Molecular Plant Advance Access originally published online on February 8, 2008
Molecular Plant 2008 1(2):295-307; doi:10.1093/mp/ssm030
| ||||||||||||||||||||||||||||||||||||||||||||||||||||
Selective Deactivation of Gibberellins below the Shoot Apex is Critical to Flowering but Not to Stem Elongation of Lolium
a CSIRO, Plant Industry, Canberra, ACT 2600, Australia
b Research School of Chemistry, Australian National University, Canberra, ACT 0200, Australia
c Department of Genetics and Biotechnology, University of Aarhus, 4200 Slagelse, Denmark
d Present Address: ENSIS, PO Box E, 4008 Kingston, ACT 2604, Australia
1 To whom correspondence should be addressed. E-mail rod.king{at}csiro.au, fax (61)262465000.
 |
Abstract |
|---|
 TOP Abstract INTRODUCTIONRESULTSDISCUSSIONMETHODS |
|---|
Gibberellins (GAs) cause dramatic increases in plant height and a genetic block in the synthesis of GA1 explains the dwarfing of Mendel's pea. For flowering, it is GA5 which is important in the long-day (LD) responsive grass, Lolium. As we show here, GA1 and GA4 are restricted in their effectiveness for flowering because they are deactivated by C-2 hydroxylation below the shoot apex. In contrast, GA5 is effective because of its structural protection at C-2. Excised vegetative shoot tips rapidly degrade [14C]GA1, [14C]GA4, and [14C]GA20 (>80% in 6 h), but not [14C]GA5. Coincidentally, genes encoding two 2β-oxidases and a putative 16–17-epoxidase were most expressed just below the shoot apex (<3 mm). Further down the immature stem (>4 mm), expression of these GA deactivation genes is reduced, so allowing GA1 and GA4 to promote sub-apical stem elongation. Subsequently, GA degradation declines in florally induced shoot tips and these GAs can become active for floral development. Structural changes which stabilize GA4 confirm the link between florigenicity and restricted GA 2β-hydroxylation (e.g. 2
-hydroxylation and C-2 di-methylation). Additionally, a 2-oxidase inhibitor (Trinexapac Ethyl) enhanced the activity of applied GA4, as did limiting C-16,17 epoxidation in 16,17-dihydro GAs or after C-13 hydroxylation. Overall, deactivation of GA1 and GA4 just below the shoot apex effectively restricts their florigenicity in Lolium and, conversely, with GA5, C-2 and C-13 protection against deactivation allows its high florigenicity. Speculatively, such differences in GA access to the shoot apex of grasses may be important for separating floral induction from inflorescence emergence and thus could influence their survival under conditions of herbivore predation. |
INTRODUCTION |
|---|
|
TOPAbstractINTRODUCTIONRESULTSDISCUSSIONMETHODS |
|---|
There are clear associations between applied and endogenous gibberellins (GAs) and flowering in dicotyledonous species including Arabidopsis (Wilson et al., 1992; Xu et al., 1997; Blázquez and Weigel, 2000; Eriksson et al., 2006), and flowering in monocotyledonous species including Lolium spp. (Evans, 1964; Pharis and King, 1985; King and Evans 2003; King et al., 2001, 2006; MacMillan et al., 2005). On transfer to a long day (LD), the levels of bioactive GAs (variously GA1, GA4, GA5, and GA6) increase rapidly in the leaf, petiole and shoot apex of both monocots and dicots (Metzger and Zeevaart, 1980; Talon and Zeevaart, 1990; Xu et al., 1997; Gocal et al., 1999, 2001; King et al., 2001, 2003, 2006; MacMillan et al., 2005).
Increase in endogenous levels of bioactive GAs is often associated with increased expression of 20-oxidase GA biosynthetic genes (Xu et al., 1997; Hisamatsu et al., 2005; MacMillan et al., 2005; King et al., 2006). However, a GA increase could also be brought about by a reduction in GA degradation. One group of such enzymes—the 2-oxidases—inactivate bioactive GA1 and GA4 by hydroxylation of Carbon-2 (C-2; see Figure 1)—an action demonstrated in a large number of species, including Phaseolus coccineus (Thomas et al., 1999), Arabidopsis (Thomas et al., 1999), Pisum sativum (Lester et al., 1999; Martin et al., 1999), and spinach (Lee and Zeevaart, 2002). As expected, GA content is elevated in the pea 2-oxidase mutant sln and shoot growth is enhanced (see ref. in Lester et al., 1999). The converse—reduced GA levels and dwarfing—result from 2-oxidase overexpression (Sakamoto et al., 2001; Busov et al., 2003; Schomburg et al., 2003; Lee and Zeevaart, 2005; Curtis et al., 2005; Kloosterman et al., 2007). A second pathway for GA deactivation involves epoxidation at C-16,17 (see Figure 1) and, in rice, overexpression of a GA 16,17-epoxidase caused dwarfing and a mutant (eui) showed enhanced stem elongation (Zhu et al., 2006).
|
In our initial studies with Lolium temulentum leaves, we found no decrease in 2-oxidase transcript when their GA content increased at the time of LD floral induction (King et al., 2006). However, deactivation of GAs could be important during their transport to the shoot apex—a suggestion supported by findings with rice, in which a 2-oxidase was found to express highly in sub-apical vascular bundles of vegetative plants and later to become undetectable when the inflorescence developed (Sakamoto et al., 2001). As an extension of this suggestion of tissue-localized effects, GA deactivation could also be selective, as the various enzymes show different substrate specificities (Thomas and Hedden, 2006).
