A Multigene Family Encoding R-SNAREs in the Ciliate Paramecium tetraurelia
Christina Schilde*
,a, Thomas Wassmer, Joerg Mansfeld, Helmut Plattner and Roland KissmehI
Chair of Cell Biology and Ultrastructure Research, University of Konstanz, PO Box 5560, 78457 Konstanz, Germany
*Corresponding author: Christina Schilde, christina.schilde@uni-konstanz.de
SNARE proteins (soluble
N-ethylmaleimide-sensitive fac-tor attachment protein receptors) mediate membrane interactions and are conventionally divided into Q-SNAREs and R-SNAREs according to the possession of a glutamine or arginine residue at the core of their SNARE domain. Here, we describe a set of R-SNAREs from the ciliate
Paramecium tetraureliaconsisting of seven families encoded by 12 genes that are expressed simultaneously. The complexity of the endomembrane system in
Parameciumcan explain this high number of genes. All
P. tetraureliasynaptobrevins (PtSybs) possess a SNARE domain and show homology to the Longin family of R-SNAREs such as Ykt6, Sec22 and tetanus toxin-insensitive VAMP (TI-VAMP). We localized four exemplary
PtSyb subfamilies with GFP constructs andantibodies on the light and electron microscopic level.
PtSyb1-1, PtSyb1-2 and PtSyb3-1 were found in the
endoplasmic reticulum, whereas
PtSyb2 is localizedexclusively in the contractile vacuole complex.
PtSyb6was found cytosolic but also resides in regularly arranged structures at the cell cortex (parasomal sacs), the cytoproct and oral apparatus, probably representing endocytotic compartments. With gene silencing, we showed that the R-SNARE of the contractile vacuole complex,
PtSyb2,functions to maintain structural integrity as well as functionality of the osmoregulatory system but also affects cell division.
Key words: ciliates, membrane trafficking,
Paramecium,regulated secretion, SNAREs, R-SNAREs, v-SNAREs Received 21 September 2005, revised and accepted for publication 10 January 2006
Intracellular vesicle trafficking between different organelles, as well as to and from the cell surface, is a hallmark of eukaryotic cells. These membrane interactions are tightly regulated, and the molecular machinery mediating these events is the subject of intense current research. Although many details are similar, some are highly divergent in different cells or from one membrane system to another.
We are particularly interested in such aspects in the
ciliated protozoan Paramecium tetraurelia that possesses elaborate membrane-trafficking systems (1). Membrane interactions, such as the fusion of intracellular vesicles with other vesicles or the cell membrane, are mediated by soluble N-ethylmaleimide-sensitive factor attachment protein receptors (SNAREs) which are characterized by a heptad repeat coiled-coil SNARE motif (2–4). SNARE pro- teins mediate membrane interactions and have been classi- fied as target-SNAREs (t-SNAREs) and vesicle-SNAREs (v-SNAREs) according to their localization (5). Before membranes fuse, v-SNAREs and t-SNAREs in the opposing membranes form a stable trans-SNARE complex with their coiled-coil SNARE domains that is thought to be prerequisite to bring about fusion of the vesicle with the target membrane (6–8). Because SNARE pairing is specific to different membrane-trafficking pathways, cognate pairing of individual v-SNAREs with corresponding members of the large family of t-SNAREs is required before fusion can occur (4,6,7). This is reflected by the distinct subcellular localizations of different SNARE isoforms (4). One of the best characterized SNARE complexes is the synaptic vesicle SNARE-pin, formed by the t-SNAREs syntaxin-1 and synaptosomal-associated protein of 25 kDa (SNAP25), which contributes two helices to the complex, and the v-SNARE vesicle-associated membrane protein-2 (VAMP-2) or synaptobrevin-2 (2). The homo- hexameric ATPase and SNARE-specific chaperone N-ethylmaleimide-sensitive factor (NSF) and its adaptor pro- tein a -soluble NSF attachment protein ( a -SNAP) disentangle the cis-SNARE complexes after membrane fusion into single SNARE proteins that can then be retrieved for another round of vesicle fusion, but NSF may also be important before vesicle fusion for vesicle tethering (5,9–12). A conserved arginine or glutamine at the centre of the SNARE domain is important for the attachment of a -SNAP and sub- sequent NSF action. Therefore, another nomenclature classes the SNAREs into R-SNAREs and Q-SNAREs, respectively (13). As a general rule, three Q-SNAREs and one R-SNARE (synaptobrevin) are participating in a SNARE complex, the so-called 3Q þ 1R SNARE rule (13). R-SNAREs are characterized by the synaptobrevin- type SNARE domain immediately preceding a transmem- brane anchor at the C-terminus. They are either lacking an amino-terminal domain (‘brevins’) or possess an amino- terminal, so-called longin domain, and are accordingly called ‘longins’ (14,15). Longins are present in all eucar- yotes (16), whereas the distribution of brevins as known so far is more restricted to animals and yeast (15). The brevins of metazoans are the only R-SNAREs sensitive to tetanus neurotoxins (17) and are characterized by a speci- fic RD motif at the centre of their SNARE domain, in contrast to the RG motif seen in all other R-SNAREs (15).
aPresent address: Institute of Biochemistry, Schafmattstraße 18, ETH Ho¨nggerberg, HPM F 8, 8093 Zu¨rich, Switzerland.
Blackwell Munksgaard doi: 10.1111/j.1600-0854.2006.00397.x
440
Konstanzer Online-Publikations-System (KOPS) URL: http://www.ub.uni-konstanz.de/kops/volltexte/2007/4317/
URN: http://nbn-resolving.de/urn:nbn:de:bsz:352-opus-43179
The Paramecium cell is a highly complex eucaryotic cell with well-defined internal organization and displays numerous clearly defined vesicle-trafficking routes, including the standard repertoire of endocytosis, phago- cytosis, stimulated exocytosis of dense core vesicles (trichocysts) and constitutive exocytosis (1,18,19).
Coated vesicle-mediated endocytosis occurs at regularly arranged parasomal sacs (20), and the early endosomes (‘terminal cisternae’) of Paramecium are located under the regularly arranged basal bodies (21,22).
Phagocytosis of food particles takes place on a special- ized invagination of the cell surface, the oral apparatus (23). The food vacuole then takes a well-defined route (cyclosis) through the cell during which it undergoes different stages of acidification and neutralization, fusion with lysosomes and retrieval of membranes, before indigestible waste materials are eventually expelled at another defined structure, the cytoproct (23–26).
Internal fusion processes involve the anterograde and retrograde vesicle-trafficking routes between the endo- plasmic reticulum (ER) and the Golgi apparatus (22), bud- ding of constitutive secretory vesicles and biogenesis of trichocysts by multiple fusion processes (27,28). The osmoregulatory contractile vacuole complex, of which there are two units in a P. tetraurelia cell, consists of a periodically fusing membrane system of collecting canals and ampullae emptying into the central contractile vacuole that periodically opens through a porus on the cell surface to expel its contents to the outside medium (29,30). The collecting canals are connected to a system of tubules and vesicles, the spongiome, which has two structurally different components: the smooth and the decorated part of the spongiome.
Recently, we described the function of the SNARE-specific chaperone, NSF, in P. tetraurelia (31,32), and here we report the presence of R-SNAREs in this organism represented by a family with 15 genes. The P. tetraurelia synaptobrevins (PtSybs) were identified as longin-type R-SNAREs by using sequence homology. Using antibodies directed against PtSybs and GFP fusion constructs, we found a distinct subcellular distribution of different members of PtSybs. We used gene silencing by feeding (33) to investigate some PtSyb functions in P. tetraurelia and found profound effects of PtSyb2 silencing on cell and contractile vacuole complex morphology as well as affect- ing its function. However, we expect that highly homolo- gous subfamily members can be cosilenced when they share a certain degree of similarity (34).
Results
A family of 12 genes encoding synaptobrevins in P. tetraurelia
Among the 722 protein-coding genes identified in the course of the pilot sequencing project of the P. tetraurelia macronuclear genome (35,36) by automated BLAST
searches (37), we found two partial sequences with homology to mammalian synaptobrevin. The respective clones, M18F03u (EMBL-Bank accession number AJ566298) and M10H02u (accession number AJ566299), were named Ptsyb1-1 and Ptsyb2-1, respectively.
Sequencing of the corresponding clones and amplification with gene-specific primers from a Paramecium cDNA library (38) revealed cDNA information about these genes. Beyond this, we could identify two other relevant genes, Ptsyb2-2 and Ptsyb3-1, on the cDNA level.
