ORIGINAL PAPER

T. Voza Æ P. Gautret Æ L. Renia Æ J.-C. Gantier

M.-N. Lombard Æ A.G. Chabaud Æ I. Landau

Variation in murid Plasmodium desequestration and its modulationby stress and pentoxifylline

Received: 2 October 2001 /Accepted: 5 October 2001 / Published online: 21 December 2001� Springer-Verlag 2001

Abstract Changes in the parasitaemia and the charac-teristics of parasitic infection for three species of rodentPlasmodium (P. chabaudi chabaudi, P. vinckei petteri andP. yoelii yoelii) were investigated under conditions ofstress and after treatment with pentoxifylline (POF), adrug that increases red blood cell deformability andcauses peripheral vasodilatation. The results indicatedthat under stress, late parasite stages became lessabundant in the tail blood of mice. These changes mightbe the consequence of parasite sequestration. Attemptsto assess sequestration intensity were made by measur-ing the release rate (RR) of late stages for 10,000 redblood cells. The RR is given by the product of theparasitaemia (P) by the percentage of old trophozoites(OT) and schizonts (S) in the peripheral blood:RR ¼ P ð%OTþ%SÞ . With all three species, RR de-creased considerably within 5 min following themanipulation of the mice. Injections of POF had theopposite effect. POF had a protective effect againstinfection by P.v. petteri, causing a delay of 48 h in thedevelopment of infection and a higher survival rate intreated mice.

Introduction

Sequestration of Plasmodium is a phenomenon of greatimportance at different levels:

1. It occurs over a variable time period at the end of theschizogonic cycle. It can last for a long time andinvolve the schizont stage as well as the late troph-ozoite as, for example, in Plasmodium falciparum forwhich only the rings and early trophozoites are usu-ally seen in blood smears. In other species, a variablenumber of schizonts are found in the blood butmorphological analysis has revealed that these stagesare still immature, the mature schizonts being locatedin the microvasculature (Coquelin et al. 1999).

2. We believe that merozoites, released in the fast bloodflow of large or medium size vessels, are unable topenetrate red blood cells (RBCs). On the other hand,in pulmonary capillaries where schizonts are betterable to complete their development, merozoites incontact with the vascular endothelium (Coquelin et al.1999) may be associated with RBCs passing slowlynearby over a longer period.

3. Sequestration is believed to play a major role in thepathogenesis of malarial infections. It may beinvolved in the generation of histological lesions andcerebral malaria (Berendt et al. 1994). Malaria isconsidered, in some ways, to be a vascular disease(Vuong et al. 1999) in which haemorrhages andthrombosis determine ischaemia and the localnecrosis responsible for malarial pathology.

Many studies have dealt with the specific patterns ofcytoadherence between parasites and endothelial cellswhich result in sequestration. In this paper, we investi-gated whether sequestration intensity, which is highlysusceptible to the host physiological state, could beexperimentally manipulated.The experimental manipulation of mice always

induces stress and we demonstrate that it also leads toabrupt changes in the intensity of sequestration. We

Parasitol Res (2002) 88: 344–349DOI 10.1007/s00436-001-0538-7

T. Voza Æ M.-N. Lombard Æ A.G. Chabaud Æ I. Landau (&)Equipe Parasitologie Comparee et modelesExperimentaux Associee a L’INSERM (U445),Museum National d’Histoire Naturelle, 61 rue Buffon,75231 Paris Cedex 05, FranceE-mail: landau@mnhn.frTel.: +33-1-40793500Fax: +33-1-40793499

P. GautretLaboratoire de Parasitologie et de Mycologie Medicale,Pavillon Camille Guerin, CHU la Miletrie,BP 577, 86021 Poitiers Cedex, France

L. ReniaINSERM U 445, Institut Cochin de Genetique Moleculaire,Hopital Cochin, Batiment Gustave Roussy,27 Rue du Faubourg Saint Jacques, Paris 75014, France

J.-C. GantierLaboratoire de Parasitologie, Faculte de Pharmacie,Rue J.B. Clement, 92290 Chatenay Malabry, France

also investigate the influence on sequestration of pen-toxifylline (POF), a methylxanthine which increasesRBC deformability and decreases plasma viscosity(Leonhardt and Grigoleit1977; Ward and Clissold1987).

