Stereoselective Synthesis of Sofosbuvir through Nucleoside Phosphorylation Controlled by Kinetic Resolution

Article abstract of DOI:10.1002/ejoc.201800158

The preparation of Sofosbuvir, the potent key component of recent Hepatitis C (HCV) infection therapies, is reported. The process is based on the dynamic kinetic resolution of the stereochemically unstable isopropyl-2-{[chloro(phenoxy)phosphoryl]-amino}propanoate (8). A high stereoselectivity was obtained when the right protective group for 3′-OH was chosen. Ester and carbonate-based protective groups gave lower stereoselectivities, but benzyl protection allowed the phosphorylation to occur with a 92:8 ratio in favour of the product with the right configuration at the P-stereogenic centre. Starting from the γ-lactone of 2-deoxy-2-fluoro-2-methylpentonic acid, the synthesis was accomplished in eight steps in 40 % overall yield using commercially available reagents, and without any enzymatic or chemical resolution technique.

Full text of DOI:10.1002/ejoc.201800158

A Journal of
Accepted Article
Title: Stereoselective synthesis of Sofosbuvir through nucleoside
phosphorylation controlled by kinetic resolution
Authors: Elena Cini, Giuseppe Barreca, Luca Carcone, Fabrizio
Manetti, Marcello Rasparini, and Maurizio Taddei
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Eur. J. Org. Chem. 10.1002/ejoc.201800158
Link to VoR: http://dx.doi.org/10.1002/ejoc.201800158
Supported by

10.1002/ejoc.201800158
European Journal of Organic Chemistry
FULL PAPER
Stereoselective synthesis of Sofosbuvir through nucleoside
phosphorylation controlled by kinetic resolution
[b]
[b]
[a]
Elena Cini,[a] Giuseppe Barreca,
Maurizio Taddei*[a]
Luca Carcone,
Fabrizio Manetti,
Marcello Rasparini,[b]‡ and
Abstract: The preparation of Sofosbuvir, the potent key component
of recent Hepatitis C (HCV) infection therapies, is reported here. The
process is based on the dynamic kinetic resolution of the
stereochemically unstable isopropyl-2-((chloro(phenoxy)phosphoryl)-
amino)propanoate (8). The high stereoselection obtained is due to the
correct choice of the protective group at 3’-OH. While ester and
carbonate-based protections gave an inferior stereoselection, benzyl
protection allowed phosphorylation to occur in 92:8 ratio in favour of
the product with the right configuration at the P-stereogenic centre.
Starting from the γ-lactone of 2-deoxy-2-fluoro-2-methyl-pentonic
acid, the synthesis was accomplished in 8 steps in 40 % overall yield
using commercially available reagents and without any enzymatic or
chemical resolution technique.
nucleosides (as Sofosbuvir) must include the stereoselective
formation of a P-stereogenic center. Amongst different strategies
employed for ProTide syntheses, the nucleophilic substitution on
a phosphorochloridate with the nucleoside nucleophile is the most
common preparative method. Unfortunately, this approach gives
a
mixture of diasteroisomers that must be separated by
chromatography or other methods, with loss of valuable
material.[4]
potential solution to this approach is the
A
transformation of the phosphorochloridate into the corresponding
pentafluorophenol derivative that is sterochemically stable at the
P-center.[5] After fractional crystallization, the correct isomer is
subject to nucleophile substitution that occurs with almost
complete inversion of configuration giving the required
stereochemistry at P. In the case of Sofosbuvir (or analogue
nucleotides) with a free 3’OH in the final structure, a selective
reaction at the 5’OH is also required. This problem was originally
approached with the use of protective groups at 3’ and recently
with the use of a Lewis acid that selectively activate the 5’OH for
reaction with the isopropyl ((S)-(perfluorophenoxy)(phenoxy)-
phosphoryl)-L-alaninate (Scheme 1).[6] Finally, a group of Merck
scientists published recently the use of a chiral catalyst for the
stereoselective assemble of ProTide nucleosides starting from
the stereochemically unstable isopropyl-2-((chloro(phenoxy)-
Introduction
Sofosbuvir is an antiviral agent approved for treatment of Hepatitis
C Virus (HCV) infections.[1] The remarkable activity of this
molecule is related to the presence of a phosphoramide moiety in
position 5’ of the nucleoside. This pro-drug approach, called
ProTide, allows: i) a substantial increase of the cell permeability
of the molecule ii) the generation of the nucleotide
monophosphate in cell after a simple metabolic transformation
phosphoryl)amino)propanoate
(Scheme
1).
Using
the
unprotected nucleoside, a tailor made enantiomerically pure
catalyst is required to reach a high level of regio- and
stereoselectivity.[7]
with subsequent increase of in vivo phosphorylation rate.[2]
A
comprehensive SAR study revealed that an enantiomerically pure
L amino acid linked to the phosphorous is required for the first
step of the ProTide metabolism, generating a cyclic intermediate
through substitution of the aryloxy moiety bonded to the
phosphorous. Final ring opening and hydrolysis of the
phoshoramide produces the monophosphate that is rapidly
transformed into the active triphosphate metabolite. This
architecture requires a chiral phosphorous and the configuration
may influence activity, metabolic transformation and even
toxicity.[3] Thus, an efficient synthesis of ProTide modified
[a]
E. Cini, F. Manetti, M.Taddei
Dipartimento di Biotecnologie, Chimica e Farmacia,
Università degli Studi di Siena
Via A. Moro2 I-53100, Siena (Italy)
E-mail: maurizio.taddei@unisi.it
[b]
‡
G.Barreca. L. Carcone, M. Rasparini
Chemessentia srl
Via Bovio 2, I-28100; Novara (Italy)
Actual address: Janssen Pharmaceutica, Turnhoutseweg 30, 2340
Beerse (Belgium).
