Plant microtubules are intrinsically more dynamic than those from animals. We know little about the dynamics of the interaction of plant microtubule-associated proteins (MAPs) with microtubules. Here, we have used tobacco and Arabidopsis MAPs with relative molecular mass 65 kDa (NtMAP65-1a and AtMAP65-1), to study their interaction with microtubules in vivo. Using fluorescence recovery after photobleaching we report that the turnover of both NtMAP65-1a and AtMAP65-1 bound to microtubules is four- to fivefold faster than microtubule treadmilling (13 seconds compared with 56 seconds, respectively) and that the replacement of NtMAP65-1a on microtubules is by random association rather than by translocation along microtubules. MAP65 will only bind polymerised microtubules and not its component tubulin dimers. The turnover of NtMAP65-1a and AtMAP65-1 on microtubules is similar in the interphase cortical array, the preprophase band and the phragmoplast, strongly suggesting that their role in these arrays is the same. NtMAP65-1a and AtMAP65-1 are not observed to bind microtubules in the metaphase spindle and their rate of recovery is consistent with their cytoplasmic localisation. In addition, the dramatic reappearance of NtMAP65-1a on microtubules at the spindle midzone in anaphase B suggests that NtMAP65-1a is controlled post-translationally. We conclude that the dynamic properties of these MAPs in vivo taken together with the fact that they have been shown not to effect microtubule polymerisation in vitro, makes them ideally suited to a role in crossbridging microtubules that need to retain spatial organisation in rapidly reorganising microtubule arrays.

Plant microtubules undergo dynamic reorganisation through the cell cycle and in response to internal and external stimuli (Nick, 1998; Hussey, 2004). Four distinct arrays appear sequentially: the interphase cortical array directs cell expansion, the preprophase band determines where the cell plate will be formed by the phragmoplast array and in common with all other eukaryotes the mitotic spindle separates the daughter chromosomes. The dynamics of plant microtubules both in vitro (Moore et al., 1997) and in vivo (Hush et al., 1994; Shaw et al., 2003) have been shown to be faster than the dynamics of animal microtubules. The major microtubule protein is a heterodimer of α/β tubulin and the structure of the tubulins shows a high degree of phylogenetic conservation (Burns and Surridge, 1994). The difference in the dynamic properties of plant and animal microtubules in vitro has been suggested to arise as a result of small differences in tertiary structure of the dimer, arising from the small number of residue differences between animal and plant tubulins (Moore et al., 1997). In animals numerous structural MAPs (e.g. tau, MAP1, MAP4) that are not present in plants modulate microtubule dynamics (Kreis and Vale, 1993), but so far no plant structural MAP has been shown to affect microtubule dynamics in vivo.

MAP65 is one of the best characterised plant MAPs and it was originally identified as a family of proteins of ∼65 kDa molecular weight that co-purified with plant microtubules (Jiang and Sonobe, 1993; Chan et al., 1999). Biochemically purified MAP65 induces microtubule bundling (Jiang and Sonobe, 1993) via formation of 25 nm crossbridges (Chan et al., 1999). A recombinant MAP65 from Arabidopsis, AtMAP65-1, is able to generate similar crossbridges and this requires AtMAP65-1 to form homodimers (Smertenko et al., 2004). Although MAP65 crossbridges microtubules it does not appear to have a direct effect on microtubule dynamics (Smertenko et al., 2000; Wicker-Planquart et al., 2004; Smertenko et al., 2004).

The tobacco MAP65, NtMAP65-1a, and the Arabidopsis MAP65, AtMAP65-1 are from the same phylogenetic clade of MAP65 isotypes (Hussey et al., 2002) and they show similar patterns of localisation on microtubules in the four microtubule arrays (Smertenko et al., 2000; Smertenko et al., 2004). They are present on subsets of microtubules in the two cortical arrays (interphase and the preprophase band), on microtubules at the anaphase spindle midzone and they persist at the midzone in the cytokinetic phragmoplast. Electron microscopy shows that microtubules in the cortical array, the preprophase band and the phragmoplast are bundled, but as yet we know nothing about the dynamic properties of the MAP65 crossbridges.

In this paper we investigate whether the affinity of NtMAP65-1a or AtMAP65-1 for microtubules is cell cycle stage specific and whether the mechanism of bundling is likely to be the same or different in the four microtubule arrays. Moreover, we ask whether this protein is translocated on microtubules or whether it is added and removed locally. To this end we have generated stable transformed Arabidopsis lines and tissue culture cells expressing either NtMAP65-1a or AtMAP65-1 fused to GFP. The interaction of the GFP-fused MAP65s with microtubules has been studied in vivo using fluorescent recovery after photobleaching (FRAP). We report that the turnover of the NtMAP65-1a and the AtMAP65-1 microtubule interaction is faster than microtubule treadmilling and faster than other known structural MAPs from animals. We provide data that strongly indicate that the properties of NtMAP65-1a (AtMAP65-1) make it ideal for temporally crossbridging microtubules in dynamic microtubule arrays so that their spatial organisation is maintained.

Tissue culture conditions

Arabidopsis seeds were germinated on the 1/2 MS basic salt medium supplemented with 3% sucrose and 7.5% agar at 23°C and 14 hours day/10 hours night light cycle.

