Arabidopsis inositol polyphosphate 6-/3-kinase (AtIpk2beta) is involved in axillary shoot branching via auxin signaling.

The Arabidopsis (Arabidopsis thaliana) inositol polyphosphate 6-/3-kinase gene (AtIpk2β) is known to participate in inositol phosphate metabolism. However, little is known about its physiological functions in higher plants. Here, we report that AtIpk2β regulates Arabidopsis axillary shoot branching. By overexpressing AtIpk2β in the wild type and mutants, we found that overexpression of AtIpk2β leads to more axillary shoot branches. Further analysis of AtIpk2β overexpression lines showed that axillary meristem forms earlier and the bud outgrowth rate is also accelerated, resulting in more axillary shoot branches. The AtIpk2β promoter/β-glucuronidase (GUS) fusion (AtIpk2β∷GUS) expression pattern is similar to that of the auxin reporter DR5∷GUS. Moreover, AtIpk2β can be induced in response to exogenous indole-3-acetic acid (IAA) treatments. In addition, AtIpk2β overexpression plants exhibit IAA-related phenotypes and are more resistant to exogenous IAA treatments. Further analysis employing reverse transcription-polymerase chain reaction shows that some genes, including auxin-biosynthesis (CYP83B1), auxin-transport (PIN4), and auxin-mediated branching genes (MAX4 and SPS), are regulated by AtIpk2β. Taken together, our data provide insights into a role for AtIpk2β in axillary shoot branching through the auxin signaling pathway.

The pattern of axillary shoot branching and the growth of axillary shoots determine to a large extent the growth and developmental status of plants. In the majority of flowering plants, the primary shoot apical meristem (SAM) is activated during embryogenesis and followed by the formation of additional meristems (Steeves and Sussex, 1989). The primary plant axis is provided by the SAM, whereas the architecture of the shoot system is further determined by the activation of axillary meristems. Compared to the SAM, axillary meristems may initiate and then develop into either a branch or a blocked axillary bud (Stafstrom and Sussex, 1992;Evans and Barton, 1997). Axillary shoot branching generally involves two developmental stages: the formation of axillary meristems and the outgrowth of axillary buds. In Arabidopsis, maize (Zea mays), tomato (Solanum lycopersicum), petunia (Petunia hybrida), and pea (Pisum sativum), various mutants with axillary shoot branching defects have recently been identified (Shimizu-Sato and Mori, 2001;Ward and Leyser, 2004;McSteen and Leyser, 2005). Although the precise mechanisms controlling axillary shoot branching are poorly understood, some factors are demonstrated to be important to this process.
Arabidopsis IP3Ks (AtIpk2a and AtIpk2b) expressed in stem, leaf, stigma, siliques, and fast-growing regions, including root tips and root hairs (Xia et al., 2003;Xu et al., 2005), which implied that Arabidopsis IP3K may play important roles in plant growth and development. As expected, AtIpk2a transgenic plants exhibited superiority in pollen germination, pollen tube growth, root growth, and root hair development (Xu et al., 2005). However, the physiological functions of AtIpk2b in higher plants remain unknown.
In this study, we demonstrate a novel role for AtIpk2b in axillary shoot branching. Moreover, we investigated the correlation between AtIpk2b expression pattern and in vivo auxin reporter DR5TGUS, as well as the responses of AtIpk2b overexpression plants to IAA treatments. Finally, we analyzed the expression of auxin-related genes in AtIpk2b overexpression plants. Our results suggest that AtIpk2b plays a role in axillary shoot branching through the auxin signaling pathway.

