PIFs get BRright: PHYTOCHROME INTERACTING FACTORs as integrators of light and hormonal signals




Light and temperature, in coordination with the endogenous clock and the hormones gibberellin (GA) and brassinosteroids (BRs), modulate plant growth and development by affecting the expression of multiple cell wall- and auxin-related genes. PHYTOCHROME INTERACTING FACTORS (PIFs) play a central role in the activation of these genes, the activity of these factors being regulated by the circadian clock and phytochrome-mediated protein destabilization. GA signaling is also integrated at the level of PIFs; the DELLA repressors are found to bind these factors and impair their DNA-binding ability. The recent finding that PIFs are co-activated by BES1 and BZR1 highlights a further role of these regulators in BR signal integration, and reveals that PIFs act in a concerted manner with the BR-related BES1/BZR1 factors to activate auxin synthesis and transport at the gene expression level, and synergistically activate several genes with a role in cell expansion. Auxins feed back into this growth regulatory module by inducing GA biosynthesis and BES1/BZR1 gene expression, in addition to directly regulating several of these growth pathway gene targets. An exciting challenge in the future will be to understand how this growth program is dynamically regulated in time and space to orchestrate differential organ expansion and to provide plants with adaptation flexibility.

I. Introduction

After germination, plant life entails a constant adaptation to the changing environment and, for this to occur, perceived external information must be integrated with endogenous developmental programs to adjust growth. This exchange of information relies on the fine control of transcription, mediated by nuclear factors that determine when and where a gene must be expressed. One of the best-characterized families of transcription factors known to modulate plant growth is the PHYTOCHROME INTERACTING FACTORS (PIFs). PIF3, the founder member of this gene family, was identified during screening for interactors of the C-terminal domain of phytochrome B (Ni et al., 1998). Mis-expression of this factor leads to impaired response to red (R) and far-red (FR) light, suggestive of a PIF3 role in the early steps of light signal transduction (Kim et al., 2003; Monte et al., 2004). The PIF family comprises 15 members (Toledo-Ortiz et al., 2003), mostly found in light signaling, although a more prevailing role has been described for PIF1, PIF3, PIF4 and PIF5 (Ni et al., 1998; Huq et al., 2000; Oh et al., 2004).

Analyses of PIF mutants and over-expression lines suggest a role of these factors in the positive regulation of plant growth (Duek & Fankhauser, 2005; Leivar et al., 2008). PIFs activate gene expression by interacting with G-box (CACGTG) motifs in the promoters of their gene targets, with a main function in cell expansion (Toledo-Ortiz et al., 2003; Leivar et al., 2008; de Lucas et al., 2008; Bai et al., 2012). Individually, these factors also regulate other processes, including chloroplast differentiation, seed germination and flowering (Oh et al., 2009; Stephenson et al., 2009; Kumar et al., 2012).

During the last few years, it has become clear that, in addition to light, other signals regulate the activity of PIFs. The circadian clock (Nozue & Maloof, 2006) and temperature (Koini et al., 2009) regulate PIF gene expression, and plant hormones, such as gibberellin (GA) (de Lucas et al., 2008) and brassinosteroids (BRs) (Bai et al., 2012), affect PIF transcriptional activity. Such a complex network of regulators underscores the pivotal role of PIFs in plant growth and development, and, in this review, we discuss the recent findings concerning the function of this family of regulators, which represent a wonderful example of how transcription is regulated by cross-talk by both internal and external cues.

II. Light and phytochrome signaling

The quantity and quality of sunlight vary according to diurnal, seasonal and local variations. Plants perceive light of different wavelengths through a diverse set of photoreceptors which, when light activated, initiate specific transduction pathways that modify growth and development according to the environment. UV-B light (280–315 nm) is sensed by the recently discovered UVR-8 protein (Rizzini et al., 2011). The phototropins PHOT1–2 (Briggs & Christie, 2002), cryptochromes CRY1–3 and Zeitlupe (ZTL) family (Kim et al., 2007) of photoreceptors perceive UV-A (315–400 nm) and blue (400–495 nm) light, whereas light in the R and FR spectrum is perceived by the phytochromes (Quail et al., 1995).

Phytochromes are present in all green plants, as well as in cyanobacteria, non-photosynthetic bacteria and fungi (Karniol et al., 2005). These photoreceptors exist in two thermally stable conformations, depending on light irradiation: an inactive 660 nm R light-absorbing form (Pr) and an active 730 nm FR light-absorbing form (Pfr). In Arabidopsis, they comprise a family of five members (PhyA–PhyE) (Clack et al., 1994). The N-terminal region of the molecule, including the GAF phytochromobilin-binding domain, functions as the light perception module, whereas the C-terminal region, encompassing the PAS and histidine kinase-related (HKRD) domains, is involved in dimerization. phyA is only stable in the dark and belongs to the light-labile or type I category, while phyB–E are light stable or type II, with phyB having a more prevalent role in R light (Rockwell & Lagarias, 2006).

In the dark, phytochromes are mainly localized in the cytosol, where they were shown to interact with PKS1 (Phytochrome Kinase Substrate 1) and NDPK2 (Nucleoside Diphosphate Kinase 2) (Choi et al., 1999; Fankhauser et al., 1999). PKS1 binds the phototropin phot1 and the NON-PHOTOTROPIC HYPOCOTYL 3 (NPH3) scaffold protein at the plasma membrane, being shown to play a critical role in phototropism (Lariguet et al., 2006). NDPK2 binds the mitogen-activated protein (MAP) kinases, AtMAPK3 and 6, with a role in reactive oxygen species (ROS) signaling (Moon et al., 2003), and in addition has been shown to function as a GTPase-activating protein (GAP), thus connecting phytochrome with G protein-mediated signaling (Shen et al., 2008). In mosses, a direct interaction of Pfr Ppphy4 and phototropins at the plasma membrane would also mediate the chloroplast relocation and tropic protonemal reactions to R light (Jaedicke et al., 2012). Interestingly, a similar interaction of Arabidopsis phyA and phot1 was observed in onion epidermal cells, but not in yeast cells, suggesting that, in higher plants, PKS1 is required for this interaction. Cytosolic Pfr phytochromes have also been reported to inhibit protochlorophyllide oxidoreductase A (PORA) mRNA translation (Paik et al., 2012), the function of these photoreceptors in the cytoplasm being covered in a recent review (Hughes, 2013).

In higher plants, light perception promotes active Pfr phytochrome migration into the nucleus, in a process assisted by different shuttle facilitators (Yamaguchi et al., 1999; Chen et al., 2004; Fankhauser & Chen, 2008). Nuclear accumulation of phyB is induced by R light and, to a lesser extent, by continuous blue light, but not by FR light. By contrast, phyA nuclear import is triggered by a brief pulse of R, FR or blue light. Translocation of phyA is very rapid (within minutes) and requires the small proteins FHY1 (FAR RED ELONGATED HYPOCOTYL 1) and FHL (FHY1 LIKE), shown to share a conserved nuclear localization signal (NLS) and nuclear exclusion signal (NES) and a C-terminal septin-related domain (SRD) (Zhou et al., 2005; Genoud et al., 2008). FHY1 and FHL interact with the Pfr form of phyA through their SRD motifs, phyA using the NLS of these adaptor proteins for its nuclear transport (Rösler et al., 2007). phyB translocation, by contrast, is much slower and does not require FHY1/FHL. The C-terminal phyB PRD domain (amino acids 594–917) harbors a putative NLS that is both necessary and sufficient for nuclear localization (Chen et al., 2005). The finding that this domain and the GAF-PHY photosensory domains interact in a light dependent manner led to the postulation that Pfr photoconversion promotes phyB nuclear transport by unmasking this NLS signal (Chen et al., 2005). However, more recent studies, performed in Acetabularia (no functional phytochrome has been identified in this alga), showed that the PIF transcription factors may play a relevant role in facilitating phyB translocation (Pfeiffer et al., 2012).

