Acidocalcisomes of Trypanosoma brucei have an inositol 1,4,5-trisphosphate receptor that is required for growth and infectivity

Abstract

Acidocalcisomes are acidic calcium stores rich in polyphosphate and found in a diverse range of organisms. The mechanism of Ca²⁺ release from these organelles was unknown. Here we present evidence that Trypanosoma brucei acidocalcisomes possess an inositol 1,4,5-trisphosphate receptor (TbIP₃R) for Ca²⁺ release. Localization studies in cell lines expressing TbIP₃R in its endogenous locus fused to an epitope tag revealed its partial colocalization with the vacuolar proton pyrophosphatase, a marker of acidocalcisomes. IP₃ was able to stimulate Ca²⁺ release from a chicken B-lymphocyte cell line in which the genes for all three vertebrate IP₃Rs have been stably ablated (DT40-3KO) and that were stably expressing TbIP₃R, providing evidence of its function. IP₃ was also able to release Ca²⁺ from permeabilized trypanosomes or isolated acidocalcisomes and photolytic release of IP₃ in intact trypanosomes loaded with Fluo-4 elicited a transient Ca²⁺ increase in their cytosol. Ablation of TbIP₃R by RNA interference caused a significant reduction of IP₃-mediated Ca²⁺ release in trypanosomes and resulted in defects in growth in culture and infectivity in mice. Taken together, the data provide evidence of the presence of a functional IP₃R as a Ca²⁺ release channel in acidocalcisomes of trypanosomes and suggest that a Ca²⁺ signaling pathway that involves acidocalcisomes is required for growth and establishment of infection.

Intracellular Ca²⁺ serves as a second messenger for a variety of cell functions, including secretion, contraction, cell division, and differentiation (1). Cells use two sources of Ca²⁺ for generating signals: Ca²⁺ release from intracellular stores and Ca²⁺ entry across the plasma membrane. Ca²⁺ release from intracellular stores of mammalian cells is controlled by at least three groups of channels: ryanodine receptors, located in the endoplasmic reticulum (ER) and stimulated by cyclic ADP ribose (cADPR) (2); inositol 1,4,5-trisphosphate receptors (IP₃R), also mainly located in the ER and stimulated by IP₃ (3); and two pore channels, preferentially located in acidic calcium stores and stimulated by NAADP (4, 5). IP₃ is generated by hydrolysis of phosphatidylinositol 4,5-bisphosphate (PIP₂) catalyzed by phosphoinositide-specific phospholipases C (PI-PLCs) (6), which also generate diacylglyerol, whereas ribosyl cyclases catalyze the generation of both cyclic ADP ribose and NAADP from NAD (7). Orthologs to genes encoding some of these enzymes and channels are present in protists (8, 9), suggesting an early appearance of Ca²⁺ signaling during evolution.

Cytosolic Ca²⁺ in trypanosomatids such as Trypanosoma cruzi, the etiologic agent of Chagas disease, and Trypanosoma brucei, which belongs to the group of parasites that causes African trypanosomiasis or sleeping sickness, is maintained through the concerted operation of distinct Ca²⁺ transporting systems located in the plasma membrane, ER, mitochondria, and the acidic calcium stores known as acidocalcisomes (10). The IP₃/diacylglycerol pathway was described in T. cruzi more than 20 y ago (11), whereas IP₃ was also detected in T. brucei (12). T. cruzi PI-PLC has been studied in more detail and was shown to be rather unusual in that it does not have a pleckstrin homology domain to bind to the plasma membrane, has a highly charged region between the catalytic X and Y domains, and is N-myristolyated and palmitoylated (13). This lipid modification is important for plasma membrane localization and for stimulation of differentiation of the infective trypomastigotes into amastigotes (14, 15). A gene ortholog to that encoding TcPI-PLC has been found in T. brucei, and the protein product appears to have similar domains to those of the T. cruzi enzyme. Despite the presence of a PI-PLC and the detection of IP₃ in both T. cruzi and T. brucei, early attempts to detect Ca²⁺ release by IP₃ in permeabilized cells were unsuccessful (12, 16).

