Chloroplasts Modulate Elongation Responses to Canopy Shade by Retrograde Pathways Involving HY5 and Abscisic Acid

Light and retrograde pathways interact to regulate the hypocotyl elongation response to shade.

Abstract

Plants use light as energy for photosynthesis but also as a signal of competing vegetation. Using different concentrations of norflurazon and lincomycin, we found that the response to canopy shade in Arabidopsis (Arabidopsis thaliana) was repressed even when inhibitors only caused a modest reduction in the level of photosynthetic pigments. High inhibitor concentrations resulted in albino seedlings that were unable to elongate when exposed to shade, in part due to attenuated light perception and signaling via phytochrome B and phytochrome-interacting factors. The response to shade was further repressed by a retrograde network with two separate nodes represented by the transcription factor LONG HYPOCOTYL 5 and the carotenoid-derived hormone abscisic acid. The unveiled connection among chloroplast status, light (shade) signaling, and developmental responses should contribute to achieve optimal photosynthetic performance under light-changing conditions.

INTRODUCTION

Life on our planet heavily relies on photosynthesis, i.e. the use of solar energy (sunlight) to fix carbon into organic matter linked to the production of oxygen from water. In plants, the quantity and quality of the incoming light strongly influence growth and development. For example, oxidative stress and eventual damage can occur if the amount of light exceeds the photosynthetic capacity of the chloroplast. By contrast, light supply and hence photosynthetic activity can be compromised by the shading of nearby plants. Under a canopy, plants might actually be exposed to moments of both excess light (e.g. sunflecks) and low light (i.e. shading) during the same day. Even in open habitats, plants are usually found in communities where competition for light might result in overgrowing and eventual shadowing by neighbors.

Light quality is an important signal that informs plants of potential competitors. Vegetation absorbs light from the visible region (called “photosynthetically active radiation” or PAR, 400–700 nm). In particular, it absorbs red light (R, 600–700 nm) but transmits and reflects far-red light (FR, 700–800 nm), therefore causing a reduction in the R-to-FR ratio (R/FR). Both PAR (light quantity) and R/FR (light quality) are greatly reduced under a plant canopy, whereas the presence of nearby plants (without direct vegetation shading) involves a more moderate reduction of R/FR without changes in PAR (Casal, 2012; Martínez-García et al., 2014; Fiorucci and Fankhauser, 2017). Independently of the PAR level, a drop in R/FR acts as a signal that strongly and differentially affects elongation of shade-avoiding plants such as Arabidopsis (Arabidopsis thaliana) and most crops (Martínez-García et al., 2014). Low R/FR signals also cause a decrease in the levels of photosynthetic pigments (chlorophylls [CHLs] and carotenoids [CRTs]) in seedlings and adult plants (Roig-Villanova et al., 2007; Patel et al., 2013; Bou-Torrent et al., 2015; Llorente et al., 2017). These and other responses triggered by a reduced R/FR are collectively known as “shade avoidance syndrome” (SAS), and aim to overgrow neighboring plants, readjust photosynthetic metabolism, and eventually launch reproductive development (Franklin, 2008; Casal, 2012; Gommers et al., 2013; Martínez-García et al., 2014).

Low R/FR signals indicative of shade are perceived by the phytochrome (phy) family of photoreceptors. Five genes encode the phy family in Arabidopsis: phyA to phyE. Whereas phyB is the major phy controlling the responses to shade, other phy members such as phyD and phyE can also redundantly contribute to the control of shade-modulated elongation growth or flowering time (Franklin, 2008; Martínez-García et al., 2014). In the case of photolabile phyA, an antagonistic negative role has been reported for the seedling hypocotyl elongation response to shade. Thus, the SAS is induced by phyB deactivation but gradually antagonized by phyA in response to high FR levels characteristic of plant canopy shade (Casal, 2012; Martínez-García et al., 2014). This intrafamily photosensory attenuation mechanism might act to suppress excessive elongation under prolonged direct vegetation shade. It remains unknown whether other SAS responses, including photosynthetic pigment decrease, are also affected by this antagonistic regulation by phyA and phyB.

In any case, the balance between positive and negative regulators of the SAS acting downstream phytochromes was found to be instrumental to regulate not only hypocotyl elongation but also CRT biosynthesis (Franklin, 2008; Casal, 2013; Bou-Torrent et al., 2015). Positive regulators of the SAS include transcription factors of the basic-helix-loop-helix family such as PHYTOCHROME-INTERACTING FACTORs (PIFs) andhomeodomain Leu zipper class II family members, whereas the basic Leu zipper transcription factor LONG HYPOCOTYL 5 (HY5) and other basic-helix-loop-helix family members have negative roles. Among them, PIFs and HY5 have also been found to participate in retrograde signaling during deetiolation, i.e. in the communication between chloroplasts and nucleus when underground seedlings sense the light and change from skotomorphogenic (i.e. heterotrophic) to photomorphogenic (i.e. photosynthetic) development (Ruckle et al., 2007; Martín et al., 2016; Xu et al., 2016).

Alterations in the physiological status of the chloroplast in light-grown plants are also signaled to the nucleus by a variety of retrograde pathways that readjust nuclear gene expression accordingly (Baier and Dietz, 2005; Gläßer et al., 2014; Chan et al., 2016). Because exposure to shade causes a decrease in the accumulation of CHLs and CRTs that can eventually compromise photosynthesis and photoprotection (Roig-Villanova et al., 2007; Cagnola et al., 2012; Bou-Torrent et al., 2015), we reasoned that the derived effects on chloroplast homeostasis might not be just a consequence but influence the response to shade itself (e.g. in terms of elongation) through retrograde signaling. The work reported here aimed to test this possibility.

RESULTS AND DISCUSSION

Functional Chloroplasts Are Required for Full Response to Simulated Shade

To initially test whether retrograde signals modulate the elongation response to plant proximity, we used two kinds of inhibitors of chloroplast function associated with retrograde signaling: norflurazon (NF), an inhibitor of CRT biosynthesis (Chamovitz et al., 1991) and lincomycin (LIN), an inhibitor of chloroplast protein synthesis (Mulo et al., 2003). Both inhibitors were present in the medium used for seed germination and seedling growth (Figure 1). This medium also contained Suc to sustain growth even in the absence of photosynthesis. As expected, a concentration-dependent bleaching was observed in wild-type Arabidopsis plants grown under white light (W) with NF or LIN (Figure 1A). The concentration of inhibitors required to obtain albino seedlings was adjusted to our experimental conditions. For example, an albino phenotype was previously observed in wild-type seedlings grown without Suc in the presence of 5 μM NF under 100 μmol m^–2 s^–1 or with 50 nM NF under 5 μmol m^–2 s^–1 (Saini et al., 2011). We used intermediate light intensity conditions (20–24 μmol m^–2 s^–1) and, most importantly, added Suc in the medium, which together made it necessary to adjust NF concentration to 200 nM to obtain completely albino seedlings (Figure 1A).

