Calcium-mediated protein secretion potentiates motility in Toxoplasma gondii

Toxoplasma gondii, the causative agent of toxoplasmosis, is a model organism for the phylum Apicomplexa, which also includes Plasmodium spp. (the causative agents of malaria) and Cryptosporidium parvum (an agent of waterborne diarrhea). Members of this phylum are named for their specialized apical end, which contains unique secretory organelles. Motility by these obligate intracellular parasites is essential for the rapid invasion of suitable host cells. Apicomplexans undergo directed motility despite a lack of locomotory organelles (King, 1988), and they invade host cells actively rather than relying on the host cell cytoskeleton (Dobrowolski and Sibley, 1996). Genetic and biochemical studies indicate that this unique form of locomotion, termed gliding motility, is driven by coupling the translocation of surface adhesins to an actin-myosin motor beneath the parasite plasma membrane (Sultan et al., 1997). The velocity and direction of gliding motility have been shown to depend on actin polymerization (Wetzel et al., 2003), but the methods by which other aspects of motility are regulated are largely unknown.

Apicomplexans contain a conserved family of proteins called thrombospondin-related anonymous proteins (TRAPs) that serve as adhesins for parasite motility. TRAP family members are type 1 transmembrane proteins that contain an integrin-like A domain and at least one thrombospondin type 1 domain (Kappe et al., 1999). Deletion of TRAP in Plasmodium sporozoites demonstrates that the adhesin encoded by this gene is required for both gliding and invasion (Sultan et al., 1997). The TRAP homologue expressed by T. gondii MIC2, is stored in secretory organelles termed micronemes. Contact with the host cell causes fusion of microneme vesicles with the plasma membrane, placing MIC2 on the apical surface of the parasite (Carruthers et al., 1999b). During cell invasion, MIC2 is transported to the posterior of the parasite (Carruthers et al., 1999a), where it is cleaved within its transmembrane domain by protease(s) to release the extracellular domains (Brossier et al., 2003; Carruthers et al., 2000). Attempts to delete MIC2 in T. gondii have been unsuccessful, suggesting that it is essential. Another micronemal protein, M2AP, aids the transport of MIC2 throughout the secretory network (Rabenau et al., 2001). Parasites lacking M2AP (m2apKO) demonstrate an 80% reduction in rapid cell invasion, indicating that secretion of MIC2 is required for efficient cell entry (Huynh et al., 2003).

Although the precise signals inducing microneme secretion are unclear, pharmacological studies indicate that parasite calcium is involved (Carruthers and Sibley, 1999). Apicomplexans have multiple calcium stores, such as acidocalcisomes, the endoplasmic reticulum and the mitochondria (Moreno and Docampo, 2003). Increasing intracellular calcium by treatment with alcohols (Carruthers et al., 1999b), calcium ionophores (Carruthers and Sibley, 1999) or ryanodine receptor agonists (caffeine and ryanodine) (Lovett et al., 2002), triggers microneme secretion in the absence of host cells. Chelation of intracellular calcium with BAPTA-AM blocks both microneme secretion and gliding motility (Lovett and Sibley, 2003; Vieira and Moreno, 2000). However, chelation of extracellular calcium with BAPTA or EGTA has no effect on either microneme secretion or parasite motility and cell invasion (Lovett and Sibley, 2003). In addition, host cell calcium signaling is not required for parasite invasion (Lovett and Sibley, 2003). Studies using fluo-4 and video microscopy demonstrate that cytosolic calcium levels undergo dramatic changes during parasite motility (Lovett and Sibley, 2003), but the relationship between these oscillations and gliding is not understood.

Here we use calmidazolium (CAL) to explore the link between calcium fluxes, MIC2 secretion and parasite gliding. Although CAL is marketed as a calmodulin inhibitor, it has also been shown to increase the levels of intracellular calcium in Dictyostelium discoidium (Schlatterer and Schaloske, 1996), Madin Darby canine kidney (MDCK) cells (Jan and Tseng, 2000), smooth muscle cells (Sunagawa et al., 1998) and platelets (Luckhoff et al., 1991). Our studies show that CAL affects oscillations in intracellular calcium that increase MIC2 secretion and stimulate gliding motility in T. gondii.

