• Environmental challenges such as low light intensity induce differential growth-driven upward leaf movement (hyponastic growth) in Arabidopsis thaliana. However, little is known about the physiological regulation of this response. Here, we studied how low light intensity is perceived and translated into a differential growth response in Arabidopsis.
• We used mutants defective in light, ethylene and auxin signaling, and in polar auxin transport, as well as chemical inhibitors, to analyze the mechanisms of low light intensity-induced differential growth.
• Our data indicate that photosynthesis-derived signals and blue light wavelengths affect petiole movements and that rapid induction of hyponasty by low light intensity involves functional cryptochromes 1 and 2, phytochrome-A and phytochrome-B photoreceptor proteins. The response is independent of ethylene signaling. Auxin and polar auxin transport, by contrast, play a role in low light intensity-induced differential petiole growth.
• We conclude that low light intensity-induced differential petiole growth requires blue light, auxin signaling and polar auxin transport and is, at least in part, genetically separate from well-characterized ethylene-induced differential growth.
The gaseous plant hormone, ethylene, is the key trigger for submergence-induced hyponasty in Arabidopsis thaliana (Millenaar et al., 2005). Other signals, such as vegetation-derived low-red : far-red ratio (Pierik et al., 2005), low light intensity (Millenaar et al., 2005; Mullen et al., 2006) and elevated temperatures (Koini et al., 2009) also induce hyponastic growth in Arabidopsis. It was shown that ethylene interacts with light-quality cues to induce shade-avoidance responses, including hyponasty, in tobacco (Nicotiana tabacum) (Pierik et al., 2004b). More recently, phytochrome-mediated petiole elongation in Arabidopsis was shown to rely on ethylene as well as on the phytohormone auxin, and ethylene-induced hypocotyl elongation was suggested to occur through the interaction with auxin (Pierik et al., 2009). Auxin and its polar transport (PAT) are fundamental players in the plant's developmental program, but are also typically associated with differential growth (Lehman et al., 1996; Harper et al., 2000; Friml & Palme, 2002; Friml, 2003) and with shade-avoidance responses (Morelli & Ruberti, 2000; Tao et al., 2008; Pierik et al., 2009). More specifically, auxin plays a role in submergence-induced hyponastic growth in Rumex palustris (Cox et al., 2004) and in the regulation of leaf movement in agar-grown Arabidopsis seedlings (Vandenbussche et al., 2003). However, it is unknown how low light intensity induces differential growth responses and if this response requires auxin. In the present study, we aim to understand how low light intensity is perceived and how this signal is translated into a hyponastic growth response.
We showed that the low blue light photon fluence rate is an important component of low light intensity to establish a rapid induction of hyponastic petiole growth. Cryptochrome 1 (cry1), cry2, phytochrome A (phyA) and phyB are the photoreceptor proteins involved in detecting reduced light intensity. In addition, photosynthesis-derived signals can also induce differential growth. Using mutant analyses, ethylene production measurements and transcript analysis, we demonstrate that low light-induced differential petiole growth does not require ethylene perception and as such is, at least in part, genetically separate from the submergence signaling route towards hyponastic growth. We furthermore provide evidence that auxin and PAT are required to induce a maximal hyponastic growth response and to maintain elevated petiole angles.
Materials and Methods
Plant material and growth conditions
Three Arabidopsis thaliana (L.) Heynh accessions were used: Columbia-0 (Col-0; N1092), Landsberg erecta (Ler; NW20) and Wassilewskija-2 (Ws-2; N1602). [The numbers within the parentheses represent the Nottingham Arabidopsis Stock Center accession numbers.] The corresponding genes for all mutants used in this study are significantly expressed in Col-0 petioles (Supporting Information Table S1). The mutants used in our studies were obtained either from the stock center or from the authors who described the mutants.
