Dynamic infrared imaging analysis of apical hook development in Arabidopsis: the case of brassinosteroids

Authors


Summary

  • Germination of Arabidopsis seeds in darkness induces apical hook development, based on a tightly regulated differential growth coordinated by a multiple hormone cross-talk. Here, we endeavoured to clarify the function of brassinosteroids (BRs) and cross-talk with ethylene in hook development.
  • An automated infrared imaging system was developed to study the kinetics of hook development in etiolated Arabidopsis seedlings. To ascertain the photomorphogenic control of hook opening, the system was equipped with an automatic light dimmer.
  • We demonstrate that ethylene and BRs are indispensable for hook formation and maintenance. Ethylene regulation of hook formation functions partly through BRs, with BR feedback inhibition of ethylene action. Conversely, BR-mediated extension of hook maintenance functions partly through ethylene. Furthermore, we revealed that a short light pulse is sufficient to induce rapid hook opening.
  • Our dynamic infrared imaging system allows high-resolution, kinetic imaging of up to 112 seedlings in a single experimental run. At this high throughput, it is ideally suited to rapidly gain insight in pathway networks. We demonstrate that BRs and ethylene cooperatively regulate apical hook development in a phase-dependent manner. Furthermore, we show that light is a predominant regulator of hook opening, inhibiting ethylene- and BR-mediated postponement of hook opening.

Introduction

Plants show a remarkably high capability to respond dynamically to various internal and environmental changes. Complex signalling pathways that integrate these endogenous and exogenous cues confer a high degree of developmental plasticity. Interactions between hormonal and environmental pathways optimize developmental patterns in a spatio-temporal manner (Casal et al., 2004; Vandenbussche et al., 2005).

Arabidopsis seedlings undergo skotomorphogenic development if seed germination occurs in complete darkness, as in case of below-ground germination (Neff et al., 2000). The development of an apical hook is a prerequisite for seedling survival (Harpham et al., 1991). It protects the cotyledons and the delicate shoot meristem when seedlings make their way through the soil (Darwin & Darwin, 1881; Guzmán & Ecker, 1990). Cotyledons remain unexpanded whereas the hypocotyl rapidly elongates to enable emergence from the soil (Gendreau et al., 1997). Under natural conditions, hook formation initiates shortly after germination. After complete formation the fully closed apical hook is maintained until exposure to light after seedling emergence from the soil (Raz & Ecker, 1999; Žádníková et al., 2010). The development of the apical hook depends on a tightly regulated differential growth driven by a multiple hormone cross-talk (Raz & Ecker, 1999; Li et al., 2004).

Genetic modification of auxin biosynthesis or exogenous administration of auxin results in defective apical hook formation (Schwark & Schierle, 1992; Boerjan et al., 1995; Celenza et al., 1995; King et al., 1995). Pharmacological treatments and mutations impairing auxin transport or signalling also result in a hookless phenotype (Lincoln et al., 1990; Lehman et al., 1996; Vandenbussche et al., 2010; Žádníková et al., 2010). Hence, an asymmetrical auxin distribution is indispensable for hook development. Exogenous treatment of etiolated Arabidopsis seedlings with ethylene, or its precursor 1-aminocyclopropane-1-carboxylic acid (ACC), exaggerates apical hook curvature (reviewed by Ecker, 1995). Apical hook formation results from the cross-talk between ethylene and auxin signalling. Ethylene activates the ethylene response-gene HOOKLESS1 (HLS1), implicated in the regulation of auxin distribution (Lehman et al., 1996). Auxin signalling functions downstream of ethylene signalling (Vandenbussche et al., 2010), with HLS1 as key link between both signalling pathways (An et al., 2012). Apart from auxin and ethylene, other plant hormones participate in the regulation of hook development. Enhanced cytokinin concentrations in amp1/hls2/cop2 hookless mutants suggest cytokinin involvement (Chaudhury et al., 1993; Hou et al., 1993). Inhibition of gibberellin (GA) biosynthesis or signalling also results in a hookless phenotype, indicating a role for GAs as well (Achard et al., 2003; Alabadí et al., 2004; Vriezen et al., 2004). GA control of hook development operates through transcriptional regulation of genes of ethylene and auxin signalling pathways (Achard et al., 2003; Vriezen et al., 2004; Gallego-Bartolomé et al., 2011). Both ethylene and GAs impinge on HLS1 expression through the EIN3/EIL1 transcription factors (An et al., 2012).

Several lines of evidence suggest brassinosteroid involvement in apical hook development. The constitutive photomorphogenic phenotype and absence of the characteristic hook in dark-grown brassinosteroid (BR) biosynthesis mutants constitutive photomorphogenesis and dwarfism (cpd), cabbage1/dwarf1 (cbb1/dwf1), deetiolated2 (det2) and brassinosteroid, light and sugar1 (bls1) (Chory et al., 1991; Kauschmann et al., 1996; Li et al., 1996; Szekeres et al., 1996; Laxmi et al., 2004), is an incentive to further study the function of BRs in hook development. Previous studies demonstrated the significance of BRs for apical hook formation (De Grauwe et al., 2005; Gendron et al., 2008). The exact function of BRs in the different phases of hook development and interaction with ethylene therein remains to be clarified.

Kinetic analysis of the dynamics of apical hook development allows dissection of the hormonal interplay which lies at the basis of the formation, maintenance and opening phases. Previous studies of Arabidopsis development were often based on end-point analyses (Binder, 2007). However, seedlings are highly sensitive to hormonal and environmental cues, resulting in a dynamic development suited for rapid adaptation to a changing environment. Hence, in end-point analyses a large quantity of dynamic information is lost (Degli Agosti et al., 2002; Binder, 2007; Miller et al., 2007). An in-depth analysis of plant development requires a method which comprises the spatial and temporal heterogeneity of plant organ differentiation (Basu et al., 2007). Kinetic image analysis has become a key tool in the study of plant development and represents a considerable improvement over previous techniques. High-resolution cameras, automation and modern computational technologies have optimized dynamic image analysis as a method for high-throughput, dynamic quantification of plant development (reviewed by Spalding & Miller, 2013). Infrared (IR) imaging enables accurate monitoring of etiolated seedling development, without stimuli of visible light (Parks & Spalding, 1999; Binder et al., 2004; Miller et al., 2007; Vandenbussche et al., 2010).