Caffeine and chlorpromazine were purchased from Sigma (St Louis, MO). BAPTA-AM [1,2-bis(o-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid, sodium salt], BAPTA and calmidazolium were obtained from EMD Biosciences (San Diego, CA). Fluo-4 AM was purchased from Molecular Probes (Eugene, OR). Trifluorazine and W7 were purchased from CalBiochem (EMD Biosciences, La Jolla, CA).

T. gondii tachyzoites were maintained by two-day passage in monolayers of human foreskin fibroblast (HFF) cells as described previously (Morisaki et al., 1995). Wild-type RH strain, RH strain parasites expressing β-galactosidase (clone 2F) (Dobrowolski et al., 1997), mutants lacking M2AP (m2apKO) and the complemented clone (IC4) (Huynh et al., 2003) were used in some experiments. Parasites were isolated after host cell lysis, passed through a 3-μm filter and washed with Hanks' balanced salt solution (Life Technologies, Gaithersburg, MD) that also contained 0.001 M EGTA, 0.01 M HEPES (HHE). All cultures were shown to be free of mycoplasma with the GenProbe mycoplasma detection system (GenProbe, San Diego, CA).

Parasites were allowed to glide on FBS-coated coverslips and trails deposited on the substrate as described previously (Håkansson et al., 1999). To determine the effects of calcium on gliding, parasites were treated with varying concentrations of calmidazolium, caffeine, chlorpromazine, W7, trifluorperazine (all from Calbiochem), BAPTA, BAPTA-AM or DMSO for 10 minutes prior to gliding. Treated parasites were allowed to settle on coated coverslips and incubated at 37°C for 15 minutes in the presence of agents. Cells were lightly fixed in 2.5% formalin-PBS and trails were visualized by staining with mAb DG52 conjugated to Alexa488 or Alexa594 (Molecular Probes, Eugene, OR). Average trail length in parasite body lengths (approximately 7 μm) was determined from five randomly selected 63× fields that contained ∼10 parasites per field from two coverslips per experiment. Results were averaged from three separate experiments (mean±s.e.m.). Wide-field fluorescence images were collected with a Zeiss Axiocam and Zeiss Axiovision software v3.0, then processed and merged with Adobe Photoshop v5.5 (Adobe Systems, Mountain View, CA).

Parasite invasion of HFF cells was quantified using a previously described two-color fluorescence assay that distinguishes extracellular from intracellular parasites (Carruthers et al., 1999a). In brief, following a 5-15 minute pulse for invasion, lightly fixed cells were stained with mAb DG52 to the parasite cell surface protein SAG1 (conjugated to Alexa594) followed by detergent permeabilization and re-staining with the same antibody conjugated to a second fluorochrome (Oregon Green). The percentage of parasites that had invaded (green) was determined from five random fields of at least ten parasites per field for two coverslips per experiment. The results from four experiments were averaged (mean±s.e.m.).

Gliding of parasites resuspended in Ringer's solution was monitored by time-lapse video microscopy as described previously (Håkansson et al., 1999). Images were recorded with a Hamamatsu ORCA ER camera (Hamamatsu Inc, Japan) using Openlab v3.0.9 (Improvision, Lexington, MA), cropped and saved as QuickTime movies (v5.0). Videos were used to calculate twirling speed, percentage of parasites gliding per time period and duration of gliding. Percentages of gliding parasites and average duration of gliding were calculated using five randomly selected fields containing at least ten parasites from each of two coverslips per condition per experiment. The results from three experiments were averaged.

Parasites were loaded with 100 nM Fluo-4 AM (Molecular Probes) for 5 minutes at 37°C, centrifuged and resuspended in warm Ringer's solution. Parallel samples were treated with 1% FBS and 0.02% DMSO, 1 μM CAL or 1 μM caffeine, and video images were recorded over a 15 minute time period. Time-lapse phase and fluorescent images were collected at two frames per second under low-light illumination. Pixel intensity was analyzed with Openlabs 3.0.9. The average duration (time between beginnings of successive cycles), peak to low (time between the highest signal and the lowest signal during a cycle) and time between (time between successive peaks) was analyzed for each parasite as described (Lovett and Sibley, 2003). A new cycle was defined if the pixel intensity of a given frame dropped below the lowest point of the previous cycle, and was followed by two frames with pixel intensity above this lowest point. The intensity of fluo-4 fluorescence was normalized by dividing the temporal fluorescence intensity (F_t) by the fluorescence intensity at the start of each cycle (F_o) as described previously (Torihashi et al., 2002).

