Plasmodium falciparum cysteine protease falcipain-1 is not essential in erythrocytic stage malaria parasites

Plasmodium falciparum cysteine protease falcipain-1 is not essential in erythrocytic stage malaria parasites Puran S. Sijwali*†, Kentaro Kato†‡, Karl B. Seydel‡, Jiri Gut*, Julie Lehman*, Michael Klemba§, Daniel E. Goldberg§, Louis H. Miller‡, and Philip J. Rosenthal*¶ *Department of Medicine, San Francisco General Hospital, University of California, San Francisco, CA 94143-0811; ‡Laboratory of Malaria and Vector Research, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD 20892; and §Howard Hughes Medical Institute and Departments of Medicine and Molecular Microbiology, Washington University, St. Louis, MO 63110 Contributed by Louis H. Miller, April 16, 2004

M

alaria is one of the most important infectious disease problems of humans, particularly in developing countries. Plasmodium falciparum, the most virulent human malaria parasite, is responsible for hundreds of millions of illnesses and more than 1 million deaths each year (1). The control of malaria is increasingly limited by the resistance of malaria parasites to available drugs. New antimalarial drugs, ideally directed against new targets, are needed. Among potential new targets for antimalarial chemotherapy are cysteine proteases. The best-characterized plasmodial cysteine proteases are falcipains, which are papain-family cysteine proteases that are expressed by erythrocytic parasites (2, 3) and share many unusual features (3). Two nearly identical copies of falcipain-2 and a single copy of falcipain-3 are encoded nearly contiguously in an ⬇12-kb segment of chromosome 11 (http:// PlasmoDB.org). These proteases are similar in sequence (68% identity between falcipain-2 and falcipain-3) and appear to function as food vacuole hemoglobinases (4, 5), as was recently confirmed for falcipain-2 (6). Falcipain-1 is encoded on chromosome 14, is less similar in sequence to the other falcipains (38–40% identical to falcipain-2 and falcipain-3), and has an uncertain function (7). Study of falcipain-1 has been hindered by the need for an adequate system for heterologous expression. Recently, a chemical genetics approach provided insight into the function of falcipain-1 (8). The falcipain-1 inhibitor YA29Eps(S,S) (YA29) was reported to specifically block erythrocyte invasion by merozoites, suggesting that falcipain-1 mediates this process. To more definitively characterize the function of falcipain-1 in erythrocytic malaria parasites, we knocked out the falcipain-1 gene and evaluated the effects of YA29 and other protease inhibitors against WT and falcipain-1 knockout parasites. We found, surprisingly, that erythrocytic stage falcipain-1 knockout www.pnas.org兾cgi兾doi兾10.1073兾pnas.0402738101

parasites developed normally. Furthermore, YA29 and other cysteine protease inhibitors had similar effects against knockout and WT parasites, with inhibition of hemoglobin hydrolysis, parasite development, and subsequent erythrocyte rupture, but no apparent inhibition of erythrocyte invasion. Materials and Methods Materials. P. falciparum strains D10, 3D7, HB3, and W2 were obtained from the Malaria Research and Reference Reagent Resource Center (Manassas, VA). The transfection plasmid pDC and compound WR99210 were gifts from David Fidock (9) and David Jacobus (Jacobus Pharmaceuticals, Princeton), respectively. Compound YA29 was a gift from Matthew Bogyo (Stanford University, Stanford, CA) and was synthesized as reported in ref. 8. All other biochemical reagents were from Sigma or Fisher. Parasite Culture. P. falciparum parasites were cultured in human

erythrocytes at 2% hematocrit in RPMI medium 1640 supplemented with 10% human serum (10). Synchronization was maintained by serial treatment with 5% D-sorbitol (11).

Construction of Transfection Plasmids and Generation of Falcipain-1 Knockout Parasites. DNA encoding truncated falcipain-1 (⌬FP1;