The growth procedures were as described in Millenaar et al. (2005). Conditions in the growth chamber were: 20°C, 70% (v/v) relative humidity, and 9 h photoperiod (200 µmol m−2 s−1 photosynthetic active radiation (PAR)). Plants were automatically tap-watered each day at the start of the photoperiod. Plants in stage 3.9, according to Boyes et al. (2001) (i.e. a vegetative plant with approx. 15 rosette leaves), were used in all experiments. All experiments started 1.5 h after the start of the photoperiod to minimize variation, which could be induced by circadian and diurnal rhythms (Salter et al., 2003). All experiments were repeated at least twice, with independent plant batches.
Light manipulation experiments
For experiments carried out under low light intensity, the light quantity was reduced by 90%, from 200 µmol m−2 s−1 to 15–20 µmol m−2 s−1 (PAR) at the start of the experiment. This was achieved by using a neutral shade cloth that did not change the relative light spectral composition (Fig. S1); this was verified using a Licor1800 spectroradiometer (Licor, Lincoln, NE, USA).
Low light was enriched in blue-light components or in red-light components by placing colored filters on top of boxes and applying forced ventilation to guarantee continuous refreshment of the air (Fig. 1c, Fig. S2; Lee Filters, Andover, UK). To filter out all colors but red, a ‘026 Bright Red’ filter was used. To filter out all colors but blue, a ‘200 Double CT Blue’ filter, in combination with a 2-cm layer of 25 g l−1 CuSO4·5 H2O solution (Merck, Darmstadt, Germany), was used. CuSO4 removes far-red light and was necessary to maintain a red : far-red ratio similar to the other treatments (data not shown). Because of the low transparency of some of the filters used, we reduced the control light intensity in order to have comparable light intensities. Therefore, the hyponastic growth response was relatively low in this experiment compared with other experiments described.
Ethylene application was as described in Millenaar et al. (2005). Pure ethylene (Hoek Loos BV, Schiedam, the Netherlands) and air (70% relative humidity) were mixed using flow meters and controllers (Brooks Instruments BV, Ede, the Netherlands) to generate a concentration of 5 µl l−1 ethylene. This was flushed continuously, at 75 l h−1, through glass cuvettes containing the plants, and was then vented away. A concentration of 1 µl l−1 ethylene was reached in the cuvettes after c. 10 min; 5 µl l−1 was reached after 40 min. The ethylene concentration was checked regularly on a gas chromatograph (GC955; Synspec, Groningen, the Netherlands), and remained constant for the duration of the experiment. Control cuvettes were flushed with air (70% relative humidity) at the same flow rate.
Ethylene release was measured in real-time on control and low light intensity-treated Col-0 plants under standard growth chamber conditions. Single plants were placed in glass cuvettes (1.5 l) that were flushed continuously with ethylene-free air (1 l h−1 flow rate) 1 d before the treatment. On the next day, plants were either subjected to standard low light-intensity conditions or remained untreated (i.e. controls). During the measurements, air that had left the cuvettes was led through scrubbers to remove all CO2 and H2O and then entered a laser-driven photo acoustic ethylene detector (SensorSense, Nijmegen, the Netherlands) to measure real-time ethylene release from the plants. The ethylene detector signal was corrected for background noise using an empty cuvette and for soil-derived ethylene, which was measured from a pot with a standard substrate, but no plant. After 24 h, plants were taken out of the cuvettes and weighed. Ethylene production rates were calculated as nl of ethylene h−1 g−1 FW.
3-(3,4-Dichlorophenyl)-1,1-dimethylurea (DCMU) (10 µm) and 2,3,5-triiodobenzoic acid (TIBA) (50 µm), in 0.1% Tween and 0.01% ethanol, were applied to the plants by spraying. Mock plants were treated with identical solutions that lacked the active components. Pretreatment with TIBA took place at 66, 42 and 18 h before the start of the experiment.
