jf. Sci. 31, 213-224 (1978) 213 Printed in Great Britain © Company of Biologists Limited igyS

OSMOREGULATION IN THE ALGA VACUOLARIA VIRESCENS. STRUCTURE OF THE CONTRACTILE AND THE NATURE OF ITS ASSOCIATION WITH THE GOLGI APPARATUS

PETER HEYWOOD Division of and Medicine, Brown University, Providence, Rhode Island 02912, U.S.A.

SUMMARY The of the chloromonadophycean alga Vacuolaria virescens is a permanent structure that possesses a specialized membrane: subunits of this membrane have a diameter of 21—24 nm and in places are arranged in a regular hexagonal pattern. The lateral walls of these subunits form regularly spaced bristles or pegs which extend inwards from the trilaminar mem- brane for a distance of 13-15 nm. The contractile vacuole is situated immediately above an extensive Golgi apparatus that covers most of the anterior surface of the nucleus. Vesicles of Golgi origin give rise to subsidiary which in turn empty into the contractile vacuole. Golgi vesicles, subsidiary vacuoles and the contractile vacuole contain similar electron-dense material. It is suggested that this material might be a highly hydrophilic substance which will attract from the into the Golgi vesicles, subsidiary vacuoles and contractile vacuole from whence it is discharged from the cell. This method of is compared to that occurring in other and .

INTRODUCTION Many algae and protozoa possess contractile vacuoles (sometimes termed 'water expulsion vesicles'). These osmoregulatory expel a hypotonic from the cell and so compensate for the influx of water from the surrounding medium due to . Although contractile vacuoles are present in some marine (Dodge, 1973; Lloyd, 1928), they are usually absent from organisms inhabiting marine or brackish water. For example, in the algal class Chloromonadophyceae contractile vacuoles are present in the freshwater (Fott, 1971; Mignot, 1967) but are absent from the closely related marine representatives (Mignot, 1976; Toriumi & Takano, 1973). A similar situation has been reported in laboratory cultures of moeivusii (Guillard, i960). A mutant of this behaved like a brackish water alga in that it required the of the culture medium to be higher than 1-5 atm. (1-013 xio2 kN m~2). Significantly, culture media of this osmotic pressure caused the contractile vacuole of the wild strain to cease functioning which led Guillard to the conclusion that ' the sole essential function of contractile vacuoles in C. moewusii is elimination of water'. The physiological aspects of contrac- tile vacuoles have received a good deal of attention and there is an extensive literature 214 P- Heywood dealing with the effects on contractile vacuole function of such factors as the osmotic pressure, pH, and of the external medium (Ahmad & Couillard, 1974; Czarska, 1964; Kitching, 1967; Organ, Bovee & Jahn, 1968; Rifkin, 1973). It has also been demonstrated that adenosine triphosphate caused isolated vacuoles of proteus to contract (Prusch & Dunham, 1970) and accelerated the contraction cycle in multimicronucleatum (Organ et al. 1968). The focus of this investigation is the freshwater chloromonadophycean alga Vacuolaria virescens Cienkowsky. The earliest account of this organism (Cienkowsky, 1870) described the contractile vacuole, a conspicuous feature of the cell which is presumably the basis for its generic name. Subsequently the contractile vacuole has been examined with the light microscope (Dangeard, 1939; Poisson & Hollande, 1943; Spencer, 1971; Tschermak-Woess, 1954) and with the electron microscope (Mignot, 1967; Schnepf & Koch, 19666). Schnepf & Koch (19666) described the contractile vacuole of Vacuolaria virescens as a temporary structure which was formed by fusion of Golgi-produced vesicles. They postulated that after discharging its contents its membrane became part of the plasmalemma. The present report does not support this idea: the evidence indicates that the contractile vacuole is a permanent structure which possesses a specialized membrane similar to that occurring in some other contractile vacuoles. The association of the contractile vacuole and the Golgi apparatus is described and it is suggested that the function of the Golgi may be to produce a hydrophilic substance which will attract water into the Golgi vesicles, subsidiary vacuoles and contractile vacuole. In this manner water which has entered the cell by osmosis will become sequestered into the contractile vacuole and can then be discharged from the cell.