lacking coding sequence for the first 179 N-terminal and the last 37 C-terminal amino acid residues, including a catalytic Asn) was PCR-amplified from the pTOP-FP1 plasmid by using Taq DNA polymerase (Invitrogen) and appropriate primers (forward: 5⬘AATGCATGCAAGCTTCGTGAAGAAGAAAAAGATGATAAAAAAGTATATC-3⬘; reverse: 5⬘-CCAGTCGACAAGCT TA ATCCA ATAGTATATA ATAT TATCATCTGG-3⬘; HindIII sites are underlined). Plasmid pDC contains PcDT 5⬘兾hDHFR兾Pfhrp2-3⬘ and cam 5⬘兾hsp86-3⬘ expression cassettes for selection with WR99210 and expression of the desired gene in P. falciparum parasites, respectively (9). The cam 5⬘兾hsp86-3⬘ expression cassette was replaced with HindIII-digested ⌬FP1 to construct the transfection plasmid pD-FP1. Ring stage D10strain parasites were transfected with 100 ␮g of pDC-FP1 (prepared with the Mobius 1000 Plasmid kit from Novagen) and selected with WR99210 as described in refs. 6 and 12. Genomic DNA (gDNA) was extracted from parasites and analyzed for chromosomal integration of the plasmid by PCR with integration (1UP: 5⬘-AAATCATCCGCGGCCGCTCGAGATAATAATAAAAATG-3⬘; MR: M13 reverse primer, Invitrogen) and WT (1UP; 1R: 5⬘-TGGTTAAGCTTTTACAAGATAGGATAGA AGAC-3⬘) specific primers. To eliminate Abbreviations: YA29, YA29-Eps(S,S); gDNA, genomic DNA; E-64, N-(trans-epoxysuccinyl)L-leucine 4-guanidinobutylamide; E-64d, (2S,3S)-trans-epoxysuccinyl-L-leucylamido-3methylbutane ethyl ester; ALLN, acetyl-leucine-leucine-norleucine aldehyde. †P.S.S.

and K.K. contributed equally to this work.

¶To

whom correspondence should be addressed at: Box 0811, University of California, San Francisco, CA 94143-0811. E-mail: [email protected].

© 2004 by The National Academy of Sciences of the USA

PNAS 兩 June 8, 2004 兩 vol. 101 兩 no. 23 兩 8721– 8726

MICROBIOLOGY

Among potential new targets for antimalarial chemotherapy are Plasmodium falciparum cysteine proteases, known as falcipains. Falcipain-2 and falcipain-3 are food vacuole hemoglobinases that may have additional functions. The function of falcipain-1 remains uncertain. To better characterize the role of falcipain-1 in erythrocytic parasites, we disrupted the falcipain-1 gene and characterized recombinant parasites. Disruption of the falcipain-1 gene was confirmed with Southern blots, and loss of expression of falcipain-1 was confirmed with immunoblots and by loss of labeling with a specific protease inhibitor. Compared with wild-type parasites, falcipain-1 knockout parasites developed normally, with the same morphology, multiplication rate, and invasion efficiency, and without significant differences in sensitivity to cysteine protease inhibitors. In wild-type and knockout parasites, cysteine protease inhibitors blocked hemoglobin hydrolysis in trophozoites, with a subsequent block in rupture of erythrocytes by mature schizonts, but they did not inhibit erythrocyte invasion by merozoites. Our results indicate that although falcipain-1 is expressed by erythrocytic parasites, it is not essential for normal development during this stage or for erythrocyte invasion.

parasites containing episomal plasmids, we subjected cultures to two 2-week cycles of incubation with and without WR99210 (5 nM). We then cloned recombinant parasites by limiting dilution and analyzed multiple clones by PCR for falcipain-1 gene disruption. Southern Hybridization. gDNA was isolated from schizont-stage parasites by using the PureGene DNA isolation kit (Gentra Systems), digested with XmnI or HincII–BsrGI restriction endonucleases, separated on agarose gels, transferred to Hybond-N membranes (Amersham Pharmacia Biosciences), and probed with [␣-32P]dATP-labeled truncated falcipain-1 (Megaprime DNA labeling system, Amersham Pharmacia Biosciences), as described in ref. 5. Immunoblotting. Parasites were harvested at 6–10% parasitemia

at 12-h intervals; parasite pellets were suspended in 200 ␮l of PBS [with 1 mM PMSF, 5 mM EDTA, 2 mM benzamidine䡠HCl, 10 ␮M pepstatin, and 1 mM 4-(2-aminoethyl)benzenesulfonyl fluoride (AEBSF)], and parasite proteins were resolved by SDS兾PAGE (13) and transferred onto poly(vinylidene difluoride) membranes (Bio-Rad). Membranes were incubated with blocking buffer (PBS兾0.2% casein兾0.1% Tween 20) containing appropriate dilutions of antibodies to purified recombinant falcipains [mature falcipain-1, truncated-profalcipain-2, and truncated-profalcipain-3 (5)], followed by alkaline phosphataseconjugated goat anti-rat IgG in blocking buffer (1兾5,000 dilution), and reactions were developed by using a Western-Star chemiluminescence kit (Tropix, Bedford, MA). Measurement of Parasite Development. Synchronized parasites