Comparison of these sequences with the genomic version allowed us to determine number, size and position of the introns in all four genes, including Ptsyb2-2 and Ptsyb3-1 (EMBL-Bank accession numbers AJ566300 and AJ566301). We performed further searches of the genomic sequence with those four genes and were able to identify a family with 12 members all encoding R-SNAREs and three related longin-like fold containing pro- teins (PtSyb4-1, PtSyb4-2 and PtSyb5-1) (Table 1). The remaining Ptsybs have been deposited by Genoscope (www.genoscope.cns.fr) at EMBL-Bank (www.embl.org) under the accession numbers shown in Table 1. Both macronuclear DNA and cDNA sequence information were experimentally verified by amplification with gene-specific oligonucleotide primers followed by sequencing, and the positions of introns were thus identified. The Ptsyb genes encode proteins between 23 and 27 kDa and contain between two and five small introns of 21–31 bp length flanked by the conventional Paramecium intron borders 5
0-GTA . . . T/A/CAG-3
0(36,39,40) (Table 1). Most Ptsyb gene families comprise two closely related isoforms.
Exceptions are the subfamilies PtSyb3, PtSyb5 and PtSyb8, which are represented by only a single member.
The identity at amino acid level between the isoforms of a subfamily lies between 64.6 and 96.6%.
PtSyb6-1 and PtSyb6-2 are 93.2% identical at amino acid level, but intron splicing of the Ptsyb6-2 gene at the first intron creates a stop mutation and would result in the expression of only a 57 amino acid fragment of the protein (Figure 1A). As the presence of spliced Ptsyb6-2 cDNA was experimentally verified, the existence of such a short peptide is possible. However, no band in the range of the expected molecular weight of a PtSyb6-2 peptide could be detected in Western blots with an antibody directed against the N-terminus of Ptsyb6-1 (data not shown), suggesting that translation of the Ptsyb6-2 mRNA does not occur (see Discussion).
An additional gene with homology to Ptsyb3-1, hence named Ptsyb3-2, was found in the genome sequence.
However, sequencing of the corresponding cDNA
revealed that splicing of the first intron causes a shift of
reading frame and results in a truncated protein even
though the sequence homology to PtSyb3-1 on the
amino acid level continues in a different reading frame
well after this termination codon. It remains unknown so
far whether this gene is functional.
The degree of sequence conservation to R-SNAREs not only to other species but even within the different synaptobrevin gene subfamilies of P. tetraurelia is relatively low ( < 30%
identity; Table 1). However, protein family homology searches revealed homology to the longin domain of Ykt6 and Sec22 and especially to the longins of plants (Arabidopsis thaliana V711, V712, V713, V714, V724, V725 and V727).
By homology to other R-SNAREs, the Paramecium synaptobrevins are R-type SNAREs, but unexpectedly not all of them actually exhibit the conserved arginine residue in the central (zero-) layer of the SNARE domain (Figure 1B). The subfamilies PtSyb8 and PtSyb9 lack the conserved arginine but do not possess a glutamine typi- cal of Q-SNAREs at this position either. However, the homology to the R-SNARE longin domain of PtSyb8 and overall sequence homology of the PtSyb9 isoforms to other PtSybs clearly place them into the longin-type family of R-SNAREs.
The majority of PtSybs possess a C-terminal stretch of 17–24 hydrophobic amino acids for membrane insertion (Figure 1A,B). There are two exceptions: the protein subfamilies PtSyb6 and PtSyb7 lack such a C-terminal hydro- phobic stretch. They might instead, in analogy to their closest mammalian and yeast homologue Ykt6, be anchored in the membrane by prenylation and palmitoyla- tion of cysteine residues of a C-terminal CCXXF/Y motif (41–43) (Figure 1B)–an aspect not pursued further here.
When we analysed the phylogenetic relationships between the different families of PtSybs, we obtained robust trees for the major branches by different methods
(Figure 2). The phylogenetic tree also revealed that Paramecium possesses synaptobrevins of the VAMP7 and Ykt6 type. Interestingly, we were not able to identify a Sec22 homologue in Paramecium, whereas in another ciliate, Tetrahymena thermophila, a Sec22 homologue clearly exists. Comparison with synaptobrevins of T. ther- mophila identified homologues close to the Paramecium synaptobrevins (Figure 2). The loss of the transmem- brane domain as a single event in the phylogeny of the synaptobrevins is illustrated by the close clustering of the PtSyb6 and PtSyb7 families. They all lack the transmem- brane domain (Figure 1A,B) and group in the constructed phylogenetic tree with Ykt6 and their transmembrane domainless T. thermophila orthologues (Figure 2). Other Paramecium R-SNAREs such as PtSyb3-1, PtSyb8 and PtSyb9 appear to have no equivalent in higher organisms and might therefore represent Paramecium- or ciliate- specific R-SNAREs with orthologues for at least some of them present in Tetrahymena. Another clade com- prises PtSyb subfamilies 1 and 2 and paralogues of plant and mammalian longin domain containing synapto- brevins, representative of vacuolar, endosomal, ER and Golgi synaptobrevins that were grouped together as VAMP7-like proteins (44). However, according to our pre- sent knowledge, Paramecium does not possess homo- logues of the VAMPs considered relevant for regulated exocytosis (8,9), the brevins (see Discussion). A tree constructed from only the SNARE domains of the R-SNAREs confirmed the Ykt6-like group and the probably ciliate-specific PtSyb8/PtSyb9 clade but failed to validate a VAMP7-like and a Sec22-like group and was supported only by low bootstrap values (Supplemental Figure 1, Supplementary Material ).
Table 1: Synaptobrevins ofParamecium tetraurelia
Gene Accession number bp Introns aa kDa % identity between % identity to % identity toPtSyb1-1 Number Size (bp)
isoforms at bp level isoforms at aa level at aa level
PtSyb1-1 AJ566298 761 3 23–28 228 26.6 – – –
Ptsyb1-2 CR855907 760 3 22–27 228 26.3 83.9 85.2 85.2
Ptsyb2-1 AJ566299 741 4 25–29 210 24.5 – – 29.9
Ptsyb2-2 AJ566300 742 4 26–29 210 24.4 81.1 69.7 28.4
Ptsyb3-1 AJ566301 690 3 21–28 205 23.8 – – 21.8
Ptsyb4-1a CR855905 648 3 22–27 191 22.6 – – 12.0
Ptsyb4-2a CR855904 645 3 23–28 188 22.2 74.8 64.6 18.0
Ptsyb5-1a CR855903 528 1 24 167 19.3 – – 14.3
Ptsyb6-1 CR855902 663 2 23þ25 204 23.6 – – 13.2
Ptsyb6-2 CR855978 667 2 24þ28 57 6.5 88.9 93.2 16.1
Ptsyb7-1 CR855901 718 4 23–29 202 23.6 – – 12.8
Ptsyb7-2 CR855900 715 4 23–29 202 23.6 87.5 96.6 13.8
Ptsyb8-1 CR855899 764 3 24–31 228 26.0 – – 15.4
Ptsyb9-1 CR855898 789 5 23–27 220 25.8 – – 15.4
Ptsyb9-2 CR855897 788 5 23–27 220 25.7 84.6 76.9 15.4
P. tetraureliasynaptobrevin genes. Accession numbers AJ566298 to AJ566301 (GenBank) and CR855897 to CR855907 (Genoscope).
Sequences were aligned by the ClustalW method (slow/accurate) and the percentage of identities between isoforms was calculated.
Numbers and sizes of introns were experimentally verified and the resulting molecular mass is given in kDa.
aThe synaptobrevin-related genes Ptsyb4-1, Ptsyb4-2 and Ptsyb5-1 contain a longin-like fold, but no recognizable SNARE domain;
however, they share sequence homology toPtsyb3-1.
