Involvement of transmembrane proteins in the association of cortical microtubules with the plasma membrane in tobacco BY-2 cells

Cortical microtubules (MTs) in the cells of higher plants are cross-linked with the plasma membrane (Hardham and Gunning, 1978; Marchant, 1978; Doohan and Palevitz, 1980; Lloyd et al. 1980). The orientation of MTs has been reported to be controlled by various factors, such as plant hormones and the developmental state of the cell (Gunning and Hardham, 1982; Hardham, 1982). Although there has been no clarification of the mechanism whereby the orientation of cortical MTs is controlled, it is likely that the. mechanism by which cortical MTs are crosslinked with the plasma membrane plays an important role in the regulation of the orientation of cortical MTs.

Marchant and Hines (1979) reported that cortical MTs in Mougeotia protoplasts were not associated with the plasma membrane for a short period of time after the preparation of the protoplasts. We repeated their experiments with a different result. We found that cortical MTs were associated with the plasma membrane even in the case of freshly prepared Mougeotia protoplasts (Kakimoto and Shibaoka, 1986). It seemed possible that the difference between Marchant and Hines’ result and ours was due to differences in the wall-digesting enzymes used for the preparation of the protoplasts. Since the levels of impurities in commercially available preparations of walldigesting enzymes have decreased from year to year, it is probable that the wall-digesting enzymes used by Marchant and Hines in 1979 contained impurities that had the ability to disrupt the association between cortical MTs and the plasma membrane, while the preparations that we used in 1986 did not contain such impurities. Since proteases seem to be the most likely candidates for such impurities, we examined the effects of proteases on the association of cortical MTs with the plasma membrane in cultured tobacco cells.

Recently, we reported that extracellular materials, such as the cell wall and exogenously applied extensin, increased the cold-stability of cortical MTs in protoplasts isolated from cultured tobacco cells (Akashi et al. 1990). These results led us to advance the hypothesis that the mechanism of interaction between extracellular materials and cortical MTs is associated with some component(s) of the plasma membrane. If such a mechanism is identical with the mechanism by which cortical MTs are crosslinked with the plasma membrane, and if treatment with proteases results in the disruption of these cross-links, treatment with proteases should also interfere with stabilization of cortical MTs by the cell wall or extensin. Therefore, we also examined whether or not pretreatment with proteases eliminates the effect of extensin on the cold-stability of cortical MTs.

Tobacco BY-2 cells (Nicotiana tabacum L. cv. Bright Yellow 2) were used. The cells were cultured in suspension in Linsmaier and Skoog’s medium (LS medium), supplemented with 3% sucrose and 0.2 mg I⁻¹ 2,4-dichlorophenoxyacetic acid (2,4-D), pH 5.8, at 26°C in the dark. Cells were subcultured at intervals of 7 days (Nagata et al. 1981).

Cells subcultured for 4 to 6 days were harvested and incubated for 2 h at 30 °C in a solution that contained 1 % Cellulase Onozuka RS (Yakult Pharmaceutical Co. Ltd, Takarazuka, Hyogo, Japan), 0.1% Pectolyase Y-23 (Seishin Pharmaceutical Co. Ltd, Nagareyama, Chiba, Japan) and 0.4 M mannitol, pH 5.5 (Nagata et al. 1981). The isolated protoplasts were washed with 0.4 M mannitol.

The isolated protoplasts were suspended in 2,4-D medium (LS medium supplemented with 3% sucrose, 0.4 M mannitol, and 0.2mgl⁻¹ 2,4-D) and cultured in Petri dishes (6cm in diameter) in the dark at 26°C for 3 days (Hasezawa and Syono, 1983), during which time the protoplasts regenerated cell walls.

