Developmental Expression of a Trypanosoma cruzi Phosphoinositide-Specific Phospholipase C in Amastigotes and Stimulation of Host Phosphoinositide Hydrolysis

Abstract

Phosphoinositide phospholipase C (PI-PLC) plays an essential role in cell signaling. A unique Trypanosoma cruzi PI-PLC (TcPI-PLC) is lipid modified in its N terminus and localizes to the outer surface of the plasma membrane of amastigotes. We show here that TcPI-PLC is developmentally regulated in amastigotes and shows two peaks of surface expression during the developmental cycle of T. cruzi, the first immediately after differentiation of trypomastigotes into amastigotes and the second before differentiation of amastigotes into trypomastigotes. Surface expression of TcPI-PLC coincides with phosphatidylinositol 4,5-bisphosphate (PIP₂) depletion in the host cell membrane and with an increase in the levels of its product, inositol 1,4,5-trisphosphate. During extracellular differentiation, PI-PLC is secreted into the incubation medium. Maximal early expression of TcPI-PLC on the surface of amastigotes and PIP₂ depletion coincide with host cytoskeletal changes, Ca²⁺ signaling, and transcriptional responses described previously. The presence of TcPI-PLC on the outer surface of the plasma membrane of the parasite and the capacity to be secreted and to alter host phospholipids are novel mechanisms of the host-parasite interaction.

Phosphoinositide-specific phospholipases C (PI-PLCs) catalyze the hydrolysis of phosphatidylinositol-4,5-bisphosphate (PIP₂) to d-myo-inositol-1,4,5-trisphosphate (IP₃) and sn-1,2-diacylglycerol (DAG) (1, 7). Both products of this reaction function as second messengers in eukaryotic signal transduction cascades. The soluble IP₃ triggers the release of calcium from intracellular stores (1, 7). The membrane-resident DAG controls cellular protein phosphorylation states by activating various protein kinase C isozymes (16).

Previous work has indicated that some of the signaling components of the inositol phosphate/diacylglycerol signaling pathway are present in Trypanosoma cruzi (5, 15, 17), the etiologic agent of Chagas disease. T. cruzi has several developmental stages: the epimastigote that is found in the insect vector and can be grown in axenic cultures; the amastigote, or intracellular form, that lives in the cytoplasm of nucleated cells; and the trypomastigote, which is the terminal differentiation stage in the vector (metacyclic form) or is found in the bloodstream of mammalian hosts (bloodstream form). The gene encoding the T. cruzi PI-PLC (TcPI-PLC) has been identified, and the protein product has been characterized biochemically (6). TcPI-PLC is more similar to the zeta-type PLCs of mammalian cells (22). However, the T. cruzi enzyme possesses N myristoylation and palmitoylation consensus sequences that have not been described in any other PI-PLC and has been shown to be N myristoylated and palmitoylated in vivo (6). It has been shown that this lipid modification is important for its plasma membrane localization and for differentiation of trypomastigotes into amastigotes (18). The enzyme has been shown recently to localize to the outer surface of the plasma membrane of amastigotes (12).

In the present study we investigated the appearance of TcPI-PLC in the plasma membrane of the parasites during differentiation of trypomastigotes to amastigotes in vitro and in vivo. We also investigated the potential role of TcPI-PLC in host cell signaling pathways.

MATERIALS AND METHODS

Chemicals and reagents.

The protein assay was from Bio-Rad (Hercules, CA). Alexa Fluor 546 and Alexa Fluor 488 were from Molecular Probes (Invitrogen, Eugene, OR). The monoclonal antibodies anti-Ssp3 and anti-Ssp4 were gifts from Norma Andrews (University of Maryland), and rabbit anti-TcPI-PLC antibody was produced in our laboratory as described previously (6). Myo- [³H]inositol (7.27 Ci/mmol) and ³H-PIP₂ (5.0 Ci/mmol) were from Perkin-Elmer (Waltham, MA). PIP₂ was from Avanti-Polar Lipids, Inc. (Alabaster, AL). All other reagents were analytical grade.

Culture methods.