On import into the nucleus, phyA and phyB localize into speckles or nuclear bodies, also known as photobodies (Van Buskirk et al., 2012). phyB forms two types of nuclear body: the early photobodies, which form within 15 min of light exposure and co-localize with PIFs, and the late photobodies, which are larger and more stable, and are observed after longer (2–3 h) R light treatments, once PIFs have been degraded (Bauer et al., 2004). Genetic screens for mutants defective in phyB photobody formation identified the HEMERA (HMR) protein which is structurally related to the multiubiquitin receptor RAD23 and is localized to the periphery of photobodies (Chen et al., 2010). Interestingly, hmr mutants are albino and exhibit a tall phenotype in R and FR light, have been shown to accumulate phyA, PIF1 and PIF3 in the light, supporting a role for nuclear bodies as sites of protein turnover.

phyA co-localizes with CONSTITUTIVE PHOTOMORPHOGENIC 1 (COP1) in early nuclear bodies, the direct interaction of these proteins via the phyA PRD and COP1 WD40-repeat domains mediating rapid ubiquitination and destabilization of phyA in R light (Seo et al., 2004; Saijo et al., 2008). The COP1/SPA E3 ligase complex targets several photomorphogenesis-promoting factors, including HY5 (ELONGATED HYPOCOTYL5), HYH (HY5-HOMOLOG), LAF1 (LONG AFTER FAR-RED LIGHT1) and HFR1 (LONG HYPOCOTYL IN FAR-RED LIGHT1), for proteasomal degradation. These factors are stabilized by phytochromes in the light via the inactivation of COP1 (Osterlung et al., 2000; Lau & Deng, 2012), although the mechanisms governing this regulation are not fully understood. Direct interaction of phytochromes and the COP1/SPA proteins is thought to cause a rapid, initial inactivation of COP1, whereas long-term inactivation is achieved by the exclusion of the COP1 protein from the nucleus. Nuclear COP1 also mediates phyB degradation by preferentially binding the N-terminal region of Pfr phyB. It is noteworthy that phyB–PIF interaction stimulates COP1-catalyzed phyB polyubiquitination and degradation (Jang et al., 2010). Consistent with these results, nuclear phyB levels are higher in pif and cop1 mutants, indicating that PIFs and phyB mutually regulate their nuclear abundance.

III. Phytochrome regulation of PIF activity

In the nucleus, phytochromes exert their biological activity by affecting the expression of c. 10% of the Arabidopsis genes (Tepperman et al., 2006; Leivar et al., 2009). This control is achieved via interaction with different families of transcription factors, of which the PIFs are the best characterized (Toledo-Ortiz et al., 2003). phyB has been reported to interact with seven members of this gene family (PIF1/PIF3-LIKE 5, PIF3, PIF4, PIF5/PIL6, PIF6/PIL2, PIF7 and PIF8), while two of these members (PIF1/PIL5, PIF3) also bind phyA (Huq et al., 2004; Khanna, 2004). The function of the other family members, that is, PIL1, SPATULA (SPT), ALCATRAZ (ALC), HFR1, bHLH23, bHLH56, bHLH119 and bHLH127 (basic helix–loop–helix), does not seem to be associated with these photoreceptors.

Binding to phytochromes promotes the phosphorylation of the PIF proteins, with 10 min of R light being sufficient for PIF3 phosphorylation (Bauer et al., 2004). PIF phosphorylation is the primary event triggered by phytochromes, priming the subsequent polyubiquitination and degradation of these proteins (Al-Sady et al., 2006). However, neither the kinase(s) responsible for phosphorylation of PIFs nor the E3 ubiquitin ligase(s) responsible for their ubiquitination have been identified. Likewise, the nature of a putative phosphatase hypothesized to dephosphorylate PIFs under shade conditions remains unknown.

PIF3 isolated from seedlings grown in R light was found to be phosphorylated in multiple serine/threonine (Ser/Thr) sites (Ni et al., 2013) and these phosphorylation events act collectively to trigger rapid ubiquitination and degradation of the protein. Mutation of these residues however, did not affect phyB interaction or the DNA-binding ability of PIF3. Formation of nuclear aggregates is not affected by the PIF3 phosphorylation state, although phosphomimic mutations result in the degradation of this factor in the dark. Phosphorylation of PIF3 is on the other hand required for negative feedback modulation of phyB levels, attenuated sensitivity to light on prolonged light exposure probably being the consequence of the co-degradation of both proteins (Ni et al., 2013). phyB was also shown to prevent PIFs from binding to their target promoters (Park et al., 2012), hence adding an additional level of regulation to these factors.

Most PIF mutants display photomorphogenic phenotypes in the dark (short hypocotyls, open cotyledons and the accumulation of chlorophyll precursors), whereas over-expression of PIFs leads to an exaggerated skotomorphogenic phenotype (long hypocotyls, negative hypocotyl gravitropic growth, unopened cotyledons, sustained apical hook and inhibition of chlorophyll biosynthesis), indicative of a negative role of these factors in phytochrome signaling. Even so, members of this gene family, such as PIF6 (Penfield et al., 2010), and some of the non-phytochrome-interacting members, have been reported to act as positive regulators of phytochrome signaling. An example is HFR1, which prevents excessive response to shade, by forming non-DNA-binding heterodimers with PIF4 and PIF5 (Hornitschek et al., 2009).

IV. Developmental processes regulated by PIFs

Arabidopsis imbibed seeds require light to germinate and this R/FR reversible response is mostly controlled by phyB (Shinomura et al., 1994). PIF1/PIL5 plays a pivotal role in mediating this response by regulating the expression of ABA- and GA-related genes (Oh et al., 2007, 2009). PIL5 represses GA signaling by binding a G-box motif in the promoters of the GAI (GA-INSENSITIVE 1) and RGA (REPRESSOR OF ga1-3) genes (Oh et al., 2006, 2007), and also indirectly promotes ABA biosynthesis and GA catabolic gene expression though activation of the SOMNUS (SOM), ABI3 and ABI5 (ABA-INSENSITIVE) gene targets (Kim et al., 2008; Oh et al., 2009). In the light, phyB destabilizes the PIL5 protein and reverses PIL5 action, the reduced ABA levels and increased GA synthesis leading to DELLA destabilization and the triggering of seed germination (Oh et al., 2007). Cold temperatures act synergistically with light in promoting seed germination, this response being controlled by the light-stable SPATULA (SPT) factor (Penfield et al., 2005). PIF6 is also expressed to high levels during seed development, with two transcript splice variants detected in seeds (Penfield et al., 2010). Although the loss of PIF6 function increases primary seed dormancy, over-expression of the shorter transcript, lacking the DNA-binding domain, results in reduced dormancy, pointing to a relevant role of this splice variant in dormancy release (Penfield et al., 2010).