were cultured in complete medium for three cycles with change of medium after every cycle. Every 48 h, when parasites were at the ring stage, triplicate aliquots were fixed with 1% formaldehyde in PBS (pH 7.4) for 48 h at room temperature and labeled with YOYO-1 (1 nM; Molecular Probes) in 0.1% Triton X-100 in PBS (14). Parasitemias were determined from dot plots (forward scatter vs. fluorescence) acquired on a FACSort flow cytometer by using CELLQUEST software (Becton Dickinson). To evaluate the effects of YA29 on erythrocyte rupture and invasion, synchronized parasites were incubated with or without 10 ␮M YA29 (in 1% DMSO) at multiple postinvasion time points, and parasitemias were determined at different intervals by microscopic counting. To compare effects on knockout and parental parasites, early schizonts (40 h postinvasion) were purified on 63% Percoll (Sigma) and cultured at 2–3% parasitemia for 12 h with N-(trans-epoxysuccinyl)-L-leucine 4guanidinobutylamide (E-64), (2S,3S)-trans-epoxysuccinyl-Lleucylamido-3-methylbutane ethyl ester (E-64d), YA29, or no inhibitor (in all cases in 0.5% DMSO). After 12 h, smears were made for microscopy, and remaining parasites were processed for flow cytometry, as described. Test of Proplasmepsin Processing. Trophozoite-stage PM2GT兾B7 parasites expressing a plasmepsin II- GFP fusion (15) were cultured with 5 ␮g兾ml brefeldin A for 2 h at 37°C, washed once in medium, and washed with and resuspended in medium containing 0.2% DMSO or 100 ␮M acetyl-leucine-leucinenorleucine aldehyde (ALLN), 10 ␮M YA29, or 100 ␮M E-64, each in 0.1–0.2% DMSO. Parasites were incubated for an additional 3–4 h, and fluorescence and phase-contrast images were collected as described in ref. 15. Effect of Protease Inhibitors. IC50 values for the inhibition of

recombinant falcipain-2 and falcipain-3 were determined as described in refs. 16 and 17. To evaluate the antiparasitic effects of protease inhibitors and antimalarial drugs, we cultured parasites for 48 h, beginning at the ring stage, with different

8722 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0402738101

Fig. 1. Schematic representation of integration into the falcipain-1 locus. Regions encoding the pro and mature domains of the protease, human DHFR (hDHFR), and PCR primers (1UP, 1R, MR) are labeled. Thin dotted, thick broken, and thick lines represent the plasmid backbone, chromosomal DNA, and regulatory sequences, respectively. Two contiguous copies of the plasmid are indicated by 2 x pD-FP1. Bent arrows indicate 5⬘–3⬘ orientation of genes, horizontal arrows primer binding sites, and vertical arrows restriction endonuclease sites. Sizes of restriction fragments produced with XmnI (X), HincII (H), and BsrGI (B) are shown in kilobases in parentheses.

concentrations of compounds. Parasitemias were determined by flow cytometry, and IC50 values were calculated as described in refs. 16 and 17. In each case, goodness of curve fit was documented by R2 values of ⬎0.95. We analyzed IC50 values by using the paired t test in PRISM (GraphPad, San Diego), and P values ⬍0.05 were considered statistically significant. Results Disruption of the Falcipain-1 Gene. To disrupt the falcipain-1 gene,

P. falciparum D10 strain parasites were transfected with a pD-FP1 plasmid carrying a truncated falcipain-1 gene and a PcDT 5⬘兾hDHFR兾Pfhrp2-3⬘ expression cassette for selection with WR99210 (Fig. 1). After 12 weeks of culture with WR99210, plasmid integration at the falcipain-1 locus was verified by PCR. Parasites then were subjected to two rounds of selection with and without drug, followed by cloning by limiting dilution to obtain a homogenous population. Two falcipain-1 knockout clones (KO2 and KO11) were studied in detail. PCR with WT and integration-specific primers indicated integration of the plasmid into the FP1 locus and absence of the intact falcipain-1 gene (Fig. 2A). Southern blotting showed genomic restriction digestion patterns for each clone consistent with disruption of the falcipain-1 locus by single-site homologous crossover-mediated insertion of three copies of the plasmid (Fig. 2B). Immunoblotting with monospecific antibodies to falcipains confirmed the complete absence of falcipain-1 expression in recombinant parasites, whereas expression of falcipain-2 and falcipain-3 remained very similar to that in WT parasites (Fig. 2C) (5). As final confirmation of the knockout, binding of the generic cysteine protease inhibitor 125I-labeled DCG04 to a protein previously identified as falcipain-1 (8) was seen in the parental strain, but not recombinant parasites (M. Bogyo, personal communication). Characterization of Falcipain-1 Knockout Parasites. The morphology

of WT and falcipain-1 knockout parasites was the same throughout the life cycle. The knockout parasites showed growth patterns very similar to those of WT parasites; after three cycles of growth parasitemias were nearly identical for the WT strain and two knockout clones (Fig. 3A). Knockout and WT parasites also exhibited nearly identical rates of invasion of erythrocytes (Fig. 3B). For all three lines, ring parasitemias increased 4- to 5-fold over the schizont parasitemias of the prior cycle. Effects of Cysteine Protease Inhibitors on Erythrocyte Rupture and Invasion. Broadly active cysteine protease inhibitors such as E-64

are known to block parasite development by inhibiting hemo-

Sijwali et al.

