Functional profiling, identification, and inhibition of plasmepsins in intraerythrocytic malaria parasites

Article abstract of DOI:10.1002/anie.200903747

Profile and eliminate! Plasmepsins (PMs), aspartic proteases required for malariaparasite growth, are promising antimalarial targets. The in situ screening of PMs with probes formed from β-hydroxyazides 1 and alkynes with a photo-cross-linking unit and a

Full text of DOI:10.1002/anie.200903747

Angewandte
Chemie
DOI: 10.1002/anie.200903747
Malaria
Functional Profiling, Identification, and Inhibition of Plasmepsins in
Intraerythrocytic Malaria Parasites**
Kai Liu, Haibin Shi, Huogen Xiao, Alvin G. L. Chong, Xuezhi Bi, Young-Tae Chang,
Kevin S. W. Tan, Rickey Y. Yada, and Shao Q. Yao*
Malaria is a global disease which affects 300–500 million
people annually and kills 1–2 million. The most deadly form
of the disease is caused by the pathogen Plasmodium
falciparum. Currently, quinolines and antifolates are the
most common antimalarial drugs.[1] The cost of the drugs, as
well as the emergence of multidrug resistance, has, however,
become a major problem. Thus, there is a need for newer and
ideally cheaper drugs against this devastating disease.[2]
P. falciparum has two stages of growth, one sexual and one
asexual. The human asexual erythrocytic phase (blood stage)
is the cause of most malaria-associated pathology. Upon their
invasion of red blood cells (RBCs), the parasites differentiate
(ring stage), metabolize hemoglobin (trophozoite stage), and
replicate (schizont stage) over the following 48 h, before
being released (by rupture of the host cell) into the blood
stream. Proteases, including cysteine (i.e. falcipains) and
aspartic proteases (i.e. plasmepsins (PMs)), are required for
parasite growth through the digestion of human hemoglobin
and the delivery of necessary nutrients. They have long been
considered promising antimalarial targets.[3] Genomic data
obtained for P. falciparum predict at least ten genes that
encode aspartic proteases, four of which (PM-I, PM-II, PM-
IV, and the histoaspartic protease or HAP) have been found
so far, mostly in the food vacuole (FV) of the parasite. The
existence of the other hypothetical aspartic proteases, how-
ever, has not been confirmed experimentally.
In the last few years, drug-discovery efforts towards
potential plasmepsin inhibitors have somewhat waned after
gene-knockout experiments showed that parasites could still
survive, albeit with a reduced growth rate, without most of the
four functionally redundant FV plasmepsins.[4] It is now
believed that the only effective way to kill the parasite with
PM drugs would be with inhibitors that could simultaneously
target as many plasmepsins as possible.[5] At present, most
inhibitors developed are only effective against selected PMs[6]
owing to difficulties associated with recombinant expression
and the insufficient biochemical characterization of certain
PMs (i.e. PM-I and HAP) in vitro,[7a,b] as well as the lack of
methods that enable the simultaneous screening of the
activity of different PMs in situ.[7c] Previously, activity-based
probes (ABPs) were used successfully for the in situ screening
of malarial cysteine proteases.[8] We report herein the first
chemical proteomics approach for the functional profiling of
all four PMs in intraerythrocytic malaria parasites. This
strategy was made possible by the development of affinity-
based probes (AfBPs) against PMs (Scheme 1).[9] The in situ
screening of PMs with these probes against a focused library
of 152 hydroxyethyl-containing small molecules has led to the
identification of G16 as an effective inhibitor against the
parasite in infected RBC cultures.