In darkness, PIF1 and PIF3 inhibit photomorphogenesis by negatively regulating chloroplast development and chlorophyll synthesis. pif1 and pif3 mutants accumulate protochlorophyllide, a phototoxic intermediate in the chlorophyll biosynthetic pathway which causes photo-oxidative damage on illumination. PIF1 directly activates the expression of the protochlorophyllide oxidoreductase PORC gene, whereas PIF1 and PIF3 repress the tetrapyrrole synthesis HEMA1 and GUN4 genes (Moon et al., 2008; Stephenson et al., 2009), and play a specific role in the diurnal regulation of chloroplast development genes. Excess light generates ROS, leading to photodamage and photoinhibition, and eventually cell death (Li et al., 2009). During de-etiolation, PIF1 and PIF3 prevent ROS production by forming heterodimers with the bZIP transcription factors HY5 and HYH, hence avoiding the activation of ROS-responsive genes by these factors (Chen et al., 2013).

The etiolated development of dark-grown seedlings is principally regulated by PIF1, PIF3, PIF4 and PIF5. Although this PIF quartet acts redundantly in driving etiolated development, the expression of their gene targets is not equally altered in all the gene mutant combinations (Zhang et al., 2013). PIF1 appears to have a more prominent contribution in dark-grown seedlings (Leivar et al., 2008; Stephenson et al., 2009), whereas, in green seedlings, elongation is mainly mediated by PIF4 and PIF5 (Lorrain et al., 2008; Hornitschek et al., 2012). In short-day-grown plants, the PIF4 and PIF5 proteins accumulate at the end of the night, coinciding with the highest hypocotyl elongation window (Nozue et al., 2007), and PIF3 was also shown to contribute to this rhythmic growth (Soy et al., 2012). PIF5 controls apical hook formation by regulating ethylene biosynthesis (Khanna et al., 2007), whereas, in the light, ethylene induces hypocotyl elongation in a process regulated by PIF3 (Zhong et al., 2012). PIFs have also been implicated in sucrose-induced hypocotyl elongation in the dark, with PIF1, PIF3, PIF4 and PIF5 transcript levels shown to be up-regulated by sucrose, in a process that depends on the presence of GAs (Liu et al., 2011). PIF7, in turn, has been reported to play a prominent role in shade avoidance (Li et al., 2012a).

PIF3, PIF4 and PIF6 are also involved in light-mediated control of stomatal development. The stomatal index is reduced in pif4 mutants, whereas a significant increase in stomatal density is observed in pif6 plants (Casson et al., 2009). pif3 and pif4 mutants, in addition, exhibit wider open stomata, implying a function of these factors downstream of phyB in the inhibition of stomatal opening (Wang et al., 2010). PIF4 and PIF5 also play a role in blue light-induced phototropism. PIF4 over-expressors display a severely reduced response to unilateral blue light illumination, whereas this response is exacerbated in the double pif4pif5 mutant (Sun et al., 2013). These factors were shown to act downstream of the blue light sensors PHOT1 and PHOT2 and negatively modulate phototropism by transcriptionally activating the IAA19 and IAA29 auxin repressors (Sun et al., 2013).

Transcriptome analysis of PIF single, double and pifq (pif1, pif3, pif4, pif5) mutants led to the identification of specific and overlapping downstream genes (Leivar et al., 2009; Shin et al., 2009). Of the PIF3-regulated genes, 74% are redundantly regulated by one or more PIFs (Zhang et al., 2013) and most of these genes are also light regulated (Leivar et al., 2009). More recently, DNA-binding assays for different PIFs have demonstrated that a G-Box (CACGTG) and the PBE variant (CACATG) are the preferred binding motifs for the PIF proteins (Hornitschek et al., 2012; Zhang et al., 2013). PIFs regulate the expression of different classes of transcription factors (bHLH, homeobox, bZIP, ARF (AUXIN RESPONSE FACTORS), AUX/IAA, AP2-EREBP (APETALA2/ethylene-responsive element binding protein), BBX (B-BOX PROTEIN) and TCP), highlighting their function as integrators of multiple signaling pathways (Fig. 1). Moreover, several of their gene targets are implicated in cell elongation and the control of photosynthetic capacity, demonstrating a direct function of PIFs in the regulation of these processes. Shared and distinct functions of these factors have been comprehensively discussed in a recent review by Jeong & Choi (2013).

Figure 1.

PHYTOCHROME INTERACTING FACTORs (PIFs) regulate multiple aspects of plant development. PIFs function as negative regulators of light signaling by sustaining seedling etiolation. PHYA and PHYB trigger PIF degradation on red light perception, and blue light represses PIF4 and PIF5 transcription in a PHOT1- and PHOT2-dependent manner. These factors facilitate nuclear translocation of PHYB and, under prolonged light, promote COP1-mediated destabilization of this photoreceptor. Members of this gene family redundantly control skotomorphogenic development, but also play specific roles in the control of seed germination, stomatal development, flowering promotion, chlorophyll biosynthesis and phototropic responses. PIF6 constitutes an exception in this family of regulators as it positively regulates photomorphogenesis, inhibiting hypocotyl growth and promoting seed germination. COP1, RING E3 ligase CONSTITUTIVE PHOTOMORPHOGENIC 1; PHOT1/PHOT2, blue light PHOTOTROPINS; PHYA/PHYB, red light photoreceptor PHYTOCHROMES. BES1/BZR1, BR-response factors; BIN2, BR-signaling GSK3 kinase; GID1, GA receptor; SLY1, SCFSLY1 E3 ligase F-box subunit; DELLA, GA-signaling DELLA repressors; IAA, auxin induced AUX/IAA proteins; SAUR, SMALL AUXIN UP RNAs; ATHB, homeodomain leucine-zipper; FT, FLOWERING LOCUS T.

V. PIFs and hormonal control of light signaling

GA, auxin and BRs play major roles in the light-mediated regulation of hypocotyl growth (Jaillais & Chory, 2010; Lau & Deng, 2010; Depuydt & Hardtke, 2011). Mutants impaired in the synthesis or response to these hormones suppress the constitutive elongation phenotype of phyB mutants and cause de-etiolated growth in the dark. Notably, the integration of these hormone pathways has been shown to occur at the transcriptional level, PIFs acting as main hubs in this growth hormonal network. The finding that PIFs play a direct role in the activation of auxin biosynthesis (Franklin et al., 2011; Hornitschek et al., 2012; Li et al., 2012a,b; Sun et al., 2012), in addition to the more recent discovery that they act in a concerted manner with the BR-related BES1 (BRI1-EMS-SUPPRESSOR1) and BZR1 (BRASSINAZOLE-RESISTANT1) factors to promote cell expansion (Oh et al., 2012), has provided a functional link between these different hormone pathways and has enabled us to start to uncover how signaling by these hormones is integrated to coordinate plant growth.

1. GA signaling

GAs promote hypocotyl elongation predominantly by signaling the degradation of the growth inhibitory DELLA proteins, a family of repressors which, in Arabidopsis, is encoded by five genes (RGA, GAI, RGL1–3 (RGA-LIKE)) of partially redundant functions (Schwechheimer, 2008; Hauvermale et al., 2012). These repressors bind the PIF DNA recognition domain and sequester these factors into an inactive complex unable to bind to DNA (Feng et al., 2008; de Lucas et al., 2008; Alabadí & Blázquez, 2009). GA perception by the soluble AtGID1 (GIBBERELLIN INSENSITIVE DWARF 1) receptor promotes GA–AtGID1 interaction with the DELLAs, favoring the recognition of these repressors by the F-box SLEEPY1 (SLY1) subunit of the SCFSLY1 E3 ligase complex (Willige et al., 2007; Gao et al., 2011; Sun, 2011). Ubiquitination by SCFSLY1 triggers proteasomal degradation of these repressors and releases PIF inhibition, allowing the activation of cell elongation genes.

