Chemoselective staudinger strategy in the practical, fit for purpose, gram-scale synthesis of an HCV RNA polymerase inhibitor

Article abstract of DOI:10.1055/s-0030-1259085

The use of a late-stage selective Staudinger reaction in the preparation of the novel HCV RNA polymerase inhibitor is described. Georg Thieme Verlag Stuttgart - New York.

Full text of DOI:10.1055/s-0030-1259085

LETTER
57
Chemoselective Staudinger Strategy in the Practical, Fit for Purpose,
Gram-Scale Synthesis of an HCV RNA Polymerase Inhibitor
C
hemoselective
o
S
taudinger
u
Strategy in th
i
e
Synt
s
hesis of an
-
H
CV R
C
N
A
P
olymerase
hInhibitor arles Campeau,* Paul D. O’Shea
Department of Process Research, Merck Frosst Canada Ltd., 16711 Trans Canada Highway, Kirkland, Québec H9H 3L1, Canada
E-mail: Louis-Charles_Campeau@merck.com
Received 20 October 2010
tions would therefore allow for late-stage introduction of
this amine and circumvention of problems often associat-
ed with a protecting group strategy.
Abstract: The use of a late-stage selective Staudinger reaction in
the preparation of the novel HCV RNA polymerase inhibitor is de-
scribed.
In the context of our discovery colleagues’ efforts in the
development of an NS5B inhibitor,4 we were faced with
the problem of developing a practical and efficient synthe-
sis of compound 1 which would allow for its preparation
on gram-scale. The synthesis posed several synthetic chal-
lenges resulting from the use of a deaza-adenine base and
a 4¢-azido 2¢-fluoroarabinose core. Herein we describe the
multigram synthesis of 1 via late-stage introduction of the
4-amino group via a mild SNAr reaction of NaN3 followed
by a chemoselective Staudinger reduction.5
Key words: Staudigner, nucleoside, 7-deaza-adenosine, azide,
HCV NS5B
7-Deaza-adenosines are a unique class of nucleosides
which have received considerable attention due to their in-
teresting biological activity as inhibitors of HCV RNA
polymerase (NS5B).1,2 Inhibitors of this enzyme block
HCV replication, making NS5B a crucial target in the de-
velopment of novel anti-HCV therapies.3 The main avail-
able building block to synthesize 7-deaza-adenosine are
based on 6-chloro-7-deazapurine which can be conve-
niently appended onto any carbohydrate derivative using
conventional anomeric center displacement chemistry.
Traditional methodology used to introduce the required 6-
amino functionality relies on SNAr reaction with ammonia
(Scheme 1). Because the reaction conditions required for
this step are harsh (>85 ºC using a large excess of ammo-
nia), this step is usually performed at an early stage of a
synthetic sequence in order to minimize side reactions and
starting material decomposition. It is therefore necessary
to protect the amino functionality during the remaining
steps of the sequence until it is to be unveiled, usually in
the final step. Methods relying on milder reaction condi-
Initially, 1 was prepared using the discovery route out-
lined in Scheme 2.4 The key features of this route were the
formation of vinyl ether 7 through a sequence of six trans-
formations from the fluoro-arabinose derivative 2. The 6-
amino functionality was installed via an SNAr reaction
with ammonia on the sensitive vinyl ether precursor. Un-
avoidable substrate decomposition and the necessity for
the use of a pivaloyl protecting group (which needed to be
removed in the final step of the synthesis) resulted in poor
overall yield (<0.1%). Installation of the key 4¢-azido
functionality was achieved using epoxidation with
DMDO and Lewis acid enabled ring opening with
TMSN3.6 Final deprotection and HPLC purification com-
pleted the end game.
O
Cl
R
N
NH
O
HO
N3
N
1) SNAr
deprotection
N
N
+ NH3
NH2
HO
2) protection
N
F
N
N
N
N
harsh conditions
early stage
1
traditional strategy
N3
Cl
O
HO
N3
N
N
N
SNAr
reduction
+ N3
NH2
HO
N
N
F
N
N
N
N
milder conditions
late stage
1
this work
Scheme 1 Strategies for the installation of the 6-amino functionality of 7-deaza-adenosine
SYNLETT 2011, No. 1, pp 0057–0060
x
x.
x
x.
2
0
1
1
Advanced online publication: 07.12.2010
DOI: 10.1055/s-0030-1259085; Art ID: S07610ST
© Georg Thieme Verlag Stuttgart · New York