T. cruzi amastigotes and trypomastigotes (Y strain) were obtained from the culture medium of L₆E₉ myoblasts by a modification of the method of Schmatz and Murray (23) as we have described before (14). Contamination of trypomastigotes with amastigotes and intermediate forms or of amastigotes with trypomastigotes or intermediate forms was always <5%. Trypomastigotes were induced to transform into amastigotes axenically as described previously (6). For intracellular imaging of the parasites, mammalian cells were seeded onto coverslips in 12-well culture plates and allowed to grow for 24 h. To semisynchronize the infection, we incubated the parasites at a ratio of 5:1 (parasite/host cell) for 30 min, unless otherwise stated, and washed the cells three times to eliminate extracellular parasites.

Fluorescence microscopy.

Extracellular parasites preparation and fixation were performed as described previously (6). Infected cells were prepared as described before (14), except for the permeabilization that was performed for 10 min with Triton X-100 in phosphate-buffered saline (pH 7.4). The dilutions used for primary antibodies were as follows: α-TcPI-PLC, 1:100; α-Ssp4 and α-Ssp3, 1:1,500; and α-PIP₂, 1:50. Secondary antibodies were diluted at 1:1,000. Coverslips were mounted by using a mounting medium containing DAPI (4′,6′-diamidino-2-phenylindole) at 5 μg/ml. Differential interference contrast (DIC) and direct fluorescence images were obtained by using an Olympus IX-71 inverted fluorescence microscope with a Photometrix CoolSnapHQ charge-coupled device camera driven by DeltaVision softWoRx3.5.1 (Applied Precision, Issaquah, WA). This same software was used to deconvolve and process the final images. The final figures were built by using Adobe Photoshop 6.0 (Adobe Systems, Inc., San Jose, CA.).

Increase in IP₃ in L₆E₉ myoblasts infected with T. cruzi.

Myoblasts (10⁷) were labeled with [³H]inositol for 24 h in medium 199 supplemented with 20% dialyzed fetal bovine serum. Cells were washed three times to remove the labeled inositol, infected with 5 × 10⁸ trypomastigotes for 3 h, and washed again three times to remove extracellular trypomastigotes. Only 1% radioactivity was found in these trypomastigotes. Control and infected cells were incubated for 6, 12, 18, 21, 24, 30, 48, and 72 h. Three hours before each extraction, LiCl (15 mM) was added to the culture medium to inhibit inositol phosphatase activities. Inositol phosphates were analyzed as described before (15) at the times indicated. The levels of inositol phosphate (IP), inositol 1,4-diphosphate (IP₂), and inositol 1,4,5-trisphosphate (IP₃) were increased only after 18 to 24 h postinfection.

Analyses of PIP₂ labeling at the plasma membrane of infected and uninfected cells.

In order to correlate the labeling of PIP₂ and the infection by T. cruzi, 200 infected and uninfected cells, from three independent experiments, were randomly counted, and the percentage of labeled cells was determined. To measure the intensity of the PIP₂ labeling, deconvolved images were analyzed using the data inspector tool of the DeltaVision softWoRx3.5.1. The fluorescence intensity in arbitrary units was used to generate x-y plots showing the position and intensity of the fluorescence across the host cells. Graphics were generated by the software SigmaPlot version 10 (San Jose, CA).

TcPI-PLC activity assay.

Samples were obtained after 4 and 18 h of trypomastigote to amastigote differentiation. Cells were lysed in radioimmunoprecipitation assay (RIPA) buffer as described previously (12). Extracellular medium was centrifuged four times at 1,000 × g for 10 min in order to eliminate cellular contamination and then concentrated 10 times by centrifugation in centrifugal filters with 3,000 Da of molecular size exclusion (Centriprep Millipore, County Cork, Ireland). TcPI-PLC activity was measured in the lysates and extracellular media as the release of water-soluble radioactivity from ³H-PIP₂ as described previously (6) with minor modifications. Briefly, organic solvents of stock solutions were dried just prior to use under a stream of nitrogen and suspended in reaction buffer by sonication for 10 s in a Branson digital sonifier. The PI-PLC activity was measured by using a published protocol (6).

RESULTS

Developmental regulation of TcPI-PLC expression during trypomastigote-to-amastigote differentiation in vitro and in vivo.

We studied the expression of TcPI-PLC during the differentiation of trypomastigotes to amastigotes in vitro and in vivo compared to that of the amastigote and trypomastigote markers Ssp4 and Ssp3, respectively.