Overexpression of AtIpk2b Generates More Axillary Shoot Branches
We have previously shown that the AtIpk2b gene (AGI locus no. At5g61760) is expressed in various Arabidopsis organs, including roots, stems, leaves, and flowers (Xia et al., 2003). To further analyze the physiological functions of AtIpk2b, we overexpressed the AtIpk2b gene in Arabidopsis. Eleven AtIpk2b overexpression plants, named OX-2, OX-3, OX-5, OX-8, OX-9, OX-15, OX-25, OX-26, OX-33, OX-35, and OX-64, were identified and confirmed by northern-blot analysis (Fig. 1A). The expression levels of the AtIpk2b overexpression lines varied (Fig. 1A). Six of the AtIpk2b overexpression lines displayed elevated levels of AtIpk2b protein as determined by western-blot analysis using an antibody raised against maltose binding protein (MBP)-AtIpK2b fusion protein (Fig. 1B). After two generations of segregation (seeds were germinated on Murashige and Skoog [MS] medium containing hygromycin), homozygous AtIpk2b overexpression plants were selected for further studies. Two overexpression lines (OX-9 and OX-26) exhibiting a relatively high AtIpk2b protein level were chosen for the following experiments (Fig.  1B). Compared to the wild type, we found that branching of axillary shoots was more pronounced in the overexpression lines ( Fig. 2A). At maturity, AtIpk2b overexpression plants had increased axillary branches.
To determine the effects of AtIpk2b on shoot architecture precisely, we examined the shoots from wildtype and AtIpk2b overexpression plants. The AtIpk2b overexpression plants did not alter leaf initiation rate and flowering time during vegetative growth (data not shown). However, AtIpk2b overexpression plants started to produce more secondary inflorescences once they flowered. As summarized in Table I, AtIpk2b overexpression lines produced twice as many total branches than the wild type (37.9 6 2.5 versus 13.4 6 0.6). Axillary shoot formation from the axils of cauline leaves of the primary bolt was also stimulated in AtIpk2b overexpression plants. Secondary branching was obviously affected as well: Overexpression lines produced approximately 3 times the wild-type secondary shoots developed from the axils of leaves of first-order shoots (Table I). To qualify higher order branching, the ratio of the total number of branches divided by the number of first-order branches was calculated (Stirnberg et al., 2002). AtIpk2b overexpression plants showed no significant difference from the wild type in this ratio (0.7 6 0.1 versus 0.6 6 0.1, n 5 10). All these data suggest that the growth of higher order branching was greatly promoted by overexpressing AtIpk2b.
We also obtained two T-DNA insertion lines of AtIpk2b, ipk2b-1, and ipk2b-2 (Fig. 1C). Homozygous T-DNA insertion lines were identified by PCR using AtIpk2b-specific primers 3 K-f and 3 K-r (Xia et al., 2003;Fig. 1C). No PCR product was found in both mutant lines, while a 900-bp fragment was produced from wild-type genomic DNA (Fig. 1D). Reverse transcription (RT)-PCR analysis did not detect AtIpk2b expression in both lines, while the expression of AtIpk2a was not suppressed (Fig. 1E). Interestingly, the T-DNA mutant lines did not show significant branching differences compared to the wild type ( Fig. 2B). However, when ipk2b-1 mutant was transformed with AtIpk2b (see ''Materials and Methods''), the expression of AtIpk2b increased in transgenic mutants, whereas the expression of AtIpk2a was not changed (Fig. 1F). The transgenic mutants produced more branches at maturity (Fig. 2B), similar to the AtIpk2b overexpression lines. These results imply that overexpressing AtIpk2b increases branching.

AtIpk2b Overexpression Lines Displayed Earlier Timing of Axillary Meristem Formation and Increased Rate of Bud Outgrowth
Axillary shoot branching includes two developmental steps: the formation of axillary meristems and bud outgrowth. To investigate if AtIpk2b is preferentially involved in either stage or both, we analyzed the early stages of axillary shoot development in plants grown in short photoperiods. It has been demonstrated that axillary shoot development at each leaf position can be There is no PCR product in homozygous mutants (ipk2b-1 and ipk2b-2), while there is a DNA band with predicted size of 900 bp in the wild type. E, RT-PCR shows the abolished AtIpk2b transcript expression in mutant lines, while the expression of AtIpk2a was not altered. F, RT-PCR experiment shows that transgenic plants (ipk2b-1/AtIpk2b) express a stronger AtIpk2b transcript level than the wild type. classified into three stages: axillary cell divisions, appearance of the axillary meristem, and formation of the first axillary leaf primordium (Stirnberg et al., 2002). In the wild type, the node positions for stage 1 ranged between 15 and 20 nodes (median 18), for stage 2 ranged between 20 and 25 nodes (median 23), and for stage 3 ranged between 25 and 30 nodes (median 28) from the apex (data not shown). However, in AtIpk2b overexpression plants, the ranges of node positions at stages 1, 2, and 3 were closer to the apex (median 13 for stage 1, median 19 for stage 2, and median 22 for stage 3). This result suggests that AtIpk2b obviously affected the timing of axillary meristem formation and increased the rate of bud outgrowth subsequent to meristem initiation, which demonstrates that overexpression of AtIpk2b affects the two stages of axillary shoot branching and results in more branching.

