• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of plntcellLink to Publisher's site
Plant Cell. Mar 2001; 13(3): 667–680.
PMCID: PMC135510

Gibberellin/Abscisic Acid Antagonism in Barley Aleurone Cells

Site of Action of the Protein Kinase PKABA1 in Relation to Gibberellin Signaling Molecules

Abstract

The antagonism between gibberellins (GA) and abscisic acid (ABA) is an important factor regulating the developmental transition from embryogenesis to seed germination. In barley aleurone layers, the expression of genes encoding α-amylases and proteases is induced by GA but suppressed by ABA. It has been shown that an ABA-induced protein kinase, PKABA1, mediates the ABA suppression of α-amylase expression. Using a barley aleurone transient expression system, we have now localized the site of action of PKABA1 relative to other signal transduction components governing the expression of α-amylase. The expression of α-amylase can be transactivated by the transcription factor GAMyb, which is itself induced by GA. A truncated GAMyb containing the DNA binding domain but lacking the transactivation domain prevents the GA induction of α-amylase, further supporting the notion that GAMyb mediates the GA induction of α-amylase expression. Although ABA and PKABA1 strongly inhibit the GA induction of α-amylase, they have no effect on GAMyb-transactivated α-amylase expression. Using a GAMyb promoter–β-glucuronidase construct, we also show that both ABA and PKABA1 repress the GA induction of GAMyb. In the slender mutant, GAMyb and α-amylase are highly expressed, even in the absence of GA. However, this constitutive expression can still be inhibited by ABA, PKABA1, or an inhibitor of cGMP synthesis. On the basis of these observations, we suggest that PKABA1 acts upstream from the formation of functional GAMyb but downstream from the site of action of the Slender gene product. Because PKABA1 inhibits the GA induction of the GAMyb promoter–β-glucuronidase construct, it appears that at least part of the action of PKABA1 is to downregulate GAMyb at the transcriptional level.

INTRODUCTION

The interaction between phytohormones, particularly that between gibberellin (GA) and abscisic acid (ABA), is an important factor controlling the transition from embryogenesis to germination in seed. During germination of cereal grains, the embryo secretes GA to the aleurone layer, where it promotes the expression of several genes encoding hydrolytic enzymes (Ritchie and Gilroy, 1998b; Lovegrove and Hooley, 2000, and references therein). Expression of these genes is blocked by ABA during seed development, in dormant seeds, and in seedlings under unfavorable germination conditions. Cereal aleurone layers, therefore, are an excellent system in which to explore the molecular mechanisms involved in hormonally regulated gene expression, particularly the antagonism between GA and ABA (Bethke et al., 1997; Lovegrove and Hooley, 2000).

The promoter sequences important for the GA induction and ABA suppression of α-amylase genes have been studied extensively (Skriver et al., 1991; Gubler and Jacobsen, 1992; Lanahan et al., 1992; Rogers and Rogers, 1992). Three regions—box1 (amylase box), GARE (for gibberellin response element), and pyrimidine box—are found in the promoters of all GA-inducible α-amylase genes. An additional region important for GA inducibility (O2S binding box) is found in low pI α-amylase gene promoters. The promoter of a GA-induced cysteine proteinase has elements similar to those found in α-amylase promoters, such as the GARE and the pyrimidine box, and a newly described upstream element necessary for GA responsiveness (Cercós et al., 1999). The characterization of these promoter elements has provided the basis for the discovery of trans-acting factors involved in the transduction of GA signals in barley aleurone layers. A DNA binding protein complex capable of interacting with the GARE box and surrounding regions has been shown to be present only in GA-treated aleurone layers (Sutliff et al., 1993). Gubler et al. (1995) cloned a GA-induced Myb-like protein (GAMyb) that binds specifically to the GARE box of an α-amylase promoter. GAMyb is able to transactivate the expression of α-amylase and other GA-regulated genes (Gubler et al., 1995, 1999; Cercós et al., 1999). Several negative regulators of the GA induction of gene expression have been identified in barley. A zinc fingerlike protein has been shown to bind the GARE box and repress the expression of α-amylase and other genes (Raventos et al., 1998). A barley homolog of the Arabidopsis SPY gene (Jacobsen et al., 1996), which encodes a putative O-linked N-acetylglucosamine transferase, strongly decreases GA induction of the α-amylase gene (Robertson et al., 1998).

Another putative negative regulator of GA signaling is the protein encoded by the Slender1 (SLN1) gene. A recessive mutation in this gene leads to pleiotropic phenotypes, such as rapid elongation of stems and leaves, sterility, and high α-amylase production, even in the presence of GA biosynthesis inhibitors (Foster, 1977; Chandler, 1988; Lanahan and Ho, 1988). The slender (sln1) mutant phenotype can be mimicked in wild-type plants by exogenous applications of a high dose of GA. However, the endogenous levels of active GA in the slender mutant are known to be lower than those in the wild type (Croker et al., 1990).

Several biochemical studies indicate that GA is perceived at the plasma membrane (Hooley et al., 1991; Gilroy and Jones, 1994) and that increased levels of cytosolic calcium (Gilroy, 1996) and calmodulin (Schuurink et al., 1996) are early events in signal transduction. G proteins and protein phosphatases also may be involved in GA signaling (Kuo et al., 1996; Jones et al., 1998). In addition, Penson et al. (1996) determined that cGMP is an important component in the transduction of the GA signal.

