Figure 7.

Model of cytokinin regulation by the shoot branching network. Similar models are presented by Foo et al. (2005) and Johnson et al. (2006), and reviewed by Beveridge (2006) and Dun et al. (2006). New data presented here are incorporated in this model together with the findings of Tanaka et al. (2006). 1, Synthesis of a mobile branching inhibitor (SMS) in the rootstock and shoot is dependent on RMS1/MAX4 and RMS5/MAX3. 2, SMS perception/transduction is dependent on RMS4/MAX2; phenotypic grafting evidence shows that this F-box protein acts predominantly in the shoot and is required for SMS action. 3, Outputs of the RMS4/MAX2 signal transduction pathway include separate inhibition of bud outgrowth and repression of RMS2; suppression of bud formation or branching does not prevent activation of the feedback signal (Fig. 4). 4, RMS2 affects long-distance feedback and exhibits graft-transmissible action (Fig. 6). 5, The long-distance feedback signal activates SMS synthesis in shoot and roots; RMS1 and RMS5 expression is minimal in rms2 mutants. 6, The feedback signal also represses xylem cytokinin from the roots; all rms mutants except rms2 have low xylem cytokinin. 7, It is possible that X-CK directly stimulates bud outgrowth (Fig. 6). 8, Shoot cytokinin stimulates bud outgrowth; direct addition of cytokinin to buds generally causes increased growth. 9, Decapitation activates IPT and enhances local shoot cytokinin content. 10, Decapitation also reduces SMS by suppressing RMS1 and RMS5 gene expression and increases X-CK independently of RMS2 action (Fig. 3). 11, Total shoot cytokinin levels in intact plants are maintained by an unknown homeostatic system (Figs. 1 and 2). Solid arrows are interpretations based on direct evidence; broken arrows are tentative or poorly understood relationships. [See online article for color version of this figure.]