The SEC6 protein is required for contractile vacuole function in ...

Research Article

2885

The SEC6 protein is required for contractile vacuole function in Chlamydomonas reinhardtii Karin Komsic-Buchmann, Lisa Marie Stephan and Burkhard Becker* Botany, Cologne Biocenter, University of Cologne, 50674 Cologne, Germany *Author for correspondence ([email protected])

Journal of Cell Science

Accepted 14 February 2012 Journal of Cell Science 125, 2885–2895 ß 2012. Published by The Company of Biologists Ltd doi: 10.1242/jcs.099184

Summary Contractile vacuoles (CVs) are essential for osmoregulation in many protists. To investigate the mechanism of CV function in Chlamydomonas, we isolated novel osmoregulatory mutants. Four of the isolated mutant cell lines carried the same 33,641 base deletion, rendering the cell lines unable to grow under strong hypotonic conditions. One mutant cell line (Osmo75) was analyzed in detail. The CV morphology was variable in mutant cells, and most cells had multiple small CVs. In addition, one or two enlarged CVs or no visible CVs at all, were observed by light microscopy. These findings suggest that the mutant is impaired in homotypic vacuolar and exocytotic membrane fusion. Furthermore the mutants had long flagella. One of the affected genes is the only SEC6 homologue in Chlamydomonas (CreSEC6). The SEC6 protein is a component of the exocyst complex that is required for efficient exocytosis. Transformation of the Osmo75 mutant with a CreSEC6-GFP construct rescued the mutant completely (osmoregulation and flagellar length). Rescued strains overexpressed CreSEC6 (as a GFP-tagged protein) and displayed a modified CV activity. CVs were larger, whereas the CV contraction interval remained unchanged, leading to increased water efflux rates. Electron microscopy analysis of Osmo75 cells showed that the mutant is able to form the close contact zones between the plasma membrane and the CV membrane observed during late diastole and systole. These results indicate that CreSEC6 is essential for CV function and required for homotypic vesicle fusion during diastole and water expulsion during systole. In addition, CreSEC6 is not only necessary for CV function, but possibly influences the CV cycle in an indirect manner and flagellar length in Chlamydomonas. Key words: Contractile vacuole, Flagella, Osmoregulation, SEC6, Exocyst, Chlamydomonas

Introduction Contractile vacuoles (CVs) are osmoregulatory organelles found in many unicellular freshwater protists without cell walls and some sponges (Allen and Naitoh, 2002). CVs are membranebound cell compartments that periodically accumulate (diastole) and expel (systole) water out of the cell, allowing cells to survive under hypotonic conditions. Based on structure and behavior about six basic types of CV have been described (Patterson, 1980). Despite this structural diversity the basic functions seem to be conserved between different eukaryotes because the same proteins and cellular processes have been found in Amoeba, Dictyostelium, Paramecium, Trypanosoma and green algae [e.g. V-ATPase (Becker and Hickisch, 2005; Fok et al., 2002; Heuser et al., 1993; Montalvetti et al., 2004; Nishihara et al., 2008; Robinson et al., 1998; Wassmer et al., 2005), aquaporin (Montalvetti et al., 2004; Nishihara et al., 2008), vesicular transport (Becker and Hickisch, 2005; Buchmann and Becker, 2009; Bush et al., 1994; Harris et al., 2001; Kissmehl et al., 2007; Schilde et al., 2006; Stavrou and O’Halloran, 2006); see KomsicBuchmann and Becker for a summary of identified proteins and cellular processes (Komsic-Buchmann and Becker, 2012)]. There are many accounts of the osmoregulatory role of CVs (Allen, 2000; Allen and Naitoh, 2002), and it has been proposed that water enters the CV by osmosis. V-ATPase and/or V-PPase drive secondary active transport systems, allowing water to follow passively through aquaporins. However, no acidification of the CV (as expected for a proton-pump-mediated uptake

system) has ever been observed. Therefore, HCO32 has been postulated to be the anion species continuously eliminated from the cell through the CV (Robinson et al., 1998; Tominaga et al., 1998). This would be similar to the situation for water transport in animal epithelia (Hoffmann, 1986; Zeuthen, 1992), but experimental evidence for a role of HCO32 in CVs has never been presented. By contrast, experimental evidence points to the involvement of phosphate in CV function in Trypanosoma and Chlamydomonas (Rohloff et al., 2004; Ruiz et al., 2001) and K+ and Cl2 have been identified as the major osmolytes in the cytosol and CV in Paramecium (Stock et al., 2002). The structure and function of the CV in Chlamydomonas have been investigated in some detail (Luykx et al., 1997a; Luykx et al., 1997b; Robinson et al., 1998). At the end of diastole the contractile vacuole of Chlamydomonas is spherical, expels the liquid into the medium and the CV fragments into smaller vacuoles (systolic phase; Fig. 1C). During diastole these smaller vacuoles swell and fuse with each other to form again the spherical vacuole at the end of a cycle (Luykx et al., 1997b) (Fig. 1A). Several questions remain regarding the situation in Chlamydomonas and more generally. (1) Exocytotic pore-like structures were identified in ciliates (McKanna, 1973) but have been very difficult to demonstrate in many green algae (Buchmann and Becker, 2009; Luykx et al., 1997b). (2) How the liquid leaves the cell in these systems is not clear, but conspicuous intra-membrane particle arrays (up to 180 nm in diameter) have been observed in the plasma membrane overlying

Journal of Cell Science 125 (12)


2886

Fig. 1. The contractile vacuole of Chlamydomonas reinhardtii CC3395. (A–C) The ultrastructure of the CVs in CC3395. The two CVs are located close to the basal body (A). At the end of diastole (B) the CV membrane forms a contact zone with the plasma membrane marked by cytosolic electron dense material between the membranes (arrows). In the systolic phase the CV fragments into smaller vesicles (C). The contact zone persists apparently until the end of systole (arrow). The ‘bracelet’, a specialized plasma membrane region at the basis of the flagellum (F), is marked by an ellipse in A. M, mitochondrion; N, nucleus. (D) Frames from a light microscope time-lapse recording. Numbers indicate the time passed since the end of last diastole. The white arrow marks the CV, scale bar: 5 mm. (E) The growth of CC3395 in four different media (TAP/2, TAP, TAP-S and TAP-SS). The strain can grow in every medium tested. (F) The relationship between the CV period, the CV volume and the efflux of each CV to the cell surface (n545). The bigger the cell surface is, the longer the CV period, the higher the CV efflux and the larger the CV volume. (G) The mean values and the standard deviation of the data set in F, given as non-normalized and normalized to the cell surface. The CV period shows higher variation in the normalized data set, whereas the normalized data set for the CV volume and the water efflux from a cell shows less variation than the non-normalized data set. Numbers above the bars indicate the coefficient of variation for the different data sets.

