Proteasome Inhibitors Block Development of Plasmodium spp.

Abstract

Proteasomes degrade most of the proteins inside eukaryotic cells, including transcription factors and regulators of cell cycle progression. Here we show that nanomolar concentrations of lactacystin, a specific irreversible inhibitor of the 20S proteasome, inhibit development of the exoerythrocytic and erythrocytic stages of the malaria parasite. Although lactacystin-treated Plasmodium berghei sporozoites are still invasive, their development into exoerythrocytic forms (EEF) is inhibited in vitro and in vivo. Erythrocytic schizogony of P. falciparum in vitro is also profoundly inhibited when drug treatment of the synchronized parasites is prior, but not subsequent, to the initiation of DNA synthesis, suggesting that the inhibitory effect of lactacystin is cell cycle specific. Lactacystin reduces P. berghei parasitemia in rats, but the therapeutic index is very low. Along with other studies showing that lactacystin inhibits stage-specific transformation in Trypanosoma and Entamoeba spp., these findings highlight the potential of proteasome inhibitors as drugs for the treatment of diseases caused by protozoan parasites.

The development of new antimalarial drugs is an urgent priority considering the increasing prevalence of drug-resistant Plasmodium falciparum parasites and the absence of effective vaccines or of vector control measures (3). Malaria infection is initiated when Plasmodium sporozoites are injected into the bloodstream of the host by an infected anopheline mosquito. Shortly after, sporozoites enter hepatocytes where they develop into exoerythrocytic forms (EEF). Each EEF contains thousands of merozoites which rupture from the hepatocyte and invade erythrocytes. In the erythrocytic cycle of P. falciparum, which lasts 48 h, merozoites mature into trophozoites and then into schizonts. A mature schizont contains between 8 and 26 merozoites, each of which is capable of infecting a new erythrocyte.

During both its hepatic and erythrocytic stages the parasite undergoes radical morphological changes and many rounds of replication, events that likely require proteasome activity. Proteasomes are major components of the eukaryotic cellular machinery (5, 18, 23, 28), mediating the normal turnover of proteins and the degradation of proteins that have been improperly folded or denatured (29). In addition to these housekeeping functions, proteasomes play a key role in cell cycle progression (20) and the regulation of numerous transcription factors (24).

Studies of proteasome function have been facilitated by the availability of lactacystin, a highly specific inhibitor of proteasome proteolytic activity. Lactacystin is a Streptomyces metabolite whose active form binds irreversibly to the catalytic threonines in the active sites of the β subunits of the proteasome (6, 9, 12, 16). Previous studies have shown that lactacystin affects the stage-specific transformation of Trypanosoma cruzi trypomastigotes into amastigotes (15) and the encystation of Entamoeba invadens (14). Here we use lactacystin to study the role of proteasomes in the life cycle of malaria parasites.

MATERIALS AND METHODS

Drugs.

Lactacystin and lactacystin analogs were synthesized as previously described (4, 13, 22). 7-Ethyl lactacystin and des-7-methyl lactacystin were synthesized in the Harvard laboratory. All drugs, except clasto-lactacystin β-lactone, were dissolved in H₂O to 1 mM and stored at 4°C until use. clasto-lactacystin β-lactone was solubilized in dimethyl sulfoxide to 10 mM and stored at −20°C until use. Lactacystin for injection into rats was dissolved in phosphate-buffered saline (PBS), pH 7.4, immediately before use.

Assay for EEF development in vitro.

This assay was performed as described previously (19) with a few modifications. Briefly, HepG2 cells (ATCC HB8065; American Type Culture Collection, Manassas, Va.) were plated in chamber slides (model 4808; Lab-tek, Naperville, Ill.) 48 h before each experiment. Plasmodium berghei sporozoites were dissected from mosquito salivary glands and resuspended in Dulbecco modified Eagle medium (Gibco BRL, Gaithersburg, Md.) supplemented with 10% fetal calf serum (HyClone Laboratories, Logan, Utah) and 20 mM HEPES (Sigma, St. Louis, Mo.). Approximately 50,000 sporozoites were added per well, and the parasites were allowed to adhere to and invade the HepG2 cells for 2 h. The wells were washed, and the cells were grown for an additional 2 days after which they were fixed with methanol. The EEF were then revealed with monoclonal antibody (MAb) 2E6 (34) followed by goat anti-mouse immunoglobulin (Ig) conjugated to horseradish peroxidase (Accurate Chemical Corp., Westbury, N.Y.) and 3,3′-diaminobenzidine (Sigma). The EEF in each well were counted microscopically with a 20× light microscope objective.

