Evaluation of triazolamers as active site inhibitors of HIV-1 protease

Article abstract of DOI:10.1016/j.bmcl.2009.09.049

Proteases typically recognize their peptide substrates in extended conformations. General approaches for designing protease inhibitors often consist of peptidomimetics that feature this conformation. Herein we discuss a combination of computational and ex

Full text of DOI:10.1016/j.bmcl.2009.09.049

Bioorganic & Medicinal Chemistry Letters 19 (2009) 6023–6026
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Evaluation of triazolamers as active site inhibitors of HIV-1 protease
*
Andrea L. Jochim, Stephen E. Miller, Nicholas G. Angelo, Paramjit S. Arora
Department of Chemistry, New York University, 100 Washington Square East, New York, NY 10003, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Proteases typically recognize their peptide substrates in extended conformations. General approaches for
designing protease inhibitors often consist of peptidomimetics that feature this conformation. Herein we
discuss a combination of computational and experimental studies to evaluate the potential of triazole-
linked b-strand mimetics as inhibitors of HIV-1 protease activity.
Received 25 August 2009
Revised 11 September 2009
Accepted 14 September 2009
Available online 17 September 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
Peptidomimetics
Protease inhibitor
b-Strand mimetics
The conformational and chemical instabilities as well as poor
pharmacological properties of peptides often limit their potential
as reagents in molecular biology and drug discovery. Biomimetic
oligomers that display protein-like functionality and adopt defined
conformations have proven to be invaluable alternatives to pep-
tides.1–8 We previously reported a new class of nonpeptidic oligo-
mers, termed triazolamers, in which the peptide bond is replaced
with 1,2,3-triazole rings but the chiral main-chain and amino acid
side chains are preserved (Fig. 1a).9 We hypothesized that these
oligomers may access a restricted set of conformations due to
the dipole–dipole interaction between neighboring triazole rings.
We have described NMR studies which indicate that short triazola-
mers adopt zigzag conformations reminiscent of b-strands
(Fig. 1b).9 Optimized solution and solid phase syntheses of triazo-
lamers, which involve iterative diazotransfer and Huisgen 1,3-
dipolar cycloaddition steps, have also been reported.10
crystallized in complex with HIV-1, which makes them useful tem-
plates for the design of triazolamers as HIVPR inhibitors. Modeling
studies suggested that triazolamers 1–2 would provide ideal starting
points for our studies. Figure 2b–d shows that triazolamer 1 superim-
poses well onto the conformation of L-700,417 in its protein-bound
state. Two key differences between triazolamers and conventional
Proteases typicallyrecognizeandcleavepeptides inextended con-
formations and, in general, four to five residues provide a majority of
binding interactions.11 To establish the potential of triazolamers as
functional b-strand mimetics,4,5,8,12 we evaluated their ability to
target HIV-1 protease (HIVPR) as a model. We began our studies by
comparing triazolamers in zigzag configuration with conformations
adoptedby several knowninhibitors ofHIVPR. HIVPRisa C2-symmet-
rical homodimer and offers four hydrophobic pockets at the inter-
face.13 Typical inhibitors of HIVPR feature bulky residues to target
thesepockets. Wedesignedtriazolamersthatmimicactiveconforma-
tions adopted by prototypical HIVPR inhibitors L-700,41714 and
A-7470415 (Fig. 2a). Both of these tetrapeptide derivatives target the
protease with low nanomolar inhibition constants and have been
Figure 1. (a) Triazolamers feature a 1,2,3-trizole ring in place of the amide bond.
R = amino acid side chain. (b) The triazolamer mimics a peptide b-strand with
similar axial distances between the i and i + 2 side chains. For clarity, side chains are
depicted as methyl groups.
* Corresponding author. Tel.: +1 212 998 8470.
E-mail address: arora@nyu.edu (P.S. Arora).
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.09.049

