Biochemical characterization of Leishmania major aquaglyceroporin LmAQP1: possible role in volume regulation and osmotaxis

Permeability of LmAQP1 in Xenopus oocytes

We have previously shown that, transfection of LmAQP1 in Leishmania promastigotes and amastigotes renders these cells sensitive to both arsenite and antimonite, and these cells accumulate higher levels of the metalloids than untransfected controls (Gourbal et al., 2004). The metalloid-permeability of LmAQP1-expressing oocytes was examined in this study. Oocytes injected with LmAQP1-cRNA accumulated 40- and 240-fold higher As(III) and Sb(III), respectively, compared with the control set (Fig. 1A). These data reaffirm our previous observation that the Leishmania LmAQP1 serves as a metalloid channel.

The water permeability of LmAQP1 was evaluated by expression in Xenopus oocytes and compared with human AQP1. After 3 days of incubation at 16°C, oocytes expressing LmAQP1 or AQP1 were subjected to hypotonic shock in 1:3 diluted ND96 medium, which established an outwardly directed osmotic gradient of 140 mOsm. As shown in Fig. 1B, oocytes injected with LmAQP1-cRNA exhibited a 20-fold (P_f ≈ 180 μm s⁻¹) increase in swelling rate compared with the water-injected control. Under similar conditions, AQP1-cRNA rendered the oocytes 30 times (P_f ≈ 300 μm s⁻¹) more water-permeable than control oocytes. LmAQP1 exhibited a P_f value similar to that of the TbAQPs (Uzcategui et al., 2004), and therefore can be classified as having intermediate water permeability, approximately 60% that of AQP1. Interestingly, treatment with 0.5 mM HgCl₂ did not affect the water permeability of oocytes injected with LmAQP1-cRNA, although water transport was inhibited in AQP1-expressing oocytes under similar conditions (Fig. 1B). Thus, in contrast to PbAQP and TcAQP, which are sensitive to HgCl₂ (Montalvetti et al., 2004; Promeneur et al., 2007), LmAQP1 is a mercurial-insensitive water channel.

The glycerol permeability of LmAQP1 was determined by an isosmotic swelling assay with a 130 mM inwardly directed gradient of glycerol. This concentration gradient led to an influx of glycerol, resulting in a secondary influx of water, and consequently swelling of the oocytes. As shown in Fig. 2, injection of oocytes with LmAQP1-cRNA increased the glycerol permeability by ∼25-fold when compared with the water-injected control. These results are comparable to that reported for other protozoal aquaglyceroporins (Hansen et al., 2002; Uzcategui et al., 2004; Promeneur et al., 2007). The above studies demonstrate that LmAQP1 is highly permeable to both water and glycerol.

Aquaglyceroporins are able to transport a wide variety of solutes. Human aquaglyceroporin-9 (AQP9), a homologue of LmAQP1, has been shown to be a promiscuous solute channel that, besides water, also allows the passage of a wide variety of non-charged solutes, including carbamides, polyols, purines and pyrimidines (Tsukaguchi et al., 1999). To determine the pore specificity of LmAQP1, we analysed the swelling rates of LmAQP1-expressing oocytes in the presence of a variety of non-ionic solutes (Fig. 2). The solute permeability decreased in the order: methylglyoxal > glycerol > dihydroxyacetone > glyceraldehyde > erythritol (C-4 sugar alcohol) > adonitol (C-5 sugar alcohol). Solute permeability was inversely proportional to the carbon chain length; C3-compounds (methylglyoxal, glycerol, dihydroxyacetone, glyceraldehyde) permeated better than C4 and C5-sugar alcohols, while the C6-sugar alcohols (mannitol and sorbitol) showed negligible transport. Additionally, there was minimal transport of urea through the LmAQP1 channel. This low transport of urea probably helps the parasite to survive within the host liver. The high permeability of methylglyoxal is surprising because methylglyoxal is known to be toxic to cells. This high uptake may explain the antileishmanial activity of methylglyoxal, reported earlier (Sarkar and Bhaduri, 1986). Interestingly, although Xenopus oocytes expressing LmAQP1 showed high uptake of methylglyoxal, Leishmania donovani promastigotes overexpressing LmAQP1 were not hypersensitive to methylglyoxal compared with control cells transfected with vector only (data not shown). This may be because Leishmania can metabolize methylglyoxal to d-lactate (Darling and Blum, 1988). The specific activity of Leishmania methylglyoxal reductase has been found to be the highest of all the catabolic enzymes (Ghoshal et al., 1989).

