Regulation of isoprene emission in Populus trichocarpa leaves subjected to changing growth temperature

Authors


T. D. Sharkey. Fax: +1 517 353 9334; e-mail: tsharkey@msu.edu

ABSTRACT

The hydrocarbon isoprene is emitted in large quantities from numerous plant species, and has a substantial impact on atmospheric chemistry. Temperature affects isoprene emission at several levels: the temperature at which emission is measured, the temperature at which leaves develop, and the temperatures to which a mature leaf is exposed in the days prior to emission measurement. The molecular regulation of the response to the last of these factors was investigated in this study. When plants were grown at 20 °C and moved from 20 to 30 °C and back, or grown at 30 °C and moved from 30 to 20 °C and back, their isoprene emission peaked within 3 h of the move and stabilized over the following 3 d. Trees that developed at 20 °C and experienced 30 °C episodes had higher isoprene emission capacities than did leaves grown exclusively at 20 °C, even 2 weeks after the last 30 °C episode. The levels and extractable activities of isoprene synthase protein, which catalyses the synthesis of isoprene, and those of dimethylallyl diphosphate (DMADP), its substrate, alone could not explain observed variations in isoprene emission. Therefore, we conclude that control of isoprene emission in mature leaves is shared between isoprene synthase protein and DMADP supply.

INTRODUCTION

Isoprene (2-methyl 1,3-butadiene) is a biogenic hydrocarbon produced by many plant species (Sharkey & Yeh 2001). There is an extremely large flux of isoprene from plants to the atmosphere [500 Tg per year (Guenther et al. 1995; Fuentes et al. 2000)], and because isoprene is reactive with hydroxyl radicals and ozone in the atmosphere, it is of substantial importance in atmospheric chemistry (Thompson 1992; Monson & Holland 2001). Development of mechanistic models of isoprene emission could help predict emission rates; for such models to be developed, the factors and processes that regulate isoprene emission must be understood.

Isoprene is synthesized by the activity of isoprene synthase (IspS) (Silver & Fall 1991), upon dimethylallyl diphosphate (DMADP). The DMADP used for isoprene synthesis is produced by the chloroplastic methylerythritol 4-phosphate (MEP) pathway (Schwender et al. 1997). Until recently, the molecular mechanisms that regulate isoprene emission were not at all understood, though it was known that the activity of IspS can be correlated with isoprene emission rate (Kuzma & Fall 1993; Schnitzler, Lehning & Steinbrecher 1997; Sharkey et al. 2005). In developing kudzu (Pueraria montana) and Populus trichocarpa leaves, isoprene emission is regulated at the level of IspS transcription (Wiberley et al. 2005; Sharkey, Wiberley & Donohue 2007). Studies of IspS and deoxyxylulose-5-phosphate reductoisomerase, the committing enzyme of the MEP pathway, have indicated that in Populus × canescens, isoprene emission is regulated at transcriptional and post-translational levels (Mayrhofer et al. 2005; Loivamäki et al. 2007). However, the molecular regulation of isoprene emission in response to a variety of factors, including changing growth temperatures for mature leaves, is not yet elucidated.

Isoprene emission is greatly affected by leaf temperature; these effects can be observed at several levels. Firstly, the leaf temperature during the measurement of emission rate is known to alter observed emission. Emission increases with increasing measurement temperature from 20 to about 40 °C and declines thereafter (Rasmussen & Jones 1973; Tingey et al. 1979; Monson & Fall 1989; Loreto & Sharkey 1990; Harley, Guenther & Zimmerman 1996, 1997; Singsaas & Sharkey 2000). This pattern is accounted for as a logarithmic response in common isoprene emission models (Guenther et al. 1993, 1995). However, over the course of a day (Geron et al. 2000; Funk et al. 2003) or a growing season (Pier & McDuffie 1997; Schnitzler et al. 1997; Goldstein et al. 1998; Fuentes & Wang 1999; Fuentes, Wang & Gu 1999; Geron et al. 2000; Xiaoshan et al. 2000; Lehning et al. 2001; Petrón et al. 2001; Rinne et al. 2002), the capacity for isoprene emission at a specified temperature can increase, which does not fit with existing models.

The temperatures at which leaves develop also affect their isoprene emission capacities. Leaves that are grown at lower temperatures have lower emission capacities than those grown at higher temperatures, and they also begin to emit isoprene later in their development. This may also partially explain the variation in isoprene emission rates that is observed throughout a growing season (Monson et al. 1994; Wiberley et al. 2005).

