Molecular Identification of a SNAP-25-Like SNARE Protein in Paramecium

Abstract

Using database searches of the completed Paramecium tetraurelia macronuclear genome with the metazoan SNAP-25 homologues, we identified a single 21-kDa Qb/c-SNARE in this ciliated protozoan, named P. tetraurelia SNAP (PtSNAP), containing the characteristic dual heptad repeat SNARE motifs of SNAP-25. The presence of only a single Qb/c class SNARE in P. tetraurelia is surprising in view of the multiple genome duplications and the high number of SNAREs found in other classes of this organism. As inferred from the subcellular localization of a green fluorescent protein (GFP) fusion construct, the protein is localized on a variety of intracellular membranes, and there is a large soluble pool of PtSNAP. Similarly, the PtSNAP that is detected with a specific antibody in fixed cells is associated with a number of intracellular membrane structures, including food vacuoles, the contractile vacuole system, and the sites of constitutive endo- and exocytosis. Surprisingly, using gene silencing, we could not assign a role to PtSNAP in the stimulated exocytosis of dense core vesicles (trichocysts), but we found an increased number of food vacuoles in PtSNAP-silenced cells. In conclusion, we identify PtSNAP as a Paramecium homologue of metazoan SNAP-25 that shows several divergent features, like resistance to cleavage by botulinum neurotoxins.

Membrane trafficking in eukaryotic cells involves budding of vesicles from a donor compartment and transport to and fusion with the acceptor compartment. The soluble N-ethylmaleimide-sensitive factor attachment protein receptors (SNAREs) are of central importance in the mediation of membrane fusions (32). The crystal structure of the synaptic SNARE complex has been resolved (70). The ternary synaptic SNARE complex consists of the SNARE motifs of synaptobrevin-2 (VAMP2) and syntaxin-1A and the two SNARE motifs from the synaptosome-associated protein of 25 kDa (SNAP-25). Structures of different SNARE complexes revealed a highly conserved four-helix structure, with the difference that the positions of the two SNARE motifs from SNAP-25 can be contributed by two different SNARE proteins (7). The highly conserved pattern of SNARE pairing has led to the so-called 3Q-plus-1R rule (21). According to this rule, fusogenic SNARE complexes always contain three SNARE motifs containing a glutamine residue in the center of the SNARE motif (Q-SNARE) and one SNARE displaying an arginine at the same position (R-SNARE). Furthermore, Qa-, Qb-, Qc-, and R-SNAREs can be recognized by specific sequence features (40).

Identification of the SNARE components of the synaptic SNARE complex and functional analysis have been greatly facilitated by the availability of specific inhibitors, e.g., by Clostridium botulinum neurotoxins (BoNTs), that specifically cleave certain neuronal SNAREs (46). BoNTs are zinc-dependent proteases which, by cleaving SNARE proteins, inhibit neurotransmitter release. The structural basis for the specificity of SNAP-25 cleavage by BoNT/A and BoNT/E has been solved, and the interacting amino acids have been mapped (13, 15).

Most SNAREs possess a carboxy-terminal transmembrane domain, whereas others, like the SNAP-25 protein and the R-SNAREs of the Ykt6 family, are attached to the membrane by fatty acid modification. Mammalian SNAP-25 is membrane attached by palmitoylation on a conserved stretch of cysteine residues situated between the two SNARE motifs (75). However, such a cysteine cluster is absent from the vertebrate proteins SNAP-29 and SNAP-47 (31, 67), as well as from all SNAP-25 homologues outside of the metazoans, and the modes of membrane attachment, if any, of those proteins remain to be determined. Homologues to mammalian SNAP-25 have been found in a variety of organisms ranging from unicellular organisms to plants, fungi, and higher eukaryotes (40).

Disassembly of the fully assembled SNARE complex is performed by the SNARE-specific chaperone NSF, an AAA-type ATPase (64), and SNAPs recruit NSF to the SNARE complex (59). The exact time point of NSF action before or after membrane fusion has been debated, and it is possible that different requirements for regulation are met in various membrane fusion events (25, 44, 63, 72, 78).

SNARE-mediated fusion is a common feature of all eukaryotic cells, and all of the above-mentioned components of the SNARE fusion machinery have also been identified in the ciliated protozoan Paramecium tetraurelia (22, 36, 37, 61). Paramecium, which must perform all of the autonomous functions of an entire organism, possesses highly diversified membrane trafficking pathways (53). P. tetraurelia is capable of a fast synchronous release of dense core vesicles, defensive organelles called “trichocysts,” that has striking similarities to dense core vesicle exocytosis of neuroendocrine cells (52, 74). Like many other ciliates, P. tetraurelia has regularly arranged cortical structures and organelles, such as ciliary bases, “alveolar sacs” (calcium stores), sites of constitutive endo- and exocytosis (“parasomal sacs”), early endosomes (“terminal cisternae”), and trichocysts, all of which are arranged in a highly regular pattern. This feature facilitates the identification of organelles and membrane interaction sites. For instance, the ∼1,000 trichocysts are predocked in a fusion-ready state at precisely predictable sites. Food vacuole uptake and processing occur in a highly ordered manner by transformation through defined stages while moving on a fixed route through the cell (“cyclosis”) (2-5). Many of the membrane interaction sites involved are endowed with different SNAREs (37). Furthermore, P. tetraurelia possesses a pair of contractile vacuole systems for osmo- and ion regulation, each consisting of a collecting system of five to seven radial canals that empty through ampullae into a central contractile vacuole (1). NSF and different SNAREs of the R- and Q-types were also found in the contractile vacuole system (37, 61).

Here, we investigated the properties and subcellular localization of a homologue of the SNARE protein, SNAP-25, in P. tetraurelia. So far, SNAP-25 homologues have been investigated only in metazoans, fungi, and plants (11, 14, 16, 30), and the present work is the first study of a SNAP-25 homologue in a unicellular organism.

MATERIALS AND METHODS

Cell culture.

Wild-type strains of P. tetraurelia were stocks of 7S and d4-2, derived from stock 51S (65). Cells were cultivated in a bacterially inoculated medium as described previously (38). For permeabilization experiments, cells were permeabilized in Dryl's buffer (2 mM sodium citrate, 1 mM NaH₂PO₄, 1 mM Na₂HPO₄, 1.5 mM CaCl₂ [pH 6.8] [19]) supplemented with 0.2% bovine serum albumin (BSA) with 0.2%, 0.5%, or 1% Triton X-100, 0.1% or 0.3% digitonin, or 0.01% saponin. To demonstrate the acidification of food vacuoles, P. tetraurelia cells were fed with pHrodo (Invitrogen, Karlsruhe, Germany) Escherichia coli bioparticles for 20 min and results were analyzed by using epifluorescence microscopy using an Axiovert 100TV microscope equipped with filter set number 9 and a plan-Neofluar ×40 oil immersion objective (numerical aperture, 1.3) and imaging with a ProgRes C10 plus camera system (Jenoptik, Jena, Germany).