the CV region (Weiss et al., 1977). These arrays apparently form only during systole and are often matched by a similar array in the CV membrane opposing the plasma membrane array (Weiss et al., 1977). Both array are connected by cytosolic electron dense material (Weiss et al., 1977) (Fig. 1A) and similar cytosolic electron dense material has also been detected in another green alga Mesostigma viride (Buchmann and Becker, 2009). (3) A role for cytoskeletal elements during the CV cycle could only be demonstrated in Dictyostelium (Taft et al., 2008), indicating that force generation during systole by cytoskeletal elements does not play any role in most systems. Changes in membrane structure have been implicated in water expulsion during systole in Paramecium (Allen and Naitoh, 2002), but whether this is a general mechanism remains to be seen. In addition, our knowledge of how the CV cycle is controlled and

adapted to the need of the cell is at best fragmentary. Calcium, protein kinases and cAMP have been implicated (Rohloff and Docampo, 2008), but in no system is the CV really understood. Chlamydomonas is a well-established protist model system (Grossman et al., 2003). The genome of Chlamydomonas has recently been sequenced (Merchant et al., 2007). Chlamydomonas can be transformed using several methods (Coll, 2006; Grossman et al., 2003). Silencing of genes using RNA interference (RNAi) has been successfully introduced in Chlamydomonas and is continuously improving (Schroda, 2006), and several proteins have been expressed as GFP-tagged constructs (Fuhrmann et al., 1999; Huang et al., 2007; Ruiz-Binder et al., 2002; Schoppmeier et al., 2005), making it possible to observe the in vivo dynamics of subcellular structures and/or proteins. For this reason we have started a forward genetic approach to analyze CV function

Contractile vacuole of Chlamydomonas in Chlamydomonas. Osmoregulatory mutants isolated after insertional mutagenesis showed defects in CV structure and function. We have analyzed a mutant in which membrane fusion events related to CV function are apparently impaired. We show that the deletion of the single Chlamydomonas SEC6 protein accounts for the observed phenotype, indicating a role for SEC6, and probably the exocyst complex, in CV function in Chlamydomonas. Results


Characterization of the contractile vacuole of Chlamydomonas reinhardtii CC3395

We used Chlamydomonas reinhardtii strain CC3395 for the mutant screen, which does not have a cell wall and is easily transformed. We first characterized the CV of this strain using light and electron microscopy (Fig. 1A–D). As in other Chlamydomonas strains, the large round CV visible at late diastole develops from small vacuoles (Fig. 1A) and forms close contact zones with the plasma membrane at the end of diastole (Fig. 1B), which apparently persist during systole (Fig. 1C). CC3395 grows in media of different osmotic strengths (Fig. 1E, see also Fig. 8A); note that TAP/2 contains only half of the mineral nutrients of the other media. Cells had two CVs at the anterior end in all media tested except TAP-SS (containing 120 mM sucrose, increasing the total osmotic strength of the medium to 204 mosM). In this medium less than 5% of the cells exhibited CVs that were visible with a light microscope. CVs are only visible with the light microscope in Chlamydomonas when it is in hypotonic medium; therefore, this result indicates that the cytosolic osmolarity of Chlamydomonas CC3395 is approximately 200 mosM. Preliminary data indicate that the cytosolic osmolarity varies with growth conditions and status of the cells (unpublished own observations), therefore only cells 4–6 days after subculturing (end of log phase, Fig. 1E) were used in our analysis (see Materials and Methods for details on cell culturing). In TAP medium the average maximum diameter of the large round vacuole at the end of the diastole was 1.7860.43 mm and the contraction interval 20.665.3 seconds (n545). The diastole lasted 19.465.0 seconds and the systole 1.360.5 seconds (supplementary material Movie 1). From these results it can be calculated that in TAP medium (64 mosM) a Chlamydomonas CC3395 cell expels approximately 11.968.75 mm3/minute (approximately 2% of the total cell volume per minute). Water uptake in a cell is directly proportional to the cell surface area; we therefore performed a linear correlation analysis between CV volume, CV period and water efflux and the cell surface area of a cell. As is evident from Fig. 1F, all three factors showed a good correlation to the cell surface area [r250.6568 (CV period–cell surface area), r250.7907 (CV volume–cell surface area), r250.7872 (efflux–cell surface area)]. Chlamydomonas cells considerably increase in size during the cell cycle. Because the cells were not synchronized in our cultures, we tried to normalize the CV data using the cell surface areas determined for each cell from the videos used to characterize the CVs. We then calculated the mean values and standard deviations for the normalized and non-normalized data set (Fig. 1G). As expected, the normalized data set showed less variation than the non-normalized data set for the CV volume and the water efflux from a cell (compare the coefficients of variation indicated above the bars in Fig. 1G). By contrast, the standard deviation obtained for the normalized data set was bigger for the contraction interval of the CV, when compared with the non-

2887

normalized data set (Fig. 1G). These results indicate that cells use mainly variation of the size of the CV to adapt to the increasing water influx during cell growth, whereas the contraction period is apparently regulated by a different factor. Mutant screen

To isolate osmoregulatory insertional mutants we used the mutant screen designed by Luykx et al. in combination with insertional mutagenesis using the hygromycin B resistance marker developed by Berthold et al. (see Materials and Methods for details) (Luykx et al., 1997a; Berthold et al., 2002). On TAP plates containing 0.06 M sucrose (TAP-S, 144 mosM) 2858 hygromycin-B-resistant clones were obtained, and these were screened for failure to grow in TAP medium. Seven mutant cell lines failed to grow at all in TAP medium. In addition, 68 cell lines showed a different growth phenotype (different growth rate, different color, etc.) than cells grown in TAP medium containing 0.06 M sucrose, but were still able to grow in TAP medium. Altogether we obtained a total of 75 potential osmoregulatory mutants (hereafter referred to as Osmo1–Osmo75). We concentrated our work on seven cell lines (Osmo12, 28, 32, 64, 66, 67, 75) showing a strong phenotype (no growth in TAP medium, growth in TAP-S medium). Osmo64, 66, 67, 75 carry the same insertion of the hygromycin B marker

Restriction enzyme site-directed amplification-PCR (RESDAPCR) (Gonza´lez-Ballester et al., 2005) was used to determine the locus of insertion of the hygromycin B resistance marker for Osmo12, 28, 32, 64, 66, 67 and 75. We obtained the 59 and 39 flanking sequences for Osmo64, 66, 67 and 75. All four isolated strains contained exactly the same 33,641 base deletion (Fig. 2A), indicating that the clones might have originated from the same insertion event (possibly by cell division after the insertion of the marker gene, during the recovery time after transformation). By contrast, we were only able to determine the 39 insert flanking sequences for the other three mutants showing a strong phenotype: Osmo28 (Fig. 2B), Osmo12 and 32 (Fig. 2C). Primer walking indicated that also in these strains large deletions (.9 kb) had occurred (Fig. 2B,C). Because the 59 insert flanking sequences are identical for Osmo12 and Osmo32, these two clones probably also originated from the same insertion event. At present the deletion size in Osmo12, 28 and 32 is not known, so all further work was carried out with Osmo75 as a representative of Osmo64, 65, 67 and 75. Characterization of Osmo75

We determined growth curves for Osmo75 in the same media used for characterization of the parental strain and characterized the mutant cell lines by video and electron microscopy (Fig. 3). Fig. 3A shows the growth curves for Osmo75 in the different media (also see Fig. 8A). The mutant was not able to grow in media of low osmolarity. Assuming that the mutant cell lines have a similar cytosolic osmolarity to that of the parental strain, Osmo75 cells are able to grow under mild hypotonic conditions (144 mosM) but fail to grow, or even die, under strong hypotonic conditions (#64 mosM). Video and electron microscopy was used to investigate whether the observed growth defect is related to CV malfunction. Video microscopy confirmed that indeed the CV