2. GA and BR cross-talk

BRs act synergistically with GA in the promotion of hypocotyl and petiole growth. BR mutants display a stronger dwarf phenotype and de-etiolated growth in the dark than do GA mutants (Tanaka et al., 2003) and actually, GAs were shown to be unable to restore growth of the BR-deficient det2-1 or BR-insensitive bri1 mutants, suggesting a role of BRs downstream of GAs (Bai et al., 2012; Gallego-Bartolomé et al., 2012). Remarkably, GA-induced destabilization of the RGA repressor is not impaired in these mutants, indicating that BR signaling is not required for the degradation of DELLAs. However, destabilization of these repressors does not promote growth in the absence of BRs (Bai et al., 2012; Gallego-Bartolomé et al., 2012). BRs are perceived by the membrane receptor kinase BRI1 (BRASSINOSTEROID INSENSITIVE 1), which initiates a signaling cascade leading to the activation of the BES1 and BZR1 HLH factors, with a main role in BR-regulated gene expression (Kim & Wang, 2010; Clouse, 2011). Recently, work by independent groups has established that BES1 and BZR1 interact at the protein level with the DELLAs (Bai et al., 2012; Gallego-Bartolomé et al., 2012; Li et al., 2012b). These repressors bind the non-phosphorylated nuclear BES1 and BZR1 proteins and block their DNA-binding activity, thus suppressing the activity of these factors via a similar mechanism to that reported for PIFs. Indeed, the genes differentially expressed in bzr1-1D seedlings, with a constitutive BZR1 activation, significantly overlap with GA-responsive genes, indicating that GA up-regulates them via a BR/BZR1-dependent mechanism (Bai et al., 2012; Gallego-Bartolomé et al., 2012). Notably, BZR1 and PIF4 have been reported to heterodimerize via their DNA-binding domains, PIFs being required for BZR1 activity and, conversely, BZR1 being required for PIF-mediated cell elongation (Oh et al., 2012). PIF4 and BZR1 bind as a complex to the promoters of their co-regulated targets, which are highly enriched in auxin-responsive and cell wall-related genes, shown to be repressed by light. Hence, the DELLA–BZR1–PIF4 module antagonizes light signaling by activating a growth regulatory program that leads to enhanced auxin signaling and up-regulated expression of genes with a role in longitudinal expansion (Fig. 2). BES1 and BZR1 also act independently of PIFs, homodimeric complexes of these factors repressing BR levels, by binding a conserved BRRE (CGTGT/CG) element in the BR biosynthetic gene promoters (Oh et al., 2012). A reduced BR level, as a result of this repressive function, probably contributes to inactivate these factors and repress PIF4–BZR1 complex formation, and thus negatively feeds back on cell elongation (Fig. 2).

Figure 2.

PHYTOCHROME INTERACTING FACTORs (PIFs) act as central hubs of a growth regulatory program integrating the light and hormonal pathways. Light and the hormones gibberellin (GA), brassinosteroids (BRs) and auxin play a prevalent role in the control of cell expansion. Signaling by these hormones is integrated at the transcriptional level. PIF expression is regulated by the circadian clock and the stability of these factors is modulated by light. GAs promote growth by inducing the degradation of the DELLAs, allowing the accumulation of the PIFs and BES1/BZR1 factors in its free active form. BRs are required for BIN2 inactivation, allowing PP2A-mediated dephosphorylation of BES1 and BZR1, and nuclear translocation of these factors. BZR1 co-activates PIF4, these factors binding as a complex the PBE or HUD element conserved in the promoters of cell wall-related, TAA1 and YUC biosynthetic and auxin-signaling genes. Indole-3-acetic acid (IAA) biosynthesis and transport are required for cell elongation, auxin signaling feeding back on this pathway by the activation of GA biosynthesis and BES1/BZR1 gene expression. BES1/BZR1, BR-signaling HLH factors; BIN2, GSK3 kinase BR-INSENSITIVE 2; DELLAs, GRAS DELLA repressors; PBE and HUD, promoter E-box and ‘Hormone Up at Dawn’ elements; PP2A, protein phosphatase 2A; TAA1 and YUC, Tryptophan aminotransferase 1 and YUCCA flavin monooxygenase IAA biosynthetic enzymes.

3. Auxin synthesis and transport

In addition to the GA and BR pathways, auxin signaling is also required for growth promotion (Tao et al., 2008; Pierik et al., 2009). Auxins activate a transcriptional cascade mediated by the TIR1/AFB (TRANSPORT INHIBITOR RESPONSE 1/AUXIN-RELATED F-BOX) family of F-box proteins, which functions as the auxin receptor, and the ARF and AUX/IAA families of transcriptional regulators (Chapman & Estelle, 2009). AUX/IAA ubiquitination by the auxin–SCFTIR1/AFB complex targets these negative regulators for proteasomal degradation and relieves their inhibition on ARFs (Leyser, 2010; Tromas & Perrot-Rechenmann, 2010; Calderón-Villalobos et al., 2012). Destabilization of these repressors depends on a conserved degron in domain II, mutations within this region leading to AUX/IAA stabilization and an insensitive response to auxin (Brunoud et al., 2012). IAA is synthesized from l-tryptophan, via a main indole-3-pyruvic acid (IPA) pathway, involving the TRYPTOPHAN AMINOTRANSFERASE OF ARABIDOPSIS (TAA) and YUCCA (YUC) flavin monooxygenase families of enzymes (Stepanova et al., 2008; Tao et al., 2008; Yamada et al., 2009; Mashiguchi et al., 2011; Won et al., 2011). PIF4, PIF5 and PIF7 have been shown to bind the TAA1 and YUC8 promoters and activate the expression of these genes (Franklin et al., 2011; Sun, 2011; Hornitschek et al., 2012; Li et al., 2012a,b). However, differential activation of these genes in response to shade or elevated temperatures suggests that PIFs regulate their expression via independent pathways. TAA1 and YUC expression is primarily induced in the cotyledons (Tao et al., 2008; Stavang et al., 2009), transport of IAA to the petioles and hypocotyl being required for the elongation of these organs (Stavang et al., 2009; Keuskamp et al., 2010). Auxin transport is coordinated by the PIN (PIN-FORMED) efflux proteins and the ABCB family of transporters, in addition to the AUX/LAX influx carriers (Zažímalová et al., 2010). PIN expression and polar membrane localization are regulated by light and GA signaling, enhanced auxin transport by these efflux carriers being reported to mediate the differential elongation of cells in the apical hook, or during shade and phototropic responses (Keuskamp et al., 2010; Zádníková et al., 2010; Ding et al., 2011; Willige et al., 2011). Blue light also promotes phot1-mediated phosphorylation of ABCB19 (Christie et al., 2011) and the D6 PROTEIN KINASE (D6PK)-dependent phosphorylation of PIN3, PIN4 and PIN7 (Willige et al., 2013), leading to a redistribution of auxin flow and the asymmetric growth response that precedes hypocotyl bending towards unidirectional light (reviewed by Haga & Sakai, 2012; Christie & Murphy, 2013).