SDS-PAGE was performed in 8% mini-gels under reducing conditions and proteins were transferred to nitrocellulose membranes as described previously (Carruthers et al., 1999b). Western blotting was performed using mouse anti-TgMIC2 monoclonal antibody 6D10 (ascites, 1:10,000) (Carruthers and Sibley, 1999), rabbit anti-TgMIC4 monoclonal antibody (1:5000) (Brecht et al., 2000), rabbit polyclonal anti-TgMIC5 antibody (1:5000) (Brydges et al., 2000) and/or rabbit polyclonal anti-TgACT1 actin antibody (1:10,000) (Dobrowolski et al., 1997).

Parasites were treated with varying concentrations of inhibitors or 1% DMSO for 5 minutes. Secretion was stimulated by transfer to 37°C for 15 minutes followed by transfer to a wet ice bath. To determine the effects of calcium on CAL-mediated secretion, samples were treated with 100 μM BAPTA-AM or 1% DMSO for 5 minutes at 18°C before incubation at 37°C for 2 minutes. Alternatively, 100 μM BAPTA was added immediately before stimulation of secretion for 2 minutes to minimize leaching of intracellular Ca²⁺ stores. Parasite supernatants were separated from pellets by centrifugation at 14,000 g at 4°C and run on SDS-PAGE gels along with dilutions of the cell pellet. Stimulation with 1% ethanol was used as a positive control for secretion (Carruthers et al., 1999b). Accidental parasite lysis was monitored by the release of actin, which is 98% soluble in T. gondii (Wetzel et al., 2003). Secretion was quantified from western blots using a Fujifilm FLA5000 phosphorimager (Fujifilm Medical Systems USA Inc, Stamford, CT) and analyzed using Fujifilm Image Gauge 4.0.

Parasites were washed twice in buffer A (116 mM NaCl, 5.4 mM KCl, 0.8 mM MgSO₄, 5.5 mM D-glucose, and 50 mM HEPES at pH 7.4) and centrifuged at 500 g for 10 minutes at room temperature. Cells were resuspended to a final density of 1×10⁹ cells/ml in loading buffer which consisted of buffer A plus 1.5% sucrose and 6 μM fura-2/AM. The suspensions were incubated for 30 minutes in a 25°C water bath with mild agitation. Subsequently, the cells were washed twice with buffer A to remove extracellular dye. Cells were resuspended to a final density of 1×10⁹ cells/ml in buffer A and were kept on ice. Parasites were viable for several hours under these conditions. For fluorescence measurements, a 50 μl aliquot of the cell suspension was diluted in 2.5 ml buffer A (2×10⁷ cells/ml final density) in a cuvette and scanned in a Hitachi F-2000 spectrofluorometer. For fura-2 measurements, excitation was at 340 and 380 nm and emission was at 510 nm. The fura-2 fluorescence response to intracellular calcium concentration was calibrated from the ratio of 340/380 nm fluorescence values after subtraction of the background fluorescence of the cells at 340 and 380 nm as described (Grynkiewicz et al., 1985). [Ca²⁺]_i was calculated by titration with different concentrations of Ca²⁺-EGTA buffers (Moreno and Zhong, 1996). Concentrations of the ionic species and complexes at equilibrium were calculated by employing an iterative computer program as described (Moreno and Zhong, 1996). Traces shown are representative of three independent experiments conducted on different cell preparations.

To examine the effects of CAL on T. gondii motility, we analyzed trail formation and cell entry after drug treatment. Parasites gliding on a substrate leave trails of surface membrane proteins and lipids that are readily visualized with immunofluorescence microscopy (Håkansson et al., 1999). Migration across the substratum was increased by CAL in a dose-dependent manner, as demonstrated by increased trail deposition by gliding parasites (Fig. 1A). The average length of trails formed after treatment with 10 μM CAL, a drug concentration consistent with that used in other systems, was double the length of trails of control parasites (Fig. 1B). Treatment with 10 μM CAL also led to a statistically significant increase in parasite invasion into host cells (Fig. 1C).