MATERIALS AND METHODS Vacuolaria virescens Cienkowsky was obtained from the Cambridge Culture Collection (number LB 1195/1). It was grown on defined medium (Heywood, 1973) at 23 ± 1 °C and was aerated with a gas mixture of 4 % COj in air. Illumination (1-94 x io3 lux) was supplied by Ecko daylight flourescent tubes either continuously or intermittently (16 h light, 8 h dark). Living cells were examined by Nomarski differential interference microscopy using a Zeips photomicroscope. The cells are extremely delicate and easily burst under the pressure of the coverslip; to prevent this the coverslip was supported on small pieces of broken coverslip which ensured a greater depth of liquid and so reduced the pressure. Cells were photographed with an exposure of i/3Oth s at x 171 magnification on Plus-X film (125 ASA). For electron microscopy cells harvested by gentle centrifugation were fixed for 1 h at 10 °C in 2 % glutaraldehyde buffered to pH 6-5 with 007 M phosphate buffer. After several washes in buffer they were postfixed in buffered 1 % osmium tetroxide; cells fixed in this manner are shown in Figs. 2-4. Alternately, cells were fixed for 6 h at 0-2 °C in 1 % chrome-osmium fixa- tive (Dalton, 1955); cells treated with this fixative are shown in Figs. 5-10. All cells were subsequently dehydrated in a graded ethanol series and embedded in Epon. Sections cut with a diamond knife were collected on copper grids coated with Formvar and carbon. Sections were stained with aqueous uranyl acetate and lead citrate, and examined using a Philips 201 S elec- tron microscope. Osmoregulation in Vacuolaria

OBSERVATIONS The contractile vacuole in Vacuolaria virescens may reach 8-10 /.cm in diameter and is a conspicuous feature of the cell. It occurs at the anterior of the organism and discharges close to the point of flagellar insertion. For a particular temperature and

Fig. 1. Two complete cycles of the contractile vacuole. These micrographs of a stationary cell were photographed at 10-s intervals using Nomarski differential inter- ference microscopy. The following stages can be distinguished: full (A, G, M), relatively empty (c, D, H, 1, N), and filling (E, F, J-L, O, P). In B the arrow indicates the anterior portion of a collapsed contractile vacuole which was photographed during the course of emptying. The small dark structure to the upper right of the figure number in Fig. 1 is thought to represent a subsidiary vacuole. The sequence of changes occurring in this region of the cell can be followed through successive frames; arrowheads in E, H, I, K indicate instances where the subsidiary vacuole is probably emptying into the con- tractile vacuole. x 1370. 2i 6 P. Hey wood growth medium the duration of the contractile vacuole cycle is relatively constant both in individual cells and between members of a population. For example, in defined medium (Heywood, 1973) at 23 °C the period was 49-62 s. Fig. 1 follows the changes occurring in the contractile vacuole through 2 complete cycles. In spite of the relatively long time between frames (10 s) the continuous nature of the contractile vacuole cycle is apparent from this series, and therefore it need hardly be pointed out that the designation of individual stages in this process is an arbitrary one. However, it is useful to recognize the following stages: full (A, G, M), relatively empty (c, D, H, I, N), and filling (E, F, J-L, O, P). One complete cycle of the contractile vacuole shows the following sequence: it is full in A, and in B it was photo- graphed during the course of emptying; in c and D the contractile vacuole is relatively empty but its presence can be distinguished as an irregular area in the cell; in E and F the contractile vacuole is filling, and in G it has attained its spherical full state. In one instance, B, the contractile vacuole was photographed during the course of emptying. These micrographs indicate that the contractile vacuole is a permanent structure which is present throughout the contraction cycle even though it is less conspicuous immediately after emptying. The empty contractile vacuole has an irregular outline but as it fills it usually becomes somewhat elliptical (for example, frames j and o) before it assumes the spherical shape which is its characteristic appearance during the latter part of the filling period. In low-power electron micrographs of Vacuolaria virescens the arrangement of the contractile vacuole and other organelles becomes more apparent (Fig. 2). The cell lacks a and is bounded by a thin . The outer layer of the cell contains chloroplasts which are held in position by endoplasmic reticulum or narrow strands of cytoplasm. There is a large nucleus surrounded by a layer of cytoplasm containing mitochondria and endoplasmic reticulum. An extensive Golgi apparatus occurs between the anterior surface of the nucleus and the contractile vacuole. This region of the cell is seen in greater detail in Figs. 3-5, in which 3 categories of membrane- bound structures with electron-dense contents can be recognized; in ascending order of size and in increasing distance from the surface of the nucleus these are Golgi vesicles, subsidiary vacuoles and the contractile vacuole. Earlier light-microscope descriptions of Vacuolaria virescens noted the presence of a 'supranuclear cap' extending over the anterior surface of the nucleus and producing vesicles which ultimately formed the contractile vacuole (Poisson & Hollande, 1943). Electron micrographs of this region indicate that the supranuclear cap is an extensive Golgi apparatus (Figs. 3-5; see also Mignot, 1967; Schnepf & Koch, 19666). During the course of examining many electron micrographs (including serial sections) through this region of the cell I have been unable to detect any other structure which might be responsible for channelling water into the contractile vacuole; for example, there is nothing comparable to the system of fluid segregation tubules which surrounds the contractile vacuole in some protozoa (McKanna, 1974, 1976). It is therefore possible to confirm earlier observations (Schnepf & Koch, 19666) that the Golgi apparatus is responsible for water secretion in this organism. Figs. 3-5 illustrate differences between the images obtained with the 2 types of Osmoregulation in Vacuolaria 217