Here, we describe a platform for the analysis of dark-grown Arabidopsis seedling development that builds upon previous IR imaging systems (Parks & Spalding, 1999; Binder et al., 2004). Our platform substantially improves imaging throughput and resolution, enabling detailed analyses of Arabidopsis skotomorphogenic development. The high-quality time-lapse imaging allows video reconstruction of individual etiolated seedlings, providing a readily conceivable, detailed view of skotomorphogenic development. By integrating an automatic light-dimmer system, seedling development can be monitored automatically from germination in darkness over skotomorphogenesis to photomorphogenesis upon gradual exposure to light in a single experimental run. Our platform is characterized by a simple low-cost design, enabling a general usability without losing quality. We employed this platform to dynamically monitor apical hook development during skotomorphogenesis and hook opening during the transition from skoto- to photomorphogenesis in Arabidopsis thaliana and to unravel the role of brassinosteroids and their interaction with ethylene in the different phases of hook development. A reciprocal regulation of both hormonal pathways is demonstrated.

Materials and Methods

Plant material and growth conditions

Wild-type (WT) Arabidopsis thaliana (L.) Heynh. Columbia (Col-0), and det2-1 and cbb1/dwf1-6 mutants were obtained from the Nottingham Arabidopsis Stock Centre. WT Landsberg erecta (Ler-0) and the phyB-1 mutant in a Ler-0 background were obtained from the Arabidopsis Biological Resource Center (ABRC). With WT we refer to Col-0, unless stated elsewhere. All the experiments were conducted using half-strength Murashige and Skoog (MS) medium (Murashige & Skoog, 1962) consis-ting of 2.15 g l−1 MS salt mixture (Duchefa, Haarlem, the Netherlands), 8 g l−1 plant tissue culture agar (LABM, Bury, UK), supplemented with 1% sucrose (for plant tissue culture; Sigma-Aldrich); pH was adjusted to 5.7 using potassium-hydroxide (KOH) before adding the agar. The medium was autoclaved under high pressure and cooled down to ± 55°C before being poured into square Petri dishes (120 × 120 × 17 mm); 45 ml of medium was used per Petri dish. Epibrassinolide (EBR) (Sigma-Aldrich) was dissolved in ethanol, and 1-aminocyclopropane-1-carboxylic acid (ACC) (Sigma-Aldrich) was preserved as aqueous stock solution. Brassinazole (BRZ) (TCI, Tokyo, Japan) was stored as dimethylsulfoxide (DMSO) stock solution. Seeds were liquid sterilized using a sterilization solution consisting of 5% NaOCl (Bleach water; S.A. Delhaize Group N.V., Brussels, Belgium) and 0.05% Tween 20 solution (VWR International, Radnor, PA, USA), and washed several times with sterile distilled water. Then, seeds were manually sown using tweezers in two rows of 14 seeds in Petri dishes, which were sealed with a single layer of micropore tape (3M MICROPORE, 1.25 cm × 9.1 m; Novolab, Geraardsbergen, Belgium) to enable aeration of the seedlings and to impede bacterial or fungal contaminations. To synchronize germination, seeds were stored at 4°C for 60 h in darkness. After exposure to white light for 6 h to stimulate germination, seeds were transferred back to darkness at 22°C for the desired growth period (until all seedlings had open apical hooks).

Infrared imaging and automatic light dimmer system

The IR imaging experiment was conducted in a dark growth chamber at 22°C. Petri dishes were placed in a vertical position next to each other in a black box (600 × 400 × 330 mm; Fig. 1). The interior of the box was coated with a layer of matt black paint to reduce light reflection. Two 18-Megapixel (5184 × 3456) digital single-lens reflex (SLR) CCD cameras (EOS 550D; Canon Inc., Tokyo, Japan) were used for digital time-lapse imaging. The area of the CMOS APS-C sized image sensor is 22.3 × 14.9 mm and the physical pixel size is 18.5 μm2. The EF-S 18–55 mm (focal length) 1 : 3.5–5.6 IS II lens mount (Canon Inc.) was used to obtain images of the Arabidopsis seedlings. Two identical IR LEDS were used in our imaging system (IR LED illuminator 850 nm; ABUS security centre, Af?ng, Germany). The IR LEDS were mounted on the black box, in the front top left and right corners, illuminating the diagonal corner and thereby creating homogenous IR light. The cameras were IR converted (Advanced Camera Services, Watton, UK) enabling IR imaging at low ISO speeds and without the need for IR-filters. The manual focus mode and automatic stabilizer were used for image acquisition. Remote control of the digital time-lapse imaging was achieved using the DSLR Remote Pro 1.9.1 software (Breeze Systems, Camberley, UK). To enable optimal image acquisition, we carefully adjusted camera and image acquisition settings in anticipation of each experiment. Images were acquired every hour over the desired time period. The use of two sets of cameras enables digital time-lapse imaging of four Petri dishes in a single experimental run. The Petri dishes were aligned perpendicularly to the optical axis of the cameras (i.e. in vertical position). Some space was left in between the aligned Petri dishes, to enhance aeration and to reduce condensation. In order to prevent condensation from accumulating on the lid between the seedlings and the cameras, Petri dishes were inversely positioned, with their lids facing away from the camera. The two cameras were mounted on a height-adjustable jack to allow optimal positioning. The distance between the cameras and the aligned Petri dishes was modified such that each camera could simultaneously obtain digital images of two adjacent Petri dishes. The throughput can be increased by adding cameras or by using a linearized motor to move a camera along a series of Petri dishes.

Figure 1.

Set-up of the IR imaging system. Four Petri dishes with seeds were placed inverted and in vertical position next to each other in the imaging box. Two digital single-lens reflex (SLR) CCD cameras (EOS 550D) were used for digital time-lapse imaging. Two identical IR LEDS were used in our imaging system (IR LED illuminator 850 nm). Remote control of the digital time-lapse imaging was achieved using the DSLR Remote Pro 1.9.1 software.

