Anthocyanin biosynthesis in maize requires two classes of regulatory proteins, C1 and R, proposed to be transcriptional activators responsible for the coordinate expression of a suite of structural genes, including Bronze2 (Bz2). Structural analysis of the Bz2 promoter was performed in electroporated BMS maize protoplasts in which 35S:R ("pR") and 35S:C1 ("pC1") plasmids activate the endogenous anthocyanin structural genes resulting in pink protoplasts within 24 hr in virtually all viable cells. These plasmids also activate expression of a Bz2:luciferase reporter construct (Bodeau and Walbot, Mol. Gen. Genet. 233: 379-387, 1992). To better understand the coordinate regulation of the maize anthocyanin structural genes, we analyzed the structure and function of the Bz2 promoter by deletional and site-directed mutagenesis of chimeric reporter gene constructs.

Deletional analysis showed that sequences necessary for regulated expression of Bz2 reside between -63 and -84 relative to the major transcription start site. Sequences from -2200 to -224 could be deleted with no effect on luciferase expression either in the presence or in the absence of co-electroporated pR and pC1. Deletion to -134 and to -84 decreased expression about 50% and 90%, respectively, but these promoters were still inducible. More severe deletions, to -63 and to -48, virtually eliminated inducibility. Thus, sequences critical for R and C1 inducibility likely reside between -63 and -84 in the Bz2 promoter, with sequences out to -224 contributing to the overall induction. There is no evidence for negative elements whose deletion would allow R- or C1- independent expression, because expression remained very low in the absence of the two regulators.

Previous analysis of the Bz1 promoter indicated that two sequence motifs were important for induction by R and C1 (Roth et al., Plant Cell 3: 317-325, 1991). The first is a TAACTG element (which we designate a "C1-motif"), the sequence bound by Myb, an animal homologue of C1. The second element is a CAGGTG sequence similar to CACGTG (which we designate an "R-motif"), the consensus binding site of many bHLH proteins, homologous to R. Site-directed mutagenesis of our Bz2:luciferase plasmid demonstrated that R- and C1-motifs located between -63 and -112 are important for Bz2 promoter activation as well. Three putative R-motifs were mutated: R:-91 (CACGAG from -91 to -86), R:-68 (CACGAC from -68 to -63), and R:+5 (GAGGTG from +5 to +10). Two putative C1-motifs were mutated: C1:-112 (CAGTTA from -112 to -107) and C1:-78 (CGGTCA from -78 to -73). Mutating either the C1:-78 element or the R:-68 element alone decreased expression of the reporter gene, but the other three motifs could be mutated without a significant change in luciferase levels. Combinations of double and triple mutants indicated that the C1:-112 element is also important for full promoter activity.

We noticed a region from -191 to -139 that contained several R- and C1-motifs, but that could be deleted with less than a 50% reduction in promoter inducibility by R and C1. To test whether this was a redundant R/C1 responsive region, we fused it to two non-inducible Bz2 promoter deletions and to a truncated 35S promoter, and assayed the inducibility of the chimeric promoters. Each of the base plasmids lacking the upstream promoter region was at most 2-fold inducible by pR and pC1. With the addition of the -191 to -139 promoter fragment, the resultant promoters were induced about 10-fold by pR and pC1. The absolute expression level was, however, only about 3% the level of the wild-type Bz2 promoter. Interestingly, the upstream fragment inserted in the reverse orientation was equally functional. Thus the -191 to -139 region may act as an enhancer-like R and C1 responsive region that is qualitatively, but not quantitatively similar to the -63 to -84 R and C1 control region.

While chimeric reporter genes are powerful tools for regulatory studies, a potential hazard is that regulatory regions present in the coding sequence of the native gene may be overlooked. To test for such control, we introduced a 4 nt insertion into a genomic Bz2 clone, in order to differentiate transcripts from the endogenous gene, and reconstructed a series of Bz2 genes with truncated and mutated promoters. We compared the mRNA expression levels in electroporated protoplasts using RNase protection. Just as we observed using reporter constructs, deletion of promoter sequences to -224 or to -134 had virtually no effect on mRNA level, while deletion to -48 resulted in very low mRNA accumulation. Mutating the C1:-78 motif also resulted in about a 75% decrease in mRNA accumulation. These results confirmed that R/C1 regulation of Bz2 is primarily transcriptional and that the regulation is mediated through the promoter elements identified in reporter gene expression assays. 

Return to the MNL 68 On-Line Index
Return to the Maize Newsletter Index
Return to the MaizeGDB Homepage