Environmental and developmental effects on the biosynthesis of UV-B screening pigments in Scots pine (Pinus sylvestris L.) needles

ABSTRACT

The major UV-B screening pigments of the epidermal layer of Scots pine (Pinus sylvestris) needles are flavonol 3-o-glycosides (F3Gs) esterified with hydroxycinnamic acids at positions 3′′ and 6′′. Acylation is the last step in biosynthesis and is catalysed by position-specific hydroxycinnamoyl transferases (3′′ and 6′′HCT). The UV-B dependence of these enzyme activities was studied in primary needles of Scots pine seedlings grown under different UV-B conditions in environmentally controlled sun simulators. 6′′HCT activity was induced upon UV-B irradiation while 3′′HCT activity was not induced but showed high constitutive values. To investigate the biosynthesis of diacylated F3Gs during needle development under natural conditions, the HCT activities and metabolite contents were analysed in needles of field-grown mature pine trees. Accumulation of diacylated compounds as well as of 6′′HCT activity occurred transiently in the first year of needle development only. In contrast, 3′′HCT activity exhibited broad maxima in two consecutive years during needle growth. The data suggest that acylated F3Gs are first formed as soluble compounds which are then translocated into the cell wall to be bound by their hydroxycinnamoyl residues.

Abbreviations

F3G

flavonol 3-O-glycoside

HCT

hydroxycinnamoyl-CoA flavonol 3-O-glucoside hydroxycinnamoyltransferase (EC 2.3.1.-)

K3G

kaempferol 3-O-glucoside

UV-B

ultraviolet B radiation (280–315 nm).

INTRODUCTION

UV-B has strong effects on plants (Frohnmeyer & Staiger 2003), for example, on DNA through the dimerization of thymidin residues, oxidation of membrane lipids and proteins (Jansen, Gaba & Greenberg 1998) and genome stability by a so far unknown mechanism (Ries et al. 2000). The effects of UV-B alone and in combination with other climatic factors on whole terrestrial ecosystems, and possible impacts of UV-B on trees and forests in particular have recently been summarized (Caldwell et al. 2003; Sullivan 2005). To avoid UV-B dependent damages, plants have developed efficient protection and repair mechanisms (Britt 1996; Greenberg et al. 1997). A frequent response is the production of UV-B screening pigments and the deposition of these metabolites in the epidermal cell layer, consequently protecting the photosynthetically active mesophyll (Reuber, Bornman & Weissenböck 1996; Chalker-Scott 1999; Bieza & Lois 2001). These metabolites are mostly flavonoids and/or hydroxycinnamic acids, which absorb in the ultraviolet (UV) range of the solar spectrum. In Scots pine and Norway spruce, the F3Gs doubly acylated by hydroxycinnamic acids are the main UV-B screening pigments (Schnitzler et al. 1996; Fischbach et al. 1999) with p-coumaric acid at position 3′′ and p-coumaric or ferulic acid at position 6′′ of the molecule. Acylation with hydroxycinnamic acids efficiently contributes to the absorption in the UV-B range. It has been demonstrated that these compounds are formed de novo upon UV-B irradiation of the plants, and that they are localized in the epidermal cell layer (Schnitzler et al. 1996, 1997; Hutzler et al. 1998). The highly protective function of the epidermal cell layer of coniferous needles has been indicated by model calculation (Schnitzler et al. 1996) and has been demonstrated by direct microfibre-optic measurements (DeLucia, Day & Vogelman 1992; Day 1993). Interestingly, diacylated F3Gs not only occur in conifers but also in some broadleaf trees, like three oak species (Romussi, Parodi & Caviglioli 1991) and Eryobotrya japonica (Kawahara, Satake & Goda 2002). This suggests that these metabolites may play an important role in UV-B screening in a variety of economically important tree species. Compounds of this structure have also been detected in leaves of the tropical climbing fern Stenochlaena palustris (Liu et al. 1999). In Scots pine so far, the F3Gs of the three basic structures, kaempferol, isorhamnetin and quercetin were found. Their biosynthesis require the key enzymes phenylalanine ammonia-lyase (PAL) and chalcone synthase (CHS) (Forkmann & Heller 1999). Surprisingly, these enzymes were not affected by UV-B, at least at high photosynthetically active radiation (PAR) conditions, suggesting that UV-B regulation would occur at a later step of the biosynthetic pathway (Schnitzler et al. 1996).

