|Home | About | Journals | Submit | Contact Us | Français|
A Giardia-specific protein family denominated as α-giardins, represents the major protein component, besides tubulin, of the cytoskeleton of the human pathogenic parasite Giardia lamblia. One of its members, α19-giardin, carries an N-terminal sequence extension of MGCXXS, which in many proteins serves as a target for dual lipid conjugation: myristoylation at the glycine residue after removal of the methionine and palmitoylation at the cysteine residue. As the first experimental evidence of a lipid modification, we found α19-giardin to be associated with the membrane fraction of disrupted trophozoites. After heterologous coexpression of α19-giardin with giardial N-myristoyltransferase (NMT) in Escherichia coli, we found the protein in a myristoylated form. Additionally, after heterologous expression together with the palmitoyl transferase Pfa3 in Saccharomyces cerevisiae, α19-giardin associates with the membrane of the main vacuole. Immunocytochemical colocalization studies on wild-type Giardia trophozoites with tubulin provide evidence that α19-giardin exclusively localizes to the ventral pair of the giardial flagella. A mutant in which the putatively myristoylated N-terminal glycine residue was replaced by alanine lost this specific localization. Our findings suggest that the dual lipidation of α19-giardin is responsible for its specific flagellar localization.
The human pathogenic parasite Giardia lamblia (syn. Giardia intestinalis), a phylogenetically basal eukaryote (41), is the causative agent of giardiasis, an intestinal disease most prevalent in developing countries (39). The protist has a simple life cycle consisting of two stages, a vegetative trophozoite dwelling in the host intestine and an infective cyst form. Proliferating trophozoites are distinguished by a complex cytoskeleton whose most striking feature are eight flagella and a ventral disk, by which the parasite attaches to the intestinal epithelium of the host (11). As a diplomonad protist, the parasite possesses four different pairs of flagella, of which only the ventral and posterolateral ones are replicated in the first round of the cell cycle; the other two pairs require two further cell divisions for their complete renewal (32).
Besides tubulin, the Giardia-specific giardins account for the major protein components of the giardial cytoskeleton (7, 36). From these 30- to 45-kDa proteins, the α-giardins have been recognized, based on sequence similarities, as annexin homologues (28). They represent a multiple set of all-helical proteins distinguished by four annexin domains each. We have previously confirmed that some α-giardins, indeed, satisfy a criterion of annexins, i.e., Ca2+-dependent association with phospholipids (1, 13, 43). Hence, these giardins have been nominated as protozoan annexins E1 to E3 (9). However, in contrast to annexins of multicellular organisms, some α-giardins contain specific sequence motifs in the fourth annexin domain, and these motifs are located at the cytoplasmic face and may make contact between the plasma membrane and the cytoskeleton within the cell (35, 43).
The genome of G. lamblia possesses a total of 21 α-giardin-encoding genes (27, 30, 47). The particular importance of their gene products for the parasite may be indicated by the following: the human genome contains only 12 annexin genes, and the genome of Saccharomyces cerevisiae has none at all. Phylogenetic analyses of the α-giardin genes revealed that 19 members of this family evolved from two widely ramified branches, whereby the genes coding for α14-giardin (annexin E1, according to the annexin nomenclature) and α19-giardin jointly form a single arm (47). α19-Giardin, the subject of the present study, carries a predicted N-terminal sequence extension with an MGCXXS motif known as a target for dual fatty acylation, i.e., myristoylation at the N-terminal glycine and palmitoylation at the cysteine residue (8). However, no experimental data for a lipid conjugation to this protein and any other giardin are currently available.
In the present study, we provide the first evidence that α19-giardin indeed can be both myristoylated and palmitoylated. In contrast to α14-giardin, which we found to be located in all giardial flagella as well as in the median body of the trophozoites (43, 46), α19-giardin appears exclusively localized to the ventral flagella of the trophozoites and is restricted to those portions protruding outside the cell body.
Trophozoites of G. lamblia strain WBC6 (ATCC 50803) were cultured in Keister's modified TYI-S-33 medium (20). Cells were harvested at the end of the logarithmic phase (after 3 to 4 days), washed two times in a buffer composed of 20 mM Tris-150 mM NaCl, pH 6.8 (TBS), and stored at −20°C in the presence of 10 μM (final concentration) trans-epoxysuccinyl-l-leucylamide-(4-guanidino)butane (E-64), a potent inhibitor of cysteine proteases. Giardia trophozoites were synchronized by a two-step procedure following the protocol of Poxleitner et al. by using nocodazole and aphidicolin (Sigma) (37).
