Hormonal Regulation of Branching in Grasses

Axillary meristems, which form in the axils of leaves, play an essential role in plant architecture and reproduction. During vegetative development, axillary meristems give rise to branches, called tillers in grasses, while during reproductive development, axillary meristems give rise to flowering branches or to flowers. The control of branching by axillary meristems is under hormonal, environmental, developmental, and genetic control. In this Update I review the role of hormones in regulation of axillary meristem initiation and outgrowth during both vegetative and inflorescence branching. Hormones play a critical role in regulating branching (McSteen and Leyser, 2005; Beveridge, 2006; Ongaro and Leyser, 2008). Auxin is required for axillary meristem initiation during both vegetative and inflorescence development. In addition, basipetal movement of auxin from the shoot apex suppresses axillary bud outgrowth in the phenomenon known as apical dominance. Cytokinin regulates meristem size and hence indirectly affects branching (Shani et al., 2006; Kyozuka, 2007). In addition, acropetal movement of cytokinin coming from the roots promotes axillary bud outgrowth. It has long been proposed that an additional hormone travels acropetally from the root to inhibit bud outgrowth. The highlight of last year was the identification of this new plant growth hormone, strigolactone (Gomez-Roldan et al., 2008; Umehara et al., 2008). The environment also plays a significant role in regulating branch outgrowth. It is commonly known, especially in grasses, that increased planting density leads to reduced branching (Doust 2007a, 2007b; Kebrom and Brutnell, 2007). This could be due to shading, which is known to inhibit branching, or competition for resources, as fertilizer and nutrients are known to promote branching. Divergent mechanisms of tillering in the grasses. A, Maize with a single axis of growth. One or two axillary shoots containing the female inflorescence (the ear) grow out. Tillers are generally suppressed. B, Several young rice plants showing that tillers grow out early in vegetative development. C, Ornamental Miscanthus. Tillers grow out from rhizomes beneath the soil surface. A triploid variety, Miscanthus × giganteus, is being investigated as a potential biofuel crop. [See online article for color version of this figure.] During inflorescence development in Arabidopsis, floral meristems arise in the axils of reduced leaves called bract leaves (Long and Barton, 2000; Grbic, 2005). In grass inflorescence development, axillary meristems give rise to branches and spikelets before they give rise to flowers (McSteen et al., 2000; Bommert et al., 2005). The identity and determinacy of these different meristem types are controlled by transcription factors (Bortiri and Hake, 2007; Update by Thompson and Hake, this issue [Thompson and Hake, 2009]). Although hormones have been implicated in regulation of inflorescence branching, the exact mechanism is unknown (Barazesh and McSteen, 2008). The differences in the activity of axillary meristems produced during development imply that all axillary meristems do not respond similarly to the same stimuli. For example, in Arabidopsis, there is acropetal outgrowth of buds during vegetative development and basipetal outgrowth of axillary buds during reproductive development (Hempel and Feldman, 1994; McSteen and Leyser, 2005). In grasses, the basal nodes are the ones from which tillers arise (Fig. 1B). Heterochronic mutations that extend the juvenile phase lead to increased tiller number (Poethig, 1988; Chuck et al., 2007). In foxtail millet (Setaria italica), branches are also produced from the upper nodes of the plant, in addition to tillers from basal nodes, and these are under separable genetic and environmental control (Doust et al., 2004; Doust, 2007a). These studies indicate that genetic, hormonal, and environmental signals intersect with developmental signals. Many genes have been identified that regulate axillary meristem initiation and outgrowth during vegetative and reproductive development (Schmitz and Theres, 2005; Bennett and Leyser, 2006; Doust 2007b). Most of these genes have been identified as mutants that have fewer branches, because axillary meristems fail to initiate, or that are bushier than normal due to constitutive outgrowth of axillary buds. Some of these genes encode integral components of hormone biosynthesis, perception or signaling pathways providing a direct link to hormone control. Here, I highlight the conservation and diversification of the mechanisms that control branching within grasses and between grasses and eudicots. Many of the genes regulating axillary meristem initiation affect both vegetative and reproductive development (Table I Genes regulating axillary meristem initiation in monocots and eudicots Genes regulating axillary meristem initiation