We recently characterized the acyltransferases acting on F3Gs (Kaffarnik et al. 2005). These HCTs catalyse the last two steps in the biosynthesis of diacylated F3Gs. These exhibit position specificity towards the hydroxyl groups at the sugar residue in a strictly sequential manner, acylating first hydroxyl group 6′′ with p-coumaric or ferulic acid (6′′HCT) followed by hydroxyl group 3′′ with p-coumaric acid (3′′HCT).

The purpose of this study was to analyse the UV-B regulation of these enzymes in Scots pine needles. We used seedlings grown both in the presence and absence of UV-B in sun simulators under realistic environmental conditions, and measured HCT activities and product accumulation in primary needles. In addition, we studied the biosynthesis of UV-B screening pigments during needle development in field-grown, mature pine trees in order to gain further insight into the formation of the UV-B screen under natural conditions.

MATERIALS AND METHODS

Sun simulator experiments

Scots pine seedlings were grown and pre-cultivated essentially as described by Schnitzler et al. (1997) at 130 µmol m⁻² s⁻² PAR (daily dose of 6.4 mol m⁻²) and 0.5 W m² UV-A irradiance. The UV-B emitted from the standard cool-white fluorescence lamps was removed [biological effective UV-B weighting (UV-B_B.E.) < 3 mW m⁻²] by 5-mm UV-absorbing Plexiglas, Type 20070 sheets (Degussa, Darmstadt, Germany). Some 4 weeks after germination, the seedlings were transferred to a sun simulator at the following climatic parameters: 14-h light period (9-step curve with maximum intensity for 1 h around noon), temperature gradient between 18 and 22 °C and relative humidity gradient between 60 and 80%. Within a 2-week acclimation period, the light intensity was increased in three steps of 2 d with daily doses of 11, 22 and 35 mol m⁻², and a final step at the seventh day to 45 mol m⁻² at a maximal irradiance of 1700 µmol m⁻² s⁻² PAR irradiance for one h at noon. During acclimation, UV-B was excluded by standard float glass (thickness of 8 mm). The UV-B treatment was started by replacing float glass with Sanalux (thickness of 4 mm; Desag, Grünenplan, Germany) in the +UV-B variant. The details of the sun simulator set-up are described elsewhere (Döhring et al. 1996; Thiel et al. 1996). In brief, a combination of the following lamp types were used: metal halide lamps (Osram HQI/D 400 W, München, Germany), quartz halogen lamps (Osram Halostar 300 and 500 W), blue fluorescent tubes (Philips TLD 18, 36 W, Eindhoven, the Netherlands) and UV-B fluorescent tubes (Philips TL 12 40 W). Excessive infrared radiation was absorbed by a water layer of approximately 20 mm thickness. The irradiation parameters were determined with a spectroradiometer (DTM 300 double monochromator; Bentham, Reading, UK) (Thiel et al. 1996). The calibration of spectral irradiance is traceable to the Physikalische-Technische Bundesanstalt (PTB, Braunschweig, Germany), the national institute for science and technology. The irradiation values are listed in Table 1.

