INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Actin-binding proteins of the superfamily of spectrins are recognized as ubiquitous proteins present in all animal and in plant cells. The high M_r proteins are largely -helical, possess actin-binding sites, interact with ligands such as phosphatidylinositol-4,5-bis-phosphate, and bind Ca²⁺ and calmodulin. The multiple functions of spectrins include signal transduction by interacting with integrins and other receptors (Burridge et al., 1988), the regulation of cell shape, mechanical properties, and regulation of the Golgi structure, and vesicle trafficking (Bennett and Gilligan, 1993; Holleran and Holzbauer, 1998; De Matteis and Morrow, 2000).

In animal cells spectrins are found in association with the plasma membrane and internal membrane compartments including the Golgi apparatus and lysosomes (Beck et al., 1994, 1997; Devarajan et al., 1996; Holleran et al., 1996; Hoock et al., 1997; De Matteis and Morrow, 2000). In erythrocytes, self-associating spectrin tetramers of two - and two -subunits form a flexible submembrane skeleton by connecting short actin filaments and linking them to integral membrane proteins by direct interactions or mediated by ankyrin and other proteins (e.g. Bennett and Gilligan, 1993).

Information on the distribution and function of spectrins in plant cells, recently reviewed by De Ruijter and Emons (1999), is sparse. Spectrin-like epitopes were localized mainly at the plasma membrane in several plant species and different cell types (Michaud et al., 1991; Wang and Yan, 1991; De Ruijter and Emons, 1993). Faraday and Spanswick (1993) found a 230-kD protein that cross-reacted with antibodies against human spectrin in plasma membrane-enriched microsome fractions from rice roots. Microinjection of fluorescing monoclonal anti-human spectrin revealed a punctuated labeling pattern associated with the endoplasmic reticulum (ER) and the periphery of epidermal cells of onion bulb scales (Reuzeau et al., 1997). By immunogold labeling, Holzinger et al. (1999) demonstrated spectrin-like epitopes at different types of secretory vesicles, dictyosomes, and plasma membrane and suggested that spectrins may be functional in exocytosis. Spectrin-like epitopes were also located in nuclei of various plant tissues (De Ruijter et al., 2000).

In unipolar cells, labeling of spectrin-like proteins with antibodies raised against animal erythrocyte spectrin was found at the plasma membrane in the growing tips of root hairs (Miller et al., 1997; De Ruijter et al., 1998), pollen tubes (Derksen et al., 1995) and fungal hyphae (Kaminskyi and Heath, 1995). Because labeling was absent from tips of non-growing root hairs (De Ruijter et al., 1998), it was proposed that an apical actin/spectrin array may be involved in the stabilization of the rapidly extending plasma membrane and the cell wall at the tip, the site of exocytosis.

In this study the localization of spectrin-like epitopes was studied in tip-growing characean rhizoids and protonemata, two very similar, but oppositely gravitropically responding cell types. Labeling of spectrin-like proteins is reported to be associated with a distinct actin-organized subdomain of endoplasmic membranes, the ER aggregate in the center of the Spitzenkörper, first identified in rhizoids as a spherical cytoplasmic clear zone by using differential interference contrast (Bartnik and Sievers, 1988). The ER aggregate is tightly associated with a dense apical actin array (Braun and Wasteneys, 1998a) and is present in tip-growing cells, but disappears when tip growth terminates (Bartnik et al., 1990; Braun and Sievers, 1993). The ER aggregate represents the structural center of the Spitzenkörper and is suggested to be essential for controlling the tip-high calcium gradient (Braun and Richter, 1999). Furthermore, it may play a crucial role in the mechanism of tip growth and the regulation of the statoliths-induced reorientation of positively gravitropic rhizoids and negatively gravitropic protonemata (Braun, 1996b, 1997; Sievers et al., 1996). Based on the localization of spectrin-like epitopes in gravistimulated or drug-treated cells, conclusions are drawn on the possible function of spectrin-like proteins in the mechanism of gravity-oriented tip growth.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Immunochemical Detection of Spectrin-Like Proteins