Fig. 2. Longitudinal section through a cell of Vacuolaria virescens showing the cell membrane, chloroplasts, nucleus (n), nucleolus {mi), Golgi apparatus (j>) and con- tractile vacuole (c). To aid in orientation the labelling is placed in the same position in both this figure and in the higher-magnification view of the anterior region of this cell (Fig. 3). X5160. 2l8 P. Heywood Osmoregulation in Vacuolaria 219 fixation employed in this investigation. Both gave similar results for most cell com- ponents but the chrome-osmium fixative resulted in a more intensive staining of cell structures; in particular, it is apparent that the contractile vacuole, subsidiary vacuoles and Golgi vesicles contain electron-dense material (Figs. 3, 4). This material is not preserved after fixation in glutaraldehyde followed by postfixation in osmium tetroxide (Fig. 5). However, in cells fixed in this manner the contractile vacuole membrane shows a regular substructure (Figs. 5-10), which is not preserved in the absence of glutaraldehyde (Figs. 3, 4). The only remaining difference between the 2 types of fixation is that the Golgi vesicles and subsidiary vacuoles are somewhat smaller after fixation in chrome-osmium than after fixation in glutaraldehyde and postfixation in osmium tetroxide. These observations obviously raise questions about fixation artifacts, in particular, whether the contractile vacuole membrane really possesses a regular substructure, and whether the Golgi vesicles, subsidiary vacuoles and con- tractile vacuole contain some material other than water. In the firstinstanc e it is unlikely that faulty fixation would produce a regular specialization of the membrane; it is more probable that this appearance corresponds to a structure found in living cells which is not preserved by fixation in chrome-osmium. This conclusion is strengthened by the fact that similar types of modified membranes, sometimes termed 'coated membranes' or 'bristle coat structures', are known to occur in other situations invol- ving transport, particularly transport of water (Aaronson & Behrens, 1974; Franke, Kartenbeck & Spring, 1976; Hoffman, 1976; McKanna, 1974, 1976; Morre", Mollen- hauer & Bracker, 1971). Similarly, it is unlikely that faulty fixation would cause electron-dense material to be present in the Golgi vesicles, subsidiary vacuoles and contractile vacuole unless there had been widespread cytoplasmic damage, in which case this material could have leaked into these regions of the cell. There is no sign of such damage nor is this material present in other regions of the cell. Furthermore, similar material occurs in the Golgi apparatus of Glaucocystis where this is also involved in water secretion. Light-microscope observations on living cells of Vacuolaria virescens surrounded by India ink indicated that in some instances the water eliminated by the contractile contained mucus (Tschermak-Woess, 1954). This presumably corresponds to the material preserved in the contractile vacuole after treatment with the chrome-osmium fixative. The presence of electron-dense material in the Golgi vesicles, subsidiary vacuoles and contractile vacuole suggests a mechanism by which water molecules can enter the osmoregulatory system and be bound there

