High-throughput analysis of the transcriptional patterns of sexual genes in malaria

Abstract

Background: Plasmodium falciparum (Pf) is the leading protozoan causing malaria, the most devastating parasitic disease. To ensure transmission, a small subset of Pf parasites differentiate into the sexual forms (gametocytes). Since the abundance of these essential parasitic forms is extremely low within the human host, little is currently known about the molecular regulation of their sexual differentiation, highlighting the need to develop tools to investigate Pf gene expression during this fundamental mechanism.

Methods: We developed a high-throughput quantitative Reverse-Transcription PCR (RT-qPCR) platform to robustly monitor Pf transcriptional patterns, in particular, systematically profiling the transcriptional pattern of a large panel of gametocyte-related genes (GRG). Initially, we evaluated the technical performance of the systematic RT-qPCR platform to ensure it complies with the accepted quality standards for: (i) RNA extraction, (ii) cDNA synthesis and (iii) evaluation of gene expression through RT-qPCR. We then used this approach to monitor alterations in gene expression of a panel of GRG upon treatment with gametocytogenesis regulators.

Results: We thoroughly elucidated GRG expression profiles under treatment with the antimalarial drug dihydroartemisinin (DHA) or the metabolite choline over the course of a Pf blood cycle (48 h). We demonstrate that both significantly alter the expression pattern of PfAP2-G, the gametocytogenesis master regulator. However, they also markedly modify the developmental rate of the parasites and thus might bias the mRNA expression. Additionally, we screened the effect of the metabolites lactate and kynurenic acid, abundant in severe malaria, as potential regulators of gametocytogenesis.

Conclusions: Our data demonstrate that the high-throughput RT-qPCR method enables studying the immediate transcriptional response initiating gametocytogenesis of the parasites from a very low volume of malaria-infected RBC samples. The obtained data expand the current knowledge of the initial alterations in mRNA profiles of GRG upon treatment with reported regulators. In addition, using this method emphasizes that asexual parasite stage composition is a crucial element that must be considered when interpreting changes in GRG expression by RT-qPCR, specifically when screening for novel compounds that could regulate Pf sexual differentiation.

Keywords: Automatization; Gametocyte; Gametocytogenesis; Gene expression; Malaria; Plasmodium falciparum; RT-qPCR.

Fig. 1

The robot-automated method yields high purity Pf RNA. a, b Spectrophotometric quantification of the RNA yield extracted from Pf-iRBC ring or trophozoite stage cultures using (i) robot-automated magnetic binding (rMB), (ii) solid phase extraction (SPE) or (iii) liquid–liquid extraction (LLE) methods. c, d Mean A₂₆₀/A₂₈₀ purity ratio determined spectrophotometrically for samples extracted with the same methods. e, f Mean RNA integrity number (RIN) calculated by capillary electrophoresis (TapeStation^® analysis). Panels a–d present the mean of three biological replicates (n = 3) with three technical replicates each. Panels e, f present the mean of three biological replicates (n = 3) with one technical replicate each. Error bars represent SD. A one-way ANOVA with Tukey’s multiple comparisons test was performed for panels a–d, and a single-value t-test of the mean RIN against the RIN value of 8.5 (considered minimum for high-quality RNA [71], represented as a horizontal line in the plot) was performed for panels e, f. ns = p ≥ 0.05 (non-significant), *p < 0.05, **p ≤ 0.01, ***p ≤ 0.001 and ****p ≤ 0.0001

Fig. 2

The robot-automated method enables precise gene expression analysis by RT-qPCR. a Gene panel selected to evaluate early gametocytogenesis-related gene (GRG) expression dynamics. The basal expression of each gene in the panel was evaluated for ring- and trophozoite-stage cultures, which continuously commit and generate a basal level of gametocytes expressing these genes. Technical repeatability was evaluated by the standard deviation value (error bars) of the Ct in three independent biological repeats (n = 3). b SPUD polymerase inhibition assay for cDNA synthesized by robot-automated methods from ring- and trophozoite-stage parasites. Artificial SPUD template at a 10^− 5 nM concentration was used as a non-inhibition control, and phenol was used as a positive inhibition control. No Amp. = no amplification (Ct ≥ 40). a–b Mean of three biological replicates (n = 3) with three technical replicates each. Error bars represent SD. A two-way ANOVA with Šidák’s multiple comparisons test was performed for Panel b. ns = p ≥ 0.05 (non-significant)