globin hydrolysis (18, 19) and possibly by action against erythrocyte rupture or invasion (18, 20–22). Recently, a competition analysis evaluating effects of a peptidyl epoxide inhibitor library against P. falciparum lysates identified a falcipain-1-specific inhibitor, YA29. This compound was reported to inhibit erythrocyte invasion by merozoites, suggesting a specific function for falcipain-1. However, the hypothesis that falcipain-1 mediates erythrocyte invasion was not consistent with prior studies that did not show the inhibition of erythrocyte invasion by cysteine protease inhibitors (20) or our result that falcipain-1 knockout parasites had normal invasion rates. To clarify the effects of YA29, we evaluated the morphologies of 3D7-strain parasites after incubations with YA29 beginning at the late ring (16 h postinvasion) or schizont (40 h postinvasion) stages (Fig. 4). In both cases, parasites developed swollen, dark-staining food vacuoles, consistent with an inhibition of hemoglobin hydrolysis, as described for other cysteine protease inhibitors (17, 18). We also closely followed the loss of schizonts and formation of rings after incubation of YA29 with early schizonts (Fig. 5A). YA29 caused a modest inhibition of the rupture of schizonts (seen as the accumulation of schizonts, compared with control), beginning at 50 h postinvasion, when schizont rupture and new ring formation were occurring in control parasites. Beginning at 48 h postinvasion, ring parasitemia was lower in treated than control cultures, consistent with the explanation that YA29 exerted a modest inhibition in new ring formation because of its inhibition of normal parasite maturation and possibly independent effects on erythrocyte rupture by mature schizonts. The inhibition of ring formation was more pronounced when YA29 Sijwali et al.

Fig. 3. Multiplication and invasion rates. (A) Multiplication. Parasitemias were measured serially for WT and falcipain-1 knockout (KO2 and KO11) parasites, cultured in identical conditions, beginning with ring-stage parasites at 0.5% parasitemia. (B) Invasion. Percoll-purified schizonts (40 h postinvasion) were grown with fresh erythrocytes (2% hematocrit), and counts of schizonts at the beginning of the assay and rings after 12 h of culture were obtained. Error bars represent the standard deviations of means from three replicates of representative experiments.

was added earlier in the life cycle (36 or 38 h postinvasion), presumably because earlier incubation exerted more pronounced inhibitory effects on hemoglobin hydrolysis (Fig. 5B). The peptidyl epoxide YA29 was previously identified as a falcipain-1-specific inhibitor (8), although a direct measure of inhibition was not possible because of the lack of recombinant active falcipain-1. However, recombinant falcipain-2 and falcipain-3 are available. To better appreciate the inhibitory speci-

Fig. 4. Effect of YA29 on morphology. 3D7-strain parasites were incubated with 10 ␮M YA29 or 1% DMSO beginning at either 16 (A and B) or 40 (C and D) h postinvasion, and photomicrographs were subsequently made at 40 (A and B) and 52 (C and D) h. PNAS 兩 June 8, 2004 兩 vol. 101 兩 no. 23 兩 8723

MICROBIOLOGY

Fig. 2. Evaluation of recombinant parasites. WT and two falcipain-1 knockout clones (KO2 and KO11) were assessed. (A) PCR. Products were amplified for the indicated gDNAs by using primers specific for integration (1UP兾MR) and WT (1UP兾1R) falcipain-1 loci. Sizes of markers (M) are shown in kilobases. (B) Southern blot. gDNA digests were probed with an [␣-32P]dATP-labeled truncated falcipain-1 gene. Patterns for XmnI (X) and HincII–BsrGI (H-B) digested gDNAs are consistent with integration of three copies of plasmid DNA into the falcipain-1 locus. (C) Immunoblot. Total parasite lysates (4 ⫻ 107 parasites per lane) of ring (R), early trophozoite (ET), late trophozoite兾early schizont (LT兾 ES), and mature schizont (S) stages of WT and knockout parasites were probed with monospecific antibodies to the three proteases.

Fig. 6. Effects of inhibitors on processing of plasmepsin II. Fluorescence (PMII-GFP) and phase-contrast images of parasites treated with DMSO (A and B), ALLN (100 ␮M; C and D), YA29 (10 ␮M; E and F), and E-64 (100 ␮M; G and H). Only ALLN produced a ring of fluorescence around the food vacuole that corresponds to a block in processing.

