Synthesis of new 4,5-dihydrofuranoindoles and their evaluation as HCV NS5B polymerase inhibitors

Article abstract of DOI:10.1021/ol203177g

The synthesis of substituted 3,4-dihydrofuranoindoles is reported. These new indole compounds were used to synthesize potent HCV NS5B inhibitors. The binding mode of the dihydrofuranoindole-derived inhibitors was established via X-ray crystallographic stu

Full text of DOI:10.1021/ol203177g

ORGANIC
LETTERS
2012
Vol. 14, No. 2
556–559
Synthesis of New 4,
5-Dihydrofuranoindoles and Their
Evaluation as HCV NS5B Polymerase
Inhibitors
‡
ꢀ
ꢀ
Francisco Velazquez,* Srikanth Venkatraman, Charles A. Lesburg, Jose Duca,
Stuart B. Rosenblum, Joseph A. Kozlowski, and F. George Njoroge
Merck Research Laboratories, 2015 Galloping Hill Road, Kenilworth,
New Jersey 07033-1300, United States
francisco.velazquez@merck.com
Received November 28, 2011
ABSTRACT
The synthesis of substituted 3,4-dihydrofuranoindoles is reported. These new indole compounds were used to synthesize potent HCV NS5B
inhibitors. The binding mode of the dihydrofuranoindole-derived inhibitors was established via X-ray crystallographic studies.
The indole ring system is one of the most important
heterocyclic structures known in medicinal chemistry.
Nature has been a continuous supplier of molecules
containing indole moieties embedded in their structures.1
More recently, numerous indole-containing molecules can
trace their origins to the synthetic laboratory. A great
number of indole-containing molecules have shown to
possess a wide variety of biological activities such as anti-
infectives, neurotransmitters, anticancer, antiinflammatory,
and others.2 Therefore, indoles are preferred molecular
pharmacophores for identification of new drug candidates.3
As part of the efforts to identify molecules that possess
antiviral activity, our research group investigated indole-based
scaffolds based on structural information obtained from lead
compounds bound to the active site of the hepatitis C (HCV)
NS5B polymerase enzyme.4 Based on information gathered
from the indole-based lead compounds, we proposed to in-
vestigate the binding mode of new indole scaffolds containing
a dihydrofuran ring fused at the 4- and 5-positions of the
indole core. Although the 4,5-dihydrofuranoindole system
is known in the literature,5 it was reported as a side product
in the preparation of regioisomeric 6,7-dihydrofuranoin-
dole. Furthermore, there was no precedent for compounds
having substitutions attached to the dihydrofuran ring
of the proposed 4,5-dihydrofuranoindole cores. Our
efforts were therefore focused on the design and execu-
tion of synthetic routes to access new dihydrofuranoin-
dole cores in order to gather important structural
information to advance our research program. In this
‡ Present address: Novartis Institutes for BioMedical Research, 100 Tech-
nology Square, Cambridge, MA 02139, United States.
(1) Kochanowska-Karamyan, A. J.; Hamann, M. T. Chem. Rev.
2010, 110, 4489–4497.
(2) (a) Gul, W.; Hamann, M. T. Life Sci. 2005, 78, 442–453. (b)
Ahmad, A.; Sakr, W. A.; Wahidur Rahman, K. M. Curr. Drug Targets
2010, 11, 652–666.
(3) Rodrigues de Sa Alves, F.; Barreiro, E. J.; Manssour Fraga, C. A.
Mini-Rev. Med. Chem 2009, 9, 782–793.
(4) Lesburg, C. A.; Cable, M. B.; Ferrari, E.; Hong, Z.; Mannarino,
A. F.; Weber, P. C. Nat. Struct. Biol. 1999, 6, 937–943.
(5) Roffey, J. R.; Davidson, J. E.; Mansell, H. L.; Hamlyn, R. J.;
Adams, D. R. PCT Int. Appl. WO 2001012602, 2001.
r
10.1021/ol203177g
Published on Web 01/05/2012
2012 American Chemical Society

