Search

We have obtained immunofluorescence and immunoblot evidence for the presence of kinesin-like protein (KLP) in pollen tubes of tobacco using an antibody generated against peptides encoded by theKATA gene ofArabidopsis. This antibody recognizes an Mr 140,000 polypeptide inArabidopsis seedlings, and stains the mitotic apparatus in this species as well as in tobacco suspension cells. In tobacco pollen tubes prepared for dual immunofluorescence localizations of KLP and β-tubulin, the antibody binds transiently to microtubule (Mt) bundles and the nucleus in premitotic generative cells; it then stains the developing mitotic apparatus as the nuclear envelope breaks down. By metaphase, fluorescence is located over kinetochore fibers and associated Mts. Localization of KLP is concentrated in the midzone during anaphase, and by early cytokinesis, it closely brackets the cell plate. Phragmoplast fluorescence then spreads along the phragmoplast distal to the cell plate. Punctate staining is also detected along vegetative Mts. No KLP localization is seen in pollen tubes treated with antibody after it had been preadsorbed to the antigenic peptides. The antibody recognizes an Mr 110,000 polypeptide in extracts of tobacco pollen tubes, and a polypeptide of somewhat lower Mr inTradescantia pollen tubes. Our results show that KLP(s) related to KatAp are present in tobacco generative cells and may play roles in the organization and/or operation of the mitotic apparatus and phragmoplast.

The inhibitory effects of benomyl on pollen tube growth have received little attention, particularly at the microscopic and immunohistochemical levels. Pollen germination and tube growth in the presence of benomyl were evaluated in Tradescantia virginiana to investigate the effects of this fungicide on pollen germination rate, tube growth and morphology, and microtubule (Mt) organization. Benomyl was incorporated in germination media at 0, 480, 600 or 720 mg l -1 . Inhibition of pollen germination, cytoplasmic streaming, and tube elongation were associated with benomyl treatments. Benomyl also induced abnormal pollen tube morphology and Mt organization. Compared to controls, Mts in the treated tubes were characteristically fewer in number, fragmented, sinuous and increasingly disorganized. At the two highest benomyl concentrations, Mts were considerably fewer or absent in apical/subapical regions of the pollen tubes. This work verifies that benomyl incorporated into germination media at concentrations lower than recommended field rates inhibit pollen germination and tube growth, and that the effects are associated with alterations of Mt organization.

The kinesin-like proteins (KLPs) are a large family of plus- or minus-end-directed microtubule motors important in intracellular transport, mitosis, meiosis, and development. However, relatively little is known about plant KLPs. We prepared an antibody against two peptides in the microtubule binding domain of an Arabidopsis KLP (KatAp) encoded by the KatA gene, one of a family of genes encoding KLPs whose motor domain is located near the C terminus of the polypeptide. Such KLPs typically move materials toward the minus end of microtubules. An immunoreactive band (Mr of 140,000) corresponding to KatAp was demonstrated with this antibody on immunoblots of Arabidopsis seedling extracts. During immunofluorescence localizations, the antibody produced weak, variable staining in the cytoplasm and nucleus of interphase Arabidopsis suspension cells but much stronger staining of the mitotic apparatus during division. Staining was concentrated near the midzone during metaphase and was retained there during anaphase. The phragmoplast was also stained. Similar localization patterns were seen in tobacco BY-2 cells. The antibody produced a single band (Mr of 130,000) in murine brain fractions prepared according to procedures that enrich for KLPs (binding to microtubules in the presence of AMP-PNP but not ATP). A similar fraction from carrot suspension cells yielded a cross-reacting polypeptide of similar apparent molecular mass. When dividing BY-2 cells were lysed in the presence of taxol and ATP, antibody staining moved rapidly toward the poles, supporting the presence of a minus-end motor. Movement did not occur without ATP, with AMP-PNP, or with ATP plus antibody. Our results indicate that the protein encoded by KatA, KatAp, is expressed in Arabidopsis and is specifically localized to the midzone of the mitotic apparatus and phragmoplast. A similar protein is also present in other species.

A polyclonal antibody directed against a 170 kDa myosin heavy chain from lily pollen tubes was employed to (a) assess the cellular distribution of the polypeptide using immunofluorescence methods, and (b) ascertain if similar polypeptides are present in pollen tubes and somatic cells of other species. Fluorescence is associated with particles of various size as well as an amorphous component, and is concentrated in the apical cytoplasm of lily and tobacco pollen tubes. Apical fluorescence is more extensive in lily than in tobacco, which may be related to different streaming patterns and apical zonation seen at the ultrastructural level. In suspension cells of tobacco andArabidopsis, fluorescence is concentrated around the nuclei. Dual localizations indicate that anti-myosin fluorescence may be associated with the presence of actin. Little or no staining was seen in controls consisting of either pre-immune serum or mono-specific IgG that had been preadsorbed with the 170 kDa polypeptide. Immunoblots show that a 170 kDa immunoreactive polypeptide is present in pollen tubes of tobacco andTradescantia virginiana in addition to lily, and in suspension culture cells of tobacco andArabidopsis and extracts of wholeArabidopsis seedlings. Our results show that a conserved 170 kDa myosin heavy chain is present in a variety of monocot and dicot cells. They are also consistent with the presence of multiple myosins in plants in general and pollen tubes in particular.