6024
A. L. Jochim et al. / Bioorg. Med. Chem. Lett. 19 (2009) 6023–6026
Figure 2. (a) Chemical structures of known HIV-1 inhibitors and their triazolamer mimics. (b and c) Triazolamer 1 and structure of L400,417 in complex with HIV-1 protease
(PDB Code: 4phv), and d) overlay of L400,417 (green) and triazolamer 1 (magenta).
peptidomimetic HIVPR inhibitors are (i) the lack of a transition state
analog unit, typically a secondary alcohol group, in the backbone of
the triazolamers, and (ii) the hydrogen bonding (amide) groups. The
triazolamer backbone does not offera b-strand’s hydrogen bondfunc-
tionality, althoughthe N-2 andN-3 electronpairs may serve as hydro-
gen bond acceptors.13 Therefore, the ability of triazolamers to target
HIVPR would allow us to gauge the relative importance of the transi-
tion state analog units andhydrogen bondingnetworks versus hydro-
phobic contacts in a system that provides a preorganized scaffold for
binding.
We synthesized and tested twenty triazolamer analogs starting
with 1 and 2; the key results are shown in Table 1. Compounds
were synthesized following reported procedures.10 We utilized
previously described FRET-based assays to determine the efficacy
of these triazolamers as HIV-1 protease inhibitors.16 These experi-
ments allow us to develop and test hypotheses about the interac-
tion of these oligomers with biological receptors. The key results
from this pilot study reveal suggestive patterns regarding the spec-
ificity of triazolamers for the target enzyme. Our best compounds 3
ligands were treated using the united-atom approximation, and
prepared using the molecular modeling program InsightII.18 Only
polar hydrogens were added to the protein, and Kollman united-
atom partial charges were assigned using AutoDock Tools.19 Atom-
ic solvation parameters were assigned to the protein atoms using
the AddSol utility in AutoDock3. The grid maps were calculated
using AutoGrid 3.0. In all cases grid maps contained 61 Â 61 Â 61
points with spacing of 0.375 Å. The partial charges of the ligands
were calculated using the Gasteiger–Marsili method.20 Rotatable
bonds in the ligand were defined by the AutoTors utility. Ten doc-
kings with the Lamarkian Genetic Algorithm search method were
performed for triazolamer compounds 3, 6, and 7.
The HIV-1 dimer has a C-2 symmetry; according to our Auto-
Dock model, the R2 and R3 residues of the triazolamer compound
make contacts with the symmetrical hydrophobic pockets on the
individual units of HIV-1 protease (Fig. 3). The computational stud-
ies reveal that the benzyl group resides close to a hydrophobic
´
pocket composed of Leu23, Ile50, Val82, and Ile-84, corresponding
15
´
to the S1 and S1binding pockets of HIV-1 protease. The docking
and 8 inhibit the enzyme activity with Ki values of roughly 25
The worst compounds 9–10 inhibit with Ki values of over 500
l
l
M.
M.
results are consistent with previous studies that show bulky
hydrophobic groups of the ligand occupying the S1 or S1 pockets.
21
´
These preliminary results imply a correlation between the number
of triazole units in oligomers (9 vs 6) and inhibitory activity. Olig-
omers composed of four triazole linkages (and five side chain
groups) are the best leads in this small library. (Uncharged triazo-
lamers are hydrophobic, but incorporation of one amino group al-
lows us to prepare solutions of triazolamers in 10% DMSO in
aqueous buffers for the protease assays.) Comparison of 3 with 6
and 7 reveals that replacement of the benzyl group in 3 (as the side
chain of second N-terminal residue in these pseudo-symmetric
compounds) with isopropyl groups has a detrimental effect on
the inhibitory activity of the compounds. To explore the possible
reasons for this effect, we utilized AutoDock3 and modeled these
compounds in the enzyme cavity (Fig. 3).17
Table 2 lists the correlation between the AutoDock results and
experimental IC50 values for triazolamers 3, 6, and 7. We find that
the potency of triazolamers can be predicted by the distance be-
tween the R2 or R3 side chain and Ile-84 of the protease.
The docking studies do not predict interactions between the
nitrogens on the triazole rings and side chain or backbone atoms
in the protease. Given the importance of hydrogen bonding net-
works in protease targeting, we expect that addition of hydrogen
bond donor groups to the triazole rings will enhance the inhibitory
activity of these compounds. Studies that explore such modified
triazolamers are currently underway.
In conclusion, these studies suggest that nonpeptidic b-strand
mimetics, termed triazolamers, offer attractive starting points for
the rational design of protease inhibitors. Docking studies were
used to gain insights into binding mode of the basic scaffold and
The A-74704/HIV-1 structure15 (PDB code 9HVP) was used as a
starting point for the AutoDock calculations: the protein and

A. L. Jochim et al. / Bioorg. Med. Chem. Lett. 19 (2009) 6023–6026
6025
Table 1
Structures of selected triazolamers and their 50% inhibitory concentrations (IC50) against HIV-1 proteasea
a
b
Compound
IC50
Compound
IC50
(
lM)
(lM)
+
NH3
N N
N N
+H3N
N
N
N
N
N
N
78
7
129 15
N N
N N
N N
N N
1
5
N N
N N
N N
N N
N N
N
N
N
N
N
+H3N
N
N
N
N
101 12
60
7
N
N
N
N
N
N
N
N
6
2
N N
N N
N N
N N
+H3N
N
N
N
N
N
N
29
9
131 19
+H3N
N
N
N
N
N
N
N N
N N
7
3
H
N
H
N
N N
N N
N N
N N
+H3N
N
N
N
N
+H3N
72 18
25
7
N
N
N
N
N N
N N
O
N N
N N
O
+
NH3
8
4
+
+
NH3
CONHMe
NH3
CONHMe
N N
N N
>500
>500
N
N N
N
N
N N
N
N N
N
N
N N
9
10
a
Values obtained from a FRET-based assay. Please see the Supplementary data for details.
Values are means of three experiments.
b