With the completion of several genome projects, it was possible to use bioinformatic methods to identify genes encoding putative IP₃ receptors (8, 17–19) and other channels (9) in protists. Interestingly, apart from one study that described a functional IP₃ receptor in Paramecium tetraurelia (18), biochemical evidence that the other genes identified encode for functional IP₃ receptors is still lacking. Proteomic data from enriched contractile vacuole fractions of T. cruzi (20) included spectra from the putative IP₃Rs, suggesting that these proteins are expressed in subcellular fractions (20).

In this work, we report the characterization of the IP₃R from T. brucei, which localizes to their acidocalcisomes. IP₃ was able to release Ca²⁺ from permeabilized cells and isolated acidocalcisomes, and UV light photolysis of caged IP₃ produced a significant increase in intracellular Ca²⁺ in live trypanosomes. Knockdown of the TbIP₃R gene by RNA interference (RNAi) in bloodstream form (BSF) trypanosomes led to growth defects and reduced infectivity in vivo underscoring the relevance of this channel in the parasite.

Results

Characterization of TbIP₃R.

One gene (Tb927.8.2770) encoding for a putative IP₃R was found in the T. brucei genome at the TriTryp database (http://tritrypdb.org/tritrypdb/) and named TbIP₃R. The full-length cDNA of TbIP₃R was cloned by PCR amplification and confirmed by sequencing as described in SI Materials and Methods. The orthologs identified in T. cruzi (TcCLB.509461.90) and Leishmania major (LmjF.16.0280) shared 41% and 19.3% amino acid identity, respectively, to TbIP₃R. Structural analysis (ELM and TMHMM servers) predicted five transmembrane domains in the C-terminal region between amino acids 2615–2637, 2649–2671, 2730–2752, 2771–2795, and 2862–2884. The ORF predicts a 3,099-amino-acid protein with an apparent molecular weight of 345 kDa. TbIP₃R possesses a series of conserved domains including putative suppressor domain-like (SD), ryanodine receptor IP₃R homology (RIH), and RIH-associated (RIAD) domains (9). A motif for a Ca²⁺-specific selectivity filter (GVGD) (21) is present in the putative intraluminal loop between transmembrane domains at the C-terminal region. Of the 10 residues that have been proposed to form a basic pocket that binds IP₃ (22, 23), 4 are conserved in TbIP₃R. Other features of trypanosomatid putative IP₃Rs have been described previously (9).

TbIP₃R Localization in T. brucei.

To investigate the localization of T. brucei IP₃R, the C terminus was tagged in procyclic (PCF) trypomastigotes with an HA tag using homologous recombination with the endogenous gene locus. Western blot analysis confirmed expression of proteins of the expected size (Fig. S1A). The TbIP₃R partially localized to acidocalcisomes, as demonstrated by colocalization with antibodies against T. brucei acidocalcisomal marker vacuolar proton pyrophosphatase (TbVP1) (24) (Fig. 1A). An additional punctate staining of TbVP1 that did not colocalize with TbIP₃R was also detected (Fig. 1A) and could correspond to trafficking vesicles. In this regard, it has been described that the adaptor protein-3 (AP-3) complex is involved in sorting proteins, like TbVP1, to acidocalcisomes from the Golgi or from endosomes in both L. major (25) and T. brucei (26). No colocalization with TbBiP, an ER marker (27) with a clear reticular labeling (Fig. 1B), was detected, ruling out ER localization of the TbIP₃R.

Fig. 1.

Localization of TbIP₃R in PCF trypanosomes. (A) TbIP₃R partially colocalizes with TbVP1 in acidocalcisomes (Pearson’s correlation coefficient of 0.874), as shown by immunofluorescence microscopy analysis. The merge images show the colocalization in yellow. (B) Lack of colocalization of TbIP₃R with TbBiP in the endoplasmic reticulm (Pearson’s correlation coefficient of 0.156). (Scale bars, 10 µm.) DIC, differential interference contrast.

TbIP₃R Expression in DT40 Cells and Response to IP₃.

