'Double-drugs' - A new class of prodrug form of an HIV protease inhibitor conjugated with a reverse transcriptase inhibitor by a spontaneously cleavable linker

Article abstract of DOI:10.1016/S0960-894X(00)00202-X

We designed and synthesized a new series of prodrug-type anti-HIV agents consisting of a peptidomimetic HIV protease inhibitor conjugated with a nucleoside reverse transcriptase inhibitor in an effort to enhance the antiviral activity. For the conjugation, a series of linkers that conjoin the two different classes of inhibitors have been investigated. Conjugates using a succinyl amino acid linker were shown to release the parent components via the spontaneous imide formation at a faster rate compared to conjugates using a glutaryl amino acid linker, as expected from the energetically favorable cyclization to the five-membered ring. Herein, we report a new 'double-drug' 4b (KNI-1039) with a glutarylglycine linker, which exhibited extremely potent anti-HIV activity compared with that of the individual components. (C) 2000 Elsevier Science Ltd. All rights reserved.

Full text of DOI:10.1016/S0960-894X(00)00202-X

Bioorganic & Medicinal Chemistry Letters 10 (2000) 1227±1231
`Double-Drugs'Ð A New Class of Prodrug Form of an HIV
Protease Inhibitor Conjugated with a Reverse Transcriptase
Inhibitor by a Spontaneously Cleavable Linker
Hikaru Matsumoto, a Tomonori Hamawaki, a Hisashi Ota, a Tooru Kimura, a
Toshiyuki Goto, b Kouichi Sano, b Yoshio Hayashi a and Yoshiaki Kiso a,*
aDepartment of Medicinal Chemistry, Center for Frontier Research in Medicinal Science, Kyoto Pharmaceutical University,
Yamashina-ku, Kyoto 607-8412, Japan
bDepartment of Microbiology, Osaka Medical College, Takatsuki-shi, Osaka 569-8686, Japan
Received 29 February 2000; accepted 27 March 2000
AbstractÐWe designed and synthesized a new series of prodrug-type anti-HIV agents consisting of a peptidomimetic HIV protease
inhibitor conjugated with a nucleoside reverse transcriptase inhibitor in an e€ort to enhance the antiviral activity. For the con-
jugation, a series of linkers that conjoin the two di€erent classes of inhibitors have been investigated. Conjugates using a succinyl
amino acid linker were shown to release the parent components via the spontaneous imide formation at a faster rate compared to
conjugates using a glutaryl amino acid linker, as expected from the energetically favorable cyclization to the ®ve-membered ring.
Herein, we report a new `double-drug' 4b (KNI-1039) with a glutarylglycine linker, which exhibited extremely potent anti-HIV
activity compared with that of the individual components. # 2000 Elsevier Science Ltd. All rights reserved.
Introduction
single molecule, taking into consideration the anity of
nucleosides towards cell membrane.8 Previously, based
on this `double-drug' strategy, we had developed potent
prodrug-type anti-HIV agents,9,10 KNI-684 and KNI-
694, in which the carboxyl group of HIV PR inhibitors,
KNI-413 and KNI-852 was directly esteri®ed with the
50-hydroxyl group of a nucleoside RT inhibitor, 30-
azido-30-deoxythimidine (AZT) (Fig. 1).11 The anti-HIV
activities of prodrug-type conjugates, KNI-684 and
KNI-694 were found to be more potent than that of
AZT and the parent protease inhibitors.9,10 These
results suggested the following: (a) the `double-drug'
strategy involving the combination of two di€erent
classes of inhibitors together would enhance the anti-
HIV ecacy synergistically, and (b) also be e€ective in
the improvement of its physicochemical characteristics.
However, this `double-drug' strategy using the direct
esteri®cation method was limited to inhibitors containing
free carboxylic acid.
