Field-Based Flow Cytometry for Ex Vivo Characterization of Plasmodium vivax and P. falciparum Antimalarial Sensitivity

Ex vivo antimalarial sensitivity testing in human malaria parasites has largely depended on microscopic determination of schizont maturation. While this microscopic method is sensitive, it suffers from poor precision and is laborious. The recent development of portable, low-cost cytometers has allowed us to develop and validate a simple, field-optimized protocol using SYBR green and dihydroethidium for the accurate and objective determination of antimalarial drug sensitivity in freshly isolated Plasmodium vivax and Plasmodium falciparum.

M icroscopic examination of ex vivo matured malaria parasites remains the gold standard method used to determine the intrinsic sensitivity of fresh Plasmodium vivax and Plasmodium falciparum isolates to antimalarial drugs (1)(2)(3)(4)(5)(6)(7)(8). Ex vivo studies involve the manipulation of primary clinical samples of Plasmodium spp. in an artificial environment for no longer than 48 h. The modified WHO microtest assay is sensitive, relatively simple, and inexpensive and continues to be applied to a range of studies (9)(10)(11)(12)(13)(14)(15)(16)(17), especially those seeking novel antimalarial therapeutics against drug-resistant malaria (18)(19)(20)(21). However, the microscopic examination of Giemsa-stained thick films central to this method is tedious and time-consuming and requires skilled microscopists. Moreover, large inter-and intraobserver variations of parasite staging are frequently observed (7). Attempts to find an alternative ex vivo method suitable for both P. vivax and P. falciparum have been largely unsuccessful due to the high background noise present in clinical isolates (caused by a number of factors, including leukocytes, red blood cell autofluorescence, gametocytes, and contaminating protein signatures in host plasma) compared with the low target signal of the maturing parasite (clinical isolates frequently have parasitemias of Ͻ0.1%) (22)(23)(24). Perhaps the most objective and direct method to determine schizont maturation is the use of flow cytometry (25)(26)(27)(28). However, the high expense and fragility of most flow cytometers significantly limit their use in field laboratories. Fortunately, the recent development of relatively cheap, portable 2-laser flow cytometers (such as the Accuri C6; Becton, Dickinson) for the first time allows flow cytometric evaluation of ex vivo susceptibility assays in areas where malaria is endemic (29). Capitalizing on this new capability, we have developed a precise, accurate, fast, and simple flow cytometry (FC) method to conduct ex vivo drug sensitivity assays of P. vivax and P. falciparum under field conditions using only 2 colors.
Forty-eight isolates of P. vivax and 15 isolates of P. falciparum with parasitemias of between 0.02% and 0.5%, predominantly at the early ring stage (Ͼ80% of the total stages present), were collected from patients attending clinics at the Thai-Myanmar border (collected under the approved ethics protocol FMT-019-10 [Mahidol University, Faculty of Tropical Medicine Internal Review Board]). The isolates were transported to the Shoklo Malaria Research Unit (SMRU) field laboratory within 6 h of collection; the stages of parasitemia were assessed, and samples were then depleted of white blood cells (WBCs) by cellulose medium fiber (Sigma catalogue no. C6288) filtration as previously described (30) and cultured in the presence of 8 to 514 ng/ml of chloroquine diphosphate (molecular weight [MW], 515.9) (CQ) or 0.3 to 19 ng/ml sodium artesunate (MW, 406.4) (AS) using the protocol described by Russell et al. (8). At harvest (ϳ42 h postculture), the 200 l of blood medium in each well was mixed, and 20 l from each well was dispensed into a small curved-bottom tube (Micronic) and stained with 2 l of dihydroethidium (Sigma) and 5 l of SYBR green (made up with 63 l of phosphate-buffered saline [PBS]) (Sigma) and incubated for 20 min at room temperature. During the staining time, thick films (3 l packed red blood cells [RBCs]) were made from each of the wells for Giemsa staining and microscopic examination. The fluorescent staining reaction was stopped after 20 min with the addition of 400 l of PBS, and the reaction products were stored on an ice brick until FC analysis. The FC analysis was conducted using an Accuri C6 (Becton, Dickinson), and the gating strategy was per the method of Malleret et al. (29) (see Fig. 1A in the supplemental material). However, two special modifications were made to this protocol. First, only 60,000 events rather than 300,000 events were counted (reducing the count time per well from ϳ1.2 min to ϳ15 s). Note that for parasitemias less than 0.1%, we suggest using 100,000 events (see Fig. 1B and C in the supplemental material). Second, no CD45 staining was necessary, as Ͼ98% of the WBCs were removed from the isolates by cellulose. Slide counts for the The target gate representing schizont development events is indicated on each plot. Underneath the plots are the corresponding micrographs of Giemsa-stained thick films collected from the same culture wells. microscopy were conducted as described by Russell et al. (8). The proportion of events in the target gate (for cytometry) or the mature schizonts (for microscopy) at each of the treatments was normalized to that in the drug-free control. The proportion of schizont maturation at each corresponding drug concentration was then entered into the online ICEstimator (http: //www.antimalarial-icestimator.net/MethodIntro.htm), and the 50% inhibitory concentration (IC 50 ) was calculated by nonlinear regression analysis (31,32).
After 42 h of culture in the drug-free controls, the schizont "target gate" on the cytometer plot corresponding to the cluster of events with the highest levels of DNA (SYBR green, y axis) and RNA (dihydroethidium, x axis) can be clearly discerned for both species (Fig. 1), with the number of events in the FC plot target gate corresponding to the presence or absence of schizonts in the thick films (Fig. 1). As it is important to ensure that later-stage parasites and gametocytes initially present in the precultured isolate (time 0) do not confound the events in the schizont gate postculture, we ran an FC analysis on these initial samples. Any background events present near the target gate were then later subtracted from the events present at gates of the control and treatments postculture.
The culture success rate for both species was good, with 95.8% (46/50) and 86.7% (13/15) of P. vivax and P. falciparum samples, respectively, reaching at least 60% schizonts in the drug-free control after at least 42 h culture. Of the 46 successful P. vivax cultures, we were unable to model the IC 50 data for one of the CQ assays.
Bland-Altman analysis indicated good agreement between the methodologies (independent of drug type used) ( Fig. 2B and D). There was a slight bias toward higher IC 50 s with the flow cytometry method for both P. falciparum (Ϫ0.03 log 10 units) and P. vivax (Ϫ0.025 log 10 units).
In summary, the antimalarial sensitivity data for the new FC assay matched those of the traditional microscopy very closely. In the one case where there was a significant difference between the IC 50 analysis of FC and microscopy, the actual mean difference in AS IC 50 for P. vivax was less than 0.1 ng/ml, which is unlikely to be of biological significance. This 0.1-ng/ml disparity should also be put in the context of interreader variability between microscopists, which in our experience can be an order of magnitude greater. It should also be noted that the time to acquire data from the FC method is only ϳ2 min per drug (8 wells), compared to 18 min by microscopy. While assay described here used an extended exposure of AS (42 h), we have also used a more physiological 2-h "pulse exposure" of AS at the beginning of the FC assay to mimic the Ͻ1-h half-life pharmacokinetic profile of this drug in vivo (this results in an ϳ10-fold increase in the AS IC 50 [data not presented]). In conclusion, our data support the use of this simple FC protocol as a precise and more objective alternative to the microscopic determination of antimalarial drug sensitivity in fresh isolates of P. vivax and P. falciparum. Further studies involving a wider range of drugs are planned.