2888


Fig. 2. Insertion of the hygromycin B marker casette caused huge deletions. The corresponding areas of the genome of Chlamydomonas reinhardtii (http://www.phytozome.net) are shown. The deletion, determined by RESDA-PCR (see Material and Methods) for these mutants, is indicated by the grey bar. (A) In Osmo75 four genes are completely deleted (Cre20.g759800.t1.2–Cre20.g759950.t1.2) and two genes are truncated (Cre20.g759750.t1.2 and Cre20.g760000.t1.2) owing to the insertion of the marker cassette in the genome. (B,C) In Osmo28 and Osmo12 and 32 only one flanking sequence of the marker cassette could be determined. By primer walking a minimal deletion of 9 kb could be detected. Vertical arrows indicate the identified insertion site at the 39 end of the marker cassette. Horizontal arrows indicate the positions of primers that failed to amplify the 59-flanking region of the marker cassette.

cycle was aberrant in Osmo75 cells (supplementary material Movies 2–4). In all hypotonic media tested, all cells of the parental strain have two CVs following a typical alternating CV cycle with a large round vacuole at the end of the diastole (Fig. 1D, Fig. 3B). By contrast, all Osmo75 cells showed CV dysfunctions (changes in the number of CVs, the size and the contraction interval of a CV) in TAP-S medium (Fig. 3B–L). However, the CV phenotype was quite variable in a given Osmo75 cell population. Many of the cells (61.3%) had multiple smaller CVs in the region close to the basal bodies (Fig. 3D,E; supplementary material Movie 3). Surprisingly 23.7% of the cells did not have any CVs that were visible using light microscopy (Fig. 3F; supplementary material Movie 4). In 8.0% of the investigated cells one enlarged CV (Fig. 3G) was visible, and in 3.7% two enlarged CVs (Fig. 3C; supplementary material Movie 2) were visible. Finally, in 3.3% of the examined cells a mixed morphotype was detected: one enlarged CV and multiple smaller CVs (Fig. 3H). Electron microscopy confirmed the light microscopy observations (Fig. 3I,J). We could also detect the typical

contact zones formed by the CV membrane with the plasma membrane during systole, although water expulsion was clearly impaired in Osmo75 (see below; Fig. 4). Often contact zones in Osmo75 appeared to contain less electron dense cystosolic material between the plasma membrane and CV membrane than in CC3395 cells (compare Fig. 1B with Fig. 3K,L; see also Fig. 8H). To investigate the behavior of individual CVs in the Osmo75 strain in more detail we selected videos of cells with two or one enlarged CVs (Fig. 3C,G) and recorded the size of the CVs every 5 seconds in TAP-S medium (Fig. 4). For comparison, the size of individual CVs in the parental strain was recorded also. CVs in CC3395 showed an oscillating pattern. The diameter of a CV increases during diastole and rapidly decreases during systole; at the end of systole generally no CV is visible in the light microscope (Fig. 4). However, it is noteworthy that in TAP-S, CVs of CC3395 cells do not always completely empty (Fig. 4), whereas in TAP medium the CV of CC3395 always completely disappears (not shown). By contrast, CVs of Osmo75 cells show irregular increases and decreases of the CV diameter or appear

2889


Contractile vacuole of Chlamydomonas

Fig. 3. The contractile vacuoles of the osmoregulatory mutant Osmo75. (A) The growth of Osmo75 in four different media (TAP/2, TAP, TAP-S and TAP-SS). The strain grew in TAP-S and TAP-SS, but failed to grow, or even died, in TAP and TAP/2. (B–H) Cells of Osmo75 show variable CV morphologies. The graph (B) shows the proportion of the various mutant and parental cells with the different CV morphologies (100 cells of each type were analyzed in triplicate). (C–H) Examples for the different CV phenotypes. CVs are indicated by white asterisk. The arrowhead in F marks the cytoplasmic region normally displaying a CV. Scale bar: 5 mm. (I–L) Electron micrographs of the CVs of Osmo75. (I) Four CVs are visible in one cell (multiple CVs per cell). Two of them show contact zones with the plasma membrane (marked by arrows and are shown enlarged in K,L). (J) The mixed phenotype of Osmo75, one enlarged CV and multiple smaller CVs.

for some time constant (Fig. 4), but total discharges rarely occurred. Taken together the observed osmoregulatory phenotype indicates that in Osmo75 membrane fusion events during the CV cycle are impaired. The multiple small vacuoles might be caused by inefficient homotypic vacuolar fusion during diastole. The enlarged CVs are possibly caused by failure to terminate systole and achieve water expulsion. Finally we noted that Osmo75 cells had a distinct flagellar length phenotype. Flagella of Osmo75 cells were much longer (9.6361.55 mm) than those of the parental CC3395 strain (6.9761.05 mm; Fig. 5).

Protein targeting to the CV is not impaired in Osmo75

To test whether protein targeting to the CV is impaired in Osmo75 we tried to develop a GFP marker system for CV in Chlamydomonas. Aquaporins have been implicated in CV function in several other organisms (Montalvetti et al., 2004; Nishihara et al., 2008). The genome of Chlamydomonas reinhardtii encodes only two putative aquaporins (Anderberg et al., 2011) CreMIP1 (Cre12.g549300; www.phytozome.net) and CreMIP2 (Cre17.g711250). RT-PCR showed that CreMIP1 but not CreMIP2 is expressed in vegetative cells (data not shown). We reasoned that CreMIP1–GFP might be a useful marker to investigate whether protein targeting to the CV is

2890



Fig. 4. Osmo75 fails to expel liquid from the cells efficiently. The diameter of individual CVs was determined every 5 seconds and used to calculate the surface area of individual CVs in the parental strain CC3395 and Osmo75. Each line represents a different CV. Whereas CVs in the parental strain show a reiterating pattern, CVs of Osmo75 cells show a completely irregular behavior.

impaired in the Osmo75 cell line. In addition, expression of CreMIP1–GFP might confirm that, as in other systems, aquaporins are localized to the CV. The full-length cDNA of CreMIP1 was cloned into the GFP expression vector pJR38 (Neupert et al., 2009). Osmo75 and the UVM4 strain (which was specifically developed for GFP expression in Chlamydomonas) (Neupert et al., 2009) were transformed with a linearized ScaI–XbaI fragment of pJR38-MIP1-GFP, coding for CreMIP1– GFP and the APHVIII protein (paromomycin resistance). Paromomycin-resistant clones were selected and screened for GFP expression. Fig. 6 shows the results of this experiment. Nontransformed cells showed some background fluorescence in the GFP channel (Fig. 6A). However, as is evident from Fig. 6B–E, CreMIP1–GFP clearly localized to the CV in the UVM4 (Fig. 6B,C) and Osmo75 (Fig. 6D,E) strains, indicating that protein targeting of the CreMIP1-GFP construct to the CV is not impaired in Osmo75. In addition, using CreMIP1-GFP in the UVM4 genetic background we always observed that the CV membrane and plasma membrane apparently did not intermingle with each other during the CV cycle (Fig. 6C). Rescue of Osmo75

Based on the available genome sequence (www.phytozome.net), the 33,641 base deletion in Osmo75 affects six gene models. Four putative proteins are deleted and two additional putative proteins are truncated. Table 1 summarizes the available information on the six gene models. Gene model Au9.Cre20.g759900.t1 encodes the only putative SEC6 protein in Chlamydomonas. SEC6 proteins have been shown to be part of the exocyst complex

Fig. 5. Flagellar length of Chlamydomonas reinhardtii CC3395, the mutant Osmo75 and three rescued strains (Osmo75-SEC6GFP). Significant differences from CC3395 are indicated by asterisks (*P#0.05, ***P#0.001). Significant differences between the rescued strains and Osmo75 are indicated by hashes (###P#0.001); n537, 37, 29, 33 and 32, respectively, left to right.

