Occurrence of Proteinaceous 10-nm Filaments throughout the Cytoplasm of Algae of the Order Dasycladales

EXPERIMENTAL CELL RESEARCH ARTICLE NO.

240, 176–186 (1998)

EX973924

Occurrence of Proteinaceous 10-nm Filaments throughout the Cytoplasm of Algae of the Order Dasycladales Sigrid Berger, Werner Wittke, and Peter Traub1 Max-Planck-Institut fu¨r Zellbiologie, Ladenburg 68522, Germany

INTRODUCTION Previously, whole-mount electron microscopy of nuclei extruded together with residual cytoplasm from the rhizoids of several algal species of the order Dasycladales has revealed the occurrence of an intra- and perinuclear network of 10-nm filaments morphologically indistinguishable from that of mammalian vimentin intermediate filaments. The present investigation demonstrates the existence of a filament system throughout the cytoplasm of the rhizoid, stalk, and apical tip of these giant cells. However, while the perinuclear 10-nm filaments interconnecting the nuclear surface with a perinuclear layer of large, electrondense bodies filled with nucleoprotein material are of smooth appearance, those continuing within and beyond the perinuclear bodies are densely covered with differently sized, globular structures and, therefore, are of a very rough appearance. The filaments in the very apical tip of the cells are mainly of the smooth type. The transition from smooth to rough filaments seems to occur in the numerous perinuclear dense bodies surrounding the large nucleus. Digestion of the rough filaments with proteinase K removes the globules from the filament surface, revealing that throughout the nonvacuolar, intracellular space the filaments have the same basic 10-nm structure. On the other hand, gold-conjugated RNase A strongly binds to the filament-attached globules but not to the smooth, perinuclear, and the proteinase K-treated, rough filaments. In addition, an antibody raised against Xp54, a highly conserved protein which in Xenopus oocytes is an integral component of stored mRNP particles, decorates the rough but not the smooth 10-nm filaments. These results support the notion that the 10nm filament system of Dasycladales cells plays a role in the transient storage of ribonucleoprotein particles in the cytoplasm and possibly fulfils a supportive function in the actomyosin-based transport of such material to various cytological destinations. q 1998 Academic Press

1

To whom reprint requests should be addressed.

176

0014-4827/98 $25.00 Copyright q 1998 by Academic Press All rights of reproduction in any form reserved.

AID

ECR 3924

/

6i2f$$$441

Since the detection of intermediate filaments (IFs) about 30 years ago, much effort has been dedicated to unveil their functions in the cytoplasm of eukaryotic cells. Keeping in mind the large number of different IF subunit proteins with unique characteristics in addition to their common propensity to polymerize into uniform 10-nm filaments, their high susceptibility to posttranslational modification, and their differentiationspecific and developmentally regulated expression [for reviews, see 1–7], it seems rather unlikely that they serve a merely structural, cytoskeletal role. Indeed, evidence is accumulating that they fulfill additional functions as constituents of a number of intracellular information transfer pathways [8, 9] and eventually in gene regulation [9]. Yet, so far IFs have unequivocally been demonstrated only in the animal kingdom. Although recent data suggest their existence in plant cells, too [for references, see 10], their occurrence throughout the plant kingdom is still doubted by many cell biologists [for a review, see 11]. A structural role, as it is assigned to IFs in animal cells, might indeed not be necessary to the same extent for plant cells since their overall morphology is stabilized by a rigid cell wall. However, the universality of basic cell structure and cell physiological processes throughout the animal and plant kingdoms makes the occurrence of IF or IF-like filament systems in plant cells highly probable. Due to their siphonous organization and large size, the unicellular and mononuclear algae of the order Dasycladales [12] lend themselves as ideal objects to search for IFs in plant cells. Their cytoplasm and nuclei can be manually isolated in a well-preserved state by cutting the cells open and carefully squeezing out their contents. Therefore, technical problems caused by a rigid, impermeable cell wall, which often places plant cells as experimental systems behind animal cells, are avoided. Taking advantage of these circumstances, extended 10-nm filament systems have been detected around nuclei extruded from the rhizoids of several species of Dasycladales that, morphologically, appeared identical