In support of a model in which GAs are selectively excluded from the apex, GA1 and GA4 are absent from the vegetative shoot apex of L. temulentum but are present in leaves, while GA5 is present in both tissues and increases significantly in LD (Gocal et al., 1999; King et al., 2001, 2003, 2006). The structural differences between these three GAs, summarized in the GA biosynthetic pathway in Figure 1, highlight the potential for GA5 to resist 2β-oxidation because of its C-2–3 double bond, while its 13-OH would tend to inhibit GA epoxidation at C-16,17 (see Zhu et al., 2006). Conversely, GA1 and GA4 would be inactivated by 2-oxidases (see Thomas and Hedden, 2006) and GA4 could also be inactivated by 16,17-epoxidases (Zhu et al., 2006).
Here, to determine the possible role in flowering of both 2-oxidases and 16,17-epoxidases, we have examined their mRNA expression patterns near the shoot apex of L. temulentum. Enzyme activity has also been assessed based on metabolism of [14C]-GAs supplied to isolated shoot tips. Lastly, we have confirmed the linkage between GA deactivation and flowering by examining the response to GAs structurally altered to give protection against GA deactivation at C-2, C-13, and C-16,17, and/or by applying Trinexapac Ethyl to inhibit 2-oxidases (see Rademacher, 2000).
|
RESULTS |
|---|
|
TOPAbstractINTRODUCTIONRESULTSDISCUSSIONMETHODS |
|---|
Differences in GA Structure which Determine their Bioactivity
Previously, we suggested that effectiveness of GAs might reflect differences in their stability (Evans et al., 1990; King et al., 2001; King and Evans, 2003). As an example, GA1 and GA5 are likely to be more stable than GA4 and they are more florigenic when applied once to the leaf of L. temulentum (Figure 2A and 2C), the untreated plants remaining vegetative in SD. The response to GA5 best matches that to a LD because it causes flowering but with limited stem elongation, whereas GA1 causes flowering but with excessive stem elongation (Figure 2D and see review in King and Evans, 2003).
|
GA5 with its C-2,3 double bond (Figure 1) should be structurally protected against 2-oxidase catalysed hydroxylation at C-2. Likewise, 2-oxidase protection by the addition of C-2 methyl groups to GA4 may account for the huge increase in florigenicity of 2,2-dimethyl GA4 compared with GA4 (Figure 2A and Evans et al., 1990, 1994).
Interestingly, for stem elongation of the same plants, the ranking of GAs changes. Although superior for flowering, GA5 is least effective for stem elongation (Figure 2D). By implication, then, GA 2-oxidation is relatively less important in stem tissue—a conclusion also supported by the smaller differential seen in stem elongation induced by 2,2-dimethyl GA4 relative to GA4 (2,2-dimethyl GA4 was 
100-fold more effective than GA4 for stem elongation but 5 000–10 000-fold more effective for flowering; Figure 1A vs 1B and c.f. Evans et al., 1994). Thus, not only is GA deactivation detrimental for flowering, but the structurally based differences in response to applied GA for stem elongation and flowering suggest that deactivation might be greater nearer the shoot apex. For this reason, we examined gene expression patterns in the shoot apex and, for comparative purposes, further down the shoot.
Localization of 2-oxidases and a Putative 16,17-epoxidase Just Below the L. temulentum Shoot Apex
To define the pattern of gene expression near the apex, we harvested 0.8-mm stem segments from all stem tissue in the top 7 mm of stem from the tip, as shown schematically in Figure 3. Harvests were made in the middle of the day for plants held continuously in SD with or without GA1 application 3 d earlier and at the same time (3 d later) for plants exposed to a single LD. All plants were 8 weeks old at the time of harvest and the very lowest segments were discarded because they were not always present and encroached on the root zone (Figure 3). Further experiments confirmed the findings of this study.
|
LtGA2ox1 expression was greatest within 2–3 mm of the apex, but, further down the stem, it was less than 50% of the maximum (Figure 3A). A similar but much magnified pattern of LtGA2ox1 expression in the lower stem is evident for shoots harvested 3 d after GA1 was applied to the leaf or the leaf was exposed to a LD (Figure 3A). In these two treatments (LD; and SD + GA1), expression of LtGA2ox1 was lowest in the apical segment and indistinguishable from the low expression in the untreated SD apical segment to which all expression levels were normalized.
The high stem and sub-apical expression of LtGA2ox1 (Figure 3A) is indicative of deactivation of GA20—the immediate precursor of bioactive GA1 and GA5. Bioactive GA1 and GA4 themselves would be deactivated by LtGA2ox5 (see Figure 1). Based on the expression profiles of these two genes, their greatest action would be at or just below the shoot apex, their expression dropping to less than 20–30% of the maximum further down the young stem tissue (Figure 3A and 3C). In addition, transport of active GA4 into the shoot apex would be affected by its inactivation by a putative LtGA16,17epoxidase (Figure 1). Later, we present evidence for endogenous epoxidation of GA4 but the gene assayed for expression (Figure 3B) is no more than a putative epoxidase. The L. temulentum gene shows 97% amino acid identity with the L. rigidum cytochrome P450 gene, which, in turn, shows 41% identity with the rice epoxidase. We were unsuccessful with the yeast functional epoxidase assay used by Zhu et al. (2006).
A characteristic of some plant 2-oxidases is GA up-regulation of their expression (Thomas et al., 1999; and see review in Thomas and Hedden, 2006). Such feed-forward regulation of LtGA2ox1 by GA1 is shown in Figure 3A. The increase in endogenous GAs within hours of the LD exposure (King et al., 2001, 2006) would also account for LD up-regulation of LtGA2ox1 (Figure 3A). Interestingly, LtGA2ox5 expression was reduced dramatically for harvests 3 d after the LD, but LD had no effect on expression of the putative LtGA16,17epoxidase (Figure 3B and 3C).