Computerized digital camera system and image analysis
To measure changes in petiole angle, a custom-built computerized digital camera system was used as described in Millenaar et al. (2005). To enable continuous photography, no dark period was included in the 24-h experimental period. Single plants were placed in glass cuvettes with the petiole under investigation positioned perpendicular to the axis of the camera. To facilitate the measurements, any leaf that was obscuring the petiole being photographed was removed. Additionally, the petiole was marked at the petiole/lamina junction with drawing ink. These preparations did not affect the response of the petiole (data not shown). Digital photographs were taken every 10 min. The angles of the petioles were measured on these images using a PC-based image-analysis system with a custom-made macro using the ks400 (version 3.0) software package (Carl Zeiss Vision, Germany). The petiole angles from the light dose–response curves and light-enrichment experiments were measured using ImageJ version 1.35o (http://rsb.info.nih.gov/ij/). The petiole angle is defined as the angle between the horizontal and a line drawn between the ink mark at the petiole/lamina junction and a fixed point at the base of the petiole. To take into account the changes in angle of control plants during the course of the experiments, we performed a pairwise subtraction. This is the difference between the angles of treated and control plants for each time point (Benschop et al., 2007). The new standard errors for the differential response were calculated by taking the square root from the summation of the two squared standard errors.
The measurements obtained of the mutant plants (Table S2) and of the wild-type plants were compared statistically by testing fitted parameters of a sigmoidal function, which was used previously by Cox et al. (2003).
Petioles of Col-0 were harvested and snap-frozen in liquid nitrogen. Subsequently, RNA was isolated from these petioles using the RNeasy extraction kit (Qiagen, USA). Genomic DNA (gDNA) removal, complementary DNA (cDNA) synthesis and real-time reverse transcriptase–polymerase chain reaction (RT-PCR) were performed as described by Millenaar et al. (2006). Primer sequences of ACO4 (At1g05010), ERS2 (At1g04310) and ACT2 (reference; At5g09810) were as described in Millenaar et al. (2006). Real-time RT-PCR data were calculated using the comparative cycle threshold (Ct) method described by Livak & Schmittgen (2001).
Characterization of low light intensity-induced hyponastic growth
To characterize the hyponastic growth response of A. thaliana to low light-intensity conditions, we studied the dose–response relationship between light intensity and petiole angles in Col-0 and Ler. Petiole angles decreased gradually with increasing light intensities, from 5 to 200 µmol m−2 s−1 PAR (Fig. 1a). Figure 1(b) demonstrates the kinetics of low light-intensity (20 µmol m−2 s−1) induced hyponastic growth. In both Col-0 and Ler the response started within 1 h, the angle change per unit time was maximal after c. 3 h and the maximal angles were reached at c. 10 h (Col-0) or 16 h (Ler).
To obtain insight into how low light intensity is sensed, the sensitivity of petioles for different wavelength regions was studied in Col-0. If low blue or low red is important for detection of low light intensity, then low light that is relatively enriched in blue or red should prevent hyponastic growth. The photon fluence rates of blue light (400–500 nm) and red light (600–700 nm) during the enrichment experiment were made comparable to control light treatment, while total light intensity (400–700 nm) and the red : far-red ratio (655–665:725–735 nm) were kept similar (Fig. 1c, Fig. S1). After 6 h in blue light-enriched low light, the observed petiole angles (−4.1 ± 1.7 degrees) were significantly (P < 0.05) different from those of plants in nonmanipulated light conditions (Fig. 1d). By contrast, growth in red light-enriched low light did not prevent hyponasty in low light. After prolonged treatment (24 and 48 h) all low light-intensity treatments had induced hyponastic growth, including the one enriched in blue light.
Figure 1(e) shows that the time-of-induction of low light intensity does not determine the ability to respond. Even when low light was administered 1.5 h after the start of the subjective night period (t = 9 h), a distinct hyponastic growth response, with normal kinetics, was observed.
cry1, cry2, phyA and phyB sense low light intensity
To test how low light is perceived, we analyzed photoreceptor mutants. Arabidopsis has four known blue-light photoreceptors: cry1, cry2, phototropin 1 (phot1) and phot2 (Lin, 2002 and references therein). The phot1 and phot2 mutants and the phot1 phot2 double mutant showed a low light intensity-induced hyponastic growth response which was similar to that of the wild type, suggesting that phototropins are not involved in this response (Fig. 2a–c). Also, the cry1/hy4 and cry2/fha1 mutants displayed a response similar to that of the wild type (Fig. 2d,e; Table S1). Interestingly, the cry1 cry2 double mutant showed a reduced response to low light (Fig. 2f), suggesting that cryptochromes are redundantly involved in low light intensity-induced hyponastic growth.
