|Home | About | Journals | Submit | Contact Us | Français|
Volatile organic compounds (VOCs) are expected to have an important role in plant adaptation to high temperatures. The impacts of increasing night-time temperature on daytime terpenoid emissions and related gene expression in silver birch (Betula pendula) and European aspen (Populus tremula) clones were studied. The plants were grown under five different night-time temperatures (6, 10, 14, 18, and 22°C) while daytime temperature was kept at a constant 22°C. VOC emissions were collected during the daytime and analysed by gas chromatography–mass spectrometry (GC-MS). In birch, emissions per leaf area of the C11 homoterpene 4,8-dimethy1-nona-1,3,7-triene (DMNT) and several sesquiterpenes were consistently increased with increasing night-time temperature. Total sesquiterpene (SQT) emissions showed an increase at higher temperatures. In aspen, emissions of DMNT and β-ocimene increased from 6°C to 14°C, while several other monoterpenes and the SQTs (Z,E)-α-farnesene and (E,E)-α-farnesene increased up to 18°C. Total monoterpene and sesquiterpene emission peaked at 18°C, whereas isoprene emissions decreased at 22°C. Leaf area increased across the temperature range of 6–22°C by 32% in birch and by 59% in aspen. Specific leaf area (SLA) was also increased in both species. The genetic regulation of VOC emissions seems to be very complex, as indicated by several inverse relationships between emission profiles and expression of several regulatory genes (DXR, DXS, and IPP). The study indicates that increasing night temperature may strongly affect the quantity and quality of daytime VOC emissions of northern deciduous trees.
Most plant species have the ability to synthesize and release volatile organic compounds (VOCs) (Owen and Peñuelas, 2005; Vickers et al., 2009). The precursors of terpenoids are dimethylallyl diphosphate (DMAPP) and isopentenyl diphosphate (IPP) (Schwarz, 1994), which are synthesized by the methylerythritol 4-phosphate (MEP) pathway [i.e. the deoxyxylulose-5-phosphate (DXP) or non-mevalonate pathway] in chloroplasts and by the mevalonate (MVA) pathway in the cytoplasm (Rodríguez-Concepción and Boronat, 2002; Rodríguez-Concepción et al., 2004; Owen and Peñuelas, 2005). These pathways are independent, but interact through metabolic cross-talk (Hemmerlin et al., 2003; Rodríguez-Concepción et al., 2004). Monoterpenes in non-storing plants and isoprene (C5) in isoprene-producing plants are volatilized after production and rapidly lost (Niinemets et al., 2004). Semi-volatile sesquiterpenes (SQTs) are easily adsorbed on vegetation surfaces such as leaves and bark, and are later observed by above-canopy flux measurement techniques after re-release from foliage (Helmig et al., 2004). There could also be delayed release of SQTs from surfaces in the morning (Schaub et al., 2010).
