| To whom correspondence should be addressed: Department of Biological Sciences, Graduate School of Science, Osaka University, 1-1 Machikaneyama-cho Toyonaka, Osaka 560-0043, Japan. Tel: +81–6–6850–6776, Fax: +81–6–6850–6765 E-mail: hotta@bio.sci.osaka-u.ac.jp Abbreviations: CBB, Coomassie Brilliant Blue; DAPI, 4',6-diamidino-2-phenylindole; DMSO, dimethylsulfoxide; HRP, Horseradish Peroxidase; MT, microtubule; MTOC, microtubule-organizing center; MTOP, microtubule-organizing particle; PIPES, piperazine-N,N-bis (2-ethanesulfonic acid). |
Microtubules are 25-nm hollow tubes made of α- and β-tubulin dimers, which are found among all eukaryotic cells. They are essential for various activities such as cell shape determination, intracellular trafficking, cell motility and cell division. Higher plant cells lack discrete microtubule-organizing centers (MTOCs) like centrosomes of animal cells; however, they show various microtubular arrays nucleated from dispersed sites through cell cycle progression. Higher plant cells have five typical microtubular arrays: the cortical array, the radial array from the nuclei, the preprophase band, barrel-shaped spindle and the phragmoplast (Mineyuki, 2007; Wasteneys, 2002; Baluska et al., 1997; Hasezawa et al., 1997). However, the mechanisms that regulate the organization of these microtubular arrays are not well understood.
In animal or fungal cells, microtubules are organized by structurally defined MTOCs such as the centrosome or the spindle pole body. Such microtubules have a uniform polarity; they are radially generated from MTOCs anchoring minus ends, and elongate by tubulin polymerization at plus ends distal to the MTOC (Doxsey, 2001). The third tubulin, γ-tubulin is considered to be a universal component of MTOC in animal and fungal cells (Doxsey, 2001). A ring-shaped multiprotein complex containing γ-tubulin is embedded in centrosome and serves as a template for microtubule nucleation (Doxsey, 2001; Zheng et al., 1995).
Also in higher plant cells, γ-tubulin is thought to work as a component of MTOC. Indeed, it has been shown that γ-tubulin is localized in dispersed sites but along microtubular arrays in plant cells (Liu et al., 1993; Liu et al., 1994). Recently, it was reported that γ-tubulin nucleates new microtubules along the existing microtubules in the cortical arrays of plant cells (Murata et al., 2005), and the essential role of γ-tubulin for the assembly of the cortical microtubules in the acentrosomal plant cells has been confirmed in vivo (Binarová et al., 2006; Pastuglia et al., 2006).
In higher plant cells, microtubular arrays radiate from the nuclear surfaces toward the cell periphery during certain stages of interphase (radial microtubules; hereafter radial MTs) (Lambert, 1980; Mizuno, 1993; Stoppin et al., 1994). Radial MTs might have important roles in nuclear positioning, import and/or export of nuclear materials. Spc98, a component of γ-tubulin complex in animal and yeast cells, was reported to localize on tobacco nuclear surface (Erhardt et al., 2002) implying that plant nuclear surfaces possess similar properties to centrosomes. Conversely, Mizuno suggested that under in vitro condition, tubulin might be incorporated into the proximal ends of radial MTs around nuclear particles isolated from tobacco cells (Mizuno, 1993). This implies the possible existence of an unknown MTOC which induces a totally reversed incorporation of tubulin on the plant nuclear surface in contrast with γ-tubulin containing MTOCs. The above paradox, including the polarity of radial MTs, remains unresolved.
We tried to purify a protein factor with microtubule organizing activity from isolated nuclei and identified to be histone H1. Histone H1 readily forms complexes with α,β-tubulin (not γ-tubulin) and organize “aster-like” microtubules in vitro (hereafter aster-like MTs) (Nakayama et al., submitted). Moreover, we found that the antibodies raised against histone H1 blocked the nucleation of radial MTs from nuclear surfaces (Nakayama et al., submitted). This strongly suggested that histone H1-tubulin complex acts as the microtubule-organizer on the nuclear surface.