Scheme 1. Most recent synthetic approaches to Sofosbuvir (ref 5-7). In blue
the ProTide appendage.
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/ejoc.201800158
European Journal of Organic Chemistry
FULL PAPER
Following our interest in the synthesis of relevant active
pharmaceutical ingredients (APIs),[8,9] engaged in developing a
convenient preparation of Sofosbuvir, we decided to investigate
the possibility to carry out a dynamic resolution of the racemic
phosphochloridate using the enantiomerically pure nucleoside
itself as the selector.
best results in terms of selectivity and conversion (entries 2 and
10 in Table 1).
The working hypothesis was to study the influence of the size of
the protective group at 3’OH and the nature of the base employed
to activate the OH on phosphorylation stereoselectivity. Some
examples of phosphorylation of protected nucleosides have been
previously described,[10–12] although with low selectivity. However,
a systematic study on the influence of the protection at the 3’
position has been never reported and also the influence of the
base employed on the stereoselectivity has been not deeply
investigated.
Results and Discussion
Thus, nucleosides 3-5 were prepared starting from nucleoside 2
through the classic transient protection at 5’ with TBDMS, further
acylation of the 3’ OH and final desilylation with TBAF (Scheme
2).[11]
Scheme 3. Phosphorylation of 3’-protected nucleosides 3-5.
Table 1. Influence of the presence of salts in phosphorylation of 3
Entry
1
Salt[a]
Conversion (%)[b]
dr (9a:9b))[c]
65:35
73:27
65:35
64:26
65:35
66:34
67:33
65:35
63:37
83:17
62:38
-
94
85
67
96
46
86
70
96
44
90
41
2
LiCl
3
NaCl
KCl
4
5
CsF
6
LiF
7
CsCl
KBr
8
9
KI
Scheme 2. Protection at 3’OH of nucleoside 2. For 3,
hydroxysuccinimide, for 5, X = Cl.
X = N-
10
11
LiCl (complex)
ZnCl2
As deprotection of compound 9 (Scheme 3) was anticipated to be
the easiest to perform, the optimization of reaction conditions
started on nucleoside 3 (Scheme 3). Phosphorylation of 3 was
attempted with chloro(phenoxy)phosphoryl)amino)propanoate 8
prepared in situ from phenyl dichlorophosphate 6 and the iso-
propyl ester of L-alanine (Scheme 3). After activation of the 5’-OH
of 3 with t-BuMgCl, different salts were added in the reaction
mixture in order to study the influence of the cation (and probably
the resulting aggregation) on the stereoselectivity. As reported in
Table 1, the presence of different salts did not have a dramatic
influence on the stereoselectivity of the reaction. Increase of the
cation size, anion size or the presence of ammonium salt in
solution did not improve the selectivity that resembled that of the
reaction without additives. The use of LiCl (added) or the use of
the complex iPrMgCl LiCl, the so-called Turbo-Grignard,[13] gave
a] Reaction conditions: [b] Determined respect to unreacted nucleoside by
HPLC analysis. [c] Determined by HPLC using isopropyl ((S)-bis-
(phenoxy)phosphoryl)-L-alaninate as internal standard.
The more hindered Boc derivative 4 gave comparable results in
terms of conversion and selectivity (Figure 1). The benzoyl
derivative 5 gave better conversion again with the preformed LiCl
complex with the Grignard, but always an almost equimolar
amount of the two diasteroisomers (Figure 1). Although not
completely satisfactory, the Cbz protection at 3’ and the
nucleoside activation at 5’ with the Turbo-Grignard were
encouraging enough to further investigate other benzyl
protections.[9] Surprisingly, the 3’O-benzyl derivative (15 in
Scheme 5) has been never reported before, we prepared it,
following a synthetic approach comparable to that applied for
compounds 3-5
This article is protected by copyright. All rights reserved.

10.1002/ejoc.201800158
European Journal of Organic Chemistry
FULL PAPER
% of 10a
conversion of 4 into 10
% of 11a
analysis of the product (SI), confirming the decisive solvent effect
in nucleoside alkylation.
Compound 15 was then submitted to reaction with 8 after
activation with iPrMgCl LiCl complex and compound 16 was
obtained with very high conversion (99%) and a still excellent dr
(92:8). This product was readily deprotected with H2 over Pd/C in
MeOH to give product 1 in almost quantitative yield (Scheme 5).
The product obtained in 89 % yield after a single crystallization,
showed the same spectroscopic and physical characteristics of
the original Sofosbuvir.
conversion of 5 into 11
reaction on 4
reaction on 5
Figure 1. Influence of the presence of salts in phosphorylation of 4 and 5.
Unfortunately, after silylation of the 5’OH, NaH in DMF was
employed in order to activate the 3’OH and exclusively the
dibenzyl derivative 13 was obtained even after increasing the ratio
between 12 and NaH /benzyl bromide (Scheme 4). However,
when compound 13 was submitted to our best reaction conditions,
the phosphorylated compound 14 was obtained with high
conversion and a very high dr (97:3 in favor of the correct
diastereoisomer 14a). [9]
Scheme 5. Transformation of 3’-benzylderivative 15 into Sofosbuvir.
Although very efficient, this procedure might be considered less
economical in terms of atoms employed and number of steps with
respect to the catalytic procedures that, starting from 2, do not
use protection at 3’OH.[6,7,15] However, all the published
syntheses of compound 2,[15,16] described starting from
commercially available lactone 17 (Scheme 6) requires the use of
protections. A first full protection of the two hydroxyl groups with
benzoyl chloride is required (18) before performing the
introduction of the nucleobase as 4-benzoylcytosine (Scheme
6).[12] Thus, after reduction of the carbonyl of 18 and further
acetylationof the OH of 19, the nucleobase 20 is silylated to
product 21 prior the reaction, to give fully protected compound 22.