Phytochromes absorb blue, red and far-red light, and the phyA and phyB mutants showed delayed low light intensity-induced hyponastic growth (Fig. 2g,h). To test the involvement of other phytochromes, we examined the phytochrome mutant, long hypocotyl2/genome uncoupled3, which contains reduced amounts of phyA–E. The delay of low light intensity induced hyponastic growth in the hy2 mutant (Fig. 2i) was comparable to that of the phyA and phyB mutants, suggesting that phyC–E do not act redundantly to phyA and phyB in low light intensity-induced hyponastic growth.
Double combinations of phyA with cry1 or cry2, and the phyA cry1 cry2 triple mutant, showed delayed responses similar to those of the phyA single mutant or of the cry1 cry2 double mutant (Fig. 2j–l). This suggests that phyA cannot substitute for cry1 and/or cry2 in sensing low light intensities.
After 24 h, all mutants tested responded to low light in a manner similar to that of the wild type. Consequently, there may be additional, unrelated signaling pathways involved in low light intensity-induced differential growth. The photosynthetic machinery can generate light intensity-dependent signals that influence nuclear gene expression. We tested involvement by inhibiting photosynthetic electron transport with DCMU. This treatment induced hyponastic growth in control light conditions (Fig. 3a). The executer1 fluorescent (ex1 flu) double mutant, disturbed in signaling from the chloroplasts to the nucleus, showed a clear delay of hyponastic growth in low light intensity (Fig. 3b). Together, these data suggest that signals derived from the photosynthetic machinery can influence hyponastic growth.
Ethylene perception is not required for low light intensity-induced differential growth
Ethylene induces a hyponastic growth response of which the kinetics shows remarkable similarity with low light intensity-induced hyponasty (Millenaar et al., 2005). It would therefore be possible that ethylene is a downstream element in low light intensity-induced hyponastic growth.
To test this, we studied low light intensity-induced hyponastic growth in the ethylene-insensitive mutants ethylene insensitive 2 (ein2), ethylene response1-4 (etr1-4) and ein4-1, and in the constitutive ethylene-response mutant constitutive triple response (ctr1). None of these showed an aberrant response to low light intensity (Fig. 4a–d).
Moreover, ethylene production measured online using photo acoustic spectroscopy did not significantly change upon exposure to low light intensity (Fig. 4e).
Real-time RT-PCR experiments confirmed that the expression of ACC-oxidase (At1g05010) and ETHYLENE RESPONSE SENSOR2 (ERS2), serving as ethylene marker genes (Wilkinson et al., 1995; Hua et al., 1998; Vriezen et al., 1999), were strongly induced by ethylene treatment (Fig. 4f, g), but not at all by low light-intensity treatment. The same conclusion could be drawn based on a broader set of ethylene-biosynthesis and signal-transduction marker genes (Fig. S2). Together, these data suggest that ethylene is not an essential component in the regulation of low light intensity-induced hyponasty.
Auxin perception and polar auxin transport are required for low light intensity-induced hyponastic growth
To test the involvement of auxin transport in low light intensity-induced hyponastic growth, we applied the polar auxin transport (PAT) inhibitor, TIBA, to Col-0 (Fig. 5a). Interestingly, inhibitory effects of TIBA on the magnitude of hyponastic growth and on the maintenance (continuation of the high leaf angle after the maximum response magnitude, up to 24 h) of an elevated leaf angle were observed in low light intensity (Fig. 5a; Table S1). Similarly, albeit stronger, inhibition of low light intensity-induced hyponastic growth was observed after treatment with the auxin transport inhibitor naphthylphtalamic acid (NPA; data not shown). These effects were not caused by toxicity of the chemicals and solvents used, as a similar treatment had no effect on ethylene-induced hyponastic growth (Fig. S3).
