Allele-specific degradation in endosperm development
--G. L. Yerk, R. L. Phillips and R. V. Kowles

Endosperm tissue is a product of the union of one sperm nucleus with two polar nuclei, forming a triploid tissue. Following a period of free nuclear division, during which the endosperm is a syncitial tissue, cell walls are laid down giving rise to a cellularized, uninucleate tissue. Thereafter, endosperm growth proceeds by two processes--increases in cell number and increases in cell size. At 8 to 10 days after pollination (DAP), there is a peak in mitotic index defined as the percentage of mitotic cells from the total endosperm cell population. After this period, mitotic activity declines rapidly, to almost zero. At the same time, DNA content per nucleus begins to increase dramatically reaching peak levels between 16 and 18 DAP (Kowles and Phillips, PNAS USA 82:7010-7014, 1985). The increase in DNA content per nucleus is the result of polytenization of the chromosomes (Kowles et al., Dev. Genet. 11:125-132, 1990). Subsequent to the peak in DNA content per nucleus, a reduction in DNA content per nucleus is observed (Kowles and Phillips, 1985). In the case of inbred A188, the reduction in DNA content per nucleus is approximately one third of the maximum DNA content per nucleus.

The purpose of this study was to further characterize the reduction in DNA content per endosperm nucleus observed in endosperm nuclei sampled after 18 DAP. Furthermore, because the reduction in DNA content per nucleus for A188 was approximately one third, experiments were constructed to test whether preferential degradation of the paternal genome had occurred.

Paired-plant reciprocal crosses were made for B73 x Mo17 and A188 x B73 in the summers of 1990 and 1991. Several individuals within each inbred were also self-pollinated. All pollinations for comparison were made on the same day. Endosperm tissue was removed from kernels and frozen for DNA extraction at a time period estimated to correspond to the peak in DNA content (early) and one estimated to correspond to the time period after which degradation had begun using the CTAB method. Endosperm DNA was digested with HindIII or EcoRI. Digested DNA was electrophoresed in 0.8% agarose and transferred to Immobilon N. Blots were probed using clones from a PstI genomic library from the University of Missouri-Columbia (umc) and a 9kb fragment containing the ribosomal genes. For situations where neither allele was absent, autoradiograms were scanned using Apple scan and ratios of band intensity within a lane were determined using Image 1.44 (NIH) to determine whether degradation of one allele was occurring.

In 1990, the maternal allele was consistently and completely lost in the B73 x Mo17 hybrid at the late sampling date but was present in the reciprocal hybrid, Mo17 x B73, at the late sampling date and in both reciprocal hybrids at the early date. This was true for all six of the maize chromosomes tested. The other four chromosomes were not evaluated due to a limited amount of endosperm DNA. Multiple probing of two of the chromosomes demonstrated that the pattern of allele loss is the same for both arms of the chromosome, as well as multiple sequences on the same arm of a chromosome. Probing with pZMR1, the 9kb ribosomal DNA repeat, demonstrated the same pattern of allele loss as seen with the UMC clones. This repeat unit contains the 18S, 5.8S, and 26S ribosomal DNA sequences.

In 1991, allele loss was not as complete. Some probes showed reduced banding intensity while others were completely absent. In addition, the allele being degraded came from B73 in some cases and from Mo17 in other cases. For example, hybridization with umc15 and umc131, both of which demonstrated complete allele loss in 1990, indicated that the restriction fragment from the Mo17 parent was reduced in intensity. In the case of umc131, the reduction was 50%, while for umc15 the reduction was 42%. The results with umc4 showed a different pattern of allele loss. In this material a reduction in banding intensity for both the Mo17 and B73 restriction fragments mapping to chromosome 2 were observed in the B73 x Mo17 hybrid at the early sampling date. At the later sampling date, the same bands were absent in both hybrids. Additional restriction fragments located on chromosome 8 were present in both hybrids at both the early and late sampling times. Probing of the same material with umc39, located adjacent to umc4, indicated that the restriction fragment from Mo17 was reduced in intensity compared to the B73 restriction fragment.

The observed pattern of allele loss was more complex for the A188 x B73 materials in 1990 compared to the Mo17 x B73 materials. Alleles were either present in both reciprocals at both the early and late sampling dates or they were completely absent in all cases. The missing allele came from A188 in some cases and from B73 in other cases. Once again, six of the ten chromosomes were evaluated. In addition, the pattern of allele loss varied within a chromosome. Thus, alleles from different parents were lost at different locations on the same chromosome.

In 1991, both reductions in intensity and complete absence of alleles were observed in the A188 x B73 materials. Again, the alleles involved came from A188 in some cases and B73 in others. If an allele was absent, it was missing in both reciprocals at both sampling dates.

Several conclusions have been drawn from these experiments. First, allele specific degradation is occurring in both sets of material examined. In 1990, the degradation was always complete. In 1991, degradation was partial in some cases and complete in others. This difference may be due to different environmental conditions in the two growing seasons. The rate of endosperm development as measured by DNA content and cell number per endosperm is extremely dependent on heat units (Schweizer et al., PNAS submitted). Second, the pattern of allele loss was not necessarily the same from one season to the next. In some instances, a probe which was completely absent in one season showed only partial reduction in the other season. In other instances, probes showing either complete or partial degradation in one season demonstrated no degradation in the other season. Third, the pattern of allele specific degradation is not universal from one pair of lines to another. The number of patterns that may exist cannot be estimated due to the small number of combinations examined. Fourth, the patterns observed within a pair of lines for a given probe are consistent from one plant to another. Finally, the loss of a particular allele cannot be attributed to artifacts from self-pollination as the maternal allele is often lost.

A number of possibilities exist which might explain the observed patterns of allele loss. These include sequence specific recognition at these sites, protein marking of these sites, conformational differences at these sites, and the possibility that these may be active transcriptional sites. The overall purpose for the degradation of these DNA sequences may be to provide nucleotides which can be stored and used by the developing and germinating embryo. Experiments are in progress to test a number of these hypotheses. 

Return to the MNL 67 On-Line Index
Return to the Maize Newsletter Index
Return to the MaizeGDB Homepage