Rewiring of jasmonate and phytochrome B signalling uncouples plant growth-defense tradeoffs

Plants resist infection and herbivory with innate immune responses that are often associated with reduced growth. Despite the importance of growth-defense tradeoffs in shaping plant productivity in natural and agricultural ecosystems, the molecular mechanisms that link growth and immunity are poorly understood. Here, we demonstrate that growth-defense tradeoffs mediated by the hormone jasmonate are uncoupled in an Arabidopsis mutant (jazQ phyB) lacking a quintet of Jasmonate ZIM-domain transcriptional repressors and the photoreceptor phyB. Analysis of epistatic interactions between jazQ and phyB reveal that growth inhibition associated with enhanced anti-insect resistance is likely not caused by diversion of photoassimilates from growth to defense but rather by a conserved transcriptional network that is hardwired to attenuate growth upon activation of jasmonate signalling. The ability to unlock growth-defense tradeoffs through relief of transcription repression provides an approach to assemble functional plant traits in new and potentially useful ways.

P lants continuously monitor external cues to tailor their growth, development and defensive capabilities in ways that optimize reproductive success, especially in environments where nutrients and light may be scarce. Theories to explain the costs and patterns of plant defense often invoke the existence of physiological tradeoffs that arise from allocation of limited resources to protective compounds and associated morphological structures [1][2][3] . Thus, a common interpretation of the 'dilemma of plants to grow or defend' 3 is that elevated expression of defense traits consumes metabolic resources at the expense of growth, whereas rapid growth, such as that triggered by shade light from competitors, diverts resources that could otherwise be invested in the defense arsenal. Although the concept of growth-defense tradeoffs is a major paradigm in ecological studies of plant resistance to herbivores and pathogens [1][2][3][4] , recent studies question the simplistic view that allocation of limited metabolic resources to one process necessarily reduces energetic expenditures in the other process [5][6][7][8][9][10] . These collective studies highlight the need to develop a more accurate conceptual framework for understanding how tradeoffs between growth and immunity optimize plant fitness in dynamic environments.
Here, we sought to exploit genetic tools in Arabidopsis thaliana to better understand how growth and defense are mechanistically integrated and to explicitly test whether the underlying signalling networks can be reconfigured to allow for simultaneous growth and defense. Our approach was based on the fact that defense responses mediated by the lipid-derived hormone jasmonate (JA) are associated with potent growth inhibition of leaves and roots [11][12][13][14] . We show that growth-defense tradeoffs are uncoupled in leaves of an Arabidopsis mutant (jazQ phyB) lacking five Jasmonate ZIM-domain (JAZ) transcriptional repressors and the photoreceptor phytochrome B (phyB). The concomitant robust growth and heightened anti-insect defense of jazQ phyB plants can be attributed, at least in part, to parallel activation of MYC and PIF transcription factors that in wild-type (WT) plants are repressed by JAZ and phyB, respectively. We further show that the capacity of jazQ phyB plants to grow and defend well simultaneously is associated with changes in leaf architecture and increased partitioning of carbon to leaf-area growth, but does not depend on increased leaf area-based photosynthesis. Thus, growth-defense antagonism in this system does not appear to be caused by constraints on the availability of metabolic resources that fuel growth and defensive processes but rather by a hormone-linked transcriptional network that is hardwired to restrict growth and upon activation of JA signalling. These collective findings highlight the importance of transcriptional repressor proteins in optimizing growth-defense balance, and further suggest that genetic modification of pathways that integrate defense and light signalling is a potential strategy to combine growth and defense traits in new ways.