04-08-98 06:55:10

eca

TEN-NANOMETER FILAMENTS IN DASYCLADALES

with the vimentin IF network of mammalian fibroblasts [10]. In addition, these studies revealed the existence of an intranuclear filament network that is indistinguishable in its electron-microscopic appearance from the perinuclear 10-nm filament system. Notwithstanding the presence of the nuclear envelope, it seems to form a filamentous continuum with the perinuclear filaments. The present investigation describes the extension of this filament meshwork into the cytoplasm of the rhizoid, stalk, and apical tip of growing Dasycladales cells, but with a morphology different from that of the nuclear and perinuclear networks in that its 10nm filaments are covered with huge quantities of what seems to be ribonucleoprotein (RNP) particles. MATERIALS AND METHODS Cells. The unicellular, siphonous marine algae of the order Dasycladales were grown in artificial sea water as described in Berger and Kaever [12]. Acetabularia acetabulum (Acetabulariaceae) and Batophora oerstedi (Dasycladaceae) cells at different developmental stages, starting from about 1 cm length up to full development before the onset of meiosis, were used. Whole-mount electron microscopy. In general, the procedure followed the method of Fey et al. [13]. Cells were cut open with scissors in 100 mM phosphate buffer, pH 7.4, 300 mM sucrose, 5 mM magnesium acetate, 25 mM EGTA, 0.1% bovine serum albumin. For studies of the structural organization of the main axis (stalk) cytoplasm, the apical tip and rhizoid of the cell were cut off and the contents of the remaining part carefully squeezed out into the buffer. Only cytoplasmic portions extruded with the same siphonous organization as in the intact cell were used for further processing. For the study of the structural organization of apical tip or gametophore cytoplasm, the respective parts were cut off from the cell and their contents squeezed out while the utmost end was held with fine forceps. For the analysis of the structural organization of the rhizoid cytoplasm remote from the nucleus, B. oerstedi cells had to be used in order to clearly separate rhizoid cytoplasm from perinuclear cytoplasm. In many of these cells, during a certain developmental stage the nucleus and perinuclear cytoplasm are located in the lower part of the main axis rather than in one of the rhizoid branches. From such cells, the rhizoid was cut off and, holding it at its utmost end with fine forceps, its contents were squeezed out. The cytoplasmic siphons were transferred with fine needles into CSK I buffer [10 mM Pipes, pH 6.8, 100 mM NaCl, 300 mM sucrose, 3 mM MgCl2 , 1 mM EGTA, 1.2 mM phenylmethylsulfonyl fluoride (PMSF), 0.5% Triton X-100]. They were centrifuged for 30 min at 20,000g and 107C onto a freshly glow-discharged, carbon-collodiumcoated electron microscopy (EM) grid sitting at the bottom of a small centrifuge tube. From here on the procedure was carried out at room temperature with the cytoplasm sticking to the EM grid. The cytoplasm was extracted with CSK II buffer (10 mM Pipes, pH 6.8, 250 mM ammonium sulfate, 300 mM sucrose, 3 mM MgCl2 , 1 mM EGTA, 1.2 mM PMSF, 0.5% Triton X-100) for 5 min, with CSK III buffer (10 mM Pipes, pH 6.8, 50 mM NaCl, 300 mM sucrose, 3 mM MgCl2 , 1 mM EGTA, 1.2 mM PMSF, 0.5% Triton X-100) for 10 min, with CSK II buffer for 5 min, and washed twice for 2 min each with CSK I buffer. It was then treated with RNase A (25 mg/ml; Boehringer Mannheim, Mannheim, Germany) in CSK I buffer for 20 min and washed once with CSK I buffer for 5 min. When the residual cytoplasm was treated with RNase–gold, the RNase digestion step was omitted. In some experiments, proteins in the samples were digested prior to RNase A digestion. The procedure included the following steps: CSK III buffer was washed off three times for 2 min