Expression and localization of LmAQP1

To examine the expression of LmAQP1, a rabbit polyclonal antibody was raised against a synthetic peptide, corresponding to amino acids 139–152 of LmAQP1. Using the affinity-purified LmAQP1 antiserum, Western blot analysis of X. laevis oocytes injected with LmAQP1-cRNA showed a ∼28 kDa protein band (Fig. 3C). In contrast, no such band was detected in oocytes injected with water. Similarly, the pre-immune serum did not give any reaction (data not shown). This 28 kDa band corresponds to the predicted molecular mass of LmAQP1.

Immunoblotting was next performed on plasma membrane-enriched fractions isolated from LmAQP1-transfected L. donovani promastigotes (Fig. 3A). A faint ∼28 kDa band was observed in cells transfected with LmAQP1 but not in vector alone controls. Western blotting was performed on flagellar fraction that was purified by shearing followed by differential centrifugation on sucrose gradients. Immunoblot of the flagellar fraction probed with the LmAQP1 antibody, clearly showed a ∼28 kDa band, suggesting that LmAQP1 was located exclusively to the flagella (Fig. 3A). Immunofluorescence experiments were performed to reveal the localization of LmAQP1 in L. donovani promastigotes and amastigotes (Fig. 3B). Fixed promastigotes, examined by fluorescence microscopy, showed selective immunostaining of the flagella. No staining was visible on the main body of the parasite. In amastigotes, staining was observed in the anterior part of the cell, most likely in the flagellar pocket. Additionally, a prominent medial band of staining was also noted, probably representing the contractile vacuole/spongiome complex (cf. Fig. 4B).

To confirm our observations of LmAQP1 localization obtained by fluorescence microscopy, we also performed immuno-electron microscopy on amastigote forms of the LmAQP1-transfected cells. The LmAQP1 primary antibody was detected with a secondary antibody labelled with gold particles. In amastigotes, particles were detected on the flagellar remnant and surrounding flagellar pocket membranes, the kinetoplast-mitochondrion, and the contractile vacuole/spongiome complex (Fig. 4). Identification of the organelles is based upon cellular morphology, as described (Linder and Staehelin, 1979; Attias et al., 1996). Labelling of the exterior plasmalemma was found to be extremely rare in amastigotes.

Effect of Arg230 mutation on LmAQP1 activity

Molecular dynamics simulation of solute permeation through AQP1 and the bacterial glycerol facilitator GlpF demonstrates that the conserved arginine residue (Arg195 in AQP1 and Arg206 in GlpF) functions as a filter for protons and other positive ions (de Groot and Grubmuller, 2001). The corresponding residue in LmAQP1 is Arg230. The role of Arg230 in solute permeation was examined by altering it to lysine, which retains the positive charge, and to alanine, which has lost the charge and is also smaller in size. The effect of mutagenesis was examined in both Leishmania promastigotes and Xenopus oocytes. Expression of the altered proteins was detected by Western blotting. Both R230A and R230K mutants were expressed on oocyte membranes at levels similar to that of wild-type LmAQP1 (Fig. 3C). Additionally, immunofluorescence staining shows that either R230A or R230K LmAQP1 is localized to the flagellum of promastigotes (Fig. 3D).

The effect of the mutations on water permeability was examined using the swelling assays in Xenopus oocytes. The R230A mutant was non-functional while the R230K mutant had only 20% of the wild-type water transport activity (Fig. 5A). The permeability of other solutes was also examined. The R230A mutant showed zero to negligible transport of glycerol, dihydroxyacetone, glyceraldehyde and adonitol (Fig. 5B). In contrast, although the R230K mutant was impermeable to adonitol, it still showed 5–10% of the wild-type activity with respect to the transport of glycerol, dihydroxyacetone and glyceraldehyde (Fig. 5B). Interestingly, both R230A and R230K mutant transported methylglyoxal at ∼40% of wild-type levels (Fig. 5B).