AtIpk2b Expression Pattern Is Similar to That of DR5TGUS
We previously generated an AtIpk2bTGUS fusion reporter gene, and it was shown that the AtIpk2b gene is expressed throughout various Arabidopsis tissues (Xia et al., 2003). Particularly, GUS activity was detected in root tips, root hairs, root vasculature, vascular bundles of young leaves, and emerging lateral root primordia with their associated vascular tissue (Fig.  3A). Close examination of 5-d-old seedlings revealed strong GUS staining at the distal ends of emerging leaf primordia, young stipules, vasculature, and the tips of the cotyledons, but not throughout emerging leaves (Fig. 3, B-D). The staining was also present in leaf axils and the basal part of axillary buds (Fig. 3E). Furthermore, we observed GUS activity in root tip more carefully. GUS staining was observed in epidermis, endodermis, stele, quiescent center, and columella cells, with a stronger expression under the quiescent center that followed the expression pattern of the auxin reporter DR5TGUS (see Supplemental In Arabidopsis, axillary shoot branching is initiated at the shoot apex with the formation of axillary meristems (Hempel and Feldman, 1994). To understand the mechanisms of how AtIpk2b contributes to axillary meristem initiation and branch bud formation, we further investigated its spatial and temporal expression patterns in mature Arabidopsis plants. Initially, AtIpk2bTGUS expression was observed at the leaf axils, before any gross morphological changes became evident (Fig. 3F). Subsequently, strong GUS activity was maintained in leaf axils during the formation of axillary shoots (Fig. 3, G-I). GUS activity was detected in the protuberance and axillary meristems, and extended to the entire branch bud, including the axillary leaf primordia and young leaves. Branching-related gene MONOCULM1 (MOC1) has been reported to play an important role in controlling rice (Oryza sativa) tillering by initiating axillary buds and promoting their outgrowth (Li et al., 2003). Our observation of GUS staining was similar to the expression pattern of MOC1 gene during rice tillering. Consistent with accelerated branching in AtIpk2b overexpression lines, these results suggest that AtIpk2b plays an important role in the initiation of axillary meristems and bud outgrowth, probably through the auxin signaling pathway.
AtIpk2b Is an Auxin-Inducible Gene and Is Important for the Auxin Signaling Pathway The coincidence of AtIpk2bTGUS expression with DR5TGUS distribution implies the connections of AtIpk2b and the auxin signaling pathway. To test this hypothesis, the effects of exogenous IAA on AtIpk2b expression were analyzed using AtIpk2bTGUS transgenic seedlings. The expression of AtIpk2b was enhanced in roots after treatment with 1 mM IAA (Fig. 4A).  Expression of AtIpk2b became stronger when the concentration of exogenous IAA was increased from 1 mM IAA to 10 mM IAA (Fig. 4A). We also examined AtIpk2b expression when treated with 40 mM IAA. Figure 6A shows that AtIpk2b expression increased after 0.5 h of treatment, whereas the expression of AtIpk2a remained virtually unaffected (Fig. 6A). These results indicate that AtIpk2b is an IAA-inducible gene, while AtIpk2a is not.
To test whether AtIpk2b is necessary for normal endogenous auxin signals, we analyzed many developmental aspects that are thought to correlate with auxin signals. Consistent with the hypothesis, the petioles of cotyledons were more elongated in lightgrown AtIpk2b overexpression plants (Fig. 4B). In addition, when grown in dark for 6 d, the hypocotyls of AtIpk2b overexpression seedlings were significantly longer than those of the wild type (OX-9: 11.9 mm; OX-26: 12.2 mm; wild type: 8.9 mm), although the differences in hypocotyl lengths were less evident in light-grown seedlings (Fig. 4C). These results suggest that AtIpk2b overexpression lines altered auxin-related processes, indicating that AtIpk2b is involved in the auxin signaling pathway.