AID

ECR 3924

/

6i2f$$$442

04-08-98 06:55:10

177

each with CSK I buffer lacking PMSF. The digestion was performed either for 15 min with 50 mg/ml proteinase K at 377C or overnight with 50 mg/ml protease type XXIV or trypsin from bovine pancreas (all from Sigma, Deisenhofen, Germany) at room temperature in CSK I buffer without PMSF, followed by three washes with CSK I buffer for 2 min each. Thereafter, the cytoplasm was fixed with 2% glutaraldehyde in 0.1 M cacodylate buffer, pH 7.4, for 30 min on ice. The fixative was removed by three 10-min washes with 0.1 M cacodylate buffer. The sample was postfixed with 1% osmium tetroxide in 0.1 M cacodylate buffer, pH 7.4, on ice, washed three times for 10 min each with 0.1 M cacodylate buffer, dehydrated at room temperature with increasing concentrations of either ethanol or acetone, critical point-dried in a Balzers CPD 030 apparatus (Balzers, Liechtenstein), and rotary shadowed with titanium at an angle of 107. Electron microscopy was done with a Philips EM 400T. Conjugation of RNase A to colloidal gold. For the preparation of colloidal gold, 100 ml of 0.01% tetrachloroauric acid was heated to boiling. Four milliliters 1% Na3-citrate was added and the solution boiled for 5 min in order to reduce the gold chloride. The formation of colloidal gold was indicated by the appearance of a light red color. After cooling to room temperature, the pH of the gold suspension was adjusted to 9.0 with 0.2 M K2CO3 . To load the gold particles with protein, 1 mg RNase A was dissolved in 1 ml of double-distilled water and 50 ml of the gold suspension was added to it with constant stirring. After 2 min, 600 ml of the suspension was mixed with 100 ml of 2.5 M NaCl. If the color of the enzyme–gold complex did not change from red to violet, this was taken as indication that all gold particles were coated with enzyme. After incubation for another 3 min, 2.5 ml of 1% polyethylene glycol (Carbowax 20M; Serva, Heidelberg, Germany) was added to stabilize the complexes. The suspension was then centrifuged for 50 min at 27,000g and 47C to remove nonbound protein. The supernatant was carefully aspirated and discarded. The sediment was resuspended in 2 ml of PBS (0.02 M phosphate buffer, pH 7.5, 0.13 M NaCl) containing 0.5 mg/ ml polyethylene glycol and the suspension centrifuged for 10 min at 1000g and 47C to remove clumped granules. Na-azide was added to a final concentration of 0.5 mg/ml. The suspension was stored at 47C for no longer than a week. For labeling of whole-mount preparations, the suspension was diluted 1:10 or 1:100. Binding of gold-conjugated RNase A to rough filaments. Cytoplasmic siphons were prepared as for whole-mount electron microscopy, up to their extraction with CSK III buffer for 10 min. They were then washed twice with PBS containing 0.05% Carbowax 20M and incubated with gold-conjugated RNase A in PBS for 45 min at room temperature. The incubation was followed by three 10-min washes with PBS. Thereafter, the samples were fixed, washed, and critical point dried as described for whole-mount electron microscopy. Antibodies and antibody labeling of whole mounts. The following mouse monoclonal antibodies were tested on whole-mount preparations for their affinity for Dasycladales cytoplasmic rough filaments: anti-actin N350 (Amersham, Braunschweig, Germany); anti-cytokeratins pan, 1-8, and 18 (Boehringer Mannheim); anti-cytokeratins 8 and 18 (Progen, Heidelberg, Germany); anti-cytokeratins AE1 and AE3 (kindly provided by Dr. T.-T. Sun, New York University Medical Center, New York, NY); anti-lamin (against all three lamins [14]; the hybridoma cells were kindly provided by Dr. G. Warren, University of Dundee, Dundee, Scotland); anti-vimentin AS12, AS13, PK17, and SK1 (directed against various epitopes of mouse vimentin, unpublished results); and anti-a-IFA ([15], produced in suspension culture of hybridoma cell line TIB 131 from the American Type Culture Collection, Rockville, MD). In addition, the following polyclonal antibodies were tested: goat anti-vimentin 259 [16], rabbit anti-lamin B [17], and rabbit anti-Xp54 [18] (kindly provided by Dr. J. Sommerville, University of St. Andrews, St. Andrews, Scotland). As second-

eca

178

AID

BERGER, WITTKE, AND TRAUB

ECR 3924

/

6i2f$$3924

04-08-98 06:55:10

eca

TEN-NANOMETER FILAMENTS IN DASYCLADALES ary antibodies, 10-nm gold-conjugated goat anti-rabbit IgG, 10-nm gold-conjugated goat anti-mouse IgG (both from Biotrend Chemikalien, Ko¨ln, Germany), and 10-nm gold-conjugated donkey anti-goat IgG (Plano, Wetzlar, Germany) were used. Antibody labeling of whole-mount preparations was as described previously [10]. Embedment-free sections. The technique employed followed the procedure described by Capco et al. [19] and Berger et al. [10]. Since the large Dasycladales cells had to be cut into pieces to facilitate penetration of buffers and fixatives, the pieces were closed with plugs of 1% low-melting agarose to prevent distortion or loss of cytoplasm. Isolation and negative staining of 10-nm filaments. Fifty A. acetabulum cells were bundled at their tips and cut open with scissors. The cytoplasm was slowly centrifuged out of the cells into 10 mM TrisrHCl, pH 7.6, 160 mM KCl, 1 mM PMSF. The cytoplasm was resuspended in buffer, loaded on top of 20% sucrose in buffer, and centrifuged for 10 min at 20,800 rpm and 47C. The resuspended pellet was adsorbed to carbon-coated mica and negatively stained according to the method of Valentine et al. [20].