Fig. 3

Dihydroartemisinin (DHA) alters the PfAP2-G expression profile and inhibits parasitic development. a Schematic illustration of the experimental setup: the transcriptional profiles of five GRG markers were built for synchronized NF54 trophozoite-stage parasites treated over the course of one 48-h cycle after a short (3-h) pulse with DHA (5 or 10 nM) or 0.1% DMSO (solvent control) or non-treated (NT) control. The flask and arrow represent treatment induction starting at the first time point (0 hpt). RNA was extracted by the robotic unit at 4, 12, 25, 30 and 48 hpt for RT-qPCR expression analysis. The predominant stage in each time span is also presented in the illustration for comparison with the transcriptional profiles. b–f Temporal transcriptional profiles of the early gametocytogenesis markers PfAP2-G (b), gexp05 (c), Pfg14.748 (d), Pfs16 (e) and sbp1 (f, a ring-stage marker). g Transcript levels of Pf GRGs: PfAP2-G, gexp05, Pfg14.744, Pfg14.748, Pfs25 and sbp1, 24 h post-DHA induction under 2 mM choline treatment. h Growth assay for NF54 Pf-iRBCs post-treatment with DHA (5 nM or 10 nM) for 3 h (trophozoite stage), monitored by flow cytometry. The parasitemia was quantified every 24 h over a 72-h period. The respective volumes of DMSO (0.05% and 0.1%) were used as solvent controls, and non-treated (NT) parasites were used as negative control. The flow cytometry gating strategy for a representative example can be observed in Additional file 14: Fig. S11. In Panels b–f and h, data represent the mean of three independent biological repeats (n = 3) with three technical repeats each, while in Panel g, data represent the mean of five independent biological repeats (n = 5) for all the genes except for sbp1 (4 biological repeats, n = 4). Mean transcript levels were calculated using the relative standard curve method and normalized to the transcript levels of uce in Panels b–f. Error bars represent SD. Two-way ANOVA with post-hoc tests were run using estimated marginal means with the R package 'emmeans' for Panels b–f and h. An unpaired t-test was performed between the DHA-treated and non-treated parasites for each gene individually in Panel g. *p < 0.05, **p ≤ 0.01 and ****p ≤ 0.0001

Fig. 4

Choline represses PfAP2-G expression early in gametocytogenesis commitment. a Schematic illustration of the experimental setup: the transcriptional profiles of nine GRG markers were built for synchronized NF54-gexp02-tdTomato trophozoite-stage parasites treated under conditions of choline removal (- Choline) or choline supplementation (+ Choline). The flask and the arrow represent treatment induction (± Choline) starting at the first time point (0 hpt). RNA was extracted by the robotic unit at 6, 12, 18, 24, 30, 36, 42 and 48 hpt for RT-qPCR expression analysis. In parallel, parasitemia, % gexp02⁺ iRBC (proxy for the percentage of sexually committed cells) and stage composition in the culture were monitored by flow cytometry and Giemsa-stained smears (Additional file 13: Fig. S10). The predominant stage in each time span is also presented in the illustration for comparison with the transcriptional profiles. b–f Temporal transcriptional profiles of the ring and schizont stage-specific markers sbp1 (b), rhoph2 (c), the sexual commitment marker PfAP2-G (d), the sexual ring markers gexp05 and gexp02 (e), and the early gametocyte markers Pfg14.744, Pfg14.748 and Pfs16 (f) at the selected time points within 48 h. Mean transcript levels were calculated using the relative standard curve method and normalized to the transcript levels of uce in panels b–f. The smoothened trend line in the data was generated using a LOESS (locally weighted scatterplot smoothing) function with a span of 0.75 for the three independent biological repeats (n = 3). Two-way ANOVAs with post-hoc tests were run using estimated marginal means, with the R package ‘emmeans’ for all genes. *p < 0.05, **p ≤ 0.01, ***p ≤ 0.001 and ****p ≤ 0.0001