Results
A jaz quintuple mutant exhibits constitutive JA responses. We devised a genetic screen to identify mutants of Arabidopsis that display enhanced JA-regulated defense against insect herbivory without an associated reduction in leaf growth. This screen was based on a signalling model 15 predicting that removal of JAZ repressor proteins would constitutively activate anti-insect defense and inhibit growth of vegetative tissues (Fig. 1a). Such a phenotype has not been previously observed for jaz loss-of-function mutants of Arabidopsis, most likely because of functional redundancy between the 13 members of the JAZ gene family 16 . We thus developed an Arabidopsis line (jaz quintuple or jazQ) with T-DNA insertion mutations in five JAZ genes (JAZ1/3/4/9/10) ( Supplementary Fig. 1). These JAZs were selected on the basis of their phylogenetic relationship, their demonstrated role in inhibiting various transcription factors (for example, MYCs) that execute defense responses, and their capacity to interact with DELLA proteins that antagonistically link JA signalling to gibberellic acid (GA)-mediated growth responses (Fig. 1a) [17][18][19][20][21][22] . Root growth assays showed that jazQ seedlings have both an increased sensitivity to exogenous JA and a constitutive short-root phenotype in comparison to WT seedlings ( Fig. 1b and Supplementary Fig. 1). We also compared the hormone sensitivity of jazQ seedlings to the jaz10-1 mutant, which is one of the few jaz single mutants to exhibit enhanced sensitivity to JA 11,23,24 . Consistent with the notion that JAZ family members serve partially redundant roles in JA signalling, jazQ roots were significantly more sensitive to JA than the jaz10-1 mutant ( Supplementary Fig. 1). These findings support a key role for multiple JAZs in the control of root growth. Our data are also consistent with recent genetic analysis of the co-repressor Novel Interactor of JAZ (NINJA), which directly interacts with JAZs to negatively regulate JA signalling in roots 14,25 . Analysis of defense-related phenotypes showed that glucosinolates and anthocyanins, whose biosynthesis in Arabidopsis is positively regulated by JA 22,26 , accumulate to higher levels in jazQ seedlings than WT (Fig. 1c,d). We also found that soil-grown jazQ plants had remarkably heightened resistance to attack by the generalist herbivore Trichoplusia ni (Fig. 1e). In contrast to these elevated defense traits, leaf area, petiole length and rosette dry weight were all reduced in jazQ relative to WT ( Fig. 1f and Supplementary Fig. 2). jazQ also delayed the time to bolting but did not affect the number of leaves at the time of bolting ( Supplementary Fig. 2). These results demonstrate that jazQ plants exhibit constitutive growth-defense antagonism (that is, reduced growth with enhanced defense) and thus provide a unique genetic model with which to interrogate how the JA branch of immunity is linked to growth.
Loss of PHYB uncouples growth-defense tradeoffs in jazQ. We visually screened an ethyl methanesulfonate (EMS) -mutagenized population of jazQ for mutants with increased rosette size and persistence of elevated leaf anthocyanin content. Among several suppressor of jazQ (sjq) mutants identified, one line (sjq11) showed a particularly striking leaf growth pattern that was heritable in subsequent generations (Fig. 2a). Importantly, bioassays performed with T. ni larvae showed that sjq11 plants also maintained heightened resistance to herbivory (Fig. 2b). Initial characterization of sjq11 revealed phenotypes similar to those described for phytochrome B (phyB) photoreceptor mutants, including early flowering time and elongated hypocotyls and petioles under continuous white light 27 . Genetic allelism tests and DNA sequencing confirmed that sjq11 harbours a null mutation in the PHYB gene ( Supplementary Fig. 3). To eliminate the possibility that additional EMS mutations contribute to the sjq11 phenotype, further studies were performed with a jazQ phyB sextuple mutant obtained by crossing the reference phyB-9 null allele into the jazQ background.
Analysis of growth and defense traits in jazQ phyB plants showed that the jazQ and phyB 'single' mutant phenotypes were largely additive and often organ specific. For example, jazQ phyB roots retained the JA-hypersensitive phenotype of jazQ, whereas jazQ phyB hypocotyls showed the red-light insensitive phenotype of phyB ( Supplementary Fig. 4). Adult jazQ phyB plants grown in soil resembled phyB in having elongated petioles, flattened rosette leaves 28 and early flowering time ( Fig. 2c and Supplementary  Fig. 5). The phyB mutation is thus epistatic to jazQ for these growth traits. Interestingly, the rosette diameter, projected leaf area and dry mass of jazQ phyB rosette leaves under our growth conditions exceeded that of the jazQ and phyB parents, suggesting that the combination of jazQ and phyB has transgressive effects on leaf growth ( Fig. 2d and Supplementary Fig. 5). Despite its robust vegetative growth, jazQ phyB plants maintained the heightened anti-insect defense and anthocyanin content of jazQ (Fig. 2e,f). The effect of combining jazQ and phyB on resistance to T. ni feeding was particularly striking because phyB alone causes high susceptibility to this herbivore ( Supplementary Fig. 6). jazQ is thus epistatic to phyB with respect to leaf defense traits. These collective data demonstrate that phyB fully suppresses the diminutive growth stature of jazQ aerial organs but, remarkably, does not compromise the heightened anti-insect resistance that is imparted by jazQ.