Materials and methods

Biological material

Swiss IOPS OF1 white mice (IFFA CREDO, France) weighing20–25 g were inoculated intraperitoneally with a cryopreservedstabilate of one of the following Plasmodiumspecies: P. chabaudichabaudi (strain 864 VD), P. yoelii yoelii (strain 265 BY) andP. vinckei petteri (strain 106 HW).

Evaluation of sequestration

Quantitative evaluation of sequestration was made by comparingthe parasitaemia and parasite stage distribution before and aftermanipulation or drug-treatment on Giemsa-stained tail-bloodsmears. When sequestration increases, the parasitaemia decreaseson tail-blood smears and the percentage of old stages also dimin-ishes. The reverse is observed when sequestration lessens.When a mouse has a parasitaemia of P% and a parasitic pattern

of R% rings, YT% young trophozoites, MT% mid term troph-ozoites, OT% old trophozoites and S% schizonts, the release rate(RR) for 10,000 RBCs is given by the product P(OT+S). When,after treatment, the RR increased in the tail-blood of the mouse, weassumed that sequestration decreased and when RR decreased thatsequestration increased.In ‘‘young or early stages’’ we included rings (R), young

trophozoites (YT) and mid-term trophozoites (MT) in ‘‘old or latestages’’, old trophozoites (OT) and schizonts (S), according to theclassification of Cambie et al. 1991.

Data interpretation

In this study we only investigated general biological trends, con-sidering mean RR values and RR variations between consecutivemeasures. Except for experiment 5 (Fig. 1) with P. vinckei petteri,each mouse manipulation must be considered as an independentexperiment. It would be hazardous to interpret the results forgroups of mice.

Results

P. c. chabaudi: RR in relation to experimentalprocedure

Nineteen mice were infected intraperitoneally with acryopreserved strain of P. c. chabaudi. At 6 days post-inoculation, when the parasitaemia ranged from 1% to18%, the following treatments were applied randomly tofour groups of four or five mice:

1. Series C: control2. Series P: intramuscular prick without injection3. Series S: intramuscular injection of 0.1 ml saline(0.9% NaCl)

4. Series POF: intramuscular injection of 0.1 ml POF(2 mg/mouse)

Blood smears from all animals were taken beforetreatment (T0), and at 5, 10 and 15 min after treatment(T5, T10 andT15).Values of RR are detailed in Table 1 . The results

show that in all groups (C, P, S, and POF) the value ofRR varied after treatment. In group C, P and S, RRdecreased 5 min post-treatment (13 from 14 mice) and atT10 the RR of 10 from 14 mice increased.The type of stress is not relevant since the decrease in

RR showed little variation in the first three groups (C, Pand S). In the POF group, however, the majority of mice(three from five) showed an increased RR atT5. At T10and T15 the values were variable within and betweengroups in comparison with the RRs estimated at T5 andT10 repectively.

P. c. chabaudi. RR changes after repeated injectionsof POF

Three mice were infected with a frozen stabilate ofP. c. chabaudi. At 4 days post-inoculation, when para-sitaemia ranged from 2% to 8%, the following treat-ments were applied:

1. One control mouse received six injections of 0.15 mlsaline (NaCl, 0.9%) everyhalf hour from12:00 to 15:00.

2. POF mice received six injections of 0.15 ml POF(0.3 mg) every half hour from 12:00 to 15:00.

Blood smears of all mice were made at T0, T5 andT10.In the control mouse, the RR decreased markedly 5 min

Fig. 1 Protective effect of pentoxifylline against Plasmodiumvinckei petteri. Mean parasitaemias from day 6 to 14 post-inoculation. Control group: daily intramuscular injections of0.1 ml saline solution at 11:00 and 16:00 from day 1 to day 13post-inoculation (at 17:00). POF group: daily intramuscularinjections of 0.1 ml POF (2 mg/mouse) at 11:00 and 16:00 fromday 1 to day 13 post-inoculation (at 17:00)

345

after each of the six injections (Table 2 ) and eitherremained low or increased again at T10. In mice treatedby POF, RR increased at T5 or remained unchanged. At10 min after treatment, the RRs of the POF mice wereback near their original value in half of the cases.