4. PIF4–BZR1 interaction and hormonal control of plant growth

Auxin and BRs exert synergistic effects on hypocotyl elongation, with at least 40% of the genes up-regulated in response to BRs also found to be induced in response to auxin, or in the IAA over-producer yucca mutants (Nemhauser et al., 2004). BRs have been reported to activate the DR5 promoter (Nakamura et al., 2003), hence implying a role of these hormones in auxin-regulated gene expression. As observed for GAs, a functional BR response pathway is required for auxin-induced growth, with the BR-insensitive bri1 mutation suppressing the elongated phenotype of the constitutive yucca mutants. Conversely, auxin signaling is required for BR-induced growth, the auxin-insensitive axr1 and tir1 mutants displaying a reduced response to BRs, whereas response to these hormones is saturated in yucca plants. Strikingly, promoter analyses of the BR- and auxin-regulated genes did not identify a conserved ARF-binding AuxRE (TGTCTC) element, but found that these promoters were enriched in a CACATG motif (Nemhauser et al., 2004), recently shown to be recognized by the PIF4–BZR1 complex (Oh et al., 2012). Several copies of the core ARF-binding TGTC element were also observed in half of these promoters, indicating a role of ARFs in the direct activation of this subclass of genes (Nemhauser et al., 2004). Indeed, recent transcriptome studies of auxin-induced elongating hypocotyls have revealed that picloram activates the expression of cell expansion-related genes via PIF- and GA-dependent, but also PIF-independent, pathways (Chapman et al., 2012). In addition, although auxin- and PIF-regulated genes overlap significantly, there is an additional group of genes that still responds to IAA treatment in pif4pif5 seedlings and thus correspond to direct ARF gene targets (Nozue et al., 2011; Chapman et al., 2012). This suggests a regulatory model, whereby GA and BRs promote hypocotyl growth via direct BZR1–PIF4-mediated activation of cell wall metabolism and auxin-responsive gene expression, increased auxin levels, as a result of activated TAA1 and YUC gene expression, further promoting cell elongation via an auxin-response pathway independent of PIFs (Fig. 2). Remarkably, this additional pathway regulates GA biosynthesis and BES1/BZR1 gene expression (Frigerio et al., 2006; Chapman et al., 2012), and thus contributes to release DELLA repression, feeding back on BZR1–PIF4 complex formation by increasing BES1/BZR1 levels (Fig. 2).

In Arabidopsis, ARFs and Aux/IAA are respectively encoded by 23 and 29 genes of partially redundant functions (Tiwari, 2004; Guilfoyle & Hagen, 2007). ARF5 and ARF7 are required for auxin-controlled cell expansion (Hardtke et al., 2004), an impaired response to auxin also being observed in arf7arf19 double mutants or the gain-of-function solitary root (slr)/iaa14 (Okushima et al., 2005), bdl/iaa12 (Hardtke et al., 2004) and msg2/iaa19 (Tatematsu et al., 2004) mutants (Table 1). Notably, the arf7 and msg2/iaa19 mutants exhibit an insensitive response to BRs, these mutations suppressing the constitutive phenotype of bzr1-1D seedlings (Gallego-Bartolomé et al., 2011). Likewise, the auxin signaling axr1-12 mutant and the shy2-2/iaa3 gain-of-function mutation have been shown to suppress the elongated phenotype of PIF4ox seedlings (Sun et al., 2012). BZR1 binds the IAA19 and ARF7 promoters, IAA19 and IAA29 also being reported to be direct targets of PIF4 (Sun et al., 2013), which suggests that BR-responsive expression of these genes is regulated via PIF4–BZR1 complex formation. Although the function of the AUX/IAAs repressors downstream of this growth-promoting complex is not well understood, in PIF4ox seedlings, DR5 expression was found to be increased at the base of the hypocotyl, but strongly repressed in the cotyledons (Sun et al., 2012, 2013). Hence, it is possible that a dynamic interplay between enhanced auxin synthesis and transport, and AUX/IAA repressor activation, controls differential hypocotyl and petiole elongation during responses to shade and elevated temperatures, and phototropic bending of the hypocotyl. An important challenge in the future will be to decipher how this growth regulatory network is integrated in space and time, to shape differential growth of the various plant organs.

Table 1. Hormone signaling and biosynthetic mutants with an affected response to light, vegetational shade or elevated temperatures
MutantAGI codeFull nameFunctionPhenotypeReferences
  1. Gain-of-function mutants are indicated with a superscript letter: aDELLA domain deletion; bmutation in the PEST domain; cactivation tagging; dmutation in the domain II degron.

GA pathway
 gai1.1 a AT1G14920GA INSENSITIVE 1DELLA repressor

GA insensitive


Peng et al. (1997)
 gai-t6    ElongatedPeng et al. (2002)
 rga-24 AT2G01570REPRESSOR OF ga1-3DELLA repressorElongatedSilverstone et al. (1998)

GA insensitive


Dill et al. (2001)


(gai-t6, rga-t2, rgl1-1, rgl2-1, rgl3-4)











DELLA repressor

GA constitutive

Very elongated

Feng et al. (2008)
 gid1 a, b, c






DELLA receptor

GA insensitive


Willige et al. (2007)

GA deficient


Sun et al. (1992)
BR pathway






BR receptor

BR insensitive


Friedrichsen et al. (2000)
bes1-D b AT1G19350bri1-EMS-SUPPRESSOR 1

HLH transcription


BR constitutive


Yin et al. (2002)
 bzr1-1D b AT1G75080



HLH transcription


BR constitutive


Wang et al. (2002)
 det2-1 AT2G38050DE-ETIOLATED 2





Li et al. (1996)
AUX pathway
 tir1-1 AT3G62980



AUX receptor

AUX insensitive


Ruegger et al. (1998)






IAA biosynthesis


IAA deficient

Short hypocotyls

in shade

Tao et al. (2008)

Stepanova et al. (2008)

 yucca c AT4G32540YUCCA 1





Zhao et al. (2001)







IAA biosynthesis

CYT P450

IAA-deficient dwarfZhao et al. (2001)
axr1-12 AT1G05180AUXIN RESISTANCE 1RUB-ACTIVATINGAUX insensitive dwarfLincoln et al. (1990)








B3-like transcription factorsAUX insensitive dwarfHardtke et al. (2004)








B3-like transcription factorsAUX insensitive dwarf, impaired lateral root formationOkushima et al. (2005)
shy2-2/iaa3 d AT1G04240



AUX/IAA transcriptional repressorAUX insensitive dwarf, short rootsTian et al. (2002)
bdl/1aa12 d AT1G04550



AUX/IAA transcriptional repressorAUX insensitive impaired primary root meristemHardtke et al. (2004)
slr/iaa14 d AT4G14550



AUX/IAA transcriptional repressorAUX/BR insensitive dwarf, impaired lateral root formationOkushima et al. (2005)
msg2/iaa19 d AT3G15540



AUX/IAA transcriptional repressorAUX/BR insensitive altered phototropism and gravitropism, no apical hookTatematsu et al. (2004)

VI. Role of PIFs in diurnal control of cell elongation

Hypocotyl growth relies on the rapid expansion of cells preformed in the embryo (Gendreau et al., 1997). Elongation of these cells is driven by the loosening of the cell wall and increased water turgor, in a process that is tightly regulated by environmental conditions. Under short-day photocycles (8 h : 16 h, light : dark), hypocotyl elongation is maximal at dawn, and this growth phase is shifted to dusk under continuous light (Dowson-Day & Millar, 1999; Nozue et al., 2007). This time of day-specific response is regulated by the circadian clock which, by anticipating daily fluctuations in light and temperature, ensures the correct timing of metabolic assimilation and growth, thus improving plant fitness (Stitt & Zeeman, 2012). In Arabidopsis, c. 30–50% of the genes exhibit rhythmic oscillation (Michael & McClung, 2003; Covington et al., 2008), with maximal peaks of expression during dark to light (‘morning’ genes) or light to dark (‘evening’ genes) transitions. Such phase-specific expression has been correlated with the presence in their promoters of an ‘evening’ (EE, AAtatcT) or GATA (CTtatcC) element, enriched in dusk-expressed genes, and a ‘morning’ (ME, NccacACN) or G-box (GccacGTG) element, which is over-represented in dawn-expressed genes (Covington et al., 2008; Michael et al., 2008a,b). Interestingly, BR, IAA and GA genes were found to be co-expressed at the time of day at which the hypocotyl growth rate is maximal (Michael et al., 2008a). This expression pattern correlates with the presence of a CACATG motif or Hormone Up at Dawn (HUD) element in the promoters of these genes, which, when multimerized, was shown to be sufficient to confer a cyclic pattern of expression to the LUC reporter gene. It is noteworthy that this motif is identical to that recently shown to be recognized by the PIF4–BZR1 complex (Oh et al., 2012), pointing to a role of this complex in the diurnal activation of these hormone pathways. The expression of these genes, in addition, was found to be up-regulated in clock mutants during the dark phase, and to be constantly elevated in phyB plants (Michael et al., 2008a), further supporting a role of PIF4 in their regulated expression. Hence, the concerted action of the PIF4 and BES1/BZR1 factors appears to play a pivotal role in linking light signaling and the circadian clock with this growth-promoting hormonal program, and thus in orchestrating rhythmic hypocotyl growth.

