Potent inhibitors of the HIV-1 protease incorporating cyclic urea P1-P2 scaffold

Article abstract of DOI:10.1016/j.bmcl.2004.08.038

We have developed an efficient synthetic approach to analogues of potent HIV-protease inhibitor (PI) 4. Key chemistry includes TFA-mediated deprotection of MOB-protected 16 to precursor 17. Alkylation of 17 enabled the synthesis of many PIs, as exemplified here by 13, 19-21, in excellent yields and purity, and thus overcoming bottlenecks of our prior synthetic approach to these PIs. These results allowed to rapidly evaluate the SAR in P1-P2 cyclic urea-based HIV-1 PIs as well as provided a strong rationale towards P1-P2 pyrrolidone scaffold-based PIs, ultimately leading to subnanomolar PIs (follow-up communication). We have developed synthetic approaches to novel analogues of 2-imidazolidinone scaffold 2, which was found to be an effective P1-P2 mimetic in HIV-1 protease inhibitor 4. This enabled a rapid synthesis of analogues of 4 and subsequently allowed us to evaluate and rationalize the SAR. Accordingly, trans relationship of P1 and P2 substituents in the P1-P2 mimetic, as found in a related 2-pyrrolidone-based scaffold 1, was found necessary for high potency against HIV-1 protease. Results of this study provided further rationale towards subsequent optimization of 2-pyrrolidone-based lead 3, which led us to potent and drug-like HIV-1 protease inhibitors described in a follow-on report (Bioorg. Med. Chem. Lett. 2004, 14, in press. doi:10.1016/j.bmcl.2004.08.039).

Full text of DOI:10.1016/j.bmcl.2004.08.038

Bioorganic & Medicinal Chemistry Letters 14 (2004) 5685–5687
Potent inhibitors of the HIV-1 protease incorporating cyclic
urea P1–P2 scaffold
Wieslaw M. Kazmierski,* Eric Furfine, Yolanda Gray-Nunez, Andrew Spaltenstein and
Lois Wright
GlaxoSmithKline Research, 5 Moore Drive, Research Triangle Park, NC 27709, USA
Received 24 June 2004; revised 17 August 2004; accepted 17 August 2004
Available online 15 September 2004
Abstract—We have developed synthetic approaches to novel analogues of 2-imidazolidinone scaffold 2, which was found to be an
effective P1–P2 mimetic in HIV-1 protease inhibitor 4. This enabled a rapid synthesis of analogues of 4 and subsequently allowed us
to evaluate and rationalize the SAR. Accordingly, trans relationshipof P1 and P2 substituents in the P1–P2 mimetic, as found in a
related 2-pyrrolidone-based scaffold 1, was found necessary for high potency against HIV-1 protease. Results of this study provided
further rationale towards subsequent optimization of 2-pyrrolidone-based lead 3, which led us to potent and drug-like HIV-1 pro-
tease inhibitors described in a follow-on report (Bioorg. Med. Chem. Lett. 2004, 14, in press. doi:10.1016/j.bmcl.2004.08.039).
Ó 2004 Elsevier Ltd. All rights reserved.
The discovery of HIV-1 protease inhibitors (PI), has
been widely accepted as a key factor contributing to
1
3
Y=H
Indinavir
Y=A Ki=0.067 nM
P1'
the decrease of AIDS-related mortality rates in the
developed world. However, the relatively large pill bur-
den of the highly active antiretroviral therapy (HAART)
might be a factor in sometimes inconsistent patient
adherence to drug regimen. This in turn limits the ben-
efits of HAART by allowing the rebound of viral load
and potential emergence of viral resistance. Since much
of the pill burden associated with PIs is a direct conse-
quence of their limited aqueous solubility, our laborato-
ries have been involved in discovery efforts towards
novel PIs with improved solubility, potency and viral
resistance profiles.1–7 We recently disclosed spiro-pyrrol-
idone3 and spiro-morpholinone4 motifs that can serve as
P1–P2 mimetics in marketed HIV protease inhibitors,
such as Amprenavir8 or Indinavir (Fig. 1). In this and
a follow-up9 report we are describing new synthetic pro-
tocols developed towards additional P1–P2 scaffolds,
mono-substituted pyrrolidine 1 and 2-imidazolidinone 2.