Annotation and characterization of the P. tetraurelia SNAP gene.

The Paramecium genome database (http://paramecium.cgm.cnrs-gif.fr) was BLASTP searched with the amino acid sequences of the SNAP-25 homologues from other organisms obtained from NCBI (http://www.ncbi.nlm.nih.gov). The “supercontigs” of positive hits were identified by BLASTN searches, and the gene sequence was manually completed, starting with an ATG start codon and terminating with a TGA stop codon. Putative introns, which, in Paramecium, are 18 to 35 nucleotides long and flanked by conserved 5′-GT and 3′-AG sequences (57), were manually annotated using MapDraw (DNA Star, Madison, WI) software. The resulting predicted protein sequence was reciprocally analyzed by BLASTP searches of the NCBI database (6). Conserved motif searches were performed with either PROSITE (9) or BLAST-RPS software, using Pfam entries of the corresponding CDD database (12, 45). We also used PSIPRED (34) and MEMSAT 2 (33, 35), two software methods for secondary structure prediction (included with the server at http://bioinf.cs.ucl.ac.uk/psipred/ [47]).

PCR of genomic DNA and cDNAs.

Total wild-type DNA from strain 7S for PCR was prepared from log-phase cultures as reported by Godiska et al. (24). The open reading frame of the P. tetraurelia SNAP (PtSNAP) gene was amplified by reverse transcriptase (RT) PCR, using total RNA prepared according to Haynes et al. (29). RT-PCR was performed in a programmable T3 model thermocycler (Biometra, Göttingen, Germany), using a 3′ oligo(dTT) primer (5′-AACTGGAAGAATTCGCGGCCGCGGAATTTTTTTTTTTTTT-3′) and a SuperScript III RT (Invitrogen) for first-strand cDNA synthesis. The subsequent PCR was performed with Advantage 2 cDNA polymerase mixture (Clontech, Palo Alto, CA) using the PtSNAP-specific oligonucleotides (Table 1) with or without the artificial SpeI/XhoI or XbaI/XhoI restriction site added at their ends. In general, amplifications were performed with one cycle of denaturation (95°C, 1 min), 40 to 42 cycles of denaturation (95°C, 30 s) and annealing (54 to 58°C, 45 s), and an extension step (68°C, 3 min), followed by a final extension step at 68°C for 5 min. PCR products were subcloned into the pCR2.1 plasmid by using a TOPO-TA cloning kit (Invitrogen) according to the manufacturer's instructions. After clones were transformed into E. coli (TOP10F′) cells, positive clones were sequenced as described below.

TABLE 1.

Oligonucleotides used for amplification and expression of PtSNAP

SNAP type	Restriction recognition sites	Oligonucleotide
Oligonucleotides for RT-PCR
Dei-1		AACTGGAAGAATTCGCGGCCGCGGAATTTTTTTTTTTTTTT
SNAP-A	Xho	CCGCTCGAGATCCTTTAATGATTTTTTTTGTTTTTTC
SNAP-B	Spe	GGACTAGTAAGCTTATGCAATAATAACAAATATAAAACAG
SNAP-C		TTAATCACACAAAAATCTCTATTAAAA
SNAP-D		GCCGCATTAAATTAAGAACAAGAA
SNAP-E	Xho	CCGCTCGAGGTTTTTTCATTCTACTTGGAC
SNAP-F	Xba	GCTCTAGAAAGATCGATTACATTTTGGATG
SNAP-G	Spe	GGACTAGTAAGCTTATGGATCTCAAGTATTCTACTATC
SNAP-H		GTTCGTCATTGGAGTTTCATCG
SNAP-I		CACATCTTATGGAGTCAAGTCTC
SNAP-K	Spe	GGACTAGTAAGCTTATGTTCTCTTATCTGTCAATTA
SNAP-L		CAGATTACTTGTTGTTCTTCG
SNAP-M	Spe	GGACTAGTAAGCTTATGTCTTATATTTAACATCTCAATA
SNAP-O	Spe	GGACTAGTAAGCTTATGTTCAGCCTCAGCAACAAAT
SNAP-P		GCGAGCTTACTAATCAATATGTG
SNAP-Q		GTGATTCGCAATTACGGATCTCC
SNAP-R		CTCCTCTTGTTCTTATTC
Oligonucleotides for fusion PCR for heterologous expression of SNAP
SNAP-1	Xho	GCGCTCGAGTCCTTTAATGATTTTTTTTGTTTTTTC
SNAP-2		GTTGCTCAGGATTTCTTGTTGTTG
SNAP-3		GGCAGATTGTTGATTTATTTGGTAC
SNAP-4		CTTATGTAATTTCTGTTGATTTTGATC
SNAP-5		GAAGAGTGCTTTAAATTGGCCCC
SNAP-6		GTATTTGTTGTTTTTGATCATCTTTC
SNAP-7		GACCTGCCTTTGGGGTGGTTGTTG
SNAP-8		CATTTGATTTGTTTGATTAATCATCTC
SNAP-9		GATTTGTTAAGAGCTGTTGGTATTTC
SNAP-10		CTTTTGGTTTATTCTATCTAATTGGGTA
SNAP-11		CATTCTGCTTGGACATTTGGACAG
SNAP-a	Nco	GCGCCATGGATCAAGCCGCATTAAATCAAGAAC
SNAP-b		CAACAACAAGAAATCCTGAGCAAC
SNAP-c		GTACCAAATAAATCAACAATCTGC
SNAP-d		GATCAAAATCAACAGAAATTACATAAG
SNAP-e		GGGGCCAATTTAAAGCAGTCTTC
SNAP-f		GAAAGATGATCAAAAACAACAAATAC
SNAP-g		CAACAACCACCCCAAAGGCAGGTC
SNAP-h		GAGATGATTAATCAAACAAATCAAATG
SNAP-i		GAAATACCAACAGCTCTTAACAAATC
SNAP-j		TACCCAATTAGATAGAATAAACCAAAAG
SNAP-k		CTGTCCAAATGTCCAAGCAGAATG

Sequencing.

Sequencing was done by MWG Biotech (Martinsried, Germany) custom sequencing service. DNA sequences were aligned by the CLUSTAL W feature integrated in the DNAStar Lasergene software package (DNAStar, Madison, WI).

Construction and microinjection of GFP expression plasmids.

