PERSISTENCE OF ATOVAQUONE IN HUMAN SERA FOLLOWING TREATMENT: INHIBITION OF P. FALCIPARUM DEVELOPMENT IN VIVO AND IN VITRO

Drug treatment of volunteers.

Permission from the Ethics Committee of the Imperial College School of Medicine at St Mary’s Hospital, London, was obtained for drug treatment of volunteers, none of whom had previously experienced malaria. Three volunteers were bled for pretreatment serum and on the same day, day 0, were treated with atovaquone/ proguanil hydrochloride (four tablets, each containing 250 mg atovaquone and 100 mg proguanil daily as a single dose for 3 days); they were subsequently bled for serum 7, 14, 21, 28, 42, and 56 days after day 0. These sera had previously been tested against P. berghei.⁷ Sera also were obtained from two volunteers (A1 and A2) treated with atovaquone only (Wellvone™ [GlaxoWellcome Research Triangle Park, NC] at a dosage equivalent to that in atovaquone/proguanil) taken on day 0 and on days 7, 14, 21, 28 or 35, and 42 or 49 after treatment.

Transmission experiments: P falciparum culture.

The 3D7 clone of the NF54 strain of P. falciparum (provided by Prof. D. Walliker) was used in all transmission experiments. This was maintained in a semiautomated culture system developed by Ponnudurai et al.⁸ Briefly, parasites from asexual cultures grown in standard tissue culture flasks (Nunc, Roskilde, Denmark) in RPMI-1640 medium containing 25 mM Hepes, 2 g/L sodium bicarbonate (Sigma-Aldrich, Poole, UK) supplemented with 10% v/v human heat-inactivated AB serum, 50 mg/L hypoxanthine (Sigma-Aldrich), and gassed with 1% oxygen, 3% CO₂, and 96% nitrogen were diluted to 1% parasitemias with fresh red cells. Then, 750 μL of these infected red cells were added to each specially designed glass culture flask⁸ held on a cam-operated table that allowed medium (10 mL) to be changed automatically twice daily without removal from the incubator. Small samples were removed for testing for exflagellation^7,9,10 from day 10 onward. When enough mature gametocytes capable of exflagellation were present in cultures, the flasks were harvested, usually on days 11–17; flask contents were centrifuged in tubes prewarmed to 37°C containing 200 μL of fresh red cells previously washed once with phosphate-buffered saline (PBS). After centrifugation for 3 mins at 400g, supernatant culture medium was removed and the cell pellet mixed with respective sera to give a 40% hematocrit, keeping cells and sera at 37°C. Then 500-μL samples of the serum/parasitized red cells were placed in membrane feeders^7,9 also kept at 37°C by the passage of warm water.

Mosquito infections.

Five- to seven-day-old female Anopheles stephensi mosquitoes were given water only for 24h before infection. Pots containing approximately 50 mosquitoes were placed beneath the membrane feeders, and mosquitoes were allowed to feed for 30 mins (at 21–22°C ambient temperature). The mosquitoes were then held at 27°C, 80% relative humidity overnight, briefly anesthetized with ^CO2 ^while unfed mosquitoes were removed, and subsequently maintained in the same conditions on fructose/para-aminobenzoic acid.^7,9 On day 7, they were killed; then they were dissected and oocyst counts made on the midguts.

Infectivity was defined as the arithmetic mean oocyst number per mosquito in batches of mosquitoes fed post-treatment sera expressed as a percentage of the equivalent number in those fed pretreatment (day 0) sera;^7,10 data on sera of three atovaquone/proguanil-treated volunteers tested in four experiments were pooled and the means for each time point were expressed as the percentage of the mean for the day 0 sera as previously described.^7,10 Data from three experiments on one atovaquone-treated volunteer (A1) were pooled but there was only one experiment on A2 sera, with data only up to day 28.

The number of infected mosquitoes in each pot also was noted and was expressed as a percentage of the total number dissected; results for different experiments were again pooled and expressed as a percentage of the day 0 value.

Asexual stage experiments.

Three lines of P. falciparum were maintained as asexual stage cultures in standard tissue culture flasks: the 3D7 clone of NF54; a chloroquine-resistant line, K1 (provided by Prof. D. Warhurst); and an atovaquone-resistant isolate (NGATV01) recently isolated from a patient at St. Mary’s Hospital, London.¹¹ The medium and gas mixture were as described above but medium was renewed only once daily, and cultures were diluted with fresh red cells two to three times a week.