PtSyb1 and PtSyb3 are localized in the ER
We created C-terminal GFP fusions of PtSyb1-1, PtSyb1-2, PtSyb-2-2, PtSyb3-1 and an N-terminal GFP fusion of PtSyb6-1 and microinjected the constructs into the macro- nucleus of P. tetraurelia cells. (An N-terminal attachment of GFP was preferred for Ptsyb6-1 because of the
expected anchorage of PtSyb6-1 to the vesicle membrane by C-terminal prenylation and/or palmitoylation and the possibility of interference of C-terminal GFP fusion with this modification.) In addition, immunolabelling of fixed cells with the available anti-PtSyb (Figure 3) and other antibodies and co-localization with organellar markers such A
B
240 Aa 228 228 210
210
Longin domain
Longin domain with low score
Synaptobrevin coiled-coil homology
Predicted transmembrane
domain 205
204 57*
202 202
220 227
220
SNARE domain Transmembrane region
PtSyb1-1
12 117
117
112
112
118
110
57
65 130
130
100 136
131
131 97
96
142 142
65
142 85
127
127
131 191
191193 213
210 193 187
210 193 187
178 186 203
202
202
202
202 221 196
191
191 197
197 216
216 216 197
131 12
7
7
6
7
7
9
8
8
PtSyb1-2 PtSyb2-1 PtSyb2-2 PtSyb3-1 PtSyb6-1 PtSyb6-2 PtSyb7-1 PtSyb7-2 PtSyb8-1 PtSyb9-1 PtSyb9-2
Figure 1: (A) Domain structure ofParamecium tetraureliasynaptobrevins (PtSybs).Domain predictions are based on homology to known proteins (Pfam, PDB) and identified longin domains with high scores (green) or with low scores of >0.01(light green) and synaptobrevin SNARE domains (blue). Potential transmembrane domains (yellow) were determined by TMpred (75).Ptsyb6-2produces a truncated protein of 57 amino acids (*) because of the creation of a stop codon at position 172 by intron splicing. (B) Amino acid alignment (ClustalW; slow, accurate) of PtSybs showing SNARE domains and the carboxy-terminal transmembrane domains. The conserved zero-layer arginine (red) is not present in allPtSyb8-1,PtSyb9-1 andPtSyb9-2, but histidine or asparagine residues occupy that position there. Conserved heptad repeats (black) of the SNARE domains are aligned. The transmembrane domains are between 17 and 24 amino acids long and thePtSyb families 6 and 7 are lacking a transmembrane anchor but instead possess a potential carboxy- terminal putative prenylation/palmitoylation motif (yellow).
as 3,3
0-dihexaoxacarbocyanine iodide [DiOC
6(3)], a marker for the endoplasmic reticulum (45,46), were performed.
A GFP fusion protein of PtSyb1-1 showed staining of the cortical ER network (Figure 4A,B). This staining pattern is remarkably similar to staining of living cells with the ER marker DiOC
6(3) (Figure 4C). A similar, but less distinct staining pattern was obtained with a PtSyb1-2-GFP fusion construct, where a membrane system stretching through- out the cell and a cortical pattern was labelled (Figure 4D,E). The staining of the ER was not so clear in
cells transfected with a PtSyb3-1-GFP construct, because a high cytoplasmic background resulting from the high overexpression level obscured any potential ER staining (Figure 4F). For this reason, staining of the endogenous proteins with specific antibodies was performed.
Immunolabelling of fixed cells with the anti-PtSyb1 anti- body showed a reticular staining that stretches throughout the cell with some concentrations near the cortex resem- bling the ER, and costaining of the anti-PtSyb1 antibody with the established ER marker DiOC
6(3) showed close overlap with anti-PtSyb1 staining (Figure 4G–I). However,
76 88 48
48
99
99 35
99 99
58 49
46 29 69
66 21
39
89 89 30
31 74
99 99 98
95
100
100
100 100
DDB0219666 HsYkt6
ScYkt6 PtSyb7-1 PtSyb7-2 Tt4410
Tt4411 PtSyb6-1
Tt14913
Tt19728 Tt2415
Tt4722
Tt11170 Tt20867 AtVAMP711
AtVAMP712 AtVAMP713 AtVAMP714
AtVAMP723 AtVAMP725
AtVAMP724 AtVAMP727 DdVamp7B
DdVamp7A
DDB0219546 DdSybA
EhSYBL1
ScNyv1 ScSnc1
ScSnc2 HsSyb1 HsVAMP2 HsVAMP3
EhSYBL2
Tt7285 ScSec22
DDB0190688 brevins
VAMP7-like ciliate-specific?
Ykt6-like
Sec22-like PtSyb1-1
PtSyb1-2
PtSyb2-1 PtSyb2-2 PtSyb3-1 HsSybl1
PtSyb8-1
100
100 100
100
100
100
100
99
99
63
0.2
100
Figure 2: Synaptobrevin family phylogeny.Neighbour joining tree with 1000 bootstrap replicates encompassingParamecium tetra- urelia synaptobrevin proteins (PtSybs; seeTable 1), Tetrahymena thermophila(Tt) synaptobrevin homologues from preliminary gene predictions 08–2004 at TIGR (gene identifiers 2415, 4410, 4411, 4722, 7285, 11170, 14913, 19728 and 20867),Arabidopsis thaliana (At) VAMPs 711–714, 723–725, 727 (GenBank accession numbers O49377, Q9SIQ9, Q9LFP1, Q9FMR5, Q8VY69, O23429, O48850, Q9M376),Homo sapiens(Hs) synaptobrevin 1 (Syb1), VAMP2, VAMP3, Ykt6 and synaptobrevin-like protein 1 (Sybl1) (GenBank accession numbers AAA60603, AAH02737, CAB63146, CAG46805, AAH56141),Saccharomyces cerevisiae(Sc) Snc1, Snc2, Nyv1, Ykt6 and Sec22 (GenBank accession numbers AAC05002, NP_014972, NP_013194, AAB32050, AAB67373), Dictyostelium discoideum (Dd) SybA, VAMP7A, VAMP7B (DDB0214903, DDB0231535, DDB0231542) and DDB0190688, DDB0219546, and DDB0219666 andEntamoeba histolytica (Eh) synaptobrevin-like proteins 1 and 2 (Sybl1 and Sybl2) (GenBank accession numbers AY256852 and AY309014).
Parameciumsynaptobrevins addressed in this work are indicated with grey diamonds. Bootstrap support values are indicated above the branches and an evolutionary distance scale is given below the figure.
the appearance of the staining with both the anti-PtSyb1 antibody and DiOC
6(3) was variable between experiments and under different growth conditions. Such variability has been described previously (46), and we conclude that both PtSyb1-1 and PtSyb1-2 are present in the ER of Paramecium. PtSyb3-1 was localized using the anti- PtSyb3-1 antibody to the ER network and also showed close overlap with DiOC
6(3) staining (Figure 4K–M). At the ultrastructural level, staining with the anti-Syb1 antibody in areas enriched in ER and near vesicles of unknown identity was observed (Figure 5A,B). PtSyb1 was enriched in ER-rich zones 3.5-fold ( 1.2), whereas all other cortical structures, such as alveolar sacs, tricho- cyst docking sites, terminal cisternae and cilia were devoid of PtSyb1 staining. Electron microscopic (EM) analyses with the anti-PtSyb3 antibody showed clusters of staining in close proximity to membranes of the ER (Figure 5C,D).
PtSyb3 was enriched 1.95 times ( 0.4) in such ER-rich domains.
PtSyb6 is cytosolic and associated with parasomal sacs and the cytoproct
When an amino-terminal GFP-PtSyb6-1 fusion construct was expressed in Paramecium cells, predominantly cytosolic staining was observed (Figure 6A). If such GFP- PtSyb6-1-transfected cells were permeabilized with 1%
Triton X-100 or 0.5% digitonin, GFP fluorescence disap- peared completely, indicating that PtSyb6-1 is predomi- nantly cytosolic (data not shown). This may reflect the absence of a transmembrane domain anchorage of the protein. After careful prefixation followed by permeabiliza- tion with digitonin, an antibody raised against PtSyb6-1 decorated the cytoproct and a regular surface pattern as well as a large population of small vesicles in the cell
(Figure 6B,C). Labelling of the cytoproct (Figure 6B) may reflect involvement of PtSyb6 in the release of spent phagosome contents by a specialized form of exocytosis (47). Furthermore, PtSyb6 staining was detected around the oral apparatus and in some cells, additionally the radial arms of the contractile vacuole complex were decorated.
Closer inspection revealed labelling between trichocyst docking sites, i.e. associated with ciliary basal bodies (Figure 6D,E). On isolated cell cortices, the anti-PtSyb6 antibody labelled a regular surface pattern in close apposi- tion to basal bodies, but PtSyb6 staining did not exactly co-localize with basal bodies (Figure 6F). From this, we conclude that PtSyb6 is located in the parasomal sacs of Paramecium. Within the cortex, immuno-EM gold labelling revealed staining of vesicles associated with the endocy- totic pathway that are situated in the vicinity of basal bodies (Figure 6G). PtSyb6 was enriched over the ER domains 7.8-fold ( 1.5). Enrichment of PtSyb6 is 15.1-fold ( 1.0) over terminal cisternae (early endosomes) with surrounding small vesicle-rich domains. Furthermore, some enrichment of PtSyb6 occurred in domains with acidosomes (approximately 2.3-fold) and discoidal vesicles (approximately 2.9-fold), but this was difficult to ascertain statistically. Particularly in the case of acidosomes, one has to consider that in contrast to all other structures mentioned, enrichment is referred to an organelle with relatively small surface to volume ratio, while SNAREs can reside only on the organelle surface.
PtSyb2 is localized in the contractile vacuole complex The fusion protein PtSyb2-2-GFP stained the entire con- tractile vacuole complex of the Paramecium cell, i.e. the vacuole porus, the contractile vacuole, the collecting ampullae and the collecting (radial) canals (Figure 7A).