PtSNAP-specific PCR products obtained with the oligonucleotides SNAP-O and SNAP-A or SNAP-K and SNAP-A (Table 1) were cloned into the enhanced green fluorescent protein (eGFP) expression plasmid pPXV-GFP (27) in front of the eGFP gene, as described by Wassmer et al. (77), between the SpeI and XhoI restriction sites of the plasmid, using conventional cloning procedures (58). Thus, because the actual start codon was unknown in the beginning, a short version and a long version of a GFP fusion protein were constructed. For microinjection of cells, the pPXV-SNAP-GFP fusion plasmids were linearized with SfiI, which cuts in between the Tetrahymena thermophila inverted telomeric repeats, thus helping to stabilize the DNA in the macronucleus after injection (28). DNA to be injected was isopropanol precipitated and resuspended to a concentration range of 1 to 5 μg/μl in MilliQ water. For microinjection, postautogamous cells were used, which were allowed to grow for three or four generations in bacterially preinoculated medium. To avoid disturbing the transformation process, we also treated cells with 0.2% aminoethyldextran (AED) to remove trichocysts (54) and equilibrated in Dryl's buffer (19) supplemented with 0.2% BSA. DNA microinjections were made with glass microcapillaries, using an Axiovert 100TV phase-contrast microscope (Zeiss, Oberkochen, Germany). Expression of GFP fusion proteins in clonal descendants of microinjected cells was analyzed after 24 to 48 h by epifluorescence microscopy with an Axiovert 100TV microscope (Zeiss) equipped with filter set 13 or 9, a plan-Neofluar ×40 oil immersion objective (numerical aperture, 1.3) and a ProgRes C10 plus camera system from Jenoptik. Excitation light was produced by a 100-W HBO lamp. Images were processed with either Axiovision software (Zeiss) or Adobe Photoshop (Adobe Systems, San Jose, CA). Confocal images were acquired with an LSM510 Meta confocal scanning microscope (Zeiss) equipped with a plan-Neofluar ×63 oil immersion objective (numerical aperture, 1.4).

Gene silencing by feeding.

The coding sequences of the PtSNAP gene, either as a ∼300-bp fragment from genomic DNA or as a full-length cDNA sequence, were amplified by PCR using the PtSNAP-specific oligonucleotides (Table 1) and cloned into the double T7 promoter plasmid pL4440 (71) over the SpeI and XhoI restriction sites. Plasmids were introduced in the E. coli Ht115 strain, and Paramecium cells were fed with these strains as described in detail by Galvani and Sperling (23) and by Wassmer et al. (77). The Paramecium cells were analyzed after 24 to 96 h of feeding. The cells' capability for trichocyst exocytosis was routinely tested with a saturated solution of picric acid (56).

Recombinant expression of PtSNAP in E. coli.

For heterologous expression of PtSNAP, we selected a part of the coding region of PtSNAP (Q11-K175; EMBL accession number CAK57530). After the mutated Paramecium glutamine codons (TAA and TAG) were substituted for the universal glutamine codons (CAA and CAG) by PCR methods (18) (Table 1 lists oligonucleotides), this region of PtSNAP was cloned into the NcoI/XhoI restriction sites of the pRV11 expression vector (79), a derivative of the pET system from Novagen (Madison, WI), which adds an eight-amino-acid peptide to the C terminus of the selected sequence, including a His₆ tag for purification of the recombinant peptides. PtSNAP_Q11-K175 was then recombinantly expressed in E. coli BL21(DE3)-pLysS cells.

Purification of the recombinant PtSNAP and preparation of polyclonal antibodies.

The recombinant PtSNAP_Q11-K175 protein was purified by affinity chromatography on Ni²⁺-nitrilotriacetate agarose under denaturing conditions, as recommended by the manufacturer (Novagen, Madison, WI). The recombinant peptide was eluted at pH 4.5 with a buffer containing 8 M urea, 100 mM NaH₂PO₄, and 10 mM Tris-HCl (pH 4.5) supplemented with 1 M imidazole. The collected fractions were analyzed on sodium dodecyl sulfate (SDS)-polyacrylamide gels, and those containing the purified recombinant protein were pooled, dialyzed against phosphate-buffered saline (PBS; pH 7.4), and used for the immunization of a rabbit. After the rabbit received several boosts, positive sera were taken and affinity purified by two subsequent chromatography steps as described previously (38).

Cell fractionation.

For subcellular fractionation, cells were grown in axenic culture medium at 25°C and harvested at the late logarithmic phase as previously described (39). Whole-cell homogenates were prepared in 20 mM phase buffer (20 mM Tris-maleate, 20 mM NaOH, 20 mM NaCl, 250 mM sucrose [pH 7.0]) as described previously (38). Soluble and particulate fractions were separated by centrifugation at 100,000 × g for 60 min at 4°C. A protease inhibitor cocktail containing 15 μM pepstatin A, 100 mU/ml aprotinin, 100 μM leupeptin, 0.26 mM N α-(p-toluene sulfonyl)-l-arginine methyl ester (TAME), 28 μM E64, and 0.2 mM Pefabloc SC (all from Sigma-Aldrich, Schnelldorf, Germany) was used throughout the preparation. Similarly, P. tetraurelia homogenates were separated on a 10 to 30% Optiprep (Axis-Shield PoC AS, Oslo, Norway) gradient at 46,000 × g for 18 h at 4°C.

BoNT treatment of cell lysates.

BoNT/A (Sigma-Aldrich) and BoNT/E (List Biological Laboratories, Campbell, CA) were reconstituted in sterile double-distilled H₂O, supplemented with 1 mg/ml BSA to 0.1 mg/ml and activated in 200 mM Tris-HCl (pH 8.0), 500 mM NaCl, and 50 μM ZnCl₂ with 5 mM dithiothreitol for 30 min at 37°C. Approximately 30 μg of protein of crude cell lysates from P. tetraurelia or PC12 cells or 5 μg of purified recombinant PtSNAP or rabbit SNAP-25 control peptide (List Biological Laboratories) was incubated with 20 ng of the respective BoNTs for 1 h at 37°C. The protein was methanol precipitated and analyzed by SDS-polyacrylamide gel electrophoresis (PAGE) (see below). Rabbit SNAP-25 was detected on Western blots with an anti-human SNAP-25 mouse monoclonal antibody (clone SP12; Upstate Biotechnology, NY).

SDS-PAGE and immunoblotting.

Protein samples were denatured by boiling for 5 min in SDS sample buffer and subjected to electrophoresis in 15% SDS-polyacrylamide gels, using a discontinuous buffer system described previously (36). Electroblotting onto nitrocellulose membranes and immunobinding were carried out as described previously (38) by using affinity-purified antibodies against PtSNAP. Bound antibodies were detected with a peroxidase-conjugated secondary antibody (anti-rabbit immunoglobulin G [IgG]), using an ECL detection system (Amersham, München, Germany). The anti-proteindisulfide-isomerase (anti-PDI) antibody was kindly provided by E. Ladenburger (University of Konstanz).

Immunofluorescence analysis.

