HIV protease-mediated activation of sterically capped proteasome inhibitors and substrates

Article abstract of DOI:10.1021/ja109377p

Strategies for selectively killing HIV-infected cells present an appealing alternative to traditional antiretroviral drugs. We show here the first example of an inactive "Trojan horse" molecule that releases a cytotoxic, smallmolecule proteasome inhibitor upon cleavage by HIV-1 protease. As a proof-of-concept strategy, the protein avidin was used to block entry of the compound into the proteasome in the absence of HIV-1 protease. We demonstrate that this strategy is also feasible without requiring an exogenous protein; a polylysine dendrimer-containing molecule is unable to enter the proteasome until cleaved by HIV-1 protease. These results demonstrate that conditional proteasome inhibitors could prove useful in the development of new tools for chemical biology and future therapeutics.

Full text of DOI:10.1021/ja109377p

COMMUNICATION
pubs.acs.org/JACS
HIV Protease-Mediated Activation of Sterically Capped Proteasome
Inhibitors and Substrates
Dennis L. Buckley,† Timothy W. Corson,‡, Nicholas Aberle,‡ and Craig M. Crews*,†,‡,§
†Department of Chemistry, ‡Department of Molecular, Cellular & Developmental Biology, and
§Department of Pharmacology, Yale University, New Haven, Connecticut 06511, United States
S Supporting Information
b
ABSTRACT: Strategies for selectively killing HIV-infected
cells present an appealing alternative to traditional antire-
troviral drugs. We show here the first example of an inactive
“Trojan horse” molecule that releases a cytotoxic, small-
molecule proteasome inhibitor upon cleavage by HIV-1
protease. As a proof-of-concept strategy, the protein avidin
was used to block entry of the compound into the protea-
some in the absence of HIV-1 protease. We demonstrate
that this strategy is also feasible without requiring an
exogenous protein; a polylysine dendrimer-containing mol-
ecule is unable to enter the proteasome until cleaved by
HIV-1 protease. These results demonstrate that conditional
proteasome inhibitors could prove useful in the develop-
ment of new tools for chemical biology and future ther-
apeutics.
he development of anti-HIV drugs, including reverse tran-
scriptase inhibitors, protease inhibitors, co-receptor inhibi-
T
Figure 1. A “Trojan horse” that releases a small molecule capable of
reaching the catalytic threonine of the proteasome only after cleavage by
HIV-1 protease.
tors, and integrase inhibitors, has changed HIV infection into a
chronic, mostly treatable condition.1 However, drug resistance is
problematic, and this is exacerbated by the need for continual anti-
retroviral treatment.2,3 As an alternative to the simple manage-
ment of long-term HIV infection, strategies have been developed
that seek to exploit cleavage by HIV-1 protease to activate “Trojan
horses” that selectively kill HIV-infected cells. Previously, these
strategies have employed proteins such as caspases,4 proteinac-
eous toxins,5 and ribonucleases6 that normally exist as inactive
zymogens or contain a destabilizing sequence. These proteins were
modified to contain an HIV-1 protease recognition sequence;
cleavage by HIV-1 protease activated or stabilized the cytotoxic
protein, leading to cell death.
1 would be released (Figure 2a), which we have found to be a
potent inhibitor of the proteasome's chymotrypsin-like activity
with low nanomolar cytotoxicity, similar to other epoxyketone-
based proteasome inhibitors.11
As a proof of concept, we synthesized 2, incorporating a biotin
moiety that, in the presence of avidin, would create a steric cap
large enough to prevent entry into the proteasome (Figure 2a).
Instead of including an epoxyketone pharmacophore, we first
incorporated a fluorogenic aminomethylcoumarin (AMC) moiety12
that gives a direct readout of the ability of 2 to enter the proteasome.
As shown in Figure 2b, when incubated with 20S proteasomes in
the absence of avidin, 2 is able to enter the proteasome, where
AMC is liberated and fluoresces. However, in the presence of
avidin, 2 cannot enter the proteasome, preventing AMC release.
Upon addition of HIV-1 protease, an increase in fluorescence is
detected, indicating that the protease-labile linker is cleaved,
allowing the released fragment to enter the proteasome and
AMCtobehydrolyzedbytheproteasomalcatalyticsites(Figure2b).
Encouraged by this result, we next synthesized 3, which con-
tained a Leu-epoxyketone13 in place of the Leu-AMC, and
To make this approach more drug-like, we sought to develop a
compound that would release a cytotoxic small molecule instead
of a protein. Our strategy involves the HIV-protease-mediated
release of a cytotoxic epoxyketone proteasome inhibitor7 that is
rendered inactive by incorporating a large “blocking” group
(Figure 1). This steric cap prevents entry of the inhibitor into the
hollow catalytic chamber of the proteasome, which has an opening
of 1.3 nm.8,9 On the basis of a comparison of SQNY/PIV, an HIV
protease consensus sequence, with structure-activity data from
proteasome inhibitor design studies,10,11 the peptide SQNY/PIVF
was chosen to form a protease-labile linker between a leucine
epoxyketone and the steric cap. Upon cleavage by HIV protease,
Received: October 18, 2010
Published: December 27, 2010
r
2010 American Chemical Society
698
dx.doi.org/10.1021/ja109377p J. Am. Chem. Soc. 2011, 133, 698–700
|