1. Clock regulation of PIF transcription

The Arabidopsis clock consists of three main negative feedback loops, interlocked in a ring structure, such that repression of a loop component by its immediate negative regulator leads to indirect activation of the third loop components (Pokhilko et al., 2012; Yamashino, 2013). The morning-expressed LHY (LATE ELONGATED HYPOCOTYL) and CCA1 (CIRCADIAN CLOCK ASSOCIATED 1) MYB factors repress the expression of the evening-phase TOC1 (TIMING OF CAB EXPRESSION 1), ELF3, ELF4 (EARLY FLOWERING) and LUX genes by binding a conserved EE in the promoters of these genes (Li et al., 2011). PRR9, PRR7 and PRR5 (PSEUDO-RESPONSE REGULATOR) play partially redundant roles in repressing CCA1 and LHY expression (Nakamichi et al., 2010) and, together with TOC1, ensure that LHY/CCA1 are expressed with sharp peaks in the morning (Gendron et al., 2012). ELF3, ELF4 and LUX, in turn, have been shown to interact at the protein level to form the so-called ‘evening complex’ (EC), which binds the PRR9 and LUX promoters (Nusinow et al., 2011; Herrero et al., 2012) and represses the day-phase PRR9 and PRR7 genes, hence completing the loop.

Mutations in any of these oscillator components result in shorter (i.e. lhy cca1 and TOC1ox plants) or longer (i.e. CCA1ox and the toc1 and prr9 prr7 prr5 mutants) hypocotyls, underscoring a role of the clock in the control of diurnal hypocotyl growth. Remarkably, these mutants show altered hypocotyl lengths under diurnal conditions, but not in continuous light, a trait that differentiates them from phyB and PIF4ox plants. An important breakthrough was the finding that abnormal hypocotyl elongation of these mutants is caused by an alteration in the PIF4 and PIF5 expression profiles (Nozue et al., 2007; Niwa et al., 2009). In wild-type plants, PIF4 and PIF5 show a robust circadian rhythm of expression, these transcripts increasing in short days during late night to peak at daytime and be repressed at dusk (Fig. 3). In CCA1ox and prr9 prr7 prr5 seedlings, PIF4 and PIF5 transcripts are also elevated during night-time and an earlier up-regulation is observed in the toc1 mutant (Niwa et al., 2009). The recent finding that the evening clock components ELF3, ELF4 and LUX repress PIF4 and PIF5 transcription at dusk (Nusinow et al., 2011) has largely improved our understanding of how the clock governs the cyclic expression of these genes, at the time that has provided a mechanistic basis for their mis-regulation in clock-defective mutants. Indeed, mutations in the ELF3, ELF4 or LUX gene lead to longer hypocotyls and up-regulated levels of expression of the PIF4 and PIF5 transcripts during early night, and this phenotype is suppressed by the pif4pif5 mutation (Nusinow et al., 2011). Thus, rhythmic hypocotyl growth can be explained by an external coincidence model, in which PIF4 and PIF5 transcription, in concert with phyB-mediated protein destabilization, restricts the accumulation of the PIF4 and PIF5 proteins to the end of the night (Nozue et al., 2007; Niwa et al., 2009; de Montaigu et al., 2010; Nomoto et al., 2012; Shin et al., 2013).

Figure 3.

PHYTOCHROME INTERACTING FACTORs (PIFs) control time of day-specific hypocotyl growth. In short days, hypocotyl elongation is maximal at dawn. The circadian clock regulates PIF4 and PIF5 transcription, such that their transcript levels increase at the end of the night and are elevated during the day, to decrease at dusk. These factors are stable in the dark and are rapidly phosphorylated by phytochromes in the light. A pool of these proteins is still observed during the day time, but inactivated by the DELLAs. Accumulation of the PIF4 and PIF5 proteins late in the night promotes PIFs–BES1/BZR1 interaction. The formation of this co-activator complex is dominant over BZR1 dimerization, and increased brassinosteroid (BR) levels lead to enhanced growth. After transition to light, PHYB represses PIFs and BES1/BZR1 can homodimerize, the repression of BR levels further contributing to arrest growth. BES1/BZR1, BR-signaling HLH factors; DELLAs, GRAS DELLA repressors; PHYB, Phytochrome B.

2. GA signaling is gated by the clock

The clock also controls the expression of the GID1a (GIBBERELLIN INSENSITIVE1a) and GID1b genes, encoding the GA receptor (Arana et al., 2011). In short-day photocycles, GID1 genes show maximal expression during the night and their transcript levels are greatly reduced early in the day. This oscillation pattern contributes to the destabilization of DELLAs during the night, such that the lowest abundance of these repressors coincides with the period of maximal hypocotyl growth (Achard et al., 2007; Arana et al., 2011). Consistent with these changes in DELLA protein levels, transient expression of the stable gai protein has been shown to have a stronger inhibitory effect at night-time, whereas application of GAs is more effective during the day. Loss of DELLA function in the global mutant, in turn, leads to continuous arrhythmic growth, highlighting a prominent role of these repressors in PIF4 and PIF5 repression during the daytime (Feng et al., 2008; de Lucas et al., 2008; Alabadí & Blázquez, 2009; Arana et al., 2011).

3. Oscillations in BR levels contribute to cyclic elongation

BR signaling mediates nuclear accumulation of the BES1 and BZR1 factors and, as such, is required for the PIF4 co-activator function of these factors (Oh et al., 2012). At the same time, BES1 and BZR1 bind a conserved BRRE (CGTG(T/C)G) motif in the CPD (CONSTITUTIVE PHOTOMORPHOGENESIS), ROT3 (ROTUNDIFOLIA3), DWF4 (DWARF4) and BR6OX2 promoters, to repress the expression of these genes (He et al., 2005; Oh et al., 2012). Hence, relative nuclear levels of the PIF4 and BES1/BZR1 factors are expected to play a relevant role in the control of diurnal oscillations in BR levels and in cyclic hypocotyl growth. Indeed, BR genes peak at the same time as maximal hypocotyl elongation, indicating that the accumulation of PIF4 at dawn favors PIF4–BZR1 interaction over BZR1 dimerization and de-represses BR synthesis, contributing to enhanced growth (Fig. 3). In this way, the dual PIF4 co-activator and BR synthesis inhibitory function of the BES1/BZR1 factors is foreseen to serve as a fine-tuning mechanism to restrict PIF function, in addition to modulating the activity of these factors in response to environmental cues.