Total RNA was extracted from trophozoites using Trizol reagent (Gibco BRL) and quantified by spectrophotometry (260 nm). For reverse transcriptase PCR (RT-PCR), cDNA was synthesized using RevertAid H Minus M-MuLV Reverse Transcriptase and DNase I (Fermentas). Both experiments were performed according to the protocol provided by the manufacturer. Primers for amplifying the α19-giardin DNA were synthesized using sequence information from the Giardia genome project (27): GTT TGT TAG AGC ATG GGT TGT GCC GCA TCA ACT CCC (sense) and GTC CAT CGA GTT CCC GGG TCA GTC GCC GCG (antisense). The thermal profile for PCR was as follows: 5 min at 95°C and 30 cycles of 30 s at 60°C, 1 min 72°C, and 15 s 94°C, followed by 2 min at 60°C and 5 min at 72°C. The PCR products were analyzed by electrophoresis in 1% agarose gels, followed by ethidium bromide staining, and photographed under UV light.
Genomic DNA was isolated from fresh trophozoites using an Elu-Quick kit (Schleicher & Schuell). The open reading frame of α19-giardin was amplified via PCR using the following primers constructed according to sequence information from the Giardia genome database (27): GCT GAT GCC AAG CTT GTA ATG GGA AAC CAC (sense) with a restriction site (shown in bold) for HindIII and GTC CAT CGA GTT CCC GGG TCA GTC GCC GCG (antisense) with an SmaI restriction site (in boldface). The obtained gene fragment was ligated in frame into the multiple cloning site of the expression vector pJC45, a derivative of pJC40, which contains a 50-nucleotide extension sequence coding for 10 histidine residues (His tag) (5). Recombinant plasmids were transformed into Escherichia coli BL21(DE3)/pAPlacIQ. Freshly transformed bacteria were inoculated into LB medium (200 μg/ml ampicillin) and grown at 37°C until the optical density at 600 nm reached 0.5. Recombinant expression was induced by induction with isopropyl-1-thio-β-d-galactopyranoside at a final concentration of 0.8 mM. Culturing was continued for 3 h. Recombinantly overproduced α19-giardin with a deletion of residues 1 to 16 (α19-giardin Δ1-16) was applied on a Ni2+-nitrilotriacetic acid column in binding buffer (5 mM imidazole, 0.5 M NaCl, 20 mM Tris-HCl, pH 7.9), eluted with 0.25 to 0.5 M imidazole in the same buffer, and analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE).
SDS-PAGE was performed in the Tris-glycine system of Douglas et al. (6), using 12% gels. Proteins were visualized by dispersion staining with Coomassie brilliant blue G-250 (31). Western blotting was performed according to Matsudaira (25) in a CAPS [3-(cyclohexylamino)-1-propanesulfonic acid]-NaOH-buffer, pH 11.0, containing 10% methanol. For immune decoration of the blots, rabbit antiserum against recombinant α1-giardin and α14-giardin in dilutions of 1:10,000, α19-giardin in a dilution of 1:10,000 to 1:100,000, and a monoclonal antibody against anti-acetylated tubulin (mouse immunoglobulin 2b [IgG2b] isotype; Sigma) in a dilution of 1:2,000 were used. For secondary antibodies, we employed goat anti-rabbit IgG (1:20,000; Pierce) and goat anti-mouse IgG (1:1,000; Biomol), both coupled to horseradish peroxidase. Blots were developed with 4-chloro-1-naphthol as a chromogenic substrate or by chemiluminescence (Pierce).
For phospholipid binding studies, the method of Boustead et al. (3) was followed. Briefly, 400 μl of 20 mM HEPES-NaOH, pH 7.4, containing 100 mM KCl, 2 mM MgCl2, 1 mM EGTA, 5 mM dithiothreitol, and 0.5 mg of multilamellar liposomes (brain extract; Sigma) was added to 10 μg of recombinant α19-giardin Δ1-16. The mixture was incubated for 40 min at room temperature under agitation and then centrifuged for 10 min at 15,000 × g. In a parallel experiment, the free Ca2+ concentration in the mixture was adjusted to 3.8 mM; in control incubations the brain extract was omitted. The reversibility of Ca2+-dependent phospholipid binding of α19-giardin was examined via successive additions of EGTA (10 mM) to the pellet fraction. The supernatant and pellet fractions were subjected to SDS-PAGE and Western blotting using the anti-α19-giardin antiserum.