Table 1.  Irradiation conditions in the sun simulator experiment

Parameter	Treatment (+UV-B)	Control (–UV-B)
The daily doses (fluences) of the main irradiation parameters are listed. The biological effective UV-B weighting (UV-BB.E.) was calculated according to Caldwell (1971) and normalized at 300 nm. The error margins for the radiometric data were in the range of ± 10% (Thiel et al. 1996).
Photosynthetic photon fluence (mol m^−2)	46.0	45.6
UV-A(315−400 nm) fluence (kJ m^−2)	1060	712
UV-B(280−315 nm) fluence (kJ m^−2)	28.000	< 0.001
Biological effective UV-B fluence (UV-BB.E.; kJ m^−2)	6.200	< 0.001

The seedlings were harvested at the beginning of the acclimation period, at time 0 and 12, 24 and 36 h, and 2, 3, 4, 6, 10 and 13 d after onset of UV-B treatment. Additional control samples were harvested at day 0, 7 and 14 from seedlings grown under pre-cultivation conditions. At any time point, three randomly chosen pots each containing approximately 200 seedlings were harvested separately to avoid fringe effects and inhomogeneities of the chamber. The cotyledons and primary needles were separated, immediately frozen in liquid nitrogen and stored at −80 °C until further use. Mean values and standard deviations were calculated from the three parallel samples per time point and treatment.

Sampling of field-grown trees

The needles were collected between 1997 and 1999 at the GSF-National Research Center for Environment and Health (GSF) campus from two 15- and 40-year-old field-grown trees, respectively. Both trees produced male and female cones in all 3 years. The sampling was performed weekly or biweekly between April and September, and every 4 weeks between October and March. The needles of three recent year classes were separately harvested from one twig per time point and cut as a whole. The samples were immediately frozen in liquid N₂ and stored at −80 °C until further use.

Field radiation measurements

Outdoor light and UV radiation was measured with spectrally integrating sensors. For the PAR range, a LI 190 sensor (Li-Cor Biosciences, Lincoln, NE, USA) and for the UV range a PMA2101 SUV detector (Solar Light, Glenside, PA, USA) were used. The UV sensor had a spectral characteristic which approximates the erythemal action spectrum. Although there is no exact procedure to convert erythemally weighted data into other UV-B units, we derived an approximation factor of 1.62 for conversion of MED to kJ m⁻² from frequent spectral outdoor measurements to estimate UV-B flux values integrated over the 280–315 nm band.

Reference substances and substrates

Kaempferol, quercetin and isorhamnetin 3-O-glucosides, as well as 6′′-p-coumaroyl-kaempferol 3-O-glucoside (tiliroside) were from Extrasynthèse (Lyon, France). Preparation of CoA-esters of p-coumaric and ferulic acids was performed as described elsewhere (Kaffarnik et al. 2005). Other chemicals used were of highest available purity and were purchased from Sigma (Steinheim, Germany).

Protein extraction

Approximately 100 mg of needle material from seedlings or pine trees was homogenized with a pestle and mortar in 1.5 mL extraction buffer [100 mm sodium phosphate, 10.0% (w/v) sucrose, 1.5% (w/v) polyethylene glycol (PEG) 1450, 5 mm 1,4-dithioerythritol (DTE), pH 6.8] with 50 mg polyrinylpolypyrrolidone (PVPP) and 3 mg Celite (Sigma, Steinheim, Germany) on ice. After two centrifugations (20 000 × g at 4 °C, 5 min each) the supernatant was desalted on a NAP-5 column (Amersham Biosciences, Freiburg, Germany) according to the manufacturer’s instructions using the extraction buffer. Protein concentration was measured by Coomassie Brilliant Blue (Serva, Heidelberg, Germany) staining (Bradford 1976) using bovine serum albumin (BSA) as standard.

Enzyme assays

Enzyme assays were performed in a total volume of 212 µL with 200 µL of protein extract, adjusted to protein concentrations between 50 and 100 µg/mL before addition, and 6 µL each of hydroxycinnamoyl-CoA (3.5 mm in H₂O) and F3G (3.5 mM in methanol) at final concentrations of 0.1 mm. The reaction was started by the addition of the acceptor substrate, the assay incubated at 37 °C for 60 min and the reaction stopped by the addition of 1 nmol pinosylvin methyl ether which served as an internal standard. The products were then extracted with two portions of 200 µL ethyl acetate, the organic phases pooled and dried under a stream of N₂ at room temperature. The residue was redissolved in 80 µL 50% (v/v) acetonitrile in H₂O, cleared by centrifugation at 20 000 × g for 5 min and analysed by high-performance liquid chromatography (HPLC).

