A multiplex assay for the sensitive detection and quantification of male and female Plasmodium falciparum gametocytes

Abstract

Background: The transmission of malaria to mosquitoes depends on the presence of gametocytes that circulate in the peripheral blood of infected human hosts. Sensitive estimates of the densities of female gametocytes (FG) and male gametocytes (MG) may allow the prediction of infectivity to mosquitoes and thus a molecular estimate of the human infectious reservoir for transmission.

Methods: A novel multiplex qRT-PCR assay with intron-spanning primers was developed for the parallel quantification of FG and MG. CCp4 (PF3D7_0903800) transcripts specific for FG and PfMGET (PF3D7_1469900) transcripts specific for MG were quantified in total nucleic acids. The assay was validated on sex-sorted gametocytes from culture material and on samples from clinical trials with gametocytocidal drugs. Synthetic RNA standards were generated for the two targets genes and calibrated against known gametocyte quantities.

Results: The limit of detection was determined at 0.1 male and 0.1 female gametocyte/µL, which was equal to the limit of quantification (LOQ) for MG, while the LOQ for FG was 1 FG/µL. Results from previously reported clinical trials that used separate gametocyte qRT-PCR assays for FG (targeting Pfs25) and MG (targeting PfMGET) were reproduced with the multiplex assay. High levels of agreement between separate assays and the multiplex approach were observed (R² = 0.9473, 95% CI 0.9314-0.9632, for FG measured by transcript levels of Pfs25 in qRT-PCR or CCp4 in multiplex; R² = 0.8869, 95% CI 0.8541-0.9197, for MG measured by PfMGET in either single or multiplex qRT-PCR). FG and MG transcripts were detected in pure ring stage parasites at 10,000- and 100,000-fold reduced frequency for CCp4 and PfMGET, respectively. The CCp4 and PfMGET transcripts were equally stable under suboptimal storage conditions.

Conclusions: Gametocyte densities and their sex ratios can be determined in the presented one-step multiplex assay with higher throughput than single assays. The interpretation of low gametocyte densities at asexual parasite densities above 1000 parasites/µL requires caution to avoid false positive gametocyte signals from spurious transcript levels in ring stage parasites.

Fig. 1

Limit of detection and variation of the gametocyte multiplex assay. Individual results of 8 technical replicates for female-sorted (a) and male-sorted (b) gametocytes of decreasing densities. Seven measurements (dots) obtained by calculating the reference from one randomly chosen dilution series. The coefficient of variation (filled bars) was calculated as standard deviation/mean of the respective calculated densities. Sample positivity in n = 26 independent experiments is indicated in the white boxes

Fig. 2

Stage-specificity of the multiplex assay. Female-specific (a) and male-specific (b) transcripts are detected in high concentrations of ring stage parasites of the F12 line and the 3D7/AP2-G-GFP-DDglmS line). Mean Ct values ± 2SD for varying numbers of independent experiments: n = 2 and n = 9 for CCp4 in ring stage parasites and female gametocytes; n = 2 and n = 17 for PfMGET in ring stage parasites and male gametocytes, respectively. NA not available/not measured (samples at these concentrations were not available), ND not detected. Lines represent fitted linear regression curves

Fig. 3

Sex-specificity of the multiplex assay. Female-specific (a) and male-specific (b) transcripts are detected in high concentrations of gametocytes of the opposite sex (empty symbols). Mean Ct values ± 2SD for varying numbers of independent experiments: n = 3 and n = 9 for CCp4 in male and female gametocytes; n = 4 and n = 17 for PfMGET in female and male gametocytes, respectively. ND not detected. Lines represent fitted linear regression curves

Fig. 4

Performance of the multiplex-assay on clinical trial samples. a, b Agreement with previously measured gametocyte densities with single assays, using Pfs25 (a) and PfMGET (b) transcripts. The linear regression was fitted as follows [95% confidence interval]: for FG y = 0.9944 [0.9497–1.040]x + 0.0751 [− 0.0627 to 0.2130] [28], y = 1.033 [0.977–1.092]x − 0.2554 [0.4619–0.04893] [15] and for MG y = 0.8626 [0.7744–0.9507]x + 0.5919 [0.2802–0.9037] [28], y = 0.9217 [0.8633–0.9800]x + 0.3508 [0.1515–0.5500] [15]. c Gametocyte prevalence as determined by single or multiplex qPCR for four treatment arms (Dicko et al. 2018), 7 days after treatment with either DP, dihydroartemisinin–piperaquine (n = 15); DP + MB, dihydroartemisinin–piperaquine + methylene blue (n = 19); SP–AQ, sulfadoxine–pyrimethamine and amodiaquine (n = 19) or SP–AQ and a single dose of primaquine (n = 19). d Sex ratios of gametocytes determined by multiplex or individual qPCR; samples as in c