Hormone-controlled UV-B responses in plants

Abstract

Ultraviolet B (UV-B) light is a portion of solar radiation that has significant effects on the development and metabolism of plants. Effects of UV-B on plants can be classified into photomorphogenic effects and stress effects. These effects largely rely on the control of, and interactions with, hormonal pathways. The fairly recent discovery of the UV-B-specific photoreceptor UV RESISTANCE LOCUS 8 (UVR8) allowed evaluation of the role of downstream hormones, leading to the identification of connections with auxin and gibberellin. Moreover, a substantial overlap between UVR8 and phytochrome responses has been shown, suggesting that part of the responses caused by UVR8 are under PHYTOCHROME INTERACTING FACTOR control. UV-B effects can also be independent of UVR8, and affect different hormonal pathways. UV-B affects hormonal pathways in various ways: photochemically, affecting biosynthesis, transport, and/or signaling. This review concludes that the effects of UV-B on hormonal regulation can be roughly divided in two: inhibition of growth-promoting hormones; and the enhancement of environmental stress-induced defense hormones.

Introduction

Plants are sessile organisms which need to modulate their growth to the constantly changing environment. One variable element is light, an imperative cue necessary for photosynthesis which also functions as a source of information about the environment, seasons, or time of the day, enabling plants to adjust their morphology, development, or metabolism. The processes involved rely on a network of interactions of signaling pathways which are elicited by endogenous and environmental signals. The light is absorbed by photoreceptors within specific spectral bands and, as a consequence, plants are able to perceive qualitative and quantitative alterations in the light that they perceive ( Jiao et al. , 2007 ; Kami et al. , 2010 ). UV-B (280–315nm) is situated at the far end of the shortwave irradiation in the solar spectrum, with wavelengths of 290–315nm reaching the surface of the earth. This spectral range received a lot of scientific attention at the end of the 20th century, when a stratospheric ozone depletion and concomitant increase in UV-B radiation were detected. Ever since, many studies on the effects of UV-B on living organisms have been published, and it has become apparent that UV-B, although initially mainly regarded as a stressor, also has regulatory effects in plants ( Jordan, 1996 ; Jansen, 2002 ; Frohnmeyer and Staiger, 2003 ; Ulm and Nagy, 2005 ; Jenkins, 2009 ). Stress effects occur mainly at high doses of UV-B in non-acclimated plants, while low levels of UV-B can induce true photomorphogenic effects acclimating the plant without any signs of stress ( Jenkins, 2009 ; Jansen and Bornman, 2012 ). As hormones are ubiquitous in the regulation of plant architecture and metabolism, it comes as no surprise that downstream of the UV-B signaling pathway, hormonal pathways are affected. Here, we provide an overview of how UV-B steers plant hormones and consequently plant morphology and defense.

UV-B photomorphogenic signaling by UVR8

UV-B signaling can be roughly divided into (i) non-specific signaling, mostly stress related, and (ii) UV-B-specific, photomorphogenesis-related signaling ( Jenkins, 2009 ; Jansen and Bornman, 2012 ). However, photomorphogenic and stress responses are not mutually exclusive, and an overlap in UV-B doses exists in which both can occur.

UV-B-specific signaling mediated by the UV RESISTANCE LOCUS 8 (UVR8) photoreceptor ( Kliebenstein et al. , 2002 ; Rizzini et al. , 2011 ) involves regulation of the function of the broad photomorphogenesis response regulators CONSTITUTIVELY PHOTOMORPHOGENIC 1 (COP1) ( Oravecz et al. , 2006 ; Rizzini et al. , 2011 ) and ELONGATED HYPOCOTYL 5 (HY5) ( Kliebenstein et al. , 2002 ; Brown et al. , 2005 ). In addition, a crosstalk with the photomorphogenesis-inhibiting phytochrome-interacting factors (PIFs) has been described ( Hayes et al. , 2014 ).

The photomorphogenic effects of UV-B rely to a large extent on UVR8 ( Ulm and Jenkins, 2015 ). Plants can perceive and distinguish UV-B in a direct way through the UVR8 photoreceptor which is expressed throughout all the plant tissues ( Brown et al. , 2005 ; Rizzini et al. , 2011 ). When UV-B is absent, the main fraction of UVR8 is located in the cytoplasm in a dimeric state. Upon exposure to UV-B, a nuclear accumulation of UVR8 monomers occurs within 5min, causing downstream UV-B-induced signaling ( Fig. 1 ). The intrinsic tryptophans Trp285 and Trp233 serve as UV-B chromophores ( Rizzini et al. , 2011 ; Christie et al. , 2012 ; Wu et al. , 2012 ). These tryptophan residues at the UVR8 dimer interface take part in photoinduced proton-coupled electron transfer reactions which cause the monomerization of UVR8 ( Mathes et al. , 2015 ). UVR8 has homology with chromatin-binding proteins, yet its capacity to bind chromatin is contested ( Binkert et al. , 2016 ). It is likely that the main mechanism of UVR8 signaling involves interaction with other proteins, such as COP1.