Immunofluorescence analyses were performed with permeabilized cells. Cells suspended in piperazine-N,N′-bis(2-ethanesulfonic acid) (PIPES)-HCl buffer (5 mM; pH 7.2) supplemented with 1 mM KCl and 1 mM CaCl₂ were fixed in 4% (wt/vol) freshly depolymerized formaldehyde in the same buffer solution. Following fixation, cells were permeabilized with 0.5% digitonin (Sigma-Aldrich) for 30 min at 20°C, washed in PBS, and then incubated twice in PBS supplemented with 50 mM glycine and finally in PBS plus 1% BSA. Samples were then exposed to affinity-purified anti-PtSNAP antibodies (1:50) or to monoclonal anti-α-tubulin antibodies (clone DM1A; Sigma-Aldrich), followed by AlexaFluor488- or AlexaFluor594-conjugated F(ab′)₂ fragments of goat anti-rabbit and goat anti-mouse IgG (Invitrogen), both diluted 1:100 in PBS plus 1% BSA. For controls, either preimmune serum was used or primary antibodies were omitted. Samples were mounted with Mowiol supplemented with N-propylgallate to reduce fading. Fluorescence was analyzed with an LSM510 Meta model confocal laser scanning microscope (Zeiss) equipped with a plan-apochromat ×63 oil immersion objective (numerical aperture, 1.4) or in a conventional epifluorescence microscope (see above). Images acquired with the LSM510 software were processed with Photoshop software (Adobe Systems).

RESULTS

Identification of PtSNAP.

The developing Paramecium genome database (http://paramecium.cgm.cnrs-gif.fr) based on the Paramecium genome project (8, 17, 66, 80) was tBLASTN searched with the amino acid sequence of SNAP-25 homologues from other organisms. The search with leech (Hirudo medicinalis) SNAP-25 (GenBank accession no. gb|AAC47499) first returned three Qc-SNAREs, PtSyx14-1 (emb|CAK58342), PtSyx14-2 (emb|CAK88055), and PtSyx15-1 (emb|CAK79412) (37) as major hits. Eventually, a single SNAP-25-like sequence could be identified on scaffold_105, and the corresponding coding region was completed using flanking sequence information of the respective supercontig (SuperContig_11387). The putative ATG start codon and the TGA stop codon were manually assigned, as well as the position of a single 25-bp conventional intron (Fig. 1A). This gene structure prediction fitted well with the automatically annotated gene model (GSPATT00028565001; emb|CAK57530) published later (8). Reciprocal BLASTP searches with the PtSNAP sequence against those of GenBank confirmed the annotation of PtSNAP as a SNAP-25-like protein, in which the closest matches were the Anopheles gambiae strain PEST AGAP001394-PA (gb|EAA01106.5), Drosophila pseudoobscura GA21816-PA (gb|EAL27731.1), the Aedes aegypti synaptosome-associated protein (EAT44027.1), and the Drosophila melanogaster SNAP-24 protein (gb|AAF73834.1). Generally, the sequence conservation between PtSNAP and homologues of other species is low (expectation values of ≥0.21). However, this holds true for many SNAREs, since the SNARE motif is structurally conserved, i.e., not necessarily with a high degree of sequence homology.

FIG. 1.

(A) Nucleotide and deduced amino acid sequences of PtSNAP. The bases are numbered referring to the position of the start ATG codon (bold). The locations of oligonucleotide primers used in this study are indicated below the underlined nucleotide sequence. The hypothetical N-terminal extended amino acid sequence is indicated in gray capital letters. The first, a Qb-SNARE motif, is marked in yellow, and the second, a Qc-SNARE motif, is marked in blue. Hyphens mark the positions of the introns, and stars mark the translation stop codons TGA. (Ba) Homology between the region containing PtSNAP on scaffold_105 (continuous red line) and the corresponding region of scaffold_121 (below). A color bar indicating the degree of sequence similarity and a nucleotide ruler are shown above. (Bb) Schematic illustration of the position of PtSNAP (blue) on scaffold_105 and the deletion in the respective region from the sister scaffold_121 below. Numbers above and below refer to the base pair number within the respective scaffold.

Owing to a recent whole-genome duplication, Paramecium genes often occur as pairs of closely related orthologues (8), and we previously described a great diversification of the Qa-SNARE and R-SNARE families (37, 61). However, we were not able to identify any other SNAP-25-like protein in the Paramecium genome. We searched the corresponding sister scaffold_121 for the presence of a PtSNAP orthologue, but in the respective region, a deletion seems to have occurred (Fig. 1B). Sequence searches of the genome for the related ciliate T. thermophila (20) revealed a gene (TTHERM_00526630) similar to that which encodes PtSNAP. So far, we were not able to identify SNAP-25 homologues in other ciliates in the Ciliate Ortholog Database (http://oxytricha.princeton.edu/COD/).

An algorithm specifically trained on SNARE motifs has been developed (40), and when the respective SNARE database was searched with PtSNAP, matches with expectation values of e⁻¹¹ for the consensus SNAP-25 Qb/c motifs were obtained (Fig. 2A). Furthermore, when reverse PSI-BLAST (rpsBLAST) was performed with PtSNAP, high similarity was found with a number of motifs from SNAP-25 homologues from different species (Fig. 2B). Importantly, conservation of the characteristic SNARE motif heptad repeats was observed for PtSNAP (Fig. 2B). In a phylogenetic tree constructed from the orthologues, PtSNAP consistently grouped within this group (Fig. 2C), and different methods of tree construction gave identical branching patterns. A hydrophilicity plot for PtSNAP shows no clear indication of membrane attachment sites (Fig. 2D).

FIG. 2.

(A) SNARE motif score for PtSNAP with the SNARE motif trained algorithm (SNARE-DB [40]). Shown are the scores for the SNAP.b and SNAP.c motifs and the homology to the consensus motifs. Conserved residues are shaded in black; similar residues are in gray. The position of the SNARE motif heptad repeats is indicated above the sequence. (B) Alignment of the Qb- and Qc-SNARE motifs of SNAP-25 with the Qb/c-SNARE motifs of other SNAP-25 homologues: Tetrahymena thermophila TTHERM_00526630 (Tt00526630; GenBank accession no. gi|89309844); Plasmodium falciparum SNAP-23 (PfSNAP23; gi|23615361); Dictyostelium discoideum GRAM-domain-containing protein (DDB0237970; gi|66827589); Homo sapiens synaptosome-associated protein 23 (HsSNAP23; gi|1374813), SNAP-25 (HsSNAP25; gi|14714976), and SNAP-29 (HsSNAP29; gi|6685982); Hirudo medicinalis SNAP-25 homologue (HmSNAP25; gi|1923252); Caenorhabditis elegans resistance to inhibitors of cholinesterase (RIC-4) family member (CeY22F5A.3; gi|32567202); protein K02D10.5 with two t-SNARE domains (CeK02D10.5; gi|17554000); Drosophila melanogaster synaptosome-associated protein 24 (DmSNAP24; gi|8163739), SNAP-25 (DmSNAP25; gi|548941); Schizosaccharomyces pombe SNAP-25 homologue (SpSNAP25; gi|3650385); Saccharomyces cerevisiae t-SNARE component Sec9 (ScSec9p; gi|730733), SNAP-25 homologue Spo20p (ScSpo20p; gi|6323659); Arabidopsis thaliana synaptosome-associated protein SNAP25-like SNAP-29 (AtSNAP29; gi|15241436), SNAP-30 (AtSNAP30; gi|15222976), and SNAP-33 (AtSNAP33; gi|15240163). The heptad amino acid repeats of the SNARE motif are shaded black, and the conserved residues are gray. Amino acid positions of the corresponding proteins are indicated on both sides. Presumptive cleavage sites for BoNT/A and BoNT/E are indicated below. (C) Neighbor-joining tree (with 1,000 bootstrap replicates) of phylogenetic relationships between SNAP-25 homologues. Species names and protein identifiers are the same as those shown in panel A. Bootstrap support values for the nodes are shown, and evolutionary distances are indicated by the scale bar below. (D) Kyte-Doolittle hydrophilicity plot of PtSNAP. Amino acid positions are indicated by the ruler above.