In this work, we determined at which end the incorporation of tubulin into radial MTs occurred. Like our previous observation of microtubule-nucleation from isolated nuclear particles (Mizuno, 1993), tubulin was incorporated at the proximal ends of the radial MTs from the surface of nuclei as well as the aster-like MTs organized by histone H1-tubulin complexes. Furthermore, the polarity of such microtubules were determined to be proximal end plus. Our results indicate that in higher plant cells, a previously uncharacterized MTOC operates on the nuclear surface in a manner unlike that of either centorosomes of animal cells and spindle pole bodies of fungal cells.
Tobacco BY-2 cells were cultured as described elsewhere (Nagata et al., 1992). Cells were synchronized at S phase by culture with 5 mg/L aphidicolin for 24 h. The nuclei were isolated without detergent as described previously (Nakayama et al., submitted).
Histone H1 extracted with 0.4 N sulfuric acid from BY-2 cells (Nakayama et al., submitted) was resolved by SDS-PAGE (Laemmli, 1970) and obtained by electro-elution from the gel. Tubulin was purified from porcine brain by two cycles of assembly and disassembly, followed by chromatography on DEAE Sephacel (GE Healthcare Bio-Science Corp., Piscataway, NJ, USA) (Shelanski et al., 1973). Rhodamine- and caged fluorescein-labeled tubulin was prepared according to the method described previously (Hyman et al., 1991).
Acid extracted proteins or purified histone H1 were resolved by SDS-PAGE and transferred onto a nitrocellulose membrane. After blocking for 1 h in phosphate buffered saline (PBS) supplemented with 5% milk powder, 1:250 primary anti-histone H1 antibody or 1:500 primary anti-histone H3 antibody (Upstate Biotechnology, Lake Placid, NY, USA) was added and incubated for 1 h. After washing by 0.05% Tween 20 in PBS, 1:1000 HRP-goat anti-rabbit IgG secondary antibody (KPL, Gaithersburg, MD, USA) or 1:1000 HRP-goat anti-mouse IgG secondary antibody (Bio-Rad, Hercules, CA, USA) was added and incubated for 1 h. After washing the bands were visualized by incubation of the membrane with the peroxidase substrate 4-chloro-1-naphthol (0.05%) and hydrogen peroxide (0.03%).
The isolated nuclei were fixed in 0.01% glutaraldehyde in acetone-methanol (1:1, v/v) at –20°C for 1 h. The fixed nuclei were then rehydrated in 50 mM PIPES buffer pH 6.9 and attached to a poly-L-lysine-coated coverslip. After being washed and blocked with 5% milk powder in 50 mM PIPES buffer (pH 6.9) for 1 h, the nuclei were incubated with affinity-purified polyclonal antibody against tobacco histone H1 (1:250 to 1:500) for 1 h. The nuclei were then washed and incubated for 1 h with rhodamine-conjugated goat anti-rabbit IgG (1:350; Molecular Probes, Eugene, OR, USA). A control experiment was performed by using mouse monoclonal anti-human histone H3 (1:500; Upstate Biotechnology) as a primary antibody and fluorescein-conjugated goat anti-mouse IgG (1:50; ICN/Cappel, Aurora, Ohio, USA) as a secondary antibody. The nuclei were treated for 5 min with a 1 mg/L solution of 4',6'-diamidino-2-phenylindole (DAPI) and washed, then embedded in a glycerol solution containing 0.01% n-propyl gallate and observed. Three-dimensional images of the nuclei were acquired with a DeltaVision deconvolution microscope system (Applied Precision, Mercer Island, WA, USA; Haraguchi et al., 1999) equipped with a DPlanApo oil immersion objective lens (100×, NA 1.30 or 60×, NA 1.40; Olympus, Tokyo, Japan), a PXL cooled CCD camera (Photometrics, Tucson, AZ, USA) and SoftWoRx software. Image stacks of the nuclei were subjected to iterative deconvolution (Agard et al., 1989). The maximum projections of the deconvolved sections are shown. Obtained images were exported to TIF format and processed by Adobe Photoshop 6.0 to give the final images.
Histone H1-tubulin complexes were formed as previously described (Nakayama et al., submitted). The complexes with/without pre-formed microtubules were then negatively stained (2% uranyl acetate) and examined with JEOL 1200EX transmission electron microscope (JEOL, Tokyo, Japan) operated at 80 kV.