In contrast to other previously known aspartic protease
probes, which were based on specific inhibitors against their
respective targets (e.g. presenilin and g-secretase),[10] we
aimed to establish a general approach that would be
applicable to a variety of aspartic proteases. The seven
probes A–G, each of which contains a hydroxyethyl-based
warhead “WH” with varying R1 and R2 groups, were
assembled by click chemistry from the corresponding azide-
containing WH and the alkyne, which contains a benzophe-
none (BP) photo-cross-linking unit and a tetraethylrhod-
amine (TER) reporter (Scheme 1, top).[11] Hydroxyethyl-
containing scaffolds are general transition-state analogues of
aspartic proteases. In probes A–G, aliphatic and aromatic
amino acid groups were chosen strategically, since they are
preferred in three of the four PMs (PM-I, PM-II, and PM-IV;
HAP is not well-characterized).[3,4,6] Other aspartic proteases
may be targeted in future by structural tuning of the WH. The
use of click chemistry for the efficient chemical assembly of
complex ABPs is well-documented.[12a] In our case, it also
provided rapid access to the 152 hydroxyethyl inhibitors A1–
H19 against the PMs (see the Supporting Information for
complete structures).[12b]
[*] K. Liu, Prof. Dr. S. Q. Yao
Department of Biological Sciences
National University of Singapore
14 Science Drive 4, Singapore 117543 (Singapore)
Fax: (+65)6779-1691
E-mail: chmyaosq@nus.edu.sg
Homepage: http://staff.science.nus.edu.sg/~syao
H. Shi, Dr. X. Bi, Dr. Y. T. Chang, Prof. Dr. S. Q. Yao
Department of Chemistry, National University of Singapore
3 Science Drive 3, Singapore 117543 (Singapore)
Dr. H. Xiao, Dr. R. Y. Yada
Department of Food Science, University of Guelph
Guelph, N1G 2W1 (Canada)
A. G. L. Chong, Dr. K. S. W. Tan
Department of Microbiology, National University of Singapore
5 Science Drive 2, Singapore 117597 (Singapore)
[**] Financial support was provided by the National University of
Singapore (R-143-000-394-112), the Agency for Science, Technology,
and Research (R-143-000-391-305), the National Research Founda-
tion (Competitive Research Programme R143-000-218-281), the
Natural Sciences and Engineering Research Council of Canada, and
the Canada Research Chairs Program.
The eight hydroxyethyl WHs were synthesized chemically.
Upon “click” assembly of the probes and the inhibitors, the
resulting compounds were further characterized and purified
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200903747.
Angew. Chem. Int. Ed. 2009, 48, 8293 –8297
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
8293

Communications
unambiguously
that
the
37 kDa band corresponds to
all four labeled PMs (magnifi-
cation in Figure 1b; see also
Figure S8 and Tables S1 and
S2 in the Supporting Informa-
tion). The probes could there-
fore be used to profile all four
known PMs and their enzy-
matic activities directly from
the parasite. The detection of
neither zymogens of the four
PMs nor other proteins indi-
cated the specificity of the
probes in targeting only
active PMs. Since it would be
highly desirable if the probes
could also detect the activity
of other previously unidenti-
fied aspartic proteases in the
malaria proteome (i.e. those
predicted by genomic data), a
deliberate effort was made in
the 2D-GE/MS experiments
to look for new protein spots.
However, these attempts were
unsuccessful; either these
hypothetical aspartic pro-
teases were present in very
low abundance or our probes
were not suitably designed to
detect/label them.
Subsequently, we deter-
mined the enzymatic activity
of the four plasmepsins by
Scheme 1. Assembly of affinity-based probes (AfBPs) and the 152-membered library of potential inhibitors
against all four FV plasmepsins in P. falciparum.
labeling both detergent-solu-
ble and detergent-insoluble
lysates from each develop-
(when necessary; see the Supporting Information). To
demonstrate the suitability of the probes for the UV-initiated
proteomic profiling of plasmepsins, we initially used recombi-
nant PM-I, PM-II, and HAP (Figure 1a; see also Figure S4 in
the Supporting Information); highly distinct labeling profiles
against different aspartic proteases were observed. This result
indicated that the variable R1 and R2 groups exerted a strong
influence over specific enzyme/probe interactions. The label-
ing was abolished by treatment with pepstatin and mutations
of key catalytic residues in the enzyme active site (see
Figures S5 and S6 in the Supporting Information); thus,
labeling was indeed dependent on enzyme activity.
The probes were subsequently used to label proteomes of
highly synchronized parasites obtained at different stages of
parasite development (ring, trophozoite, and schizont; Fig-
ure 1b). A 37 kDa protein band, which corresponds to the
molecular weight of the four known PMs, was highly visible
across probes A–G, although the labeling intensities varied.
