Synthesis, antiviral activity and pharmacokinetics of P1/P1′ substituted 3-aminoindazole cyclic urea HIV protease inhibitors

Article abstract of DOI:10.1016/S0960-894X(02)01064-8

A series of P1/P1′ substituted cyclic urea analogues were prepared in an attempt to increase the intra-cellular antiviral potency of the nonsymmetrical 3-aminoindazoles DMP 850 and DMP 851. The effect of alkyl substitution of the P1/P1′ residues on cellul

Full text of DOI:10.1016/S0960-894X(02)01064-8

Bioorganic & Medicinal Chemistry Letters 13 (2003) 605–608
Synthesis, Antiviral Activity and Pharmacokinetics of
P1/P10 Substituted 3-Aminoindazole Cyclic Urea HIV
Protease Inhibitors
Robert F. Kaltenbach, III,a,* Mona Patel,a Robert E. Waltermire,b Gregory D. Harris,b
Benjamin R. P. Stone,b Ronald M. Klabe,a Sena Garber,a Lee T. Bacheler,a
Beverly C. Cordova,a Kelly Logue,a Matthew R. Wright,a Susan Erickson-Viitanena
and George L. Trainora
aBristol-Myers Squibb Company, Experimental Station, Wilmington, DE 19880, USA
bProcess R&D Department, Bristol-Myers Squibb Company, Process Research Facility, Deepwater, NJ 08023, USA
Received 28 October 2002; accepted 2 December 2002
Abstract—A series of P1/P10 substituted cyclic urea analogues were prepared in an attempt to increase the intra-cellular antiviral
potency of the nonsymmetrical 3-aminoindazoles DMP 850 and DMP 851. The effect of alkyl substitution of the P1/P10 residues on
cellular antiviral potency, protein binding, resistance profile and pharmacokinetics are described.
# 2003 Elsevier Science Ltd. All rights reserved.
There has been an intense effort to find new therapeutics
that inhibit the human immunodeficiency virus protease
(HIV-Pr). In clinical trials, inhibition of this enzyme was
shown to reduce the levels of infectious virus and
increase CD4 cell counts in patients infected with HIV.1
Recently, we described the nonsymmetrical 3-amino-
indazole cyclic ureas DMP 850 and DMP 851.2 These
clinical candidates are potent inhibitors of HIV-Pr and
showed a good pharmacokinetic profile in dogs.
view of the excellent enzymatic potency, we felt an
increase in cellular potency could be obtained by modi-
fications designed to improve cell penetration. We have
previously shown that, in general, the symmetrical bis-3-
aminoindazole substituted cyclic ureas are too polar and
have limited cellular activity. Increasing the lipophilicity
of the 3-aminoindazole P2/P20 residues3 or the benzyl
P1/P10 residues4 by substituting small alkyl substituents
resulted in improved antiviral activity. However, with
either approach there was a limit on the size of the alkyl
groups tolerated. Larger branched substituents sig-
nificantly decreased binding due to unfavorable steric
interactions and limited further improvements in whole
cell antiviral activity. We predicted that increasing the
lipophilicity of the benzyl P1/P10 residues in the DMP
850 and DMP 851 series would increase the cellular
potency. In addition, since DMP 850 and DMP 851 had a
Clog P5 of 5.2 an increase of only one or two additional
carbons on each P1 residue should be sufficient.
DMP 850 and DMP 851 are highly potent inhibitors of
the HIV-Pr enzyme with a Ki of 0.031 and 0.021 nM,
and show modest translation in a whole cell antiviral
assay with an IC90 of 62 and 56 nM, respectively. In
Chemistry
Previous syntheses of the P1/P10 substituted cyclic ureas
4a–e required the stereoselective reduction of
a
substituted oxime derived from l-tartaric acid.6
Although this route was useful for our preliminary
*Corresponding author. Tel.: +1-609-252-4000; e-mail: robert.
kaltenbach@bms.com
0960-894X/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0960-894X(02)01064-8