The activated UVR8 monomer physically interacts with COP1, a key component in light signaling, through the C-terminal 27 amino acid region of UVR8 ( Oravecz et al. , 2006 ; Rizzini et al. , 2011 ; Cloix et al. , 2012 ). This interaction with COP1 does not lead to UVR8 degradation but abolishes the E3 ligase function of the COP1/SPA (SUPPRESSOR OF PHYTOCHROME A) complex, causing accumulation of the E3 targets. An example of such a target is the transcription factor HY5 ( Ulm et al. , 2004 ; Brown et al. , 2005 ; Huang et al. , 2013 ). HY5 is a basic leucine zipper transcription factor (bZIP) and is a well-characterized and a crucial component in light signaling of various wavelengths ( Jiao et al. , 2007 ). The role of HY5 in UV-B-induced signaling was confirmed in studies performed on hy5 mutant seedlings which are UV hypersensitive ( Kliebenstein et al. , 2002 ; Brown et al. , 2005 ). The HY5 protein drives its own transcription and that of its homolog HY5 HOMOLOG ( HYH ) ( Binkert et al. , 2014 ). Furthermore, the relationship between HY5 and COP1 proteins is remarkable. In darkness, COP1 ubiquitylates HY5, leading to proteasomal degradation ( Osterlund et al. , 2000 ; Lau and Deng, 2012 ). When exposed to UV-B, the interaction of UVR8 monomers and COP1 inhibits HY5 degradation and, as a consequence, HY5 accumulates ( Oravecz et al. , 2006 ). These findings make HY5 an excellent UV-B-responsive marker gene when also taking into account its central function in other light signaling pathways. Based on a meta-analysis of transcriptome data, Fierro et al. (2015) showed an overlap in gene regulation of >70% between the UVR8 and phytochrome B signaling pathways. Indeed, UVR8 and phyB pathways co-operate to fine-tune plant growth and defense ( Mazza and Ballare, 2015 ). Although not described as a central part of UVR8 signaling, the basic helix–loop–helix (bHLH) PIFs connect with the UVR8 pathway. PIFs are repressor proteins involved in the shade avoidance syndrome (SAS) under low re:far-red light conditions and controlled by PHYB (reviewed by de Lucas and Prat, 2014 ). There is increasing evidence that UV-B-induced response pathways also act via an interaction between the UVR8 photoreceptor and the SAS pathway which involves PIF4, PIF5, and PIF7 ( Lorrain et al. , 2008 ; Hornitschek et al. , 2012 ; Hayes et al. , 2014 ). Indeed, the PIF4 and PIF5 transcription factors are degraded upon UV-B exposure ( Hayes et al. , 2014 ), whereas PIF7 is activated via dephosphorylation upon shading, suggested to be controlled by UV-B ( Leivar et al. , 2008 ; Li et al. , 2012 ). Based on the response of hy5hyh mutants compared with wild-type plants, this signaling pathway is UVR8 dependent but is believed to be HY5/HYH independent ( Hayes et al. , 2014 ). Furthermore, UV-B signaling and PIFs have contrasting effects with respect to leaf curling ( Fierro et al. , 2015 ). In addition, PIF transcription factors and HY5 bind the same G-box promoter elements, competing for regulation of gene expression ( Toledo-Ortiz et al. , 2014 ). In most cases, this regulation is considered antagonistic, yet the exact activation and repression mechanisms remain elusive.

UV-B-related UVR8-independent signaling

Parallel UV-B-regulated mechanisms which are clearly distinct from the UVR8 photomorphogenic pathway exist ( Jenkins, 2009 ). One of the pathways co-ordinating the response to acute stress has been characterized in more detail. The involved perception mechanism remains unknown, but the pathway depends on the activity of mitogen-activated protein kinases (MAPKs; MPKs) ( Besteiro et al. , 2011 ). In Arabidopsis, MPK3 and MPK6 are activated upon UV-B radiation and are negative regulators of acute stress resistance, as mpk3 and mpk6 mutants are more resistant to UV-B stress. The MPK3- and MPK6-interacting MAPK phosphatase 1 (MKP1) is rapidly stabilized upon UV-B exposure ( Besteiro and Ulm, 2013 ) and is necessary for UV-B stress resistance in the wild type ( Besteiro et al. , 2011 ).

Although some reactive oxygen species (ROS; e.g. H ₂ O ₂ ) production upon UV-B radiation is mediated by UVR8, for instance during stomatal closure ( Zhu et al. , 2014 ), there is possibly also a UVR8-independent pathway. Non-UVR8-mediated UV-B-increased ROS may be a consequence of disruption of metabolic activities ( Hideg et al. , 1993 ; Hideg and Vass, 1996 ; Zhang et al. , 2003 ; Lidon et al. , 2012 ) or of the action of membrane-localized NADPH oxidases ( Kalbina and Strid, 2006 ; Tossi et al. , 2009 ). ROS-scavenging enzymes can be induced by UV-B, especially at high doses [3.3W m ⁻² for 18h ( Esringu et al. , 2016 )], whereas glutathione peroxidase, glutathione transferases, and glutaredoxins were up-regulated upon exposure to short periods of high intensity UV-B ( Ulm et al. , 2004 ; Brown et al. , 2005 ). Even at low UV-B levels (0.093–0.127W m ⁻² ), exposure leads to the expression of oxidative stress-related defense genes such as glutathione reductase and accumulation of organic defense compounds such as pyridoxine ( Strid, 1993 ; Kalbin et al. , 1997 ). Despite the fact that ROS production has often been associated with irreversible stress or distress, elastic stress ‘eustress’ could also be ROS dependent ( Hideg et al. , 2013 ). Furthermore, ROS can be part of the signaling cascades involved, especially in combination with nitric oxide (NO) ( Tossi et al. , 2014 ).