Trophozoites of G. lamblia were attached to polylysinated coverslips at 37°C, fixed for 7 min with methanol, and permeabilized for 10 min with acetone, both at −20°C. In the following, the cells were rehydrated for 10 min with TBS and then incubated for 30 min with blocking buffer (3% bovine serum albumin [BSA] in TBS). After reacting with the rabbit anti-α19-giardin antiserum for 1 h or with the antitubulin antibody (both at 1:500), the samples were washed three times with TBS and then incubated in the dark for 1 h with Cy3-conjugated anti-rabbit F(ab)2 fragment from sheep (1:100; Sigma) or Cy2-conjugated goat anti-mouse IgG F(ab)2 fragment (1:100; Dianova). After three washes with TBS again, the cells were analyzed using a fluorescence microscope (Leica DM 5500 B; Wetzlar, Germany).
The complete open reading frame (GL-50803_4026) (27) was PCR amplified from genomic DNA using the oligonucleotide primers Alpha19Gi-HA-PacI (antisense) CGT TAA TTA ATC AAG CGT AGT CTG GGA CAT CGT ATG GGT AAG CGC CGC GGG GAG TCG AGG ATTC and Alpha19Gi-XbaI (sense) CGT CTA GAG GTG GCA CGA ACA CCT TTAG introducing a hemagglutinin (HA) epitope tag at the 3′ end. The ~1,450-nucleotide fragment was cloned into the XbaI and PacI restriction sites of the pPacV-integ expression vector (19). The vector was linearized by restriction digestion with SwaI and electroporated into trophozoites as described previously (14). Transgenic parasites were selected by resistance against puromycin. Selection was discontinued as soon as the resistant population emerged (5 to 7 days). Alternatively, to increase the signal in immunofluorescence of the epitope-tagged G2A (substitution of alanine for glycine at position 2) variant, which had lost the ability to localize to the ventral flagella, the circular plasmid was electroporated and maintained episomally under constant selection. Mutant variants (G2A and C3A) were generated by site-directed mutagenesis of the wild-type gene. A chimeric protein comprised of full-length α19-giardin fused to green fluorescent protein (α19-giardin::GFP) was constructed in the pPacV-integ expression vector.
Cells were harvested by cooling and centrifugation at 900 × g for 10 min. Fixation and preparation for fluorescence microscopy were done as described previously (24). Briefly, cells were washed with cold phosphate-buffered saline (PBS) and fixed with 3% formaldehyde in PBS for 40 min at 20°C, followed by a 5-min incubation with 0.1 M glycine in PBS. Cells were permeabilized with 0.2% Triton X-100 in PBS for 20 min at room temperature and blocked overnight in 2% BSA in PBS. Incubations of fixed cells with a mouse monoclonal Alexa Fluor 488-conjugated anti-HA antibody (dilution, 1:30; Roche Diagnostics, Mannheim, Germany) were done in 2% BSA-0.2% Triton X-100 in PBS for 1 h at 4°C. Washes between incubations were done with 0.5% BSA-0.05% Triton X-100 in PBS. Labeled cells were embedded with Vectashield (Vector Laboratories, Inc., Burlingame, CA) containing the DNA intercalating agent 4′-6-diamidino-2-phenylindole (DAPI) for detection of nuclear DNA. Immunofluorescence analysis was performed on a Leica SP2 AOBS confocal laser scanning microscope (Leica Microsystems, Wetzlar, Germany) equipped with a glycerol objective (Leica HCX PL APO CS 63× objective/1.3 numerical aperture Corr). Image stacks were collected with a pinhole setting of 1 Airy unit and twofold oversampling. Image stacks of optical sections were further processed using the Imaris software suite (Bitplane, Zurich, Switzerland).
For live-cell microscopy, cells expressing the α19-giardin::GFP chimeric protein were harvested at 12 h postinduction and transferred to 24-well plates at a density of 6 × 106/ml. After incubation on ice for 5 to 8 h, oxygenated cells were sealed between microscopy glass slides and warmed to 21°C or 37°C. For fluorescence recovery after photobleaching and time-lapse series, images were collected on a Leica SP2 AOBS confocal laser scanning microscope (Leica Microsystems, Wetzlar, Germany) using a water immersion objective (Leica, HCX PL APO CS 63× objective/1.2 numerical aperture W Corr). The pinhole was set to 2 Airy units in order to increase the thickness of the optical sections to accommodate the moving ventral flagella in the z-plane. Quantifiable criteria for cell viability were active attachment to substrate and continuous beating of the ventral and anterolateral flagellum pairs. Video and surface rendering of the raw time-lapse series was generated using the Imaris software suite (Bitplane, Zürich, Switzerland).