To study whether Ca²⁺ is released on stimulation of the TbIP₃R by IP₃, we stably transfected the gene encoding this receptor and that of the rat IP₃R1 (RnIP₃R1), used as a positive control, in a chicken B-lymphocyte cell line (DT40) in which the genes for all three vertebrate IP₃Rs were stably ablated (DT40-3KO) (28). DT40-3KO cells were used as negative controls. Heterologous expression of the TbIP₃R and RnIP₃R1 was confirmed by immunocytochemistry (Fig. 2C) and by Western blot analyses (Fig. S1 B and C). The results shown in Fig. 2C are consistent with the localization of both TbIP₃R and RnIP₃R1 to the ER of DT40 cells. A low-affinity Ca²⁺ indicator (Mag-Fluo-4) was trapped within the ER to measure luminal free [Ca²⁺] in saponin-permeabilized DT40 cells. Addition of MgATP (1.5 mM) stimulated Ca²⁺ uptake until a steady-state Ca²⁺ loading was reached (Fig. 2A). The time course of Ca²⁺ loading was similar for DT40-3KO, DT40-WT, DT40-3KO(RnIP₃R1), and DT40-3KO(TbIP₃R) cells, but only the last three responded to IP₃ with a decrease in Mag-Fluo-4 fluorescence. In DT40-3KO(RnIP₃R1) and DT40-WT cells, a maximal concentration of IP₃ (10 µM) caused the fluorescence to decrease by 81.2 ± 4.9%, whereas the same IP₃ concentration caused the fluorescence of DT40-3KO(TbIP₃R) cells to decrease slowly and by 61.7 ± 3.6%. There was a concentration-dependent release of Ca²⁺ by IP₃ in all cases (Fig. 2B), and TbIP₃R was considerably less sensitive to IP₃ than RnIP₃R1 (Fig. 2B). In addition, the IP₃ agonist adenophostin A (0.5 µM) showed comparable Ca²⁺ release to 10 µM IP₃ in the TbIP₃R-expressing DT40-3KO cells (Fig. 2B, Inset). In contrast, addition of 10 µM cADPR or 1 µM NAADP did not stimulate Ca²⁺ release (Fig. 2B, Inset).

Fig. 2.

IP₃-evoked Ca²⁺ release in permeabilized DT40 cells. (A) Fluoresence changes evoked by addition of MgATP (1.5 mM) and IP₃ (10 µM) to DT40-WT (wt), DT40-3KO (ko), DT40-3KO(TbIP₃R) (TbIP₃R), and DT40-3KO(RnIP₃R1) (RnIP₃R) cells. An increase in fluorescence indicates decreasing medium Ca²⁺ or increasing vesicular Ca²⁺. (B) Effect of different concentrations of IP₃ on Ca²⁺ release by cells described in A. Inset shows Ca²⁺ release by IP₃ (10 µM), adenophostin A (Ada, 0.5 µM), cADPR (10 µM), or NAADP (1 µM). Values in B and Inset are means ± SD of three independent experiments. (C) TbIP₃R (DT40TbIP₃R)- and RnIP₃R1 (DT40RnIP₃R)-transfected DT40-3KO cells were stained with anti-HA (1:250) and anti-rat IP₃R1 (1:200), respectively, and analyzed by immunofluorescence microscopy. (Scale bar, 10 µm.) Left images are by DIC.

IP₃ Releases Ca²⁺ from Permeabilized T. brucei Cells and Isolated Acidocalcisomes.