Staining with the anti-PtSyb2 antibody displayed the same distribution as was found for the GFP fusion con- struct (Figure 7B). Close inspection showed that the stain- ing along the radial canals was slightly irregular, possibly representing a number of small vesicles and/or tubules (Figure 7B, inset). Immunogold labelling of cells expres- sing PtSyb2-2-GFP with an anti-GFP antibody (48) showed staining of the membraneous network closely apposed to the radial canals (‘smooth spongiome’) along the radial canal, the ampulla and of the membrane of the contractile vacuole (Figure 7C). The same staining was obtained in non-transfected cells with the anti-PtSyb2 antibody (Figure 7D). The enrichment of PtSyb2 over the smooth spongiome was 2.55-fold ( 0.1). With none of the anti- bodies used, we have seen any enrichment of pA-Au
5labelling in the decorated spongiome.
Effects of homology-dependent gene silencing of PtSyb2
Because of the very confined localization of PtSyb2 to the contractile vacuole complex, we choose it as an example to study its function by gene silencing. Cells were fed with Escherichia coli expressing double-stranded RNA coding for PtSybs, which elicits a small interfering RNA
α-PtSyb1homogenatesuper natant
pellet microsomescor tices
α-PtSyb2
α-PtSyb3
α-PtSyb6
kDa 25 25
25 25
Figure 3: Western blots of Paramecium tetraurelia cell fractions with the R-SNARE-specific antibodies used for immunolocalizations.The antibodies recognize the monomeric PtSybs in homogenates, the 100 000gpellet and microsome fractions when they are boiled in 10% SDS for 10 min.PtSyb1 is also detected in isolated cortices with the anti-PtSyb1 antibody.
(si-RNA)-mediated silencing process of the target genes.
To exclude any effect of the silencing on its own, vector only containing the gfp gene was used for mock silencing.
In fixed gfp-silenced cells, normal staining pattern was seen with anti-PtSyb2 and with an antibody against the a-subunit of the V
0part of the Paramecium V-ATPase (anti- PtV
0aSU) which is localized in the rough spongiome apposed to the smooth spongiome (49,50) (Figure 8A,B).
As a positive control, the nd7 gene, which is essential for trichocyst exocytosis (34,51,52), was silenced in parallel experiments and 99% discharge-deficient cells were obtained after 24 h (data not shown). When cells were
silenced for 24 h with Ptsyb2-1 and Ptsyb2-2 silencing constructs, staining of the contractile vacuole complex with the anti-PtSyb2 antibody disappeared completely, whereas the same could still be decorated with the anti- V
0-ATPase a-subunit antibody (Figure 8C–F). This under- lines the specificity of the anti-PtSyb2 antibody for the PtSyb2 family proteins. However, in Ptsyb2-silenced cells, the contractile vacuole complex often appeared mal- formed, with inflated radial canals (Figure 8F). Silencing of Ptsyb2-1, and in a more pronounced way of Ptsyb2-2, or with a double-silencing construct of both genes led, after 40 h, to a morphological defect of cells that also affected
Syb1-1-GFP liveDiOC
DiOC
Syb1
merge merge
Syb3 Syb1-2-GFP fixed
Syb1-2-GFP live Syb3-1-GFP live
Syb1-1-GFP live DiOC
A B C
D
G H I
K L M
E F
Figure 4: Localization ofParamecium tetraureliasynaptobrevin families 1 and 3 in the endoplasmic reticulum (ER) 1.(A) Surface and (B) median view of a live cell expressing aPtSyb1-1-GFP fusion protein showing a cortically enriched tubular-reticular staining resembling structures of the ER. (C) Non-transfected, living cell stained with DiOC6(3), cortical view. (D) Median view of a live cell expressing aPtSyb1-2-GFP fusion protein showing weak fluorescence in the internal ER network. (E) A fixedPtSyb1-2-GFP-expressing cell shows a similar cortical staining pattern like PtSyb1-1-GFP; however, a higher fluorescence in the internal ER network is observed. (F) Median view of a live cell expressing aPtSyb3-1-GFP fusion protein showing strong cytoplasmic fluorescence that masks any possible ER staining. (G – I) Confocal image of a fixed non-transfected cell stained with DiOC6(3) (green channel) and anti-PtSyb1 (red channel) shows co-localization of both markers in the ER, but not in the food vacuoles that were only labelled with DiOC6(3). (K – M) Confocal image of a fixed non-transfected cell stained with DiOC6(3) (green channel) and anti-PtSyb3 (red channel) shows co-localization of both markers in the ER, but not in the food vacuoles that were only labelled with DiOC6(3). Scale bars¼10mm.
their swimming behaviour (compare Figure 9A,B and Supplementary Material). Most cells failed to divide prop- erly and gave rise to doublet cells of various shapes (Figure 9C – E). The often rounded or boomerang-shaped cells were rotating quickly around their axis and after 72 h eventually stopped feeding and died. The osmoregulatory system of silenced cells was still working, but the con- tractile vacuole was pumping at much reduced frequency (Figure 9F). Trichocyst exocytosis triggered by picric acid was not affected in the silenced cells indicating a specific effect of Ptsyb2 gene silencing on the osmoregulatory system and cell division (Figure 9D,E). Interestingly, the effects of silencing Ptsyb2-2 or of both isoforms with the tandem silencing construct Ptsyb2-1-Ptsyb2-2 were more pronounced compared to silencing of Ptsyb2-1 alone indicating a functional non-redundance of both isoforms.
Discussion
In the present work, we describe 12 genes coding for R-SNAREs in P. tetraurelia. Exhaustive searches of the P. tetraurelia macronuclear genome presumably revealed the complete set of R-SNAREs in this organism. By com- parison, the yeast genome contains only five R-SNARE
genes (53), the human genome nine (54), whereas in the plant A. thaliana 14 R-SNAREs have been identified (44).
When we consider the multitude of membrane inter- actions in the Paramecium cell (1), this high number of R-SNAREs is not at all surprising, and if each gene were specific to a particular membrane interaction an even higher number of synaptobrevins would be expected.
From other systems, it is known that R-SNAREs can par- ticipate with different binding partners in two or more different SNARE complexes (55–58). For instance, in yeast, the SNARE complex involved in transport from the ER to the cis-Golgi consists of Sec22p (R-SNARE), Sed5p (Qa-SNARE), Bos1p (Qb-SNARE) and Bet1p (Qc-SNARE), respectively (59–62). In mammalian cells, the homologous proteins include Sec22b (R-SNARE), syntaxin 5 (Qa- SNARE), membrin (Qb-SNARE) and Bet1 (Qc-SNARE) (63). In both, yeast and mammals, Sed5p/syntaxin 5 is also involved in trafficking steps within the Golgi apparatus where it is assumed to interact with a homologous set of SNAREs including Ykt6p/Ykt6 (R-SNARE), Gos1p/GS28 (Qb-SNARE) and Sft1p/GS15 (Qc-SNARE) (43,64–66).
Thus, the occurrence of promiscuous R-SNARE pairing in Paramecium is a possibility and could account for the high number of specific membrane interactions in this cell (1).
A
C D
B
ERER
mi
mi
Figure 5: Electron micrographs of immunogold labellings with anti- PtSyb1 and anti-PtSyb3 antibodies.
(A) Overview of an electron micro- graph showing immunogold labelling of ER-rich domains with anti-PtSyb1.
Scale bar¼500 nm. (B) At higher magnification from (A) (box) gold labelling close to ER membranes is visible (circles). Scale bar¼500 nm.
(C) Overview of an electron micro- graph showing immunogold labelling of ER-rich domains with anti-PtSyb1.
Scale bar¼500 nm. (D) At higher magnification (box), gold labelling of a region enriched in ER is visible (circles). Scale bar¼100 nm. ER, endoplasmic reticulum; mi, mitochon- drium.