Harvested and washed cells from 60-ml cultures were disrupted by sonication, and the homogenate was centrifuged for 35 min at 100,000 × g and 4°C. The precipitated membrane fraction was resuspended in 100 μl of PBS, and from this 5-μl portions were incubated for 30 min on ice with the following solutions: (i) 100 μl of PBS containing 1% Triton X-100, (ii) 100 μl of the same buffer containing 23 mM EGTA, (iii) 500 μl of 0.2 M Na2CO3 in 20 mM HEPES-KOH (pH 7.4), (iv) 500 μl of 6 M urea in 20 mM Tris-HCl (pH 7.4), and (v) 500 μl of 1 M NaCl in 20 mM Tris-HCl (pH 7.4). All samples contained 1× protease inhibitor cocktail (consisting of a 1 mM phenylmethylsulfonyl fluoride, 1 mM EDTA, 1 mM benzamidine, and 10 μg of pepstatin). After centrifugation at 100,000 × g for 30 min at 4°C, proteins in the pellet and in the supernatant (after trichloroacetic acid precipitation) were analyzed by SDS-PAGE and Western blotting (45).
An expression system in E. coli consisting of two plasmid constructs, one encoding α19-giardin and the other encoding the giardial N-terminal myristoyltransferase (gNMT), was used to examine N-terminal myristoylation of α19-giardin. The open reading frame of the α19-giardin gene was amplified via PCR using the following primers: GTT TGT TAG CAT ATG GGT TGT GCC GCA TCA ACT CCC (sense) and GTT CAC AAG CTT GTC GCC GCG GGG AGT CGA GGT TC (antisense); the amplificate, after digestion with NdeI-HindIII, was cloned into the pET23a(+) vector encoding a C-terminal His6 tag (Novagen). The G2A mutant of α19-giardin was produced by site-directed mutagenesis. For this purpose, the construct pET23a(+)α19wt was amplified using the following primers encoding the desired amino acid exchange (underlined): GTT TGT TAG CAT ATG GCC TGT GCC GCA TCA ACT CCC (sense) and GGG AGT TGA TGC GGC ACA GGC CAT ATG CAT ACA AAC (antisense). Primers used for the amplification of the gene encoding gNMT were designed based on sequence information obtained from the Giardia genome project, sequence CH991783, locus GL50803_5772: GAA TAA AAA AGC CAG GCT AGC ATG CCT GAT CAC (sense) and GTG TTT TCT ATT CTG CAG TCA TAT CAA CAC GAC (antisense). The amplification product, after digestion with NheI and Xho, was cloned into pET24d(+).
E. coli BL21(DE3)pLysS cells were cotransformed with pET23a(α19wt) and pET23a(α19G2A) and pET24d(gnmt), either singly or in combination. Cultures (2 ml) were grown to an optical density at 600 nm of 0.5 in LB medium containing the appropriate antibiotics, and expression of recombinant proteins was induced by the addition of 1 mM isopropyl-1-thio-β-d-galactopyranoside. Immediately afterwards, the cultures were subdivided into two 1-ml aliquots; one aliquot contained 100 μCi of [9,10(n)-3H]myristate (GE Healthcare) in 6.6 μl of ethanol, and the other contained 6.6 μl of ethanol alone (23). The cells were then incubated at 37°C for a 3-h period, harvested (10,000 × g for 5 min), and washed three times each with 1 ml of PBS. Bacteria were recovered by centrifugation, resuspended in 0.1 volume of cracking buffer (4% SDS-0.08 M Tris, pH 6.8), and lysed by heating for 5 min. Cellular debris was removed by spinning the lysate at 10,000 × g for 1 min. The amount of protein in the supernatant was determined using a bicinchoninic kit (Pierce). Samples containing 100 μg of total protein were boiled after the addition of SDS sample buffer and separated by electrophoresis through a 12% SDS-polyacrylamide gel. The gels were treated with 2,5-diphenyloxazole (Fluka), dried in vacuo, and exposed to Konica-Minolta X-ray film at −80°C for 24 to 48 h.
