A role for ethylene in the phytochrome-mediated control of vegetative development

Authors


*(fax +61 362262698; e-mail jim.weller@utas.edu.au).

Summary

Members of the phytochrome family of photoreceptors play key roles in vegetative plant development, including the regulation of stem elongation, leaf development and chlorophyll accumulation. Hormones have been implicated in the control of these processes in de-etiolating seedlings. However, the mechanisms by which the phytochromes regulate vegetative development in more mature plants are less well understood. Pea (Pisum sativum) mutant plants lacking phytochromes A and B, the two phytochromes present in this species, develop severe defects later in development, including short, thick, distorted internodes and reduced leaf expansion, chlorophyll content and CAB gene transcript level. Studies presented here indicate that many of these defects in phyA phyB mutant plants appear to be due to elevated ethylene production, and suggest that an important role of the phytochromes in pea is to restrict ethylene production to a level that does not inhibit vegetative growth. Mutant phyA phyB plants produce significantly more ethylene than WT plants, and application of an ethylene biosynthesis inhibitor rescued many aspects of the phyA phyB mutant phenotype. This deregulation of ethylene production in phy-deficient plants appears likely to be due, at least in part, to the elevated transcript levels of key ethylene-biosynthesis genes. The phytochrome A photoreceptor appears to play a prominent role in the regulation of ethylene production, as phyA, but not phyB, single-mutant plants also exhibit a phenotype consistent with elevated ethylene production. Potential interactions between ethylene and secondary plant hormones in the control of the phy-deficient mutant phenotype were explored, revealing that ethylene may inhibit stem elongation in part by reducing gibberellin levels.

Introduction

Light plays a central role in plant development, and the phytochrome photoreceptors are important regulators of a wide range of light responses. Phytochromes are well known for their role in seedling de-etiolation, with response to red light mediated by the phytochrome A (phyA) and phytochrome B (phyB)-type photoreceptors, and response to far-red light mediated solely by phyA (Neff et al., 2000; Takano et al., 2001; van Tuinen et al., 1995; Weller et al., 1995, 2000, 2001b; Whitelam et al., 1993). Phytochromes also play a role in seedling establishment under blue light, along with the cryptochrome and phototropin families of blue light-specific receptors (Liscum et al., 2003; Neff and Chory, 1998; Platten et al., 2005; Weller et al., 2001a,b).

Phytochromes also regulate later stages of plant development. For example, the phytochromes influence flowering in response to day length in photoperiodic species such as pea (Pisum sativum) and Arabidopsis (Childs et al., 1997; Franklin et al., 2003; Reed et al., 1993; Valverde et al., 2004; Weller et al., 1997, 2001b). In most species examined to date, the phytochromes also regulate vegetative development, including suppressing stem and petiole elongation and promoting leaf development. phyB and phyB-type mutants in a number of species, including pea, are characterized by elongated stems and reduced leaf development (Figure 1; Childs et al., 1997; Devlin et al., 1998, 1999; Lòpez-Juez et al., 1992; Reed et al., 1993; Weller et al., 1995, 2000, 2001b). Although mutant phyA plants in Arabidopsis, rice and tomato exhibit essentially wild-type (WT) vegetative development, analysis of plants lacking both phyA and phyB photoreceptors indicates that phyA also contributes to the suppression of stem elongation in Arabidopsis and tomato plants grown under white light (Devlin et al., 1996; Franklin et al., 2003; Takano et al., 2001; van Tuinen et al., 1995; Weller et al., 2000; Whitelam et al., 1993; J.L. Weller, unpublished data). In contrast to these other species, phyA mutants in pea are characterized by dramatically shorter, thickened internodes compared with WT plants (Figure 1; Weller et al., 1997, 2001b). This stem phenotype is even more severe in plants lacking both phyA and phyB (Weller et al., 2001b), the only two phytochromes that appear to be present in this species (Platten et al., 2005). The stems of phyA phyB mutant plants are pale, short and thickened, and severely distorted (Weller et al., 2001b; Figure 1). This unique phenotype identifies a potentially novel role for phytochromes in the promotion of stem elongation in older plants that contrasts with their well characterized role in inhibiting elongation under a range of light conditions.

Figure 1.

