Effects of interruption of apicoplast function on malaria infection, development, and transmission

Molecular and Biochemical Parasitology 109 (2000) 17–23

Effects of interruption of apicoplast function on malariainfection, development, and transmission

Margery Sullivan a, Jun Li a, Sanjai Kumar b, M. John Rogers a,Thomas F. McCutchan a,*

a Growth and De6elopment Section, Laboratory of Parasitic Diseases, National Institute of Allergy and Infectious Diseases,National Institutes of Health, Bldg. 4, Rm. B1-28, Bethesda, MD 20892-0425, USA

b Malaria Program, Na6al Medical Research Center, Bethesda, MD 20814-5055, USA

Received 18 November 1999; received in revised form 3 February 2000; accepted 4 March 2000

Abstract

A chloroplast-like organelle is present in many species of the Apicomplexa phylum. We have previouslydemonstrated that the plastid organelle of Plasmodium falciparum is essential to the survival of the blood-stagemalaria parasite in culture. One known function of the plastid organelle in another Apicomplexan, Toxoplasmagondii, involves the formation of the parasitophorous vacuole. The effects of interruption of plastid function onsporozoites and sexual-stage parasites have not been investigated. In our previous studies of the effects ofthiostrepton, a polypeptide antibiotic from streptococcus spp., on erythrocytic schizongony of the human malaria P.falciparum, we found that this antibiotic appears to interact with the guanosine triphosphatase (GTPase) bindingdomain of the organellar large subunit ribosomal RNA, as it does in bacteria. We investigate here the effects of thisdrug on life-cycle stages of the malaria parasite in vivo. Preincubation of mature infective sporozoites withthiostrepton has no observable effect on their infectivity. Sporozoite infection both by mosquito bite and sporozoiteinjection was prevented by pretreatment of mice with thiostrepton. Thiostrepton eliminates infection with erythrocyticforms of Plasmodium berghei in mice. Clearance of infected red blood cells follows the delayed kinetics associated withdrugs that interact with the apicoplast. Thiostrepton treatment of infected mice reduces transmission of parasites bymore than ten-fold, indicating that the plastid has a role in sexual development of the parasite. These results indicatethat the plastid function is accessible to drug action in vivo and important to the development of both sexual andasexual forms of the parasite. © 2000 Elsevier Science B.V. All rights reserved.

Keywords: Anti-malarial drugs; Guanosine triphosphatase site; Malaria transmission; Plastid; Thiopeptides; Thiostrepton

www.elsevier.com/locate/parasitology

1. Introduction

A number of parasites from the phylumApicomplexa contain an unusual organelle referredto as an apicoplast [1,2]. Evolutionarily, the

Abbre6iations: GTPase, guanosine triphosphatase.* Corresponding author. Tel.: +1-301-4966149; fax: 1-301-

4020079.E-mail address: [email protected] (T.F. Mc-

Cutchan)

0166-6851/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved.

PII: S0166-6851(00)00226-7

M. Sulli6an et al. / Molecular and Biochemical Parasitology 109 (2000) 17–2318

organelle appears to have arisen as a result ofsecondary endosymbiosis and is related mostclosely to the green alga organelle [3]. Inhibitionof plastid function leads to the death of theparasite both in Toxoplasma tachyzoites [4] and inPlasmodium falciparum asexual blood forms [5].

Drugs associated with the treatment ofprokaryotic organisms, such as clindamycin andmacrolipid antibiotics, have been proposed to se-lectively target the plastid [4–6]. The kinetics ofparasite clearance reveal a characteristic delayedpattern of death, unlike the more immediate ef-fects seen with such drugs as chloroquine orpyrimethamine. Investigation of the effects ofdrugs that interact specifically with the plastidprovides both a promising approach to the dis-covery of antiparasitic agents and the toolsneeded to understand plastid function.

2. Materials and methods

2.1. Drugs and preparation

Thiostrepton (500 mg, 1530 U mg−1; Cal-biochem, La Jolla, CA) was dissolved in 1 mlN,N-dimethylacetamide (Sigma, St. Louis, MO)and gradually added to 4 ml of normal salinesolution and mixed completely. Diphosphate pri-maquine and diphosphate chloroquine (Sigma)were dissolved directly into normal saline. All thedrug solutions were sonicated at room tempera-ture for 30 min prior to serial dilution.