BY-2 cells were cultured as described (Jiang and Sonobe, 1993). For the disruption of microtubules in the BY-2 lines expressing NtMAP65-1a:GFP 10 μM oryzalin was added to the medium and the cells were incubated for 2 hours at 25°C with continuous shaking. The 10 mM stock solution of oryzalin was prepared in DMSO and stored at –20°C.

NtMAP65-1a:GFP and AtMAP65-1:GFP constructs and plant transformation

Full-length smGFP (accession number U70495) (Davis and Vierstra, 1998) was subcloned into NotI and XhoI sites of pGreen II vector. Then full-length NtMAP65-1a (Smertenko et al., 2000) or AtMAP65-1 (Smertenko et al., 2004) was sub-cloned downstream of GFP into SalI and XbaI or XhoI and EcoRI sites respectively. The sequences of GFP and NtMAP65-1a or AtMAP65-1 chimeras and their linker region in the pGreenII vector were confirmed by sequencing. The constructs were co-transformed with pSoup vector into Agrobacterium tumifaciens strain C58C3 for the transformation of Arabidopsis thaliana plants and into LBA4404 for the transformation of tobacco BY-2 cells. Arabidopsis plants were transformed using the floral dipping method (Clough and Bent, 1998) and BY-2 cells were transformed by the coincubation method (Geelen and Inzé, 2001). The seeds were germinated as described above except that the medium was supplemented with 50 mg/l kanamycin. BY-2 transformants were selected on medium containing 200 mg/l kanamycin and 500 mg/l carbenicillin. The kanamycin-resistant colonies expressing MAP65:GFP fusions were subcultured and maintained in liquid medium containing 200 mg/l kanamycin.

Microscopy and FRAP measurements

For taking the static MAP65-1:GFP images, 7-day-old Arabidopsis seedlings were collected from agar plates and mounted in distilled water. For FRAP or time-lapse studies, seedlings were partially immobilised by mounting in 1% low gelling temperature agarose at 37°C and observed immediately. BY-2 tissue culture cells were mixed with an equal volume of 1% low gelling point agarose (Sigma, Dorset, UK) solution in BY-2 medium. The samples were observed using a Zeiss 510 inverted confocal microscope, with an Argon/Krypton laser, equipped with a Plan-Neofluor 40 ×/1.3 oil immersion objective, a HFT488 dichroic mirror and a 505 nm long path emission filter. The laser power was used at a maximum of 6% for samples expressing either GFP:NtMAP65-1a or GFP:AtMAP65-1. The imaging was performed within the first 20 minutes of the sample lifetime. In the FRAP experiments the frames were collected every 3.6 seconds for the analysis of A. thaliana seedling cells, every 2 seconds for BY-2 cortical microtubules, phragmoplast and preprophase band and every 0.3 seconds for mitotic spindle and oryzalin-treated cells.

For the statistical analysis of the FRAP data, the readings from the first 25 scans were selected then the background value was measured outside the cell and subtracted from the experimental FRAP values. The data collected from numerous cells (the number is indicated for each of the experiments) were averaged and normalised by division by the mean value. The level of recovery was expressed as a percentage of the fluorescence before photobleaching.

To calculate the time of 50% signal recovery, t½, the FRAP curve from each of the experiments was fitted to an exponential recovery curve: F(t) = Finf – (FinfF0)*exp(–t*koff) (Bulinski et al., 2001), where F0 is the fluorescence after the photobleaching, Finf is the fluorescence when the recovery reached the plateau stage, t is time and koff is the first-order rate constant that describes the rate of recovery. To estimate the fraction of MAP65:GFP initially bound to microtubules and the real koff of signal recovery, the experimental data was fitted into a two-component model consisting of a free-diffusing component with fast recovery time and a second component describing the koff of bound MAP65:GFP, F(t) = Finf – (FinfF0)*[(1–A)*exp(–t*0.402) + A*exp(–t*koff)] where A is the fraction of MAP65:GFP initially bound to the microtubules, 0.402 is the recovery rate constant of the freely diffusing unbound fraction of MAP65:GFP estimated in the experiment where the cells were treated with oryzalin, Finf and F0 are as described above. The curve fit and calculation of koff was done with Abscissa software version 3.2.1 (http://iapf.physik.tu-berlin.de/DZ/bruehl/). The t½ was calculated as ln(2)/koff.

Surface plasmon resonance

Surface plasmon resonance measurements were made using a Biacore 3000 system (Biacore AB, Uppsala, Sweden) running at 10 μl/minute at 25°C. Microtubule stabilising buffer (MTSB; 0.1 M PIPES, 2 mM EGTA, 50 mM NaCl, pH 6.8) was used as running buffer. Proteins were immobilised on a CM5 chip (Biacore AB) by carbodiimide coupling to amine groups using EDC/NHS and standard protocols, with subsequent ethanolamine treatment to block unreacted sites as recommended by the manufacturer. NtMAP65-1a recombinant protein (Smertenko et al., 2000) at a final concentration of 50 μg/ml in 10 mM sodium citrate buffer (pH 4.5) was covalently coupled on the first channel giving a final response of 9300 relative units (RU) and the second channel of the chip was treated similarly but without protein, for use as a no protein control. Tubulin was purified using phosphocellulose chromatography according published methods (Shelanski et al., 1973) and stored in liquid nitrogen. A fresh tubulin aliquot was quickly thawed at 37°C and kept on ice for 15 minutes prior to the experiment. Then part of the aliquot was diluted to 0.2 mg/ml in MTSB and all tubulin solutions were centrifuged at 100 000 g for 20 minutes at 2°C to pellet tubulin aggregates. After the centrifugation, 10 μM taxol was added to the non-diluted tubulin solution and the mixture was incubated at 30°C for 15 minutes to promote microtubule polymerisation. The final concentration of protein in both solutions was adjusted to 0.1 mg/ml and verified by a Bradford assay. The tubulin dimers and microtubules were injected over the NtMAP65-1a and control surfaces to test for binding. Taxol alone did not give a signal when added to the MTSB buffer.