Protoplasts or cells with regenerated cell walls were incubated at 26°C in 20 mM potassium phosphate buffer (KP buffer), pH 6.0, that contained 0.4 M mannitol and was supplemented with 1 to 100mgl⁻¹ trypsin, 10 to 400mgl⁻¹ chymotrypsin, or 40units ml⁻¹ trypsin attached to cross-linked beaded agarose (trypsin beads, 78 units per ml packed gel; Sigma Chemical Co., St Louis, MO, USA). Protoplasts or cells were also incubated in the abovementioned solution supplemented with 500 mg I⁻¹ leupeptin and 25 mg I⁻¹ (p-aminophenyl) methanesulfonyl fluoride (APMSF; Wako Pure Chemical Industries, Ltd, Osaka, Japan) and these cells were used as controls. The proteolytic reaction was stopped by addition to the reaction mixture of 10 volumes of a solution of 0.4 M mannitol that contained 500 mgl⁻¹ leupeptin and 25 mg I⁻¹ APMSF. Trypsin beads were washed with distilled water just before use. Trypsin beads at 40 units ml^−-1 were equivalent to 4.6 mg I⁻¹ trypsin in proteolytic activity.

Membrane ghosts of tobacco BY-2 cells were prepared by the method of Marchant (1978). A small drop of a suspension of protoplasts was placed on a coverslip that had been treated with an aqueous solution of 0.5mgml⁻¹ poly-L-lysine (average M_r, 120000) and protoplasts were allowed to settle for about 2 min. The coverslips with adhering protoplasts were soaked in a MT-stabilizing buffer (25 mM Pipes, pH 6.9, that contained 2 HIM EGTA and 1 mM MgSO₄) for 5 min. This procedure resulted in the bursting of protoplasts, with membrane ghosts remaining on the coverslip.

Protoplasts were suspended in KP buffer, pH 6.0, that contained 0.4 M mannitol alone, 0.4 M mannitol plus 0.1 mg ml⁻¹ extensin or 0.4M mannitol plus 0.1 mg ml⁻¹ poly-L-lysine, and chilled at 0°C for 30 min. Cells with regenerated cell walls were suspended in KP buffer, pH 6.0, that contained 0.4 M mannitol, and chilled at 0°C for 30 min.

MTs were visualized by immunofluorescence staining by the method of Wick et al. (1981). Details of the procedures for staining were described in a previous paper (Akashi et al. 1990).

Extensin isolated from culture filtrates of tobacco XD-6S cells was used. Details of the procedures for isolation were described in a previous paper (Akashi et al. 1990). The amino acids and sugars in extensin from tobacco XD-6S were analyzed by Kawasaki (1989).

Abundant cortical MTs were present both in protoplasts isolated from tobacco BY-2 cells (Fig. 1A) and on membrane ghosts prepared from the protoplasts (Fig. 2A), as judged by immunofluorescence after reaction with a tubulin-specific antibody. A 10-min treatment of protoplasts with 30 mg I^-1 trypsin had no apparent effect on cortical MTs: abundant MTs were present in the trypsin-treated protoplasts (Fig. IB). However, only a few MTs were present on the membrane ghosts prepared from the trypsin-treated protoplasts (Fig. 2B), suggesting that cortical MTs in the trypsin-treated protoplasts were not firmly associated with the plasma membrane. Trypsin at Imgl⁻¹ had no effect. Trypsin at 10 mgl⁻¹ or higher was effective and its effectiveness increased with increases in concentration. A 10-min treatment with 100 mg I⁻¹ trypsin caused bursting of some protoplasts. A 5-min treatment with 30 mg I⁻¹ trypsin was also effective, but the effectiveness varied from protoplast to protoplast. A 30min treatment with 30mgl⁻¹ trypsin, which was only as effective as a 10-min treatment with trypsin at the same concentration in decreasing the number of MTs on the membrane ghosts, caused bursting of some protoplasts.

Inhibitors of proteases eliminated the effects of trypsin. Abundant MTs were present both in protoplasts that had been treated with 30 mg I⁻¹ trypsin together with 500mgl⁻¹ leupeptin and 25mgl⁻¹ APMSF (Fig. 1C) and on the membrane ghosts prepared from protoplasts pretreated in this way (Fig. 2C), suggesting that it is the proteolytic activity of trypsin that disturbs the association between cortical MTs and the plasma membrane.