The neuronal SNAP-25 and SNAP-23 homologues are normally membrane attached by means of palmitoylation on a stretch of four conserved cysteine residues (41, 75). However, such a palmitoylation site is absent from the other mammalian SNAPs, SNAP-29 and SNAP-47. Likewise, we found no palmitoylation signal in PtSNAP. In fact, there is not a single cysteine residue in the amino acid sequence of PtSNAP on which fatty acid modification could occur.

Experimental verification of PtSNAP by PCR and RT-PCR methods.

To verify the existence of the in silico-identified PtSNAP gene and its in vivo expression, the genomic and cDNA sequences of PtSNAP were amplified (Fig. 3A) with specific PCR primers (Table 1 and Fig. 1, SNAP-B plus SNAP-A), subcloned, and fully sequenced. Thus, the expression of the gene, as well as the predicted intron position, was verified. Since initially there were several possibilities for the position of the ATG start codon, we also tried to obtain PCR products from cDNA with primers covering an ATG further upstream (SNAP-G plus SNAP-A) (Fig. 3B). The amplification products were checked for the presence or absence of the intron by sequencing or digestion with the restriction enzyme NsiI which cuts within the intron sequence. Surprisingly, amplifications from cDNA could be obtained with primers lying as far as 184 bp upstream of the predicted translation start point (SNAP-H plus SNAP-A) (Fig. 3B). No RT-PCR products were obtained with primers lying more than 184 bp upstream from the assumed starting ATG codon (SNAP-K/L/M/O/P plus SNAP-A) (Fig. 3C). Thus, there were only two possible localizations of the ATG start codon: at bp position 1 or at bp position −116 (Fig. 1), resulting in a 20.8-kDa or a 25.3-kDa protein, respectively. To address this question, an antibody was raised against amino acids Q11 to K175 of PtSNAP.

FIG. 3.

Amplification of the PtSNAP gene from genomic DNA and cDNA. The downstream primer is always SNAP-A. For the position of the SNAP primers used, refer to Fig. 1A. (A) Amplification of PtSNAP from genomic DNA (gDNA) and cDNA, with upstream oligonucleotide primers SNAP-B or SNAP-C, amplifies products of the expected size. (B) Amplification of PtSNAP from gDNA and cDNA, with upstream oligonucleotide primer SNAP-G or SNAP-H, also amplifies products of the expected size. (C) Upstream primers SNAP-K, SNAP-L, and SNAP-M are able to amplify products from gDNA but not from cDNA. The DNA size marker used throughout all experiments is a 1-kb ladder, and band sizes (in bp) are indicated to the left.

Detection of PtSNAP in Western blots.

PtSNAP_Q11-K175 was recombinantly expressed in E. coli cells. This required substituting 19 TAA and TAG glutamine codons of Paramecium for the CAA and CAG codons of the universal genetic code by PCR methods (18). The recombinantly expressed PtSNAP_Q11-K175 containing a C-terminal hexahistidine tag was purified by affinity chromatography on Ni²⁺-nitrilotriacetate agarose under denaturing conditions and was used for immunization of a rabbit. Polyclonal antibodies were affinity-purified from the final serum.

When used in Western blots against P. tetraurelia cell lysates, the anti-PtSNAP antibody recognized two major bands with apparent molecular masses of 20 and 21 kDa (Fig. 4A), confirming the predicted ATG start position at the second possible start codon. An additional immunoreactive band of about 46 kDa was present only when the lysates had been boiled for 5 min at 95°C and probably represents aggregates of PtSNAP (Fig. 4A), as such irreversible aggregation of membrane proteins in SDS at ≥50°C has been described before (60). When P. tetraurelia cell lysates were fractionated into soluble and insoluble fractions, the 20-kDa band preferentially stayed in the 16,000 × g supernatant, whereas the 21-kDa band went with the pellet fraction. PtSNAP could be extracted from the pellet with 1% Triton X-100, 2 M NaCl, and 4 M urea or 100 mM NaCO₃ but not by treatment with 1 M hydroxylamine (Fig. 4B), a deacylating reagent that attacks thioester bonds of palmitoylated proteins (48, 51). These data suggest that the higher molecular weight form of PtSNAP is not palmitoylated and probably not myristoylated but is bound to membranes by means of protein-protein interactions. However, we cannot exclude the possibility that the smaller molecular weight form represents a degradation product of full-length PtSNAP.

FIG. 4.

Western blot detection of PtSNAP in P. tetraurelia cell lysates. (A) An affinity-purified anti-PtSNAP antibody recognizes two bands of 20 and 21 kDa (white and gray arrowheads) each. An additional PtSNAP-cross-reactive band (black arrowhead) of ∼46 kDa is induced by boiling (+) the samples and is not present when boiling is omitted (−; right lane). Asterisks indicate probable degradation products of PtSNAP. (B) Distribution of the 20- and 21-kDa PtSNAP-immunoreactive bands is indicated by white and gray arrowheads in cell fractionations (L, lysate; S1, supernatant; P1, pellet) and in samples treated (S2, supernatant; P2, pellet) with 1% Triton X-100, 2 M NaCl, 4 M urea, 100 mM NaCO₃ and 1 M hydroxylamine. (C) Differential distribution of the 20- and 21-kDa PtSNAP immunoreactive bands in a 10 to 30% Optiprep gradient after equilibrium centrifugation. Dense membranes segregate to the top of the gradient (left); less dense membranes and soluble material accumulate at the bottom (right). Arrowheads indicate distribution as described in the legend to panel A. (D) Treatments of P. tetraurelia cell lysates (L, left) or control reactions of PC12 cell lysates (L, right) with BoNT/A, mutated inactive BoNT/A* (E224Q), and BoNT/E are shown. The mutated BoNT/A* gave rise to some higher-molecular-weight bands that are immunoreactive with anti-PtSNAP antibody. (E) Coomassie blue-stained gels of recombinant PtSNAP (rPtSNAP, top) and a recombinant mammalian SNAP-25 test substrate (rSNAP-25, bottom) treated with BoNT/A, mutated inactive BoNT/A* (E224Q), and BoNT/E.

When P. tetraurelia cell lysates were separated on 10 to 30% Optiprep gradients (55), the 21-kDa band segregated with membrane fractions to the top of the gradient, whereas the 20-kDa PtSNAP immunoreactive band segregated to the bottom of the gradient, where soluble material accumulates (Fig. 4C). The boiling-induced 46-kDa PtSNAP immunoreactive band was situated in the middle of the gradient (Fig. 4C). We conclude that the two forms of PtSNAP have distinct distributions in the cell and possibly also function in different complexes. However, the type of modification (or degradation) of PtSNAP remains unknown, as with many PtSNAP-25 homologues from other organisms.

