Salamini (MGNCL 41:100; 42:91; 43:91) first described a new bronze (bz) mutable allele in maize, now referred to as AcBz. McClintock determined that the bz mutable phenotype of AcBz was due to the presence of Ds at the bz2 locus, and that there was an Ac element between the sh and bz loci (McClintock, pers. comm.). McClintock also observed that new bz mutable kernels frequently arose in the AcBz strain. The AcBz-derived bz mutable alleles are collectively called "bz-m5". Four independently derived bz-m5 alleles (A through D; obtained from McClintock) were analyzed. They differ from one another in the frequency at which Ac excises from the bz locus, restoring the bz phenotype. The bz-m5-A allele shows the highest frequency of excision, followed by B, C and D. Of interest in the bz-m5 alleles is the possible correlation between Ac excision patterns and the site of insertion of Ac. To date, differences in excision pattern in other mutable alleles have been correlated with alterations within the element itself (Fedoroff and Kingsbury, in prep.; J. W Schiefelbien, et al., PNAS 82:4783, 1985; Zs. Schwarz-Sommer, et al., EMBO J. 4:2439, 1985). The present alleles, therefore, provide a unique opportunity for determining what effect sequences surrounding the Ac element have on its ability to excise.

Experiments were conducted to determine (1) if the Ac element of each allele is inserted at different positions within the bz gene, and (2) if the Ac element is the same in all four mutable alleles. To determine the site of insertion of Ac, DNA from plants of each strain (A through D) was digested with several restriction enzymes and analyzed by Southern blot hybridization. A 2.2 kb Pst fragment containing the entire bz coding sequences (D. Furtek, pers. comm.) was used as a probe. The sites of insertion of Ac, determined from such analysis, are given in Fig. 1. There appears to be a complete 4.5 kb Ac insertion in all alleles, and the insertions are at different sites within the bz coding region. All four are within 2.0 kb of each other. For all four alleles, the orientation of Ac relative to the bz gene is the same. There were no insertions within the bz gene in DNA isolated from AcBz plants.

To investigate the similarity of the Ac elements in the four alleles, we have undertaken the cloning of the bz locus of the bz-m5 alleles. Analysis of one clone with restriction enzymes indicates that the Ac element is indistinguishable from Ac elements isolated from two other mutable alleles, bz-m2 and wx-m9. Ac elements from other bz-m5 alleles are currently being isolated in a similar manner for further analysis and comparison.

From these results, we conclude that each bz-m5 arises from the independent insertion of Ac as a result of short-range transposition of the Ac element into the bz locus. Furthermore, the differences in sites of insertion of Ac within the bz gene correlate with differences in excision pattern of Ac from the gene. There are several possible explanations for the latter observation. First, because the bz gene contains introns (J. Schiefelbein, pers. comm.), differences in the excision frequency of Ac could correspond to the presence of Ac in an exon or an intron sequence. A revertant allele of wx-m9 contains a 6 bp duplication where Ac was originally inserted (Pohlman et al., Cell 37:635, 1984). This small insertion would maintain the correct reading frame of wx, allowing a wild-type wx phenotype. Excision of Ac from an intron, for example, may not have to be as precise as excision from an exon for bz activity to be restored. Second, if the sequences immediately flanking the Ac element are part of the substrate recognized by the putative Ac "transposase" (the transacting transposition function, Pohlman et al., Cell 37:636, 1984), altering these sequences could alter the rate at which Ac excises to restore gene activity. To further investigate these possibilities, we are analyzing the bz sequences flanking the Ac element in all four bz-m5 alleles, as well as characterizing sites of Ac insertion in seven new, independently derived bz-m5 alleles.

Figure 1.

M.A. Johns¹, N. Fedoroff and J. Banks

¹Present address: Northern Illinois University, Dekalb, Illinois
 
 

Return to the MNL 60 On-Line Index
Return to the Maize Newsletter Index
Return to the MaizeGDB Homepage