Co-expression of MYC and PIF regulons in jazQ phyB. A potential mechanistic explanation for the enhanced growth and defense attributes of jazQ phyB plants comes from recent studies that implicate crosstalk between the JA, phyB and GA signalling pathways in the modulation of growth-defense balance.
Within this signalling network, GA stimulates cell extension growth by promoting the degradation of DELLA proteins that repress PIF transcription factors 29 (Fig. 1a). Reciprocal antagonism between the JA and GA pathways involves JAZ-DELLA interactions that prevent these repressors from inhibiting their cognate transcription factors 19,20,30 . JA-GA crosstalk is integrated with the shade avoidance response through phyB-mediated perception of changes in the ratio of red to far red (R:FR) light. Low R:FR ratios indicative of leaf shading reduce phyB activity to relieve repression on PIFs, thereby promoting rapid growth through the concerted action of growth hormones such as auxin and brassinosteroid ( Fig. 1a) [31][32][33] . Concurrent with this growth response to plant competitors, inactivation of phyB by low R:FR (or phyB mutation) is associated with depletion of DELLA proteins, increased JAZ stability, accelerated turnover of MYC transcription factors, and suppression of JA-triggered immune responses 5,9,34 .
These considerations led us to test the hypothesis that the combination of jazQ and phyB causes concomitant derepression of the MYC and PIF transcriptional programs to promote growth T. ni larval weight (mg) and defense simultaneously. In support of this idea, we found that overexpression of PIF4 in the jazQ background partially rescued the small rosette size and short petiole length of jazQ without affecting anthocyanin accumulation and resistance to T. ni feeding ( Supplementary Fig. 7). This finding indicates that increased PIF4-mediated growth does not attenuate the defense stature of jazQ plants. The inability of PIF4 overexpression to fully recapitulate the jazQ phyB phenotype indicates that additional regulatory factors contribute to the growth vigour of jazQ phyB.
To further test the hypothesis that MYC and PIF transcriptional modules are simultaneously activated in jazQ phyB plants, we used mRNA sequencing to compare the transcript profile of WT, jazQ, phyB, and jazQ phyB seedlings. 'Secondary metabolism' and 'response to stress' and were among the biological processes most significantly overrepresented in ontologies of 257 genes expressed to higher levels in jazQ than WT ( Fig. 3 Fig. 8), as well as genes involved in the synthesis of triterpenoids, jasmonates, and various defensive proteins ( Supplementary Fig. 9). In agreement with their enhanced defense stature, jazQ phyB plants maintained increased expression of the majority (68%) of genes that are upregulated in jazQ (Fig. 3, Supplementary Figs 8 and 9). Analysis of growthrelated genes revealed that the set of 235 genes upregulated in both phyB and jazQ phyB is enriched for functional classes involved in responses to auxin, shade avoidance, cell wall organization and light stimulus (Fig. 3). Several genes within this group were previously shown to be direct targets for PIF transcription factor binding 33,35,36 (Supplementary Fig. 9). These data provide evidence that the combination of jazQ and phyB promotes concomitant expression of defense and growth-related genes, at least in part, via the concurrent activation of MYC and PIF transcriptional modules.
Among the 576 transcripts whose abundance was significantly increased in jazQ phyB but not jazQ or phyB, there was a strong over-representation of GO terms related to secondary metabolism, cell wall organization, growth and auxin transport (Fig. 3). These data suggest that the combination of jazQ and phyB has synergistic effects on the expression of certain growth and defense responses in jazQ phyB. In support of this idea, quantitative PCR analysis showed that wound-induced expression of select JA-response genes was significantly higher in jazQ phyB than WT leaves, which may contribute to the heightened defense of jazQ phyB plants relative to WT ( Supplementary Fig. 10). It is possible that these synergistic effects of jazQ and phyB on gene expression result from functional interaction between MYCs and PIFs at the level of protein-protein interaction or altered binding to common cis-regulatory elements in target genes 21,33,34,37 . Supporting this hypothesis, PIF4 targets previously identified by ChIP-seq 33 include an over-representation of genes associated with the terms 'jasmonate stimulus' and 'response to wounding' (Supplementary Table 1). Moreover, several PIF4 targets are also targets of MYC2 (Supplementary Table 2). We cannot exclude the possibility that concurrent loss of phyB and JAZ1/3/4/9/10 affects  the activity of transcriptional regulators other than PIFs and MYCs to contribute to the gene expression profile and other general phenotypes of jazQ phyB plants.