2.2. Preincubation of sporozoites with drug

Sporozoites prepared as described below wereisolated and then incubated for 1–2 h with eachdrug (thiostrepton, chloroquine, and primaquine).The groups, of four mice each, were intravenouslyinoculated with 500 sporozoites in 0.2 ml RPMI.Fifty sporozoites consistently result in 100% infec-tion in the controls. Mice were then observed byblood smear every other day.

2.3. Drug treatment of mice prior to sporozoiteinfection

Infection by sporozoites was also investigatedafter pretreatment of mice with drugs. Groups offour mice each were treated with drug for the 2days before infection as well as at 3 and 24 h aftersporozoite challenge. Infection was done eitherdirectly by subjecting mice to the bite of infectedmosquitoes or intravenous injection as describedabove. Drug treatment was terminated after day 1and the mice were then observed by blood smearevery other day [7]. All untreated mice becameinfected (Table 1).

In the intravenous procedure, the sporozoiteswere isolated from the salivary glands of Anophe-les stephensi mosquitoes that had been fed 21 daysearlier on infected mice. The salivary glands weredissected into complete RPMI, triturated in aglass grinder, and the released sporozoitescounted in a hemocytometer. Sporozoites werediluted in RPMI medium to a final concentrationof 500 per 0.2 ml. After injection, mice wereobserved by blood smear every other day. Alluntreated mice became infected.

Table 1Prophylactic activity of thiostrepton on mice bitten by mosquitoes infected with P. berghei

ParasitemiaCompounds Dose per mg kg−1 i.p.

Day 14Day 10Day 8Day 6

0/70/70/70/7500 mgThiostrepton50 mg 0/8Chloroquine 3/8 4/8 4/8

Primaquine 0/80/80/850 mg 0/810% DMAa 1/8 5/8 7/8 8/8Control

a DMA, N,N-dimethylacetamide.

M. Sulli6an et al. / Molecular and Biochemical Parasitology 109 (2000) 17–23 19

Table 2Dose-related suppressive effects of thiostrepton on mice intravenously infected with P. berghei blood stages

Dose per mg kg−1 i.p.Compounds Parasitemia

Day 4 Day 6 Day 8 Day 10

Thiostrepton 500 mg 8/8 3/8 0/7 0/78/8 5/8 3/8 0/8250 mg8/8 8/8125 mg 8/8 4/8

6.25 mg 8/8 8/8 8/8 8/80/8 0/8 0/8 0/8Chloroquine 15 mg8/8 8/8 8/810% DMAa 8/8Control


Table 3Prophylactic activity of thiostrepton on mice intravenously injected with P. berghei sporozoites

Dose per mg kg−1 i.p.Compounds Parasitemia

Day 6 Day 8 Day 10 Day 14

Thiostrepton 500 mg 0/4 0/4 0/4 0/40/4 0/4 0/4Primaquine 0/450 mg0/4 4/4 4/410% DMAa 4/4Control


2.4. Treatment of malaria infected mice

The 4-day suppressive assay was used to testthe effect of thiostrepton on erythrocytic stages ofthe rodent malaria, Plasmodium berghei, ANKAstrain [7]. Parasites from a donor mouse with10–20% parasitemia, were collected into a hep-arin solution. The infected blood was diluted withsaline to a final concentration of 106 per 0.2 mlblood suspension. On the first day (D0), groups ofeight female BALB/c mice, age 6–8 weeks andweighing 2092 g, were intravenously inoculatedwith 106 infected erythrocytes by the tail vein.Infection was followed by an intraperitoneal injec-tion of anti-malarial drug. Drug treatment wasalso administered on the 3 days following infec-tion (D+1, D+2, and D+3). On D+4, bloodsmears were taken, and anti-malarial activity wasassessed by microscopy (Table 2).

2.5. Mosquito transmission

Infected mice were started on a 4-day regime of

drug treatment when the parasitemia reached 1%.When parasitemia reached 6% in control animals,mosquitoes were allowed to feed on groups ofmice that had received the different drug treat-ments. Fourteen days after blood feed,mosquitoes were dissected. The presence of eggswas used as an indication that a mosquito hadtaken a normal blood feed. Mosquitoes that hadnot taken a normal feed were dropped from con-sideration. The mosquito mid-guts were thenstained in mercurochrome, and oocysts werecounted using a light microscope (Table 3).