To confirm our hypothesis, we further analyzed the phenotypes of light-and dark-grown seedlings in response to exogenous IAA treatments (Fig. 5). Exogenous IAA has been shown to inhibit the elongation of the primary root (Evans, 1984) and to stimulate lateral root formation (Katsumi et al., 1969). Figure 5A illustrates the effects of IAA application on primary root elongation in wild-type and AtIpk2b overexpression plants. The root lengths of AtIpk2b overexpression seedlings were longer than those of wild type at different IAA concentrations. As IAA concentrations increase, root length is reduced in both wild-type and AtIpk2b overexpression plants (Fig. 5A). Compared with the wild type, whose roots were evidently inhibited by exogenous auxin, the AtIpk2b overexpression plants exhibited much less inhibition (Fig. 5A). At an IAA concentration of 0.1 mM, the root lengths of wild-type seedlings were approximately 71.4% of those of untreated seedlings, whereas AtIpk2b overexpression roots displayed much less inhibition, with root lengths being approximately 94.7% (OX-9) and 90.5% (OX-26) of those of untreated controls. When auxin was supplemented at a concentration of 10 mM IAA, wild-type roots were severely inhibited to approximately 25% of the length of untreated seedlings. By contrast, AtIpk2b overexpression roots showed less inhibition, with root lengths being approximately 36.8% (OX-9) and approximately 28.6% (OX-26) of those of untreated controls, respectively (Fig. 5A). We also examined the lateral root density of wild-type and AtIpk2b overexpression seedlings at different concentrations of IAA (Fig. 5B). AtIpk2b overexpression plants showed more lateral roots and higher relative lateral density than the wild type (Fig. 5B). However, Figure 4. AtIpk2b is induced by IAA and AtIpk2b overexpression lines show auxin-related phenotypes. A, GUS staining shows that expression of AtIpk2b was enhanced in roots after treatment with IAA. Five-dayold seedlings were exposed to different concentrations of IAA for 3 h before staining. B, The petioles of cotyledons (1 week old) are more elongated in AtIpk2b overexpression line (OX-26) than in the wild type. C, Hypocotyl length measurement shows the wild type and AtIpk2b overexpression lines (OX-9 and OX-26) have similar hypocotyl length under light-grown conditions; however, AtIpk2b overexpression lines have much longer hypocotyls than the wild type under dark-grown conditions. Error bars represent SE (n . 12). on media containing increasing concentrations of IAA, the relative lateral density became lower in AtIpk2b overexpression plants than in the wild type (Fig. 5B). All of these results suggest that AtIpk2b inhibits IAA responses in roots to some extents. We also analyzed the outgrowth of cauline lateral inflorescences from excised nodes as described by Chatfield et al. (2000), and found that AtIpk2b overexpression lines also altered exogenous IAA inhibition of excised lateral inflorescence outgrowth (data not shown).

AtIpk2b Regulates the Expression of Auxin-Related Genes
The expression level of some auxin-related genes was also detected in AtIpk2b overexpression line OX-26 and the wild type. RT-PCR experiments showed that the expression of auxin-biosynthesis genes NIT1, TRP, YUCCA, and CYP79B2 was not changed, while the expression of CYP83B1 was decreased in AtIpk2b overexpression plants (Fig. 6B). In addition, the expression of some auxin-transport genes was also detected. Compared to the unchanged expression of PIN1, PIN2, PIN3, and PIN7, the expression of PIN4 was much stimulated in AtIpk2b overexpression plants (Fig. 6B). The change of CYP83B1 and PIN4 demonstrated that auxin biosynthesis and transport were modified in AtIpk2b overexpression plants, consistent with their altered IAA responses (Fig. 5).
Finally, we compared the expression of other auxinrelated genes. Figure 6C shows that the expression of MAX4 and SPS was decreased in overexpression plants, while the expression of AXR1 was not changed.