In rice, one 2-oxidase expresses in vascular bundles just below the shoot apex (Sakamoto et al., 2001) and, in Arabidopsis, a 2-oxidase reporter gene construct expresses highly but diffusely just below the shoot apex (Jasinski et al., 2005). Similarly, for L. temulentum, LtGA2ox1 expression was localized to vascular parenchyma in stem cross-sections taken 1 mm below the shoot apex (Figure 4A). At a 4.5 greater magnification, a repeat hybridization with an adjacent section showed the same pattern (Figure 4B). At an equal concentration of riboprobe, the control (sense probe) barely hybridized (Figure 4C).
|
The increased LtGA2ox1 expression with GA1 or LD treatment (Figure 3) is associated with the onset of flowering of L. temulentum. To further examine this relationship between GA, flowering, and 2-oxidases, we applied various doses of GA5 and GA1 to plants in SD. These two GAs differ in effectiveness for flowering and stem elongation. GA5 is more florally effective and GA1 is more effective for stem elongation (Figure 5B and see also Figure 2C and 2D), as we reported previously (Evans et al., 1990; King and Evans, 2003). Of these two GAs, GA5 was far less effective for inducing LtGA2ox1 expression, especially at doses of 1–5 µg per plant (Figure 5A).
|
) (25 µg per plant) on Expression of LtGA2ox1 in the Top 3.5 mm of the Shoot Tip of L. temulentum.In a repeat study, effectiveness of GAs for inducing expression of LtGA2ox1 was GA1 > GA4 > GA5 at a dose of 5 µg per plant (Table 1) but effectiveness is reversed for flowering (i.e. GA5 > GA1
GA4; e.g. Figure 2 and see King and Evans, 2003). For stem elongation, the ranking of GAs is the same as for LtGA2ox1 expression (data not shown). Although transcript of LtGA2ox1 does not increase when GA5 causes flowering (Figure 5), it increases in association with flowering induced by GA1 (Figure 3). As we discuss later, these contradictory results could reflect differences in both GA deactivation and in receptor binding affinity.
|
GA Degradation Studies
As an independent test of the relationship between flowering and GA deactivation near the shoot apex, we supplied [14C]GAs to excised vegetative shoot apices which included up to 2 mm of apex/stem base (cf. diagram in Figure 3). Based on tissue extraction and radiocounting of HPLC fractions, within 6 h, [14C]GA20 was converted almost completely to its 2β-hydroxylated product [14C]GA29 (Figure 6). Previously, we had confirmed this product as [14C]GA29 by GCMS after HPLC and also showed similar rapid and complete 2β-oxidation of GA1 to GA8 (Junttila et al., 1997).
|
Given their different substrate specificities in functional assays (see Materials and Methods), LtGA2ox1 would regulate GA20 deactivation in the shoot tip and LtGA2ox5 would metabolize GA1 and GA4. Previously, we showed that shoot tips rapidly converted all applied GA1 to its 2-hydroxlated product, GA8 (Junttila et al 1997). Here, when [17-14C]GA4 was applied to shoot tips, the most abundant metabolite chromatographed with GA34, the expected 2-hydroxylated metabolite of GA4 (Figure 6). A more polar peak (fraction 20–21) chromatographed where we would expect to find [17-14C]GA34 catabolite (Pearce et al., 2002). The most polar peak chromatographed with a retention time on C-18 reverse phase HPLC, which is characteristic for a 16,17-epoxide (S. Yamaguchi, personal communication). Our GCMS examination of this peak showed the presence of a product with the same GC retention time as GA4 16,17-diol, with a strong ion at m/z 493 (M+-17-CH2OTMS; cf. Zhu et al., 2006, for GCMS information) and which, because of the fragmentation pattern during MS, is characteristic for both [17-14C] labelled and endogenous [12C] GA4 substrate. The 16,17-diol analyzed here is formed rapidly when the epoxide is exposed to low concentrations of acetic acid, as employed during HPLC (see evidence in Zhu et al., 2006). Thus, we infer from these data that the 16,17-epoxide is one GA4 metabolite and that both 2-oxidases and 16,17-epoxidases contribute to GA4 deactivation in and just below the shoot tip.
Despite the rapid metabolism of [14C]GA20 (Figure 6) to its 3β-hydroxylated product GA29, in the same experiment, [14C]GA5 was hardly catabolised (Figure 6). The small amount of a GA6-like product in fraction 18 fits with our previous in-vitro evidence that GA5 is converted by a 3-oxidase to GA6 (King et al., 2004). As GA5 and GA20 only differ structurally in their C-2,3 bond (see Figure 1), it is presumably the double bond of GA5 which protects it from 2β-oxidation.
To examine the effects of LD on GA deactivation, we supplied [14C]GA20 or [14C]GA4 for 6 h to batches of shoot tips isolated in SD or at various times after a single florally inductive LD. We report deactivation as percent remaining substrate, since multiple metabolites may form in this assay (see Figure 6). Initially, metabolism was rapid in shoot tips from vegetative plants or for plants exposed to a LD. After 6 h, only 15–30% of extractable counts remained as non-metabolized [14C]GA20 or [14C]GA4 (Table 2). Two or more days after the LD, the rate of metabolism had dropped by 30%, which matches the relatively low-level LtGA2ox5 expression seen 3 d after the LD (Figure 3). The late decrease in [14C]GA4 metabolism (Table 2) may reflect its potential for 16,17-epoxidation as well as 2-oxidation, but the spread of metabolites detected on HPLC was too great to allow us to make this conclusion with any confidence.
|
Endogenous GA1 and GA4 become detectable in the L. temulentum shoot apex several days after a single LD (King et al., 2001), which coincides with decreased metabolism of GA4 and GA20 by this time (Table 2). Additional assays with [14C]GA4 (but in duplicate, not triplicate) confirmed these findings and its metabolism was further reduced following repeated LD exposures (data not shown). Such repetition of the LD causes both more rapid flowering and an earlier and more dramatic increase (by up to 10-fold) in shoot apex GA1 and GA4 content (King et al., 2001).