Microscopic assay for quantification of sporozoite invasion.

This assay was conducted according to the method described by Renia et al. (27) with a few modifications. HepG2 cells were plated in chamber slides as described above. P. berghei sporozoites were pretreated with 3 μM lactacystin in Dulbecco modified Eagle medium-fetal calf serum for 1 h at room temperature, washed, and then added to the HepG2 cells. Controls were pretreated with medium alone. The parasites were incubated with the HepG2 cells for 1 h at 37°C in 5% CO₂. The unattached sporozoites and medium were then removed, and the cells were fixed with 4% paraformaldehyde. The extracellular parasites were revealed by incubation with MAb 3D11 followed by goat anti-mouse Ig conjugated to rhodamine (Boehringer Mannheim, Indianapolis, Ind.). The HepG2 cells were then permeabilized with methanol, and all parasites (intra- and extracellular) were revealed with MAb 3D11 followed by goat anti-mouse Ig conjugated to fluorescein isothiocyanate (FITC; Boehringer Mannheim). MAb 3D11 binds to the repeats of the P. berghei circumsporozoite protein (38), found on both sporozoites and EEF (32). The slides were mounted, and each field was counted with two different UV filters so that both FITC-labeled and rhodamine-labeled sporozoites could be counted. Between 40 and 50 fields were counted per well. The percent invasion for each well was calculated from the following equation: [(total number of parasites − number of extracellular parasites)/total number of parasites] × 100 = % invasion, where the total number of parasites is the number of FITC-labeled sporozoites and the number of extracellular parasites is the number of rhodamine-labeled sporozoites.

Assessment of C-type rRNA switching to A-type rRNA.

HepG2 cells (2.5 × 10⁵ cells/well) were plated in 24-well plates (Falcon; Becton Dickinson, Franklin Lakes, N.J.) and allowed to grow for 2 days. P. berghei sporozoites were incubated with 3 μM lactacystin or without lactacystin for 15 min at room temperature, and then 20,000 sporozoites were added to each well. After 2 h the medium was removed and fresh medium without inhibitor was added. At 5 and 21 h after infection, the cells from each well were trypsinized, spun at 300 × g, and resuspended in 1 ml of Tri-Reagent (Sigma) and total cellular RNA was extracted according to the manufacturer’s instructions. Reverse transcriptase PCR (RT-PCR) was performed with an RT-PCR kit (Perkin-Elmer, Branchburg, N.J.). Total RNA was quantified by measuring the absorbance at 260 nm, and RT reactions were performed with 0.1 μg of RNA and random hexamers supplied by the manufacturer. PCR for this cDNA was performed with primers specific for either C- or A-type rRNA. These primers were designed based on published sequences (17) and included a 5′ primer common to both types of rRNA (5′-GCCTGAGAAATAGCTACCACATC-3′) and a 3′ primer specific for either A-type rRNA (5′-CATGAAGATATCGAGGCGGAG-3′) or C-type rRNA (5′-GGATAAAAGCAGTGACAGAAGTC-3′). Relative amounts of C- and A-type rRNA in each starting sample were estimated by performing PCR with serial dilutions of the cDNA.

Culture of erythrocytic stages.

P. falciparum 3D7 erythrocytic stages were cultured by standard methods (25, 33) except that the culture medium contained 0.5% Albumax I (Gibco) in place of human serum.

[³H]hypoxanthine uptake assay.