HPLC analyses

Analysis of enzyme products

HPLC separation was performed according to Kaffarnik et al. (2005). A 250 × 4.6 mm Spherisorb ODS2 5.0 µm column (Bischoff, Leonberg, Germany) was run for 3 min with solvent A [1.9% (v/v) formic acid, 0.1% (w/v) ammonium formate in water] followed by a gradient for 7 min to 35% solvent B [1.9% (v/v) formic acid, 0.1% (w/v) ammonium formate, 9.6% (v/v) water in acetonitrile], 7 min to 44% B, 5 min to 79% B and 1 min to 100% B, detection was performed at 314 nm.

Quantification of soluble and cell wall-bound phenolics from needles

Analysis of soluble phenolic compounds was performed by reversed-phase (RP)-HPLC as described elsewhere (Schnitzler et al. 1996, 1997; Turunen et al. 1999). Cell wall-bound phenolics were determined according to Strack et al. (1988) with the following modifications: residues of soluble metabolite analysis from approximately 30 mg plant material were exhaustively washed with methanol and then incubated with 150 µL 1 m NaOH, 100 mm ascorbate and 53 mm sodium borohydride overnight at room temperature in the dark. After acidification with 150 µL 1.5 m formic acid and addition of 300 µL methanol, the mixture was centrifuged at 9000 × g for 10 min, and 20 µL of the clear supernatant was analysed with RP-HPLC as described for soluble metabolites.

Calculation of smoothing curves

Smoothing curves including the overall data given in Figs 1 and 2 were calculated using the non-parametric local regression (LOESS) procedure of the SAS Version 8.2 L (SAS Institute GmbH, Heidelberg, Germany). The fitting procedure was performed with linear functions and smoothing parameters of 0.15 for the curve in Fig. 1a and 0.30 for the curves in the remaining figures. The case numbers were n = 114 in Fig 1a, n = 113 in Fig. 2b and n = 97 in the remaining figures.

Figure 1. Seasonal course of HCT activities and content of diacylated F3Gs in field-grown adult Scots pine trees. Samples from three needle age classes from a mature pine tree growing at the GSF-National Research Center for Environment and Health (GSF) site were collected from 1997 to 1999 (needle age class: 1997, circles; 1998, squares; 1999, triangles) and tested for 6′′HCT activity (a), 3′′HCT activity (b) and content of diacylated F3Gs (c). Closed symbols in (a) and (b) represent data obtained with isorhamnetin 3-O-glucoside (I3G), open symbols in (a) with K3G, in (b) with tiliroside as the flavonol substrates, in all cases coumaroyl CoA was used as the donor substrate. Similar measurements were performed on a second tree with comparable results. The curves represent predicted values computed by a non-parametric smoothing procedure (LOESS). x axis: 0 = 1 January of the year of needle emergence.

Figure 2. Seasonal course of cell wall-bound compounds and soluble diacylated K3G in field-grown adult Scots pine trees. As in Fig. 1, samples from three needle age classes (1997, circles; 1998, squares; 1999, triangles) from a mature field-grown tree were analysed. The sum of methanol-soluble, diacylated kaempferol derivatives (a) consisting of approximately 90% 3′′,6′′-di-p-coumaroyl kaempferol and 10% 3′′-p-coumaroyl-6′′-feruloyl K3Gs, are shown. The kaempferol derivatives are the main diacylated F3Gs in Scots pine needles (see Fig. 1c). Treatment of the methanol-washed residue with NaOH released wall-bound F3Gs and hydroxycinnamic acids. The major F3G was K3G (b). The total content of kaempferol derivatives is also shown (c). Only minor amounts of quercetin 3-O-glucoside and no isorhamnetin 3-O-glucoside were found. The main hydroxycinnamic acid released by alkaline hydrolysis was p-coumaric acid (d), but minor amounts of ferulic acid were also detected (approximately 5% of total hydroxycinnamic acids). Similar measurements were performed on a second tree with comparable results. The curves represent predicted values computed by a non-parametric smoothing procedure (LOESS).