A
f
m
oa oa
oa
oa
oa t*
t*
t
t
t oa
cp
D
F G
E
B C
as
bb
Figure 6: Localization of Paramecium tetraureliasynaptobrevin family 6.(A) GFP fluorescence in a live cell expressing a GFP- PtSyb6-1 fusion protein shows mainly cytoplasmic staining with the exception of the oral apparatus (oa), food vacuoles (f) and the macronucleus (m). Bright objects represent crystals in the cell. Scale bar¼10mm. (B) Immunostaining of a fixed non-transfected cell with a polyclonal anti-PtSyb6-1 antibody shows strong staining of the cytoproct (cp) and a punctate cortical staining (B – surface focus and C – median focus). Cells were carefully prefixed in 2% formaldehyde and permeabilized after prefixation with 0.5% digitonin. Scale bar¼10mm. (D, E) Detail from (D) confocal image slice (thickness 1mm) of a fixed cell expressing the GFP-PtSyb6-1 fusion protein. The oral apparatus (oa) and trichocyst bodies (t) are not stained. Internally decondensed trichocysts (t*) that trapped free GFP-PtSyb6-1 during fixation appear labelled. A regular pattern of GFP-PtSyb6-1 in the cell cortex can be seen (arrows) with labelling occurring at the level of trichocyst bodies (t), which in cross-section appear as dark circles. Label is enriched around the oral apparatus. Scale bar¼10mm. (F) Confocal image slice of an isolated cell cortex showing close apposition, but no co-localization of basal bodies labelled with anti-tubulin (red) and parasomal sacs labelled with the anti-PtSyb6 antibody (green). Scale bar¼10mm. (G) Immuno-EM gold labelling (circles) of a cluster equivalent to fluorescently labelled spots between trichocysts seen in (D) and (E), representing vesicles of the endocytotic pathway near ciliary basal bodies (bb). as, alveolar sac; e, endocytic vesicle. Scale bar¼100 nm.
All Paramecium R-SNAREs are of the longin type similar to the plant R-SNAREs (44) underlining the closer relation- ship of ciliates to plants than to yeast and animals, which also contain R-SNAREs of the brevin type responsible for regulated fast exocytosis. We identified two PtSyb families (PtSyb6 and PtSyb7) with homology to the longin-type protein Ykt6, but surprisingly no Sec22 homo- logue was found in P. tetraurelia. However, it is possible that potential genes escaped the data mining approach due to high evolution rates or if they were disrupted by many or unconventional introns.
The Paramecium SNAREs also may not follow the con- vention of the ‘3Q þ 1R-rule’ (9,13), because three out of
12 R-SNAREs do not exhibit the conserved arginine resi- due in the central (zero-) layer of the SNARE domain (Figure 1B). This is so far known only for a small number of syntaxins (Q-SNAREs) such as Giardia intestinalis SynPM2, Encephalitozoon cuniculi SynPM and Plasmodium falciparum SynPM. These findings could be explained by the high AT content of those organisms and a resulting CGX (R) to AAY (N) codon bias (67). The Paramecium genome also has a high AT-content of 72%
(35,40), and similar exchanges of the arginine residue were observed in the T. thermophila R-SNAREs. The yeast Qc-SNARE Bet1, like other fungal homologues, also carries a serine instead of a glutamine at the con- served position; however, it functions normally in SNARE
A B
C
D
ss
ss ss
ds
ds ds
ss rc
rc
rc
v a
r r
r
r v
v
Figure 7: Localization ofParamecium tetraureliasynaptobrevin family 2 in the contractile vacuole complex.(A) GFP fluorescence in cells expressing aPtSyb2-2-GFP fusion construct. Staining was observed along radial arms (r), the contractile vacuole (v) including ampullae (a). (B) An anti-PtSyb2 antibody also stains the whole contractile vacuole complex. Staining was observed on the contractile vacuole (v) and radial arms (r). Higher magnification (inset) shows irregular staining (arrowheads) along the radial arms. Scale bar¼10mm.
(C) Electron micrograph of a cell expressingPtSyb2-2-GFP detected with an anti-GFP antibody shows gold labelling along the radial canal (rc) and of the smooth spongiome (ss). Scale bar¼500 nm. (D) Electron micrograph of a non-transfectedP. tetraureliacell stained with the polyclonal anti-PtSyb2 antibody. Gold labelling occurs on the smooth spongiome and close to the radial canal, whereas the decorated spongiome (ds) is devoid of label. Scale bar¼500 nm.
complexes (68,69). It is interesting to note that P. tetraurelia contains one R-SNARE, PtSyb2-1, with a potentially tetanus neurotoxin-cleavable RD signature in its SNARE motif.
We found three PtSybs localized in the ER: PtSyb1-1, PtSyb1-2 and PtSyb3-1. We noticed subtle differences in the distribution over the ER of the two PtSyb1 isoforms in which PtSyb1-1 shows a reticulate cortical staining whereas PtSyb1-2 was more confined to the internal ER network. The antibody directed against isoform PtSyb1-1 probably does not discriminate between its two isoforms, which are 84% identical, and the antibody stains both, the cortical and internal ER network. The distinct distribution of the two isoforms may represent functional differentia- tion in different subsections of the ER (70,71). Because of the similar staining patterns obtained from the C-terminal GFP fusion constructs of PtSyb1-1 and PtSyb1-2 com- pared to the ER marker DiOC
6(3) (Figure 4A,C,E), a block in ER exit of the fusion proteins can be excluded.
Furthermore, co-localization was obtained for DiOC
6(3) and the relevant specific antibodies (Figure 4G – M), and we conclude that both isoforms of PtSyb1 and PtSyb3 reside in the ER network of Paramecium.
As predicted by the localization of PtNSF1 there (32), the contractile vacuole complex of Paramecium is likely to be
a place of intense membrane interactions and PtSyb2 is the candidate R-SNARE for membrane interactions in the smooth spongiome. As was demonstrated by homology- dependent gene silencing, PtSyb2 is required for the structural integrity and function of the organelle, but silencing Ptsyb2 also revealed a global morphological defect affecting swimming behaviour and cell division.
How PtSyb2 is affecting cell division and whether there is some kind of crosstalk between the osmoregulatory system and the cell division machinery remain a tantalizing question. Silencing of the Ptsyb2 gene subfamily demonstrated the specificity of the anti- PtSyb2 antibody for this subfamily, because staining disappeared completely after knockdown of Ptsyb2 (Figure 8C,E).
In our semiquantitative immunogold analysis, we referred pA-Au
5labelling events to unit area of the respective organelle analysed. The local enrichment determined was well compatible with our observations by fluores- cence labelling, but EM analysis allowed us to identify more clearly labelling of organelles, such as early endo- somes (with associated vesicles), and organellar substruc- tures, such as the smooth spongiome of the contractile vacuole complex, as well as absence of any label enrich- ment from the decorated spongiome.
A B
C D
E F
pPD-gfp
anti-PtSyb2 anti-PtV0aSU
pPD-syb2-1
pPD-syb2-2
Figure 8: Silencing ofPtsyb2genes by feeding.Cells silenced with pPD- gfp(A and B; negative control), pPD- syb2-1(C and D) and pPD-syb2-2(E and F) were fixed and stained with an anti-PtSyb2 polyclonal antibody (A, C, E) or, as a control, with an antibody recognizing theParamecium tetraure- lia V-ATPase V0-part a-subunit (V0aSU), which stains the contractile vacuole complex (B, D, F). (E, F) pPD- syb2-2-silenced cells of abnormal, rounded shape with severely disfig- ured radial arms (arrows) of the con- tractile vacuole complex (F).
Specificity of the anti-PtSyb2 antibody is demonstrated by the complete dis- appearance of staining in the contrac- tile vacuole complex after silencing of Ptsyb2-1 or Ptsyb2-2 (C and E).
Scale bars¼10mm.
The role of the PtSybs lacking any homologues outside the ciliates is still under investigation, and they may have a role in the elaborate phagosome-trafficking pathway of ciliates or other ciliate-specific trafficking pathways. We are currently also investigating the role of other PtSybs but so far have not been able to unequivocally identify an R-SNARE responsible for the fast triggered exocytosis of trichocysts. It is, however, possible that if a particular R-SNARE is silenced, another PtSyb (especially its closely related isoform) may be able to take over its function, as has been found with the functional redundancy of VAMP-2 and VAMP-3 in human cells (72). Thus, the results from double-silencing experiments will be important in providing an answer to the question whether the fast regulated exocytosis of trichocysts is SNARE-dependent or not.
Materials and Methods
Computational analysis
BLAST searches were performed at NCBI (37). Domain predictions were performed with PFAM (73), PDB (74) and at GenBank at NCBI. Potential transmembrane anchors were determined with TMpred (75). Molecular modelling was performed with 3DJIGSAW (76–78). Amino acid alignment was performed with ClustalW (79). Phylogenetic and molecular evolution- ary analyses were performed using MEGA version 3.0 (80).