Figure 1.

Hypocotyl Elongation in Response to Shade Requires Functional Chloroplasts.

(A) Wild-type (Col) plants were germinated and grown under W for 7 d on media with or without the indicated concentrations of NF or LIN. Graphs represent the mean ± se values of total CHL and CRT contents of at least n = 8 independent samples (pools of seedlings) from two different experiments. Pigments were quantified by spectrophotometric methods and represented relative to the levels found in the absence of inhibitors. Pictures show the phenotype of representative seedlings.

(B) Hypocotyl elongation in 7-d-old seedlings germinated and grown on media supplemented with the indicated concentrations of NF or LIN under W or exposed to W+FR during the last 5 d. Graphs in the left represent the length of the hypocotyls (mean ± se of n ≥ 100 seedlings grown in different plates in at least two independent experiments) relative to the value in samples grown under W in the absence of inhibitors. Graphs in the right represent the elongation response to shade of the same samples. They show the ratio of hypocotyl length under W+FR relative to that under W. A value of “1” means no growth differences between W and W+FR; values > 1 indicate higher growth under W+FR; and values < 1 indicate lower growth under W+FR. Asterisks mark values statistically > 1 (t test, P ≤ 0.05), i.e. responsive to shade by increasing hypocotyl elongation.

The presence of inhibitors had no significant effect on hypocotyl length under W (Figure 1B). However, exposure to FR-enriched W (W+FR) to simulate canopy shade progressively impaired elongation as levels of photosynthetic pigments (CHLs and CRTs) decreased. Importantly, inhibition of shade-triggered elongation growth was observed at concentrations of NF or LIN that only slightly reduced the levels of photosynthetic pigments and had no visual impact on seedling pigmentation (e.g. 25 nM NF or 5 μM LIN), suggesting that even moderate alterations in chloroplast function might influence the response to shade. Hypocotyl elongation in response to shade was completely blocked at concentrations of NF causing >80% loss of CHLs, whereas an even lower reduction (50%) was required for a lack of response in LIN-treated seedlings (Figure 1B). In both cases, completely bleached seedlings did not elongate at all when exposed to W+FR compared with W controls (Figure 1B), suggesting that functional chloroplasts are required for the elongation response to canopy shade.

We next aimed to confirm that the disrupted elongation response to W+FR observed in bleached seedlings was not due to energetic constraints. If nonphotosynthetic seedlings lacking CHLs maintain an intrinsic capacity to grow, it would be expected that their hypocotyls would elongate when treated with growth-promoting hormones such as brassinosteroids, auxins, or gibberellins. In agreement, seedlings grown in the presence of NF concentrations that completely blocked photosynthetic development (2 μM) were able to elongate very similarly to control green seedlings when treated with any of these hormones (Supplemental Figure 1). The same hormone treatments caused a similar growth response in the case of mutant hdr-3 seedlings, which are unable to produce the precursors for CHL and CRT biosynthesis in chloroplasts and hence display an albino phenotype (Pokhilko et al., 2015). We therefore conclude that functional chloroplasts are not required for hormone-mediated hypocotyl elongation (at least in Suc-supplemented media), but are necessary for growth in response to shade signals.

Defective Chloroplast Function Impairs Phy-Mediated Shade Signaling

Phytochromes are the main photoreceptors involved in shade perception and signal transduction, with phyB having a predominant role in Arabidopsis (Casal, 2012; Martínez-García et al., 2014; Fiorucci and Fankhauser, 2017). To address whether treatment with NF or LIN had an impact on phy signaling, we used transgenic 35S:PHYB-GFP plants expressing a biologically active version of the phyB protein fused to the green fluorescent protein (GFP; Yamaguchi et al., 1999). Under W, the phyB-GFP fusion protein shows a characteristic distribution in nuclear speckles, presumably the site where active photoreceptor proteins interact with other nuclear factors to mediate light signaling (Yamaguchi et al., 1999). We observed that only minutes after exposing 35S:PHYB-GFP seedlings to an end-of-day FR treatment to simulate shade, the green fluorescence associated to the phyB-GFP reporter became more disperse in the nuclei of epidermal hypocotyl cells (Figure 2), likely reflecting phyB inactivation. Strikingly, this shade-mediated inactivation process was clearly delayed in albino seedlings germinated and grown in the presence of NF (5 μM) or LIN (1 mM). While mock (i.e. green) seedlings displayed evenly distributed nuclear phyB-GFP fluorescence in all analyzed cells 90 min after the light treatment, inhibitor-grown (i.e. albino) seedlings still showed nuclear speckles in some cells after 210 min. These results suggest that functional chloroplasts are required for proper phyB inactivation in response to shade signals. Consistent with this conclusion, the shade-triggered and phy-dependent stabilization of photolabile PIF3 was attenuated in NF-bleached seedlings (Supplemental Figure 2).

Figure 2.

Retrograde Signals Prevent phyB Deactivation in Response to Shade.

Transgenic 35S:PHYB-GFP plants were germinated and grown under W for 7 d on media with or without 5 μM NF or 1 mM LIN. After a 5-min irradiation with FR to simulate shade, the distribution of the phyB-GFP reporter protein in nuclei from epidermal hypocotyl cells was analyzed by confocal laser scanning microscopy. Ten seedlings from three independent experiments were observed for each treatment and time point. Graph shows mean ± se values of the percentage of cells showing nuclear speckles in a total of 90 nuclei per treatment and time point. Bottom images: Representative images of nuclei from seedling hypocotyl cells 0 and 90 min after the light treatment.

To further test whether phy function was altered in albino seedlings, we next analyzed the expression of rapidly shade-induced phy primary target genes in wild-type plants either treated or not with NF (Figure 3). In particular, we chose genes PIL1, ATHB2, HFR1, YUCCA8, and PAR1 (Roig-Villanova et al., 2006). Wild-type plants grown under W for 7 d were exposed to W+FR for 1 h and then samples were collected and used for RNA extraction and reverse transcription-quantitative PCR (RT-qPCR) analysis. As expected, comparison of W-grown controls and shade-exposed (1 h W+FR) samples showed that all genes analyzed were induced by shade in green seedlings, ranging from twofold (PAR1) to 80-fold (PIL1). In NF-grown seedlings, however, the induction was much reduced (Figure 3). HFR1, YUCCA8, and PAR1 gene expression hardly changed after W+FR treatment in albino seedlings, whereas PIL1 induction was only 10% compared with that detected in green seedlings and ATHB2 upregulation was less than half. From these results, we conclude that the absence of functional chloroplasts somehow prevents normal light (i.e. shade) perception and signal transduction by phytochromes.