RESULTS

In analogy to the technique applied to visualize the perinuclear 10-nm filament system of Dasycladales cells [10], these were cut open and the cytoplasm of their main axis was squeezed out to be subjected to the whole-mount extraction procedure. The extrusion of the cytoplasm from the stalk was necessary since the impermeable cell wall surrounding it prevented an efficient extraction of the cell segments. As illustrated in Fig. 1, a dense filament system was detected to permeate the entire cytoplasmic space of the Dasycladales cells. The filaments were of the same straight, branching and intertwining structure as those occurring in the immediate vicinity of the nucleus (Fig. 2; see also [10]). However, in contrast to these perinuclear filaments those remote from the nucleus were densely covered with globular material of different size. While some of the particles have the size of ribosomes, most of them are much larger. Often the filaments appeared to be in close contact with organelles (Fig. 3), mostly via their rough but sometimes also their smooth regions. Occasionally, filaments extending from the nuclear surface into the cytoplasm carried only a few globular structures (Fig. 4). Since they appear to represent a transition form between smooth and rough filaments and measured approximately 10 nm in diameter, the core structure of the rough filaments very likely is of

179

the same 10-nm type as the smooth filaments. Indeed, when long stretches of rough filaments were sometimes interrupted by smooth regions, their diameter could also be measured to be approximately 10 nm. Sometimes, the smooth filaments were locally untwined, suggesting a helical arrangement of subfilament strands in the filaments (Fig. 4, inset). The frequency and length of undecorated filament sections increased toward the apical tip of the cell. The cytoplasm of the apical tip was often occupied by an extended network of only smooth filaments (Fig. 5). Since it was possible that this extended network of undecorated 10-nm filaments is characteristic not only of the apical part of the Dasycladales cells but also occurs at the other pole, in the rhizoid, the cytoplasm of the latter was also subjected to the whole-mount extraction procedure. Such a possibility had to be taken into consideration because the nucleus located in one of the intertwined branches of the rhizoid is always embedded in a dense meshwork of smooth 10-nm filaments [10]; it might therefore occupy the whole cytoplasm of the rhizoid. In order to clearly differentiate between perinuclear filaments and filaments in more remote areas of the rhizoid, B. oerstedi cells were employed for specimen preparation. Before the onset of gametophore formation, the nuclei of these cells move to the lower part of the main axis so that the entire cytoplasm of the rhizoid should be free of those filaments normally adhering to the nucleus. The filaments detected in the rhizoid cytoplasm turned out to be of the same rough type as those in the main axis (data not shown). A large number of perinuclear dense bodies with unknown function surround the nuclei of Dasycladales cells [22, 23]. Sometimes, relics of these bodies could be recognized in whole-mount preparations. They were filled with large quantities of very rough filaments (Figs. 6 and 7) and connected to the nucleus via cytoplasmic bridges occupied by masses of smooth 10-nm filaments (Fig. 6). The rough filaments entering the stalk and rhizoid cytoplasm apparently originate from the perinuclear dense bodies. To refute the argument that the filaments observed are artifacts of the whole-mount extraction procedure, cytoplasm was carefully centrifuged out of the cell and

FIG. 1. Whole-mount preparation of A. acetabulum stalk cytoplasm. A dense meshwork of straight, branching, and intertwining filaments is revealed after high-salt, nonionic detergent extraction and digestion with RNase A. The filaments are densely covered with globules. Although the globules are mostly of the same size and shape, some of them clearly differ. Bar, 110 nm. FIG. 2. Whole-mount preparation of an A. acetabulum nucleus. The micrograph shows a section of the perinuclear area which is occupied by a dense network of 10-nm filaments associated with the residual nuclear envelope. The filaments have a smooth surface. Bar, 100 nm. FIG. 3. Stalk cytoplasm of A. acetabulum after its processing for whole-mount electron microscopy. Residual organelles are often seen in close association with the filaments. Bar, 55 nm. FIG. 4. Whole-mount preparation of perinuclear cytoplasm of A. acetabulum. Occasionally, 10-nm filaments sparsely covered with globular structures are observed in the immediate vicinity of the nucleus and throughout the main axis cytoplasm. The inset shows a local region of untwining of a 10-nm filament with visualization of a helical arrangement of subfilament strands. Bar, 50 nm; inset bar, 25 nm.