 Phenotype of WT and phytochrome-deficient mutant plants.
(a) Stem (internode 8) of untreated WT, phyA, phyB and phyA phyB mutant plants (leaves have been removed).
(b) Shoot tip of phyA phyB mutant plants untreated or following application of 100 μg of the ethylene-synthesis inhibitor aminoethoxyvinyl glycine (AVG).
(c) Internode length of WT and phytochrome-deficient mutants treated with 100 μg AVG or 10 μg gibberellin (GA3). Values are mean ± SE, n = 16–25.

We previously noted similarities between the internode phenotype of phytochrome-deficient mutant pea plants and the effects of ethylene (Weller et al., 2001b). Gibberellin (GA), indole-3-acetic acid (IAA), brassinosteroids (BR) and ethylene all influence stem elongation (Kende et al., 1998; Pierik et al., 2004; Ross et al., 2005), and have been reported to participate in light-regulated stem elongation. In de-etiolating seedlings of pea and Arabidopsis, GA has emerged as a key mediator of phytochrome-regulated stem elongation (Alabadíet al., 2004; Symons and Reid, 2003). In more mature plants, studies in a number of species have also highlighted potential roles for hormones, including ethylene, GA, IAA and BR, in light regulation of stem elongation, for example in stem elongation responses to shade (reviewed by Vandenbussche et al., 2005).

Here the roles of plant hormones in phytochrome-mediated vegetative development of adult pea plants were explored. Elevated ethylene production emerges from these studies as a major factor in the development of the phytochrome-deficient phenotype. We found little evidence for the involvement of BR or IAA, although ethylene may act, in part, through suppressing GA production. Our results suggest that in pea, ethylene may participate in a broad subset of light-regulated developmental responses including suppressing stem elongation, gene expression, chlorophyll accumulation and leaf development, and can be an important mediator of phytochrome-regulated development, at least under some circumstances.

Results

Phytochrome-deficient mutants of pea grown under glasshouse conditions are characterized by a number of changes in growth and development. Both phyA and phyA phyB mutants develop significantly shorter internodes than WT plants (P < 0.001; Figure 1a,c; Weller et al., 2001b), while the internodes of phyB plants are considerably longer than WT (Figure 1a,c; Weller et al., 1995). The internodes of phyA phyB mutant plants are also considerably thicker than WT, due to increases in the size of epidermis, cortex and stele cells, and are often severely distorted (Figure 1a; Weller et al., 2001b). Increased thickness of phyA phyB stems is reflected in a significant increase in stem fresh weight (FW) per unit length (Figure 2b). The extremely pale appearance of the internodes of phyA phyB plants (Figure 1a; Weller et al., 2001b) is due to a significant reduction in chlorophyll content in this tissue, and is mirrored by a dramatic reduction in the transcript level of the CAB9 gene (P < 0.001; Figure 2c,d), a representative of several light-regulated Lhcb1 genes in pea (White et al., 1992). Leaflet area is also considerably reduced in phyA phyB plants compared with WT plants (P < 0.001; Figure 2e).

Figure 2.

 Phenotype of 28-day-old WT and phyA phyB mutant plants treated with 100 μg aminoethoxyvinyl glycine (AVG) or 10 μg gibberellin (GA3).
(a) WT and phyA phyB mutant plants treated with AVG or GA3; (b) stem weight per unit length; (c) chlorophyll content of uppermost mature internode; (d) relative CAB9 transcript level in uppermost expanding internode; (e) leaflet area. Values are average ±SE, (b,c) n = 5; (e) n = 16–25. (d) Values normalized for actin, mean ± SE, n = 3.

Application of an ethylene-synthesis inhibitor rescues many aspects of the phyA phyB mutant phenotype

A number of hormones have been implicated in the control of stem and leaf development and regulation of the photosynthetic apparatus. We observed previously that the short, thickened stems of phyA phyB plants are reminiscent of plants treated with ethylene (Weller et al., 2001b), and a similar appearance has also been described for genotypes with elevated ethylene levels or constitutive ethylene response (Collett et al., 2000; Hall and Bleeker, 2003; Larsen and Chang, 2001; Ross and Reid, 1986; Sharp et al., 2000). Reduced leaf expansion, chlorophyll content and CAB transcript level have also been reported in plants treated with exogenous ethylene (Grbic and Bleecker, 1995; Ross and Reid, 1986; Zacarias and Reid, 1990).