Figure 3.

Shade-Triggered Induction of Phy Primary Target Genes Is Attenuated in Bleached Seedlings.

Wild-type plants were germinated and grown under W for 7 d on media with or without 5 μM NF. Before and after 1 h of exposure to W+FR, RNA was isolated from seedlings and used to analyze the transcript levels of the indicated genes by RT-qPCR. Graph represents the induction response (transcript levels in W versus those after exposure to W+FR). Mean ± sd of n = 3 pools of seedlings, from independent experiments, are shown.

Functional phy holoproteins require the covalent attachment of a phytochromobilin (PΦB) chromophore to each phy apoprotein monomer (Rockwell et al., 2006). The synthesis of PΦB occurs in the plastid and the early steps are shared with those required to synthesize heme and CHLs (Figure 4). To test whether the observed reduction in shade-triggered phy inactivation (and hence hypocotyl elongation) in albino seedlings could result from impaired accumulation of PΦB, we analyzed the elongation response to shade of Arabidopsis mutants defective in PΦB synthesis (Parks and Quail, 1991). In particular, we used the hy1-1 allele, which was isolated from a fast-neutron mutagenized population of Landsberg erecta (Ler) and carries a short deletion that disrupts its function (Davis et al., 1999). As shown in Figure 4, elongation in response to W+FR was still observed in the hy1-1 mutant (Figure 4B). We therefore conclude that treatment with bleaching inhibitors interferes with phy-dependent signaling by mechanisms other than defective chromophore availability.

Figure 4.

gun Mutants Show Different Elongation Responses to Shade.

(A) Roles of GUN proteins in retrograde signaling and production of CHLs, heme, and the phy chromophore.

(B) Elongation responses to shade in mutants defective in some of the GUN proteins represented in (A). Mutants and their respective wild-type backgrounds (Ler for hy1/gun2 and Col for the rest) were germinated and grown as indicated with or without 5 μM NF. Graph represents the mean ± se values of at least two independent experiments with n ≥ 25 seedlings each.

Plastid retrograde signaling has been previously shown to interact with components of light signaling networks to coordinate chloroplast biogenesis with both the light environment and development (Larkin and Ruckle, 2008; Lepistö and Rintamäki, 2012; Ruckle et al., 2012; Martín et al., 2016; Xu et al., 2016). In fact, mutants defective in the PΦB biosynthetic enzymes HY1/GENOMES UNCOUPLED 2 (GUN2) and HY2/GUN3 (Figure 4A) were isolated in a screen for GUN mutants that retained partial expression of genes encoding photosynthesis-related plastidial proteins after NF treatment (Mochizuki et al., 2001). Other GUN proteins such as GUN5 (Mochizuki et al., 2001) participate in a different branch of the tetrapyrrole pathway that leads to the production of CHLs (Figure 4A). Unlike other GUN proteins, GUN1 is not an enzyme but a central integrator of retrograde signaling pathways that was proposed to coordinate photomorphogenesis with chloroplast function (Koussevitzky et al., 2007; Ruckle et al., 2007; Ruckle and Larkin, 2009). Similar to wild-type plants, mutants gun1-101 (Ruckle et al., 2007) and gun5-1 (Mochizuki et al., 2001) elongated in response to W+FR under normal growth conditions (i.e. when chloroplasts are functional) but not when chloroplast development was blocked with NF (Figure 4). Together, the described results suggest that alteration of chloroplast function influences a retrograde signaling pathway independent of GUN proteins that modulates the phy-mediated response to shade.

Retrograde Pathways Repressing Shade-Triggered Hypocotyl Elongation Involve HY5 but Not GUN1

To identify components of the chloroplast-modulated transduction pathway involved in the response to shade, we next tested the possible role of SAS-related transcription factors known to be involved in both light and retrograde signaling: PIFs (Martín et al., 2016) and HY5 (Ruckle et al., 2007; Xu et al., 2016). A role for PIFs as positive regulators of the response to shade (including hypocotyl elongation) is well established (Lorrain et al., 2008; Leivar et al., 2012; Bou-Torrent et al., 2015). However, under our experimental conditions the quadruple pifQ mutant defective in PIF1, PIF3, PIF4, and PIF5 activities showed a wild-type phenotype in terms of shade-triggered hypocotyl elongation in both green and albino seedlings (Figure 5).

Figure 5.

HY5 Represses Shade-Triggered Hypocotyl Elongation in a GUN1-Independent Manner.

Wild-type (Col, WT) as well as single (hy5), double (hy5 gun1), and quadruple (pifQ) mutant lines were germinated and grown as indicated with or without 5 μM NF or 1 mM LIN. Graph represents the mean ± se elongation values of at least two independent experiments with n ≥ 25 seedlings each. Asterisks mark statistically significant responses to shade (t test, P ≤ 0.05).

HY5 has been proposed to have a function in the adaptation to prolonged shade and the response to sunflecks, i.e. exposure to sunlight through gaps in the canopy (Sellaro et al., 2011; Ciolfi et al., 2013). The role of this elongation-repressing transcription factor in controlling the shade-promoted growth of seedling hypocotyls, however, remains unclear. Our previous work (Bou-Torrent et al., 2015) showed that complete loss of HY5 activity in the null hy5-2 mutant (referred to as hy5 from now on) hardly had an influence on the elongation of Arabidopsis seedlings exposed to a W+FR treatment mimicking vegetation proximity (R/FR = 0.05). As shown in Figure 5, however, hy5 seedlings displayed increased hypocotyl elongation compared with the wild type when illuminated with light of a lower R/FR (0.02), reminiscent of canopy shade. These results suggest that HY5 is a repressor of hypocotyl elongation in green seedlings exposed to low or very low R/FR conditions. Consistent with this conclusion, shade-triggered hypocotyl growth was inhibited in transgenic seedlings overaccumulating HY5 in a hy5 background (Supplemental Figure 3).

Similar to wild-type plants, the elongation response to canopy shade of hy5 seedlings was almost completely blocked with LIN (Figure 5). However, the growth response of HY5-deficient seedlings was not abolished but just attenuated in NF-supplemented medium. Similar results were obtained in medium lacking Suc, but the effects of HY5 gain or loss of function on the elongation response of green or NF-treated seedlings, respectively, were much more obvious in the presence of Suc (Supplemental Figure 3). We therefore kept using Suc-supplemented media for the rest of the work. Double hy5 gun1-101 mutants were also found to display a partial elongation response to shade in NF but not in LIN, similar to that found for the single hy5 mutant (Figure 5). Together, the described results show that HY5 is a repressor of canopy-shade–triggered hypocotyl elongation. When this negative regulator is lost, the elongation response to shade can still be blocked by a GUN1-independent retrograde pathway that is active in LIN-treated but not in NF-treated albino seedlings.