We also tested PtSNAP for susceptibility to cleavage by BoNTs. Whereas the cleavage site for BoNT/E (15) is conserved in PtSNAP, the site for BoNT/A (13) is not. When we tested with cell extracts (Fig. 4D) or recombinantly expressed PtSNAP (Fig. 4E), we could not find any cleavage of PtSNAP, either by BoNT/A or by BoNT/E. Activity of the respective botulinum toxins was demonstrated by the cleavage of endogenous SNAP-25 of PC-12 cells, detected with an anti-human SNAP-25 antibody, or by the cleavage of recombinant mammalian SNAP-25. Using a negative control for BoNT/A cleavage, we also tested mutated BoNT/A* (E224Q), which is unable to cleave SNAP-25 (Fig. 4D and E). The mutated BoNT/A* was also not active with PtSNAP but gave rise to some higher-molecular-weight bands that are immunoreactive with anti-PtSNAP, as if it were irreversibly binding to the protein (Fig. 4D).

PtSNAP is distributed ubiquitously over the cell.

Since initially there were two possibilities for the localization of the ATG start codon of the PtSNAP gene, we cloned two versions with a C-terminal GFP tag, one starting at ATG at bp position 1 and the other one starting at ATG bp position −116. When they were microinjected into P. tetraurelia macronuclei, both versions resulted in identical localization patterns, and there was no effect on cell viability. Both constructs gave a high cytosolic GFP fluorescence, with exclusion of the macronucleus and the food vacuole lumen (Fig. 5A and B). Above the strong cytosolic signal, staining of food vacuole membranes and smaller vesicles and along the radial canals of the contractile vacuole system was observed (Fig. 5A and B, enlargement). Attempts to reduce the strong cytosolic GFP fluorescence by permeabilizing the cells with Triton X-100, digitonin, or saponin resulted in a complete loss of GFP fluorescence. Thus, the majority of PtSNAP appears to be (detergent) soluble.

FIG. 5.

GFP fluorescence in live cells microinjected with a long version (PtSNAP-lv-GFP) (A) and a short version (PtSNAP-sv-GFP) (B), with enlarged details of stained vacuole membranes (middle) and corresponding bright field image (far right). Note that the stained food vacuole (fv) has moved during the objective lens change in the enlargement shown at the right compared to that shown at the left. mac, unstained macronucleus. (B, middle panel) A vacuole (vac) is located on top of the dark appearing macronucleus. The radial canals (rc) and ampullae (amp) of the contractile vacuole system are also weakly stained. (C to F) Confocal image slices (thickness, 1 μm) of fixed PtSNAP-sv-GFP-expressing cells. (C) Median slice showing staining of the membrane of food vacuoles, in the vicinity of trichocysts (tr; the dark, carrot-shaped cortical objects), on cilia (ci) and inside the macronucleus. (D) Median slice showing staining of radial canals and the central contractile vacuole of the contractile vacuole system, between trichocysts and inside the macronucleus. (E) Superficial slice showing staining of dot-like structures and the whole cell surface. cs, cytostome. (F) Enlarged image of a superficial slice showing staining of the whole cell surface and on the regularly arranged parasomal sacs (ps; encircled, between dark trichocysts) but not on trichocyst tips (trt) (indicated by arrows), whose positions can be extrapolated from their regular pattern. Scale bars = 10 μm.

To visualize internal membrane structures, we fixed PtSNAP-GFP expressing cells and analyzed them by confocal microscopy. This reduced the cytosolic background fluorescence, and the staining of internal membranes became visible. By using this method, we were able to visualize the food vacuole membranes, the cell surface membranes, the radial canals, and the central vacuole of the contractile vacuole system (Fig. 5C and D). Unexpectedly, there was also signal from cilia and from within the macronucleus. The presence of this signal contrasts with that observed from living cells, where the macronucleus was devoid of GFP fluorescence (Fig. 5A, B), while staining of cilia in living cells could not be resolved due to their movement, which was faster than the camera frame-grabbing rate. In both cases, we suspect a redistribution of soluble PtSNAP upon fixation.

We also found PtSNAP-GFP staining between docked trichocysts but not on trichocyst tips (Fig. 5E and F). Enhanced staining at a position diagonal and posterior to trichocysts possibly represents parasomal sacs or other vesicles of the endosomal system (Fig. 5F).

To consolidate the data obtained from GFP overexpression, we used the affinity-purified anti-PtSNAP antibody for localization of PtSNAP by immunostaining. Staining of food vacuole membranes (Fig. 6A) and along the radial canals and of the central vacuole of the contractile vacuole system (Fig. 6B) could be confirmed. Staining peripherally between trichocysts (Fig. 6A) was also found and probably represents endoplasmic reticulum (ER) subdomains. Furthermore, we also observed staining with anti-PtSNAP in the macronucleus, confirming the results obtained from fixed PtSNAP-GFP-expressing cells (Fig. 6B). Staining of the sites of constitutive endo- and exocytosis (parasomal sacs) with anti-PtSNAP is visible when we focused on the cell surface (Fig. 6C). To correctly address the punctate surface staining pattern, we also performed confocal microscopy imaging with cells double stained for PtSNAP and α-tubulin (Fig. 7). We observed PtSNAP antibody staining at the cytostome (Fig. 7B), where a great number of parasomal sacs are located (R. D. Allen, electron micrograph [http://www5.pbrc.hawaii.edu/allen/ch10/14-pca740125-18.html]), and on the cell surface in very close apposition to basal bodies (Fig. 7C and D). However, discriminating between the 20- and 21-kDa forms of PtSNAP was not possible with this method. In summary, we found PtSNAP in a regular cortical pattern, at food vacuoles, between trichocysts, and on the radial arms and central vacuole of the contractile vacuole system.

FIG. 6.

Immunostaining with an affinity-purified anti-PtSNAP antibody. Left panels show the whole cells, and right panels show an enlargement of the indicated areas. (A) Median view showing staining of food vacuole membranes (fv) alongside the cytostome (cs) and between trichocysts. (B) Median view showing staining of the radial canals (rc) and central pulsating vacuole of the contractile vacuole system, as well as staining of the macronucleus (mac). (C) Surface focus showing staining of regularly arranged parasomal sacs. Occasional doublets of parasomal sacs, indicated by arrows, may possibly represent division situations. Scale bars = 10 μm.

FIG. 7.

Confocal microscopy image slices (0.9 μm, thickness) of a P. tetraurelia cell double stained with anti-PtSNAP (green) and anti-α-tubulin (red) antibodies. (A) Overview of a slice from the cortical region. The outline of the cell is indicated by a thin white oval line, with the anterior end of the cell orientated at the top. (B, C, and D) Enlarged details from the boxed regions of panel A. The regular staining pattern probably represents parasomal sacs (green), generally one juxtaposed to duplicate basal bodies (red, arrows). Scale bar = 10 μm.