The ABA signal transduction pathway in aleurone layers is largely unknown (for recent reviews, see Leung and Giraudat, 1998; Lovegrove and Hooley, 2000). Both plasma membrane and internal receptors have been postulated (Gilroy and Jones, 1994; Gilroy, 1996). Phospholipase D has been proposed as an intermediate in the propagation of ABA signaling (Ritchie and Gilroy, 1998a). The levels of cytosolic calcium decrease in response to ABA treatment (Gilroy, 1996), suggesting a possible role of calcium as a second messenger. The importance of protein phosphorylation in transducing the ABA signal also has been shown in several reports (Sheen, 1996; Bethke et al., 1997; Grill and Himmelbach, 1998; Li et al., 2000). Specifically, it has been demonstrated that ABA's induction of genes such as HVA1 and suppression of α-amylase genes diverge in two different signal transduction pathways, with the induction branch being sensitive to a protein phosphatase 2C (Shen, 1996) and the suppressive pathway being modulated by an ABA-responsive serine/threonine protein kinase, PKABA1 (Anderberg and Walker-Simmons, 1992; Gómez-Cadenas et al., 1999). PKABA1 transcript levels increase in response to ABA in scutellar, root, and shoot tissues (Holappa and Walker-Simmons, 1995). As ABA levels increase in drying seeds, PKABA1 transcript levels increase, reaching their peak at seed maturity. PKABA1 transcript levels remain high in ABA-treated aleurone layers but decrease below detectable levels in GA-treated aleurone.

Although it is known that PKABA1 represses the GA induction of genes coding for hydrolytic enzymes, the molecular mechanism underlying the GA/ABA antagonism has not been fully explored. In this work, using particle bombardment coupled with the transient expression of transgenes in barley aleurone layers, we studied how the ABA and GA signal transduction cascades interact with each other. We show that PKABA1 acts as an intermediary of the ABA signal transduction pathway, repressing the GA induction of GAMyb and therefore of hydrolytic enzymes. The involvement of the slender mutant is also addressed, and our observations allow us to locate SLN1 in the GA signal transduction pathway acting upstream of GAMyb and PKABA1.

RESULTS

The Transcription Factor GAMyb Transactivates α-Amylase Promoters

The GA induction of α-amylase gene expression has been studied extensively (Ritchie and Gilroy, 1998b, and references therein). The role of the transcription factor GAMyb as a regulator of α-amylase expression has been shown (Gubler et al., 1995). To test this effect in our experimental system, we cotransformed barley embryoless half-seed (i.e., without an endogenous source of GA) with the GAMyb cDNA driven by the constitutive Ubi-1 promoter along with a low pI α-amylase–β-glucuronidase (GUS) reporter construct and subsequently incubated the transformed tissue for 24 hr in the absence of hormones (Figure 1). When the α-amylase reporter construct was bombarded alone (Figure 1B, 0 on the x axis), low basal GUS activity levels were detected. Cotransformation of the reporter construct with the GAMyb effector construct (Figure 1A) strongly increased the GUS activity level in the absence of gibberellic acid (GA3) treatment (Figure 1B). This transactivation of α-amylase expression (as measured by GUS activity) followed a dosage-dependent pattern and eventually reached a level comparable to that of GA3-treated half-seed bombarded only with the reporter construct (Figure 1B, dotted line). Similar results were obtained with a high pI α-amylase reporter construct (data not shown). The maize C1 protein, a different Myb-like protein (Paz-Ares et al., 1987) driven by the same constitutive promoter, failed to transactivate the α-amylase promoters (data not shown; Gubler et al., 1995).

Figure 1.
The Transcription Factor GAMyb Transactivates the α-Amylase Promoter in the Absence of GA3.

A Truncated Version of GAMyb Blocks the GA Induction of α-Amylase Genes

To elucidate the role of GAMyb as a specific component of the GA signal transduction pathway leading to the induction of the α-amylase genes, a truncated version of GAMyb (GAMyb-BD) was made by mutating the codon for Asn-159 (at the end of the DNA binding domain) to a stop codon, deleting the presumptive transactivation domain. Although incubation with GA3 or cobombardment with the full-length GAMyb induced the expression of the α-amylase reporter construct (Figure 2B), coexpression of GAMyb-BD not only failed to transactivate the α-amylase promoter but also blocked the response of this reporter construct to GA3 treatment. Similar results were obtained with a high pI α-amylase reporter construct (data not shown).

Figure 2.
A Truncated Version of GAMyb (GAMyb-BD) Blocks the GA Induction of α-Amylase.

The Transactivation of the α-Amylase Promoter by GAMyb Cannot Be Repressed by ABA or the Protein Kinase PKABA1

We have shown previously that PKABA1 mimics the role of ABA as a repressor of the GA induction of hydrolytic enzymes (Gómez-Cadenas et al., 1999). To further localize the site of action of PKABA1, we cobombarded embryoless half-seed with PKABA1 and GAMyb effector constructs along with the α-amylase–GUS reporter construct (Figure 3). The GA induction of α-amylase was repressed by either ABA treatment or cobombardment with PKABA1. However, both treatments failed to inhibit the transactivation of the same reporter construct by GAMyb (Figure 3B). The level of GUS produced as the result of GAMyb transactivation was virtually unchanged when the aleurone layers were treated with ABA or cobombarded with PKABA1.

Figure 3.
The GAMyb Transactivation of the α-Amylase Promoter Is Not Suppressed by ABA or the Protein Kinase PKABA1.

The Expression of GAMyb Is Hormone Responsive

The data shown in Figure 3 suggest that the ABA signaling cascade antagonizes the GA signal transduction pathway upstream of the formation of a functional GAMyb. It has also been reported that the expression of GAMyb is regulated at the level of transcription (Gubler et al., 1995). To clarify the interaction between PKABA1 and GAMyb, a genomic clone of GAMyb was isolated and the hormonal regulation of GAMyb expression was studied. To determine the 5′ boundary of the GAMyb promoter, a series of 5′ terminal deletions was made (Figure 4).

Figure 4.
The Expression of GAMyb Is Hormonally Regulated.