Studies with an Inhibitor of GA Metabolism and with GA Structural Variants
The evidence above of differences in metabolism of GAs (GA4 and GA20 > > > GA5; Figure 6) and of distinctive effects of GA structure on florigenic activity (Figure 2) highlights the possibility of using differences in GA structure to assess the relationship between GA deactivation and flowering. Therefore, we examined effects of further structural changes to GAs along with application of Trinexapac Ethyl (TNE)—a 2-oxidase inhibitor (Rademacher, 2000).
TNE and GA4 caused flowering when applied together to the leaf of Lolium perenne in SD (Figure 7). Given individually, neither TNE nor GA4 had any effect on flowering. Thus, their effectiveness, when applied together, indicates specificity of TNE as an inhibitor of 2-oxidases and, potentially, with no side effects, of TNE. Consistent with this conclusion, there was no additional flowering in response to TNE applied with 2,2-dimethyl GA4—a finding we expected because this synthetic GA is already protected against 2-oxidation (Figure 7). However, this latter evidence is not compelling because the dose of 2,2-dimethyl GA4 may have been saturating and this may have hidden any effect of TNE.
|
In similar studies with L. temulentum, TNE applied simultaneously with GA1 or GA4 increased their florigenicity (Figure 8). Untreated plants were vegetative in SD and TNE alone caused no flowering. Only plants with a score of 2 and above are considered floral (cf. McDaniel et al., 1991) so GA4 was only florigenic when both 2-oxidation was restricted by co-application of TNE and its epoxidation was hindered, as in 16,17-dihydro GA4, or by adding a 13-OH group, as in GA1 (Figure 8). Protection against 16,17 epoxidation by C-13 hydroxylation was clear in the study of Zhu et al. (2006) and this protection may explain the greater florigenicity of GA1 relative to GA4, the latter GA lacking a C-13 hydroxyl (Figure 1). Structural change at C-16,17 (as in 16,17-dihydro GAs) had less of an effect on florigenicity than the presence of a C-13 hydroxyl.
|
To help localize the site of action of 2-oxidases, TNE was injected in water into the air space above the shoot apex of L. temulentum and, at the same time, GA4 was applied in ethanol to the leaf. These plants were also exposed to a single LD to allow better visualization of the weak response to GA4. When given alone, neither applied chemical had any effect on flowering relative to an untreated control (mm floral shoot apex length: LD, 2.1 ± 0.08; LD plus GA4, 2.23 ± 0.07; LD plus TNE, 2.17 ± 0.07; n = 14). However, flowering was promoted following co-application of GA4 to the leaf and TNE to the apex (mm floral shoot apex length: 2.68 ± 0.11). This result is consistent with TNE acting at the shoot tip to restrict 2β-oxidation and inactivation of GA4 during its transport from the leaf to the shoot apex.
The importance of protection against epoxidation and 2-oxidation was further supported by our application of structural variants of GA1 and GA4. Presence or absence of a C-13 hydroxyl was combined with an - or β-hydroxyl at C-2. The presence of the C-13 hydroxyl gave consistently better flowering in paired comparisons across three GAs (Figure 9A). Also, the addition of a 2β hydroxyl depressed GA activity, (GA8 < GA1 or GA34 
GA4) but more so for stem elongation than for flowering (Figure 9). Most striking is the greater flowering in response to 2-hydroxylation, as in GA56 and GA47 (Figure 9A), but their florigenicity is no better than for GA5 (data not shown).
|
Our observations raise questions about the extent that 2β-oxidation inactivates a GA (reviewed in Thomas and Hedden, 2006). For stem elongation of L. temulentum, the 2β-oxidation products of GA1 and GA4 (i.e. GA8 and GA34, respectively), were not completely inactive (Figure 9). It is also not clear why GA56 had such a weak effect on stem elongation (Figure 9B), but the same result was found in a repeat experiment. Nevertheless, when taken together, these results show that protection against deactivation at C-2 by a 2
hydroxyl in combination with protection at C-13 creates a more florigenic GA, but also that, for stem elongation, such protection is less important. |
DISCUSSION |
|---|
|
TOPAbstractINTRODUCTIONRESULTSDISCUSSIONMETHODS |
|---|
GA synthesis, deactivation, pool sizes, receptor affinities, and specificities as substrate for deactivation enzymes will all determine their effectiveness for flowering. Here, we show that GA deactivation is important in GA-regulated flowering. The two 2β-hydroxylases we examined and a putative 16,17 epoxidase would all deactivate bioactive GAs and they express most highly in or just below the shoot apex and probably in the vascular tissue, where GAs would be unloaded in their transport from the leaf to the shoot apex. We show how differences in substrate preference of these enzymes make various GAs more or less stable and, so, more or less florigenic. Ready deactivation of GA1 and GA4 in isolated shoot tips but not of structurally protected GA5 shows why GA5 is an effective endogenous LD floral signal in the grass Lolium. Later, during floral development, we note that GA deactivation is reduced and then GA1 and GA4 can take on a regulatory role in flowering.
Localized Enzyme Expression and Substrate Preferences
Of the enzymes examined here, LtGA2ox5 with its high activity in SD would rapidly degrade GA1 and GA4 and so interfere with their transport into the vegetative shoot apex. The supply of GA20 would be affected by LtGA2ox1 and more so in LD than in SD. Additionally, the 16,17-epoxidases would inactivate GA4 and, if our putative Lolium gene is a true epoxidase, its expression indicates activity in all day lengths. Further enhancing selectivity for deactivation, GA1 and GA4 induce 2-oxidase transcription in shoot tips (cf. Figures 2 and 5, Table 1 and evidence cited in Thomas and Hedden, 2006) but GA5 causes less feed-forward induction of GA 2-oxidase (Figure 5, Table 1). A further class of GA deactivation enzymes were reported recently for Arabidopsis—the GA-methylases (Varbanova et al., 2007). However, the published Northern blot assays show that their expression is restricted to siliques, so they may not be relevant in the control of GA import into the shoot apex.