A special case in the UV-B signaling mechanism is the regulation of the ARIADNE12 E3 ligase. This gene is induced by UV-B via the UVR8 pathway, yet, at higher fluence rates, the induction is independent of UVR8 but is still COP1 dependent ( Xie et al. , 2015 ), even though other COP1-regulating photoreceptors do not appear to stimulate this induction ( Xie and Hauser, 2012 ). To date, the underlying cause of COP1 regulation at higher fluence rates of UV-B remains unknown.

NO is well known as a signaling molecule and regulator of redox balance, and is often associated with oxidative stress ( Correa-Aragunde et al. , 2015 ; Domingos et al. , 2015 ). In general, UV-B light causes increased production of NO in plants (e.g. Mackerness et al. , 2001 ), with multiple physiological and morphological outputs as a result. UV-B perceived by UVR8 leads to an increase in NO that stimulates stomatal closure ( Ge et al. , 2014 ; Tossi et al. , 2014 ). Interestingly, UV-B induction of H ₂ O ₂ , derived from O ₂⁻ produced by AtrbohD and AtrbohF, and NO accumulation do not occur in Gα protein knockout mutants ( gpa1 ), and exogenous application of H ₂ O ₂ or NO rescues the defect ( He et al. , 2013 ). However, a link between UVR8 and Gα proteins has not been described. Regarding morphological outputs, UV-B-stimulated NO production was suggested to be the basis of growth reduction in maize ( An et al. , 2005 ). In addition, NO is an intermediate in the stimulatory effect on accumulation of photoprotective pigments in maize. Mechanistically, lack of NO blocks the UV-B-induced expression of maize P transcription factor (ZmP) and its targets CHALCONE SYNTHASE and CHALCONE ISOMERASE , suggesting that NO plays a key role in the UV-B-regulated phenylpropanoid biosynthetic pathway ( Tossi et al. , 2011 , 2012 ). In Arabidopsis, a similar mechanism exists where expression of HY5 and MYB12 transcription factors, and the target C4H , all regulated by UV-B, depends on the presence of NO ( Tossi et al. , 2011 ).

Consequently, the effects on NO are dual. On the one hand, UV-B increases the NO concentration, which converts UV-B-induced ROS such as superoxide anions (O ₂⁻ ) into peroxynitrite (ONOO ⁻ ). It is known that NO mitigates oxidative damage effects ( Siddiqui et al. , 2011 ; Shi et al. , 2005 ; Arora et al. , 2016 ), and NO is up-regulated by H ₂ O ₂ activation of NO synthase ( Zhang and Zhao, 2008 ) in excised leaves of kidney bean ( Phaseolus vulgaris ). Such an effect is also visible upon UV-B stress ( Shi et al. , 2005 ; Siddiqui et al. , 2011 ). Protection from UV-B-induced oxidative damage by NO has been observed in Arabidopsis, lettuce, and soybean ( Zhang et al. , 2009 ; Santa-Cruz et al. , 2014 ; Esringu et al. , 2016 ). On the other hand, NO, downstream of UV-B, is involved in the up-regulation of the expression of HY5, MYB12 (Arabidopsis), and ZmP (maize), and the consequent activation of phenylpropanoid and flavonoid biosynthesis pathways ( Mackerness et al. , 2001 ; Tossi et al. , 2011 ). The resulting NO-regulated accumulation of photoprotective pigments was also shown in birch (M. Zhang et al. , 2011 ), indicating a widespread response in higher plants.

UV-B and hormones

UV-B and auxin

Auxins are involved in nearly all developmental processes, such as cell elongation and differentiation, leaf development, stem elongation, root growth, and photo- and gravitropic growth. They are mainly biosynthetically derived from the amino acid tryptophan and are perceived by two types of receptors, a membrane-associated AUXIN BINDING PROTEIN (ABP)1 and TRANSPORT INHIBITOR RESPONSE (TIR)1/AUXIN SIGNALING F-BOX (AFB) receptors. The latter participate in marking AUX/IAA proteins, which are inhibitors of AUXIN RESPONSE FACTORS (ARFs), for degradation. Hence, the presence of auxin indirectly regulates ARF transcription factor activity. Plants exposed to UV-B display a dwarf phenotype, small thick leaves with short petioles, leaf curling, a short inflorescence, and an increased root/shoot ratio (reviewed by Jansen, 2002 ). These UV-B-induced changes in plant morphology point towards specific auxin-regulated processes, suggesting that auxins are involved in UV-B-induced acclimation processes under moderate UV-B doses and also in the stress response induced by high UV-B.

Indeed auxins have been frequently associated with UVB-regulated growth merely based on (i) the phenotypic similarities between UV-B-acclimated plants and phytohormone mutants ( Jansen, 2002 ) and (ii) the UV-B-mediated changes in the expression of auxin-related genes in seedlings and mature leaves ( Hectors et al. , 2007 , 2012 ; Brown and Jenkins, 2008 ; Favory et al. , 2009 ; Pontin et al. , 2010 ; Vandenbussche et al. , 2014 ). The involvement of auxins in this UV-B-induced morphogenesis can occur at distinct levels. First, auxin homeostasis can be affected via photo-oxidative damage, biosynthesis, conjugation, and/or degradation ( Curry et al. , 1956 ; Dezeeuw and Leopold, 1957 ; Jansen et al. , 2001 ). Furthermore, photomorphogenic regulation can occur at the level of redistribution (via transport, influx, and efflux) ( Naqvi, 1976 ; Wargent et al. , 2009 ; Ge et al. , 2010 ; Yu et al. , 2013 ; Fierro et al. , 2015 ) or at the level of a potential signal crosstalk between auxin and components of the light signaling pathway, changing auxin sensitivity ( Cluis et al. , 2004 ). The result yields a strong representation of genes regulated by UV-B and auxin in an opposite way ( Hectors et al. , 2007 ; Vandenbussche et al. , 2014 ; Fierro et al. , 2015 ).