Commercially available human erythrocyte spectrin, as reference, was blotted and recognized by anti-human spectrin (S1515) and anti-chicken spectrin (S-1390) at molecular masses of 220 and 240 kD (Fig. 1, A and B). The staining of the 220-kD band with anti-chicken spectrin is only very faintly visible (Fig. 1B). Extracts of Chara rhizoids contained total protein concentrations of 2 to 3 µg µL¹. Immunodetection of spectrin-like proteins with antibodies against spectrins from chicken and human erythrocytes resulted in a staining of bands at about 170 and 195 kD (Fig. 1, C-E). Staining of the bands was slightly improved by adding 0.1% (w/v) Triton X-100 to the homogenization buffer (Fig. 1, compare D with E). The anti-human erythrocyte spectrin antibody recognized an additional band at 110 kD (Fig. 1C), which are probably breakdown products. Human spectrin, however, shows fragments at about 70 kD and trypsin digestion also produces fragments of the same molecular mass (Speicher et al., 1980; Holzinger et al., 1999). Increasing the concentration of protease inhibitors did not result in a noticeable reduction of 110-kD proteins on the blot. Pre-absorbing anti-chicken and anti-human spectrin strongly reduced staining of the two bands (only the first shown in Fig. 1F).

View larger version (64K): [in this window] [in a new window]   Figure 1.   Western blots (7.5% [w/v] SDS gel) of human erythrocyte spectrin (A and B) and a protein extract of Chara rhizoids (C-F). For immunodetection, antibodies raised against human erythrocyte spectrin (A and C) and chicken erythrocyte spectrin (B and D-F) were used. In A and B, the antibodies recognized - and -spectrin at 220 and 240 kD, whereas in the rhizoid extract, the anti-spectrin antibodies detected bands at 170 and 195 kD. The use of homogenization buffer that contained 0.1% (w/v) Triton X-100 (D) resulted in a slightly stronger staining of the bands (compare with E). Pre-absorbing anti-chicken antibodies with human spectrin strongly reduced staining of the bands (F).

Immunofluorescent Localization of Spectrin-Like Epitopes

Antiserum raised against spectrin from chicken erythrocytes localized spectrin-like epitopes in Chara rhizoids and protonemata (Fig. 2). Confocal immunofluorescence microscopy of freeze-shattered cells that were chemically fixed in normal vertical orientation (non-gravistimulated cells) revealed an intense, spherically-shaped labeling pattern symmetrically positioned in the apical dome close to the cell tips, but no labeling at the plasma membrane. The basal zone with the large vacuole and the subapical cytoplasmic zone containing the large nucleus and numerous organelles gave only weak, non-specific background fluorescence. A polyclonal antibody and two monoclonal antibodies raised against spectrin from human erythrocytes gave no specific immunofluorescence labeling. Double labeling with antibodies against actin and spectrin from chicken erythrocytes showed that the spherical spectrin-fluorescence pattern spatially coincided with the position of the apical, dense actin array (Figs. 2, A-C, E-G) and the cytoplasmic area, which excludes organelles and vesicles, but contains the ER aggregate in the center of the Spitzenkörper. It, therefore, appears as a spherical clear zone in the differential interference contrast image (compare Figs. 2, D and H with 5A).

View larger version (93K): [in this window] [in a new window]   Figure 2.   Localization of actin and spectrin-like epitopes in the apex of a Chara rhizoid (A-D) and a Chara protonema (E-H) by immunofluorescence double labeling. A and E, Labeling with anti-actin; B and F, labeling with anti-chicken spectrin; C and G, overlay of A and B, E and F, respectively. The position of the ER aggregate indicated by the spherical clear zone in the corresponding differential interference contrast (DIC) image (D and H) is marked by the arrows in each image. Optical section images of 1-µm thickness. Bar = 5 µm.