gamma-Tubulin-specific antibodies stain the microtubule (Mt) arrays of Arabidopsis suspension cells in a punctate or patchy manner. During division, staining of kinetochore fibers and the phragmoplast is extensive, except in the vicinity of the plus ends at the metaphase plate and cell plate. gamma-Tubulin localization responds to low levels of colchicine, with staining receding farther toward the minus (pole) ends of kinetochore fibers. At higher drug concentrations, gamma-tubulin also associates with abnormal Mt foci as well as with the surface of the daughter nuclei facing the phragmoplast. During UV-induced recovery from colchicine, gamma-tubulin increases along the presumptive minus ends of mitotic Mts as well as the phragmoplast near the daughter nuclei. With CIPC, immunostaining is concentrated around the centers of focal Mt arrays in multipolar spindles. In the presence of taxol, Mts are more prominent but the mitotic apparatus and phragmoplast are abnormal. As with CIPC, gamma-tubulin is concentrated at focal arrays. Increased punctate staining is also present in interphase arrays, with fluorescent dots often located at the ends of Mts. These results support a preferential association between gamma-tubulin and Mt minus ends, but are also consistent with more general binding along the walls of Mts. Thus, minus ends (and Mt nucleation sites) may be present throughout plant Mt arrays, but gamma-tubulin may also serve another function, such as in structural stabilization.

gamma-Tubulin is a protein associated with microtubule (Mt)-organizing centers in a variety of eukaryotic cells. Unfortunately, little is known about such centers in plants. Genomic and partial cDNA clones encoding two gamma-tubulins of Arabidopsis were isolated and sequenced. Comparisons of genomic and cDNA sequences showed that both genes, TubG1 and TubG2, contain nine introns at conserved locations. The sequences of the two genes both predict proteins containing 474 amino acids, with molecular masses of 53,250 and 53,280 D, respectively. The predicted gamma 1- and gamma 2-tubulins exhibit 98% amino acid identity with each other and approximately 70% amino acid identity with the gamma-tubulins of animals and fungi. RNA gel blot results demonstrated that both genes are transcribed in suspension culture cells, seedlings, and roots and flowers of mature plants. Immunoblots of Arabidopsis proteins using an antibody specific to a conserved peptide of gamma-tubulin showed a major cross-reacting polypeptide with an M(r) of 58,000. The same antibody stained all Mt arrays in tissue and suspension culture cells of this species. Binding was inhibited by the homologous oligopeptide in the gamma-tubulins predicted by the two Arabidopsis gene sequences. Antibody staining avoided the plus ends of Mts at the kinetochores and cell plate, but unlike the case in animal cells, seemed to be localized over broad stretches of the kinetochore fibers and phragmoplast toward the minus ends. We concluded that at least two gamma-tubulin protein homologs are present in Arabidopsis and that at least one of them is localized along Mt arrays. Its distribution is correlated with and may help explain unique characteristics of Mt organization in plants.

Barry A. Palevitz I used to be a bench scientist. Starting as a graduate student in the heyday of cell ultrastructure and ending 30 years later with immunofluorescence and a little molecular biology (thanks to great collaborators), I studied the plant cytoskeleton. About 5 years ago, for a variety

An antibody specific for a conserved gamma-tubulin peptide identifies a plant polypeptide of 58 kDa. gamma-Tubulin antibody affinity purified from this polypeptide recognizes the centrosome in mammalian cells. Using immunofluorescence microscopy, we determined the distribution of this gamma-tubulin-related polypeptide during the complex changes in microtubule arrays that occur throughout the plant cell cycle. We report a punctate association of gamma-tubulin-related polypeptide with the cortical microtubule array and the preprophase band. As cells enter prophase, gamma-tubulin-related polypeptide accumulates around the nucleus and forms a polar cap from which early spindle microtubules radiate. During metaphase and anaphase, gamma-tubulin-related polypeptide preferentially associates with kinetochore fibers and eventually accumulates at the poles. In telophase, localization occurs over the phragmoplast. gamma-Tubulin-related polypeptide appears to be excluded from the plus ends of microtubules at the metaphase plate and cell plate. Its distribution during the cell cycle may be significant in light of differences in the behavior and organization of plant microtubules. The identification of gamma-tubulin-related polypeptide could help characterize microtubule organizing centers in these organisms.

In order to gain a more complete understanding of the organization of the mitotic apparatus (MA) in the generative cells (GCs) of flowering plants, pollen tubes ofNicotiana tabacum were examined using tubulin immunocytochemistry and Hoechst fluorescence. The observations were then compared with previously published information onTradescantia GCs and the MA of somatic cells. At the onset of division, the prominent microtubule (Mt) bundles characteristic of GCs are reorganized into a more random Mt network. At late prophase/prometaphase, kinetochores appear to interact with this network, resulting in the formation of K-fibers that frequently link in tree-like aggregates. The GC MA takes the form of a distinct spindle and often has pointed, focused poles; the metaphase plate is usually oblique. Karyokinesis involves both anaphase A and B; lengthening of interzonal Mts is accompanied by elongation of the spindle. In late anaphase/early telophase, phragmoplast Mts are formed in association with the proximal face of the sperm nuclei. The phragmoplast remains prominent for some time, so that its Mts as well as another population generated from the distal face of the sperm nuclei constitute the initial sperm cytoskeleton. Comparisons indicate that the spindle in tobacco GCs falls on a continuum of organization between that of somatic cells and the MA ofTradescantia GCs.

In dividing generative cells ofTradescantia, kinetochore pairs do not line up on a typical metaphase plate, but instead are distributed along the length and depth of the cell prior to anaphase onset. Kinetochore (K) fibers are linked to each other and to a system of axial microtubule (Mt) bundles in an arrangement that makes discrete half spindles, if present, not immediately obvious. Because such arrangements may have important implications for the conduct of the remainder of division, anaphase events were closely scrutinized using a combination of tubulin and kinetochore immunocytochemistry (the latter with CREST serum). Anaphase appears to consist of three principal processes. Around the time of anaphase onset, K-fibers and surrounding Mts become reorganized into two large superbundles. To each superbundle is attached a set of nonfilial kinetochores bound for one end of the cell. The K-fibers then appear to shorten to varying degrees; in many cases, kinetochores become linked directly to the superbundles. The superbundles then separate in an anaphase B-like process, further moving the kinetochores toward opposite ends of the cell. The superbundles themselves shorten, and distances within the bundles also decrease, such that the kinetochores cluster closer together. The results indicate that reorganization of Mts into superbundles (and the consequential manifestation of bipolarity) is important for orderly chromosome separation.