HIV-1 protease (HIV PR) and reverse transcriptase
(RT) are major targets for anti-HIV drug discovery, and
their inhibitors are widely used in the chemotherapy of
AIDS.1 Combination of HIV PR inhibitors and RT
inhibitors has become the standard clinical practice to
maintain the antiviral e€ect and to prevent emergence
of a drug-resistance virus.2,3 We have studied the struc-
ture±activity relationship of dipeptide-based HIV PR
inhibitors.4 Among them, KNI-413 (Fig. 1) and KNI-
727 (Fig. 2) exhibited potent HIV PR inhibition but
poor antiviral activity.4 6
The antiviral ecacy of inhibitors depends not only on
the enzyme inhibitory activity, but also on their intra-
cellular concentrations closely related to the membrane
permeability.7 The low antiviral activities of KNI-413
and KNI-727 were probably due to their insucient cell
membrane permeability. These ®ndings prompted us to
introduce the `double-drug' strategy that combines an
HIV PR inhibitor and a nucleoside RT inhibitor in a
The objective of the present study was to apply the
concept of `double-drug' strategy to another dipeptide
HIV PR inhibitor, KNI-727 which is de®cient of a car-
boxyl group. The poor water solubility due to the high
lipophilic nature of KNI-727 might restrict its penetra-
tion across the cell membrane, thus resulting in poor
*Corresponding author. Tel.: +81-75-595-4635; fax: +81-75-591-
9900; e-mail: kiso@mb.kyoto-phu.ac.jp
0960-894X/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(00)00202-X

1228
H. Matsumoto et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1227±1231
anti-HIV activity. To enhance the anti-HIV activity and
improve the physicochemical characteristics of KNI-
727, we designed and synthesized new hybrid-type pro-
drugs of KNI-727 conjugated with AZT by a series of
linkers (Fig. 2). Here we can report a new `double-drug'
4b (KNI-1039) with a glutarylglycine linker (Fig. 3),
which exhibits 920 and 62 times more potent anti-HIV
activity than KNI-727 and AZT, respectively.
Conjugation Strategy
As for the linkers, dicarboxylic acid derivatives were
used in the study to covalently link both the hydroxyl
groups of KNI-727 and AZT. At ®rst, we incorporated
succinic acid as a linker, but the resulting compound
KNI-935 (Fig. 4) was ine€ective (EC50=188.9 nM,
HIV-1IIIB/CEM-SS) in the anti-HIV assay. The prob-
able reason could be due to the high stability of the
succinyl ester linkage towards enzymatic cleavage as
described by Calenbergh and Nair,12 and hence the
parent drugs were not regenerated inside the cell. The
major consideration in the design of these prodrugs is
focused not only on the transportation across the cell
membrane, but also on its reversion to the parent com-
pounds in the cytoplasm to act on their respective tar-
gets.13 The essential criteria in the design of such
prodrugs are that (i) the `double-drugs' should contain a
linker that is stable outside the target cell and (ii) once
in the cytoplasm, it should regenerate the parent com-
pounds. Based on these criteria, a series of linkers which
have the ability to cleave spontaneously in physiological
environment were studied (Fig. 3).
As the esters of succinamic acid are easily hydrolysed via
succinimide formation under mild alkaline conditions,14
Figure 1. Structures of dipeptide HIV PR inhibitors, KNI-413 and
KNI-852, and the conjugates, KNI-684 and KNI-694.
Figure 2. Design and proposed mechanism of hybrid-type Anti-HIV agents.
Figure 3. Conversion of double-drugs to KNI-727 and AZT.

