Plant growth in dense vegetation can be strongly affected by competition for light between neighbours. These neighbours can not only be detected through phytochrome-mediated perception of a reduced red:far-red ratio, but also through altered blue light fluence rates. A reduction in blue light (low blue) induces a set of phenotypic traits, such as shoot elongation, to consolidate light capture; these are called shade avoidance responses. Here we show that both auxin and brassinosteroids (BR) play an important role in the regulation of enhanced hypocotyl elongation of Arabidopsis seedlings in response to blue light depletion. Only when both hormones are experimentally blocked simultaneously, using mutants and chemical inhibitors, will the response be fully inhibited. Upon exposure to low blue several members of the cell wall modifying XYLOGLUCAN ENDOTRANSGLUCOSYLASE/HYDROLASE (XTH) protein family are regulated as well. Interestingly, auxin and BR each regulate a subset of these XTHs, by which they could regulate cell elongation. We hypothesize that auxin and BR regulate specific XTH genes in a non-redundant and non-synergistic manner during low-blue-induced shade avoidance responses of Arabidopsis seedlings, which explains why both hormones are required for an intact low-blue response.
Plants often grow in dense vegetation where shade-intolerant plants compete for available light as the canopy progressively closes through the growing season. Plants have evolved various phenotypically plastic traits to help sustain light capture and avoid being overgrown, and thus shaded, by neighbouring plants. These traits make up the so-called shade avoidance syndrome (SAS) and include enhanced elongation of stems and petioles, upward leaf movement (hyponasty) and increased apical dominance (Ballaré, 1999; Franklin, 2008; Keuskamp et al., 2010a).
The shade avoidance responses are induced upon detection of neighbouring competitors. A classic signal is the reduced ratio between red (R) and far-red (FR) light (low R:FR) that results from the selective absorption of R light for photosynthesis and the reflection of FR light. This change in the R:FR ratio is sensed by the phytochrome photoreceptors (Ballaréet al., 1990). When plants continue to grow, the canopy closes and actual shading occurs. In addition to red light, blue light is also depleted, as both are absorbed in equal amounts by chlorophyll to fuel photosynthesis (Ballaré, 1999; Vandenbussche et al., 2005).
Here, we study the hormonal control of low-blue-induced hypocotyl elongation in Arabidopsis seedlings, a response that is controlled through combined regulation of CRY1 and CRY2 (Pierik et al., 2009). We show that auxin transport and signalling constitute an important control point of this response. However, this is different from the low-R:FR-induced shoot elongation since inhibition of the auxin pathway does not inhibit the entire elongation response. We demonstrate that next to auxin, BR also regulate this elongation response to blue light depletion and that both hormones target XTH genes. We suggest that the combined regulation of auxin and BR is the main route through which low-blue-induced shade avoidance in Arabidopsis seedlings is controlled.
The reduction of blue light fluence rates (low blue), achieved by filtering blue light out of standard white light with the use of a yellow filter (Pierik et al., 2009), stimulated Arabidopsis hypocotyl elongation by about threefold (Figure 1a). Epidermal cell length measurements indicated that this elongation response to low blue light involved enhanced cell expansion and probably not cell proliferation, although the latter was not directly measured (Figure 1b,c). Furthermore, the strongest elongation was observed in the middle of the hypocotyls (Figure 1b).
We showed previously that polar auxin transport (PAT) partly regulates low-blue-induced hypocotyl elongation (Pierik et al., 2009), which is confirmed here by using the PAT inhibitor 1-naphthylphthalamic acid (NPA) (Figure 2a). We investigated whether the auxin transport-associated PIN-FORMED (PIN) proteins regulate low-blue-induced hypocotyl elongation. We tested mutants for PINs that are known to be expressed in the hypocotyl, and of these only pin3-3 and pin7 showed a reduced, but not absent, low-blue response (Figure 2a). The pin3-3 pin7 double mutant showed the same extent of reduction in the low-blue response as did pin3-3, indicating that these two PINs do not act redundantly. Reporter lines (promoter:GUS) for these PINs were used to analyze the expression of these genes in response to low-blue conditions. Both pPIN3:GUS and pPIN7:GUS showed enhanced GUS staining throughout the hypocotyl in response to low-blue treatment (Figure 2b–i) indicating increased expression of these two genes, located particularly to the stele.