In order to assess the possible role of microfilaments (Mfs) in events preceding plant cell division, actin was localized in root cells of Allium cepa and Tradescantia virginiana by immunofluorescence microscopy. The distribution of Mfs was compared to that of microtubules (Mts) by means of dual localizations employing both antiactin and antitubulin. Cycling interphase cells contain Mfs that extend into all regions of the cytoplasm in random fashion. Prior to the rearrangement of the cortical Mt array into the initial broad preprophase band (PPB), the number of Mfs in the cytoplasm decreases, while a new population appears in the cortex. The cortical Mfs, which usually occupy the entire cell surface, are aligned parallel to the cortical Mts. When the initial PPB appears, these Mfs still cover the cortex or are arranged as a broad band encompassing the PPB. As the PPB narrows, the Mfs are also confined to an increasingly restricted zone usually wider than the PPB.than the PPB. When the PPB reaches its narrowest, densest configuration, aligned Mfs are excluded from the band proper, while others appear in flanking regions of the cortex. From prometaphase through anaphase, cortical Mfs are largely restricted to the ends of the cell overlying the spindle poles; they also tend to become more randomly oriented. Little or no actin is present in the spindle. During telophase, the two zones of aligned cortical Mfs over the ends of the cell gradually disappear and are replaced by new interphase networks. These changes provide additional data on the possible control of PPB organization by actin, and in addition indicate that the cortex may be the origin of the actin that aggregates at the spindle poles during cytochalasin treatment. © 1992 Wiley-Liss, Inc.

Division of the generative cell of Tradescantia virginiana is unusual in that a typical bipolar spindle is not present Instead, the cell contains an axial system of microtubuie (Mt) bundles, with kinetochores distributed along the length and depth of the cell. Kinetochore fibers appear to be derived from and remain attached to the Mt bundles. Localizations with both anti-tubulin and CREST serum were performed in order to probe this relationship further. Pairs of CREST-positdve, fluorescent dots presumed to be kinetochores are initially oriented transverse to the cell axis and appear to be associated with the Mt bundles via lateral interactions. Adjacent pairs are often joined to the same bundles, like rungs on a ladder. Lateral interactions are then converted into or replaced by end-on kinetochore fibers similar in morphology to those seen in other cells. This conversion is accompanied by a realignment of the kinetochore pairs along the long axis of the cell. In addition, the center-to-center spacing between filial kinetochores doubles. Interconnections between kinetochore fibers and surrounding Mts appear to be maintained during the transition. These results may provide a general insight into the manner in which kinetochores interact with the division apparatus in eukaryotes.

A collaborative effort was initiated to resolve differences in two recent papers on the effects of cytochalasins in root cells. While both studies reported similar effects on interphase cells (i.e., replacement of microfilaments by many small specks and rods), Palevitz (Cell Motil. Cytoskeleton 9:283-298, 1988) maintained that cytochalasins B and D induce actin aggregation at the poles of dividing Allium root cells at a concentration of 10 μM with rhodamine phalloidin as a reporter probe, whereas McCurdy and Gunning (Cell Motil. Cytoskeleton 15:76-87, 1990) could not find these aggregates following antiactin immunocytochemistry in Triticum roots treated with CB at 50 μM. Employing identical methods and materials in the same laboratory, we found that CD induces polar actin aggregates in dividing cells of both species. However, the aggregates in Triticum are smaller and occur less frequently than those in Allium. A similar pattern is seen with CB.

Tobacco BY-2 suspension cultures were synchronized with aphidicolin in order to assess the relationship between microtubules (MTs), microfilaments (MFs), and the nuclear envelope (NE) at different stages of the cell cycle. Using immunofluorescence techniques, ordered MT arrays were found in the cortex in G1; few MTs are evident deeper in the cytoplasm or near the nucleus. However, MTs radiate from the surface of the nucleus during S and G2 as the interphase cortical array is replaced by the preprophase band. Perinuclear fluorescence is also visible at the end of cytokinesis but does not overlap with new ordered cortical arrays early in G1. When isolated nuclei are examined, associated MTs are again evident in S and G2, but not in G1. Microfilaments are colocalized with the MTs in the radiating arrays, as ascertained by dual staining of cells with rhodamine phalloidin. Propyzamide treatment leads to the loss of MTs at all stages, while cytoplasmic and perinuclear MF networks persist. Conversely, cytochalasin D disrupts MFs, including those radiating from the nucleus during S and G2, without any apparent effect on MTs. The results cast doubt on a proposed role for the NE in the generation of cortical MTs in plants. A universal role for MFs in the deployment of MTs is also in question.