Three different constructs were cobombarded into embryoless half-seed along with an internal control construct, UBI–luciferase. The basal levels of expression of the GAMyb constructs (Figure 4, open bars) were relatively high compared with those of the α-amylase reporter construct in the absence of any effector construct (Figures 1 and and2).2). Incubation with GA3 for 12 hr induced the expression of all three constructs (Figure 4, gray versus open bars). The construct with 327 bp of 5′ flanking sequence (pRZ135) displayed the highest GA induction. On the other hand, the GA induction of these constructs was still repressed by the presence of ABA, although deleting to −168 (pRZ141) resulted in the partial loss of ABA suppression, that is, no significant difference between +GA3 and +GA3+ABA samples anymore. However, the levels of basal expression were still high (Figure 4). In light of these observations, the pRZ135 construct was chosen for subsequent experiments.

To analyze the time course of the GA induction and ABA suppression of GAMyb, the construct pRZ135 was bombarded into embryoless half-seed, which were then incubated for different times with no hormones, GA3, or a mixture of GA3 and ABA (Figure 5). The maximal GA induction was reached by 12 hr. ABA repressed the GA induction, although this repression was only partial after 24 hr of incubation with hormones (Figure 5). Similar results were obtained when the pRZ126 construct was used for the same experiment (data not shown). RNA gel blot analysis was used in a study that yielded comparable results showing an increase in GAMyb transcript levels in response to GA3 (Gubler et al., 1995). These transient expression results are consistent with our own RNA gel blot findings, which showed that ABA partially blocked the GA induction of GAMyb transcript levels after 24 hr of incubation with hormones (data not shown).

Figure 5.
Time Course of GA Induction of GAMyb.

PKABA1 Specifically Represses the GA Induction of GAMyb

To examine the possible role of PKABA1 in mediating ABA suppression of the GAMyb promoter, we cobombarded embryoless barley half-seed with the effector construct UBI–PKABA1 along with the GAMyb reporter construct (Figure 6). In the absence of the effector construct, the expression of GAMyb was induced after 12 hr of incubation with GA3. However, coexpression of PKABA1 along with the GAMyb reporter construct blocked the response to GA3. Similar results were obtained using PKABA1 cDNA driven by the cauliflower mosaic virus 35S promoter (35S) (data not shown). As controls, cobombardment of a different protein kinase, CDPKci (Harper et al., 1994), or a null version of PKABA1 (without the glycine-rich loop that is thought to be necessary for nucleotide binding and protein kinase activity) (Gómez-Cadenas et al., 1999) driven by the 35S or Ubi-1 promoter, failed to repress GA induction of the reporter construct (Figure 6).

Figure 6.
PKABA1 Specifically Represses the GA Induction of GAMyb.

PKABA1 Blocks the Constitutive Expression of α-Amylase and GAMyb in the slender Mutant

The slender mutation in barley is a single-locus recessive mutation that causes a plant to appear as if it had been grown in saturating concentrations of GA, even in the presence of GA biosynthesis inhibitors (Foster, 1977; Chandler, 1988; Lanahan and Ho, 1988). This phenotype is observed despite the fact that endogenous levels of active GA are lower in the slender mutant than in the wild type (Croker et al., 1990). Thus, SLN1 is probably involved in the repression of the GA signal transduction pathway. We studied the role of SLN1 in GA-mediated gene expression in aleurone layers using transient expression analysis. As shown in Table 1, in the wild-type aleurone layers, expression of α-amylase and GAMyb reporter constructs was induced by GA3 treatment. However, in the slender aleurone layers, both reporter constructs showed increased levels of expression in the absence of GA3 treatment. GA3 treatment of slender aleurone layers did not enhance the expression of α-amylase or GAMyb reporter constructs.

Table 1
Levels of Expression of α-Amylase and GAMyb in the Wild Type and the slender Mutant Barleya

To understand the relationship among PKABA1, SLN1, and GAMyb, two experiments were performed in the slender mutant background. First, to further investigate the role of GAMyb as a key element in the activation of α-amylase genes, the effect of GAMyb-BD was tested on the slender aleurone layers. As expected, the expression of the α-amy-lase reporter construct was high in the absence of hormones (Figure 7). Incubation with GA3 or cobombardment with GAMyb slightly increased the already high levels of expression of α-amylase. However, cobombardment with GAMyb-BD reduced the constitutive expression of α-amylase by 50%. Second, to test the role of PKABA1 as a putative repressor of the constitutive expression of α-amylase genes in the slender mutant, several effector constructs were cobombarded into slender embryoless half-seed along with the α-amylase reporter construct (Figure 8). Coexpression of either 35S–PKABA1 (Figure 8B) or UBI–PKABA1 (Figure 8C) strongly repressed the constitutively high levels of expression of the α-amylase construct. However, coexpression of another protein kinase (CDPKci) or a null version of PKABA1 did not significantly alter the high levels of expression of α-amylase in slender aleurone layers. Figure 8B also shows that incubation with ABA effectively blocked the expression of α-amylase in the mutant tissue.

Figure 7.
A Truncated GAMyb Blocks the Constitutive Expression of α-Amylase in the slender Mutant.
Figure 8.
PKABA1 Specifically Represses the Constitutive Expression of α-Amylase in the slender Mutant.

When GAMyb–GUS was used as a reporter construct (data not shown), ABA treatment or cobombardment with PKABA1 significantly repressed the constitutive expression of the GAMyb construct by 40% in the aleurone layers of slender mutant barley.