In vitro microtubule-organization assay of the isolated nuclei or of the histone H1-tubulin complexes was performed as follows. The isolated nuclei or pre-formed histone H1-tubulin complexes (as previously described (Nakayama et al., submitted)) were attached to a poly-L-lysine-coated coverslip and incubated at 37°C for 45 to 60 min with 0.4 mg/mL rhodamine-tubulin (a mixture of rhodamine-labeled tubulin and unlabeled tubulin at 1:4) in assay buffer (50 mM PIPES, 1 mM EGTA, 1 mM MgCl2, 0.1% n-propyl gallate, 10% DMSO, and 1 mM GTP, pH 6.9).
Initially, radial MTs or aster microtubules were generated as above. After a 15-min incubation, the buffer on the coverslip was replaced with new buffer containing 0.4 mg/mL dimly labeled rhodamine-tubulin (a mixture of rhodamine-labeled tubulin and unlabeled tubulin at 1:10) that had been pre-warmed for 2 min at 37°C. After 30 to 45 min of incubation, 0.5 μM Oregon Green 488 conjugated paclitaxel (Molecular Probes) was added and the result immediately observed. Oregon Green 488 conjugated paclitaxel binds to the total length of each microtubule, and we confirmed that the result was not influenced by the use of this microtubule-stabilizing reagent. Images of radial MTs were obtained by using a Fluoview confocal laser scanning microscope (Olympus) equipped with an UPlanSApo oil immersion objective lens (100×, NA 1.40; Olympus). Oregon Green 488 and rhodamine fluorescences were monitored using a 510 to 550-nm band pass emission filter and a 585-nm long pass emission filter respectively, and a 488/568-nm excitation line of a Kr/Ar laser. Images of aster-like MTs were obtained by using a DeltaVision system (Applied Precision) equipped with a DPlanApo oil immersion objective lens (100×, NA 1.30; Olympus), a PXL cooled CCD camera (Photometrics) and SoftWoRx software. All the images were exported to TIF format and processed by Adobe Photoshop 6.0 to give the final images.
Radial MTs were generated from nuclei using a mixture of tubulin (rhodamine-labeled tubulin plus caged fluorescein-labeled tubulin plus unlabeled tubulin at 5:1:20), and after 15 min the tips of the elongating microtubules were activated by a 406-nm QLM laser equipped with a DeltaVision-RT microscope system (Applied Precision). Time-lapse images were taken every 4 to 20 min for about 140 min after activation using the DeltaVision with a PlanApo oil immersion objective lens (60×, NA 1.40; Olympus) and a CoolSNAP HQ cooled CCD camera (Photometrics). To avoid inhibiting the elongation by light irradiation, the exposure time was minimized.
Aster microtubules were generated from the histone H1-tubulin complex using rhodamine-tubulin (a mixture of rhodamine-labeled tubulin and unlabeled tubulin at 1:4) and the middle regions of the microtubules were photobleached by irradiation with a 568-nm Kr/Ar laser equipped with a Fluoview confocal microscope. The laser power was maximized, and scanning was performed six times for a total of 4 s. A series of images were taken every 3 to 10 min for 30 to 60 min using a DeltaVision deconvolution microscope equipped with an SPlanApo oil immersion objective lens (60×, NA 1.40; Olympus) and a PXL cooled CCD camera, and selected images were shown.
Polarity-marked microtubules (rhodamine-labeled) were prepared as previously described (Howard and Hyman, 1993). Polymerized polarity-marked microtubules were laid on the top of glycerol cushion (100 mM PIPES, 1 mM EGTA, 1 mM MgCl2, 60% glycerol, pH 6.9), and precipitated by centrifugation for 20 min at 100,000×g at 30°C. The precipitated polarity-marked microtubules were suspended in the microtubule-stabilizing buffer (100 mM PIPES, 1 mM EGTA, 1 mM MgCl2, 5 μM paclitaxel pH 6.9). Pre-formed complexes and polarity-marked microtubules were mixed. Then, the polarity-marked asters were observed. Single-section images were taken under the Fluoview scanning confocal microscope.
