Disubstituted bis-THF moieties as new P2 ligands in nonpeptidal HIV-1 protease inhibitors

Article abstract of DOI:10.1021/ml2000356

A series of darunavir analogues featuring a substituted bis-THF ring as P2 ligand have been synthesized and evaluated. High affinity protease inhibitors (PIs) with an interesting activity on wild-type HIV and a panel of multi-PI resistant HIV-1 mutants co

Full text of DOI:10.1021/ml2000356

LETTER
pubs.acs.org/acsmedchemlett
Disubstituted Bis-THF Moieties as New P2 Ligands in Nonpeptidal
HIV-1 Protease Inhibitors
Konrad Hohlfeld,† Cyrille Tomassi,† J€org Kurt Wegner,‡ Bart Kesteleyn,‡ and Bruno Linclau*,†
†School of Chemistry, University of Southampton, Highfield, Southampton SO17 1BJ, United Kingdom
‡Tibotec, Turnhoutseweg 30, 2340 Beerse, Belgium
S Supporting Information
b
ABSTRACT: A series of darunavir analogues featuring a substituted bis-THF ring as
P2 ligand have been synthesized and evaluated. High affinity protease inhibitors
(PIs) with an interesting activity on wild-type HIV and a panel of multi-PI resistant
HIV-1 mutants containing clinically observed, primary mutations were identified
using a cell-based assay. A number of PIs have been synthesized that show equivalent
and greater activity for HIV-1 mutant strains as compared to wild-type HIV-1. The
activity on the purified enzyme was confirmed for a selection of analogues.
KEYWORDS: HIV protease, inhibitors, darunavir, bis-THF bis-diol
he advent of protease inhibitors (PIs) into highly active
antiretroviral therapy (HAART) regimens in 1996 marks a
Ghosh reported the successful incorporation of an oxatricyclic
tris-THF ligand into the PI structure (2a, Figure 1), in which the
third THF moiety enables the formation of a water-mediated
hydrogen bond with several conserved residues on the dimer
T
major innovation in AIDS chemotherapy.1 Inhibition of the
HIV-1 protease by PIs leads to immature, noninfectious viral
particles and suppresses HIV-1 replication in the patient.2
However, high replication rates, lack of HIV reverse transcriptase
proofreading capacity, and therapeutic pressure produce mutant
strains that are resistant to one or more PIs.3 Other limitations
arise from the peptidelike character of all first generation PIs. A
major breakthrough in subsequent structure-based design was
the development of a stereochemically defined fused bis-tetra-
hydrofuran (bis-THF) moiety as a nonpeptidal high-affinity P2
ligand, leading to the selection of darunavir (1, DRV, Figure 1) as
a conceptually new HIV-1 PI.4,5 DRV has proven to be highly
effective against a broad spectrum of multi-PI resistant HIV
mutant strains.6 The X-ray crystal structure of DRV-bound
HIV-1 protease revealed an extensive network of ligandꢀprotein
hydrogen-bonding interactions involving the enzyme backbone.
Both oxygen atoms of the bis-THF moiety form strong hydrogen
bonds with the backbone NH groups of Asp 29 and Asp 30.7
Because the structure of the bis-THF ligand is very significant for
the superior resistance profile of DRV, further research mainly
focused on DRV analogues with alterations around the sulfon-
interface as well as direct and water-mediated CH O interac-
3 3 3
tions with Gly48.11 Recently, Ghosh et al. also reported the
activity of the C4-alkoxylated bis-THF derivatives 2b,c (as well
as their corresponding diastereoisomer, not shown) on HIV-
protease (HIV-1LA1).12 This prompts us to report our results
on the synthesis and biological evaluation of novel DRV analo-
gues with C4-exo substituted bis-THF ligands (3aꢀj13 and 5aꢀi,
Figure 1).
While the C4-substituted DRV analogues could in principle
be obtained via direct functionalization of the corresponding
PI containing an OH group at that position (not shown), a
conventional convergent strategy was chosen in which the PI
analogues were obtained by coupling of a functionalized bis-THF
derivative 6 with a number of amines 7 (Scheme 1). A few cases
(analogues a) required functionalization postcoupling to arrive
at the desired PI (see below). At this stage, the DRV para-
aminophenyl-sulfonamide 7a5 and the (2-methylamino)-ben-
zoxazole-sulfonamide 7b8 were selected as P20 ligands based
on their pronounced hydrogen-bonding abilities in the S20
subsite.