Several controls were performed to test the specificity of the spectrin antibody and to demonstrate the correlation between the labeling of spectrin-like proteins, the actin organization, and the presence of the ER aggregate. In cells that were double-labeled with anti-actin and anti-spectrin pre-absorbed with human spectrin as antigen, the typical actin labeling (Fig. 3A) and the cytoplasmic clear zone of the ER aggregate (Fig. 3C) were present, but only very faint labeling of spectrin-like proteins (Fig. 3B) was found. Replacing anti-spectrin with normal rabbit serum produced no fluorescence labeling (Fig. 3E).

View larger version (118K): [in this window] [in a new window]   Figure 3.   A through C, Immunofluorescence double labeling of a rhizoid with anti-actin (A) and with immunodepleted anti-spectrin (B). The position of the ER aggregate (arrows) is recognizable as a spherical clear zone in the DIC image (C) and in the form of a dense actin array (A), but is only faintly visualized by immunodepleted anti-spectrin (B). Projections of five serial images taken at 0.8-µm z-steps. D through F, Immunofluorescence labeling of a rhizoid with anti-actin (D) and with rabbit serum replacing anti-spectrin (E). The position of the ER aggregate (arrows) is recognizable in the DIC image (F) and in D, but no specific labeling was produced by the rabbit serum (E). Projections of three serial images taken at 0.8-µm z-steps. G through I, Imunofluorescence labeling of a rhizoid with anti-actin (G) and anti-spectrin (H) after application of 10 µM cytochalasin D for 15 min. The ER aggregate has become disintegrated and is no longer recognizable in the DIC image (I), actin microfilament bundles are strongly fragmented (G), and no spectrin-like epitopes are labeled (H). Projections of five serial images taken at 1-µm z-steps. J through L, Immunofluorescence labeling of a rhizoid, which was treated with 2 µM A23187 for 10 min, with anti-actin (J) and anti-spectrin (K). The ER aggregate is not visible in the DIC image (L). Actin microfilaments form thick, randomly oriented bundles in the apex and appear fragmented in the subapical zone (J). Spectrin-like epitopes are not detected (K). Projections of five serial images taken at 1-µm z-steps. Bars = 5 µm.

Tip Growth Correlates with the Presence of Spectrin-Like Epitopes

The results of the experiments were identical in rhizoids and protonemata and, therefore, the presented images of rhizoids are representative for both cell types. Cytochalasin D was used to destroy the complexly organized actin microfilament system in the apices of rhizoids and protonemata. The fragmentation and eventual complete depolymerization of the actin microfilaments (Fig. 3G) resulted in the disintegration of the apical aggregation of ER membranes (Fig. 3I) concurrent with a rapid termination of tip growth within about 5 min (n = 48). Spectrin-like epitopes were not detected in these cells (Fig. 3H). The resumption of tip-growth activity in 75% of the cells after removal of the inhibitor, however, was accompanied by the reorganization of the actin cytoskeleton and the reappearance of the ER aggregate. Then, the spherical labeling of spectrin-like proteins could again be demonstrated (not shown).

Spectrin-like epitopes were also absent in the apices of cells whose tip-high gradient of cytoplasmic calcium had been disturbed by the calcium ionophore A23187 (n = 56). Application of 2 µM A23187 for 10 min resulted in a rapid termination of tip growth within 2 to 5 min and the disappearance of the ER aggregate (compare Figs. 3L with 5). In contrast to cytochalasin D, the ionophore did not cause a complete breakdown of the actin system, but resulted in a major reorganization and bundling into randomly oriented actin microfilaments in the apical zone and partial fragmentation in the subapical zone (Fig. 3J). Spectrin-like epitopes were not immunofluorescently localized in these non-growing cells (Figs. 3K and 4B). However, 1 to 2 h after removal of A23187, the cell tips increased in diameter and, subsequently, tip-growth activity was resumed in about 70% of the cells. In the newly forming tip, which grew out with its original diameter, the actin microfilaments became refocused in that area of the apical dome where the ER aggregate had reassembled by then (Fig. 4D') and where the spectrin epitopes gradually reappeared (Fig. 4, C and D). The distribution of spectrin-like epitopes and the corresponding growth rates prior to and after treating rhizoids with 2 µM A23187 is summarized in Figure 4.