(Bro¨cker et al., 2010). The exocyst complex belongs to the group of multi-subunit tethering factors required for efficient membrane fusion events and is involved in polarized secretion in many eukaryotic systems. Membrane fusion events occur during the diastole and at the beginning of systole during a CV cycle. Therefore the deletion of a protein similar to SEC6, possibly affecting exocyst function, seems a probable molecular cause for the observed phenotype of Osmo75. For this reason we concentrated our work on this protein, referred to as CreSEC6. RT-PCR revealed that CreSEC6 is expressed in the parental strain (Fig. 7, lanes 1–4) under all tested osmotic conditions, whereas we failed to amplify the same PCR fragment from the Osmo75 strain (Fig. 7, lane 5). Based on the Augustus gene model in Phytozome (www.phytozome.net) we expected the fulllength cDNA to be 2019 bases (672 aa) long excluding both UTRs. However, the isolated full-length cDNA was 2439 bases (812 aa) long. Comparison of the full-length cDNA with the transcript and protein sequence in Phytozome indicates that the predicted protein sequence in Phytozome misses one exon. This exon is also present in the published Arabidopsis sequence. Blast analysis of the full-length cDNA sequence showed that CreSEC6 is 31% identical (47% similar) to the SEC6 from Arabidopsis thaliana. To confirm that CreSEC6 is indeed responsible for the observed phenotype of Osmo75 we tried to rescue the mutant with a CreSEC6–GFP fusion protein again using pJR38 as the expression vector. On TAP plates 312 clones resistant to hygromycin B and paromomycin were obtained. Osmo75 cells do not grow in TAP medium, so this already indicated a successful rescue. For all clones investigated no defects in CV function could be observed by light microscopy of living cells, and the long flagellar phenotype was nearly completely rescued when analyzed in three randomly selected rescue cell lines (Fig. 5). To characterize the rescued strains in more detail, ten strains were randomly selected (Osmo75-A5, A9, C11, D10, E3, F6, F9, G6, H5 and H7). RT-PCR analysis showed that the CreSEC6GFP construct is expressed in all ten strains (Fig. 7, lanes 6–15) and at a higher level than the endogenous SEC6 in CC3395. Thus all rescue cell lines are SEC6 overexpressors. Although all ten rescue strains grew at all osmotic conditions tested, some strains (e.g. A5 and D10) did not grow as well on strong hypotonic media and isotonic media, possibly because of different insertion sites of the CreSEC6-GFP construct in the various strains. Detailed light microscope analyses of the CV cycle revealed interesting differences between the ten rescued Osmo75 strains investigated and the parental strain. Whereas the contraction interval was not significantly altered in all ten rescue strains; the ratio of CV:cell surface area and water efflux:cell surface area



2891

Fig. 6. Expression of CreMIP1-GFP in the UVM4 and Osmo75 background. (A) Untransformed UVM4 cell. (B,C) UVM4 cell transformed with CreMIP1-GFP. (D,E) Osmo75 cell transformed with CreMIP1-GFP. (B,D) Overview of the whole cell. (C,E) Time-lapse images of the CV region of the cells shown in B and D, respectively. Numbers indicate the time (in seconds). Exposure time for individual frames was 2.2 seconds. PH, phase contrast; Ex 460500, excitation wave length. Scale bars: 5 mm.

increased significantly (CV volume:cell surface area, P#0.001 for seven of the ten rescue strains; water efflux:cell surface area, P#0.001 for all ten rescue strains). Electron microscopy confirmed that the CV structure was completely restored (Fig. 8E–I). The ultrastructure of the CV in the rescued strains examined (G6, Fig. 8E–I; C11 not shown) was indistinguishable from that of the parental strain. Discussion The molecular mechanisms of CV function are still poorly understood. Over the last year several proteins have been implicated in CV function in several systems [for a recent summary, see Komsic-Buchmann and Becker (KomsicBuchmann and Becker, 2012)]. Generally, proton pumps, SNAREs, Rab proteins and calcium signaling have been shown to be important for CV function. The current models suggest that water uptake into the CV is by osmosis, energized by proton pumps, and that aquaporins facilitate this process. Although our knowledge about water uptake into the CV has greatly increased in recent years, the mechanism of water expulsion has not been so well studied in most systems. To increase our knowledge on CV function in green algae and in general we choose a forward genetic approach using Chlamydomonas as a model system and

investigated the cellular localization of a Chlamydomonas aquaporin. Early genomic analyses indicated only a single aquaporin in the genome of Chlamydomonas. A recent detailed analysis of algal MIPs indicated the presence of at least a second isoform (Anderberg et al., 2011). However, RT-PCR indicated that CreMIP2 is not expressed in C. reinhardtii, whereas CreMIP1 could be easily detected. Using a CreMIP-GFP construct we could clearly show that MIP1 is localized to the CV. In vivo observations of the CV indicated the CV membrane to be a stable compartment with no intermixing with the plasma membrane. These results suggest that similar to the CV in other systems, the membrane of the Chlamydomonas CV contains an aquaporin (Montalvetti et al., 2004; Nishihara et al., 2008). In addition, as in many other systems the CV membrane and the plasma membrane do not intermingle during the CV cycle (Patterson, 1981; Zanchi et al., 2010), suggesting that potential membrane fusion events follow the kiss-and-run mechanism. For the mutant screen we selected Chlamydomonas CC3395, which has no cell wall. Analyses of the CV cycle in CC3395 showed that, overall, the situation is very similar to strain 137c (average diameter at end of systole, contraction interval), indicating that the cell wall has only a minor effect on water

Table 1. (Putative) proteins of Chlamydomonas affected in the Osmo75 strain Name of model Cre20.g759750.t1 Cre20.g759800.t1 Cre20.g759850.t1 Cre20.g759900.t1 Cre20.g759950.t1 Cre20.g760000.t1

Annotation Phytoene dehydrogenase SpoU rRNA methylase family CGI-12 protein related Exocyst complex component Sec6 Protein of unknown function (DUF789) None

Deletion/Truncation 139 b truncation of the 39 end of the 39UTR Deletion Deletion Deletion Deletion 1089 base truncation of the 59 end

Expression + + +

2892



Fig. 7. Expression of SEC6 in various Chlamydomonas strains. RT-PCR was performed using the parental strain in four different media of different osmotic strengths, and Osmo75 (upper panel) and 10 rescue strains of Osmo75 (two lower panels). The genes targeted by the primers used are indicated at the top. Lane numbers refer to the different templates used and are explained on the right. The centrin gene was used as loading control. Expected length: SEC6, 126 bp; SEC6GFP, 222 bp; centrin, 94 bp.