The biological function of many volatiles is still poorly understood, although several experiments have indicated clear functions for some of these volatiles within a plant (Holopainen, 2004). The effect of biotic interactions on VOC emissions from plants and long-distance signalling in plant defence has been studied intensively during the last two decades (Gouinguene and Turlings, 2002; Vallat et al., 2005; Gershenzon, 2007; Heil and Ton, 2008). Studies of VOC emission response to abiotic factors have traditionally focused on isoprene (Sharkey et al., 2008), but in recent years increasing research concerning C6, C10, and C15 volatiles has been published (for reviews, see Laothawornkitkul et al., 2009; Vickers et al., 2009; Yuan et al., 2009). The ecophysiological functions of VOCs include plant–plant communication and plant–herbivore interactions (De Moraes et al., 2001; Ibrahim et al., 2005; Blande et al., 2007) as well as attraction of natural enemies to their herbivore prey (Holopainen, 2004; Vuorinen et al., 2004; Mumm et al., 2008). Abiotic factors such as temperature, light, and the availability of carbon for the synthesis of VOCs are known to control the short-term emission of isoprenoids (Grote and Niinemets, 2008; Sharkey et al., 2008). The emission of stored monoterpene (MTs) in ducts, glands, or trichomes is controlled mainly by temperature and/or light, while the emission of isoprene, which is not stored, is also clearly light and temperature dependent (Staudt and Bertin, 1998; Kesselmeier and Staudt, 1999; Dindorf et al., 2006; Loreto et al., 2007). However, recent studies have also shown that SQT emission increases exponentially with temperature (Tarvainen et al., 2005; Helmig et al., 2007). In addition, rising temperature can also affect the quality of volatile emissions from plants (Gouinguene and Turlings, 2002). There is increasing evidence that VOCs, especially isoprene, provide plants with protection against high temperature (Rennenberg and Schnitzler, 2002; Velikova and Loreto, 2005; Peñuelas and Munne-Bosch, 2005; Sharkey et al., 2008; Wiberley et al., 2008). In addition to thermotolerance, isoprene emissions provide tolerance to ozone and other reactive oxygen species (Vickers et al., 2009) and may function as a ‘safety valve’ to remove unwanted metabolites and dissipate excess energy in high light conditions (Rosenstiel et al., 2004). Besides this, biogenic VOCs play a very important role in atmospheric chemistry as they participate in the formation of ozone in NOx-polluted atmospheres and quench ozone in unpolluted environments (Lerdau and Slobodkin, 2002).
HMGR (hydroxymethylglutaryl CoA reductase), DXS (1-deoxy-D-xylulose 5-phosphate synthase), and DXR (1-deoxy-D-xylulose 5-phosphate reductoisomerase) are important enzymes in producing the first intermediates specific to cytosolic MVA and plastidial MEP pathways. IPP isomerase mediates the conversion of IPP (produced in the mevalonate pathway) to DMAPP, which is the precursor required for isoprenoids (Hoeffler et al., 2002), while isoprene synthase (IspS) produces isoprene from DMAPP through the plastidial MEP pathway (Schwender et al., 1997). The function of several enzymes and related genes has been documented for both MVA and MEP pathways (McGarvey and Croteau, 1995; Rodriguez-Concepción and Boronat 2002; Sasaki et al., 2007). For example, the expression of genes encoding DXS and DXR that act on the MEP pathway produces DMAPP, which is the substrate for IspS which produces isoprene (Sasaki et al., 2007). Both IspS activity and isoprene emissions were increased by heating in transgenic Arabidopsis, providing evidence that one of the physiological functions of isoprene emission could be the protection of the plants from heat stress (Sasaki et al., 2007).
For this study, two predominant species in the northern environment, silver birch (Betula pendula) and European aspen (Populus tremula), were selected for their different emission profiles. Betula pendula is an MT emitter and P. tremula is an MT and isoprene emitter (Hakola et al., 1998), while both species emit SQTs (Zhang et al., 1999). So far, very little is known about intraspecific variation in VOC emissions for these species (Vuorinen et al., 2005), although large variation in several stress adaptations (e.g. against oxidative stress) and leaf metabolism has been reported in birches and aspens (Häikiö et al., 2007; Oksanen et al., 2007).
The aim of this study was to investigate the influence of night-time warming on the VOC emission rates of birch (B. pendula) and aspen (P. tremula), originating from areas with low night temperatures. Current global warming has increased surface humidity in the Northern hemisphere (Dai, 2006), which may alleviate the minimum night temperatures. An attempt was also made to identify some of the genetic trends associated with VOC emission rates by comparing expression levels of key genes from the plastidic MEP pathway and the cytoplasmic MVA pathway. In addition to temperature treatment comparisons, differences in emission profiles and related gene expression between the genotypes were studied within these species. It is hypothesized that increasing night-time temperature stimulates the biosynthesis of plant VOCs through activation of the above-mentioned key genes, and this activation is reflected in the daytime emissions of these compounds. Furthermore, it is hypothesized that the ratio of specific VOCs in the emission profiles could be altered and that there is variation among the genotypes in warming responses in both species.