The enantioselective synthesis of the substituted bis-THF
fragments 6bꢀj is shown in Scheme 2 and relies on the
regioselective protection of a bis-THF diol 8, as described
earlier.13,14 From 10, functionalization with benzyl bromide or
with a dimethyl sulfoxide (DMSO)/acetic anhydride/acetic acid
mixture15 led to fragments 12b13 and 12j. The synthesis of the
phenoxy-containing ligand 12c was achieved via a three-step
sequence starting with nucleophilic aromatic substitution with
8ꢀ10
ꢀ
amide P2 substituent and the P1 phenyl group.
However,
our studies of the X-ray crystal structure of HIV-1 protease in
complex with DRV and several recognition sequences (see the
Supporting Information) led to the conclusion that substituents
on the bis-THF B-ring could lead to additional interactions of the
PI with the enzyme backbone. In particular, hydroxy, alkoxy, or
alkoxyalkyl groups on the C4-position of the bis-THF B ring
could form hydrogen-bonding interactions with the Gly48 amide
group, with or without a water molecule. In addition, phenyl or
benzyl groups were predicted to be able to fill a hydrophobic
pocket in the proximity of the bis-THF moiety.
Received: February 7, 2011
Accepted: March 27, 2011
Published: March 31, 2011
r
2011 American Chemical Society
461
dx.doi.org/10.1021/ml2000356 ACS Med. Chem. Lett. 2011, 2, 461–465
|

ACS Medicinal Chemistry Letters
LETTER
Scheme 3. Synthesis of Inhibitors 1 and 2a
a Reagents and conditions: (a) Et3N, CH2Cl2. (b) Compound 7a or 7b,
Et3N, CH2Cl2. (c) TBAF, THF (only for analogues b).
desilylation, which was avoided by using the endo-PMB ether 11.
Fluorobenzylation and alkylation afforded the bis-THF fragments
12dꢀi. Deprotection of 12bꢀj to bis-THF alcohols 6bꢀj was
achieved via TBAF-mediated desilylation, DDQ oxidation, or Pd-
catalyzed hydrogenolysis. Silyl ether 9 also served as the precursor
for the synthesis of PI analogues 3a and 5a (see below).
Finally, the synthesized ligands 6bꢀj and silyl ether 9 were
converted into the corresponding mixed carbonates 14 using dis-
uccinimidyl carbonate (13, DSC)16 and triethylamine (Scheme 3).
Treatment of known amines 7a5 or 7b8 with the activated carbonates
14 in the presence of triethylamine afforded the PIs 3 and 5. For the
synthesis of PIs 3/5a additional treatment with TBAF was required to
remove the silyl protection on the exo-hydroxy group.
Figure 1. Structure of DRV (1) and PIs 2ꢀ5.
Scheme 1. Retrosynthetic Analysis of Inhibitors 3 and 5
All compounds were tested for their antiviral activity against
wild-type HIV-1 and a panel of recombinant clinical isolates,
which were selected because of their high level of cross-resistance
against several approved PIs.3 Every isolate has 6ꢀ8 amino acid
substitutions in the protease gene that are associated with
resistance to PIs, including the primary mutations Met46 f Ile
(M46I), G48V, I50V, L76V, V82A, I84V, and L90M (see the
footnote of Table 1). Isolates M1 and M2 are of particular
interest, as they were identified with primary mutations (I50V,
I84V, and L76V) that are associated with resistance to DRV.3
Atazanavir (entry 1), lopinavir (entry 2), DRV (1, entry 3), and
its previously described benzoxazole analogue 48 (entry 10) serve
as reference compounds.