Parasites were synchronized with 5% sorbitol (Sigma) (21) by two treatments, 30 h apart, resulting in approximately 90% synchrony. Parasites were used 18 h after the second treatment. In the standard assay (7), [³H]hypoxanthine (Amersham, Arlington Heights, Ill.), 0.5 μCi/well, and drugs at the concentrations indicated in Fig. 5 were added at the time of plating. For the time course assay, the [³H]hypoxanthine, with or without 0.6 μM lactacystin, was added at the time points indicated in Fig. 6. Two sets of negative controls were included in the experiments; one contained uninfected erythrocytes and label, and the other contained label and medium alone. Plates were incubated at 37°C for 24 h and harvested onto glass fiber filters (Wallac Oy, Turku, Finland) with a 1295-001 cell harvester (Wallac Oy), and the filters were counted in a 1205 Betaplate (Wallac Oy) liquid scintillation counter. All treatments were performed in triplicate wells.

FIG. 5.

Lactacystin analogs inhibit P. falciparum erythrocytic stages and isolated human proteasomes similarly. (a) Chemical structures of lactacystin and the analogs studied. (b and c) Synchronized trophozoites at 18 h of the erythrocytic cycle were plated in 96-well microtiter plates with [³H]hypoxanthine and the concentrations of inhibitors indicated. After 24 h, cells were harvested and incorporation of the label was measured. Shown are the means of triplicate wells ± standard deviations. (d) Proteasomes isolated from normal human erythrocytes, plus inhibitors at the indicated concentrations, were incubated with fluorogenic substrate N-succinyl-Leu-Leu-Val-Tyr-7-amido-4-methyl-coumarin to measure chymotrypsin-like activity. Each point is the mean of the fluorescences of duplicate wells ± the standard deviation. Symbols (panels b, c, and d): solid circles, lactacystin; open circles, clasto-lactacystin β-lactone; open upright triangles, β-acetylaminoethyl lactacystin; open diamonds, de-N-acetyl lactacystin; solid triangles, clasto-lactacystin dihydroxy acid; open hexagons, 7-ethyl lactacystin; open inverted triangles, des-7-methyl lactacystin; open squares, methyl ester lactacystin.

FIG. 6.

Lactacystin inhibits DNA synthesis in a stage-specific manner. Synchronized trophozoites were plated in 96-well microtiter plates at 18 h into the erythrocytic cycle. [³H]hypoxanthine with (open circles) or without (closed circles) lactacystin was added to wells at the times indicated. All cells were harvested at 48 h in the cycle, and incorporation of the label was measured. Shown are the means of triplicate wells with standard deviations.

Parasite growth in erythrocytes pretreated with lactacystin.

Three milliliters of packed, washed human erythrocytes was resuspended at a 50% hematocrit in RPMI 1640 and incubated with 10 μM lactacystin or without lactacystin for 1 h at 37°C. Cells were washed three times in 10 volumes of RPMI 1640 at 37°C for 2 h per wash. Untreated schizonts were concentrated to a parasitemia of ∼80% (10) and added to control and lactacystin-treated target cells, so that the starting parasitemias were ∼0.1%. Each treatment was performed in triplicate. Parasitemias were measured daily by blind counting of the number of infected erythrocytes per 2,000 cells on Giemsa-stained blood smears from each flask.

Erythrocyte proteasome isolation.

Lactacystin-treated and control erythrocytes were treated and washed as for the growth assay described above and then were washed once in 10 volumes of ice-cold 10 mM Tris–150 mM NaCl, pH 7.5. The cells were then lysed in 6 ml of ice-cold 10 mM Tris, pH 7.5, and incubated on ice for 5 min. Three milliliters of 10 mM Tris–800 mM NaCl, pH 7.5, was added to each tube before ultracentrifugation at 10,000 × g for 30 min at 4°C with a Beckman SW-41 rotor and an L8-80 ultracentrifuge. The supernatants were loaded onto a 1-ml HiTrap Q (Pharmacia) column for anion exchange fast protein liquid chromatography. Samples were eluted with an NaCl gradient from 200 mM to 1 M in 10 mM Tris, pH 7.5, and 1.2-ml fractions were collected on ice.

Enzymatic assay for proteasome activity.

The chymotrypsin-like activities of HiTrap Q fractions were measured with the fluorescent substrate N-succinyl-Leu-Leu-Val-Tyr-7-amido-4-methyl-coumarin (Sigma), as described by Gonzalez et al. (15).