Birch (B. pendula Roth) and aspen (P. tremula L.) seedlings from clones (Bp8, Bp17, Bp26, and Pt2.2, Pt5.2, Pt6.0, respectively) were used in this study. The clones were randomly selected from naturally regenerated birch and mixed forests from south-eastern Finland, and represent natural genotypic variation within populations. The birch material was micropropagated from three genotypes growing in a stand in Punkaharju (61°48′N, 29°18′E), used for intensive biodiversity studies. Plant material for aspen was selected from three different locations in central Finland. Since aspen individuals that grow close to each other might well be of the same genotype (Suvanto and Latva-Karjanmaa, 2005), the material was selected from distant locations, but within the same latitude to ensure comparable climatic conditions between the locations.
The experiment started 1.5 months after the planting of micropropagated plantlets in 2006. Plantlets (20–40cm in height at the beginning of the experiment) were grown at varying night-time temperatures from 13 October to 24 November. One genotype of each species (Bp26 and Pt2.2) was selected to study the daytime volatile emissions. To compare genotypes of both species, all clones were grown at a night-time temperature of 14°C.
The plants were planted into plastic pots (Ø 15cm) containing peat and vermiculite (1:1) and grown in controlled-environment plant growth chambers (Conviron PGW36, Controlled Environments Limited, Winnipeg, Manitoba, Canada) for 6 weeks. Five different night-time temperatures from 6°C to 22°C were used with intervals of 4°C (Fig. 1). The daytime temperature was constant (22°C) in all treatments between 10:00 am and 4:00 pm. Temperatures were set to decrease in the evening between 4:00 pm and 9:00 pm, stabilizing at each of the night-time temperatures. In the morning, temperatures rose from 5:00 am to 10:00 am back to the daytime temperature. Relative humidity was a constant 60% in all treatments. Light was increased in steps during 2h in the morning (between 4:00 am and 6:00 am) and decreased similarly between 8:00 pm and 10:00 pm. Light was provided by fluorescent lamps and light bulbs. Daytime photosynthetically active radiation (PAR) was ~300μmol m−2 s−1.
To avoid the consequences of possible differences between the individual chambers, the seedlings were moved weekly from one chamber to another, where the growth conditions were re-programmed. In addition, the positions of the pots inside the chambers were changed twice a week. Seedlings were watered daily, and fertilizer (Kekkilä, Superex nro5, N:P:K 11:4:25) was added weekly in solutions approximating a total N input of 35kg ha−1 year−1.
The tops of the main shoot with young and mature leaves (~26 for birch) and (~13 for aspen) were enclosed inside pre-cleaned (120°C) Multi-Purpose Cooking Bags (Polyethylene terephthalate, 25×55cm in size, Look, Terinex Ltd, UK) (Ibrahim et al., 2008). Before inserting the seedling into the bag, one of the top corners of the bag was cut with scissors and an air flow of 750–800ml min−1 (Stewart-Jones and Poppy, 2006; Ibrahim et al., 2008) was passed through a charcoal filter and used to flush out residual contamination for 15–20min. Thereafter, the test seedling and the Teflon tube were inserted into the bag through its mouth and the bag was tightened around the stem with gardening wire. After enclosure of the foliage, the bags were flushed for a further 15min (Mäntyla et al., 2008; Ibrahim et al., 2010). The air flow was reduced to 300ml min−1 before the start of sampling. The sampling tube was inserted into the bag through the top corner. The filtered replacement air with ambient CO2 concentration was scrubbed with MnO2 to remove ozone and pumped into the bag.