Fig. 5. The effects of YA29 on schizont maturation over time. (A) Progression from schizonts to rings. Synchronized 3D7-strain parasites were treated with (continuous lines) or without (dashed lines) 10 ␮M YA29 40 h after erythrocyte invasion, and the percentages of erythrocytes infected with schizonts (filled symbols) and rings (open symbols) were counted microscopically every 2 h. (B) Effect of the time of initiation of treatment. Parasites in a representative experiment were treated with (filled bars) or without (open bars) 10 ␮M YA29 at the indicated time after invasion, and the subsequent ring parasitemia 52 h after the initial invasion was determined microscopically.

essed fluorescent plasmepsin II in the food vacuole membrane, which appeared as a fluorescent ring surrounding the food vacuole (15). This unique distribution of GFP-tagged plasmepsin II was caused by the peptidyl inhibitor ALLN, but not E-64 (16). To determine whether the effects of YA29 were due, at least in part, to inhibition of plasmepsin II processing activity, we treated parasites expressing GFP-tagged plasmepsin II with ALLN, E-64, and YA29, and evaluated the morphology of live parasites by fluorescence microscopy. As with E-64, YA29 did not cause the morphology that is indicative of an inhibition of plasmepsin processing (Fig. 6). Therefore, it does not appear that the inhibitory effect of YA29 was because of its action on the processing of plasmepsins. Sensitivity of Falcipain-1 Knockout Parasites to Protease Inhibitors.

ficity of YA29, we assessed the actions of it and of other cysteine protease inhibitors against recombinant enzymes and on the development of three P. falciparum strains. YA29 showed modest activity against falcipain-2, but not falcipain-3 (Table 1). Results with E-64 and YA29 were similar for three different strains of P. falciparum, including those sensitive (3D7, HB3) and resistant (W2) to chloroquine and other antimalarial drugs (Table 1). Curiously, the more membrane-permeant E-64 analog, E-64d, was 4 times more effective against W2-strain parasites than against drug-sensitive strains. Another predicted cysteine protease activity of erythrocytic parasites is the processing of plasmepsin aspartic proteases (15, 23, 24). Inhibition of this activity in parasites transfected with GFP-tagged plasmepsin II led to the accumulation of unproc-

We next compared the effects of cysteine protease inhibitors and other compounds on the development of WT and knockout parasites. Cysteine protease inhibitors were somewhat more active against falcipain-1 knockout parasites than against WT, although in all cases changes in sensitivity were not statistically significant (Fig. 7). Interestingly, knockout clone KO11 was more sensitive than clone KO2 to E-64 and E64d. The aspartic protease inhibitor pepstatin was equally active against knockout and parental parasites. The antimalarial drug chloroquine was

Table 1. Inhibition of recombinant protease activity and parasite multiplication Protease activity, ␮M* Compound E-64 E-64d YA29 Pepstatin Chloroquine

Parasite multiplication, ␮M†

Falcipain-2

Falcipain-3

3D7

W2

HB3

0.012 4.9 86.4 ND ND

0.032 6.3 ⬎400 ND ND

2.9 1.9 9.4 8.3 0.01

2.6 0.5 8.8 6.8 ⬎0.10

2.1 1.8 21.1 6.3 0.009

ND, not determined. *All values are mean IC50, based on three replicates for each inhibitor, with all experiments performed with the same concentration of enzyme. †Mean IC 50 values were determined from two experiments, each done in duplicate. 8724 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0402738101

Fig. 7. Parasite sensitivity to inhibitors. WT and falcipain-1 knockout parasites (KO2 and KO11) were incubated with multiple concentrations of inhibitors for one full life cycle, beginning at the ring stage. IC50s were calculated [for WT: E-64, 2.8 ␮M; E-64d, 1.5 ␮M; YA29, 11.2 ␮M; pepstatin (Pep), 5.9 ␮M; and chloroquine (CQ), 17 nM], and values for knockout parasites were compared with those for WT. Error bars indicate standard deviations from means of three experiments, each done with two replicates.

Sijwali et al.

