Inhibition of cell expansion enhances cortical microtubule stability in the root apex of Arabidopsis thaliana

Cortical microtubules regulate cell expansion by determining cellulose microfibril orientation in the root apex of Arabidopsis thaliana. While the regulation of cell wall properties by cortical microtubules is well studied, the data on the influence of cell wall to cortical microtubule organization and stability remain scarce. Studies on cellulose biosynthesis mutants revealed that cortical microtubules depend on Cellulose Synthase A (CESA) function and/or cell expansion. Furthermore, it has been reported that cortical microtubules in cellulose-deficient mutants are hypersensitive to oryzalin. In this work, the persistence of cortical microtubules against anti-microtubule treatment was thoroughly studied in the roots of several cesa mutants, namely thanatos, mre1, any1, prc1-1 and rsw1, and the Cellulose Synthase Interacting 1 protein (csi1) mutant pom2-4. In addition, various treatments with drugs affecting cell expansion were performed on wild-type roots. Whole mount tubulin immunolabeling was applied in the above roots and observations were performed by confocal microscopy. Cortical microtubules in all mutants showed statistically significant increased persistence against anti-microtubule drugs, compared to those of the wild-type. Furthermore, to examine if the enhanced stability of cortical microtubules was due to reduced cellulose biosynthesis or to suppression of cell expansion, treatments of wild-type roots with 2,6-dichlorobenzonitrile (DCB) and Congo red were performed. After these treatments, cortical microtubules appeared more resistant to oryzalin, than in the control. According to these findings, it may be concluded that inhibition of cell expansion, irrespective of the cause, results in increased microtubule stability in A. thaliana root. In addition, cell expansion does not only rely on cortical microtubule orientation but also plays a regulatory role in microtubule dynamics, as well. Various hypotheses may explain the increased cortical microtubule stability under decreased cell expansion such as the role of cell wall sensors and the presence of less dynamic cortical microtubules.

Even though the "alignment hypothesis" has been proposed decades ago, a mechanism explaining how cortical microtubules, located inside of the plasma membrane, control the orientation of cellulose microfibrils, outside of the protoplast, remained ambiguous. Several models have been proposed to answer the above-mentioned question [16][17][18]. The direct observation of CESA complexes sliding on the plasma membrane over the cortical microtubules [5] directly supported the model of Heath [16]. Additionally, it has been recently found that cortical microtubules are connected to CESA complexes via the Cellulose Synthase Interacting 1 (CSI1) protein [19][20][21] and Companion of Cellulose Synthase protein 1 and 2 (CC1 and CC2), respectively [22].
Apart from the role of cortical microtubules on cellulose microfibril orientation, several authors supported that CESA activity and cellulose synthesis also exert an effect on cortical microtubule orientation, suggesting thus that cortical microtubule-cellulose microfibril relationship is bi-directional. In particular, experiments with cellulose synthesis inhibitors [23] and genetic analysis of CESA mutants (cesa) revealed that CESA impairment and cellulose deficiency affect cortical microtubule stability and organization (among others [24][25][26]). In addition, Le et al. showed that exposure to ethylene inhibited cell elongation in A. thaliana root tips and consequently altered cortical microtubule orientation [27]. Thorough experiments by Panteris et al. have revealed that the primary factor for microtubule disorientation in mutants with defective cellulose synthesis and plants affected by cellulose synthesis inhibitors is the suppression of cell expansion, which is a direct consequence of cellulose shortage [28,29].
Even though plenty of data about cortical microtubule orientation and cell wall mechanics are now available, information about the possible relationship of cellulose synthesis and/or cell expansion with cortical microtubule stability, manifested as sensitivity or resistance to depolymerizing drugs, remains scarce. Paredez et al. have shown that prc1-20, an A. thaliana cesa6 mutant, is more sensitive to the microtubule depolymerizing drug oryzalin, than the wild-type [25]. A similar observation has been also reported for the CSI1 mutant, csi1-2, which is defective in cellulose synthesis [30].
In order to further investigate if, apart from prc1-20 [25], mutations in other A. thaliana CESA genes affect the sensitivity of cortical microtubules to oryzalin, cesa1, cesa3 and cesa6 mutants were examined in the present study. Surprisingly, the cortical microtubules, in the root cells of the mutants reported here, appeared more persistent against oryzalin, compared to the microtubules of the wild-type. Furthermore, to dissect if this increased persistence is due to defective cellulose synthesis or to decreased cell expansion, chemical treatments with DCB (2, 6 dichlorobenzonitrile; inhibitor of cellulose synthesis, [31]) or Congo red (cellulose binding stain; [32,33]) were also performed. The observations support that the increased persistence of cortical microtubules is associated with inhibition of cell expansion rather than cellulose deficiency.