The volatile emissions from the enclosed seedlings were collected into steel sample tubes filled with a 50%:50% combination of Tenax TA and Carbopack B (100mg of each mesh 60/80, Supelco, Bellofonte, PA, USA) adsorbents. The sampling air flow was 200ml min−1. The sampling time was 1h and all samples were collected in a time period between 9:30 am and 4:30 pm with replicates from all the treatments distributed evenly across this period. Compounds were thermally desorbed at 250°C for 10min and cryofocused in a cold trap at –30°C. The compounds were injected onto an HP-5MS capillary column (length 50m, id 0.2mm, film thickness 0.33μm, J&W Scientific USA, Agilent Technologies). The column temperature was held first at 40°C for 1min, increasing to 210°C at a rate of 5°C min−1 and rising to a final temperature of 250°C at a rate of 20°C min−1. The carrier gas was helium with constant pressure of 20.7psi. The samples were analysed by gas chromatography–mass spectrometry (GC-MS; Hewlett Packard GC 6890, MSD 5973). The spectra of external standards for available compounds and the Wiley library (John Wiley & Sons, Ltd, Chichester, UK) were used to identify the adsorbed compounds (Ibrahim et al., 2008; Himanen et al., 2009). Commercially available reference substances were used to quantify the emissions. The concentrations of the MTs α-terpinene, α-terpinolene, and (E)-β-ocimene were quantified assuming that the responses were the same as that of the MT α-pinene in the standard (Vuorinen et al., 2007; Himanen et al., 2009). The SQTs β-bourbonene, 3,7-guaiadiene, (Z,E)-α-farnesene, α-amorphene, and (E,E)-α-farnesene were quantified using the response of the SQT α-humulene in the standard (Himanen et al., 2009).
After each volatile collection, gas exchange was measured with a Li-Cor 6400 photosynthesis system (LI-COR Biosciences, Lincoln, NE, USA). Photosynthesis was measured from one mature leaf of each seedling used for VOC collections. The temperature in the cuvette was 22°C and a light-emitting diode (LED) light source was used to provide light at the level of PAR 400μmol m−2 s−1. CO2 concentration was 360ppm and air flow through the cuvette was a constant 500μmol s−1. After measurements, the leaves used for VOC collection were scanned and their areas were analysed (Adobe Photoshop Elements 3.0, Adobe Systems Inc., San Jose, CA, USA). Leaves were oven dried (70°C for 2d) for determination of dry weights. Four seedlings from each genotype and treatment were used to obtain the other growth data. Leaf areas of the whole seedlings were estimated based on area analyses of every second leaf of the seedling, taking into account a mass relationship of analysed and non-analysed areas. Specific leaf area (SLA; leaf area per unit dry weight) was calculated for the whole seedling.
For gene expression analysis, samples were taken the day after VOC measurements from three Pt2.2 and three Bp26 seedlings growing under each of the five temperature treatments. Genotypic comparison was made of three genotypes grown at a night-time temperature of 14°C. The fifth and eighth leaves from the top of the seedling were collected, frozen in liquid nitrogen, and stored in a deep freezer (–70°C) until RNA extractions. The mRNA levels were analysed using quantitative real-time PCR (qRT-PCR). Total RNA was isolated from the leaf samples using the cetyltrimethylammonium bromide (CTAB) method (Chang et al., 1993). The DNase I-treated total RNA was reverse transcribed using an oligo(dT18) primer and DyNAmo 2 step SYBR Green qPCR kit (Finnzymes). QRT-PCRs were carried out using a Dynamo HS SYBR Green kit (Finnzymes) in a 20μl reaction volume with 0.5μM gene-specific primers and 2μl of diluted cDNA, deriving from 5–15ng of total RNA, as a template.