MAX4 and SPS are required for auxin-mediated bud inhibition and outgrowth (Tantikanjana et al., 2001;Bainbridge et al., 2005). The decreased expression of MAX4 and SPS indicates that bud initiation and outgrowth are improved, which is consistent with the increased branching phenotype in AtIpk2b overexpression lines. Furthermore, decreased expression of MAX4 suggests that AtIpk2b functions upstream of MAX4. AXR1 is one of the auxin-response genes (Stirnberg et al., 1999). AXR1-mediated auxin signaling plays important roles in auxin signal transduction. To illustrate if AtIpk2b was involved in AXR1-mediated auxin signaling, we analyzed the expression level of AtIpk2b in axr1-3 mutant. Figure 6D shows that there was no significant difference between axr1-3 and the wild type (Fig. 6D). Furthermore, Figure 6C shows that the transcriptional level of AXR1 was also not changed in AtIpk2b overexpression line (Fig. 6C). These results indicate that there is no correlation between AtIpk2b and AXR1 in transcriptional level.

DISCUSSION
We previously reported the isolation and identification of Arabidopsis AtIpk2b, a homolog of animal IP3K genes (Xia et al., 2003). However, little is known about its physiological functions in higher plants. The data presented here provide evidence for a role of AtIpk2b in axillary shoot branching through the auxin signaling pathway.
Auxin-related mutants fall into three classes on the basis of whether they affect meristem initiation, outgrowth (e.g. max), or both (e.g. sps/bushy, moc1; Tantikanjana et al., 2001;Li et al., 2003;Ward and Leyser, 2004). Our results demonstrate that AtIpk2b affects both axillary meristem initiation and outgrowth. Overexpressing AtIpk2b in wild-type and mutant plants yields more axillary shoot branches (Fig. 2). We found that more shoot buds are generated at axils of cauline leaves and rosette leaves in AtIpk2b overexpression lines (Table I). In addition, higher order branches from nodes on the inflorescence are also promoted, while the wild type lacks axillary meristem at these higher order nodes (data not shown). The ranges of node positions at which axillary cell divisions and appearance of the axillary meristem were seen closer to the apex in AtIpk2b overexpression lines than in the wild type. These observations indicate that enhanced axillary shoot formation in AtIpk2b overexpression plants is determined to a large extent by earlier formation of axillary meristems. This conclusion is further supported by AtIpk2bTGUS studies (Fig. 3). High levels of GUS activity are detected in axils of leaf primordia before any morphological alteration and during axillary bud formation. This expression pattern is similar to that of SPS, which negatively regulates axillary meristem formation and growth (Tantikanjana et al., 2001). RT-PCR analysis showed that the expression of SPS was decreased in Figure 6. RT-PCR analysis of auxin-related genes. One-week-old seedlings were used for all experiments. A, Wild-type seedlings were treated with 40 mM IAA for 0, 0.5, 1, and 2 h, respectively. AtIpk2b expression was induced by IAA after 0.5 h, while AtIpk2a expression was not induced by IAA. B, Compared with the wild type, CYP83B1 expression was decreased and PIN4 expression was increased in AtIpk2boverexpressing plant (OX-26), whereas expression of NIT1, TRP, YUCCA, CYP79B2, PIN1, PIN2, PIN3, and PIN7 was not changed. C, Transcript levels of MAX4 and SPS decreased in AtIpk2b-overexpressing plant (OX-26), while levels of AXR1 were not changed. D, Transcript levels of AtIpk2b were not changed in the wild type (Columbia), axr1, and max4.
AtIpk2b overexpression plants. These results support the involvement of AtIpk2b in axillary meristem initiation. The rate of bud outgrowth was increased in AtIpk2b overexpression plants compared to the wild type. GUS staining was also observed after axillary meristem initiation (Fig. 3, H and I). High GUS activity was also detected in entire axillary buds during their growth and extended to young leaves. This spatial and temporal expression patterns are similar to those of the MOC1 gene in rice (Li et al., 2003). MOC1 is an important controller of rice tillering and its mutant shows defects in the formation of tiller buds and bud outgrowth (Li et al., 2003). Another critical gene involved in axillary outgrowth is MAX4, and max4 mutants exhibit more branching (Sorefan et al., 2003). Interestingly, we observed a decrease in MAX4 transcription level in AtIpk2b overexpression plants. Although the detailed pathways are not known, our data suggest that MAX4 is regulated by AtIpk2b during bud outgrowth. Further analysis of branching shows that the proportion of vegetative nodes that are potential sites to produce a new branch was higher in AtIpk2b overexpression plants than in the wild type ( Table I).