Chemical treatments
Wild-type seedlings, than, pcesa3 and mre1 were treated by pouring oryzalin solution over the dishes as well as by transplanting in dishes with solid medium supplemented with oryzalin. The rest of the seedlings were treated only via transplanting in dishes with drug-supplemented medium. After germination, 5-7 day-old seedlings were treated for 4 or 6 h with oryzalin, either by pouring an aqueous solution of the drug over the seedlings in the Petri dishes, which were continuously shaken on a rocking platform, or by transplanting the seedlings in dishes with medium supplemented with working oryzalin concentration. The final oryzalin concentration was 200 nM (for longer treatment) or 400 nM (for short treatment; readily diluted from a stock solution of 20 mM in dimethyl-sulfoxide). Seedlings of the wild-type as well as of all the mutants, except rsw1, were treated at room temperature. Since rsw1 expresses its cellulose-deficient phenotype at 30 °C, wild-type and rsw1 seedlings remained for 1 h at a 30 °C chamber before treatment with oryzalin, which was also performed at 30 °C. In addition, wildtype, than, prc1-1 and any1 seedlings were transplanted to medium containing 2 mM colchicine (a concentration usually applied on plant cells; Panteris et al. and references therein [35]) for 1.5 h.
Furthermore, wild-type seedlings were treated with combined applications of 400 nM DCB (diluted from stock 10 mM in dimethyl-sulfoxide) and 200 nM oryzalin. Specifically, wild-type seedlings grown in Petri dishes were transplanted to medium containing 400 nM DCB for 22 h and afterwards were transplanted to medium containing 400 nM DCB + 200 nM oryzalin for another 6 h. Wild-type seedlings transplanted to medium with 400 nM DCB for 28 h as well as wild-type seedlings transplanted to medium containing 200 nM oryzalin for 6 h were used as control samples for the combined treatments.
The same procedure was followed for the combined treatment with 10 mg L − 1 Congo red (G. Grubler & Co., Berlin, Germany) and 200 nM oryzalin. Wild-type seedlings transplanted to medium containing 10 mg L − 1 Congo red for 28 h were used as control samples.

Immunolocalization
Immunolabelling was performed on seedlings untreated or treated as above, with monoclonal rat anti-α-tubulin (YOL1/34, AbD Serotec, Kidlington, UK) and FITCanti-rat (developed in goat, Invitrogen, Carlsbad, CA) as primary and secondary antibody, respectively, diluted in PBS (Phosphate buffered saline, pH 7.4) at 1:40 for wholemount immunofluorescence microscopy. The protocol described in Panteris et al. was applied [29]. The samples were examined with a Zeiss LSM780 confocal laser scanning microscope (CLSM), images were acquired with ZEN2011 software (Carl Zeiss, Munich, Germany), following the manufacturer's instructions. Imaging and processing settings were the same for control and treated samples, both wild-type and mutants.

Fluorescence intensity measurements
Fluorescence intensity quantifications were performed in maximum projections of serial CLSM sections of root tips, in the transition and fast elongation zone [36], using the ImageJ software [37] according to Pappas et al., using the formula for corrected total cell fluorescence (CTCF) [36,38]. A total of at least thirty individual cells from three root tips per treatment were used for the fluorescence intensity calculation (10 cells (n) were measured in each root tip). All experiments were performed in three technical and biological repeats. Statistical significance was analyzed by non-parametric Student's t-test, with significance level set at p < 0.005.

