Katanin-Dependent Microtubule Ordering in Association with ABA Is Important for Root Hydrotropism

Root hydrotropism refers to root directional growth toward soil moisture. Cortical microtubule arrays are essential for determining the growth axis of the elongating cells in plants. However, the role of microtubule reorganization in root hydrotropism remains elusive. Here, we demonstrate that the well-ordered microtubule arrays and the microtubule-severing protein KATANIN (KTN) play important roles in regulating root hydrotropism in Arabidopsis. We found that the root hydrotropic bending of the ktn1 mutant was severely attenuated but not root gravitropism. After hydrostimulation, cortical microtubule arrays in cells of the elongation zone of wild-type (WT) Col-0 roots were reoriented from transverse into an oblique array along the axis of cell elongation, whereas the microtubule arrays in the ktn1 mutant remained in disorder. Moreover, we revealed that abscisic acid (ABA) signaling enhanced the root hydrotropism of WT and partially rescued the oryzalin (a microtubule destabilizer) alterative root hydrotropism of WT but not ktn1 mutants. These results suggest that katanin-dependent microtubule ordering is required for root hydrotropism, which might work downstream of ABA signaling pathways for plant roots to search for water.

The regulatory mechanism of the dynamic instability of cortical microtubules, switching between shrinkage and growth by polymerization, is conserved in eukaryotes [14,15]. In higher plants, the rearrangements of cortical microtubule arrays occur during cell growth and division, and the cortical microtubule arrays direct the movement of the plasma membrane-localized cellulose synthase for cell-wall patterning [16][17][18][19][20][21]. Previous studies have demonstrated that cortical microtubule dynamics are important for the directional growth of roots upon environmental stimuli, such as light, temperature, and gravity [22][23][24]. After gravistimulation, auxin is rapidly redistributed to the lower side of the root, leading to the reorientation of microtubule arrays toward a longitudinal direction of the root, further resulting in growth inhibition in the lower side and, thus, root bending [23,25]. Endogenous or exogenous application of auxin leads to a rapid rearrangement of cortical microtubules toward a longitudinal direction in etiolated hypocotyls and roots, resulting in growth inhibition in higher plants [26]. In gravistimulated Physcomitrella patens, a microtubule reoriented and gathered mainly on the lower flank of the tips of caulonemal filaments [27]. Taken together, the rearrangement of microtubule arrays is proposed as a way to direct growth toward the gravity vector via the regulation of auxin gradients in plant roots [22,26]. However, the role of cortical microtubules in controlling root hydrotropic responses remain an issue to be elucidated.
Microtubule-associated proteins (MAPs) bind directly to microtubules and regulate microtubule organization and the direction of cell expansion in plants [28]. MAPs control and fine tune microtubule dynamic instability, which is essential for the orientation of cell and directional growth [29]. The well-characterized microtubule-severing protein KATANIN, an AAA ATPase protein, disassembles microtubule minus-ends from the nucleation sites to depolymerize the minus-ends of cytoskeletal polymers and mediates microtubule spatial organization [28,29]. KATANIN1 (KTN1) maintains cortical microtubule ordering in Arabidopsis, and the loss-of-function mutant ktn1-2 shows disordered cortical microtubule arrays and reduced cell division [30]. KTN1 is also critical for cells in response to mechanical stress, which amplifies the differences in the rate of cell elongation between neighboring cells in Arabidopsis [31].
In this study, we show that KTN1 plays an important role in root hydrotropism. The cortical microtubule arrays in wild-type (WT) roots were rearranged after hydrostimulation treatment, but the cortical microtubule arrays in the ktn1 mutant were not able to be reorganized, resulting in a defect in root hydrotropic bending. In addition, ABA signaling enhanced the root hydrotropism of WT but not ktn1 mutants, indicating that cortical microtubule ordering might work downstream of ABA signaling pathways for root hydrotropic responses.