Finally, the temperatures that mature leaves have experienced in the days prior to measurement alter their isoprene emission. Isoprene emission capacities correlate with the temperatures that the leaves have experienced during the past 6 h to 2 d (Sharkey et al. 1999; Geron et al. 2000; Hanson & Sharkey 2001), or up to 15 d (Petrón et al. 2001), but the proposed time range is variable and a molecular explanation for this phenomenon is lacking.

In this study, newly available tools for molecular analysis of isoprene synthesis were used to examine the regulation of isoprene emission in leaves subjected to changing growth temperatures. Plants that developed at 20 and 30 °C were moved back and forth between these two temperatures; their isoprene emission capacities and photosynthesis rates, IspS mRNA and protein quantities, and DMADP levels were monitored. Though plants were grown at different temperatures, all leaves were held at 30 °C while isoprene emission was being assayed, so that the effects of differing growth temperatures on capacity for isoprene emission at the same temperature could be observed.

MATERIALS AND METHODS

Plant growth conditions

Poplar trees [P. trichocarpa (Torr. & Gray)] were grown from 23 cm stem cuttings obtained from Segal Ranch Hybrid Poplars (Grandview, WA, USA), in 19 L pots containing a vermiculite/peat moss-based growth medium (Metro-Mix 360, The Scotts Company, Marysville, OH, USA). The trees were grown in temperature-controlled greenhouses in the Biotron at the University of Wisconsin-Madison. Fifteen were grown at 20/16 °C day/night temperature without extended day length, while 20 were grown at 30/20 °C day/night temperature with day length extended to 16 h using high-pressure sodium vapour lamps that provided 300 µmol m–2 s–1. Lights came on at 0500 h and were turned off at 2100 h. Plants were watered with 1/10 strength Hoagland's solution (Hoagland & Arnon 1938); they were also fertilized with 14-14-14N-P-K Osmocote, applied according to the manufacturer's instructions (The Scotts Company).

Leaf sample collection

The first leaf longer than 1 cm at the shoot apex of each tree was designated ‘leaf 1’, and the numbers of all other leaves were determined by counting down the stem from leaf 1. It was previously determined that leaves 10 through 16 have indistinguishable isoprene emission (data not shown), so sampled leaves were selected from this range. Also prior to this experiment, isoprene emission from leaves of trees grown at 20 and 30 °C was sampled every 3 h from 0600 through 1500 h, to determine the times at which isoprene emission was fairly constant. This was done on 7 and 8 June 2005. Samples for the temperature-change experiment were collected between 21 June and 15 July 2005. All tissue samples were frozen in dry ice immediately after collection and then stored at –80 °C until use.

The initial temperature-change sample collection included leaves from three trees grown at 20 °C and three trees grown at 30 °C. This was designated time 0 for move A (A0). Ten of the trees were then moved from the 20 °C room into the 30 °C room, and 12 from the 30 °C room into the 20 °C room. Samples were collected 3, 6, 24, 48 and 72 h after this move (designated A3, A6, etc.). After 72 h, the plants were moved back to their original temperature; time point A72 also served as the zero time point for this second move (A72/B0). Again, samples were collected 3, 6, 24, 48 and 72 h after the second move (B3, B6, etc.). This was repeated for a third move, which took the 20 °C-grown trees to 30 °C for a second time, and the 30 °C-grown trees to 20 °C for a second time (time points B72/C0, C3, C6, etc.). After time point C72/D0, the trees were moved back to their original growth temperatures; samples were then collected every 3 d for 2 weeks (D72, D144, D216, D288 and D360). At some time points, samples were also collected from 20 °C- and 30 °C-grown trees that were never moved, but stayed at one set of growth temperatures constantly. These trees were designated ‘20U’ and ‘30U’, indicating the temperature at which they developed and the fact that they were unmoved. Trees that were moved were designated ‘20M’ and ‘30M’, indicating the temperature at which they developed and the fact that they were moved between temperatures.

Samples at 0, 24, 48 and 72 h after moving, as well as all the ‘D’ time points, were collected between 0900 and 1100 h; samples at 3 h after moving, between 1200 and 1330 h; samples at 6 h after moving, between 1500 and 1630 h. These times were selected because preliminary experiments had shown isoprene emission in this experimental set-up to be fairly constant between 0900 and 1600 h (Fig. 1). Collected leaf punches were 113 mm2 in area. Each time any treatment (20U, 20M, 30U, 30M) was sampled, one leaf from each of three different trees was collected for that treatment and the leaves’ data averaged.