somewhat more active against the knockout parasites, although this difference was not statistically significant (Fig. 7). To assess the action of cysteine protease inhibitors on erythrocyte rupture and invasion in falcipain-1 knockout parasites, synchronized schizont-stage parasites (40 h postinvasion) were treated with different concentrations of inhibitors, and counts of schizonts and rings were performed 12 h later. E-64, E-64d, and YA29 caused decreases in ring formation at the same concentrations at which schizonts were seen to accumulate, consistent with our hypothesis that the inhibition of new ring formation was because of a block in development, and possibly an additional independent effect on schizont rupture (Fig. 8; results with E-64 are not shown). Activities against WT and falcipain-1 knockout parasites were nearly identical, suggesting that the action of YA29 was not due only to the specific inhibition of falcipain-1. Discussion New transfection technologies allow us to precisely examine the functions of P. falciparum proteins. Knockout strategies are particularly attractive for defining the functions of enzymes, such as falcipain-1, for which heterologous expression systems are not well developed. The knockout of falcipain-1 was readily achieved, in part because the loss of this protein did not confer a significant disadvantage to the parasite. Falcipain-1 knockout parasites had normal morphologies and multiplied at the same rate as WT parasites. Thus, falcipain-1 is not essential for erythrocytic stage parasites. Our results do not exclude a potential role for falcipain-1 in erythrocytic parasites, however, because up-regulation of other proteases may have compensated for loss of falcipain-1 in the knockout parasites. P. falciparum likely requires protease activity for multiple processes during the erythrocytic cycle, including hemoglobin degradation, the hydrolysis of host and parasite proteins that accompanies erythrocyte rupture and then invasion, and the processing of plasmodial proteins throughout the life cycle. Attempts to dissect the functions of plasmodial proteases have included studies of the effects of protease inhibitors on cultured Sijwali et al.

PNAS 兩 June 8, 2004 兩 vol. 101 兩 no. 23 兩 8725

MICROBIOLOGY

Fig. 8. Effect of inhibitors on maturation of schizonts and invasion. Percollpurified WT (open bars) and knockout (KO2, filled bars; KO11, hatched bars) schizonts (40 h postinvasion) were combined with fresh erythrocytes (2% hematocrit, 2–3% parasitemia) and grown with 0.5% DMSO or indicated concentrations of the inhibitors for 12 h. Counts of schizonts (A and B) and rings (C and D) then were made. Error bars indicate standard deviations from means of two replicates. Counts of schizonts were corrected by subtracting the number of unruptured schizonts in a DMSO control from those in treated cultures and are expressed as percentage of the number of schizonts at the beginning of treatment.

parasites. Cysteine protease inhibitors block hemoglobin hydrolysis and development by trophozoites, leading to a characteristic morphological abnormality in which the food vacuole fills with undegraded hemoglobin (18, 19). In addition, because of the inhibition of development or independent effects, cysteine protease inhibitors block the release of merozoites from schizontinfected erythrocytes at the completion of the erythrocytic cycle (18, 20–22). In in vitro studies, falcipain-2 and falcipain-3 hydrolyzed hemoglobin (4, 5) and falcipain-2 also hydrolyzed the erythrocyte cytoskeletal proteins ankyrin and band 4.1 (25, 26), but until recently, specific roles of cysteine proteases in the parasite life cycle have been uncertain. We now appreciate that four typical papain-family cysteine proteases are present in P. falciparum: falcipain-1, two nearly identical copies of falcipain-2, and falcipain-3. Falcipain-2 and falcipain-3 have been well characterized recently, and appear to participate in hemoglobin degradation by trophozoites, as both proteases have been localized to trophozoite food vacuoles, possess acidic pH optima, and degrade hemoglobin in vitro (4, 5). In addition, it was demonstrated recently that disruption of the falcipain-2 gene led to decreased hemoglobin degradation by trophozoites, proving a role for falcipain-2 in this process (6). Attempts to knock out falcipain-3 have to date been unsuccessful (P.S.S., unpublished observation), suggesting that this protease, which unlike falcipain-2 is encoded by a single-copy gene, plays a key role in erythrocytic parasites. Despite falcipain-1 being the first plasmodial cysteine protease gene identified (7), its detailed biochemical characterization has not been possible because of inadequate systems for the production of recombinant protease (27). Recently, a chemical genetics approach was used to circumvent this problem by identifying a falcipain-1 inhibitor (YA29) by using an inhibitor competition assay and then evaluating the biological effects of this inhibitor. This compound appeared to specifically inhibit the invasion of erythrocytes by P. falciparum merozoites, leading to a proposed role for falcipain-1 in this process. However, it remained unclear whether the specificity of YA29 for falcipain-1 was complete. It also is a concern that these results conflicted with prior studies with isolated merozoites that showed no inhibition of erythrocyte invasion by broadly active cysteine protease inhibitors (20, 28). To further examine the role of falcipain-1 in erythrocytic parasites, we disrupted the falcipain-1 gene. Falcipain-1 knockout parasites developed normally, clearly indicating that this protease is not essential for erythrocytic malaria parasites. The selection of falcipain-1 knockout parasites may have been accompanied by the elaboration of an invasion mechanism different from that in WT parasites. To consider this possibility, the effects of YA29 were studied in detail. Compared with untreated parasites, those treated with YA29 demonstrated an accumulation of unruptured schizonts. Earlier treatment led to a more pronounced inhibitory effect, presumably because of more pronounced inhibition of hemoglobin hydrolysis in trophozoites and early schizonts than in more mature parasites. The concentration of YA29 required to block development and schizont rupture was in the micromolar range, but varied somewhat between experiments, perhaps because of evaluation of different parasite strains and兾or minor differences in the stages at which parasites were treated. However, it is important to note that our studies included the same three P. falciparum strains that were previously demonstrated to show a block in erythrocyte invasion after incubation with YA29 (8). The effects of the broadly active cysteine protease inhibitors E-64 and E-64d did not differ notably from those of YA29, which was identified as a specific falcipain-1 inhibitor but also has modest activity against recombinant falcipain-2. All of the cysteine protease inhibitors blocked hemoglobin hydrolysis in trophozoites and early schizonts and inhibited subsequent erythrocyte rupture by mature schizonts. Our results are consistent

