The Role of Hemocytes in Anopheles gambiae Antiplasmodial Immunity

Abstract

Hemocytes synthesize key components of the mosquito complement-like system, but their role in the activation of antiplasmodial responses has not been established. The effect of activating Toll signaling in hemocytes on Plasmodium survival was investigated by transferring hemocytes or cell-free hemolymph from donor mosquitoes in which the suppressor cactus was silenced. These transfers greatly enhanced antiplasmodial immunity, indicating that hemocytes are active players in the activation of the complement-like system, through an effector/effectors regulated by the Toll pathway. A comparative analysis of hemocyte populations between susceptible G3 and the refractory L3-5 Anopheles gambiae mosquito strains did not reveal significant differences under basal conditions or in response to Plasmodium berghei infection. The response of susceptible mosquitoes to different Plasmodium species revealed similar kinetics following infection with P. berghei,P. yoelii or P. falciparum, but the strength of the priming response was stronger in less compatible mosquito-parasite pairs. The Toll, Imd, STAT or JNK signaling cascades were not essential for the production of the hemocyte differentiation factor (HDF) in response to P. berghei infection, but disruption of Toll, STAT or JNK abolished hemocyte differentiation in response to HDF. We conclude that hemocytes are key mediators of A. gambiae antiplasmodial responses.

Introduction

Hemocytes are important immune effector cells that are thought to participate in pathogen recognition, mediate fat body production of immune peptides and mediate and coordinate the mosquito systemic immune response [1, 2, 3]. Recently, hemocyte differentiation has also been shown to play an important role in innate immune memory responses in Anopheles gambiae mosquitoes that enhance the response to subsequent Plasmodium infections [4].

The Toll, Imd and JNK pathways have been shown to mediate antiplasmodial immune responses that target the ookinete stage of the parasite [5, 6, 7, 8] while the STAT pathway limits oocyst survival through activation of nitric oxide synthase [9]. For example, overactivation of the Toll pathway, by silencing the suppressor cactus, has been shown to result in a strong immune response that great ly reduces Plasmodium berghei infection in A. gambiae mosquitoes [5, 6]. The mosquito complement-like system mediates this response, because cosilencing of the thioester-containing protein 1 (TEP1) and cactus prevents the antiplasmodial effect of silencing cactus alone [5]. TEP1 is produced by hemocytes and is released into the mosquito hemolymph [10]; however, because systemic injection of double-stranded RNA (dsRNA) into the hemocoel silences cactus expression in many organs, the mechanism by which cactus silencing promotes TEP1 activation and the identity of the cells or tissue mediating this response have not been established.

The genetically selected A. gambiae L3-5 strain is refractory to infection with most Plasmodium parasites [11] and expresses a dominant allele of TEP1 that mediates the refractory phenotype [10, 12]. The refractory strain is in a chronic state of oxidative stress that is exacerbated after blood feeding [13], and recent studies revealed that refractory females have reduced longevity, faster utilization of lipid reserves, impaired mitochondrial state 3 respiration, increased rate of mitochondrial electron leak and higher expression levels of several glycolytic enzyme genes [14]. The status of hemocytes in the refractory strain and their potential participation in the antiplasmodial responses of this refractory strain has not been investigated.

An innate immune memory priming response is triggered in A. gambiae mosquitoes when Plasmodium ookinetes invade the midgut and disrupt the barriers that normally prevent direct contact between the microbiota and midgut epithelial cells. Plasmodium infection, in the presence of gut bacteria, triggers the release of a hemocyte differentiation factor (HDF) in the mosquito hemolymph that increases the granulocyte population of challenged mosquitoes and confers enhanced immunity to subsequent Plasmodium infections [4]. Furthermore, the transfer of cell-free hemolymph from challenged mosquitoes (which contains HDF) induces hemocyte differentiation in recipient mosquitoes and renders them more resistant to Plasmodium infection [4]. The signaling cascades involved in the induction of HDF synthesis or the response of hemocytes to HDF have not been established.