Journal of the American Chemical Society
COMMUNICATION
Figure 2. (a) Structures of biotin-containing compounds and the released epoxyketone proteasome inhibitor. (b) Activity of 2 as a proteasome
substrate with or without pretreatment with HIV-1 protease and in the presence of avidin (0.72 mg/mL, 10 U/mL) or bovine serum albumin (BSA).
Activity of proteasome substrates is determined by fluorescence of released AMC, normalized against starting fluorescence and monitored over time.
(c) Activity of 3 as a proteasome inhibitor with or without pretreatment with HIV-1 protease and in the presence of avidin (0.72 mg/mL, 10 U/mL) or
BSA. (d) Dose-dependent response of 3 with avidin (0.72 mg/mL, 10 U/mL), with or without pretreatment with HIV-1 protease. Activity of
proteasome inhibitors is measured by inhibition of fluorescence of AMC released from Suc-LLVY-AMC, normalized against starting fluorescence and
monitored over time.
Figure 3. (a) Structures of generations 0-3 of 4 and generations 1-3 of 5. (b) Activity of generations 0-3 of 4 as proteasome substrates. (c) Activity of
generations 1-3 of 5 as proteasome substrates. (d) Activity of generation 3 of 5 with or without pretreatment with HIV-1 protease. Activity of
proteasome substrates is determined by fluorescence of released AMC, normalized against starting fluorescence and monitored over time.
demonstrated that it inhibits proteasome activity (as measured
by the cleavage of the fluorogenic substrate LLVY-AMC) in the
absence of avidin. The addition of avidin decreased the inhibitory
activity of 3, while the addition of HIV-1 protease liberated 1,
restoring potent proteasomal inhibition (Figure 2c). The dose
dependency of this effect was studied by measuring proteasome
activity in the presence of various concentrations of 3 with or with-
out exposure to HIV protease (Figure 2d). The chymotrypsin-like
activity of the proteasome was inhibited by 3 with an IC50 of 58 nM
in the presence of avidin after incubation with HIV-1 protease, a
7.7-fold increase in potency compared to the control lacking
HIV-1 protease (IC50 = 450 nM).
699
dx.doi.org/10.1021/ja109377p |J. Am. Chem. Soc. 2011, 133, 698–700