LY83583, a Guanylyl Cyclase Inhibitor, Blocks GA Signal Transduction Upstream of GAMyb but Downstream of SLN1

The role of cGMP in hormone-regulated gene expression was established by Penson et al. (1996), who showed that cGMP levels increased in response to GA treatment in barley aleurone layers. Furthermore, the presence of a guanylyl cyclase inhibitor, LY83583 (LY), blocked the GA-induced accumulation of α-amylase and GAMyb transcripts. However, cGMP alone was not sufficient to increase either of these transcripts (Penson et al., 1996). We also have tested the site of action of cGMP relative to other regulatory molecules. Embryoless half-seed (wild type and slender) were bombarded with two GA-inducible reporter constructs, α-amylase–GUS and GAMyb–GUS, and with an ABA-inducible construct, HVA1–GUS. The half-seed were subsequently incubated with different combinations of hormones and LY (Figure 9). The expression of the α-amylase reporter construct was highly induced by GA3 in wild-type aleurone layers (Figure 9A), whereas it was constitutively high in the slender mutant (Figure 9B). However, the addition of 200 μM LY strongly inhibited α-amylase expression in both genotypes. The pattern of expression of the GAMyb–GUS reporter construct (Figures 9C and 9D) was very similar to that of the α-amylase gene, although the basal levels of GAMyb–GUS were considerably higher. Similarly, LY incubation repressed both the GA3 induction of GAMyb–GUS in wild-type aleurone layers and the constitutive expression of the same reporter construct in slender aleurone layers. To determine the specificity of cGMP as part of the GA signal transduction pathway, the HVA1 gene, which is highly induced by ABA but insensitive to GA, was used as an additional reporter construct. Figures 9E and 9F show that the slender mutation did not affect the ABA upregulation of gene expression. Furthermore, incubation with LY did not significantly repress the ABA induction of HVA1 in wild-type or slender aleurone layers.

Figure 9.
The cGMP Biosynthesis Inhibitor LY83583 Represses the Expression of GAMyb and α-Amylase.

DISCUSSION

The interaction between GA and ABA has been the subject of decades of study. Early reports by Chrispeels and Varner (1966) and Jacobsen (1973) indicated that GA and ABA do not compete for a common site, yet little is known about the antagonism between these two hormones. In this work, we found that the basis of interaction between ABA and GA is at least in part the repression of GAMyb by PKABA1, which is induced by ABA. We also used the slender mutant and an inhibitor of cGMP synthesis to locate the sites of action of PKABA1 and GAMyb. A scheme summarizing our findings is presented in Figure 10.

Figure 10.
Diagram Summarizing the Molecular Events Involved in the Interaction between ABA and GA Signal Transduction Pathways.

The role of a GA-regulated transcription factor, GAMyb, as an activator of downstream GA-regulated genes encoding α-amylases and proteinases has been well documented. It has been shown that GAMyb binds specifically to the GARE box of α-amylase promoters (Gubler et al., 1995) and that the constitutive expression of GAMyb transactivates the expression of α-amylase and other GA-inducible genes in barley aleurone layers (Cercós et al., 1999; Gubler et al., 1999). In this work, we present further evidence to suggest that the GAMyb effect on α-amylase expression is specific and physiologically relevant. GAMyb transactivation followed a dosage-dependent pattern (Figure 1), and more importantly, GAMyb-BD, a truncated protein containing only the DNA binding domain of GAMyb, not only failed to transactivate the α-amylase promoter but also blocked the GA signal transduction pathway in aleurone layers (Figure 2). We suggest that GAMyb-BD interferes with the binding of the endogenous protein to the GARE sequence, impairing the transcription of the α-amylase gene.

The role of protein phosphorylation/dephosphorylation in GA/ABA signal transduction in barley aleurone layers has been suggested elsewhere (Bethke et al., 1997; Leung and Giraudat, 1998; Ritchie and Gilroy, 1998b). We have shown previously that constitutive expression of an ABA-inducible protein kinase, PKABA1, strongly suppressed the GA induction of α-amylase and proteinase genes (Gómez-Cadenas et al., 1999). The data shown in Figure 3 provide additional evidence to support that report. Either incubation with ABA or the constitutive expression of PKABA1 effectively repressed the GA induction of an α-amylase reporter construct. However, the GAMyb transactivation of the same promoter construct was not affected by either of these treatments (Figure 3). When a UBIGAMyb construct was introduced into the aleurone cells, treatment of ABA did not repress the GAMyb transactivation of α-amylase, even with nonsaturating amounts of the GAMyb effector (data not shown). These observations strongly suggest that ABA and PKABA1 act upstream of the formation of a functional GAMyb.

To further investigate the interaction between GA and ABA, the effects of hormonal treatments and PKABA1 on GAMyb expression were studied. Deletion experiments indicated that the region up to −327 (construct pRZ135; Figure 4) of the GAMyb promoter is sufficient to confer hormone responsiveness. Deletion of the region between −327 and −168 (construct pRZ141) led to a partial loss of ABA repressibility of the GAMyb promoter. Time-course experiments (Figure 5) also confirmed previous data showing that the GA induction of GAMyb reaches a maximal level by 12 hr after GA3 treatment (Gubler et al., 1995), which slightly precedes the full induction of α-amylase. Quantitatively, the GA induction of GAMyb was moderate, mainly due to a high basal level of expression. It is possible that because of the high stability of the GUS protein used as a reporter in this study, the basal levels of GAMyb expression are overstated. However, these data are consistent with our own observations and with previously published data (Penson et al., 1996) showing elevated background levels of GAMyb transcripts by RNA gel blot analyses, even in the absence of GA. The ABA treatment was able to completely repress the GA induction of GAMyb at 12 hr (Figure 5). The constitutive expression of PKABA1 had the same consequence as the expression of ABA, whereas neither an unrelated protein kinase (CDPKci) nor a null version of PKABA1 (Gómez-Cadenas et al., 1999) was able to repress the GA inducibility of GAMyb (Figure 6). However, both ABA and PKABA1 failed to reduce the basal expression of the reporter construct (Figures 5 and and6).6). Two possible explanations could account for the lack of correlation between the tight regulation of α-amylase expression by hormones (Figures 2 and and3)3) (Lanahan et al., 1992) and the relatively high level of GAMyb transcript in the absence of hormones and in the presence of both GA3 and ABA. First, it is possible that both GA and ABA exert an additional post-transcriptional control on GAMyb expression. Second, it is likely that additional factors contribute to the regulation of α-amylase expression. A zinc finger–like DNA binding factor, HRT, has been reported as a potential negative modulator of the expression of α-amylase because it interacts with GARE in the α-amylase promoters (Raventos et al., 1998). Thus, the ratio of HRT and GAMyb also could be important for the regulation of α-amylase gene expression.