To examine the role of ethylene in the development of phytochrome-deficient plants, we monitored the effect of applying the ethylene-synthesis inhibitor aminoethoxyvinyl glycine (AVG) to WT and phytochrome mutants. Treatment with AVG had little effect on the development of WT or phyB plants (Figures 1 and 2; data not shown). In contrast, AVG application had a dramatic effect on the development of phyA phyB plants, fully or partially rescuing many aspects of the mutant phenotype (Figure 1b). Application of AVG almost completely restored the length and radial expansion of phyA phyB internodes to WT (Figures 1c and 2b). Application of AVG also resulted in a significant fourfold increase in chlorophyll content of the internodes of phyA phyB plants (P < 0.05; Figure 2c) that was clearly visible to the naked eye. A similar increase in the transcript level of CAB9 was also observed in phyA phyB plants following AVG application (P < 0.05; Figure 2d). Leaflet expansion was also significantly increased in phyA phyB plants following AVG application (P < 0.05; Figure 2e). The internode length of phyA single-mutant plants was also almost fully restored to WT by application of the ethylene-synthesis inhibitor (Figure 1c, data not shown).

Ethylene production is elevated in phyA phyB plants

Given the partial rescue of the phyA phyB mutant phenotype by an ethylene-synthesis inhibitor, we next examined the possibility that phytochrome-deficient pea plants overproduce ethylene. The WT and phyA phyB plants were grown under glasshouse conditions until 12–14 leaves were expanded and the whole plant, together with the pot, was transferred to a perspex chamber, sealed and kept under glasshouse conditions for a further 24 h. A sample of headspace gas was removed, and ethylene evolution by the whole plant was determined by GC–MS. The phyA phyB mutant plants evolved approximately twofold more ethylene than WT plants of a similar developmental stage on a dry weight basis (P < 0.05; Figure 3). Headspace gas samples were also taken from chambers containing only pots filled with soil, and did not indicate any significant production of ethylene from non-plant sources.

Figure 3.

 Ethylene evolution from WT and phyA phyB mutant plants. Values are mean ± SE, n = 6.

Expression of ethylene-biosynthesis genes in phytochrome mutants

An important determinant of ethylene production in a number of species, including pea, is the transcript level of genes encoding enzymes in the ethylene biosynthesis pathway (Peck et al., 1998; Wang et al., 2002). Two key enzymes in this pathway are ACC synthase (ACS), which catalyses the production of the ethylene precursor ACC (1-aminocyclopropane-1-carboxylate) and ACC oxidase (ACO), which converts ACC to ethylene (Wang et al., 2002). In pea, two ACS and one ACO enzymes have been characterized and are expressed in expanding shoot tissue (Peck and Kende, 1998; Peck et al., 1998). The transcript level of all three genes was monitored in seedlings and in more mature WT and phytochrome mutant plants. Unlike ethylene measurements, which were necessarily taken at the whole-plant level, ethylene biosynthesis transcript levels were monitored only in young expanding shoot tissue, as in phyA phyB plants this tissue will go on to develop the characteristics associated with ethylene overproduction.

The transcript levels of both ACS1 and ACO1 were significantly higher in the expanding internodes of 46-day-old phyA phyB plants compared with WT plants (P < 0.05; Figure 4). The transcript levels of these ethylene-biosynthesis genes were up to 30-fold higher in phyA phyB plants compared with similar WT plants. Compared with WT plants, mature phyA plants also exhibited a significant 6-fold increase in ACS1 transcript level in apical tissue, and a similar increase in ACO1 transcript level in expanding internodes (P < 0.05). The ACS1 transcript level was also somewhat elevated in the expanding internode of mature phyB mutant plants compared with WT plants (P < 0.05). The expression of ACS2 did not differ significantly between genotypes (data not shown). Interestingly, the transcript levels of both ACS1 and ACO1 were also markedly elevated (up to 60-fold) in 7-day-old phyA phyB seedlings (Figure 4).

Figure 4.

 Relative transcript level of ethylene-biosynthesis genes in WT and phytochrome-deficient mutant plants.
(a) ACS1; (b) ACO1 transcript levels in apical tissue and expanding internode of 7-day-old seedlings and 46-day-old plants. Transcript levels are normalized to actin, values are mean ± SE, n = 2–3.