The protective groups of 22 are removed with ammonia in MeOH
and finally the 4-benzoylcytosine is transformed into uracil with
hot acetic acid[16] to give nucleoside 2 (Scheme 6, reactions inside
the red frame). Compound 2 may then be transformed in a single
step using selective phosphorylation at 5’OH cited in Scheme 1 to
give Sofosbuvir in a 7 steps procedure from 17 .[5–7] Consequently,
as protection at the OH is required for nucleobase insertion, we
decided to introduce also our different protective groups at 3’ and
5’ OHs on lactone 18, obtaining compound 23 in 67% yield
(Scheme 6). Reduction of the lactone followed by acetylation and
cytosine introduction gave acceptable yields (71% overall) of
product 25. The acid environment required for transformation of
the benzoyl-cytosine into uracil allowed the contemporary
Scheme 4. Preparation and phosphorylation of the dibenzyl derivative 13
Unfortunately, deprotection of 14a to Sofosbuvir
1 was
troublesome and conversion to pure 1 was possible only after a
multistep purification from several by-products.[9]
Thus, we tried to find a method for the selective protection of the
OH in 3’ without affecting the uracil NH. We came across a paper
that describes the selective 3’-O- or 3-N-protection of 5’-O-
TBDMS-thymidines using aprotic polar solvents with low or high
dielectric constant, respectively.[14] The reaction was described
under ultrasound or microwave activation, but we were pleased to
discover that reacting 12 with 3 eq of NaH in THF and 1 eq of
benzyl bromide at -10°C for 12 h, exclusively the monobenzylated
compound 15 was obtained in 63% overall yield. The correct
position of the benzyl was independently established by X-ray
This article is protected by copyright. All rights reserved.

10.1002/ejoc.201800158
European Journal of Organic Chemistry
FULL PAPER
removal of the TBDMS group at 5’ position, giving compound 15
in good yields.
be formed, favoring the formation of the P(R) configuration on the
final phosphorylated product.
Conclusions
Phosphoramidate nucleotides are the key constituents of the
ProTide approach to prodrugs successfully employed in therapies
against viral pathologies as HCV infections. While previous
procedures are based on standard racemate resolution, use of
chiral auxiliaries or complex catalysts specifically designed for the
process, we have developed a stereoselective method for the
introduction of the P-stereogenic center via kinetic dynamic
resolution of the stereochemically labile chloro(phenoxy)-
phosphoryl)amino)propanoate,[18] using the nucleoside protected
at 3’ with a benzyl group as chiral selector. Only this expedient is
required to obtain high stereoselectivity and all our synthesis
proceeds with commercially available reagents. Extension of this
process to other therapeutically relevant ProTide nucleotides are
currently under investigation.
Experimental Section
Benzyl ((2R,3R,4R,5R)-5-(2,4-dioxo-3,4-dihydropyrimidin-1(2H)-yl)-4-
fluoro-2-(hydroxymethyl)-4-methyltetrahydrofuran-3-yl)
carbonate
(3). General procedure. To a solution of (2'R)-2'-deoxy-2'-fluoro-2'-
methyluridine 2 (1.87 g,7.2 mmol) in anhydrous pyridine (15 mL) at 0 °C
under N2, imidazole (0.62 g, 9,2 mmol) and tert-butyldimethylsilyl chloride
(1.20 g, 7.85 mmol) were added. The reaction mixture was stirred at r.t. for
12 h and then quenched with methanol (5 mL). After 1 h of additional
stirring at r.t. the solvent was removed under vacuum. The residue was
purified by silica gel column chromatography (petroleum ether/EtOAc 1:4
v/v) to give 12 (2.46 g) in 93% yield. This product was dissolved in dry
CH2Cl2 (50 mL) and N-(benzyloxycarbonyloxy)succinimide (1.97 g, 7.92
mmol) was added under N2. The reaction mixture was cooled to 0 °C and
DMAP (176 mg, 1.44 mmol.) followed by Et3N (1.42 mL, 10.8 mmol,) was
added. The reaction mixture was slowly warmed to room temperature and
stirred for 16 h. A 1 M HCl solution was added (20 mL), the phases
separated, the aqueous phase extracted with CH2Cl2 and the collected
organic phases were washed with brine, dried over Na2SO4 and
concentrated. The residue was purified by flash column chromatography
(petroleum ether/EtOAc 8:2) to give the O-Cbz protected product (1.11 g,
75% yield) as a colourless oil. This compound was solubilized in THF (22
mL), treated with TBAF (2.07 g, 3.51 mmol,) and stirred for 3 h at room
temperature. The solvent was evaporated and the product purified by
column chromatography (EtOAc) to give 3 (1.45 g, 68%) as a white solid
(m.p. 153-154 °C) with the same spectroscopic data described in the
literature.[11] Anal. calcd for C18H19FN2O7: C 54.82, F 4.82, H 4.86, N 7.10,
O 28.40; found: C 54.79, H 4.83, N 7.07.
Scheme 6. Comparison of our Sofosbuvir synthesis with the previously
described one (in the red dotted line frame) starting from the common
intermediate lactone 17.
Finally, the transformation into Sofosbuvir, as described in
Scheme 6, required additional two reactions. Overall, our
approach, based on the dynamic resolution of the easily
accessible chloride 8, employs commercially available (low cost)
reagents and provides Sofosbuvir in 40% isolated yield starting
from available lactone 17. The process is realized in 8 steps, only
one step more (the debenzylation at 3’)[17] if compared with the
previous synthesis based on the selective phosphorylation at 5’
with a tailor made chiral catalyst.[7]
In order to rationalize the variation in stereoselectivity observed
changing the protection at 3’OH, phosphorylation of 15 was
monitored by HPLC analysis in different conditions. The
selectivity did not change during reaction time, with a value of 95:5
at 5% conversion. The amount of 8 employed had no influence on
selectivity; however, in the presence of less than 1 eq of 8, no
trace of 16ab was detected in the reaction medium.