To specify the involvement of auxin in low light intensity-induced hyponastic growth, we subsequently tested mutants with disturbed auxin perception or PAT. The TRANSPORT INHIBITOR RESPONSE1 (TIR1) protein is an auxin receptor (Dharmasiri et al., 2005b; Kepinski & Leyser, 2005). The tir1 mutant in Col showed a low light-intensity response that was comparable to that of Col-0 plants treated with TIBA (i.e. reduced maximal angle, and angle maintenance defects) (Fig. 5b). Five close TIR1 homologous F-box proteins are named AFB1–5. The tir1 afb1 afb2 afb3 quadruple mutant had a significantly lower maximal angle, and maintenance defects were visible (Fig. 5c), similarly to tir1.
The tir3-1 mutant in Col, in which auxin transport is hampered, and mutants in the auxin efflux-associated proteins, pin-formed3 (pin3) and pin7 all showed angle-maintenance defects similar to those of TIBA-treated Col plants and tir1. pin3 and pin7, but not tir3, were also affected in their maximum angle (Fig. 5d–f).
In conclusion, both pharmacological and genetic data demonstrate that auxin perception and PAT are involved in the regulation of the magnitude of low light intensity-induced differential growth as well as in the maintenance of elevated petiole angles in low light. Auxin and PAT do not seem to be required for a fast induction of the low light intensity-induced hyponastic growth response.
Hyponastic growth upon detection of light quality changes is a component of the shade-avoidance syndrome, a suite of traits that direct plant growth towards the better-lit zones of a vegetation, away from the shade imposed by neighbors (Franklin, 2008). Next to these morphological responses to light quality, reductions in light quantity also induce a variety of acclimations. Shade exposure, for example, induces strong acclimations at the photosynthetic level, including a re-allocation of photosynthetic machinery from the old, shaded, leaves to the younger ones that are higher in the canopy where light conditions are more favorable (Boonman et al., 2006). We propose that the low light intensity-induced hyponastic growth response described here contributes to these shade-acclimation responses because it facilitates the exposure of particularly young leaves to less-shaded conditions. This adds light intensity to the palette of signals that induce plant responses to neighbors. We show here that this low light-intensity response could be a response to particularly the reduced blue light fluence rate, which has been shown previously to induce shade-avoidance responses (Ballaréet al., 1991; Pierik et al., 2004b; Sasidharan et al., 2008). Interestingly, not only cryptochromes, but also phyA and phyB, appear to be involved in the hyponastic growth response to low light intensity. The phytochrome involvement is different from the low blue light-mediated hypocotyl elongation in Arabidopsis seedlings that was recently shown to be mediated by cry1 and cry2 (Pierik et al., 2009) without involvement of phytochromes. It is unlikely that this important role for phytochromes in low light intensity-induced hyponastic growth is caused by the red light component in the light treatments because red-light enrichment could not prevent hyponasty. The latter finding differs from the results of a study on hyponastic growth by Mullen et al. (2006), who showed that red light could prevent hyponasty. However, to show this, these authors used a red-light treatment of 100 µmol m−2 s−1, which is even higher than the red light component used in the controls in the present study.
We therefore hypothesize that the phytochrome involvement found here might be a result of its sensitivity to blue-light wavelengths (Whitelam et al., 1993; Neff & Chory, 1998). However, as phyB is a constitutively shade-avoiding mutant, with constitutively high petiole angles (phyB-5, 42.2 ± 1.5; Ler, 31.2 ± 0.5 degrees in control light conditions), we cannot exclude that the reduced response to low light intensity is the result of mechanical constraint and/or saturation of the signaling route caused by this higher initial angle.
Prolonged exposure to low light intensity led to the induction of hyponastic growth, even in combined multiple loss-of-function photoreceptor mutants. Alternative signals might come from the photosynthetic machinery, as it can generate light intensity-dependent signals that influence nuclear gene expression. Possible signals are the redox state of the plastoquinone pool, thioredoxine, oxygen radicals, chlorophyll synthesis and sugar status (Danon & Mayfield, 1994; Escoubas et al., 1995; Pfannschmidt et al., 1999; Strand et al., 2003; Geigenberger et al., 2005; Piippo et al., 2006). Indeed, we observed hyponastic growth after addition of the photosynthesis inhibitor DCMU under control light conditions that normally do not induce hyponastic growth (Fig. 3a). Furthermore, we observed a decreased hyponastic response to low light in the ex1 flu mutant (Fig. 3b), which is disturbed in signaling from the chloroplasts to the nucleus (Wagner et al., 2004; Lee et al., 2007).