P. y. yoelii. RR changes after a single injectionof POF

Eleven mice were sub-inoculated intraperitoneally withthe blood of a mouse infected with P. y. yoelii. The

experiment began 5 days post-inoculation when para-sitaemia ranged from 0.8% to 2.3%:

1. Five control mice received an intramuscular injectionof 0.3 ml saline (0.9% NaCl).

2. Six POF mice received an intramuscular injection of0.3 ml POF (6 mg/mouse).

Blood smears from all animals were performed justbefore the injections (T0) then 5 and 10 min later (T5and T10).The percentages of six different blood cell categories

were considered: parasite-free reticulocytes, reticulocyteswith a young stage (ring, young and mid-term troph-ozoites), reticulocytes with a single late stage (RO),polyparasitized reticulocytes (PR), normocytes with asingle young stage, and normocytes with a late stageparasite (NO).The RR of the following stages was calculated:

1. Multiple-infected reticulocytes:RRp ¼ Pð%PRÞ (whereP is the percentage parasitaemia),

2. Multiple-infected reticulocytes and reticulocytesor normocytes with an old stage parasite:RRo ¼ P ð%PRþ%ROþ%NOÞ . Results are pre-sented in Table 3.

The RRs (RRp as well as RRo) of all the controlmice decreased within the 5 min following the salineinjection whereas the RRs of half of the mice treatedwith POF increased. In neither group did the RRs showany clear pattern at T10.These values are compatible with the following

interpretation: the RBCs or the hypertrophiedpolyparasitized reticulocytes which are less deformableare more intensely sequestered in the deep microcircu-latory system under stress conditions. This sequestrationis reduced and often reversed after a POF injection.

Table 1 Release rate ofPlas-modium chabaudi chabaudi.Release rate: RR ¼ Pð%OTþ%SÞ: P=parasitaemia,OT=old trophozoites, S=schi-zonts. Release rates of micebefore (T0), 5 min (T5), 10 min(T10) and 15 min (T15) aftertreatment. C1–C5: control gro-up. P1–P4: intramuscular prickwithout injection. S1–S5: intra-muscular injection of saline(0.9% NaCl). POF1–POF5:intramuscular injection of 2 mgPOF. Data from one individualof the P group are not presentedbecause of a very weak para-sitaemia

Mouse Release rates at times:

T0 T5 T10 T15 T0–T5

C1 9 8 12 14 1C2 2 1 2 1 1C3 6 8 10 12 –2C4 66 55 48 77 11C5 52 33 49 39 19P1 11 8 14 13 3P2 112 103 87 134 9P3 17 11 17 15 6P4 32 29 23 22 3S1 290 262 304 321 28S2 4 2 6 7 2S3 265 222 243 247 43S4 60 28 31 46 32S5 178 149 116 139 29POF1 16 19 22 24 –3POF2 108 50 24 70 58POF3 12 16 17 18 –4POF4 98 125 83 73 –27POF5 148 60 158 94 88

Table 2 Release rate of P. c. chabaudi. Release rate: RR ¼Pð%OT þ%SÞ: P=parasitaemia, OT=old trophozoites, S=schi-zonts. Release rates of mice before (T0), 5 min (T5) and 10 min(T10) after treatment. Intramuscular injections of saline into controlmouse mice (C12–C15) or 0.3 mg POF (mice POF6 and 7) everyhalf-hour from 12:00 to 15:00

Mouse Release rates at times:

T0 T5 T10

C12.00 169 103 173C13.00 229 138 355C13.30 328 203 353C14.00 256 218 209C14.30 318 200 264C15.00 376 254 248POF1 12.00 35 72 40POF1 13.00 261 313 136POF1 13.30 307 307 238POF1 14.00 255x 187 258POF1 14.30 384 280 160POF1 15.00 358 371 353POF2 12.00 189 314 312POF2 13.00 257 286 198POF2 13.30 291 299 273POF2 14.00 334 336 282POF2 14.30 308 251 258POF2 15.00 296 345 275

346

P. vinckei petteri. RR changes after a single injectionof POF

Thirteen mice were infected intraperitoneally at 16:30with a cryopreserved strain of P. vinckei petteri. On day 5post-inoculation, the following treatments were applied1 h before schizogony, at 14:45:

1. Five control mice received an intramuscular injectionof 0.15 ml saline (0.9% NaCl).

2. Eight POF mice received an intramuscular injectionof 0.15 ml POF (0.3 mg/mouse).

Blood smears from all animals were taken beforetreatment (T0), and 5, 10 and 15 min later (T5, T10andT15).The values of parasitaemias and RR aredetailed in Table 4 .The results show clearly that, 5 min after the POF

injection, the RR is increased while in the control mice,the opposite occurs. At T10 and T15 RR did not followany clear pattern in either group.