The organization of microtubules (MTs) in the cortex of cells at interphase is an important element in morphogenesis. Mechanisms controlling the initiation of MTs and their spatial ordering, however, are largely unknown. Our recent study concerning the generation of a radial array of MTs in stomatal guard cells inAllium showed that the MTs initiate in a cortical MT-organizing zone adjacent to the ventral wall separating the two young guard cells (Marc, Mineyuki and Palevitz, 1989, Planta179, 516, 530). In an attempt to detect MT-ordering mechanisms separate from the sites of MT initiation, we now employ various drugs to manipulate the geometry and integrity of the ventral wall and thereby also the associated MT-organizing zone. In the presence of cytochalasin D the ventral wall is tilted away from its normal mid-longitudinal anticlinal alignment, while treatments with the herbicide chloroisopropyl-N-phenylcarbamate (CIPC) induce the formation of a branched ventral wall. Nonetheless, in either case the MTs still form a radial array, although this is asymmetric as it is centered in accordance with the misaligned or branched ventral wall. Since the MTs maintain their original course undisturbed as they extend beyond the abnormal ventral wall, there is no evidence for the presence of an inherent MT-ordering mechanism at locations remote from MT-initiation sites. Following treatments with caffeine, which abolishes the formation of the ventral wall, the MTs revert to a transversely oriented cylindrical array as in normal epidermal cells. Thus the presence of the ventral wall, and presumably also the associated MT-organizing zone, is essential for the establishment of the radial array. The MT-organizing zone is therefore involved not only in the initiation of MTs, but also in determining their spatial order throughout the cell cortex.

The preprophase band (PPB) of microtubules (Mts), which appears in the G2 phase of the cell cycle in higher plants but disappears well before the end of karyokinesis, is implicated in the determination of the division plane because its location marks the site at which the phragmoplast/cell plate will fuse with the parental plasmalemma during cytokinesis. The PPB first appears as a rather wide array, which progressively narrows before or during prophase. Actin-containing microfilaments (Mfs) have recently been reported in the PPB, but the role of these elements in PPB organization and/or function remains unclear. The present study employed fluorescence and pharmacological methods in symmetrically and asymmetrically dividing epidermal cells of Allium to probe this problem. Our results show that PPBs in cells treated with 2–200μM cytochalasin D (CD) are still transversely aligned but remain two to three times wider than mature bands in control cells. Treatment for 0.5 h at 20 μM is sufficient to make the PPBs abnormally widel Premitotic nuclear migration in asymmetrically dividing cells is also inhibited by CD, as is the positioning of the mitotic apparatus and the new cell plate. The plate is still transverse, however. Band-like arrays of cortical Mfs become evident in most interphase cells by prophase. The band remains quite wide compared to the final dimensions of the Mt PPB, and clearly encompasses it. Levels of CD as high as 200μM decrease the number of cells with transverse actin bands, although a majority still retain them. Other F-actin arrays are disrupted by the drug. Thus, while CD does not inhibit the formation of an initial, broad, transverse PPB in most cells, it does prevent the narrowing process that defines the precise division site. The role of actin in this effect is discussed.

Structural observations provide persuasive evidence for the existence of a cortical network of endoplasmic reticulum (ER) in a large number of plant and animal cells. The network in plants generally possesses a polygonal pattern in which smooth, tubular elements are joined by intervening lamellar segments. The individual elements of ER are often positioned extremely close to the plasma membrane (PM), and may form appositional contacts, but fusion does not occur. The network arises at cytokinesis and establishes continuity between the cortical ER of daughter cells in the form of tightly furled membrane tubules that traverse the plasmodesmata. The specific function of the cortical ER complex is unknown but different possibilities seem attractive. It may serve key roles in anchoring the cytoskeleton and in facilitating secretion. The cortical ER might also participate in the communication of signals between the exterior of the cell and cytoplasm. As a consequence of its ability to release and/or sequester Ca, the ER could control the cytoplasmic activity of this ion and thus a host of physiologically and developmentally important reactions.

Previous observations indicate that division of the generative cell inTradescantia virginiana is characterized by several unusual features, including persistence of surrounding microtubule (Mt) bundles during karyokinesis, lack of a distinct metaphase plate and direct contribution by mitotic Mts to the cytoskeleton of young sperm. We have further probed karyokinesis in these cells using additional antitubulin and chromosome staining, as well as kinetochore visualizations with CREST serum. The CREST antibodies reveal kinetochores as paired and single fluorescent dots similar to those seen in other species stained with this preparation. Double localizations show that the dots are located at the ends of Mt bundles previously identified as kinetochore fibers (Palevitz and Cresti 1989). Before anaphase, paired kinetochores are distributed along the length of the cell. They also tend to be located at the cell periphery or are directly connected to peripheral Mt bundles by their kinetochore (K)-fibers. Twelve pairs of dots can be counted per cell, equal to the expected number of chromosomes. During anaphase, kinetochore separation starts at various positions along the length of the cell, producing single, relatively uniformly distributed kinetochores in the crotches of forks formed by K-fiber trunks and elongating Mt branches attached to the base of the trunks. Eventually, K-fibers with attached kinetochores aggregate in stepwise fashion on thick Mt bundles at both ends of the cell. This pattern is reflected in the cytoskeleton of young sperm. These results further document the unusual distribution of chromosomes and kinetochores inTradescantia generative cells and the origin of the Mt cytoskeleton in sperm cells.

In animal cells, microtubule assembly is usually initiated at one specialized structure, the centrosome. By contrast, in plant cells, microtubule assembly begins at a variety of locations within the cell. A member of the tubulin gene family, gamma-tubulin, is localized to the centrosome in animal cells and is important in the assembly of microtubules in vivo. Recent reports have identified gamma-tubulin genes in plants and have described the complex intracellular distribution of the encoded polypeptides. Here, Harish Joshi and Barry Palevitz comment upon how this information may help elucidate the organizing principles of the complex arrays of microtubules in plant cells.