At 4 h (4 ± 1.5 h, n = 5) after induction of differentiation in vitro at pH 5.0, staining with antibodies against TcPI-PLC had a punctate appearance in amastigotes (Fig. 1, TcPI-PLC, 4 h, arrow) and was completely absent in trypomastigotes. Surface staining with Ssp4 was less evident (Fig. 1A, Ssp4, left panel). The surface expression of TcPI-PLC reached a peak at ∼12 h (Fig. 1A, 12 h) (peak at 12 ± 4 h, n = 5) and disappeared by 72 h (Fig. 1A, right panel) (peak disappearance at 72 ± 12 h, n = 5). This pattern of expression was similar to that of Ssp4. Interestingly, prolonged incubation of amastigotes in acidic medium (96 h) resulted in their differentiation into epimastigotes, and this was preceded by surface expression of TcPI-PLC in the amastigotes (see Fig. S1 in the supplemental material).

FIG. 1.

Surface expression of TcPI-PLC and Ssp4 during differentiation of trypomastigotes to amastigotes. (A) Trypomastigotes were incubated at pH 5 for 4 h and then transferred to medium at pH 7.4 for different times. By 12 h, the surface expression of Ssp4 and TcPI-PLC is evident. The surface expression of TcPI-PLC decreases by 48 h and disappears by 72 h. The two top rows show immunofluorescence pictures, and the bottom row shows merge images of fluorescence and DIC pictures. (B) Immunofluorescence analysis of surface expression of TcPI-PLC in intracellular amastigotes in HeLa cells. Intracellular trypomastigotes upon entry into HeLa cells (3 h, arrow) or after completion of the intracellular cycle (96 h) are not labeled with antibodies against TcPI-PLC. Labeling starts to be visible by 6 h, becomes more evident by 12 and 24 h after infection, disappears by 48 h, and starts to appear again by 72 h, becoming very evident by 80 h. Bars, 10 μm.

When analyzing the expression of TcPI-PLC in T. cruzi trypomastigotes during in vivo infection of host cells, we observed that the expression of TcPI-PLC started by 5 to 6 h postinfection of HeLa cells (Fig. 1B) (appearance of the signal at 5 ± 2 h, n = 5) and became prominent by 12 to 24 h (Fig. 1B) (peak at 18 ± 6 h, n = 5). Cell surface labeling of TcPI-PLC gradually disappeared until it appeared again by 72 to 80 h (second peak at 72 ± 10 h, n = 5), before disappearing again after final differentiation of amastigotes into trypomastigotes by 96 h (disappearance at 90 ± 10 h, n = 5). Figure S2 in the supplemental material shows the same panels (Fig. 1A and B) examined only by fluorescence microscopy. Similar results were observed when L₆E₉ myoblasts were used instead of HeLa cells (see Fig. S3 in the supplemental material).

When macrophages were infected with trypomastigotes, there were also two peaks of expression of TcPI-PLC, one after 12 h of infection (Fig. 2 A) (peak at 12 ± 4 h, n = 4) and the second by 60 h (Fig. 2A, 60 h) (second peak at 60 ± 10 h, n = 4), which is somewhat earlier than that occurring in HeLa cells or L₆E₉ myoblasts. Labeling with antibodies against Ssp4 persisted up to 60 h (Fig. 2A, red) (disappearance at 60 ± 10 h, n = 4). Labeling with Ssp3 was only detected in infecting trypomastigotes (Fig. 2B, red, 1 h) or once the parasites have differentiated into trypomastigotes (Fig. 2B, 84 h). Figure 2C shows an analysis of the increase in TcPI-PLC expression with time. Figure S4 in the supplemental material shows the same panels (Fig. 2A and B) but examined by fluorescence microscopy only.

FIG. 2.

Surface expression of TcPI-PLC, SSp4 (A) and Ssp3 (B) in intracellular amastigotes from macrophages infected with trypomastigotes. Two peaks of surface expression of TcPI-PLC occurred. The first started at ∼12 h and the second started at ∼60 h after infection. (A) Labeling of Ssp4 was clear after 36 h and up to 60 h (red). Trypomastigotes (84 h) were not labeled with either TcPI-PLC or Ssp4 antibodies. (B) Expression of Ssp3 was detected in trypomastigotes upon invasion (1 h) or after differentiation from amastigotes (84 h). (C) Changes in expression of TcPI-PLC at the plasma membrane with time, as measured by fluorescence intensity in arbitrary units ± the standard error of the mean (SEM; n = 4).