The quantitative PCR primers (Table 5) were designed from P. tremula sequences obtained in this study by amplifying with primers designed from an alignment of Populus expressed sequence tags (ESTs) that had homology to P. alba×P. tremula ispS (AJ294819), P. trichocarpa DXS (EU693019), Arabidopsis isopentenyl-diphosphate delta-isomerase (IPP1; At5g16440), and Arabidopsis 3-hydroxy-3-methylglutaryl COA reductase (HMG1, At1g76490). DXR primers were designed from P. tremula ESTs that have a high homology to P. trichocarpa DXR (EU693020) as well as to Arabidopsis DXR (At5g62790). The quantitative PCR primers for B. pendula DXS, DXR, and HMGR were designed using a silver birch EST library (Aalto and Palva, 2006). The Q-PCR amplicons were sequenced.
α-Tubulin was chosen as a reference gene in aspen and in birch (GI: 6723477, GenBank). The amplicon lengths and the primer sequences are given in Table 5. The reactions were performed in triplicate in an iCycler iQ Real-time PCR (Bio-Rad). The PCR conditions were: 95°C for 15min, followed by 35 cycles of 95°C for 15s, 57°C for 20s, and 72°C for 20s. After final annealing (72°C, 5min) and redenaturation (95°C, 1min), a melt curve analysis was done by decreasing the temperature from 95°C to 60°C at 0.5°C intervals. The fold change in gene expression was calculated using the comparative Ct method (2–ΔΔCt) (Livak and Schmittgen, 2001).
Statistical analyses were performed using the statistical package SPSS for Windows (14.0 and 15.0). Analysis of variance (ANOVA) followed by Tukey's multiple comparisons was used for the analysis of normally distributed volatile compounds. Compounds which were not normally distributed were analysed with the non-parametric Kruskal–Wallis test followed by Mann–Whitney test. Leaf growth and net photosynthesis data were also analysed by ANOVA/Tukey, and linear regressions for leaf area and SLA were calculated with Microsoft Office Excel 2007.
In birch, the emission of the homoterpene DMNT increased significantly with increasing night temperature, while the monoterpene β-ocimene peaked at 18°C. The emission of SQTs β-bourbonene, (E,E)-α-farnesene, and γ-cadinene increased significantly (Table 1) with elevated night-time temperature, while the emission of α-copaene, α-humulene, δ-cadinene, (E)-caryophyllene, and α-amorphene increased clearly at night temperatures of 18–22°C compared with lower temperatures (Table 1). Among the non-terpenes (NTs), emissions of the green leaf volatile (GLV) (Z)-3-hexenyl acetate was significantly higher at 22°C than at lower (6–14°C) night temperatures (Table 1). Emissions of the MTs α-pinene, β-pinene, β-myrcene, limonene, 1,8-cineole, and α-terpinolene differed between the treatments, but did not show any correlation with night temperature (Table 1). Total SQTs were significantly increased at higher night temperatures (Table 1). Results from the three birch clones indicated that there were no significant differences between the clones in terms of individual compounds; however, clone Bp8 emitted higher total MTs, while Bp17 emitted higher total NTs than the other two clones (Table 2).
In aspen, the isoprene emission did not differ between the night temperature treatments (Table 3). Emissions of the homoterpene DMNT and the MT β-ocimene showed constant increases from 6°C to 14°C, while the MT β-pinene and the MT alcohol linalool increased from 6°C to 18°C. The SQT (Z,E)-α-farnesene was emitted at a significantly higher rate at 18°C while the SQT (E,E)-α-farnesene showed consistent increases from 6°C to 18°C. Among the NTs, emissions of the GLVs (Z)-3-hexenyl acetate, (Z)-3-hexenyl isovalerate, and (Z)-3-hexenyl butyrate were increased consistently with night temperatures of 6°C to 14°C (Table 3). The total SQTs were increased consistently across the range of 6–18°C, while NTs were increased across the range of 6–14°C (Table 3).