A
computational protocol consisting in B3LYP/6-31G** calculations
using THF and PBF as the solvent and solvation model,
respectively, in combination with a systematic pseudo Monte
Carlo conformational routine[7] was attempted, but the locations of
Mg and Li ions made results difficult to rationalize. We observed
that the presence of a carbonyl in position 3’ of the nucleoside
stabilizes an intramolecular complex with Mg, while, with a benzyl
ether, this arrangement is not possible. Probably, with benzyl
ether 15 an intermolecular chelated transition state with 8 might
tert-Butyl ((2R,3R,4R,5R)-5-(2,4-dioxo-3,4-dihydropyrimidin-1(2H)-yl)-
4-fluoro-2-(hydroxymethyl)-4-methyltetrahydrofuran-3-yl) carbonate
(4): Compound 4 was prepared following the general procedure above
using in Boc2O in place of CbzOSu; 43% overall yield as a white solid m.p.
1
159-161 °C. H NMR (400 MHz, CDCl3) δ = 9.95 (s, 1H), 7.96 (d, J = 8.1
Hz, 1H), 6.15 (d, J = 18.2 Hz, 1H), 5.73 (d, J = 8.1 Hz, 1H), 4.94 (dd, J =
22.3, 9.3 Hz, 1H), 4.16 (d, J = 9.2 Hz, 1H), 4.05 (dd, J = 12.7, 3.8 Hz, 1H),
3.75 (d, J = 12.7 Hz, 1H), 3.33 (s, 1H), 1.56 – 1.23 (m, 12H) . 13C NMR
(100 MHz, CDCl3): δ = 163.66, 153.11, 150.57, 140.00, 102.94, 99.80,
89.31, 84.39, 79.80, 73.18 (d, J = 15.9 Hz), 59.40, 27.56 (3C), 16.98 (d, J
This article is protected by copyright. All rights reserved.

10.1002/ejoc.201800158
European Journal of Organic Chemistry
FULL PAPER
= 25.0 Hz). Anal. calcd for C15H21FN2O7: C 50.00, F 5.27, H 5.87, N 7.77,
O 31.08; found: C 49.05, H 5.85, N 7.74
Hz, 1H), 7.49 – 7.13 (m, 5H), 5.97 (d, J = 19.1 Hz, 1H), 5.66 (d, J = 8.0 Hz,
1H), 5.38 (s, 1H), 4.69 (s, 2H), 4.06-3.84 (m, 3H), 3.67 (d, J = 12.5 Hz,
1H), 1.28 (d, J = 22.5 Hz, 3H) . 13C NMR (100 MHz, DMSO) δ 162.88,
150.62, 139.77, 138.00, 128.33 (2C), 127.87 (3C), 102.16, 100.47 (d, J =
181.3 Hz), 88.48 (d, J = 39 Hz), 80.92, 77.78 (d, J = 16.3 Hz), 73.14, 58.83,
17.23 (d, J = 17.23 Hz). Anal. calcd for C17H19FN2O5: C 58.28, F 5.42, H
5.47, N 8.00, O 22.83; found: C 58.23, H 5.45, N 7.97.
(2R,3R,4R,5R)-5-(2,4-Dioxo-3,4-dihydropyrimidin-1(2H)-yl)-4-fluoro-
2-(hydroxymethyl)-4-methyltetrahydrofuran-3-yl
benzoate
(5):
Compound 5 was prepared following the general procedure above using
benzoyl chloride in place of CbzOSu: 47% overall yield as a white solid
1
m.p. 248-250 °C. H NMR (400 MHz, DMSO) δ = 11.51 (s, 1H), 8.00 (m,
3H), 7.68 (t, J = 7.4 Hz, 1H), 7.54 (m, 2H), 6.10 (d, J = 19.1 Hz, 1H), 5.72
(d, J = 8.1 Hz, 1H), 5.55 – 5.18 (m, 2H), 4.28 (d, J = 8.9 Hz, 1H), 3.94 –
3.52 (m, 2H), 1.33 (d, J = 23.0 Hz, 3H). 13C NMR (100 MHz, DMSO) δ
164.92, 162.90, 150.58, 139.86, 134.01, 129.58 (2C), 128.92 (2C), 128.56,
102.49, 100.30 (d, J = 183 Hz), 88.97, 79.94, 71.74 (d, J = 13 Hz), 58.67,
17.39 (d, J = 24 Hz). Anal. calcd for C17H17FN2O6: C 56.04, F 5.21, H 4.70,
N 7.69, O 26.35; found: C 56.00, H 4.67, N 7.66,
(2S)-Isopropyl
2-(((((2R,3R,4R,5R)-3-(benzyloxy)-5-(2,4-dioxo-3,4-
dihydropyrimidin-1(2H)-yl)-4-fluoro-4-methyltetrahydrofuran-2-
yl)methoxy)(phenoxy)phosphoryl)amino)propanoate (16). To a flask
provided with a mechanical stirrer, reflux condenser, thermometer and
under N2, (L)-alanine (5.0 g, 56.1 mmol) was mixed with a solution of
hydrogen chloride in i-propanol (1 1 .1 % w/w, 73.8 g, 224.5 mmol). The
reaction mixture was heated to boiling (80-85 °C) for 4 hours. Once the
conversion was complete, (monitoring in TLC by eluting with 7:3 ethanol-
water and developing with ninhydrin) the solution was concentrated to
residue in vacuum to give (L)-alanine isopropyl ester hydrochloride as a
dense oil (9.4 g quantitative yield). To a flask provided with mechanical
stirrer, reflux condenser, thermometer and under N2, (L)-alanine isopropyl
ester hydrochloride (9.4 g, 56.1 mmol) was mixed with methyl t-butyl ether
(80 mL) and the resulting mixture was stirred at room temperature until a
homogeneous suspension was obtained. The mixture was then cooled to
-55 °C and Et3N (13.0 g, 128.5 mmol) was added, looking to maintain the
system at a temperature below -50 °C. At the end of the addition, the