In order to enable dynamic image analysis of light-induced apical hook opening during the transition from skoto- to photomorphogenesis an automatic light dimmer system was developed. A microcomputer-controlled automatic lamp dimmer (Model BIGv1, 220–240 V, AC 50 Hz, 2 A, 400 W; Europese Cultuurvogel Shop E.C.S., Oss, the Netherlands) was used to simulate a natural sunrise and sunset or gradual exposure to light when seedlings emerge from the soil. The automatic lamp dimmer allows linear (12-bit) dimming of the light with an adjustable fade time between 15 s and 2 h. The device was coupled to two frosted light bulbs with an intensity of 0.7 μmol m−2 s−1 (250 V, 40 W), mounted outside the black box, to reduce direct light reflection in the cameras. An analogue timer clock was used to remotely activate the automatic light dimmer system for the desired time. When light was applied, camera settings (AV, TV) in DSLR Remote Pro 1.9.1 were remotely adjusted using the TeamViewer software (TeamViewer GmbH, Göppingen, Germany). In a parallel experiment, light treatments were performed as follows. A white fluorescent light lamp with an intensity of 150 μmol m−2 s−1 (F40CW; Philips, Amsterdam, the Netherlands) was mounted inside the black box, above the cameras, to directly illuminate the Petri dishes with seedlings. The light duration was computer-controlled using Q Light Controller software (http://qlc.sourceforge.net/). Gradual light exposure (linear increase in light intensity from 0 to 0.7 μmol m−2 s−1 over 2 h) and light pulse treatments (10, 30 s and 1 min) were applied in the maintenance phase of apical hook development, when the majority of hypocotyl hooks were closed.

Kinetic analysis of apical hook development

Manual measurements were preferred over automatic measurements because they provide more insight in the dynamics of apical hook development, as well as in the individual variation between seedlings per treatment, and between treatments. ImageJ (NIH, Bethesda, MD, USA) was used to measure the angles of apical hook curvature. By convention, the angle of curvature α is defined as 180° minus the angle formed by the tangential of the apical part and the axis of the lower part of the hypocotyl (Supporting Information Fig. S1a,b). In the case of hook exaggeration, the angle of curvature is defined as 180° plus α (Fig. S1c,d; Vandenbussche et al., 2010). Apical hook development is considered to consist of three consecutive phases: formation, maintenance and opening. We adopted the definitions of the three phases, as postulated by Vandenbussche et al. (2010). The end of the hook formation phase was defined as the time point at which the angle of hook curvature reached 95% of its maximum value. The maintenance phase comprised the plateau in which hook angles differed at most 5% from the maximum angle of curvature. This was then succeeded by the hook opening phase. Time-point zero represents the time of germination, defined as radicle protrusion. The images were cropped per row of seedlings per Petri dish with Photolapse software (Softonic International S.L., Barcelona, Spain) to facilitate image processing in ImageJ. The cropped images were stacked in ImageJ to facilitate image analysis and hook measurements.

Statistics

Quantitative differences were analysed using student t-tests and deemed statistically significant for < 0.05. The mean number of biological samples is 20–30. Error bars represent the standard error of the mean.

Results

An improved dynamic infrared imaging system for kinetic studies of skoto morphogenesis

In order to dissect the hormonal interplay in apical hook development, we established a platform for dynamic imaging of etiolated Arabidopsis seedling development from germination onward. Our CCD-camera based system employs infrared (IR) light to enable imaging in darkness, whilst avoiding the induction of photomorphogenesis (Fig. S2). The system was first used to monitor the kinetics of apical hook development in etiolated seedlings both treated and untreated with ACC. In untreated etiolated seedlings hook formation was initiated shortly after germination until the curvature reached an angle of c. 170°, c. 36 h later, comprising the formation phase. During the maintenance phase the hook is kept closed for c. 24 h under standard conditions. The unfolding of the hook marks the end of the maintenance phase, and heralds the beginning of the opening phase, which is completed c. 150 h later (Fig. S3). These results corroborate previous studies (Binder, 2007; Vandenbussche et al., 2010; Žádníková et al., 2010). Two stages of opening could be discerned: a first stage of rapid hook opening initiated just after the maintenance phase (60 h) and continued until 108 h, followed by a slow opening stage. The mean slope of the rapid stage (−2.4093) was significantly different (< 0.0001) from the mean slope of the slow stage (−0.2556). Wild-type (WT)seedlings treated with ACC showed a significant exaggeration of apical hook curvature (< 0.0001) compared to untreated seedlings, which can be ascribed to an extension of the formation phase (with 12 h) at the beginning of the second day after germination (Fig. S3), in line with published data (Vandenbussche et al., 2010; Žádníková et al., 2010). Furthermore, the maintenance phase of ACC-treated seedlings (12 h, from 48 to 60 h) was shorter than in untreated seedlings (24 h, from 36 to 60 h). The kinetics of hook opening were, however, very similar, with a fast and slow opening stage. The results indicate that ethylene prolongs the formation phase and thereby induces an exaggeration of apical hook curvature. We conclude from this experiment that our dynamic imaging system is capable of rendering high-quality, IR images suitable for detailed studies of highly dynamic and micro-scale developmental processes, such as apical hook development (Fig. S4). Apart from the high quality, our platform also enables an unprecedented high-throughput IR imaging of 112 seedlings in a single experimental run.

Brassinosteroids are required for apical hook formation and prolong hook maintenance

In order to investigate the involvement of BRs in apical hook development, we first analysed the response to epibrassinolide (EBR), and brassinazole (BRZ), an inhibitor of BR biosynthesis (Asami et al., 2000), of WT Arabidopsis thaliana Col-0 ecotype seedlings. Seedlings treated with EBR showed an extension of the maintenance phase (48 h, from 36 to 84 h), compared to the untreated seedlings (24 h, from 36 to 60 h; Fig. 2a). By contrast, in BRZ-treated seedlings hook formation stopped prematurely and proceeded into hook opening without entering a conspicuous maintenance phase (Fig. 2a). Compared to untreated seedlings the maximum hook curvature was significantly less (= 0.0222) for BRZ-treated seedlings, but not influenced by EBR treatment (= 0.3120). Furthermore, the kinetics of hook formation and opening were not noticeably affected by EBR or BRZ treatment. Fast and slow stages of hook opening could again be distinguished (Fig. 2a). However, complete hook opening is reached later upon EBR treatment (c. 264 h) and earlier upon BRZ treatment (c. 168 h), compared to untreated seedlings (c. 216 h). Together, these results suggest that BRs regulate the transition between the maintenance and opening phase, prolonging the maintenance phase and thereby delaying hook opening. Moreover, these results also indicate that BRs are required to develop a completely closed apical hook.

Figure 2.