In this study, we explore two different aspects of the participation of hemocytes in antiplasmodial immunity: (1) the initial activation of the complement-like system in response to Plasmodium infection, and (2) the activation of hemocyte differentiation as part of a priming response that enhances immunity to subsequent infections. The strength of the priming response to infections with different Plasmodium species that have variable levels of compatibility with a given mosquito strain were compared, and a functional screen was carried out to establish the participation of different signaling pathways in HDF synthesis and on hemocyte responses to HDF.

Materials and Methods

Mosquito Maintenance and Plasmodium spp. Infections

A. gambiae (G3 and L3-5 strains) mosquitoes were reared at 27°C, 80% humidity with a 12-hour light/dark cycle, following standard laboratory conditions. P. berghei and P. yoelli infections were conducted using the P. berghei GFP-CON transgenic 259cl2 strain and GFP P. yoelii yoelii 17X nonlethal transgenic strain [15], respectively. Infections were maintained by serial passage in 3- to 4-week-old female BALB/c mice, and parasitemias were assessed by light microscopy from methanol-fixed blood smears that were stained with 10% Giemsa and air dried prior to visualization. Female mosquitoes were exposed to GFP P. berghei- or P. yoelii-infected mice that had a 3–6% parasitemia and with 1-2 exflagellations/field. In turn, P. falciparum infections were conducted as previously described [4]. Briefly, P. falciparum NF54 strain gametocyte culture was maintained in O⁺ human erythrocytes using RPMI-1640 medium supplemented with 25 mM, 50 mg/l hypoxanthine, 25 mM NaHCO3 and 10% (v/v) heat-inactivated type O⁺ human serum. A. gambiae (G3 strain) mosquitoes were fed on gametocyte cultures (stages IV and V, 14-16 days old) using artificial membrane feeders at 37°C for 30 min. Mosquitoes were then kept at 27°C and 80% relative humidity until the day of hemolymph collection.

Hemocyte Collection and Quantification

Hemocytes were collected as previously described [4]. Briefly, mosquitoes were injected into the thorax with an anticoagulant solution consisting of 60% Schneider's Insect medium, 10% fetal bovine serum and 30% citrate buffer. A fine incision was made between the last 2 abdominal segments using sterile fine-tip forceps, and additional anticoagulant buffer was added until 10 µl was recovered using a pipettman with a sterile pipette tip. For gene expression of the total hemocyte population, perfusions were collected in TRIzol reagent (Invitrogen-Life Technologies) and subjected to the method for gene expression analysis (see below). For quantification, the collected sample was deposited in a sterile disposable hemocytometer slide (10 µl capacity, Neubauer Improved, iNCYTO C-Chip DHC-N01). Hemocyte numbers were then counted using a light microscope (40× objective), differentiating the three hemocyte cell types (granulocytes, prohemocytes and oenocytoids) [16]. Total number of cells and proportions were then determined from these numbers for each individual mosquito. To evaluate any differences in hemocytes between the susceptible (G3) and refractory (L3-5) A. gambiae strains, hemolymph was collected at 2, 4, 6, 8, 10, 12 and 16 days after emergence in sugar-fed mosquitoes. To evaluate any species-specific effects of Plasmodium infection on hemocyte differentiation, female A. gambiae G3 strain mosquitoes were infected with either P. berghei,P. yoelii or P. falciparum, and their hemolymph was collected and hemocytes counted as described above.

Hemolymph (Cell-Free) and Hemocyte Transfers

For transfer of cells or cell-free hemolymph from silenced donors into naïve recipients, 1- to 2-day-old donor mosquitoes were injected with dsRNA against cactus or the non-target LacZ control. At 2-3 days after silencing, 2 µl hemolymph was perfused from each of 25-40 silenced donors using the modified anticoagulant buffer, collected using a siliconized pipet tip into siliconized microcentrifuge tubes on ice. Each sample was monitored by light microscopy for the presence of bacteria and discarded if bacteria were observed. Samples were centrifuged at 4°C, 6,000 rpms for 20 min, and the supernatant was moved to a new tube. Cells were resuspended in modified anticoagulant buffer. Then, 1- to 2-day-old recipient female mosquitoes were immediately injected with 207 μl of either cell-free hemolymph or cells in fresh buffer. Recipient mosquitoes were allowed to recover at 28°C for 3 days after transfer and were then either fed on a P. berghei-infected mouse, maintained, dissected and counted as indicated above or homogenized in RNAlater (Ambion-Life Technologies) and subjected to RNA extraction for gene expression analysis indicated below (fig. 1b, c).