The detection of palmitoylation of α19-giardin was performed using a yeast system containing the palmitoyl transferase gene PFA3 under the control of a galactose promoter as described by Hou et al. (18). For this purpose, the yeast strain CUY2171 (MATaleu2 trp1 ura3-52 prb1-122 pep4-3 PFA3::TAP-kanMX PFA3::GAL TRP) was transformed with the glucose-controlled plasmid pRS406-Nop1pr-GFP-α19-giardin or pRS406-Nop1pr-GFP-α19-giardin (C3A) and grown in SDC-Ura medium. Confocal fluorescence microscopy was performed using a Leica DM5500 fluorescence microscope (Leica AG, Solms, Germany). The biotinylation assay was conducted according to Hou et al. (17). Briefly, yeast cells were broken in a buffer containing 1% Triton X-100 and free thiol groups were quenched by the addition of 25 mM N-ethylmaleimide. The palmitoyl thioesters were then cleaved by hydroxylamine, the liberated thiol groups were coupled to biotin-BMCC [1-biotinamido-4-(4′-[maleimidoethyl-cyclohexane]-carboxamido)butane)], and the biotinylated protein was captured by a neutravidin pull-down. After desorption by boiling, proteins were analyzed by SDS-PAGE and Western blotting. Subcellular fractionation was done by differential centrifugation of spheroblasts as described by Subramanian et al. (42). Pellets obtained at 13,000 × g and the supernatant fraction obtained at 100,000 × g were analyzed by SDS-PAGE and Western blotting. As controls, antibodies directed against the transaldolase Tal1 as a marker for a soluble protein (40) and the ATPase subunit Vma6 as a marker for the vacuolar fraction (kindly provided by C. Ungermann) were employed.
As a first approach to provide evidence for the existence of a transcription product of the α19-giardin gene in trophozoites of G. lamblia, we performed RT-PCR on first-strand cDNA using primers constructed according to sequence information from the Giardia genome project (27). This way, we got a clear signal with the expected size of ~1.3 kb in an agarose gel (Fig. (Fig.1A).1A). Control sequencing of the amplification product revealed the predicted N-terminal sequence, MGCAAS, which serves as a signature sequence for dual acylation (2). As derived from its complete nucleotide sequence, α19-giardin consists of 438 amino acid residues with a calculated molecular mass of 47.7 kDa and a weakly acidic isoelectric point of 5.5. We cloned a gene fragment encoding an N-terminally truncated α19-giardin that included the complete four annexin repeats but lacked the potential acylation motif (see below). We heterologously overexpressed this gene in E. coli strain BL21 and used the affinity chromatographically purified recombinant protein as an antigen for raising polyclonal antibodies in rabbits. In a Western blot assay employing a trophozoite extract, the antiserum reacted with a protein of about 47 kDa mainly from the pellet fraction (Fig. (Fig.1B),1B), suggesting that native α19-giardin is probably membrane associated in the trophozoites.
As reported previously, the most closely related α14-giardin localizes to all flagella and, to a smaller extent, is found in the median body of the trophozoites (43, 46, 47). To find out whether α19-giardin is also immunologically detectable in these organelles, we mechanically detached all the flagella from the cell body and separated them from cellular debris by density gradient centrifugation in a Percoll gradient (46). In immunoblots of this flagellar protein fraction, strong signals emerged when antibodies were directed against α19-giardin and α14-giardin; tubulin was used as a control (Fig. (Fig.1C).1C). By contrast, polyclonal antibodies toward α1-giardin, which is known to be located in the plasma membrane of the cell body (36), did not react, demonstrating both the purity of the flagellum preparation and the specificity of the antibodies used. To find out in which flagellum type α19-giardin resides, we performed immunocytochemical colocalization studies of α19-giardin together with tubulin on fixed and permeabilized cells. For these experiments, we employed the anti-α19-giardin antibodies as well as antibodies against tubulin and secondary antibodies conjugated with different fluorophores (Cy3 for α19-giardin and Cy2 for tubulin). As shown in Fig. Fig.2A,2A, a red fluorescence signal indicated that the α19-giardin protein emerged solely from the ventral flagella, whereas the antitubulin antibody recognized all the flagella (46). Yellow signals indicating colocalization in merged images conspicuously occurred solely at the distal part of the ventral flagella, suggesting that α19-giardin is located only in that part of the axonema protruding outside the cell body. Arrest of the cell cycle in G1 by subsequent treatment with nocodazole and aphidicolin according to the protocol of Poxleitner et al. (37) revealed that α19-giardin permanently remains in the ventral flagella during cell division (Fig. (Fig.2B2B).