The role of cry1 in the phyA mutant internode phenotype

One interpretation of the phyA phyB phenotype is that it could result from the specific loss of phytochrome signalling. Alternatively, it could represent the more general effect of a reduction in the overall state of de-etiolation. We recently showed that the blue-light photoreceptor cryptochrome 1 (cry1) also plays a role in regulating internode elongation of white light-grown pea plants (Platten et al., 2005) and, as for phyB, the loss of cry1 in an otherwise WT background results in an increase in internode elongation. We therefore compared the effects of mutation of cry1 and phyB in a phyA mutant background. The loss of cry1 largely overcomes the short-internode phyA phenotype (Figure 5a; Platten et al., 2005) in contrast with the loss of phyB, which enhances it (Figure 1; Weller et al., 2001b). Similar opposing effects of phyB and cry1 on ACO1 transcript level in a phyA background were also observed (Figures 4 and 5b). While the ACO1 transcript level was elevated in the expanding internodes of both phyA and phyA cry1 mutant plants compared with WT, ACO1 transcript levels were reduced in phyA cry1 mutant plants approximately threefold compared with phyA plants (Figure 5b). Thus for both internode length and transcript level of an ethylene-biosynthesis gene, the effect of the phyA mutation was reduced in a cry1 mutant background and increased in a phyB mutant background (Figures 1, 4 and 5).

Figure 5.

 Phenotype and relative ACO1 transcript level of 35-day-old WT, phyA, cry1 and phyA cry1 plants.
(a) Internode length, mean ± SE, n = 15.
(b) Relative transcript level of ACO1 in apical tissue and expanding internodes of 35-day-old plants. Transcript levels are normalized to actin, values are mean ± SE, n = 2–3.

The phyA phyB mutant phenotype does not appear to be due to altered IAA or BR levels

Studies in a number of species have revealed that ethylene production can be regulated by other plant hormones, including IAA and BR. For example, elevated IAA has been shown to lead to elevated ethylene production in a number of species (Peck and Kende, 1998; Woeste et al., 1999; Yi et al., 1999). Elevated ethylene production has been observed in plants with both elevated and reduced BR levels (Arteca et al., 1988; Nomura et al., 2004; Ross and Reid, 1986; Woeste et al., 1999). However, studies presented here indicate that it is unlikely that the ethylene overproduction in phyA phyB plants is due to changes in the endogenous level of IAA or BR. In contrast to the prediction that IAA levels may be elevated in phyA phyB plants, endogenous IAA levels are actually somewhat reduced in the expanding internodes of phyA phyB plants (Figure 6). Furthermore, as AVG did not have a clear effect on endogenous IAA levels in WT or phyA phyB plants, it seems unlikely that ethylene influences the development of phyA phyB plants through changes in IAA levels (Figure 6). Similarly, BR does not appear to play a central role in the development of phyA phyB plants, as application of either the bioactive BR, brassinolide, or the BR synthesis inhibitor brassinazole to phyA phyB plants did not have a significant effect on development (data not shown).

Figure 6.

 Gibberellin (GA) and indole-3-acetic acid (IAA) levels in expanding internode of 38 d old WT and phyA phyB plants treated with aminoethoxyvinyl glycine (AVG). Values are mean ± SE, n = 3.

An interaction between ethylene and GA in the development of phyA phyB plants

A number of studies indicate that ethylene may itself regulate growth by secondarily influencing another plant hormone, GA. Ethylene is thought to influence apical hook formation in Arabidopsis and stem elongation in several semi-aquatic plants by regulating GA levels and/or response (Kende et al., 1998; Rijnders et al., 1997; Vriezen et al., 2004). Similarly, in tobacco ethylene is unable to promote stem elongation in the presence of GA-synthesis inhibitors (Pierik et al., 2004), suggesting that ethylene action requires GA.

We examined the possibility that ethylene may influence endogenous GA levels in phyA phyB plants. Endogenous GA levels were examined in the expanding internodes of WT and phyA phyB plants treated with AVG. The expanding internodes of untreated phyA phyB plants contained significantly lower levels of the bioactive GA (GA1) than found in comparable WT tissue (P < 0.05; Figure 6). Untreated phyA phyB mutant plants also contained significantly less of the direct precursor to GA1 (GA20; P < 0.001). In contrast, the level of GA8, the inactive 2-oxidation product of GA1, and of GA29, the 2-oxidation product of GA20, did not differ significantly between WT and phyA phyB plants (data not shown). Treatment with AVG did not have a significant effect on GA levels in WT plants. However, blocking ethylene synthesis resulted in a fourfold increase in GA1 levels in phyA phyB plants, restoring GA1 in mutant plants to a level not significantly different from WT plants. This increase in GA1 levels in AVG-treated phyA phyB plants does not appear to be due to up-regulation of GA20 production, as GA20 levels were essentially unchanged following AVG application. In addition, there is no evidence that ethylene influences 2-oxidation of GA1, as AVG application did not have a clear effect on GA8 levels in phyA phyB plants (data not shown).

