We pay tribute to the seminal paper 'A microtubule in plant cell fine structure' by Myron C. Ledbetter and Keith R. Porter (1963) by summarizing the very limited knowledge of plant cell ultrastructure that we had prior to that publication, and, by way of our three retrospective accounts, show how this paper stimulated and influenced subsequent research on plant microtubules. Micrographs of historical interest are presented that are either previously unpublished or from primary publications.

We discuss models for production of tubulin flux in kinetochore microtubules. Current models concentrate solely on microtubules and their associated motors and enzymes. For example, in some models the driving force for flux is enzymes at the poles and the kinetochores; in others the driving force is motor molecules that are associated with a stationary spindle matrix. We present a different viewpoint, that microtubules are propelled poleward by forces arising from the spindle matrix, that the forces on the microtubules "activate" polymerising and depolymerising enzymes at kinetochores and poles, that matrix forces utilise actin, myosin, and microtubule motors, and that the matrix itself may not necessarily be static.

During investigations into whether diatoms might have some form of cytoplasmic spatial determinants, we unexpectedly encountered what appears to be an exception to the rule that valve formation must be preceded by a mitosis. Cells of Ditylum brightwellii were plasmolysed so that the cytoplasm was reduced to a small spherical cytoplast unattached to the frustule. Recovery was followed using time-lapse video microscopy. In interphase cells, recovery was initiated by filopodia growing from the cytoplast and attaching to the inside corners of the frustule. Later, the protoplast reinflated and recovered completely. If cells were plasmolysed during valve formation, the partially formed valves were shed from the cytoplast. During subsequent recovery, the filopodia completely ignored the partially formed valves, extending past them. After some hours during which they were unsuccessful in locating the corners of a mature valve, the filopodia retracted and the two cytoplasts first shrank, then slowly expanded so that daughter cells almost touched each other. Then they slowly retracted again, but this time, secreted a new valve. These cells subsequently divided normally. We found no evidence that this second phase of valve secretion is preceded by mitosis. Our interpretation is that the cytoplasts, after being disassociated from their normal spatial cues, somehow detect that a normal valve is not present in one half of the daughter cells, and that this stimulates the cells to restart the whole morphogenetic cycle that produces a new valve.

Mitosis and cytokinesis in the pennateDiatoma is described. Prior to division, a doubled “Persistent Polar Complex” (PPC), the focus of numerous cytoplasmic microtubules, migrates from near the nucleus to one side of the cell near the girdle bands, followed by the nucleus. The central dense core of the doubled PPC breaks down as a central spindle grows between the two PPCs, which now each become characteristically associated with a small vacuole and other features of unknown significance. The tubules of the central spindle terminate in a layer, the “Spindle Insertion” (SI), close to the PPCs; other, “polar” tubules radiate from each SI, mostly toward the nucleus, which becomes increasingly deformed by them until the nuclear envelope is ruptured. The elongating spindle enters one side of the nucleus laterally; as shown previously, the central spindle consists of two interdigitated half spindles, but the polar tubules which diverge laterally from the SI, by metaphase form a complex, cone-shaped array emanating from each pole. Some polar tubules penetrate the chromatin and may represent true kinetochore tubules; others, which persist conspicuously throughout mitosis, extend tangentially past the chromatin, out into the cytoplasm, intersecting with those from the other pole. The core structure of the PPCs alters from being plate-shaped at prophase, to being rod-like by metaphase. The chromosomes form a donut-shaped mass of chromatin penetrated by the central spindle throughout meta-and anaphase. Chromosomal separation is accomplished in two stages: the chromatin splits and moves up to the SI, and then the central spindle elongates, concurrent with its overlap region decreasing markedly in extent. Thus, this latter part of anaphase movement could be generated by microtubule sliding past microtubule. Then each PPC distinctly separates from its SI, and moves away from the daughter nucleus during cytokinesis; it again becomes the focus of numerous tubules and often ends up in one corner of the daughter cell. Meanwhile, the central spindle and the SIs, now surmounted by the reforming telophase nuclei, slowly disintegrate during cleavage.

Fertilization in Murrayella periclados was followed, using time-lapse videomicroscopy, from spermatial release to nuclear migration along the trichogyne. The localization of actin filaments in male and female gametes is shown during stages of fertilization using fluorescence microscopy techniques. These observations are compared to fertilization and actin localization in another red alga, Bostrychia moritziana. The results suggest that the fertilization events are generally the same in both species and that the actin cytoskeleton partly governs sexual plasmogamy and nuclear migration.