Interaction between NtMAP65-1a and microtubules is dynamic

We generated Arabidopsis lines expressing GFP tagged to the N-terminus of NtMAP65-1a by stably transforming A. thaliana ColR with a GFP:NtMAP65-1a transcriptional fusion inserted under the control of the 35S promoter in the vector, pGreenII. T2 seed was used throughout for the visualisation of GFP. The germination rate, the growth and morphology of these Arabidopsis lines were similar to those harbouring vector alone, pGreeenII.

GFP:NtMAP65-1a localised to microtubules in all cells examined including epidermal cells within the cotyledon, hypocotyl, root including guard cells and trichomes (Fig. 1A-D). However, the density of the microtubule network visualised in this way appears to be less then that observed using tubulin:GFP (Ueda and Matsuyama, 2000).

We used the Arabidopsis lines expressing GFP:NtMAP65-1a to study the dynamics of this MAP65 interaction with microtubules in vivo using FRAP approach. Root, cotyledon and hypocotyl cells were assessed for their suitability for these experiments. Hypocotyl epidermal cells were chosen as these have a very low level of autofluorescence and, because of their large flattened shape, many microtubules could be observed in a single focal plane. Moreover they are easily immobilised in low gelling temperature agarose. For the FRAP experiments, a narrow patch of the cell containing several fluorescing microtubules (Fig. 2A) was bleached with a laser pulse and the time required for the recovery of 50% of the signal (t½) was estimated. This t½ value represents the time taken for half of the GFP:NtMAP65-1 to be replaced and therefore is a numerical estimation of the protein turnover. The selected area (outlined as a white rectangle in Fig. 2A) was bleached for approximately 4.8 seconds. Subsequently, images were collected every 3.6 seconds (Fig. 2B) and the intensity of the GFP signal in the selected area was measured and plotted against time (Fig. 2C). The t½ for the GFP:NtMAP65-1a was found to be 8.96 seconds (n=24). Considering the fact that tubulin:GFP in similar experiments has a t½ of 58.95 seconds (Table 1), these data demonstrate that the GFP:NtMAP65-1a signal recovers faster than GFP:tubulin, hence the exchange of GFP:NtMAP65-1a on the microtubule surface must be independent of microtubule treadmilling. As a further comparison we determined the t½ of another GFP chimera that interacts in vivo with MAP4:GFP (Granger and Cyr, 2000), and found it to be 2.91±0.21 seconds (n=20).

Table 1.

Analysis of FRAP data

Fusion Cell cycle stage/treatment Cell type koff*t½ (seconds) nkoff Bound (%)
NtMAP65-1a  Interphase   Hypocotyl   0.077   8.96±0.85   24    
NtMAP65-1a  Interphase   BY-2   0.100   6.95±0.91   23   0.094   80.7  
NtMAP65-1a  PPB   BY-2   0.117   5.92±0.64   19   0.083   85.8  
NtMAP65-1a  Metaphase   BY-2   0.960   0.72±0.21   23   n.d.   n.d.  
NtMAP65-1a  Phragmoplast   BY-2   0.146   4.83±0.64   20   0.117   75.0  
NtMAP65-1a  Oryzalin   BY-2   0.402   1.71±0.24   23    
AtMAP65-1  Interphase   BY-2   0.117   5.93±0.86   20    
MBD-GFP   Interphase   Hypocotyl   0.238   2.91±0.21   20    
Tubulin   Interphase  Hypocotyl    58.95     
Tubulin   Interphase  Stamen hair    67.0±3.3     
Tubulin   Metaphase  Stamen hair    31.4±6.1     
Tubulin   Phragmoplast  Stamen hair    60.0±8.1     
Fusion Cell cycle stage/treatment Cell type koff*t½ (seconds) nkoff Bound (%)
NtMAP65-1a  Interphase   Hypocotyl   0.077   8.96±0.85   24    
NtMAP65-1a  Interphase   BY-2   0.100   6.95±0.91   23   0.094   80.7  
NtMAP65-1a  PPB   BY-2   0.117   5.92±0.64   19   0.083   85.8  
NtMAP65-1a  Metaphase   BY-2   0.960   0.72±0.21   23   n.d.   n.d.  
NtMAP65-1a  Phragmoplast   BY-2   0.146   4.83±0.64   20   0.117   75.0  
NtMAP65-1a  Oryzalin   BY-2   0.402   1.71±0.24   23    
AtMAP65-1  Interphase   BY-2   0.117   5.93±0.86   20    
MBD-GFP   Interphase   Hypocotyl   0.238   2.91±0.21   20    
Tubulin   Interphase  Hypocotyl    58.95     
Tubulin   Interphase  Stamen hair    67.0±3.3     
Tubulin   Metaphase  Stamen hair    31.4±6.1     
Tubulin   Phragmoplast  Stamen hair    60.0±8.1     

The dissociation constants koff and t1/2 were estimated for GFP fusions in various cell types and at various cell cycle stages. n column shows the number of cells analysed in each experiment and Bound indicates the fraction of NtMAP65-1a:GFP initially bound to the microtubules, expressed as a percentage of the total fusion protein.