The importance of localized, sub-apical deactivation of GA1/GA4 but not of GA5 in vegetative plants is shown by three independent lines of our evidence. First, during the 2 d after a florally-inductive LD, expression of GA deactivation genes is greater in the sub-apical vasculature than it is some millimetres further down the stem. This expression analysis confirms and considerably extends in-situ expression studies with rice (Sakamoto et al., 2001) and 2-oxidase promoter-GUS expression studies in Arabidopsis (Jasinski et al., 2005). Second, our activity assays with isolated shoot tips showed stability of GA5 but rapid deactivation of GA1, GA4, or GA20 (Figure 6, and see Junttila et al., 1997). Third, access of endogenous GAs to the shoot apex matches their stability. GA5 is present in vegetative shoot apices but GA1 and GA4 are absent or barely detectable (King et al., 2001). Similarly, the shoot apex content of GA20, which is the precursor of GA1, is less than 10% of that in sub-apical stem tissue (GA20 1.0 ± 0.1 ng g–1 dry weight in the shoot apex but 13.9 ± 2.2 ng g–1 dry weight 3–6 mm below the shoot apex; King and Moritz, unpublished data). Thus, we now have a basis for understanding why GA5 is most effective for flowering of L. temulentum.
GA Deactivation and Flowering
A number of our findings address the question of the relationship between sub-apical GA deactivation and flowering. First, GAs susceptible to deactivation—GA1 and GA4—are weakly florigenic (Evans et al., 1990) and not detected in the shoot apex of vegetative plants or during the first days after exposure to a florally inductive LD. In contrast, GA5 is not deactivated and, when applied, it is strongly florigenic. Furthermore, following the rapid (<4 h) increase in GA5 in a leaf exposed to a florally inductive LD (King et al., 2006), its content in the shoot apex doubles within hours (King et al., 2001). Second, structural change to protect GA1 and GA4 at sites crucial for attack by 2-oxidases and 16,17-epoxidases (C-2; C-13 and C-16,17; see Figure 1) enhances their florigenicity and this is especially apparent for 2,2-dimethyl GA4—a GA form which is not metabolized in vitro by a plant 2-oxidase (Yamauchi et al., 2007). Less dramatic but similar protection at C-2 is evident from the increased florigenicity of the 2-hydroxylated derivatives, GA56 and GA47: both of these GAs are more active than their C-2 β-hydroxylated forms, GA8 and GA34, respectively (Figure 9). This set of structural comparisons (Figure 9) also highlights the value of the C-13 hydroxyl—an effect attributable to protection against C-16,17-epoxidases (Zhu et al., 2006) by reduced nucleophilicity at C-16,17 associated with electron withdrawal. Third, GA4 becomes florigenic when 2-oxidase activity is blocked by simultaneous application of the 2-oxidase inhibitor, TNE (Figures 7 and 8). This interaction was apparently near the shoot apex because TNE applied to the shoot tip enhanced the effectiveness of GA4 applied to the leaf. Thus, sensitivity/insensitivity to metabolism is central for florigenicity of a GA.
Taken together, selective deactivation of GA1/GA4 limits their florigenicity. In contrast, the greater stability of GA5 and GA6, endogenous GAs of Lolium (see King et al., 2001, 2003) allows them to act ‘by default’, as LD up-regulated transported signals that control early events of floral initiation of Lolium (see King et al., 2003, 2006). This ‘by default’ early action of GA5 is consistent with the evidence that GA4 and GA1 can be intrinsically florigenic if protected against deactivation (Figures 2, 6, 7 and 8
). It is also supported by our evidence that GA4 becomes florigenic later in floral development (King et al., 2001). At this time, there is an associated reduction of 30% in the rate of GA4 metabolism (Table 2) and endogenous GA1 and GA4 become detectable in the shoot apex (King et al., 2001). Interestingly, a rice 2-oxidase studied by Sakamoto et al. (2001) shows reduced sub-apical expression at this time and, in a preliminary experiment, we have detected a similar reduction in expression of LtGA2ox1 (King, unpublished).
Although not central to our focus on flowering, GA1 and GA4 stimulate stem elongation much more effectively than GA5 (e.g. Figures 2 and 5). This is consistent with the shoot tip tissues showing a dominant effect involving protection against deactivation. Further from the shoot apex (Figure 3: basal stem segments 4–5 mm below the apex and just above the zone of lateral root formation), reduced deactivation of GA1 and GA4 would allow them to dominate by virtue of their greater receptor-binding effectiveness. This claim is supported by evidence that GA1 and GA4 are100–500 times more effective than GA5 for amylase production by barley half-seeds, where neither GA synthesis nor degradation are likely to be important (King et al., 2004 and references therein). Overall, reduced GA1 and GA4 deactivation allows them to act on stem elongation and as the most important effectors because of their high receptor-binding capacity. By contrast, in the same plant, GA5 can dominate shoot apex responses because it resists degradation and so is effective despite its potentially low receptor-binding capacity.