606
R. F. Kaltenbach, III et al. / Bioorg. Med. Chem. Lett. 13 (2003) 605–608
Scheme 1. Reagents and conditions: (a) CH3C6H4R1, s-BuLi, THF, ꢀ15–0 ꢁC, 60–99%; (b) RaNi, MeOH, H2 (250 psi), 100 ꢁC, 82–93%; (c) 1,10-
CꢁDI, 1,1,2,2-tetrachloroethane, reflux, 46–60%; (d) BnBr (5a–e) or BuI (9a–e), KOt-Bu, THF, 0 ꢁC, 62–76%; (e) 3-CN-4-F-BnBr, KOt-Bu, THF,
.
0 C, 72–96%; (f) H2NNH2 H2O, n-butanol, reflux, 80–99%; (g) HCl, MeOH/H2O, 81–96%.
screening, the cyclic ureas were generally obtained in
moderate yields and a more efficient synthesis was
desired for scaleup for in vivo studies. One option was
an alternate tartrate derived route developed by our
chemical process research and development group. In
this methodology, the P1/P10 residues were introduced
via a stereoselective addition of benzyllithium to the
dihydrazone 1, followed by reduction to the diamine
and cyclization.7 By modifying this synthetic approach
by incorporating appropriately substituted benzyl-
lithiums the P1/P10 substituted cyclic ureas could be
obtained in improved yields as shown in Scheme 1.
DMP 850 and DMP 851 series, mono-substitution of
the P1/P10 residues was tolerated giving only a slight
decrease in enzyme binding, and a 2–5-fold improve-
ment in cellular antiviral activity (IC90s). The p-methyl
analogues 8b and 12b improved the IC90 over 5-fold to
12 nM and the m-methyl analogues 8c and 12c
improved the IC90s about 2-fold. Similar improvements
were seen with the bulkier p-ethyl analogues, however,
losses in binding were apparent which limited the
increase in cellular potency. The 3,5-dimethyl sub-
stituted analogues significantly decreased binding and
showed no improvement in IC90 compared to the
unsubstituted parents.
For example, by metalating p-xylene with s-BuLi (1.3 M
in cyclohexane) in THF at ꢀ15 ꢁC the resulting
p-methyl-benzyllithium was formed. Addition of the
dihydrazone 1 in THF followed by warming to 0 ꢁC
gave the bis-hydrazine in 95% yield. Reduction with
Raney nickel and hydrogen (250 psi) in methanol gave
the diamine 3b in 83% yield. Similarly, the correspond-
ing m-methyl, p-ethyl and 3,5-dimethyl substituted
diamines 3c–e could be obtained from m-xylene,
p-ethyltoluene or mesitylene. Cyclization with 1,10-carbo-
nyldiimidazole in refluxing tetrachloroethane gave the
substituted cyclic ureas 4a–e. Alkylation with 1.1
equivalents of benzyl bromide or butyl iodide gave the
mono-alkylated cyclic ureas 5a–e and 9a–e. The 3-ami-
noindazole was introduced by alkylation with 3-cyano-
4-fluoro-benzylbromide followed by treatment with
hydrazine hydrate in refluxing n-butanol. Finally,
deprotection of the acetonide provided the non-
symmetrical 3-aminoindazoles 8a–e and 12a–e.
In addition to antiviral potency against wild type viru-
ses, the effect of binding to human plasma proteins and
activity against drug resistant viral mutants were
important factors in our selection criteria. The relative
degree of plasma protein binding was determined by a
functional assay where the IC90s were determined in the
absence or presence of physiological levels of two
human plasma proteins (45 mg/mL human serum albu-
min and 1 mg/mL alpha-1-acid glycoprotein). The
results are reported as the fold increase in IC90 (fold
shift). The mono-substituted analogues all showed an
increase in plasma protein binding over DMP 850 and
DMP 851. This, in effect, lowers the amount of free
drug available in plasma. By calculating the protein
binding adjusted IC90s (IC90ꢂfold shift) it was shown
that the substituted analogues were still more potent
than DMP 850 and DMP 851 but the increased protein
binding partially offsets the increases in antiviral
potency.
Adequate antiviral potency to inhibit common drug
resistant strains of HIV is an additional factor in the
selection of our protease inhibitors. Despite the efficacy
of the current protease containing antiretroviral regi-
mens, some failures occur due to the development of
Results and Discussion
The P1/P10 substituted cyclic ureas were tested in our
enzyme binding affinity8 and whole cell antiviral9
assays. The results are shown in Table 1. In both the