Previous investigations have suggested that the phototactic behaviour of diatoms, as observed by the accumulation of diatoms in low/moderate intensity light spots and avoidance of high intensity light, is primarily due to a change in cell direction at light/dark boundaries. In order to better determine the nature of this direction reversal and the manner in which it is generated, we used a high energy microbeam coupled to a monochrometer to irradiate specific areas of moving diatoms. When the leading end of moving Craticula cuspidata (Kützing) D.G. Mann cells were irradiated with a 1–3 s exposure of high irradiance (about 500 to 2000 μmol photons m−2 s−1), they quickly reversed their direction (100% of cells tested), while unirradiated control cells rarely changed direction. Similar irradiation of the trailing half of the cell caused no change in direction, but greatly reduced the effect of a subsequent irradiation of the leading half. Using a monochrometer to vary the wavelength of irradiation and a shutter to adjust the exposure time, the irradiation time needed to induce a direction change in 50% of the cells (I50) was determined for each of five wavelengths. Concurrent measurement of the energy output of the light source at each wavelength showed the most efficient wavelength to be 500 nm, close to the wavelength previously determined to be most effective in inducing a direction change at low light level light/dark boundaries in C. cuspidata. By adjusting the area of irradiation to a small spot size, we compared the effectiveness of irradiating cells at the tip, middle, and edge of the leading half of the cell. While irradiations at the tip of moving cells still resulted in 100% of cells changing direction, irradiating the edge (where the chloroplast was located) or the middle of the leading half (near the cytoplasmic storage granules) had greatly reduced effect. Such evidence suggests that phototaxis in C. cuspidata requires a photodetection system that is most sensitive to approximately 500 nm light and is located at the tips of the cells.

Like other diatoms, living cells of Chaetoceros decipiens Cleve expand lengthwise before they divide. During prophase, the nucleolus disappears in about 30 s. The spindle is very small but anaphase chromosome separation can be seen. Following rapid cleavage, the protoplasts contract, plasmolyzing slightly and transforming the cleavage furrow into a lens‐shaped opening between daughter cells. During valve initiation, the surface of the furrow is molded slightly into the shape of the mature valve face. Then daughter cells expand further, becoming fully turgid as they open the slots in the girdle bands through which the setae will grow. Soon, delicate protrusions push through the girdle bands and develop into the setae, which are very sensitive: any disturbance will immediately stop their steady growth. Healthy setae display soft, mobile tips and tiny organelles (mitochondria) actively move along the lumen. Their curvature and uniform diameter is controlled during growth with exquisite precision, and in optimal conditions, they can become very long. At their initiation, cells appear fully turgid; however, many cells soon become slightly plasmolyzed during seta growth. This observation strongly suggests that turgor pressure cannot be responsible for driving extension; the possible mechanism is discussed in the following paper.

Variable numbers of bivalents and sex chromosomes do not attach to the spindle when prophase or early prometaphase cranefly spermatocytes (2n=8) are treated with cytochalasin D or latrunculin. The unattached bivalents lie in the cytoplasm or at the spindle pole, and they do not delay onset of autosomal anaphase; sometimes they disjoin at the same time as the attached bivalents, so they respond to the global signals that initiate anaphase. Unattached sex chromosomes do not delay autosomal anaphase, either. Of various interpretations of these data, we think the best explanation is that the checkpoint system responds to physical rather than chemical cues; we think that the spindle is a “tensegral” structure, that chromosomes need to interact with the spindle in order to be recognised by the anaphase-onset “checkpoint control”, and that the physical interaction of chromosomes with spindle acts as a signalling network. Cytochalasin D and latrunculin treatments delay onset of sex chromosome anaphase (which normally occurs about 15 min after autosomal anaphase) and cause altered patterns of sex-chromosome segregation.