To reveal the subnuclear localization of histone H1, we performed immunofluorescence microscopy using anti-histone H1 antibody established previously (Nakayama et al., submitted). The specificity of antibodies was checked by immunoblot analysis shown in Fig. 1A and previous report (Nakayama et al., submitted). We used the nuclei isolated from suspension-cultured tobacco BY-2 cells highly synchronized in ‘S phase’, at which most of the cells show the radial microtubular array. We found that histone H1 localized not only in the nucleoplasm but also on the nuclear rim (Fig. 1B, left panel). The former is presumably the linker histone, since the staining totally overlaps with that of DNA (Fig. 1B, center and right panels). Conversely, the latter does not overlap with the staining of DNA (Fig. 1B, right panel). Since it was clear that histone H3 localized only in the nucleoplasm (Fig. 1C, all panels), the staining of histone H1 on the nuclear rim is specific but not caused by a technical artifact. No staining of the nucleoplasm and nuclear rim was observed using pre-immune serum.
 View Details | Fig. 1. Histone H1 localizes on the nuclear rim separately from chromatin as well as inside the nucleus. (A) Immunoblot analysis. Acid extracted histones (lane 1, 3 and 5) and purified histone H1 (lane 2 and 4) were resolved by SDS-PAGE and immunoblotted by anti-histone H1 antibody (lane 3 and 4) or anti-histone H3 antibody (lane 5). Gel was stained by CBB (lane 1 and 2). Both antibodies are specific. (B) Histone H1 (red in left panel) localizes in the nucleoplasm and on the nuclear rim. Histone H1 on the nuclear rim does not co-localize with chromatin (blue in center panel). The right panel is the merged image. The large round hole inside the nucleus is the nucleolus. (C) Control experiment using anti-histone H3 antibody. Histone H3 (green in left panel) totally co-localizes with chromatin (blue in center panel). The right panel is the merged image. Scale bars, 5 μm. |
We used a label-dilution experiment to examine at which end the incorporation of tubulin occurs in the organization of radial MTs on the intact nuclei. Definition of the terms, ‘proximal’ and ‘distal’ are shown in Fig. 2A. We separately prepared brightly and dimly labeled rhodamine-tubulin. After primary incubation of the nuclei with brightly labeled rhodamine-tubulin, they were incubated with dimly labeled rhodamine-tubulin (Fig. 2B). Fluorescent taxol (Oregon Green 488 paclitaxel, Molecular Probes) was added just before observation to allow visualization of the total length of each microtubule (Fig. 2C, and merged image D). The nuclear surfaces were visualized by highly accumulated tubulin onto them during early step (Fig. 2B and E, asterisks). Such tubulin probably forms complexes with histone H1, and serves as a MTOC. Radial MTs generated from the nuclear surface had a bright outer region (Fig. 2B, E arrowhead) and a dim inner region which was polymerized earlier and later, respectively (Fig. 2B). This suggests that tubulin is incorporated into the proximal ends of the microtubules on the nuclear surface.
 View Details | Fig. 2. Tubulin is incorporated into the radial MTs at the proximal end. (A) Sketch of the experimental background and the definitions of ‘proximal’ and ‘distal’. (B to D) Images obtained by label-dilution experiment. Radial MTs (green, C) emanating from the nuclear surface show proximal dim fluorescence and distal bright fluorescence (red, B). The merged image is shown in (D). A white asterisk and a white arrowhead in (B) show the nuclear surface and bright region, respectively. (E) Quantification of the fluorescence intensity on the white line in (D). The fluorescence of rhodamine-tubulin (red) have two peaks indicating the nuclear surfaces (black asterisks) and two peaks indicating microtubules organized early (black arrowheads). (F) Time-lapse images of photoactivation experiment. During elongation of the radial MTs (red), the uncaged fluorescent marks on them (green) moved outward. White arrowheads indicate the movement of green marks. Scale bars, 10 μm. |
We conducted another experiment using a photoactivation technique to confirm the above result by time-lapse observation. For this purpose, we prepared caged fluorescein-labeled tubulin. Caged fluorescein is a photoactivatable fluorescent dye whose activity is masked until ultraviolet light converts the dye into an active state (Mitchison, 1989). Radial MTs were generated from a mixture of tubulin labeled with caged fluorescein and rhodamine. The tips of the elongating microtubules were irradiated with a near-ultraviolet laser and uncaged green fluorescent marks appeared on them (Fig. 2F, white arrowheads). Time-lapse observation after uncaging showed that the green marks moved outward during elongation of the radial MTs (Fig. 2F). The green mark moved at an average of 0.035 μm min–1 (s.d.=0.015, n=7) for about 70 to 80 min after photoactivation. Occasionally, some populations of green marks moved more slowly, but no obvious incorporation of tubulin at the distal ends was observed. These results suggest that nucleation and elongation of the radial MTs are achieved not by the incorporation of tubulin into the distal ends (as is the case in the centrosome of animal cells) but by continuous incorporation of tubulin into the proximal ends on the nuclear surfaces.
