The regulatory anthocyanin loci, an1, an2, an4 and an11 of Petunia hybrida, and r and c1 from Zea mays, control transcription of different sets of target genes. Both an2 and c1 encode a MYB-type protein. This study reports the isolation of a P. hybrida gene, jaf13, encoding a basic helix-loop-helix protein that, on the basis of sequence homology and intron/exon structure, represents the P. hybrida orthologue of the Z. mays r genes. Ectopic expression of an2 and jaf13 is sufficient for activation of the dihydroflavonol 4-reductase-A (dfrA) promoter and enhanced pigment accumulation in P. hybrida. This indicates that an2 and jaf13 play a key role in determining the tissue-specific expression pattern of structural genes. However, because chalcone synthase (chs) and flavanone-3-hydroxylase (f3h) are not activated, the pattern of pigmentation is not fundamentally altered. Expression of an2 in Z. mays complements a mutation in pl, a c1 paralogue, indicating that an2 activates a wider set of target genes in this host. Transient expression assays in Z. mays and P. hybrida tissues showed that C1 and R or AN2 and JAF13 can activate the promoter of the c2 gene, encoding Z. mays CHS, but not the chsA promoter from P. hybrida. These results indicate that regulatory anthocyanin genes are conserved between species and that divergent evolution of the target gene promoters is responsible for the species-specific differences in regulatory networks.

The shape and color of flowers are important for plant reproduction because they attract pollinators such as insects and birds. Therefore, it is thought that alterations in these traits may result in the attraction of different pollinators, genetic isolation, and ultimately, (sympatric) speciation. Petunia integrifolia and P. axillaris bear flowers with different shapes and colors that appear to be visited by different insects. The anthocyanin2 ( an2 ) locus, a regulator of the anthocyanin biosynthetic pathway, is the main determinant of color differences. Here, we report an analysis of molecular events at the an2 locus that occur during Petunia spp evolution. We isolated an2 by transposon tagging and found that it encodes a MYB domain protein, indicating that it is a transcription factor. Analysis of P. axillaris subspecies with white flowers showed that they contain an2 Ϫ alleles with two alternative frameshifts at one site, apparently caused by the insertion and subsequent excision of a transposon. A third an2 ؊ allele has a nonsense mutation elsewhere, indicating that it arose independently. The distribution of polymorphisms in an2 ؊ alleles suggests that the loss of an2 function and the consequent changes in floral color were not the primary cause for genetic separation of P. integrifolia and P. axillaris. Rather, they were events that occurred late in the speciation process, possibly to reinforce genetic isolation and complete speciation. INTRODUCTIONFlowers are the structures containing the male and female sex organs of angiosperms. Flowers of diverse species display a wide range of different morphologies and pollination strategies. For instance, flowers of wind-pollinated species usually possess small and inconspicuous petals or no petals at all, whereas flowers of insect-pollinated plants usually possess large, brightly colored, and patterned petals that serve as visual signals and a landing site for visiting insects.Recent experiments suggest that the wide variety of plant and flower morphologies may have depended on the evolution of a relatively small number of genes. First, mutations at single loci, usually isolated by breeders or researchers, are sufficient to cause fundamental alterations in inflorescence architecture (Doebley et al., 1997;Souer et al., 1998). Similarly, the different shapes, colors, and color patterns of naturally occurring Mimulus (monkeyflower) spp are due to alterations at only a few (major) loci (Bradshaw et al., 1995).Second, even very different inflorescence and flower architectures appear to be determined by genes that often encode conserved proteins but that differ in their expression patterns (reviewed in Doebley and Lukens, 1998).The isolation of key regulatory loci and analysis of the molecular alterations that have taken place in them will provide new insights into the evolution and diversification of flower morphology. The biosynthesis of anthocyanin flower pigments is particularly suited for such studies, because it is a well-defined biochemical pathway that is being...

In this study, we demonstrate that in petunia at least four regulatory genes (anthocyanin-1 [an1], an2, an4, and an11) control transcription of a subset of structural genes from the anthocyanin pathway by using a combination of RNA gel blot analysis, transcription run-on assays, and transient expression assays. an2- and an11- mutants could be transiently complemented by the maize regulatory genes Leaf color (Lc) or Colorless-1 (C1), respectively, whereas an1- mutants only by Lc and C1 together. In addition, the combination of Lc and C1 induces pigment accumulation in young leaves. This indicates that Lc and C1 are both necessary and sufficient to produce pigmentation in leaf cells. Regulatory pigmentation genes in maize and petunia control different sets of structural genes. The maize Lc and C1 genes expressed in petunia differentially activate the promoters of the chalcone synthase genes chsA and chsJ in the same way that the homologous petunia genes do. This suggests that the regulatory proteins in both species are functionally similar and that the choice of target genes is determined by their promoter sequences. We present an evolutionary model that explains the differences in regulation of pigmentation pathways of maize, petunia, and snapdragon.

Establishment of loss-of-function phenotypes is often a key step in determining the biological function of a gene. We describe a procedure to obtain mutant petunia plants in which a specific gene with known sequence is inactivated by the transposable element dTphl. Leaves are collected from batches of 1000 plants with highly active dTphl elements, pooled according to a three-dimensional matrix, and screened by PCR using a transposon-and a gene-specific primer. In this way individual plants with a dTphl insertion can be identified by analysis of about 30 PCRs. We found insertion alleles for various genes at a frequency of about 1 in 1000 plants. The plant population can be preserved by selfing all the plants, so that it can be screened for insertions in many genes over a prolonged period.offspring and found to be due to insertions of dTphl elements (E.S., A.v.H., D.K., J.N.M.M., and R.K., unpublished data). Thus, the high incidence of mutations in the W138 line is mainly due to transposition of dTphl elements.We show how plants with a dTphl insertion in a specific gene can be identified in a large plant population by a polymerase chain reaction (PCR) assay. Such plants are heterozygous for the insertion allele unless the insertion occurred in a previous generation. After selfing, progeny homozygous for the insertion allele are obtained that can be analyzed for an altered phenotype. The screened plant population is maintained by selfing all plants, so that semipermanent libraries are obtained that can be screened at any time. These libraries are open to the scientific community.