Although the UV-B effects are usually associated with the shoot, a clear inhibition of root elongation and ectopic root hair formation can also be observed upon exposure to high UV-B doses ( Ge et al. , 2010 ; Krasylenko et al. , 2012 ). Roots express the UVR8 UV-B receptor ( Rizzini et al. , 2011 ) as well as two other root-specific UV-B-sensing proteins, ROOT UV-B SENSITIVE 1 (RUS1) and RUS2 ( Tong et al. , 2008 ; Leasure et al. , 2009 ), which may act in favor of a light-escape phototropic root growth ( Yokawa and Baluska, 2015 ).

The regulation of auxins is linked to the transcription factors central to UV-B signaling. Mapping of HY5-targeted genes revealed that HY5 is, among others, a transcription factor in auxin signaling and auxin transport ( Cluis et al. , 2004 ; Wargent et al. , 2009 ; Zhang et al. , 2011 ; Hayes et al. , 2014 ; Galvao and Fankhauser, 2015 ). The transcription regulators PIF4, PIF5, and PIF7 regulate the expression of a number of genes with PIF-binding sites in their promoter, including the AUX/IAA genes IAA19 and IAA29, and the auxin biosynthesis genes YUC2, YUC5, YUC8, and YUC9, indicating that PIFs modulate elongation growth by directly regulating the auxin-controlled responses at multiple levels ( Hornitschek et al. , 2012 ; Sun et al. , 2013 ) ( Fig. 1 ). All these facts provide good evidence that the HY5/HYH-dependent pathway couples the light signal with a direct control on both auxin biosynthesis and signaling ( Cluis et al. , 2004 ; Sibout et al. , 2006 ; Halliday et al. , 2009 ).

UV-B and brassinosteroids

Brassinosteroids (BRs) are plant-specific steroidal hormones with strong growth-promoting properties, essentially playing a role in nearly all phases of plant development ( PengZhu et al. , 2013 ). BRs and white light antagonistically regulate the developmental change from etiolation in darkness to photomorphogenesis in light conditions ( Luo et al. , 2010 ). In brief, BR signaling relies on the following mechanism: binding of BR to BRASSINOSTEROID INSENSITIVE 1 (BRI1) ( Li and Chory, 1997 ) leads to activation of the intracellular kinase domain and triggers a downstream signaling cascade eventually leading to positive regulation of the BR response by nuclear localization and dephosphorylation of two transcription factors BRASSINAZOLE RESISTANT1 (BZR1) and BRASSINOSTEROID INSENSITIVE1-ETHYL METHANESULFONATE-SUPPRESSOR (BES1) ( He et al. , 2005 ; Sun et al. , 2010 ; Yu et al. , 2011 ).

Transcriptomic data on whole rosettes indicate UV-B regulation of some genes involved in the BR pathways such as DWARF1 ( DWF1 ), BRASSINOSTEROID INSENSITIVE 1 SUPRESSOR 1 ( BRS1 ), BES-INTERACTING MYC-LIKE PROTEIN ( BIM1 ), and RELATED TO ABA-INSENSITIVE3/VIVIPAROUS1 (RAV1 ), suggesting that UV-B effects depend on BR signaling ( Ulm et al. , 2004 ; Hectors et al. , 2007 ). Recently, a direct interaction between BZR1 and COP1 has been reported ( Kim et al. , 2014 ). In darkness, COP1 resides in the nucleus and is part of an E3 ligase that regulates degradation of HY5, leading to inhibition of photomorphogenesis ( Lau and Deng, 2012 ). In darkness, COP1 captures the phosphorylated (inactive) form of BZR1 which leads to its degradation. This increases the ratio of dephosphorylated BZR1 (active form) compared with phosphorylated BZR1. Thus, COP1 action increases the formation of active BZR1 homodimers which positively steers BR signaling and promotes elongation growth in the dark ( Kim et al. , 2014 ). UVR8-mediated sequestration of COP1 abolishes its E3 ligase activity. A logical consequence could be that COP1 is no longer degrading the inactive BZR1; the inactive BZR1 levels increase and consequently lower the pool of active homodimeric BZR1, decrease BR signaling, and therefore inhibit elongation. Furthermore, when UVR8 is activated, HY5 protein accumulates and is found to interact directly with the dephosphorylated (active) form of BZR1 in Arabidopsis seedlings. During cotyledon opening in darkness, this interaction leads to an attenuation of the transcriptional activity of BZR1, therefore reducing BR signaling, suggesting that HY5 reduces skotomorphogenic growth in part by inhibiting BZR1 activity ( Li and He, 2016 ). Thus, in light conditions and under UV-B light, one can expect HY5 to inhibit BR signaling. In addition to the above, it has been demonstrated that membrane steroid-binding protein 1 (MSBP1), a negative regulator of BR signaling by enhancing the endocytosis of BAK1 ( Song et al. , 2009 ), directly interacts with and is activated by HY5 and HYH, thus further promoting photomorphogenesis and inhibiting elongation ( Kliebenstein et al. , 2002 ; Shi et al. , 2011 ; Yin et al. , 2015 ).