Enhanced auxin biosynthesis through the TAA1 pathway is required for low-R:FR-induced hypocotyl elongation (Tao et al., 2008), as is auxin perception through the TRANSPORT INHIBITOR RESPONSE1 (TIR1) auxin receptor (Keuskamp et al., 2010b). Figure 3 shows that both TAA1-induced auxin biosynthesis (knocked out in wei8-1) and the auxin receptor family TRANSPORT INHIBITOR RESPONSE1/AUXIN-BINDING F-BOX PROTEIN (TIR1/AFB) (knocked out in tir1-1 and tir1afb1afb2afb3) are also needed for the complete low-blue response found in wild-type Col-0 seedlings. However, both mutants still retained a significant low-blue response. The auxin receptor inhibitor α-(phenylethyl-2-one)-IAA (PEO-IAA) was used to block auxin perception through TIR1 and its homologues (Hayashi et al., 2008). Seedlings treated with PEO-IAA had a reduced but not completely absent low-blue response (Figure 3a). Consistent with these observations, low-blue stimulated the pIAA19:GUS signal intensity, which is indicative of increased auxin action and which was completely abolished when treated with PEO-IAA under low blue (data not shown). Taken together, these data confirm that auxin is up-regulated upon low-blue exposure and contributes to the observed hypocotyl elongation response to this light treatment. In addition, we show that other regulators are also likely to be involved since blocking auxin production, transport or signalling could never fully block the low-blue response.
Brassinosteroids control low-blue-induced hypocotyl elongation
Despite the overlap between auxin and BR regulatory pathways, distinct differences remain (Goda et al., 2004; Nemhauser et al., 2004). We tested whether BRs were required for low-blue-induced hypocotyl elongation and found that the BR receptor mutant bri1-1 (brassinosteroid insensitive 1) (Wang et al., 2001) had a strongly reduced low-blue response (Figure 4a). Furthermore, inhibition of BR biosynthesis, as in the rot3-1 (rotundifolia 3) mutant (Tsuge et al., 1996) or via the application of brassinazole (Brz), an inhibitor of BR biosynthesis (Asami et al., 2000), led to impaired low-blue-induced hypocotyl elongation (Figure 4). Importantly, and similar to the findings for auxin inhibition, neither in bri1-1 and rot3-1 nor in seedlings treated with Brz was the low-blue response completely absent (Figure 4).
Auxin and brassinosteroids are both required for low-blue-induced hypocotyl elongation
Both exogenously applied auxin (namely IAA) and BR (namely epibrassinolide, BL) stimulated hypocotyl elongation under control white light conditions in a dose-dependent manner (Figure S1). However, these single hormone treatments at saturating concentrations did not induce an elongation response that was as strong as the low-blue response. Interestingly, when applied simultaneously, BL and IAA stimulated hypocotyl elongation to a much larger extent than either hormone alone, leading to hypocotyl lengths that were remarkably similar to those of low-blue-exposed seedlings (Figure 5). This suggests that the combined action of these two hormones suffices to induce the elongation response seen under low-blue conditions.
Since the inhibition of auxin or BR biosynthesis and/or perception could only partly reduce low-blue-induced hypocotyl elongation, we tested whether their combined inhibition would completely block the elongation response to low blue. Col-0 seedlings still showed a (reduced) low-blue response when treated separately with brassinazole (Brz) or PEO-IAA as shown in Figures 3 and 4. When Brz and PEO-IAA were applied simultaneously, the hypocotyl length in low blue was reduced to that observed in control white light conditions (Figure 6a). Furthermore, treatment of the auxin mutants wei8-1, pin3-3 pin7 and tir1afb1afb2afb3 or NPA-treated Col-0 seedlings with Brz resulted in a complete lack of elongation response to low blue (Figure 6b). Similarly, BR mutants (bri1-1 and rot3-1) treated with PEO-IAA showed no response to low blue at all (Figure 6c). These data imply that the combined action of auxin and BR is required and sufficient for low blue-induced hypocotyl elongation.