Dissection of PtSNAP function by gene silencing.

Owing to its homology to SNAP-25, the SNARE involved in stimulated exocytosis in neuronal cells, and because Paramecium is capable of stimulated exocytosis of dense core vesicles, we first concentrated on the effects of the PtSNAP posttranscriptional gene silencing on the exocytosis of trichocysts. Surprisingly, however, we could find no such effect for PtSNAP. Exocytosis stimulated with picric acid (a fixing agent used for rapid genetic screening) or with the secretagogue AED occurred to the same extent as that of the wild-type control cells (Fig. 8A, B). Also, neither the docking of trichocysts nor their ability to decondense their contents was affected in PtSNAP-silenced cells. However, when those cells were examined with a light microscope, they appeared completely filled with food vacuoles (Fig. 8D and G). There was also no effect of PtSNAP silencing on cell viability. We even observed a consistent, although not statistically significant, increase in the division rate of PtSNAP-silenced cells compared to that of controls (Fig. 8E and F). The number of food vacuoles was increased (P < 0.013) after 72 to 96 h of silencing compared to that of control cells that were fed with the same strain of bacteria, while the number of acidified food vacuoles, as determined by feeding with pH-sensitive fluorophore-labeled bacteria, was unchanged (Fig. 8G). Efficient silencing was demonstrated by the downregulation of PtSNAP levels after 72 h of silencing, as probed in Western blots with the specific anti-PtSNAP antibody (Fig. 8H).

FIG. 8.

Posttranscriptional gene silencing of PtSNAP (PtSNAP-RNAi). Stimulation of trichocyst release with picric acid in a control (A) and a typical PtSNAP-silenced cell (B), both showing complete discharge of trichocysts. (C and D) Bright field images of a typical control cell (C) and a PtSNAP-silenced cell (D) showing moderate enrichment of vacuoles in the latter. Scale bars = 10 μm. (E) Division rates of controls (black) and PtSNAP-silenced cells (gray) from one set of experiments. Asterisks indicate a reduced division rate during the first 24 h of silencing due to a lag effect after transfer from normal medium to feeding solution. Note the increased division rate of PtSNAP-silenced cells from 48 h onward. (F)The averaged percentage difference in division rate between the control and PtSNAP-silenced cells is statistically significantly increased. Bar, standard error of the mean (SEM); P value, from paired t test. (G) Increase in the total number of food vacuoles in PtSNAP-silenced cells. Shown are averages of the number of food vacuoles per cell in control (black) and PtSNAP-silenced cells (gray). No change in the number of acidified vacuoles was found (hatched columns). Bars, SEM P value, from unpaired t test. (H) Demonstration of successful PtSNAP gene silencing by Western blotting of lysates from cells with different durations of silencing detected with the anti-PtSNAP antibody. In PtSNAP-silenced cells, PtSNAP becomes highly reduced from the third day of silencing onward (bottom). No decrease is seen in the loading control detected with an anti-proteindisulfide-isomerase (anti-PtPDI) antibody (top).

These results were surprising because of the central role of mammalian SNAP-25 homologues in stimulated exocytosis and because PtSNAP is the only candidate for a SNAP-25 gene-like gene identified in Paramecium so far. Additionally, PtSNAP posttranslational gene silencing in exocytosis-deficient nd9-1 cells, where the trichocyst docking sites are not formed, did not lead to a morphological undocking of trichocysts (data not shown).

According to the localization of PtSNAP in parasomal sacs, we suspected it might have a function in the constitutive exocytosis of surface antigens. However, we could find no differences between the presence and expression patterns of surface antigens A, B, D, and H of PtSNAP-silenced cells compared to that of control cells (data not shown).

DISCUSSION

Number of SNAP-25 genes.

The SNAP-25-like proteins belonging to the class of Qb/Qc-SNAREs are the only examples known so far of dual-SNARE-motif-containing proteins (40). Here, we identify and characterize a single SNAP-25 homologue in the ciliate P. tetraurelia. Like all SNAREs of ciliates (37, 61), it shares only a low degree of overall sequence homology with mammalian homologues. However, a gene similar to the SNAP-25 gene (TTHERM_00526630) exists in the related ciliate T. thermophila, and it will be interesting to see if there are similar homologues found in other ciliates. This would be important to ascertaining an evolutionary origin of SNAP-25-like genes before the emergence of multicellular organisms. Three SNAP-25 homologues have also been identified in the plant Arabidopsis thaliana (30), a genus that branched off in the phylogenetic tree well before the fungus/animal split (10). There is, however, no evidence so far for a role of those SNAP-25 homologues in stimulated exocytosis outside the animal kingdom. So, if SNAP-25-like genes were part of the original gene repertoire of the last common eukaryotic ancestor, what was their exact role? Were they originally involved in membrane fusion or associated with other cellular processes? A more comprehensive sampling of SNAP-25-like genes from other taxa will be necessary to answer these questions.

The PtSNAP gene apparently has retained no sister isoform from the recent genome duplication (8). Instead, there is a deletion in the corresponding region of the sister scaffold_121. Similarly, there is only a single SNAP-25 gene homologue present in the genome of T. thermophila (TTHERM_00526630) (20). This finding was surprising because mammals contain at least four SNAP-25 homologues, SNAP-23, SNAP-25, SNAP-29, and SNAP-47 (40), which can be functionally diversified further by alternative splicing. Ciliates, however, possess no alternative splicing, and, therefore, all Qb/c-SNARE functions have to be performed by a single PtSNAP gene product.

Posttranslational modification.

All plant SNAP-25-like proteins lack the conserved cysteine cluster of mammalian SNAP-25 that could act as attachment points for palmitate residues. However, the A. thaliana SNAP-33 (AtSNAP-33) protein, which is also devoid of a central cysteine cluster, at least was shown to localize to the plasma membrane (30), although the mechanism of its membrane attachment is also not known. There is evidence for an N-myristoylation sequence motif (G83-L88) at an equivalent position of the cysteine cluster in PtSNAP, but this localization between the two SNARE motifs does not agree with conventional N-terminal co- or posttranslational myristoylation. On the other hand, it has been reported that myristoyl residues can be posttranslationally attached to lysine residues (68, 69), so it is possible that myristoylation on one or several of the numerous lysine residues of PtSNAP could occur. Likewise, palmitoylation of lysine residues had been found in adenylate cyclase toxin by mass spectrometry (26). At this point, we cannot exclude the possibility that this modification pathway is used in Paramecium. Because myristoylation or palmitoylation on lysine residues is through O-ester and not through thioester bonds, the treatment with 1 M hydroxylamine at a neutral pH would not necessarily have hydrolyzed these bonds. Therefore, we cannot with certainty exclude fatty acid modification of PtSNAP. Another possibility is that the smaller PtSNAP immunoreactive band simply represents a proteolytic degradation product of the full-length protein, because the relative ratios detected between those two bands showed some variability between experiments.