The effect of Arg230 mutation on metalloid sensitivity and transport was also examined in both L. donovani promastigotes and amastigotes. Promastigotes transfected with the wild-type LmAQP1 gene exhibited 60- and 360-fold higher sensitivity to As(III) and Sb(III) respectively (Fig. 6A and B), and accumulated high levels of both metalloids (Fig. 6C and D). In contrast, cells expressing either the R230A or R230K LmAQP1 were resistant to both As(III) and Sb(III) (Fig. 6A and B), and accumulated low levels of the metalloids, similar to the vector alone control (Fig. 6C and D). Similar observations were made when either the wild-type or mutant LmAQP1 was expressed in axenic amastigotes and Xenopus oocytes. Axenic amastigotes, overexpressing wild-type LmAQP1, accumulated higher amounts of metalloids compared with R230A and R230K mutants and the vector alone control (Fig. 6E and F). Oocytes expressing R230A LmAQP1 did not accumulate either As(III) or Sb(III) (Fig. 6G and H). However, oocytes injected with R230K LmAQP1-cRNA transported low levels of As(III) but not Sb(III) (Fig. 6G and H).

Role of LmAQP1 in volume regulation

Upon exposure to hypoosmotic stress, several mammalian cell types as well as a number of protozoa, swell initially but later return to their normal volume. This phenomenon, dubbed as the regulatory volume decrease, is accomplished by the efflux of various osmolytes to the extracellular environment followed by passive water flow (Rohloff et al., 2003). L. donovani promastigotes and amastigotes transfected with either the wild-type or R230A LmAQP1 were subjected to a 50% reduction in osmolarity (from 300 to 150 mOsm). As cell swelling leads to a decrease in the absorbance, the process of volume recovery was monitored by following the absorbance of the cell suspension at 550 nm. Subsequent to a hypoosmotic shock, L. donovani promastigotes expressing wild-type LmAQP1 showed a drop in absorbance, indicating cell swelling, followed by a steady rise in absorbance to near original levels, signifying volume recovery (Fig. 7A). The volume recovery process was essentially complete over a 3 min period. In contrast, promastigotes transfected with either the R230A LmAQP1 gene or the vector control showed increased cell swelling, as reflected by a higher drop in absorbance, and slower volume recovery than cells expressing the wild-type protein (Fig. 7A). Similar results were observed with axenic amastigotes. Figure 7B shows that following a hypoosmotic stress, amastigotes overexpressing wild-type LmAQP1 swelled up as reflected by a decrease in absorbance, and recovered their volumes much faster than the R230A mutant or the vector alone control. A similar decline in volume regulation was also observed in cells expressing R230K LmAQP1 (data not shown). Although in all cases, the volume recovered cells were virtually identical in terms of motility and morphology from control cells maintained under osmotic conditions, cells expressing wild-type LmAQP1 seem to be better equipped in combating abrupt osmotic changes in the environment.

Role of LmAQP1 in osmotaxis response of promastigotes

Leslie et al. (2002) have shown that Leishmania promastigotes move towards concentrations of sugar substrates and this taxis requires the presence of an osmotic gradient. This phenomenon of osmotaxis has been proposed to be involved in the migration of Leishmania promastigotes from the midgut to the proboscis of the sandfly, during the insect phase of the parasite life cycle, which is essential for successful transmission of the parasite to the vertebrate host (Oliveira et al., 2000; Leslie et al., 2002). The exclusive flagellar localization of LmAQP1 in Leishmania promastigotes prompted us to examine the osmotaxis response in cells overexpressing LmAQP1. Leishmania promastigotes transfected with vector, wild-type LmAQP1, R230A LmAQP1 and R230K LmAQP1 were subjected to osmotaxis assay as described in Experimental procedures. In this assay, a capillary tube is filled with agarose gel containing 100 mM glucose, while leaving 1 cm of the end of the tube filled with buffer only. The diffusion of glucose from the agarose filled to the open end of the capillary tube establishes a concentration gradient that simulates the solute gradient within the insect gut. The number of promastigotes that move over a certain interval from a buffered cell suspension into the open end of a capillary tube was determined. A significant higher number of promastigotes, overexpressing wild-type LmAQP1, were isolated from the capillary indicating that these cells moved faster towards higher concentrations of glucose, compared with the vector alone control (Fig. 8). Thus, promastigotes expressing wild-type LmAQP1 exhibited faster osmotaxis than the vector control. In contrast, promastigotes expressing altered R230A and R230K LmAQP1 showed similar osmotaxis response as the vector control. The low level of osmotaxis observed in Leishmania promastigotes transfected with either the vector or altered LmAQP1 is most likely due to expression of the chromosomal copy of LmAQP1 in these cells. These results clearly indicate that LmAQP1 plays a role in osmotaxis.