As CAL stimulated motility and invasion in static assays, we next examined parasites in real time to determine how this drug potentiated gliding. Parasite motility can be divided into three behaviors: twirling, circular gliding and helical gliding (Håkansson et al., 1999). T. gondii treated with 10 μM CAL were found to twirl (Table 1) and glide in a circular pattern (data not shown) at similar speeds to controls, which indicated that gliding velocity was not increased by CAL treatment. However, CAL treatment increased the length of time parasites spent gliding by 78% as well as increasing the actual number of parasites moving by 39% (Table 1).

Previous studies have indicated that microneme secretion is essential for motility (Carruthers et al., 1999a). To determine if CAL directly stimulated microneme secretion by T. gondii, parasites were treated with varying concentrations of CAL. Cells were removed by centrifugation and the presence of secreted proteins MIC2 and MIC4 in supernatants was examined by western blotting. MIC2 is a convenient marker for microneme secretion because the secreted 95-100 kDa form (sMIC2) is released into the supernatant, whereas the uncleaved 115 kDa form remains in intact cells (Carruthers et al., 2000). Micromolar concentrations of CAL stimulated secretion of both MIC2 and MIC4 when compared with parasites treated with DMSO alone as shown by western blot (Fig. 2A) and quantitative phosphorimager analysis (Fig. 2B). As actin is 98% globular in these parasites, its presence in the supernatant was monitored to control for inadvertent cell lysis. Cell lysis was typically 1% or less, as indicated by comparison with dilutions of a parasite cell standard (% pellet).

To confirm that the effects of calmidazolium on gliding were due to its action on MIC2 and not on other adhesins we used the m2apKO strain, which has a specific defect in MIC2 secretion (Huynh et al., 2003). CAL induced microneme secretion by m2apKO at similar levels to RH strain parasites, as indicated by the release of the microneme protein MIC5 (Fig. 3A,B). However, MIC2 secretion was not stimulated by micromolar concentrations of CAL in the m2apKO strain (Fig. 3A). Quantification of secreted MIC2 indicated that m2apKO parasites released tenfold less MIC2 than wild-type (RH strain) parasites treated with CAL (Fig. 3B). This result is consistent with previous findings that in the absence of M2AP, secretion of MIC2 is specifically impaired (Huynh et al., 2003). Actin release due to accidental cell lysis did not significantly vary between the wild type (RH) and m2apKO parasites and was less than 5% (data not shown).

To examine the effects of CAL on motility by m2apKO parasites, we analyzed trail formation after drug treatment. Migration across the substratum by m2apKO parasites was not altered by CAL and trail lengths remained similar regardless of drug treatment (Fig. 3C). However, trail formation by IC4, the complemented clone of m2apKO, increased with CAL treatment to a similar degree as in the wild type (data not shown). Interestingly, m2apKO parasites were found to have a slight but statistically significant defect in trail formation. Collectively, these results indicate that CAL potentiated gliding motility in wild-type parasites by increasing MIC2 secretion.

CAL is marketed as a calmodulin inhibitor. To determine if this mechanism was operating in T. gondii, we tested the effects of several calmodulin inhibitors including chlorpromazine, trifluoroperazine and W7. Treatment with these agents did not increase the length of trails made by gliding parasites (Table 2), indicating that CAL probably affected motility by another mechanism.