All of these data demonstrate that AtIpk2b also functions in axillary bud outgrowth.
Apical dominance is another broadly known mechanism that controls axillary bud growth (Thimann and Skoog, 1934;Thimann, 1937). In Arabidopsis, the growth of axillary buds is inhibited by auxin derived from the apical bud. However, auxin does not enter lateral buds and a second signal is involved in the repression of lateral shoot outgrowth. Auxin distribution analysis shows that the distal ends of leaf primordia and stipules of 4-d-old seedlings have high levels of IAA (Avsian-Kretchmer et al., 2002). IAA also accumulates in the leaf tip. GUS activity presents the coincidence between the expression patterns of AtIpk2b and IAA distribution. Moreover, AtIpk2b gene expresses throughout development with the highest level in root, inflorescences, and flowers, suggesting that AtIpk2b has a broad role in many auxin-regulated processes (Xia et al., 2003). At present, although we do not know how AtIpk2b affects these processes, several observations raise the possibility that AtIpk2b participates in the auxin signaling pathway. First, AtIpk2b can be induced by exogenous IAA (Figs. 4A and 6A). Second, AtIpk2b overexpression lines display altered IAA responses. Root elongation, lateral root formation, and excised lateral inflorescence outgrowth under exogenous IAA are affected in AtIpk2b overexpression plants (Fig. 5). Third, AtIpk2b overexpression lines display decreased expression of CYP83B1 and enhanced expression of PIN4. CYP83B1 decreases the level of IAA by distributing indole-3acetaldoxime to the glucosinolate pathway. It has been shown that CYP83B1-deficient plants overaccumulate IAA, indole-3-acetaldehyde, and IAA-Asp, whereas CYP83B1 overexpression results in high indole glucosinolates and low IAA (Bartel et al., 2001). PIN4dependent auxin transport actively maintains the auxin gradient in roots, stabilizing it through a feedback loop (Friml, 2003). The pin4 mutant affects endogenous auxin gradients and root pattern. These results indicate that overexpression AtIpk2b affects auxin distribution and accumulation. Furthermore, the expression level of MAX4 and SPS was decreased in overexpression lines (Fig. 6). These two genes are required for auxin-mediated shoot branching (Tantikanjana et al., 2001;Sorefan et al., 2003;Bainbridge et al., 2005). The decreased expression of MAX4 and SPS indicates an important role for AtIpk2b in axillary shoot branching by negatively regulating MAX4 and SPS, although it is also possible that AtIpk2b additionally affects other regulators of axillary shoot branching. Our data indicate that AtIpk2b is an early responsive gene in regulating branching by the auxin signaling pathway.
Although AtIpk2b is involved in the auxin signaling pathway, its exact functions in branching systems are not clear. Classical branching hypothesis states that auxin regulates shoot branching in conjunction with secondary messengers, and the candidates for this signal include cytokinin and SMS (Sachs and Thimann, 1967;Li et al., 1995;Foo et al., 2005;Beveridge, 2006). The SPS gene acts as a modulator of cytokinin metabolism and MAX4 is required for the production of SMS (Tantikanjana et al., 2001;Beveridge, 2006). The decreased expression of MAX4 and SPS in AtIpk2b overexpression plants suggested that AtIpk2b is involved in the negative regulation of these two signals in branching systems. Auxin and the MAX-dependent hormone interact to inhibit branching. However, the MAX-dependent hormone is a novel regulator of auxin transport that regulates bud outgrowth independent of AXR1-mediated auxin signaling (Bennett et al., 2006). AtIpk2b is involved in auxin signaling and affects the expression of CYP83B1 and PIN4, which are involved in auxin biosynthesis and auxin transport, respectively. Therefore, AtIpk2b should also be independent of AXR1-mediated auxin signaling. RT-PCR showed that the expression of AXR1 was indeed not changed in AtIpk2b overexpression plants. Moreover, we also detected unchanged expression of AtIpk2b in the axr1-3 mutant (Fig. 6). These data indicate that AtIpk2b may be independent of the AXR1-mediated auxin signaling pathway, at least at the transcriptional level.