Comparisons between the three clones of aspen indicated that clone Pt5.2 emitted significantly higher amounts of the MTs α-pinene, β-myrcene, and 1,8-cineole than clones Pt2.2 and Pt6.0, although clone Pt2.2 emitted more total MTs than Pt5.2 and Pt6.0 (Table 4). The MT β-ocimene and the homoterpene DMNT were emitted at significantly higher rates by clone Pt2.2 than the other two clones, while the highest emission rate of the SQT (E)-caryophyllene was from clone Pt6.0, differing significantly from clone Pt5.2 (Table 4). The highest isoprene emission was from clone Pt5.2 (Table 4).
Daytime average photosynthesis related to VOC emissions was practically unaffected by night-time temperature, as it was not possible to detect any significant differences between the treatments (Fig. 2a). There was also no significant change in total leaf biomass in either species (Fig. 2b), although total leaf area increased due to warmer night-time temperatures in both species (Fig. 2c). In aspen, the increase in total leaf area was 53% from 6°C to 22°C (linear regression, r2=0.99). In birch, there was no change in total leaf area up to 14°C, but after that a similar linear increase (linear regression, r2=0. 97) resulted in a 32% increase from 6°C to 22°C. In both species SLA also increased with temperature (Fig. 2d). Although a minor effect in SLA was observed between 6°C and 10°C and the strongest impact was observed between 10°C and 14°C, the increase in general was quite linear, despite the differences between the genotypes (data not shown). Increase in SLA: aspen, 3.5 (cm2g−1)/1°C, r2=0.956; birch, 3.5 (cm2g−1)/1°C, r2=0.916. SLA was greater in birches than aspens.
In birch, the highest expression of DXR, DXS, and IPP was found at 10°C, while down-regulation of these genes occurred at higher night-time temperatures (Fig. 3). No major differences in these gene activities were found between the genotypes (Fig. 3). In aspen, expression of DXR and IspS was down-regulated with increasing night-time temperature, whereas DXS and HMGR were not clearly affected by temperature treatments (Fig. 4). Genotypic comparisons showed that IPP and IspS expression was highest in clone Pt5.2 (correlating with the highest isoprene emissions of this clone) and that the lowest expression for all genes was found in clone Pt2.2 (Fig. 4).
The results from the present study indicate that elevated night-time temperature enhances daytime emissions of volatiles, especially SQTs and the homoterpene DMNT in birch. Interestingly, DMNT, which is considered to be primarily induced by biotic stresses (Dicke, 2009) including feeding by spider mites (Pinto et al., 2007), mining fly larvae (Wei et al., 2006), and moth larvae (Vuorinen et al., 2007; Mäntylä et al., 2008) and which has ecological significance, for example in attraction of predatory mites (Pinto et al., 2007) and birds (Mäntylä et al., 2008), showed a strong night-temperature dependence in this study. An earlier study by Gouinguene and Turlings (2002) on maize showed that emissions of DMNT and several SQTs from plants stressed by simulated herbivore damage were independent of growth and sampling temperatures ranging from 17°C to 37°C. However, exposure to C6 GLVs, particularly to (Z)-3-hexenyl acetate, can trigger DMNT emissions from uninfested maize plants (Yan and Wang, 2006) and prime DMNT emissions from Populus deltoides×nigra (Frost et al., 2008). These results suggest that increasing night-time temperature may represent a stress factor for northern trees.