Quantitative PCR assay for sporozoite infectivity.

Plasmodium yoelii sporozoites were incubated with or without 5 μM lactacystin for 1 h at room temperature and injected intravenously (i.v.) into Swiss Webster mice. Two thousand sporozoites were injected into each mouse, and 40 h later the mice were sacrificed and their livers were removed. Total RNA extraction from livers and RT-PCR were performed as described by Briones et al. (2) with 1 μg of RNA. PCR analysis of this cDNA was performed with parasite rRNA primers that recognize P. yoelii-specific sequences within the A-type and C-type 18S rRNA. These reactions were performed in the presence of a competitor template, constructed by insertion of a 66-bp DNA fragment into the cloned 393-bp rRNA parasite amplification product. Mouse hypoxanthine phosphoribosyltransferase (HPRT) primers and competitor were used as positive controls to assess the efficiency of RT reactions as described previously (26).

Assessment of the effects of lactacystin on malaria infection in rats.

Groups of six Sprague-Dawley rats (Taconic, Germantown, N.Y.), each weighing approximately 60 g, were infected with P. berghei blood stages and then monitored by taking blood smears. Treatment with lactacystin started when the average parasitemia was approximately 1%. The rats were then distributed into experimental and control groups, such that the parasitemias of the groups were comparable. Experimental groups were injected with lactacystin diluted in PBS, while control groups were injected with PBS alone. Blood smears were counted blindly at the indicated times after treatment. Data were analyzed by a repeated-measure analysis of variance (one-tailed test) with SAS software (30). In other experiments, P. yoelii sporozoites were incubated with 3 μM lactacystin or without lactacystin for 1 h at room temperature and then injected into mice i.v. at the doses indicated. Blood smears were taken from the mice starting at day 3 and assessed for the presence of parasites as described above.

RESULTS

Lactacystin inhibits the exoerythrocytic development of P. berghei in vitro.

Table 1 shows that P. berghei sporozoites treated with lactacystin did not develop normally into EEF. This inhibition occurred whether the sporozoites were plated onto HepG2 cells in the presence of the drug (Table 1, experiment 1) or preincubated with the drug and then washed before being plated onto target cells (Table 1, experiment 2). Preincubation with clasto-lactacystin dihydroxy acid, the inactive product of lactacystin hydrolysis (9), had no effect. Lactacystin treatment performed 24 h after the addition of parasites to the cells (Table 1, experiment 3) still resulted in a significant inhibition of EEF development, and those EEFs that developed were smaller than normal.

TABLE 1.

Lactacystin inhibits development of sporozoites into EEF

Expt	Inhibitor	Concn (μM)	No. of EEFa	% Inhibition
1b	None		448 ± 25
	Lactacystin	1	109 ± 10	75
		3	5 ± 2	99
		9	0	100
2c	None		291 ± 5
	Lactacystin	1	156 ± 6	46
		3	34 ± 5	88
		9	0	100
	clasto-Lactacystin  dihydroxy acid	9	280 ± 32	6
3d	None		816 ± 9
	Lactacystin	1	574 ± 13	29
		3	458 ± 17	43
		9	408 ± 3	50

^aNumber of EEF per 100 high-powered fields. 

^bInhibitor present for the first 3 h of the experiment. 

^cParasites were preincubated with inhibitor for 1 h, washed, and then plated on cells. 

^dInhibitor was added 24 h after parasites were plated on cells. The EEF in the lactacystin-treated wells were smaller than those in the control wells. 

We then tested whether the reduction in the number of EEF with lactacystin treatment was due to an effect of the drug on sporozoite invasion of target cells. Sporozoites were preincubated with lactacystin and then added to HepG2 cells. After 1 h, the cells were fixed and a double-staining immunofluorescence assay was used to distinguish between intracellular and extracellular parasites in order to calculate invasion rates. There was no difference between the invasion rate for control sporozoites and those for sporozoites preincubated with up to 9 μM lactacystin for 1 h (Table 2). Invasion of host cells by malaria and other Apicomplexa parasites is an active process (11, 35, 36). Thus, the observed inhibition of EEF development by lactacystin was not due to a lethal effect of the drug on sporozoites.