in monocots and eudicots Auxin plays a role in axillary meristem initiation in the inflorescence. A, Arabidopsis inflorescence with secondary branches (br) and flowers (fl). B, pid mutant in Arabidopsis showing the pin-shaped inflorescence due to lack of flower production. C, Normal maize tassel with long branches (br) at the base of the main spike. Short branches called spikelet pairs (sp) cover the branches and the central spike. D, bif2 mutant of maize with very few branches and spikelets. [See online article for color version of this figure.] Three PIN1 loci have been identified in maize and rice (Xu et al., 2005; Carraro et al., 2006; Gallavotti et al., 2008b). The maize ZmPIN1a gene complements the Arabidopsis pin1 mutant, restoring its ability to make flowers, indicating that ZmPIN1a likely functions in auxin transport in maize (Gallavotti et al., 2008b). ZmPIN1a is localized in all axillary meristems and lateral organ primordia in maize similar to Arabidopsis (Carraro et al., 2006; Gallavotti et al., 2008b). Therefore, even though maize makes multiple types of axillary meristems in the inflorescence, all are characterized by expression of PIN1. Antisense knockdown of the rice OsPIN1 gene supports the conserved function of PIN1 genes in auxin transport in monocots (Xu et al., 2005). barren inflorescence2 (bif2) mutants in maize have a phenotype analogous to the pin phenotype in Arabidopsis (Fig. 2, C and D). In bif2 mutants, fewer branches and spikelets form; however, these arise from the first axillary meristems produced by the inflorescence (McSteen and Hake, 2001). The maize BIF2 gene encodes a Ser/Thr protein kinase co-orthologous to PID (McSteen et al., 2007). BIF2 phosphorylates ZmPIN1a and affects ZmPIN1a localization, indicating that BIF2 regulates PIN1 in maize, similar to the function of PID in Arabidopsis (A. Skirpan and P. McSteen, unpublished data; Friml et al., 2004; Michniewicz et al., 2007). Overexpression of rice OsPID has effects on seedling development similar to the effects of treatment with an auxin transport inhibitor, providing further evidence for a function in auxin transport (Morita and Kyozuka, 2007). Therefore, it is likely that PID-like proteins have conserved roles in regulating auxin transport in monocots and eudicots. On the other hand, BIF2 is localized in the nucleus as well as at the cell periphery, indicating that it plays additional roles in development (Skirpan et al., 2008). Yeast two-hybrid screening with PID identified calcium-binding proteins that act upstream of PID (Benjamins et al., 2003). In contrast, a yeast two-hybrid screen with BIF2 identified BARREN STALK1 (BA1), a nuclear localized basic helix-loop-helix putative transcription factor as an interacting partner with BIF2 (Gallavotti et al., 2004; Skirpan et al., 2008). In vitro kinase assays showed that BIF2 phosphorylates BA1 (Skirpan et al., 2008). bif2 and ba1 mutants both have defects in axillary meristem initiation, indicating that the in vitro interaction may have in vivo relevance (McSteen and Hake, 2001; Ritter et al., 2002; Skirpan et al., 2008). Therefore, in maize, BIF2 also functions in the nucleus. Whether the same is true for PID in Arabidopsis or OsPID/OsBIF2 in rice is not yet known. Although PID has not been reported to be nuclear localized, other PID-like proteins in Arabidopsis have been reported to be localized in the nucleus (Zegzouti et al., 2006). There is an ortholog of BA1 in rice, LAX PANICLE (LAX; Komatsu et al., 2003), but an ortholog of BA1 has not been reported in Arabidopsis. While the exact molecular mechanism of the interaction of BA1/LAX with auxin is not clear, there does appear to be a connection with auxin (Gallavotti et al., 2008b; Skirpan et al., 2008). As already mentioned, BA1 has been shown to physically interact with BIF2 (Skirpan et al., 2008). BA1 may be auxin induced, as it is not expressed following treatment with an auxin transport inhibitor (Wu and McSteen, 2007). In addition, the auxin reporter DR5 indicates that auxin maxima do not form in the ba1 mutant (Gallavotti et al., 2008b). Overexpression of LAX in rice leads to defects in development, indicating that LAX functions in auxin-mediated development (Komatsu et al., 2003). Therefore, it is likely that BA1/LAX plays a role in auxin-mediated axillary meristem initiation. Localized auxin biosynthesis is also required for axillary meristem initiation in maize as in Arabidopsis, but, interestingly, there is a higher level of redundancy in Arabidopsis than in maize. The sparse inflorescence1 (spi1) mutant of maize has fewer branches and spikelets due to the absence of axillary meristems (Gallavotti et al., 2008a). SPI1 encodes a monocot-specific YUC gene family member required for localized auxin biosynthesis (Gallavotti et al., 2008a). Knockout of four YUC genes is required in Arabidopsis to generate a phenotype as severe as the spi1 mutant in