RESULTS

Response to UV-B of 3″- and 6″HCT activities and diacylated F3Gs in Scots pine seedlings

To investigate whether enzymatic activities of 3′′ and 6′′HCT activities are controlled by UV-B, the Scots pine seedlings were studied in a sun simulator under defined irradiation and controlled climate conditions (Table 1). The seedlings were germinated and grown under exclusion of UV-B (final UV-B_B.E. < 0.001 kJ m⁻² d⁻¹) and increasing PAR levels from 6.4 to 45.0 mol m⁻² d⁻¹ between the age of 4 and 6 weeks. They were then exposed to ambient UV-B (28 kJ m⁻², UV-B_B.E. of 6.2 kJ m⁻² d⁻¹). HCT activities as well as contents of diacylated F3Gs were measured at regular intervals before and after UV-B treatment (Fig. 3). The specific activity of 6′′HCT increased 3-fold during the change from low PAR at pre-cultivation to the highest PAR condition finally applied in the sun simulator from 1 to 3 µkat kg⁻¹ protein. UV-B was excluded during this period with UV-B-absorbing Plexiglas (Degussa, Darmstadt, Germany) at pre-cultivation, and with float glass sheets in the sun simulator. Subsequent UV-B irradiation of the seedlings caused an additional 2.7-fold induction of 6′′HCT activity within 2 d of up to 8 µkat kg⁻¹ protein (Fig. 3a). After a total of 13 d, the activity returned to nearly control levels indicating the UV-B acclimation of the plants. In seedlings grown at high PAR conditions under continuous UV-B exclusion, no change of 6′′HCT activity was apparent. In control seedlings at low PAR, a slight increase of activity from 1.1 to 1.7 µkat kg⁻¹ protein was observed within the same time period. In contrast to 6′′HCT the 3′′HCT activity did not significantly differ between presence and absence of UV-B at high PAR. The levels measured ranged continuously around 12 µkat kg⁻¹ protein, and exceeded the maximal 6′′HCT activity about 1.5-fold (Fig. 3b). At low PAR, the activity ranged between 6 and 10 µkat kg⁻¹ protein throughout the cultivation period, suggesting a relatively high constitutive activity level, which responded to PAR and/or UV-A but not to UV-B.

Figure 3. HCT activities and content of diacylated F3Gs in primary needles of Scots pine seedlings during exposure to different light conditions. Scots pine seedlings were germinated and pre-cultivated for 4 weeks under UV-B exclusion at 130 µmol m⁻² s⁻² photosynthetically active radiation (PAR). After transfer to the sun simulator, the seedlings were acclimated for another 2 weeks to increasing light intensities until finally reaching 1700 µmol m⁻² s⁻² without UV-B. At time zero of UV-B treatment, the exposure was started by replacing float glass by Sanalux in the +UV-B condition, and the plants were grown for 14 d at a daily dose of 28.000 kJ m⁻² and < 0.001 kJ m⁻² UV-B in the +UV-B (▴) and the –UV-B treatment (▪), respectively. Additional control seedlings were grown at pre-cultivation conditions (○). Samples were tested for 6′′HCT activity with K3G (a), and 3′′HCT activity with tiliroside (b) as the flavonol substrates and coumaroyl CoA as the donor substrate. The content of diacylated F3Gs (c) was also determined. Mean values ± SD of three randomly sampled pools are shown. In (a) and (c), the differences between –UV-B and +UV-B were significant with P ≤ 0.01 using paired Student’s t-test except at timepoints marked (ns). SD, standard deviation.