*

koff estimated by single exponential fit; koff estimated by double exponential fit; Shaw et al., 2003; §Hush et al., 1994; n.d., not determined as values were below the threshold for diffusion seen in the oryzalin-treated control cells.

Fig. 1.

NtMAP65-1a N-terminal GFP fusion protein binds to microtubules in the cells of various organs of 7-day-old Arabidopsis seedlings. (A-D) GFP signal detected with a scanning confocal microscope in root tip (A), hypocotyl (B), leaf epidermis (C) and trichome (D). Bar, 10 μm.

Fig. 1.

NtMAP65-1a N-terminal GFP fusion protein binds to microtubules in the cells of various organs of 7-day-old Arabidopsis seedlings. (A-D) GFP signal detected with a scanning confocal microscope in root tip (A), hypocotyl (B), leaf epidermis (C) and trichome (D). Bar, 10 μm.

NtMAP65-1a does not interact with tubulin dimers in vitro

In the FRAP experiments we assume that GFP:NtMAP65-1a molecules from the non-photobleached areas of the cell have to diffuse into the bleached region and exchange with the GFP:NtMAP65-1a molecules bound to the microtubules. Therefore, the rate of GFP:NtMAP65-1a recovery after photobleaching will depend not only on the dynamics of the GFP:NtMAP65-1a interaction with microtubules, but also on the mobility of the chimera protein in the cytoplasm. However, it is possible that the diffusion of GFP:NtMAP65-1a and the interaction with polymerised microtubules will be affected if there is any interaction of GFP:NtMAP65-1a with the cytoplasmic pool of tubulin dimers. Here, we have used surface plasmon resonance (SPR) to determine if NtMAP65-1a binds the tubulin dimer. SPR is `an electron charge density wave phenomenon that arises at the surface of a metallic film when light is reflected at the film under specific conditions' (www.biacore.com/pdf/TN/TN_1.pd). This phenomenon allows measurement of the interaction between two molecules if one of them is attached to the surface of the reaction chamber bearing a metallic film. We have exploited this phenomenon to analyse the interaction between microtubules or tubulin and MAP65.

The experiment was performed using a Biacore 3000 system equipped with a CM5 chip. NtMAP65-1a was covalently coupled to one channel whereas a second channel was treated in the same way except that NtMAP65-1a recombinant protein was omitted. We also attempted to couple tubulin and taxol-stabilised microtubules to chip surfaces, but both aggregated under the low pH conditions required for coupling. Once prepared, the chips were sequentially loaded with MTSB buffer, a tubulin dimer solution, MTSB buffer, a suspension of taxol stabilised microtubules and again MTSB and the surface plasmon resonance was measured and plotted (Fig. 3). The data show that the injection of the tubulin solution showed no differential interaction between the two channels indicating that there is no interaction between NtMAP65-1a and tubulin dimers. In contrast, when a similar concentration of the microtubule suspension was injected, binding to the NtMAP65-1a channel was significantly greater compared to the control channel and the bound microtubules were slowly dissociated in the subsequent MTSB wash taking approximately 150 seconds for the signal to return to the baseline. This indicated that microtubules interact with the NtMAP65-1a coupled to the chip surface. The assay was repeated three times with different NtMAP65-1a batches and similar results were recorded in all cases. Based on these data we conclude that in the FRAP experiments no interaction occurs between NtMAP65-1a and tubulin and that the speed of recovery is due to the diffusion of the chimera molecules in the cytoplasm and the dissociation rate of bleached molecules off the microtubule surface.

Fig. 2.

GFP:NtMAP65-1a signal recovery after photobleaching. (A) GFP:NtMAP65-1a signal in the hypocotyl of 7-day-old Arabidopsis seedlings. The white rectangle outlines the photobleached area. (B) Time series (top to bottom) of GFP:NtMAP65-1a signal recovery during a FRAP experiment within the area indicated by the rectangle in A. The numbers on the right-hand side indicate the time in seconds when each of the frames was collected with 0 corresponding to the image before the photobleaching onset and 4.766 just after the photobleaching. (C) The first FRAP measurement was taken just before the photobleaching and corresponds to point 0. The grey sector represents the duration of photobleaching, after which 14 images were collected and measured at approximately 4.14-second intervals. The fluorescence signal was measured in 24 cells and expressed as the mean percentage of the signal before photobleaching. The error bars indicate s.d. of the mean. Bar, 5 μm.

Fig. 2.

GFP:NtMAP65-1a signal recovery after photobleaching. (A) GFP:NtMAP65-1a signal in the hypocotyl of 7-day-old Arabidopsis seedlings. The white rectangle outlines the photobleached area. (B) Time series (top to bottom) of GFP:NtMAP65-1a signal recovery during a FRAP experiment within the area indicated by the rectangle in A. The numbers on the right-hand side indicate the time in seconds when each of the frames was collected with 0 corresponding to the image before the photobleaching onset and 4.766 just after the photobleaching. (C) The first FRAP measurement was taken just before the photobleaching and corresponds to point 0. The grey sector represents the duration of photobleaching, after which 14 images were collected and measured at approximately 4.14-second intervals. The fluorescence signal was measured in 24 cells and expressed as the mean percentage of the signal before photobleaching. The error bars indicate s.d. of the mean. Bar, 5 μm.