letter, we report our efforts in the synthesis of new
dihydrofuran indole systems and their application as
molecular probes.
Scheme 2. Synthesis of Dihydrobenzofuran-7-carboxaldehyde 8
Scheme 1. New 4,5-Fused Dihydrofurano Indole Systems and
Proposed Retrosynthetic Analysis for Their Synthesis
The synthesis of previously unknown aldehyde 14
was accomplished using an approach similar to that
employed by Spoors and co-workers for preparation
of 8. We extended the utility of this approach to form
the required stereogenic center present in 14 during the
cyclization step.
The 4,5-fused dihydrofuranoindole based compounds
1ꢀ3 (Scheme 1) were deemed important probes that
would provide valuable structural information about the
manner in which these molecules bind to the active site
of the HCV NS5B polymerase enzyme. We postulated a
synthetic approach based on the HemetsbergerꢀKnittel
indole synthesis for their preparation.6 A general retro-
synthetic analysis for indoles 1ꢀ3 is outlined in Scheme 1.
We proposed to assemble the desired dihydrofuran indole
systems via CꢀH insertion of a nitrene formed from a
vinyl azido intermediate such as 5. The required vinyl
azido intermediate 5 could be obtained from condensa-
tion of dihydrobenzofuran carboxaldehyde 6 and ethyl
azido acetate. Aldehyde intermediates of type 6 were also
unknown in the literature and synthetic routes for their
preparation had to be devised as well.
Scheme 3. Synthesis of (S)-3-Methyldihydrobenzofuran-7-car-
boxaldehyde 14
A. Synthesis of Dihydrobenzofuran Carboxaldehydes
and Assembly of Indole Cores. The synthesis of dihydro-
furanoindole 1 commenced with preparation of aldehyde
8 (Scheme 2).7 A one-pot, two-step approach previously
described by Spoors using 2,6-dibromophenol (7) as
starting material was reported to deliver 8 in 76% yield.
The first step of this one-pot transformation is based
on a modified Parham cyclization.8 An alternative two-
step approach was used in our laboratory to obtain 8
in slightly higher yield (85% overall). Commercially
available dihydrobenzofuran-7-carboxylic acid (9)
was converted into Weinreb amide 10 followed by
lithium aluminum hydride reduction to deliver 8 in
high yield.
Thus, synthesis of 14 began with preparation of the
optically active alcohol 11 from 2,6-dibromophenol (7) as
outlined in Scheme 3. Having an efficient method to
obtain 11 was very important because the Parham cycli-
zation would depend on the absolute configuration of
the stereogenic center in 11 to install the 30-methyl with
complete stereocontrol during the intramolecular SN2
process. Different approaches for synthesis of 11 were
considered including the asymmetric reduction of a
ketone precursor or the enantioselective addition of a
methyl nucleophile to an aldehyde. Ultimately, a simpler
methodology was devised using the regioselective open-
ing of (R)-2-methyloxirane to obtain alcohol 11 in ex-
cellent yield with the required R-configuration at the
stereogenic center. Conversion of 11 into bromide 12
delivered the substrate needed for the critical cyclization
step.
(6) (a) Hemetsberger, H.; Knittel, D.; Weidmann, H. Monatsh.
Chem. 1970, 101, 161–165. (b) Knittel, D.; Hemetsberger, H.; Leipert,
R.; Weidmann, H. Tetrahedron Lett. 1970, 17, 1459–1462.
(7) Plotkin, M.; Chen, S.; Spoors, P. G. Tetrahedron Lett. 2000, 41,
2269–2273.
(8) Bradsher, C. K.; Reames, D. C. J. Org. Chem. 1981, 46, 1384–
1388.
Org. Lett., Vol. 14, No. 2, 2012
557