For the tests on sera from volunteers, parasitized red cells were diluted to a 1% parasitemia with fresh red cells (previously washed twice with PBS) and incubated for 48 h at a 2–4% hematocrit with 10% v/v day 0 or post-treatment sera in 100-μL cultures in 96-well trays. The trays were held in gas-tight chambers in the usual gas mixture. Parasitemias were counted blind on Giemsa-stained thin blood films, and the number of morphologically normal and abnormal rings, tro-phozoites, and schizonts was noted.

Sera were tested in triplicate and mean parasitemias expressed as a percentage of those on day 0; data from two to three experiments per strain were pooled and the final mean level of parasite growth expressed as a percentage of that at day 0.

Statistical treatment of results.

Data (mean oocyst counts, mosquito infections, or percentage parasitemias) from post-treatment sera were compared with data from day 0 sera using the Mann-Whitney U test.

Transmission experiments.

The numbers of fed mosquitoes per pot dissected 7 days after feeding usually ranged from about 20–40. In pots containing mosquitoes fed control day 0 sera from the atovaquone/proguanil-treated volunteers, the mean oocyst count per mosquito was 13.2 (range 0.3–33.2) and for the atovaquone serum controls, it was 7.1 (range 0.2–13.2). The mean percentage of infected mosquitoes for day 0 sera was 60.3 for atovaquone/proguanil treatment (range 13–83) and for atovaquone treatment was 53.8 (range 14–91).

Sera taken 4–28 days after atovaquone/proguanil treatment totally blocked infection of the mosquitoes, and significant inhibition of oocyst development (P = 0.05; Mann-Whitney U test) was still present on day 42. With day 42 sera, some oocysts were found (mean: 12.5% SEM ± 1.94, of day 0 value), and for day 56 sera, mean oocyst numbers were approaching those of day 0 (87.0% SEM ± 3.70 of day 0; Table 1). The percentage of infected mosquitoes (prevalence) was much reduced in pots containing post-treatment atovaquone/ proguanil sera up to day 42 (Table 1). Of a total of 315 mosquitoes dissected from all pots fed atovaquone/proguanil sera from day 4 to day 28, no oocyst-positive mosquito was found.

The prolonged inhibition of infectivity was due to atovaquone: sera taken from both volunteers treated with atovaquone alone totally inhibited oocyst development up to and including day 28 (Table 1). No infected mosquitoes were found in pots fed post-treatment atovaquone sera up to day 49 for volunteer A1, although a total of 309 were dissected. Sera from volunteer A2 inhibited infectivity up to day 28. (Ninety-five mosquitoes from post-treatment sera dissected were all negative, compared with 39 of 43 positive for day 0 serum.)

Asexual stage experiments.

After initiation of cultures at 1% and 48-hour culture at 37°C, parasitemias in wells containing day 0 sera ranged from 3–11% (mean 7.6% SEM ± 0.79) for all culture experiments. For the 3D7 clone, post-treatment atovaquone/proguanil sera were inhibitory up to day 28 (mean 11.4% SEM ± 0.3) of growth in day 0 sera, see Figure 1; P = 0.05 by Mann-Whitney U test in all experiments comparing parasitemias in wells containing sera up to and including day 28 with day 0 sera). The level of inhibition was less than that seen with the transmission experiments. This may reflect the 10% v/v of sera in the culture wells compared with the 60% v/v in membrane feeders, but it also should be noted that there is a natural decline in parasite numbers in the mosquito of 40- to 316-fold between macrogamete and ooki-nete.¹² However, on day 42, the mean parasitemia still was only 43.9% SEM ± 17.7 of the day 0 value. Of parasites present in wells containing day 7 to day 28 sera, 80–90% were abnormal (pycnotic, irregular cytoplasm, etc.). By day 42, this percentage had fallen to 16–22%. (Data not shown.) Tested against the chloroquine-resistant K1 strain of P. falciparum, a similar pattern of inhibition was observed, i.e., mean growth in wells containing day 7 to day 28 sera was less than 20% of the day 0 value, and parasites in these wells also exhibited abnormal morphology. (Data not shown.) Post-treatment sera from volunteer A1 were inhibitory up to and including day 49 when tested against asexual parasites of the 3D7 clone of P. falciparum (Figure 2): for day 49 sera, growth was a mean of 15.3% SEM ± 3.07 of day 0 sera, Figure 2. In all experiments comparing parasitemias in wells containing sera up to and including day 49 with day 0 sera, the differences were statistically significant (P = 0.05, Mann-Whitney U test). Sera from A2 were less inhibitory by day 28 (40% ± of growth in day 0 sera, Figure 3), though the difference was still significant (P = 0.05, Mann-Whitney U test).

Tested against resistant clone NGATV01, despite minor inhibition by some sera, there was no significant inhibition of growth by post-treatment sera from either of the atovaquone-treated donors (Figures 2 and 3), except that a late serum (day 42) from A2 reduced growth by 51%.