View larger version (54K): [in this window] [in a new window]   Figure 4.   Graph showing the rates of elongation growth of a representative Chara rhizoid prior to and after incubation with 2 µM A23187 for 10 min (area of lighter gray color) and the corresponding spectrin-immunolabeling images (A-D). The result of spectrin immunolabeling is demonstrated before (A) and 30 min after the treatment (B). The reappearance of spectrin-like epitopes (C and D) and the reorganization of the actin cytoskeleton (D') is shown during the formation (C) and outgrowth of the new tip (D and D') after resumption of tip-growth activity. Spectrin fluorescence reappears in the form of a small patch close to the apical membrane, and later resumes its original position and size in the center of the Spitzenkörper. Projections of six serial images taken at 1-µm z-steps.

Electron microscopic examination of rhizoids and protonemata confirmed that the ER aggregate was present only in actively tip-growing cells (Fig. 5A), but disappeared in cells that had stopped tip growth after cytochalasin treatment (see Bartnik and Sievers, 1988) or the application of A23187 (Fig. 5B). In Figure 5B, the highly organized aggregation of ER membranes in the rhizoid tip is replaced by a loose arrangement of randomly oriented cisternae after tip growth was stopped by the application of 2 µM A23187. In some apices of non-growing cells ER membranes were completely missing (not shown).

View larger version (99K): [in this window] [in a new window]   Figure 5.   Electron microscopic images of the apex of an untreated (A) and a A23187-treated Chara rhizoid (B). In the actively tip-growing cell (A), the ER aggregate (arrows) is located in the center of the Spitzenkörper with its abundant vesicles. Incubation with 2 µM A23187 for 15 min caused a dispersion of the ER aggregate and resulted in a random distribution of ER membranes in the rhizoid apex. Bars = 5 µm.

Spectrin Immunolocalization in Gravistimulated Cells

To initiate the oppositely gravitropic responses in rhizoids and protonemata, cells were rotated to a horizontal position for approximately 15 min prior to fixation and immunolabeling. The gravitropic response is about to start after that time in both cell types. The shape and signal intensity of the spherical anti-spectrin fluorescence array remained unchanged, but its localization differed considerably at the beginning of the graviresponse in both cell types. In rhizoids that had just started downward bending after 15 min in a horizontal position, the apical fluorescence labeling was still localized in a symmetrical position in the bending tip of all successfully labeled cells (n = 15; Fig. 6A) and kept that stable position during all phases of the graviresponse.

View larger version (76K): [in this window] [in a new window]   Figure 6.   Localization of spectrin-like proteins in a Chara rhizoid (A) and a Chara protonema (B) at the beginning of the opposite graviresponses after 15 min in a horizontal position. A, In the rhizoid, the labeling of spectrin-like proteins, indicating the position of the ER aggregate, is still located close to the growth center at the tip. B, In the protonema, the labeling of spectrin-like proteins is clearly displaced toward the upper flank where the future outgrowth starts with the formation of a bulge. Broken lines outline the outermost tip region and indicate the median line of the cells. Optical section images of 1.2-µm thickness. Bars = 5 µm.

In the representative protonema shown in Figure 6B, however, the fluorescing array was asymmetrically positioned, clearly displaced toward the upper flank, already before the cell had started to bend upward (n = 9). The displaced spectrin-fluorescence array pointed to that site of the upper flank representing the site of future outgrowth that starts with the appearance of a bulge. During the first drastic upward shift of the protonema tip, the spherical spectrin labeling returned into a symmetrical position in the apical dome and remained there during the later stages of gravitropic curvature characterized by much slower bending rates.