Amplification of Ptsyb genes by PCR from genomic DNA and cDNA
Total wild-type DNA from strain 7S for PCR was prepared from log-phase cultures as described previously (81). The open reading frames (ORFs) of Ptsybs were amplified with gene-specific oligonucleotide primers (Supplemental Table 1) from aP. tetraureliad4-2 cDNA library (38) or freshly prepared cDNA using SuperScriptTMIII reverse transcriptase (Invitrogen,
A
C
E
D
F B
45 40 35 30 25 20
Time between contractions (s)
15 10 5 0
n=60
pPD-gfp pPD-syb2-1 pPD-syb2-2 n=37 n=24
Figure 9: Morphology of Ptsyb2gene-silenced Paramecium tetraureliacells.(A) Fixed control (pPD-gfpmock-silenced) cell. (B) Fixed and (C) livePtsyb2-2gene-silenced cell showing abnormal round cell shape. (D)Ptsyb2-2gene-silenced cell shows normal ability to discharge trichocysts. (E) Immunostaining of aPtsyb2-1-Ptsyb2-2 double-silenced cell, 40 h after beginning of silencing by feeding, stained with an anti-tubulin antibody and Hoechst 33342 DNA stain. Arrows indicate the positions of contractile vacuole complexes of which there are four per cell. Scale bars¼10mm. (F) Pumping activity of the contractile vacuole inPtsyb2-1 andPtsyb2-2-silenced cells, respectively, and a pPD-gfpcontrol with normal contraction period. A marked decrease in the pumping frequency of the contractile vacuole inPtsyb2-silenced cells is noticeable with a stronger effect for the silencing construct pPD-syb2-2or a double-silencing construct pPD-syb2-1-syb2-2compared to silencing with pPD-syb2-1alone. Sample sizes are indicated as N numbers below and bars represent standard deviations.
Karlsruhe, Germany) fromP. tetraurelia7S with 2mL of cDNA or cDNA library template under identical PCR conditions.
Cloning procedures
Ptsyb-specific PCR products were subcloned into pCR2.1-TOPO (Invitrogen) and sequenced from the M13 universal and reverse primer- binding sites of the plasmid (MWG, Ebersberg, Germany).Ptsyb-specific PCR products were purified using the QIAquick PCR Purification Kit (Qiagen, Hilden, Germany) and double-digested with the appropriate restriction enzymes in the buffer suggested by the manufacturer.
Double-digested macronuclear DNA or cDNA was then separated on TAE agarose gels, extracted using the QIAquick gel extraction kit (Qiagen) and ligated into the accordingly cut pPD (33) and pPXV-GFP (82) plasmids with T4 DNA ligase (NEB, Frankfurt, Germany).
Cell culture and test of exocytosis capacity
The wild-type strains ofP. tetraureliaused were strains7Sandd4-2. Cells were cultivated as described previously (83). For subcellular fractionation, axenic cultures were used (84). Capability of trichocyst exocytosis was routinely tested with a saturated solution of picric acid (85).
Microinjection of cells and microscopy
For microinjection of cells, pPXV-GFP plasmids were linearized withSfiI or NotI which both cut in between the T. thermophilainverted telomeric repeats of the plasmids, thus protecting the linearized product from degra- dation in cells (86). The DNA was isopropanol-precipitated and resus- pended to a concentration of 1–5mg/mL in MilliQ water. All cells used for microinjection were after induced autogamy. Cells were kept for 3–4 gen- erations after autogamy in salad medium supplemented with 0.8mg/mL of stigmasterol and fed withEnterobacter aerogenes. Before microinjection, the cells were treated with 0.01% aminoethyldextran (87) to stimulate trichocyst exocytosis, thus avoiding any further discharge that could dis- turb microinjection. Cells were then washed twice in Dryl’s bufferþ0.02% BSA (2 mM sodium citrate, 1 mM NaH2PO4, 1 mM
Na2HPO4and 1.5 mMCaCl2) (88). DNA microinjections were made with glass microcapillaries under an Axiovert 100TV phase-contrast microscope (Zeiss, Oberkochen), using a micromanipulator with a manually controlled air-pressure microinjector. The expression of the GFP fusion constructs in cells clonally derived from microinjected cells was followed after 24 and 48 h of growth inE. aerogenessuspension at 27C by epifluorescence micro- scopy in an Axiovert 100TV microscope (Zeiss) with a GFP filter (488 nm).
Fluorescence staining was analysed with a conventional LM Axiovert 100TV (Zeiss), or a confocal laser scanning microscope LSM 510 (Zeiss) equipped with a Plan-Apochromat63 oil immersion objective (NA 1.4).
Homology-dependent gene silencing by feeding
Targeted silencing ofPtsybgenes by feeding was performed as described (33). The full-length ORF ofPtsyb2-1was cloned withHindIII andXhoI into the pPD vector (33), and a 523-bp macronuclearXbaIDNA fragment of Ptsyb2-2was cloned into pPD. For the tandem silencing construct, the cDNA from subcloning of Ptsyb2-2 was excised from pCR2.1-Ptsyb2- 2(cDNA) withKpnI andXhoI, and the fragment was ligated intoKpnI and XhoI-digested pPD-syb2-1(cDNA), leaving a small fragment (45 bp) from the pCR2.1 cloning vector as spacer between the two genes. Production of dsRNA inE. colistrain HT115 was induced at an OD600¼0.4 with 0.5 mM
isopropyl-thio-galactoside.
Expression and purification of Paramecium synaptobrevin-specific peptides in E. coli
For heterologous expression ofPtSyb-specific peptides, we selected aa 1–194 of PtSyb1-1, aa 1–192 of PtSyb2-1 and aa 1–185 of PtSyb3-1 (excluding the hydrophobic transmembrane domains), an N-terminal frag- ment (aa 1–74) and a C-terminal fragment (aa 69–204) ofPtSyb6-1. After mutating all deviantParameciumglutamine codons (TAA and TAG) of the PtsybORF into universal glutamine codons (CAA and CAG) by fusion PCR methods with gene-specific mutation oligonucleotide primers (Supplemental Table 1, Supplementary Material), the resulting fragments were cloned into theNcoI/XhoI restriction sites of a pRV11a expression vector derived
from the pET System (Novagen, Bad Soden, Germany) which contains a C-terminal His6tag for purification of the recombinant polypeptides (89).
RecombinantPtSyb polypeptides were purified by affinity chromatography on Ni2þ-NTA agarose under denaturing conditions, as recommended by the manufacturer (Novagen). Native expression ofPtSybs to a higher level could not be obtained under the conditions tested. The recombinant pep- tides were eluted with a pH step gradient, pH 8.0–4.5 in 8Murea (in 100 mM NaH2PO4and 10 mMTris – HCl). The fractions containing the recombinant peptides were pooled, brought to neutral pH and used for immunization of rabbits and mice.
Antibodies
We generated rabbit polyclonal antibodies against PtSyb1-1 (aa 1–194), PtSyb2-1 (aa 1–192),PtSyb3-1 (aa 1–185) andPtSyb6-1 (aa 69–204) and mouse polyclonal antibodies against the N-terminus (aa 1–74) ofPtSyb6-1.
Antibodies against the recombinant PtSyb polypeptides were raised in rabbits or mice. After several boosts of the animals, positive sera were taken at day 60 and purified by two subsequent chromatography steps, a first step on a His-tag peptide column (24-amino acid peptide, to remove His-tag-specific antibodies), followed by an affinity step on the correspond- ingPtSyb polypeptide. A further step of purification with the non-cognate PtSyb polypeptides was carried out for the anti-PtSyb1-1, anti-PtSyb2-1 and anti-PtSyb3-1 antibodies in order to decrease the degree of cross- reactivity between different R-SNAREs. Mouse polyclonal antibodies against the amino terminus ofPtSyb6-1 were not affinity-purified.
The polyclonal rabbit antibody directed against the a1–1 subunit of the V0
part of the Paramecium V-ATPase (anti-PtV0aSU) stains the contractile vacuole complex (90).
Anti-a-tubulin was a mouse monoclonal antibody (clone DM1A; Sigma- Aldrich, Schnelldorf, Germany) and was used diluted 1:500 in immunostainings.
Immunofluorescence labelling and staining with dyes
Monoxenically grown cells were concentrated by centrifugation for 2 min at 900gand washed twice in 5 mMPipes buffer, pH 7.0, containing 1 mMKCl and 1 mMCaCl2. Cells were fixed in 2–8% (w/v) freshly prepared formaldehyde in PBS for 20 min at room temperature and permeabilized by addition of 0.5% digitonin (Sigma). Cells were washed twice in PBS, blocked for 10 min with 50 mMglycine in PBS and resuspended in 1%
BSA in PBS. Rabbit and mouse antibodies diluted 1:50 and 1:100 in PBS (þ1% BSA), respectively, were applied to cells overnight at 4C. After 45 min washes in PBS, AlexaFluor488and AlexaFluor568-coupled Fab20 fragments of goat anti-rabbit antibodies (Molecular Probes, Karlsruhe, Germany) and FITC-conjugated anti-mouse antibodies (Dianova, Hamburg, Germany) or AlexaFluor594coupled Fab20 fragments of goat anti-mouse antibodies (Molecular Probes) were applied for 60 min, fol- lowed by 45 min washes in PBS. Samples were shaken gently during all incubation and washing steps. After the final wash, samples were resuspended in PBSþ1% BSA. Staining of single cells was performed as described previously (91) with the difference that the prepermeabiliza- tion step with 1% Triton-X100 was omitted.