Unlike L. temulentum, in which only endogenous GA5 and GA6 increase at the shoot apex to regulate its flowering, with Arabidopsis, GA1 and GA4 increase in its shoot tip when it flowers in SD (Eriksson et al., 2006) and LD increases GA4 levels in the shoot (Gocal et al., 2001). Also, unlike Lolium, in which applied GA4 is ineffective for flowering (Figures 7 and 8), in SD, it causes flowering of Arabidopsis (Xu et al., 1997; Gocal et al., 2001; Eriksson et al., 2006) and of other LD-responsive species, including Matthiola incana (Hisamatsu et al., 1998) and Thlaspi arvense (Metzger, 1990). Thus, GA deactivation is apparently unimportant for flowering of these other LD species. As a further contrast, some annual dicots, including Pisum sativum (Reid et al., 1977) and Sinapis alba, do not require GA for their flowering and, not surprisingly, there is no evidence for a florigenic role for GA5 (Corbesier et al., 2004). For another LD dicot—Fuchsia hybrida—GAs actually inhibit flowering, but this involves a unique response to GA associated with an over-stimulation of stem elongation and a diversion of photosynthate away from the shoot apex (King and Ben-Tal., 2001).
Overall, while GAs are clearly not a universal florigen, they play an important role in the flowering of a wide range of higher plant species (Pharis and King, 1985; King and Evans, 2003; Eriksson et al., 2006). Our focus here on their deactivation has provided a more substantial but more complex understanding and one which may be more relevant to flowering of grasses and cereals. The scenario of a localized deactivation of the most growth-active GAs allows for GA regulation of flowering while restricting/delaying stem elongation. Such a mechanism could be critical for grasses to survive extreme winter cold and grazing by animals. Interestingly, the late attenuation of GA4 deactivation is consistent with enhanced stem elongation later in floral development. In contrast, dicot annuals, including Arabidopsis, which would have evolved without this selection pressure, flower in response to GA4 and bolt at the same time as they initiate flowers. Unlike the grasses, herbaceous dicots may not have needed to acquire a mechanism for restricting access of highly growth-active GAs to the shoot apex.
|
METHODS |
|---|
|
TOPAbstractINTRODUCTIONRESULTSDISCUSSIONMETHODS |
|---|
Plant Material and Growing Conditions
As described previously (MacMillan et al., 2005; King et al., 2006), plants of Lolium temulentum L. strain Ceres and a clone of Lolium perenne, CPM-1, were grown vegetatively in 8-h, short day (SD) photoperiods in sunlit controlled environment cabinets. Both species remained vegetative in SD and could be induced to flower by a single LD for L. temulentum or by two LD for L. perenne after it had been vernalized at 8°C for 8 weeks in SD. We report flowering as either shoot apex length or floral score because these measures are equivalent (see McDaniel et al., 1991). There were 10–14 replicates per treatment and values for flowering and stem length are the mean ± S.E.
Leaf applications of GAs or TNE, Trinexapac-ethyl (Primo TM), ethyl[4-(1'-cyclopropyl{1'-hydroxy}methylene)-3,5-dioxocyclohexane-1-carboxylate], were in 95% (v/v) ethanol/water, as outlined previously (King et al., 2006). The culturing of excised shoot apices for analysis of GA metabolism was described in Junttila et al. (1997).
Expression of L. temulentum GA-2-oxidases and a 16,17-epoxidase
Common steps of GA deactivation often involve enzymes from conserved, multigene families but with differing substrate specificities (e.g. Thomas et al., 1999; Lester et al., 1999; Spielmeyer et al., 2004; Thomas and Hedden, 2006). Thus, Lolium likely has several GA 2-oxidases and we have examined two of these. LtGA2ox1 will metabolize GA20 to GA29 but not GA1 to GA8 (King et al., (2006). The second gene, LtGA2ox5 (GenBank # EF687858
[GenBank]
), was identified in a L. perenne EST collection using a barley 2-oxidase (Hv2ox5: AY551433
[GenBank]
) as a query sequence in a Blastn search. As for Hv2ox5, when transiently expressed in a reticulocyte assay (Spielmeyer et al., 2004), in 2 h, the protein of LpGA2ox5 almost completely converted [14C]GA1 to its 2β-hydroxylated product [14C]GA8. [14C]GA4 and [14C]GA20 were also metabolized by this enzyme (results not shown).
To analyze expression of a L. temulentum GA 16,17-epoxidase, we began with a L. rigidum cytochrome P450 gene, Lol-31-j (GenBank # AF321861 [GenBank] ; Fischer et al., 2001). This gene shows 41% amino acid homology to a rice GA 16,17-epoxidase, CYP714D1 (Zhu et al., 2006). Primers based on L. rigidum sequence were able to amplify the predicted epoxidase product from L. temulentum cDNA and there was 98% amino acid identity for a 197-bp product. Similarly, genes from L. temulentum and L perenne show very high amino acid identity (e.g. 97% for GA20oxidase1 and CONSTANS: King et al., 2006, and 97% between LtGA2ox5 and LpGA2ox5 (not shown)).
The primer pairs used for LtGA2ox1 were as described before (King et al., 2006); those for LtGA2ox5 and LtGA16,17epoxidase studies were:
LtGA2ox5: Forward: CAGGGCTTCTTCAAGGTGAC; Rev: GTGAGGCAGAGCAGGAGGTA
Putative LtGA16,17epoxidase: Forward: GATGGAGGAAGCAGGGTGTA; Rev:CAAGCAAGTGAGGGAAGAGG
For gene expression studies, hand-cut stem cross-sections 0.83 ± 0.05 mm thick, excluding leaf primordia, were collected sequentially from the shoot apex down and immediately frozen in tubes held on dry ice and combined with successive harvests of matching sections from 10–14 plants. The bulked samples were stored at –80°C until ground for RNA extraction. In related expression studies, the shoot tip (0–3.5 mm) was taken. Total RNA was isolated using an RNeasy mini kit (Qiagen USA, Valencia, CA).
The conditions for quantitative PCR assays were as detailed in King et al. (2006). For normalization, we included a commercially available mouse liver RNA. The Q-PCR analysis was performed using the ‘Comparative Quantification’ method with Rotogene 4.5 or 5.0 Software (Corbett Research).