Treatment of 12 with 1 equiv of n-BuLi at ꢀ78 °C
triggered a 5-exo-tet cyclization which occurred with
inversion of configuration at the stereogenic center to
assemble the (S)-3-methyl dihydrobenzofuran core. Ad-
dition of a second equivalent of n-BuLi followed by DMF
delivered the desired aldehyde 14. This one-pot procedure
allowed preparation this valuable intermediate in large
quantities.
described in the literature (method A).9 Although this
methodology delivered the desired vinylazido com-
pounds 23 ꢀ 25 (as methyl esters), it suffered from
low yields and extended reaction times. In order to
solve this problem, a two-step approach for this syn-
thetic transformation was investigated (method B).
First, aldehydes 8 and 17 were reacted with ethyl azido
acetate in the presence of DBU and lithium chloride
to obtain the azidohydroxy intermediates 20 and 22
in good yields (65ꢀ86%). The azidohydroxy inter-
mediates were then converted to their corresponding
mesylate derivatives followed by an in situ β-elimina-
tion to give the vinylazido intermediates 23 and 25 in
excellent yields (Scheme 6).
Scheme 4. Initial Attempts for the Synthesis of 3,3-Dimethyl-
dihydrobenzofuran-7-carboxaldehyde 17
Scheme 6. HemetsbergerꢀKnittel Synthesis of 30-Substituted
Dihydrofuranoindoles
Different methods were investigated for the synthesis of
previously unknown aldehyde 17. The first attempt in-
volved the free radical cyclization of 16 which was pre-
pared in quantitative yield from 2-hydroxy-3-bromo-
benzaldehyde (15) (Scheme 4). As expected, the reaction
gave the desired cyclization product 17 via a 5-exo-trig
cyclization process but the yield was disappointingly poor
(10%). We then investigated an alternative approach
using palladium to promote the cyclization. Once again,
the desired product 17 was obtained in suboptimal
yield (23%).
The final assembly of the dihydrofurano indoles 1ꢀ3
was expected to occur from vinylazido intermediates
23ꢀ25 under HemetsbergerꢀKnittel conditions. Recent
work has been reported in the development of new
rhodium based catalysts to promote the CꢀH insertion
at lower temperatures.10 In our research, we found that
these catalysts would partially convert the vinylazido
intermediates to the required dihydrofurano indoles. On
the other hand, the thermally promoted process gave the
desired dihydrofuranoindoles 1, 2, and 3 in moderate
yields (42, 20, and 37%, respectively).
Scheme 5. Synthesis of 3,3-Dimethyldihydrobenzofuran-7-car-
boxaldehyde 17
At this point, we were concerned about the possibility
that the aldehyde moiety of 16 was unfavorably interfer-
ing in the reaction. Therefore, we investigated the cycliza-
tion of 18 (prepared from 2-bromophenol) with the
assumption that the aldehyde moiety could be introduced
at a later stage (Scheme 5). Thus, palladium-catalyzed
cyclization of 18 gave the desired 3,3-dimethyldihydro-
benzofuran (19) in 60% yield. Deprotonation of 19 with
sec-BuLi in the presence of TMEDA followed by addition
of DMF gave aldehyde 17 in quantitative yield.
Having on hand the required aldehydes 8, 14, and 17,
we proceeded to convert them into the required 30-sub-
stituted dihydrofuranoindoles. The initial attempts in-
volved the condensation of the aldehydes with ethyl azido
acetate in the presence of sodium methoxide as previously
(9) For selected examples, see: (a) Chezal, J.-M.; Paeshuyse, J.;
Gaumet, V.; Canitrot, D.; Maisonial, A.; Lartigue, C.; Gueiffier, A.;
Moreau, E.; Teulade, J.-C.; Chavignon, O.; Neyts, J. Eur. J. Med. Chem.
2010, 45, 2044–2047. (b) Henn, L.; Hickey, D. M. V.; Moody, C. J.; Rees,
C. W. J. Chem. Soc., Perkin Trans. 1 1984, 2189–2196. (c) Roy, P.; Boisvert,
M.; Leblanc, Y. Org. Synth. 2007, 84, 262–271.
(10) Stokes, B. J.; Dong, H.; Leslie, B. E.; Pumphrey, A. L.; Driver,
T. G. J. Am. Chem. Soc. 2007, 129, 7500–7501.
(11) See the Supporting Information for experimental details. Also
see the following references for information about structureꢀactivity of
functional groups in these compounds: Anilkumar, G.; Lesburg, C.;
Selyutin, O.; Rosenblum, S.; Zeng, Q.; Jiang, Y.; Chan, T.-Y.; Pu, H.;
Vaccaro, H.; Wang, L.; Bennett, F.; Chen, K.; Duca, J.; Gavalas, S.;
Huang, Y.; Pinto, P.; Sannigrahi, M.; Velazquez, F.; Venkatraman,
S.; Vibulbhan, B.; Agrawal, S.; Butkiewicz, N.; Feld, B.; Ferrari, E.;
He, Z.; Jiang, C.; Palermo, R.; Mcmonagle, P.; Huang, H.-C.; Shih,
N.-Y.; Njoroge, G.; Kozlowski, J. Bioorg. Med. Chem. Lett. 2011, 21,
5336–5341.
558
Org. Lett., Vol. 14, No. 2, 2012