with prior observations that the trophozoite stage is most sensitive to treatment with cysteine protease inhibitors (29) and with the conclusion that the principal activity of these agents is to block hemoglobin hydrolysis. Falcipain-2 and falcipain-3 seem to be the principal hemoglobinases that are acted on by cysteine protease inhibitors, as supported by correlations in many classes of inhibitors between action against these enzymes, inhibition of hemoglobin hydrolysis, and prevention of parasite development (16, 17, 30, 31). It remains unclear whether effects of cysteine protease inhibitors on erythrocyte rupture were due only to the consequences of the inhibition of hemoglobin hydrolysis, or whether inhibition of other activities of cysteine proteases also played a role. Cysteine protease-mediated activation of plasmepsin aspartic proteases also may contribute indirectly to hemoglobin hydrolysis, although the proteases responsible for this processing are unknown (16, 22). If cysteine proteases independently mediate erythrocyte rupture, potential candidates include falcipain-2, which cleaves the erythrocyte cytoskeletal proteins ankyrin and band 4.1 in vitro (25, 26); falcipain-3, which is maximally expressed late in the life cycle (5); serine repeat antigen family proteins, which share homology with cysteine proteases and are expressed in mature schizonts (28); and other putative cysteine proteases (32). Falcipain-1, which is expressed throughout the erythrocytic cycle and has been local-

We thank Matthew Bogyo and Doron Greenbaum for providing compound YA29 and for critical reviews of this manuscript; Photini Sinnis for sharing unpublished results; and the Malaria Research and Reference Reagent Resource Center for providing parasite strains D10 (MRA201), 3D7 (MRA-102), HB3 (MRA-155), and W2 (MRA-157), contributed by Y. Wu, D. J. Carucci, T. E. Wellems, and D. E. Kyle, respectively. This work was supported by a grant from the Medicines for Malaria Venture (to P.J.R.) and by National Institutes of Health Grants AI35800 (to P.J.R.), RR01081 (to P.J.R.), and AI41718 (to D.E.G.).

1. Breman, J. G. (2001) Am. J. Trop. Med. Hyg. 64, 1–11. 2. Rosenthal, P. J. & Miller, L. H. (2001) in Antimalarial Chemotherapy: Mechanisms of Action, Resistance, and New Directions in Drug Discovery, ed. Rosenthal, P. J. (Humana, Totowa, NJ), pp. 3–13. 3. Rosenthal, P. J., Sijwali, P. S., Singh, A. & Shenai, B. R. (2002) Curr. Pharm. Des. 8, 1659–1672. 4. Shenai, B. R., Sijwali, P. S., Singh, A. & Rosenthal, P. J. (2000) J. Biol. Chem. 275, 29000–29010. 5. Sijwali, P. S., Shenai, B. R., Gut, J., Singh, A. & Rosenthal, P. J. (2001) Biochem. J. 360, 481–489. 6. Sijwali, P. S. & Rosenthal, P. J. (2004) Proc. Natl. Acad. Sci. USA 101, 4384–4389. 7. Rosenthal, P. J. & Nelson, R. G. (1992) Mol. Biochem. Parasitol. 51, 143–152. 8. Greenbaum, D. C., Baruch, A., Grainger, M., Bozdech, Z., Medzihradszky, K. F., Engel, J., DeRisi, J., Holder, A. A. & Bogyo, M. (2002) Science 298, 2002–2006. 9. Fidock, D. A., Nomura, T., Talley, A. K., Cooper, R. A., Dzekunov, S. M., Ferdig, M. T., Ursos, L. M., Sidhu, A. B., Naude, B., Deitsch, K. W., et al. (2000) Mol. Cell 6, 861–871. 10. Trager, W. & Jensen, J. B. (1976) Science 193, 673–675. 11. Lambros, C. & Vanderberg, J. P. (1979) J. Parasitol. 65, 418–420. 12. Fidock, D. A. & Wellems, T. E. (1997) Proc. Natl. Acad. Sci. USA 94, 10931–10936. 13. Laemmli, U. K. (1970) Nature 227, 680–685. 14. Barkan, D., Ginsburg, H. & Golenser, J. (2000) Int. J. Parasitol. 30, 649–653. 15. Klemba, M., Beatty, W., Gluzman, I. & Goldberg, D. E. (2004) J. Cell Biol. 164, 47–56. 16. Lee, B. J., Singh, A., Chiang, P., Kemp, S. J., Goldman, E. A., Weinhouse, M. I., Vlasuk, G. P. & Rosenthal, P. J. (2003) Antimicrob. Agents Chemother. 47, 3810–3814. 17. Shenai, B. R., Lee, B. J., Alvarez-Hernandez, A., Chong, P. Y., Emal, C. D., Neitz, R. J., Roush, W. R. & Rosenthal, P. J. (2003) Antimicrob. Agents Chemother. 47, 154–160. 18. Dluzewski, A. R., Rangachari, K., Wilson, R. J. & Gratzer, W. B. (1986) Exp. Parasitol. 62, 416–422.