Fig. 1.

Effect of activation of the Toll pathway in hemocyte antiplasmodial immunity. n.s. = Not significant; Cact = cactus-silenced control; LacZ = dsLacZ controls. a Total hemocyte cell count and proportions of granulocytes, oenocytoids and prohemocytes in individual mosquitoes from cactus-silenced and dsLacZ controls. Error bars indicate SEM. *** p < 0.001. b, c P. berghei oocyst numbers in mosquito midguts that received either washed cells or cell-free supernatant from cactus-silenced or dsLacZ control donors at 12 h before feeding (b) or 3 days before feeding (c). Horizontal red lines indicate median infection intensity. SN = Supernatant; Prev. = prevalence.

To assess whether the innate immune signaling pathways are required for the production of the HDF, four groups of mosquitoes were allowed to feed on the same P. berghei-infected mouse 3 days after silencing of the main pathway components. Two groups of mosquitoes (one control and one experimental) were placed at 28°C soon after blood feeding to prevent infection (naïve set) while the two other groups (challenged set) were maintained at 21°C for 48 h to proceed with normal midgut infection and then transferred to 28°C to reduce parasite load. Seven days after infection (dpi), the cell-free hemolymph was collected from the naïve and challenged sets and transferred into four recipient sugar-fed mosquito groups. Hemocyte collection and counts were then performed 4 days after transfer (fig. 4).

Fig. 4.

Effect of disrupting the Toll, Imd, STAT or JNK pathways on HDF production and release into the mosquito hemolymph. a Model depicting the process of parasite invasion of the midgut epithelial cells and the release of HDF following midgut injury and exposure to gut microbiota. b Proportion of granulocytes in sugar-fed mosquitoes that received either cell-free hemolymph from naïve (Nv) or challenged (Ch) dsLacZ controls or cactus, Rel1, Rel2 or JNK silenced mosquitoes. Hemolymph was collected 7 days after P. berghei infection and transferred into recipient mosquitoes to assess hemocyte differentiation. Proportion of granulocytes in mosquitoes that received cell-free hemolymph from naïve or challenged donors of dsLacZ or immune pathway-silenced backgrounds. Error bars indicate SEM. *** p < 0.001, ** p < 0.01, * p < 0.05.

To assess the effect of immune signaling pathway silencing on the hemocyte responses to HDF, cell-free hemolymph transfers were conducted as previously described [4]. In short, hemolymph was collected from naïve and challenged mosquitoes 7 days after priming by perfusion using a modified anticoagulant buffer (95% Schneider's medium and 5% citrate buffer). The collected hemolymph was then centrifuged at 4°C, 10,000 rpms for 10 min, and the cell-free supernatant was transferred to a new microcentrifuge tube. The cell-free hemolymph was then injected in a total volume of 210 μl per mosquito. Hemocytes in recipient mosquitoes were counted 4 days after transfer following the methodology described above (fig. 5).

Fig. 5.

Participation of immune signaling pathways in hemocyte responses to HDF. a Model depicting the effect of cell-free hemolymph transfer from challenged mosquitoes, which contains HDF (red dots), on the proportion of granulocytes (blue oval cells). b Proportion of granulocytes in dsLacZ controls or silenced-recipient mosquitoes after transfer of cell-free hemolymph from naïve (Nv) or challenged (Ch) donors. Error bars indicate SEM. *** p < 0.001, ** p < 0.01.