Moreover, there is also a link between BRs and PIFs. The BZR1 and PIF4 transcription factors interact directly, causing a co-regulation of light- and BR-responsive target genes ( Oh et al. , 2012 ). It is not known how this interaction will exactly influence the PIFs, but we could speculate that PIFs and HY5 are mutually exclusive for binding to BZR1 in view of their contrasting roles in photomorphogenesis. This suggests that BR may influence the balance between HY5 and PIF action by enhancing BZR1 effects.

Apart from the aforementioned photomorphogenic connections between UV-B signaling components and BR, there is also a crosstalk between UV-B and BR that affects defense signaling. Interestingly, BRs have been suggested as regulators of plant defense ( Malinowski et al. , 2009 ; De Bruyne et al. , 2014 ). In wild-type plants, pathogenesis-related protein 5 (PR-5) expression is clearly up-regulated by UV-B, while BR-deficient cpd mutants do not show this induction ( Savenstrand et al. , 2004 ). Thus, UV-B inducibility of certain genes is hampered in BR-deficient mutants. It is not known whether this is due to constitutive interference with the photomorphogenic program.

UV-B and gibberellins

Gibberellins (GAs) are mainly known as hormones that induce germination and flowering, and, in the context of this review, as positive regulators of expansion growth. Their signaling involves the breakdown of central DELLA proteins that cause growth inhibition (reviewed in Hauvermale et al. , 2012 ). UV-B can slightly alter the spectrophotometric characteristics of GAs, yet without affecting their growth-promoting activities ( Baltepe and Mert, 1974 ). This makes GAs an unlikely direct target for UV-B damage with plant physiological consequences. Therefore, regulation of GA pathways by UV-B is probably genetically controlled, and is supported by interactions between UV-B signaling components and GA signaling. In pea, the LIP1–LONG1 (orthologs of COP1–HY5) module regulates GA biosynthesis. LONG1, the ortholog of HY5 in pea, was shown to be responsible for diminishing the levels of the active GA ₁ . The presumed cause of this is a LONG1-dependent increase of expression of the GA-inactivating enzyme GA2OX2 in wild-type pea in white light ( Weller et al. , 2009 ). It is not known whether LIP1 and LONG1 are also fully functional orthologs for the Arabidopsis COP1 and HY5 genes in UV-B signaling. Nevertheless there are indications that this module provides UV-B-induced photoprotection by flavonoid accumulation ( Dukker, 2013 ). In addition, the downstream processes regulated by the pea LIP1–LONG1 module and UV-B signaling in Arabidopsis share factors related to GA biosynthesis and catabolism. Transcriptomic data in seedlings indicated induction of expression of GA-inactivating enzyme-encoding genes GA2OX2 and GA2OX8 by UV-B irradiation ( Ulm et al. , 2004 ; Weller et al. , 2009 ). A negative effect of UV-B on GA levels indeed appears somewhat widespread, as both in soybean and rice, UV-B reduced the GA content ( Lin et al. , 2002 ; Peng and Zhou, 2009 ). Also in Arabidopsis rosettes, GAs were suggested to be under UV-B control ( Hectors et al. , 2007 ). Recently, a more detailed study linking UV-B and GA pathways suggested GA2OX1 as a central component in the regulation of elongation growth by the UVR8 pathway ( Hayes et al. , 2014 ). Thus, it appears that there is a general mechanism in which UV-B causes inactivation of GA by enhancing the expression of GA2OX genes. In addition, confirming the suppressing effect on GA signaling, downstream, DELLA proteins accumulate upon UV-B radiation ( Hayes et al. , 2014 ). Della quintuple mutants, gai-t6rga-24rgl1-1rgl2-1rgl3-1 have constitutive GA signaling-mediated cell expansion, yet still respond to UV-B by growth inhibition. This supports the existence of a parallel UV-B-mediated growth-regulating pathway ( Hayes et al. , 2014 ). This may be the auxin pathway (see above) or another as yet unidentified component. It is noteworthy that there is evidence for a reciprocal regulation of GA signaling to the UV-B signaling, as in dark-grown plants with reduced GA levels, HY5 accumulates. This is probably a consequence of reduced COP1 action in the absence of GA ( Alabadi and Blazquez, 2009 ). Thus, lack of GA may further enhance the photomorphogenic UV-B effects. Finally, both the HY5 and PIF signaling pathways cross-react via DELLA proteins as PIFs are known to be repressed by abundant DELLA proteins ( Hayes et al. , 2014 ; Mazza and Ballare, 2015 ). UV-B, perceived by the photoreceptor UVR8, probably via a HY5/HYH-dependent pathway, leads to GA2OX expression and DELLA accumulation, suppressing PIF function. On the other hand, in parallel, a HY5/HYH-independent pathway leads to the degradation of PIF4 and PIF5, resulting in decreased auxin biosynthesis and activity, and limited elongation growth ( Hayes et al. 2014 ) ( Fig. 1 ).

UV-B and abscisic acid

Abscisic acid (ABA) is the main hormone in directing drought stress responses in plants. As drought is often accompanied by intense solar radiation, a number of studies on the interactions between UV-B and ABA are available ( Bandurska et al. , 2013 ). Despite relatively low absorption in the UV-B region ( Gao et al. , 2016 ) ( http://home.cc.umanitoba.ca/~adam/lab/hplc/aba.shtml ), a direct effect of UV-B light on ABA exists. Work in Rumex patientia showed that although ABA is a minor target for photolysis, solar UV-B is capable of isomerizing ABA to 50% cis , trans and 50% trans , trans ( Lindoo et al. , 1979 ; Bangerth, 1982 ). Cis , trans -ABA is the active form in plants, and supposedly the prevalent one; therefore, in tissues where sufficient UV-B penetrates, the active ABA pool may be negatively affected. Indeed, trans , trans ABA has been found in UV-B-treated plants ( Rakitina et al. , 1994 ).