B. Synthesis of HCV NS5B Inhibitors and Macromole-
cular X-ray Diffraction Studies. To exemplify the utility of
the new dihydrofuranoindoles synthesized in this study,
appropriate functional groups were attached to com-
pounds 1 and 2 in order to convert them into effective
HCV NS5B inhibitors.11 Thus, acylsulfonamides 26 and
27 were obtained from 1 and 2, respectively, and had their
inhibitory activity measured,12 Figure 1. It was deter-
mined that the 30-methyl-substituted compound 27 (IC50
=
6 nM) increased the potency by 2-fold compared to the
unsubstituted analogue 26 (IC50 = 15 nM).
Figure 2. X-ray structures of dihydrofuranoindoles 26 (green)
and 27 (yellow) bound to HCV NS5B. Hydrogen bonds to the
backbone atoms of Ile-447 and Tyr-448 are shown as dots. Also
labeled are residues Pro-197 and Met-414, which line the sub-
pocket into which the 30-methyl group of 27 protrudes. The
interior surface of the HCV NS5B protein is shown in light gray.
This figure was prepared using PyMOL.
The 30-methyl-substituted dihydrofuranoindole 27 (yellow,
IC50 = 6 nM) showed the same interactions as compound
26. However, the 30-methyl group projects more deeply
into the hydrophobic cavity. The 2-fold increase in potency
was attributed to the additional van der Waals interac-
tions made by the 30-methyl group.
In conclusion, a series of new dihydrofuranoindoles were
synthesized as core structures for HCV NS5B polymerase
inhibitors. Synthesis of the dihydrofuranoindoles required
preparation of previously unknown 3-substituted dihydro-
benzofuran carboxaldehydes and their conversion to dihydro-
furano indoles was reported. The 30-substituted dihydro-
furanoindoles were converted to potent HCV NS5B inhibitors
and used as molecular probes to obtain new information
about the binding mode of our initial lead compounds in
order to design inhibitors with improved potency.
Figure 1. Enzymatic activity of dihydrofuranoindole-based
HCV NS5B inhibitors 26 and 27.
The primary goal for the preparation of these dihydro-
furanoindoles was their utilization as molecular probes to
investigate the mode of binding in the HCV NS5B poly-
merase active site. To this end, compounds 26 and 27 were
soaked into preformed crystals of HCV NS5B protein.13
The resulting structures showed how the unsubstituted
inhibitor 26 (Figure 2, green color, IC50 = 15 nM) binds
to the enzyme. Important interactions include H-bonding
of the pyridone moiety with the protein backbone of Tyr-
448 and Ile-447. The dihydrofuran ring of 26 rests in a
hydrophobic pocket deep within the active site cavity
lined in part by the side chains of Pro-197 and Met-414.
Supporting Information Available. Experimental pro-
cedures, spectroscopical data, and copies of spectra for
selected compounds. This material is available free of
charge via the Internet at http://pubs.acs.org.
(12) Ferrari, E.; He, Z.; Palermo, R. E.; Huang, H.-C. J. Biol. Chem.
2008, 283, 33893–33901.
(13) Macromolecular crystallographic data have been deposited at
the Protein Data Bank (http://www.pdb.org) with accession codes 3uph
and 3upi.
Org. Lett., Vol. 14, No. 2, 2012
559