P. vinckei petteri. Protective effect of POF

In order to investigate the possible therapeutic effects ofPOF, 12 mice were infected at 17:00 with a frozenstabilate of P. v. petteri. Synchronous species schizogonywas thus expected around 16:00.Two groups of six mice received daily intramuscular

injections at 11:00 and 16:00, from day 1 to day 13 post-inoculation, of 0.1 ml of either saline (control group) orof POF solution (2 mg). Tail-blood smears were takendaily from day 6 post-inoculation in order to monitorthe course of the infection.POF injections induced a 48-h-delay in the devel-

opment of infection (Fig. 1). Mean parasitaemiadiffered significantly between groups on days 7, 8 and 9(one way ANOVA, P<0.001). All six treated micesurvived and became negative but only two mice of thecontrol group survived at the end of the experiment onday 14.

Intensity of the sequestration and desequestration

The importance of the release and retention of old stageparasites is better understood when the whole bloodpicture of the mouse is considered.Experiment 2 (P. c. chabaudi: RR changes after

repeated injections), which has the most data (22 POFinjections and six control injections), was analysed andthe number of sequestered or desequestered parasitesevaluated. The mean number of late stages in the POFgroup was 283.7 (±93.3 SD) for 10,000 RBCs at T0.This increased to 330.1(±120.0 SD) at T5. For a20-gmouse (blood volume 2 ml) the total differ-ence between T0 and T5 was approximately(330:1� 283:7Þ: 5:10 and 186 (±50.5 SD) at T5. Hence,within 5 min, about 90 million late-stage parasites were

Table 3 Release rate of P. yoelii yoelii. Release rate:RRp ¼ Pð%PRÞ: P= parasitaemia, PR=polyparasitized reticulo-cytes. Release rate: RRo ¼ Pð%PRþ %ROþ%NOÞ : RO=reticu-locytes with old trophozoites or schizonts, NO=normocytes withold trophozoites or schizonts. Release rates of mice before (T0),5 min (T5) and 10 min (T10) after treatment. Intramuscular injec-tions of saline (control mice C1–C5) or 6 mg POF ( POF mice 1–6)

Mouse Release rates at times:

T0 T5 T10

C1 P 0.8 0.9 1.3RRp 26 22 52Rro 39 19 71

C2 P 1.7 2 2.4RRp 62 50 74Rro 101 96 119

C3 P 2.3 2.9 2.7RRp 0 0 0RRo 92 90 85

C4 P 1.7 1.4 1.8RRp 29 22 31RRo 53 43 55

C5 P 1.5 1.1 1.4RRp 21 12 11RRo 62 46 38

POF1 P 1.4 1.4 1.2RRp 27 24 37RRo 60 50 46

POF2 P 1 1.3 1RRp 19 45 22RRo 46 68 36

POF3 P 1.5 1.3 1.6RRp 35 51 55RRo 40 59 77

POF4 P 12.4 0.7 0.8RRp 55 30 28RRo 77 63 43

POF5 P 1.3 2.2 2.2RRp 22 37 57RRo 43 61 42

POF6 P 1 1.3 1.5RRp 28 33 47RRo 45 84 57

Table 4 Release rate ofP. vinckei petteri. Release rate:RR ¼ Pð%OT þ%SÞ: P=parasitaemia, OT=old trophozoites,S=schizonts. Release rates (for 10,000 red blood cells) of micebefore (T0), 5 min (T5), 10 min (T10) and 15 min (T15) after treat-ment. Variation of RR within 5 min after treatment (T5)T0).C1–C5: control mice, POF1–POF8: intramuscular injection of0.3 mg POF

Mouse Release rates at times:

T0 T5 T10 T15

C1 117 93 98 57C2 296 252 175 324C3 451 430 373 393C4 242 220 184 184C5 203 110 118 87POF1 533 650 658 544POF2 287 480 344 262.5POF3 144 222 140 169POF4 85 122 145 140POF5 78 86 62 46POF6 72.5 72.6 50.6 78POF7 30 52 35 37POF8 84 88 82 79

347

retained in the deep circulatory system under stressconditions and 50 million were released in the peripheralcirculatory system of this mouse after a 0.3 mg POFinjection.