MAP65 molecules interchange at any site along the length of the microtubule

So far our data suggest that the rate of GFP:NtMAP65-1a turnover in its association with microtubules is not dependent on microtubule treadmilling (as this rate is much faster than microtubule treadmilling) nor is this rate affected by an interaction of NtMAP65-1a with free tubulin dimers. Therefore the signal may recover as the result of either active transport of MAP65 along microtubules or the exchange of the MAP65 molecules randomly along the length of the microtubules. To address these possibilities we examined the pattern of GFP:NtMAP65-1a recovery. A total of 18 microtubule bundles were photobleached in different cells and their recovery monitored (shown for a representative microtubule bundle in Fig. 4). The microtubule recovers completely ∼30 seconds after photobleaching (Fig. 4A,B) indicating that this is sufficient time for the exchange of most, if not all, the attached GFP:NtMAP65-1a molecules. Moreover, the intensity of the GFP:NtMAP65-1a signal along the length of the microtubule at each time point was measured (Fig. 4C). The results show that the signal recovers randomly suggesting that GFP:NtMAP65-1a is exchanged along the microtubule length rather than actively moved along the microtubule.

Fig. 3.

NtMAP65-1a does not interact with tubulin dimers. NtMAP65-1a was covalently attached to the matrix of the Biacore chip, then solutions containing tubulin, buffer alone, taxol-stabilised microtubules and again buffer alone (as indicated at the top of the chart) were sequentially passed over the chip and the binding response or surface plasmon resonance (SPR) was measured over time and plotted (black line). The empty chip surface was used as negative control (grey line). The concentration of the tubulin dimers and microtubules was 0.1 mg/ml.

Fig. 3.

NtMAP65-1a does not interact with tubulin dimers. NtMAP65-1a was covalently attached to the matrix of the Biacore chip, then solutions containing tubulin, buffer alone, taxol-stabilised microtubules and again buffer alone (as indicated at the top of the chart) were sequentially passed over the chip and the binding response or surface plasmon resonance (SPR) was measured over time and plotted (black line). The empty chip surface was used as negative control (grey line). The concentration of the tubulin dimers and microtubules was 0.1 mg/ml.

Fig. 4.

Random recovery of GFP:NtMAP65-1a signal over the microtubule bundle length. (A) The first measurement was taken just before the photobleaching and corresponds to point 0 (the image of the microtubule at this time point is represented in part B point 0). The grey sector represents the duration of photobleaching, after which 35 images were collected and measured at approximately 2.5-second intervals. The fluorescence signal on the chart is expressed in arbitrary values. (B) Time series (left to right) of GFP:NtMAP65-1a signal recovery after photobleaching the microtubules shown in A. The numbers at the top of each image indicate time in seconds when each of the frames was collected with 0 corresponding to the image before the photobleaching onset and 12.0 just after the photobleaching. The images shown follow the microtubule to complete recovery and this occurred after 34.5 seconds. (C) Changes in time (x axes) of the distribution of the fluorescence signal (z axes) along the length of the microtubule (y axes) shown in A and B. The signal intensity is expressed in arbitrary values.

Fig. 4.

Random recovery of GFP:NtMAP65-1a signal over the microtubule bundle length. (A) The first measurement was taken just before the photobleaching and corresponds to point 0 (the image of the microtubule at this time point is represented in part B point 0). The grey sector represents the duration of photobleaching, after which 35 images were collected and measured at approximately 2.5-second intervals. The fluorescence signal on the chart is expressed in arbitrary values. (B) Time series (left to right) of GFP:NtMAP65-1a signal recovery after photobleaching the microtubules shown in A. The numbers at the top of each image indicate time in seconds when each of the frames was collected with 0 corresponding to the image before the photobleaching onset and 12.0 just after the photobleaching. The images shown follow the microtubule to complete recovery and this occurred after 34.5 seconds. (C) Changes in time (x axes) of the distribution of the fluorescence signal (z axes) along the length of the microtubule (y axes) shown in A and B. The signal intensity is expressed in arbitrary values.

Interaction of MAP65 with microtubules during mitosis

Previous immunostaining data have shown that NtMAP65-1a or AtMAP65-1 binds to a subset of microtubules not just in the interphase cortical array but also in the mitotic arrays: in late G2/prophase microtubules of the preprophase band were decorated with this MAP65, no detectable binding of this MAP65 to microtubules was observed in the prometaphase and metaphase spindles but its localisation was concentrated in the midzone of the anaphase spindle and the phragmoplast (Smertenko et al., 2000; Smertenko et al., 2004). We have generated tobacco BY-2 cell lines stably transformed with either GFP:NtMAP65-1a or GFP:AtMAP65-1 and we have followed the localisation of the chimera proteins through mitosis. Generally, localisation of GFP:NtMAP65-1a (Fig. 5A-C) and GFP:AtMAP65-1 (data not shown) was indistinguishable and resembles that shown by immunostaining (Smertenko et al., 2000; Smertenko et al., 2004). That is the chimera proteins bound microtubules in both cortical arrays, a mostly cytoplasmic signal in and around the metaphase spindle and a specific binding to the midzone of the anaphase spindle (Fig. 5C); a dramatic accumulation of the signal in the anaphase spindle midzone occurred during sister chromatid polewards movement (see Movie 1 in supplementary material for the GFP:NtMAP65-1a-transformed cell line) which suggests that the signal for NtMAP65-1a accumulation at the midzone is generated during anaphase B. GFP:NtMAP65-1a also concentrated at the midzone of the phragmoplast where cell plate synthesis is initiated. However, this concentration was most apparent at the early stages of phragmoplast formation (Fig. 5C, 82-132 seconds), later having a more broad distribution on phragmoplast microtubules (224 seconds) and eventually disappearing from the cell plate region as microtubules disassembled (398 seconds).