Nevertheless, increased ABA accumulation seems to occur predominantly as part of a UV-B stress response in the model plant A. thaliana ( Rakitina et al. , 1994 ). More ABA accumulated upon moderate to high UV-B levels ( Table 1 ) in soybean ( Peng and Zhou, 2009 ), rice ( Lin et al. , 2002 ), tobacco ( Dinh et al. , 2013 ), lettuce ( Esringu et al. , 2016 ), and maize ( Tossi et al. , 2009 ). In this case, ABA appears to increase upon stress and cause a photoprotective effect as ABA biosynthesis mutants are more sensitive to UV-B with respect to leaf injury ( Tossi et al. , 2009 ). Further indirect support for this notion comes from the analysis of supersensitive to ABA and drought 2 ( sad2 ) mutants of Arabidopsis ( Verslues et al. , 2006 ). The sad2-1 mutant has enhanced ABA accumulation upon UV-B radiation ( Chen et al. , 2013 ), possibly leading to a feedforward mechanism, as sad2-1 is more sensitive to ABA ( Verslues et al. , 2006 ). The lack of nuclear importin function in sad2 mutants and consequent reduced nuclear accumulation of MYB4, an inhibitor of CINNAMATE-4-HYDROXYLASE expression, results in accumulation of more photoprotective pigments ( Zhao et al. , 2007 ). This leads to increased UV tolerance. However, it is not known whether ABA is directly involved in the nuclear localization of MYB4.

Also downstream of the biosynthesis, increased ABA signaling is recorded upon UV-B signaling in Artemisia ( Pan et al. , 2014 ). In conclusion, in many plants, ABA and UV-B signaling seem to join forces to enhance resistance to drought and UV-B ( Tossi et al. , 2012 ; Bandurska et al. , 2013 ) even extending the co-action of these signals beyond the plant kingdom, suggesting a general stress response. Along the same lines, hy5 mutants are resistant to ABA, suggesting that HY5 is potentiating the effects of ABA ( Chen et al. , 2008 ). Mechanistically, HY5 binds to the ABI5 promoter and is required for the expression of the bZIP transcription factor ABI5, and ABI5 target LEA (late embryogenesis abundant) genes ( Chen et al. , 2008 ; Xu et al. , 2014 ).

Yet, the effects of UV-B on ABA accumulation and signaling appear to be to some degree species dependent. Even in closely related species, significant differences can exist. While in Populus cathayana drought-induced ABA accumulation is enhanced by UV-B, in Populus kangdingensis ABA levels are not affected by UV-B ( Ren et al. , 2007 ). Furthermore, in R. patientia , no increase in ABA levels was detected in high UV-B intensities ( Lindoo et al. , 1979 ). Moreover, UV-B diminishes ABA levels in soybean developing seeds ( Liu et al. , 2013 ). In grapevine (Malbec), transcriptomics point to up-regulation of genes involved in the major ABA catabolic pathway, including ABA 8'-hydroxylase, by low UV-B ( Pontin et al. , 2010 ). For the latter observation, it is unclear whether this leads to less ABA, or whether this reflects a feedback inhibition pathway to overcome an overload of ABA.

Finally, it is likely that the increase in ABA is stress related, and that what is perceived as UV-B stress depends on the species, with different thresholds of UV-B intensity causing their effect differently in different species.

UV-B and ethylene

Ethylene is a gaseous hormone that has functions in both morphogenesis and stress signaling. Few data have been published on the interaction between ethylene and UV-B signaling. It was shown that the absence of the central component and positive regulator of ethylene signaling EIN2 leads to enhanced protection from UV-B stress due to elevated flavonoid levels ( Sun et al. , 2011 ). These results are in line with ethylene inhibiting photomorphogenesis by stimulating COP1 function ( Yu et al. , 2013 ). Upon UV-B exposure, flavonoids usually accumulate to high levels in the epidermis. Interestingly, ethylene mediates a specific epidermal stomatal closure response to UV-B in Vicia faba L., acting upstream of H ₂ O ₂ and NO ( He et al. , 2011 ). Notably, most of the 1-aminocyclopropane-1-carboxylic acid (ACC) synthase genes of Arabidopsis, encoding the rate-limiting step in ethylene biosynthesis, are expressed in the epidermis, but not in the mesophyll ( Tsuchisaka and Theologis, 2004 ). UV-B radiation at various, mainly high, intensities ( Table 1 ) stimulates ethylene production in many plant species including pear ( Predieri et al. , 1995 ), Arabidopsis ( Mackerness et al. , 1999 ; Rakitina et al. , 2001 ), tobacco ( Nara and Takeuchi, 2002 ; McLeod et al. , 2008 ), maize ( Wang et al. , 2006 ), and canola ( Qaderi et al. , 2010 ). However, it is not clear whether this production is enzymatic or not. It has been suggested that UV-B-induced ethylene is derived from pectin in the leaves ( McLeod et al. , 2008 ), while production through the canonical ethylene biosynthesis pathway with ACC as intermediate has also been documented ( Nara and Takeuchi, 2002 ; Wang et al. , 2006 ). The accumulation of ACC was UV-B dose dependent in pea, with more ACC accumulating at higher UV-B doses ( Katerova et al. , 2009 ). Part of the underlying molecular mechanism may lie in the transcriptional up-regulation of ACC synthases. In tomato, ACS genes are up-regulated in the presence of enhanced UV-B ( An et al. , 2006 ). In Malbec grapevine plantlets, up-regulation of ethylene biosynthesis and transcription factor transcripts were seen under relatively mild UV-B conditions ( Pontin et al. , 2010 ) ( Table 1 ). However, this seems an exception. Many data point to stress signaling enhancing ethylene biosynthesis, rather than the UVR8 pathway.