Auxin and BR regulate partly similar and partly different XTH genes during low-blue-induced hypocotyl elongation
Both auxin and BR are known to regulate XTH expression (e.g. Yokoyama and Nishitani, 2001; Goda et al., 2004; Nemhauser et al., 2004; Sun et al., 2010). We examined whether XTH genes could be targets for auxin and BR during low-blue-induced hypocotyl elongation in order to understand how these two hormones are both required for the elongation response to occur. We first measured XTH enzyme activity (XDA; xyloglucan degrading activity) and found that low blue caused a strong increase in XDA (Figure 7). Those XTHs that are known to be light regulated were selected with the use of Genevestigator (Zimmermann et al., 2004) and tested for transcriptional regulation in low blue. We identified eleven XTH genes that were differentially regulated in low blue relative to control (Table 1). Next, we studied whether auxin and BR were functional to XTH regulation in low blue and if these two hormones had specific XTH targets under low-blue conditions (Figure 8).
Table 1. The relative transcript abundance (log2 scale) of different XTHs in seedlings after 1 day of low-blue treatment relative to 1 day in control light conditions
Log2 fold change
*Significant differences relative to the expression of control treated seedlings (P <0.05).
NS, not significant.
Data are mean ± SE (n =3–4).
The expression in low blue light of XTH2, XTH3, XTH9 and XTH25 was more strongly affected by BR inhibition (Brz) than by auxin inhibition (PEO-IAA), whereas the expression of XTH13, XTH21 and XTH33 was more strongly affected by auxin inhibition than by BR inhibition. This indicates that different XTHs respond differently to auxin versus BR control during low-blue-induced hypocotyl elongation. Some of the low-blue-induced XTHs (XTH15 and XTH16) were not affected at all by Brz or PEO-IAA, whereas XTH8 and XTH17 were affected to the same extent by the two inhibitors. Interestingly, only two XTH genes (XTH8 and XTH16) showed a stronger sensitivity to the combined inhibition of auxin and BR, relative to inhibition of either pathway alone.
We studied the regulation of hypocotyl elongation in response to blue-light depletion in Arabidopsis seedlings and found that the combined action of auxin and BR could account for the complete seedling response to low blue. Addition of exogenous IAA and BL to seedlings grown in control light strongly stimulated hypocotyl elongation, much more than either hormone alone. Furthermore, simultaneous inhibition of the auxin and BR pathways led to a complete inhibition of the low-blue response, whereas hypocotyl length under control light was not affected by these treatments. This joint regulation of a shade avoidance response by BR and auxin is comparable to the recent observations on low-R:FR-induced petiole elongation (Kozuka et al., 2010). Interestingly, these control mechanisms are different from low-R:FR-induced hypocotyl elongation or low-blue-induced petiole elongation (Keuskamp et al., 2010b; Keller et al., 2011), indicating that different combinations of developmental stage and light quality recruit different combinations of physiological regulators of plasticity.