Fig. 5.

Interaction of NtMAP65-1a with microtubules during cell division. GFP:NtMAP65-1a signal in the preprophase band (A), mitotic spindle (B) and anaphase to telophase transition (C). The numbers in the images in C indicate time in seconds from the beginning of image acquisition. Bar, 10 μm.

Fig. 5.

Interaction of NtMAP65-1a with microtubules during cell division. GFP:NtMAP65-1a signal in the preprophase band (A), mitotic spindle (B) and anaphase to telophase transition (C). The numbers in the images in C indicate time in seconds from the beginning of image acquisition. Bar, 10 μm.

We have analysed the interaction of NtMAP65-1a with the microtubules during the cell cycle in BY-2 cells using FRAP on the tobacco cell lines expressing GFP:NtMAP65-1a (Table 1). The t½ was found to be similar for interphase microtubules (6.95±0.91 seconds, n=23) and for microtubules in the preprophase band (5.92±0.64 seconds, n=19). The turnover of GFP:AtMAP65-1 was similar to that of GFP:NtMAP65-1a in the cortical array (5.93±0.86 seconds, n=20 and 6.95±0.91 seconds; Table 1) and in the mitotic microtubule arrays (data not shown). In the phragmoplast midzone the t½ of GFP:NtMAP65-1a was slightly faster (4.83±0.64 seconds, n=23) than for the cortical arrays and the fraction of GFP fusion protein bound to the microtubules was less than in the preprophase band (75.0% compared to 85.8%; Table 1). In the metaphase spindle the t½ was over sevenfold lower than in all other arrays (0.72±0.21 seconds, n=20). We have compared the recovery time in the metaphase spindle with the recovery time in cells where microtubules were depolymerised with the anti-microtubular drug oryzalin. No filamentous structures were visible in the majority of the cells following a 2-hour treatment with 10 μM oryzalin (data not shown). Disruption of microtubules should cause MAP65 to become cytoplasmic so only free diffusion of GFP:NtMAP65-1a can occur. Indeed, the t½ value in this case was found to be 1.71±0.24 seconds (n=23). This t½ value is greater than that in the metaphase spindle indicating that in this array the majority of the NtMAP65-1a is not bound to microtubules and does not form complexes that can slow its mobility.

We have used tobacco and Arabidopsis MAP65:GFP chimera proteins to study their interaction with microtubules in vivo. Our results show that NtMAP65-1a interacts with polymerised microtubules only and not with the tubulin dimer and that the rate of turnover of the MAP65s on microtubules is high (see Table 1). Rather than being translocated along microtubules, they are added and replaced randomly along the microtubule length. Moreover, the MAP65/microtubule interaction is regulated in a cell cycle dependent manner.

The rate of GFP:NtMAP65-1a turnover on interphase cortical microtubules was found to be similar in Arabidopsis hypocotyl epidermal cells and in rapidly proliferating tobacco suspension culture cells indicating that the interaction of NtMAP65-1a is the same in both homologous and heterologous plant cells, and that the interaction is similar in the cortical microtubules of differentiated and rapidly dividing tissues. The turnover of the GFP:AtMAP65-1 chimeric protein, the Arabidopsis homologue of NtMAP65-1a, was found to be very similar to that of GFP:NtMAP65-1a (see Table 1) suggesting that they share the same mechanism of microtubule interaction. The recombinant AtMAP65-1 and NtMAP65-1b bundle, but neither promote polymerisation nor prevent catastrophes of microtubules in vitro (Smertenko et al., 2004; Wicker-Planquart et al., 2004). Also, the turnover of GFP:NtMAP65-1a in all cell types and microtubule arrays examined was found to be at least fourfold greater than plant tubulin itself. The high rate of microtubule association/dissociation for this MAP65 isotype correlates with it having no impact on microtubule dynamics in vitro. Therefore this MAP will not significantly affect the dynamics of microtubules in vivo and is ideally suited for bundling and crossbridging of plant microtubules during microtubule array formation and reorganisation.

Microtubules have an increased growth and catastrophe rate in the transition from interphase to preprophase band formation. However, the shrinkage rate and the rescue frequency are not affected. The net result is that during this transition the microtubules become shorter and more dynamic (Dhonukshe and Gadella, 2003). We have shown that GFP:NtMAP65-1a localises to the preprophase band throughout its development. Moreover, MAP65 turnover in the preprophase band is similar to that in the interphase cortical array. Together these data suggest that changes in microtubule dynamics do not affect the dynamics of the MAP65 microtubule interaction. Also, if binding of MAP65 is a prerequisite to microtubule bundling, which is the case in in vitro studies, then the same bundling mechanism is likely to occur in the two cortical arrays.