3-Benzyl-1-((2R,3R,4R,5R)-4-(benzyloxy)-3-fluoro-5-(hydroxymethyl)-
3-methyltetrahydrofuran-2-yl)pyrimidine-2,4(1H,3H)-dione (13):
A
solution of 12 (500 mg, 1.34 mmol prepared as described in the above
general procedure) in dry DMF (3 mL) was cooled to 0 °C . NaH (110 mg
of 60% dispersion in mineral oil, 2.7 mmol) was slowly added having care
to maintain the temperature below 5 °C. After 10 min of stirring at 5°C,
benzyl bromide (0.320 mL, 2.7 mmol) was added dropwise. When the
addition was complete, the temperature was raised to 25 °C and the
mixture was kept under stirring until the conversion was complete,
monitoring by TLC with petroleum ether /EtOAc 7:3 (about 15 h). The
reaction mixture, cooled to 0 °C, was diluted with diethyl ether. Phases
were separated, and the aqueous phase was extracted with diethyl ether;
the organic phases were combined and dried over Na2SO4, filtered and
concentrated under vacuum. The residue was purified by flash column
chromatography (petroleum ether/ EtOAc 7:3) to give O,N-di-Bn protected
product (638 mg, 86% yield) as a colourless oil that was solubilized in THF
(15 mL) and treated with TBAF (543 mg, 1.72 mmol,) for 3 h at room
temperature. The solvent was evaporated, and the product purified by
column chromatography (EtOAc) to give 13 as a white solid (320 mg,
62%). m.p. 97-99 °C. 1H NMR (400 MHz, CDCl3) δ 7.68 (d, J = 6.6 Hz,
1H), 7.49 – 7.15 (m, 10H), 6.05 (d, J = 17.0 Hz, 1H), 5.75 (d, J = 8.1 Hz,
1H), 5.14-5.02 (AB system, 2H), 4.77-4.65 (AB system, 2H), 4.18 – 3.87
(m, 3H), 3.71 (dd, J = 12.2, 4.2 Hz, 1H), 1.94 (s, 1H), 1.26 (d, J = 22.0 Hz,
3H). 13C NMR (100 MHz, CDCl3) δ 161.87, 150.60, 137.58, 136.82,
136.05, 128.42 (2C), 128.21 (2C), 128.08, 128.00 (2C), 127.91 (2C),
127.27, 101.95, 99.6 (d, J = 182.9 Hz), 90.52 (d, J = 28.0 Hz), 80.13, 76.59
(d, J = 16.3 Hz), 73.67, 59.42, 43.79, 17.00 (d, J = 25.5 Hz). HRMS (ESI):
calcd. for C24H25FN2NaO5 [M+Na]+ 463,1645; found 463,1649.
mixture was maintained at -55 °C and
a solution of phenyl
dichlorophosphate (10.8 g, 51.2 mmol) in methyl t-butyl ether (9.5 mL) was
added, keeping the temperature below -50 °C. The reaction mass was
stirred at -55 °C until the conversion was complete (about 2 hours). The
reaction mixture was heated to -20 °C and filtered under N2 atmosphere.
The filtrate was stored at -20 °C under N2 supposing a quantitative
formation of chloride 8. Inside a flask equipped with mechanical stirrer,
reflux condenser, thermometer and under N2, compound 15 (8.9 g, 25.4
mmol) was dissolved in dry THF (140 mL). This solution was cooled to -25
°C and i-propylmagnesium chloride lithium chloride complex 1:1 (1.3 M in
THF, 39 mL, 50.7 mmol), was slowly added making sure that the
temperature did not exceed -20 °C. When the addition was complete, the
reaction mixture was stirred at -20 °C for 2 h and then a solution of the
previously prepared phosphorylchloride 8 was slowly added over about 2
h, keeping the temperature below -20 °C. Once the conversion was
complete (approximately 2 h), the reaction mixture was poured in a mixture
of isopropyl acetate (170 mL), water (90 mL) and acetic acid (4 mL) cooled
to 0 °C. The mixture was slowly heated to room temperature (20-25 °C),
and the phases were separated. The organic phase was washed with
water and concentrated to residue under vacuum, removing the residual
solvents by co-evaporation with methanol, having care to maintain the
temperature of the hating bath below 40 °C. As amorphous residue was
obtained (15.7 gr, 99% conversion) that was analysed by HPLC; column:
Symmetryâ C18 (4.6 mm x 25 mm, 5 µm); eluent: H2O (+0.1%
H3PO4)/CH3CN 75:25; flow: 1 mL/min; detector: UV lamp, l 254 nm;
conversion: 98%, r.d.: 93:7. The sample was crystallized from isopropyl
acetate/methanol to give the analytical sample. 1H NMR (400 MHz, CDCl3)
δ 9.90 (s, 1H), 7.59 – 6.99 (m, 11H), 6.09 (d, J = 18.8 Hz, 1H), 5.63 (d, J
= 8.1 Hz, 1H), 4.94 (dt, J = 12.6, 6.3 Hz, 2H), 4.82 – 4.40 (m, 3H), 4.36 –
3.69 (m, 4H), 1.52–0.92 (m, 12H). HRMS (ESI) calcd. for C29H35FN3O9PNa
642.1993 [M+Na]+; found 642.1990.