Scheme 2. Synthesis of Building Blocks 6bꢀja
All three approved PIs (entries 1ꢀ3) inhibit HIV-1IIIB at low
nanomolar concentrations (EC50 = 6.6ꢀ13.9 nM) but exhibit a
great variation in their resistance profile. Atazanavir and lopinavir
(entries 1 and 2) showed a decreased activity against the majority
of mutants. Interestingly, with atazanavir, an increased inhibition
of mutant M2 was observed. DRV (1, entry 3) is equally potent
against isolate M1. However, while mutant M3 was found to be
hypersusceptible to DRV, a loss in activity was noted with isolate
M2, which bears two primary mutations known to be associated
with decreased susceptibility to DRV (I50V and L76V). Hyper-
susceptibility is defined as a lower fold-change on a mutant
virus when being compared to the wild-type virus (FC < 1) and
has only been reported for a few cases.10,17 The benzoxazole
analogue 4 (entry 8) is slightly more potent against HIV-1IIIB
and shows a better resistance profile against the mutant panel
(EC50 = 0.9ꢀ3.2 nM). All newly synthesized compounds, apart
from the free hydroxy analogues 3/5a, inhibit wild-type HIV-1 at
low nanomolar or even subnanomolar concentrations and show
good cytotoxicity profiles (CC50 g 9.8ꢀ32.3 μM), producing an
acceptable selectivity index (CC50/EC50) of over 2000. The
activity on HIV-1IIIB was also measured in the presence of 50%
human serum (HIV-1IIIB þ 50% HS) as an indication for the
tendency of the compounds to bind to plasma proteins. The data
show that there is no extensive plasma protein binding. PIs 3a
and 5a (entries 4 and 9) were found to be only weak inhibitors of
a Reagents and conditions: (a) NaH, DMF/THF; R-X, TBAI.
(b) DMSO, Ac2O, AcOH. (c) (i) FꢀC6H4ꢀNO2, KHMDS, THF;
(ii) cyclohexene, Pd(OH)2, EtOH; (iii) tBuONO, BF3 OEt2, CH2Cl2;
3
Cu powder, EtOH. (d) NaHMDS, THF; PMBCl, TBAI. (e) TBAF,
THF. (f) DDQ, CH2Cl2/H2O. (g) H2, Pd/C, EtOH.
4-fluoronitrobenzene followed by removal of the nitrogroup. How-
ever, the NaH-mediated benzylation suffered from competitive
462
dx.doi.org/10.1021/ml2000356 |ACS Med. Chem. Lett. 2011, 2, 461–465

ACS Medicinal Chemistry Letters
LETTER
Table 1. Antiviral Activity (EC50) of Compounds 3 and 5 against Wild-Type (HIV-1IIIB) and Mutant Virusesa
a The following PI resistance-associated mutations were identified in the isolates. M1: L10I, M46I, I64V, I84V, L90M, I93L. M2: L10I, I13V, M46I,
I50V, L63P, L76V. M3: L10I, K20R, M36I, G48V, I62V, A71V, V82A, I93L. Bold figures represent primary resistance mutations, and underlined figures
are associated with resistance to DRV according to the IAS (ref 3). b The EC50 values were determined using MT4 cells, and all assays were conducted in
quadruplicate. The numbers in parentheses represent the fold change in EC50 value for each isolate relative to the wild-type HIV-1IIIB
.
HIV-1IIIB but showed a good resistance profile toward the
mutant panel. Indeed, all three isolates were found to be
hypersusceptible to both compounds. For the other C4-substi-
tuted PIs, some general trends could be established. More
lipophilic compounds (entries 5, 6, and 10ꢀ14) exhibit good
activity against wild-type virus and clinical isolates M1/M3, but a
loss in activity (FC = 2.7ꢀ28) was observed with mutant M2.
The benzylated PI 3b (entry 5) has also been tested previously
against a different mutant panel, whereas similar resistance
patterns were observed.13 Although being slightly less active
against HIV-1IIIB when compared to DRV or the benzylated
analogues, a more balanced activity profile was observed for the
methoxyethoxy-substituted PIs 3/5h (entries 7 and 15). The
clinical isolates M1 and M3 were found to be hypersusceptible to
inhibitors 3/5h, and only a 1ꢀ5-fold increase of EC50 was noted
against isolate M2. A further analysis of the relationship between
chain length and antiviral activity showed that a methyl linker
(compounds 5i and 5j, entries 16 and 17) is slightly favored over
the ethyl linker (compound 5h) on HIV-1IIIB. However, the
methyl linker exhibits more variation in activity on the mutants,
463
dx.doi.org/10.1021/ml2000356 |ACS Med. Chem. Lett. 2011, 2, 461–465

ACS Medicinal Chemistry Letters
LETTER
Figure 2. Three-dimensional depiction of possible binding modes of 3a (A) and 3h (B).
and the hypersusceptibility is lost. A direct interpretation of the
hypersusceptibility remains challenging, since the observed effect
might have multiple reasons. The most likely structural reason
might be a changed protease or folding dynamics of the protease
and substrates due to interference of the inhibitors with the flap
region (containing Gly48 and Gly480).
