Fig. 4: Fidelity and speed of microtubule band formation depends on KATANIN.
a–c Maximum projection of dual-labelled mCh-TUA5 and GFP-KTN seedlings in non-induced cells (b) and during PX formation (b, c). KTN1-GFP localizes majorly to forming microtubule bands during early (b) and mid (b) stages of PX formation. Scale bars = 10 µm. d Maximum projections of mCh-TUA5 VND7 during PX formation in wild-type (left) and in ktn1-2 (right). Arrowheads in c and d point toward regions with incomplete band separation. Scale bar = 5 µm. e Fraction of cells classified into early, mid and late stages of PX formation for induced mCh-TUA5 VND7 (left) and ktn1-2 VND7 (right). Dots depict fractions of early and mid-stages of nine seedlings for each genotype. f Relative progression of VND7 (left) and ktn1-2 VND7 (right). Microtubule arrays in VND7 are significantly more progressed (2.1 ± 0.7, mean ± SD, 174 cells from 9 seedlings) compared to the ktn1-2 VND7 (1.7 ± 0.6, 127, 9, **p = 0.0042, Welch’s unpaired, two-sided t-test). Dots depict VND7 progression for nine seedlings for each genotype. g Maximum projection of stained secondary walls of induced wild type (left) and ktn1-2 (right) 48 h after induction. Scale bar = 10 µm. h–k Basic Fuchsin-stained wild-type (h) and ktn1-2 (i-j) xylem files in 6-day-old roots. In ktn1-2, an altered arrangement of PX (i) and MX (j) vessels is observed, PX vessel thicknesses vary (compare vertical lines in h and i) and cell wall bands are wider and more disordered (k, right) as compared to wild type (k, left). Band shapes are indicated by yellow lines. Scale bars = 10 µm. l Band separation is fastest for initially co-aligned transverse arrays (cyan and blue) and slowest for initially longitudinal arrays (red). Simulations assuming default parallel nucleation and no KTN activity (rx = 0 s−1). Lines and margins represent medians ± 16% and 84% percentiles.