The effects of propiconazole on germination and tube growth of Tradescantia virginiana pollen when incorporated in germination media at 0, 102, 136, or 170 μl l−1 were evaluated using light microscopy and immunocytochemistry. Propiconazole inhibited pollen germination, cytoplasmic streaming, and tube elongation. Treatments also induced abnormal tube morphology and cytoskeletal distribution. Tubes treated with propiconazole displayed weaker microfilament (Mf) signals along the pollen tubes, with amorphous staining. Microtubule (Mt) distribution was also severely affected. In treated tubes, the proximal portions had characteristically fragmented Mts. Fewer Mt bundles were seen in the subapical region, and these were located further from the apex. Propiconazole effects were generally concentration dependent. The results indicate that propiconazole affects both Mfs and Mts; however, the effects may be an indirect result of the drug's influence on membranes.

γ-Tubulin was localized in tobacco pollen tubes using an antibody raised against a peptide conserved in all known γ-tubulins. Antibody staining occurs in a primarily punctate pattern along the length of the microtubule bundles in generative cells and along cortical microtubules in the vegetative cytoplasm. During generative cell division, γ-tubulin is localized in the forming mitotic apparatus. By metaphase, it is present along kinetochore fibers except at their plus ends located at the kinetochores. By telophase, staining is observed in the phragmoplast, where it again avoids the plus ends of microtubules at the cell plate. γ-Tubulin is also present at the periphery of the sperm nuclei. A patch of intense staining on the distal side of each nucleus marks the site of assembly of a new population of sperm microtubules. No specific fluorescence is present in control pollen tubes treated with preimmune IgG. These localization patterns bear similarities to those seen in somatic cells and in addition may help explain changes in microtubule arrays between generative cells and sperm.

Fluorescence microscopy was used to visualize microtubules (Mts) and chromatin in an effort to further clarify the relationship between the generative cell (GC) and vegetative nucleus (VN) in pollen tubes of tobacco. Prominent Mt bundles are present in one or more GC extensions that can be finger-like or lamellar in form. While the VN is positioned distal to the GC in most cases, it can also straddle the cell or lie proximal to it. In all cases, however, extensions embrace, penetrate or clasp the VN. GC Mts are reorganized during the formation of the mitotic apparatus, and cell extensions are fully or partially withdrawn. By telophase in many pollen tubes, the VN shifts to a more proximal position and appears to adhere to the region of the GC containing the phragmoplast. Application of oryzalin leads to the disorganization of Mts, changes in cell shape, including the loss or alteration of cell extensions, and separation of the GC and VN in some cases. However, the position and polarity of the VN is maintained in most pollen tubes. The results indicate that GC Mts and cell extensions play a role in the association with the VN. However, the relationship appears to be controlled by other factors as well. Attention should now be directed at potential interactions involving the VN envelope, vegetative plasma membrane, GC plasma membrane and extracellular matrix.

A key event in the differentiation of elliptically shaped guard cells such as those in Allium is the formation of a radial array of cortical microtubules (Mts) which, by controlling the orientation of wall microfibrils, plays an important role in cell shaping. Previous experiments strongly indicated that the array is nucleated in a zone adjacent to the new ventral wall soon after cytokinesis. In order to further clarify the function of this zone, we performed dual immunolocalizations on Allium guard cells with anti-beta-tubulin, to detect Mts, and an antibody to gamma-tubulin, a protein known to be present at Mt-organizing centers in other species and recently identified in plants as well. gamma-Tubulin antibody stained the cortical zone adjacent to the ventral wall, while little or no fluorescence was present elsewhere along the radial Mt array or at other sites in the cell. The antibody also stained the mitotic poles and phragmoplast in guard mother cells, as it does in other material. No staining was seen when the primary antibody was omitted. The results are consistent with nucleation of the radial array at a cortical-Mt-organizing zone next to the ventral wall, and set the stage for more in-depth studies on the spatial and temporal control of Mt formation in differentiating cells.

Disagreement has arisen over the presence of actin-containing microfilaments (Mfs) in angiosperm generative cells and sperm (GSP). In order to address this issue, we subjected GSP of Tradescantia virginiana, Nicotiana tabacum and Rhododendron laetum to a series of localizations using different antiactins, rhodamine phalloidin and antimyosin. Coordinate staining with antitubulin and Hoechst 33258 defined the status of the microtubule (Mt) cytoskeleton and stages of generative cell division. Additional experiments utilized cytochalasin D (CD). In no instance could Mfs be detected in GSP of the three species. Instead, Mfs seen at the periphery of GSP appear to be continuous with vegetative Mfs and thus are in the vegetative cytoplasm. Mfs are not seen in the constriction zone of dividing T. virginiana generative cells, nor are they indicated in the phragmoplast of N. tabacum and R. laetum. Myosin localizations reveal punctate staining in the vegetative cytoplasm and a thin line of fluorescence around the the outside of the generative cell. While CD seems to delay generative cell division, cytokinesis still takes place. CD-induced Mf fragments are evident in the vegetative cytoplasm but not in GSP. The weight of evidence therefore indicates that GSP do not contain Mfs. The implications of this conclusion for the behavior of GSP and the mechanism of cytokinesis in dividing generative cells are considerable.

Publisher Summary This chapter focuses on the organization, composition, and function of the generative cell and sperm cytoskeleton. Microtubule (MT) organization in generative and sperm cells (GSP) contrasts sharply with that in the surrounding vegetative cytoplasm. MT is composed of heterodimers of α - and β -tubulin, which share sequence homology at the DNA and protein levels with counterparts in animals. Vegetative MTs in hydrated pollen grains and pollen tubes do not occur in similar thick bundles but instead comprise much thinner elements organized in various patterns. In pollen tubes MTs are arranged primarily in helical to longitudinal arrays in the outer cortex. The cortical elements consist of single MTs or planar groups that are cross-bridged to each other and to the plasma membrane. The MT bundles in GSP are also unlike somatic cell arrays. While cortical MTs are linked to the plasma membrane and associate with each other in somatic cells, the kind of large, flexuous bundles that ramify through the cytoplasm of GSP are not observed. Instead, the MTs are largely limited to a single layer next to the plasma membrane.