1-((2R,3R,4R,5R)-4-(Benzyloxy)-3-fluoro-5-(hydroxymethyl)-3-
methyltetrahydrofuran-2-yl)pyrimidine-2,4(1H,3H)-dione
(15).
A
solution of 12 (1.85 g, 4.95 mmol prepared as described in the above
general procedure) in dry THF (20 mL) was cooled to -10 °C. NaH (593
mg of the 60% dispersion in mineral oil 14.8 mmol) was added over 10 min
in portions of approximately 50 mg each. After stirring the mixture at -10
°C for 30 min, benzyl bromide (0.588 mL, 4.95 mmol) was added dropwise.
When the addition was complete, the temperature was raised to r.t. and
kept under stirring for 16 hours. The reaction mixture, cooled to 0 °C, was
diluted with diethyl ether (40 mL) and water (20 mL) was slowly added.
The phases were separated, the aqueous phase extracted with Et2O and
the collected organic phases were dried over Na2SO4 and concentrated.
The residue was purified by flash column chromatography (petroleum
ether/EtOAc 8:2) to give O-Bn protected product (1.33 g, 2.87 mmol, 58%
yield) as a colourless oil that was solubilized in THF (15 mL) and treated
with TBAF (1.36 g, 4.30 mmol, 1.5 equiv) for 3 h at room temperature. The
solvent was evaporated and the product purified by column
chromatography (EtOAc) to give 15 (634 mg, 63%) as a white solid m.p.
180-182 °C. 1H NMR (400 MHz, CDCl3) δ = 11.45 (s, 1H), 7.92 (d, J = 8.0
Sofosbuvir (1). Product 16 (16.9 g, 25.4 mmol) was dissolved in methanol
(200 mL) and treated with H2 (1 atm) at r.t. using Pd/C as a catalyst (10%
w/w, 50% wet, 2.7 g, 1.3 mmol). After 12 h of stirring, the conversion was
complete, and the mixture was filtered on a Celite pad to remove the Pd
and the filtered cake washed with methanol. The solution was
concentrated to residue under vacuum and residual methanol was
removed by repeated co-evaporation with dichloromethane. 14.1 g of
residue are obtained (quantitative conversion). After crystallisation from
This article is protected by copyright. All rights reserved.

10.1002/ejoc.201800158
European Journal of Organic Chemistry
FULL PAPER
dichloromethane /acetonitrile, compound 1 (12.5 gr) was obtained in 89%
yield. HPLC analysis of the residue (carried out by means of a Symmetry®
3H), 0.05 (s, 3H). 13C NMR (100 MHz, CDCl3) δ 169.11, 137.72, 128.49
(2C), 128.09, 128.03 (2C), 99.75 (d, J = 179.3 Hz), 98.51 (d, J = 36.3 Hz),
79.16 (d, J = 15.6 Hz), 73.78, 62.33, 25.91 (3C), 21.11, 18.38, 16.83 (d, J
= 24.6 Hz), -5.33, -5.45. HRMS (ESI): calcd. for C21H33FNaO5Si [M+Na]+
435,1979; found 435,1984.
C18 column (4.6 mm
x 25 mm, 5 μιη) as eluent H2O (+0.1 %
H3PO3)/CH3CN 45:55 at 1 mL/min and UV 254 nm detector) shows a
quantitative conversion and a diastereomeric purity higher than 99%. m.p.
97-99 °C. 1H NMR (400 MHz, DMSO) δ 11.47 (s, 1H), 7.53 (d, J = 7.4 Hz,
1H), 7.36-7-32 (m, 2H), 7.28 – 7.09 (m, 2H), 6.22 – 5.92 (m, 2H), 5.81 (d,
J = 4.5 Hz, 1H), 5.51 (d, J = 7.9 Hz, 1H), 5.04 – 4.61 (m, 1H), 4.51 – 4.25
N-(1-((3R,4R,5R)-4-(Benzyloxy)-5-(((tert-butyldimethylsilyl)oxy)-
methyl)-3-fluoro-3-methyltetrahydrofuran-2-yl)-2-oxo-1,2-dihydro-
pyrimidin-4-yl)benzamide (25) and conversion to nucleoside 15: The
silylated cytosine base was prepared by refluxing a suspension of N-
benzoylcytosine (125 mg, 0.585 mmol) and ammonium sulfate (4 mg,
0.029 mmol, 0.05 equiv.) in hexamethyldisilazane (1.3 mL) for 2 h and
concentrating the resulting solution by vacuum distillation, followed by
drying the residue under vacuum (0.2 mmHg) for 2 h at r.t. The oily residue
was dissolved in chlorobenzene (2 mL). To this solution was added the
acetate 24 (160 mg, 0.39 mmol) and neat tin tetrachloride (199 µL, 1.72
mmol). After stirring under a nitrogen atmosphere for 2 h at r.t., the reaction
was heated to 70 °C for 19 h. The reaction mixture was cooled to 0 °C and
solid sodium bicarbonate (0.94 g) and EtOAc (50 mL) were added. To the
stirred solution was slowly added water (20 mL) (caution: vigorous
evolution of carbon dioxide). The mixture was vigorously stirred for 30 min
at r.t., the suspension was filtered and the collected solid was washed with
EtOAc (12 mL). The solution was concentrated, and the residue was
purified by flash column chromatography (petroleum ether/EtOAc 6:4) to
give 25 as colourless oil (169 mg, 0.26 mmol, 77%). Data for the major
isomer: 1H NMR (400 MHz, CDCl3) δ 8.41 (d, J = 7.2 Hz, 1H), 7.92 (d, J =
6.7 Hz, 2H), 7.61 – 6.99 (m, 10H), 6.36 (d, J = 17.6 Hz, 1H), 4.75 – 4.57
(m, 2H), 4.20 (dd, J = 18.2, 10.6 Hz, 2H), 3.88-3-99 (m, 2H), 1.34 (d, J =
20.0 Hz, 3H), 0.98 (s, 9H), 0.16 (s, 6H). The so obtained compound 25
was suspended in 80% aqueous AcOH (5 mL) and heated at reflux for 12
h. The clear solution was concentrated, and the residue was purified by
flash column chromatography (petroleum ether/ethyl acetate 6:4) to give
the nucleoside 15 (confirmed by H1NMR and HPLC) as a white solid (57
mg, 93%).