Probe G consistently gave the strongest labeling profile and
thus was chosen for further studies. Analysis by two-dimen-
sional gel electrophoresis/MS and western blotting confirmed
ment stage (Figure 1c). The results indicated that PM activity
was highly regulated and peaked at the trophozoite stage for
the insoluble fraction. This observation is consistent with the
role of these PMs in hemoglobin degradation, which occurs in
the food vacuole. The distinct profile of soluble-fraction
activities peaked at the schizont stage; this peak probably
indicates a change in the subcellular localization of the PMs
(PM-II was previously shown to be released from the FV in
the late schizont stage). Western blotting with antibodies
against the four PMs also indicated that their absolute protein
expression levels corroborated well with labeling profiles
observed with probe G (see Figure S9 in the Supporting
Information). Thus, the utility of our probes for the accurate
reporting of the activities of PMs from crude malaria
proteomes was further confirmed.
One of the key advantages of these AfBPs is the ability to
use them for the simultaneous detection of the activities of
multiple PMs in their native environment. This so-called in
situ screening method, originally described by Cravatt and co-
workers for other enzymatic systems,[13] enabled us to identify
potential inhibitors against all four PMs directly from the
8294 www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 8293 –8297

Angewandte
Chemie
potency was confirmed in a dose-dependent in situ screening
assay (Figure 2a; see also Figure S10 in the Supporting
Information). One of the most potent inhibitors identified,
Figure 1. Profiling, identification, and proteome characterization of
P. falciparum aspartic proteases. a) Characterization of recombinant
PMs with the seven AfBPs A–G. Left: labeling profile of recombinant
PMs with the AfBPs; right: tree-view representation of labeled bands
(on the basis of their intensity). b) Parasite extracts at different stages
were labeled with the seven AfBPs (the TER–BP–alkyne was used as a
negative control). Top: in-gel fluorescence scanning showing the
specific labeling of the 37 kDa band across different stages; bottom
left: spectral counts of the labeled bands; bottom right: schizont-stage
parasite extracts were labeled with probe G and then subjected to 2D-
GE/MS analysis to identify the four PMs (spots 1–4 were identified as
PM-II, PM-I, HAP, and PM-IV, respectively). pI=isoelectric point.
c) Characterization of plasmepsin activities from total lysates (left),
NP-40-soluble fractions (middle), and insoluble fractions (right) across
different intraerythrocytic stages of P. falciparum. R=ring stage,
T=trophozoite stage, S=schizont stage. Top: spectral counts of the
respective bands observed by in-gel fluorescence scanning.
Figure 2. Inhibition of P. falciparum aspartic proteases. a) In situ
screening assay and determination of the IC50 values of G15 and G16
against all four PMs in the parasite proteome (for full details, see
Figure S10 in the Supporting Information). inh=inhibitor. b) Represen-
tative images of parasite-infected RBCs treated with G16 (10 mm; with
dimethyl sulfoxide (DMSO) and G15 as controls); arrows show the
abnormal development of parasites. Scale bar: 600 mm. c) Dose-
dependent-inhibition results from (b).
G16, showed an IC50 value of 937.5 nm. In contrast, G15, a
“false positive” identified from standard enzymatic assays by
using selected recombinant PMs, showed significantly weaker
inhibition (IC50 = 21.6 mm). This result underscores the impor-
tance of our in situ screening assay for the future discovery of
general PM inhibitors.
malaria proteome without the recombinant production of
every active PM.
The inhibitory effect of these candidate compounds was
tested with parasite-infected RBC cultures. The distinct
activity/solubility profiles of PMs (Figure 1c) prompted us
to test the inhibitory effect of the inhibitors against schizont-
stage PMs, which showed the highest activity in the detergent
nonyl phenoxylpolyethoxylethanol (NP-40). RBCs were
treated with the inhibitors 40 h after parasite invasion (i.e.