We next analyzed the levels of HY5 transcripts before and after exposure to our shade conditions (Figure 6). In green wild-type plants (grown without inhibitors), the levels of HY5 transcripts were similar under W and up to 8 h of our W+FR treatment (Figure 6A). By contrast, immunoblot analysis of a HY5-GFP reporter in complemented hy5 35S:HY5-GFP plants showed increased protein levels after the simulated shade treatment (Figure 6B). Chromatin immunoprecipitation (ChIP) experiments also detected increased levels of HY5-GFP bound to target promotors in shade-exposed green seedlings (Figure 6C). Although the endogenous HY5 protein might not behave exactly as the overexpressed GFP-tagged version of the protein, our results are in agreement with previous studies using a different reporter (HY5-myc) that concluded that the low R/FR treatment stabilizes HY5 (Pacín et al., 2016). Post-transcriptional HY5 accumulation when R/FR is low or very low in natural environments (such as in deep or canopy shade) might help to prevent seedlings from exhibiting excessive elongation.

Figure 6.

HY5 Levels are Regulated by Shade and Retrograde Signals.

(A) Levels of HY5-encoding transcripts in wild-type and hy5 plants germinated and grown under W for 7 d and then exposed to W+FR for the indicated times. Transcript levels were quantified by RT-qPCR and represented relative to those in W-grown wild-type plants (mean ± se of n = 3 samples corresponding to pools of whole seedlings grown in different experiments).

(B) Levels of HY5-GFP protein in hy5 35S:HY5-GFP plants germinated and grown under W for 7 d with or without 5 μM NF or 1 mM LIN and then exposed to W+FR for the indicated times. Protein levels were quantified from immunoblot analysis with a commercial anti-GFP serum. Mean ± se values (n ≥ 3 samples from pools of whole seedlings grown in different experiments) are represented relative to those in plants grown without inhibitors before exposure to shade. Asterisks mark statistically significant differences relative to the 0 h time point (t test, P ≤ 0.05).

(C) ChIP analysis of HY5-GFP binding to the promoters of the indicated genes. After germinating and growing plants of the hy5 35S:HY5-GFP line on media without inhibitors as indicated, ChIP experiments were done using commercial anti-GFP serum. Chromatin from these samples and from no-antibody controls was then used for RT-qPCR amplification of HY5 binding sites in the promoter of the genes. Enrichment was calculated as the ratio of anti-GFP versus no-antibody values after normalization with input samples (i.e. before ChIP). Graph shows mean ± se values of n = 2 samples from seedlings grown in different experiments. Asterisks mark statistically significant differences in shade-treated samples (t test, P ≤ 0.05).

(D) Levels of HY5-encoding transcripts in wild-type plants germinated on medium with or without 5 μM NF or 1 mM LIN and grown under W for 2 d followed by five additional days under W or under W+FR. Transcript levels were quantified by RT-qPCR and represented relative to those in plants grown under W without inhibitors (mean ± se of n = 3 samples corresponding to pools of whole seedlings grown in different experiments).

Both HY5-encoding transcripts (Figure 6D) and HY5-GFP protein (Figure 6B) were higher in albino seedlings grown with LIN or NF independent of the light treatment, suggesting that these inhibitors promote HY5 function by increasing gene expression (or/and transcript stability) and decreasing protein turnover. The observation that hypocotyl length is not reduced in W-grown seedlings in the presence of inhibitors (Figure 1B) despite accumulating higher HY5 levels (Figure 6B) suggests that hypocotyl elongation is suppressed to a saturating level by multiple pathways under W and hence it would not be further repressed by increasing HY5 function. In response to W+FR, however, enhanced HY5 activity together with reduced light signaling in bleached wild-type seedlings would result in no hypocotyl elongation. Only when the repressor activity of HY5 is removed (i.e. in HY5-defective mutants), a second pathway that inhibits the elongation response of albino seedlings becomes apparent in the presence of LIN but not in the presence of NF (Figure 5).

CRT-Derived Products Repress Shade-Triggered Hypocotyl Elongation

The distinct mode of action of LIN and NF and particularly their differential effect on CRT levels is illustrated by their concentration-dependent impact on photosynthetic pigment accumulation (Figure 1A). High-performance liquid chromatography analysis of CRT contents (Supplemental Figure 4) confirmed that albino LIN-treated seedlings accumulated low but detectable levels of lutein and violaxanthin as well as traces of β-carotene and neoxanthin. By contrast, NF blocks the desaturation of phytoene, the first committed intermediate of the CRT pathway (Figure 7). As expected, NF-treated seedlings accumulated phytoene (which is colorless and hence not detected in the spectrophotometric assay used in Figure 1A) but were virtually devoid of downstream CRTs (Supplemental Figure 4). Similar to that observed with LIN, other bleaching inhibitors that prevent chloroplast development and cause albinism without specifically blocking the production of CRTs, such as the plastid protein synthesis inhibitor chloramphenicol (CAP) or the nitrogen assimilation inhibitor phosphinotricin (PPT), were found to prevent shade-triggered elongation growth in wild type and HY5-defective mutants (Figure 7B). By contrast, inhibition of the CRT pathway downstream of lycopene by blocking the activity of lycopene cyclases with 2-(4-chlorophenylthio)-triethylamine chloride (CPTA) resulted in albino seedlings that were able to respond to shade and elongate when HY5 function was lost (Figure 7).

Figure 7.

Blockage of the CRT Pathway Derepresses Shade-Triggered Elongation of Bleached HY5-Defective Seedlings.

(A) Pathways for the biosynthesis of CRTs and derived hormones. The steps targeted by NF and other inhibitors and the reactions catalyzed by enzymes that determine metabolic flux to CRTs (PSY) and ABA (ABA2) are shown. Xanthophylls are boxed in yellow.

(B) Elongation responses to shade in wild-type and mutant plants defective in HY5, PSY, or both. Wild-type (WT) and single hy5 mutant plants were germinated and grown as indicated on media either supplemented or not with concentrations of NF, CPTA, FSM, LIN, CAP, or PPT producing albino seedlings. Single psy-1 and double hy5 psy-1 mutants were only grown without inhibitors. Graph represents the mean ± se values of n ≥ 30 seedlings in a representative experiment.