Hydrotropic Bending of WT Col-0 Roots Is Reduced by Oryzalin Treatment
Root hydrotropism is the directional growth of roots bending toward soil moisture. Baral et al. [32] recently showed that tissue bending requires microtubule array rearrangements in Arabidopsis. In this study, we first examined the effects of the microtubule destabilizer oryzalin (100 nM and 200 nM), which depolymerizes microtubules and decreases the number of microtubule arrays [33], on the root hydrotropic bending of wild-type (WT) ecotype Col-0 in the hydrotropism assay [5]. During a time course of 24 h, the hydrotropic bending curvature of WT roots reached its maximum after 11 h of hydrostimulation in our assay ( Figure 1A). We observed the average root-bending angles of approximately 19, 29, and 18 • after 5, 11, and 24 h of hydrostimulation, respectively. Interestingly, when 100 nM of oryzalin was applied, the hydrotropic curvature of WT roots reached its maximum at 5 h and then remained the same until the end of our assays. The average root-bending angles of 16, 15, and 15 • were observed after 5, 11, and 24 h of hydrostimulation, respectively. As a result, the oryzalin-treated roots showed significantly reduced hydrotropic bending at 11 h compared with the roots under mock treatment condition ( Figure 1A). In addition, there were no significant differences in root growth between the 100 nM oryzalin-treated and untreated roots within 11 h, suggesting that attenuated hydrotropic bending was not caused by retarded root growth. The inhibition effect of oryzalin treatment on root growth was observed only after 11 h of oryzalin treatment, at which time no significant difference on the hydrotropic curvatures between treatments can be observed ( Figure 1A). These results indicated that the inhibitory effects of oryzalin on the root hydrotropic response occurred before root growth was attenuated, suggesting that proper organization of microtubule arrays is essential for achieving the full bending of hydrostimulated roots. On the other hand, the gravitropic bending of WT roots was not affected in the presence of 100 nM oryzalin ( Figure 1B), which is consistent with previous publications [34][35][36]. The untreated and oryzalin-treated roots displayed similar bending angles, which are on average approximately 50, 80, and 80 • after 5, 11, and 24 h of gravistimulation. Additionally, root growth was severely attenuated upon the 200 nM oryzalin treatment, which is not a favorable condition to justify the role of oryzalin in either root hydrotropism or gravitropism ( Figure 1A,B). Taken together, these results demonstrated that microtubule depolymerization caused by oryzalin treatment had greater effects on root hydrotropism than root gravitropism in Arabidopsis.

Katanin Is Required for Root Hydrotropism in Arabidopsis
Since katanin, a microtubule severing enzyme, is known to play an important role as a catalyzer of tissue bending, we performed hydrotropic assays by using Arabidopsis KTN1 mutants ktn1-4 and ktn1-6, which exhibit compromised microtubule turnover [30,33], to examine if katanin-dependent microtubule organization is involved in root hydrotropism. The results showed that in the 5 days after germination (dag) ktn1-4 and ktn1-6 seedlings displayed significantly reduced root hydrotropic responses compared with that of WT ( Figure 2A), but the root gravitropic curvatures in ktn1-4 and ktn1-6 seedlings were not significantly influenced ( Figure 2B). The root growth of ktn1-4 and ktn1-6 seedlings is also much slower than that of WT ( Figure 2A,B). These results suggested that microtubule dynamics is critical for hydrotropic responses but not for gravitropic responses in Arabidopsis roots.
To understand how cortical microtubules are rearranged during root hydrotropic responses, we next analyzed cortical microtubule ordering in roots of WT and the ktn1 mutant. The transgenic Arabidopsis lines expressing α-tubulin 6 fused to green fluorescent protein driven by a 35S promoter in Col-0 (GFP-TUA6/Col-0) and ktn1 background (GFP-TUA6/ktn1) were used to visualize cortical microtubules by using confocal microscopy [30,37]. After hydrostimulation treatments, the cortical microtubules in the cells of the root elongation zone were rearranged from perpendicular arrays into an oblique orientation along the root growth axis, whereas the orientation of the cortical microtubule arrays in the root maturation and transition zones remained the same (Figures 3A,C,E and S1A,B). On the contrary, cortical microtubules in the root cells of the ktn1 mutant consistently showed disordered organization before and after hydrostimulation and were even more transverse in the hydrotropism assay ( Figure 3B,D,F). Taken together, with the reduced hydrotropic responses in the ktn1 mutant, these data suggest that cortical microtubule ordering is required for root hydrotropism.

Katanin Is Required for Root Hydrotropism in Arabidopsis
Since katanin, a microtubule severing enzyme, is known to play an important role as a catalyzer of tissue bending, we performed hydrotropic assays by using Arabidopsis KTN1 mutants ktn1-4 and ktn1-6, which exhibit compromised microtubule turnover [30,33], to examine if katanin-dependent microtubule organization is involved in root hydrotropism. The results showed that in the 5 days after germination (dag) ktn1-4 and ktn1-6 seedlings displayed significantly reduced root hydrotropic responses compared with that of WT (Figure 2A), but the root gravitropic curvatures in ktn1-4 and ktn1-6 seedlings Error bars indicate the SE; n ≥ 36. Asterisk indicates significant difference from the WT at * p < 0.05 or ** p < 0.01 (Student's t-test).