In line with the transient UV-B induction of 6′′HCT activity, diacylated F3Gs accumulated from 1.0 to 2.5 µmol g⁻¹ fw within 4 d (Fig. 3c). During the following 9 d, a further minor increase to 3 µmol g⁻¹ fw was observed. Under UV-B exclusion, these metabolites only slightly accumulated from 1.0 to 1.5 and 1.9 µmol g⁻¹ fw at low and high PAR conditions, respectively, reflecting low activities of 6′′HCT (Fig. 3c). Taken together, these results show that 6′′HCT activity depends on UV-B while 3′′HCT activity is constitutively high.

Developmental dependence of HCT activities and diacylated F3Gs in needles of mature Scots pine trees

In order to study the role of HCTs in the biosynthesis of UV-B screening pigments during needle development of Scots pine under field conditions, emerging needles and 1-year-old specimens were collected simultaneously throughout for 3 years from mature field-grown trees. These needles were tested for 3′′- and 6′′HCT activities and contents of diacylated F3Gs (Fig. 1).

The activity of 6′′HCT was high in young needles during needle emergence and the early developmental stages in May and June (Fig. 1a). Later, from the middle of August, and in 1-year-old needles, no 6′′HCT activity was detected (Fig. 1a). In contrast, 3′′HCT activity was present throughout the first year of needle development showing a broad maximum in summer (June and July), and lower but measurable values in the following winter (Fig. 1b). In the second year, the activity increased again in April with a distinct maximum in May (Fig. 1b). A slight increase above background values was even observed in spring of the third year (data not shown).

The content of diacylated F3Gs increased sharply during early needle development in line with high activities of the HCTs (Fig. 1c). Highest values of up to 3–4 µmol g⁻¹ fw were reached at the end of June of the first year and then declined during the following months, consistent with the disappearance of 6′′HCT activity. No increase of these metabolites was observed in the second year, and values below 1 µmol g⁻¹ fw were finally measured at the end of the second year.

Radiation conditions in the field

The findings, that 6′′HCT was induced by UV-B in seedlings, but was active only during early needle development in the field, raised the question about the factors responsible for 6′′HCT induction in the field. To compare 6′′HCT activities with radiation conditions at the natural site, radiation measurements on the GSF site were conducted. As an example, the annual course of 1999 for daily doses averaged over 7 d of PAR and UV-B integrated over the 280–315 nm range is shown in Fig. 4. Values increased during April and May and were highest between June and August reaching up to 44 mol m⁻² and 26 kJ m⁻² for PAR and UV-B, respectively. Non-averaged data for single days were up to 50 mol m⁻² and 30 kJ m⁻² for PAR and UV-B, respectively. In September and October, the daily doses decreased, and reached minimum values of 3.4 mol m⁻² of PAR and 1.4 kJ m⁻² of UV-B in winter. These data demonstrate that the highest 6′′HCT activities in developing needles coincide with high radiation conditions in the field.

Figure 4. Radiation conditions at the GSF-National Research Center for Environment and Health (GSF) site (48.23°N, 11.6°E, 495 m above sea level) during 1999. The annual course for 1999 of UV-B (280–315 nm, solid line) and PAR (dashed line) measured with a SUV detector or a LI 190 sensor are shown, respectively. Light intensities were measured automatically every minute. For clarity, 7 day doses calculated from the original data are shown. For details, see Materials and Methods.