Discussion

From our results, it appears that sequestration is highlysensitive to physiological variations in the host such asthose occurring during stress. Stress is a complex processwhich involves peripheral vasoconstriction and centralvasodilatation, particularly in the internal organs. Theincrease of parasite sequestration induced by stresscaused by the manipulation of mice, such as pricks,injections or just tail-blood sampling, may be related tothe circulatory modification in the mice. Blood smearswere performed with blood drawn from the peripheralcirculation where, under stress conditions, older stageparasites are less numerous than in the better irrigatedinternal organs.The limit between MT and OT may be determined

differently by different investigators, but in this type ofexperiment these differences are of little importancesince RR varies in the same way whatever the chosenlimit between TM and TA.

Effect of POF on the circulation and the sequestrationof rodent malaria parasites

POF, which causes peripheral vasodilatation, decreasesblood viscosity and improves RBC deformability, isused for the treatment of peripheral vascular disease,cerebrovascular disease and a number of other condi-tions involving defective regional microcirculation(Ward and Clissold 1987; Bowton et al. 1989).In our study, this drug seems to counteract the effects

of stress but also results in a sudden release of old par-asitic stages into the peripheral blood. The responses tostress and POF injections are the same for single andrepeated manipulation, are always observed almostimmediately and only for a short period (10 min).An acute increase in the flow rate of capillaries, where

late-stage parasites are located, and impairment of theadhesion of parasitized cells by POF may explain therelease of parasites in the circulating blood.From our results, POF also shows a long-term effect

on the course of infection and survival in P. vinckeipetteri infected mice.Several workers have assessed the effectiveness of POF

against severe malaria due to impaired microcirculatoryflow by sequestration, and loss of RBC deformability.Studies with the murine cerebral malaria model revealedthat a daily intraperitoneal POF treatment (1 mg, fromthe day of infection) prevented neurological symptoms inP. bergheiANKA-infectedCBA/Camice (Kremsner et al.1991) or neuronal cell damage inP. bergheiK173 infected

C57/Bl6 mice (Stoltenburg-Didinger et al. 1993) but didnot affect parasitaemia and survival.

POF and tumour necrosis factor

Tumour necrosis factor (TNF) increase has often beenincriminated in the pathology of severe malaria and ithas been hypothesized that POF inhibits the productionof TNF (and other cytokines). TNF is also known toup-regulate the expression of adhesion molecules onendothelial cells (Springer 1990). Several authors haveinvestigated the relationships between POF and TNF inboth human and murine malaria.Injections of TNF to non-infected mice produced the

same pathological changes observed in severe P. vinckeimalaria (Clark et al. 1987). These studies as well aspreliminary observations on the therapeutic effects ofPOF on P. falciparum cerebral malaria (Graninger et al.1991; Landau and Attali 1993) led to a therapeutic trialwith children with cerebral malaria. POF in a continu-ous intravenous infusion of 10 mg/kg per day in com-bination with quinine significantly decreased themortality and the recovery time from coma when com-pared to a quinine only treatment. This beneficial effectof POF was associated with a decline in the serum levelsof TNF (Di Perri et al. 1995).POF is also able to disrupt P. falciparum rosettes in

vitro (Lehman et al. 1997), suggesting that this drug notonly acts via its effect on cytokine production.

Conclusions

Most of our experiments were performed during the15 min following stress or POF injection and involveonly the immediate sequestration or release of old-stageparasites. These rapid changes in the pattern of parasiticinfection are, in our opinion, likely to be determined bythe rheology of blood and the degree of flexibility of theRBCs.Our preliminary observations lead to several impor-

tant questions:

1. Does sequestration of Plasmodium old stages occur inall of the vascular system or only in some specificregions?

2. Apart from the immunological aspects which havebeen studied by other workers, what is the impor-tance of rheological parameters in sequestration? Isthe rigidity of parasitized RBCs, which hinders theircirculation, the primary mechanism which leads tocytoadherence?