We also used the slender mutant to further dissect GA signal transduction in barley aleurone cells. Because the slender mutation is recessive, the wild-type Slender gene (SLN1) is likely to encode a negative regulator of the GA signal cascade (Foster, 1977; Chandler, 1988; Lanahan and Ho, 1988). More importantly, the effect of the slender mutation on α-amylase expression can be blocked completely by treatment with 20 μM ABA (Lanahan and Ho, 1988), which suggests that the regulatory function of this gene is exerted early in the GA signal transduction cascade. The expression of both the GAMyb and α-amylase reporter constructs was increased in the slender aleurone layers, which is comparable to the expression in GA3-treated wild-type aleurone layers (Figures 7 to to9).9). Moreover, GA3 treatment of slender aleurone layers enhanced α-amylase and GAMyb expression only slightly beyond the already high levels. Our data reinforce the importance of GAMyb as a central regulator of GA signal transduction, even in the slender mutant background, because GAMyb-BD still blocked the expression of the α-amylase reporter construct in the slender mutant (Figure 7). Furthermore, the constitutive expression of α-amylase and GAMyb can be repressed by ABA treatment or cotransformation with PKABA1 (Figure 8).

The specificity of PKABA1 mimicking the ABA effect is suggested by the fact that a different protein kinase or a null version of PKABA1 did not have a significant effect on the expression of GAMyb or α-amylase. Moreover, the ABA signal transduction pathway does not appear to be affected by the slender mutation (Chandler, 1988; Lanahan and Ho, 1988). In fact, Figures 9E and 9F show that the ABA upregulation of the HVA1 gene was not affected by the slender mutation. Collectively, these data suggest that the point of interaction between the ABA (mediated by PKABA1) and GA signal transduction pathways occurs downstream of the action of SLN1 but upstream of the action of GAMyb. The molecular nature of SLN1 has been the subject of intense investigation. By means of restriction length polymorphism analysis, it has been shown that the SLN1 and HvSPY genes are two different loci (Robertson et al., 1998). Because the slender mutation is recessive, it is likely that SLN1 encodes a repressor and that the role of GA is to inhibit the repressive action of SLN1, thus allowing the downstream signaling to proceed, similar to what has been proposed for the action of GAI and RGA in Arabidopsis (Peng et al., 1999; Sun, 2000).

Finally, the data in Figure 9 show that LY, a guanylyl cyclase inhibitor, is able to repress the GA induction of GAMyb and α-amylase reporter constructs in both wild-type and slender aleurone layers. However, the same inhibitor does not affect the ABA induction of HVA1 in both genetic backgrounds. These data, together with the observation that GA induces a rapid accumulation of cGMP in barley aleurone layers (Penson et al., 1996), reinforce the model suggesting that cGMP acts as a second messenger in the GA signal transduction pathway (Penson et al., 1996). cGMP alone cannot substitute for the GA regulation of α-amylase synthesis, which suggests cooperation among cGMP and other elements (probably Ca2+ and/or calmodulin) in the transduction of the GA signal (Penson et al., 1996). Our results suggest that SLN1 acts upstream from or at the same level as guanylyl cyclase(s) in GA signal transduction. The relationship between guanylyl cyclase(s) and PKABA1 is not yet clear, and more investigation is needed. However, ABA treatment does not modify the steady state levels of cGMP in aleurone layers (Penson et al., 1996). Therefore, we speculate that ABA (and PKABA1) blocks the GA signaling cascade downstream of the point of action of cGMP. Guanylyl cyclases appear to be a central component in other plant signal transduction networks, including light detection (Brown et al., 1989) and phytochrome A responses (Bowler and Chua, 1994). However, the molecular mechanism of cGMP action in plants is not yet well understood. In mammalian systems, cGMP can alter calcium levels that in turn control protein kinase activities and gene transcription (Butt et al., 1993; Wang and Robinson, 1997).

In summary, the data presented here provide new insights into the molecular interaction between the ABA and GA signal transduction pathways. We propose that protein phosphorylation, and particularly PKABA1, are important elements connecting the ABA cascade of events to the suppression of α-amylase gene expression. PKABA1 blocks the GA upregulation of GAMyb and therefore the expression of α-amylase in both wild-type and slender mutant aleurone layers. Our data also reinforce the role of cGMP as an important intermediary in GA signal transduction, most likely acting downstream of SLN1. Although additional regulatory molecules or different levels of regulation of PKABA1 and/or GAMyb cannot be excluded, we have demonstrated that at least part of the action of PKABA1 is to downregulate GAMyb at the transcriptional level.

METHODS

Plant Material

Barley (Hordeum vulgare cv Himalaya) seed were used for all of the experiments reported in this work. For the experiments involving slender seed, due to the sterility of the homozygous mutant, the progeny had to be maintained as a heterozygous population. Half-seed–containing embryos were planted to determine their phenotypes, and only the slender homozygous embryoless half-seed were used for transient expression.

Preparation of the DNA Constructs

Reporter constructs were prepared as follows. (1) α-Amylase– β-glucuronidase (GUS) was constructed by linking the promoter (up to −331), the entire 5′ untranslated sequence, and the first intron of the low pI α-amylase gene, Amy 32b, to the GUS coding sequence and the 3′ untranslated region of the same α-amylase gene (Lanahan et al., 1992). (2) HVA1–GUS was made by ligating the 68-bp ABRC3 (abscisic acid [ABA] response complex 3, derived from the HVA1 promoter sequence) to the GUS coding sequence (Shen et al., 1996). (3) GAMyb–GUS constructs are described below.