AID

ECR 3924

/

6i2f$$$442

04-08-98 06:55:10

eca

180

AID

BERGER, WITTKE, AND TRAUB

ECR 3924

/

6i2f$$3924

04-08-98 06:55:10

eca

TEN-NANOMETER FILAMENTS IN DASYCLADALES

then freed of large organelles by sucrose density gradient centrifugation. In the remaining cytoplasm, smooth 10-nm filaments as well as filaments densely covered with particles could be visualized after negative staining (Fig. 8). In a further experiment, an attempt was made to preserve the natural distribution of the 10-nm filaments in the cytoplasm of Dasycladales cells during cell extraction. For this purpose, the technique of embedment-free sectioning was employed. Sectioning through the main axis also showed a dense network of rough filaments permeating the cytoplasm, although their decoration with globular structures was not as clearly seen as that in whole-mount preparations (Figs. 9 and 10). Possibly, the thick cell wall still surrounding the cytoplasm prevented an optimal extraction of the stalk segments. Nevertheless, the surface structure of the filaments was clearly different from that of the smooth, perinuclear 10-nm filaments described previously [10]. The filaments were also seen in association with organelles (Fig. 9). Since the binding of the globules to the filaments is resistant to extraction with high salt, it was attempted to remove them by proteolytic digestion and thus to reveal with more clarity the core structure of the rough filaments. As shown in Fig. 11, digestion of the wholemount preparations with proteinase K produced smooth 10-nm filaments with outlines identical to those of the perinuclear 10-nm filaments. The digestion did not affect the association of the filaments with large organelles. Digestion with protease type XXIV or trypsin completely destroyed the filaments, demonstrating that they are proteinaceous structures throughout. To investigate the chemical nature of the globular structures attached to the 10-nm core filaments, wholemount preparations of A. acetabulum cytoplasm were incubated with gold-conjugated RNase A. This experiment was suggested by the fact that during their vegetative growth phase Dasycladales cells accumulate and store large quantities of RNP material in their cytoplasm. Figure 12A indeed shows strong binding of the RNase A–gold complex to the globular structures but not to the smooth filaments occasionally observed within the dense network of rough 10-nm filaments (Fig. 12B). Gold-conjugated RNase A also did not associate with the smooth 10-nm filaments of the perinuclear region or with the core filaments stripped of the globules by proteinase K digestion. Preincubation of

181

the rough filaments with free RNase A totally suppressed the binding of the RNase A–gold conjugate. Decoration of the rough filaments was also achieved with the polyclonal antibody anti-Xp54 [18] which is directed against an evolutionarily highly conserved mRNP protein from Xenopus (Fig. 13). The antibody did not react with the smooth perinuclear 10-nm filaments. Finally, an attempt was undertaken to characterize the filaments immunologically and thus to gain information on their protein composition. A series of antibodies directed against mammalian IF proteins (see Materials and Methods) and an anti-actin antibody which recognizes actin in Acetabularia totally failed to react with the filaments of whole-mount preparations. Monoclonal anti-lamin A/B/C and polyclonal antilamin B antibodies raised against mammalian lamins were also completely inactive on the rough 10-nm filaments, including their occasionally exposed smooth sections, but the anti-lamin B antibody efficiently labeled the perinuclear, smooth 10-nm filaments [10]. DISCUSSION

Complementing previous data showing the occurrence of proteinaceous 10-nm filaments in the perinuclear area of the cytoplasm and in the interior of the nucleus of Dasycladales cells [10], the results of the present investigation demonstrate that this filament system is not restricted to a distinct region within the rhizoid but extends from the nucleus to the very apical tip of the cell; i.e., the total, nonvacuolar, intracellular space is occupied by a filamentous continuum. With respect to a potential biological role of the 10-nm filaments in Dasycladales cells, the characteristics of their intracellular distribution suggest that they may fulfill a structure-stabilizing function as sturdy cytoskeletal elements. Although the highly differentiated shape of the algae is maintained primarily by a rigid cell wall, the 10-nm filaments may stabilize the thin cytoplasmic layer underneath the plasma membrane toward the huge vacuole which occupies the largest part of intracellular space in these cells. On the other hand, the 10-nm filaments may also participate in the organization of the cytoplasm, positioning a plethora of organelles and other cytoplasmic constituents in order to allow their optimal perfor-

FIG. 5. Cytoplasm of the apical tip of an A. acetabulum cell which had been processed for whole-mount electron microscopy. Extended regions of smooth 10-nm filaments occur between densely covered filaments. Bar, 100 nm. FIG. 6. Whole-mount electron microscopy of the perinuclear region of an A. acetabulum cell. Sometimes residues of the perinuclear dense bodies (PB), a large number of which surround the nucleus, survive the extraction procedure. An extended meshwork of smooth 10nm filaments spans the space between the nuclear surface and the perinuclear dense bodies. Bar, 200 nm. FIG. 7. Whole-mount electron microscopy of the perinuclear region of an A. acetabulum cell. In the residues of perinuclear dense bodies, 10-nm filaments densely covered with globular structures are seen. Bar, 150 nm. FIG. 8. Isolated and negatively stained rough and smooth 10-nm filaments from the main axis of A. acetabulum. Bar, 100 nm.