Strains and media

Leishmania donovani strain LdBob was a kind gift from Professor Stephen M. Beverley, Washington University School of Medicine. LdBob was cycled between the promastigote and axenic amastigote forms using an established protocol (Goyard et al., 2003). Leishmania promastigotes and amastigotes were grown in culture media as described earlier (Goyard et al., 2003). Promastigotes were grown at 25°C while the axenic amastigotes were cultivated at 37°C with 5% CO₂. X. laevis were maintained in our animal facility and oocytes were harvested periodically.

Oligonucleotide-directed mutagenesis

The cloning of wild-type LmAQP1 into pGEM-T Easy vector (LmAQP1/pGEMEasy) has been described previously (Gourbal et al., 2004). Mutations in LmAQP1 were introduced by site-directed mutagenesis using the QuikChange Site-Directed Mutagenesis Procedure (Stratagene), as described previously (Mukhopadhyay et al., 2003). The mutagenic oligonucleotides used for both strands and the respective changes introduced (underlined) are as follows: R230A, 5′-GC CTC AAC CCC GCA GCC GAT TTG TCA CCT C-3′ (sense) and 5′-G AGG TGA CAA ATC GGC TGC GGG GTT GAG GC-3′ (antisense); R230K, 5′-GGC CTC AAC CCC GCA AAG ATT TGT CAC CTC GC-3′ (sense) and 5′-GCG AGG TGA CAA ATC CTT TGC GGG GTT GAG GCC-3′ (antisense). Each mutation was confirmed by sequencing the entire gene using a CEQ2000 DNA sequencer (Beckman Coulter).

Expression of LmAQP1 in oocytes

For expression in Xenopus oocytes, wild-type and mutant LmAQP1 in pGEM-T Easy vector were subcloned into the EcoRI site of the pL2-5 vector (Seyfang et al., 1997) (a generous gift from Professor Scott M. Landfear, Oregon Health Sciences University). Following linearization with SmaI, cRNAs were transcribed in vitro using the mMessage mMachine^® kit (Ambion). Stage V and VI defolliculated X. laevis oocytes were injected with 50 nl of water (control), 50 nl of water containing 10 ng of cRNA (wild-type, R230A or R230K LmAQP1), 50 nl of water containing 5 ng of AQP1-cRNA, or combination of 10 ng (wild-type, R230A or R230K LmAQP1) and 5 ng of AQP1-cRNA. The oocytes were maintained at 16°C for 3 days in ND96 buffer containing 5 mM HEPES, pH 7.4, 96 mM NaCl, 2 mM KCl, 1.8 mM CaCl₂, 1 mM MgCl₂, 2.5 mM sodium pyruvate, 0.5 mM theophylline and 2 μg ml⁻¹ of gentamicin sulphate.