In pea, GA1 is an important positive regulator of internode elongation, and the restoration of GA1 levels in AVG-treated phyA phyB plants indicates that at least part of the short-internode phenotype of phyA phyB plants may be due to altered GA levels. To examine the effect of restoring GA levels in phyA phyB plants, WT and phyA phyB plants were treated with GA3 (Figures 1c and 2). Like AVG application, GA3 treatment restored internode length and stem FW per unit length of phyA phyB plants to near WT levels (Figures 1c and 2a,b). Application of GA3 also resulted in a small increase in internode length in WT plants (Figures 1c and 2a). However, GA3 treatment failed to prevent distortion of the stems of phyA phyB plants (Figure 2a). In addition, GA3 application had little effect on the chlorophyll content or CAB9 transcript level in the internodes of phyA phyB plants, and actually inhibited leaflet expansion in both WT and phyA phyB plants compared with untreated plants (Figure 2c–e).

Discussion

Light perception is essential for appropriate plant growth and development at all stages of the life cycle, and studies with phytochrome-deficient mutants in a number of species have revealed important roles for individual phytochromes in a range of light responses (Neff et al., 2000; Takano et al., 2001; van Tuinen et al., 1995; Weller et al., 1995, 2000, 2001b; Whitelam et al., 1993). Analysis of mutants lacking more than one phytochrome has also been crucial in highlighting roles for the phytochromes in plant development, including the maintenance of the rosette habit of Arabidopsis and control of inflorescence and fruit development in tomato (Devlin et al., 1996, 1998; Weller et al., 2001a). To date, pea and Arabidopsis are the only two species in which mutant combinations that completely lack phytochrome function can be constructed, and studies with these plants potentially provide new insights into the role of the phytochromes in plant development. In pea, mature phy-deficient phyA phyB mutant plants exhibit a striking vegetative phenotype. The phyA phyB double mutant of pea shares many of the characteristics of multiple phytochrome-deficient mutants of Arabidopsis, such as reduced leaflet expansion, chlorophyll content and CAB transcript level, but also develops a striking vegetative phenotype later in development, with a short, thickened stem unique to pea.

Here we report that ethylene overproduction plays a central role in the development of the phyA phyB mutant phenotype, not only in the control of stem elongation, but also in the regulation of the photosynthetic apparatus. This provides evidence that ethylene may participate in a broad subset of phytochrome-regulated processes, and suggests that an important role of the phytochromes in the vegetative development of white light grown pea is to restrict ethylene production.

Ethylene production is elevated in glasshouse-grown phyA phyB plants (Figure 3), and clearly contributes to many aspects of the phyA phyB mutant phenotype. Elevated ethylene production appears to be the major factor contributing to the short, distorted internodes of phyA phyB plants, as application of the ethylene-synthesis inhibitor AVG fully restored elongation and radial expansion of phyA phyB internodes to WT (Figures 1 and 2). Indeed, application of ethylene to WT pea plants results in dramatically shortened internodes (Ross and Reid, 1986; data not shown), and increased radial expansion and some stem distortion is observed in pea phyB mutant plants following ethylene application (data not shown). Similar reductions in shoot elongation and increases in radial swelling of the stem have also been associated with elevated ethylene levels and/or responsiveness in other species (Collett et al., 2000; Hall and Bleecker, 2003; Larsen and Chang, 2001; Sharp et al., 2000). Ethylene also contributes to the reduction in chlorophyll content, CAB transcript level and leaflet expansion in phyA phyB plants, as AVG application had clear positive effects on these processes in mutant plants (Figure 2). This is consistent with previous reports of a reduction in leaf expansion in plants with elevated ethylene level and/or response, including pea (Hall and Bleecker, 2003; Ross and Reid, 1986; Vandenbussche et al., 2003); and reductions in chlorophyll content and CAB transcript level in Arabidopsis plants following ethylene application (Grbic and Bleecker, 1995; Zacarias and Reid, 1990). However, factors independent of ethylene may also play a role, as AVG was unable fully to restore these aspects of phyA phyB phenotype to WT levels, although this may also be attributable to an incomplete restoration of normal ethylene production in all tissues of AVG treated plants.

