--R. V. Kowles, G. Yerk, R. L. Phillips and F. Srienc
A group of 28 defective kernel (dek) mutants obtained from M. G. Neuffer, University of Missouri, Columbia, MO were studied with regard to DNA content per nucleus and cell number per endosperm. Defective kernels are phenotypically distinct from normal kernels. An obvious characterization is their pronounced smaller size on a segregating ear. The overall mean kernel weight for normal kernels was 215.6 mg compared to 49.2 mg for defective kernels constituting a 4.4-fold difference.
Developmentally normal and defective kernels were collected and fixed at 16 days after pollination (dap). The fixative was replaced with 70% ethanol after 24 hours, and the samples were stored at -20 C. Nuclear suspensions were made in such a way that each preparation essentially consisted of all of the nuclei of the endosperm from a single kernel. This was accomplished by carefully dissecting the endosperm from the pericarp, nucellus, and embryo tissues, and gently forcing it through a stainless steel screen with several successive washes of a grinding medium. By low-speed centrifugation, the grinding medium was replaced with mithramycin buffer which, in turn, was replaced with the fluorochrome mithramycin A. The number of nuclei in each endosperm was calculated by removing samples from the nuclear suspension for use with standard hemocytometric procedures. The nuclei were counted with a combination of fluorescence and low-illiumination phase contrast microscopy.
DNA content per nucleus and endoreduplication patterns were determined by flow cytometry. Nuclei were classified according to small angle light-scattering signals, indicative of the size of the nuclei, and according to fluorescence signals as a measure for the DNA content of the nuclei. Because of the large variation of the properties of the nuclei, the signals were collected on a logarithmic scale. Nuclei classified according to peak fluorescence and light-scattering signals appeared as clustered populations permitting convenient exclusion of debris by gating from further analysis. Gated signals were displayed as one-parameter histograms for analysis.
DNA content per nucleus, indicative of the degree of endoreduplication, and the number of cells per endosperm, indicative of mitotic activity, are presented in Table 1. In all cases, a greatly reduced number of cells per endosperm was observed in the defective kernel versus the normal kernel from segregating ears. The mean cell number for the normal kernels in the 28 strains was 160,000 at 16 dap, and the mean for the defective kernels in these strains was only 46,000. Most defective kernels were also found to have considerably less DNA per nucleus than the normal kernels. The mean DNA content per nucleus in arbitrary units of fluorescence for the normal kernels in the 28 strains at 16 dap was 469 A.U., and the mean for the defective kernels was 401 A.U.
In most of the strains, the defective kernels show a greater percentage of nuclei in the 3C and 6C levels, approximately the same percentage in the 12C level, and a much lesser percentage in the 24C, 48C, and 96C levels. Notable differences in DNA replication patterns, however, were noted among several dek strains. A comparable pattern between the defective and normal kernels is seen in strain 1009 (Fig. 1a, b) and almost a complete absence of endoreduplication in strain 868A (Fig. 1c, d). One example in which fewer rounds of endoreduplication are apparent in the endosperm of the defective kernel is seen in strain 1387A (Fig. 2a, b). In another example, the defective kernel undergoes the same number of rounds of endoreduplication but a lesser percentage of nuclei are found with the higher C levels of DNA (strain 1112, Fig. 2c, d).
Strain 1009 proved to be an interesting exception to the other dek strains in that it shows a similar pattern of endoreduplication between the defective and normal kernels from a segregating ear. The normal kernel had a mean of 512 A.U. of DNA per nucleus, and the defective kernel had a mean of 511 A.U. A comparison of their histograms, which reveal the distribution of nuclei in each of the C level regions, also shows a striking similarity. The cell number in the normal endosperm, however, is 158,000 to only 47,000 in the defective kernel endosperm. Apparently, this mutation affects mitotic activity without having an effect upon the endoreduplication process. This strain may be useful for other studies of endosperm development and cellular kinetics.
Figure 1. Endoreduplication patterns displayed as histograms in which the frequency (Y axis) of nuclei are recorded relative to their fluorescence intensity (X axis). The fluorescence intensity, in turn, is indicative of the DNA content of the nucleus. (a) Strain 1009 normal kernel; (b) Strain 1009 defective kernel; (c) Strain 868A normal kernel; (d) Strain 868A defective kernel.
Figure 2. Endoreduplication patterns of (a) Strain 1387A normal kernel; (b) Strain 1387A defective kernel; (c) Strain 1112 normal kernel; (d) Strain 1112 defective kernel.
Table 1. Cell number and DNA content per nucleus in normal and defective kernel endosperm at 16 DAP.
to the MNL 65 On-Line Index
Return to the Maize Newsletter Index
Return to the Maize Genome Database Page