The Parasite Reduction Ratio (PRR) Assay Version 2: Standardized Assessment of Plasmodium falciparum Viability after Antimalarial Treatment In Vitro

Abstract

With artemisinin-resistant Plasmodium falciparum parasites emerging in Africa, the need for new antimalarial chemotypes is persistently high. The ideal pharmacodynamic parameters of a candidate drug are a rapid onset of action and a fast rate of parasite killing or clearance. To determine these parameters, it is essential to discriminate viable from nonviable parasites, which is complicated by the fact that viable parasites can be metabolically inactive, whilst dying parasites can still be metabolically active and morphologically unaffected. Standard growth inhibition assays, read out via microscopy or [³H] hypoxanthine incorporation, cannot reliably discriminate between viable and nonviable parasites. Conversely, the in vitro parasite reduction ratio (PRR) assay is able to measure viable parasites with high sensitivity. It provides valuable pharmacodynamic parameters, such as PRR, 99.9% parasite clearance time (PCT_99.9%) and lag phase. Here we report the development of the PRR assay version 2 (V2), which comes with a shorter assay duration, optimized quality controls and an objective, automated analysis pipeline that systematically estimates PRR, PCT_99.9% and lag time and returns meaningful secondary parameters such as the maximal killing rate of a drug (E_max) at the assayed concentration. These parameters can be fed directly into pharmacokinetic/pharmacodynamic models, hence aiding and standardizing lead selection, optimization, and dose prediction.

Keywords: Emax; PCT; PRR; Plasmodium falciparum; lag phase; limiting dilution; malaria; parasite reduction ratio; parasite viability; pharmacodynamics.

Figure 1

Comparison of drug activity after 72 h of drug exposure as determined by [³H] hypoxanthine incorporation (3HH, dark blue), microscopy (mic, blue), and by Parasite Reduction Ratio (PRR) assay (PRR, light blue) from three biological replicates with one, two, and four technical replicates for the microscopic-, [³H] hypoxanthine incorporation-, and the PRR assay, respectively. Resolution of drug activity is highest when measuring parasite viability using the PRR assay with compounds in order of decreasing activity: artemisinin, chloroquine, pyrimethamine, atovaquone. Artemisinin, chloroquine and pyrimethamine were examined at ten times the 50% inhibitory concentration (IC₅₀), atovaquone at a concentration of 100 nM.

Figure 2

The in vitro PRR assay version 2 (V2). (A) Experimental setup. Cultures of P. falciparum strain NF54 are incubated with compound at 10× IC₅₀, which was predetermined following 72 h of drug incubation. Drug is replenished every 24 h. At 0, 24, 48, 72, 96, and 120 h, a culture aliquot is taken and washed three times. Drug-free parasites are serially diluted and incubated in 96-well plates for 14 days in order to allow the viable parasites to resume growth to a measurable culture. Finally, the number of viable parasites present after washing can be extrapolated from the number of dilutions still yielding parasite-positive wells. An automated data analysis calculates pharmacodynamics (PD) parameters from the normalized raw data. In parallel, drug stability and washing efficiency are monitored with two bioassay-based quality controls and untreated parasite cultures serve as growth control. (B) Validation of the new experimental protocol with a shorter assay duration based on the reference compounds artemisinin, chloroquine, pyrimethamine and atovaquone. Killing profiles and PD parameters were obtained using the original PRR assay (V1, red) [6] or the PRR assay V2 (V2, blue) with a shorter assay duration. Drugs were tested at 10 × IC₅₀ (or 100 nM for atovaquone in V2) in ≥three independent experiments; error bars represent standard error of the mean (SEM) of ≥four technical replicates; Lag = lag time (h), PRR = parasite reduction ratio, PCT = 99.9% parasite clearance time (h).

Figure 3

Automated data analysis method. (A) Schematic representation of the model. The model considers a lag, linear and tail phase. The lag time (lag) is chosen from a set of potential lags based on the best model fit and marks the end of the lag phase. The slope ‘a’ of the linear phase is used to determine the log₁₀ (PRR) through multiplying by 48 h. The 99.9% parasite clearance time (PCT_99.9%) is the time when 99.9% of the parasites are killed, i.e., when the log₁₀ number of viable parasites is 2. Note that not all killing profiles possess a lag and/or a tail phase. (B) Killing profiles and PD parameters for the reference compounds artemisinin, chloroquine, pyrimethamine and atovaquone as determined by the new analysis method of PRR assay V2. Drugs were tested at 10 × IC₅₀ (or 100 nM for atovaquone in V2) in ≥three independent experiments. Error bars represent SEM of ≥four technical replicates and ribbons represent the 95% confidence interval; Lag = lag time (h), PRR = parasite reduction ratio, PCT = 99.9% parasite clearance time (h), E_max = maximal killing rate (h⁻¹); square brackets indicate the 24 h range (for lags) or the 95% confidence interval.

Figure 4

Validation of the new analysis method. PD parameter as obtained with PRR assay V1 or V2 analysis were compared for 232 molecules. (A) Comparison for lag time. (B) Comparison for PRR and (C) PCT_99.9%: the dashed, grey line indicates identity of the two methods, the red line is the linear regression with the equation and R² value displayed in the plot. (D) Demonstration of the effect on PRR values when extrapolating the slope between 0 and 24 h measurement to 48 h (blue) versus when not extrapolating (red). Not extrapolating results is underestimating the observed drop in parasites. (E) Linear regression for PRR when excluding all molecules with extrapolated PRR (PRR > 5) from the data set (n = 178).