maize (Cheng et al., 2006). Phylogenetic analysis shows that this is due to massive expansion of the gene family in both monocots and eudicots. Interestingly, the expression of SPI1 differs from the expression of OsYUC1, the ortholog of SPI1 in rice (Yamamoto et al., 2007). Furthermore, antisense knockdown of OsYUC1 in rice has severe defects in root and stem elongation, but a defect in the inflorescence was not reported (Yamamoto et al., 2007). Therefore, there has been functional diversification of the YUC gene family even within the grasses. BIF2, BA1/LAX, and SPI1 also play a role in axillary meristem initiation during vegetative development (Ritter et al., 2002; Komatsu et al., 2003; McSteen et al., 2007; Gallavotti et al., 2008a). As maize does not normally tiller, these studies took advantage of the teosinte branched1 (tb1) mutant of maize (discussed further in the outgrowth section). tb1 mutants are highly tillered, because all axillary buds that are normally suppressed grow out producing a bushy plant (Doebley et al., 1997). Double mutant combinations between ba1 and tb1 produced no tillers, indicating that BA1 is required for axillary meristem initiation during vegetative development (Ritter et al., 2002). bif2;tb1 and spi1;tb1 double mutants also produced fewer tillers than tb1 single mutants, indicating that they too function in axillary meristem initiation (McSteen et al., 2007; Gallavotti et al., 2008a). In rice, double mutants between lax and small panicle (spa) lead to an absence of tillers, indicating that LAX and SPA play overlapping roles in tiller development (Komatsu et al., 2003). In contrast, pid1, pin1, and yuc mutants in Arabidopsis do not apparently produce fewer side branches (Bennett et al., 1995; Cheng et al., 2006). In fact, plants containing a knockout of multiple YUC genes have a bushy appearance due to reduced apical dominance. These differences could be due to genetic redundancy, but it is interesting that the initiation of vegetative axillary meristems in Arabidopsis does not appear to be as sensitive to reductions in auxin biosynthesis or transport as in maize. Other transcription factors that regulate axillary meristem initiation in monocots and eudicots include the GRAS-type transcription factor LAS1/LS/MOC1 (Schumacher et al., 1999; Greb et al., 2003; Li et al., 2003) and the HD ZIP class III transcription factor REV/OsHB3 (Otsuga et al., 2001; Itoh et al., 2008; Table I). The NAC domain transcription factors CUC1,2,3 play a role in axillary meristem initiation in Arabidopsis, but in rice, overexpression of OsTIL1 enhances axillary meristem outgrowth rather than initiation (Vroemen et al., 2003; Hibara et al., 2006; Mao et al., 2007; Raman et al., 2008). R2 R3 Myb transcription factors RAX1,2,3/BLIND (Schmitz et al., 2002; Keller et al., 2006; Muller et al., 2006) play a role in eudicots, but a homolog in monocots has not yet been identified. Therefore, the roles of these transcription factors still need to be clarified in monocots. Furthermore, the relationship of these transcription factors with hormonal regulation is not clear. Cytokinin also regulates branch and spikelet number in grasses, but in this case, the effect on branching is a secondary effect due to a defect in the shoot apical meristem. Cytokinin plays a fundamental role in regulation of apical meristem size (Shani et al., 2006; Kyozuka, 2007; Zhao, 2008). In Arabidopsis, increased cytokinin levels lead to increased meristem size and reduced cytokinin levels lead to reduced meristem size (Nogue et al., 2000; Werner et al., 2003). Two recent articles have demonstrated the fundamental importance of cytokinin in branching and, hence, yield in rice. CYTOKININ OXIDASE (CKX) is an enzyme that degrades cytokinin (Sakakibara, 2006). Mutations in the rice CKX gene led to increased panicle branch and spikelet number in the inflorescence and increased yield (Ashikari et al., 2005). Conversely, mutations in the rice LONELY GUY gene, which encodes an enzyme that catalyzes the last step in cytokinin biosynthesis, led to the production of fewer branches and spikelets and decreased yield (Kurakawa et al., 2007). Therefore, the role of cytokinin in meristem size to be conserved in monocots and eudicots. for and factors bud outgrowth. The an axillary bud in the of a and factors or not the axillary bud out. hormones regulate branching. Auxin bud cytokinin acropetally promotes while acropetal movement of strigolactone bud outgrowth. factors as and density inhibit bud while fertilizer and nutrients promote bud outgrowth. The role of auxin in apical is also by mutants with defects in auxin biosynthesis, and For example, mutants in Arabidopsis, as auxin are bushy et al., of auxin biosynthesis by multiple yuc mutants reduced apical (Cheng et al., 2006). reductions in auxin transport led to increased tillering in monocots. OsPIN1 knockdown mutants the number of tillers, indicating