Of all the sequences of the α-giardins in the Giardia genome data base, α19-giardin is the only one containing an N-terminal signature sequence for dual lipid conjugation with myristate and palmitate (30). Correspondingly, α19-giardin appeared in the membrane fraction after centrifugation of a trophozoite homogenate (Fig. (Fig.1B).1B). To analyze the strength and mode of the membrane association, we subjected the pellet fraction of a crude cell extract to extraction experiments with sodium carbonate, sodium chloride, urea, and Triton X-100. This method serves as a standard protocol to assess the membrane association of a protein via a lipid anchor (45). Analyses of the samples by SDS-PAGE as shown in Fig. Fig.3A3A revealed that only the nonionic detergent Triton X-100 was able to release α19-giardin into the soluble fraction in significant amounts. This observation suggests that α19-giardin behaves like an integral membrane protein because neither unspecific adsorption nor precipitation of unfolded protein is apparently responsible for its membrane association. Moreover, in a partition experiment with Triton X-114, we recovered the native protein in the detergent phase although a portion with lower molecular mass emerged in the aqueous phase as well (Fig. (Fig.3A).3A). This may be due to peptidolytic splitting off of the N terminus containing the attached fatty acid moiety through a temperature shift to 30°C that provokes a phase separation during this experiment. A typical behavior of annexins from higher eukaryotes is their capability to associate with negatively charged phospholipids in the presence of Ca2+ ions. In order to assess whether α19-giardin likewise possesses this characteristic property of annexins, we performed phospholipid-binding assays using a bovine brain extract according to the method of Boustead et al. (3). Because the full-length recombinant α19-giardin was insoluble in detergent-free aqueous solutions, we employed the N-terminally truncated recombinant protein mentioned above in these experiments. Western blot analyses showed that the protein associated with the phospholipids in the presence of Ca2+ ions (Fig. (Fig.3B).3B). In the absence of free Ca2+, the protein remained in the supernatant fraction, in both the presence and absence of phospholipids. After addition of the Ca2+-chelator EGTA, the bound protein detached from the phospholipids, which demonstrates that this mode of phospholipid association, like that of annexins, is reversible.
Both the N-terminal motif and the biochemical properties of α19-giardin argue for the existence of a lipid anchor on α19-giardin. To test directly the acylation of α19-giardin with a myristoyl moiety, we employed a dual expression system in which we cotransformed an appropriate expression strain of E. coli with different expression vectors, with one encoding gNMT and the other one encoding wild-type α19-giardin and the G2A mutant. Protein expression was induced by isopropyl-β-d-thiogalactopyranoside in the presence of [9,10(n)-3H]myristate (4), and the products were analyzed by both immunoblotting and fluorography. As demonstrated by Western blotting, the expression rates of the α19-giardin genes used in all samples were stable and comparable in amounts. However, in the fluorographs the incorporation of the radioisotope was visible only in those transformants containing the wild-type α19-giardin gene (Fig. (Fig.4).4). Neither omission of the gNMT nor the use of the G2A mutant as a reaction partner of the gNMT resulted in a signal. This clearly proves the N terminus of α19-giardin to be a signature sequence for an acylation with a myristoyl moiety. To test whether the posttranslational modification of the glycine in position 2 with a myristoyl group is necessary for the characteristic localization of α19-giardin, we mutated this position to alanine (G2A) in the construct α19-giardin-HA. Transgenic parasites expressing a single integrated copy of the α19-giardin-HA (G2A) gene by the pPacV-integ expression vector did not show a detectable signal in immunofluorescence microscopy (data not shown), whereas the C-terminally HA-tagged wild-type α19-giardin showed the expected exclusive localization in the ventral pair of flagella (Fig. (Fig.5A).5A). Additional localization to other flagellum types could be observed only in transgenic parasites with strong overexpression due to multiple copies if the circular plasmid was maintained episomally (data not shown). In the same way, we overexpressed the mutant. These cells showed a completely cytoplasmic localization of α19-giardin-HA mutated in this single position, consistent with abrogation of the specific targeting of this variant (Fig. (Fig.5B).5B). To test whether localization of a tagged copy of α19-giardin impaired the function of the ventral flagella, we expressed an α19-giardin::GFP chimera in transgenic trophozoites. The cells showed a flagellar localization identical to that of the HA-tagged variant (Fig. (Fig.5C),5C), and beating of the ventral flagellum pair was unimpaired.