Ethylene levels are strongly influenced by the transcript level of biosynthesis genes ACS and ACO (Peck et al., 1998), and the elevated ethylene levels of mature phyA phyB mutant plants are probably due, in part, to the elevated transcript level of ACO1 and ACS1 in expanding shoot tissue (Figure 4). Indeed, expression of these two genes in expanding internodes of mature phyA phyB plants was up to 30-fold higher than observed in WT plants. In contrast, ethylene production from whole adult phyA phyB plants was only approximately double that of WT (Figure 4). This apparent disparity between increases in transcript level and ethylene production could be due to the fact that transcript level was measured in expanding shoot tissue and ethylene production was measured at the whole-plant level. If ethylene overproduction in phyA phyB plants is most pronounced in certain tissues, such as the expanding shoot, large increases in ethylene production in this tissue may not lead to large differences in ethylene production at the whole-plant level.

Suppression of ethylene production by the phytochrome system in pea is consistent with reports of phytochrome action in a number of other species. Red light has been shown to down-regulate ethylene production in etiolated pea, rice, Impatiens balsamina and bean seedlings, and in some cases this can be reversed by far-red light (Imaseki et al., 1970; Michalczuk and Rudnicki, 1993; Steed et al., 2004; Vangronsveld et al., 1988). Elevated ethylene levels have also been reported in sorghum and tobacco plants grown under far-red enriched and/or low light (Finlayson et al., 1998; Pierik et al., 2004).

In pea, several pieces of evidence indicate that phyA may play a prominent role in the regulation of ethylene levels. The short-internode phenotype of phyA mutant plants was clearly rescued by AVG application (Figure 1c), and the transcript levels of several ethylene-biosynthesis genes are elevated in the expanding internodes of phyA mutant plants (Figure 4). In Arabidopsis and sorghum, where ethylene can promote stem elongation, an important role for phyB in suppressing ethylene production has emerged. The phyB mutants in these species produce more ethylene than WT plants (Finlayson et al., 1998; Vandenbussche et al., 2003), although it is not clear if elevated ethylene contributes to the partially etiolated phenotype of these plants. In contrast, we have little evidence that ethylene levels are significantly elevated in single phyB mutant pea plants. As ethylene is known to inhibit stem elongation in pea (Ross and Reid, 1986), elevated ethylene levels are not consistent with the elongated internodes of pea phyB plants (Figure 1a,c; Weller et al., 1995). In addition, AVG application had little effect on stem elongation of phyB mutant plants (Figure 1c). The severe phenotype of phyA phyB double-mutant pea plants does indicate, however, that a lesion in phyB contributes to the development of the ethylene-related double-mutant phenotype.

Comparisons with the cry1 mutant have given two additional perspectives on the phyA phyB short-internode phenotype. First, they show that this phenotype is not simply the indirect effect of a general reduction in the state of de-etiolation, because the phyA cry1 mutant actually has longer internodes than the phyA mutant (Figure 5a; Platten et al., 2005). Second, they show that the development of the ethylene-related short-internode phenotype of phyA is largely dependent on the presence of cry1. This could be interpreted as an antagonistic effect of phyA on cry1-mediated inhibition of elongation, and the short-internode phyA and phyA phyB phenotypes as the result of enhanced cry1-dependent responsiveness to blue light. We are currently investigating this possibility using artificial light sources. The reduction in internode length in phyA mutant plants and associated increase in ACO1 transcript level was reduced in a cry1 mutant background (Figure 5), consistent with a negative association between elongation and ethylene-biosynthesis gene expression, suggesting that cry1 may actually promote ethylene biosynthesis.

A similar interaction between the phytochromes and cry1 has been observed in pea seedlings (Platten et al., 2005; Weller et al., 2001b). Mutant phyA seedlings grown under red and far-red light have longer internodes than corresponding PHYA plants, whereas under high irradiance blue or white light, they can actually have shorter internodes. This apparent blue light-specific promotion of internode elongation by phyA is evident in either a WT or a phyB background, but not in a cry1 background, where phyA clearly inhibits elongation (Platten et al., 2005). As with older plants, this suggests that in seedlings, phyA may also act, in part, by antagonizing cry1 action, a specific interaction not reported previously from any other system. Whether this similarity between seedlings and more mature plants extends to the involvement of ethylene in regulating stem elongation requires examination, especially given the elevated transcript levels of several ethylene-biosynthesis genes in phyA phyB seedlings (Figure 5).