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

The ER aggregate is a unique structure symmetrically positioned in the apices of tip-growing rhizoids and protonemata of characean algae, two very similar, but oppositely graviresponding cell types. The distinctive, spherically shaped subdomain of ER was first demonstrated by differential interference contrast and electron microscopy in the apex of growing cells (Bartnik and Sievers, 1988). It was interpreted as the structural center of the vesicle-rich Spitzenkörper involved in vesicle guidance and the control of exocytosis at the tip (Bartnik et al., 1990). Previous inhibitor studies and cytoskeleton staining have shown that actin microfilaments, but not microtubules (Braun and Sievers, 1994), are tightly associated with the ER aggregate (Sievers et al., 1991; Braun and Wasteneys, 1998a). Cytochalasin-induced disruption of the actin cytoskeleton was reported to cause the disappearance of the ER aggregate and to inhibit tip growth (Bartnik and Sievers, 1988; Braun and Sievers, 1993). The actin cytoskeleton is complexly organized in the apex of both cell types; numerous fine actin microfilaments focus in the central actin-rich area that is occupied by the ER aggregate (Braun and Wasteneys, 1998a, 1998b). The apical actin organization was generally less well preserved in immunolabeled cells as compared with cells stained with fluorescently conjugated phalloidin (Braun and Wasteneys, 1998a, 1998b). The fine actin bundles that run from the central actin array toward the apical plasma membrane are scarcely visualized by immunofluorescence labeling.

Freeze shattering and immunofluorescence double-labeling using anti-actin and antiserum raised against spectrin from chicken erythrocytes produced a prominent, spherically shaped fluorescence pattern that coincided strongly with the dense actin array and the position of the ER aggregate in the apices of rhizoids and protonemata. These findings provide evidence that a spectrin-like protein is a major component of the ER aggregate-associated cytoskeleton. Furthermore, it was demonstrated that labeling of spectrin-like proteins was absent after cytochalasin-induced disruption of the actin cytoskeleton, which caused the disappearance of the ER aggregate and termination of tip growth. It must be concluded from these results that the structural integrity of the ER aggregate and the presence of spectrin-like epitopes strongly depend on an intact actin microfilament system.

Association of spectrin with specific subcompartments of the ER is already known from animal cells. The submicrovillar network of ER membranes of honeybee photoreceptor cells is associated with actin and also shows distinct -spectrin immunolabeling, suggesting a role in stabilizing and maintaining this functional endomembrane subregion (Baumann, 1998). There is evidence that different isoforms of spectrin and ankyrin are also associated with the ER of neuronal cells (Zagon et al., 1986; Malchiodi-Albedi et al., 1993) and other endomembranes such as the Golgi apparatus and lysosomes (Beck et al., 1994, 1997; Devarajan et al., 1996; Holleran et al., 1996; Hoock et al., 1997; De Matteis and Morrow, 2000).

A tip-high gradient of cytoplasmic free calcium has recently been shown to be a prerequisite for tip growth in characean rhizoids and protonemata (Braun and Richter, 1999), as was reported for most other tip-growing cell types (for review, see Sanders et al., 1999 and refs. therein). Interfering with the calcium gradient of rhizoids and protonemata by the application of the calcium ionophore A23187 also caused the disintegration of the ER aggregate and prevented labeling of spectrin-like proteins, as was described for cytochalasin D treatment. By disturbing the gradient of cytoplasmic free calcium, the complex actin microfilament system was not completely destroyed, but reorganized to randomly oriented, thick actin bundles. With the resumption of tip-growth activity and the formation of a new tip, spectrin-like epitopes gradually reassembled concurrently with the reformation of the actin arrangement and the ER aggregate. These results indicate that tip growth is closely correlated with the structural integrity of the ER aggregate that is organized by an actin/spectrin scaffold that in turn appears to be regulated by the calcium gradient. Effects of A23187 on gravitropic tip growth and the ultrastructure of rhizoids and protonemata will be presented in more detail elsewhere.