To test whether the ability to respond to shade of hy5 seedlings grown in the presence of NF or CPTA was specifically due to the blockage of the CRT pathway, we next used Arabidopsis mutants. The enzyme PHYTOENE SYNTHASE (PSY) produces phytoene in the first committed step of the CRT pathway (Figure 7A). Because PSY is encoded by a single gene in Arabidopsis (Ruiz-Sola and Rodríguez-Concepción, 2012), the knock-out mutant psy-1 (Pokhilko et al., 2015) does not produce phytoene and hence cannot feed the pathway for the biosynthesis of downstream CRTs (Supplemental Figure 4). As a consequence, the mutant displays an albino phenotype undistinguishable from that observed in wild-type seedlings treated with NF or CPTA (Pokhilko et al., 2015). Similar to that described for wild-type seedlings grown in the presence of CRT biosynthesis inhibitors, psy-1 seedlings were unable to elongate when exposed to W+FR (Figure 7B). However, the elongation response was rescued when both HY5 and CRTs were missing in double hy5 psy-1 mutant seedlings (Figure 7B).

Pharmacological or genetic blockage of the CRT pathway prevents the biosynthesis of CRTs and derived products, but it might also cause an accumulation of upstream metabolites. Among them, methylerythritol cyclodiphosphate, an intermediate of the pathway that supplies the metabolic precursors of CRTs (Figure 7A), has been shown to act as a retrograde signal in response to stress (Xiao et al., 2012). Blockage of methylerythritol cyclodiphosphate production with the inhibitor fosmidomycin (FSM; Figure 7A), however, did not prevent the elongation response to shade of NF-treated hy5 seedlings (Figure 7B). We therefore conclude that what allows hy5 seedlings to respond to shade is not the accumulation of a metabolite upstream PSY but the depletion of a CRT-derived product synthesized after the step blocked by CPTA, i.e. downstream of lycopene (Figure 7A).

As represented in Figure 7A, lycopene cyclization leads to the production of CRTs with two β-rings (β,β-CRTs such as β-carotene and derived xanthophylls) with one β- and one ε-ring (β,ε-CRTs such as lutein). The production of β,ε-CRTs in Arabidopsis is completely blocked in the green lut2 mutant (Supplemental Figure 4; Emiliani et al., 2018), which is defective in the only gene encoding LYCOPENE ε-CYCLASE in this plant species (Figure 7A; Ruiz-Sola and Rodríguez-Concepción, 2012). Loss of β,ε-CRTs did not change the elongation response to shade of single lut2 (versus wild type) or double hy5 lut2 (versus hy5) seedlings (Supplemental Figure 5). We therefore concluded that the effect observed with CPTA (Figure 7B) is not due to the absence of β,ε-CRTs but most likely to defects in the β-carotene branch of the CRT pathway (Figure 7A). Considering all these data together, we speculated that unidentified products derived from β,β-CRTs can repress shade-induced elongation growth in seedlings bleached with LIN and other inhibitors that do not target the CRT biosynthesis pathway. The absence of these products in seedlings treated with NF or CPTA, or in the psy-1 mutant, allows hypocotyl elongation in response to shade but only when the growth-inhibitory effect of HY5 is released.

Abscisic Acid Represses the Elongation Response to Shade

Among the biologically active metabolites derived from β,β-CRTs (Hou et al., 2016), we decided to evaluate the role of abscisic acid (ABA) as this plant hormone was found to participate in the transduction of chloroplast-derived reactive oxygen species/redox signals (Baier and Dietz, 2005; Gläßer et al., 2014; Chan et al., 2016), to modulate hypocotyl growth (Lau and Deng, 2010; Humplík et al., 2017) and to act together with HY5 in the regulation of several plant cell responses (Chen et al., 2008; Xu et al., 2014). Furthermore, treatment with low R/FR was reported to induce ABA production and signaling in tomato (Solanum lycopersicum) and Arabidopsis (Cagnola et al., 2012; González-Grandío et al., 2013; Holalu and Finlayson, 2017). Indeed, a tendency to increase ABA contents was observed in green seedlings after only 1 h of exposure to our simulated canopy shade conditions (Supplemental Figure 6).

As expected, ABA was absent in NF-treated seedlings but it could be detected in LIN-grown seedlings (Supplemental Figure 6). If the presence of ABA in LIN-treated hy5 seedlings contributed to inhibit their response to shade, it would be expected that preventing the formation of this hormone would be sufficient to rescue their response to shade. In agreement, a genetic blockage of the last step of ABA biosynthesis, catalyzed by the ABA2 protein (Figure 7A), allowed LIN-treated double hy5 aba2 seedlings to elongate in response to shade (Figure 8). Single hy5 and double hy5 aba2 seedlings had a very similar response to shade in NF-supplemented medium. By contrast, in green seedlings grown in the absence of inhibitors (i.e. with functional chloroplasts), the double mutant elongated more than single hy5 seedlings when exposed to shade (Figure 8A). We therefore concluded that HY5 and ABA likely repress shade-induced hypocotyl elongation by independent pathways.

Figure 8.

ABA Represses Shade-Triggered Hypocotyl Elongation Independent of HY5.

(A) Elongation responses to shade in wild-type (WT) and mutant plants defective in HY5, ABA2, or both. Plants were germinated and grown as indicated with or without 5 μM NF or 1 mM LIN. Graph represents the mean ± se values of a total of n ≥ 25 seedlings from at least two independent experiments. Asterisks mark statistically significant responses to shade (t test, P ≤ 0.05).

(B) Effect of ABA on the elongation of NF-treated seedling hypocotyls in response to shade. Wild-type (WT) and hy5 plants germinated with or without 5 μM NF plus the indicated concentrations of ABA were grown under W for 2 d followed by five additional days under W or under W+FR. Graphs represent hypocotyl length (mean ± se of n ≥ 25 seedlings in a representative experiment).

This conclusion was further supported by the results of treating NF-grown wild type and hy5 seedlings with increasing concentrations of ABA (Figure 8B). Although no effect was observed in the wild type, the ability of hy5 seedlings to elongate in response to W+FR exposure was progressively repressed as ABA concentration increased. At concentrations of the hormone of 200 nM or higher, which are within the physiological range (Waadt et al., 2014), NF-treated hy5 seedlings did not respond to shade (Figure 8B), similar to that observed with the LIN treatment.

Exogenous ABA treatment was also able to repress shade-promoted hypocotyl elongation in green wild-type seedlings grown without inhibitors (Figure 9). We used this phenotype to identify ABA-related transcription factors involved in this response. Mutants defective in ABA-INSENSITIVE 3 (ABI3) and ABI4 elongated slightly more than wild-type seedlings when illuminated with W+FR and this response was not repressed by ABA. By contrast, ABI5-defective seedlings showed a wild-type phenotype in terms of sensitivity of shade-triggered elongation to ABA treatment (Figure 9A). These results suggest that ABI3 and ABI4 but not ABI5 are required for ABA to inhibit hypocotyl elongation.

Figure 9.

ABI3 and ABI4 But Not ABI5 Participate in the ABA-Mediated Repression of Shade-Induced Hypocotyl Elongation.