Fig. 3. Higher magnification of the contractile vacuole region of the cell in Fig. 2. The micrograph includes an oblique section through the base of the flagellar (/r) which makes contact with the anterior surface of the nucleus (n). Electron-dense material is present in vesicles produced by the Golgi apparatus (g), and in the subsidiary vacuole (s) and the contractile vacuole (c). The arrowhead indicates a region where the sub- sidiary vacuole was about to empty into the contractile vacuole. x 14850. Fig. 4. The proximal face of the Golgi apparatus Cg) is situated close to the surface of the nucleus (n). Golgi vesicles show an increasing gradation in size towards the distal face. Electron-dense material is present in the Golgi vesicles, subsidiary vacuoles (s) and the contractile vacuole (c). Large numbers of membrane profiles and small vesicles occur in this region of the cell, x 24000. P. Heytoood Osmoregulation in Vacuolaria for sufficient time to allow for their expulsion from the cell. The latter is an important consideration since in the absence of any restraint water molecules contained within vacuoles have a strong tendency to cross the limiting membrane and to re-enter the surrounding cytoplasm (McKanna, 1976). It is postulated that the Golgi apparatus is producing a strongly hydrophilic substance which will attract water molecules into

10 Figs. 7-10. Sections through contractile vacuole membrane. Arrowheads in Figs. 8-10 indicate the arrangement of bristles on the cytoplasmic surface of the membrane. Figs. 7, 8 contain surface views of these membrane subunits. x 100 000.

Fig. 5. Longitudinal section through the nucleus (JI), Golgi apparatus (g), subsidiary vacuoles (J), and contractile vacuole (c). The (/ ) and portions of the flagellar root (/>) are also present. The contractile vacuole which was relatively empty at the time of fixation is seen in oblique section with the result that 2 compartments appear to be present. The intervening region of contractile vacuole membrane is present in sur- face view and is shown in greater detail in Fig. 6. x 16 500. Fig. 6. Higher magnification of the region delimited by asterisks in Fig. 5. The sub- unit structure of the membrane can be seen in this view of the surface. An arrange- ment of 4 subunits (arrows) around a central subunit is present. The hexagonal nature of this packing arrangement was demonstrated by photographic rotational enhance- ment, x 150000. 222 P. Heytvood Golgi vesicles, subsidiary vacuoles and contractile vacuole and maintain them there until they are eliminated from the cell. It is not known whether the subsidiary vacuoles are permanent structures into which the Golgi vesicles empty their contents and which, in turn, discharge into the contractile vacuole, or whether they are transient structures formed by fusion of the Golgi vesicles, and which will disappear after they have emptied into the contractile vacuole. In both instances there will be considerable recycling of membranes and membrane components: the number of small vesicles and membrane profiles in this region (Figs. 3-5) indicate that such a process is occurring. At present there is little information about the mechanism of contractile vacuole discharge. For example, it is not known whether this structure empties to the exterior at a specific site. The cause of the contraction is uncertain: close examination of the cytoplasm around the vacuole has failed to reveal any structures (for example, microfilaments) which might be implicated in this process. However, in the course of these investigations new information was gathered on the structure of the contractile vacuole membrane. Perpendicular sections indicate that regions of the trilaminar membrane possess regularly spaced bristles or pegs which extend from the membrane into the cytoplasm (Figs. 8-10). These bristles extend inwards from the membrane for a distance of 13-15 nm and are spaced at a distance of 21-24 nm. In surface views (Figs. 6—8) the contractile vacuole membrane is observed to consist of regular subunits. Photographic rotational enhancement of selected areas (for example, the subunits indicated by arrows in Fig. 6) indicates that the subunits are arranged in a hexagonal pattern. The diameter of the subunit corresponds to the 21-24 nm distance between adjacent bristles in sections through the membrane, thus indicating that the bristles represent sections through the lateral walls of the membrane subunits.