AID

ECR 3924

/

6i2f$$$443

04-08-98 06:55:10

eca

182

AID

BERGER, WITTKE, AND TRAUB

ECR 3924

/

6i2f$$3924

04-08-98 06:55:10

eca

TEN-NANOMETER FILAMENTS IN DASYCLADALES

mance in the scope of a great variety of developmental and cell physiological processes. Since the Dasycladales are physically gigantic, unicellular, mononuclear organisms with distinct morphological features, a high morphogenetic capacity, and a pronounced polarity [12], the 10-nm filaments will likely fulfill different functions in the various parts of the cell. This supposition receives support from the regionally different surface morphology of the filaments in whole-mount preparations of Dasycladales cells. While in the immediate vicinity of the nucleus and in the very apical tip of the cell the filaments have a smooth appearance, in the remaining parts of the rhizoid and throughout the stalk they are densely covered with a multitude of differently sized, globular structures, although in central parts of the stalk rather smooth filament sections occasionally have also been detected. Since the rough filaments have been observed in whole-mount preparations as well as in embedment-free sections of well-fixed cytoplasmic material, an artifactual covering of smooth filaments with globular, subcellular structures or drying artifacts is highly improbable. The natural appearance of mainly rough but also smooth filaments throughout the cytoplasm is further supported by the fact that these structures are also detected by negative staining after gently centrifuging out the cytoplasmic contents from cells which have been cut open. The fact that variation in surface texture does not result from the presence of different filament types could be substantiated by the observation that brief treatment of the rough filaments with proteinase K completely removed the particles from the filament surface and produced filaments morphologically indistinguishable from those associated with the nucleus or occurring in the apical tip of the cell. It was interesting to note that the filaments themselves largely resisted limited proteolytic degradation. They thus behaved very similar to mouse epidermal cytokeratin filaments which had been shown to physically survive limited chymotryptic digestion although the terminal, glycinerich polypeptide regions of their constituent subunit proteins had been removed from the a-helix-enriched core structure [21]. The question of whether the subunit proteins of the Dasycladales 10-nm filaments undergo a similar proteolytic truncation can only be answered after their molecular characterization. Yet, the resistance of the filaments to proteinase K points to a

183

compact packaging of the subunit proteins in the filament structure. Extensive digestion of the filaments with trypsin or other proteases, however, completely destroyed them, demonstrating their proteinaceous nature. The variability in structure and probably functionality of the 10-nm filaments in different parts of the cell may be intelligible on the basis of the organization of the cytoplasm in the perinuclear region and in the basal part of the stalk. In thin-section electron microscopy, the nucleus appears surrounded by a thin, organelle-free rim of cytoplasm which is connected via cytoplasmic bridges to a layer of numerous, electron-dense, perinuclear bodies. Based on whole-mount [10] and thin-section [22, 23] electron-microscopic observations, the smooth, perinuclear 10-nm filaments occupying the cytoplasmic channels have been suggested to provide a communication route between the nucleus and the perinuclear bodies. As further demonstrated in the present study, the perinuclear 10-nm filament network extends into the interior of the perinuclear dense bodies and then, farther along, into the cytoplasm of the stalk and rhizoid, except that now it is decorated with globular structures. Since the perinuclear dense bodies harbor large quantities of nucleoprotein material with cytochemical properties of chromatin [24], it is reasonable to assume that the globular structures which are associated with the 10-nm filaments emerging from the perinuclear bodies are RNP particles. This assumption is substantiated by the results of proteinase K digestion and gold–RNase A binding studies which have shown that the globules contain protein as well as RNA. Because of RNase treatment of the cytoplasmic specimens during their processing for electron microscopy, the particles have presumably lost a substantial fraction of their RNA moiety. The question arises of why the filaments interconnecting the perinuclear dense bodies with the nucleus in whole-mount preparations are virtually free of particles whereas those within the bodies and the stalk and rhizoid cytoplasm are heavily loaded with them. On the assumption that the perinuclear bodies serve as a first reservoir and distributor of RNP particles, they must be continuously supplied with RNP material by the nucleus. Since by fluorescence microscopy a large amount of actin has been found in the perinuclear area and coalignment of actin cables and 10-nm filaments

FIG. 9. Transmission electron microscope image of an embedment-free section through detergent-extracted and RNase A-digested stalk cytoplasm of A. acetabulum. A dense filament system interlaces the cytoplasm and contacts large organelles (arrow points to a residual chloroplast). Bar: 500 nm. FIG. 10. Transmission electron microscope image of an embedment-free section through detergent-extracted and RNase A-digested stalk cytoplasm of A. acetabulum. The filaments are densely covered with globular structures although there are always sections which are devoid of particles (arrow). Bar, 100 nm. FIG. 11. Whole-mount preparation of the stalk cytoplasm of A. acetabulum after proteinase K and RNase A digestion. Most of the filaments have a diameter of 10 nm and are free of the globular particles. Bar, 50 nm.