(m, 1H), 4.22-4-20 (m, 1H), 3.97-3.75 (m, 3H), 1.25 – 1.11 (m, 12H). 13
C
NMR (100 MHz, DMSO) δ 172.64, 162.76, 150.77, 150.47, 139.59, 129.68
(2C), 124.61, 120.11(2C), 102.29, 100.32 (d, J = 179.9 Hz), 88.42, 79.51,
71.57 (d, J = 16.8 Hz), 68.02, 64.77, 49.83, 21.42 (2C), 19.81 (d, J = 6.3
Hz), 16.56 (d, J = 25.1 Hz). 31P NMR (162 MHz, DMSO-d6) δ 4.82. Anal.
calcd for calcd. for C22H29FN3NaO9P: C 49.91, F 3.59, H 5.52, N 7.94, O
27.20, P 5.85; found C 49.96, H 5.50, N 7.97.
(3R,4R,5R)-4-(Benzyloxy)-5-(((tert-butyldimethylsilyl)oxy)methyl)-3-
fluoro-3-methyldihydrofuran-2(3H)-one (23): To a solution of 17 (900
mg, 5.49 mmol) in dry CH2Cl2 (10 mL) was added dry pyridine (1.02 mL,
12.6 mmol) under N2 and the reaction mixture was cooled to 0 °C.
Imidazole (403 mg, 6.31 mmol) and tert-butyldimethylsilyl chloride (910
mg, 6.04 mmol) dissolved into dry CH2Cl2 (3 mL) were added. The reaction
mixture was slowly warmed to r.t. and stirred for 16 h then a 1 M HCl
solution was added (10 mL), the phases separated, the aqueous phase
extracted with CH2Cl2 and the collected organic phases were washed with
brine, dried over Na2SO4 and concentrated. The residue was purified by
flash column chromatography (petroleum ether/EtOAc 7:3) to give the 3’
silylated product (1.34 g, 4.82 mmol, 88% yield) as a white solid. This
product was solubilized in THF (40 mL), cooled to –10 °C and NaH (578
mg of 60% dispersion in mineral oil, 14.46 mmol) was added in portions
over 10 min and the mixture was stirred at -10 °C for 30 min, then benzyl
bromide (573 µL, 4.82 mmol,) was added dropwise. When the addition was
complete, the mixture was slowly warmed to r.t. and kept under stirring for
16 h. To the reaction mixture, cooled to 0 °C, diethyl ether and water were
added, the phases separated, the aqueous phase extracted with diethyl
ether, the collected organic phases dried over Na2SO4 and concentrated.
The residue was purified by flash column chromatography (petroleum
ether/EtOAc 9:1) to give 23 (1.35 g, 76% yield) as a colourless oil. 1H NMR
(400 MHz, CDCl3) δ 7.43 – 7.29 (m, 5H), 4.77 (d, J = 11.7 Hz, 1H), 4.66
(d, J = 11.7 Hz, 1H), 4.42 (dt, J = 6.5, 2.3 Hz, 1H), 4.05 (dd, J = 8.0, 4.5
Hz, 1H), 3.95 (dd, J = 12.2, 2.4 Hz, 1H), 3.73 (dd, J = 12.2, 2.4 Hz, 1H),
1.54 (d, J = 23.2 Hz, 3H), 0.85 (s, 9H), 0.05 (d, J = 7.1 Hz, 6H). 13C NMR
(101 MHz, CDCl3) δ 170.91 (d, J = 22.1 Hz), 136.93, 128.65 (2C), 128.45,
128.19 (2C), 91.56 (d, J = 184.6 Hz), 81.71, 76.85 (d, J = 17.1 Hz), 73.74,
60.48, 25.79 (3C), 19.44 (d, J = 24.9 Hz), 18.28, -5.40, -5.52. HRMS (ESI):
calcd. for C19H29FNaO4Si [M + Na]+ 391,1717; found 391,1711.
Acknowledgements
Support from MIUR (Rome) through the grant Dipartimento di
Eccellenza 2018-2022 is gratefully acknowledged.