CAL is also known to increase intracellular calcium in a variety of cells. To determine whether CAL increased the [Ca²⁺]_i in T. gondii, we used fura-2/AM to monitor calcium levels in the parasite (Grynkiewicz et al., 1985). Increasing the concentration of CAL stimulated progressively greater increases in [Ca²⁺]_i when parasites were maintained in a buffer containing 1 mM CaCl₂ (Fig. 4A). In contrast, when parasites were incubated in a buffer containing 1 mM EGTA (to prevent Ca²⁺ entry), a modest rise in [Ca²⁺]_i was observed at low concentrations of CAL (1-3 μM) whereas at higher concentrations of CAL (>5 μM), [Ca²⁺]_i decreased after a slight initial rise (Fig. 4B). This result suggests that the drug made the membrane permeable to calcium, resulting in calcium loss from the cell. However, CAL-treated cells were not labeled by propidium iodide, indicating that the plasma membrane bilayer integrity was preserved (data not shown). Although this indicates that calcium efflux was not due to nonspecific membrane permeabilization, it is possible that CAL induced ion exchange across the membrane, either by influencing channels or pumps. To monitor the changes in [Ca²⁺]_i directly during the influx of extracellular calcium, parasites were initially incubated in 1 mM EGTA (Fig. 4C). The slight increase in [Ca²⁺]_i observed after addition of CAL, was greatly augmented by further addition of 2 mM CaCl₂ to the extracellular medium (Fig. 4C, trace b). Collectively, these results indicate that at low concentrations of CAL, [Ca²⁺]_i was increased due to release of calcium from intracellular stores, whereas at higher drug concentrations calcium entered across the plasma membrane from the extracellular medium.

To determine the source of elevated calcium affecting secretion and motility in CAL-treated parasites, we incubated parasites concurrently with CAL and BAPTA to chelate extracellular calcium, or BAPTA-AM to sequester intracellular calcium. First, we examined whether the presence of extracellular calcium affected CAL-stimulated motility as monitored by trail formation. We found that trail length no longer increased with high doses of CAL if parasites were treated simultaneously with BAPTA (Fig. 5A). However, BAPTA can cause leakage of calcium from intracellular stores if cells are exposed to the chelator for prolonged lengths of time (Lovett and Sibley, 2003). Therefore, we used a rapid secretion assay to examine whether extracellular calcium was partly responsible for the effects of CAL. First, parasites were treated with DMSO or varying concentrations of CAL at temperatures that were non-permissive for secretion. Next, BAPTA was added and parasites were immediately placed at 37°C for 2 minutes to stimulate secretion. Cells were removed by centrifugation and secretion was evaluated by detection of MIC2 in the supernatants by western blotting (Fig. 5B). Treatment of parasites with BAPTA blocked CAL stimulation of secretion, demonstrating that the influx of extracellular calcium was indeed responsible for the effects of CAL (Fig. 5C). BAPTA-AM treatment blocked microneme secretion and gliding regardless of CAL concentration (data not shown), reflecting the known requirement for elevated intracellular calcium for these events (Lovett and Sibley, 2003).

To determine if other agents that increase intracellular calcium levels and microneme secretion potentiated gliding by T. gondii, we examined parasites treated with caffeine. Caffeine is thought to cause microneme secretion by stimulating release of calcium from IP₃ or ryanodine-like receptors in T. gondii (Lovett et al., 2002). We found that caffeine did not increase trail length in gliding assays (Fig. 5D). Thus, not all reagents that increased intracellular calcium and microneme secretion stimulated gliding.

To explore why calcium-mediated protein secretion stimulated gliding in CAL-treated but not caffeine-treated parasites, we used fluo-4 to visualize qualitative changes in calcium levels during gliding by T. gondii. Calcium fluxes correlate with gliding motility and often precede initiation of gliding (Lovett and Sibley, 2003). We found that gliding by control and CAL-treated parasites was generally associated with brightly fluorescent cells that underwent cycles of oscillating fluorescence that gradually decreased in intensity, as reported previously (Lovett and Sibley, 2003). After treatment with 1 μM CAL, parasites fluoresced even more brightly (Table 3) and the frequency of oscillations increased (Fig. 6B and Table 3). Plotting kinetic changes in fluo-4 fluorescence suggested that CAL-treated parasites had quicker fluxes in intracellular calcium (Fig. 6B, Table 3). Conversely, parasites loaded with fluo-4 and treated with 1 μM caffeine remained constantly bright (Fig. 6A,B and Table 3) and ceased to glide after ∼10 minutes (data not shown). Collectively, these data suggested that calcium transients, and not simply increases in intracellular calcium levels, are required for efficient gliding by T. gondii.