DISCUSSION The timing of the contractile vacuole cycle (49-62 s at 23 °C) is in agreement with that observed by earlier investigators, for example, 36-42 s at 18-20 °C (Spencer, 1971), 80 s at 18 °C (Poisson & Hollande, 1943), and 48 s at an unspecified temperature (Schnepf & Koch, 19666). The latter authors concluded that the contractile vacuole is a temporary structure formed by fusion of vesicles produced by the Golgi apparatus. They suggested that after discharging its aqueous contents the contractile vacuole membrane became part of the plasmalemma. This is a reasonable speculation since a flow of membrane material from the Golgi apparatus to the cell surface is known to occur in other instances (Dauwalder, Whaley & Kephart, 1972; Morre' et al. 1971). However, the distinctive structure of the contractile vacuole membrane indicates that a similar mechanism is not applicable to Vacuolarta virescens since this type of membrane specialization is not present in the membranes of the Golgi vesicles, subsidiary vacuoles or in the plasmalemma. Moreover, light microscopy using Normarski differential interference microscopy indicates that the contractile vacuole is a perma- nent organelle. The regularly spaced bristles or pegs extending from the contractile vacuole membrane into the cytoplasm resemble 'bristle coat structures'. This type of 'coated Osmoregulation in Vacuolaria 223 membrane' is known to occur in other situations involving transport of water and (McKanna, 1974, 1976). It is possible that this specialization may confer additional strength on the membrane to enable it to withstand the pressure changes occurring during the contractile vacuole cycle. Alternatively, the bristle elements may have an active role in effecting the emptying of the contractile vacuole. McKanna (1976) has termed the 'coated membrane' a 'permeability modulating membrane coat' and has described the occurrence of this coat in the fluid segregation organelles of some protozoa. In these organisms the Golgi apparatus is not associated with the contractile vacuole. Instead the contractile vacuole is surrounded by a net- work of fluid segregation tubules which are confluent with the contractile vacuole. McKanna (1974, 1976) has discussed how this type of arrangement can account for the entry of water molecules into the osmoregulatory system and for their remaining within this system until their discharge from the cell. In contrast, the contractile vacuole system of Vacuolaria virescens possesses an extensive Golgi apparatus in place of the network of fluid segregation tubules. The presence of electron-dense material within the Golgi vesicles, subsidiary vacuoles and contractile vacuole suggests an alternate method for fluid segregation. If the Golgi produces a highly hydrophilic substance this will attract water from the cytoplasm into the Golgi vesicles, subsidiary vacuoles and contractile vacuole and will retain water in these compartments until its discharge from the cell. It is known that some extracellular material has a high affinity for water, for example, the acid mucopolysaccharide hyaluronic acid can bind 68-520 ml of water per g (Fitton Jackson, 1964). This observation led to the suggestion that 'similar materials while still in the Golgi cisternae or vesicles could influence intra- cellular hydration' (Dauwalder et al. 1972). In order to substantiate this model for secretion in Vacuolaria virescens it will be necessary to isolate the substance produced by the Golgi apparatus (for example, from the culture medium in which cells were grown) and determine its physical characteristics. In the absence of this type of information there is evidence from, comparative studies that a similar mechanism may be employed by other algae. For example, the Golgi apparatus of Glaucocystis which is also involved in water secretion produces vesicles containing electron-dense material; the possibility that this organelle may secrete together with the water has been recognized (Schnepf & Koch, 1966 a). Furthermore, the observation by Dodge (1973) that the contractile vacuole frequently occurs between the Golgi apparatus and the plasmalemma suggests that this mechanism of water secretion may be widespread in the algae.

I am most grateful to Ms Amy Davidoff for her careful and skilled assistance in preparing the illustrations and in typing the manuscript. This investigation was supported by a Bio- medical Research Support Grant from Brown University.