To demonstrate the response of TbIP₃R in T. brucei, we used permeabilized PCF previously incubated with 0.1 mM PP_i (to acidify acidocalcisomes) and 1 mM ATP (to stimulate Ca²⁺ uptake by the Ca²⁺/H⁺ countertransporting ATPase of acidocalcisomes (29). Uptake and release of Ca²⁺ were measured with Calcium Green-5N. Digitonin addition caused a drop in Ca²⁺ due to ATP-driven Ca²⁺ uptake until a steady state was reached. Under these conditions, Ca²⁺ release was detected on addition of IP₃ (Fig. 3A) and was concentration dependent (Fig. 3B). In the absence of PP_i, no significant IP₃-evoked Ca²⁺ release was detected, as reported previously (12), suggesting that acidification of acidocalcisomes by TbVP1 was required for Ca²⁺ uptake by the Ca²⁺-ATPase located in the organelles, as reported before (29). Adenophostin A (Ada, 0.5 µM) had a comparable effect to IP₃ (10 µM), whereas ryanodine (Ry, 10 µM) had no effect of Ca²⁺ release (Fig. 3B). Fig. 3A shows that addition of nigericin, which is a K⁺/H⁺ exchanger that alkalinizes acidocalcisomes, released the amount of Ca²⁺ taken up in response to ATP addition, whereas further addition of ionomycin, which is a Ca²⁺ ionophore that releases Ca²⁺ from alkaline or neutral compartments, was able to release additional Ca²⁺, probably present in acidocalcisomes and the ER. A similar experiment was done with isolated acidocalcisome fractions of PCF. H⁺ transport activity by the H⁺-PPase was detectable in the purified acidocalcisome fraction using acridine orange, indicating that the organelles were intact (Fig. S2). In this assay, acridine orange accumulates and dimerizes in vesicles as they acidify, leading to changes in its fluorescence properties (30). As reported before (29), Ca²⁺ uptake using ATP was more efficiently detected after previous addition of PP_i to acidify the organelles (Fig. 3C). When 0.1 mM PP_i was added, there was an initial sharp drop in fluorescence due to Ca²⁺ chelation and then a steady-state rise as pyrophosphate was hydrolyzed and chelated Ca²⁺ was released. Subtraction of this rising background allowed the detection of Ca²⁺ uptake when ATP was added. Addition of 10 µM IP₃ led to a rise in the medium Ca²⁺, followed by return to a similar rate of Ca²⁺ uptake to that before IP₃ addition. The effect of IP₃ on Ca²⁺ release was concentration dependent (Fig. 3D). Addition of nigericin plus ionomycin released the Ca²⁺ taken up and additional Ca²⁺ stored in acidocalcisomes (Fig. 3C). It has been shown before that previous alkalinization with nigericin is needed for ionomycin to be able to release Ca²⁺ from acidocalcisomes (29).

Fig. 3.

Ca²⁺ uptake and release by digitonin-permeabilized T. brucei procyclic trypomastigotes and by isolated acidocalcisomes. (A) The reaction medium contained 125 mM sucrose, 65 mM KCl, 10 mM Hepes-KOH buffer, pH 7.2, 1 mM MgCl₂, 2.5 mM potassium phosphate (reaction buffer), 2 mM succinate, 0.1 mM PP_i, 1 mM ATP, and 1 µM Calcium Green-5N. Cells (5 × 10⁷) were added to the reaction medium, and the reaction was started adding 50 µM digitonin. IP₃ (5 µM), nigericin (Nig, 2 µM), and ionomycin (IO, 1 µM) were added where indicated. A decrease in fluorescence indicates decreasing medium Ca²⁺ or increasing vesicular Ca²⁺. (B) Dose–response effect of IP₃ and effect of adenophostin A (Ada, 0.5 µM) or ryanodine (Ry, 10 µM) on Ca²⁺ release. The results are means ± SD of data from three independent experiments. (C) Assay mixtures contained reaction buffer, 1 µM Calcium Green 5-N, and acidocalcisome fraction (50 µg protein). Data after ATP addition were corrected for steadily rising background by subtraction of data points for control assays without ATP addition, using Microsoft Excel. PP_i (0.1 mM), ATP (1 mM), IP₃ (10 µM), nigericin (Nig, 2 µM), and ionomycin (IO, 1 µM) were added at the points indicated. (D) Effect of different concentrations of IP₃ on Ca²⁺ release from isolated acidocalcisomes. The results are means ± SD of two independent experiments, each one with five replicates.

TbIP₃R Is Required for Optimal Growth In Vitro.

Knockdown of TbIP₃R by induction of double-stranded RNA resulted in growth defects in both BSF (Fig. 4A) and PCF trypanosomes (Fig. 4B). The effects were most pronounced with BSF trypanosomes, with up to a 30–43% reduction in the number of cells. Northern blot and ImageJ analyses showed that the mRNA was down-regulated by 49.3–79.6% after 2–8 d of tetracycline addition to PCF trypanosomes (Fig. 4D). Similar results were observed after RNAi of BSF trypanosomes (Fig. 4C), although the mRNA recovered from 20.9% to 44% after 6 d. All further phenotypic analyses were performed on day 4 of RNAi induction for both PCF and BSF trypanosomes, unless indicated otherwise.