Keywords: antiviral agents • kinetic resolution • nucleotides • API
synthesis • protecting groups
(3R,4R,5R)-4-(Benzyloxy)-5-(((tert-butyldimethylsilyl)oxy)methyl)-3-
fluoro-3-methyltetrahydrofuran-2-yl acetate (24). Protected lactone 23
(220 mg, 0.60 mmol) was dissolved in dry THF (4 mL) and the solution
was cooled to -78 °C under a nitrogen atmosphere. DIBAL 1.0 M in THF
(1.20 mL, 1.20 mmol, 2 equiv) dropwise. After 3 h, the reaction was
complete at TLC analysis. Water (300 µL was added), temperature was
raised to room temperature and kept under stirring for 30 min. Dry Na2SO4
(1 g) was added and after 30 min of stirring at r.t., the mixture was filtered
on Celite, the filtrate was evaporated to give the crude lactol that was
solubilized in CH2Cl2 and treated with DMAP (7 mg, 0.06 mmol,) and acetic
anhydride (357 µL, 3.78 mmol) and the reaction was stirred at room
temperature for 16 h. The reaction was concentrated, and the residue was
purified by flash column chromatography (petroleum ether/EtOAc 8:2) to
give the acetate 24 as colorless oil (230 mg, 93%) 4:1 mixture of anomers
as determined by NMR in CDCl3. Data for the major isomer: 1H NMR (400
MHz, CDCl3) δ 7.46 – 7.13 (m, 5H), 6.01 (d, J = 11.2 Hz, 1H), 4.70 (m,
2H), 4.14 (dd, J = 7.7, 3.4 Hz, 1H), 3.99 (dd, J = 24.1, 8.1 Hz, 1H), 3.89 –
3.65 (m, 2H), 2.03 (s, 3H), 1.37 (d, J = 22.0 Hz, 3H), 0.90 (s, 9H), 0.06, (s,
[1]
D. R. Nelson, J. N. Cooper, J. P. Lalezari, E. Lawitz, P. J. Pockros,
N. Gitlin, B. F. Freilich, Z. H. Younes, W. Harlan, R. Ghalib, et al.,
Hepatology 2015, 61, 1127–1135.
[2]
[3]
[4]
[5]
[6]
[7]
Y. Mehellou, H. S. Rattan, J. Balzarini, J. Med. Chem. 2017,
acs.jmedchem.7b00734.
P. J. Thornton, H. Kadri, A. Miccoli, Y. Mehellou, J. Med. Chem. 2016,
59, 10400–10410.
U. Pradere, E. C. Garnier-Amblard, S. J. Coats, F. Amblard, R. F.
Schinazi, Chem. Rev. 2014, 114, 9154–9218.
B. S. Ross, P. Ganapati Reddy, H. R. Zhang, S. Rachakonda, M. J.
Sofia, J. Org. Chem. 2011, 76, 8311–8319.
B. Simmons, Z. Liu, A. Klapars, A. Bellomo, S. M. Silverman, Org.
Lett. 2017, 19, 2218–2221.
D. A. DiRocco, Y. Ji, E. C. Sherer, A. Klapars, M. Reibarkh, J.
This article is protected by copyright. All rights reserved.

10.1002/ejoc.201800158
European Journal of Organic Chemistry
FULL PAPER
Dropinski, R. Mathew, P. Maligres, A. M. Hyde, J. Limanto, et al.,
[14]
[15]
K. Teste, L. Colombeau, A. Hadj-Bouazza, R. Lucas, R. Zerrouki, P.
Krausz, Y. Champavier, Carbohydr. Res. 2008, 343, 1490–1495.
P. Wang, B. K. Chun, S. Rachakonda, J. Du, N. Khan, J. Shi, W. Stec,
D. Cleary, B. S. Ross, M. J. Sofia, J. Org. Chem. 2009, 74, 6819–
6824.
Science (80-. ). 2017, 356, 426–430.
[8]
E. Cini, L. Banfi, G. Barreca, L. Carcone, L. Malpezzi, F. Manetti, G.
Marras, M. Rasparini, R. Riva, S. Roseblade, et al., Org. Process
Res. Dev. 2016, 20, 270–283.
[9]
G. Barreca, M. Rasparini, A. Poggiali, L. Carcone, R. R. Tufaro, G.
Marras, M. Taddei, E. Cini, WO2016151542A1, 2016.
M. Lewandowska, P. Ruszkowski, K. Chojnacka, N. Kleczewska, M.
Hoffmann, K. Kacprzak, L. Celewicz, Bioorganic Med. Chem. 2016,
24, 2330–2341.
[16]
M. J. Sofia, D. Bao, W. Chang, J. Du, D. Nagarathnam, S.
Rachakonda, P. G. Reddy, B. S. Ross, P. Wang, H. R. Zhang, et al.,
J. Med. Chem. 2010, 53, 7202–7218.
[10]
[17]
[18]
C. A. Busacca, D. R. Fandrick, J. J. Song, C. H. Senanayake, Adv.
Synth. Catal. 2011, 353, 1825–1864.
[11]
[12]
[13]
J. H. Cho, F. Amblard, S. J. Coats, R. F. Schinazi, Tetrahedron 2011,
67, 5487–5493.
J. Steinreiber, K. Faber, H. Griengl, Chem. - A Eur. J. 2008, 14, 8060–
8072.
P. G. Reddy, B. K. Chun, H. R. Zhang, S. Rachakonda, B. S. Ross,
M. J. Sofia, J. Org. Chem. 2011, 76, 3782–3790.
R. Li-Yuan Bao, R. Zhao, L. Shi, Chem. Commun. 2015, 51, 6884–
6900.
This article is protected by copyright. All rights reserved.

10.1002/ejoc.201800158
European Journal of Organic Chemistry
FULL PAPER
Entry for the Table of Contents (Please choose one layout)
Layout 2:
FULL PAPER
*API synthesis
Elena Cini, Giuseppe Barreca, Luca
Carcone, Fabrizio Manetti, Marcello
Rasparini, and Maurizio Taddei*
Page No. – Page No.
A new method for the introduction of the P-stereogenic center via dynamic kinetic
resolution of the stereochemically labile phosphorylchloride, employs the nucleoside
protected at 3’ with a benzyl group as the chiral selector. Only this expedient is
required to obtain high stereoselectivity in an 8-step synthesis of the anti-HCV drug
Sofosbuvir.
Stereoselective
Sofosbuvir through
phosphorylation controlled by kinetic
resolution
synthesis
of
nucleoside
This article is protected by copyright. All rights reserved.