Insensitivity of PtSNAP to botulinum toxins.

Using biochemical methods, we find PtSNAP is not cleaved by BoNT/A or BoNT/E, even though the site of BoNT/E cleavage is conserved in the primary amino acid sequence of PtSNAP. However, because the recognition motif of BoNTs is a conformational rather than an amino acid motif (13, 15), the great evolutionary distance to mammals may entail that PtSNAP is not a substrate for those toxins. Earlier analyses in our laboratory showed that injection of BoNT/A into Paramecium cells had no effect on wild-type cells (75a), while it prevented redocking of trichocysts after chemically induced undocking with cytochalasin B in nd9-1 cells at nonpermissive temperatures, where trichocysts are attached to the cortical Ca²⁺ stores, but not at the plasma membrane (50). These effects of BoNTs on the redocking of detached trichocysts in nd9-1 cells may be explained by unspecific cleavage of other proteins.

Localization of PtSNAP.

We found that on Western blots, PtSNAP appears in two different forms and that the higher-molecular-weight form clearly behaves as a membrane-associated protein, even though any possible type of modification on PtSNAP remains so far unknown. However, we cannot tell which one of the two forms is posttranslationally modified or whether both forms are posttranslationally modified. Both PtSNAP forms sediment with different fractions on a density gradient. We also found evidence for a dynamic distribution of PtSNAP between a soluble cytosolic and a membrane-bound pool, whereas the functional significance of this is still unclear. We could localize PtSNAP on a number of internal membranes, i.e., the membranes of food vacuoles, the contractile vacuole system, and the internal ER subdomains and parasomal sacs, as well as on the plasma membrane (Fig. 9). Furthermore, there is a large cytosolic pool of PtSNAP. This suggests the involvement of PtSNAP in a number of membrane fusion processes. We could not detect any accumulation of PtSNAP on trichocyst tips, where exocytic fusion sites are preformed. However, we saw an overall labeling of the cell surface in fixed PtSNAP-GFP-expressing cells equivalent to the localization of SNAP-25 in neuronal and neuroendocrine cells. Labeling of PtSNAP-GFP in the vicinity of trichocysts probably represents peripheral ER extensions. The pronounced labeling of the sites of constitutive endo- and exocytosis, the parasomal sacs, with both the PtSNAP-GFP construct and the anti-PtSNAP antibody suggests the involvement of PtSNAP in membrane trafficking there. Because several other SNAREs were found in those compartments (37, 61; C. Schilde, unpublished results), we expect that PtSNAP is a SNARE partner in several different SNARE complexes there. A challenging finding is the occurrence of PtSNAP in the contractile vacuole system. Again, several other SNAREs (37, 61; C. Schilde, unpublished data), as well the SNARE-specific chaperone NSF (36), localize to the contractile vacuole system as if there was a high extent of membrane trafficking. At this time, we can only speculate about the function of SNAREs in the osmoregulatory system.

FIG. 9.

Paramecium trafficking network (based on data from R. D. Allen and A. K. Fok [3]) superimposed with PtSNAP distribution (green). Dotted lines mark the path of organelles, whereas continuous arrows mark vesicle delivery pathways. Question marks indicate putative trafficking pathways for which PtSNAP involvement has not been demonstrated so far. Abbreviations: as, acidosome; ci, cilium; cp, cytoproct; cph, cytopharynx; cs, cytostome; cvc, contractile vacuole complex; ds, decorated spongiome of the cvc; dv, discoidal vesicle; ee, early endosome (terminal cisterna); er, endoplasmic reticulum; fv, food vacuole; ga, Golgi apparatus; gh, ghost; pm, plasma membrane; ps, parasomal sac (coated pit); rv, recycling vesicles; ss, smooth spongiome of the cvc; tr, trichocyst; trp, trichocyst precursor.

The observation of macronuclear PtSNAP after fixation of cells but not before needs further explanation. Most likely there is a redistribution of soluble PtSNAP during fixation. Native PtSNAP (molecular mass, 20.8 kDa), as well as the GFP-fused molecule (molecular mass, 46.8 kDa), are small enough to diffuse freely through nuclear pore complexes. We assume an active mechanism for the retention of PtSNAP in the cytosolic compartment, which becomes inactivated upon fixation.

Functional aspects.

Unlike the role expected from its homology to mammalian SNAP-25, we could not find a role for PtSNAP in the stimulated exocytosis of dense core vesicles (trichocysts). This was unexpected, because PtSNAP exists as a single transcript and successful gene silencing could be demonstrated by Western blotting with the specific anti-PtSNAP antibody. However, Paramecium contains several other Qb- and Qc-SNAREs (C. Schilde, unpublished results), so there could be redundancy of function. Such a functional redundancy has been observed for SNAREs in many other cases (43, 62, 76). Accordingly, in certain mammalian cell types, posttranscriptional gene silencing or expression of a dominant-negative mutant form of SNAP-23 has not led to any phenotypic defects in secretion, even though SNAP-23 is the only SNAP-25 homologue normally present in those cells (49). In conclusion, from our data, we cannot exclude the possibility that redundancy of function masked a possible effect of PtSNAP on trichocyst exocytosis.

We observed an increase in the number of food vacuoles per cell in PtSNAP-silenced Paramecium cells. Feeding of silenced cells with pH indicator Congo red-stained yeast cells showed that this is due to an increased uptake of food vacuoles (data not shown), not to a defect in food vacuole processing and/or defecation, and we could exclude a defect in the acidification of food vacuoles. Another possibility is that the total capacity of the digestive system is limited by the availability of acidosomes. The slightly enhanced division rate of PtSNAP-silenced cells could point to an increased energy supply from an increased number of food vacuoles. The localization of PtSNAP observed at the cytostome could indicate a role there in food uptake.

Attenuation of SNARE expression does not always have to be deleterious, as shown by the improved salt tolerance of A. thaliana plants depleted of AtVAMP714 (42). Also, a role for so-called inhibitory SNAREs in fine-tuning membrane fusion specificity by engagement in nonproductive SNARE complexes has been suggested (73). Thus, the lack of a deleterious effect of PtSNAP silencing could be explained by a release of an inhibition state, if PtSNAP would act as an inhibitory SNARE. A closer investigation of the effects of PtSNAP gene silencing on food vacuole processing will be needed to clarify the exact role of PtSNAP in this process.

Conclusions.

In summary, the present work is the first investigation of a SNAP-25 homologue in protists and opens the exciting opportunity to study the role of such dual-SNARE-motif-containing proteins outside the animal kingdom. The results from the glutamine-rich PtSNAP of Paramecium are important because a similar asparagine-rich SNAP-25 homologue exists in the malaria parasite Plasmodium falciparum (gi|23619154), an apicomplexan related to ciliates, both of which are contained in the phylum Alveolata. Although it is difficult to assign a precise role to PtSNAP in the phagocytic cycle, it evidently plays a role in this complex process.