(A) Effect of ABA on the elongation responses to shade of wild-type (WT) seedlings and mutants defective in ABI3, ABI4, or ABI5. Plants were germinated and grown as indicated on media with or without 0.2 μM ABA. Graph represents the mean ± se values of a total of n ≥ 25 seedlings from two independent experiments.

(B) Elongation responses to shade of double mutants defective in HY5 and either ABI3 or ABI4. Plants of the indicated genotypes were germinated and grown as illustrated with or without 1 mM LIN. Graph represents the mean ± se values of a total of n ≥ 25 seedlings from two independent experiments. Asterisks mark statistically significant responses to shade (t test, P ≤ 0.05). WT = wild type.

If these transcription factors also transduce the ABA signal in the molecular pathway that blocks elongation in shade-exposed albino seedlings, it would be expected that double mutants lacking both HY5 and ABI3 or ABI4 and grown in the presence of LIN would be able to elongate when exposed to W+FR. Indeed, these double mutants elongated more than their parental lines when both mock (green) and LIN-treated (albino) seedlings were grown under simulated shade (Figure 9B). Shade-triggered elongation of LIN-treated hy5 abi3 and hy5 abi4 seedlings, however, was reduced compared with that of ABA-defective hy5 aba2 seedlings (Figure 9B). These results suggest that ABI3 and ABI4 might not participate in the same ABA signaling pathway eventually repressing hypocotyl elongation, but have partially redundant roles in this process (Figure 10).

Figure 10.

Model for the Modulation of Shade Elongation Responses by Retrograde Signals.

In green plants with functional chloroplasts, low R/FR (i.e. canopy shade) signals promote accumulation of growth-promoting PIFs (via phyB deactivation) and also of growth-repressing HY5 (via phyA), likely to prevent an excessive elongation response. Persistent shading or other environmental factors challenging chloroplast function (including exogenous treatment with LIN or NF) can repress phyB inactivation, enhance HY5 expression, and likely promote HY5 stability, eventually resulting in decreased elongation growth. An independent pathway involves ABA, a CRT-derived hormone that represses shade-triggered hypocotyl elongation via ABI3 and ABI4. NF (but not LIN) prevents the production of ABA. As a result, loss of both HY5 and ABA in NF-treated hy5 seedlings allows them to elongate when exposed to low R/FR, whereas this hypocotyl response is blocked by low but detectable levels of ABA in LIN-treated mutants.

A Mechanistic Model for the Modulation of Shade Elongation Responses by Plastid-Dependent Signals

A model generated based on the described results is shown in Figure 10. In high-plant–density environments, like those found in forests, prairies, or orchard communities, a set of R/FR-dependent adaptive responses are unleashed in shade-avoiding plants. Compared with plant proximity (without direct vegetative shading), canopy shade in nature involves lower R/FR values associated with a reduction in the amount of PAR. Although phyB is the major phy controlling these responses, the photolabile phyA has an antagonistic negative role in the shade-mediated regulation of hypocotyl elongation (Ciolfi et al., 2013; Martínez-García et al., 2014; Wang et al., 2018; Zhang et al., 2018). Independently of the PAR level, phyB is deactivated by shade of intermediate, low, and very low R/FR, whereas phyA signaling is activated by shade of low and very low R/FR. As a result, hypocotyl elongation is derepressed under conditions mimicking vegetation proximity (a response aimed at overgrowing neighbors for optimal light exposure). Under R/FR values typical of canopy shade, however, phyA activation prevents seedlings from exhibiting excessive elongation (Figure 10). Our results reported here and elsewhere (Bou-Torrent et al., 2015) suggest that HY5 represses the hypocotyl elongation response more strongly under canopy shade. As previously proposed, HY5 might be principally involved in the phyA-dependent pathway (Ciolfi et al., 2013; Wang et al., 2018; Zhang et al., 2018) whereas other transcription factors, including growth-promoting PIFs, would be mostly associated to the phyB-dependent pathway (Figure 10). These antagonistic phyB/PIF and phyA/HY5 pathways likely provide young seedlings with the capacity to rapidly elongate when impending competition is nearby but also to attenuate excessive growth when growing under a canopy.

During seedling deetiolation, the phyB/PIF pathway converges with a GUN1-dependent retrograde pathway to antagonistically regulate the transcriptional photomorphogenic network (Martín et al., 2016). The GUN1-mediated retrograde signal involved in this particular process was proposed to attenuate photomorphogenesis when chloroplast function is challenged, and to be independent of ABI4 and HY5 (Martín et al., 2016). Our results reported here suggest that in shade-exposed seedlings, a completely different retrograde network that is independent of GUN1 but does depend on HY5, ABI3, and ABI4 modulates the antagonistic action of phyA/HY5 and phyB/PIFs signaling pathways (Figure 10).

Prolonged exposure to shade causes a decrease in the accumulation of CHLs and CRTs that can eventually compromise photosynthesis and photoprotection (Roig-Villanova et al., 2007; Cagnola et al., 2012; Bou-Torrent et al., 2015). Our results suggest that such a challenge to the chloroplast functional status might in turn feed-back to regulate the response to shade (Figure 10). Treatment with low concentrations of NF or LIN (i.e. those causing weak to moderate reduction in the level of photosynthetic pigments) was sufficient to repress the hypocotyl elongation response to low R/FR (Figure 1), likely due to delayed phyB deactivation after a reduction in R/FR (Figure 2). Decreased phyB deactivation correlated with impaired PIF accumulation (Supplemental Figure 2) and attenuated gene expression changes (Figure 3). NF or LIN treatments also caused an enhanced accumulation of HY5 transcripts and increased the stability of the HY5-GFP reporter protein (Figure 6). Together, our findings suggest that retrograde signals inhibit the SAS by repressing the (positive) phyB/PIFs pathway and by promoting the (negative) phyA/HY5 pathway (Figure 10).

Our work further unveiled ABA as a component of the feedback mechanism. This CRT-derived hormone was found to repress shade-triggered hypocotyl elongation (Figure 8), likely through the action of the transcription factors ABI3 and ABI4 (Figure 9). ABI4 has been proposed to participate in GUN1-dependent retrograde signaling (Koussevitzky et al., 2007; Sun et al., 2011; Guo et al., 2016; Xu et al., 2016). However, the results supporting this claim have been repeatedly challenged (Kacprzak et al., 2018). Our data suggest that ABI4 (and ABI3) may act redundantly to transduce the ABA-dependent signal that represses shade-triggered hypocotyl elongation in response to chloroplast dysfunction (Figure 9). While HY5 was previously shown to directly bind and activate the promoter of ABI5 to promote light-induced hypocotyl inhibition during deetiolation (Chen et al., 2008; Xu et al., 2014), our results suggest that this mechanism does not participate in the control of shade-dependent hypocotyl growth. First, HY5 and ABA appear to repress hypocotyl growth by independent pathways (Figure 8). Second, ABI5 is not required to transduce the ABA signal eventually repressing the response to shade (Figure 9).