In the present study, several SQTs emitted by birch showed a steady increase in emission rate with increasing night-time temperature. In line with these results, B. pendula was reported to be a high SQT emitter, with emissions equalling or even exceeding MT emissions (Vuorinen et al., 2005, 2007). Exponential dependencies of both MT and SQT emissions on sampling temperature have previously been found for intact pine trees in branch enclosure experiments (Helmig et al., 2006, 2007). It was reported that SQT emission rates of loblolly pine (Pinus taeda L) had a stronger temperature dependency than MT emission rates (Helmig et al., 2007). Helmig et al. (2007) were also able to show that a 10°C increase in night-time temperature increased the night-time emission rate to a level close to the daytime emission rate (at the same temperature), indicating that SQT emissions are strongly modulated by temperature conditions regardless of light. In their study, SQTs were the dominating compounds at temperatures >30°C, while MTs were more abundant at lower temperatures, indicating that emissions of SQTs are triggered by high temperatures, probably as a direct stress response. Hartikainen et al. (2009) did not detect SQT emissions from P. tremula saplings, when grown in normal boreal field conditions with monthly mean temperatures below +20°C. Based on this information, it is suggested that in some tree species such as birch, SQT emissions are temperature dependent and that this may have a positive effect on the thermotolerance of the plants. Accordingly, maximum VOC emissions have also been measured during the warmest summer season in coniferous trees, as reported by Pressley et al. (2004) and Kim et al. (2005) in the case of MTs, and Tarvainen et al. (2005) and Hakola et al. (2006) in the case of SQTs. Night-time sampling of VOC emissions and tissue pools was not conducted in the present study and consequently it cannot be estimated how much of the increasing SQT emissions was due to increased biosynthesis or delayed release from plant surfaces (Helmig et al., 2004; Schaub et al., 2010) at higher temperatures.
In the present study, the isoprene emission measured during the daytime was lowest in plants grown at the highest night temperature. This is consistent with the recent findings of Hartikainen et al., (2009) who found a decrease of isoprene emission at high temperatures in field-growing aspen clone Pt2.2. Although, isoprene responses to night-time temperatures have not been previously reported, the present findings suggest that the temperature dependency of isoprene cannot be generalized to all plants. However, the results are inconsistent with several previous studies reporting that isoprene emissions at daytime temperatures are strictly temperature dependent in addition to light dependent (Singsaas et al., 1999; Hanson and Sharkey 2001; Sharkey et al., 2008). Isoprene emission is also known to be induced by several other abiotic factors, such as increased tropospheric ozone concentrations, as demonstrated in Phragmites australis and Quercus pubescens by Velikova et al. (2005), and by UV-B enhancement, as reported in Q. gambelii by Harley et al. (1996) and peatland plants by Tiiva et al. (2007). Isoprene might play a role in thermoprotection of leaves and especially photosynthetic apparatus (Sharkey et al., 2008, and references therein). Velikova et al. (2006) found that exogenous isoprene fumigation reduced the negative effect of high temperatures (38°C) on photosynthetic activity in Platanus orientalis leaves. In the present experiment, photosynthesis was practically unaffected by increased night-time temperature, but the temperatures used were below normal heat stress temperatures. Therefore, the protective role of isoprene cannot be proved in this case.
The larger total leaf areas of the seedlings with increasing night temperature could mean a consistent increase of all measured emissions. At the forest stand level this kind of change would suggest a considerable increase of daytime reactive VOC emissions. However, calculations of total VOC emissions from whole seedlings cannot be made from these data, because emission profiles probably change both quantitatively (Blande et al., 2007; Frost et al., 2007; Brilli et al., 2009) and qualitatively along the seedling profile from top to base, as is the case with MTs (Brilli et al., 2009) and isoprene in poplars (Sharkey et al., 2008; Brilli et al., 2009). With unaffected net photosynthesis, increasing VOC emissions would indicate changes in carbon balance through enhanced carbon release back to the atmosphere. However, this hypothesis needs to be studied more extensively before further conclusions are drawn.
Increased SLA with night-time temperature is in accordance with the meta-analysis of Poorter et al. (2009) reporting that leaf mass per area is lower in plants grown at higher temperatures. Recent results of Hartikainen et al. (2009) also demonstrated significantly thinner leaves in the same aspen clones growing in field conditions with diurnally elevated temperature. In that study, thinner leaves had a thinner epidermis, palisade, and spongy mesophyll. A thinner leaf could possibly enable a more rapid diffusion and release of newly synthesized SQTs. Some evidence for this comes from considering the reverse of an example by Vuorinen et al. (2004): thicker leaves of elevated CO2-grown cabbage plants emitted a marginally reduced amount of the SQT (E,E)-α-farnesene and the homoterpene DMNT, when VOCs were sampled in ambient air (Vuorinen et al., 2004). However, thinner leaves may need more thermal protection.