Journal of the American Chemical Society
COMMUNICATION
At this point, we began to explore options that did not require
an ectopic protein, which led us to dendrimers as possible steric
caps. In particular, we decided to base our steric cap on dendritic
polylysines, which were developed as cell-permeable transfection
reagents.14-17 WhilepreviousworkbyDeMartinoandco-workers18
has shown conclusively that the proteasome can accommodate
more than one peptide chain, we felt it would be possible to
create a large enough branched dendrimer to prevent entry into
the proteasome. We synthesized generations 0 through 3 of den-
dritic polylysine compound 4 in order to determine the minimal
size of the dendrimer needed to prevent proteasome entry
(Figure 3a). We first used a shortened peptide linker, allowing us
to ignore the possibility that an overly long linker might enable
the AMC substrate to reach the active site inside the proteasome
even if the dendrimeric end of the molecule cannot enter. Our
assays showed that the ability of these compounds to act as
proteasome substrates decreased sharply with increasing den-
drimeric size. While the second generation of 4 (containing four
terminal lysine residues) still showed moderate activity as a
proteasome substrate, the third generation of 4 (containing eight
terminal lysines) had no detectable activity (Figure 3b).
’ ACKNOWLEDGMENT
This work is funded by a grant from the Bill & Melinda Gates
Foundation through the Grand Challenges Exploration Initiative.
T.W.C. was the Canadian Institutes of Health Research Jean-Franc-ois
St-Denis Fellow in Cancer Research and a Bisby Fellow. N.A. was
the American-Australian Association's Alcoa Foundation Fellow.
The reagent EVA630, HIV protease, was obtained from the Program-
me EVA Centre for AIDS Reagents, NIBSC, UK, supported by
the EC FP6/7 Europrise Network of Excellence, AVIP and
NGIN consortia, and the Bill and Melinda Gates GHRC-CAVD
Project and was donated by Dr. Iva Pitchova from the Institute of
Organic Chemistry and Biochemistry of the Czech Academy of
Science, Praha, Czech Republic.
’ REFERENCES
(1) Mehellou, Y.; De Clercq, E. J. Med. Chem. 2010, 53, 521.
(2) Pomerantz, R. J.; Horn, D. L. Nat. Med. 2003, 9, 867.
(3) Geeraert, L.; Kraus, G.; Pomerantz, R. Annu. Rev. Med. 2008,
59, 487.
(4) Vocero-Akbani, A. M.; Heyden, N. V.; Lissy, N. A.; Ratner, L.;
Dowdy, S. F. Nat. Med. 1999, 5, 29.
Generations 1-3 of compound 5, which contained the full
SQNY/PIVF linker, were then synthesized. These compounds
showed size dependence similar to that of the various generations
of 4, with the third generation again showing no activity (Figure 3c).
We then tested the third generation of 5 after incubation with HIV-
1 protease. While the protease-free control had no activity and
was identical to the DMSO control, the third generation of 5 show-
ed significant activity as a proteasome substrate after protease
cleavage. This demonstrates the ability of HIV-1 protease to
rescue the activity of our compound (Figure 3d).
We have shown the first example of a cytotoxic small-molecule
released from a “Trojan horse” compound by HIV-1 protease. We
demonstrated the validity of our strategy using biotin and avidin
as a steric cap, which prevents access of the cytotoxic inhibitor to
the proteasome active site until activation by HIV-1 protease. Next,
we demonstrated the feasibility of this strategy without exogen-
ous proteins through the use of polylysine dendrimers. We showed
that entry into the proteasome could be effectively blocked through
the use of a third-generation lysine dendrimer, while activity re-
mained in the first and second generations, in agreement with
previous work showing that the proteasome can incorporate at
least two peptide chains.18 We were able to rescue the activity of
our proteasome substrates through the addition of HIV-1 protease
to our compounds containing a protease-labile linker. This work
provides the foundation for the design of future conditional pro-
teasome inhibitors for use in chemical biology studies of the pro-
teasome and in the treatment of HIV and other diseases.
(5) Falnes, P. O.; Welker, R.; Kr€ausslich, H. G.; Olsnes, S. Biochem. J.
1999, 343 (Pt. 1), 199.
(6) Turcotte, R. F.; Raines, R. T. AIDS Res. Hum. Retroviruses 2008,
24, 1357.
(7) Meng, L.; Mohan, R.; Kwok, B. H.; Elofsson, M.; Sin, N.; Crews,
C. M. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 10403.
(8) Groll, M.; Kim, K.-B.; Kairies, N.; Huber, R.; Crews, C. M. J. Am.
Chem. Soc. 2000, 122, 1237.
(9) L€owe, J.; Stock, D.; Jap, B.; Zwickl, P.; Baumeister, W.; Huber, R.
Science 1995, 268, 533.
(10) Moore, M. L.; Bryan, W. M.; Fakhoury, S. A.; Magaard, V. W.;
Huffman, W. F.; Dayton, B. D.; Meek, T. D.; Hyland, L.; Dreyer, G. B.;
Metcalf, B. W. Biochem. Biophys. Res. Commun. 1989, 159, 420.
(11) Elofsson, M.; Splittgerber, U.; Myung, J.; Mohan, R.; Crews,
C. M. Chem. Biol. 1999, 6, 811.
(12) Zimmerman, M.; Ashe, B.; Yurewicz, E. C.; Patel, G. Anal.
Biochem. 1977, 78, 47.
(13) Laidig, G. J.; Radel, P. A.; Smyth, M. S. U.S. Patent 2005/
0256324 A1, November 17, 2005.
(14) Kaneshiro, T. L.; Wang, X.; Lu, Z.-R. Mol. Pharmaceutics 2007,
4, 759.
(15) Ohsaki, M.; Okuda, T.; Wada, A.; Hirayama, T.; Niidome, T.;
Aoyagi, H. Bioconjugate Chem. 2002, 13, 510.
(16) Okuda, T.; Sugiyama, A.; Niidome, T.; Aoyagi, H. Biomaterials
2004, 25, 537.
(17) Watanabe, K.; Harada-Shiba, M.; Suzuki, A.; Gokuden, R.;
Kurihara, R.; Sugao, Y.; Mori, T.; Katayama, Y.; Niidome, T. Mol. BioSyst.
2009, 5, 1306.
(18) Liu, C.-W.; Corboy, M. J.; DeMartino, G. N.; Thomas, P. J.
Science 2003, 299, 408.
’ ASSOCIATED CONTENT
S
Supporting Information. Chemical and biochemical meth-
b
ods and characterization of novel compounds. This information
is available free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
craig.crews@yale.edu
Present Addresses
Department of Ophthalmology, Indiana University School of
Medicine.
700
dx.doi.org/10.1021/ja109377p |J. Am. Chem. Soc. 2011, 133, 698–700