When amastigotes, instead of trypomastigotes, were used to infect macrophages high levels of TcPI-PLC were observed at the plasma membrane since early internalization (30 min) and labeling decreased earlier (Fig. 3 A) compared to trypomastigotes-infected macrophages (Fig. 2A). The initial surface expression of TcPI-PLC already disappeared by 24 h (Fig. 3A) (disappearance at 24 ± 4 h, n = 4). These amastigotes were able to differentiate into trypomastigotes earlier than when trypomastigotes were used as inoculum, as revealed by labeling with antibody against Ssp3, a trypomastigote marker (Fig. 3B), and, accordingly, the second peak of TcPI-PLC expression occurred by 48 to 60 h postinfection (Fig. 3B) (peak at 54 ± 6 h, n = 4). Figure 3C shows an analysis of the increase in TcPI-PLC expression with time. Figure S5 in the supplemental material shows the same panels (Fig. 3A and B), as examined only by fluorescence microscopy.

FIG. 3.

Surface expression of TcPI-PLC, SSp4 (A), and Ssp3 (B) in intracellular amastigotes from macrophages infected with amastigotes. Surface expression of TcPI-PLC occurred earlier than with trypomastigotes-infected host cells (compare to Fig. 2). The second peak of surface expression occurred at 48 h after infection. (A) Labeling of anti-TcPI-PLC and anti-Ssp4. (B) Expression of Ssp3 was detected in trypomastigotes after differentiation from amastigotes (60 to 72 h). Bars, 10 μm. (C) Changes in expression of TcPI-PLC at the plasma membrane with time, as measured by fluorescence intensity in arbitrary units ± the SEM (n = 4).

In conclusion, surface expression of TcPI-PLC appears to peak after trypomastigotes differentiate into amastigotes and before differentiation of amastigotes into trypomastigotes or epimastigotes, and the time of its appearance depends on the infective stage and the host cell used.

Hydrolysis of phosphatidylinositol 4,5-bisphosphate and generation of IP₃ in myoblasts infected with T. cruzi.

Since TcPI-PLC is exposed to the cytoplasm of the host cell and could potentially affect host phospholipids, we first studied whether changes in PIP₂ levels can be induced by T. cruzi invasion. Control L₆E₉ cells (Fig. 4 A, uninfected) showed strong plasma membrane staining with antibodies against PIP₂. In cultures inoculated with trypomastigotes (Fig. 4A, infected) there was a clear difference between host cells containing or not amastigotes. Those that contained amastigotes labeled with TcPI-PLC after 18 to 24 postinfection showed much lower staining with antibodies against PIP₂ than those that did not contain amastigotes (Fig. 4B and C).

FIG. 4.

Correlation between higher levels of TcPI-PLC expression and host PIP₂ hydrolysis. (A) Immunofluorescence analysis of L₆E₉ myoblasts after infection with T. cruzi. After 18 to 24 h of infection, amastigotes were labeled with antibodies against TcPI-PLC (red), and host cells were labeled with antibodies to PIP₂. The photos are merge images of DIC and fluorescence (left panels) or fluorescence (right panels). Bars, 10 μm. (B) Percentage of host cells labeled with antibodies against PIP₂ in infected and uninfected cells 18 and 24 h after infection. Values are means ± the SEM from three different experiments. (C) Line scan analysis of PIP₂ labeling in L₆E₉ myoblasts. Plots corresponding to six cells representative of each condition are shown. (D) Increase in IP₃ in L₆E₉ myoblasts infected with T. cruzi. Myoblasts were labeled with [³H]inositol, washed, infected with trypomastigotes, and incubated for different times. Levels of inositol phosphate (IP), inositol 1,4-diphosphate (IP₂), and inositol 1,4,5-trisphosphate (IP₃) were increased only at 18 to 24 h postinfection. The graph shows the increase in inositol phosphates as means ± SEM of triplicate determinations.