A number of studies have indicated that ethylene may interact with other hormones, including IAA, GA and BR, to regulate aspects of growth and development (Peck and Kende, 1998; Ross and Reid, 1986; Vriezen et al., 2004). Studies presented here indicate that IAA and BR do not appear to play a central role in the development of the phenotype of pea phyA phyB mutant plants. However, at least part of the effect of elevated ethylene on stem elongation of phyA phyB plants may be mediated through a reduction in GA1. Low GA1 levels in the expanding internodes of phyA phyB were restored to WT levels by AVG application (Figure 6), and application of exogenous GA to phyA phyB plants restored internode length and width to WT levels (Figures 1c and 2a,b). Indeed, application of ethylene to WT pea plants, which reduces internode elongation (Ross and Reid, 1986; data not shown), also reduces GA1 levels (data not shown). However, unlike AVG application, GA application had little effect on other aspects of the phyA phyB mutant phenotype, including stem distortion, leaflet expansion, CAB transcript level and chlorophyll content (Figure 2a,c–e). Therefore ethylene clearly influences many aspects of the growth of phyA phyB plants independently of GA.

It is important to note that application of an ethylene-synthesis inhibitor to WT plants had little effect on the processes that were monitored (Figures 1, 2 and 6). Although elevated ethylene production clearly inhibited stem elongation, leaflet expansion, chlorophyll content, CAB transcript level and GA1 levels in phyA phyB plants, application of AVG to WT plants had almost no effect on these processes. This suggests that the ethylene levels produced in WT plants under these conditions do not limit the growth or GA1 content of WT plants. A number of other studies have indicated that baseline ethylene levels produced in WT plants may not limit vegetative growth. For example, although elevated ethylene levels or sensitivity have been shown to inhibit leaf expansion (Hall and Bleecker, 2003; Ross and Reid, 1986; Vandenbussche et al., 2003), Tholen et al. (2004) reported that transgenic Arabidopsis, tobacco and petunia plants with reduced ethylene sensitivity had a similar leaf area and relative growth rate to WT plants when grown in a well ventilated area. Similarly, ethylene-insensitive mutants in tobacco and tomato have also been reported to display relatively normal leaf growth under selected conditions (Pierik et al., 2004; Whitelaw et al., 2002).

In white light-grown pea, a functional phytochrome system appears to be essential to maintain ethylene at levels that do not disrupt vegetative development. In pea, stem elongation and radial expansion, CAB transcript level, chlorophyll accumulation and leaflet expansion are regulated by the phytochromes, at least in part, through suppressing ethylene production. It will be of interest to investigate the role of ethylene in the phenotype of phytochrome-deficient mutants in other species (including Arabidopsis, where ethylene can act as a promoter of stem elongation) to determine if this is a general role for phytochromes in the regulation of vegetative plant development.

Experimental procedures

Plant material and growth conditions

Pea lines were derived from cv. Torsdag. The phyA, phyB and phyA phyB mutants have been described by Weller et al. (1997, 2001b), and the cry1 and phyA cry1 mutants are described by Platten et al. (2005). All plants were grown in a 1:1 mixture of dolerite chips and vermiculite topped with potting mix and, if grown to maturity, received nutrient solution weekly. Plants were grown under glasshouse conditions as described previously (Weller et al., 1997) under an 18-h photoperiod comprising a natural photoperiod extended before dawn and after dusk with mixed white fluorescent and white incandescent lights.

Application studies

Aminoethoxyvinyl glycine (Sigma, St Louis, USA) was applied at a rate of 100 μg per plant in 5 μl 80% ethanol; GA3 was applied at a rate of 10 μg per plant in 10 μl 100% ethanol. For experiments shown in Figures 1 and 2, AVG and GA3 were applied to the uppermost expanded leaf of plants with five to six leaves expanded. For the experiment shown in Figure 6, AVG was applied to the uppermost expanded leaf plants with seven to nine leaves expanded. A similar volume of 100% ethanol was applied to control plants.

For experiments shown in Figures 1 and 2, various characteristics were measured 1 month after treatment. For chlorophyll measurements, a 2-cm segment was removed from the uppermost fully expanded internode and weighed, and chlorophyll was extracted by incubating in dimethyl sulfoxide overnight at 60°C. Chlorophyll content was determined from the absorbance at 649 and 665 nm according to Hiscox and Israelstam (1979). Leaflet area was estimated as the length × width of the largest leaflet three nodes above the site of application. Internode length was measured once the plants had matured.