To examine localization below the apex of LtGA2ox1 mRNA, 10-µm-thick cross-sections were prepared from fixed, paraffin-embedded shoot tips. In vitro transcribed DIG-labelled riboprobes were synthesized for a 3’ gene-specific 360-bp fragment (nucleotide 1118–1477). Using EcoRI:HindIII to excise the nucleotide fragment inserted in pBluescript SK(+) or pBluescript KS(+), sense and antisense riboprobes were generated from the T7 promoter. Hybridizations were at a high stringency, with an equivalent sense or anti-sense probe concentration and were performed simultaneously on each tissue set. The basic in-situ hybridization protocol was as described in Ferrándiz et al. (2000).
|
Acknowledgements |
|---|
Bruce Twitchin (RSC ANU Canberra Australia) is thanked for synthesizing many of the novel GA derivatives used here. Peter Chandler (CSIRO) is thanked for much discussion and support with the gene functional assays. No conflict of interest declared.
-
Blázquez MA, Weigel D. Integration of floral inductive signals in Arabidopsis. Nature (2000) 404:889–892.[CrossRef][ISI][Medline]
Busov VB, Meilan R, Pearce DW, Ma C, Rood SB, Strauss SH. Activation tagging of a dominant gibberellin deactivation gene (GA 2-oxidase) from Poplar that regulates tree stature. Plant Physiol. (2003) 132:1283–1291.
Corbesier L, Kustermans G, Périlleux C, Melzer S, Moritz T, Havelange A, Bernier G. Gibberellins and the floral transition in Sinapis alba. Physiol. Plantarum (2004) 122:152–158.[CrossRef]
Curtis IS, Hanada A, Yamaguchi S, Kamiya Y. Modification of plant architecture through the expression of GA 2-oxidase under the control of an estrogen inducible promoter in Arabidopsis thaliana L. Planta (2005) 222:957–967.[CrossRef][ISI][Medline]
Eriksson S, Bohlenius H, Moritz T, Nilsson O. GA4 is the active gibberellin in the regulation of LEAFY transcription and Arabidopsis floral initiation. Plant Cell (2006) 18:2172–2181.
Evans LT. Inflorescence initiation in Lolium temulentum L.V.: the role of auxins and gibberellins. Aust. J. Biol. Sci. (1964) 17:10–23.
Evans LT, King RW, Chu A, Mander LN, Pharis RP. Gibberellin structure and florigenic activity in Lolium temulentum, a long-day plant. Planta (1990) 182:97–106.[ISI]
Evans LT, King RW, Mander LN, Pharis RP. The relative significance for stem elongation and flowering in Lolium temulentum of gibberellin 3β-hydroxylation. Planta (1994) 192:130–136.[ISI]
Ferrándiz C, Gu Q, Martienssen R, Yanovsky M. Redundant regulation of meristem identity and plant architecture by FRUITFULL, APETALA1 and CAULIFLOWER. Development (2000) 127:725–734.[Abstract]
Fischer TC, Klattig JT, Gierl A. A general cloning strategy for divergent plant cytochrome P450 genes and its application in Lolium rigidum and Ocimum basilicum. Theor. Appl. Genet. (2001) 103:1014–1021.[CrossRef][ISI]
Gocal GFW, et al. GAMYB-like genes and gibberellin signalling in Arabidopsis. Plant Physiol. (2001) 127:1682–1693.
Gocal GF, Poole AT, Gubler F, Watts RJ, Blundell C, King RW. Long-day up-regulation of a GAMYB gene during Lolium temulentum inflorescence formation. Plant Physiol. (1999) 119:1271–1278.
Hisamatsu T, King RW, Helliwell CA, Koshioka M. The involvement of gibberellin 20-oxidase genes in phytochrome-regulated petiole elongation of Arabidopsis. Plant Physiol. (2005) 138:1106–1116.
Hisamatsu T, Koshioka M, Kubota S, King RW. Effect of gibberellin biosynthesis inhibitors on growth and flowering of Stock (Matthiola incana (L.). R. BR.). J. JPN Soc. Hortic. Sci. (1998) 67:537–543.
Jasinski S, Piazza P, Craft J, Hay A, Woolley L, Rieu L, Phillips A, Hedden P, Tsiantis M. Knox action in Arabidopsis is mediated by coordinate regulation of cytokinin and gibberellin activities. Curr. Biol. (2005) 15:1560–1565.[CrossRef][ISI][Medline]
Junttila O, King RW, Poole A, Kretschmer G, Pharis RP, Evans LT. Regulation in Lolium temulentum of the metabolism of gibberellin A20 and gibberellin A1 by 16,17-dihydro GA5 and by the growth retardant, LAB 198 999. Aust. J. Plant Physiol. (1997) 24:359–369.
King RW, Ben-Tal Y. A florigenic effect of sucrose in Fuchsia and its inhibition by gibberellin-induced assimilate competition. Plant Physiol. (2001) 125:1–9.
King RW, Evans LT. Gibberellins and flowering of grasses and cereals: prizing open the lid of the ‘florigen’ black box. Annual Review of Plant Biology. (2003) 54:307–328.[CrossRef][Medline]
King RW, Evans LT, Mander LM, Moritz T, Pharis RP, Twitchin B. Synthesis of gibberellin GA6 and examination of its role in flowering of Lolium temulentum. Phytochemistry (2003) 62:77–82.[CrossRef][ISI][Medline]
King RW, Junttila O, Mander LM, Beck E. Gibberellin structure and function: biological activity and competitive inhibition of gibberellin 2- and 3-oxidase enzymes. Physiol. Plantarum (2004) 120:287–297.[CrossRef][Medline]
King RW, Moritz T, Evans LT, Junttila O, Herlt AJ. Long day induction of flowering in Lolium temulentum involves sequential increases in specific gibberellins at the shoot apex. Plant Physiol. (2001) 127:624–632.