3. Results

3.1. Pre-incubation of sporozoites in athiostrepton solution

We incubated batches of mature sporozoites ineach of the following, thiostrepton (500 mgml−1), primaquine (50 mg ml−1) or chloroquine(50 mg ml−1). Primaquine kills sporozoites and


therefore was included as a control for drug activ-ity. Chloroquine, which has no effect on exoery-throcytic liver stages of the parasite but iseffective against blood-stage parasites, was in-cluded as a control of sporozoite viability. Fi-nally, a batch of sporozoites that had beenincubated in saline without drug was used asanother control. Mice were divided into groups offour, after which members of the different groupswere injected with 500 sporozoites from one of thecategories described above. Incubation with pri-maquine prevented infections in the mice, whilesporozoites treated with chloroquine or thiostrep-ton remained infective, as did the untreated con-trols. Unlike primaquine, thiostrepton appearedto have no effect on mature sporozoites whentreated before injection (data not shown).

Treatment of the mice with each drug prior toinoculation of sporozoites gave a different result.The period prior to emergence of infective mero-zoites from the liver is between 42 and 72 h afterinfection [7]. In these experiments, mice weretreated for each of the 2 days prior to inoculation,at the time of inoculation, and 1 day after inocu-lation (see Section 2). Sporozoites were deliveredboth by the bite of an infected mosquito (Table 1)and by manual inoculation of sporozoites (Table4), as described above. This regimen of thiostrep-ton treatment prevented infection in both situa-tions. The infection rate of chloroquine-treatedmice after the bite of a mosquito was about 50%and probably resulted from residual chloroquinekilling parasites after their emergence from theliver. The injection of sporozoites intochloroquine-treated mice was not attempted. Theprimaquine treatment also eliminated infectionsby both routes, while the control mice became

infected with parasites in both situations. Theelimination of parasites by thiostrepton in thiscase most probably occurred during developmentafter invasion but, as in the case with chloroquine,could have occurred after emergence into theblood.

3.2. Treatment of P. berghei-infected mice withthiostrepton

Mice were inoculated intravenously with 106

parasites and then separated into six groups ofeight mice each. The high numbers of parasites inthe inoculum routinely allowed parasitemias to bepatent by the third day after infection. Groupswere then treated on a 4-day regimen as describedin Section 2 and in Table 2. Mice were treatedwith either chloroquine or one of a series ofdilutions of thiostrepton. Chloroquine treatmentcleared parasites rapidly as expected, and none ofthe eight mice developed a parasitemia. Thiostrep-ton acted more slowly, as drugs targeted to theplastid are known to do. Mice treated with thetwo higher doses of thiostrepton controlled theparasitemia (Table 2). All eight mice in the high-dose group cleared parasites by day 10, whileseven of eight cleared parasites in the second levelof dosing. The parasites in these mice appeared tobe highly vacuolated, and the characteristic bluecolor of the cytoplasm after Giemsa stain waseither missing or confined to small areas (Fig. 1).At lower doses of thiostrepton and in the controlgroup, parasites were seen on day 4 and continueto rise in number until day 10. This indicates thatearlier studies on the thiostrepton treated para-sites grown in culture were relevant to in vivotreatment as well. The delayed effect of the drug isconsistent with that found when using drugsthought to specifically interact with the plastid-like organelle.

3.3. Sexual de6elopment of the parasite in themosquito is disrupted by treatment withthiostrepton

Gametocytes develop in the blood of the verte-

Table 4Transmission blocking activity of thiostrepton on P. bergheioocyst development

Compounds Dose per mg Number of mid-gutswith oocystskg−1 i.p.

500 mg 1/12Thiostrepton12/12Control Saline


Fig. 1. Morphology of treated and untreated infected red blood cells. WC indicates white blood cells.

brate and release gametes within the gut of themosquito, while zygote formation takes place inthe mosquito gut and proceeds through its insectcycle. Thus, gametocyte development is funda-mentally different from the asexual parasite, andoften drugs that have an effect on one do nothave an effect on the other. Here, we tested the

effects of thiostrepton and chloroquine on trans-mission of P. berghei (Table 3). Chloroquine doesnot block parasite transmission.