58
L.-C. Campeau, P. D. O’Shea
LETTER
Cl
N
N
N
H
O
1) KOH, TDA-1
O
HO
O
BzO
BzO
MeCN, 50%
N
NH3
1) I2, Ph3P
N
Br
Cl
Cl
N
MeOH, 110 °C
16% over 3 steps
HO
2) NH3 (7 N)
MeOH, 70%
HO
2) DBU, MeCN
F
N
F
N
F
N
N
2
O
O
O
O
DMDO (3 equiv)
acetone, –30 °C
dr > 25:1
1) TBSCl, py
N
N
NHPv
N
TBSO
NH2
NHPv
HO
2) PvCl, DIPEA
60% over 2 steps
TBSO
F
N
F
N
F
N
N
N
7
TMSN3
SnCl4 (3 equiv)
O
1) TBAF, THF
2) NH3–MeOH, 40 °C
HO
O
HO
N3
N
N3
N
–78 °C, 11% (2 steps)
dr 1:1
NHPv
NH2
TBSO
3) HPLC separation
9% over 3 steps
HO
F
N
F
N
N
N
1
Scheme 2 Discovery route for the synthesis of 1
To realize our objective, several key areas for optimiza- competing saponification of the benzoyl protecting
tion were identified. First, we envisioned the replacement groups. With careful monitoring of the reaction, we were
of the early stage introduction of the 6-amino functional- able to achieve high conversion to the desired product
ity to the 7-deazapurine ring with a later stage alternative with minimal attendant hydrolysis. The crude product
using milder conditions. This would preclude the need for could then be deprotected using ammonia in methanol.
a protection–deprotection strategy for the 6-amino group. Activation of the primary alcohol with MsCl–Et3N result-
Second, DMDO oxidation would need to be replaced with ed in the formation of a significant amount of bismesylate.
a more scalable alternative. Finally, the introduction of the On the other hand, diol 3 could be selectively activated
azide at C4¢ led to 50% loss of this advanced intermediate with TsCl at the primary position. This simple alternative
due to the lack of selectivity observed for this transforma- allowed us to avoid the use of Ph3P/I2 which complicated
tion. A more diastereoselective approach would therefore purification of the activated species. We sought to use to-
be desirable.
sylate 4 directly in the elimination reaction but were sur-
prised to find it was quite stable and no conversion to the
vinyl ether was observed. Finkelstein displacement of the
tosylate group with NaI also proved difficult with only
We began our investigation with the SN2 reaction of 4-
chloro deazapurine with bromide 2 (Scheme 3). We found
that the low yield in the discovery route was a result of
4-chloro
deazapurine
KOH, TDA-1
O
O
O
BzO
BzO
HO
BzO
BzO
NH3 (40 equiv)
TsCl (1.6 equiv)
N
N
Br
MeCN, 1.5 h
82%
MeOH–PhMe
o/n, 90–95%
Cl
O
Cl
py, 1.5 h
HO
F
F
F
N
N
2
N
N
3
O
TBSOTf
2,6-lutidine
TsO
1) NaI (2.2 equiv)
3-pentanone, 90 °C
TsO
TBSO
N
N
Cl
HO
CH2Cl2, 0 °C to r.t.
87% (2 steps)
Cl
2) DBU (2.3 equiv)
MeCN, r.t.
90% over 2 steps
F
N
N
F
N
N
4
O
I2 (1.15 equiv)
N
O
I
1) Bz2O, DMAP, TBAF
THF, 70–99%
BnNEt3N3 (1.2 equiv)
O
ClBzO
Cl
N3
N
TBSO
N3
N
NMM (0.3 equiv)
THF, 70%
F
N
Cl
TBSO
N
2) MCPBA, MCBA
Bu4NHSO4, CH2Cl2–H2O
1 h, 30–45%
Cl
BzO
5
F
N
dr = 6:1
N
F
N
N
6
Scheme 3 Elaboration of 4¢-azido core
Synlett 2011, No. 1, 57–60 © Thieme Stuttgart · New York