H. Matsumoto et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1227±1231
1229
we employed succinylglycine as a linker. The compound
with this linker could disintegrate spontaneously to
release KNI-727 and the ester intermediate 1. The
enzymatic hydrolysis of the ester intermediate 1 within
the cell could release AZT. In order to study the spon-
taneous cleavage rate, conjugates with various linkers
such as glutarylglycine, succinyl-b-alanine, and glutaryl-
b-alanine were synthesized and analyzed.
tected with 4 N HCl/dioxane to yield the corresponding
ester of AZT (2a) (90% yield). KNI-727 (0.18 mmol,
100 mg) was coupled with succinic anhydride (0.2
mmol, 26 mg) in THF-ether in the presence of dicyclo-
hexylamine (DCHA) (0.2 mmol, 43 mL) to yield the
half-ester of KNI-727 (3a) (90% yield). Condensation of
3a (0.15 mmol, 100 mg) and 2a (0.3 mmol, 108 mg)
using 1-ethyl-3-(3-(dimethylamino) propyl)carbodiimide
.
hydrochloride (EDC HCl) (0.15 mmol, 29 mg) in the
presence of 1-hydroxybenzotriazole (HOBt) (0.17 mmol,
25 mg) resulted in the prodrug (4a) which was ®nally pur-
i®ed by silica gel chromatography (70% yield). The
other prodrugs (4b, 4c, 4d) were also synthesized in a
similar manner.
Chemistry
The synthesis of these prodrugs is summarized in
Scheme 1. Brie¯y, AZT (1 mmol, 267 mg) was coupled
with Boc-Gly-OH (1.1 mmol, 193 mg) using DCC (1.1
mmol, 227 mg) in the presence of dimethylaminopyr-
idine (DMAP) (0.1 mmol, 12 mg). The product was
puri®ed by column chromatography and later depro-
Biological Evaluation and Discussion
The inhibitory potencies of these prodrug-type con-
jugates against HIV PR were evaluated as reported.6 All
of the conjugates exhibited poor enzyme inhibition,
whereas the parent HIV PR inhibitor, KNI-727, had a
good enzyme inhibitory activity (Table 1). This could be
attributed to the blockage of the hydroxyl moiety in
KNI-727, which plays a signi®cant role in the interac-
tion with the catalytic aspartyl residues of the HIV PR.4
The anti-HIV activities of the prodrug-type conjugates
were evaluated in in vitro conditions. Brie¯y, HIV-1LAI
(100 TCID50) and each 4-fold diluted prodrug were
inoculated to 5000/well Molt-4 cells in a microplate and
cultured for 7 days. EC50 values were calculated from the
ratios of surviving cells which were quantitated by WST-8
assay kit15 (Dojin Lab., Kumamoto, Japan). All of the
prodrug-type conjugates were found to be more potent
than the parent HIV PR inhibitor, KNI-727.
Figure 4. Structure of the conjugate KNI-935 using a succinic acid
linker.
The key feature of these prodrugs are the linkers which
disintegrate spontaneously in the aqueous media to
their parent compounds once inside the cell (Fig. 3). The
disintegration behavior of these prodrugs in phosphate
bu€ered saline (PBS, pH 7.4, at 37 ꢀC) was determined
by HPLC. For all the prodrugs studied, the sponta-
neous imide formation with the release of KNI-727 and
ester intermediate (1) were observed. The releasing cap-
ability of KNI-727 varied depending upon the linkers.
Compound 4a conjugated by a succinylglycine linker
had the faster releasing rate (t1=2=1.4 h) compared with
compound 4b containing
a gutarylglycine linker
(t1=2=23.3 h), as expected from the energetically favor-
able cyclization to the ®ve-membered ring. The t1=2 of
compounds (4c, 4d) containing b-alanine in the linker
were longer than those of compounds (4a, 4b) which
contain glycine.
Compound 4b (KNI-1039) containing a glutarylglycine
linker exhibited a remarkable anti-HIV activity (EC50=0.1
nM) and low cytotoxicity (therapeutic index>2000). This
antiviral activity was 920 and 62 times more potent than
KNI-727 and AZT, respectively. This result suggested
that the conjugates would penetrate the cell membrane
and then the two classes of enzyme inhibitors generated
could attack di€erent targets in the infected cells, thus
Scheme 1. Reagents and conditions: (a) i. Boc-Gly-OH or Boc-b-Ala-
OH, DCC, DMAP, DMF; ii. 4N HCl/ dioxane; (b) Succinic anhydride
.
or glutaric anhydride, DCHA, THF-ether (1:2). (c) EDC HCl, HOBt,
DMF.