Not only are scaffolds 1 and 2 more synthetically acces-
Ki=0.07 nM
OH
N
OH
N
H
N
Y
N
N
O
O
O
N
H
2
Y=H
4
Y=A Ki=1.2 nM
P2'
A
N
N
Y
O
Figure 1. P1–P2 mimetics 2-pyrrolidone 1 and 2-imidazolinone 2 and
associated HIV-1 protease inhibitors 3 and 4 incorporating the
indanolamine-based P10–P20 scaffold A.
sible than the spiro-molecules, but these motifs are also
potentially more attractive synthetic targets due to high
potency of lead compounds 3 and 4 (Fig. 1).1
Our initial efforts focused on 2-imidazolidinone
scaffold 2,8 the synthesis of which involved conversion
of Boc-phenylalanine to amide 5, which was then
reduced with borane to diamine 6 and cyclized to the
Keywords: HIV-1 protease inhibitor; Aspartyl protease inhibitor;
AIDS; Peptide mimetic; Peptidomimetic.
*
Corresponding author. Tel.: +1 919 483 9462; fax: +1 919 483
6053; e-mail: wieslaw.m.kazmierski@gsk.com
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.08.038

5686
W. M. Kazmierski et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5685–5687
Bn
Bn
Bn
OH
a, b
d
Bn
H
N
a
O
17
Boc-Phe
Bn
A
7
MOB
c
N
H2N
N
N
b
N
NH
MOB
X
Z
HN
N
O
O
O
16
Z=O
Z=H2
5
6
O
X= MOB
X= H
7
8
c
e
O
O
N
d
f
O
Bn
Ki [nM]
R
Bn
Bn
OH
12
Bn
OH
13, 3-picolyl
18.0
100.0
300.0
3.2
H
N
18, 3,4,5-trimethoxy Bn
19, 4-methylsulfonyl Bn
20, 3-F Bn
N
N
N
N
N
N
N
NH
R
A
21, 4-methoxy Bn
O
O
O
5.6
Ki [nM]
O
g, h
18.0
41.0
3-py 13
2-py 14
4-py 15
3-py 9
2-py 10
4-py 11
190.0
Figure 3. Reagents and conditions: (a) NaH/DMF, epoxide 12, 95%;
(b) neat TFA followed by basic workup, 65.6%; (c) t-Bu-OK, DMF,
Bn–X, 60–85%; (d) 4N HCl in dioxane/water (95:5, v/v), 93%.
Figure 2. Reagents and conditions: (a) EDCI, HOBt, p-meth-
oxybenzylamine; (b) TFA; (c) BH3ÆTHF in THF; (d) CDI, THF; (e)
CAN; (f) 2-, 3- and 4-picolyl chlorides, KO-t-Bu; (g) NaH/DMF, 12;
(h) 4N HCl in dioxane/water (95:5, v/v).
In contrast to the original route in Figure 2, the new
approach to 13 and related compounds described in Fig-
ure 3 eliminates extensive and laborious purification,
yielding ample quantities of products. These inhibitors
were then evaluated in the HIV-1 protease enzyme as-
say, but none proved more potent than the parent 4.
This SAR was also confirmed in the antiviral MT-4 cell
assay in which compounds 13, 16 and 18–21 were less
potent (2.2–10lM) than 4 (1.4lM). Molecular model-
ling of inhibitors of the type 18–21 in the context of
the HIV-1 protease active site led us to rationalize that
the relatively planar projection of the P2 substituent
from N1 in 2-imidazolidonone 2 precluded strong inter-
actions with the S2 pocket in the HIV-1 protease. Effec-
tive P1–P2 mimicry requires a more angular (sp3)
projection of P2 into the S2, as identified in 2-pyrrol-
idone motif 1. Consistent with this observation, adding
a cis-Me groupin 3 resulted in a largely inactive HIV-
1 protease inhibitor.14
p-methoxybenzyl (MOB)-protected urea 7 (Fig. 2). Sub-
sequent deprotection of 7 and alkylation of 2-imidazol-
inone 8 with 2-, 3- and 4-picolyl chlorides yielded the
desired 2-, 3- and 4-picolyl derivatives 9–11, albeit in
low yields. Considerable quantities of the unwanted
N3-alkyl regioisomers and of N1-, N3-bis-alkylation
products were also formed, necessitating laborious chro-
matographic purifications in order to secure small quan-
tities of 9–11.10 Anions derived from 9 to 11 were then
alkylated with epoxide 12,11 and the resulting intermedi-
ate acetonides were deprotected under acidic condi-
tions,12 yielding target inhibitors 13–15 (Fig. 2).