When Stephen Jay Gould died on May 20, he left behind unfinished business. The matter occupied a lot of his time but had nothing directly to do with land snails, punctuated equilibrium, or the spandrels of adaptation. Nor was it mentioned in staid journals or annual meetings where scientists usually trade data. No, Gould left behind a festering controversy of another sort, one involving public perceptions of science rather than the latest fossil find or molecular phylogeny. In fact, the increasingly boisterous brouhaha has more to do with religious and political agendas than it does evolution. I'm speaking of course of creationism in its latest mutation/incarnation, intelligent design or ID. Most biologists treat creationism like a dead skunk, dismissing the subject as too smelly to touch. Or, being much ado about nothing, it's not worth their time. But Gould knew better. If ID creationism has its way, the prospects for science as currently practiced are far direr than many reading this essay can conjure.

More than 100 years after the discovery of mitosis, the role of the mitotic apparatus (MA) in cell and organismal development is still the subject of lively debate (e.g., Kaplan, 1992). The developmental impact of cell division looms perhaps even larger in plants than in other organisms because plant cells are not free to migrate during morphogenesis. Therefore, plants place great stock in modulating division planes to produce new directions for growth (Sinnott, 1960). In addition, precisely placed walls serve to set off unequal compartments that then differentiate in divergent pathways. Thus, although their principal function, partitioning two equal sets of chromosomes to separate cytoplasmic compartments, remains the same, mitosis and cytokinesis assume an additional dimension in plants. Whereas animals also depend on precise division planes (Strome, 1993), the process occupies an especially prominent role in plant development (Gunning, 1982). My main purpose in this review is not to address division plane determination per se but to explore a characteristic of plant mitosis that has received much less attention. The subject of my discussion is the morphological plasticity of the MA, that is, its ability to be tilted, squeezed, or otherwise deformed under the confining influence of the cell wall and still work. Given new information about a number of proteins, we may be on the verge of understanding the molecular basis of mitotic architecture. Furthermore, with a fuller appreciation of mitotic plasticity, division plane determination can be seen from another perspective-that of a mechanism to compensate for MA asymmetry or deformation. During my discussion, I'll also make a brief foray into the evolution of microtubule organization patterns in plants.

In order to study developmental changes in microtubule organization attending the formation of a longitudinally oriented preprophase band, the guard mother cells ofAvena were examined using a new procedure for anti-tubulin immunocytochemistry on large epidermal segments. We found that the interphase band (IMB) of transverse cortical microtubules present in these cells following asymmetric division is replaced after subsidiary cell formation by mesh-like to radial microtubules that extend throughout the cytoplasm. Many of the Mts are also grouped in bundles. Gradually, this intermediate array is succeeded by longitudinal elements of the PPB. Thus, preprophase band formation is accompanied by a 90° shift in Mt orientation, with a radial arrangement serving as an intermediate stage. The micrographs are most consistent with the rearrangement of intact Mts, although changes in Mt assembly are possible as well. The role of the IMB in guard mother cells is also discussed.

The location of calmodulin in rabbit and guinea pig spermatozoa was determined by indirect immunofluorescence techniques. Spermatozoa that had not undergone the acrosome reaction exhibited four distinct regions of calmodulin-specific immunofluorescence: around the acrosome, in a band across the lower third of the head, and in two localized areas at the base and tip of the flagellum. In contrast, after the acrosome reaction, although other features of calmodulin distribution remained the same, the fluorescence associated with the anterior half of the head was notably absent. Instead, fluorescence was associated with the membranes that had separated from the sperm head. These findings suggest a potential role for calmodulin in the Ca2+-dependent control of sperm activation, in sperm-egg fusion, and in microtubule disassembly processes in the flagellum.

F-actin has been identified in the preprophase band of Allium cepa. Cells attached to subbed slides were obtained from formaldehyde-fixed root tips digested in EGTA and Cellulysin. The air-dried cells were extracted in Triton X-100, treated with rhodamine-phalloidin, rinsed briefly in PBS, and viewed in the fluorescence microscope. Interphase cells contain a network of actin fibers that extends into all areas of the cytoplasm. During preprophase, the network is replaced by a band of fibers aligned in the position of the preprophase band. Colocalization of F-actin with rhodamine-phalloidin and microtubules with tubulin immunocytochemistry confirms that the two bands are coincident. The actin appears to comprise a thin layer of fibers next to the plasmalemma. Like the microtubule preprophase band, the actin band narrows as preprophase progresses and disappears by midprophase. Fluorescent actin bands are not seen in fixed cells pretreated with excess unlabeled phalloidin before staining. They are also absent in roots exposed to cytochalasins B and D before fixation, but preprophase band microtubules at all stages of aggregation are still present. Colchicine treatment leads to the loss of both preprophase band microtubules and actin. The possible function of preprophase band actin is discussed.

The organization of kinetochore fibers was examined inAllium root cells processed for tubulin immunocytochemistry. Metaphase fibers consist of a core or trunk of Mts to which are attached numerous branches, yielding a bottle-brush of fir-tree pattern similar to that reported inHaemanthus endosperm cells. Many of the branches cross the midzone and extend into the opposite half-spindle. In addition, branch Mts associate with more than one kinetochore fiber. During anaphase, branch Mts elongate while the trunks shorten and fuse into polar caps. Our results are discussed in terms of spindle fiber organization and Mt polarity.

