A human-serum-free medium can induce more infectious P. falciparum gametocytes than a conventional human-serum-containing medium

Abstract

Malaria remains a global health problem, and the standard membrane feeding assay (SMFA) is a key functional assay for development of new interventions to stop malaria transmission from human to mosquito. For SMFA, media with ~ 10% of human serum has been used for infectious gametocyte cultures, however, there are multiple challenges to obtain a suitable human serum. Here we show a human-serum-free culture medium (HSF), which was a mixture of two stem cell culture media and AlbuMAX, supported infectious gametocyte growth. Moreover, the HSF-induced gametocytes elicited significantly higher numbers of oocysts compared to gametocytes cultured with conventional human serum medium (Conv). While some caution is required when comparing percent transmission reducing activity data generated from HSF-SMFA and Conv-SMFA, the HSF method can facilitate the establishment of gametocyte cultures or SMFA by bypassing the need for human serum. Thus, this study will support future development of P. falciparum transmission-blocking interventions.

Keywords: Plasmodium Falciparum: malaria; Gametocyte; Gametocyte culture; SMFA; Serum free; Standard membrane-feeding assay; Transmission-blocking.

Fig. 1

Gametocytes cultured with human-serum-free medium (HSF) or conventional 10% human serum medium (Conv). (a) HSF and Conv medium conditions for gametocyte cultures were compared using Plasmodium falciparum NF54 parasites. The asexual culture which was maintained using Conv was split into two conditions when gametocyte cultures were initiated, then the two groups of parasites were maintained either with HSF or Conv for ~ 17 days. The two cultures were fed separately to two groups of the same batch of Anopheles stephensi mosquitoes. Eight days after the feed, n = 20 mosquitoes were dissected per group to enumerate oocysts in each mosquito. Oocyst number in each mosquito (dots), average numbers (bars and numbers above the graph), and a statistical difference calculated by a zero-inflated negative binomial (ZINB) model are shown. *, p < 0.05; **, p < 0.01; ***, p < 0.001. The results from two independent experiments are presented. (b) A similar experiment was conducted as done in a. The only difference was that in a, SR3A and MSC media were kept separately at 4 °C, and the two media were freshly mixed just prior to daily medium change, while in b, SR3A and MSC media were mixed on day 0 of gametocyte culture and the mixture was maintained at 4 °C until used for daily medium change.

Fig. 2

Difference between HSF and Conv. (a) Average number of oocysts in each experiment (dots) and median number of average oocysts from 10 independent experiments (numbers above the graph) are shown. Conv and HSF data tested in the same experiment are connected by the line. The total gametocytemia (b), stage V gametocytemia (c), female and male stage V gametocyte ratio (d), and number of exflagellation centers (e) in each group in each experiment on the day of feed are shown with median values from 10 experiments. In each panel, a statistical difference between two groups was calculated by a Wilcoxon matched-pairs signed rank test. n.s., not significant; **, p < 0.01.

Fig. 3

Difference in inhibitory activity between HSF and Conv. (a) 16 different antibodies were tested by SMFA using HSF and Conv gametocyte cultures side-by-side. Each antibody was tested at one to three concentrations in one or two independent assays, and a total of 34 paired results were generated. Each dot represents one paired result (error bars show 95% confidence intervals), and the red dotted line shows a y = x line. While the figure is drawn using a log of mean oocyst ratio (LMR; Log₁₀ (mean oocyst in control/mean oocyst in test)) scale, corresponding %TRA levels are shown for convenience. (b) bland Altman plots are shown. For each paired data set, an average LMR of the HSF and Conv conditions (x-axis), and a difference between the average and HSF (Dif LMR; y-axis) was calculated. A negative Dif LMR means that LMR in HSF was smaller than that in Conv. The same data are grouped by experiment (left) or by target antigen (right).