Brassinosteroid function and cross-talk with ethylene signalling in apical hook development in Arabidopsis thaliana. Kinetics of apical hook development in: (a) wild-type (WT) Col-0 seedlings in response to 100 nM EBR and 2 μM BRZ, with untreated WT seedlings as control. (b) Untreated cbb1/dwf1-6 and det2-1 seedlings, and in WT Col-0 seedlings in response to 2 μM BRZ, with untreated WT seedlings as control. The inset provides a comparison of a 1-h and 12-h interval time-lapse image analysis of apical hook development in det2-1 seedlings. (c) WT Col-0 seedlings in response to 1 μM ACC, 2 μM BRZ and the combination of 2 μM BRZ and 1 μM ACC, with untreated WT seedlings as control. Inset provides a comparison of a 1 h and 12 h interval time-lapse image analysis of apical hook development in WT seedlings upon 1 μM ACC and 2 μM BRZ treatment. (d) WT Col-0 seedlings in response to 1 μM ACC, 100 nM EBR and the combination of 100 nM EBR and 1 μM ACC, with untreated WT seedlings as control. (e) cbb1/dwf1-6 and det2-1 in response to 1 μM ACC and in WT Col-0 seedlings in response to the combined treatment with 1 μM ACC and 2 μM BRZ with untreated WT seedlings as control. (f) WT seedlings, and in ein2-5 in response to 1 μM ACC, 100 nM EBR, with untreated WT seedlings as control. Red dotted lines define the formation, maintenance and opening phase of apical hook development of Col-0 seedlings. The first line depicts the transition from formation (F) to maintenance phase (M). The second line depicts the transition from maintenance to opening phase. The third line depicts the transition from the fast (FO) to the slow (SO) stage of apical hook opening. Seedlings were photographed every 12 h (> 20). Error bars, ± SEM.

In order to support the above-mentioned results, apical hook development was examined in the BR biosynthesis mutants cbb1/dwf1-6 and det2-1. In the cbb1/dwf1-6 seedlings the maximum angle of hook curvature was similar to that in untreated WT seedlings (= 0.2515; Fig. 2b). This is a rather unexpected result, given the significantly lower maximum angle of hook curvature in BRZ-treated seedlings (= 0.0222). However, comparable to BRZ-treated seedlings, the cbb1/dwf1-6 seedlings were characterized by the absence of a maintenance phase (Fig. 2b). The formation phase also immediately proceeded into hook opening. However, in det2-1 seedlings, the formation phase was terminated prematurely. Hence, the maximum angle of curvature in det2-1 was significantly different from that in untreated WT seedlings (< 0.0001). In a 1-h interval time-lapse image analysis (Fig. 2b inset), a short maintenance phase of maximum 6 h was apparent. However, based on a careful analysis of individual det2-1 seedlings, this apparent maintenance phase is merely the result of an almost immediate hook opening after germination. Furthermore, this is in strong contrast to the 24 h maintenance phase of untreated WT seedlings. The kinetics of hook opening in the cbb1/dwf1-6 and det2-1 mutants were roughly similar to that of BRZ-treated WT seedlings, with a sequential fast and then slow opening stage. However, complete hook opening was reached even earlier (c. 96 h) in the det2-1 mutants compared to BRZ-treated seedlings (c. 168 h), whereas in the cbb1/dwf1-6 seedlings the apical hook fully opened later (c. 264 h). The kinetics of hook development in both BR biosynthesis mutants confirm that BRs postpone the transition between the maintenance phase and opening phases. Furthermore, these results emphasize the necessity of BRs for the development of a completely closed apical hook.

Ethylene-induced exaggeration of the apical hook curvature depends on brassinosteroids

The interaction of BRs with ethylene in apical hook development was examined by analysing the response of etiolated WT A. thaliana Col-0 seedlings to ACC and EBR or BRZ. Upon inhibition of BR biosynthesis with BRZ, ACC-treated seedlings developed no exaggerated apical hook (Fig. 2c). The kinetics of hook development were generally comparable with seedlings treated solely with BRZ. Apical hook formation stopped early and hence, no ACC effect on hook formation could be distinguished when BR biosynthesis was inhibited. The maximum angle of hook curvature was not significantly different (= 0.8095) from BRZ-treated seedlings. However, in contrast with BRZ-treated seedlings, the formation phase was followed by a short maintenance phase (12 h, from 24 to 36 h), confirmed by the 1-h interval time-lapse image analysis (Fig. 2c inset). In the absence of BRs, ACC slowed down hook opening, especially the fast opening stage, hampering the distinction between the two stages and resulting in a later hook opening (c. 240 h) compared to seedlings treated solely with BRZ (c. 168 h). These results indicate that the ethylene-induced hook exaggeration and shortening of the maintenance phase require normal BR biosynthesis (Fig. 2c). However, BRs fail to cause an exaggeration of the apical hook in the absence of exogenous ethylene (Fig. 2a).

When, by means of exogenous ACC, the concentration of ethylene in WT seedlings treated with EBR was enhanced, a clear effect was noticed (Fig. 2d). Seedlings developed a significant exaggeration of apical hook curvature compared to both untreated seedlings (< 0.0001) and seedlings treated solely with EBR (< 0.0001). However, the exaggeration was significantly less than in seedlings solely treated with ACC (= 0.0008). This apical hook exaggeration in the presence of EBR and ACC can again be ascribed to an extension of the formation phase (with 24 h, compared to EBR-treated seedlings), at the second day after germination. Furthermore, the maintenance phase was similar (24 h, from 60 to 84 h) to untreated seedlings (24 h, from 36 to 60 h) but reduced when compared to EBR-treated seedlings (48 h, from 36 to 84 h), as was the case for ACC-treated seedlings (Fig. 2d). In the presence of elevated concentrations of EBR, ACC accelerates hook opening (Fig. 2d), consistent with the deceleration of hook opening upon ACC administration in the absence of BR biosynthesis (Fig. 2c), resulting again in fading the distinction between a stage of fast and slow hook opening. Hence, despite the exaggeration of hook curvature, opening is completed at about the same time (c. 264 h) as seedlings treated solely with EBR. These results indicate that normal BR concentrations are required for the ethylene-induced exaggeration of apical hook curvature and that elevated concentrations of BR diminish the ACC-induced extension of the formation phase.

