
The authors declare no competing financial interest.
F
https://dx.doi.org/10.1021/acs.oprd.0c00487
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
pubs.acs.org/OPRD
Communication
their limited availability or corrosive nature proved prohibitive for
multikilogram use.
(8) (a) Eastman, J. F.; Teranishi, R. Δ4-Cholesten-3-one. Org.
Synth., Coll. Vol. 1963, 4, 192. (b) Burgstahler, A. W.; Nordin, I. C.
The Secondary Structures of Nucleic Acids in Organic Solvents. J. Am.
Chem. Soc. 1961, 83, 198. (c) King, A. O.; Corley, E. G.; Anderson, R.
K.; Larsen, R. D.; Verhoeven, T. R.; Reider, P. J.; Xiang, Y. B.; Belley,
M.; Leblanc, Y. An Efficient Synthesis of LTD4 Antagonist L-699,392.
J. Org. Chem. 1993, 58, 3731.
(9) To protect the scientists executing the procedure, fire-resistant
aluminum lab coats (Fisher Part Number 19-049676) were used as an
added precaution.
(10) For instance, the impurity increased from 1% with 0.75
equivalents of lutidine to 2% with 1.5 equivalents of lutidine after
aging at 35 °C for 36 h.
G
https://dx.doi.org/10.1021/acs.oprd.0c00487
Org. Process Res. Dev. XXXX, XXX, XXX−XXX