uptake in Chlamydomonas. However, our results indicate that the cytosolic osmolarity of the cells is slightly higher in CC3395 than in 137c. Insertional mutants were generated and screened for CV dysfunction. Four of the obtained mutants (Osmo64, 65, 67 and 75) show the same 33,641 base deletion, indicating that the clones might have originated from the same insertion event (possibly by cell division after the insertion of the marker gene during the recovery time). Such large deletions are not uncommon in Chlamydomonas after transformation (Gonzalez-Ballester et al., 2011) and have been proposed to depend on the type of marker (large size, full plasmid) and the transformation method used (Gonzalez-Ballester et al., 2011). However, as we obtained large deletions using a small linear DNA fragment, most probably the transformation method is more important in this respect. The 33,641 base deletion in Osmo75 includes the only SEC6 protein, encoded in the Chlamydomonas genome. Characterization of the phenotype indicated that membrane fusion events during diastole (homotypic vacuolar fusion) and systole (exocytosis) do not operate efficiently in the CV in Osmo75, leading to hypotonic sensitivity of the cells. In addition, cells had long flagella. Rescue of the Osmo75 phenotype with a CreSEC6-GFP construct confirmed that indeed the deletion of CreSEC6 is responsible for the observed defect in CV function and flagellar length in Osmo75. The SEC6 protein is part of the exocyst complex, which belongs to the multi-subunit tethering factors (MTCs) (Bro¨cker et al., 2010). MTCs are ancient facilitators of membrane fusion events, and current knowledge indicates that every membrane fusion event requires its own tethering factor (Koumandou et al., 2007). The exocyst complex has been shown to be required for efficient exocytosis in various systems (Bro¨cker et al., 2010; Zhang et al., 2010). In this respect the observed phenotype in Osmo75 is surprising in two aspects. First, exocytosis is inefficient (leading to enlarged CVs) and homotypic vacuolar fusion (leading to many smaller CVs) does not take place efficiently. Work in the yeast system indicates that homotypic vacuolar fusion is mediated by the HOPS complex (Bro¨cker et al., 2010). The HOPS complex in yeast consists of six different subunits (Bro¨cker et al., 2010); two subunits have so far not been found in the Chlamydomonas genome (Koumandou et al., 2007). Given the observed phenotype in Osmo75 it is tempting to speculate that in Chlamydomonas SEC6 is also involved in HOPS complex-mediated homotypic vacuolar fusion. Second, the SEC6 deletion mutant in

Chlamydomonas is viable, whereas SEC6 is essential for growth in yeast (Potenza et al., 1992), and Arabidopsis T-DNA insertion lines that disrupt SEC6 expression fail to produce homozygous progenies (Ha´la et al., 2008). The latter is caused by defects in pollen germination and growth, indicating a major role in polar secretion in plants. However, in Chlamydomonas polar secretion seems not completely impaired, as the cells are able to form longer flagella. This is in striking contrast to the requirement of the exocyst in ciliogenesis in animals (Das and Guo 2011; Zuo et al., 2009). Exocyst localizes to the base of primary cilia in MDCK epithelial cells (Rogers et al., 2004) and deletion of SEC10 abolishes ciliogenesis (Zuo et al., 2009), whereas overexpression of SEC10 led to elongated primary cilia. By contrast, deletion of SEC6 in Chlamydomonas caused elongated flagella, whereas overexpression of SEC6–GFP did not change the flagellar length. The reason for the difference in behavior between these two systems is currently not clear. To our knowledge this is the first report of an involvement of SEC6 (and probably the exocyst complex) in CV function. Recently, Zanchi et al. reported that a secA mutant in Dictyostelium discoideum developed a large vacuole, which was shown to be derived from the CV (Zanchi et al., 2010). SecA is the Dictyostelium homologue of the yeast SEC1 and the mammalian Munc18 proteins (SM proteins), which are involved in vesicle docking during exocytosis and have been shown to interact with the exocyst complex, pointing to a role of the exocyst complex in CV function in Dictyostelium. However, in contrast to the SEC6 deletion in Osmo75, the SecA mutation in Dictyostelium leads only to an enlarged CV and not to a multiple CV phenotype, supporting the idea that the phenotype of Osmo75 indicates a dysfunction of two different cellular processes (homotypic vacuolar fusion and exocytosis). Interestingly, in a recent study Morgera et al. showed that SEC6 regulates exocytosis by interaction with SEC1 (Morgera et al., 2012). SEC1 [the yeast plasma membrane SM protein (plasma membrane sec1/Munc18-like proteins)] binds to the tSNARE SEC9, inhibiting the formation of the SNARE complex required for exocystosis. SEC6 releases SEC1 from SEC9, thus allowing exocytosis to proceed (Morgera et al., 2012). Given the function of the exocyst complex in other systems and these new findings, it seems plausible that exocyst in Chlamydomonas is required for the formation of the close contact zones between the plasma membrane and CV membrane and polar secretion in

2893



Fig. 8. Characterization of the rescued strains Osmo75-SEC6GFP. (A) The growth of CC3395, Osmo75 and ten randomly selected rescued strains Osmo75SEC6GFP (A5 to H7, as listed) on agar plates with different osmotic strengths ranging from strong hypotonic (32 mosM, TAP/2) to isotonic (204 mosM, TAP-SS). (B) Comparison of the CV period of CC3395 (n545) with the CV periods of the ten rescued strains Osmo75-SEC6GFP A5 to H7 (n520). Only E3 and F6 have significantly different CV periods from that of CC3395 (*P#0.05). (C) Comparison of the CV volume relative to the cell surface area of CC3395 (n545) and the rescued strains Osmo75-SEC6GFP A5 to H7 (n520). Only two rescued strains, C11 and D10, have similar CV volumes to CC3395, the others all differ significantly (*P#0.05, ***P#0.001). (D) The CV efflux relative to the CV surface area of CC3395 (n545) and the rescued strains Osmo75-SEC6GFP A5 to H7 (n520). The efflux of all rescued strain is significantly higher than the efflux of CC3395 (***P#0.001). (E–I) Electron micrographs of one rescued strain, Osmo75-SEC6GFP-G6. At the end of diastole the CV forms contact zones with the plasma membrane (E, arrows, and enlarged in H). Two CVs are visible in the cell shown in F, the left CV at mid diastole and the right CV in early diastole. The contact zones seem to persist until the end of the systolic phase (G arrows, and enlarged in I).

flagellar biogenesis. SEC6 might be required for water expulsion to proceed efficiently in the CV cycle by releasing a similar block as the SEC1 block observed in yeast. Materials and Methods Cell cultures

The following strains were used in this study: Chlamydomonas CC 3395 (arg7-8 cwd mt1) (Shimogawara et al., 1998) and UVM4 (cwd mt+ arg7) (Neupert et al.,

2009). Cells were cultured in TAP medium (Gorman and Levine, 1965), the medium for CC3395 and all derived mutant cell lines were supplemented with additional arginine. To achieve different osmolarities the medium was either diluted with aqua dest. (TKA X-CAD, Thermo Electron LED GmbH, Niederelbert, Germany) (for TAP/2), or 60 mM or 120 mM sucrose was added for TAP-S and TAP-SS, respectively. The osmolarity of all media was determined using a freezing point depression osmometer (Osmomat 010, Gonotec, Berlin, Germany). All transformants were always kept under selection pressure by addition of antibiotics to the medium and transferred into new media at least every 6 weeks.