Chymotrypsin had an effect similar to that of trypsin, but was less effective. At 10 mg I⁻¹, chymotrypsin was not effective. Chymotrypsin at 100 or 400 mg I⁻¹ was as effective as trypsin at 30 mgl⁻¹, but often caused bursting of the protoplasts (data not shown).

Non-permeating proteolytic enzyme was also effective, but the pattern of MTs remaining on the membrane ghosts prepared from protoplasts pretreated with non-permeating enzyme varied from ghost to ghost. A 90-min treatment with 40 units ml⁻¹ trypsin beads caused the dissociation of MTs from the plasma membrane in some protoplasts (Fig. 2D), while it showed no effect in the majority of the protoplasts (Fig. 2E). Usually, the dissociation occurred only in a limited region of the ghost. The lack of uniformity in the pattern of the remaining MTs seems to result from the inability of non-permeating trypsin to make uniform contact with the surface of the protoplast. Trypsin attached to a bead, which is a little larger than the protoplast, can make contact with a protoplast only when the bead happens to attach to the protoplast, and even when the bead attaches to the protoplast, trypsin on the bead can make contact only with a limited region of the surface of the protoplast. A 60-min treatment with 40 units ml⁻¹ trypsin beads was also effective, but the effect of a 30-min treatment with the same activity of trypsin beads was hardly observable. Trypsin beads showed no effect in the presence of 500mgl⁻¹ leupeptin and 25mgl⁻¹ APMSF (data not shown).

As has been reported before (Akashi et al. 1990), exogenously applied extensin protected cortical MTs in tobacco BY-2 protoplasts from disruption by low-temperature treatment. As shown in Figs 3A and 4A, MTs were not present in protoplasts chilled at 0°C for 30 min in the absence of extensin (Fig. 3A), but they were present in protoplasts chilled in the presence of 0.1 mg ml⁻¹ extensin (Fig. 4A). However, extensin did not protect cortical MTs from disruption by low-temperature treatment in protoplasts pretreated with 30 mg I⁻¹ trypsin for 10 min. Cortical MTs in trypsin-treated protoplasts were disrupted by low-temperature treatment both in the absence (Fig. 3B) and in the presence of extensin (Fig. 4B). The effect of trypsin was nullified by the simultaneous application of inhibitors of proteases. Cortical MTs in protoplasts treated with 30 mg I⁻¹ trypsin together with 500 mg I⁻¹ leupeptin and 25 mg I⁻¹ APMSF were not disrupted by low-temperature treatment in the presence of 0.1 mg ml⁻¹ extensin (Fig. 4C), but they were disrupted in the absence of extensin (Fig. 3C). Cortical MTs in protoplasts treated with 100 or 400 mg I⁻¹ chymotrypsin were disrupted by low-temperature treatment both in the absence and in the presence of 0.1 mg ml⁻¹ extensin (data not shown).

The MT-stabilizing effect of poly-L-lysine, which we reported previously (Akashi et al. 1990), was not observed in protoplasts pretreated with 30 mg I⁻¹ trypsin for 10min. In the presence of 0.1 mg ml⁻¹ poly-L-lysine, low-temperature treatment did not cause the disruption of cortical MTs in protoplasts that had not been treated with trypsin (Fig. 5A), but it did cause disruption of MTs in trypsin-pretreated protoplasts (Fig. 5B).

Abundant MTs were present in cells with regenerated cell walls. They were arranged in parallel to one another and ran transversely or almost transversely to the cell’s long axis (Fig. 6A). After low-temperature treatment, cortical MTs were still present, although they seemed to be fragmented and bundled (Fig. 7A). A 10-min treatment with 30 mg I⁻¹ trypsin did not disrupt cortical MTs but often disturbed the parallel arrays of MTs (Fig. 6B). Treatment with trypsin decreased the cold-stability of cortical MTs in cells with regenerated cell walls: no MTs were present in trypsin-pretreated cells after chilling at 0°C for 30 min (Fig. 7B).