Genetic modulation of photosynthesis and leaf architecture.
The robust growth of well-defended jazQ phyB leaves led us to investigate whether jazQ and phyB interact to modulate leaf photosynthetic efficiency. Indeed, PIF activity is known to repress chloroplast development and photosynthetic competency 37 , and we observed that 'photosynthesis' was the term most significantly overrepresented among genes that are repressed in both phyB and jazQ phyB seedlings (Supplementary Fig. 11 and Supplementary Data 1). We employed non-invasive, whole-plant chlorophyll fluorescence imaging 13,38 to determine how genetic perturbations within the phyB-GA-JA signalling network affect photosystem II efficiency (F II ) under various light regimes, including those designed to simulate natural environments ( Fig. 4a and Supplementary Fig. 12 39 . A decrease in F II was also observed in a Col-0 transgenic line (35S:PIF4) that overexpresses PIF4. Interestingly, the negative effect of phyB and 35S:PIF4 on photosynthetic efficiency was rescued by jazQ, which alone had little (or very weak positive) effect on F II (Fig. 4a and Supplementary Fig. 12). Consistent with the role of PIFs in repressing photosynthesis, a pif1/3/4/5 quadruple mutant (pifq) showed increased F II under fluctuating light conditions, whereas loss of DELLAs in the della quintuple mutant (dellaQ) reduced F II under these conditions. That photosystem II efficiency was lower in phyB than in dellaQ leaves suggests that phyB exerts stronger repression on PIF activity than DELLA proteins under these growth conditions.
To obtain additional insight into physiological processes that underlie growth-defense vigour of jazQ phyB, we investigated the relationship between photosynthesis and leaf growth to obtain an estimate of leaf construction costs 40 . Gas exchange experiments showed that phyB leaves have significantly lower photosynthetic rate per unit leaf area, whereas photosynthetic capacity of jazQ on a leaf area or dry weight basis was comparable to WT (Fig. 4b), consistent with our chlorophyll fluorescence measurements. phyB leaves also contained less area-based chlorophyll and Rubisco (D-ribulose-1,5-bisphosphate carboxylase/oxygenase) than WT. Modelling of photosynthetic parameters showed that the reduced photosynthetic capacity of phyB at high light results in part from a limitation in Rubisco activity. jazQ partially rescued the low photosynthetic capacity of phyB leaves, as well as the low area-based Rubisco and chlorophyll content of phyB (Fig. 4c,d). The mechanism by which jazQ partially restores the reduced photosynthetic performance of phyB leaves in unknown but deserves further attention.
We also found that phyB leaves were thinner than those of WT and jazQ and that this trait was retained in jazQ phyB (Fig. 4e). Because of the greater projected leaf area available to intercept light (due to longer petioles and flatter, thinner leaves), the whole-plant photosynthetic rate in jazQ phyB plants was similar to WT. Thus, costs associated with construction of jazQ phyB leaves may be lowered through increased partitioning of carbon to leaf area at the expense of leaf thickness 40 . These data suggest that changes in leaf architecture rather than increased efficiency of the photosynthetic apparatus may contribute to the growth-defense vigour of jazQ phyB relative to WT plants.

Discussion
A key innovation of our study was the development of a JA signalling mutant (jazQ), in which the removal of multiple JAZ repressors causes hyperactivation of JA responses. As a consequence, jazQ plants exhibit both enhanced resistance to insect herbivory and diminished growth of leaves and roots. The heightened defense of jazQ plants most likely involves increased activity of MYC transcription factors, which are well characterized for their role in promoting secondary metabolism and defense 21,22 . Our results also provide evidence that transcriptional changes account for the growth phenotypes of jazQ, but the molecular basis of growth restraint remains to be determined. DELLA proteins have previously been implicated in JA-mediated growth inhibition of Arabidopsis roots 19 and hypocotyls 20 . Although these studies raise the possibility that removal of JAZs in jazQ increases the growth-inhibiting activity of DELLAs, previous studies showed that DELLA proteins are not required for wound-and JA-induced growth stunting of leaves 12 . Given the high connectivity of JAZs within protein interaction networks and their emerging role as integrators of various signalling pathways 41,42 , it is possible that JAZ-interacting transcriptional regulators other than MYCs and DELLAs also contribute to general phenotypes of jazQ. Further analysis of jazQ should provide new avenues to study how JA controls plant growth and development.