In agreement with these results, Fig. 4D shows that when [³H]inositol-labeled L₆E₉ myoblasts were infected with T. cruzi trypomastigotes there was an increase in inositol phosphates 18 to 24 h postinfection. No significant changes were detected before 18 h (data not shown). The increase in all inositol phosphates is a result of IP₃ formation and degradation to IP₂ and IP by the inositol phosphatases present in the cells (7). Because [³H]inositol was incorporated by the myoblast lipids before infection, IP₃ was most probably coming from the hydrolysis of host lipids.

TcPI-PLC secretion during trypomastigote-to-amastigote extracellular differentiation.

TcPI-PLC labeling was temporarily found at the surface of intracellular amastigotes (Fig. 1B and 2) and then disappeared, while PIP₂ depletion was observed at the host plasma membrane (Fig. 4), suggesting that the enzyme is released from the parasite during its differentiation to reach the host plasma membrane. However, we could not detect labeling of the host cell using antibodies against TcPI-PLC in infected cells (Fig. 1B), probably as a consequence of dilution of the enzyme in the host cytosol. We therefore investigated whether TcPI-PLC activity could be measured in the extracellular medium during differentiation of trypomastigotes into amastigotes in vitro. Trypomastigotes were incubated at acidic pH to stimulate their differentiation and the PI-PLC activity of cell lysates and their supernatants was measured at different times using PIP₂ as substrate. Figure 5 shows that PI-PLC activity in the lysates was detected at 4 and 18 h of incubation. PI-PLC activity was also detected in the supernatants but was not stimulated by Ca²⁺. The activity was not significantly different between 4 and 18 h, suggesting a constitutive secretion once the differentiation process is stimulated. No significant lysis occurred during the differentiation process, as detected by counting the cells at time zero, and after 4 and 18 h of incubation (data not shown). The activity in the supernatants was ca. 50% of the basal activity found in lysates of equivalent amount of cells.

FIG. 5.

PI-PLC activity detected in extracellular media and in lysates obtained from amastigotes after 4 and 18 h of in vitro differentiation. Trypomastigotes (5 × 10⁷ cells/ml) were resuspended in 4 ml of Dulbecco modified Eagle medium (pH 5.0) containing 0.4% albumin and 20 mM MES (morpholineethanesulfonic acid) and then incubated at 37°C. Aliquots corresponding to 1.2 × 10⁷ amastigotes were taken at 4 and 18 h and centrifuged as described in Materials and Methods. The pellet was lysed in RIPA buffer and used to measure cellular PLC activity in the absence or presence of 100 μM Ca²⁺, while supernatants of equivalent amount of parasites were used to measure the activity released into the medium in the absence or presence of 100 μM Ca²⁺. Activity without lysates was negligible. The results are the average from four independent experiments ± the SEM. *, P < 0.01; **, P < 0.001 (one-way analysis of variance).

DISCUSSION

TcPI-PLC was found to be developmentally regulated, showing two peaks of surface expression during the life cycle of T. cruzi in the host, the first immediately after differentiation of trypomastigotes into amastigotes and the second before differentiation of amastigotes into trypomastigotes. The appearance of two peaks of expression was detected during differentiation in vitro or in vivo. This phenomenon was independent of the host cell invaded, suggesting that it is determined by the parasite and not influenced by the host. Variations in the time of maximal expression observed when amastigotes were used instead of trypomastigotes or when macrophages were used instead of nonphagocytic cells could be due to differences in the ability of the different parasite stages to invade and establish a successful infection in different hosts.