2894


Cells were cultured at 21 ˚C with a photon flux of 70 mmol/m2 s and a 14 hour:10 hour light:dark cycle. In all experiments 5-day-old cultures (6 1 day) were used with a cell density of 106–107cells/ml. Transformation of Chlamydomonas cells

All transformations were performed using Kindle’s glass bead method (Kindle, 1990). For the insertional mutagenesis Chlamydomonas CC3395 cells were transformed with the HindIII cassette of pHyg3 (Berthold et al., 2002). Cells were allowed to recover for 2 hours in TAP followed by 16 hours in TAP-S in the dark and plated on TAP-S plates containing 10 mg/ml hygromycin B (Roth, Karlsruhe, Germany). After transformation of UVM4 and Osmo75 with the CreMIP1-GFP fusion construct cells were recovered in the appropriate media and plated onto plates containing paromomycin (Sigma, St. Louis, MO; 10 g/ml). Transformed cells were screened for GFP fluorescence using a fluorescence microscope (see below). Screening for osmoregulatory mutants

Individual clones were picked and transferred into 96-well plates containing, in each well, 200 ml TAP-S. After 2–3 weeks aliquots were transferred into new microtiter plates containing either TAP-S or TAP medium. Cell lines that showed a different growth in TAP compared with TAP-S were selected and the screening process was performed in triplicate. Determination of the insertion site


The insertion flanking regions were determined using the RESDA-PCR protocol of Gonza´les-Ballester et al. and specific primers for the HindIII fragment of pHyg3 developed by Matsuo et al. (Gonza´les-Ballester et al., 2005; Matsuo et al., 2008).

Nikon Eclipse 800 (Nikon GmbH, Du¨sseldorf, Germany) microscope equipped with a mercury short lamp (Osram, Du¨sseldorf, Germany), Uniblitz shutter control (Vincent Associates, Rochester, NY), a GFP filter set (480/40; 505; 535/50) and a Spot RT CCD digital camera (Diagnostic Instruments, Sterling Heights, MI). The images and videos were analyzed with Metamorph imaging software, version 6.3r4 (Universal Imaging, Corp., Bedford Hills, NY). Electron microscopy

Cells were concentrated by centrifuging at 500 g at 20 ˚C for 15 minutes and resuspended in high salt medium (HSM) with an appropriate amount of sucrose added to reach the respective osmotic strength, and additional HEPES (3 mM final concentration) before fixation simultaneously with glutaraldehyde and aqueous osmium tetroxide (final concentration 1.25% and 1%). The first minute of fixation was at room temperature and the additional 30 minutes on ice. After fixation the cells were washed once with fresh medium. To allow easier handling during the dehydration procedure, the cell pellets obtained by centrifugation were cross linked with BSA as follows. Cells were resuspended in BSA solution (30% in medium) and transferred into BEEM capsules (Plano, Marburg, Germany), pelleted (500 g, at room temperature for 15 minutes) and overlaid with glutaraldehyde solution (2.5% in medium). Samples were incubated for 30 minutes on ice before removal of the pellets. The cell pellets were incubated overnight in a 1% aqueous uranyl acetate solution at 4 ˚C. Samples were washed and dehydrated in an ethanol series and embedded in Epon 812. Ultrathin sections (60 nm) were cut with a Leika microtome EM UC7 and a diamante knife (diatome, 45 ˚ angle). Sections were stained with 2% aqueous uranyl acetate and lead citrate (Reynolds, 1963). Micrographs were taken with a transmission electron microscope (CM 10, Phillips, Eindhoven, The Netherlands) and a digital camera (Orius SC200W 1; Gatan, Pleasanton, CA). Images were analyzed with Digital Micrograph and Adobe Photoshop CS4.

GFP fusion constructs

Total RNA was isolated using TRI REAGENT (MRC, Cincinnati, OH) following the manufacturer’s instructions. cDNA was synthesized with the Revert Aid First Strand cDNA Synthesis Kit (Fermentas, Burlington, Canada). In all PCR reactions DreamTaq (Fermentas) was used in combination with an enhancer for GC-rich templates (Ralser et al., 2006). The complete cDNA of CreSEC6 (Cre20.g759900) and CreMIP1 (Cre12.g549300) was amplified and cloned into pGEM-T-easy (Promega, Madison, WI). Subsequently, NdeI restriction sites were added to the coding sequences by PCR, and ligated into the NdeI restriction site of pJR38 (Neupert et al., 2009) resulting in CreSEC6-GFP and CreMIP1-GFP fusion constructs (primer, sequences and vector maps are presented) (supplementary material Table S1; Figs S1, S2, S3).

Acknowledgements

The authors thank R. Bock (Golm, Germany) for providing plasmid pJR38 and the UVM4 strain, and W. Mages (Regensburg, Germany) for the plasmid pHyg3 and M. Schroda (Golm, Germany), K.-F. Lechtreck (Athens, GA, USA) and J. Brown (Worcester, MA, USA) for helpful discussions. In addition, we thank the following students: D. Langenbach, R. M. Benstein, A.-K. Alteko¨ster and K. Kehl, who helped to characterize the Osmo75 mutant. Funding

Mutant rescue

For rescue of the mutant phenotype, Osmo75 cells were transformed with CreSEC6-GFP. After transformation cells were allowed to recover in TAP medium before plating the cells on solid TAP medium without additional sucrose. RT-PCR

Total RNA was isolated using the peqGOLD Plant RNA kit (Peqlab, Erlangen, Germany). cDNA synthesis was performed with the Revert Aid First Strand cDNA Synthesis kit (Fermentas) using 1 mg total RNA. In the PCR reactions DreamTaq (Fermentas) was used in combination with an enhancer for GC-rich templates (Ralser et al., 2006). The primers were designed using quantprime (http://www. biomedcentral.com/1471-2105/9/465) and were specific for cDNA (primers are listed in supplementary material Table S1). Growth in different media

Cells for the determination of growth curves where cultured under reduced light (20–30 mMol/m2/second). Two biological replicates were counted four times each using a Neubauer hematocytometer. Growth on media of different osmotic strengths was also analyzed using a plate assay. The number of cells in a culture was counted using a Neubauer hematocytometer. Cells were then diluted with TAP medium to a concentration of 0.376106 cells/ml. The diluted cell suspension (3 ml) was dropped onto agar plates with different osmotic strength (without antibiotics), in triplicate. Cells were grown for 21 days. Light microscopy

Light microscopy was performed as described by Buchmann and Becker, except that we used 7 ml cell suspension on each slide (Buchmann and Becker, 2009). For video microscopy images were taken every 0.5 seconds. At least three cycles were analyzed per CV. The surface and volume of the CVs and cells were calculated as a prolate spheroid. The efflux rate per cell was calculated using the size and period of the cell investigated. To measure the flagella length, cells were fixed with 5% Lugol’s iodine. Linear regression analysis and significance tests (Student’s t-test with Welch correction) were done using the GraphPad Prism 5 software (GraphPad Software Inc., La Jolla, CA). Fluorescence microscopy was performed using a