Effector constructs were prepared as follows. (1) UBIGAMyb consisted of the maize Ubi-1 promoter linked to the coding region of GAMyb and to the 3′ untranslated region of the nopaline synthetase gene, NOS-T (Gubler et al., 1995). (2) GAMyb-BD was made by mutating the Asn at position 159 to a stop codon in the UBIGAMyb construct by site-directed mutagenesis (Cercós et al., 1999). (3) UBI–PKABA1 was made by cloning a reconstituted PKABA1 cDNA (Anderberg and Walker-Simmons, 1992; Gómez-Cadenas et al., 1999) into the plant expression vector pAHC17 (Christensen and Quail, 1996). (4) UBI–null-PKABA1 is identical to UBIPKABA1 but lacks the glycine-rich loop (GSGNFG, amino acids 11 to 16), which is part of the nucleotide binding site (Gómez-Cadenas et al., 1999). (5) 35SCDPKci was made by linking the cauliflower mosaic virus 35S promoter to the coding region of a calcium-independent mutant CDPK gene of Arabidopsis thaliana (Harper et al., 1994; J.F. Harper, unpublished observations) and the 3′ untranslated region of the NOS gene. (6) 35SPKABA1 was made by substituting the Ubi-1 promoter in UBIPKABA1 for the 35S promoter used in the 35SCDPK construct.

Isolation of a GAMyb Genomic Clone

A genomic library of barley (cv Igri) in the vector Lambda FIX II was obtained from Stratagene (La Jolla, CA). A KpnI fragment of 1.6 kb containing the entire coding region of GAMyb (Gubler et al., 1995), which was kindly provided by Drs. Frank Gubler and John Jacobsen (Commonwealth Scientific and Industrial Research Organization [CSIRO], Australia), was used as a probe to screen 106 colony-forming units. One clone of ~16.5 kb containing the entire coding region and at least 4 kb upstream and downstream of the open reading frame was isolated. A 3.5-kb BamHI fragment containing the region 1031 to +2177 of the GAMyb gene was subcloned into the BamHI site of pBluescript KS+ (Stratagene), generating the plasmid pRZ118. The GAMyb genomic sequence was submitted to GenBank with accession number AY008692. The clone pRZ118 was subsequently cut with SapI, blunt ended with the Klenow fragment of the DNA polymerase I, and then recut with KpnI. The fragment KpnI/SapI was subcloned into pJR268b (kindly provided by Dr. John C. Rogers, Washington State University, Pullman) previously cut with BamHI, blunt ended with the Klenow fragment, and then recut with KpnI. This new plasmid (pRZ126) contains 1031 bp of the promoter, the first intron, and the sequence encoding the first seven amino acids of the GAMyb protein fused in frame to the coding region of GUS and the 3′ untranslated region of the Amy6-4 gene from barley.

The −327 deletion construct (pRZ135) was generated by cutting pRZ126 with XhoI and BglII, filling in with the Klenow fragment, and religating the plasmid. The −168 deletion construct (pRZ141) was obtained by cutting pRZ126 with XhoI and partially with MfeI, treating with the Klenow fragment, and religating the plasmid. All constructs were created and maintained in the Escherichia coli strain XL1-Blue MRF′ (Stratagene).

Particle Bombardment and Transient Expression Assays

Detailed descriptions of the transient expression procedure with the barley aleurone system and the particle bombardment technique have been published (Lanahan et al., 1992; Shen et al., 1993). Briefly, the mixture (in equal molar ratio) of a reporter construct and a maize UBIluciferase internal control construct, pAHC18 (Christensen and Quail, 1996), was bombarded into barley embryoless half-seed (four replicates per test construct).

To test the effect of regulatory molecules such as PKABA1 and GAMyb, different amounts of effector constructs were included, as indicated in the figures. After incubation in the presence or absence of hormones for different times, the bombarded half-seed in sets of four were homogenized with the aleurone still attached to the starchy endosperm in 800 μL of grinding buffer (Shen et al., 1993). After centrifugation at 12,000g for 10 min at 6°C, 100 μL of the supernatant was assayed for luciferase activity using a luminometer (Moonlight 2010; Analytic Luminescence Laboratory, San Diego, CA). For GUS assays, 50 μL of the supernatant was diluted with 200 μL of GUS assay buffer (Shen et al., 1993) and incubated at 37°C for 20 hr. Fifty microliters of the reaction mixture was diluted in 2 mL of 0.2 M Na2CO3, and the resulting fluorescence was measured in a Sequoia-Turner model 450 fluorometer (Unipath, Mountain View, CA) that was adjusted to read 1000 units for 1 μM 4-methylumbelliferone. The GUS activities of all samples were normalized against those of the luciferase internal control. The normalized GUS activity represents the total number of fluorescence units from an aliquot of extract that contained 2 × 106 relative light units of luciferase activity. The data reported in each figure represent the results of four independent transformations in a typical experiment. Each experiment was repeated at least two times with similar results. Statistical analysis was performed with the program Statgraphics Plus, v.2.1.

Acknowledgments

We thank Drs. Frank Gubler and John Jacobsen (CSIRO, Canberra City, Australia) for the barley GAMyb cDNA clone, Dr. Karen Cone (University of Missouri, Columbia) for the maize C1 Myb cDNA, Dr. John Rogers (Washington State University, Pullman) for the α-amylase clones, Dr. Peter Chandler (CSIRO) for the slender barley seed, Dr. Jeff Harper (Scripps Institute, La Jolla, CA) for the CDPK cDNAs, Dr. Lynn Hollapa (U.S. Department of Agriculture/Washington State University) for the PKABA1 null mutant, and Dr. Daisuke Yamauchi (Washington University, St. Louis, MO) for help in the initial screening for barley GAMyb genomic clones. This work was supported by grants from the U.S. Department of Agriculture (No. NRICGP 97-35100-4228) and the National Science Foundation (No. IBN-9983126).