RNAi Silencing

RNAi-directed silencing of immune signaling pathways was performed as previously described [17]. In short, 2- to 3-day old female A. gambiae mosquitoes were cold anesthetized and injected with 69 μl of a 3-µg/µl dsRNA. The MEGAscript RNAi kit (Ambion) was used to synthesize the dsRNA solution using as a template cDNA from A. gambiae mosquitoes. Primers used for gene silencing have been previously designed and are listed in online supplementary table S1 (for all online suppl. material, see www.karger.com/doi/10.1159/000353765). RNAi-treated mosquitoes were then used for subsequent assays at 3 dpi. The whole-body silencing efficiency of cactus, Rel1, Rel2,STAT and JNK after dsRNA injection was, 80, 64, 45, 75 and 75%, respectively, relative to dsLacZ controls.

Gene Expression

Hemocytes were collected in TRIzol as indicated above and processed for RNA according to the manufacturer's instructions. Similarly, whole bodies were homogenized in RNAlater and processed for RNA using the RNeasy RNA extraction kit (Qiagen) according to the manufacturer's instructions. RNA from either method was assessed for concentration and quality using a NanoDrop (Thermo Scientific) and used in synthesis of cDNA using the QuantiTect reverse transcription kit with DNA Wipeout (Qiagen). Generated cDNA was used as template for real-time quantitative PCR assays using the DyNAmo SYBR green qPCR kits (Thermo Scientific) with gene-specific primers given in online supplementary table S1 and run using a CFX96 Real Time PCR Detection System (Bio-Rad). Data were collected and analyzed using the CFX96 software, and statistical significance was determined using Student's t test (GraphPad, La Jolla, Calif., USA).

Statistical Analysis

All statistical analyses were performed using GraphPad Prism 5 (GraphPad). Analyses derived from at least two independent biological replicates. Hemocyte numbers and proportion analyses were performed using Student's t test, and error bars represent the SEM. Significance was assessed at p < 0.05.

Results

Activation of the Toll Pathway in Hemocytes and Antiplasmodial Immunity

We explored the role of the Toll pathway specifically in the hemocyte compartment. The effect of overactivating the Toll pathway, by silencing the suppressor cactus, on the relative abundance of different hemocyte populations was investigated. cactus silencing significantly increased the proportion of oenocytoids from 34 to 51% (p < 0.0001, t test), with a concomitant reduction in the prohemocyte population from 62 to 46% (p < 0.0001, t test), relative to dsLacZ-injected controls (fig. 1a). However, no significant changes in the proportion of granulocytes or the total number of hemocytes were observed (fig. 1a). The injection of dsLacZ did not affect the proportion or the total number of hemocytes relative to uninjected controls (online suppl. fig. S1).

The hypothesis that overactivation of Toll signaling in hemocytes mediates the enhanced antiplasmodial response that is observed when cactus expression is silenced by systemic dsRNA injection was explored. Hemolymph from cactus-silenced mosquitoes was collected 2-3 days after dsRNA injection using the low-pressure flushing method previously described [4]. Cells were collected by centrifugation, washed with buffer and injected into recipient mosquitoes that received the equivalent of half of the hemocytes present in one mosquito. The cell-free hemolymph was also injected into recipient mosquitoes, and they received the equivalent of 1/10 of the hemolymph present in one mosquito. The effect of transferring cells or cell-free hemolymph from donors that had been injected with either dsLacZ or dsCactus on the susceptibility of the recipients was evaluated. The number of oocysts present at 7 dpi was significantly lower in mosquitoes that received the cell-free hemolymph from cactus-silenced donors 12 h before feeding, relative to the dsLaZ controls, but the transfer of cells had no effect on infection (fig. 1b). However, the transfer of either cell-free hemolymph or cells from cactus-silenced donors 3 days before feeding on an infected mouse greatly reduced P. berghei infection (fig. 1c). We confirmed that there was no difference in cactus expression levels in the mosquitoes that received either hemolymph or cells from a cactus-silenced donor (online suppl. fig. S2).