In our hands, transmission of P. berghei to themosquito is most efficient when fed at para-sitemias between 6 and 8%. We therefore inocu-lated mice with parasites and monitored their


parasitemias. When the mice reached a 1% para-sitemia, we treated half with thiostrepton as de-scribed in Section 2. We then waited until theuntreated control mice reached parasitemias ofbetween 6 and 8%. We allowed mosquitoes tofeed on either saline- or thiostrepton-treated mice.After 14 days, we dissected mosquitoes andlooked for eggs, which would indicate a normalblood feed. Using the mosquitoes that tookblood, we proceeded to measure transmission by acount of oocysts (Table 3). All 12 mosquitoes hadover 100 oocysts per mid-gut in the saline-treatedcontrols. Only one of 12 mosquitoes that fed onthe thiostrepton-treated mice developed oocysts.The infected mosquito, in this single case, hadfewer than ten oocysts. This experiment was re-peated five times with the same or very similarresults.

Collections of 50 mosquitoes from each of theabove groups were allowed to develop further.Our data indicate that of the mosquitoes thattook a blood meal from thiostrepton-treated mice,only 10% could develop sporozoites. On day 28,each group of mosquitoes fed on two mice, re-spectively. One of two mice from the drug-treatedgroup developed parasitemia, while both micefrom the untreated group developed parasitemia.Infective mature sporozoites were thereforepresent on or before day 28. Hence, althoughthiostrepton has a very significant effect on thenumber of infected mosquitoes resulting from ablood feed, ‘breakthrough’ parasites can developat an essentially normal rate. The effect ofthiostrepton may be reversible after a certainpoint in parasite development.

4. Discussion

A number of the organisms included in thephylum Apicomplexa, including malaria and Tox-oplasma, contain a plastid-like organelle whosefunction is not totally understood. Treatment ofToxoplasma gondii tachyzoites with an inhibitorof plastid genome replication, ciprofloxicin, leadsto poisoning of the parasite with delayed kinetics.It has been shown that intracellular replication oftachyzoites is unaffected by ciprofloxacin while

the parasite remains in the host cell. A malforma-tion of the parasitophorous vacuole, however,correlates with entry into a new host cell and thesubsequent death of the parasite. Although theplastid may be essential for other functions suchas heme biosynthesis, the observable effect ofinterrupting its function lies with membrane func-tion [4].

Sporozoites develop normally if incubated withthiostrepton prior to injection. Mice pretreatedwith the drug, however, do not become infected asthe result of either the bites of infectedmosquitoes or direct injection of sporozoites.Since the sporozoites contact with the drug alonedoes not affect its potential for further develop-ment, the action of thiopeptides is likely to occurafter liver cell invasion. Parasite growth inhibitionwithin the liver cell could account for parasitedeath.

Parasites were cleared from the blood of miceby treatment with thiostrepton. The rate of clear-ance was delayed in comparison with chloroquine(Table 2). Treated parasites became highly vacuo-lated, and the cytoplasm did not stain in a charac-teristic manner (Fig. 1). The appearance of thecells is reminiscent of parasites that have beentreated with drugs that inhibit hemoglobin diges-tion. There appear to be large food vacuoles andsmall separated regions, which stain blue, perhapsrepresenting cytoplasm containing undigestedhemoglobin. This may be an effect of interruptedplastid function, but this cannot be presently re-solved. Further, the inoculation of a million para-sites into the mouse, in combination with thelarger doses of the drug (250 and 500 mg kg−1),caused mice to have a ruffled fur appearance. Thebehavior and appearance of the mice returned tonormal within a few days after cessation of drugtreatment.

Parasite transmission to the mosquito was alsodrastically reduced by pretreatment of mice withthiostrepton. The experiment was performed onfive independent occasions. Mosquitoes thatshowed a normal clutch of eggs were assumed tohave taken a normal blood meal and were furtherexamined for developing parasites. The percent-age of mosquitoes developing egg clutches was


approximately equal in both drug-treated andcontrol batches. Each time, the test showed dra-matic and similar results. Approximately 10% ofthe mosquitoes taking drug-treated blood devel-oped oocysts, while essentially all of themosquitoes taking untreated blood developedoocysts. The number of oocysts developing fromtreated blood was also much lower than in con-trols. One set of results is shown in Table 3.Blockage occurred somewhere between thegametocyte stage and formation of the oocyst.Once the oocyst was formed, the development ofinfective sporozoites could occur at a normal rate;hence, no observable effects of prior treatmentwith thiostrepton were observed. This contrastswith the delayed response seen in blood stages. Itis clear that transmission can be significantly af-fected by the presence of thiostrepton.