Unfortunately a direct comparison of these results with the
data recently published by Ghosh11,12 is difficult, as PIs having a
para-methoxy sulfonamide-based P20 ligand (as in 2a) and para-
amino sulfonamide- or benzoxazole-based P20 ligands (as in 3
and 5) have different interactions in the S20 pocket of the
protease resulting in a different resistance profile. In addition,
the panel of mutants investigated by Ghosh for PI 2a was focused
on mutations mainly associated with resistance against other PIs,
not including DRV as in our case. Although the tris-THF PI 2a is
approximately 10-fold more active against wild-type HIV-1 than
DRV, activity was reduced against all of the selected multidrug-
resistant mutants. A similar reduction of activity of DRV against
their mutant panel was observed.
The enzymatic activity against wild-type HIV-1 protease was
determined for the promising analogues 3h (IC50 = 2.5 nM) and
5h (IC50 = 1.5 nM) as well as the hydroxy analogue 3a (IC50 = 2.9
nM) and benzyl derivative 3b (IC50 = 3.5 nM).18 All four PIs are
highly active on the purified enzyme and show similar potency as
DRV (1, IC50 = 3.8 nM). In fact, the enzymatic activity of PI 3a is
over 250-fold higher than in the cellular assay. This leads to the
conclusion that additional ligandꢀprotein interactions are present
with PIs 3/5a. However, the higher polar surface area and the
increased number of hydrogen bond donors in the hydroxy
analogues seem to limit the uptake into the cell, which explains
the weak potency of 3/5a against HIV-1IIIB in the cellular assay.
In Figure 2, a 3D depiction of X-ray crystal structure-based
modeling studies involving3a and 3h is shown. On the basisof the
2HS119 coordinates, a R group extension in combination with a
conformational search and subsequent energy minimizations was
performed. In addition to the bonding distances, a percentage
score is given, indicating the likeliness of a hydrogen bond based
on statistical analysis as described by Clark and Labute.20 The
hydroxy group of PI 3a appears to form a hydrogen bond with the
backbone carbonyl of Gly48. This interaction is also observed in
natural substrates of the PR (see the Supporting Information),
which could contribute to the hypersusceptibility characteristic
and competition against natural substrates of 3a. Furthermore,
the hydroxy group interacts with a water network, which itself is
hydrogen bonding to amide NH of residue Gly48 and then
extending to water-mediated interactions to the side chains of
Asp29 and Asp30 (see Figure 2A). This analysis corresponds with
the recently published crystal structure of the 2bꢀHIV-1 protease
complex,11 in which the methoxy oxygen atom forms a water-
mediated hydrogen bond with the amide NH of Gly48.
For 3/5h, the methoxyethoxy moiety possibly displaces the water
molecule and the second oxygen could form a direct hydrogen bond
to the backbone NH of Gly48 (see Figure 2B). This interaction can
also be found in some natural substrates, adding to the hypersus-
ceptibility characteristic of those compounds. These direct back-
bone interactions, which are less influenced by residue mutations
(side chain might change completely), furthermore explain the
hypersusceptibility of the compounds 3/5h and 3a. The hydrogen-
bonding distance is in both cases 1.7 Å but has in the case of 3/5h a
better orientation, which results in an 80% optimal hydrogen bond
as calculated by Clark/Labute, while the contact in 3a is only 39%
optimal. A similar modeling study with fluorobenzyl analogue 5e
suggests weak CꢀH π interactions between Gly48 and the
3 3 3
aromatic ring (see the Supporting Information).
In conclusion, we designed and synthesized a series of
disubstituted bis-THF containing HIV-1 PIs, which show
464
dx.doi.org/10.1021/ml2000356 |ACS Med. Chem. Lett. 2011, 2, 461–465

ACS Medicinal Chemistry Letters
LETTER
interesting in vitro antiviral activity against wild-type virus and
three multi-PI resistant virus strains. Many analogues display low
nanomolar activity, and for some analogues, the hypersuscept-
ibility of the mutants in a cellular context is noteworthy. On the
basis of our modeling studies, it could be shown that PIs 3/5h
successfully extend the “backbone binding” concept,21 which is
crucial in combating multidrug resistance. Structural studies on
the characterization of potential new ligandꢀenzyme interac-
tions as well as further structureꢀactivity relationship studies on
disubstituted bis-THF ligands are currently underway.