1230
H. Matsumoto et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1227±1231
Table 1. Anti-HIV activities and physicochemical characteristics of `double-drugs'.
Compound
m
n
Inhibition of
HIV proteasea
EC50 HIV-1LAI
Molt-4
/
Relative
potency
t1=2
(h)b
Water solubility
(mg/mL)
4a (KNI-1038)
4b (KNI-1039)
4c (KNI-1046)
4d (KNI-1047)
KNI-727
1
1
2
2
2
3
2
3
37%
55%
7%
12%
100%
Ð
5.3 nM
0.1 nM
1.0 nM
3.4 nM
92.0 nM
6.2 nM
1.2
62
6.2
1.8
0.1
1.0
1.4
23.3
3.8
40.0
Ð
290
371
373
235
91
AZT
Ð
Ð
a% of HIV-1 protease inhibition in the presence of 5 mM of inhibitors.
bt1=2 is the time required for 50% release of KNI-727 at 37 ꢀC in phosphate bu€ered saline (pH 7.4).
exhibiting synergistic anti-HIV activity. On the other hand,
compound 4a (KNI-1038) containing a succinylglycine
linker showed similar antiviral activity as that of AZT,
which might be due to the instability of compound 4a
(t1=2=1.4 h). Compound 4a might disintegrate outside
the cell, resulting in the release of AZT intermediate
which may penetrate into the cell to inhibit RT. The t1=2
of 4c (KNI-1046) was longer than that of 4a, that is, the
anti-HIV activity of 4c was 5 times more potent than
that of 4a. Among all the prodrugs, compound 4d
(KNI-1047) exhibited longer half life (t1=2=40 h), how-
ever, its antiviral activity was poor. The poor antiviral
activity of 4d may be attributable to the high stability
which prevents the disintegration to its parent com-
pounds. These results show that the antiviral activity
may be in¯uenced by the stability of conjugates.16 The
antiviral ecacy of conjugated compounds depends on
many factors, such as enzyme inhibition, cell membrane
permeability, extracellular stability, intracellular disin-
tegration, and the correlation between them is very
complicated.
viral activity was 920 and 62 times more potent than that
of KNI-727 and AZT, respectively. We believe that this
`double-drug' approach in the ®eld of medicinal chemistry
would pave the way in the development of more potent
drugs. The extension of this `double-drug' strategy to
other compounds is in progress.
Acknowledgements
This research was supported in part by the Frontier
Research Program of the Ministry of Education,
Science and Culture of Japan, and grants from the
Ministry of Education, Science and Culture of Japan
and Japan Health Science Foundation. We thank Dr. C.
V. Ramesh for useful discussion and Ms. Y. Matsui for
technical assistance.
References and Notes
1. (a) Helm, K. Biol. Chem. 1996, 377, 765. (b) De Clercq, E. J.
Med. Chem. 1995, 28, 2491. (c) Kiso, Y. Biopolymers 1996, 40,
235.
2. Craig, J. C.; Whittaker, L.; Duncan, I. B.; Roberts, N.
Antiviral Chemistry Chemotherapy 1994, 5, 380.
The water solubility of these prodrugs was determined
by RP-HPLC. Aliquots of these prodrug solutions were
injected and the concentrations were determined by
comparison of the peak area (Table 1). All the prodrugs
exhibited better water solubilities (290±373 mg/mL) com-
pared to the parent compound KNI-727 which was less
water-soluble (91 mg/mL). These results indicate that the
prodrug-type conjugates satis®ed one of the essential cri-
teria in improving the physicochemical characteristics of
the compounds with a subtle balance between hydro-
philicity and lipophilicity, and thus would enhance the
cell membrane permeability.
3. (a) Gulick, R.; Mellors, J.; Havlir, D.; Eron, J.; Gonzalez,
C.; McMahon, D.; Richman, D.; Valentine, F.; Jonas, L.;
Meibohm, A.; Emini, E.; Chodakewitz, J. New Engl. J. Med.