TABLE 2.

Lactacystin does not affect sporozoite invasion

Lactacystin concn (μM)a	% Invasionb
	Expt 1	Expt 2	Expt 3
0	44	26	27
0	35	28	29
1	46		33
1	51		30
3	50	27	33
3	45	28	33
9			30
9			27

^aSporozoites were incubated with or without lactacystin for 1 h before addition to HepG2 cells. 

^bCalculated based on the number of intracellular parasites in 40 to 50 fields/slide, from the double-staining immunofluorescence assay, 1 h after addition to HepG2 cells. 

Since lactacystin did not inhibit cell invasion by sporozoites, we performed experiments to test whether development of the parasite was affected by the drug. Characteristic morphologic changes accompany the development of EEF from sporozoites. Within 4 h after invasion, the middle of the sporozoite expands into a characteristic bulb-like structure (Fig. 1a) (1). After 15 h the parasite has a spherical shape (Fig. 1c). If the sporozoites were treated with lactacystin, none of the parasites had this bulb-like structure at 4 h after invasion (Fig. 1b). At 15 h, approximately half of the parasites remained as slender sporozoites inside the cell (Fig. 1d). The other half consisted of a mixture of round parasites with normal morphology and parasites with a pyknotic appearance (Fig. 1e).

FIG. 1.

Lactacystin alters the normal development of sporozoites into EEF in vitro. P. berghei sporozoites were incubated in 3 μM lactacystin or medium alone for 15 min and then added to HepG2 cells and allowed to invade and begin their development into EEF. Four and fifteen hours later the cells were fixed and stained with MAb 3D11 by using the double-staining assay, which allows the distinction between intracellular and extracellular sporozoites. Magnification, ×100.

We also examined the effect of lactacystin on another developmental marker, namely, the switch in rRNA expression from C-type sporozoites to A-type EEF. Lactacystin-treated and control sporozoites were plated on HepG2 cells, and 5 and 21 h later the cells were harvested for quantitative RT-PCR using C- and A-type-specific rRNA primers. At 5 h, there was little A-type rRNA in either the lactacystin-treated or control sporozoites (Fig. 2a). However, at 21 h only control sporozoites showed an increase in the amount of A-type rRNA (Fig. 2b). During this time there was no apparent change in the amounts of C-type rRNA in lactacystin-treated and control sporozoites (data not shown).

FIG. 2.

Lactacystin inhibits the switch to A-type rRNA of P. berghei in vitro. Sporozoites were incubated with 3 μM lactacystin or without lactacystin for 15 min and then plated on HepG2 cells. After 2 h, the medium was removed and fresh medium without inhibitor was added. At 5 (a) and 21 h (b), total RNA was extracted and RT reactions were performed with 0.1 μg of RNA. Quantitative PCR of the cDNA was performed with primers specific for A-type rRNA and serial dilutions of cDNA. The first, second, and third (panel b only) lanes in each panel show the results of PCR performed with 2, 0.4, and 0.2 μl, respectively, of cDNA.

Effects of lactacystin on P. yoelii EEF development in vivo.

Sporozoites of P. yoelii are, for unknown reasons, significantly more infectious to mice than those of P. berghei (2) and were therefore used for in vivo experiments. P. yoelii sporozoites were preincubated in medium with or without 3 μM lactacystin and then injected into mice. The mice injected with lactacystin-treated sporozoites showed an increase in the prepatent period versus controls (Table 3). As shown, injection of 10,000 or 1,000 lactacystin-treated sporozoites results in the same prepatent period as the injection of 100 control sporozoites, suggesting that the drug inhibited development by 90 to 99% (31).

TABLE 3.

Lactacystin inhibits sporozoite infectivity in vivo

No. of sporozoites injected	Inhibitora	No. of mice positive for parasites/no. of mice injectedb on day:
		3	4	5
100,000	None	1/1
	Lactacystin	2/2
10,000	None	2/2
	Lactacystin	0/3	3/3
1,000	None	3/3
	Lactacystin	0/3	3/3
100	None	0/3	3/3
	Lactacystin	0/3	0/3	2/3c

^aSporozoites were preincubated in medium with 3 μM lactacystin or without lactacystin for 1 h and then injected into mice i.v. 

