Interpolation of microtubules into cortical arrays during cell elongation and differentiation in roots of Azolla pinnata

Longitudinal sections of roots of Azolla pinnata R. Br. were prepared for electron microscopy so that cortical microtubules could be counted along the longitudinal walls in cell files in the endodermis, pericycle, and inner and outer cortex, and in sieve and xylem elements. With the exception of the xylem, where there are no transverse cell divisions, each file of cells commences with its initial cell and then possesses a zone of concomitant cell expansion and transverse cell division, followed, after completion of the divisions, by a zone of terminal cell differentiation. The cells augment their population of cortical microtubules as they elongate and divide, showing a net increase of up to 0.6 micron of polymerized microtubule length per min. Two main sub-processes were found: (i) When a longitudinal wall is first formed it is supplied with a higher number of microtubules per unit length of wall than it will have later, when it is being expanded. This initial quota becomes diluted as the second sub-process commences. (ii) The cells interpolate new microtubules at a rate which is characteristic of the cell, and, in the endodermis, of the face of the cell, while the cell elongates. Most cell types thus maintain a set density of cortical microtubules while they elongate and divide. Comparisons of endodermal cells in untreated controls, and roots that had been treated with colchicine, low temperature, or high pressure indicate that the initial quota of microtubules, and the later interpolations, and differentially sensitive to microtuble perturbations. Three types of behaviour, all related to changes in the cell walls, were noted as cortex, xylem and sieve element cells entered their respective phases of cell differentiation. The cortical cells expanded in all dimensions, and the interpolation of microtubules diminished or ceased. The sieve elements continued to elongate, and interpolated at a high rate, reaching unusually high densities of microtubules when the cell walls were being thickened. During this period a net increase of 2.0 micron of polymerized microtubule length per min was calculated. Thereafter interpolation ceased and the density of microtubules declined. The sample applied to developing xylem except that, because wall-thickening is localized rather than widespread, the rise and subsequent fall in the density of microtubules was less marked. The data are discussed in relation to the participation of microtubules in wall deposition and to the hypothesis that cortical microtubules arise in discrete zones along the edges of cells.