Support for the above-mentioned effect of BRs on ethylene-induced exaggeration of apical hook curvature was sought by examining the effect of ethylene on hook development in the cbb1/dwf1-6 and det2-1 mutants. As for BRZ-treated seedlings, cbb1/dwf1-6 mutants developed no exaggerated hook upon ACC application (Fig. 2e). Therefore, the maximum angle of hook curvature was significantly less (< 0.0001) than in seedlings treated solely with ACC. In ACC-treated cbb1/dwf1-6 seedlings, the formation phase was followed by a maintenance phase (24 h, from 36 to 60 h), unlike non-treated seedlings, which lacked a maintenance phase (Fig. 2b). Two stages of hook opening (fast and slow) were apparent. The failure of hook exaggeration upon ACC treatment was also observed in det2-1 seedlings (Fig. 2e). The formation phase was even more reduced compared to the cbb1/dwf1-6 seedlings, similar to WT seedlings treated with both ACC and BRZ (Fig. 2c). The maximum angle of hook curvature was significantly less (< 0.0001) than for ACC-treated cbb1/dwf1-6 seedlings. In ACC-treated det2-1 seedlings the formation phase was followed by a conspicuous maintenance phase (24 h, from 12 to 36 h), in contrast with untreated det2-1 seedlings (Fig. 2b). Apical hook opening also proceeded in distinct fast and slow stages. As was the case for the BRZ-treated seedlings, ACC slowed down the opening phase, resulting in a later opening of the hook (c. 132 h) compared to untreated det2-1 seedlings (c. 96 h). This was, however, not the case for the cbb1/dwf1-6 seedlings. These results confirm that the effect of ethylene on apical hook development requires BRs and corroborate the involvement of ethylene in prevention of hook opening, in the absence of BR biosynthesis.

Ethylene and brassinosteroid signalling are essential for normal apical hook development

In order to further investigate the relationship between BRs and ethylene in apical hook development, a kinetic analysis of the ethylene insensitive2-5 (ein2-5) mutant was conducted. The apical hook opened before full formation was achieved, without proceeding into a maintenance phase and irrespective of the presence of ACC (Fig. 2f). Therefore, ein2-5 seedlings both treated with ACC and untreated, differed significantly (< 0.001) from the untreated WT seedlings in the maximum angle of hook curvature. Again, the opening phases proceeded in a distinct fast and slow stage. These results confirm the indispensable function of ethylene in apical hook development. When treated with EBR, hook development was severely hampered. The formation phase of ein2-5 seedlings proceeded at a substantially slower rate compared to untreated ein2-5 seedlings and terminated even earlier (Fig. 2f). Therefore, the maximum angle of curvature was significantly smaller compared to untreated WT seedlings (< 0.0001). The incompletely closed apical hook started opening without entering a maintenance phase. However, given the large error bars associated with the first two measurements, these results should be interpreted cautiously. Inhibition of BR biosynthesis in the ein2-5 mutant, upon administration of BRZ, caused a dramatic disruption of early seedling development. Germination was impeded in the majority of ein2-5 seedlings. The seedlings that successfully germinated lacked an apical hook. Root and hypocotyl development were merely visible. Hence, these results suggest that together, ethylene and BRs are essential for normal seedling development.

Photomorphogenic control of apical hook opening

Our dynamic imaging system was equipped with an automatic light dimmer to kinetically study the effect of light on apical hook opening. The dimmer enables a gradual exposure of etiolated seedlings to white light, simulating the progressive exposure to sunlight when seedlings emerge from the soil. WT seedlings were grown in the dark and apical hook development was monitored using the above-described IR-imaging system. Seedlings remained in complete darkness until a fully closed apical hook was formed. Subsequently, the automatic light dimmer system was employed to gradually expose (linear increase in light intensity from 0 to 0.7 μmol m−2 s−1 over 2 h) the seedlings to light; to induce the transition from skoto- to photomorphogenesis. The results indicate that in > 80% of the seedlings the hook had completely opened in < 30 h after light exposure, whilst in dark-grown WT seedlings with a completely closed apical hook, this was only the case after > 150 h (Figs 3a, S3). The hook remained closed for 8 h after the exposure to light (lag-time), followed by rapid opening. Moreover, unlike in etiolated seedlings, no fast and slow stage of hook opening could be distinguished. Immediate exposure of WT seedlings to light with an intensity of 0.7 μmol m−2 s−1 resulted in largely similar results as upon gradual exposure (Fig. 3a), except that hook opening proceeded slightly more slowly. In > 80% of the seedlings the hook had completely opened in < 40 h after light exposure. However, the hook remained closed for 7 h after light exposure, again followed by rapid opening. These results indicate that light dominates hormones in apical hook development and that under natural conditions upon (low-intensity) light exposure a fast opening of the apical hook is advantageous for seedling growth and survival.

Figure 3.

Photomorphogenic control of apical hook opening in Arabidopsis thaliana. Kinetics of apical hook opening: (a) in darkness and upon gradual and immediate light exposure in wild-type (WT) Col-0 seedlings. (b) In darkness in WT Ler-0 and mutant phyb-1 seedlings. (c) In darkness and upon gradual light exposure in Ler-0 WT seedlings. (d) In darkness and upon gradual light exposure in phyb-1 seedlings. (e) In darkness and upon a light pulse of 10, 30 s and 1 min in WT Col-0. The green line and light bulb depict the time after germination upon which etiolated seedlings were exposed to light after apical hook formation in the dark. Apical hook curvature was measured every 12 h in darkness and every hour in light (> 15). In kinetic analyses (e) apical hook curvature was measured every hour in both light and darkness (> 10). Error bars, ± SEM.

The role of PHYB in apical hook opening

Given that phytochrome B (PHYB) is essential for de-etiolation upon exposure to white light (reviewed by Franklin & Quail, 2010), the role of PHYB in apical hook development was studied both in darkness and upon gradual exposure to white light using our automatic light dimmer. Therefore, the dynamics of apical hook development were analysed in etiolated phyB-1 mutant seedlings and compared with WT Ler-0 seedlings (Fig. 3b). In darkness, WT Ler-0 seedlings formed a fully closed apical hook that was maintained for c. 24 h (from 24 to 48 h), whereas in phyB-1 seedlings, hook formation stopped prematurely (with a maximal curvature of 145°) and proceeded in a shorter maintenance phase (12 h, from 24 to 36 h). Complete hook opening was reached earlier in phyB-1 (c. 144 h) compared to the Ler-0 WT seedlings (c. 168 h), which is largely attributed to the incomplete hook formation and shorter maintenance phase of the photoreceptor mutant. These results indicate an important role for PHYB in apical hook development in darkness. Upon gradual exposure to white light after apical hook formation in darkness, hook opening was completed in > 80% of the Ler-0 seedlings in < 30 h (Fig. 3c), comparable to WT Col-0 seedlings (Fig. 3a). Furthermore, a fast and slow stage of hook opening could be discerned. However, the hook immediately started to open and was not maintained for several hours after light exposure as in Col-0. In phyB-1 mutant seedlings, the hook also immediately started to open upon gradual light exposure (Fig. 3d). Hook opening was more irregular and in the second phase also slower in phyB-1 seedlings (mean slope = −3.2292) compared to Ler-0 WT seedlings (mean slope = −5.8666). Overall, despite the smaller maximum hook curvature of phyB-1 seedlings, hook opening was completed later (c. 40 h) than in Ler-0 WT seedlings (c. 30 h).