6026
A. L. Jochim et al. / Bioorg. Med. Chem. Lett. 19 (2009) 6023–6026
derivatives. This analysis provides a persuasive explanation for the
differential activities of triazolamers and offers a platform for
continuing studies.
Acknowledgments
We thank Professor Celia Schiffer, UMass Medical School Center
for AIDS Research, for generous donation of the wild type HIV-1
protease. We are grateful to the NSF (CHE 0848410) for financial
support of this work, and also thank the NSF for equipment Grants
MRI-0116222 and CHE-0234863, and the NCRR/NIH for Research
Facilities Improvement Grant C06 RR-16572.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2009.09.049.
References and notes
1. Hill, D. J.; Mio, M. J.; Prince, R. B.; Hughes, T. S.; Moore, J. S. Chem. Rev. 2001, 101,
3893.
2. Horne, W. S.; Gellman, S. H. Acc. Chem. Res. 2008, 41, 1399.
3. Seebach, D.; Gardiner, J. Acc. Chem. Res. 2008, 41, 1366.
4. Smith, A. B.; Hirschmann, R.; Pasternak, A.; Akaishi, R.; Guzman, M. C.; Jones, D.
R.; Keenan, T. P.; Sprengeler, P. A.; Darke, P. L.; Emini, E. A.; Holloway, M. K.;
Schleif, W. A. J. Med. Chem. 1994, 37, 215.
5. Reichelt, A.; Martin, S. F. Acc. Chem. Res. 2006, 39, 433.
6. Stigers, K. D.; Soth, M. J.; Nowick, J. S. Curr. Opin. Chem. Biol. 1999, 3, 714.
7. Davis, J. M.; Tsou, L. K.; Hamilton, A. D. Chem. Soc. Rev. 2007, 36, 326.
8. Phillips, S. T.; Rezac, M.; Abel, U.; Kossenjans, M.; Bartlett, P. A. J. Am. Chem. Soc.
2002, 124, 58.
9. Angelo, N. G.; Arora, P. S. J. Am. Chem. Soc. 2005, 127, 17134.
10. Angelo, N. G.; Arora, P. S. J. Org. Chem. 2007, 72, 7963.
11. Tyndall, J. D.; Nall, T.; Fairlie, D. P. Chem. Rev. 2005, 105, 973.
12. Tsai, J. H.; Waldman, A. S.; Nowick, J. S. Bioorg. Med. Chem. 1999, 7, 29.
13. Brik, A.; Alexandratos, J.; Lin, Y. C.; Elder, J. H.; Olson, A. J.; Wlodawer, A.;
Goodsell, D. S.; Wong, C. H. ChemBioChem 2005, 6, 1167.
Figure 3. (a) Chemical structure of triazolamers 3, 6 and 7 with the residue
positions numbered, (b) full view of compound 3 docked into active site (PDB code
9HVP), (c) front view of the residues of HIV-1 protease composing the hydrophobic
cavity around R2 and R3 of compound 3, (d) side view of residues of HIV-1 protease
composing the hydrophobic cavity around R2 and R3 of compound 3, and (e) view of
the hydrophobic cavity.
14. Bone, R.; Vacca, J. P.; Anderson, P. S.; Holloway, M. K. J. Am. Chem. Soc. 1991,
113, 9382.
15. Erickson, J.; Neidhart, D. J.; VanDrie, J.; Kempf, D. J.; Wang, X. C.; Norbeck, D.
W.; Plattner, J. J.; Rittenhouse, J. W.; Turon, M.; Wideburg, N.; Kohlbrenner, W.
E.; Simmer, R.; Helfrich, R.; Paul, D. A.; Knigge, M. Science 1990, 249, 521.
16. Matayoshi, E. D.; Wang, G. T.; Krafft, G. A.; Erickson, J. Science 1990, 247, 954.
17. Morris, G. M.; Goodsell, D. S.; Halliday, R. S.; Huey, R.; Hart, W. E.; Belew, R. K.;
Olson, A. J. J. Comp. Chem. 1998, 19, 1639.
Table 2
Correlation of AutoDock calculations and experimental results
Triazolamer
Closest enzyme residue
Distancea (Å)
IC50 (lM)
18. Insight II, version 2000 ed.; Accelrys: San Diego, CA.
19. Sanner, M. F. J. Mol. Graphics Modell. 1999, 17, 57.
20. Gasteiger, J.; Marsili, M. Tetrahedron 1980, 36, 3219.
21. Nakatani, S.; Hidaka, K.; Ami, E.; Nakahara, K.; Sato, A.; Nguyen, J. T.; Hamada,
Y.; Hori, Y.; Ohnishi, N.; Nagai, A.; Kimura, T.; Hayashi, Y.; Kiso, Y. J. Med. Chem.
2008, 51, 2992.
7
6
3
Ile-84
Ile-84
Ile-84
5.50
3.49
3.19
131
60
29
a
Distance between side chains of residues R2 and R3 on the triazolamers and Ile-
84 of the enzyme.