(9) Miller, J. F.; Andrews, C. W.; Brieger, M.; Furfine, E. S.; Hale,
M. R.; Hanlon, M. H.; Hazen, R. J.; Kaldor, I.; McLean, E. W.; Reynolds,
D.; Sammond, D. M.; Spaltenstein, A.; Tung, R.; Turner, E. M.; Xu,
R. X.; Sherrill, R. G. Ultra-potent P1 modified arylsulfonamide HIV
protease inhibitors: The discovery of GW0385. Bioorg. Med. Chem. Lett.
2006, 16, 1788–1794.
(10) Cihlar, T.; He, G.-X.; Liu, X.; Chen, J. M.; Hatada, M.;
Swaminathan, S.; McDermott, M. J.; Yang, Z.-Y.; Mulato, A. S.; Chen,
X.; Leavitt, S. A.; Stray, K. M.; Lee, W. A. Suppression of HIV-1 Protease
Inhibitor Resistance by Phosphonate-mediated Solvent Anchoring.
J. Mol. Biol. 2006, 363, 635–647.
(11) Ghosh, A. K.; Xu, C.-X.; Rao, K. V.; Baldridge, A.; Agniswamy,
J.; Wang, Y.-F.; Weber, I. T.; Aoki, M.; Miguel, S. G. P.; Amano, M.;
Mitsuya, H. Probing Multidrug-Resistance and ProteinꢀLigand Inter-
actions with Oxatricyclic Designed Ligands in HIV-1 Protease Inhibi-
tors. ChemMedChem 2010, 5, 1850–1854.
K.H. and C.T. thank Tibotec for funding. We thank Tibotec
for antiviral activity determination.
’ ASSOCIATED CONTENT
(12) Ghosh, A. K.; Martyr, C. D.; Steffey, M.; Wang, Y.-F.;
Agniswamy, J.; Amano, M.; Weber, I. T.; Mitsuya, H. Design, Synthesis,
and X-ray Structure of Substituted Bis-tetrahydrofuran (Bis-THF)-
Derived Potent HIV-1 Protease Inhibitors. ACS Med. Chem. Lett.,
published online January 27, 2011; DOI: 10.1021/ml100289m.
(13) Linclau, B. Protease Inhibitor Precursor Synthesis. WO2006/
0899422 (compound 3b).
(14) Linclau, B.; Jeffery, M. J.; Josse, S.; Tomassi, C. Enantioselective
synthesis and selective monofunctionalization of (4R,6R)-4,6-Dihy-
droxy-2,8-dioxabicyclo[3.3.0]octane. Org. Lett. 2006, 8, 5821–5824.
(15) Pojer, P. M.; Angyal, S. J. Methylthiomethyl ethers: General
synthesis and mild cleavage. Protection of hydroxyl groups. Tetrahedron
Lett. 1976, 17, 3067–3068.
S
Supporting Information. Description of inhibition assay
b
and experimental procedures and spectral data for compounds 3,
5, 6, 9, 11, and 12. This material is available free of charge via the
Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*Tel: þ44(0)23 8059 3816. Fax: þ44(0)23 8059 3781. E-mail:
bruno.linclau@soton.ac.uk.
(16) Takeda, K.; Akagi, Y.; Saiki, A.; Tsukahara, T.; Ogura, H.
Studies on activating methods of functional groups. Part X. Convenient
methods for syntheses of active carbamates, ureas and nitrosoureas
using N,N0-disuccinimidyl carbonate (DSC). Tetrahedron Lett. 1983,
24, 4569–4572.
(17) De Meyer, S.; Descamps, D.; Van Baelen, B.; Lathouwers, E.;
Cheret, A.; Marcelin, A.-G.; Calvez, V.; Picchio, G. Confirmation of the
negative impact of protease mutations I47V, I54M, T74P and I84V and
the positive impact of protease mutation V82A on virological response
to darunavir/ritonavir. Poster presented at XVIIIth IHDRW, Florida,
2009; for the abstract, see Antiviral Ther. 2009, 14 (Suppl. 1), A147.