Protoplasts were prepared from the guard cells ofA. cepa. Epidermal peels taken from expanding green leaves and largely free of mesophyll were treated with Cellulysin, and protoplasts were harvested after 18 h of digestion. That the protoplasts were derived from guard cells was ascertained from their characteristic vacuolar autofluorescence and from observations showing that all other epidermal cells are killed in the peeling procedure. The protoplasts proved to be a good system with which to view the cell cortex and inner surface of the plasmalemma. The lysis of cells adhering to polylysine-treated, Formvar-coated grids, followed by negative staining in uranyl acetate, showed that many microtubules normally present in ordered arrays in situ remain closely applied to the inner surface of the plasmalemma in protoplasts. In addition, numerous vesiculate elements including coated vesicles and/or pits are present amongst the microtubules. Similar vesicles are evident in thin sections of fixed, embedded guard cells and protoplasts. The significance of these structures in the cell cortex is discussed.

Changes in microtubule organization were monitored in the stomatal complexes of Avena sativa using tubulin immunocytochemistry. Radial arrays of cortical microtubules, previously thought to be characteristic of guard cells, also appear in adjacent subsidiary cells early in development. The subsidiary cell arrays are evident even before guard cells form via division of precursor guard mother cells. Thus, before the stomatal pore opens between sister guard cells, each complex contains four similar microtubule arrays. As the pore opens, however, the subsidiary cell system is reorganized into a network of microtubules distributed along the length of the cell. A similar change is effected in the guard cells after the pore opens. Subsidiary cells and guard cells elongate during later stages of differentiation, and a thickened wall is deposited int he narrow midzone of the latter. At the same time, microtubules in both cells assume a more axial orientation. The results are discussed in terms of developmental symmetry and the control of microtubule organization and cell wall deposition.

The development of autofluorescent vacuoles in the stomatal cells ofAllium cepa andA. vineale was investigated using fluorescence microscopy of live cells, low light level television, cytochemistry and electron microscopy. During cell differentiation, the vacuole undergoes two major changes in morphology. In an intermediate form, it consists of a reticulum or network of interlinked tubules and small chambers. The network is formed from globular cisternae in very young GMCs and is maintained as a reticulum until it is transformed back into a globular form later in the differentiation of guard cells. The network thus remains intact through the course of one cell division. During its existence, the reticulum undergoes complex movements and rearrangements. The significance of these changes in the vacuole is discussed in terms of vacuole ontogeny and function and the mechanisms that control motility in plant cells.

The organization of microtubule (MT) arrays in the guard mother cells (GMCs) of A. cepa was examined, focussing on the stage at which a longitudinal preprophase band (PPB) is established perpendicular to all other division planes in the epidermis. In the majority of young GMCs, including those seen just after asymmetric division, MTs are distributed randomly throughout the cortex and inner regions of the cytoplasm. Few MTs are associated with the nuclear surface. As the GMCs continue to develop, MTs cluster around the nucleus and a PPB appears as a wide longitudinal band. Microtubules also become prominent between the nucleus and the periclinal and transverse walls, while they decrease in number along the radial longitudinal walls. The PPB progressively narrows by early prophase, and a transversely oriented spindle gradually ensheaths the nucleus. These observations indicate that the initial, broad PPB is organized by a rearrangement of the random cytoplasmic array of MTs. Additional reorganization is responsible for MTs linking the nucleus and the cortex in the future plane of the cell plate, and for narrowing of the PPB.

The initiation and development of a radial array of microtubules (MTs) in guard cells of A. cepa was studied using immunofluorescence microscopy of tubulin in isolated epidermal layers. Soon after the completion of cytokinesis, MTs originate in the cortex adjacent to a central strip of the new, anticlinically oriented ventral wall separating the two guard cells. Cortical MTs extend from the mid-region of the central strip toward the cell edge where the ventral wall joins the inner periclinal wall. They then spread in a fan-like formation along the periclinal wall and gradually extend along the lateral and end walls as well. Many MTs criss-cross at various angles as they arc past the edge formed by the junction of the ventral and periclinal walls, but they do not terminate there, indicating that, contrary to previous report, the edge is not involved in MT initiation. Instead, the mid-region of the central strip appears to function as a planar MT-organizing zone. Initially, MTs radiate from this zone through the inner cytoplasm as well as the cortex. During cell expansion, however, the cortical MTs increasingly predominate and consolidate into relatively thick, long bundles, while the frequency of non-cortical MTs diminishes. The apparent density of MTs per unit surface area is maintained as the cells expand and gradually flex into an elliptical shape. The guard cells eventually separate completely at the pore site. The entire process is accomplished within about 12 h.

Cytoskeletal organization and chromosome behavior were studied inTradescantia generative cells prior to and during sperm formation using in vitro grown pollen tubes and fluorescence staining methods. Before pollen germination, the crescent-shaped generative cell contains a reticulate microtubule (Mt) system. The cell elongates dramatically after germination, and its Mts assume a helical to longitudinal arrangement. Chromosome condensation is evident approximately 3hr after germination. Kinetochores appear as dark interruptions in the Mt array, and thus seem to attach directly to interphase fibers. No metaphase plate typical of other cells is observed with either DAPI or anti-tubulin staining. Instead, the chromosomes adopt a twisted or braided arrangement, with kinetochores distributed along the length of the cell and kinetochore fibers linked to each other and to surrounding fibers. Anaphase is characterized by a staggered, overlapping separation of chromosomes and by elongation of Mt branches connecting opposing kinetochore fibers. Cytokinesis appears to utilize a furrowing process; a phragmoplast or cell plate was never seen. As a result of these events, the sperm directly inherit their cytoskeleton from generative cell Mts involved in division. No actin fibers are observed at any stage using rhodamine-phalloidin staining. The results are discussed in terms of other reports on sperm formation, possible mitotic and cytokinetic mechanisms, and past distinctions between Mt arrays in higher plant somatic cells.