Effects of oryzalin on cortical microtubules in cesa3 mutants
To determine whether mutations in CESA3 subunit affect cortical microtubule stability, microtubule depolymerization by oryzalin was assessed. In this study, A. thaliana root apex was classified into four distinct zones, the meristematic, transition (also referred as distal elongation zone), fast elongation (also known as elongation zone) and growth terminating zone, according to Verbelen et al. [39]. Epidermal cells of the transition and fast elongation zone were studied for cortical microtubule stability, since cell divisions in them are scarce and microtubules are uniformly distributed. Besides, as one of the first effects of oryzalin is the disruption of the cell division-specific microtubule arrays [40][41][42][43], the meristematic zone was also monitored to verify whether seedlings were affected by the drug. The features revealing cortical microtubule stability or susceptibility in the transition and fast elongation zone were their integrity, as exhibited by the degree of fragmentation, and their density in each cell, as revealed by the relevant fluorescence intensity measurement. The orientation and density of cortical microtubules in wild-type seedlings, as well as in the heterozygous than and homozygous, rsw1 and prc1-1 mutants, have been shown in Panteris et al. [28,29].
Firstly, we investigated cortical microtubule stability in heterozygous than (cesa3; referred to as than/+) seedlings. Cortical microtubules of than/+ roots, under any procedure of treatment (see Methods), appeared more stable than those of the wild-type ( Figs. 1 and 2). In the meristematic zone of both wild-type and than/+, abnormal cell divisions, disoriented phragmoplasts and incomplete cell walls could be observed (Figs. 1a and c  and 2a, b). In transition and fast elongation zone cells of than/+ roots, cortical microtubules were more integral and densely arranged, compared to those of the wild-type (Figs. 1b, d and 2c, d). The integrity of cortical microtubules was further verified by the fluorescence intensity measurements (Fig. 3). Taken together, cortical microtubules appear more stable in than/+ root cells than in those of the wild-type. This indicates that there might be a correlation between cortical microtubule stability and CESA function or cellulose content, as than/+ is characterized by reduced cellulose content and whole plant dwarfism [14].
Subsequently, we investigated whether the increased resistance to oryzalin is restricted to the than/+ mutant only, or it is a general feature of cesa3 mutants. Therefore, the mre1 as well as pcesa3, a transgenic plant expressing the than allele [14], were treated with 400 nM oryzalin for 4 h. Both mutants exhibited cortical microtubules more persistent against oryzalin than the wild-type. Specifically, in transition and fast elongation zone cells of the mutants, cortical microtubules appeared longer and more integral (Fig. 4b, c) than those of wild-type roots (Fig. 4a). In addition, higher fluorescence intensity, as a result of more tubulin polymers, was observed in the mutants, compared to the wild-type (Additional file 1: Fig. S1), which confirms the presence of more and longer microtubules. These observations indicate that cortical microtubule stability is a common feature among cesa3 mutants.

The effect of oryzalin on cortical microtubules in cesa1 and cesa6 mutants
In order to examine whether cortical microtubule stability is limited to cesa3 mutants or it might be a general feature of primary cell wall-related cesa mutants, rsw1 and any1 (cesa1 mutants), as well as prc1-1 (cesa6 mutant) seedlings were treated with 200 nM oryzalin for 6 h. In the transition and fast elongation zones of any1 and prc1-1, more integral microtubules could be observed (Fig. 5b, c), compared to those of wild-type roots. At 30 °C, root cells of rsw1 showed reoriented but integral cortical microtubules, after treatment with oryzalin, while wild-type roots exhibited fewer and shorter microtubules under the same treatment ( Fig. 5e; cf. 5d). The increased persistence of cortical microtubules against oryzalin in the mutants was also confirmed by fluorescence intensity quantification (Additional file 2: Fig. S2). It can, therefore, be concluded that increased cortical microtubule persistence against oryzalin is a common feature of primary cell wall-related cesa mutants.
Our results are different from those of Paredez et al., who described the prc1-20 (cesa6 mutant) as hypersensitive to oryzalin [25]. This statement was supported by reorientation of cortical microtubules in root cells, as well as by root swelling when exposed to oryzalin. The differences between our findings and the above may be due to different observation approaches and/or the specific criteria for assessing stability. In the above work [25], a GFP-MAP4 marker was used to visualize microtubules, while here α-tubulin immunostaining was performed. Microtubule-Associated Protein 4 (MAP4) decorates microtubules and its fusion to GFP allows microtubule visualization [24]. It has been considered that overexpression of GFP-MAP4 fusion proteins may produce an altered developmental phenotype of microtubule dynamics and orientation [24,44]. In contrast to GFP markers, immunostaining provides more precise and exact microtubule phenotype as it does not interfere with expression and/or function.
In addition, Paredez et al. did not observe any microtubule disassembly but reorientation [25]. In our study, however, cortical microtubules underwent depolymerization without reorientation. In terms of the properties of anti-microtubule drugs, this behavior appears more expected [40] than microtubule reorientation.