AID

ECR 3924

/

6i2f$$$443

04-08-98 06:55:10

eca

184

AID

BERGER, WITTKE, AND TRAUB

ECR 3924

/

6i2f$$3924

04-08-98 06:55:10

eca

TEN-NANOMETER FILAMENTS IN DASYCLADALES

is seen in the stalk cytoplasm (unpublished observations), it is conceivable that the 10-nm filaments play a supportive role in the microfilament-based transport of RNPs from the nucleus to the perinuclear bodies and, to a lesser extent, in the stalk. As the RNP particles must be highly mobile along the filament tracks for the transport function, their binding to the filaments is relatively weak and they might therefore have been washed out together with actin during preparation of the whole-mount specimens. Not infrequently, particles bound to the smooth, perinuclear 10-nm filaments could be observed. Beyond that, to cope with the heavy traffic between the nucleus and the perinuclear bodies, the filaments should not be obstructed by tightly bound, bulky material. Moreover, the perinuclear area is a zone of extensive protein synthesis providing the nucleus with structural proteins and proteins for the assembly of ribosomes and mRNP particles constantly exported into the cytoplasm. Consequently, most of the constituents of the protein-synthesizing machinery are assembled into polyribosomes which in turn are associated with microfilaments [25–27]. Since the microfilaments are destroyed during whole-mount preparation, the polyribosomes are also solubilized and removed from the remaining cytoskeleton. These circumstances also apply to the apical tip of the cell where because of active protein synthesis the 10-nm filament network is deprived of RNP particles as well and probably supports microfilament-based transport. In contrast to their transport-related functions in the perinuclear area and the apical tip, in the perinuclear dense bodies and the stalk segment between these and the tip of the cell the 10-nm filaments are proposed to largely fulfill a storage function. During the vegetative phase, considerable quantities of RNP material have to be stored with somewhat greater stability at specific sites in the cytoplasm. This suggestion is based on the facts that young Dasycladales cells from which the rhizoid containing the nucleus has been amputated are still capable of species-specific morphogenesis and that different fragments of a cell have different morphogenetic potentials [12]. Consequently, genetic information in the form of mRNPs must be differentially distributed and stored in the cytoplasm and ribosomes for its translational expression must also be present in sufficient amounts. The 10-nm filament system might play an important part in this scenario. Support for this notion comes from the reactivity of the rough but not the smooth 10-nm filaments with an antibody raised

185

against the Xenopus protein Xp54 which in Xenopus oocytes is an integral component of stored mRNP particles and which is highly conserved from yeast to man [18]. Furthermore, in contrast to the actin filament system, the 10-nm filaments may exhibit great structural variability as a consequence of changes in subunit composition and posttranslational modification of the subunit proteins and also as a prerequisite for differential storage of RNP particles. It is probably both of these factors which distinguish the rough from the smooth 10-nm filaments. The major transition between both forms seems to take place in the perinuclear dense bodies, but posttranslational modification of the components to be transported or deposited on the 10-nm filaments may also contribute to their differential distribution. On the basis of this working hypothesis, present and future studies focus on the molecular characterization of the subunit proteins of the smooth and rough 10nm filaments in order to determine whether the two systems indeed have different subunit compositions or whether their different capacities to interact with cytoplasmic components rest on differential posttranslational modification of their protein subunits. They will also provide an answer to the question of whether the 10-nm filaments of Dasycladales cells are close relatives of the IFs of animal cells, with which they share the same morphology [10], or whether they represent a new class of filaments. Moreover, these investigations will necessarily have to include the biochemical characterization of the globular structures associated with the 10-nm filaments in the perinuclear dense bodies and in the cytoplasm of the rhizoid and stalk. Because in animal cells prosomes, facultative RNP particles associated with untranslated mRNPs, have been found to follow the distribution of IFs [28] and such particles also have been detected in a variety of higher plant cells [29], much attention will be given to their detection in Dasycladales cells and their association with the 10nm filaments. Finally, with respect to a possible involvement of the 10-nm filaments in transport processes, their distribution in comparison with that of actin filaments and myosin is of paramount interest. We thank Mrs. Wilfriede Menrad and Mr. Klaus Burger for providing us with cultured cells, Mrs. Dagmar Massoth and Ms. Sabine Gru¨b for excellent technical assistance, Mrs. Annegret Gawenda for the photographs, and Mrs. Yvonne Klein for secretarial work. We are also grateful to Dr. Robert L. Shoeman for critically reading the manuscript.