A significant increase in the turnover of GFP:NtMAP65-1a was observed in the metaphase spindle: the t½ value decreased sevenfold. As we have shown that NtMAP65-1a does not interact with tubulin dimers, depolymerisation of microtubules with oryzalin will remove the major binding site for this MAP65 in the cell. In the absence of microtubules the t½ value was greater than in metaphase, indicating that in the metaphase spindle GFP:NtMAP65-1a does not bind to any structure and nor does it make any complexes. Its diffusion is presumably increased owing to the dynamic cytoplasmic flow within mitotic spindles. However, there must be some interaction with as yet unknown factors that inhibits the ability of NtMAP65-1a to bind microtubules. For example, the presence of proteins that compete for the same binding site or post-translational modification of NtMAP65-1a that can alter its structure.

In anaphase and telophase GFP:NtMAP65-1a is observed to concentrate mainly in the midzone of the spindle and the phragmoplast, as was previously shown by antibody staining in tobacco and Arabidopsis tissue culture cells (Smertenko et al., 2000; Smertenko et al., 2004). However, the video of the GFP fluorescence (Movie 1 in supplementary material) shows how dramatic this concentration preceding polewards movement of the chromosomes actually is. We have measured FRAP and found no significant differences in the turnover of NtMAP65-1a on microtubules in the phragmoplast, interphase cortical microtubules and preprophase band. These data indicate that the exchange rate of the NtMAP65-1a or the lifetime of the bridges between microtubules does not depend on the stage of the cell cycle. It would appear that the interaction of NtMAP65-1a with microtubules is regulated on an `on or off' principle and wherever bundling occurs the dynamics are the same. Other proteins known to be involved in the control of plant microtubule organisation in vivo such as katanin (Stoppin-Mellet et al., 2002), MOR1/GEM1 (Twell et al., 2002) or EB1 (Chan et al., 2003) must be able to work in concert with MAP65 to regulate the dynamics and spatial pattern of microtubules bundled by MAP65.

Interestingly, the turnover of the yeast homologue of MAP65, Ase1p was found to be 7.5 minutes (Schuyler et al., 2003), which is almost 100 times slower than NtMAP65-1a and AtMAP65-1. It is known that Ase1p is an important component of the spindle midzone matrix, responsible for the stabilisation and the maintenance of the spindle midzone. In Arabidopsis there are nine MAP65 genes and they form a divergent gene family (Hussey et al., 2002). Moreover, the two isoforms AtMAP65-1 (Smertenko et al., 2004) and AtMAP65-3 (Muller et al., 2004) show differential localisation in the four microtubule arrays, with AtMAP65-3 being restricted to only the mitotic arrays. We have shown previously that AtMAP65-3 is essential for the maintenance of phragmoplast structure (Muller et al., 2004) and it is tempting to speculate that this isoform is more closely related in function to Ase1p than to NtMAP65-1a. Perhaps AtMAP65-3 stabilises subsets of microtubules responsible for anchoring and maintaining the integrity of the spindle and phragmoplast midzone but that AtMAP65-1 plays a more active role in the bundling of dynamic microtubules, helping them retain spatial organisation in dynamic microtubule arrays.

Recently, it has been shown that PRC1, the mammalian homologue of MAP65, binds several kinesin motor proteins: KIF4, MKLP and CENP-E. In KIF-4-deficient cells proteins that normally localise to the midzone, which includes PRC1, were dispersed (Kurasawa et al., 2004). No midzone appears in PRC-1-deficient cells. These results suggest that KIF 4 and PRC1 are essential for the organisation of the spindle midzone and one possibility is that KIF4 translocates PRC1 to the midzone. The FRAP analysis presented in this paper is consistent with the conclusion that NtMAP65-1a associates and dissociates randomly along microtubules rather than being translocated.

In summary these data show that the interaction of the MAP65, NtMAP65-1 (and its Arabidopsis relative AtMAP65-1) with microtubules is very dynamic, that it occurs randomly along the length of microtubules and that it is cell cycle stage dependent. Taken together with the previous finding that NtMAP65-1 and AtMAP65-1 proteins form crossbridges but have no effect on microtubule dynamics in vitro (Chan et al., 1999; Wicker-Planquart et al., 2004; Smertenko et al., 2004), we conclude that both proteins are ideally suited to a role in forming crossbridges in microtubules that need to retain spatial organisation in rapidly reorganising microtubule arrays.