Notes

This paper is dedicated to Maarten Chrispeels and Joe Varner, who first reported the GA/ABA antagonism in the mid sixties.

References

  • Anderberg, R.J., and Walker-Simmons, M.K. (1992). Isolation of a wheat cDNA clone for an abscisic acid-inducible transcript with homology to protein kinase. Proc. Natl. Acad. Sci. USA 89, 10183–10187. [PMC free article] [PubMed]
  • Bethke, P.C., Schuurink, R., and Jones, R.L. (1997). Hormonal signalling in cereal aleurone. J. Exp. Bot. 48, 1337–1356.
  • Bowler, C., and Chua, N.H. (1994). Emerging themes of plant signal transduction. Plant Cell 6, 1529–1541. [PMC free article] [PubMed]
  • Brown, E.G., Newton, R.P., Evans, D.E., Walton, T.J., Younis, L.M., and Vaughan, J.M. (1989). Influence of light on cyclic nucleotide metabolism in plants: Effects of dibutyl cyclic nucleotides on chloroplast components. Phytochemistry 28, 2559–2563.
  • Butt, E., Geiger, J., Jarchau, T., Lohmann, S.M., and Walter, U. (1993). The cGMP-dependent protein kinase: Gene, protein, and function. Neurol. Res. 18, 27–42. [PubMed]
  • Cercós, M., Gómez-Cadenas, A., and Ho, T.-H.D. (1999). Hormonal regulation of a cysteine proteinase gene, EPB1, in barley aleurone layers: Cis- and trans-acting elements involved in the coordinated gene expression regulated by gibberellins and abscisic acid. Plant J. 19, 107–118. [PubMed]
  • Chandler, P.M. (1988). Hormonal regulation of gene expression in the “slender” mutant of barley (Hordeum vulgare L.). Planta 175, 115–120. [PubMed]
  • Chrispeels, M.J., and Varner, J.E. (1966). Inhibition of gibberellic acid induced formation of α-amylase by abscisin II. Nature 212, 1066–1067.
  • Christensen, A.H., and Quail, P.H. (1996). Ubiquitin promoter-based vectors for high-level expression of selectable and/or screenable marker genes in monocotyledonous plants. Transgenic Res. 5, 213–218. [PubMed]
  • Croker, S.J., Hedden, P., Lenton, J.R., and Stoddart, J.L. (1990). Comparison of gibberellins in normal and slender seedlings. Plant Physiol. 94, 194–200. [PMC free article] [PubMed]
  • Foster, C.A. (1977). Slender: An accelerated extension growth mutant of barley. Barley Genet. Newslett. 7, 24–27.
  • Gilroy, S. (1996). Signal transduction in barley aleurone protoplasts is calcium dependent and independent. Plant Cell 8, 2193–2209. [PMC free article] [PubMed]
  • Gilroy, S., and Jones, R.L. (1994). Perception of gibberellin and abscisic acid at the external face of the plasma membrane of barley (Hordeum vulgare L.) aleurone protoplasts. Plant Physiol. 104, 1185–1192. [PMC free article] [PubMed]
  • Gómez-Cadenas, A., Verhey, S.D., Holappa, L.D., Shen, Q., Ho, T.-H.D., and Walker-Simmons, M.K. (1999). An abscisic acid-induced protein kinase, PKABA1, mediates abscisic acid-suppressed gene expression in barley aleurone layers. Proc. Natl. Acad. Sci. USA 96, 1767–1772. [PMC free article] [PubMed]
  • Grill, E., and Himmelbach, A. (1998). ABA signal transduction. Curr. Opin. Plant Biol. 1, 412–418. [PubMed]
  • Gubler, F., and Jacobsen, J.V. (1992). Gibberellin-responsive elements in the promoter of a barley high-pI α-amylase gene. Plant Cell 4, 1435–1441. [PMC free article] [PubMed]
  • Gubler, F., Kalla, R., Roberts, J.K., and Jacobsen, J.V. (1995). Gibberellin-regulated expression of a myb gene in barley aleurone cells: Evidence for Myb transactivation of a high-pI α-amylase gene promoter. Plant Cell 7, 1879–1891. [PMC free article] [PubMed]
  • Gubler, F., Raventos, D., Keys, M., Watts, R., Mundy, J., and Jacobsen, J.V. (1999). Target genes and regulatory domains of the GAMYB transcriptional activator in cereal aleurone. Plant J. 17, 1–9. [PubMed]
  • Harper, J.F., Huang, J.-F., and Lloyd, S.J. (1994). Genetic identification of an autoinhibitor in CDPK, a protein kinase with a calmodulin-like domain. Biochemistry 33, 7267–7277. [PubMed]
  • Holappa, L.D., and Walker-Simmons, M.K. (1995). The wheat abscisic acid-responsive protein kinase mRNA, PKABA1, is up-regulated by dehydration, cold temperature, and osmotic stress. Plant Physiol. 108, 1203–1210. [PMC free article] [PubMed]
  • Hooley, R., Beale, M.H., and Smith, S.J. (1991). Gibberellin perception at the plasma membrane of Avena fatua aleurone protoplasts. Planta 183, 274–280. [PubMed]
  • Jacobsen, J.V. (1973). Interactions between gibberellic acid, ethylene, and abscisic acid in control of amylase synthesis in barley. Plant Physiol. 51, 198–202. [PMC free article] [PubMed]
  • Jacobsen, S.E., Binkowski, K.A., and Olszewski, N.E. (1996). SPINDLY, a tetratricopeptide repeat protein involved in gibberellin signal transduction in Arabidopsis. Proc. Natl. Acad. Sci. USA 93, 9292–9296. [PMC free article] [PubMed]