This work was supported by the Deutsche Forschungsgemeinschaft [grant number Be1779/12-1 to B.B.]. Note added in proof

After acceptance of this paper a new release of the Chlamydomonas genome became available (v5.3, 8 June 2012). In the new release the former scaffold 20 has been mapped to chromosome 17 and therefore all gene IDs have been changed as follow: Cre20.g759750.t1 Phytoene dehydrogenase; new gene ID: g18016.t1; Cre20.g759800.t1 SpoU rRNA methylase family, new gene ID: g1801t.t1; Cre20.g759850.t1 CGI-12 protein related, new gene ID: g18014.t1; Cre20.g759900.t1 Exocyst complex component Sec6, new gene ID: g18013.t1; Cre20.g759950.t1 Protein of unknown function (DUF789), new gene ID: g18012.t1, and Cre20.g760000.t1 No annotation, new gene ID: g18011.t1. Supplementary material available online at http://jcs.biologists.org/lookup/suppl/doi:10.1242/jcs.099184/-/DC1

References Allen, R. D. (2000). The contractile vacuole and its membrane dynamics. Bioessays 22, 1035-1042. Allen, R. D. and Naitoh, Y. (2002). Osmoregulation and contractile vacuoles of protozoa. Int. Rev. Cytol. 215, 351-394. ˚ . and Johanson, U. (2011). Algal MIPs, high Anderberg, H. I., Danielson, J. A diversity and conserved motifs. BMC Evol. Biol. 11, 110. Becker, B. and Hickisch, A. (2005). Inhibition of contractile vacuole function by brefeldin A. Plant Cell Physiol. 46, 201-212. Berthold, P., Schmitt, R. and Mages, W. (2002). An engineered Streptomyces hygroscopicus aph 70 gene mediates dominant resistance against hygromycin B in Chlamydomonas reinhardtii. Protist 153, 401-412. Bro¨cker, C., Engelbrecht-Vandre´, S. and Ungermann, C. (2010). Multisubunit tethering complexes and their role in membrane fusion. Curr. Biol. 20, R943-R952.


Contractile vacuole of Chlamydomonas Buchmann, K. and Becker, B. (2009). The system of contractile vacuoles in the green alga Mesostigma viride (Streptophyta). Protist 160, 427-443. Bush, J., Nolta, K., Rodriguez-Paris, J., Kaufmann, N., O’Halloran, T., Ruscetti, T., Temesvari, L., Steck, T. and Cardelli, J. (1994). A Rab4-like GTPase in Dictyostelium discoideum colocalizes with V-H(+)-ATPases in reticular membranes of the contractile vacuole complex and in lysosomes. J. Cell Sci. 107, 2801-2812. Coll, J. M. (2006). Methodologies for transferring DNA into eukaryotic microalgae. Span. J. Agric. Res. 4, 316-330. Das, A. and Guo, W. (2011). Rabs and the exocyst in ciliogenesis, tubulogenesis and beyond. Trends Cell Biol. 21, 383-386. Fok, A. K., Yamauchi, K., Ishihara, A., Aihara, M. S., Ishida, M. and Allen, R. D. (2002). The vacuolar-atpase of Paramecium multimicronucleatum: gene structure of the B subunit and the dynamics of the V-ATPase-rich osmoregulatory membranes. J. Eukaryot. Microbiol. 49, 185-196. Fuhrmann, M., Oertel, W. and Hegemann, P. (1999). A synthetic gene coding for the green fluorescent protein (GFP) is a versatile reporter in Chlamydomonas reinhardtii. Plant J. 19, 353-361. Gonza´lez-Ballester, D., de Montaigu, A., Galvań, A. and Fernańdez, E. (2005). Restriction enzyme site-directed amplification PCR: a tool to identify regions flanking a marker DNA. Anal. Biochem. 340, 330-335. Gonzalez-Ballester, D., Pootakham, W., Mus, F., Yang, W. Q., Catalanotti, C., Magneschi, L., de Montaigu, A., Higuera, J. J., Prior, M., Galvań, A. et al. (2011). Reverse genetics in Chlamydomonas: a platform for isolating insertional mutants. Plant Methods 7, 24. Gorman, D. S. and Levine, R. P. (1965). Cytochrome f and plastocyanin: their sequence in the photosynthetic electron transport chain of Chlamydomonas reinhardi. Proc. Natl. Acad. Sci. USA 54, 1665-1669. Grossman, A. R., Harris, E. E., Hauser, C., Lefebvre, P. A., Martinez, D., Rokhsar, D., Shrager, J., Silflow, C. D., Stern, D., Vallon, O. et al. (2003). Chlamydomonas reinhardtii at the crossroads of genomics. Eukaryot. Cell 2, 1137-1150. Ha´la, M., Cole, R., Synek, L., Drdova´, E., Pecenkova´, T., Nordheim, A., Lamkemeyer, T., Madlung, J., Hochholdinger, F., Fowler, J. E. et al. (2008). An exocyst complex functions in plant cell growth in Arabidopsis and tobacco. Plant Cell 20, 1330-1345. Harris, E., Yoshida, K., Cardelli, J. and Bush, J. (2001). Rab11-like GTPase associates with and regulates the structure and function of the contractile vacuole system in Dictyostelium. J. Cell Sci. 114, 3035-3045. Heuser, J., Zhu, Q. and Clarke, M. (1993). Proton pumps populate the contractile vacuoles of Dictyostelium amoebae. J. Cell Biol. 121, 1311-1327. Hoffmann, E. K. (1986). Anion transport systems in the plasma membrane of vertebrate cells. Biochim. Biophys. Acta 864, 1-31. Huang, K., Diener, D. R., Mitchell, A., Pazour, G. J., Witman, G. B. and Rosenbaum, J. L. (2007). Function and dynamics of PKD2 in Chlamydomonas reinhardtii flagella. J. Cell Biol. 179, 501-514. Kindle, K. L. (1990). High-frequency nuclear transformation of Chlamydomonas reinhardtii. Proc. Natl. Acad. Sci. USA 87, 1228-1232. Kissmehl, R., Schilde, C., Wassmer, T., Danzer, C., Nuehse, K., Lutter, K. and Plattner, H. (2007). Molecular identification of 26 syntaxin genes and their assignment to the different trafficking pathways in Paramecium. Traffic 8, 523-542. Komsic-Buchmann, K. and Becker, B. (2012). Contractile vacuoles in green algae structure and function. In Advances in Algal Cell Biology (ed. K. Heimann and C. Katsaros). Berlin: Walter de Gruyter. Koumandou, V. L., Dacks, J. B., Coulson, R. M. R. and Field, M. C. (2007). Control systems for membrane fusion in the ancestral eukaryote; evolution of tethering complexes and SM proteins. BMC Evol. Biol. 7, 29. Luykx, P., Hoppenrath, M. and Robinson, D. G. (1997a). Osmoregulatory mutants that affect the function of the contractile vacuole in Chlamydomonas reinhardtii. Protoplasma 200, 99-111. Luykx, P., Hoppenrath, M. and Robinson, D. G. (1997b). Structure and behavior of contractile vacuoles in Chlamydomonas reinhardtii. Protoplasma 198, 73-84. Matsuo, T., Okamoto, K., Onai, K., Niwa, Y., Shimogawara, K. and Ishiura, M. (2008). A systematic forward genetic analysis identified components of the Chlamydomonas circadian system. Genes Dev. 22, 918-930. McKanna, J. A. (1973). Fine structure of the contractile vacuole pore in Paramecium. J. Protozool. 20, 631-638. Merchant, S. S., Prochnik, S. E., Vallon, O., Harris, E. H., Karpowicz, S. J., Witman, G. B., Terry, A., Salamov, A., Fritz-Laylin, L. K., Marećhal-Drouard, L. et al. (2007). The Chlamydomonas genome reveals the evolution of key animal and plant functions. Science 318, 245-250. Montalvetti, A., Rohloff, P. and Docampo, R. (2004). A functional aquaporin colocalizes with the vacuolar proton pyrophosphatase to acidocalcisomes and the contractile vacuole complex of Trypanosoma cruzi. J. Biol. Chem. 279, 38673-38682. Morgera, F., Sallah, M. R., Dubuke, M. L., Gandhi, P., Brewer, D. N., Carr, C. M. and Munson, M. (2012). Regulation of exocytosis by the exocyst subunit Sec6 and the SM protein Sec1. Mol. Biol. Cell 23, 337-346.