Staining with 3,30-dihexaoxacarbocyanine iodide [DiOC6(3); Sigma) was performed at 0.1mg/mL after fixation for 1 h at room temperature, and cells were washed 35 min in PBS. Staining of nuclei was performed with 10 nMHoechst 33342 (Molecular Probes) for 10 min in PBS/BSA or TBST*/BSA (0.15MNaCl in 10 mM Tris– HCl pH 7.4, 0.1% Tween-20, 2 mMMgCl2and 1 mMEGTA containing 3% BSA). Cells were mounted with Mowiol or mixed with CitiFluor antifade ragent (Citifluor, Canterbury, UK) to reduce fading and analysed with the appropriate filter sets as described above.
Cell fractionation
Cell surface complexes (‘cortices’) were isolated as described (92). Other cell fractions were prepared from axenic cell cultures as previously described (93). Protein concentrations were determined colorimetrically
with BSA as a standard. A protease-inhibitor cocktail containing 15mM pepstatin A, 100 mU/mL aprotinin, 100mM leupeptin, 28mM E64 and 0.2 mM Pefabloc SC (Roche, Mannheim, Germany) was used in all preparations.
Electrophoretic techniques and Western blot analysis
Protein samples were denatured by boiling for 3 min in sample buffer and subjected to electrophoresis on SDS polyacrylamide gels using the discon- tinuous buffer system of Laemmli (1970). Protein molecular weight stand- ards were used in accordance with the manufacturer’s directions. Protein blotting was performed at 1 mA/cm2for 1.5 h using a semidry blotter (Bio- Rad, Munich, Germany). Antibodies diluted 1:1000 in 5% (w/v) non-fat dry milk and Tris – buffered saline pH 7.5 were applied overnight at 4C.
Primary antibodies were detected by goat anti-rabbit IgG coupled to horse- radish peroxidase (ICN Pharmaceuticals, Eschwege, Germany) and detected with 5-bromo-4-chloro-3-indolyl phosphate and Nitro Blue tetra- zolium (Roche) or ECLTMWestern Blotting System (Amersham, Freiburg, Germany), respectively.
Immunogold labelling and EM analysis
With most samples, fixation was followed by embedding in LR Gold, as described previously (94), while occasionally a lower concentration of 2%
and 0.15% of the respective aldehydes was used in conjunction with embedding in LR White (London Resin, London, UK) according to the manufacturer’s instructions. Ultrathin sections collected on nickel grids were processed with 3% BSA in PBS, and samples were analysed in a Zeiss EM10 electron microscope. The anti-GFP antibody (48) was used at a dilution of 1:20. From here, samples were treated and analysed as described previously (94).
The postembedding immuno-EM method and quantitiative analysis has been performed precisely as described by Kissmehl et al. (94). The only exception was the use of protein A-gold conjugates of 5 nm diameter (pA- Au5) for detecting primary antibodies, i.e. IgGs from rabbits. The pA-Au5
samples were obtained from Medical School, Utrecht (NL). Also as described, prints of 80 000 magnification have been used to count gold granules and to determine the areas of cell sections to which pA-Au5
counts have to be referred. With each of the synaptobrevins analysed, our EM analysis focused on cell regions whose preferential labelling had been documented by immunofluorescence and/or after expression of GFP fusion proteins. (Note that for EM analysis, only wild-type non-transfected cells have been analysed.) The labelling density (gold granules per unit area) over ER-rich cell regions, the smooth spongiome and around the endocytotically active regions near ciliary basal bodies, respectively, have thus been determined, also as described (94). To express objectively the enrichment of label over regions of interest, we also deter- mined the labelling density over an invariable reference structure (tricho- cysts) to which the labelling densities (SEM) over the respective structures have been referred. The respective enrichment factors deter- mined have thus been determined from up to 12 cell sections for the respective structures.
Acknowledgments
We are grateful for technical support by J. Hentschel, L. Nejedli, K. Nu¨hse, C. Danzer, B. Kottwitz and O. Traub and thank R. Vo¨gele for the gift of the pRV11 expression vector and E. Ferrando-May and P. Grote for use of the LSM510 facilities (all University of Konstanz, Germany). We also express our gratitude to J. Beisson and F. Ruiz for their help and use of their equipment and J. Cohen (all CNRS, CGM, Gif-Sur-Yvette, France) for early access to theParameciumgenome. This work was supported by DFG grant PL78/20/2 to H. P and R.K.
Supplementary Material
Video 1: Movie file showing the phenotype and swimming behaviour of PtSyb2 gene-silenced cells. One mock-silenced cell was added for com- parison.PtSyb2-silenced cells at 48 h of silencing have a rounded shape and perform quick rotations around their own axis before they eventually die.
Supplemental Figure 1: Synaptobrevin family phylogeny based on only the SNARE domain. Neighbour joining tree with 1000 bootstrap replicates encompassing the SNARE domains ofParamecium tetraurelia synaptobrevin proteins (PtSybs; see Table 1).Tetrahymena thermophila (Tf) synaptobrevin homologues from preliminary gene predictions 08- 2004 at TIGR (gene identifiers 2415, 4410, 4411, 4722, 7285, 11170, 14913, 19728, 20867).Arabidopsis thaliana (At) VAMPs 711–714, 723–
725, 727 (GenBank accession numbers O49377, Q9SIQ9, Q9LFP1, Q9FMR5, Q8VY69, O23429, O48850, Q9M376). Homo sapiens (Hs) synaptobrevin 1 (Syb1), VAMP2, VAMP3, Ykt6 and synaptobrevin-like pro- tein 1(Sybl1) (GenBank accession numbers AAA60603, AAH02737, CAB63146, CAG46805, AAH56141), Saccharomyces cerevisiae (Sc) Snc1, Snc2, Nyv1, Ykt6 and Sec22 (GenBank accession numbers AAC05002, NP_014972, NP_013194, AAB32050, AAB67373).Dictyostelium discoideum(Dd) SybA, VAMP7A, VAMP7B (DDB0214903, DDB0231535, DDB0231542) and DDB0190688, DDB0219546, and DDB0219666 and Entamoeba histolytica (Eh) synaptobrevin-like proteins 1 and 2 (Sybl1 and Sybl2) (GenBank accession numbers AY256852 and AY309014).Parameciumsynaptobrevins addressed in this work are indi- cated with gray diamonds. Bootstrap support values are indicated above the branches and an evolutionary distance scale is given below the figure.
Supplemental Table 1:Oligonucleotide primers used for amplification and cloning ofParameciumtetraurelia synaptobrevin genes
These materials are available as part of the online article from http://
www.blackwell-synergy.com
References
1. Plattner H, Kissmehl R. Molecular aspects of membrane trafficking in Paramecium. Int Rev Cytol 2003;232:185–216.
2. Jahn R, Su¨dhof TC. Membrane fusion and exocytosis. Annu Rev Biochem 1999;68:863–911.
3. Chen YA, Scheller RH. SNARE-mediated membrane fusion. Nat Rev Mol Cell Biol 2001;2:98–106.
4. Hong W. SNAREs and traffic. Biochim Biophys Acta 2005;1744:120–144.
5. So¨llner T, Bennett MK, Whiteheart SW, Scheller RH, Rothman JE. A protein assembly-disassembly pathway in vitro that may correspond to sequential steps of synaptic vesicle docking, activation, and fusion.
Cell 1993;75:409–418.
6. Weber T, Zemelman BV, Mcnew JA, Westermann B, Gmachl M, Parlati F, So¨llner TH, Rothman JE. SNAREpins: minimal machinery for membrane fusion. Cell 1998;92:759–772.
7. Hu C, Ahmed M, Melia TJ, So¨llner TH, Mayer T, Rothman JE. Fusion of cells by flipped SNAREs. Science 2003;300:1745–1749.
8. So¨llner TH. Regulated exocytosis and SNARE function (Review). Mol Membr Biol 2003;20:209–220.
9. Jahn R, Lang T, Su¨dhof TC. Membrane fusion. Cell 2003;112:519–533.
10. Malhotra V, Orci L, Glick BS, Block MR, Rothman JE. Role of an N-ethylmaleimide-sensitive transport component in promoting fusion of transport vesicles with cisternae of the Golgi stack. Cell 1988;54:221–227.
11. Clary DO, Griff IC, Rothman JE. SNAPs, a family of NSF attachment proteins involved in intracellular membrane fusion in animals and yeast. Cell 1990;61:709–721.