Standard oocyte swelling assays

Oocyte swelling assays were performed as described by Hansen et al. (2002). Water permeability was measured by transferring oocytes into 1:3 diluted ND96 medium. Solute permeability was measured in an isoosmotic ND96 in which 65 mM NaCl was substituted with 130 mM of a non-ionic test solute. While determining the permeability of methylglyoxal, NaCl was entirely replaced with 115 mM of methylglyoxal to maintain buffer isoosmoticity. For inhibition experiments, oocytes were pre-incubated in 0.5 mM HgCl₂ in ND96 buffer, washed twice in the same buffer, before subjecting to swelling assay. The swelling assays were performed at room temperature and video-monitored. The relative oocyte volume was calculated from the covered area. Osmotic water permeability (P_f, μm s⁻¹) and solute permeability (Ps) were calculated using the following equation:

(1)

(2)

where, S is the oocyte surface area (= 0.045 cm²), V₀ is the initial oocyte volume (= 9 × 10⁻⁴ cm³), V_w represents the molecular water volume (= 18 cm³ mol⁻¹), osm_in − osm_out is the osmotic gradient, osm_total is the total osmolarity of the system (200 mOsm), sol_out − sol_in is the osmotic solute gradient, and d(V/V₀)/dt in s⁻¹ is measured from the initial slope of the relative volume increase.

Cloning of LmAQP1 in Leishmania expression vector pSP72αhygroα

pSP72αhygroα was created on the backbone of pSP72 (Promega) and pGEM7αhygroα vectors. pGEM7αhygroα was created as described previously (Papadopoulou et al., 1992; Gourbal et al., 2004). Either the wild-type or mutant LmAQP1/pGEMEasy was digested with XbaI and HindIII and cloned into the same sites of pSP72αhygroα.

Transfection of LmAQP1 in Leishmania

Transfection of pSP72αhygroα bearing either the wild-type or mutant LmAQP1 gene into the Leishmania promastigotes was accomplished as described previously (Ouellette et al., 1990). All transfectants were maintained in the presence of 0.3 mg ml⁻¹ hygromycin (Invitrogen).

Determination of EC₅₀ of metalloids in promastigotes

Metalloid sensitivity of the promastigote transfectants were determined as described previously (Gourbal et al., 2004). Briefly, log-phase promastigote cultures were diluted to 10⁶ cells ml⁻¹ in a culture medium containing various concentrations of As(III) in the form of sodium arsenite, or Sb(III) in the form of potassium antimonyl tartrate. Growth was monitored from the absorbance at 600 nm after 5 days using a microplate reader (Spectramax 340, Molecular Devices). EC₅₀ was calculated from the sigmoidal analysis of percentage live cells versus metalloid concentrations using the Origin Graphing Software. Each assay was performed at least three times. Error bars were calculated from the mean ± SE.

Uptake assays

Log-phase Leishmania promastigotes or amastigotes were washed twice with phosphate-buffered saline (PBS), pH 7.4 (Invitrogen) and suspended in PBS at a density of 10⁸ cells ml⁻¹. The promastigotes were then incubated with 10 μM As(III) or Sb(III) for 10 min at room temperature. A 0.2 ml portion was filtered through a 0.22 μm nitrocellulose filter and the filter washed once with 5 ml of ice-cold PBS. The filters were digested with 0.5 ml of concentrated HNO₃ (69–70%) (EM Science) for 1 h at 70°C, allowed to cool to room temperature, and diluted with high pressure liquid chromatography grade water (Sigma) to produce a final concentration of HNO₃ of approximately 3%, and then analysed by inductively coupled plasma mass spectrometry (ICP-MS) with a PerkinElmer ELAN 9000. Standard solutions were made in the range of 0.5–10 p.p.b. in 3% HNO₃ using arsenic and antimony standards (Ultra Scientific). Each transport experiment was repeated at least three times with duplicate samples. Error bars were calculated from the mean ± SE.

For measurement of metalloid transport in oocytes, five oocytes were incubated with either 1 mM As(III) or Sb(III) in ND96 buffer, for 90 s, at room temperature. The oocytes were washed four times with 1 ml of ice-cold ND96 buffer. Each oocyte was then transferred to a separate tube and digested with 0.1 ml of concentrated HNO₃ as described above. Each sample was diluted with high pressure liquid chromatography grade water (Sigma) to produce a final concentration of HNO₃ of approximately 3%, and the metalloid concentration was determined by ICP-MS. Each transport experiment was repeated at least three times. Error bars were calculated from the mean ± SE.