Arabidopsis mutants defective in phyB were found to accumulate greater amounts of ABA under well-watered conditions and to be less sensitive to exogenous ABA treatments (González et al., 2012). Further supporting a negative role of light in ABA synthesis, dark treatment of previously light-grown plants resulted in increased ABA contents (Weatherwax et al., 1996). A shade-triggered increase in ABA production was reported here (Supplemental Figure 6) and elsewhere (Cagnola et al., 2012; González-Grandío et al., 2013; Holalu and Finlayson, 2017). It is possible that W+FR treatment might promote ABA production to repress the elongation response to shade as part of the mechanism that prevents a too-intense commitment (Figure 10). These results together support ABA as a central signal connecting the functional status of the chloroplast with light responses.

Interestingly, the plastid-synthesized metabolite 3ʹ-phosphoadenosine 5ʹ-phosphate (PAP), which functions as a retrograde signal during oxidative stress caused by high light exposure and drought, was recently shown to act in concert with ABA signaling in guard cells to mediate stomatal closure and in seeds to mediate dormancy and germination (Pornsiriwong et al., 2017). PAP accumulates when the SAL1 phosphatase that normally degrades this metabolite is inactivated during oxidative stress (Estavillo et al., 2011). SAL1-defective mutants show a short hypocotyl phenotype in the light, indicating that accumulation of PAP can repress hypocotyl elongation (Kim and von Arnim, 2009; Chen and Xiong, 2011). This phenotype is rescued (at least partially) in double sal1 phyB and sal1 hy5 mutants (Kim and von Arnim, 2009; Chen and Xiong, 2011), suggesting that functional phyB and HY5 are required for the PAP-promoted and light-dependent repression of hypocotyl growth. Further experiments should explore whether PAP is the retrograde signal deduced from our data to attenuate the response to shade in terms of hypocotyl elongation by independently inhibiting phyB deactivation, increasing HY5 accumulation, and promoting ABA signaling (Figure 10).

Besides ABA, it is possible that other CRT-derived products might also contribute to the repression of shade-triggered hypocotyl elongation detected in hy5 seedlings bleached with LIN, CAP, or PPT but not with NF or CPTA (Figure 7). In particular, strigolactones are hormones derived from β-carotene (Figure 7A) that inhibit hypocotyl elongation in the light by a mechanism requiring phytochromes and involving upregulation of HY5 expression and protein (Tsuchiya et al., 2010; Jia et al., 2014). Other metabolites produced after cleavage of CRTs include β-cyclocitral, and unknown compounds that modulate developmental and stress responses (Hou et al., 2016). While β-cyclocitral is a relatively well-established retrograde signal associated with oxidative stress (Ramel et al., 2012), its contribution to hypocotyl elongation is unknown. Similarly, no hypocotyl growth alterations have been reported in mutants lacking CRT-derived signals that do have an influence on leaf development (Van Norman and Sieburth, 2007; Avendaño-Vázquez et al., 2014). Whether any of these CRT-related metabolites participate in the elongation response to shade remains to be investigated.

Collectively, our data support the notion that chloroplasts are plant cell compartments with fundamental roles not only for photosynthesis and metabolism but also for environmental (light) sensing and signaling. Here we show that HY5 and ABA (via ABI3 and ABI4) are nodes of a plastid-modulated network that attenuates the response to shade in terms of hypocotyl elongation. In green plants with functional chloroplasts, light signals associated with canopy shade rapidly promote hypocotyl elongation via the phyB/PIF pathway. Exposure to low R/FR also triggers negative (growth-repressing) circuits involving the phyA/HY5 pathway and the CRT-derived hormone ABA, likely to prevent an excessive response and facilitate the return to non-shade conditions if the low R/FR signal disappears (e.g. if a commitment to the shade-avoidance lifestyle is unnecessary). When maintained, shade further causes a decrease of CHL and CRT contents, which might eventually disrupt chloroplast homeostasis. Such situation would be then signaled to feedback-regulate the response to the light signal by independently inhibiting phyB deactivation, increasing HY5 accumulation, and promoting ABA signaling. This mechanism connecting the metabolic status of the chloroplast with light (shade) signaling and developmental responses likely contributes to achieve optimal photosynthetic performance.

METHODS

Plant Material

All mutants used in this work are listed in Supplemental Table 1. Arabidopsis (Arabidopsis thaliana) lines used here were in the Columbia (Col) background with the only exception of hy1-1, a Landsberg erecta (Ler) mutant (Rodríguez-Concepción et al., 2004). Some of those lines were already available in our lab and previously used in published works, including hdr-3 (Pokhilko et al., 2015), gun1-101 (Llamas et al., 2017), gun5-1 (Llamas et al., 2017), pifQ (Toledo-Ortiz et al., 2010), hy5-2 (Bou-Torrent et al., 2015), psy-1 (Pokhilko et al., 2015), lut2 (Emiliani et al., 2018), aba2 (Ruiz-Sola et al., 2014), and hy5 35S:HA-HY5 (Toledo-Ortiz et al., 2014). Lines abi3-8 (Nambara et al., 2002), abi4-1 (Finkelstein et al., 1998), abi5-7 (Tamura et al., 2006), and 35S:GUS-PIF3 (Monte et al., 2004) were requested. For generation of double mutants, single homozygous plants were crossed and the F2 progeny was first screened for the characteristic long hypocotyl phenotype associated to the hy5 mutation in homozygosis. Long individuals were then PCR-genotyped to identify homozygous mutants for the second gene and confirm that they were also homozygous for hy5. For the generation of the 35S:HY5-GFP construct, the full coding region of the Arabidopsis HY5 complementary DNA (cDNA) was PCR-amplified using primers HY5‐attB1-F and HY5‐attB2-R (Supplemental Table 2) and cloned into Gateway pDONR-207. Cloning into Gateway pGWB405 eventually generated the construct for the 35S promoter-driven expression of a C-terminal fusion of the sGFP reporter protein to HY5. This construct was used to transform the hy5-2 mutant by floral dipping. The hy5 35S:HY5-GFP line used for the experiments reported here was selected based on complete complementation of the long hypocotyl phenotype associated with the hy5 mutation and high levels of nuclear GFP fluorescence. Line 35S:PHYB-GFP was generated by transforming Col-0 plants with the same construct previously found to work in an Arabidopsis phyB mutant in the Ler background (Yamaguchi et al., 1999). From the resulting transformants, we selected for further experiments one of the lines showing a clearer accumulation of the phyB-GFP protein in nuclear bodies under W.