Furthermore, the MPK3/MPK6-dependent pathway, involved in UV-B stress responses ( Besteiro et al. , 2011 ), has control of ethylene biosynthesis as one of its outputs. MPK3/MPK6 action leads to increased activity of stress-related ACC synthases ( Han et al. , 2010 ), and ethylene-related transcripts increase at UV-B levels associated with stress ( Ulm et al. , 2004 ). In contrast, the UVR8 pathway is an unlikely candidate for stimulating ethylene biosynthesis. Transcriptome data point towards down-regulation of the ethylene signal by a low level of UV-B ( Hectors et al. , 2007 ), and HY5 reduces ethylene biosynthesis. HY5 binds to the AtERF11 promoter to activate its transcription. AtERF11 is a transcriptional repressor and interacts with the dehydration-responsive element in the ACS2/5 promoters, resulting in decreased ethylene biosynthesis ( Li et al. , 2011 ). In addition, PIF-controlled reduction of the cytosolic enzyme METHIONINE GAMMA LYASE (MGL), degrading the ethylene precursor methionine to α-ketobutyrate and further to isoleucine ( Rebeille et al. , 2006 ), has an effect on ethylene production as well ( Fig. 1 ). Again, the effects at low levels of UV-B may be species dependent, as in Artemisia the ethylene signal is increased ( Pan et al. , 2014 ).

UV-B and jasmonates

Jasmonate (JA) is a well-known hormone in the regulation of defense against herbivores and necrotrophs, and also has a role in thigmomorphogenesis and wounding responses. JA absorbs very poorly in the UV-B region of the spectrum ( http://home.cc.umanitoba.ca/~adam/lab/hplc/jasmonicacid.shtml ). At high intensities, UV-B has been shown to increase the production of JA in Arabidopsis ( Mackerness et al. , 1999 ). JA accumulates more in the mung bean cultivar HUM12, in the presence of supplemental UV-B ( Choudhary and Agrawal, 2014 ) Solar UV-B is able to increase JA biosynthesis in Nicotiana sp. ( Izaguirre et al. , 2003 ; Dinh et al. , 2013 ).

Yet, exceptions exist, as in tobacco, UV-B does not stimulate JA production, yet enhances the JA-dependent induction of trypsin protease inhibitors and defense ( Demkura et al. , 2010 ). Similarly, in tomato, supplementary UV-B potentiates the defense by increasing trypsin protease inhibitor gene expression ( Stratmann et al. , 2000 ). Therefore, UV-B has been suggested to co-opt defense signaling ( Stratmann, 2003 ). This UV-B effect is believed to contribute to the resistance of plants against herbivores, but has only been documented for Solanaceae. In Arabidopsis, the UV-B effect on defense was suggested to be independent of JA ( Demkura and Ballare, 2012 ). Nevertheless, UV-B decreases the attractiveness of Arabidopsis plants for the diamondback moth, a process which depends on JA signaling ( Caputo et al. , 2006 ). Available microarray data indicate that UV-B radiation increases the expression of some JA-related genes ( Mazza and Ballare, 2015 ). In contrast, supplementary moderate levels of UV-B on broccoli sprouts increased the expression of JA signaling genes, while negatively affecting the performance of Pieris brassicae caterpillars ( Mewis et al. , 2012 ).

The up-regulation of JA derivatives by UV-B in some species may be something more than a beneficial side effect on plant defense. MethylJA protects barley seedlings from UV-B stress, by enhancing antioxidant and free radical-scavenging abilities, suggesting that endogenous elevation of JA by UV-B is a response that protects plants from high level UV-B stress ( Fedina et al. , 2009 ) ( Table 1 ).

UV-B and salicylate

Salicylic acid (SA), a defense hormone against biotrophs, has been frequently associated with ROS ( Herrera-Vasquez et al. , 2015 ). Since light in the UV range can cause ROS production, it is no surprise that UV light has a link with SA. Numerous observations indicate an increase in SA and plant defense under UV-C stress in plants, including tobacco ( Yalpani et al. , 1994 ; Yao et al. , 2011 ), Arabidopsis ( Yao et al. , 2011 ; Mintoff et al. , 2015 ), and pepper ( Mahdavian et al. , 2008 a ). Some of these typical UV-C effects also apply to UV-B.