FIG. 12. Whole-mount electron microscopy of the stalk cytoplasm of A. acetabulum. The specimen was incubated with gold-conjugated RNase A. The complex binds to a large number of the globular structures attached to the 10-nm filaments (A and B). It never binds to the filaments themselves (B). Bars, 100 nm. FIG. 13. Whole-mount preparation of stalk cytoplasm and decoration of the rough 10-nm filaments with an antibody against Xp54, which in Xenopus oocytes is an integral component of stored mRNP particles [18]. Bar, 110 nm.

AID

ECR 3924

/

6i2f$$$443

04-08-98 06:55:10

eca

186

BERGER, WITTKE, AND TRAUB

REFERENCES 1. Traub, P. (1985). ‘‘Intermediate Filaments. A Review,’’ Springer-Verlag, Berlin/Heidelberg/New York/Tokyo. 2. Franke, W. W. (1987) Cell 48, 3–4. 3. Steinert, P. M., and Roop, D. R. (1988) Annu. Rev. Biochem. 57, 593–625. 4. Klymkowsky, M. W., Bachant, J. B., and Domingo, A. (1989) Cell Motil. Cytoskeleton 14, 309–331. 5. Albers, K., and Fuchs, E. (1992) Int. Rev. Cytol. 134, 243–279. 6. Fuchs, E., and Weber, K. (1994) Annu. Rev. Biochem. 63, 345– 382. 7. Quinlan, R., Hutchinson, C., and Lane, B. (1994) in ‘‘Protein Profile’’ (P. Sheterline, Ed.), Vol. 1, Issue 8. 8. Skalli, O., and Goldman, R. D. (1991) Cell Motil. Cytoskeleton 19, 67–79. 9. Traub, P., and Shoeman, R. L. (1994) Int. Rev. Cytol. 154, 1– 103. 10. Berger, S., Shoeman, R. L., and Traub, P. (1996) Protoplasma 190, 204–220. 11. Menzel, D. (1992) Bot. Acta 106, 294–300. 12. Berger, S., and Kaever, M. J. (1992). ‘‘Dasycladales, an Illustrated Monograph of a Fascinating Algal Order,’’ Thieme, Stuttgart. 13. Fey, E. G., Krochmalnic, G., and Penman, S. (1986) J. Cell Biol. 102, 1654–1665. 14. Burke, B., Tooze, J., and Warren, G. (1983) EMBO J. 2, 361– 367.

15. Pruss, R. M., Mirsky, R., Raff, M. C., Thorpe, R., Dowding, A. J., and Anderton, B. H. (1981) Cell 27, 419–428. 16. Giese, G., and Traub, P. (1986) Eur. J. Cell Biol. 40, 266–274. 17. Traub, P., Scherbarth, A., Willingale-Theune, J., Paulin-Levasseur, M., and Shoeman, R. L. (1988) Eur. J. Cell Biol. 46, 478– 490. 18. Ladomery, M., Wade, E., and Sommerville, J. (1997) Nucleic Acids Res. 25, 965–973. 19. Capco, D. G., Krochmalnic, G., and Penman, S. (1984) J. Cell Biol. 98, 1878–1885. 20. Valentine, R. C., Shapiro, B. M., and Stadtman, E. R. (1968) Biochemistry 7, 2143–2152. 21. Steinert, P. M., Rice, R. H., Roop, D. R., Trus, B. L., and Steven, A. C. (1983) Nature 302, 794–800. 22. Franke, W. W., Berger, S., Falk, H., Spring, H., Scheer, U., Herth, W., Trendelenburg, M. F., and Schweiger, H. G. (1974) Protoplasma 82, 249–282. 23. Berger, S., and Schweiger, H. G. (1975) in ‘‘Molecular Biology of Nucleocytoplasmic Relationships’’ (S. Puiseux-Dao, Ed.), pp. 243–250, Elsevier, Amsterdam. 24. Berger, S., and Schweiger, H. G. (1986) J. Cell Sci. 80, 1–11. 25. Pryme, I. F., Johannessen, A., and Vedeler, A. (1995) Adv. Mol. Cell. Biol. 12, 41–74. 26. Hesketh, J. E. (1996) Exp. Cell Res. 225, 219–236. 27. Zak, E. A., Sokolov, O. I., Greengauz, O. K., Bocharova, M. A., and Klyachko, N. L. (1997) J. Exp. Bot. 48, 1019–1026. 28. Scherrer, K., and Bey, F. (1994) Prog. Nucleic Acid Res. 49, 1– 64. 29. Schliephacke, M., Kremp, A., Schmid, H.-P., Ko¨hler, K., and Kull, U. (1991) Eur. J. Cell Biol. 55, 114–121.

Received December 3, 1997

AID

ECR 3924

/

6i2f$$$444

04-08-98 06:55:10

eca