that reduced auxin transport in outgrowth of tillers even in rice, which has a number of tillers (Xu et al., 2005). OsPIN1 knockdown mutants also tiller that tillers at a than This phenotype is also in other mutants that affect auxin transport in rice and indicates that auxin transport is required for tiller as well as tiller outgrowth et al., 2007). As auxin does not the bud to inhibit bud a was proposed et al., 2003). Cytokinin was a for the as cytokinin travels from the root to the shoot and the bud it promotes outgrowth (Fig. Ongaro and Leyser, and auxin has been shown to inhibit cytokinin biosynthesis et al., 2004; et al., 2006). there is also evidence that cytokinin of auxin et al., that cytokinin are bushy et al., 2001; et al., 2001). The identification of class of bushy mutants led to the of the new branching hormone, strigolactone (Fig. A required for regulation of branching was proposed on the identification of the axillary meristem mutants in Arabidopsis, mutants in and decreased apical mutants in Beveridge, 2006; Ongaro and Leyser, 2008; Table Genes regulating axillary bud outgrowth in monocots and eudicots Genes regulating axillary bud outgrowth in monocots and eudicots encodes et al., 2004; et al., and encode et al., 2003; et al., 2005). These are proposed to act in to an unknown et al., encodes a enzyme that of and in the of this unknown et al., 2005). of a number of highly mutants from rice showed that this is conserved in grasses. encodes an ortholog of et al., 2006) and encodes an ortholog of et al., 2007). The in the rice mutants is a secondary effect of the of tillers, as of tillers plant et al., 2006). The was on for the The was the was I first as a of from plant roots that promote the of of plants et al., 2007). monocots and are a of in eudicots. also to be in of roots by et al., 2005). In this case, the interaction is to the plant, to increased of to the was the of an article showing that are from the et al., 2005). The that are in shoot branching from two on the of roots by to the genes regulating the biosynthesis of As the mutants are in they the on the The first that the genes may be in of a strigolactone it was that the mutants to interact with (Gomez-Roldan et al., 2008). from a the rice that the rice mutants in et al., 2008). In both and rice, it was shown that root of the mutants are in (Gomez-Roldan et al., 2008; Umehara et al., 2008). the rice also showed that mutants have levels of strigolactone et al., 2008). In the of is of the that are (Gomez-Roldan et al., while in rice, and are et al., 2008). The of a of the The direct addition of to the buds of or Arabidopsis mutants branching (Gomez-Roldan et al., 2008). The rice to that of to rice or Arabidopsis mutants of branching et al., 2008). the it was shown that root from the mutants to promote of (Gomez-Roldan et al., while the rice mutants to be by et al., 2008). a strigolactone may be the for auxin in the control of apical dominance. are Some of the mutants not respond to and proposed to be in perception of the encode an protein that is a of the et al., 2002; et al., 2006). The rice ortholog is et al., 2005). The to the which promotes the and of are in multiple hormone signaling they a transcription factor required for signaling (McSteen and Zhao, 2008). A that is the of It be a a transcription that is in bud outgrowth that its by the bud outgrowth. that does the the protein in the same that auxin the auxin protein et al., interesting is the strigolactone by and there an protein expressed in or in The of may also of the environmental effects of branching. As strigolactone biosynthesis is by and it is proposed that this to these nutrients et al., 2007). The rice that strigolactone a mechanism for the plant to to the to branching nutrients are not or to branching nutrients are et al., 2008). Furthermore, is also in perception of signals providing a mechanism to signals with branching et al., 2007). Although strigolactone to be a conserved of branching in monocots and eudicots, differences in between components have been identified et al., 2005; et al., 2006). For example, is by auxin in the shoot in and is by auxin in the shoot in rice et al., 2005; et al., but is by auxin in the root in Arabidopsis et al., 2005). Furthermore, in and in rice are in other mutants et al., 2005; et al., while is not in the mutants et al., 2005). As rice, and Arabidopsis have different with to branch it is not that there be differences in the of the interaction between analysis in these may which nodes are conserved and which of is required to differences in the of between and strigolactone could differences in plant One of the of in plants is the that on the expression of the was in the of maize from its teosinte plants are highly branches grow out at the last in a very bushy on the other hand, has a single axis of growth (Fig. encodes a putative transcription factor expressed in