^bOn days 3 to 5 postinjection blood smears were taken from mice and analyzed for the presence of parasites. 

^cThe mouse that was negative on day 5 remained negative for the next 7 days.  

A more direct quantification of sporozoite infectivity was performed by means of a competitive RT-PCR assay. When the amounts of rRNA in the livers of mice injected with control and lactacystin-treated sporozoites were compared, we found approximately a 10-fold decrease in the parasite rRNA in mice injected with treated sporozoites (Fig. 3).

FIG. 3.

Lactacystin decreases sporozoite infectivity in vivo. P. yoelii sporozoites were incubated in 5 μM lactacystin or medium alone for 1 h. Two thousand sporozoites were then injected i.v. into each mouse, and 40 h later the mouse livers were harvested for isolation of RNA. Sporozoite infectivity was quantified by measuring the amount of parasite rRNA in a quantitative RT-PCR assay. The top two panels show PCRs performed with P. yoelii rRNA primers and 1 and 0.1 pg of a P. yoelii rRNA competitor. The parasite target band is 393 bp, and that of the competitor is 459 bp. The bottom panel shows control PCRs performed with the same RT reaction mixtures containing HPRT primers and 0.04 pg of an HPRT competitor; the HPRT target band is 352 bp, and that of the competitor is 450 bp. M = markers (1,000, 750, 500, 300, and 150 bp).

Lactacystin inhibits the growth of P. falciparum erythrocytic stages in vitro.

Figure 4 shows that lactacystin inhibited erythrocytic schizogony, as observed by light microscopy. Normal trophozoites have a single nucleus (Fig. 4, top left) that divides a variable number of times to produce the 8 to 26 nuclei contained in the mature schizont (Fig. 4, top right). Approximately 90% of the parasites appeared developmentally arrested when treated with 1.25 μM lactacystin (Fig. 4, bottom left) and persisted for at least 24 h with a morphology that was indistinguishable by light microscopy from that of normal trophozoites. At higher concentrations, i.e., 10 μM, however, many of the parasites showed degenerative changes (Fig. 4, bottom right).

FIG. 4.

Lactacystin inhibits the development of P. falciparum erythrocytic stages in vitro. Photomicrographs were taken of synchronized trophozoites at 18 h (top left), after which the trophozoites were incubated for another 24 h in either medium alone (top right) or 1.25 (bottom left) or 10 μM (bottom right) lactacystin.

To exclude the possibility that any of the effects of lactacystin on erythrocytic stage development were due to inhibition of the erythrocyte’s, rather than the parasite’s, proteasomes, uninfected erythrocytes were treated with 10 μM lactacystin and washed extensively. By this method, the chymotrypsin-like activity of proteasomes isolated from the treated erythrocytes was totally inhibited (data not shown). Nevertheless, parasites grew normally in the lactacystin-treated erythrocytes (data not shown).

Lactacystin and several lactacystin analogs were also shown to inhibit the nucleic acid synthesis that occurs during erythrocytic schizogony, as measured by incorporation of [³H]hypoxanthine (Fig. 5). Lactacystin spontaneously undergoes lactonization to form clasto-lactacystin β-lactone, the sole intermediate and active form of the drug, which is then hydrolyzed to become the inactive clasto-lactacystin dihydroxy acid (Fig. 5a) (9). clasto-Lactacystin β-lactone binds irreversibly to the amino-terminal threonine of the β subunits, in the active sites of the proteasome (9, 12). Significant inhibition occurred at nanomolar concentrations of lactacystin, with approximately 50% inhibition seen at 300 nM (Fig. 5b). As shown in Fig. 5b and c, clasto-lactacystin β-lactone and the methyl ester analog displayed inhibitory activities against parasite schizogony identical to that of lactacystin on a molar basis, but clasto-lactacystin dihydroxy acid was inactive. The 7-ethyl and β-acetylaminoethyl analogs were slightly more active than lactacystin, while the activities of the des-7-methyl and the de-N-acetyl analogs were greatly reduced.