By exploiting jazQ phenotypes in a genetic suppressor screen, we identified jazQ phyB as a novel genotypic combination that uncouples growth-defense tradeoffs mediated by the JA signalling pathway. This major conclusion is based on the finding that phyB suppresses the growth restriction but not enhanced defense of jazQ leaves. Previous studies of phyB-JA crosstalk showed that decreased levels of the active form (Pfr) of phyB, as a consequence of either phyB mutation or exposure to shade light, compromises JA-mediated resistance to a broad spectrum of biotic attackers 5,31,32 . Recent molecular evidence further indicates that shade-triggered repression of JA-mediated defense in Arabidopsis involves stabilization of JAZs and reduced MYC activity 9,34 . Given these findings, a plausible mechanistic explanation for why jazQ-mediated leaf defense remains elevated in the presence of phyB is that one or more of the mutated JAZ genes in jazQ (JAZ1/3/4/9/10) are critical for suppressing JA-mediated defenses when levels of active phyB are reduced (for example, in the phyB genetic background). Consistent with this, analysis of the jaz10-1 single mutant has shown that JAZ10 is required for attenuation of JA-mediated defense by shade light 9,43,44 . Future studies are needed to ascribe specific functions to individual JAZ family members and, more generally, to better understand how light and defense perception pathways are integrated to achieve growth-balance in specific tissues.
A central premise underlying current views of growth-defense balance is that the production of defensive compounds and associated morphological structures is energetically costly. The ability of jazQ phyB leaves to grow and defend well at the same time provides evidence that JA-mediated growth-defense antagonism is not simply a consequence of 'metabolic competition' that shunts resources to defense at the expense of growth 6,7,45 . Our data support an alternative (but not mutually exclusive) model, in which tradeoffs between growth and defense are orchestrated by a hormone-based transcriptional network that is hardwired to restrict growth upon activation of the JA signalling pathway. This interpretation is bolstered by an increasing number of molecular studies that have addressed to role of hormones in the intersection of growth and immunity 4,20,46,47 . In light of current theories regarding the costs of induced defense 45 , it remains to be determined whether growth inhibition resulting from genetic removal (for example, jazQ) or natural JA-induced degradation of JAZs provides a fitness benefit to the plant. It is conceivable, for example, that JA-mediated growth restriction enhances the plant's capacity to anticipate future threats or to cope with multiple-stress conditions 48 . This hypothesis predicts that simultaneous loss of phyB and JAZ, as major drivers of phenotypic plasticity, may reduce the fitness of jazQ phyB plants in natural environments because these plants lack the regulatory potential to adapt to complex and rapidly changing environments. Support for this idea comes from our RNA-seq data showing that the expression of genes involved in abiotic stress response is repressed in jazQ phyB ( Supplementary Fig. 11). An important future direction will thus be to assess the reproductive fitness of these genotypes in the presence of various predators, competitors and abiotic stressors.
The ability to unlock growth-defense tradeoffs through relief of transcription repression may provide an approach to assemble plant traits in ways that have practical application in biotechnology and agronomy. One example concerns the use of exogenous JA to elicit the biosynthesis of commercially valuable plant secondary metabolites 49,50 . A significant bottleneck in JA-elicited production of these compounds, including the anti-cancer drug paclitaxel (taxol), is the loss of plant biomass resulting from JA-mediated growth inhibition 50,51 . The design of plants with increased flux through JA-responsive metabolic pathways without concomitant growth restriction could provide a solution to this problem. Similarly, unlinking growth-defense antagonism may be useful in cropping systems that increasingly depend upon high planting densities to maximize yield. Because the ratio of R:FR light perceived by canopy leaves decreases as planting densities increase, densely planted crops may become more susceptible insect pests and pathogens that are sensitive to JA-mediated defenses 31,32 . In demonstrating that jazQ and phyB can be combined to produce what is essentially a well-defended shade leaf, our results suggest a strategy to cultivate densely planted crops with less dependence on pesticides. The evolutionary conservation of light and defense signalling pathways suggest that the findings described here can readily be applied to most plant species using modern genetic approaches.
Root growth assays. The effect of exogenous JA on seedling root growth inhibition was determined by growing seedlings on square Petri plates (Fisher) containing MS medium supplemented with the indicated concentration of methyl-JA (MeJA; Sigma-Aldrich) 57 . Plates were incubated vertically in a growth chamber maintained at 21°C under continuous light for 8 days. Primary root length was measured using the ImageJ software (http://imagej.nih.gov/ij/). WT and mutant lines were grown on the same plate to control for plate-to-plate variation.