The effect of colchicine on cortical microtubules of cesa1, cesa3 and cesa6
In order to further investigate whether the increased cortical microtubule stability of the above cesa mutants is limited to oryzalin or it is generally manifested against anti-microtubule drugs, treatments with colchicine, another microtubule-depolymerizing drug were administered. Colchicine is the most "classic" and "notorious" drug that disrupts microtubules (among others [45,46]). Wild-type, any1, than/+ and prc1-1 seedlings were treated with 2 mM colchicine for 1.5 h. The concentration of the drug applied was much higher than that of oryzalin, since colchicine was reported to be effective on plants at higher concentrations (among others [45,47]). All the seedling roots affected by colchicine showed extensive microtubule depolymerization (Fig. 6). The microtubule remnants were scarce and short. However, the fragments of microtubules appeared longer in the mutants than in the wild-type (Fig. 6b-d, cf. 6a). These results were further verified by the fluorescence intensity measurements (Additional file 3: Fig. S3). Increased cortical microtubule stability may thus be considered as a general feature among cesa mutants, irrespective of the drug or the CESA subunit mutation.

Cortical microtubule stability in the pom2-4 mutant
So far, mutants with deficient CESA subunits and decreased cellulose synthesis, except any1, which shows reduced cell wall crystallinity and CESA velocity but physiological cellulose content [15], have been studied. In order to discriminate if the increased microtubule stability resulted from malfunction of the CESAs per se or from defective cellulose synthesis, the csi1 mutant pom2-4 was treated with oryzalin. In this csi1 mutant, defects in growth and cell expansion occur due to decreased cellulose content, although the CESA subunits are normal [19,20,48]. After treatment with 200 nM oryzalin for 6 h, cell divisions appeared affected in the meristematic zone of both wild-type and pom2-4 roots (Fig. 7a, c). In transition and fast elongation zone of roots treated as above, cortical microtubules of pom2-4 appeared longer and more integral, in comparison with those of the wildtype (Fig. 7d, cf. 7b). Increased fluorescence intensity was observed in pom2-4, as well (Additional file 2: Fig. S2). Contrarily to our results, Mei et al. reported that cortical microtubules in hypocotyl cells of csi1-2 were more sensitive to oryzalin than those of the wild-type [30]. However, their observations were made by GFP-MAP4 imaging, which may affect microtubule properties and behavior (see above). In addition, root cells may react differently to oryzalin than hypocotyl cells. Taken together, our observations support that increased cortical microtubule stability, which is common in cesa and csi1 mutants, should rather be attributed not to CESA malfunction per se, but to decreased cellulose content in the cell wall.

Inhibition of cellulose synthesis increases cortical microtubule stability in wild-type seedlings
To address the question whether defective cellulose biosynthesis may improve cortical microtubule stability, wild-type roots were examined after combined treatment with DCB and oryzalin. DCB is a synthetic herbicide causing cessation of CESA complex movement, resulting in inhibition of cellulose synthesis and deposition (among others [49,50]). For this experiment, 4-dold wild-type seedlings were transferred to a medium containing 400 nM DCB for 22 h and then transferred to medium with 400 nM DCB and 200 nM oryzalin for 6 h. Seedlings treated for 6 h with 200 nM oryzalin alone were used as control (Fig. 8). Root tip bulging and inhibition of seedling growth was observed in DCB and combined DCB and oryzalin treatment (Additional file 4: Fig. S4). Abnormal cell divisions and incomplete cell walls in oryzalin and in the combined treatment prove the effect of oryzalin in the roots (Fig. 8a, b). The length of all root cell types appeared affected after treatment with DCB, and the cells appeared shorter and flattened, compared to untreated root cells ( Fig. 8b; cf. 8a, 8d, 8e; cf. 8c). As shown in Fig. 8d, DCB did not affect cortical microtubules. Wild-type roots subjected to combined treatment exhibited cortical microtubules as integral as roots treated with DCB only in the transition and fast elongation zones ( Fig. 8e; cf. 8d), not depolymerized as in oryzalin-treated roots (Fig. 8c). Increased fluorescence intensity was observed in DCB and combined DCB and oryzalin treatments compared to the control (Fig. 9).