In other tip-growing plant cells such as root hairs (De Ruijter and Emons, 1998), pollen tubes (Derksen et al., 1995), and fungal hyphae (Kaminskyi and Heath, 1995), spectrin-like epitopes have been reported to be mainly associated with the apical plasma membrane where the spectrin-associated cytoskeleton may stabilize the rapidly expanding area of exocytosis. These results are not in accordance with the results presented in this study; labeling of spectrin-like proteins was not detected at the plasma membrane. However, an ER aggregate as prominent as in characean rhizoids and protonemata has not been described in any other tip-growing cell type so far. In less polarly organized plant cells, i.e. protoplasts, embryonic cells, epidermal and root-tip cells, and others, spectrin immunofluorescence was predominantly localized at the plasma membrane, but also at the periphery of plastids and in nuclei (De Ruijter and Emons, 1993; De Ruijter et al., 2000), in certain types of secretory vesicles and dictyosomes (Holzinger et al., 1999), as well as covisualized with other endomembranes (Reuzeau et al., 1997). The function of spectrin in plant cells, however, remains unclear; stabilization of the plasma membrane and a possible function in the exocytotic process have been discussed.

The initiation of the positive gravitropic responses of characean rhizoids and the negative gravitropic response of protonemata were shown to depend on the sedimentation of statoliths (for review, see Braun, 1997). The actomyosin system plays an important, but different role in the positioning and the sedimentation of the statoliths in both cell types (Buchen et al., 1993; Braun, 1996a, 1996b). It was recently demonstrated that the negatively gravitropic upward bending of protonemata following statolith sedimentation involves a repositioning of calcium channels and the calcium gradient toward the upper flank (Braun and Richter, 1999). According to a hypothetical model, this results in the reorientation of the growth direction by a displacement of the Spitzenkörper and, thus, a shift of the growth center from the very tip to the upper flank, which may be mediated by actin and the activity of calcium-dependent, actin-binding proteins. This hypothesis is strongly supported by the observation that the spherical spectrin-fluorescence pattern that represents the position of the ER aggregate was drastically displaced toward the upper flank during the initiation of the graviresponse of protonemata, indicating the new direction of growth (Hodick, 1994). In rhizoids, however, there is evidence from centrifugation experiments (Braun, 1996a; Hodick and Sievers, 1998) that the actin-mediated anchorage of the Spitzenkörper, including its central ER aggregate, at the tip is more stable than in protonemata. As a consequence, statolith sedimentation to the physically lower cell flank results in differential flank growth due to different growth rates on the opposite subapical flanks of the apical dome, but does not cause a major repositioning of the Spitzenkörper. In accordance with this, the gradient of cytoplasmic free calcium and calcium channels (Braun and Richter, 1999) and the spherical spectrin-fluorescence labeling (this study) in gravistimulated rhizoids was always found to be positioned symmetrically in the apical dome.

In addition to the stabilizing and cell-shaping function of the submembraneous spectrin meshworks originally identified in erythrocytes (Bretscher, 1991; Bennett and Gilligan, 1993), spectrin-like molecules are involved in a recruiting system for integral membrane proteins (Devarajan and Morrow, 1996; De Matteis and Morrow, 2000), thus, creating a discrete functional subdomain with a specific set of proteins such as receptors, channel proteins, and ATPases. In the apex of characean rhizoids and protonemata, the presence of the ER aggregate is intimately correlated with active tip growth. Therefore, the ER aggregate may represent a discrete functional subcompartment required for the establishment and maintenance of the complex mechanism and the physiological environment for gravitropic tip growth. The results indicate that proteins are recognized that may have similar functions as animal spectrins; however, the findings that spectrin-like epitopes were detected by an antiserum raised against spectrin from chicken erythrocytes, but not by antibodies against human-erythrocyte spectrin; that immunodepletion of the diluted antibody with human spectrin as antigen still resulted in a faint labeling in most cells; and that spectrin-like proteins in rhizoid extracts exhibited bands at about 170 and 195 kD, which is remarkably lower than the 220 and 240 kD of - and -spectrins, indicate that they may differ considerably from animal spectrins. A spectrin-like protein with lower molecular masses (about 206 kD and lower) have already been reported in the single celled alga Chlamydomonas reinhardtii (Lorenz et al., 1995). Molecular analysis of the spectrin-like protein in characean algae is in progress. Investigating binding properties and other functional studies would help to further characterize the protein.