Living crane-fly spermatocytes were treated with 10–20 μg/ml cytochalasin D (CD) or 0.3 μg/ml latrunculin (LAT) at various stages of meiosis I. The drugs had the same effects on chromosome behaviour, but CD effects were reversible and LAT effects generally were not. When applied in mid-prometaphase to metaphase, both drugs altered subsequent anaphase poleward movements: half-bivalents either moved more slowly than normal, or moved more slowly after a brief period of movement at normal rate or stalled for 10 min or more immediately after disjunction. CD effects were reversible: within 1 min after washing out the CD, stopped chromosomes started moving and slowed chromosomes sped up. When applied in anaphase, both drugs stopped or slowed poleward chromosome movements, usually reversibly. When applied near the end of prophase, both drugs often prevented one or more bivalents in the cell from attaching to the spindle. Attached bivalents behaved as in cells treated with drugs at later stages, as described above. Unattached bivalents in the same cells moved to poles or cytoplasm in early prometaphase, where they remained motionless; at anaphase they sometimes did not disjoin, but when they did disjoin the half-bivalents did not move, either in the continued presence of the drug or when CD was washed out, confirming that they were not atttached. When CD or LAT prevented all bivalents in the cell from attaching, spindles kept in the drug were invaded by granules at about the time of normal anaphase. Conversely, when CD was washed out during late prometaphase, chromosomes often attached to spindle fibres and later entered anaphase. As CD and LAT are different anti-actin drugs, but have the same effect on chromosome behaviour, the results implicate actin in early interactions of chromosomes with spindle fibres and in anaphase chromosome movements.

Summary Mitosis in living cells ofOedogonium observed by time-lapse, was blocked by cytochalasin D (CD; 25–100 µg/ml). Normal prometaphase to anaphase takes 10–15 min; blockage of entry into anaphase by CD was reversible up to 2–2.5 h in CD and washout was followed within 10–20 min by normal anaphase and cytokinesis. After 3–6 h in CD, unseparated chromatids segregated randomly into two groups as the spindle slowly elongated considerably, becoming distorted and twisted. During this “pseudoanaphase”, chromatids sometimes split irregularly and this was stimulated by late washout of CD. CD affected chromosomal attachment to the spindle. If applied at prophase and prometaphase, spindle fibres entered the nucleus; chromosomes moved vigorously and irregularly. A few achieved metaphase only briefly. Treatment at metaphase caused chromosomes to irregularly release and after random movement, all slowly gathered at either pole. Upon removal of CD, chromosomes rapidly achieved metaphase and anaphase A and B soon followed. If CD took effect during anaphase, chromatids detaching from the spindle oscillated rapidly along it; anaphase and cytokinesis (phycoplast formation) were delayed as the cell attempted to correct for abnormal chromosomal behaviour. Thus, CD prevents normal kinetochore attachment to the spindle and actin may be the target for this response.

In Spirogyra, cytokinesis is initiated by a cleavage furrow. When the furrow contacts the elongating interzonal fibers of the telophase spindle, a small phragmoplast appears at the zone of contact, inside of which a cell-plate appears to complete cytokinesis. The use of two sequential mechanisms for achieving cytokinesis may be an intermediate, and possibly a transitional, state in the evolution of the phragmoplast found in higher plants. We tested the relative contribution of the two types of cytokinetic mechanisms by observing the effects of certain drugs on living cells, recorded in time-lapse. The antimicrotubule drug oryzalin causes the spindle to break down rapidly. When applied during early to midcytokinesis, the interzonal spindle collapses onto the cleavage furrow, which continues to grow inward. However, cytokinesis is not complete; examination of such cells 8-12 h later, after the nuclei separate, reveals a small aperture in the center of the cross-wall. This result shows that the phragmoplast/cell-plate is essential for complete cytokinesis, which apparently cannot be accomplished by cleavage alone, and that inward growth of the cleavage furrow is not dependent on microtubules. The antiactin drug cytochalasin D stops cleavage quite rapidly, and the eventual consequences of treatment depend on the stage reached in cleavage. If the cytochalasin is applied early in cytokinesis, cell division is unable to proceed further. Between telophase nuclei the cell does generate a large mass of cytoplasm that contains typical proliferating phragmoplast fibers; cell-plate formation is initiated in this mass but is never able to proceed further, and no coalescence of material into a cross-wall has been recorded. If cytochalasin is applied later, after the cleavage furrow has contacted the interzonal spindle fibers, cleavage stops but now cell-plate formation proceeds and cytokinesis is completed, albeit slowly. This result demonstrates that an interaction between the cleavage furrow and the forming phragmoplast is necessary for the latter to operate normally. These results support our suggestion that Spirogyra illustrates an intermediate stage in the evolution of the phragmoplast, since both cleavage and cell-plate formation are required for cytokinesis. (This suggestion does not require Spirogyra to be on the line of evolution leading to higher plants; rather, Spirogyra appears to be undergoing an evolutionary process similar to what might have occurred in their progenitors,) Furthermore, interaction of the cleavage furrow with the forming cell-plate/phragmoplast is required for cross-wall completion. These results may have significance in considering mechanisms by which higher plants locate their cell-plates properly, an ability associated with the evolution and possible function(s) of the preprophase band of microtubules.