Because TEP1, a key component of Plasmodium lysis mediated by the complement-like system, is known to be synthesized by hemocytes [10], we explored whether the observed immune enhancement in the recipients could be due to an overall increase in TEP1 expression. However, cactus silencing significantly reduced TEP1 expression in hemocytes collected at 3 dpi, the time when they were transferred, relative to the dsLacZ controls (p < 0.05, t test; online suppl. fig. S3), indicating that the phenotype of transferring cactus-silenced hemocytes was not due to higher steady-state levels of TEP1 expression.

Hemocyte Populations in Susceptible and Refractory A. gambiae Strains

TEP1 is known to be a critical mediator of Plasmodium elimination in the A. gambiae refractory strain, and susceptible and refractory mosquitoes are known to express different alleles of TEP1 [10]. Furthermore, genetic studies revealed that TEP1-R1, the allele present in the refractory strain, has a dominant effect in P. berghei melanization. Because TEP1 is synthesized in hemocytes, we investigated whether the refractory (L3-5) and susceptible (G3) A. gambiae strains may differ in the distribution of their hemocyte populations or in the total number of cells. No significant difference in the proportion of hemocytes was observed between these mosquito strains in samples collected from sugar-fed females during the first 16 days after emergence (fig. 2a, online suppl. fig. S4). However, the total number of hemocytes was significantly lower in the refractory strain at 8 and 12 days after emergence (fig. 2b).

Fig. 2.

Hemocyte population in sugar-fed susceptible (S) and refractory (R) strain of A. gambiae and in response to P. berghei infection. a Proportion of granulocytes (Gr), prohemocytes (Pr) and oenocytoids (Oe) in the sugar-fed susceptible and refractory strain. b Total number of hemocytes in the susceptible and refractory strain of sugar-fed females during the first 16 days after emergence. c Granulocyte and prohemocyte proportions in the susceptible and refractory strain during the first 12 days after P. berghei infection. C = Uninfected controls; I = infected. Error bars indicate SEM. *** p < 0.001, ** p < 0.01, * p < 0.05.

Changes in hemocyte populations in response to P. berghei infection were also compared between susceptible and refractory females. Both refractory and susceptible strains exhibited a similar pattern of hemocyte population changes between infected and non-infected mosquitoes: a prominent increase in granulocytes beginning at 2 dpi that persisted up to 12 dpi (fig. 2c), concurrent with a significant reduction in prohemocytes that was more prominent at 2 and 6 dpi (fig. 2c). Neither refractory nor susceptible mosquitoes displayed a drastic difference in the proportion of oenocytoids in response to infection, and a slight reduction in the total number of hemocytes in the infected group was observed for both strains (online suppl. fig. S5).

A. gambiae Hemocyte Differentiation in Response to Infection with Different Plasmodium Species

We investigated whether the immune response to an initial infection with Plasmodium affects the strength of the hemocyte differentiation response that mediates immune priming and enhances the immune response to a subsequent infection. To this end, we compared the hemocyte proliferation response of susceptible G3 females following infection with P. berghei,P. yoelii and P. falciparum, three strains that differ in their compatibility with this mosquito strain. It has been previously established that silencing TEP1 has a dramatic effect on P. yoelii infection (increasing the median number of oocysts by 32 fold) [18], an intermediate effect on P. berghei infection (increasing the median number of oocysts by 4-6 fold) [19], while silencing TEP1 has no effect on infection with P. falciparum NF54 strain [20]. We found that infection with P. yoelii or P. falciparum induced a robust increase in granulocytes, very similar to what we observed with P. berghei infection, which was already apparent at 2 dpi and was still present at 12 dpi (fig. 2c, 3a). In general, the increase in granulocytes was more robust and consistent when susceptible females were infected with P. yoelii. Both P. berghei and P. yoelii infection also resulted in a significant reduction in prohemocytes, while only a slight reduction was observed 2 days after P. falciparum infection (fig. 3b). P. yoelii infection also resulted in a drastic and persistent reduction in the total number of hemocytes, while a milder and occasional reduction was observed after infection with P. berghei or P. falciparum (fig. 3c). Slight, occasional increases in oenocytoids were observed after P. yoelii infection (online suppl. fig. S6), but no significant changes were observed after P. berghei or P. falciparum infection (online suppl. fig. S5, S6).