Surface expression of TcPI-PLC coincided with phosphatidylinositol 4,5-bisphosphate (PIP₂) depletion in the host cell membrane and with an increase in the level of its product, inositol 1,4,5-trisphosphate (IP₃). PI-PLC enzymatic activity could be detected in the extracellular medium of the cells during differentiation, suggesting that the parasite enzyme could be stimulating phosphoinositide hydrolysis in the host cell. A critical test of this hypothesis would necessarily depend on generation of a TcPI-PLC knockout or on the availability of specific inhibitors of this enzyme. However, our previous attempts to knockout TcPI-PLC gene were unsuccessful, and we concluded that the enzyme is essential for T. cruzi (18). In addition, T. cruzi lacks the machinery for RNA interference, which is present in T. brucei (4). To further complicate things, no comparable experiments could be done using T. brucei conditional knockout parasites because, although this parasite possesses an orthologue of TcPI-PLC, it lacks an intracellular stage. Although U73122 is a good inhibitor of PI-PLCs, it does not inhibit TcPI-PLC, probably due to the absence of heterotetrameric G proteins in this parasite and to the mechanism of inhibition by U73122, which is through G protein inhibition (24). Finally, expression of full-length TcPI-PLC in mammalian cells was lethal (data not shown), precluding the investigation of its effects on host plasma membrane PIP₂. TcPI-PLC is the only PI-PLC present in the parasite with the ability to hydrolyze PIP₂ (6), and although we cannot rule out the involvement of a host PI-PLC in PIP₂ changes, our results are highly suggestive that TcPI-PLC is the enzyme involved in this phenomenon.

The presence of TcPI-PLC in the outer surface of the plasma membrane of amastigotes in contact with the cytoplasm of mammalian cells would be sufficient to stimulate its activity. TcPI-PLC has high similarity to the PI-PLC expressed in mammalian sperm that is known as PI-PLCζ (PI-PLC zeta). As PI-PLCζ (9), TcPI-PLC has high Ca²⁺ sensitivity (6). PI-PLCζ has been proposed to be the “sperm factor” that triggers Ca² release from the endoplasmic reticulum when introduced into the ooplasm at fertilization (9, 22). It is interesting that the amount of PLCζ that is required to trigger Ca²⁺ oscillations in the oocyte is within the range of PLCζ in a single sperm (22). We measured a PI-PLC activity in the supernatant of 1.2 × 10⁷ parasites of around 375 pmol of PIP₂ hydrolyzed in 20 min, which translates into 3.12 × 10⁻⁵ pmol of PIP₂ hydrolyzed in 20 min per parasite. Taking into account that the concentration of PIP₂ in a mammalian cell is ∼10 μM (13, 27) and assuming a cell volume of 520 fl (26), the amount of PIP₂ in a cell would be around 52 × 10⁻⁵ pmol, which could be hydrolyzed by the PI-PLC released by a single amastigote in about 300 min. The lack of stimulation by Ca²⁺ of the secreted PI-PLC activity could be due to partial proteolysis and elimination of the Ca²⁺-binding domains of the enzyme. This could also explain the lack of reaction with the antibody in the cytoplasm of the infected host cells, in addition to a dilution effect. We cannot completely rule out that the enzyme activity detected in the supernatants of differentiating trypomastigotes is due to another enzyme with the same substrate specificity as TcPI-PLC. However, TcPI-PLC is the only PI-PLC gene identified in the genome of this parasite, and although a protein with homology to the glycosylphosphatidylinositol phospholipase C (GPI-PLC) of T. brucei (2) was found in T. cruzi (20), the enzyme from T. brucei catalyzes only the hydrolysis of GPI or phosphatidylinositol (2).

The two peaks of expression in the plasma membrane of amastigotes during development could be related to cell signaling in the host. For example, myoblasts have been shown to possess an area of elevated Ca²⁺ surrounding intracellular amastigotes (14), and it has been reported that intracellular calcium increases after 24 h of infection of BSC-1 fibroblasts (10). Cytoskeletal alterations, more notably disarray of vimentin fibers, also occur early during infection (10). Significant changes in transcript abundance are induced in the host cells 24 h after infection, suggesting that changes in host cell gene expression may correlate with a particular parasite-dependent event (3, 25). All of these changes occur at around 24 h after infection, which is the time of maximal outer surface expression of TcPI-PLC, and are compatible with global changes induced by PIP₂ hydrolysis and IP₃-dependent Ca²⁺ signaling. For example, the role of the IP₃-dependent Ca²⁺ signals in regulation of gene expression in mammalian cells is well established (8), as is the role of PIP₂ in the formation and stability of vimentin filaments (19). PIP₂ breakdown also plays a role in regulating myoblast differentiation (11), and it has been shown that T. cruzi infection inhibits muscle differentiation (21).

In summary, targeting of TcPI-PLC to the plasma membrane and its secretion support a proposed role in cell signaling in the host and suggests a novel mechanism for host-parasite interaction.