in the late schizont stage). Upon cell rupture, we measured
the percentage of parasites in the ring and schizont stages
(Figure 2b,c; see also Figure S11 and Table S3 in the Sup-
porting Information). Compound G16, but not G15, caused a
marked decrease in the number of newly formed ring-stage
parasites, and at the same time an increase in free extrac-
ellular merozoites. Thus, as well as blocking parasite develop-
To identify the most potent inhibitors against all four PMs,
we preincubated each compound of the 152-membered
hydroxyethyl-based library with the parasite lysate (see
Figure S10 in the Supporting Information) and then added
probe G and subjected the samples to UV irradiation. We
determined the relative potency of each inhibitor by measur-
ing the decrease in fluorescence intensity in the 37 kDa
labeled band. A total of eight candidate “hits” were
identified. (Four of these candidates were identified by
screening with individual recombinant PMs by using standard
enzymatic assays; the results are not shown.) These eight
compounds were resynthesized and purified, and their
Angew. Chem. Int. Ed. 2009, 48, 8293 –8297
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
8295

Communications
ment at the trophozoite/schizont stage, as one might expect,
enzyme active site (Figure 3e). A similar interaction was
observed previously between the hydroxy group in pepstatin
(a general aspartic protease inhibitor) and HAP.[15,16]
the inhibition of PM activity also caused the blockage of
either the escape of the parasites from RBCs, or their
reinvasion of RBCs. This speculation is further supported by
previous findings that PM-II was able to digest an RBC
membrane-skeleton protein in the late schizont stage at
neutral pH, and that the invasion of P. falciparum merozoites
was affected by treatment with pepstatin (a general aspartic
protease inhibitor).[14] The inhibition of parasites by G16 was
dose-dependent, whereby the estimated EC50 value of 1.04 mm
was similar to that obtained from the in situ proteomic
screening (IC50 = 937.5 nm). In contrast, G15 showed little or
no inhibition towards infected RBCs, even at the highest
concentration tested. Finally, G16 showed no apparent
cytotoxicity against common mammalian cell lines (see
Figure S12 in the Supporting Information).
In summary, we have developed the first affinity-based
probes for the functional profiling of all four PMs in
intraerythrocytic malaria parasites. Subsequent in situ screen-
ing of parasites with these probes led to the identification of a
compound, G16, which showed good inhibition against all
four PMs and parasite growth in infected RBCs. Molecular
modeling indicated that this inhibitor binds to the active site
of the plasmepsins tested, as originally designed. Our results
indicate the feasibility of using AfBP approaches for the
identification of inhibitors of less-characterized enzymes
(such as HAP) and inhibitors of multiple targets. We
anticipate that these new chemical tools should facilitate the
discovery of unknown parasite biology and new antimalarial
drugs. In the current study, we were unable to detect
previously predicted but unidentified aspartic proteases
from the malaria proteome. Thus, G16 might have further
targets, other than the four plasmepsins, in the malaria
proteome. Future research will focus on the development of
new chemical probes to address these issues.
Lastly, to gain insight into the binding mode of G16 with
FV PMs, we carried out molecular-docking studies with three
of the four PMs (Figure 3; the X-ray crystal structure of PM-I
Received: July 8, 2009
Published online: September 25, 2009
Keywords: activity-based profiling · affinity-based profiling ·
.
chemical proteomics · click chemistry · malaria
[1] R. W. Snow, C. A. Guerra, A. M. Noor, S. I. Hay, Nature 2005,
434, 214 – 217.
[2] C. Boss, O. Corminboeuf, C. Grisostomi, S. Meyer, A. F. Jones, L.
Prade, C. Binkert, W. Fischli, T. Weller, D. Bur, ChemMedChem
2006, 1, 1341 – 1345.
[3] a) D. A. Fidock, R. T. Eastman, S. A. Ward, S. R. Meshnick,
Trends Parasitol. 2008, 24, 537 – 544; b) P. J. Rosenthal, Int. J.
Parasitol. 2004, 34, 1489 – 1499; c) D. E. Goldberg, Curr. Top.
Microbiol. Immunol. 2005, 295, 275 – 291.
[4] a) J. Liu, E. S. Istvan, I. Y. Gluzman, J. Gross, D. E. Goldberg,
Proc. Natl. Acad. Sci. USA 2006, 103, 8840 – 8845; b) M. E.
Drew, R. Banerjee, E. W. Uffman, S. Gilbertson, P. J. Rosenthal,
D. E. Goldberg, J. Biol. Chem. 2008, 283, 12870 – 12876.
[5] a) F. Hof, A. Schutz, C. Fah, S. Meyer, D. Bur, J. Liu, D. E.
Goldberg, F. Diederich, Angew. Chem. 2006, 118, 2193 – 2196;
Angew. Chem. Int. Ed. 2006, 45, 2138 – 2141; b) A. Nezami, T.
Kimura, K. Hidaka, A. Kiso, J. Liu, Y. Kiso, D. E. Goldberg, E.
Freire, Biochemistry 2003, 42, 8459 – 8464.
Figure 3. Molecular docking of G16 in the active site of PM-II, HAP,
and PM-IV. G16 binds to the active-site binding pocket of a) PM-II,
b) HAP, and c) PM-IV. The hydroxy group of G16 interacts closely with
d) D34 and D214 in PM-II, e) H32 and D214 in HAP, and f) D34 and
D214 in PM-IV. g–i) Preferred conformation of G16 in the binding
pocket of PM-II (g), HAP (h), and PM-IV (i).