Palmitoylation of proteins is a reversible process, which may be why we were unable to detect this modification directly in Giardia lysates. To test whether the α19-giardin molecule is a substrate of a palmitoyl transferase, we transfected the yeast strain CUY2171 with a GFP construct encoding the wild-type and a C3A mutant of α19-giardin. This yeast strain overexpresses the palmitoyl transferase Pfa3, which is known to add palmitoyl residues to its target proteins with a broad specificity and independently of previous myristoylation (18). As shown in the fluorescence micrographs of Fig. Fig.6A,6A, the wild-type α19-giardin associated with the membrane of the main vacuole in the presence of Pfa3, whereas in the absence of Pfa3 it distributed over the entire cytosol. This result is supported by cell fractionation: α19-giardin in the presence of Pfa3 occurred in the vacuolar fraction, whereas in the absence of Pfa3 it remained in the supernatant after centrifugation (Fig. (Fig.6B).6B). The C3A mutant also delocalized throughout the whole cytosol in both the presence and absence of Pfa3. This finding is consistent with the results of a biotin switch experiment. Here, all free thiol groups of the yeast lysate were quenched with N-ethylmaleimide, and the remaining thioesters were cleaved by hydroxylamine. The liberated thiol groups were coupled to biotin, and the biotinylated protein was captured by a neutravidin pull-down assay (17). As shown in Fig. Fig.6C,6C, wild-type α19-giardin was palmitoylated in the presence of Pfa3. These findings clearly verify that α19-giardin is a substrate for this rather unspecific palmitoyl transferase activity.
In this study, we report on a novel annexin-homologous protein, α19-giardin, one of 21 α-giardins of the human pathogenic parasite G. lamblia. Although the overall sequence identities among the α-giardins (varying from 15.3% to 19.6%) are rather low, the presence of four homologous annexin-like domains supports their common evolutionary ancestry (29). α19-Giardin differs from most other giardins in carrying an N-terminal sequence extension that comprises 34 residues prior to the start of the first annexin repeat (see protein sequence in the NCBI GenBank, accession number AY781315). This is the largest sequence extension within α-giardins, such as α1-giardin and α2-giardin (2 residues) or α14-giardin (23 residues), and it is larger than that of human annexin A5 (15 residues). Furthermore, the α19-giardin molecule contains a C-terminal stretch that extends the fourth annexin repeat for about 70 residues; the importance of this part of the molecule is still unclear. However, both an integrin binding motif (RGD motif) and a glycosylation site (NDT) in this part of the molecule may argue for an extracellular function; whether and how the protein may arrive at the exterior of the cell has not yet been discovered, however. Recently published tertiary structure determinations of α11-giardin and α14-giardin by X-ray crystallography support the notion that the α-giardins fold to all-helical structures typical for annexins of other eukaryotes (9, 34, 35).
The α19-giardin molecule starts with the N terminus MGCXXS that in many proteins serves as a signature sequence for dual lipid conjugation with a myristoyl group at the N-terminal glycine (after cleavage of the methionine residue) and a palmitoyl moiety at the following cysteine residue (38). According to our database searches, α19-giardin is the only annexin, besides human annexin A13, that contains a myristoyl moiety and, as far as we know, the only annexin homolog at all that contains a dual lipid conjugation motif. Other examples of proteins with dual lipid conjugations near their N termini are the Src protein kinases and the guanine nucleotide-binding protein-α (Gα) subunits (2). The attachment of a myristoyl residue to the N-terminal glycine occurs cotranslationally and is generally catalyzed by an NMT. By database searches, we found a single gene in the Giardia genome encoding a putative NMT. We amplified and cloned this gene and showed by heterologous coexpression with the α19-giardin gene in the presence of [9,10(n)3H]myristate that this giardin indeed can be acylated by gNMT. The same result was obtained when the gene encoding human NMT1 was employed (data not shown), supporting the notion that the first six amino acid residues are most important because their fitting into the enzyme's substrate binding pocket is a prerequisite for the acylation reaction (26).