Photosensitivity of apical hook opening

In order to investigate the sensitivity of apical hook opening to light, three different light treatments were employed after hook formation was completed (Fig. 3e). Surprisingly, a brief light pulse of merely 10 s (fluence 1.5 mmol m−2) resulted in a noticeable acceleration of hook opening (mean slope = −1.0212) compared to etiolated seedlings (mean slope = −0.6170). Lengthening of the light pulse to 30 s (fluence 4.5 mmol m−2) resulted in a small but visible acceleration of hook opening (mean slope = −1.1690). Both the 10- and 30-s light treatments were characterized by a linear hook opening. When extended to 1 min (fluence 9 mmol m−2), the rate of hook opening was substantially increased (mean slope = −3.0597). However, 32 h after the light pulse, hook opening stagnated, resulting in an opening rate similar to that of etiolated seedlings (mean slope = −0.5746). Together, these results indicate that dark-grown seedlings are highly photosensitive, requiring a minimal light pulse for the onset of apical hook opening, but that the effectiveness of pulsed light treatment is limited, indicating the need for continuous light exposure.

Discussion

Dynamic IR imaging and automatic light dimmer system to study pathway networks in apical hook development

Plant development is a dynamic process, directed by multiple internal and environmental cues. Static end-point analyses are insufficient to dissect developmental mechanisms. Kinematic image analysis has therefore been introduced as a method to monitor plant development (Leister et al., 1999). Furthermore, it provides a solution for the extensive manual measurements and human interference, characteristic for previous methods, and hence, limits human error and inaccuracy (Arvidsson et al., 2011; Men et al., 2012). Dynamic imaging techniques to resolve open questions in plant development have markedly improved over the years. Current methods use high-resolution cameras and automation for high-throughput monitoring of development (Arvidsson et al., 2011; Sappl & Heisler, 2013; Spalding & Miller, 2013). Furthermore, insights into the regulation of plant development have increased due to the incorporation of mutants in dynamic imaging experiments (Arvidsson et al., 2011). However, the association of developmental changes with a mutation in the gene of interest often remains difficult, especially when only developmental progression is altered (Boyes et al., 2001). Dynamic imaging is an effective tool to study developmental gene function because it enables us to elucidate when and where deviations in development occur, and is therefore highly appropriate for dissecting the hormonal interplay which lies at the basis of the formation, maintenance and opening phases of apical hook development. Here, we described an imaging system that enables quantification of the dynamics of hook development of dark-grown seedlings under IR irradiation. Plants lack photoreceptors able to sense wavelengths longer than 750 nm (reviewed by Sullivan & Deng, 2003). Hence, the IR light (850 nm) used in our system cannot be perceived by plants and therefore simulates complete darkness. This was confirmed by the fact that there was no interference of IR light with apical hook development during skotomorphogenesis (Fig. S2).

Our imaging platform represents an improvement over previous systems to image dark-grown Arabidopsis seedlings under IR irradiation (Parks & Spalding, 1999; Binder et al., 2004, 2006). It is characterized by an uncomplicated and user- and budget-friendly design. All parts (including the automatic light dimmer) are publicly available and require no expensive technologies.

The use of two IR LEDs, illuminating opposite angles, creates an homogenous lighting environment and, together with the inverse positioning of the Petri dishes, minimalizes the reflection of the IR light during image acquisition. The vertical position of the Petri dishes against the back of the black box and IR illumination in line with the cameras further enhance the contrast between the seedlings and the background. In combination with the high-resolution cameras the latter adjustments further improve imaging resolution and throughput (Binder et al., 2004, 2006; Miller et al., 2007; Wang et al., 2009; Men et al., 2012). In Binder et al. (2004) such an IR imaging system was used to analyze the growth kinetics of Arabidopsis seedlings in response to ethylene. Quantification of this rapid response requires short-interval time-lapse imaging. Therefore, images were captured every 5 min. The rate of apical hook development is, however, substantially lower compared to the rapid growth response to ethylene. Therefore, based on the comparison of the kinetics in a 1-h and 12-h interval time-lapse image analysis (Fig. S5), a time interval of 12 h was chosen for the study of the kinetics of hook development. Thermal imaging is an alternative noninvasive imaging technique to study skotomorphogenesis (Belin et al., 2011). This technique is, however, ill-suited to study the dynamics of apical hook development. First, the pixel resolution of thermal images is at least 10 times lower than high-resolution CCD images (e.g. the system used in Belin et al., 2011, pixel resolution 320 × 240 vs 5184 × 3456 for the system presented here). This precludes the study of a small-scale structure such as the apical hook. Furthermore, high-resolution thermal cameras are much more expensive than high-resolution CCD cameras. Last but not least, thermal imaging prevents the adoption of sterile growth conditions, because it detects radiation emitted by an object, excluding the use of a lid. Vertical Petri dishes cannot be employed in the thermal imaging system. However, a vertical agar surface as used in our system, constraints seedling development in a 2D plane, facilitating focus on small structures such as the apical hook.

In order to dynamically study the transition to photomorphogenesis our IR imaging platform was equipped with an automatic light dimmer system. Previous studies exposed seedlings to continuous white light to initiate photomorphogenesis and study light-induced apical hook opening (Liscum & Hangarter, 1993; Miller et al., 2007; Alabadí et al., 2008). However, using an automatic light dimmer, the gradual exposure of emerging seedlings to light under natural conditions is more accurately reproduced. Upon adjustment of the fading time the photosensitivity of apical hook opening to light can be assessed. Furthermore, integration of the automatic light dimmer system reduces human intervention and stress exerted on the seedlings upon relocation into the light.

