Pii: s0020-7519(98)00230-6

International Journal for Parasitology 29 (1999) 619±625 The e€ect of chloroquine treatment on the infectivity of Institute of Cell, Animal and Population Biology, Division of Biological Sciences, University of Edinburgh, Edinburgh EH9 3JT, UK Received 19 October 1998; received in revised form 3 December 1998; accepted 3 December 1998 The antimalarial drug chloroquine has been reported to increase the infectivity of the forms of blood-stage malaria parasites (gametocytes) that are capable of infecting mosquito vectors. This e€ect has been demonstrated convincingly in the short term (12 h post treatment), although several authors have suggested infectivity enhancement a week or more after treatment. We carried out experiments to investigate the e€ects of chloroquine on the longer-term infectivity of gametocytes of the rodent malaria parasite, Plasmodium chabaudi, to Anopheles stephensi mosquitoes. Gametocytes of chloroquine-treated infections were signi®cantly more infectious than untreated infections 6 and 7 days post-treatment, although not on days 8 and 9. However, this e€ect was most likely the result of a reduction in infectivity in untreated infections, caused by immune activity which was not so pronounced in chloroquine-treated infections. Gametocytaemia (gametocytes per r.b.c.) showed a strong positive and linear relationship with infectivity. Infectivity was not in¯uenced by either asexual parasitaemia, asexual density or anaemia. Parsimonious interpretations of the e€ect of chloroquine on gametocyte infectivity are discussed.
# 1999 Australian Society for Parasitology Inc. Published by Elsevier Science Ltd. All rights reserved.
Keywords: Anopheles stephensi; Chloroquine; Gametocyte; Infectivity; Malaria; Plasmodium chabaudi; Transmission the transmission stages of a Plasmodium infec- failed to signi®cantly reduce malaria prevalence, tion, gametocytes [2, 3]. However, two other and resistance to many antimalarial drugs has spread at an alarming rate [1]. This is partly enhance transmission have been suggested. First, because some of the commonly used antimalarial chemotherapy may induce greater gametocyte drugs have little parasite killing activity against production, both through genotypic [4±6] and short-term phenotypic changes [7]. Given the generally positive correlation between gametocyte densities and infectivity to mosquitoes [8, 9], such * Corresponding author. Present address: Department of changes are likely to increase transmission.
Plant Sciences, University of Oxford, South Parks Road, Second, the antimalarial drug chloroquine (CQ) Oxford OX1 3RB, UK. Tel: 01865-275106; Fax: 01865- 275074; e-mail: angus.buckling@plant-sciences.oxford.ac.uk has been reported to increase the infectiousness 0020-7519/99/$20.00 # 1999 Australian Society for Parasitology Inc. Published by Elsevier Science Ltd. All rights reserved.
A.G.J. Buckling, A.F. Read / International Journal for Parasitology 29 (1999) 619±625 of gametocytes, independent of gametocyte cyte infectivity with increasing eciency for 21 days following treatment'', using cultured Data on infectivity enhancement of gameto- gametocytes of P. falciparum and of the rodent cytes by CQ are ambiguous. It is clear that CQ malaria, Plasmodium berghei. They report a sig- per se does not increase gametocyte infectivity: ni®cant increase in infectivity through time fol- the addition of CQ to cultures prior to feeding mosquitoes did not increase either the pro- gametocytes, and a non-signi®cant increase with portions of mosquitoes infected, or the number P. falciparum gametocytes, although the validity of these analyses is questionable. Infectivity of developed [1, 10, 11]. Enhanced gametocyte infec- gametocytes obtained from naturally infected tivity 12-h post sub-curative CQ treatment has CQ-treated volunteers was not signi®cantly been demonstrated in vivo for the rodent malaria greater when mixed with the patients' own (CQ- containing) plasma compared with European Plasmodium yoelii nigeriensis [14]. However, there control plasma [17]. A ®eld study investigating was no reported increase in the infectivity of factors in¯uencing human infectivity to mosqui- Plasmodium falciparum gametocytes when mixed toes also found no P. falciparum infectivity- with sera collected 24 h after CQ had been taken, compared with sera before treatment [15].