VII. Response to vegetational shade

Sunlight filtered through a dense canopy is enriched in FR and depleted in the blue light spectrum, as a result of absorption by the photosynthetic pigments of the leaves. Plants perceive low R/FR light as indicative of the presence of other competing plants and initiate a ‘shade avoidance’ response characterized by the rapid elongation of the hypocotyl and the petioles, and a more erect position of the leaves seeking light. When vegetational shade persists, apical dominance is increased and flowering is accelerated, ensuring the production of seeds (Sessa et al., 2005; Franklin, 2008; Casal, 2013). Shade is mainly sensed by phyB, FR-induced Pfr to Pr reversion, allowing PIF4 and PIF5 protein accumulation and increased expression of genes with a role in cell expansion (Lorrain et al., 2008; Hornitschek et al., 2009; Leivar & Quail, 2011). In agreement with the external coincidence model, this response is gated by the clock, FR light-induced Pfr reversion leading to PIF4 and PIF5 stabilization only when these conditions coincide with high transcript levels (Salter et al., 2003; Lorrain et al., 2008; Hornitschek et al., 2009; Kunihiro et al., 2011; Casal, 2013). Short time-scale analyses in Arabidopsis showed that elongation of the hypocotyl follows a biphasic pattern, with growth observed as early as 45 min after shade treatment, to slow down during the following 2–3 h and re-start at a faster rate (Cole et al., 2011). It is noteworthy that pif4pif5 mutants are not affected in the first growth phase, suggesting that other PIFs are implicated in this rapid response. PIF7 was recently implicated in this response (Li et al., 2012a), in addition to playing a redundant function with PIF3 and PIF4 in modulating phyB levels under prolonged R light (Leivar et al., 2008). HFR1 is also rapidly induced on exposure to shade, this protein being part of a brake mechanism that precludes excessive PIF4 and PIF5 function (Duek & Fankhauser, 2003; Sessa et al., 2005).

1. Hormonal control of the response to vegetational shade

Shade-induced hypocotyl elongation requires new auxin biosynthesis as a severely impaired shade response is observed in sav3 (SHADE AVOIDANCE 3) mutants, lacking tryptophan aminotransferase (TAA1) activity (Tao et al., 2008). During the first hour of low R/FR exposure, auxin levels increase in wild-type plants, but not in sav3 or pif7 mutants. Indeed, these mutants are impaired in shade-induced gene expression, although they respond to IAA or picloram treatment (Li et al., 2012a,b). Shade does not induce SAV3 or PIF7 gene expression, but activates the expression of the flavin monooxygenase YUCCA 2, 4, 8 and 9 genes (Zhao, 2010). Notably, PIF7 is not light destabilized as other PIFs (Leivar et al., 2008), yet it is phosphorylated in the light. During early shade avoidance, PIF7 is rapidly dephosphorylated, the activation of this factor leading to a rapid increase in free IAA in the cotyledons (Li et al., 2012a). This new IAA is transported into the hypocotyl and other tissues for shade-induced elongation (Morelli & Ruberti, 2000; Sassi et al., 2013) and this response is suppressed by the auxin transport inhibitor N-1-naphthylphthalamic acid (NPA) (Petrášek et al., 2003; Pierik et al., 2009). Low R/FR light also induces the expression of PIN3 in the hypocotyl, with pin3 mutants displaying a much reduced response to shade (Keuskamp et al., 2010). Shade redirects PIN3 from the basal to the lateral membrane side of the endodermal hypocotyl cells, leading to increased auxin signaling and IAA19 expression in the outer cortex and epidermal tissues (Keuskamp et al., 2010). Plants over-expressing the multidrug resistance-like membrane protein ABCB19 (Wu et al., 2010), tir1 loss-of-function mutants and the iaa17/axr3 and axr1-2 mutants are all impaired in shade response (Pierik et al., 2009; Keuskamp et al., 2010; Sellaro et al., 2012), thus demonstrating that enhanced auxin transport and signaling are essential for elongation growth.

Low R/FR light has also been shown to induce GA synthesis and to promote the destabilization of DELLAs (Djakovic-Petrovic et al., 2007). Breakdown of these repressors is required for shade-induced hypocotyl and petiole elongation, this response being largely suppressed in the GA-deficient ga1-3 and gain-of-function gai mutants. The gait6 and rga24 mutations restore the shade response of ga1-3 plants, whereas the tetra (gait6 rga24 rgl1-1 rgl2-1) mutant shows a constitutive shade-elongated phenotype. However, shade still induces normal petiole elongation in this multiple della knockout (Djakovic-Petrovic et al., 2007), indicating that an additional DELLA-independent pathway is involved in this response. Indeed, NPA treatment abolishes low R/FR-induced elongation to the same extent in the tetra DELLA knockout as in wild-type plants (Pierik et al., 2009), whereas GA is unable to rescue the shade avoidance response of auxin-resistant axr2-1 mutants, supporting a role of auxin signaling downstream of DELLAs. These observations are largely explained by DELLA inhibition of PIF transcriptional activity (Feng et al., 2008; de Lucas et al., 2008), destabilization of DELLAs being required for PIF activation and enhanced auxin synthesis.

PIF4 and PIF5 play a more relevant role in the second growth phase after shade treatment and during prolonged shade (Lorrain et al., 2008; Cole et al., 2011). These factors are stabilized within 15 min on FR-induced phyB Pfr to Pr reversion, pif4pif5 loss of function being shown to lead to a partial reversion of the constitutive shade avoidance response of phyB mutants (Lorrain et al., 2008). PIF4 and PIF5 directly activate the expression of ATHB2, a positive regulator of hypocotyl growth (Kunihiro et al., 2011). ATHB2 expression is rapidly and reversibly induced by low R/FR (Roig-Villanova et al., 2006), transgenic lines expressing elevated or reduced levels of ATHB2 (HAT4) or HAT2 showing longer and shorter hypocotyls, respectively, and an altered response to auxin (Morelli & Ruberti, 2002; Sawa et al., 2002). The identification of genome-wide PIF5-binding sites in seedlings subjected to low R/FR light actually showed that PIF5 regulates SAS gene expression directly (Hornitschek et al., 2012) and that, like PIF7, binds the YUC8 promoter and the promoters of several transport and auxin signaling genes (Hornitschek et al., 2012). Global expression analyses of hypocotyl growth-associated PIF4- and PIF5-regulated genes also showed a significant overlap with shade avoidance-responsive genes (Nozue et al., 2011), but, surprisingly, this group of genes was found to be highly enriched in the auxin pathway, but not in GA biosynthesis and signaling genes (Nozue et al., 2011), suggesting that PIF4 and PIF5 modulate plant growth mainly by direct regulation of auxin-related gene expression.

PIF4 and PIF5 are also the main factors involved in the reduced lamina growth (lamina : petiole ratio) and increased leaf angle response induced by attenuated blue light (Keller et al., 2011). Notably, the function of the SAV3 and PIN3 genes, which are critical for low R/FR hypocotyl elongation, is not essential to this response. GA-induced DELLA turnover is also not involved, as both gain-of-function and knock-out DELLA mutants show a normal response to reduced blue light (Keller et al., 2011). However, this response is impaired in the BR-deficient det2-1 mutants and in bri1-301 seedlings, carrying a weak mutant allele of the BRI1 receptor, although the latter retain the hyponastic response to attenuated blue light. It is noteworthy that BRs have also been reported to be required for petiole elongation and leaf morphology responses to end-of-day FR treatments (Kozuka et al., 2010), and for hypocotyl elongation in response to attenuated blue light (Keuskamp et al., 2010). Thus, it is possible that BRs control these responses by activating BES1 and BZR1, and promoting the interaction of these factors with PIF4. Reduced HFR1 levels may also play a relevant role in this response, as expression of this gene is induced by cry1 and blue light stabilizes the protein (Duek & Fankhauser, 2003; Yang et al., 2005).

