Table 1 Properties of synthesized compounds
latter is achiral, conformational enantiomers (or diastereomers
in the case of 7b) are likely, especially upon binding to the
C2-symmetric HIV-protease. The dipeptide 10b incorporating
vinylogous valine was synthesized analogously.
In order to test the biological effect of extending the peptidic
side-arms, the tetrapeptide 11b was prepared in a direct manner
Compound
Solubility/m
M
IC50/mM
(R,S,S)-2b
(S,S,S)-2b
(R,S,S)-5b
(S,S,S)-5b
7b
> 65
> 65
> 50
> 50
> 100
~ 30
47
48
46
13
40
8
2.8
24
240
by NHS–DCC-mediated coupling of 6 with
L
-valinyl- -valine
L
ethyl ester. Here, as in all previous cases, the compounds were
purified by HPLC and characterized by standard spectroscopic
and analytical means.
10b
11b
13b
14
> 10
> 100
> 100
O
N
H
CO2R
CO2R
OH
OH
units (i.e. prevention of homo-dimer formation). In order to
shed some light on these aspects, kinetic studies using the model
of Zhang were carried out on select compounds, i.e. dissociative
inhibition constants (Ki) and competitive inhibition constants
(Kc) were measured.5 Accordingly, in the case of the most
active compound 11b, the Ki and Kc values turned out to be 6.9
CH2
H
N
O
and 2.0 m , respectively. This means that active site (com-
M
petitive) inhibition dominates, although dissociative inhibition
plays some role. Mixed inhibition also pertains to the related
10a R = Me (32%)
b R = H (80%)
dipeptide 7b (Ki = 15.8 m
M
; Kc = 6.2 m ). In contrast, the
M
mechanism of action of the tyrosin derivative 13b appears to be
O
based primarily on the inhibition of dimerization of the
H
N
CO2R
monomeric HIV-protease units (Ki = 6.8 m
M
; Kc = 229 m ).
M
N
H
This still needs to be studied more closely, e.g. using light
scattering. However, preliminary molecular modelling is in line
with these conclusions.
O
OH
OH
CH2
In summary, we have designed and prepared new HIV-
1-protease inhibitors based on naphtholic and phenolic units to
which amino acids or dipeptides are attached. Although the
respective activities are lower than those of the most potent
drugs currently known,1,2 the discovery of these new lead
structures allows for the (combinatorial) synthesis of analogs
which may show improved performance.
We thank H.-J. Schramm, J. Büttner and T. Wenger (group of
R. Huber at Max-Planck-Institut für Biochemie, Martinsried)
for help in the determination of IC50 values and kinetic data and
for stimulating discussions.
O
H
N
N
H
CO2R
O
11a R = Et (46%)
b R = H (16%)
Finally, the biphenol derivative 13 was prepared by oxidative
coupling3 of N-benzyloxycarbonyl-
-tyrosine methyl ester 12.
Deprotection delivered the dipeptide 14 (Scheme 4).
L
HO OH
OH
HO OH
Notes and References
† E-mail: reetz@mpi-muelheim.mpg.de
i,ii
iii
1 Reviews: G. Moyle and B. Gazzard, Drugs, 1996, 51, 701; J. W.
Erickson, Nat. Struct. Biol., 1995, 2, 523; E. K. Wilson, Chem. Eng.
News, July 29, 1996, p. 42; C. Perez, M. Pastor, A. R. Ortiz and F. Gago,
J. Med. Chem., 1998, 41, 836.
CO2Me
H N
CO2H
2
ZNH
CO2R
NH2
CO2H
ZNH
CO2R
ZNH
12
2 See for example: M. V. Hosur, T. N. Bhat, D. J. Kempf, E. T. Baldwin,
B. Lui, S. Gulnik, N. E. Wideburg, D. W. Norbeck, K. Appelt and J. W.
Erickson, J. Am. Chem. Soc., 1994, 116, 847; G. T. Wang, S. Li, N.
Wideburg, G. A. Krafft and D. J. Kempf, J. Med. Chem., 1995, 38, 2995;
T. N. Bhat, E. T. Baldwin, B. Liu, Y.-S. E. Cheng and J. W. Erickson,
Struct. Biol., 1994, 1, 552; C. N. Hodge, P. Y. S. Lam, C. J. Eyermann,
P. K. Jadhav, Y. Ru, C. H. Fernandez, G. V. De Lucca, C.-H. Chang,
R. F. Kaltenbach III, E. R. Holler, F. Woerner, W. F. Danecker, G.
Emmett, J. C. Calabrese and P. E. Aldrich, J. Am. Chem. Soc., 1998, 120,
4570; W. W. Wilkerson, S. Dax and W. W. Cheatham, J. Med. Chem.,
1997, 40, 4079; P. K. Jadhav, P. Ala, F. J. Woerner, C. H. Chang, S. S.
Garber, E. D. Anton and L. T. Bacheler, J. Med. Chem., 1997, 40, 181.
3 A. G. Brown and P. D. Edwards, Tetrahedron Lett., 1990, 31, 6581; S. M.
Kupchan, O. P. Dhingra and C.-K. Kim, J. Org. Chem., 1978, 43,
4076.
13a R = Me (52%)
b R = H (94%)
14 (90%)
(Z = benzyloxycarbonyl)
Scheme 4 Reagents: i, VOF3; ii, LiOH, H2O; iii, H2, Pd-C
In order to screen the ability of the compounds to inhibit the
HIV-1-protease, the IC50 values were measured using standard
procedures.4 Table 1 shows that several compounds have
activities similar to a number of other HIV-protease inhibitors
which have been reported in recent years.2 It is interesting to
note that in the case of 5b (but not 2b) the absolute
configuration of the binaphthol backbone plays a significant
role in the degree of HIV-protease inhibition. Specifically, the
(R,S,S)-compound is considerably more active than the (S,S,S)-
diastereomer. However, the conformationally more flexible
compounds based on embonic acid 6 are more active, especially
the tetrapeptide 11b.
4 H.-J. Schramm, J. Boetzel, J. Büttner, E. Fritsche, W. Göhring, E. Jaeger,
S. König, O. Thumfart, T. Wenger, N. E. Nagel and W. Schramm,
Antiviral Res., 1996, 30, 155.
5 Z.-Y. Zhang, R. A. Poorman, L. L. Maggioria, R. L. Heinrikson and F. J.
Kézdy, J. Biol. Chem., 1991, 266, 15 591.
Theoretically, the mode of action of the above compounds
can either be due to active-site inhibition of the HIV-protease or
to a possible inhibition of dimerization of the two 99-amino acid
Received in Cambridge, UK, 15th July 1998; 8/05489D
2076
Chem. Commun., 1998