It is clearly important to determine whether Similarly, there was no infectivity enhancement CQ enhances gametocyte infectivity for more than a matter of hours. We therefore carried out between 0 and 72 h post-treatment [3].
experiments to investigate gametocyte infectivity of the rodent malaria, Plasmodium chabaudi, in enhance per gametocyte infectivity a week or vivo 6±9 days after CQ treatment, at the time of more after treatment, but here the data are even more ambiguous. Curative CQ-treatment of P.
Details of methodology are given in Buckling falciparum-infected humans resulted in greater et al. [7]. Male C57/BL/6J mice were infected by median parasite burdens in Anopheles balabacen- i.p. inoculation with 106 parasites on day 0.
sis mosquitoes (but not in Anopheles dirus) fed 2 Parasite clones and sample sizes are shown in and 7 days after treatment, compared with that Table 1. On day 5 or 6 p.i. (1±2 days before the before treatment. However, mean oocyst burdens average peak asexual parasite density), half the and the proportions of mosquitoes infected did mice were orally treated with a sub-curative dose not change [16]. Hogh et al. [17] recently reported of chloroquine sulphate (12 mg kgÀ1); the others that ``serum from chloroquine-treated, unin- received distilled water (controls). In each exper- fected, non-immune volunteers enhanced gameto- iment, half the control and half the CQ-treated a One mouse died from severe anaemia on day 8 post-infection.
A.G.J. Buckling, A.F. Read / International Journal for Parasitology 29 (1999) 619±625 mice were picked randomly and fed to H40 4±5- day-old Anopheles stephensi on day 12 p.i. (except in the control group in experiment 1, where two animals out of ®ve were fed to mosquitoes). The other mice were fed to mosquitoes on day 14 p.i.
Days 12 and 14 p.i. are the days of peak gameto- cyte densities in CQ-treated and untreated infec- tions, respectively [7]. Asexual parasites and gametocytes per r.b.c. and r.b.c. density were determined for each mouse on both feed days; parasite densities were calculated from the pro- duct of these variables. Approximately 25 mos- quitoes per mouse were dissected 8±9 days after Fig. 1. Mean21 S.E.M. gametocyte densities in chloroquine- feeding and the number of parasites (oocysts) on treated and untreated infections on days 12 and 14 p.i.
All analyses were carried out using GLIM 4.
within each analysis [19, 20]. Data from individ- Proportions of mosquitoes infected per mouse ual mosquitoes were nested within mouse to were analysed by logistic regression with a bino- avoid pseudo-replication. Non-signi®cant inter- mial error structure. Over-dispersion in the data actions were combined with the error term to was corrected by using a heterogeneity factor determine signi®cance of main e€ects. Di€erences (HF) of 3.05, determined from the ratio of Pearson's w 2 to the degrees of freedom in the between days and treatments were determined by minimal model [19]. Day p.i. of feed (12 or 14), ANOVA, using stepwise deletion as above, with treatment (CQ + , CQÀ) and experiment (1, 2 or treatment, day and experiment ®tted as factors.
3) were ®tted as factors, and square-root arcsin- Gametocyte densities were 3.8 times higher on transformed gametocytes per r.b.c. (gametocytae- day 12 p.i. than day 14 p.i. in CQ-treated infec- mia) ®tted as a covariate. The minimal model tions, and 4.4 times higher on day 14 p.i. than was determined by stepwise deletion, starting day 12 p.i. in untreated infections (Fig. 1, day by with the highest order interactions, with non-sig- treatment interaction: F1,29 = 5.03, P < 0.05). In ni®cant terms discarded [19]. The explanatory the full logistic regression model, gametocytaemia power of remaining terms was determined by del- explained 46% of the deviance associated with etion from the minimal model. Additional covari- ates [log10-transformed gametocyte densities, asexual densities and r.b.c. densities, and square- root arcsin-transformed asexuals per r.b.c.