LETTER
Chemoselective Staudinger Strategy in the Synthesis of an HCV RNA Polymerase Inhibitor
59
Table 1 Strategies for Introduction of 6-Amino Group
85% conversion achieved after several days at 40 °C in ac-
etone.7 Conversely, we investigated the option of install-
ing the TBS protecting group prior to the activation/
elimination sequence. The TBS group could be easily in-
stalled on tosylate 4 and Finkelstein displacement of this
intermediate proceeded cleanly without decomposition in
3-pentanone. The iodide could then be eliminated in two
hours to afford vinyl ether 5 in high yield.8 Overall this
key intermediate was prepared in seven steps and 63%
yield on 100-g scale (Scheme 3).
O
ClBzO
O
N3
ClBzO
N
N3
N
amine
Cl
BzO
source
R
BzO
F
N
N
F
N
N
Entry
Amine Source
NH3 in MeOH
NH3 in dioxane
BnNH2
R
Yield
1
2
3
4
NH2
–
NH2
–
Having established a robust and scalable route to vinyl
ether 5, we sought an alternative to the DMDO epoxida-
tion–ring-opening sequence used for the azide installation
given the risks associated with synthesizing and using
DMDO on scale,9 as well as the low diastereomeric ratio
(dr) observed for the epoxide opening.10 Inspired by
recent patent literature,11 we investigated the use of an
azido-iodination reaction of the electron-rich alkene.
Gratifyingly, this provided the expected iodoazide 6 in
70% yield with 6:1 dr in favor of the desired diastereoiso-
mer. Displacement of the iodide using MCPBA was at-
BnNH
NH2
<10%
<10%
LiHMDS, Pd
NH
5
6
NH2
N3
23%
83%
, Pd
Ph
Ph
NaN3
Bisazido 8 was treated under a variety of reduction condi-
tions (Scheme 4).14 We were gratified to see that
Staudinger reaction with PPh3 occurred at the desired
azide.15,16 The intermediate aminophosphorane was stable
and could be observed by HPLC. After complete forma-
tion of aminophosphorane was observed, the solvent was
then switched to MeCN and treatment with aq 1 M AcOH
at 60 °C delivered the free amine.
tempted on
6 but resulted in starting material
decomposition. A protecting group swap for a benzoyl al-
lowed for the reaction to proceed in modest yield.11 At this
point the diastereoisomers were separated via reverse-
phase chromatography.
The final challenge we faced was installation of the 6-
amino group (Table 1). As expected SNAr reactions of
amine nucleophiles resulted in significant decomposition
of the starting material (entries 1–3). Looking for a milder
alternative, Pd-catalyzed cross-coupling using the
Buchwald12 and Hartwig13 protocols for aniline synthesis
resulted in low yield or significant decomposition. The
benzoyl protecting groups were, for the most part, unsta-
ble to these reactions. However, we found that azide was
a good nucleophile in the SNAr reaction and that this reac-
tion could occur without decomposition affording the
bisazide in 83% yield. We envisioned that a selective
azide reduction would result in the desired amino group.
To simplify the purification process, we opted for the use
of the SCG-PPh3 reagent (Scheme 4) developed by the
Charette group.17 This allowed for easy purification of the
penultimate intermediate.18 Final global deprotection was
achieved using sodium methoxide affording 1 in high
yield.
Overall, the preparation of 1 was achieved in 12 steps in
9% yield allowing for gram-quantities to be prepared. The
present late-stage SNAr reaction of NaN3 followed by the
selective azide reduction offers a convenient procedure
for the installation of the 6-amino group in the presence of
the 4¢-azido group. The use of this SNAr–reduction strate-
gy constitutes a powerful alternative to traditional meth-
O
ClBzO
O
ClBzO
O
NaN3
ClBzO
SCG-PPh3, CH2Cl2
N3
N
N3
N
N3
N
Cl
DMF, 50 °C
then AcOH–MeCN
(60%, 2 steps)
N3
BzO
BzO
NH2
BzO
F
N
F
N
N
N
F
N
N
8
Ph
O
HO
+
NaOMe
Ph3P
P
N
N3
HO
Ph
NH2
–
MeOH
90%
ClO4
F
N
N
SCG-PPh3
1
Scheme 4 4-Amino group installation and end game
Synlett 2011, No. 1, 57–60 © Thieme Stuttgart · New York