Fig. 4.

Effect of inhibition of TbIP₃R expression by tetracycline-inducible RNAi on cell growth. A and B show the growth of BSF and PCF trypanosomes in the absence (black circles) or presence (red circles) of 1 µg/mL tetracycline for the indicated number of days. Values are mean ± SD (n = 4). (C and D) Northern blot analysis of TbIP₃R RNAi of BSF and PCF trypanosomes grown in the absence (Tet−) and presence (Tet+) of tetracycline. Total RNA was subjected to gel electrophoreses before transfer to a nylon membrane and then hybridized with the ³²P-labeled probe corresponding to the TbIP₃R coding sequence (Upper). Tubulin is shown as a loading control (Lower). Molecular markers are shown on Right. The transcript of TbIP₃R, including the 5′- and 3′-UTR was ∼10 kb in length.

Ca²⁺ Release by IP₃ Is Decreased in Mutants Deficient in TbIP₃R.

Knockdown of the expression of TbIP₃R in PCF by RNAi reduced the ability of IP₃ to release Ca²⁺ from permeabilized cells (Fig. 5 A and C). Ca²⁺ release by IP₃ was inhibited by heparin (Fig. 5 B and C), a commonly used inhibitor of the IP₃ receptors of mammalian cells (31). The combination of TbIP₃R RNAi induction and heparin almost completely blocked IP₃-induced Ca²⁺ release.

Fig. 5.

Deficient Ca²⁺ release by IP₃ in permeabilized PCF after RNAi knockdown and heparin inhibition of TbIP₃R. (A and B) Ca²⁺ uptake and release and additions were as described in Fig. 3. The blue line was without further additions after digitonin. (A) Ca²⁺ release by IP₃ addition (10 µM) in PCF grown in the presence (+Tet) or absence (−Tet) of tetracycline to induce RNAi and (B) in the presence (+Hep) or absence (−Hep) of heparin (100 µg/mL). Other aditions as in Fig. 3. (C) Results are means ± SD of data from three independent experiments. *,**Differences are statistically significant compared with respective controls, P < 0.001 (Student t test).

IP₃ Releases Ca²⁺ from Intracellular Stores in Live Trypanosomes.

To investigate Ca²⁺ release by IP₃ in live trypanosomes, we loaded TbIP₃R RNAi-induced and uninduced PCF trypanosomes with Fluo4-AM with or without caged IP₃ and recorded changes in cytosolic Ca²⁺. IP₃ was released from caged IP₃ by UV light. Fig. 6 A and B shows representative traces of Ca²⁺ responses to UV flashes (three pulses at 120 s and then six pulses at 300 s) in uninduced cells in the absence and presence of caged IP₃, respectively. In cells loaded with caged IP₃, there were rapid Ca²⁺ increases after the first (lower level) or second (higher level) flash to release free IP₃ (Fig. 6B), with a few cells responding to both UV flashes. However, the data analysis was complicated by spontaneous Ca²⁺ transients that occurred in some cells even in the absence of caged IP₃ or the UV flash (Fig. S3). Fig. 6C shows an example of a spontaneous Ca²⁺ response occurring before the IP₃-dependent Ca²⁺ release elicited by IP₃ uncaging. Spontaneous Ca²⁺ transients have also been reported in other protists (18, 32, 33). Because spontaneous Ca²⁺ transients were observed at similar rates in both uninduced and TbIP₃R RNAi-induced cells (Fig. S3), the response to IP₃ uncaging was analyzed by quantitating the Ca²⁺ responses occurring within 30 s after the UV flash (Fig. 6 A–C, shaded gray area). Data shown in Fig. S4 demonstrate that this encompasses the Ca²⁺ signals elicited by IP₃ uncaging. Fig. 6D shows the percentage of cells in which a Ca²⁺ transient occurred within 30 s of a UV flash. A significantly higher proportion of cells responded in the uninduced group loaded with caged IP₃ than in the TbIP₃R RNAi-induced group or either group without caged IP₃ (data for each experimental group are from >500 cells from four independent experiments).

Fig. 6.