(asexual parasitaemia)] were substituted with gametocytaemia in the minimal model to deter- mine if they explained a signi®cant amount of Further analyses were carried out in the same way within days (HF = 3, 3.6 for days 12 and 14 p.i., respectively) and treatments (HF = 1.8, 4.1 for CQ+ and CQÀ, respectively), using only covariates that were signi®cant in the previous model. Analyses of oocyst burdens were also car- ried out within days and treatments using a nega- Fig. 2. Relationship between proportion of mosquitoes infected and gametocytaemia (gametocytes per 1000 r.b.c.) for tive binomial error structure, with the over- all data. Small numbers in body of ®gure show multiple data dispersion parameter, k, estimated separately A.G.J. Buckling, A.F. Read / International Journal for Parasitology 29 (1999) 619±625 proportions of mosquitoes infected (Fig. 2, Because of the signi®cant treatment by day in- w 21 = 25.12, P < 0.001), which was signi®cantly teraction in the full model, data were analysed greater than the deviance explained by gameto- separately within days and treatments. When the cyte density (di€erence between gametocytaemia data from day 12 p.i. alone were analysed, infec- and gametocyte density: w 21 = 4.87, P < 0.05).
tivity per gametocyte was 3.2 times greater (at The relationship between (untransformed) game- the mean gametocytaemia) in CQ-treated than untreated infections (Fig. 3a, main e€ect of treat- infected was approximately linear (Fig. 2); allow- ing the best ®t regression to curve, by the ad- 1 = 4.3, P < 0.05). Mean oocyst burdens (per gametocyte) were 3.1 times greater (at the dition of the quadratic function, did not explain mean gametocytaemia) in CQ-treated compared signi®cantly more deviance (P > 0.1).
To determine the e€ect of CQ on per gameto- with untreated infections (main e€ect of treat- cyte infectivity, gametocytaemia had to be con- ment: F1,10 = 12.12, P < 0.01). The rates of trolled for. When this was done, the e€ect of increase of both proportion of mosquitoes CQ-treatment on proportions of mosquitoes infected and mean oocyst burdens with gameto- infected di€ered on days 12 and 14 p.i. (treat- cytaemia did not di€er between CQ-treated and untreated infections (treatment by gametocytae- No main e€ects of the factors (day, experiment mia interaction: P > 0.1). Neither proportions of or treatment), covariates (gametocyte, asexual mosquitoes infected or mean oocyst burdens (per and r.b.c. densities, and asexual parasitaemia) or gametocyte) on day 14 p.i. di€ered between CQ- their other interactions explained a signi®cant ad- treated and untreated infections (Fig. 3b, P > 0.1 ditional amount of deviance (P > 0.05, in all for main e€ects of treatment, and treatment by Fig. 3. Relationship between logit-transformed proportions of mosquitoes infected and square-root arcsin-transformed gametocy- taemia. Lines show OLS regressions from statistical models. Labels in bold type correspond to solid symbols. Small numbers in body of ®gures represent multiple data points. (a) Data from day 12 p.i. (b) Data from day 14 p.i. (c) Data from all untreated infections. (d) Data from all chloroquine-treated infections.
A.G.J. Buckling, A.F. Read / International Journal for Parasitology 29 (1999) 619±625 treated infections because of the greatly reduced sidered, the proportion of mosquitoes infected peak asexual parasite densities [7].
and mean oocyst burdens (per gametocyte) were, In untreated infections, reduced gametocyte at the mean gametocytaemia, respectively 3.4 and infectivity on day 12 p.i. was also associated with 2.9 times greater on day 14 p.i. than day 12 p.i.
low gametocyte densities. There was a similar (Fig. 3c, main e€ects of day: w 21 = 7.94, P < 0.01, pattern on day 14 p.i. in CQ-treated infections.
F1,12 = 10.89, P < 0.01, respectively; day by The infectivity of individual gametocytes may gametocytaemia interactions: P > 0.1, in both therefore be a positive function of gametocyte cases). However, within CQ-treated infections, density, as would be the case if the probability of gametocyte infectivity as measured by both pro- gamete fusion in the mosquito midgut was con- portions of mosquitoes infected and mean oocyst siderably reduced at low gametocyte densities.
burdens, did not di€er between days 12 and 14 There is, however, no evidence that per gameto- p.i. (Fig. 3d, P > 0.1 for both day and day by cyte infectiousness increased non-linearly at high gametocytaemia interaction, in both cases).