60
L.-C. Campeau, P. D. O’Shea
LETTER
(7) At higher temperatures, aromatic Finkelstein displacement
of the Cl group was observed and elimination of the alkyl
iodide proved sluggish requiring 2 d to proceed to 100%
conversion. Increased heating resulted in decomposition of
the sensitive vinyl ether product.
ods for the installation of the 6-amino group in 7-deaza-
adenosine derivatives. To our knowledge, this is the first
example of a selective Staudinger reaction on a nucleoside
bearing multiple azido groups.
(8) The beneficial effect of the large TBS protecting group on
reactions at C4¢ might result from a preferred conformation
alignment. Further mechanistic investigation is ongoing and
will be reported in due course.
(9) For a discussion on the limitations of DMDO use on scale,
see: Ferrer, M.; Gibert, M.; Sánchez-Baeza, F.; Messeguer,
A. Tetrahedron Lett. 1996, 37, 3585.
(10) Epoxidation of 4¢,5¢-unsaturated nucleosides exhibits high
selectivity. The low dr in this case results from ring opening,
see: Haragushi, K.; Takeda, S.; Tanaka, H. Org. Lett. 2003,
5, 1399.
Acknowledgment
We thank Ravi Sharma for analytical assistance. Ernest Lee and
Austin Chen are thanked for helpful discussions and critical reading
of the manuscript. Maria Emilia Di Francesco and Marco Pompei
are thanked for providing an authentic sample of 1 and for helpful
discussions.
References and Notes
(11) Connolly, T. J.; Durkin, K.; Sarma, K.; Zhu, J. PCT Int.
Appl., WO 2005/000864, 2005.
(1) For a general review on HCV therapeutics, see: (a) Brown,
N. A. Expert Opin. Invest. Drugs 2009, 18, 709. (b) Tan,
S. L.; Pause, A.; Shi, Y.; Sonenberg, N. Nat. Rev. Drug
Discovery 2002, 1, 867.
(12) Wolfe, J. P.; Åhman, J.; Sadighi, J. P.; Singer, R. A.;
Buchwald, S. L. Tetrahedron Lett. 1997, 38, 6367.
(13) This protocol is known to be sensitive to ortho substitution:
(a) Lee, S.; Jørgensen, M.; Hartwig, J. F. Org. Lett. 2001, 3,
2729. (b) Huang, X.; Buchwald, S. L. Org. Lett. 2001, 3,
3417.
(14) Palladium-catalyzed hydrogenolysis and metal-mediated
reductions (Zn, Mg, Fe) gave unsatisfactory results.
Reduction with TMS2S resulted only in partial conversion.
(15) Gololobov, Y. G.; Kasukhin, L. F. Tetrahedron 1992, 48,
1353.
(16) The exquisite selectivity might arise from both steric and
electronic influences. For recent work dealing with relative
rates of azides in Staudinger reactions, see: Lin, F. L.; Hoyt,
H. M.; Van Halbeek, H.; Bergman, R. G.; Bertozzi, C. R.
J. Am. Chem. Soc. 2005, 127, 2686.
(2) (a) Smith, D. B.; Kalayanov, G.; Sund, C.; Winqvist, A.;
Pinho, P.; Maltseva, T.; Morisson, V.; Leveque, V.;
Rajyaguru, S.; Le Pogam, S.; Najera, I.; Benkestock, K.;
Zhou, X.-X.; Maag, H.; Cammack, N.; Martin, J. A.;
Swallow, S.; Johansson, N. G.; Klumpp, K.; Smith, M.
J. Med. Chem. 2009, 52, 219. (b) Klumpp, K.; Leveque, V.;
Le Pogam, S.; Ma, H.; Jiang, W.-R.; Kang, H.; Granycome,
C.; Singer, M.; Laxton, C.; Hang, J. Q.; Sarma, K.; Smith,
D. B.; Heindl, D.; Hobbs, C.; Merrett, J. H.; Symons, J.;
Cammack, N.; Martin, J. A.; Devos, R.; Najera, I. J. Biol.
Chem. 2006, 281, 3793. (c) McCown, M. F.; Rajyaguru, S.;
Le Pogam, S.; Ali, S.; Jiang, W.-R.; Kang, H.; Symons, J.;
Cammack, N.; Najera, I. Antimicrob. Agents Chem. 2008,
52, 1604.
(3) For a review on multiple classes of NS5B inhibitors, see:
Beaulieu, P. L. Expert Opin. Ther. Pat. 2009, 19, 145.
(4) Avolio, S.; Di Francesco, M. E.; Pompei, M.; Summa, V.
PCT Int. Appl., WO 2010/026153, 2010.
(5) (a) Li, J.; Chiang, F.-I.; Chen, H.-N.; Chang, C.-W. T.
J. Org. Chem. 2007, 72, 4055. (b) Li, J.; Chen, H.-N.;
Chang, H.; Wang, J.; Chang, C.-W. T. Org. Lett. 2005, 7,
3061. (c) Chapyshev, S. V. Synlett 2009, 1. (d) Nyffeler, P.
T.; Liang, C.-H.; Koeller, K. M.; Wong, C.-H. J. Am. Chem.
Soc. 2002, 124, 10773.
(17) Poupon, J.-C.; Boezio, A. A.; Charette, A. B. Angew. Chem.
Int. Ed. 2006, 45, 1415.
(18) Separation of the product from triphenylphosphine oxide
was very tedious via traditional silica gel chromatography.
With SCG-PPh3 a simple workup and precipitation
procedure was devised: The reaction was concentrated and
taken into CH2Cl2. This solution was then washed with sat.
aq NaHCO3 and brine. The crude CH2Cl2 solution was then
concentrated to a minimum amount of solvent required for
complete solubilization. To this stirring solution was then
added MTBE dropwise which resulted in precipitation of the
SCG-phosphine oxide. The filtrate was concentrated to
afford the reduced amine in >90% purity by HPLC.
(6) Perrone, P.; Daverio, F.; Valente, R.; Rajyaguru, S.; Martin,
J. A.; Leveque, V.; Le Pogam, S.; Najera, I.; Klumpp, K.;
Smith, D. B.; McGuigan, C. J. Med. Chem. 2007, 50, 5463.
Synlett 2011, No. 1, 57–60 © Thieme Stuttgart · New York

Copyright of Synlett is the property of Georg Thieme Verlag Stuttgart and its content may not
be copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for
individual use.