Quantification of secondary metabolites. Anthocyanins were quantified using established procedures 58 , with the following modifications. Petioles were excised from 4-week-old plants and extracted in 1 ml MeOH containing 1% (v/v) HCl. Samples were incubated overnight at 4°C with constant agitation. Anthocyanin pigments in the resulting extract were measured spectrophotometrically and calculated as A 530 À 0.25(A 657 ) g À 1 fresh weight. Glucosinolates were quantified using 8-day-old seedlings grown on solid MS medium 59 . Seedlings were collected into 2 ml tubes (B50 seedlings per tube) and immediately frozen in liquid nitrogen. WT and mutant lines were grown on the same plate to avoid plate-to-plate variation. Frozen tissue was lyophilized, ground to a fine powder and extracted with 1 ml 80% MeOH containing an internal standard (25 nmol sinigrin, Sigma-Aldrich). Samples were briefly vortexed, incubated at 75°C for 15 min, and then centrifuged at 23°C at 10,000g for 10 min. Resulting supernatants were applied to Sephadex A-25 columns (Amersham). Glucosinolates were released from the columns as desulfoglucosinolates with a solution containing 30 ml of aryl sulfatase (3.0 mg ml À 1 ; Sigma) and 70 ml water (high-performance liquid chromatography grade). Following an overnight incubation in the dark at 23°C, samples were eluted from the columns with 200 ml 80% MeOH and 200 ml water. Samples were then lyophilized to complete dryness and re-dissolved in 100 ml water. Desulfoglucosinoaltes were detected by high-performance liquid chromatography 59 . Compound abbreviations in Fig. 1c correspond to the following: 3MSP, 3-methylsulfinylpropylglucosinolate; 4MSB, 4-methylsulfinylbutylglucosinolate; 7MSH, 7-methylsulfinylheptylglucosinolate; 4MTB, 4-methylthiobutylglucosinolate; 8MSO, 8-methylsulfinyloctylglucosinolate; I3M, indol-3-ylmethylglucosinolate; 4MI3M, 4-methoxyindol-3-ylmethylglucosinolate; 1MI3M, 1-methoxyindol-3-ylmethylglucosinolate and 8MTO, 8-methylthiooctylglucosinolate.
Insect-feeding assays. Insect-feeding assays were performed with soil-grown plants maintained in a growth chamber at 19°C and a photoperiod of 8 h light (120 mE m À 2 s À 1 ) and 16 h dark. Neonate Trichoplusia ni larvae (Benzon Research) were transferred to the centre of fully expanded rosette leaves of 6-week-old plants 60 . Four larvae were reared on each of 12 plants per genotype. Plants were then covered with a transparent dome and returned to the chamber for 10 days, after which larval weights were measured.
Growth and flowering time measurements. Three-to-four week-old soil-grown plants were used for all measurements (typically 10 plants per measurement), unless indicated otherwise. Petiole length of the third true leaf was measured with a caliper after leaf excision. Bolting time was measured in a separate set of plants by counting the number of true leaves on the main stem and the number of days from sowing until bolting (that is, floral buds visible in the centre of the rosette). The same set of plants was subsequently used to assess the length of time to opening of the first flower. Rosette diameter and leaf area were determined by photographing rosettes from the top with a Nikon D80 camera. The resulting images were used to calculate Feret diameter using ImageJ analysis. Total leaf area was determined with GIMP software (http://www.gimp.org). Leaf dry weight was determined by weighing excised rosettes (without roots) after freeze drying for 2 days in a lyophilizer.