In conclusion, spectrin-like proteins codistribute with the actin-organized ER aggregate in the apex of characean rhizoids and protonemata, indicating that they play a role in anchoring and maintaining the structural organization providing mechanical stability to this distinct ER subdomain. In addition, these proteins may also provide a mechanism for the recruitment of specific membrane proteins determining the characteristic functions that have been discussed for the ER aggregate, i.e. the control of the calcium homeostasis and the regulation of the oppositely gravitropic tip growth in rhizoids and protonemata.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Plant Material

Young thalli of Chara globularis Thuill. were collected from a pond (Botanischer Garten, Universität Bonn, Bonn) and cut into short segments. To induce formation of rhizoids, the side branches originating from each node were cut and the shoot segments were placed in culture chambers containing Forsberg medium modified after Wasteneys et al. (1996). Rhizoids developed after 3 to 5 d at room temperature under continuous illumination at 150 to 200 µmol m² s¹. For the production of protonemata, shoot segments were embedded in a thin layer of agar (1.2% [w/v] in distilled water) on a microscope slide, and were covered with long coverslips. These cuvettes were placed in staining jars filled with modified Forsberg medium. Protonemata developed within 10 to 20 d in darkness at room temperature (22°C).

Preparation of Protein Extracts and Immunoblotting

Rhizoids of C. globularis were collected, frozen in liquid nitrogen, and homogenized in a low-salt buffer containing 1.5 mM Tris, 0.5 mM EDTA, 1 mM phenylmethyl-sulfonylfluoride, 1% (w/v) polyvinylpyrrolidone, 2% (w/v) protease inhibitor cocktail (P-9599, Sigma, Deisenhofen, Germany), and 10 mM freshly added dithiothreitol, pH 7.4. The homogenate was centrifuged for 10 min at 20,000g and 4°C to remove cell debris and large organelles. The resulting supernatant was subjected to SDS-PAGE using 7.5% (v/v) mini-slab gels at about 15 µg of protein per lane. Protein concentrations were determined according to Bradford (1976). Gels were wet-blotted onto nitrocellulose, which was used for incubation with antibodies against spectrins from chicken and human erythrocytes (S-1390 and S-1515, Sigma) in a dilution of 1:400 and 1:800 in Tris-buffered saline (15 mM Tris, 150 mM NaCl, and 0.05% [v/v] Tween 20, pH 7.4) plus commercially available milk powder (4% [w/v]; Frema Reform) or 3% (w/v) bovine serum albumin (Sigma) for blocking. For control, pre-absorption of anti-spectrin with human spectrin (S-3644, Sigma) were carried out by incubating 0.5 mL of diluted antibody with 10 µg of human spectrin overnight at 4°C. The immunoprecipitate was spun at 12,000g for 15 min and the supernatant was used for immunolabeling.

Drug Treatment

Rhizoids and protonemata were treated with 2 µM A23187 (Sigma) or 10 µM cytochalasin D (Sigma) diluted with artificial pond water (1 mM NaCl, 0.1 mM KCl, 0.1 mM CaCl₂, and 2 mM HEPES [4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid], pH 7.2) for 10 to 15 min prior to fixation to manipulate the tip-growth activity by interfering with the gradient of cytoplasmic free calcium and with the actin cytoskeleton, respectively. Sequences were recorded on video tapes and growth rates were determined by analyzing photographs taken from the monitor.