Katanin-Regulated Microtubule Ordering Might Work Downstream of ABA in Root Hydrotropism
To investigate the correlation between ABA signaling pathways and cortical microtubule dynamics during root hydrotropic response, we analyzed the root hydrotropic response of WT, Qabi2-2 (hab1-1abi1-2abi2-2pp2ca-1 quadruple mutant), ktn1-4, and miz1 mutants in the presence of 1 µM ABA and 100 nM oryzalin upon the hydrotropism assay. MIZU-KUSSEI1 (MIZ1) was essential for root hydrotropism [1], and the Qabi2-2 quadruple mutant showed increased root hydrotropic responses [2,6]. We observed that the root curvature of the WT, but not the Qabi2-2 mutant, was enhanced by 1 µM ABA treatment. This may be because Qabi2-2 has already stayed in an ABA active status, and extra ABA (1 µM) application cannot promote root curvature further in the hydrotropism assay ( Figure 4D). Nevertheless, root hydrotropic bending in WT and the Qabi2-2 mutant was significantly reduced by the application of oryzalin ( Figure 4D). Interestingly, the supplements with indicated concentrations of ABA or oryzalin had no effect on root gravitropic bending in WT, Qabi2-2, ktn1-4, and miz1 mutants ( Figure 4E). These results indicated that cortical microtubule ordering mainly functions in ABA-enhanced root hydrotropism but not in root gravitropism. were not significantly influenced ( Figure 2B). The root growth of ktn1-4 and ktn1-6 seedlings is also much slower than that of WT (Figure 2A,B). These results suggested that microtubule dynamics is critical for hydrotropic responses but not for gravitropic responses in Arabidopsis roots. To understand how cortical microtubules are rearranged during root hydrotropic responses, we next analyzed cortical microtubule ordering in roots of WT and the ktn1 mutant. The transgenic Arabidopsis lines expressing α-tubulin 6 fused to green fluorescent protein driven by a 35S promoter in Col-0 (GFP-TUA6/Col-0) and ktn1 background (GFP- the root maturation and transition zones remained the same (Figures 3A,C,E and S1A,B). On the contrary, cortical microtubules in the root cells of the ktn1 mutant consistently showed disordered organization before and after hydrostimulation and were even more transverse in the hydrotropism assay ( Figure 3B,D,F). Taken together, with the reduced hydrotropic responses in the ktn1 mutant, these data suggest that cortical microtubule ordering is required for root hydrotropism.    To confirm the roles of cortical microtubule arrays in regulating ABA signaling pathways during root hydrotropic responses, we performed high-throughput RNA sequencing by a BGISEQ platform by using Arabidopsis root tips (<5 mm from the root cap) of Col-0 WT and ktn1-4 mutant seedlings in the absence or presence of 1 µM ABA as materials, which produced more than 1.08 G clean reads mapped onto the Arabidopsis genome sequence, and at least 25,868 gene expressions were detected from root tips of the WT and ktn1-4 mutant. A total of 44 and 700 differentially expressed genes (DEGs) (probability ≥ 0.95) were found in roots of WT and the ktn1-4 mutant on the split agar-based hydrotropism assay, respectively ( Figure 4A-C). Among these genes, 24 and 429 of DEGs displayed higher expressions, whereas 20 and 271 genes demonstrated lower expression in roots of WT and the ktn1-4 mutant under the hydrotropism assay, respectively ( Figure 4A,B). A Venn diagram analysis of these DEGs exhibited 13 common DEGs between Col-0-NM (NM: normal medium) vs. Col-0-HT (HT: hydrostimulation treatment) and ktn1-NM vs. ktn1-HT ( Figure 4C). In the presence of 1 µM ABA, the ABA-related DEGs between WT Col-0 and ktn1-4 root tips were observed after hydrostimulation treatment ( Figure 4F). Among them, interestingly, the application of 1 µM ABA reduced the expression levels of the negative regulators (i.e., CLO4 (AT1G70670)) but enhanced the expression levels of the positive regulators (i.e., FBA2 (AT4G38970)) in the ABA signaling pathway in WT, compared with those of the ktn1-4 mutant in the hydrotropism assay ( Figures 4F and 5A,B), indicating that ABA treatment stimulated root hydrotropism through the ABA signal transduction pathway in WT but not the ktn1-4 mutant. On the one hand, the mRNA level of CLO4 (Caleosin 4) is downregulated in response to ABA and salt stress [38]. The clo4 mutants were more tolerant to drought than the wild type, whereas CLO4-overexpressing (Ox) lines showed hyposensitivity to those abiotic stresses [38]. On the other hand, the mRNA level of FBA2 (fructose 1,6-biphosphate aldolase 2) is upregulated following exposure to exogenous ABA, high salinity, and drought and may play key roles in ABA and stress signaling in plants [39]. Moreover, the Tyr phosphorylation level of FBA is modulated by ABA [40]. These results suggest that the lack of katanin might inhibit the gene expression of ABA-related signaling. Taken together, these data suggested that katanin-regulated microtubule ordering is essential for root hydrotropism and might work downstream of the ABA signaling pathway during root hydrotropic bending.