Palmitoylation of proteins is a posttranslational, enzymatic process catalyzed by the so-called DHHC-cysteine rich-domain family of proteins, which apparently are involved in trafficking events within a cell (2). The reversibility of this reaction could be the reason why we were unable to detect this modification in Giardia extracts. However, in an experimental approach in which we coexpressed the α19-giardin gene together with the Pfa3 gene from yeast, a gene encoding a palmitoyl transferase with relatively broad substrate specificity (18), α19-giardin changed its localization and associated with vacuolar membranes, which suggests palmitoylation under particular physiological conditions. According to our database searches, the genome of Giardia contains nine genes encoding putative palmitoyl acyl transferases (PATs). One of them, designated gPAT, catalyzes the palmitoylation of the variant surface protein VSPH7 in the trophozoites (33, 44). The substrates of the other giardial PATs, including those named gDHHC2 and gDHHC3, are still unknown. One of their target proteins could be α19-giardin.
The expression of a vast number of α-giardins in trophozoites of G. lamblia and their putative subcellular localizations have already been reported by Weiland et al. (47). However, statements concerning their specific functions are rather speculative. As annexin homologs, these proteins may establish a connection between the plasma membrane and the cytoskeleton. In the case of α19-giardin, its exclusive location in the plasma membrane covering the ventral flagella argues for an involvement of this protein in the specific tasks of these flagellum types. Particular functions of the ventral flagella are supposed to be the anchoring of the parasite to the intestinal epithelium of the host and providing it with nutrients from the intestinal lumen (7). For instance, during attachment of the parasite to the intestinal wall, only the ventral pair, which emerges from the posterior end of the disk, continuously beats in an approximately sinusoidal swinging movement in the plane of the disk. This action may generate negative pressure below the contact zone of the ventral disk at the intestinal epithelium and, like a suction cup, fix the trophozoites to the intestinal wall (16). The fluid transport generated by the ventral flagella may then sweep matter under the body of the parasite, providing it with nutrients (10). These particular functions of the ventral flagella may be directly concerned with the local occurrence of the α19-giardin.
Generally, the components of flagella are synthesized in the cytoplasm and transported as large proteinaceous particles to the distal areas of the flagella. Assembly and maintenance of these continuous movements from the cytosol to the flagella tips along the microtubules outside the cell body are realized by intraflagellar transport (IFT) via so-called rafts that ensure the delivery of axonemal devices to the top of the flagellum (21). The components of IFT and their assembling in the flagella seem to be highly conserved during evolution; their encoding genes also exist in the Giardia genome (15). Since IFT is supposed to cover the area behind the exit point and the distal part of the giardial axonemes, the specific localization of α19-giardin in this region of the flagella may indicate an involvement of this protein in dynamic processes of IFT. In humans, the myristoylated annexin A13b is known to interact with other proteins and helps to direct them to their destination points (22). In epithelial cells, this annexin is suggested to play a role in the formation of transport vesicles and their fusion with the apical membrane (22). Likewise, the flagellar calcium binding protein, FCaBP, from Trypanosoma cruzi, the trigger of Chagas disease, possesses two fatty acid modifications, with myristoyl and palmitoyl residues, in an N-terminal stretch. Interestingly, the specific targeting of this protein to the flagellum is a calcium-dependent process (12). We could not yet prove a direct influence of calcium on the localization of α19-giardin in the ventral flagella, but our finding that the N-terminally truncated protein reversibly binds to phospholipids in a Ca2+-dependent manner may implicate Ca2+ in the specific targeting of α19-giardin to the ventral flagella prior to its palmitoylation. This reversible binding may ensure that the α19-giardin molecule, despite the complete reorganization of the flagella during cell division, remains in the ventral flagella throughout the cell cycle, as indicated in our cell synchronization experiments (32, 37).
Taken together, our data suggest that the annexin-homologous α19-giardin from trophozoites of G. lamblia is dually acylated. These nonproteinaceous modifications may help to fulfill two functions in this intestinal parasite: (i) anchoring of cytoskeletal proteins to distinct sites at the plasma membrane within the ventral flagella and (ii) involvement of α19-giardin in intracellular transport processes of proteins to their functional sites. By this means, the α19-giardin protein may take on a role as a specific mediator between the cytoskeleton and plasma membrane, particularly in the ventral flagella. A deeper insight into the biochemical properties of α19-giardin and its dynamic within the cells should shed light on its actual intracellular function.
We thank Andreas Kellersmann, Anna Ludwig, and Arun T. J. Peter for valuable contributions to this study.
This work was supported by the research training grant 612 of the Deutsche Forschungsgemeinschaft (to H.S.) and by the Swiss National Science Foundation (to A.B.H.).
Published ahead of print on 14 August 2009.
†Supplemental material for this article may be found at http://ec.asm.org/.