King RW, Moritz T, Evans LT, Martin J, Andersen CH, Kardailsky I, Blundell C, Chandler PM. Gibberellin biosynthesis and the regulation of flowering in the long day grass, Lolium temulentum L. by gibberellins and the gene FLOWERING LOCUS T. Plant Physiol. (2006) 141:498–507.
Kloosterman B, Navarro C, Bijsterbosch G, Lange T, Prat S, Visser RGF, Bachem CWB. Stga2ox1 is induced prior to stolon swelling and controls GA levels during potato tuber development. Plant J. (2007) 52:362–373.[CrossRef][ISI][Medline]
Lee DJ, Zeevaart JAD. Differential regulation of RNA levels of gibberellin dioxygenases by photoperiod in spinach. Plant Physiol. (2002) 130:2085–2094.
Lee DJ, Zeevaart JAD. Molecular cloning of GA 2-oxidase3 from Spinach and its ectopic expression in Nicotiana sylvestris. Plant Physiol. (2005) 138:243–254.
Lester DR, Ross JJ, Smith JJ, Elliott RC, Reid JB. Gibberellin 2-oxidation and the SLN gene of Pisum sativum. Plant J. (1999) 19:65–73.[CrossRef][ISI][Medline]
MacMillan CP, Blundell CA, King RW. Flowering of the grass Lolium perenne L.: effects of vernalization and long days on gibberellin biosynthesis and signalling. Plant Physiology (2005) 138:1794–1806.
Martin DN, Proebsting WM, Hedden P. The SLENDER gene of pea encodes a gibberellin 2-oxidase. Plant Physiol. (1999) 121:775–781.
McDaniel CN, King RW, Evans LT. Floral determination and in vitro floral differentiation in isolated shoot apices of Lolium temulentum L. Planta (1991) 185:9–16.[ISI]
Metzger JD. Comparison of biological activities of gibberellin and gibberellin precursors native to Thlaspi arvense L. Plant Physiol. (1990) 94:151–156.
Metzger JD, Zeevaart JAD. Effect of photoperiod on the levels of endogenous gibberellins in spinach as measured by combined gas chromatography selected ion current monitoring. Plant Physiol. (1980) 66:844–846.
Pearce DW, Hutt OE, Rood SB, Mander LN. Gibberellins in shoots and developing capsules of Populus species. Phytochemistry (2002) 59:679–687.[CrossRef][ISI][Medline]
Pharis RP, King RW. Gibberellins and reproductive development in seed plants. Annu. Rev. Plant Phys. (1985) 36:517–568.[CrossRef][ISI]
Rademacher W. Growth retardants: effects on gibberellin biosynthesis and other metabolic pathways. Annu. Rev. Plant Phys. (2000) 51:501–531.[CrossRef][ISI]
Reid JB, Dalton PJ, Murfet IC. Flowering in Pisum: does gibberellic acid directly influence the flowering process? Aust. J. Plant Physiol. (1977) 4:479–483.
Sakamoto T, Kobayashi M, Itoh H, Tagiri A, Kayano T, Tanaka H, Iwahori S, Matsuoka M. Expression of a gibberellin 2-oxidase gene around the shoot apex is related to phase transition in rice. Plant Physiol. (2001) 125:1508–1516.
Schomburg FM, Bizzell CM, Lee DJ, Zeevaart JAD, Amasino RM. Overexpression of a novel class of gibberellin 2-oxidases decreases gibberellin levels and creates dwarf plants. Plant Cell (2003) 15:151–163.
Spielmeyer W, Ellis M, Robertson M, Ali S, Lenton JR, Chandler PM. Isolation of gibberellin metabolic pathway genes from barley and comparative mapping in barley, wheat and rice. Theor. Appl. Genet. (2004) 109:847–855.[CrossRef][ISI][Medline]
Talon M, Zeevaart JAD. Gibberellin and stem growth as related to photoperiod in Silene armeria L. Plant Physiol. (1990) 92:1094–1100.
Thomas SG, Hedden P. Gibberellin metabolism and signal transduction. In: Plant Hormone Signalling—Hedden P, Thomas SG, eds. (2006) Oxford, UK: Blackwell Publishing. 147–184.
Thomas SG, Phillips AL, Hedden P. Molecular cloning and functional expression of gibberellin 2-oxidases, multifunctional enzymes involved in gibberellin deactivation. Proc. Natl Acad. Sci. U S A (1999) 96:4698–4703.
Varbanova M, et al. Methylation of gibberellins by Arabidopsis Gamt1 and Gamt2. Plant Cell (2007) 19:1040–4651.
Wilson RN, Heckman JW, Somerville CR. Gibberellin is required for flowering in Arabidopsis thaliana under short days. Plant Physiol. (1992) 100:403–408.
Xu Y-L, Gage DA, Zeevaart JAD. Gibberellins and stem growth in Arabidopsis thaliana. Effects of photoperiod on expression of the GA4 and GA5 loci. Plant Physiol. (1997) 114:1471–1476.[Abstract]
Yamauchi Y, Takeda-Kamiya N, Hanada A, Ogawa M, Kuwahara A, Seo M, Kamiya Y, Yamaguchi S. Contribution of gibberellin deactivation by AtGA2ox2 to the suppression of germination of dark-imbibed Arabidopsis thaliana seeds. Plant Cell Physiol. (2007) 48:555–561.
Zhu YY, et al. Elongated uppermost internode encodes a cytochrome P450 monooxygenase that epoxidizes gibberellins in a novel deactivation reaction in rice. Plant Cell (2006) 18:442–456.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||