REFERENCES AARONSON, S. & BEHRENS, U. (1974). Ultrastructure of an unusual contractile vacuole in several chrysomonad phytoflagellates. J. Cell Set. 14, 1-9. AHMAD, M. & COUILLARD, P. (1974). The contractile vacuole in Amoeba proteus: temperature effects. J. Protozool. 21, 330-336. 15 CEL 31 224 P- Heywood CIENKOWSKY, L. (1870). tJber Palmellaceen und einige Flagellaten. Arch, mikrosk. Anat. EntwMech. 7, 431-436. + a+ CZARSKA, L. (1964). Role of the K and Ca in the excitability of protozoan cell. Chemical and electric stimulation of contractile vacuoles. Acta protozool. 2, 287-296. DALTON, A. J. (1955). A chrome-osmium fixation for electron microscopy. Anat. Rec. 121, 281. DANGEARD, P. A. (1939). Nouvelles observations sur la Vacuolaria virescens Cienk. Botaniste 29, 3H-334- DAUWALDER, M., WHALEY, W. G. & KEPHART, J. E. (1972). Functional aspects of the Golgi apparatus. Sub-cell. Biochem. 1, 225-275. DODGE, J. D. (1973). The Fine Structure of Algal Cells. London: Academic Press. FITTON JACKSON, S. (1964). Connective cells. In The Cell, vol. 6 (ed. J. Brachet & A. E Mirsky), pp. 387-520. New York & London: Academic Press. FOTT, B. (1971). Algenkunde. Jena, East Germany: Fischer. FRANKE, W. W., KARTENBECK, J. & SPRING, H. (1976). Involvement of bristle coat structures in surface membrane formations and membrane interactions during coenocytotonic cleavage in caps of Acetabularia mediterranea. J. Cell Biol. 71, 196-206. GUILLARD, R. R. L. (i960). A mutant of Chlamydomonas ntoetuusii lacking contractile vacuoles. J. Protozool. 7, 262-268. HEYWOOD, P. (1973). Nutritional studies on the Chloromonadophyceae: Vacuolaria virescens and Gonyostomum semen. J. Phycol. 9, 156-159. HOFFMAN, L. R. (1976). Fine structure of Cylindrocapsa zoospores. Protoplasma 87, 191-219. KITCHING, J. A. (1967). Contractile vacuoles, ionic regulation, and . In Researdi in Protozoology, vol. 1 (ed. T. T. Chen), pp. 307-336. New York: Pergamon. LLOYD, F. E. (1928). The contractile vacuole. Biol. Rev. Camb.phil. Soc. 3, 329-358. MCKANNA, J. A. (1974). Permeability modulating membrane coats. I. Fine structure of fluid segregation organelles of peritrich contractile vacuoles. J. Cell Biol. 63, 317-322. MCKANNA, J. A. (1976). Fine structure of fluid segregation organelles of Paramecium contrac- tile vacuoles. J. Ultrastruct. Res. 54, 1-10. MIGNOT, J. P. (1967). Structure et ultrastructure de quelques Chloromonadines. Protistologica 3. 5-23- MIGNOT, J. P. (1976). Complements a l'^tude des Chloromonadines. Ultrastructure de Chat- tonella subsalsa Biecheler flagell6 d'eau saumatre. Protistologica 12, 279-293. MORRE, D. J., MOLLENHAUER, H. H. & BRACKER, C. E. (1971). Origin and continuity of Golgi apparatus. In Origin and Continuity of Cell Organelles (ed. J. Reinert & H. Ursprung), pp. 82-126. New York: Springer. ORGAN, A. E., BOVEE, E. C. & JAHN, T. L. (1968). Adenosine triphosphate acceleration of the nephridial apparatus of Paramecium multiinicronucleatum. J. Protozool. 15, 173-176. POISSON, R. & HOLLANDE, A. (1943). Considerations sur la cytologie, la mitose et les affinities des Chloromonadines. fitude de Vacuolaria virescens Cienk. Annls Sci. nat. (Z00L), Se> II, 5, 147-160. PRUSCH, R. D. & DUNHAM, P. B. (1970). Contraction of isolated contractile vacuolea from Amoeba proteus. J. Cell Biol. 46, 431-434. RIFKIN, J. L. (1973). The role of the contractile vacuole in the osmoregulation of Tetrahymena pyriformis. J. Protozool. 20, 108-114. SCHNEPF, E. & KOCH, W. (1966 a). Golgi-Apparat und Wasserausscheidung bei Glaucocystis. Z. PflPhysiol. 55, 97-109- SCHNEPF, E. & KOCH, W. (19666). Uber die Entstehung der pulsierenden Vacuolen von Vacuolaria virescens (Chloromonadophyceae) aus dem Golgi-Apparat. Arch. Mikrobiol. 54, 229-236. SPENCER, L. (1971). A study of Vacuolaria virescens Cienkowski. J. Phycol. 7, 274-279. TORIUMI, S. & TAKANO, H. (1973). Fibrocapsa, a new genus in Chloromonaophyceae from Atsumi Bay, Japan. Bull. Tokai. reg. . Res. Lab. 76, 25-35. TSCHERMAK-WOESS, E. (1954). Das sogenannte Alveolarplasma und die Schleimbildung bei Vacuolaria virescens. Ost. bot. Z. IOI, 328-333.

(Received 2 September 1977)