Through this system the whole developmental process from germination over skotomorphogenesis to photomorphogenesis upon gradual exposure to light, can be automatically monitored using a single experimental set-up. Our dynamic imaging system renders automated, high-quality images and hence, reduces human error and enables accurate observations and measurements of several aspects of skotomorphogenic development. The minimal human intervention and automation of image acquisition allow long-term studies. Furthermore, the spatiotemporal resolution of the images enables generation of high-resolution videos, providing a good overview of full skotomorphogenic seedling development and the involvement of hormones therein (Videos S1–S4). The use of two cameras supports digital imaging of 112 individuals within a single run, allowing multiple repeats under the same experimental conditions and improving the reproducibility of the conducted experiments. To the best of our knowledge, the platform described offers the highest throughput of currently available dynamic IR imaging methods. Here we show that this high-throughput and -resolution imaging system allows detailed studies of Arabidopsis skotomorphogenic development and dissection of the function of BRs and interaction with ethylene in the three phases of apical hook development.

Brassinosteroid function and cross-talk with ethylene signalling in apical hook development

Skotomorphogenic development is aimed at advancing the exposure to light when seedlings push through the soil in complete darkness. The apical hook is an unique structure that protects the cotyledons and shoot apical meristem from damage (Darwin & Darwin, 1881; Guzmán & Ecker, 1990). Upon exposure to light, photomorphogenic development is induced, the differential growth responsible for apical hook establishment is altered, and the hook opens (Liscum & Hangarter, 1993).

Plant hormones also contribute to a large extent to apical hook development, as hook formation, maintenance and opening are tightly regulated by multiple hormones in a complex cross-talk. The function of auxin, ethylene and gibberellins, and their interaction in the regulation of hook development have been extensively studied and their regulatory mechanisms postulated (reviewed by Abbas et al., 2013). However, the function of BRs and cross-talk with other plant hormones in apical hook development remains largely unclear. In plants BRs promote cell elongation, together with GAs, cytokinins and auxin (Mandava et al., 1981; Yopp et al., 1981; Katsumi, 1985; Clouse & Sasse, 1998). Multiple co-regulated genes have been found between BRs and GA, auxin and ethylene, three hormones also involved in the regulation of cell elongation (Nemhauser et al., 2004; Sun et al., 2010). Apical hook development relies on differential cell elongation and, hence, a function of BRs in hook development and interaction with these hormones can be expected. The constitutive photomorphogenic phenotype of BR biosynthesis mutants cpd, det2, cbb1/dwf1-6 and bls1 reinforces the presumptive function of BRs in apical hook development (Chory et al., 1991; Kauschmann et al., 1996; Li et al., 1996; Szekeres et al., 1996; Laxmi et al., 2004).

Our results underscore the essential role of BRs in hook development. Inhibition of BR biosynthesis upon BRZ administration prevents normal apical hook formation (Fig. 2a). This was confirmed by the kinetic analysis of the BR biosynthesis mutants det2-1 and cbb1/dwf1-6 (Fig. 2b). Nevertheless, exogenous BRs fail to induce an exaggerated hook in the absence of exogenous ACC (Fig. 2a). Conversely, exogenous ACC fails to prolong the formation phase and induce an exaggeration of hook curvature in the absence of BR biosynthesis upon BRZ administration, as expected given the above-mentioned BR biosynthesis-dependent hook formation (Fig. 2c). This was again confirmed in the BR biosynthesis mutants det2-1 and cbb1/dwf1-6 (Fig. 2e). Consistent with previous research cbb1/dwf1-6 seedlings (Takahashi et al., 1995; Kauschmann et al., 1996; Klahre et al., 1998; Müssig et al., 2002) were characterized by a weaker phenotype (Fig. 2b,e) in comparison with other BR-deficient mutants, such as cpd (Kauschmann et al., 1996; Szekeres et al., 1996) and det2 (Chory et al., 1991). Incompletely abolished concentrations of endogenous BRs and partial insensitivity to BRs have been postulated as explanations for the weaker phenotype of the cbb1/dwf1-6 mutant (Kauschmann et al., 1996). Nevertheless, based on our results and previous studies (De Grauwe et al., 2005; Gendron et al., 2008) we conclude that BR biosynthesis is indispensable during the formation phase and for the ethylene-induced exaggeration of apical hook curvature. Hence, ethylene probably controls hook formation partly through BR signalling (Gendron et al., 2008). However, exogenous BRs restrain the exaggeration of hook curvature induced by ACC (Fig. 2d). Hence, the ethylene effect on hook formation is inhibited by BRs, suggesting an inhibition of ethylene biosynthesis or signalling. Although stimulation of ethylene biosynthesis by BRs was observed in etiolated Arabidopsis seedlings (Woeste et al., 1999; Hansen et al., 2009), the EBR effect may result from a time-window dependent induction of a specific ACO (ACO2, Deng et al., 2007) that is only expressed or active in the formation phase. Enhanced biosynthesis is indeed not necessarily correlated with an increased physiological response. Alternatively, the EBR effect may impinge on ethylene signalling rather than biosynthesis. Consistent with previous studies (Harpham et al., 1991; Vandenbussche et al., 2010; Žádníková et al., 2010), ethylene signalling proved necessary for normal apical hook formation (Fig. 2f). The more severely disturbed hook formation upon administration of exogenous BRs agrees with the study of Vandenbussche et al. (2013), where ethylene insensitivity caused a loss of gravitropism and the treatment with exogenous BRs aggravated the effect of deficient ethylene signalling. Inhibition of BR biosynthesis in the absence of ethylene signalling substantially impaired germination and early seedling development, underscoring the vital importance of both hormones, not only for apical hook formation but also for skotomorphogenic development in general. Apart from the formation phase, BRs also appear to be implicated in hook maintenance, as inhibition of BR biosynthesis substantially reduces this phase (Fig. 2a). BR biosynthesis mutants det2-1 and cbb1/dwf1-6 corroborate the latter result (Fig. 2b). Therefore, BRs seem to directly regulate the transition from the maintenance to the opening phase. The latter was confirmed by the prolonged maintenance phase upon administration of exogenous BRs (Fig. 2a). Apart from BRs, other hormones are also involved in the regulation of hook opening. In the absence of ethylene signalling, apical hook opening follows the formation phase without proceeding into the maintenance phase (Fig. 2f). Furthermore, in the absence of BR biosynthesis, elevated concentrations of ethylene cause a prolongation of the maintenance phase (Fig. 2c,e). This indicates that ethylene prevents apical hook opening; however, BRs seem to inhibit this ethylene effect. Hence, BR regulation of hook maintenance probably functions partly through ethylene signalling. How BRs interact with ethylene and other hormones, such as gibberellins, in the prevention of apical hook opening remains to be determined (Vandenbussche et al., 2010; Gallego-Bartolomé et al., 2011).