Inhibition of cell expansion results in more stable cortical microtubules in the transition and fast elongation zones
A common feature of cellulose-deficient mutants, as well as of DCB-treated wild-type seedlings, is the inhibition of cell elongation [9,10,14,34,49]. In addition, although the any1 mutant of cesa1 is characterized by normal cellulose content, it exhibits decreased cell elongation due to defective cellulose crystallinity [15]. In order to clarify whether inhibition of cell elongation might be the main cause for increased microtubule stability, the effect of Congo red, a dye that inhibits cell expansion, was investigated (Fig. 10). Congo red, binds to cellulose without affecting its biosynthesis but preventing glucan chain crystallization and, consequently, typical microfibril formation [32,33,51]. Wild-type seedlings were treated either with 200 nM oryzalin for 6 h, or they were first transplanted for 22 h to medium supplemented with 10 mg L − 1 Congo red and then transplanted to medium containing 10 mg L − 1 Congo red and 200 nM oryzalin for 6 h. Root tips of seedlings treated with Congo red or Congo red and oryzalin exhibited intense curling (Additional file 4: Fig.  S4). Abnormal cell divisions observed in both oryzalin and combined treatments revealed the depolymerizing effect of oryzalin (Fig. 10a, b). Congo red application resulted in decreased cell elongation in the root apex developmental zones without affecting cortical microtubules (Fig. 10d). After the combined Congo red and oryzalin treatment, cortical microtubules in transition and fast elongation zone cells appeared integral and almost unaffected by oryzalin ( Fig. 10e; cf. 10d), in contrast to those of roots treated with oryzalin alone, in which cortical microtubules appeared fragmented and partially depolymerized (Fig. 10c). The increased integrity in Congo red and combined Congo red and oryzalin treatments, compared to the control, was verified by fluorescence intensity measurements (Fig. 9) and was found statistically significant (p < 0.0001). According to these findings, it may be concluded that inhibition of cell expansion, irrespective of the cause, results in increased microtubule stability. Fisher and Cyr were the first to describe a bidirectional relationship between cortical microtubules and the cell wall [23]. According to them, the mechanical properties of cell wall, rather than cellulose biosynthesis, affect cortical microtubule stability and organization. When Nicotiana tabacum BY-2 protoplasts were treated with isoxaben (a cellulose biosynthesis inhibitor), microtubules were randomly organized. Cultivation in the absence of isoxaben resulted in highly organized cortical microtubules [23]. Since then, it was supported by several authors that cellulose biosynthesis provides spatial cues for cortical microtubule organization. Chu et al. and Paredez et al. reported alterations in cortical microtubule organization and stability in cesa mutants, concluding that cellulose synthesis may play a regulatory role in microtubule organization [24,25]. On the other hand, Le et al. supported that inhibition of cell elongation in ethylene-treated roots results in microtubule reorientation [27]. In agreement with the latter work, Panteris et al., have demonstrated that chemical, mechanical or genetic inhibition of cell expansion affects cortical microtubule orientation in A. thaliana root apex, supporting the bidirectional relationship of cell wall and cortical microtubules [28,29].
Various hypotheses may explain the increased cortical microtubule stability under decreased cell expansion. It has been reported that cell wall sensors allow cortical microtubules to detect and respond to stress signals [52,53]. Such sensors include the THESEUS kinase, which is responsible for detecting cellulose deficiency, irrespective the cause, and acting as a cell wall integrity sensor [54]. In addition, another kinase, FERONIA, was reported to be important for cell wall integrity perception, in relation to cell and whole plant growth [55]. According to another hypothesis, it may be possible that, due to decreased cell expansion, cortical microtubules become more stationary. New cellulose microfibrils are deposited during cell elongation, between the existing ones [56]. Concurrently, newly-formed microtubules are recruited to maintain the density of cortical microtubule array. When cell expansion is decreased, the above procedure pertains at slow rate, possibly exerting a pausing effect on cortical microtubules, which remain stable, also interconnected by MAP bridges [57]. As a result, tubulin dimers are not released, while new ones are not added: consequently, tubulin-binding anti-microtubule drugs, such as oryzalin and colchicine exert a less pronounced effect on microtubules.

Conclusions
This study provides evidence that inhibition of cell expansion, irrespective of the cause, enhanced the stability of cortical microtubules. It is therefore supported that Cortical microtubules appear integral, not disrupted as in roots treated with oryzalin alone (e compare to c). Scale bars: 20 μm