Cell volume measurements

Relative changes in cell volume following the induction of hypoosmotic shock were followed as described earlier (Rohloff et al., 2003). Briefly, log-phase promastigotes/amastigotes were washed twice in PBS and resuspended at a density of 10⁹ cells ml⁻¹. One hundred microlitre portions of the cell suspension were transferred to a microtiter plate. Hypoosmotic shock was induced by dilution of the isotonic cell suspension with an equal volume of deionized water and the absorbance at 550 nm was recorded every 15 s for 3 min in a microplate reader (Spectramax 340, Molecular Devices). A decrease in absorbance corresponds to an increase in cell volume. Isosmotic control experiments consisted of dilution of cell suspensions with appropriate volumes of isosmotic buffer. Unless otherwise noted, all hypoosmotic shock experiments were conducted at a final osmolarity of 150 mOsm (1:1 dilution of isosmotic buffer and water). Each experiment was repeated at least three times. Error bars were calculated from the mean ± SE.

Osmotaxis assay

Migration of promastigotes through osmotic gradients was measured as described earlier (Oliveira et al., 2000; Leslie et al., 2002). Glass capillary tubes (75 mm long × 1 mm internal diameter) were loaded with a 1% agarose solution, prepared with either PBS or PBS containing 100 mM d-glucose, such that exactly 1 cm at the end of each tube remained unoccupied by the gel (Leslie et al., 2002). Late-log-phase LdBob promastigotes, overexpressing wild-type LmAQP1 or altered LmAQP1 or vector controls, were washed once with PBS, resuspended in the same buffer (5 ml) at a density of 2.5 × 10⁷ cells ml⁻¹, and placed in a microcentrifuge tube. The open end of each capillary tube was filled with PBS, and the capillary tubes were suspended vertically into the microcentrifuge tube, so that the open end of each capillary was immersed in the promastigote suspension. The capillary tubes were incubated at room temperature for 1 h. The capillary tubes were removed carefully from the cell suspension and the buffer (approximately 6 μl) was collected from the open end of each tube using a fine pipette-tip and mixed with 5 μl of PBS. The density of the promastigotes in each sample was determined using a haemocytometer and the mean density and standard error was calculated for both control and test samples. Interexperimental variability was eliminated by expressing the mean number of cells moving into the capillary tube containing 100 mM glucose relative to the mean number of cells moving towards the capillary tube in absence of glucose (attraction coefficient). An attraction coefficient of 1 thus indicates no specific movement towards glucose. Each experiment was repeated at least three times. Error bars were calculated from the mean ± SE.

Cell lysates and western blots

Plasma membrane-enriched fraction from Leishmania promastigotes was prepared as described previously (Mukhopadhyay et al., 1996). All procedures were performed at 4°C. Briefly, after washing the cells in Buffer A (75 mM Tris-HCl, pH 7.4, 140 mM NaCl, 11 mM KCl), cells were resuspended in Buffer B [10 mM HEPES, pH 7.4, 10 mM KCl, 1 mM MgCl₂, 400 mM mannitol, and Complete Protease Inhibitor Cocktail (Roche)] and broken by sonication using five pulses of 10 s each with 50% amplitude. Unbroken cells and cell debris were removed by centrifugation at 3000 g for 10 min. The supernatant was centrifuged at 17 000 g for 40 min to remove organelles and other intracellular membranes followed by centrifugation at 140 000 g for 1 h. The pellet, which is highly enriched in plasma membrane fraction, was washed and suspended in 1 ml of Buffer C (75 mM HEPES, pH 7.4, 150 mM KCl, 2 mM MgCl₂).