Evaluation of 13–15 in the HIV-1 protease enzyme assay
indicated that these picolyl P2 substituent-bearing inhib-
itors were substantially less potent than benzyl 4, sug-
gesting that the solubilizing picolyl moiety was
incompatible with the enzyme active site. We desired
to explore additional benzyl-substituted derivatives of
2. To overcome the synthetic hurdles encountered in
the synthesis of 13–15, we explored schemes, which did
not allow the formation of N3-alkyl and N1-, N3-bis-
alkyl products. The new route, starts with the coupling
of intermediates 7 and 12, to produce MOB-protected
16. We then intended to selectively deprotect the MOB
groupin 16 under conditions, which would preserve
the N-,O-acetonide moiety, and then N1-alkylate such
a derivative with a wide variety of benzyl halides. How-
ever, all attempts to deprotect 16 involving cerium
ammonium nitrate (CAN), including those previously
employed in the synthesis of 8 (Fig. 2), failed to produce
the desired product and instead yielded complex mix-
tures. Further experimentation led us to an observation
that the treatment of 16 with neat TFA for 5h, followed
by acid removal and basic workup13 resulted in interme-
diate 17 in good yield and purity. Rewardingly, 17
underwent facile alkylations with benzyl halides, allow-
ing us to rapidly synthesize a number of analogues,
including inhibitors 13 and 18–2010 (Fig. 3). In addition,
deprotection of acetonide 16 yielded 4-methoxy deriva-
tive 21 (Fig. 3).
This communication describes the development of a
robust synthetic protocol, which enabled rapid synthesis
of analogues of 2-imidazolone-based 4. Attempts to
optimize P2 substituents in this series and the emerging
SAR strongly suggest that successful P1–P2 mimicry
requires trans relationshipof both P1 and P2 substitu-
ents in the P1–P2 scaffold in order to maximize the inter-
actions with both S1 and S2 subsites.
Among the putative P1–P2 mimetics examined in the
course of this work, such as morpholinone, 1,2,5-thia-
diazolidine 1,1-dioxide, isothiazolidine 1,1-dioxide,
2-imidazolidinone and 2-pyrrolidone scaffolds,1–7 we
conclude that the latter is the most promising P1–P2
9
mimetic. Our follow-upreport discloses chemistry and
SAR developed in the 2-pyrrolidone scaffold 1 series,
leading to potent HIV-1 protease inhibitors with
enhanced drug-like properties.
Acknowledgement
We thank Richard Hazen for generating the MT4 data.

W. M. Kazmierski et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5685–5687
5687
13. Excessive (longer than 5h) exposure to the trifluoroacetic
acid results in decomposition of 16 to products of the type
B and C (Ref. 12). While shorter reaction time prevents
formation of B and C, we found out that the desired
product is obtained as a complex mixture of trifluoroacetyl
esters and trifluorocetylated amides. These can be con-
verted to the desired 17 by workupwith aqueous sodium
carbonate. Compound 17: 0.30g of 16 was dissolved in
10mL of TFA and stirred for 5h at room temperature.
After in vacuo rotary evaporation, the solid residue was
treated with 10% Na2CO3 in methanol/water for 10min,
yielding 0.15g of 17 (yield 65.6%). 1H NMR of 17 (CDCl3,
300MHz): d 8.10 (1H, d, J = 8.4Hz), 7.24 (10H, m), 7.05
(5H, m), 5.28 (1H, m), 4.10 (1H, m), 3.97 (1H, m), 3.53
(1H, m), 3.39 (2H, m), 2.95 (5H, m), 2.69 (2H, m), 2.54
(1H, m), 2.17 (1H, m), 1.92 (1H, m), 1.78 (1H, m).