Two IP3K isoforms (AtIpk2a and AtIpk2b) exist in Arabidopsis. They exhibit high homology in sequences (84% similarity and 73% identity). Both of them have inositol polyphosphate 6-/3-kinase activities on IP 3 and are involved in regulating Arg metabolism in yeast (Stevenson-Paulik et al., 2002;Xia et al., 2003). They also show similar expression patterns in seedling, leaf, root, and floral organs (Xia et al., 2003;Xu et al., 2005). In addition, ipk2b mutants and ipk2a antisense lines display no significant branching differences compared to the wild type (Z.-B. Zhang and H.-J. Xia, unpublished data; Q. Xu, unpublished data). All the evidence suggests that AtIpk2a and AtIpk2b may have potential redundant (or partially redundant) genetic function. This hypothesis is supported by DR5TGUS studies. We analyzed DR5TGUS activity in ipk2b mutants and mutant seedlings showed the similar GUS expression pattern to the wild type (see Supplemental Fig. S2). York's group (Stevenson-Paulik et al., 2005) previously pointed out the redundancy between AtIpk2a and AtIpk2b. AtIpk2a was able to fully compensate the phytate production in ipk2b mutant tissues (Stevenson-Paulik et al., 2005). However, AtIpk2a and AtIpk2b display differential roles in expression patterns in pollen tubes (Xu et al., 2005) and in phytate synthesis (Stevenson-Paulik et al., 2005). Our experiment shows the different expression between AtIpk2a and AtIpk2b. AtIpk2b was induced by IAA, but AtIpk2a was not (Fig. 6A). We also detected the different responses between the two isoforms under other hormone and stress treatments (Z. Chen, Y. Li, and H.-J. Xia, unpublished data). Thus, AtIpk2a and AtIpk2b have not only overlapping but also unique functions.
In this report, we studied the physiological functions of AtIpk2b in axillary shoot branching. It is known that auxin may influence shoot branching via multiple pathways (Bennett et al., 2006). It seems likely that AtIpk2b regulates axillary shoot branching in Arabidopsis, at least in part, by integrating the auxin signaling pathway. More detailed analysis will lead to further understanding of the molecular mechanism of signaling pathway integration.

Plant Material and Growth Conditions
Arabidopsis (Arabidopsis thaliana) ecotypes C24 and Columbia were used in this study. SALK_025091 (ipk2b-1) and SALK_104995 (ipk2b-2) mutant lines were obtained from the Arabidopsis Biological Resource Center (Alonso et al., 2003). Vector constructs and the production of AtIpk2b transgenic plants are described below.
Seeds were surface sterilized with 70% ethanol for 5 min, and then washed three times with sterile water and plated on solidified MS medium (Murashige and Skoog, 1962). After 3 d of cold treatment at 4°C, seeds were germinated at 22°C with a 16-h-light/8-h-dark cycle at 60% humidity.

Vector Constructions and Arabidopsis Transformation
To create an AtIpk2b (gene locus no. At5g61760) overexpression construct, the AtIpk2b cDNA was cut from plasmid pmIP3K with XbaI-XhoI (Xia et al., 2003) and inserted via XbaI-SalI sites into pBinAR-HPT vector behind the cauliflower mosaic virus 35S promoter. The resulting plasmid (pBin-IP3K) was first introduced into Agrobacterium tumefaciens by electroporation and then transformed Arabidopsis ecotype C24 by vacuum infiltration (modified from Grant et al., 1995). Transgenic plants were selected on MS medium containing 20 mg/L hygromycin (Calbiochem).

RNA Extraction and Northern-Blot Analysis
Total RNA from leaves of 2-week-old overexpression plants and the wild type was prepared according to the protocol of Logemann et al. (1987). RNA (35 mg) was separated electrophoretically on denaturing 15% (v/v) formaldehyde and 1.5% (w/v) agarose gels. Northern hybridizations were performed as described by Xia et al. (2003).