Flagella were isolated as follows. Promastigotes from a 4-day-old, 100 ml culture were harvested, washed twice in ice-cold Buffer D (25 mM Tris-HCl, pH 7.4, 0.2 mM EDTA, 5 mM MgCl₂, 12 mM β-mercaptoethanol, and 0.32 M sucrose) and resuspended in 2 ml of ice cold buffer E [Buffer D plus 1% BSA, 0.1 mM CaCl₂ and Complete Protease Inhibitor Cocktail (Roche)]. Flagella were detached without lysing the body of the cells by sonication on ice using three pulses of 1 s each, with 25% amplitude. Flagellar preparation was isolated from the supernatant by centrifuging at 700 g for 15 min followed by pelleting at 6780 g for 20 min. The pellet was resuspended in buffer A and loaded onto 0.8 M sucrose cushion and centrifuged at 1080 g for 20 min. The upper layer was removed and overlaid on a discontinuous gradient of sucrose (1.65 M and 1.85 M) and ultracentrifuged at 130 000 g for 3 h. The interphase flagellar layer was removed and examined by phase-contrast microscopy to confirm that the flagella were largely free of cell bodies. Protein contents of either the plasma membrane enriched or flagellar fractions were determined by the BCA protein assay kit (Pierce).

A polyclonal anti-peptide antibody against the synthetic peptide, SHFDEAEKRLLLNE, corresponding to amino acids 139–152 of L. major, was developed at Covance Immunology Services (Denver, PA) and affinity purified. For Western blot analysis, protein samples in 1× SDS gel-loading buffer (Sambrook and Russell, 2001) were heated to 50°C for 1 h, electrophoresed into a 12% SDS-polyacrylamide gel, and electro-blotted onto nitrocellulose membrane (Whatman). The blots were incubated with affinity-purified LmAQP1 antiserum (1:500) for 1 h at room temperature, and the labelling was visualized with horseradish peroxidase-conjugated goat anti-rabbit antiserum (Sigma) using the Western Lightning Chemiluminescence Reagent Plus system (PerkinElmer).

Immunofluorescence analysis

Late-log-phase promastigotes or amastigotes (∼5 × 10⁵ cells) were fixed with 4% paraformaldehyde/PBS overnight at 4°C. The cells were washed twice with PBS, transferred to Teflon-coated Immunofluorescence slides (Polysciences), and permeabilized with 0.1% Triton X-100, 50 mM Na₂PO₄, 50 mM glycine, pH 7.2, for 2 min. After blocking with 1% bovine serum albumin in PBS for 30 min, cells were incubated for 1 h at room temperature with LmAQP1 antiserum (1:300 in 1% BSA). The primary antibody was removed by washing three times with 1% BSA followed by incubation with the monoclonal anti-α-tubulin antibody (Sigma) 1:500 for another hour. After three washes with 1% BSA, the slides were incubated with Alexa Fluor 488-conjugated goat anti-rabbit IgG (1:1000 in 1% BSA) and Alexa Fluor 594-conjugated goat anti-mouse IgG (1:2000 in 1% BSA) for 1 h. Finally, the cell nuclei were stained with DAPI (0.1 μg ml⁻¹ final concentration) (Fluka) for 10 min, the slides washed three times with 1% BSA, and once with sterile water. The slides were mounted in Fluoromount-G (SouthernBiotech) and examined by Zeiss Axiophot Photomicroscope with a 63× water immersion objective lens.

Immuno-electron microscopy

LdBob promastigotes or amastigotes were processed for immuno-electron microscopy as described previously (Russell et al., 1992; Piper et al., 1995). Briefly, cells were fixed in 4% formaldehyde in HEPES saline (30 mM HEPES, 100 mM NaCl, 0.4 mM CaCl₂), pH 7.0, and rinsed in HEPES saline alone. They were then pelleted, embedded in gelatin, and infiltrated with 2.3 M sucrose/20% polyvinyl pyrrolidone in PBS. The blocks were trimmed, frozen and sectioned in a Reichert FC4E cryoultramicrotome. The cryosections were placed on carbon-coated formvar grids, blocked with 10% fetal calf serum (FCS) in PBS, incubated with LmAQP1 antiserum (1:150 in 10% FCS in PBS), washed in the same solution without antibody, followed by incubation with gold-conjugated secondary antibody. The grids were then washed sequentially in 10% FCS in PBS, PBS, and double-distilled water, prior to staining with uranyl acetate. All labelling experiments were conducted, in parallel, with controls omitting the primary antibody, and in the presence of non-immune normal rabbit serum or unrelated mouse monoclonal antibody against other Leishmania-specific antigens. These controls were consistently negative.