Microtubule distribution during wild type sporogenesis has been characterized using indirect immunofluorescence microscopy (Dev. Biol., in press). Briefly, sporocytes from the inbred lines A344, B73, F86, and 1S2P were extruded from cut anthers and fixed in a microtubule stabilizing buffer. A short digestion with wall degrading enzymes made these cells permeable to antibodies. Microtubules were stained using a monoclonal antibody against sea urchin tubulin and a secondary antibody coupled to fluorescein. Using such methods, microtubule arrays can be visualized from early prophase I stages until pollen wall formation begins around the young microsporocyte. Modification of these procedures will be necessary to obtain adequate staining of microtubules arrays during mitotic pollen divisions and pollen development.

During normal meiosis, microtubule arrays undergo predictable temporal and spatial changes. Cytoplasmic microtubule arrays usually emanate from the nuclear envelope and radiate towards the cortex during prophase, interphase and telophase of both meiotic divisions. Immediately preceding and after meiotic divisions, microtubules are found to be associated predominantly with the nucleus. It has been postulated that a site closely associated with the nuclear envelope serves as a microtubule organizing center (MTOC). Meiotic spindles of both divisions have highly focused poles and a specific orientation within the sporocytes and in the anther locule. Cytokinesis occurs after both meiotic divisions and is accomplished by a typical phragmoplast that is initiated in the spindle midzone during late anaphase and telophase. The parallel arrays of phragmoplast microtubules propagate centrifugally forming a ring around the newly formed cell plate. Cytokinesis is always completed before the next division ensues. An isobilateral tetrad of microspores is the ultimate product of this controlled pattern of meiotic divisions.

Examination of several existing maize meiotic mutants suggests that abnormal meiosis can be correlated with disruption of the cytoskeleton (Dev. Biol., in press: J. Cell Sci, submitted). Mutants we have examined include dv, ms28, ms43, Mei025 (Mei1), and ms17. The meiotic mutant dv is defective in the transition from a prophase microtubule array to a metaphase spindle. Instead of converging to form focused poles, the metaphase spindle poles remain diffuse as in prometaphase. This defect correlates with several abnormalities in subsequent developmental events. These results suggest that dv is a mutation that affects MTOC organization (Dev. Biol., in press). Breakdown of the anaphase spindle and failure to progress through the cell cycle are characteristic of the mutant ms43 . We speculate that this disruption could be caused by a defective spindle structural component, possibly one that mediates lateral interaction between microtubules. Or more likely, ms43 + is a cell cycle gene necessary for progression through meiosis. A dominant mutant, Mei025, has super-condensed metaphase chromosomes that cannot be separated to daughter cells. However, cytoplasmic radial microtubule arrays characteristic of interphase reform, but are not associated with typical nuclei. Analysis of the mutant ms28 lends further evidence to the hypothesis of Golubovskaya and coworkers (Adv. Genet. 26:149-192) that abnormal anaphase chromosome segregation and partial or complete failure of cytokinesis is due to altered microtubule dynamics. We have observed spindles with an unusual proliferation of astral microtubules, as well as spindle-shaped phragmoplasts that remain in the cytoplasm as late as the second meiotic division. Interestingly, examination of ms17 showed many of the same abnormalities observed in ms28. Most notable are the phragmoplast arrays which fail to propagate and remain in the cytoplasm until division II. However, ms17 exhibits other abnormalities suggesting it is either a stronger allele or a different gene than ms28. Based on the similar meiotic behavior, and the fact that both have been mapped to chromosome 1S, we suspect that these male sterile mutations are allelic. Genetic tests of this are currently underway, but we would be interested to learn of any other attempts to check allelism between these mutations.

Ultimately we would like to analyze meiosis and the cytoskeleton at the molecular level. We have generated and partially characterized several meiotic mutants induced by Mu mutagenesis. During the summer of 1987, we examined 2000 Mu F2 families (20 plants each) for segregation of a male sterile phenotype. Sixty-six families contained plants with defective anther exsertion, shriveled anthers, and either no pollen shed or greater than 10% of the pollen shriveled. Twenty plants for each of these families were regrown in the field during the summer of 1988. Sporocyte samples were collected from each plant, and examined for meiotic defects using temporary DAPI stained preparations. Seventeen families, representing 9 Mu F1 lines, contained plants with meiotic abnormalities. Sterile plants from each of these lines were outcrossed using B73 as the pollen parent. Heritability of these putative ms lesions was examined by selfing presumed ms/+ plants and examining the progeny for meiotic defects. Of the 9 original classes, 1 class failed to segregate any meiotic disruption and 2 others were too incompletely expressed for further study. Furthermore, 2 classes of meiotic lesions reduced female sterility to the extent that outcrosses of sterile plants are impossible. Analysis of heterozygous sibs is being pursued to determine if these two classes are heritable. The remaining four classes segregate meiotic, male sterile lesions as simple monogenic recessive traits. The most interesting of these is similar in phenotype to ms17. Although we have not yet examined microtubule distribution in this new mutant, the earliest detectable defects, based on chromosome behavior, are observed during metaphase I. This mutant, designated mms23*, is unable to align its chromosomes normally on a metaphase plate, and often forms multiple spindles. The end-product of meiosis is typically polyads of microspores, but other products are possible. Allelism tests between this mutant and ms17 are in progress.

Return to the MNL 64 On-Line Index
Return to the Maize Newsletter Index
Return to the MaizeGDB Homepage