Plant Material and Growth Conditions
Seeds of Arabidopsis thaliana wild-type (WT) ecotype Columbia-0 (Col-0) and ktn1 mutants (ktn1-4 and ktn1-6; AT1G80350), which were acquired from Dr. Deshu Lin and Dr. Rishikesh P. Bhalerao, respectively, were surface-sterilized with~5% (v/v) sodium hypochlorite for 3 min. Subsequently, seeds were washed five times with sterile water and sown on plates containing 1/2 Murashige and Skoog (MS) medium with 0.8% (w/v) and agarose 1% (w/v) sucrose. ktn1-4 and ktn1-6 mutants are in the Col-0 background. Plates were kept at 4 • C for 2 days for vernalization. After that, the seeds were grown vertically for 5 days and then transferred to 1/2 MS medium in the absence or presence of 100 nM oryzalin or 1 µM ABA for vertical growth lasting 11 or 24 h in the hydrotropism assay, as described by Miao et al. [5]. The Arabidopsis seedlings were placed in growth chambers at light dark cycles (16 h light (23 • C) and 8 h dark (21 • C)).

RNA-Sequencing Assay
For RNAsequencing, WT and ktn1-4 seedlings were germinated in 1/2 MS medium for 5 days and then transferred to the hydrotropism assay in the absence or presence of 1 µM ABA for vertical growth lasting for 11 h. Approximately 1.0 g of root tips (<5 mm from the root cap) was harvested, and each accession included three biological replicates. The collected samples were immediately frozen in liquid nitrogen. Then, total RNA was extracted from root tissues by a TRIzol reagent (Invitrogen, Carlsbad, CA, USA). The quality and quantity of the total RNA were determined by an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). After that, first-strand cDNA was synthesized and sequenced according to Miao et al. [5]. Six libraries, Col-0-NM [seedlings grown in the normal medium (NM)], Col-0-HT (hydrostimulation treatment), ktn1-NM, and ktn1-HT were constructed. The RNA-sequencing data generated in this study have been deposited at NCBI in the Sequence Read Archive (SRA) database under the accession number PRJNA610935 (https://www.ncbi.nlm.nih.gov/sra/PRJNA610935; accessed on 18 January 2022).

Quantitative (Q) RT-PCR
The RNA was isolated for 5 days after germination (dag) with WT and ktn1 seedlings after transferring them to NM or the hydrotropism assay for vertical growth lasting 11 h by using the TRIzol reagent (Product No. 93289, Sigma-Aldrich, St. Louis, MO, USA). The firststrand cDNA was generated by 1 µg RNA, the oligo(dT) 15 primer, and M-MLV reverse transcriptase (Promega, Madison, MI, USA). Each reaction contains the same aliquot of the first-strand cDNA as a template with gene-specific primers. Actin2 was used as an internal control and analyzed by the same cDNA and SYBR green, which includes four technical replicates. The primers are listed in Table S1.

Drug Treatments
Stock solutions (20 mM) of the microtubule-disrupting drug oryzalin were dissolved in dimethyl sulfoxide (DMSO) and added to MS medium at the final concentrations indicated in the text. The A. thaliana ecotype Col-0 and ktn1 loss-of-function mutant were sown in 1/2 MS medium (1% (w/v) sucrose and 0.8% (w/v) agarose) in vertically positioned Petri dishes containing the microtubule-disrupting drug oryzalin and grown for the indicated times before the following experiments. DMSO treatment was used as a control.

Microscopy
The transgenic A. thaliana lines expressing GFP-TUA6 were used to visualize microtubules in root cortical cells. The ktn1 mutant was crossed with GFP-TUA6 (Lin et al. [30]). In the split/agar-based hydrotropism assays, the cortical microtubules in roots of both genotypes were fixed for 1 h at the indicated time in a microtubule-stabilizing buffer including 8% (v/v) formaldehyde according to Hou et al. [35] with a minor revision and observed by using a Leica SP8 confocal laser-scanning microscope (Leica, Heidelberg, Germany) with an excitation at 488 nm and emission at 535 nm.