As the kinetic changes in fluo-4 fluorescence indicated that CAL-treated parasites had more rapid but less intense fluxes in intracellular calcium (Fig. 6B, Table 3), we wished to explore this relationship further. Therefore, we normalized the data shown in Fig. 6B by dividing the fluorescence intensity at each point by the intensity at the start of each cycle (Torihashi et al., 2002). Transformation of fluo-4 intensities demonstrated clear and synchronized oscillations in intracellular calcium during gliding by both control and CAL-treated parasites (Fig. 6C). Furthermore, normalization shows that levels of intracellular calcium in CAL-treated parasites oscillate at twice the frequency but half the amplitude of calcium levels in untreated parasites (Fig. 6C). Calcium oscillations in caffeine-treated parasites were not apparent even after normalization (data not shown).

Calmidazolium provides a useful tool for exploring the connection between calcium fluxes, microneme secretion and gliding motility in the Apicomplexa. We show that CAL potentiates gliding motility and cell invasion by increasing the duration of gliding by T. gondii. The effect of CAL on gliding specifically depends upon stimulation of MIC2 secretion. Treatment with CAL increased intracellular calcium levels both through release of calcium from intracellular stores and entry of calcium from the extracellular milieu. Finally, fluo-4 studies demonstrated that whereas rapid oscillations in [Ca²⁺]_i caused by CAL stimulated gliding, constantly elevated intracellular calcium in parasites treated with caffeine resulted in microneme secretion without altering gliding. Collectively, our results indicate that oscillations in intracellular calcium are necessary for efficient gliding by T. gondii.

Motility and cell invasion by apicomplexans are dependent on a family of conserved protein adhesins typified by TRAP. Disruption of TRAP, an orthologue of MIC2 in Plasmodium berghei, results in nonmotile sporozoites (Sultan et al., 1997). In addition, replacement of the penultimate tryptophan and final asparagine of TRAP with alanine and serine results in mutants that undergo non-productive `pendulum' gliding, in which sporozoites briefly move forward and then return to their original position (Kappe et al., 1999). MIC2 is a TRAP homologue expressed by T. gondii and deletion of its binding partner m2ap compromises MIC2 secretion and rapid cell invasion (Huynh et al., 2003). However, its involvement in gliding motility in T. gondii has never been demonstrated directly. Here we demonstrate that m2ap mutants also have a defect in trail formation and fail to upregulate gliding in response to CAL, demonstrating that MIC2 is involved in promoting efficient gliding by T. gondii.

The precise calcium-regulated trigger of microneme secretion by T. gondii has not been identified. In neural transmission, the synaptic vesicle protein synaptotagmin acts as the calcium sensor for triggering vesicle release (Sudhof, 2002). Although an analogous mechanism may operate in parasites, homologues of synaptotagmin have not yet been described in the Apicomplexa, despite the fact that several complete genomes are now available (Gardner et al., 2002). Evidence also exists that activated calmodulin can serve as a calcium sensor in specialized secretory cells (Coppola et al., 1999; Park et al., 1997; Quetglas et al., 2002). Future studies describing the upstream and downstream signaling events involved in calcium-mediated protein secretion by T. gondii are required to better understand the regulation of this unique process.

In most eukaryotic cells, regulated exocytosis requires mobilizing calcium from internal stores, often combined with influx of calcium from the external medium (Burgoyne and Clague, 2003). Our previous studies indicated that intracellular stores within the parasite are largely responsible for the rise in intracellular calcium that accompanies normal motility and invasion (Lovett and Sibley, 2003). Here we show that CAL can augment this process by also inducing influx of extracellular calcium across the membrane (particularly at higher drug concentrations). Because this influx of extracellular calcium is not due to non-specific membrane permeabilization, our results suggest that CAL affects calcium channels in the parasite plasma membrane. In most eukaryotic cells, calcium gains access across the plasma membrane through receptor-operated, voltage-gated and/or store-operated calcium channels (Moreno and Docampo, 2003). In general, eukaryotic cells actively export calcium through Na⁺/Ca²⁺ exchangers and a plasma membrane Ca²⁺-ATPase (PMCA) (Moreno and Docampo, 2003). A PMCA has been localized to both the plasma membrane and the acidocalciosome of T. gondii (Luo et al., 2001). Thus, CAL may mediate the entry of extracellular calcium into T. gondii by disrupting the activity of this pump at the plasma membrane.