axillary buds (Doebley et al., et al., 2002). is expressed in that are suppressed in both maize and teosinte et al., 2002). In maize, there is an in the expression of to buds to be by the cell and et al., 2002; Li et al., 2005). The in expression is proposed to be due to at a upstream of the et al., 2006). of the from rice has shown that also tillering in rice even though rice is already et al., 2003). of function of the mutant, which has tillers and a stem than normal et al., 2003). Whether the stem is a secondary effect of increased tillering is not known, but overexpression of also leads to reduced tiller number and increased Overexpression of maize also leads to a in tiller number in et al., 2008). is by environmental planting shading, and fertilizer treatment 2007b; Kebrom and Brutnell, 2007). mutants in rice are still by planting indicating that is not for this et al., 2003). In it has been shown that to signaling et al., 2006). is by plants and is plants are at there is a decreased to This leads to the of which is decreased branching and Brutnell, 2007). The to is by the In mutants have reduced branching and increased expression et al., 2006). Therefore, signals as of the There are genes to in Arabidopsis, but to play a similar role in Arabidopsis as in grasses et al., 2007; 2007). to maize, is expressed in axillary and mutants have branches in from Furthermore, analysis of showed that as an of hormonal and environmental signals to regulate or not a branch out et al., 2007; 2007). of the as the levels of are reduced in mutants et al., 2007; 2007). Furthermore, double mutants between and mutants are as bushy as single mutant, indicating that in the same as the It was proposed that the regulates at the level to branching. also with mutants that YUC have fewer branches, and double mutants have branches, indicating that is required for auxin-mediated apical dominance. The effect of auxin on levels was not as clear. One not a significant in levels in the or mutants et al., but that is auxin on buds 2007). levels also by and planting in eudicots, is proposed to act of auxin and the the interaction of with auxin and the strigolactone differs in rice. Arabidopsis, the expression of was not in the first of mutants even though tillers grow out et al., 2007). Therefore, is proposed to act of the in rice et al., 2007). double mutants between the mutants and have not yet been Furthermore, is not by the mutant was to be to auxin et al., 2007). It is that is an of hormonal and environmental signals in both monocots and eudicots. the between components to be the different growth of maize, rice, and Arabidopsis. As on the has in maize, it be interesting to the of in different plant to these differences in are due to in the of There is at the in the development of grasses as et al., 2004; 2007). There are multiple mechanisms for but the by grasses is the production of an stem called the The from an axillary meristem from the basal of the Axillary meristems on the give rise to tillers (Fig. the of the nutrients from the plant are to the which has the to The following tillers from the of the regulation of axillary meristem initiation and outgrowth from rhizomes be critical for to grasses as and Miscanthus for in biofuel production. Interestingly, it that of the mechanisms function are similar to already known to regulate axillary Auxin is in initiation as auxin is required to produce rhizomes in and treatment of plants with auxin transport the initiation of rhizomes and 2006). The levels of auxin and in the during the of the plant et al., 2007). For example, during rhizomes are there are auxin and cytokinin while during the rhizomes are there are cytokinin and auxin Therefore, hormones play a fundamental role in the control of tillering by Furthermore, even though tillers in grasses arise from they are still under the control of the same environmental that regulate tillering in grasses et al., 2001; et al., For example, increased planting density tiller number and increased fertilizer treatment tiller studies that is controlled by few genes et al., 2003; and The identification of these genes be essential for the regulation of branching in grasses. plants have different architecture by the of branching from axillary on monocots and eudicots has shown that similar mechanisms control branching in these As axillary meristems at the base of the it is not to in monocots and eudicots. to be to the function of additional components and the components are with In an of axillary meristems at different of development respond to genetic, environmental, and hormonal factors is A for the be to in the or of interaction between components have led to the of plant of be to axillary meristems by the function of these genes in plant I and of for axillary meristems, for providing the rice plants to and two for on the This review is to the of of a of plant to the of for plant development.