When tested on proteasomes isolated from human erythrocytes, several of these lactacystin analogs showed inhibitory activities relative to lactacystin that paralleled their activities against Plasmodium schizogony (Fig. 5d). Specifically, the inhibitory activity of 7-ethyl-lactacystin was increased compared to that of lactacystin, while that of des-7-methyl lactacystin was greatly reduced. However, the β-acetylaminoethyl analog had activity which was identical to that of lactacystin in this assay, in contrast to its greater potency in the inhibition of [³H]hypoxanthine uptake by the parasites.

We then tested the effect of lactacystin on the parasite at different stages of the erythrocytic cycle. Synchronized parasites at 18 h of the cycle were plated into triplicate wells and [³H]hypoxanthine, with or without lactacystin, was added at successive time points. All cells were harvested at the end of the 48-h cycle, and incorporated radioactivity was measured. The value for each time point, therefore, represents the amount of nucleic acid synthesis that occurred from that time until the end of the erythrocytic cycle. Figure 6 shows that when lactacystin was added during the first 30 h of the parasite’s cycle, [³H]hypoxanthine incorporation was inhibited. In contrast, inhibition of [³H]hypoxanthine incorporation was no longer seen when parasites were treated later, after 30 h and throughout schizogony.

Effects of lactacystin on P. berghei erythrocytic stages in vivo.

Figure 7 shows that, for rats infected with P. berghei erythrocytic stages in erythrocytic stages, treatment with one dose of 1.6 mg of lactacystin resulted in a significant (P = 0.05) reduction of parasitemia in comparison with that of controls, as calculated by a one-tailed repeated-measure analysis of variance. Treatment with a total of 4 mg of lactacystin per rat, given as three i.v. injections of 1.3 mg each, 8 h apart, cleared infection (data not shown). However, none of the five rats treated survived this regimen.

FIG. 7.

Lactacystin significantly reduces parasitemia in vivo. Six P. berghei-infected rats were paired into two groups of three rats; all rats had comparable parasitemias. Each rat in the experimental group (open circles) received 1.6 mg of lactacystin in 1 ml of PBS, given as one intraperitoneal injection of 0.5 ml and one i.v. injection of 0.5 ml at the same time. Each rat in the control group (solid circles) received identical injections of PBS alone. Giemsa-stained blood smears, taken at the time points indicated, were blindly counted to measure parasitemias. Each point represents the mean of parasitemias from three rats ± the standard deviation.

DISCUSSION

We show here that lactacystin, a proteasome inhibitor, blocks the development of the preerythrocytic and erythrocytic stages of Plasmodium spp. Lactacystin covalently modifies the catalytic N-terminal threonines of the active sites of proteasomes, inhibiting the activities of all proteasomes examined including those of mammalian cells (6, 9, 12, 16), protozoa (14, 15), and archaea (23). It is therefore highly unlikely that proteasomes of Plasmodium are not inhibited by lactacystin. In addition, lactacystin is an exquisitely specific drug: it does not inhibit any other known proteases (12). When [³H]lactacystin was incubated with whole-mammalian-cell extracts or with crude brain extracts, radioactivity was associated exclusively with proteasome subunits (12). For these reasons, we chose lactacystin, and analogs thereof, to examine the role of proteasomes in the development of Plasmodium.

Lactacystin-treated sporozoites, although invasive in vitro, did not round up normally and acquire the characteristic EEF morphology. This finding was supported by using a molecular marker for EEF development: the switch from C-type to A-type rRNA expression. Plasmodium spp. are unique in that they express different rRNAs in different stages of their life cycles (37). Although small amounts of all rRNA types can always be found in each stage, the vast majority of rRNA in sporozoites is C-type rRNA. After the parasites invade hepatocytes, they begin to synthesize A-type rRNA, and by 20 h after invasion this is the predominant rRNA associated with the parasites. When sporozoites were treated with lactacystin, no increase in A-type rRNA could be detected at 20 h. In addition, sporozoite development in vivo appeared to be inhibited by lactacystin, as treated sporozoites were at least 10-fold less infective in mice.