VIII. PIF control of responses to elevated temperatures

High temperatures (28°C) induce rapid hypocotyl and petiole elongation, in addition to leaf hyponasty and early flowering, in a response that mimics that of shade avoidance and also depends on increased auxin signaling (Koini et al., 2009; de Wit et al., 2013). PIF4 plays a critical role in this response, as it has been shown to be impaired in pif4 but not in pif5 mutants. The histone variant H2A.Z has been established to play a major role in warm temperature gene expression, arp6 mutants impaired in H2A.Z deposition displaying constitutive activation of the warm-induced transcriptome (Kumar et al., 2012). PIF4 mediates flowering promotion by binding the FT florigen promoter in a temperature-dependent manner (Kumar et al., 2012). H2A.Z nucleosomes have been shown to preclude PIF4 accessibility to this recognition site, these nucleosomes being evicted at elevated temperatures (Kumar et al., 2012).

1. Auxin, GA and BR signaling are required for the response to elevated temperatures

At warm temperatures, PIF4 binds the promoters of the YUC8, YUC9 (Stavang et al., 2009; Sun et al., 2012), TAA1 and CYP79B2 (Franklin et al., 2011) genes, leading to increased free IAA levels. Such an increase in auxin levels and enhanced DR5::GUS activity at 28°C is abolished in pif4 seedlings, although picloram restores the response to elevated temperatures in these mutants (Franklin et al., 2011). The IAA3-stabilizing shy2-2 mutation has also been reported to impair this response and to partially suppress the elongated hypocotyl of PIF4 over-expressors (Sun et al., 2012). Remarkably, high temperature induces the auxin-regulated SAUR 19-24 and SAUR61-68 gene families in the hypocotyl (Stavang et al., 2009 ), with SAUR19 over-expression being sufficient to overcome the pif4 elongation defects, hence unveiling a main function of these small proteins in cell expansion (Franklin et al., 2011; Chae et al., 2012).

GA signaling is also required for this response, elevated temperatures leading to increased GA biosynthetic gene expression and DELLA destabilization (Stavang et al., 2009). Paclobutrazol (PAC) application or the ga1-3 mutation impairs the response to warm temperatures, although enhanced growth is still observed in the global mutant. It is noteworthy that elongation of these seedlings is abolished by the inhibitors NPA and brassinazole (BRZ), indicating that auxin and BR signaling act downstream of the DELLA repressors (Stavang et al., 2009). In line with this observation, response to warm temperatures is impaired in det2-1 mutants (Gray et al., 1998; Koini et al., 2009) and abolished by the BR inhibitor propiconazole (PPZ) in wild-type plants, but not in the constitutive bzr1-1D mutant. Interestingly, PIF loss of function in the pifq bzr1-1D mutant impairs the constitutive response to elevated temperatures of bzr1-1D seedlings, in agreement with an essential role of PIF4 in this response (Oh et al., 2012). PIF4 is actually stabilized at high temperatures (Oh et al., 2012; Yamashino et al., 2013), increased PIF4 protein levels promoting hypocotyl growth in a BZR1-dependent manner.

IX. Concluding remarks and future perspectives

Control by the circadian clock, light, warm temperatures and vegetational shade converge into the PIF factors, which play a central role in growth elongation by triggering auxin biosynthesis and the modulation of auxin signaling and transport (Franklin et al., 2011; Hornitschek et al., 2012; Sun et al., 2012). Binding of PIFs to the promoters of auxin signaling and cell wall remodeling enzymes involves the formation of a co-activator complex with the BES1 and BZR1 factors (Oh et al., 2012), and, as such, the activation of BR signaling is also required for cell expansion. PIF expression is regulated by the endogenous clock (Nozue et al., 2007; Nusinow et al., 2011) and these factors are destabilized by phytochromes in the light (Khanna, 2004; Lorrain et al., 2008; Shin et al., 2009). This restricts growth to the end of the night and provides a sensitive response to changes in light quality (Casal, 2013). Moreover, PIF and BES1/BZR1 activity is regulated by the DELLAs (Feng et al., 2008; de Lucas et al., 2008; Bai et al., 2012; Gallego-Bartolomé et al., 2012; Li et al., 2012a,b), growth promotion requiring enhanced GA synthesis to destabilize these repressors. Hence, although initially identified as components of phytochrome signaling, PIFs are now accepted to function as central hubs for circadian, light, BR and GA signal integration, to modulate auxin levels and signaling, hence linking plant growth to multiple environmental stimuli (Leivar & Quail, 2011; Zhang et al., 2013).

Under free-running light conditions, the maximum rate of hypocotyl growth occurs around subjective dusk (Dowson-Day & Millar, 1999; Nozue et al., 2007) and a similar shift from dawn to dusk is observed for auxin responsiveness (Covington & Harmer, 2007; Michael et al., 2008a,b). Such a shift is not observed in the pifq mutant, highlighting a role of PIFs in gating the auxin response (Nozue et al., 2011). Although it remains unclear how PIFs regulate growth oscillations under continuous light, recent evidence indicates that these factors are not completely destabilized in the light (Yamashino et al., 2013; our results). This remaining pool of protein is likely to play a role in the response to elevated temperatures and to be activated by auxin application. Picloram has been shown to induce GA synthesis and BES1 and BZR1 gene expression (Chapman et al., 2012), the destabilization of DELLAs, together with BES1/BZR1 accumulation, probably playing a role in PIF activation during the daytime. PIFs have also been reported to mediate the shift in the daily growth rhythm induced by sucrose, with sucrose shown to increase PIF5 protein levels in the light and PIF5 over-expression partially mimicking sucrose effects (Stewart et al., 2011; Lilley et al., 2012). These observations point to a relevant function of PIFs in the light, a feature that certainly deserves further investigation.

Another key aspect still poorly understood is how the growth of the different plant organs is coordinated. Changes in environmental conditions do not affect the growth of all organs in the same manner, as the hypocotyl and the petioles elongate rapidly during vegetational shade, whereas leaf and root growth is inhibited. Thus, interaction of the individual pathways implicated in the regulation of PIF activity and cell expansion may differ between organs or even cell types. In studies in which GA and BR signaling were specifically altered in different tissues, it was observed that the growth-promoting effects of these hormones are not equivalent in all cells. Although the expression of a stable form of the DELLAs in the endodermal root cells blocks root growth and reduces the meristem size (Ubeda-Tomás et al., 2008, 2009), BR signaling appears to be more relevant in epidermal cells, as epidermis-specific expression of the BRI1 receptor in the bri1 mutants has been shown to be sufficient to restore the growth of both roots and aerial tissues of these plants (Savaldi-Goldstein et al., 2007; Hacham et al., 2011). This would indicate that signaling by these hormones activates the expression of non-cell autonomous signals which coordinate the growth of the entire organ. Auxins are the most likely candidates to play this role, but certainly a major challenge in the future will be to understand how the PIF growth regulatory network is modulated in different cell types and to elucidate how these different signaling pathways are integrated in time and space to drive the differential growth rates of the various plant organs during development.


We apologize to the colleagues whose work or original publications could not be cited because of space limitations. Research by the authors was funded by grants BIO2011-30546 and the CONSOLIDER TRANSPLANTA project from the Spanish Ministry of Science and Innovation (MICIIN).