jazQ suppressor screen and identification of sjq11. Approximately 50,000 jazQ seeds were mutagenized by immersion in a solution of 0.1% or 0.2% (v/v) EMS (Sigma-Aldrich) for 16 h at room temperature, with constant agitation. Seeds (M 1 generation) were thoroughly washed with H 2 O, stratified in the dark at 4°C for 2 days and then immediately sown on soil. M 2 seed was collected from 16 pools of self-pollinated M 1 plants (approximately 1,000 M 1 plants/pool). Soil-grown M 2 plants (B2,000 plants/pool) were visually screened for individuals having a larger rosette size than jazQ. Putative sjq (suppressors of the jazQ) mutants were rescreened in the M 3 generation to confirm heritability of phenotypes. Insight into the causal mutation in sjq11 came from the observation that sjq11 seedlings grown on MS medium in continuous white light for 3 days have elongated hypocotyls. Subsequent hypocotyl growth assays in monochromatic red light 61 confirmed a defect in red-light signalling. Briefly, sjq11 (M 3 generation) and control seeds were plated on MS medium lacking sucrose and stratified at 4°C in dark for 2 days. Mutant and control lines were grown on the same plate to control for plate-to-plate variation. A 3 h pulse of white light was then administered to improve synchronous seed germination. Plates were then returned to darkness for one d at 21°C and then transferred to a monochromatic light-emitting diode chamber outfitted to emit red light (670 ± 20 nm; 25 mE mE m À 2 s À 1 ). As a control, a set of plates containing each genotype was maintained in darkness. Following 3 days of growth, seedling hypocotyls were measured by ImageJ software analysis of scanned images. Allelism tests performed with F 1 seedlings (obtained from the cross between sjq11 and phyB-9) revealed a lack of genetic complementation. Finally, sequencing of the NATURE COMMUNICATIONS | DOI: 10.1038/ncomms12570 ARTICLE PHYB gene (AT2G18790) in sjq11 revealed a C-T transition that introduces a stop codon in a region of the gene that encodes the chromophore-binding domain of phyB.
Gene expression profiling. Global gene expression profiling in 8-day-old whole seedlings (Col-0 WT, jazQ, phyB-9, jazQ phyB-9) was assessed by mRNA sequencing (RNA-seq) performed on the Ilumina HiSeq 2000 platform. Seedlings were grown in continuous light on solid MS medium supplemented with sucrose. For each replicate sample, B200 seedlings were harvested for RNA extraction. WT and mutant seedlings were grown on the same plate to minimize plate-to-plate variation. Three independent RNA samples (biological replicates) were sequenced per genotype. Total RNA was isolated as described above and RNA integrity was assessed with a 2100 Bioanalyzer (Agilent Technologies). All samples utilized had an integrity score of at least 7.0. Single-end (50 bp)  For wounding experiments, 3-week old soil-grown seedlings were wounded twice across the midvein of four leaves (leaves 3-6, counted from first rosette leaf). After 1 h, the wounded leaves of two plants were pooled and immediately frozen in liquid nitrogen. Equivalent leaves of two unwounded plants were pooled and collected as controls. The experiment was independently replicated twice, with each experiment consisting of 3-4 biological replicates. Frozen tissue was homogenized with a TissueLyser II (Qiagen) and RNA was extracted using an RNeasy kit (Qiagen) with on-column DNase (Qiagen) treatment, as per the manufacturer's instructions. RNA quality was assessed by A 260 /A 280 ratios using a ND-1000 UV Nanodrop spectrophotometer (Thermo Scientific). cDNA was reverse transcribed using a High-Capacity cDNA Reverse Transcription kit (Applied Biosystems, ABI), as per the manufacturer's instructions, and cDNA was diluted to 0.5 ng ml À 1 with nuclease-free water. Quantitative real-time PCR reactions consisted of 5 ml of 2 Â Power SYBR Green (ABI) master mix, 2 ul diluted cDNA template (1 ng total), 1 ml 5 uM forward and reverse primers, and nuclease-free water for 10 ml total reaction volume. The forward and reverse primers used were the following: PP2A, 5 0 -AAGCAGCGTAATCGGTAGG-3 0 and 5 0 -GCACAGCAATCGGGTATAAA G-3 0 ; AOS, 5 0 -GGAGAACTCACGATGGGAGCGATT-3 0 and 5 0 -GCGTCGTGGC TTTCGATAACCAGA-3 0 ; LOX3, 5 0 -GCTGGCGGTTCGACATG-3 0 and 5 0 -GCC ATTCCTCTGCGAATTAGA-3 0 ; and MYC2, 5 0 -AGAAACTCCAAATCAAG AACCAGCTC-3 0 and 5 0 -CCGGTTTAATCGAAGAACACGAAGAC-3 0 . Reactions were run on an ABI 7500 Fast qPCR instrument with the following conditions: 95°C for 10 min, then 40 cycles of 15 s at 95°C for denaturation and 60 s at 60°C for annealing and polymerization. A dissociation curve was performed at the end of each reaction to confirm primer specificity using default parameters (15 s at 95°C, 60 s at 60°C-95°C in 1°C increments, and 15 s at 95°C). Target gene expression was normalized to the expression of PP2a, which is stable under JA-inducing conditions 13 . The normalization incorporated primer efficiencies determined for each primer pair using LinRegPCR v2012.0 (ref. 65) from the log-linear phase of each amplification plot.