Starting between 30 and 40 h after erythrocyte invasion, trophozoites go through several rounds of DNA replication and nuclear division as they develop into merozoite-containing schizonts. Lactacystin-treated trophozoites, however, did not transform into schizonts. They maintained an arrested but apparently normal morphology for extended periods (Fig. 4, lower left), similar to what was observed with lactacystin-treated sporozoites within HepG2 cells (Fig. 1d). The addition of lactacystin up to 30 h after erythrocyte infection strongly inhibited [³H]hypoxanthine incorporation, but the drug had no effect after the beginning of schizogony (Fig. 6). Thus, only the initiation of DNA synthesis was prevented by lactacystin, and not DNA synthesis per se.

A trivial explanation for the lack of inhibitory effect of lactacystin treatment during schizogony might be that the proteasomes of late stages of the parasite are inaccessible to the drug. This is unlikely, however, since the lactacystin-treated schizonts were also developmentally arrested: although they appeared normal by light microscopy and incorporated [³H]hypoxanthine normally, they did not rupture (data not shown). The mechanisms by which lactacystin inhibits Plasmodium development are not known, but an attractive possibility is that the drug affects the control of cell cycle progression in the parasite. Proteasome activity is required for transition through the G₁/S boundary and for exit from M phase in a number of cell types (20). Cell cycle control in Plasmodium is poorly understood. However, the inhibition of the onset of DNA synthesis and the lack of schizont rupture seen with lactacystin treatment may be due to similar requirements for proteasome activity in the cell cycle progression of Plasmodium.

In an attempt to find a potent parasiticidal drug, we tested several analogs of lactacystin. As shown by the [³H]hypoxanthine uptake assay, two analogs were consistently more potent inhibitors than clasto-lactacystin β-lactone in this assay: 7-ethyl lactacystin and β-acetylaminoethyl lactacystin (Fig. 5b and c). An analysis of the structure of the yeast proteasome cocrystalized with lactacystin revealed that, in addition to the presence of several hydrogen bonds between the drug and the β5/PRE2 subunit, the isopropyl group on C-10 of lactacystin is inserted into the S1 (“specificity”) pocket of the enzyme active site (16). 7-Ethyl lactacystin has one more carbon on the γ-lactam ring than lactacystin, i.e., it has an ethyl group instead of a methyl group at C-7. Since des-7-methyl lactacystin lacks the methyl group at C-7 (Fig. 5a) and shows greatly decreased activity (Fig. 5b), our findings highlight the importance of C-7 side chains for acylation of the proteasome.

The other analog with increased activity, β-acetylaminoethyl lactacystin, is modified only on the N-acetylcysteine moiety. This is unexpected since the N-acetylcysteine moiety is lost during lactonization into the active compound, clasto-lactacystin β-lactone. The reasons for the increase in activity are therefore not clear. It is thought that cells are impermeable to lactacystin and that only the β-lactone enters cells (8). One possibility is that the β-acetylaminoethyl analog can enter cells prior to lactonization. The increased hydrophobicity resulting from the removal of the carboxyl group of lactacystin to form β-acetylaminoethyl lactacystin might facilitate passage through the plasma membranes of cells.

When tested on proteasomes isolated from human erythrocytes, several analogs showed activities relative to that of lactacystin that paralleled those found in the P. falciparum [³H]hypoxanthine uptake assay (Fig. 5d). The only exception was the β-acetylaminoethyl analog, which as expected had the same potency as lactacystin on isolated proteasomes (see above). The observation that most lactacystin analogs do not discriminate between the mammalian and parasite proteasomes suggests that the active sites of these enzymes are similar. In agreement with these in vitro observations, lactacystin effectively inhibited parasite growth in vivo, but its therapeutic index precludes clinical usefulness. Several other classes of potent proteasome inhibitors exist (reviewed in reference 5), and it is hoped that those currently being developed will be more selective for the parasite proteasome. The identification of drugs that can exploit differences between the parasite and host proteasomes should be facilitated by the isolation and structural characterization of the proteasome of Plasmodium.