[ Schematic representation of the signal transduction steps in the process of low blue induced shade avoidance syndrome. Auxin and brassinosteroids (BR) are regulated upon low-blue treatment and mediate the expression of various XTH genes in order to stimulate hypocotyl elongation. Each of these hormones regulate the expression of a subset of XTHs. XTHs are listed under auxin if 1-naphthylphthalamic acid (NPA) significantly affected their expression in low blue, and under brassinosteroids if brassinazole (Brz) significantly affected their expression in low blue. XTHs are listed under combined auxin and brassinosteroid control if combined Brz and NPA treatment gave a stronger effect than either inhibitor alone. XTHs in bold indicate XTHs that are regulated specifically in that pathway alone. This results in hypocotyl elongation which is one of the shade avoidance responses. ]
Auxin and brassinosteroids both regulate XTHs in their control of developmental plasticity
The XTHs were recently shown to regulate shade avoidance responses of Arabidopsis petioles to low R:FR, and accordingly XTH enzymatic activity and gene expression were enhanced by low R:FR (Sasidharan et al., 2010). It was not yet known if XTHs were regulated by blue light depletion nor was it known how such regulation would depend on plant hormones. Here we show that both auxin and BR are needed for the full response of seedlings to low blue. We also show that BR and auxin act together to induce hypocotyl elongation and that, despite the transcriptional overlap (Goda et al., 2004; Nemhauser et al., 2004), many XTHs are pre-dominantly regulated by one of the two hormones during the low-blue treatment. Only a few are regulated by both hormones similarly (e.g. Yokoyama and Nishitani, 2001; Goda et al., 2004; Nemhauser et al., 2004; Sun et al., 2010). Interestingly, when the low-blue treatment was combined with a double treatment of both Brz and PEO-IAA, inhibitors of BR and auxin, respectively, the inhibitor-induced suppression of XTH gene expression showed an additive effect only in the case of XTH8. For most others, the XTH suppression by combined Brz and PEO-IAA treatment under low-blue-light conditions was at the same level as for the seedlings treated only with PEO-IAA, irrespective of the effect of the Brz-only treatment. This apparently dominant effect of auxin inhibition is not likely to be due to a hypothetically higher efficacy of the auxin inhibitor relative to the BR inhibitor, since for at least four XTH genes Brz gave a stronger effect than PEO-IAA (Figure 7). Only two XTH genes (XTH8 and XTH16) showed an interaction effect of the combined treatment with Brz and PEO-IAA compared with the single treatments with these inhibitors.
Although auxin and BR are thought to act synergistically (e.g. Nemhauser et al., 2004), the XTH expression data shown here do not suggest such a synergistic interaction between auxin and BR under blue light depletion. In fact, we show that the two hormones regulate specific subsets of XTH gene family members, which could explain why the regulated activity of both hormones is required for a complete hypocotyl elongation response to blue light depletion (Figures 5 and 6a). Auxin and BR are involved in other shade avoidance responses as well (e.g. Keuskamp et al., 2010b; Kozuka et al., 2010), but different XTHs are regulated during these responses and the specific regulation of XTHs therefore seems to be dependent on light quality and plant organ (Devlin et al., 2003; Kozuka et al., 2010; Sasidharan et al., 2010).
We thus hypothesize that auxin and BR regulate specific XTH genes in a non-redundant and non-synergistic manner. Future studies testing this hypothesis should include functional testing of the importance of different XTHs from both the auxin- and BR-regulated subsets, using single and multiple xth knockouts. The hypothesis predicts that XTHs from both the auxin and BR subsets should be knocked out simultaneously in order to maximally inhibit low-blue-induced hypocotyl elongation. Alternatively, low-blue-induced XTHs that did show a stronger suppression by the combined inhibition of auxin and BR action, notably XTH8, could be dominant regulators of the elongation response. It should be noted that, in addition to XTHs, various other targets may be controlled by the two hormones in order to stimulate elongation growth. Such targets for growth control might include other cell wall modifying proteins like expansins (e.g. Goda et al., 2004; Nemhauser et al., 2004, 2006; Guo et al., 2009; Park et al., 2010). In Arabidopsis petioles, expansin activity was not enhanced during low-R:FR-induced shade avoidance (Sasidharan et al., 2010), but it remains possible that hypocotyls do show such regulation.