3. It would also be interesting to know if the old stagesreleased by the drug are still viable or if, as thetherapeutic effect of the drug suggests, they have lostthe capacity to be sequestered and are destroyed?

Finally, the sequestration of the infective gametocytesof rodent malaria was demonstrated to facilitate their

348

uptake by the mosquito vector (Gautret and Motard1999). It would therefore be interesting to investigate theeffect of POF on parasite transmission.

Acknowledgements We are very grateful to Drs. Carol Herve fromHoechst Roussel Veterinary Laboratories, Georges Snounou fromthe Institut Pasteur and Bernard Leblond, cardiologist at Arcueil,for their comments and information on POF. We thank ProfessorH. Ginsburg for very helpful comments and criticism.

References

Berendt AR, Turner GDH, Newbold CI (1994) Cerebral malaria:the sequestration hypothesis. Parasitol Today 10:412–414

Bowton DL, Stump DA, Prough DS, Toole JF, Lefkowitz DS,Coker L (1989) Pentoxifylline increases cerebral bloodflow in patients with cerebrovascular disease. Stroke 20:1662–1666

Cambie G, Caillard V, Beaute-Lafitte A, Ginsburg H, Chabaud A,Landau I (1991) Chronotherapy of malaria: identification ofdrug-sensitive stage of parasite and timing of drug delivery forimproved therapy. Ann Parasitol Hum Comp 66:14–21

Clark IA, Cowden WB, Butcher GA, Hunt NH (1987) Possible roleof tumor necrosis factor in the pathology of malaria. AmJ Pathol 129:192–199

Coquelin F, Boulard Y, Mora-Silvera E, Richard F, Chabaud AG,Landau I (1999) Final stage of maturation of the erythrocyticschizonts of rodent Plasmodium in the lungs. C R Acad Sci III322:55–62

Di Perri G, Di Perri IG, Monteiro GB, Bonora S, Hennig C,Cassatella M, Micciolo R, Vento S, Dusi S, Bassetti D (1995)

Pentoxifylline as a supportive agent in the treatment of cerebralmalaria in children. J Infect Dis 171:1317–1322

Gautret PH, Motard A (1999) Periodic infectivity ofPlasmodiumgametocytes to the vector. A review. Parasite 6:103–111

Graninger W, Thalhammer F, Locker G (1991) Pentoxifylline incerebral malaria. J Infect Dis 164:829

Kremsner PG, Grundmann H, Neifer S, Sliwa K, Sahlmuller G,Hegenscheid B, Bienzle U (1991) Pentoxifylline prevents murinecerebral malaria. J Infect Dis 164:605–608

Landau Z, Attali M (1993) Use of pentoxifylline in the treatment ofcerebral malaria. Clin Infect Dis 17:137

Lehman LG, Vu-Quoc B, Carlson J, Kremsner PG (1997) Plas-modium falciparum: inhibition of erythrocyte rosette formationand detachment of rosettes by pentoxifylline. Trans R Soc TropMed Hyg 91:74–75

Leonhardt H, Grigoleit HG (1977) Effects of pentoxifylline on redblood cell deformability and blood viscosity under hyperosm-olar conditions. Naunyn Schmiedebergs Arch Pharmacol299:197–200

Springer TA (1990) Adhesion receptors of the immune system.Nature 346:425–434

Stoltenburg-Didinger G, Neifer S, Bienzle U, Eling WM, KremsnerPG (1993) Selective damage of hippocampal neurons in murinecerebral malaria prevented by pentoxifylline. J Neurol Sci114:20–24

Vuong PN, Richard F, Snounou G, Coquelin F, Renia L,Gonnet F, Chabaud AG, Landau I (1999) Development ofirreversible lesions in the brain, heart and kidney followingacute and chronic murine malaria infection. Parasitology119:543–553

Ward A, Clissold SP (1987) Pentoxifylline. A review of its phar-macodynamic and pharmacokinetic properties, and its thera-peutic efficacy. Drugs 34:50–97

349