Effects of Plasmodium falciparum mixed infections on in vitro antimalarial drug tests and genotyping

Abstract

Studying drug resistance in Plasmodium falciparum requires accurate measurement of parasite response to a drug. Factors such as mixed infection of drug-resistant and -sensitive parasites can influence drug test outcome. Polymorphic DNA sequences are frequently used to detect mixed infections; infections with a single genotype or having a minor allele smaller than a subjectively selected cut-off value are often considered single infection. We studied the effects of mixed parasite populations containing various ratios of parasites resistant and sensitive to chloroquine on outcomes of drug tests and how ratios of parasite mixtures correlated with genotypes using polymerase chain reaction-based methods. Our results show that a mixture with a resistant population as low as 10% could greatly impact a drug test outcome. None of the genotyping methods could reliably detect minor DNA alleles at < or = 10%. Mixed infection presents a serious problem for drug tests, and genotyping using microsatellite or other methods may not reliably reflect true ratios of alleles.

Figure 1

Growth inhibition drug assays of parasite mixtures. Chloroquine-resistant parasites (Dd2, Dd2¹⁹², and FCB) were mixed with chloroquine-sensitive parasite HB3 at various ratios. Inhibition data of Dd2:HB3 ratios from 10:90 to 50:50 were plotted in (A); from ratios of 1:99 to 7.5:92.5 in (B); Dd2¹⁹²(D’):HB3 were plotted in (C) and (D); and FCB:HB3 were plotted in (E) and (F), respectively. All data points were repeated at least three times, and standard error bars indicate variations between tests performed at 48, 96, and 144 hr. Negative values were generated after subtraction from controls (no drug). A mixture of 10% of resistant parasites can greatly affect IC₅₀ of a sensitive parasite, whereas a mixture of 5% or less has minimum affects.

Figure 2

Typing mixed DNA samples using microsatellite (MS) markers. DNA samples from mixed parasite populations of Dd2/HB3, Dd2¹⁹²/HB3, and FCB/HB3, respectively, were typed with six MS markers (BM17, B5M5, B5M124, C14M17, C13M13, and C1M4) three times each. Signal ratios of Dd2, Dd2¹⁹² or FCB (minor alleles) over HB3 were plotted against ratios of mixed parasite populations (Dd2/Dd2¹⁹²/FCB:HB3; 1:99, 2.5:97.5, 5:95, 7.5:92.5, 10:90, 20:80, 30:70, 40:60, and 50:50). Different MS produced different estimates of DNA ratios.

Figure 3

Estimating mixed DNA samples using real-time quantative polymerase chain reaction (PCR) and pyrosequencing. Known DNA ratios from various parasite mixtures (Dd2 or FCB:HB3 at ratio: 1:99, 2.5:97.5, 5:95, 7.5:92.5, 10:90, 20:80, 30:70, 40:60, and 50:50) and the ratios estimated using real-time qPCR (A) and pyrosequencing (B) were plotted. ‘Standard’ in (A) is the ratios of known parasite proportions; ‘RT-qPCR’ indicates ratios estimated from real-time qPCR. The plots in (B) are from pyrosequencing of DNA mixtures of FCB:HB3 and Dd2:HB3, with ratios of Dd2:HB3 estimated twice ( Dd2-1 and Dd2-2).