Fig. 3.

Hemocyte differentiation in response to P. yoelii and P. falciparum infection. Granulocyte proportions (a), prohemocyte proportions (b) and total number of hemocytes (c) between uninfected controls (C) and infected (I) A. gambiae mosquitoes during the first 12 days following P. yoelii or P. falciparum infection. Error bars indicate SEM. *** p < 0.001, ** p < 0.01, * p < 0.05.

Immune Signaling Pathways and Induction of HDF in Mosquito Hemolymph

To assess whether innate immune signaling pathways are required for the induction of HDF synthesis following a challenge with P. berghei, we disrupted signaling of the Toll, Imd, STAT and JNK pathways (online suppl. fig. S7) by silencing Rel1, Rel2, STAT-A or JNK, respectively, before mosquitoes were infected. Two groups of dsLacZ-injected controls and silenced mosquitoes were fed on a P. berghei-infected mouse and placed at either a non-permissive temperature (naïve group) or permissive (challenged group) for parasite development in the mosquito. Cell-free hemolymph from each of the four silenced treatments was transferred into recipient mosquitoes (fig. 4a). Transfer of hemolymph from all the dsLacZ-injected control groups that were challenged resulted in a significant increase in granulocytes relative to the naïve controls (fig. 4b). Disruption of the Toll, Imd, STAT or JNK pathways did not affect the proliferation response (fig. 4b, online suppl. table S2), indicating that signaling from these pathways is not essential for the induction of HDF production and release into the mosquito hemolymph.

Effect of Immune Signaling Pathway Silencing on Hemocyte Responses to HDF

To assess whether the immune signaling pathways were essential for hemocyte differentiation in response to HDF, each of them was disrupted in the recipient mosquitoes by dsRNA-mediated silencing (fig. 5a). A robust response to HDF was observed in control recipients that were injected with dsLacZ (fig. 5b). However, mosquitoes in which the Toll, STAT or JNK pathways have been compromised no longer responded to the transfer of hemolymph from challenged mosquitoes that contained HDF, and the proportion of granulocytes was not significantly different from those that received hemolymph from naïve mosquitoes (fig. 5b, online suppl. table S3). In contrast, when hemolymph from challenged donors was transferred into mosquitoes with compromised Imd signaling (via Rel2 silencing), a robust increase in the proportion of granulocytes, similar to that of the dsLacZ-injected controls, was observed (fig. 5b, online suppl. table S3), indicating that the Imd pathway is not essential for hemocytes to respond to HDF.

Discussion

Prohemocytes are thought to be precursor cells that differentiate into granulocytes and oenocytoids [21, 22]. The concomitant reduction in prohemocytes when oenocytoids increase in response to cactus silencing suggests that the overactivation of Toll signaling is favoring the differentiation of the oenocytoid lineage. The fact that transfer of a small amount of hemolymph (1/10 of the hemolymph in the donor) from a mosquito in which cactus has been silenced greatly enhances antiplasmodial immunity suggests that overactivation of the Toll pathway results in the release of a soluble factor into the mosquito hemolymph that promotes activation of the complement-like system. This factor is probably produced by hemocytes, because if mosquitoes are infected 12 h after the transfer, there is no effect on infection, but if more time is given (3 days), the recipients of hemocytes with enhanced Toll signaling also mount a stronger antiplasmodial response. Taken together, this indicates that besides their role in TEP1 synthesis, hemocytes also play an active role in regulating activation of the complement-like system.