Abundant MTs were present both in BY-2 protoplasts and on the membrane ghosts prepared from the protoplasts (Figs IA, 2A), suggesting that cortical MTs in the protoplasts are associated with the plasma membrane. However, only a few MTs or none at all were observed on the membrane ghosts prepared from protease-pretreated protoplasts (Fig. 2B) in which abundant MTs were present (Fig. IB). These results seem to indicate that treatment with proteases disrupts the association of cortical MTs with the plasma membrane. Since treatment with proteases did not disrupt MTs and non-permeating protease also disrupted the association of cortical MTs with the plasma membrane, it is unlikely that proteases enter the protoplasts and directly digest the protein(s) that cross-links MTs and the plasma membrane. Thus, we present the hypothesis that transmembrane protein(s) is involved in the association of cortical MTs with the plasma membrane (Fig. 8A) and that proteases cause the dissociation of MTs from the membrane by digesting the extracellular portion of such transmembrane protein(s) (Fig. 8B). The possibility that cross-bridges between MTs and the plasma membrane are associated with extracellular molecules has already been suggested by Lloyd et ai. (1980).

As has been reported previously (Akashi et al. 1990), exogenously applied extensin and poly-L-lysine increased the cold-stability of cortical MTs in BY-2 protoplasts (Figs 4A, 5A). If exogenously applied substances are to stabilize cortical MTs in the protoplasts, either they must enter the protoplasts or some factor that interacts with extracellular substances and intracellular MTs must be associated with the plasma membrane. Our observation that neither extensin nor poly-L-lysine stabilized MTs in the protease-pretreated protoplasts (Figs 4B, 5B) seems to eliminate the possibility that extensin or poly-L-lysine enters the protoplasts and stabilizes MTs directly, but it does suggest the presence in the plasma membrane of a mediating factor. These results, together with the result that treatment with proteases disrupts the association of cortical MTs with the plasma membrane, seem to indicate the possibility that the system that cross-links cortical MTs and the plasma membrane is identical with the system that mediates interactions between extracellular substances and MTs. We suggest that extensin or poly-L-lysine stabilizes cortical MTs in BY-2 protoplasts by interacting with the extracellular portion of the transmembrane protein(s) that is involved in the association of cortical MTs with the plasma membrane (Fig. 8C).

The parallel arrangement of cortical MTs, which were present in cells with regenerated cell walls (Fig. 6A), was disturbed by treatment with proteases (Fig. 6B). This result indicates the involvement of transmembrane protein(s) or extracellular protein(s) in the maintenance of the parallel array of cortical MTs. It is tempting to speculate that cortical MTs are immobilized by being cross-linked to the cell wall via a cross-bridge-transmembrane protein-extensin system, as illustrated schematically in Fig. 8D. The observation that treatment with proteases decreased the cold-stability of cortical MTs in cells with regenerated cell walls (Fig. 7B) indicates the possibility that this system is also involved in the stabilization of cortical MTs by the cell wall. We reported some years ago that gibberellin increased the coldstability of cortical MTs in onion leaf-sheath cells (Mita and Shibaoka, 1984) but decreased it in dwarf-pea epicotyl cells (Akashi and Shibaoka, 1987) in which abscisic acid increased the cold-stability of MTs (Sakiyama and Shibaoka, 1990). It seems probable that the cross-bridge-transmembrane protein-extensin system is also involved in the mechanism whereby plant hormones increase or decrease the cold-stability of cortical MTs. To examine whether or not such a system is really operative in the arrangement and stabilization of cortical MTs, we must isolate and characterize the transmembrane protein(s) with the ability to interact both with extensin and with MTs or MT-associated proteins.

We thank Dr S. Kawasaki of the National Institute of Agrobiological Resources for supplying us with a sample of extensin. This work was supported in part by a Grant-in-Aid for Special Project Research (no. 63110006) from the Ministry of Education, Science and Culture of Japan.