Although SA has a small absorption peak at 303nm, ( http://home.cc.umanitoba.ca/~adam/lab/hplc/salicylicacid.shtml ), the effect of UV-B on SA stability in plants has, to the best of our knowledge, not been documented. However, analysis of photodegradation in water by UV-B indicated a minor but significant effect ( Carlotti et al. , 2007 ), which in a plant physiological context may be negligible. In fact, in plants, on many occasions, an increase in SA has been observed upon UV-B exposure. UV-B induces SA accumulation in roots and leaves of barley ( Bandurska and Cieslak, 2013 ) and wheat ( Kovacs et al. , 2014 ). In tobacco, supplemental UV-B increased SA, probably by an increase in phenylalanine ammonia lyase. Consequently, downstream PATHOGENESIS RESISTANCE (PR)-1, PR-3, and PR-5 defense proteins accumulate ( Fujibe et al. , 2000 ). At least the UV-B induction of PR-1 in tobacco is dependent on ROS, yet SA can induce PR-1 without ROS accumulation ( Green and Fluhr, 1995 ). Something very similar occurs in A. thaliana leaves. UV-B increases SA levels, and downstream PR-1 , PR-2 , and PR-5 ( Surplus et al. , 1998 ). This response is severely attenuated by antioxidants and in NahG lines in which SA is catabolized, suggesting a role for ROS and SA mediating the response ( Surplus et al. , 1998 ). Furthermore, in Arabidopsis, PR-1 and GRXC-9 are induced by UV-B and SA ( Herrera-Vasquez et al. , 2015 ). Transcription is induced from as1 -like elements, which are targets for the SA pathway. Further evidence suggests that this response is conserved in more plant species. Transcriptome data in broccoli sprouts indicated increased expression of PR-1 , PR-2 , and PR-4 homologs of Arabidopsis genes associated with SA pathways by supplementary UV-B ( Mewis et al. , 2012 ). In addition, aphids performed less well on these plants ( Mewis et al. , 2012 ). This suggests that the regulation of SA by UV-B may be a way to enhance plant defense. The above results indicate the usefulness of UV-B in regulating the response to biotrophic pathogens, yet crosstalk with abiotic stresses also exists. In combination with the osmolyte polyethyleneglycol, UV-B induced more SA and prevented wilting in wheat ( Kovacs et al. , 2014 ). However, the UV-B–defense connection appears complex. In a study on the natural variation in Arabidopsis accessions in the resistance against UV-B stress, mutations in the defense gene ACTIVATED DISEASE RESISTANCE 2 ( ADR2 ) conferred higher UV stress resistance. In contrast, UV-B-sensitive accessions had strong ADR2 transcriptional activation and accumulated more SA, while these processes were much less affected in resistant plants ( Piofczyk et al. , 2015 ).

Perhaps of more importance is to know whether SA is useful in protecting against UV-B stress. SA accumulates more in mung bean cultivar HUM1, in the presence of supplemental UV-B, correlating with its higher sensitivity to UV-B stress than HUM12 ( Choudhary and Agrawal, 2014 ). So it seems that plants try to cope with UV-B stress by accumulating more SA, especially when they are stressed in general. Accumulating SA upon UV-B exposure may indeed be useful for plants. Spraying with SA protects pepper and Kentucky blue grass plants from UV-B stress ( Ervin et al. , 2004 ; Mahdavian et al. , 2008 b ). At this time, it is not known whether this protection occurs only in ‘UV-sensitive’ lines or cultivars.

UV-B and cytokinins and strigolactones

To finish, some hormones have been poorly studied regarding their UV-B regulation. Data on cytokinin (CK) and strigolactone (SL) regulation in particular are very scarce. UV-B at very high levels diminishes the CK levels in cucumber ( Kataria et al. , 2005 ) ( Table 1 ). In pea, depending on the cultivar, UV-B positively or negatively affects CK oxidase activity in plants with high and low CK levels, respectively ( Vaseva-Gemisheva et al. , 2004 ; Todorova et al. , 2006 ). The transcription regulators PIF4, PIF5, and PIF7 are known to regulate the expression of the CK oxidase genes, CKX5 and CKX6, both having a PIF-binding site in their promoter ( Hornitschek et al. , 2012 ), suggesting CKX genes are indirectly controlled by UV-B. In rice, the effect of UV-B on CK shows temporal variation ( Lin et al. , 2002 ). In addition, CKs crosstalk with HY5 in regulating flavonoid biosynthesis ( Vandenbussche et al. , 2007 ), suggesting that an interplay with UVR8 signaling may exist.

For SL, studies on the effect of UV-B are currently missing, but, as with CKs, they may make a connection with HY5 ( Jia et al. , 2014 ). SL increases the level of HY5 in white light (UV not mentioned) ( Jia et al. , 2014 ). Both SL and CKs regulate shoot branching. Since UV-B increases branching ( Hectors et al. , 2007 ), an interplay with these hormones is likely, and may be a subject for future studies.

Concluding remarks

Clearly, as for other components of the solar spectrum, the effects of UV-B on plant physiology are mediated by hormones. Some of these hormones (ABA, JA, SA, and NO) are up-regulated in their action, and appear mainly stress associated and involved in photoprotection ( Fig. 2 ). Other hormonal pathways (GA and auxins) are inhibited and mainly confer morphological alterations. Ethylene can be either up-regulated, probably as part of a stress response to high UV-B levels, or down-regulated, probably as a morphogen. The data on BR and SL, in relation to UV-B, are very scarce, yet a functional association with components of the UVR8 pathway has been described, thus also making them candidates for interaction with UV-B-controlled photomorphogenic processes. It is expected that the current scarcity of information will quickly decrease thanks to the increasing interest in this topic.

Acknowledgements

This work was supported by the Research Foundation Flanders (project no. G.0005.15N) to FV and EP, and grants from Ghent University and the Research Foundation Flanders (G.0656.13N) to DVDS.

References