Interestingly, multiple lines of evidence indicate that low light intensity-induced differential petiole growth in Arabidopsis acts independently of ethylene (Fig. 4, Fig. S2). However, downstream of low light intensity and ethylene, the two pathways may still merge. Both signals induce a very similar response (Millenaar et al., 2005), and low light intensity and ethylene affect the expression of a shared pool of 453 genes (165 up-regulated and 288 down-regulated; Pierik et al., 2005). Furthermore, combined treatment with ethylene and low light intensity induced a hyponastic response that is not different from that induced by either treatment alone (Millenaar et al., 2005), which might indicate that the two pathways saturate shared downstream components. It is thus possible that basic regulatory machinery, required for petiole movement, which controls localized growth responses, is regulated by a variety of environmental signals.
The results we present contrast those of Vandenbussche et al. (2003), who showed abolishment of low light intensity-induced hyponastic growth in ethylene-insensitive mutants. Furthermore, ACC synthase 6 (ACS6) and ACS8 were up-regulated when grown under conditions of shade. This contradicts our micro-array results that showed no change in ACC synthases (Fig. S2) and with the lack of regulation of ethylene production by low blue light as shown in Pierik et al. (2009). In the experiments by Vandenbussche et al. (2003), agar-grown seedlings were used and treatment for 6 d was required to observe a change in leaf angle, rather than the very rapid response (within hours) that we describe here. Therefore, these are probably different responses that are based on different mechanisms.
Lowering the auxin content by de-blading leaves of the semi-aquatic species R. palustris suggested a stimulatory role for auxin in all aspects of submergence-induced hyponastic growth (Cox et al., 2004). More gentle manipulation of petiolar auxin levels, using the PAT inhibitors TIBA and NPA, demonstrated that auxin and PAT are important for the rapid induction of hyponastic growth in this species, but not for the magnitude of the response and the maintenance of elevated petiole angles (Cox et al., 2004). This is in direct contrast to the auxin control of low light-induced hyponasty in Arabidopsis. Ethylene is the key trigger of submergence-induced hyponasty in both R. palustris and in Arabidopsis (Cox et al., 2003; Millenaar et al., 2005). Interestingly, treatment with TIBA did not obviously alter aspects of ethylene-induced hyponastic growth in Arabidopsis (Fig. S3). This suggests that in both species the magnitude and maintenance of hyponastic growth induced by ethylene are not under the control of PAT, whereas, in Arabidopsis, auxin and PAT are important for these hyponastic growth components upon exposure to low light intensity. This involvement of auxin and PAT is consistent with their roles in light quality-mediated elongation responses. Thus, auxin appears to be controlled by both light intensity and light quality and this regulation is functional in differential petiole growth (Fig. 5), as well as in the nondifferential elongation of petioles and hypocotyls (Morelli & Ruberti, 2000; Tao et al., 2008; Pierik et al., 2009). Auxin is known to be a key control factor in many developmental processes in plants. It thus appears that phenotypic adjustments to environmental signals, such as low light intensity-induced hyponasty, are at least partly mediated through modulation of these developmental programmes.
This work was supported by a PIONIER grant (800.84.470 to L.A.C.J.V.) and a VENI grant (no. 86306001 to R.P.) of the Netherlands organization for scientific research (NWO), and by a NIELS STENSEN grant to F.F.M. We thank Maarten Terlou, Yvonne E. M. de Jong-van Berkel and Joke van Elven for technical assistance, Diederik Keuskamp for comments on the manuscript and our colleagues G. E. Schaller, M. Koornneef, M. Ahmad, C. Lin, K. Apel, Nagatani, J. J. Casal, T. Sakai, J. Friml and M. Estelle who shared mutant lines.