Without nuclear architecture, purified histone H1 or recombinant histone H1 has the ability to form complexes with tubulin in vitro, and to organize microtubules into an aster-like shape (Nakayama et al., submitted). Therefore, the mechanism of the aster-like MT formation could be regarded as a minimal, simple model of assembly of the radial MTs. Transmission electron microscopy revealed that the complex formed in vitro was not an amorphous aggregation but an ordered ring structure whose diameter was about 35-nm stained negatively (Fig. 3A). These complexes readily self-gather and facilitate the assembly of aster-like MTs around them (Nakayama et al., submitted). We performed a label-dilution experiment under the same conditions used to examine the radial MTs emanating from the nuclear surface to check that tubulin incorporation occurs at proximal ends of the aster-like MTs formed by histone H1-tubulin complexes (Fig. 3B). The distal regions of the aster-like MTs were clearly labeled more brightly than the proximal regions (Fig. 3C and D, arrowheads). The bright cluster in the middle of the aster (Fig. 3C and D, asterisks) was assumed to be an accumulation of complexes that consisted of brightly labeled rhodamine-tubulin and histone H1. This suggests that incorporation of tubulin into the aster-like MTs occurred at the proximal ends, as was the case in the radial MTs on the nuclear surfaces.
 View Details | Fig. 3. Tubulin incorporation into the aster-like MTs also occurs at the proximal end. (A) Histone H1-tubulin complex is a ring-like structure stained negatively. (B) Sketch of the experimental background and the definitions of ‘proximal’ and ‘distal’. (C) Image obtained by label-dilution experiment. The aster-like MT has a bright distal region (a white arrowhead), a dim proximal region and a bright cluster (a black asterisk). (D) Quantification of the fluorescence intensity on the white line in (C). The fluorescence of rhodamine-tubulin have one peak indicating the bright cluster (a black asterisk) and two peaks indicating microtubules organized early (black arrowheads). (E) Images selected from the photobleaching experiment. White lines indicate the distance between the center of aster-like MTs and the photobleached line. Scale bars in (A), 100 nm; inset in (A), 10 nm; in (C) and (E), 5 μm. (F) Kymograph extracted from the same time-lapse images in (E). Vertical scale bar, 2 μm; horizontal scale bar, 6 min. (G) Quantification of the outward movement of photobleached area in (F). The length of black line in (F) was quantified. |
A photobleaching experiment was performed to confirm this result by time-lapse observation. Time-lapse images were obtained after creation of a line of photobleaching at the middle of the elongating aster-like MTs. The photobleached line moved outward during elongation of the aster-like MTs (Fig. 3E, F and G). The photobleached area moved at an average of 0.041 μm min–1 (s.d.=0.016, n=10). This result corresponds well with the elongation velocity of the tips of radial MTs (Fig. 2F). In conclusion, the histone H1-tubulin complexes organize the aster-like MTs by microtubule nucleation and continuous polymerization at the proximal ends, in the same way that they organize the radial MTs on the nuclear surfaces.