Fig. 5

Severe malaria metabolites do not immediately alter gametocytogenesis regulation. Gene expression analysis during the late schizont stage for the gametocytogenesis master regulators PfAP2-G (a), gexp02 (b), sbp1 (c) and rhoph2 (d). In the case of PfAP2-G, the NF54-gexp02-tdTomato line was used as a control to illustrate choline treatment success (a, right panel). NF54 Pf parasites were cultured under choline removal (− Choline) or supplementation (+ Choline) conditions and exposed to lactate 5 mM or kynurenic acid 250 nM for 16 h. RNA was extracted at this time point only and gene expression analyzed by robot-automatized RT-qPCR. Data were normalized to uce. Error bars represent SD. Two-way ANOVA with post-hoc tests were run using estimated marginal means with the R package 'emmeans.' The significance presented in the figure corresponds to the two-way ANOVA for the “choline presence” factor. All the individual post-hoc contrasts were not significant, and not shown. ns = p ≥ 0.05 (non-significant), *p < 0.05 and ****p ≤ 0.0001

Fig. 6

Schematic illustration of the robot-automated RT-qPCR platform for monitoring Pf gene expression profiles. The goal of the automated RT-qPCR platform is to simultaneously evaluate multiple conditions that may affect gene expression regulation. First, Pf-iRBCs are cultured under different treatments induced by robot-automated culturing (work in progress), incubation and Giemsa smear monitoring (Step 1). RNA is then extracted from Pf-iRBCs by a completely automated high-throughput method, up to 96 samples simultaneously (Step 2). cDNA is next prepared from up to 96 samples simultaneously by robot-automated RNA dilution and reverse-transcription master mix aliquoting (Step 3). Finally, 384-well RT-qPCR plates are prepared using the robotic system by aliquoting the cDNA solutions and the RT-qPCR reaction mix (Step 4)

Fig. 7

Proposed model of the mixed effects of choline on GRG expression and growth. In choline-depleted parasites (upper panel), a certain percentage of asexual parasites commit to sexual differentiation (green lines) and express higher levels of PfAP2-G, mainly at the late schizont stage (12–18 hpt), compared to choline-supplemented parasites (lower panel), in which the expression of PfAP2-G is effectively repressed. This was correspondingly observed in the RT-qPCR assay performed over the extracted RNA at this stage (Fig. 4b). After schizont rupture at 24–36 hpt, the sexual rings (green lines) derived from these PfAP2-G⁺ committed schizonts were initially expected to express their specific GRGs (gexp02 and gexp05), as if choline would affect PfAP2-G expression only. However, choline supplementation increases merozoite productivity during schizogony (represented by schizonts with a higher number of daughter merozoites in the diagram), yielding a significantly higher level of parasitemia and a significantly higher proportion of ring-stage parasites in the culture at 24–36 hpt (Additional file 13: Fig. S10). Since bulk techniques such as RT-qPCR analyze the differences in transcript abundance of the total parasitic population at a given time point, RNA extracted from the pool of cells is an overall mixture of the proportional contributions of each one of the developmental stages present in the sample. Assuming choline increases merozoite productivity equally in asexual and sexual parasites, the extracted RNA at 24–36 hpt had a much higher proportion of both sexual and asexual ring RNA in the choline-supplemented samples. Thus, RT-qPCR evaluation of GRG expression 24–36 hpt demonstrated a higher abundance of GRG transcripts under conditions of choline supplementation. The choline-mediated % gexp02⁺ iRBC reduction was not presented in this model, as it was only found to be significant after 48 hpt (Additional file 13: Fig. S10b)