New drugs designed to take advantage of themechanism(s) of parasite resistance (e.g.chloroquine) represent a promising approach tothe development of a new generation of anti-malarial drugs; however, basing so much futurehope on drugs that do not also interrupt thetransmission cycle may not completely serve theneeds of developing nations. In this light, onemust ask whether the use of apicoplast-directeddrugs as a front line for less severe cases would bea useful approach not only to disease interventionbut also to reducing transmission pressure. Al-though drugs that interact with the plastid resultin delayed clearance of parasites from the blood,they do arrest parasite growth and, as we showhere, act over a broader range of the parasite lifecycle than do the faster-acting drugs. The effect ofthe extended presence in the blood of parasitesattenuated by thiopeptides on the development ofimmunity is also a factor worth investigating. Oneexample of a plastid-directed drug being success-fully used in the field is clindamycin. This drughas been tested in field trials in the Sudan andbeen shown to be an effective anti-malarial [8].The effects on parasite transmission were nottested but are likely to exist if the drug workssimilarly to thiostrepton. Although clinical effectsof clindamycin are generally mild, there were sideeffects in a few participants. Thiostrepton itself is

probably not the ideal drug to be included in drugregimens against malaria because of its potentialtoxicity at high concentrations. Other commer-cially unavailable thiopeptides, such as micrococ-cin, are effective in vitro at concentrationscomparable to those used for chloroquine [9].Taking advantage of prior work in this area ofnatural products research, which resulted in defin-ing the organisms and the procedures for produc-ing these chemicals, might now be a cost-effectiveapproach to defining a number of anti-malarialdrugs. Other types of drugs targeting the api-coplast are equally supported by this work. Thesedrugs are unlikely to replace the use ofchloroquine, the 8-aminoquinolines, and quininederivatives, for rapid disease intervention buttheir distinctly different mechanism of action maymake them useful alternative or companion drugs.

References

[1] Wilson RJM, Denny PW, Preiser PR, Rangachari K,Roberts K, Roy A, Whyte A, Strath M, Moore DJ, MoorePW, Williamson DH. Complete gene map of the plastid-like DNA of the malaria parasite Plasmodium falciparum. JMol Biol 1996;261:155–72.

[2] McFadden GI, Reith ME, Mulholland J, Lang-UnnaschN. Plastid in human parasites. Nature 1996;381:482.

[3] Kohler S, Delwiche CF, Denny PW, Tilney LG, WebsterP, Wilson RJ, Palmer JD, Roos DS. A plastid of probablegreen algal origin in Apicomplexan parasites. Science1997;275:1485–9.

[4] Fichera ME, Roos D. A plastid organelle as a drug targetin Apicomplexan parasites. Nature 1997;390:407–9.

[5] McConkey GA, Rogers MJ, McCutchan TF. Inhibition ofPlasmodium falciparum protein synthesis: targeting theplastid-like organelle with thiostrepton. J Biol Chem1997;272:2046–9.

[6] Pfefferkorn ER, Nothnagel RF, Borotz SE. Parasiticidaleffect of clindamycin on Toxoplasma gondii grown in cul-tured cells and selection of a drug-resistant mutant. An-timicrob Agents Chemother 1992;36:1091–6.

[7] Killick-Kendrick R. Rodent Malaria. New York: Aca-demic Press, 1978.

[8] el Wakeel ES, Homeida MM, Ali HM, Geary TG, JensenJB. Clindamycin for the treatment of falciparum malaria inSudan. Am J Trop Med Hyg 1985;34:1065–8.

[9] Rogers MJ, Cundliffe E, McCutchan TF. The antibioticmicrococcin is a potent inhibitor of growth and proteinsynthesis in the malaria parasite. Antimicrob AgentsChemother 1998;42:715–6.

.

Documents

Effects of interruption of apicoplast function on malaria infection, development, and transmission