From the above results, we propose two properties of the histone H1-tubulin complex: the first is microtubule nucleation; the second is continuous association with growing microtubule ends. In other words, histone H1-tubulin complexes should have an affinity to the elongating microtubule ends although we can not necessarily consider the growing proximal end to be plus end. Preformed polarity-marked microtubules were mixed with histone H1-tubulin complexes to reveal which microtubule end they have an affinity for, the plus end or minus end. Large numbers of aster-like structures consisting of polarity-marked microtubules and histone H1-tubulin complexes were readily formed even without adding GTP (hereafter called polarity-marked asters; Fig. 4A, B). They had microtubules with their plus ends proximal to the center of the aster and their minus ends distal. The histone H1-tubulin complexes localized in the center of each aster. Transmission electron microscopy (negative stain) revealed that histone H1-tubulin complexes attached to the ends of microtubules but not the walls of them, and such a structure is assumed to be a minimal unit of the polarity-marked aster (Fig. 4C). These results indicate that the histone H1-tubulin complexes have the affinity to the plus ends of the microtubules, and the affinity to the plus ends doesn’t require GTP hydrolysis. Added to the isolated nuclei, the polarity-marked microtubules adhered laterally to the nuclear surfaces, and we could not generate “polarity-marked radial MTs”. The histone H1-tubulin complexes on the nuclear surfaces might be unable to catch the pre-formed microtubules due to the structural problem.
 View Details | Fig. 4. Polarity-marked asters are organized by histone H1-tubulin complexes. (A and B) Polarity-marked microtubules were arranged by histone H1-tubulin complex (center) with their minus ends outward. Various intermediate structures with the same microtubule polarity are also seen in (B). (C) Histone H1-tubulin ring complexes attached to the ends of microtubules. Histone H1-tubulin complexes were mixed with pre-formed microtubules, and stained negatively. Histone H1-tubulin ring complexes (black arrows) have the affinity to the ends of microtubules. Polarity-marked asters would be assembled from such ring complex-microtubule units. (D) A pathway of the assembly of polarity-marked aster. In this pathway, histone H1-tubulin complexes attach onto the walls of polarity-marked microtubules. Then these complexes move towards microtubule plus ends by an assumed motor-like activity. These microtubule-complex units assemble, and the polarity-marked asters are organized. (E) Another pathway of the assembly of polarity-marked aster. In this pathway, histone H1-tubulin complexes attach to the microtubule plus ends directly. These microtubule-complex units simply assemble each other, and then polarity-marked asters are organized. Scale bars in (A) and (B), 5 μm; in (C), 500 nm. |
In the former report, we revealed that histone H1 possessed microtubule-nucleation activity, and regarded it to be a strong candidate of the organizing factor of radial MTs on the nuclear surface (Nakayama et al., submitted). Since histone H1 has been understood to be a so-called ‘linker histone’ that binds the nucleosome and stabilizes the chromatin structure (Thoma and Koller, 1977; Thomas, 1999), we tried to reveal the detailed subnuclear localization of histone H1. Immunofluorescence microscopy demonstrated that histone H1 localized on the nuclear rim in addition to the presence in the nucleoplasm. Histone H1 on the nuclear rim seems not to co-localize with DNA. This confirms the presence of two types of histone H1: one the linker histone and the other a component of MTOC, with each histone H1 presumably localizing separately in accordance with its function. The nuclear rim binding site for histone H1 remains unidentified. The nuclei which had been treated with detergent (Nonidet P-40 or Triton-X100) to remove membrane lipids, could still be stained by anti-histone H1 antibody on their rims, and retained the ability to organize the radial MTs in vitro (Nakayama et al., submitted). This implies the association of histone H1 to certain structures such as the nuclear pore complexes associating with stable nuclear matrices, which might be resistant to detergent, and in the same time, be facing to the cytoplasm.
Tubulin was incorporated into the proximal ends of both radial MTs (on the nuclei) and aster-like MTs (on the histone H1-tubulin complexes). This result supports the notion that histone H1 serves as a component of MTOC on the nuclear surface. The microtubule-nucleating factor on the nuclear particle (microtubule organizing particle, MTOP) purified in our previous report (Mizuno, 1993) might contain histone H1. The incorporation of tubulin at the proximal end during the microtubule elongation suggests that the elongation of microtubules is achieved by ‘pushing out’ of microtubules, and probably this property is realized by histone H1-tubulin complex. Polarity-marked asters were obtained by mixing histone H1-tubulin complexes with polarity-marked microtubules but without GTP. We suggest two possible processes for the assembly of the polarity-marked aster as shown in Fig. 4D, E. One is the organization of microtubule-complex units (shown in Fig. 4C) transported by the assumed plus-end-directed motor-like property of histone H1-tubulin complexes (Fig. 4D). It seems, however, difficult to explain this process without considering the energy source. The other is the gathering of microtubule-complex units caused by the self-gathering-property of histone H1-tubulin complexes (Fig. 4E). In this process, GTP would not be required. The plus-end-directed association of histone H1-tubulin complexes might occur regardless of the presence or the absence of energy source. The histone H1-tubulin complexes might have a special affinity to the microtubule plus ends, for instance to the GTP-cap of microtubules.