Previous studies demonstrated that increased intracellular calcium stimulates microneme secretion by T. gondii (Carruthers et al., 1999a; Carruthers et al., 1999b; Carruthers and Sibley, 1999; Lovett et al., 2002); however, the agents used in these studies lead to a refractoriness in further secretion and block motility and invasion. A major advantage of CAL treatment is that it elevates calcium yet still allows motility and invasion to be examined for an extended period. Our data using other calmodulin inhibitors indicate that the effect of CAL in enhancing T. gondii motility is unlikely to be due to inhibition of calmodulin. Monitoring intracellular calcium levels during gliding revealed that more rapid fluxes in intracellular calcium levels are associated with the increase in gliding duration seen after CAL treatment of T. gondii. As these oscillations are eliminated by caffeine stimulation of secretion and as caffeine-treated parasites do not form extended trails in a gliding assay, calcium fluxes appear to be necessary for the continuation of motility. Periodic increases of calcium levels, such as those stimulated by CAL, are likely to lead to repeated rounds of microneme secretion and continuation of gliding. At present, we are not able to monitor secretion directly in live cells, and thus it is impossible to determine if each oscillation is accompanied by a separate round of microneme secretion. Prolonged calcium increase, such as one caused by caffeine, results in the rapid release of microneme proteins, followed by a refractory period where gliding is inhibited. Consistent with this hypothesis, pre-treatment of parasites with agonists that strongly stimulate microneme secretion decreases subsequent cell invasion (Carruthers et al., 1999a).

Rapid oscillations in intracellular calcium levels that were observed during gliding indicate that intracellular calcium release and reuptake mechanisms are very active in T. gondii. Unfortunately, very little is known about calcium pools or how calcium release is regulated in the Apicomplexa. In many eukaryotes, calcium is stored mainly in the endoplasmic reticulum, and its release is mediated by IP₃ or ryanodine channels. T. gondii is sensitive to antagonists and agonists of both of these channels (Lovett et al., 2002), suggesting they are present and active in these parasites. SERCA-type ATPases are calcium transporters that refill ER stores of calcium in other eukaryotes. T. gondii is sensitive to thapsigargin (Carruthers and Sibley, 1999; Moreno and Zhong, 1996), an inhibitor of these transporters (Thastrup et al., 1990; Thastrup et al., 1989), although SERCA-type ATPases have not yet been characterized in this organism. In addition, calcium ATPases on the acidocalcisome, a unique calcium storage organelle in lower eukaryotes, could participate in intracellular calcium release and reuptake in T. gondii (Docampo and Moreno, 2001).

Thus far, only one other drug, jasplakinolide (JAS), has been shown to increase gliding motility in the Apicomplexa. JAS causes unregulated actin polymerization in T. gondii leading to increased velocity and random directional changes that result in non-productive locomotion (Wetzel et al., 2003). CAL treatment does not affect filamentous actin levels in the parasite (data not shown), but instead stimulates productive gliding by increasing intracellular calcium and leading to microneme secretion. Oscillations in intracellular calcium levels appear to promote prolonged microneme secretion and lengthen the duration of gliding by T. gondii. Thus, regulation of gliding motility in these parasites occurs on two levels: actin polymerization and microneme secretion, each of which directs different aspects of motility in the Apicomplexa.

We thank the students in the Biology of Parasitism Course (Marine Biological Laboratory, Woods Hole, MA) who performed the preliminary experiments that led to this study. We also thank Naomi Morrissette for helpful discussions, Vern Carruthers (Johns Hopkins University) for antibodies to MIC2AP and the m2ap knockout and complemented strains and Julie Suetterlin for expert cell culture assistance. Supported by NIH Grant AI34036 (L.D.S.), the Burroughs Wellcome Fund (L.D.S.), NIH Institutional Training Grant AI017172-19 (D.M.W.), and the Medical Student Summer Research Program (L.A.C.) at Washington University School of Medicine.