Compound 13: 20mg of 17 (0.039mmol) in 1mL DMF
were stirred with potassium t-butoxide (26.3mg,
0.234mmol, 6equiv) for 10min, followed by addition of
6.3mg of 3-picolyl chloride in 1mL DMF. The reaction
was stirred for 20min, solvents removed in vacuo and the
product purified on HPLC yielding 14.2mg of 13 (yield
60.2%). 1H NMR of 13 (d6-acetone, 400MHz): d 8.57 (1H,
d, J = 5.3Hz), 8.42 (1H, s), 8.01 (1H, d, J = 8.0Hz), 7.80
(1H, m), 7.20 (14H, m), 6.92 (1H, d, J = 8.8Hz), 5.23 (1H,
m), 4.29 (1H, d, J = 16.2Hz), 4.29 (1H, m), 4.11 (1H, d,
J = 16.2Hz), 3.98 (2H, m), 3.48 (1H, dd), 3.18 (2H, m),
3.00 (4H, m), 2.75 (3H, m), 1.93 (1H, m), 1.88 (1H, m),
1.66 (1H, m).
References and notes
1. Spaltenstein, A.; Cleary, D.; Bock, B.; Kazmierski, W.;
Furfine, E.; Wright, L.; Hazen, R.; Andrews, W.; Almond,
M.; Tung, R.; Salituro, F. Bioorg. Med. Chem. Lett. 2000,
10, 1159.
2. Kazmierski, W.; Andrews, W.; Baker, C.; Bock, B.;
Cleary, D.; Furfine, E.; Gray-Nunez, Y.; Hale, M.; Hazen,
R.; Salituro, F.; Schairer, W.; Spaltenstein, A.; Tung, R.;
Wright, L. Book of Abstracts, 219th ACS National
Meeting, San Francisco, CA, March 26–30, 2000; poster
MEDI-126.
3. Kazmierski, W. M.; Furfine, E.; Spaltenstein, A.; Wright,
L. L. Bioorg. Med. Chem. Lett. 2002, 12, 3431.
4. Fritch, P. C.; Kazmierski, W. M. Synthesis 1999, 1, 112.
5. Kazmierski, W. M.; Fritch, P. C. In Methods in Molecular
Medicine, 1999; 23 (Peptidomimetics Protocols), pp 281–
291.
6. Kazmierski, W. M.; Bevans, P.; Furfine, E.; Spaltenstein,
A.; Yang, H. Bioorg. Med. Chem. Lett. 2003, 13, 2523.
7. Kazmierski, W. M.; Bevans, P.; Furfine, E.; Porter, D.;
Spaltenstein, A.; Yang, H. 223rd ACS National Meeting,
Orlando, FL, United States, April 7–11, 2002; Poster
MEDI-160.
8. Salituro, F. G.; Baker, C. T.; Court, J. J.; Deininger, D.
D.; Kim, E. E.; Li, B.; Novak, P. M.; Rao, B. G.;
Pazhanisamy, S.; Porter, M. D.; Schairer, W. C.; Tung, R.
D. Bioorg. Med. Chem. Lett. 1998, 8, 3637.
9. Kazmierski, W. M.; Andrews, W.; Furfine, E.; Spalten-
stein, A.; Wright, L. Bioorg. Med. Chem. Lett. 2004, 14,
in press. doi:10.1016/j.bmcl.2004.08.039.
14. For example, analogue E with the 3-methyl-substit-
uent (P2) cis to 5-benzyl (P1) was much less potent
(Ki = 170nM) than 3 (Ki = 0.05nM).
10. 1H and 13C NMR spectral and mass spectroscopy data for
all compounds support the proposed structures.
11. Askin, D.; Eng, K. K.; Rossen, K.; Purick, R. M.; Wells,
K. M.; Volante, R. P.; Reider, P. J. Tetrahedron Lett.
1994, 35, 673.
12. The deprotection of N-,O-acetonide A with TFA results in
substantial amounts of cyclitive cleavage products, lacton
B and indanolamine C. Additional experimens revealed
that under conditions of 4N HCl in dioxane (95%) and
Bn
Bn
e, f
a, b, a, c, d
N
NH
Boc
78%
OH
74%
Bn
O
O
D
Bn
OH
H
N
water (5%) formation of
suppressed.
B
and
C
is dramatically
N
O
O
E
O
Bn
O
Bn
Bn
O
OH
acid
N
Bn
N
N
(a) LHMDS ꢀ78°C; (b) MeI; (c) Bn–Br; (d) TFA; (e) Ref.
O
O
O
1; (f) 4N HCl in dioxane/water (95:5, v/v).
+
OH
B
H2N
A
C