(18) The substrate for the enzymatic assay reflects only the MA-CA
cleavage site.
(19) Kovalevsky, A. Y.; Liu, F.; Leshchenko, S.; Ghosh, A. K.; Louis,
J. M.; Harrison, R. W.; Weber, I. T. Ultra-high resolution crystal
structure of HIV-1 protease mutant reveals two binding sites for clinical
inhibitor TMC114. J. Mol. Biol. 2006, 363, 161–173.
(20) Clark, A. M.; Labute, P. 2D depiction of protein-ligand com-
plexes. J. Chem. Inf. Model. 2007, 47, 1933–1944.
’ REFERENCES
(1) Sepkowitz, K. A. AIDS—The First 20 Years. N. Engl. J. Med.
2001, 344, 1764–1772.
(2) Kohl, N. E.; Emini, E. A.; Schleif, W. A.; Davis, L. J.; Heimbach,
J. C.; Dixon, R. A. F.; Scolnick, E. M.; Sigal, I. S. Active human
immunodeficiency virus protease is required for viral infectivity. Proc.
Natl. Acad. Sci. U.S.A. 1988, 85, 4686–4690.
(3) Johnson, V. A.; Brun-Vezinet, F.; Clotet, B.; Gunthard, H. F.;
Kuritzkes, D. R.; Pillay, D.; Schapiro, J. M.; Richman, D. D. Update of the
drug resistance mutations in HIV-1: December 2009. Top. HIV Med.
2009, 17, 138–145.
(4) Ghosh, A. K.; Thompson, W. J.; Fitzgerald, P. M.; Culberson,
J. C.; Axel, M. G.; McKee, S. P.; Huff, J. R.; Anderson, P. S. Structure-
based design of HIV-1 protease inhibitors: Replacement of two amides
an a 10π-aromatic system by a fused bis-tetrahydrofuran. J. Med. Chem.
1994, 37, 2506–2508.
(5) Surleraux, D. L. N. G.; Tahri, A.; Verschueren, W. G.; Pille,
G. M. E.; de Kock, H. A.; Jonckers, T. H. M.; Peeters, A.; De Meyer, S.;
Azijn, H.; Pauwels, R.; de Bethune, M.-P.; King, N. M.; Prabu-Jeyabalan,
M.; Schiffer, C. A.; Wigerinck, P. B. T. P. Discovery and selection of
TMC114, a next generation HIV-1 protease inhibitor. J. Med. Chem.
2005, 48, 1813–1822.
(21) Ghosh, A. K.; Chapsal, B. D.; Weber, I. T.; Mitsuya, H. Design
of HIV protease inhibitors targeting protein backbone: An effective
strategy for combating drug resistance. Acc. Chem. Res. 2008, 41, 78–86.
(6) De Meyer, S.; Azijn, H.; Surleraux, D.; Jochmans, D.; Tahri, A.;
Pauwels, R.; Wigerinck, P.; de Bethune, M.-P. TMC114, a novel human
immunodeficiency virus type 1 protease inhibitor active against protease
inhibitor-resistant viruses, including a broad range of clinical isolates.
Antimicrob. Agents Chemother. 2005, 49, 2314–2321.
(7) Tie, Y.; Boross, P. I.; Wang, Y.-F.; Gaddis, L.; Hussain, A. K.;
Leshchenko, S.; Ghosh, A. K.; Louis, J. M.; Harrison, R. W.; Weber, I. T.
High resolution crystal structures of HIV-1 protease with a potent
non-peptide inhibitor (UIC-94017) active against multi-drug-resistant
clinical strains. J. Mol. Biol. 2004, 338, 341–352.
(8) Surleraux, D. L. N. G.; de Kock, H. A.; Verschueren, W. G.; Pille,
G. M. E.; Maes, L. J. R.; Peeters, A.; Vendeville, S.; De Meyer, S.; Azijn,
H.; Pauwels, R.; de Bethune, M.-P.; King, N. M.; Prabu-Jeyabalan, M.;
Schiffer, C. A.; Wigerinck, P. B. T. P. Design of HIV-1 protease
inhibitors active on multidrug-resistant virus. J. Med. Chem. 2005, 48,
1965–1973.
465
dx.doi.org/10.1021/ml2000356 |ACS Med. Chem. Lett. 2011, 2, 461–465