1997, 337, 734. (b) Mathez, D.; Bagnarelli, P.; Gorin, I.;
Katalama, C.; Pialoux, G.; Saimot, G.; Tubiana, P.; De Tru-
chis, P.; Chauvin, J.-P.; Mills, R.; Rode, R.; Clementi, M.;
Leibowitch, J. Antiviral Ther. 1997, 2, 175. (c) Autran, B.;
Carcelain, G.; Li, T. S.; Blanc, C.; Mathez, D.; Tubiana, R.;
Katalaba, C.; Debre, P.; Leibowitch, J. Science 1998, 277, 112.
4. Kiso, Y.; Matsumoto, H.; Mizumoto, S.; Kimura, T.; Fuji-
wara, Y.; Akaji, K. Biopolymers 1999, 51, 59.
5. Kiso, Y.; Yamaguchi, S.; Matsumoto, H.; Mimoto, T.;
Kato, R.; Nojima, S.; Takaku, H.; Fukazawa, T.; Kimura, T.;
Akaji, K. Arch. Pharm. Pharm. Med. Chem. 1998, 331, 87.
6. Mimoto, T.; Kato, R.; Takaku, H.; Nojima, S.; Terashima,
K.; Misawa, S.; Fukazawa, T.; Ueno, T.; Sato, H.; Shintani,
M.; Kiso, Y.; Hayashi, H. J. Med. Chem. 1999, 42, 1789.
In conclusion, based on the prodrug concept, novel
hybrid-type anti-HIV agents were developed. Among the
prodrugs studied, compound 4b (KNI-1039), a `double-
drug' of an HIV PR inhibitor KNI-727 conjugated with
a nucleoside RT inhibitor AZT by a glutarylglycine lin-
ker, exhibited a remarkable anti-HIV activity. This anti-

H. Matsumoto et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1227±1231
1231
7. Bilello, J. A.; Bilello, P. A.; Stellrecht, K.; Leonard, J.;
Norbeck, D. W.; Kempf, D. J.; Robins, T.; Drusano, G. L.
Antimicrob. Agent. Chemother. 1996, 40, 1491.
8. For reviews of nucleoside transporters and receptors, see: (a)
Cass, C. E.; Young, J. D.; Baldwin, S. A. Biochem. Cell Biol. 1998,
76, 761. (b) Ralevic, V.; Burnstock, G. Pharmacol. Rev. 1998, 50,
413.
9. Kimura, T.; Matsumoto, H.; Matsuda, T.; Hamawaki, T.;
Akaji, K.; Kiso, Y. Bioorg. Med. Chem. Lett. 1999, 9, 803.
10. Kiso, Y.; Matsumoto, H.; Yamaguchi, S.; Kimura, T.
Letters Peptide Science 1999, 6, 275.
11. Huang, Q.-Q.; Yao, S. Y. M.; Ritzel, M. W. L.; Paterson, A.
R. P.; Cass, C. E.; Young, J. D. J. Biol. Chem. 1994, 269, 17757.
12. Calenbergh, S. V.; Nair, V. Nucleosides Nucleotides 1999,
18, 737.
13. Camplo, M.; Niddam, V.; Barthelemy, P.; Faury, P.;
Mourier, N.; Simon, V.; Charvet, A. S.; Trabaud, C.; Graciet,
J. C.; Chermann, J. C.; Kraus, J. L. Eur. J. Med. Chem. 1995,
30, 789.
14. Bodanszky, M.; Kwei, J. Z. Int. J. Peptide Protein Res.
1978, 12, 69.
15. Ishiyama, M.; Miyazono, Y.; Sasamoto, K.; Ohkura, Y.;
Ueno, K. Talanta 1997, 44, 1299.
16. In the cases of KNI-684 and KNI-694, the t1=2 were 2.9
and 3.8 h, respectively (the time required for 50% release of
AZT at 37 ꢀC in phosphate bu€ered saline at pH 7.4).