Protein Extraction and Western-Blot Analysis
Leaves of 2-week-old plants were ground in liquid nitrogen and suspended in extraction buffer containing 0.1 M Tris-HCl, pH 8.0, 18% (v/v) glycerol, 10 mM MgCl 2 , 14 mM b-mercaptoethanol, 1 mg/mL pepstatin, 1 mg/mL leupeptin, and 1 mg/mL aprotinin. After a brief vortexing, the suspension was centrifuged at 14,000 rpm for 10 min and the supernatant collected. Proteins were separated on 10% SDS-polyacrylamide gels and transferred to a nitrocellulose membrane. Western blot was carried out essentially as described by Xia et al. (2003) using an antibody against MBP-AtIpk2b fusion protein at a 1:1,000 dilution.

Identification of T-DNA Insertion Mutants and Arabidopsis Transformation
Seeds of ipk2b-1 and ipk2b-2 were first screened on the MS plates containing 50 mg/mL kanamycin (Duchefa Biochemie). The homozygous T-DNA insertion mutants were identified by a PCR genotyping assay with the specific primers 3 K-f and 3 K-r (Xia et al., 2003) for each individual plant. RT-PCR was used to analyze the expression of AtIpk2b, and the actin gene served as a control in mutant and the wild type.
To overexpress AtIpk2b in ipk2b-1 mutant, the coding region of AtIpk2b was amplified by PCR using PfuTurbo DNA polymerase (Stratagene) with primers IP3K-f (5#-CCGCTCGAGAAGATGCTCAAGGTCCCTGAACACC-AAG-3#) and IP3K-r (5#-ATGGGCCCGCGCCCGTTCTCAAGTAGGGAAG-TATCG-3#) from wild-type Columbia genomic DNA. The primers added XhoI (5# end) and ApaI (3# end) restriction sites to the amplified fragment. The PCR product was inserted via XhoI-ApaI sites into vector pGreenLC201 containing a 35S promoter followed by a 63 haemagglutinin tag. The resulting plasmid pGLC201-AtIpK2b was first introduced into A. tumefaciens by electroporation and then was transferred to Arabidopsis ipk2b-1 mutant by vacuum infiltration. Transgenic plants were selected using Basta solution (Invitrogen).

Root Length and Lateral Root Density Measurement
Seeds (n . 12) were germinated vertically in light for 11 d and in dark for 5 or 6 d, respectively, and hypocotyl length of seedlings was measured. As for root length measurement, 4-d-old seedlings (n . 12) were transferred from unsupplemented plates to vertical plates containing 0, 0.1, 1, and 10 mM IAA, and primary root length was measured and calculated after 3 d of growth. Experiments were repeated three times. As for lateral root density assay, 4-d-old seedlings (n . 12) were transferred from IAA-free plates to vertical plates containing different concentrations of IAA (0, 0.01, 0.1, and 1 mM). After additional 3 d, the number of emerged lateral roots was counted and divided by the length of the primary root and then compared with the seedlings grown on unsupplemented medium.

GUS Staining Assay
The AtIpk2bTGUS transgenic plants were described by Xia et al. (2003). Samples from 5-d-old seedlings treated with 0, 1, 10, and 100 mM IAA were fixed under vacuum for 10 min in 50 mM sodium phosphate buffer, pH 7.0, 0.005% (v/v) Tween 80, and 0.3% (v/v) formaldehyde. Then, tissues were washed three times in 50 mM sodium phosphate and stained for GUS activity in 50 mM sodium phosphate, 10 mM EDTA, 0.01% (v/v) Tween 80, and 0.5 mg mL 21 5-bromo-4-chloro-3-indolyl-b-D-glucuronide. After staining overnight, the samples were washed three times with 50 mM sodium phosphate and fixed for 30 min at room temperature in 20% (v/v) ethanol, 5% (v/v) acetic acid, and 5% (v/v) formaldehyde. Chlorophyll-containing tissues were cleared in a series of ethanol:water mixtures up to 80% (v/v) ethanol, in which samples were stored. Microscopic analysis was performed using an Olympus SZX-ILLB200 microscope with bright-field optics. All images were obtained with the same modifications and intensity parameters. Images were photographed using a CCD camera and processed using Adobe Photoshop (Adobe Systems).