Gene-expression studies

Apical tissue and the uppermost expanding internode (approximately 20–30% expanded) was collected from six to eight plants into two to three pools for each genotype, and frozen in liquid nitrogen. RNA was extracted from approximately 100 mg tissue with RNeasy Mini Kit (Qiagen, Hilden, Germany). The RNA was quantified by spectrophotometry, and cDNA was synthesized from 4.5 μg RNA with Superscript II (Invitrogen, Carlsbad, USA) according to the manufacturer's instructions. cDNA was diluted and real-time PCR reactions were carried out with Dynamo SYBR Green Master Mix (Geneworks, Hindmarsh, Australia) in a Rotor Gene 2000 (Corbett, San Francisco, USA). PCR was carried out with 100–200 pmol of each primer. Primers were designed to flank an exon–exon boundary; PsACS1 F (5′-TCT CGC CGA AAA TCA GCT GT-3′); PsACS1 R (5′-GGC TTC TCC TTT TTC AGC AAG A-3′); PsACO1 F (5′-TTG GTG ATC CAC TCG AAG TGA T-3′); PsACO1 R (5′-TGG AAC TTG AGT CCC ATT GAG A-3′). PsCAB9 primers were designed to amplify this representative of the light-regulated Lhcb1 genes in pea (White et al., 1992), and are described by Platten et al. (2005). Actin was monitored as a control using primers described by Foo et al. (2005). For ACS1, ACO1 and actin, PCR was carried out under the following conditions: 94°C, 15 min; 60 cycles of 94°C, 15 sec; 58–59°C, 20 sec; 72°C, 30 sec; 75°C, 15 sec. For CAB9, PCR was carried out under the following conditions: 94°C, 3 min; 50 cycles of 94°C, 5 sec; 52°C, 40 sec. Relative transcript levels were calculated as described by Foo et al. (2005). Briefly, for each sample and gene of interest, duplicate PCR reactions were performed and the average PCR cycle number at which the fluorescence reached a preset threshold (CT) was recorded. A normalized CT value was obtained by comparing with the CT value of actin for this sample. A mean normalized CT was then calculated from biological replicates and presented together with the corresponding standard error.

Hormone analysis

For ethylene analysis, WT and phyA phyB plants were grown two per pot under glasshouse conditions until 12–14 leaves expanded, then placed in sealed 25-l perspex chambers. After a further 24 h under glasshouse conditions, three replicate 500-μl samples of headspace gas were withdrawn with a gas-tight syringe. As only four chambers were available, two pots of each genotype were analysed on three different days and the values in Figure 3 are average ethylene levels from all six replicates.

GC–MS was performed using a Hewlett Packard 5890 GC coupled to a Kratos Concept ISQ mass spectrometer (described by Hasan et al., 1994). Acetylene gas (0.4 nl, 4 ppm in 100 μl) was mixed with each sample to account for variation between GC–MS runs. Splitless injections (600 μl) were made onto a cryo-trap made from a 1-m length of megabore HP-1 column (J&W Scientific, Folsom, CA, USA). The cryo-trap was placed in liquid N outside the oven and connected to a 30-m × 0.53-mm internal diameter GC-Q PLOT column (J&W Scientific) held inside the oven at 60°C. Two minutes after injection, the trap was removed from the liquid N to release the hydrocarbon gases and the MS acquisition commenced. The He carrier gas had a head pressure of 7 psi and the system was operated in medium resolution (R = 2000, 10% valley definition), selected ion-monitoring (SIM) mode. Ethylene and acetylene were resolved by the column and the ions monitored were 27.0235 and 28.0313 (ethylene); and 26.0156 (acetylene). A standard curve of pure ethylene samples of known concentration were analysed each time samples were run.

For IAA and GA analysis, the uppermost expanding internode (approximately 20–30% expanded) of control and AVG-treated 38-day-old WT and phyA phyB plants was excised, weighed and placed into ice-cold 80% MeOH with butylated hydroxytoluene. For each genotype and treatment combination, internodes from four to seven plants were harvested into three pools. Appropriate amounts of GA and IAA standards were added to each sample, and IAA and GA were extracted and analysed as described by Jager et al. (2005) and are expressed as ng per g FW.

Acknowledgements

We thank Mike Oates for technical assistance and Ian Cummings and Tracey Winterbottom for plant husbandry. We also thank Professor L. M. Mander for labelled GAs, and the Australian Research Council for financial support.

Ancillary