Annual course of wall-bound metabolites in needles of Scots pine trees

The rapid decrease of diacylated F3G levels during needle maturation of Scots pine needles (Fig. 1c) suggested the further metabolism of these compounds. To test whether diacylated F3Gs might be translocated and bound to the cell wall, methanol-insoluble cell wall-conjugated flavonol derivatives and hydroxycinnamic acids were analysed. Because only minor amounts of quercetin and no isorhamnetin 3-O-glycosides were detected after alkali treatment, we focused our study on kaempferol derivatives. Figure 2a and b show the annual course of diacylated and cell wall-bound kaempferol derivatives, respectively. Although kaempferol-based derivatives were only about 50% of total diacylated F3Gs, the annual course closely followed the course of the sum of F3Gs (compare Fig. 2a with Fig. 1c). The levels steeply increased during early needle development, reached highest values of approximately 2.0 µmol g⁻¹ fw in June and then decreased continuously. In contrast, wall-bound K3G accumulated more gradually up to a level of 2.5–3.0 µmol g⁻¹ fw in October of the first year, and further increased in the second year up to 4.0 µmol g⁻¹ fw (Fig. 2b), suggesting a translocation in both first and second year of needle development and ageing. The sum of diacylated and cell wall-bound K3G was calculated to indicate the course of total content of the UV-B-absorbing F3Gs (Fig. 2c). As expected, the curve sharply increased during early needle development, reached values close to 4 µmol g⁻¹ fw and slightly increased again in the second year. The p-coumaric and ferulic acids were the only hydroxycinnamic acids detected in the hydrolysates, where p-coumaric acid was the major component with up to 1.7 µmol g⁻¹ fw (Fig. 2d). Ferulic acid was also detected at concentrations not higher than 0.15 µmol g⁻¹ fw (not shown).

Detailed analysis of the soluble, diacylated F3Gs, and wall-bound K3G and hydroxycinnamic acids by model calculation showed that the fraction of wall-bound metabolites still increased considerably when 6′′HCT activity was no more detectable. Furthermore, the decrease of the diacylated metabolites could not account for the incorporated F3G fraction (not shown). The data therefore suggested that metabolites other than the diacylated F3Gs, for example, 3′′-mono-acylated compounds, may be incorporated into the cell wall which would be in line with the fact that 3′′HCT is high throughout needle development.

DISCUSSION

In previous experiments with Scots pine seedlings, it had been shown that UV-B irradiation under controlled environmental conditions led to an increase of diacylated F3G levels in primary needles (Schnitzler et al. 1996, 1997). Although the de novo synthesis of these flavonoid metabolites was detected under these conditions, induction of both PAL and CHS – key enzymes of flavonoid biosynthesis, have not been observed (Schnitzler et al. 1997). In the present study, we therefore analysed the UV-B-dependent regulation of HCTs responsible for the final steps of the pathway in more detail. 6′′HCT was not only induced by UV-B, but also by high PAR and UV-A to some extent. Surprisingly, 3′′HCT was not at all induced by UV-B but showed high activities at all treatments, although monoacylated compounds never accumulated during the experiment. This indicates an additional function of 3′′HCT besides its contribution to the biosynthesis of diacylated compounds. Moreover, 3′′HCT activity exceeded that of 6′′HCT throughout the UV-induction experiment and would therefore not be a limiting factor in diacylated F3G formation.

To analyse HCT activities and formation of UV-B screening pigments during needle development under natural conditions, we chose field-grown trees in their natural environment. This is especially important, because information is scarce about the mechanisms that mediate plant responses to ambient, solar UV-B (Ballaré 2003). Here, we observed again differences between 3′′ and 6′′HCT activities. While 6′′HCT activity was detectable only during the first 3 months of needle development, broad activity peaks of 3′′HCT were detectable during the first 2 growth periods and some activity even in winter months. These profiles indicate a basically different regulation of the two enzymes, where 6′′HCT activity is light-inducible but only in combination with the early stages of needle development. However, it has to be taken into account that increasing concentrations of newly formed F3Gs may increasingly shield the needle tissue from UV-B, thus regulating 6′′HCT activity and consequently the biosynthesis of diacylated products. Based on enzyme kinetic data for the two HCTs, we had proposed a sequential acylation of F3Gs first at position 6′′ followed by position 3′′ (Kaffarnik et al. 2005). Therefore, a tight control of 6′′HCT activity would result in the limited formation of the diacylated metabolites.