Photomorphogenic control of apical hook opening

Plants rely on light for growth and development and have evolved a set of photosensory systems to detect light and to adapt their development accordingly (Kendrick & Kronenberg, 1994; Deng & Quail, 1999; Neff et al., 2000). Seedlings adopt skotomorphogenic development after below-ground germination, in which resource allocation is directed to hypocotyl elongation, to hasten seedling emergence from the soil and exposure to light, at the expense of root and cotyledon development (Gendreau et al., 1997; Josse & Halliday, 2008). Dark-grown seedlings are characterized by unexpanded cotyledons in the apical hook and nonphotosynthetic etioplasts (Kendrick & Kronenberg, 1994). Upon exposure to light it is of utmost importantance for a young plant to become photosynthetically active as soon as possible. This is supported by the rapid, light-induced opening of the apical hook, which enables fast cotyledon expansion and the production of photosynthetically active chloroplasts, compared to the slow skotomorphogenic opening (Fig. 3a). However, hook opening was substantially slower compared to previous studies in Arabidopsis (Liscum & Hangarter, 1993; light intensity 50 μmol m−2 s−1) and in sunflower (Rabe and Kutschera, 1999; light intensity 100 μmol m−2 s−1) which could be due to the considerably lower intensity of white light used in the present study. Also, the lag-time for hook opening reported here is substantially longer. Neither finding can be ascribed to the gradual light exposure. The rate of apical hook opening upon immediate light exposure was comparable (if not slightly slower) to that under gradual light exposure (Fig. 3a). The same goes for the lag-time of hook opening after light exposure. Hence, the most plausible explanation for the aberrant dynamics of hook opening compared to previous studies is the low-light intensity used in this study. In conclusion, light of low intensity is sufficient to trigger apical hook unfolding. Moreover, merely a short pulse of light (10–30 s; fluence 1.5–4.5 mmol m−2) hastens hook opening (Fig. 3e). An extended exposure to light (1 min; fluence 9 mmol m−2) substantially accelerates hook opening. However, hook opening stagnates without continuous light, resulting in an incompletely opened hook (Fig. 3e). The latter is possibly a defence mechanism to prevent the cotyledons and shoot apical meristem from being damaged upon premature hook opening.

In dark-grown seedlings, skotomorphogenic development is maintained through the repression of the photomorphogenic program. The phytochrome interacting factor (PIF) transcription factors act as constitutive repressors of photomorphogenesis in the dark. De-etiolation depends on the coordinated action of the phytochrome and cryptochrome photoreceptors and is indispensable for seedling survival (Fankhauser, 2001; Wang, 2005). Upon light exposure, PIFs are rapidly degraded, enabling the transition from skoto- to photomorphogenesis (Leivar et al., 2008). Intranuclear binding with phytochromes leads to PIF degradation (Bauer et al., 2004; Shen et al., 2005, 2007, 2008; Al-Sady et al., 2006). Phytochrome B is the most important phytochrome photoreceptor regulating de-etiolation upon white light exposure (reviewed by Franklin & Quail, 2010). The irregular and slower apical hook opening upon gradual light exposure supports the importance of PHYB for the induction of photomorphogenesis (Fig. 3d). However, our results imply that PHYB also has an effect in darkness. Etiolated phyB-1 mutant seedlings were characterized by limited hook formation (Fig. 3b), which is in agreement with the more pronounced hook unfolding in phyB in darkness in the end-point analysis of Zheng et al. (2013). Overall, the present data support a role for PHYB in sustaining skotomorphogenesis in the dark. A PHYB activity in darkness is furthermore supported by PHYB regulation of gene expression (Mazzella et al., 2005).

Model of ethylene–BR interaction in apical hook development

A model on the interaction between ethylene, BRs and light in apical hook development is presented in Fig. 4. BRs and ethylene are indispensable for and closely interact in apical hook development. However, this hormonal interplay is highly phase dependent. Both ethylene and BR signalling are indispensable for normal apical hook formation. Ethylene partly acts through BRs to regulate hook formation. BRs also inhibit ethylene action in hook formation. Both BRs and ethylene are indispensable in preserving the maintenance phase of apical hook development and, hence, keeping the hook closed. BRs inhibit the ethylene-induced postponement of hook opening. Hence, BRs partly act through ethylene to regulate apical hook maintenance. Light dominates in the regulation of apical hook development and inhibits BRs, ethylene and GAs to induce hook opening.

Figure 4.

Model of the interaction between ethylene and brassinosteroids (BRs) in apical hook development in Arabidopsis thaliana. (T) indicates an inhibitory interaction, (↑) indicates a stimulatory interaction. F, formation; M, maintenance; O, opening phase. The interaction between ethylene and BRs depends on the phase of apical hook development. BRs and ethylene are indispensable both for the formation of a completely closed apical hook and for exaggeration of apical hook curvature. Ethylene regulation of apical hook formation functions partly through BRs. Ethylene action is feedback-inhibited by BRs. BRs and ethylene are also required to preserve the maintenance phase and, hence, to keep the apical hook closed. BR regulation of the maintenance phase partly functions through ethylene, with the ethylene-mediated postponement of hook opening inhibited by BRs. Light plays a dominant role in apical development, through PHYB, and inhibits BRs and ethylene to trigger apical hook opening.

In conclusion, the high-throughput dynamic imaging system described here allows a rapid mapping of hormone and light interactions in seedling development. Future research will focus on combined automated length measurements of hypocotyls and roots, and on genetic and molecular studies to support the interactions detected by in vivo imaging.

Acknowledgements

D.S. gratefully acknowledges Yuming Hu (Laboratory of Functional Plant Biology, Ghent University) for introducing him to the imaging methodology. This work was supported by Ghent University and the Research Foundation Flanders (grant G065613N) to D.V.D.S.; by a post-doctoral fellowship of the Research Foundation Flanders to F.V.; by the European Research Council with a Starting Independent Research grant (ERC-2007-Stg-207362-HCPO) and the project CZ.1.07/2.3.00/20.0043 (to Central European Institute of Technology, CEITEC) to P.Z. and E.B.

Ancillary

Advertisement