  • Jones, H.D., Smith, S.J., Desikan, R., Plakidou-Dymock, S., Lovegrove, A., and Hooley, R. (1998). Heterotrimeric G proteins are implicated in gibberellin induction of α-amylase gene expression in wild oat aleurone. Plant Cell 10, 245–253. [PMC free article] [PubMed]
  • Kuo, A., Cappelluti, S., Cervantes-Cervantes, M., Rodriguez, M., and Bush, D.S. (1996). Okadaic acid, a protein phosphatase inhibitor, blocks calcium changes, gene expression, and cell death induced by gibberellin in wheat aleurone cells. Plant Cell 8, 259–269. [PMC free article] [PubMed]
  • Lanahan, M.B., and Ho, T.-H.D. (1988). Slender barley: A constitutive gibberellin-response mutant. Planta 175, 107–114. [PubMed]
  • Lanahan, M.B., Ho, T.-H.D., Rogers, S.W., and Rogers, J.C. (1992). A gibberellin response complex in cereal α-amylase gene promoters. Plant Cell 4, 203–211. [PMC free article] [PubMed]
  • Leung, J., and Giraudat, J. (1998). Abscisic acid signal transduction. Annu. Rev. Plant Physiol. Plant Mol. Biol. 49, 199–222. [PubMed]
  • Li, J., Wang, X.Q., Watson, M.B., and Assmann, S.M. (2000). Regulation of abscisic acid–induced stomatal closure and anion channels by guard cell AAPK kinase. Science 287, 300–303. [PubMed]
  • Lovegrove, A., and Hooley, R. (2000). Gibberellin and abscisic acid signalling in aleurone. Trends Plant Sci. 5, 102–110. [PubMed]
  • Paz-Ares, J., Ghosal, D., Wienand, U., Peterson, P., and Saedler, H. (1987). The regulatory locus c1 of Zea mays encodes a protein with homology to myb proto-oncogene products and with structural similarities to transcriptional activators. EMBO J. 6, 3553–3558. [PMC free article] [PubMed]
  • Peng, J., et al. (1999). ‘Green revolution’ genes encode mutant gibberellin response modulators. Nature 400, 256–261. [PubMed]
  • Penson, S.P., Schuurink, R.C., Fath, A., Gubler, F., Jacobsen, J.V., and Jones, R.L. (1996). cGMP is required for gibberellic acid-induced gene expression in barley aleurone. Plant Cell 8, 2325–2333. [PMC free article] [PubMed]
  • Raventos, D., Skriver, K., Schelein, M., Karnahl, K., Rogers, S.W., Rogers, J.C., and Mundy, J. (1998). HRT, a novel zinc finger transcriptional repressor from barley. J. Biol. Chem. 273, 23313–23320. [PubMed]
  • Ritchie, S., and Gilroy, S. (1998. a). Abscisic acid signal transduction in the barley aleurone is mediated by phospholipase D activity. Proc. Natl. Acad. Sci. USA 95, 2697–2702. [PMC free article] [PubMed]
  • Ritchie, S., and Gilroy, S. (1998. b). Gibberellins: Regulating genes and germination. New Phytol. 140, 363–383.
  • Robertson, M., Swain, S.M., Chandler, P.M., and Olszewski, N.E. (1998). Identification of a negative regulator of gibberellin action, HvSPY, in barley. Plant Cell 10, 995–1007. [PMC free article] [PubMed]
  • Rogers, J.C., and Rogers, S.W. (1992). Definition and functional implications of gibberellin and abscisic acid cis-acting hormone response complexes. Plant Cell 4, 1443–1451. [PMC free article] [PubMed]
  • Schuurink, R.C., Chan, P.V., and Jones, R.L. (1996). Modulation of calmodulin mRNA and protein levels in barley aleurone. Plant Physiol. 111, 371–380. [PMC free article] [PubMed]
  • Sheen, J. (1996). Ca2+-dependent protein kinases and stress signal transduction in plants. Science 274, 1900–1902. [PubMed]
  • Shen, Q. (1996). Involvement of PP2C in the signal transduction of ABA action in barley aleurone cells. In Phosphorylation-Dephosphorylation of Plant Proteins: Current Topics in Plant Biochemistry, Physiology and Molecular Biology, D.D. Randall, ed (Columbia, MO: University of Missouri), pp. 42–43.
  • Shen, Q., Uknes, S.J., and Ho, T.-H.D. (1993). Hormone response complex of a novel abscisic acid and cycloheximide inducible barley gene. J. Biol. Chem. 268, 23652–23660. [PubMed]
  • Shen, Q., Zhang, P., and Ho, T.-H.D. (1996). Modular nature of abscisic acid (ABA) response complexes: Composite promoter units that are necessary and sufficient for ABA induction of gene expression in barley. Plant Cell 8, 1107–1119. [PMC free article] [PubMed]
  • Skriver, K., Olsen, F.L., Rogers, J.C., and Mundy, J. (1991). Cis-acting DNA elements responsive to gibberellin and its antagonist abscisic acid. Proc. Natl. Acad. Sci. USA 88, 7266–7270. [PMC free article] [PubMed]
  • Sun, T.-P. (2000). Gibberellin signal transduction. Curr. Opin. Plant Biol. 3, 374–380. [PubMed]
  • Sutliff, T.D., Lanahan, M.B., and Ho, T.-H.D. (1993). Gibberellin treatment stimulates nuclear factor binding to the gibberellin response complex in a barley α-amylase promoter. Plant Cell 5, 1681–1692. [PMC free article] [PubMed]
  • Wang, X., and Robinson, P.J. (1997). Cyclic GMP-dependent protein kinase and cellular signalling in the nervous system. J. Neurochem. 68, 443–456. [PubMed]

Articles from The Plant Cell are provided here courtesy of American Society of Plant Biologists
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...