A liquid phase, competition-binding radioimmunoassay for wheat germ agglutinin, with a detection limit of 10 nanograms, was developed in order to determine the distribution of this lectin in young wheat plants. Affinity columns for wheat germ agglutinin removed all antigenically detectable activity from crude extracts of wheat tissue; thus, the antigenic cross-reactivity detected by the assay possesses sugar-binding specificity similar to the wheat germ-derived lectin. The amount of lectin per dry grain is approximately 1 microgram, all associated with the embryo. At 34 days of growth, the level of lectin per plant was reduced by about 50%, with approximately one-third in the roots and two-thirds in the shoot. The data also indicate that actively growing regions of the plant (the bases of the leaves and rapidly growing adventitious roots) contain the highest levels of lectin. Half of the lectin associated with the roots could be solubilized by washing intact roots in buffer containing oligomers of N-acetylglucosamine, whereas the remainder is liberated only upon homogenization of the tissue.

The chitin-binding lectin wheat germ agglutinin (WGA) is found at the periphery of wheat embryos, and a similar lectin is present at the root tips of older plants (Mishkind et al. 1982). Although a ferritin-conjugated secondary antibody is adequate for localizing WGA in embryos, native electron-opaque particles make the electron microscope identification of added label equivocal in other wheat tissues. As reported here, however, unambiguous ultrastructural localization of WGA-like lectin in adult wheat roots can be obtained with rabbit anti-WGA followed by colloidal gold-labeled goat anti-rabbit (GAR) IgG. Colloidal gold (CG) was prepared by the reduction of gold chloride with citrate, ascorbate or phosphorous. GAR IgG, prepared from serum by antigen affinity chromatograhy, was adsorbed to the gold particles to produce a stabilized suspension of GAR-CG. Localization was performed on 8–12 μM frozen sections of tissue fixed in 4% paraformaldehyde, 0.3% glutaraldehyde, and 0.75% acrolein in phosphate-buffered saline containing 1M sucrose. Localization with GAR-CG was first compared to that ascertained in embryos using other probes and was then extended to the roots of adult plants. An advantage of the GARCG method is that it permits the visualization of antigen at both the light and electron microscope levels in the same section. At the light level, the anti-WGA-GAR-CG complex appears as a red stain that is localized in specific tissues of embryos and in the caps and outer layers of adult roots. Sections in which lectin was detected at the light microscope level were embedded in plastic and sectioned for subcellular examination. Electron dense gold particles indicative of WGA are found at the periphery of protein bodies in wheat embryos and in vacuoles of the roots of adult plants. Sections incubated with control IgG lack reaction product.

Thick sections of fixed, embedded stomatal cells ofPhleum pratense were examined using high voltage electron microscopy and stereological procedures. The cortex of guard cells and subsidiary cells throughout differentiation contains numerous microtubules adjacent to the plasmalemma. Although microtubules are usually aligned in one net direction, individual microtubules may diverge from this orientation in various ways, producing anastomosing or crossed arrays. Also present in the cortex of both guard and subsidiary cells are collections of membranous elements and amorphous material upon which microtubules seem to focus, terminate or overlap. Such structures may constitute microtubule nucleation centers. The significance of these observations is discussed in terms of the control of microtubule development, wall microfibril deposition and cell morphogenesis.

Microtubules (MTs) participate in several processes of fundamental importance to growth and development in higher plants, yet little is known about the proteins with which they associate. Information about these molecules is important because they probably play a role in mediating functional and structural differences between various MT arrays. As a first step in gaining insight into this problem, we have isolated, from suspension-cultured cells of carrot (Daucus carota L.), non-tubulin proteins which bind to and affect microtubules (MTs) in vitro. These proteins were isolated using taxol-stabilized neuronal MTs as an affinity substrate. They cause MT bundling at substoichiometric concentrations, support the assembly of tubulin in vitro, and at low concentrations, decorate single MTs in a periodic fashion. The bundled MTs formed in vitro share similarities with those seen in situ in a variety of plant cells, including a center-center spacing of 34 nm, cold stability, resistance to anti-microtubule drugs, and sensitivity to calcium. The bundling activity is specific; other cationic proteins, as well as poly-L-lysine, do not behave in a similar manner. The bundling activity is insensitive to ATP. By assaying bundling activity with dark-field microscopy and employing standard biochemical procedures, a small number of polypeptides involved in the bundling process were identified. Affinity-isolated antibodies to one of these polypeptides (Mr=76000) were found to co-localize with MTs in the cortical array of protoplasts. Our findings are discussed with reference to the importance of these proteins in the cell and to their relationship to microtubule-associated proteins in other eukaryotes.

Cytochalasin B (CB), thought to disaggregate F-actin in animal cells, and phalloidin (Phal), known to stabilize F-actin in vivo and in vitro, have nearly identical effects on cotyledon epidermal cells of Allium cepa. Both drugs rapidly induce cessation of streaming and both, by preventing normal telophase reorientation movement, lead to abnormal division planes in dividing guard mother cells. Neither, however, prevents normal microtubule deposition, wall thickening, and cellulose orientation during guard cell differentiation. Furthermore, both drugs have no effect on spindle formation and anaphase chromosome motion. Examination of Nitella and Chara cells, in which streaming had been stopped by either agent, shows that microfilament cables are still present. With both drugs, the minimum effective concentrations were routinely used (CB, 2 μM; Phal, 100–200 μM). Our results are discussed in terms of the mode of action of these drugs and their possible role in host-fungus interactions. Implications for the mechanisms underlying cell plate alignment, cellulose orientation, and cytoplasmic streaming are discussed.