R. F. Kaltenbach, III et al. / Bioorg. Med. Chem. Lett. 13 (2003) 605–608
Table 1. Binding, protein shift and cellular activity of P1/P10 substituted cyclic ureas
607
Compd
R1
R2
Ki
(nM)a
RF IC90
(nM) b
HSA+AAG
(fold shift)c
PB adj.
IC90 (nM)d
Virus IC90 (nM)
V82F/I84Ve
HXB2b
I84Ve
Abbottf
Vertexg
DMP850
H
Bn
Bn
Bn
Bn
Bn
0.031
0.047
0.062
0.072
0.20
62
12
25
19
50
8ꢂ
16ꢂ
12ꢂ
13ꢂ
7ꢂ
496
192
300
247
350
92
17
26
81
370
310
104
410
2100
>1700
2680
1520
1880
1760
12,600
>7300
257
24
8b
8c
8d
8e
4-Me
3-Me
4-Et
>1600
3,5-Me
100
DMP851
12b
12c
12d
12e
H
n-Bu
n-Bu
n-Bu
n-Bu
n-Bu
0.021
0.037
0.045
0.094
0.19
56
12
31
10
66
17ꢂ
35ꢂ
25ꢂ
22ꢂ
952
420
775
220
69
16
54
22
56
390
260
1790
1750
1600
>1800
1470
1700
1600
4030
7540
87
7
4-Me
3-Me
4-Et
3,5-Me
aValues were measured by cleavage of a fluorescent peptide substrate using HPLC.
bDetermined by measuring the accumulation of viral RNA transcript after infection of cells with HIV-1.
cFold shift of the RNA IC90 with added HSA (45 mg/mL) and AAG (1 mg/mL).
dProtein binding adjusted IC90 (IC90ꢂ fold shift).
eValues were calculated based on quantitation of viral yield by plaque assay in MT-2 cells.10
fHIV mutant virus 25-fold resistant to ritonavir (M46I/L63P/A71V/V82F/I84V).
gHIV mutant virus 19-fold resistant to amprenavir (M46I/I47V/I50V).
In order to examine the effects of P1/P10 substitution on
bioavailability we dosed the para-methyl 8b, 12b and
meta-methyl 8c analogues orally in dogs. The pharma-
cokinetic parameters are shown in Table 2. Compared
to DMP 850 and DMP 851 the substituted analogues
showed a significantly lower Cmax and lower bioavail-
ability. In addition, the 8 h trough blood levels were
much lower and not sufficient to cover the protein
binding adjusted IC90s for the I84V mutant. Based on
the overall pharmacokinetics, it was determined
that these substituted analogues had insufficient oral
bioavailability for further development.
Table 2. Pharmacokinetics of selected compounds in dogs
Compd
R1
R2
Cmax (mM)a C8 h (mM) T1/2 (h) F%
DMP850
8b
8c
H
4-Me
3-Me
Bn
Bn
Bn
9.6
3.6
1.0
2.8
0.35
0.24
3.5
3.9
3.8b
60
29
20
DMP851
12b
H n-Bu
4-Me n-Bu
11.7
1.2
7.2
0.07
7.7
1.6
57
23
aOral bioavailability in dogs dosed at 10 mg/kg.
bCalculated from iv dosing at 5 mg/kg.
viral resistance. Our P1/P10 analogues were tested
against a panel of HIV mutant strains created by site
directed mutagenesis including viruses with single, dou-
ble, triple or quintuple amino acid changes in the pro-
tease gene.11 Although similar improvements in potency
were seen with a second wild type viral isolate HXB2,
the substituted analogues did not show significantly
increased potency against mutations that enlarge the
S1/S10 subsite (I84V, V82F/I84V).12 For example, the
p-methyl substituted derivative 8b is 5-fold more potent
against the wild type viruses RF and HXB2 but is only
slightly better than DMP 850 against the I84V single
mutant. Other viral mutants that contain the common
I84V mutation show a similar trend. The meta-methyl
and para-methyl analogues are approximately equipo-
tent to DMP 850 and DMP 851 against the V82F/I84V
double and the ritonavir resistant virus. Interestingly,
viruses without mutations in the S1/S10 pockets did not
show increased resistance to the P1/P10 variants. The
p-methyl analogues 8b and 12b gave an 11ꢂ improve-
ment in potency compared to DMP 850 and DMP 851
against anamprenavir resistant virus which is mutated
in both the S2/S20 pockets and the hinge region of the
flap.13
In conclusion, we have shown that increasing the lipo-
philicity of the nonsymmetrical 3-aminoidazoles DMP
850 and DMP 851 by modification of the P1/P10 resi-
dues improved the antiviral potency against wild type
HIV. Based on the protein binding adjusted IC90s and
the resistance profile against HIV mutants, the substituted
meta-methyl or para-methyl analogues are approximately
equipotent to DMP 850 and DMP 851. Unfortunately, the
pharmacokinetics outweigh the improvements in the anti-
viral potency, eliminating these substituted analogues as
potential development candidates.
References and Notes
1. (a) Vacca, J. P.; Condra, J. H. Drug Discov. Today 1997, 2,
261. (b) Danner, S. A.; Carr, A.; Leonard, J. M.; Leman,
L. M.; Gudiol, F.; Gonzales, J.; Raventos, A.; Rubio, R.;
Bouza, E.; Pintado, V.; Aguado, A. G.; De Lomas, J. G.;
Delgado, R.; Borleffs, J. C. C.; Hsu, A.; Valdes, J. M.; Bou-
cher, C. A. B.; Cooper, D. A. N. Engl. J. Med. 1995, 333,
1528. (c) Markowitz, M.; Saag, M.; Powderly, W. G.; Hurley,