19. Rosenthal, P. J., McKerrow, J. H., Aikawa, M., Nagasawa, H. & Leech, J. H. (1988) J. Clin. Invest. 82, 1560–1566. 20. Hadley, T., Aikawa, M. & Miller, L. H. (1983) Exp. Parasitol. 55, 306–311. 21. Salmon, B. L., Oksman, A. & Goldberg, D. E. (2001) Proc. Natl. Acad. Sci. USA 98, 271–276. 22. Wickham, M. E., Culvenor, J. G. & Cowman, A. F. (2003) J. Biol. Chem. 278, 37658–37663. 23. Francis, S. E., Banerjee, R. & Goldberg, D. E. (1997) J. Biol. Chem. 272, 14961–14968. 24. Banerjee, R., Francis, S. E. & Goldberg, D. E. (2003) Mol. Biochem. Parasitol. 129, 157–165. 25. Dua, M., Raphael, P., Sijwali, P. S., Rosenthal, P. J. & Hanspal, M. (2001) Mol. Biochem. Parasitol. 116, 95–99. 26. Dhawan, S., Dua, M., Chishti, A. H. & Hanspal, M. (2003) J. Biol. Chem. 278, 30180–30186. 27. Salas, F., Fichmann, J., Lee, G. K., Scott, M. D. & Rosenthal, P. J. (1995) Infect. Immun. 63, 2120–2125. 28. Breton, C. B., Blisnick, T., Jouin, H., Barale, J. C., Rabilloud, T., Langsley, G. & Pereira da Silva, L. H. (1992) Proc. Natl. Acad. Sci. USA 89, 9647–9651. 29. Shenai, B. R., Semenov, A. V. & Rosenthal, P. J. (2002) Biol. Chem. 383, 843–847. 30. Rosenthal, P. J., Lee, G. K. & Smith, R. E. (1993) J. Clin. Invest. 91, 1052–1056. 31. Rosenthal, P. J., Olson, J. E., Lee, G. K., Palmer, J. T., Klaus, J. L. & Rasnick, D. (1996) Antimicrob. Agents Chemother. 40, 1600–1603. 32. Wu, Y., Wang, X., Liu, X. & Wang, Y. (2003) Genome Res. 13, 601–616. 33. Brooks, S. R. & Williamson, K. C. (2000) Mol. Biochem. Parasitol. 106, 77–82. 34. Le Roch, K. G., Zhou, Y., Blair, P. L., Grainger, M., Moch, J. K., Haynes, J. D., De La Vega, P., Holder, A. A., Batalov, S., Carucci, D. J., et al. (2003) Science 301, 1503–1508. 35. Florens, L., Washburn, M. P., Raine, J. D., Anthony, R. M., Grainger, M., Haynes, J. D., Moch, J. K., Muster, N., Sacci, J. B., Tabb, D. L., et al. (2002) Nature 419, 520–526.

8726 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0402738101

ized to merozoites (8), also may play a role in erythrocyte rupture and invasion, but our data indicate that it is not required for these processes. Falcipain-1 likely also plays a role in other plasmodial life cycle stages. The processing of gametocyte (33) and sporozoite (P. Sinnis, personal communication) proteins is inhibited by cysteine protease inhibitors, and genomic (34) and proteomic (35) surveys identified falcipain-1 in sporozoites. Our data indicate that falcipain-1 is not essential in erythrocytic-stage malaria parasites. It clearly will be of interest to extend studies with falcipain-1 knockouts to other stages. Regarding the development of new drugs directed against essential processes in erythrocytic parasites, it appears that falcipain-2 and falcipain-3 are the major cysteine proteases involved in hemoglobinolysis, and that among the falcipains, they should be primary targets for drug discovery.

Sijwali et al.