Bulinski, J. C., Odde, D. J., Howell, B. J., Salmon, T. D. and Waterman-Storer, C. M. (
2001
). Rapid dynamics of the microtubule binding of ensconsin in vivo.
J. Cell Sci.
114
,
3885
-3897.
Burns, R. G. and Surridge, C. D. (
1994
). Tubulin: conservation and structure. In
Microtubules
(ed. J. S. Hyams and C. W. Lloyd), pp.
3
-31. New York: Wiley-Liss.
Chan, J., Calder, G. M., Doonan, J. H. and Lloyd, C. W. (
2003
). EB1 reveals mobile microtubule nucleation sites in Arabidopsis.
Nat. Cell Biol.
5
,
967
-971.
Chan, J., Jensen, C. G., Jensen, L. C. W., Bush, M. and Lloyd, C. W. (
1999
). The 65-kDa carrot microtubule-associated protein forms regularly arranged filamentous cross-bridges between microtubules.
Proc. Natl. Acad. Sci. USA
96
,
14931
-14936.
Clough, S. J. and Bent, A. F. (
1998
). Floral dip: a simplified method for Agrobacterium-mediated transformation of Arabidopsis thaliana.
Plant J.
16
,
735
-743.
Davis, S. J. and Vierstra, R. D. (
1998
). Soluble, highly fluorescent variants of green fluorescent protein (GFP) for use in higher plants.
Plant Mol. Biol.
36
,
521
-528.
Dhonukshe, P. and Gadella, T. W. J. (
2003
). Alteration of microtubule dynamic instability during preprophase band formation revealed by yellow fluorescent protein-CLIP170 microtubule plus-end labeling.
Plant Cell
15
,
597
-611.
Geelen, D. N. V. and Inzé, D. G. (
2001
). A bright future for the BrightYellow-2 cell culture.
Plant Physiol.
127
,
1375
-1379.
Granger, C. L. and Cyr, R. J. (
2000
). Microtubule reorganization in tobacco BY-2 cells stably expressing GFP-MBD.
Planta
210
,
502
-509.
Hush, J. M., Wadsworth, P., Callaham, D. A. and Hepler, P. K. (
1994
). Quantification of microtubule dynamics in living plant cells using fluorescence redistribution after photobleaching.
J. Cell Sci.
107
,
775
-784.
Hussey, P. J. (
2004
).
The plant cytoskeleton in cell differentiation and development
. Oxford: Blackwell Publishing.
Hussey, P. J., Hawkins, T. J., Igarashi, H., Kaloriti, D. and Smertenko, A. (
2002
). The plant cytoskeleton: recent advances in the study of the plant microtubule-associated proteins MAP-65, MAP-190 and the Xenopus MAP215-like protein, MOR1.
Plant Mol. Biol.
50
,
915
-924.
Jiang, C. J. and Sonobe, S. (
1993
). Identification and preliminary characterization of a 65kDa higher-plant microtubule-associated protein.
J. Cell Sci.
105
,
891
-901.
Kreis, T. and Vale, R. (
1993
).
Guide Book To The Cytoskeletal And Motor Proteins.
Oxford: Oxford University Press.
Kurasawa, Y., Earnshaw, W. C., Mochizuki, Y., Dohmae, N. and Todokoro, K. (
2004
). Essential roles of KIF4 and its binding partner PRC1 in organized central spindle midzone formation.
EMBO J.
23
,
3237
-3248.
Moore, R. C., Zhang, M., Cassimeris, L. and Cyr, R. J. (
1997
). In vitro assembled plant microtubules exhibit a high state of dynamic instability.
Cell Motil. Cytoskel.
38
,
278
-286.
Müller, S., Smertenko, A., Wagner, V., Heinrich, M., Hussey, P. J. and Hauser, M.-T. (
2004
). The plant microtubule associated protein, AtMAP65-3/PLE, is essential for cytokinetic phragmoplast function.
Curr. Biol.
14
,
412
-417.
Nick, P. (
1998
). Signaling to the microtubular cytoskeleton in plants.
Int. Rev. Cyt.
184
,
33
-80.
Schuyler, S. C., Liu, J. Y. and Pellman, D. J. (
2003
). The molecular function of Ase1p: evidence for a MAP-dependent midzone-specific spindle matrix.
J. Cell Biol.
160
,
517
-528.
Shaw, S. L., Kamyar, R. and Ehrhardt, D. W. (
2003
). Sustained microtubule treadmilling in arabidopsis cortical arrays.
Science
300
,
1715
-1718.
Shelanski, M. L., Gaskin, F. and Cantor, C. R. (
1973
). Microtubule assembly in absence of added nucleotides.
Proc. Natl. Acad. Sci. USA
70
,
765
-768.
Smertenko, A., Saleh, N., Igarashi, H., Mori, H., Hauser-Hahn, I., Jiang, C. J., Sonobe, S., Lloyd, C. W. and Hussey, P. J. (
2000
). A new class of microtubule-associated proteins in plants.
Nat. Cell Biol.
2
,
750
-753.
Smertenko, A. P., Chang, H. Y., Wagner, V., Kaloriti, D., Fenyk, S., Sonobe, S., Lloyd, C., Hauser, M. T. and Hussey, P. J. (
2004
). The Arabidopsis microtubule-associated protein AtMAP65-1: Molecular analysis of its microtubule bundling activity.
Plant Cell
16
,
2035
-2047.
Stoppin-Mellet, V., Gaillard, J. and Vantard, M. (
2002
). Functional evidence for in vitro microtubule severing by the plant katanin homologue.
Biochem. J.
365
,
337
-342.
Twell, D., Park, S. K., Hawkins, T. J., Schubert, D., Schmidt, R., Smertenko, A. and Hussey, P. J. (
2002
). MOR1/GEM1 has an essential role in the plant-specific cytokinetic phragmoplast.
Nat. Cell Biol.
4
,
711
-714.
Ueda, K. and Matsuyama, T. (
2000
). Rearrangement of cortical microtubules from transverse to oblique or longitudinal in living cells of transgenic Arabidopsis thaliana.
Protoplasma
,
213
,
28
-38.
Wicker-Planquart, C., Stoppin-Mellet, V., Blanchoi, L. and Vantard, M. (
2004
). Interactions of tobacco microtubule-associated protein MAP65-1b with microtubules.
Plant J.
39
,
126
-134.