The present gene expression studies indicated that VOC emissions are not directly regulated by the transcription level, but that the regulation is very complex. In birch, maximum expression of DXR, DXS, and IPP at a night-time temperature of 10°C coincided with the maximum emissions of total MTs at the same temperature, suggesting that these genes are temperature controlled. However, increasing emission of DMNT with increasing night-time temperature was accompanied by down-regulation of DXR, DXS, and IPP, operating in the MEP pathway. Similarly, in aspen, DXR was slightly suppressed when MTs were increasing at warmer night-time temperatures while IpsS showed a clearer down-regulation with warming. The high complexity of regulation of volatile emissions was previously demonstrated by Nogués et al. (2006) who reported that pool sizes of the immediate precursors (DMAPP and geranyldiphosphate) are inversely related to the emission fluxes of isoprene and MTs through a strong negative control. In addition, Rasulov et al. (2009) reported that isoprene emissions were more dependent on DMAPP pool size than IpsS activity. Mayrhofer et al. (2005) have also reported that the regulation of isoprene biosynthesis in Grey poplar (Populus×canescens) occurs at transcriptional, post-translational, and metabolic levels, and is highly variable with respect to diurnal and seasonal changes in light and temperature. Anyhow, several investigations support that DXS and DXR play an important role in the control of plant isoprenoid biosynthesis. The rate-limiting step is still unknown, although molecular engineering revealed that overexpression of DXS (Estévez et al., 2001) or of DXR (Mahmoud and Croteau, 2001) resulted in an enhanced accumulation of isoprenoid end-products.
In the present experiments, IspS proved to be more responsive than IPP to increasing temperature, showing the highest expression at the lowest night-time temperatures. In contrast to this study, a strong increase in enzymatic activity as a result of the transcriptional up-regulation of the IspS gene was reported to be induced by heat (40°C) in P. alba leaves (Sasaki et al., 2005) and (60°C) in transgenic Arabidopsis (Sasaki et al., 2007). Previous studies with Grey poplar reported that the expression rates of DXR (PcDXR) and IspS (PcISPS) were correlated with each other (Mayrhofer et al., 2005), which was also evident in the present study. A specific role for HMGR in temperature-dependent VOC production could not be demonstrated.
The changes in transcription levels of these selected genes could not explain the differences in VOC emission between the aspen and birch clones. The birch clones were rather similar in their gene expression patterns, although large differences were found in many volatile compounds. Even though aspen showed clearer differences among the clones in terms of DXR, DXS, IPP, and IspS expression, these changes were poorly correlated with related VOC emission patterns. Only the highest isoprene emission in clone Pt5.2 matched the highest expression of IPP and IspS in genotypic comparison.
The experiment demonstrated that increasing night-time temperatures strongly increase daytime emissions of the homoterpene DMNT, several SQTs, and total SQTs (at higher temperatures) in birch and the emission of several MTs up to 18°C in aspen, while isoprene emissions were not correlated with temperature increase. Although expression of DXR, DXS, IPP, and IspS was temperature dependent, further research is needed to understand the complex nature of regulation, which seems to involve inverse relationships with VOC emissions.
This study was funded by the Academy of Finland (grants 116981 to MAI, 109933 to EJO, 122338 to VH, and 111543 to JKH), University of Eastern Finland (Centre for Applied Plant Biology to VH). We would like to thank Dr James Blande for critical reading of the manuscript and linguistic suggestions. Dr Markku Aalto is thanked for providing the sequences from the Betula EST library, and Timo Oksanen for technical assistance.