12. Morgan A, Burgoyne RD. Is NSF a fusion protein? Trends Cell Biol 1995;5:335–339.
13. Fasshauer D, Sutton RB, Brunger AT, Jahn R. Conserved structural features of the synaptic fusion complex: SNARE proteins reclass- ified as Q- and R-SNAREs. Proc Natl Acad Sci USA 1998;95:
15781–15786.
14. Filippini F, Rossi V, Galli T, Budillon A, D’Urso M, D’Esposito M.
Longins: a new evolutionary conserved VAMP family sharing a novel SNARE domain. Trends Biochem Sci 2001;26:407–409.
15. Rossi V, Picco R, Vacca M, D’Esposito M, D’Urso M, Galli T, Filippini F.
VAMP subfamilies identified by specific R-SNARE motifs. Biol Cell 2004;96:251–256.
16. Dietrich LE, Boeddinghaus C, Lagrassa TJ, Ungermann C. Control of eukaryotic membrane fusion by N-terminal domains of SNARE pro- teins. Biochim Biophys Acta 2003;1641:111–119.
17. Humeau Y, Doussau F, Grant NJ, Poulain B. How botulinum and tetanus neurotoxins block neurotransmitter release. Biochimie 2000;82:427–446.
18. Plattner H, Lumpert CJ, Knoll G, Kissmehl R, Ho¨hne B, Momayezi M, Glas-Albrecht R. Stimulus-secretion coupling inParameciumcells. Eur J Cell Biol 1991;55:3–16.
19. Vayssie´ L, Skouri F, Sperling L, Cohen J. Molecular genetics of regu- lated secretion inParamecium. Biochimie 2000;82:269–288.
20. Allen RD, Schroeder CC, Fok AK. Endosomal system ofParamecium:
coated pits to early endosomes. J Cell Sci 1992;101:449–461.
21. Patterson DJ. Membranous sacs associated with cilia ofParamecium.
Cytobiologie 1978;17:107–113.
22. Allen RD. Cytology. In: Go¨rtz HD, editor. Paramecium. Berlin, Heidelberg, New York: Springer-Verlag; 1988, pp. 4–40.
23. Allen RD. Food vacuole membrane growth with microtubule-associated membrane transport inParamecium. J Cell Biol 1974;63:904–922.
24. Allen RD, Fok AK. Phagosome fusion vesicles ofParamecium. I. Thin- section morphology. Eur J Cell Biol 1983;29:150–158.
25. Allen RD, Fok AK. Phagosome fusion vesicles of Paramecium. II.
Freeze-fracture evidence for membrane replacement. Eur J Cell Biol 1983;29:159–165.
26. Allen RD, Fok AK. Membrane trafficking and processing in Paramecium. Int Rev Cytol 2000;198:277–318.
27. Momayezi M, Habermann AW, Sokolova JJ, Kissmehl R, Plattner H.
Ultrastructural and antigenic preservation of a delicate structure by cryopreparation: identification and immunogold localization during bio- genesis of a secretory component (membrane-matrix connection) in Parameciumtrichocysts. J Histochem Cytochem 1993;41: 1669–1677.
28. Gautier MC, Garreau de Loubresse N, Madeddu L, Sperling L.
Evidence for defects in membrane traffic inParameciumsecretory mutants unable to produce functional storage granules. J Cell Biol 1994;124:893–902.
29. Allen RD. The contractile vacuole and its membrane dynamics.
Bioessays 2000;22:1035–1042.
30. Allen RD, Naitoh Y. Osmoregulation and contractile vacuoles of proto- zoa. Int Rev Cytol 2002;215:351–394.
31. Froissard M, Kissmehl R, Dedieu JC, Gulik-Krzywicki T, Plattner H, Cohen J. N-ethylmaleimide-sensitive factor is required to organize functional exocytotic microdomains in Paramecium. Genetics 2002;161:643–650.
32. Kissmehl R, Froissard M, Plattner H, Momayezi M, Cohen J. NSF regulates membrane traffic along multiple pathways inParamecium.
J Cell Sci 2002;115:3935–3946.
33. Galvani A, Sperling L. RNA interference by feeding inParamecium.
Trends Genet 2002;18:11–12.
34. Ruiz F, Vayssie´ L, Klotz C, Sperling L, Madeddu L. Homology-dependent gene silencing inParamecium. Mol Biol Cell 1998;9:931–943.
35. Dessen P, Zagulski M, Gromadka R, Plattner H, Kissmehl R, Meyer E, Betermier M, Schultz JE, Linder JU, Pearlman RE, Kung C, Forney J,
Satir BH, Van Houten JL, Keller AM et al.Parameciumgenome survey:
a pilot project. Trends Genet 2001;17:306–308.
36. Sperling L, Dessen P, Zagulski M, Pearlman RE, Migdalski A, Gromadka R, Froissard M, Keller AM, Cohen J. Random sequencing ofParameciumsomatic DNA. Eukaryot Cell 2002;1:341–352.
37. Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W, Lipman DJ. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 1997;25:3389–3402.
38. Klumpp S, Hanke C, Donella-Deana A, Beyer A, Kellner R, Pinna LA, Schultz JE. A membrane-bound protein phosphatase type 2C from Paramecium tetraurelia. Purification, characterization, and cloning. J Biol Chem 1994;269:32774–32780.
39. Russell CB, Fraga D, Hinrichsen RD. Extremely short 20–33 nucleotide introns are the standard length in Paramecium tetraurelia. Nucleic Acids Res 1994;22:1221–1225.
40. Zagulski M, Nowak JK, Le Mouel A, Nowacki M, Migdalski A, Gromadka R, Noel B, Blanc I, Dessen P, Wincker P, Keller AM, Cohen J, Meyer E, Sperling L. High coding density on the largest Paramecium tetraurelia somatic chromosome. Curr Biol 2004;14:1397–1404.
41. Mcnew JA, Sogaard M, Lampen NM, Machida S, Ye RR, Lacomis L, Tempst P, Rothman JE, So¨llner TH. Ykt6p, a prenylated SNARE essen- tial for endoplasmic reticulum-Golgi transport. J Biol Chem 1997;272:17776–17783.
42. Fukasawa M, Varlamov O, Eng WS, Sollner TH, Rothman JE.
Localization and activity of the SNARE Ykt6 determined by its regula- tory domain and palmitoylation. Proc Natl Acad Sci USA 2004;101:4815–4820.
43. Hasegawa H, Yang Z, Oltedal L, Davanger S, Hay JC. Intramolecular protein-protein and protein-lipid interactions control the conformation and subcellular targeting of neuronal Ykt6. J Cell Sci 2004;117:4495–4508.
44. Uemura T, Ueda T, Ohniwa RL, Nakano A, Takeyasu K, Sato MH.
Systematic analysis of SNARE molecules in Arabidopsis: dissection of the post-Golgi network in plant cells. Cell Struct Funct 2004;29:49–65.
45. Terasaki M, Song J, Wong JR, Weiss MJ, Chen LB. Localization of endoplasmic reticulum in living and glutaraldehyde fixed cells with fluorescent dyes. Cell 1984;38:101–108.
46. Ramoino P, Diaspro A, Fato M, Beltrame F, Robello M. Changes in the endoplasmic reticulum structure ofParamecium primaureliain relation to different cellular physiological states. J Photochem Photobiol B 2000;54:35–42.
47. Allen RD, Wolf RW. The cytoproct ofParamecium caudatum: structure and function, microtubules, and fate of food vacuole membranes. J Cell Sci 1974;14:611–631.
48. Wassmer T, Froissard M, Plattner H, Kissmehl R, Cohen J. The vacuo- lar proton-ATPase plays a major role in several membrane-bounded organelles inParamecium. J Cell Sci 2005;118:2813–2825.
49. Fok AK, Aihara MS, Ishida M, Nolta KV, Steck TL, Allen RD. The pegs on the decorated tubules of the contractile vacuole complex of Parameciumare proton pumps. J Cell Sci 1995;108:3163–3170.
50. Fok AK, Yamauchi K, Ishihara A, Aihara MS, Ishida M, Allen RD. The vacuolar-ATPase ofParamecium multimicronucleatum: gene structure of the B subunit and the dynamics of the V-ATPase-rich osmoregula- tory membranes. J Eukaryot Microbiol 2002;49:185–196.
51. Lefort-Tran M, Aufderheide K, Pouphile M, Rossignol M, Beisson J.
Control of exocytotic processes: cytological and physiological studies of trichocyst mutants in Paramecium tetraurelia. J Cell Biol 1981;88:301–311.
52. Skouri F, Cohen J. Genetic approach to regulated exocytosis using functional complementation inParamecium: identification of the ND7 gene required for membrane fusion. Mol Biol Cell 1997;8:1063–1071.