Caged IP₃-dependent Ca²⁺ release in T. brucei. (A and B) Representative traces of Ca²⁺ responses to UV flash in control cells in the absence (A) and presence (B) of caged IP₃, respectively (first flash, three pulses; second flash, six pulses). Nigericin (Nig; 5 μM) and Ionomycin (IO; 5μM) were added where indicated. (C) Ca²⁺ response to UV flash in a control cell loaded with caged IP₃ in which a spontaneous transient occurred before IP₃-dependent Ca²⁺ release. (D) Percentage of cells in which a Ca²⁺ transient occurred within 30 s of a UV flash (indicated by gray shaded areas). Data for each experimental group are means ± SEM from >500 cells from four independent experiments. ***Differences are statistically significant, P < 0.001 (ANOVA with Bonferroni posttest).

TbIP₃R Is Essential for Infectivity of T. brucei in Mice.

To investigate whether TbIP₃R is essential for the establishment of an infection in animals, we inoculated groups of five to six mice with WT and RNAi transgenic BSF trypanosomes for TbIP₃R. Induction of RNAi was obtained by feeding mice with water containing doxycycline. Mice infected with the WT and noninduced cells died 5–7 d after infection with high parasitemias (Fig. 7 A and B). This result was in contrast to the doxycycline-fed mice, where most mice survived more than 10 d. By 18 d, all doxycycline-fed mice died, suggesting that a small number of parasites (RNAi escape mutants) are able to survive in the presence of doxycycline and subsequently outgrow and kill the mice. In agreement with this suggestion, tetracycline addition in vitro reduced but did not completely eliminate the expression of TbIP₃R (Fig. 4C). Doxycycline alone is known to have no effects on mice survival under the conditions used (26).

Fig. 7.

Effect of silencing of TbIP₃R expression on virulence in mice. (A) Groups of five mice were infected with WT T. brucei (wt) or trypanosomes transfected with the construct for RNAi of TbIP₃R. Doxycycline [Doxy(+); 200 µg/mL] was given in the drinking water throughout a 18-d period. (B) Parasitemia levels were monitored at 2-d intervals after infection.

Discussion

Our work clearly establishes the presence of a functional IP₃ receptor in T. brucei and, together with our previous reports (11–16), the function of a complete IP₃/diacylglycerol pathway in trypanosomes. Our results support the early emergence of this Ca²⁺ signaling pathway, preceding the origin of animals. Of the six supergroups in which eukaryotes have been divided (34), orthologs to putative IP₃Rs have been reported in protists of five supergroups [Amoebozoa (17), Archaeplastida (35), Ophistokonta (8), Chromoalveolate (18), and Excavata (9, 19, 20), and this work], although their functional roles in Ca²⁺ release have been studied only in P. tetraurelia (18) and in T. brucei (this work). As discussed before (18), putative IP₃R from protists differ significantly from mammalian IP₃ receptors. Only 4 of the 10 conserved amino acids that have been reported to be important for IP₃ binding to mouse IP₃R1 are conserved in TbIP₃R, which in addition, has only five instead of six transmembrane domains. However, the pore region is well conserved. Interestingly, the functional IP₃R of P. tetraurelia is located in the contractile vacuole complex of this Chromoalveolata (18), whereas we found that in T. brucei the receptor is localized to acidocalcisomes. Proteomic analyses of subcellular fractions have also suggested the localization of the IP₃R in the contractile vacuole (20) of T. cruzi.

Most IP₃Rs reside in ER membranes, although it has been reported that IP₃ also stimulates Ca²⁺ release from the Golgi complex (36), from the nucleus (37), and from secretory granules (38) of mammalian cells. In some mammalian cells, IP₃Rs are also targeted to the plasma membrane, where they contribute to Ca²⁺ entry (39). Secretory granules of chromaffin cells (40), pancreatic β-cells (41), acinar (42) cells, secretory granules of airway goblet cells (43), and mast cells (44) have been shown to possess IP₃Rs, although some of these results have been disputed (as discussed in ref. 45). However, it is interesting to note that some of these secretory granules, such as the mast cells granules, have similarities with acidocalcisomes (46). Although acidocalcisomes were initially described more than 18 y ago in T. brucei (47), the mechanism for Ca²⁺ release from these organelles was unknown until now. Our results are consistent with previous work using aequorin targeted to the mitochondrion of T. brucei (48), which showed that the intramitochondrial Ca²⁺ concentration can reach values much higher than cytosolic Ca²⁺ rises when Ca²⁺ release from acidocalcisomes is stimulated, suggesting a close proximity of these organelles and the presence of microdomains of high Ca²⁺ concentration in the vicinity of acidocalcisomes (49).