The results imply that if CQ does increase the infectiousness in this study, but did CQ enhance infectiousness of P. chabaudi gametocytes, the per gametocyte infectiousness? On day 12 p.i., e€ect is lost by 6±7 days post-treatment. This is gametocytes in CQ-treated mice were more infec- before the vast majority of gametocytes were pro- tious than those in untreated infections, both in terms of proportion of mosquitoes infected and infections [7]. The plasma half-life of CQ in ma- mean oocyst burdens. Two days later, any e€ects laria-infected mice is approximately 7 h [24], and of CQ on gametocyte infectivity were no longer thus would be at very low concentrations after detectable. These data can be interpreted in two 6±7 days. The current results are therefore incon- ways. First, CQ enhanced infectivity on day sistent with the suggestion that long-term metab- 12 p.i., but the e€ect was lost by day 14 p.i.
olites of CQ are responsible for the reported Second, infectivity was suppressed on day 12 p.i.
infectivity-enhancing e€ects of CQ [1, 17].
in untreated infections, but not on day 14 p.i.
Ultimately, our data are consistent with most That infectivity was greater on day 14 p.i. than previously published data: there is no strong evi- day 12 p.i. in untreated infections, but there was dence to suggest that CQ has long-term infectiv- no di€erence in infectivity between days in CQ- ity-enhancing e€ects on gametocytes.
treated infections, leads to acceptance of the lat- ter hypothesis. Thus, any e€ect of CQ on infec- demonstrated convincingly only in the short term (12 h post-treatment), in controlled experiments The reduction in infectivity in untreated infec- using rodent malaria parasites in vivo [12±14]. A tions relative to CQ-treated infections on day 12 plausible explanation for these data is an p.i. is consistent with the infection dynamics. In immuno-suppressive e€ect of CQ. Chloroquine is untreated infections a phenomenon called ``crisis'' occurs. This is a rapid reduction in parasite num- presentation [25], and leukocyte function in bers associated with low r.b.c. densities and general [26]. Such mechanisms could conceivably strong immune activity [21], during which game- reduce anti-gametocyte immune activity, resulting tocyte infectivity is suppressed [22] (and see in increased gametocyte infectivity in the short Carter et al. [23] for a related phenomenon in the term. Infectivity-enhancing e€ects would be human malaria parasite, P. vivax). Crisis expected to last only as long as CQ (or the major occurred between days 8 and 10 p.i. in untreated antimalarial metabolite, desethyl-CQ) was at suf- infections in these experiments, and it is entirely ®cient concentrations to have an immuno- plausible that gametocyte infectivity-suppressing suppressive e€ects. Infectivity enhancement ``crisis factors'' are still present by day 12 p.i., would be expected to increase with CQ dose, up but not day 14 p.i. Crisis did not occur in CQ- to a point where growing gametocytes were being A.G.J. Buckling, A.F. Read / International Journal for Parasitology 29 (1999) 619±625 destroyedÐmature gametocytes are una€ected by [7] Buckling AGJ, Taylor LH, Carlton JM-R, Read AF.
CQ [2]. This prediction is supported by inter- Adaptive changes in Plasmodium transmission strategies mediate doses of CQ having the greatest infectiv- following chloroquine chemotherapy. Proc R Soc Lond ity-enhancing e€ect [12]. Although a short-term [8] Carter R, Graves PM. Gametocytes. In: Wernsdorfer immuno-suppressive e€ect of CQ could be of po- WH, McGregor I, editors. Malaria. Principles and prac- tential importance to malaria epidemiology and tice of malariology. Edinburgh: Churchill Livingstone, the spread of drug-resistant genotypes, longer- term e€ects, if they occur, would be of great [9] Taylor LH, Read AF. Why so few transmission stages? signi®cance. We ®nd no evidence of such long- Reproductive restraint by malaria parasites. Parasitol term e€ects in P. chabaudi. Further work on P.
[10] Chutmongkonkul M, Maier WA, Seitz HM. Plasmodium falciparum is required to resolve this important falciparum: e€ect of chloroquine, halofantrine and pyri- methamine on the infectivity of gametocytes for Anopheles stephensi mosquitoes. Ann Trop Med Parasitol [11] Rosario VE, Vaughn JA, Murphy M, Harrod V, Coleman R. E€ect of chloroquine on the sporogonic cycle of Plasmodium falciparum and Plasmodium ber- ghei in Anopheline mosquitoes. Acta Leid 1988;57:53± Research Fellowship to A.R. A.B. was supported [12] Ramkaran AE, Peters W. Infectivity of chloroquine re- by an MRC studentship. We are grateful to sistant Plasmodium berghei to Anopheles stephensi David Walliker for the use of his insectaries.
enhanced by chloroquine. Nature 1969;223:635±6.