Auxin and brassinosteroids versus other hormone pathways
We show how auxin- and BR-mediated regulation of hypocotyl elongation can account for the low-blue-induced shade avoidance response. Interestingly, we showed previously that gibberellins (GA) are also a control point for shade avoidance (Djakovic-Petrovic et al., 2007; Pierik et al., 2009) implying the involvement of a third hormone. The function of GA is pre-dominantly to degrade growth-inhibiting DELLA proteins. DELLA degradation alone led to only marginal elongation growth, and shade avoidance responses could still be fully induced in multiple DELLA loss-of-function mutants. We therefore proposed a mode of regulation where low blue light induces degradation of DELLA proteins (Djakovic-Petrovic et al., 2007) which allows other growth control pathways, those shown here to be controlled by auxin and BR, to induce enhanced elongation growth.
In tobacco, an intact signalling pathway for the gaseous hormone ethylene is required for low-blue-induced internode elongation (Pierik et al., 2004). However, low-blue-induced hypocotyl elongation in Arabidopsis appears to act independently of the ethylene pathway since ethylene insensitivity does not affect the low blue response and ethylene production is not affected by low blue either (Pierik et al., 2009). Nevertheless, the two regulators identified in the present study, BR and auxin, can induce ethylene production (Arteca and Arteca, 2008). Since ethylene is required for low-R:FR-induced petiole elongation in Arabidopsis rosette plants, it would be interesting to study how this involvement relates to elongation control through auxin and BR, especially since ethylene can interact with both (e.g. De Grauwe et al., 2007; Stepanova et al., 2008; Pierik et al., 2009). Possibly a more complex network of hormone interactions would be activated to control light-quality-induced shade avoidance responses of rosette plants as compared with hypocotyl elongation in seedlings. Such a putative greater regulatory complexity in petioles compared with hypocotyls might be associated with the higher developmental complexity in more mature plants. Although hypocotyl elongation, such as in the current datasets, occurs entirely through enhanced cellular expansion, elongation of petioles and/or internodes may involve a combination of cell expansion and cell proliferation, and thus might involve more regulatory partners to control plasticity. This might also explain why Keller et al. (2011) find less explicit involvement of, for example, auxin and GA to control low-blue-induced petiole elongation and angle, relative to the study discussed here.
Plant material and growth conditions
The following Arabidopsis mutant lines were used: pin1-1 (Okada et al., 1991), pin3-3 (Friml et al., 2002a), pin4-2 (Friml et al., 2002b), pin7 (SALK_048791), tir1-1 (Ruegger et al., 1998), tir1afb1afb2afb3 (Dharmasiri et al., 2005), wei8-1 (Stepanova et al., 2008), bri1-1(ABRC) and rot3-1(ABRC). All mutants were in a Col-0 background (which was used as the wild-type line) except for tir1afb1afb2afb which is in a mixed background with Ws-2 and pin1-1 which is in an EN-2 background. Seeds were surface sterilized, stratified for 3 days in the dark at 4°C and germinated on solid agar plates as described in Pierik et al. (2009). Twenty-four hours after germination the plates were placed in the low-blue treatment or in control light conditions (see description below). Hypocotyl measurements were taken after 5 days of treatment by photographing the seedlings and measuring hypocotyl lengths from these images using imagej (http://rsbweb.nih.gov/ij/).
Standard growth chamber light (Philips HPI 400 W, 200 μm m−2 sec−1; 16 h light, 8 h dark) was filtered to reach similar light intensities between the two light treatments. Control light was standard white light filtered through spectrally neutral shade cloth leading to 130 μm m−2 sec−1 photosynthetically active radiation (PAR), containing ∼25 μm m−2 sec−1 blue light (400–500 nm). Low-blue-light treatment was standard white light filtered through two layers of LEE Medium Yellow 010 filter, giving 132 μm m−2 sec−1 PAR containing <1 μm m−2 sec−1 blue light (see supplemental information in Pierik et al., 2009, for spectral scans of these light conditions).
Epidermal cell length measurements
After 5 days of low-blue treatment, Nomarski optics on a Zeiss photomicroscope (http://www.zeiss.com/) was used to photograph the epidermal cells of the hypocotyl, after which the cell lengths were measured using imagej.