The sensitivity of TbIP₃R to IP₃ was lower than that of rat IP₃R1. One possible explanation was that TbIP₃R lacked 6 of the 10 basic residues required for high-affinity binding of IP₃ in rat IP₃R1. Another was that TbIP₃R might respond to a different intracellular messenger (such as cADPR, NAADP, and ryanodine). However, no response to those signaling molecules was detected in TbIP₃R-expressing DT40-3KO cells or permeabilized trypanosomes, excluding the latter possibility. The mammalian type III IP₃R also has low affinity and is poorly activated by IP₃ (50). However, it has been shown that the low affinity for IP₃ does not necessarily dictate the concentration at which the type III IP₃R opens (50). Rather, the IP₃ concentration needed to initiate Ca²⁺ release from intracellular stores is determined by the concentration of both Ca²⁺ and IP₃ because Ca²⁺ increases the sensitivity of the type III IP₃R to IP₃ (50). Interestingly, type III IP₃R has been localized to insulin and somatostatin secretory granules (41), as well as in the region containing secretory granules of salivary gland and pancreatic acinar cells (50), which has been called the trigger zone. This name is because Ca²⁺ signals originate in this region in each of these cell types (50).

Although Ca²⁺ signaling appears to be important for several functions in T. cruzi such as host cell invasion (51), multiplication and differentiation (52), osmoregulation (53), and programmed cell death (54), there is much less information on the role of Ca²⁺ in T. brucei. Based on the use of calcium ionophores, roles for Ca²⁺ in the release of the bloodstream stage surface coat (55) and in the maintenance of the cytoskeleton (56) have been proposed. Changes in cytosolic Ca²⁺ levels have also been reported during differentiation from bloodstream to procyclic stages of T. brucei (57). Our results indicate that Ca²⁺ signaling through the TbIP₃R has roles in growth in vitro and in vivo.

In conclusion, we identified and characterized at the molecular and biochemical levels, an IP₃ receptor in T. brucei, which is essential for growth and establishment of an efficient animal infection. The localization of this receptor in acidocalcisomes provides the long sought mechanism for Ca²⁺ release from these organelles.

Materials and Methods

Functional Analyses of IP₃Rs in DT40 Cells.

Uptake of Ca²⁺ into intracellular stores of permeabilized DT40 cells and its release by IP₃ were measured using a low-affinity Ca²⁺ indicator (Mag-Fluo-4) trapped within the ER as described previously (31) with some modifications. Confluent cells (50 mL, 2 × 10⁶ cells/mL) were collected by centrifugation at 180 × g for 10 min and suspended in 3 mL Hepes-buffered saline (HBS) containing 1 mg/mL BSA, 0.02% (wt/vol) Pluronic F127, and 20 µM Mag-Fluo-4 AM as described (31). After incubation at 20 °C for 1 h in the dark with gentle shaking, cells were centrifuged and resuspended in 10 mL Ca²⁺-free cytosol-like medium (CLM) (31) (140 mM KCl, 20 mM NaCl, 1 mM EGTA, 2 mM MgCl₂, and 20 mM Pipes, pH 7.0) containing 50 µg/mL saponin (Sigma). Cells were incubated with shaking at 37 °C for 4 min. Cells (20 µL) were sampled to confirm permeabilization of cells by using 0.1% Trypan Blue, and the cells were centrifuged and gently resuspended in 2.5 mL Mg²⁺-free CLM supplemented with 375 µM CaCl₂ and 10 µM FCCP. Ca²⁺ uptake was initiated by adding 250 µL of permeabilized cells into a cuvette containing 2.25 mL Mg²⁺-free CLM supplemented with 375 µM CaCl₂ and 1.5 mM Mg²⁺-ATP. Various concentrations (1–10 µM) of IP₃ were tested. The free Ca²⁺ concentration of the solution was monitored continuously under constant agitation at room temperature in a F-4500 Fluorescence Spectrophotometer (Hitachi), with excitation at 490 nm and emission at 525 nm.