[6] K. Ersmark, B. Samuelsson, A. Hallberg, Med. Res. Rev. 2006,
26, 626 – 666.
has not been reported). The results showed that the molecule
binds well in the active site confined by the D34–D214
aspartic acid pair in the structure of PM-II and PM-IV, and by
H32 and D214 in the structure of HAP (Figure 3a–c).
Detailed analysis of the top 100 docked models identified
the preferred conformations of G16 in PM-II and PM-IV, as
well as in nonclassical HAP (Figure 3g–i). The results also
showed that the hydroxy group of G16 interacts closely with
D34 and D214 in PM-II and PM-IV, as originally designed
(Figure 3d,f). In binding to HAP, G16 assumed a position in
which its hydroxy group fitted between H32 and D214 in the
[7] a) H. Xiao, A. F. Sinkovits, B. C. Bryksa, M. Ogawa, R. Y. Yada,
Protein Expression Purif. 2006, 49, 88 – 94; b) H. Xiao, T. Tanaka,
M. Ogawa, R. Y. Yada, Protein Eng. Des. Sel. 2007, 20, 625 – 633;
c) D. C. Greenbaum, Trends Pharmacol. Sci. 2008, 29, 51 – 56.
[8] a) D. C. Greenbaum, A. Baruch, M. Grainger, Z. Bozdech, K. F.
Medzihradszky, J. Engel, J. DeRisi, A. A. Holder, M. Bogyo,
Science 2002, 298, 2002 – 2006; b) S. Arastu-Kapur, E. L. Ponder,
U. P. Fonovic, S. Yeoh, F. Yuan, M. Fonovic, M. Grainger, C. I.
Phillips, J. C. Powers, M. Bogyo, Nat. Chem. Biol. 2008, 4, 203 –
213.
[9] We previously proposed the use of the term AfBP for this kind of
active-site affinity probe: E. W. S. Chan, S. Chattopadhaya, R. C.
8296 www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 8293 –8297

Angewandte
Chemie
Panicker, X. Huang, S. Q. Yao, J. Am. Chem. Soc. 2004, 126,
14435 – 14446.
[13] D. Leung, C. Hardouin, D. L. Boger, B. F. Cravatt, Nat.
Biotechnol. 2003, 21, 687 – 691.
[10] a) A. Weihofen, K. Binns, M. K. Lemberg, K. Ashman, B.
Martoglio, Science 2002, 296, 2215 – 2218; b) Y. Takahashi, Y.
Konno, N. Watanabe, H. Miyashita, M. Sasaki, H. Natsugari, T.
Kan, T. Fukuyama, T. Tomita, T. Iwatsubo, ACS Chem. Biol.
2007, 2, 408 – 418.
[11] a) V. V. Rostovtsev, L. G. Green, V. V. Fokin, K. B. Sharpless,
Angew. Chem. 2002, 114, 2708 – 2711; Angew. Chem. Int. Ed.
2002, 41, 2596 – 2599; b) C. Christensen, M. Meldal, J. Org.
Chem. 2002, 67, 3057 – 3062.
[14] a) S. Le Bonniec, C. Deregnaucourt, V. Redeker, R. Banerjeei, P.
Grellier, D. E. Goldbergi, J. Schrevel, J. Biol. Chem. 1999, 274,
14218 – 14223; b) P. Dejkriengkraikhul, P. Wilairat, Z. Para-
sitenkd. 1983, 69, 313 – 317.
[15] P. Phaumik, H. Xiao, C. L. Parr, Y. Kiso, A. Gustchina, R. Y.
Yada, A. Wlodawer, J. Mol. Biol. 2009, 388, 520 – 540.
[16] According to the docking results, an S configuration was favored
for the carbon center attached to the hydroxy group in G16; see
the Supporting Information for details.
[12] a) M. J. Evans, B. F. Cravatt, Chem. Rev. 2006, 106, 3279 – 3301;
b) K. A. Kalesh, P. Y. Yang, R. Srinivasan, S. Q. Yao, QSAR
Comb. Sci. 2007, 26, 1135 – 1144.
Angew. Chem. Int. Ed. 2009, 48, 8293 –8297
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 8297