Regulation of enzymes of the phenylpropanoid metabolism by external (i.e. radiation) as well as internal (i.e. developmental) factors has often been observed (Dixon & Paiva 1995; Weisshaar & Jenkins 1998) and has earlier been studied in detail, for example, in parsley cell cultures (Heller et al. 1979). Relationship between developmental stage and induction of flavonoid biosynthesis by UV-B has also been demonstrated in other plant species. Rye primary leaves responded most sensitively within 4 d after emerging from the coleoptile (Reuber et al. 1996; Burchard, Bilger & Weissenböck 2000), and UV-B-dependent flavonoid accumulation in Arabidopsis was restricted to the early stages of leaf development (Lois 1994). HCT activities, involved in phytoalexin biosynthesis, were also development-dependent in various systems, for example, wheat coleoptiles (Louis & Negrel 1991) and Equisetum sporophytes (Hohlfeld, Veit & Strack 1996).

Continuous increase of cell wall-bound metabolite levels throughout the 2 years of needle development studied concomitant with the early transient maximum of the soluble, diacylated F3Gs (Figs 1 & 2) suggests that soluble metabolites are translocated and covalently bound to the cell wall. Similar observations exist for Norway spruce and other Pinaceae species (Strack et al. 1989; Heilemann & Strack 1990; Schnitzler et al. 1996; Fischbach et al. 1999) as well as some Mediterranean evergreen broad-leaved trees (Liakoura, Manetas & Karabourniotis 2001). Based on fluorescence microscopic studies with trichomes of Quercus ilex and Olea europaea, it had been suggested that polyphenols are translocated from the perinuclear and cytoplasmatic space to cell walls during leaf ageing (Karabourniotis et al. 1998). The possible involvement of diacylated F3Gs in the process of cell-wall deposition has been earlier indicated but was not further analysed (Strack et al. 1989; Heilemann & Strack 1990; Schnitzler et al. 1996; Fischbach et al. 1999). The fact that total cell wall-bound hydroxycinnamic acids released upon alkaline hydrolysis are only about half the molar amount of F3Gs released is in favour of a model where acylated intermediates are covalently bound to the cell wall via their hydroxycinnamic acid residues.

Considering radiation conditions in the sun simulator and those in the field, daily PAR doses of about 46 mol m⁻² measured in the simulator (Table 1) and 7-day-average values between 30 and 40 mol m⁻² in May and June measured in the field were comparable (Fig. 4). The respective UV-B doses were 28 kJ m⁻² (Table 1), and between 16 and 24 kJ m⁻² (Fig. 4). On some days, daily values of up to 50 mol m⁻² PAR and 30 kJ m⁻² UV-B were observed in the field. Therefore, the radiation conditions used in our sun simulator experiment with Scots pine seedlings are in good agreement with those measured in the field. Our data therefore demonstrate that sun simulators can be successfully used in the study of the UV-B-dependent accumulation of plant secondary metabolites. Comparable data obtained from primary needles of Scots pine seedlings in the sun simulator experiments and from secondary needles of mature trees in the field also show that the seedlings are useful model systems. Application of seedlings furthermore allows to study pools of large numbers of individuals, thus reducing the usually considerable variation observed between different individuals of mature trees.

In summary, we have shown that one of the two HCTs, namely 6′′HCT, involved in the biosynthesis of UV-B screening pigments in Scots pine needles only occurs in early stages of needle development and is regulated by UV-B. On the other hand, 3′′HCT activity does not respond to UV-B but shows high constitutive activity. The overall role of this enzyme still remains elusive. However, the results of metabolite analysis indicate that its monoacylated products, which do not accumulate in soluble form, may be directly transferred into the cell wall. Furthermore, the decrease of diacylated F3Gs coincides with increase of cell wall-bound compounds in needles of field-grown trees. This suggests a translocation of these pigments into the cell wall, where they will act as a persistent UV-B screen.

ACKNOWLEDGMENTS

We thank Susanne Stich for her excellent technical assistance. We also thank Jörg-Peter Schnitzler and Thomas Vogt for providing helpful comments on a draft of the manuscript.