No significant differences in the proportion of the different hemocyte populations were observed between susceptible and refractory mosquitoes between 2 and 12 days after emergence. However, in general, the total number of hemocytes was lower in the refractory strain, and this difference was highly significant 12 days after emergence. The refractory strain is known to be in a state of chronic oxidative stress and to have reduced longevity [13, 14], suggesting these markers of accelerated aging/reduced fitness may also be correlated with the reduced number of hemocytes observed in refractory females. Both susceptible and refractory A. gambiae females responded to P. berghei infection with a strong and persistent increase in granulocytes and a modest decrease in prohemocytes, suggesting that P. berghei infection favors differentiation of the granulocyte linage. The lack of any differences in hemocyte lineage proportions between the susceptible (G3) and refractory (L3-5) strain of A. gambiae in the context of P. berghei infection suggests that the refractory phenotype is due to mechanisms other than the relative abundance of a subpopulation of hemocytes.

Comparative infections of susceptible females with different species of Plasmodium parasites revealed that P. berghei,P. yoelii or P. falciparum all triggered a significant and sustained increase in granulocytes and a reduction in prohemocytes. However, the magnitude of the responses appears to be proportional with the compatibility, that is, the extent to which the complement-like system limits infection, between each Plasmodium species and susceptible A. gambiae females [7]. The most compatible combination, P. falciparum-A. gambiae susceptible strain, had the mildest response, especially in terms of changes in the proportion of prohemocytes and total hemocytes, while the least compatible pair (P. yoelii in A. gambiae susceptible strain) exhibited the strongest and most persistent response. This suggests that, besides the disruption of the midgut barriers in the presence of the gut microbiota, the vector-parasite compatibility, which is hallmarked by the degree of activation of the complement-like system, also affects the hemocyte differentiation response.

Recent experiments revealed that the Pfs47 gene allows some P. falciparum parasites to evade the mosquito immune system [17, 20]. Expression of Pfs47 in wild-type NF54 parasites, the same line as used in this study, prevents the activation of epithelial responses to Plasmodium invasion, such as the induction of NOX5 and HPX2 expression, two enzymes that potentiate epithelial nitration and promote complement activation [20]. The observation that the priming response with P. falciparum NF54 was the weakest suggests that the ability to mount an effective immune response to the initial infection is key to establishing a robust priming response. Interestingly, P. falciparum infection in humans is a chronic disease, in which infections are usually not completely resolved, and multiple reinfections are common, due to an inability to mount effective immune memory responses. It is possible that other proteins from P. falciparum could also be modulating the initial immune response and limiting the ability of the human immune system to ‘remember’ this pathogen.

Our functional studies indicate that none of the immune signaling pathways tested (Toll, Imd, STAT and JNK) are essential for the induction of HDF synthesis and release into the mosquito hemolymph. However, the Toll, STAT and JNK pathways are key players in the hemocyte differentiation response to HDF. These signaling cascades are known to participate in hemocyte differentiation in Drosophila. For example, the Toll pathway has been shown to be involved in hemocyte proliferation and differentiation during larval hematopoiesis and following an immune challenge in Drosophila melanogaster flies [23, 24]. Constitutive activation of the STAT pathway results in overproliferation of hemocytes [25], and hemocytes release upd3, a cytokine that activates the STAT pathway in the fat body [26]. Furthermore, increasing reactive oxygen species levels in Drosophila hematopoietic progenitor cells activates JNK signaling and triggers premature differentiation into all mature cell types [27].

Whether HDF is activating these immune pathways directly in hemocytes and/or it is activating signaling in other organs that produce cytokines that, in turn, modulate hemocyte differentiation remains to be defined. Each immune signaling pathway could also control different and specific aspects of hemocyte differentiation. For instance, some of these pathways might be involved in controlling hematopoietic proliferation and differentiation [23], while others might regulate genes involved solely in hemocyte differentiation [28, 29].

We conclude that hemocytes play a central role in mediating and modulating antiplasmodial immunity. Besides their role in the synthesis of TEP1 and other key components of the complement-like system, they also play an active role in modulating activation of this immune effector mechanism. The strength of the hemocyte differentiation response to Plasmodium infection appears to be dependent on the level of compatibility between particular parasite-vector combinations. The synthesis and release of HDF in response to infection is under regulation of a new immune signaling pathway that remains to be defined, and the differentiation response to HDF is complex and involves multiple signaling cascades.

Supplementary Material

Supplementary data

Supplementary data

Supplementary data