Petunia hybrida CAROTENOID CLEAVAGE DIOXYGENASE7 Is Involved in the Production of Negative and Positive Branching Signals in Petunia

One of the key factors that defines plant form is the regulation of when and where branches develop. The diversity of form observed in nature results, in part, from variation in the regulation of branching between species. Two CAROTENOID CLEAVAGE DIOXYGENASE (CCD) genes, CCD7 and CCD8, are required for the production of a branch-suppressing plant hormone. Here, we report that the decreased apical dominance3 (dad3) mutant of petunia (Petunia hybrida) results from the mutation of the PhCCD7 gene and has a less severe branching phenotype than mutation of PhCCD8 (dad1). An analysis of the expression of this gene in wild-type, mutant, and grafted petunia suggests that in petunia, CCD7 and CCD8 are coordinately regulated. In contrast to observations in Arabidopsis (Arabidopsis thaliana), ccd7ccd8 double mutants in petunia show an additive phenotype. An analysis using dad3 or dad1 mutant scions grafted to wild-type rootstocks showed that when these plants produce adventitious mutant roots, branching is increased above that seen in plants where the mutant roots are removed. The results presented here indicate that mutation of either CCD7 or CCD8 in petunia results in both the loss of an inhibitor of branching and an increase in a promoter of branching.The dynamic process that leads to a plant''s architecture is regulated by developmental factors and by environmental conditions. Whether or not axillary meristems grow to form branches is one key component of plant architecture. Plants with altered architecture have been important in agronomy since the earliest selections were made by humans. More recent examples are vital to the productivity of our current farming systems. The domestication of maize (Zea mays) and the dwarfing of wheat (Triticum aestivum) and rice (Oryza sativa; as part of the Green Revolution) involved alterations to plant height and branch number that dramatically improved productivity (for review, see Sakamoto and Matsuoka, 2004).Arabidopsis (Arabidopsis thaliana), rice, pea (Pisum sativum), and petunia (Petunia hybrida) are important model plants in which axillary branching has been studied. The growth habits of these plants show differences when grown under standard floral inductive conditions. This is due, in part, to the differing developmental programs controlling the outgrowth of axillary branches. Petunia (inbred genetic stock V26) produces basal axillary branches between nodes two and eight that begin their growth during the vegetative growth phase (Snowden and Napoli, 2003). Axillary branches may also form in the nodes immediately below the first flower after the floral transition (Napoli et al., 1999). Arabidopsis generally produces axillary branches after flowering, releasing axillary meristems in the rosette and also from cauline leaves (Hempel and Feldman, 1994). Wild-type, tall pea cultivars such as Parvus are very unlikely to produce basal axillary branches at any stage of growth but do branch at the nodes immediately below the first flower (Stafstrom, 1995). Cultivated rice produces basal axillary branches, called tillers, during vegetative growth. The tillers formed early in plant development will produce panicles (flowering branches), and the remainder will senesce (Hanada, 1993). How these differences in development arise is yet to be understood.Although the overall architecture of plants varies considerably, the genes so far identified that control branching are frequently conserved between species. In particular, two CAROTENOID CLEAVAGE DIOXYGENASE (CCD) genes, CCD7 and CCD8, appear to be well conserved among the plant species studied. Mutations in these two genes result in increased branching phenotypes in every species studied to date (Sorefan et al., 2003; Booker et al., 2004; Snowden et al., 2005; Zou et al., 2005; Johnson et al., 2006; Arite et al., 2007). One interesting line of enquiry is to consider whether differences in the regulation or activity of these two genes are involved in the diversity of architecture seen in plants.Grafting experiments have provided insight into the control of axillary branching, in particular the discovery that signals move from roots to shoots. In petunia, Arabidopsis, and pea, some of the increased branching mutants (ccd7 and ccd8 mutants in particular) can be reverted to a wild-type phenotype by grafting mutant scions onto wild-type rootstocks (for review, see Drummond et al., 2009). Additionally, ccd8 mutant plant lines have been reverted to the wild type by the insertion of a small piece (approximately 2 mm) of wild-type hypocotyl into the hypocotyls of mutant petunia or by insertion of a small piece of epicotyl into the epicotyl of mutant pea (Napoli, 1996; Foo et al., 2001). In Arabidopsis, the ccd7 mutant has been similarly reverted using hypocotyl interstock grafts (Booker et al., 2004). Together, these results suggest the presence of a mobile branch inhibitor produced in wild-type tissue. However, an observation by Napoli (1996) suggested that decreased apical dominance1 (dad1) mutant roots may also have a branch-inducing effect in certain circumstances. A similar result was observed for pea in Parvus by Foo et al. (2001). The discussion presented by Napoli (1996) did not exclude either a branch-inducing or a branch-suppressing signal, although current models generally only consider the presence of a branch inhibitor, and recent efforts have focused on the identification of inhibitors of branching.Strigolactones have recently been identified as signaling molecules that inhibit axillary branch outgrowth in plants (Gomez-Roldan et al., 2008; Umehara et al., 2008). Strigolactones were previously identified as signal molecules secreted from roots. When arbuscular mycorrhizae detect strigolactones, they undergo a preinfection hyperbranching response that is thought to aid fungal colonization of the roots, frequently leading to improved nutrient uptake by the plant (Akiyama et al., 2005). The seeds of the parasitic plants Orobanche species and Striga species are also induced to germinate upon detection of strigolactones in the soil, resulting in significant yield losses for some crops (Cook et al., 1966; Siame et al., 1993; Yokota et al., 1998). The production of strigolactones in rice and pea has been shown to require the action of both CCD7 and CCD8 (Gomez-Roldan et al., 2008; Umehara et al., 2008). The discovery that strigolactones can alter branching confirmed a new layer of regulatory complexity in the control of branching that has long been hidden beneath the global plant growth regulators of auxin and cytokinin.In this study, we have focused on the role of the CCD7 gene in the control of branching in petunia. We have isolated a petunia CCD7 ortholog (PhCCD7) and show that the increased branching phenotype of the dad3 mutant is caused by a lesion in this gene. The phenotype of the dad3 mutant is less severe than that of the petunia ccd8 mutant (dad1), and the double ccd7ccd8 mutant is shown to be additive. These observations are contrasted with what has been observed for other plant species. We show that the regulation of PhCCD7 is similar to that of the PhCCD8 gene, with expression predominantly in root and stem tissue (although at a reduced level) and up-regulation of expression in plants with increased numbers of branches. We also provide evidence for the presence of a branch-promoting signal in mutant roots of petunia. These results suggest that there is an added layer of complexity to the control of branching that is not fully described by current models and indicate that the CCD7 gene may have a role in the diversity of plant architecture.