IP₃-Mediated Ca²⁺ Release from Isolated Acidocalcisomes.

Acidocalcisomes were isolated and purified from PCF of T. brucei Lister strain 427 by using iodixanol gradient centrifugation as previously described (58). After gradient centrifugation, the acidocalcisome pellet was collected and washed twice with isolation buffer (125 mM sucrose, 50 mM KCl, 4 mM MgCl₂, 0.5 mM EDTA, 3 mM DTT, and 20 mM Hepes-KOH, pH 7.2) containing protease inhibitor mixture (Roche) and then resuspended in 1 mL of reaction buffer (125 mM sucrose, 65 mM KCl, 2 mM MgCl₂, 2.5 mM potassium phosphate, and 20 mM Hepes-KOH, pH 7.2). The protein concentration was determined by using Pierce BCA protein assay kit with a SpectraMax. The isolated acidocalcisomes were acidified by addition of 0.1 mM PP_i, and Ca²⁺ uptake was initiated by adding 50 µg of isolated acidocalcisomes into a cuvette containing 1.9 mL reaction buffer, 1 µM Calcium Green-5N, and 1 mM Mg²⁺-ATP. Various concentrations (1–10 µM) of IP₃ were tested. Fluorescence changes were monitored in a fluorometer, with excitation at 506 nm and emission at 531 nm.

IP₃-Evoked Ca²⁺ Release by Digitonin-Permeabilized T. brucei.

The uptake and release of Ca²⁺ from intracellular stores by permeabilized T. brucei PCF were assayed by fluorescence measurements using Calcium Green-5N as described above with some modifications. Midlog phase PCF of T. brucei was collected by centrifugation at 1,000 × g for 7 min and washed twice with cold buffer A, which contained 116 mM NaCl, 5.4 mM KCl, 0.8 mM MgSO₄, 5.5 mM d-glucose, and 50 mM Hepes at pH 7.0. Cells were resuspended to a final density of 1 × 10⁹ cells/mL in buffer A and kept on ice. A 50-µL aliquot (5 × 10⁷ cells) of the cell suspension was added to the reaction buffer (2.45 mL) containing 1 µM Calcium Green-5N, 0.1 mM PP_i, 1 mM ATP, and 2 mM succinate. Ca²⁺ uptake by the cells was initiated by the addition of 50 µM digitonin, and Ca²⁺ release was stimulated by the sequential additions of IP₃, nigericin, and ionomycin.

Photorelease of Caged IP₃ and Ca²⁺ Imaging.

WT- or TbIP₃R RNAi–treated T. brucei PCFs were washed in Ringer buffer [10 mM Hepes, 155 mM NaCl, 3.0 mM KCl, 3.0 mM KH₂PO₄, 1.0 mM MgCl₂, 2.0 mM CaCl₂, 10 mM glucose, 40 µM probenecid, and 0.25% (wt/vol) fatty acid-free BSA, pH 7.4] and loaded, in suspension, with caged IP₃ (10 µM; 40 min) followed by Fluo4-AM (10 µM; 20 min), and then plated on CellTak-coated coverslips (10 min) and mounted on the stage of an Axiovert 2000 (Zeiss) spinning disk confocal microscope. Fluo4-AM fluorescence images (excitation, 488 nm; emission, 510 nm) were acquired at 5 Hz with a cooled charge-coupled device CCD camera using the data acquisition software Piper Control (Stanfordphotonics). Photo release of caged IP₃ was achieved by light flashes (burst of three to six pulses of 1-ns duration with a wavelength of 337 nm and 1.45 mJ of energy) from a nitrogen-charged UV flash lamp (Photon Technology International) guided through the objective (C-Achromatx40/1.2). Data analysis was performed using ImageJ (NIH). Spontaneous Ca²⁺ transients observed in some cells were also characterized using Fura-2 AM (5 µM; 20-min dye loading). Fura 2 fluorescence images (excitation, 340 and 380 nm; emission, 420–600 nm) were acquired at 3-s intervals with a wide-field epifluorescent microscope coupled to a CCD camera.