Overexpression of PIF4 in the jazQ background. The 35S::PIF4-TAP overexpression construct 66 was kindly provided by Dr Michael Thomashow (Michigan State University). Transformation of jazQ plants with Agrobacterium tumefaciens (strain C58C1) was performed using the flower dip method 67 . Multiple independent transformed lines (T1 generation) were selected on MS plates containing gentamycin and transferred to soil for subsequent analysis. Homozygous lines were selected by testing the T3 progeny for gentamycin resistance.
Photosynthesis measurements. Gas exchange measurements were performed with plants grown in plastic containers ('Cone-tainers', Steuwe and Sons, Tangent, OR, USA) on an 8 h light (19°C)/ 16 h dark (16°C) photoperiod and 120 mmol m À 2 s À 1 photosynthetic photon flux density (PPFD) 40,68 . Single mature rosette leaves (attached) from 8-to 10-week-old plants were used to obtain CO 2 response curves on a LI-6400XT system (LI-COR Biosciences, Lincoln, NE, USA) outfitted with a standard leaf chamber (chamber area ¼ 6 cm 2 ). Leaves were supplied with an artificial air mixture consisting of 20% O 2 , 80% N 2 , and 400 ppm CO 2 at intensity of light 500 mmol m À 2 s À 1 . Leaf temperature was maintained at B20°C (block temperature set to 18°C). Leaves were acclimated under this condition for at least 30 min before the start of each experiment. Assimilation rates were normalized to projected leaf area as measured by image analysis with the GIMP software. Area-and whole-plant-based photosynthesis and respiration was determined at four time points of the Arabidopsis growth cycle using plants grown under short-day conditions 40 .
In situ chlorophyll fluorescence measurements were performed in a Percival AR41L2 (Geneva Scientific, http://www.geneva-scientific.com) refitted as a Dynamic Environment Photosynthesis Imager (DEPI) system 13,38 . Images were processed using visual phenomics software 69 . The quantum yield of PSII (FII) was calculated as (F' M À F S )/F' M , where F S is the steady-state fluorescence and F' M is the fluorescence maximum at steady state. Statistical analysis of the FII data is provided in Supplementary Fig. 12.
Leaf thickness measurements. Leaf cross sections obtained from the 5th leaf of 22-day old rosette leaves were examined under an Olympus FluoView FV1000, Confocal Laser Scanning Microscope (Olympus, NJ, USA) in the Center for Advanced Microscopy, Michigan State University. Leaf thickness was measured as the distance between the abaxial and adaxial surfaces of the leaf 40 .
Measurement of total chlorophyll and Rubisco. Chlorophyll was extracted from 54-d old Arabidopsis rosette leaves with 96% ethanol. Absorbance of the extracted chlorophyll was measured spectrophotometrically at 665 nm and 649 nm and the total chlorophyll was calculated using the following equation: Chl a þ Chl b ¼ (13.95 A 665 -6.88A 649 ) þ (24.96A 649 -7.32A 665 ) 70 . Total protein was extracted from 54-d old Arabidopsis rosette leaves using a Plant Total Protein Extraction Kit (Sigma-Aldrich, MO, USA). A modified Lowry Assay was performed to measure the total protein concentration in the extract and the purity and quality of the extracted protein were determined by denaturing polyacrylamide gel electrophoresis. Equal amounts of total protein were loaded onto an automated capillary-based size western blotting system (ProteinSimple Wes System, San Jose CA, USA). All procedures were performed with manufacturer's reagents according to their user manual. Protein separation and immunodetection were performed automatically on the individual capillaries using the default settings. Antibodies raised against the large subunit of Rubisco (rabbit, AS03 037; Agrisera, Sweden; dilution used 1:650) were used to detect Rubisco in each protein sample. For quantification, all subsequent data generated was analysed with Compass Software provided by manufacturer (ProteinSimple, San Jose CA). Peak heights of the fluorescence signals were used to calculate relative differences of Rubisco concentration between samples. Rubisco concentration per unit leaf area was calculated based on the total protein concentration and measurements of leaf area per unit mass.
Data Availability. RNA sequencing data is deposited at the National Center for Biotechnology Information Gene Expression Omnibus (GEO) as series record GSE79012. The authors declare that all other data supporting the findings of this study are available within the article and its Supplementary Information files or are available from the corresponding author upon request.