2895

Neupert, J., Karcher, D. and Bock, R. (2009). Generation of Chlamydomonas strains that efficiently express nuclear transgenes. Plant J. 57, 1140-1150. Nishihara, E., Yokota, E., Tazaki, A., Orii, H., Katsuhara, M., Kataoka, K., Igarashi, H., Moriyama, Y., Shimmen, T. and Sonobe, S. (2008). Presence of aquaporin and V-ATPase on the contractile vacuole of Amoeba proteus. Biol. Cell 100, 179-188. Patterson, D. J. (1980). Contractile vacuoles and associated structures: their organization and function. Biol. Rev. Camb. Philos. Soc. 55, 1-46. Patterson, D. J. (1981). Contractile vacuole complex behavior as a diagnostic character for free-living amebas. Protistologica (Paris) 17, 243-248. Potenza, M., Bowser, R., Mu¨ller, H. and Novick, P. (1992). SEC6 encodes an 85 kDa soluble protein required for exocytosis in yeast. Yeast 8, 549-558. Ralser, M., Querfurth, R., Warnatz, H. J., Lehrach, H., Yaspo, M. L. and Krobitsch, S. (2006). An efficient and economic enhancer mix for PCR. Biochem. Biophys. Res. Commun. 347, 747-751. Reynolds, E. S. (1963). The use of lead citrate at high pH as an electron-opaque stain in electron microscopy. J. Cell Biol. 17, 208-212. Robinson, D. G., Hoppenrath, M., Oberbeck, K., Luyks, P. and Ratajczak, R. (1998). Localization of pyrophosphatase and V-ATPase in Chlamydomonas reinhardtii. Bot. Acta 111, 108-122. Rogers, K. K., Wilson, P. D., Snyder, R. W., Zhang, X., Guo, W., Burrow, C. R. and Lipschutz, J. H. (2004). The exocyst localizes to the primary cilium in MDCK cells. Biochem. Biophys. Res. Commun. 319, 138-143. Rohloff, P. and Docampo, R. (2008). A contractile vacuole complex is involved in osmoregulation in Trypanosoma cruzi. Exp. Parasitol. 118, 17-24. Rohloff, P., Montalvetti, A. and Docampo, R. (2004). Acidocalcisomes and the contractile vacuole complex are involved in osmoregulation in Trypanosoma cruzi. J. Biol. Chem. 279, 52270-52281. Ruiz, F. A., Marchesini, N., Seufferheld, M., Govindjee and Docampo, R. (2001). The polyphosphate bodies of Chlamydomonas reinhardtii possess a proton-pumping pyrophosphatase and are similar to acidocalcisomes. J. Biol. Chem. 276, 4619646203. Ruiz-Binder, N. E., Geimer, S. and Melkonian, M. (2002). In vivo localization of centrin in the green alga Chlamydomonas reinhardtii. Cell Motil. Cytoskeleton 52, 4355. Schilde, C., Wassmer, T., Mansfeld, J., Plattner, H. and Kissmehl, R. (2006). A multigene family encoding R-SNAREs in the ciliate Paramecium tetraurelia. Traffic 7, 440-455. Schoppmeier, J., Mages, W. and Lechtreck, K. F. (2005). GFP as a tool for the analysis of proteins in the flagellar basal apparatus of Chlamydomonas. Cell Motil. Cytoskeleton 61, 189-200. Schroda, M. (2006). RNA silencing in Chlamydomonas: mechanisms and tools. Curr. Genet. 49, 69-84. Shimogawara, K., Fujiwara, S., Grossman, A. and Usuda, H. (1998). High-efficiency transformation of Chlamydomonas reinhardtii by electroporation. Genetics 148, 18211828. Stavrou, I. and O’Halloran, T. J. (2006). The monomeric clathrin assembly protein, AP180, regulates contractile vacuole size in Dictyostelium discoideum. Mol. Biol. Cell 17, 5381-5389. Stock, C., Grønlien, H. K., Allen, R. D. and Naitoh, Y. (2002). Osmoregulation in Paramecium: in situ ion gradients permit water to cascade through the cytosol to the contractile vacuole. J. Cell Sci. 115, 2339-2348. Taft, M. H., Hartmann, F. K., Rump, A., Keller, H., Chizhov, I., Manstein, D. J. and Tsiavaliaris, G. (2008). Dictyostelium myosin-5b is a conditional processive motor. J. Biol. Chem. 283, 26902-26910. Tominaga, T., Allen, R. D. and Naitoh, Y. (1998). Electrophysiology of the in situ contractile vacuole complex of Paramecium reveals its membrane dynamics and electrogenic site during osmoregulatory activity. J. Exp. Biol. 201, 451-460. Wassmer, T., Froissard, M., Plattner, H., Kissmehl, R. and Cohen, J. (2005). The vacuolar proton-ATPase plays a major role in several membrane-bounded organelles in Paramecium. J. Cell Sci. 118, 2813-2825. Weiss, R. L., Goodenough, D. A. and Goodenough, U. W. (1977). Membrane particle arrays associated with the basal body and with contractile vacuole secretion in Chlamydomonas. J. Cell Biol. 72, 133-143. Zanchi, R., Howard, G., Bretscher, M. S. and Kay, R. R. (2010). The exocytic gene secA is required for Dictyostelium cell motility and osmoregulation. J. Cell Sci. 123, 3226-3234. Zeuthen, T. (1992). From contractile vacuole to leaky epithelia. Coupling between salt and water fluxes in biological membranes. Biochim. Biophys. Acta 1113, 229-258. Zhang, Y., Liu, C. M., Emons, A. M. C. and Ketelaar, T. (2010). The plant exocyst. J. Integr. Plant Biol. 52, 138-146. Zuo, X., Guo, W. and Lipschutz, J. H. (2009). The exocyst protein Sec10 is necessary for primary ciliogenesis and cystogenesis in vitro. Mol. Biol. Cell 20, 2522-2529.