Growth Conditions and Treatments

Seeds were surface-sterilized and germinated on solid Murashige and Skoog medium supplemented with 10 mg/mL of Suc to provide carbon and energy for albino seedlings to grow. When indicated, the medium was further supplemented with different concentrations or NF (Zorial), LIN (Sigma-Aldrich), or ABA (Sigma-Aldrich). Other chemicals added to the medium included epibrassinolide (1 μM), gibberellic acid (10 μM), picloram (5 μM), CPTA (25 μM), FSM (500 μM), CAP (50 μM), or PPT (100 μM). When comparing different lines (e.g. wild type versus mutant), they were grown together on the same plate instead of growing each line on a different plate. After stratification for at least 3 d at 4°C in the dark, plates were incubated in growth chambers at 22°C under W of 20 to 24 μmol m^–2 s^–1 PAR (R/FR = 1.6). When indicated, W was supplemented with FR provided by GreenPower LED module HF far-red (Philips) QB1310CS-670-735 light-emitting diode hybrid lamps (Quantum Devices) to simulate canopy shade (20–24 μmol m^–2 s^–1 PAR, R/FR = 0.02). Fluence rates were measured using a Spectrosense 2 meter associated with a 4-channel sensor (Skye Instruments) as described in Martínez-García et al. (2014). Grown seedlings were laid out flat on the growth media and digital images were taken to quantify hypocotyl length using the software ImageJ (NIH).

Microscopy

Whole 35S:PHYB-GFP seedlings germinated and grown under W for 7 d on media with or without 5 μM NF or 1 mM LIN were exposed to a 5-min pulse of FR (735 nm, 60 μmol/m2/s) and then kept in the dark. At different time points, treated seedlings were placed on glass slides under a safety green light and kept in the dark until observation with a BX60 FLUOVIEW FV300 microscope (Olympus). Confocal laser scan images of the hypocotyl area closer to the cotyledons were obtained at different time points in the dark with a combination of 488-nm laser excitation and 515-nm long-pass filter (LP515; Carl Zeiss). For each time point, three sequential images from different focus planes were recorded automatically.

ChIP

Approximately 800 μL of seeds from hy5 35S:HY5-GFP plants were plated on 8-square (10 × 10 cm) plates of Suc-supplemented medium. After growth for 2 d under W, four plates were left under W and four were transferred to W+FR for five additional days. For ChIP (Moon et al., 2008), each sample was divided in three aliquots after crosslinking and sonication: one input, one to be incubated with a 1:1000 dilution of anti-GFP antibody (A-11122; Invitrogen), and the last one to be processed similarly but without antibody. After DNA isolation, the three samples were used for RT-qPCR analysis of promoter sequence abundance with the primers shown in Supplemental Table 2. After normalization with the input, enrichment was calculated as the ratio of the signal with, versus without, antibody.

Gene Expression and Immunoblot Analyses

Total RNA was extracted from whole seedlings and used for RT-qPCR analysis as described in Llamas et al. (2017). Briefly, RNA was extracted with the Maxwell 16 LEV Plant RNA Kit (Promega) and quantified with a NanoDrop (Thermo Fisher Scientific). After producing cDNA with the cDNA Synthesis Kit (Roche), RT-qPCR was performed using LightCycler 480 SYBR Green I Master (Roche) with the gene-specific primers listed in Supplemental Table 2. Expression of target genes was normalized using UBC21 as the endogenous reference gene. Protein extraction, immunoblot analysis, and quantification of protein abundance were performed as described in Llamas et al. (2017) using a 1:1,000 dilution of anti-GFP serum (A-11122; Invitrogen).

Quantification of Metabolite Levels

Whole seedlings were frozen in liquid nitrogen, lyophilized, and ground in a mortar for extraction and quantification of photosynthetic pigments and ABA. CHL and CRT levels were measured either by spectrometric methods or by high-performance liquid chromatography (Bou-Torrent et al., 2015). ABA content was quantified by liquid chromatography-electrospray ionization-tandem mass spectrometry as described in Ruiz-Sola et al. (2014).

Accession Numbers

Sequence data from this article can be found in the Arabidopsis Genome Initiative or GenBank/European Molecular Biology Laboratory databases under the following accession numbers: ABA2 (AT1G52340), ABI3 (AT3G24650), ABI4 (AT2G40220), ABI5 (AT2G36270), GUN1 (AT2G31400), GUN5 (AT5G13630), HDR (AT4G34350), HY1 (AT2G26670), HY5 (AT5G11260), LUT2 (AT5G57030), PIF1 (AT2G20180), PIF3 (AT1G09530), PIF4 (AT2G43010), PIF5 (AT3G59060), PSY (AT5G17230).

Supplemental Data

• Supplemental Figure 1. Albino seedlings are able to elongate when treated with growth regulators.

• Supplemental Figure 2. NF treatment prevents shade-dependent accumulation of the GUS-PIF3 protein.

• Supplemental Figure 3. HY5 represses hypocotyl elongation in response to shade.

• Supplemental Figure 4. CRT levels in inhibitor-treated wild-type plants and CRT-defective mutants.

• Supplemental Figure 5. Shade-induced elongation responses in the absence of lutein.

• Supplemental Figure 6. ABA levels in shade-exposed and inhibitor-treated wild-type plants.

• Supplemental Table 1. Mutants used in this work.

• Supplemental Table 2. Primers used in this work.

Dive Curated Terms

The following phenotypic, genotypic, and functional terms are of significance to the work described in this paper:

• PAR1 Gramene: AT2G42870

• PAR1 Araport: AT2G42870

• HFR1 Gramene: AT1G02340

• HFR1 Araport: AT1G02340

• ATHB4 Gramene: AT2G44910

• ATHB4 Araport: AT2G44910

• HAT1 Gramene: AT4G17460

• HAT1 Araport: AT4G17460

• HAT2 Gramene: AT5G47370

• HAT2 Araport: AT5G47370

• HAT3 Gramene: AT3G60390

• HAT3 Araport: AT3G60390

• NIH Gramene: AT1G06670

• NIH Araport: AT1G06670

• SAL1 Gramene: AT5G63980

• SAL1 Araport: AT5G63980

• lut2 Gramene: AT5G57030

• lut2 Araport: AT5G57030

• ABA2 Gramene: AT1G52340

• ABA2 Araport: AT1G52340

• PIL1 Gramene: AT2G46970

• PIL1 Araport: AT2G46970

• hy5 Gramene: AT5G11260

• hy5 Araport: AT5G11260

• phyB Gramene: AT2G18790

• phyB Araport: AT2G18790

• HYPOCOTYL AmiGo: PO:0020100