Thanks to two anonymous referees for comments [13] Hamidi AN. The e€ect of chloroquine on the infectivity of rodent malaria parasites to Anopheles stephensi. Iran J [14] Ichimori K, Curtis CF, Targett GAT. The e€ects of chloroquine on the infectivity of chloroquine-sensitive and -resistant populations of Plasmodium yoelii nigerien- sis to mosquitoes. Parasitology 1990;100:377±81.
[15] Ponnudurai T, Lensen AH, Van Gemert GJ, Bensink [1] Butcher GA. Antimalarial drugs and the mosquito MP, Bolmer M, Meuwissen JH. Infectivity of cultured transmission of Plasmodium. Int J Parasitol 1997;27:975± Plasmodium falciparum gametocytes to mosquitoes.
[2] Smalley ME, Sinden RE. Plasmodium falciparum gameto- [16] Wilkinson RN, Noeypatimanondh S, Gould DJ.
cytes: their longevity and infectivity. Parasitology Infectivity of falciparum malaria patients for Anopheline mosquitoes before and after chloroquine treatment.
[3] Klein TA, Tada MS, Lima JB, Katsuragawa TH.
Trans R Soc Trop Med Hyg 1976;70:306±7.
Infection of Anopheles darlingi fed on patients infected [17] Hogh B, Gamage-Mendis A, Butcher GA et al. The with Plasmodium vivax before and during treatment with di€ering impact of chloroquine and pyrimethamine/sul- chloroquine in Costa Marques, Rondonia, Brazil. Am J fadoxine upon the infectivity of malaria species to the mosquito vector. Am J Trop Med Hyg 1998;58:176± [4] Bishop A. The e€ect of sulphadiazine, proguanil and 2:4- diamino-6:7-diisopropylpteridine upon gametocyte pro- [18] Tchuinkam T, Mulder B, Dechering K et al.
duction in Plasmodium gallinaceum. Parasitology with Plasmodium falciparum of naturally infected [5] Lines JD, Wilkes TJ, Lyimo EO. Human malaria infec- gametocyte carriers in Cameroon: factors in¯uencing the tiousness measured by age-speci®c sporozoite rates in [19] Crawley M. GLIM for ecologists. Oxford: Blackwell [6] Handunnetti SM, Gunewardena DM, Pathirana PPSL, Ekanayake K, Weerasinghe S, Mendis KN. Features of [20] Wilson K, Grenfell BT. Generalized Linear Modelling recrudescent chloroquine-resistant Plasmodium falciparum for parasitologists. Parasitol Today 1997;13:33±8.
infections confer a survival advantage on parasites and [21] Jarra W, Brown KN. Protective immunity to malaria: have implications for disease control. Trans R Soc Trop studies with cloned lines of rodent malaria in CBA/Ca mice. IV. The speci®city of mechanisms resulting in crisis A.G.J. Buckling, A.F. Read / International Journal for Parasitology 29 (1999) 619±625 and resolution of the primary acute phase parasitaemia [24] Cambie G, Verdier F, Gaudebout C, Clavier F, of Plasmodium chabaudi chabaudi and P. yoelii yoelii.
Ginsburg H. The pharmacokinetics of chloroquine in healthy and Plasmodium chabaudi-infected mice: impli- [22] Wery M. Studies on the sporogony of rodent malaria cations for chronotherapy. Parasite 1994;1:219±26.
parasites. Ann Soc Belge Med Trop 1968;48:1±138.
[25] Fox RI, Kang H-I. Mechanism of action of antimalarial [23] Carter R, Wijesekera SK, Karunaweera ND, Mendis drugs: inhibition of antigen processing and presentation.
KN. Mediators and mechanisms associated with parox- ysm in Plasmodium vivax malaria. J Pharm Pharmacol [26] Targett GAT. Drug use and the immune response.

Source: http://www.thereadgroup.net/wp-content/uploads/Buckling_IJP.pdf