608
R. F. Kaltenbach, III et al. / Bioorg. Med. Chem. Lett. 13 (2003) 605–608
A. M.; Hsu, A.; Valdes, J. M.; Henry, D.; Sattler, F.; La
Marca, A.; Leonard, J. M.; Ho, D. D. N. Eng. J. Med. 1995,
333, 1534. (d) Kitchen, V. S.; Skinner, C.; Ariyoshi, K.; Lane,
E. A.; Duncan, I. B.; Burckhardt, J.; Burger, H. U.; Bragman,
K.; Pinching, A. J.; Weber, J. N. Lancet 1995, 345, 952.
2. Rodgers, J. D.; Lam, P. Y. S.; Johnson, B. L.; Wang, H.;
Ko, S. S.; Seitz, S. P.; Trainor, G. L.; Anderson, P. S.; Klabe,
R. M.; Bacheler, L. T.; Cordova, B. C.; Garber, S.; Reid, C.;
Wright, M. R.; Chang, C.-H.; Erickson-Viitanen, S. Chem.
Biol. 1998, 5, 597.
3. Rodgers, J. D.; Johnson, B. L.; Wang, H.; Erickson-Viita-
nen, S.; Klabe, R. M.; Bacheler, L.; Cordova, B. C.; Chang,
C. H. Bioorg. Med. Chem. Lett. 1998, 8, 715.
4. Kaltenbach, R. F., III; Klabe, R. M.; Cordova, B. C.; Seitz,
S . P.Bioorg. Med. Chem. Lett. 1999, 9, 2259.
5. Computer generated Clog P was calculated using Med-
Chem Software Release 3.32, Pomona College, Claremont,
CA, USA.
6. Nugiel, D. A.; Jacobs, K.; Worley, T.; Patel, M.; Kalten-
bach, R. F., III; Meyer, D. T.; Jadhav, P. K.; DeLucca, G. V.;
Smyser, T. E.; Klabe, R. M.; Bacheler, L. T.; Rayner, M. M.;
Seitz, S. P. J. Med. Chem. 1996, 39, 2156.
7. Rossano, L. T.; Lo, Y. S.; Anzalone, L.; Lee, Y.-C.; Mel-
oni, D. J.; Moore, J. R.; Gale, T. M.; Arnett, J. F. Tetrahedron
Lett. 1995, 36, 4967.
8. Erickson-Viitanen, S.; Klabe, R. M.; Cawood, P. G.;
O’Neal, P. L.; Meek, J. L. Antimicrob. Agents Chemother.
1994, 38, 1628.
9. Bacheler, L. T.; Paul, M.; Jadhav, P. K.; Otto, M.; Stone,
B.; Miller, J. Antiviral Chem. Chemother. 1994, 5, 111.
10. Otto, M. J.; Reid, C.; Garber, S.; Lam, P. Y. S.; Scaranti,
H.; Bacheler, L. T.; Rayner, M. M.; Winslow, D. Antimicrob.
Chemother. 1993, 37, 2606.
11. (a) King, R. W.; Garber, S.; Winslow, D. L.; Reid, C.;
Bacheler, L. T.; Anton, E.; Otto, M. J. Antiviral Chem. Chemo-
ther. 1995, 6, 80. (b) Jadhav, P. K.; Ala, P.; Woerner, F. J.;
Chang, C.-H.; Garber, S. S.; Anton, E. D.; Bacheler, L. T. J.
Med. Chem. 1997, 40, 181.
12. Ala, P. J.; Huston, E. E.; Klabe, R. M.; Jadhav, P. K.;
Lam, P. Y. S.; Chang, C.-H. Biochemistry 1998, 37, 15042.
13. Partaledis, J. A.; Yamaguchi, K.; Tisdale, M.; Blair, E. E.;
Falcione, C.; Maschera, B.; Myers, R. E.; Pazhanisamy, S.;
Futer, O.; Cullinan, A. B.; Stuver, C. M.; Byrn, R. A.;
Livingston, D. J. J. Virol. 1995, 69, 5228.