Immunofluorescence Labeling and Confocal Microscopy

The freeze-shattering procedure was modified after Braun and Wasteneys (1998a). Rhizoids and protonemata were fixed for 20 min with a freshly prepared fixation solution containing 1% (v/v) formaldehyde, 1% (v/v) glutaraldehyde, PIPES (50 mM 1,4-piperazinediethanesulfonic acid), 5 mM EGTA, and 5 mM MgSO₄, pH 7.2. Cells, which were treated with A23187 or cytochalasin D, were fixed 30 min after removing the inhibitor and 10 and 40 min after resumption of tip growth was observed. After several rinses in fixation buffer without aldehydes, the buffer was gradually replaced with phosphate-buffered saline (PBS: 137 mM NaCl, 2.7 mM KCl, 4.9 mM Na₂HPO₄, and 1.5 mM KH₂PO₄, pH 7.4), incubated with freshly prepared 1 mg mL¹ NaBH₄, washed again in PBS, and rinsed with PBS containing 50 mM Gly. After permeabilization with 1% (v/v) Triton X-100 in PBS/Gly for 30 min, the cells were transferred to polyethyleneimine-coated microscope slides and plunged into liquid nitrogen for approximately 1 min. Frozen cells were carefully squashed with a second pre-chilled microscope slide. After thawing and washing three times in PBS/Gly, the cells were incubated with polyclonal rabbit anti-chicken spectrin (1:200; S-1390, Sigma), polyclonal rabbit anti-human spectrin (1:200; S-1515, Sigma), monoclonal mouse anti-human spectrin (1:100; S-3396, Sigma), and monoclonal mouse anti-human spectrin (1:100; ICN, Eschwege, Germany) for 2 h at 37°C or overnight at 20°C. After three rinses in the same buffer, the cells were incubated with fluorescein isothiocyanate- (FITC) conjugated goat anti-rabbit (F-9887, Sigma), goat anti-mouse (F-9006, Sigma), and goat anti-mouse (F-9259, Sigma), respectively, for 2 h at 37°C. Stained cells were rinsed three times with PBS and mounted in 0.1% (w/v) para-phenylene diamine and 50% (w/v) glycerol to minimize fading of the fluorescent conjugate.

By partially cracking off the frozen cell wall, the freeze-shattering method facilitates direct access for the antibodies. Although the rate of successfully labeled cells varies, the procedure results in reliable and reproducible staining. The cytoplasmic organization and the cytoskeletal arrangement in most fragments are remarkably stable (Braun and Sievers, 1994; Braun and Wasteneys, 1998a); the use of digesting enzymes is avoided.

For double-labeling of spectrin and actin, cell fragments were sequentially incubated with rabbit anti-chicken spectrin (1:200) for 2 h, with mouse anti-actin (1:400; clone C4; ICN) overnight, with FITC-conjugated goat anti-rabbit (1:200; F-9887, Sigma) for 2 h, and with Alexa 546-conjugated goat anti-mouse (1:100; Molecular Probes, Eugene, OR) for 2 h. Washing was performed after each incubation step. Stained samples were rinsed three times with PBS/Gly and were mounted in the anti-fading solution. Images of immunofluorescently labeled samples were collected using a confocal microscope TCS4D (Leica, Heidelberg) with an argon-krypton ion laser. Images were collected in dual channel mode (excitation 488/568, dichroitic mirror DD488/568, barrier filter BP FITC/LP 590). Digital images were processed with Adobe Photoshop (Adobe Systems, Mountain View, CA) and Corel Draw (Corel Corporation, Dublin) and printed on glossy photo paper with a printer (Stylus Photo 870, Epson, Tokyo). Controls included replacing the first antibody with 1% (v/v) normal rabbit serum and pre-absorbing the antibody with its antigen following the procedure described above.

Preparation for Electron Microscopy

Rhizoids and protonemata were pre-incubated in artificial pond water with and without 2 µM A23187 (Sigma) for 15 min and fixed for 20 min with 2% (v/v) glutaraldehyde in 0.1 M PIPES and 5 mM CaCl₂, pH 7.0. After thoroughly washing with fixation buffer, cells were contrasted for 1 h with 1% (w/v) OsO₄ and 1% (w/v) K₃Fe(CN)₆ in 0.1 M cacodylate and 5 mM CaCl₂. Following washing in the contrasting buffer without OsO₄ and K₃Fe(CN)₆, the samples were dehydrated in an aceton series and infiltrated with epoxy resin (Spurr's resin). Ultrathin sections were post-stained with 2% (w/v) uranyl acetate and 2% (w/v) lead citrate and examined with the transmission electron microscope EM 10 (Zeiss, Oberkochen, Germany).