Auxin transport was inhibited using NPA (25 μm) (Petrášek et al., 2003) and auxin perception was blocked using PEO-IAA (50 μm) (Hayashi et al., 2008). Brassinazole (0.5 μm) was used to inhibit BR biosynthesis (Asami et al., 2000) and BL (0.1 μm) was used to induce a BR response (Grove et al., 1979). All chemicals were added to the agar medium at the start of the light treatment. One hundred and fifty microlitres of a concentrated, sterile solution was administered as a film on top of the agar and allowed to diffuse through the medium.
GUS histochemical staining
Transgenic promoter GUS lines were used to study the expression patterns of PIN1, PIN3, PIN4 and PIN7 (Friml et al., 2002a,b; Friml et al., 2003; Benkováet al., 2003). The transgenic pIAA19:GUS line was used as a marker for auxin activity (Tatematsu et al., 2004). GUS activities were determined by incubating freshly harvested seedlings overnight in a staining solution containing 1 mm X-Gluc (5-bromo-4-chloro-3-indolyl β-d-glucoronide) in 100 mm sodium phosphate buffer pH 7.0, with 0.1 mm EDTA, 0.1% Triton X-100, 1 mm K4Fe(CN)6 and 1 mm K3Fe(CN)6), following which the seedlings were bleached in 70% ethanol for at least 48 h.
Seedlings grown under control or low-blue-light conditions were harvested after 2 days in these light treatments. For every biological replicate, approximately 200 seedlings were pooled. Harvested material was immediately frozen and stored at −80°C until further analyses. Total RNA from these seedlings was extracted using the RNeasy Kit (Qiagen, http://www.qiagen.com/). Genomic DNA contamination was eliminated using on-column DNase digestion (Qiagen). One microgram of total RNA was then reverse transcribed with random hexamers using the SuperScript III Reverse Transcriptase kit (Invitrogen, http://www.invitrogen.com/). Real-time RT-PCR was performed using a Bio-Rad single-colour real-time PCR detection system (http://www.bio-rad.com/), using gene-specific primers and the SYBR Green Supermix dye system (Bio-Rad). Relative transcript abundance was calculated using the comparative 2−ΔΔCt method using UBQ10 as an internal reference gene. Primer sequences and annealing temperatures used for all primer pairs are as listed in Table S1.
Measurement of xyloglucan-degrading activity
After 2 days of control and low-blue treatment, whole seedlings were harvested and snap frozen in liquid nitrogen. Approximately 250 seedlings were pooled per biological replicate. The preparation of crude enzyme extracts from these seedlings was as described in Sasidharan et al. (2008). These enzyme extracts were then used for the measurement of XTH activity. The XTH activity was measured as the XDA according to the method of Sulováet al. (1995) with modifications as mentioned in Sasidharan et al. (2008). The XDA is a measure of both the transglycosylating and hydrolytic activity of XTHs and could also include the hydrolytic activity of non-specific endoglucanases. However, in the incubation period used for this assay, there was a negligible contribution from the hydrolytic activity. The XDA values obtained here are therefore a measure of the transglucosylating activity and are expressed as XDA per microgram of cell wall protein. Protein estimation was performed according to the method of Bradford with a ready to use Bradford reagent from Bio-Rad.
Data were analyzed using a one-way anova followed by a Tukey’s b post-hoc test to allow comparisons among all means or with a Student’s t-test when only two means were compared.
We would like to thank Carlos Ballaré and Mercedes Keller for stimulating discussions and helpful comments on the manuscript. Seeds for mutants and reporters were kindly shared by J. Alonso, I. Blilou, J. Friml and K. Yamamoto or obtained from the Nottingham Arabidopsis Stock Centre. The α-(phenylethyl-2-one)-indole-3-acetic acid was kindly provided by K.-I. Hayashi. RP is funded by VENI grant no. 86306001 from the Netherlands Organisation for Scientific Research.