Our results suggest that the complexes on the nuclear surfaces are processively (continuously) associated with the proximal ends of the growing microtubules by their plus-end-directed affinity, and this processivity might produce the power to push out the polymerized microtubules, and results in the elongation of microtubules (Fig. 5). So, the histone H1-tubulin complex identified here is not only a novel MTOC but also a novel plus end-tracking protein (+TIP) (Carvalho et al., 2003). The two feature of histone H1-tubulin complex i.e. microtubule nucleation and the processive association with growing plus end, is similar to the action of formin which promotes the nucleation of F-actin and keeps binding processively to the elongating barbed end (Higgs, 2005; Kovar, 2006), although it is unknown whether the radial MT is polymerizing through the ring of histone H1-tubulin complex like formin and F-actin. If histone H1-tubulin complex had a motor-like activity which was coupled with the polymerization of tubulin like the case of formin and G-actin, this motor-like activity would be much lower than other motor proteins like kinesin, dynein and myosin (Howard, 1997; Mallik et al., 2005), because the elongation velocity of aster-like MTs or radial MTs was extremely low. Neither histone H1 nor tubulin has both the microtubule nucleation activity and the plus-end-directed association. Interestingly, the formation of a complex from two components might provide it such special features. Previous reports that histone H1 stabilizes sea urchin flagellar microtubules (Multigner et al., 1992) or has the ability to interact with tubulin (Mithieux et al., 1984), suggest that there is a close relationship between histone H1 and tubulin (Kaczanowski and Jerzmanowski, 2001).
 View Details | Fig. 5. A model for the organization of the radial MTs on the nuclear surface. Histone H1-tubulin complex (red ring) on the nuclear surface, facilitates the microtubule nucleation and the following continuous incorporation of tubulin at the proximal end of radial MT (green). During the tubulin incorporation, histone H1-tubulin complex pushes out the radial MT using its plus-end-directed affinity. This process seems to be coupled with the polymerization of tubulin. As a result, radial MT elongates with its minus end outward. |
Tubulin incorporation end or the polarity of radial MTs should be determined also in vivo. As known as a minus end associating protein (Zheng et al., 1995; Wiese and Zheng, 2000), γ-tubulin might function at the plasma membrane where the distal ends of radial MT reach.
Our results do not exclude the possibility that γ-tubulin containing MTOC may function on the nuclear surface as several reports showed. Tobacco BY-2 cells show radial microtubular arrays in the early G1 phase as well as S and G2 phase on which we are focusing. Other MTOCs, such as γ-tubulin containing MTOC might be functional in the early G1 phase.
We speculate that higher plants have evolved a unique MTOC to compensate the loss of the centrosome during their early evolution. During mitosis and meiosis in acentrosomal higher plant cells, microtubules growing from around the chromosome (chromosomal microtubules) self-organize into a spindle as well as in oocytes of some animal species in the absence of centrosome (Gard, 1992; Theurkauf and Hawley, 1992; Waters and Salmon, 1997). The assembly of chromosomal microtubules is observed even in some cells that naturally have centrosomes, when their centrosomes are destroyed (Khodjakov et al., 2000). So, chromosomal microtubules might be essential in spindle formation (Karsenti and Vernos, 2001; Karsenti and Nédélec, 2004). It is possible that histone H1-tubulin complexes act as MTOC in such process assembling microtubules with their polarity of distal end minus. Further investigation into the novel MTOC, histone H1-tubulin complex, will help to elucidate the molecular mechanisms of microtubule organization in both plant and animal cells.
We thank I. Meier and T. Horio for helpful comments. We thank L. J. Irving for critical reading of the manuscript. This work was supported by the Japan Society for the Promotion of Science (T. Hotta).
|
Index