Figure 1.

Basal isoprene emission rates for poplar leaves grown at 20 and 30 °C, over the course of a day. Emission rates were measured at 30 °C and 1000 µmol m–2 s–1 light. Each value is the average of three leaves’ emission rates ± SE. Light bars, 20 °C; dark bars, 30 °C.

Gas-exchange and isoprene emission measurement

Gas-exchange and isoprene emission measurements were made using a LI-COR 6400 (NB Li-Cor Inc., Lincoln, NE, USA) and a Fast Isoprene Sensor (Hills Scientific, Boulder, CO, USA), as described by Hanson & Sharkey (2001), with minor modifications. All isoprene emission and photosynthesis rates were measured at 30 °C leaf temperature and 1000 µmol photons m–2 s–1.

RNA extraction and quantitative polymerase chain reaction (PCR) analysis

Total RNA was extracted as described by Haruta et al. (2001) and quantitated using a Beckman DU 640 spectrophotometer (Beckman Coulter, Inc., Fullerton, CA, USA). All chemicals were obtained from Sigma-Aldrich (St. Louis, MO, USA) unless otherwise noted. RNA was reverse transcribed and analysed by quantitative PCR as described by Wiberley et al. (2005). The primers used were forward: 5′-ACAAATGCTGTTGAGAGATGG-3′ and reverse: 5′-CCATTTTGCTTCTTGTAGGAATG-3′. These span the fourth and fifth introns of the poplar IspS gene, and the forward primer itself spans the fourth intron [prediction from the JGI P. trichocarpa genome sequence, version 1.1 (http://genome.jgi-psf.org/Poptr1_1/Poptr1_1.home.html)], so any contaminating genomic DNA in the RNA preparation would not be amplified. The target sequence had previously been amplified from a reverse-transcribed poplar RNA extract, and the resulting DNA was quantitated as previously described. Dilutions of this, containing known numbers of copies of the target sequence, were used to prepare standard curves that were used to determine the copy numbers of the plant samples. Plant sample reactions were spiked with 1000 copies of the target amplicon to eliminate the formation of primer dimers, and this number was subtracted from the copy number calculated by the software of the Mx3000P (Wiberley et al. 2005).

Protein extraction and Western blot analysis

To extract total protein, one sample from each leaf was ground in 200 µL of SDS sample buffer (1 m Tris-HCl, pH 6.8; 10% SDS; 10% β-mercaptoethanol; 20% glycerol; 0.004% bromphenol blue), heated at 70 °C for 10 min, and centrifuged for 5 min at 12 000 g to pellet out cell wall fragments. Supernatants were separated by sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS– PAGE) and transferred to Hybond-P PVDF membrane (Amersham Biosciences, Piscataway, NJ, USA) as described by Wiberley et al. (2005). A polyclonal primary antibody for IspS was made using P. trichocarpa IspS cDNA expressed in a pET28a vector (Novagen, Madison, WI, USA) in Escherichia coli (Calfapietra et al. 2007). The primary antibody was diluted 1:5000 and the secondary antibody, which had horseradish peroxidase coupled to a donkey anti-rabbit antibody (Amersham Biosciences), was diluted 1:2500. Membranes were exposed to the antibodies using the protocol of Birkett et al. (1985), and immunoreactive proteins were detected using ECL Western blotting substrate according to the manufacturer's instructions (Pierce Biotechnology, Rockford, IL, USA). Blots were exposed to Kodak (Rochester, NY, USA) Biomax MS X-ray film for 30 s–30 min. The IspS standard was purified from E. coli as previously described (Calfapietra et al. 2007); its concentration was determined by Bradford assay (Bradford 1976). Exposed X-ray films were scanned and the protein bands quantitated using ImageJ (US National Institutes of Health, http://rsb.info.nih.gov/ij/).

Isoprene synthase activity assays

Tissue samples were ground in liquid nitrogen and homogenized in a pH 8.0 buffer of 100 mm HEPES, 20 mm MgCl2, 100 mm CaCl2, 10 mm mannitol, 30 mm NaCl, 2 mm ethylenediamine tetraacetic acid (EDTA), 0.1% (w/v) bovine serum albumin (BSA), 10% (v/v) glycerol and 0.1% (v/v) Tween 80, with 20 mm dithiotreitol (DTT), 50 mg mL−1 polyvinylpolypyrrollidone (PVPP) and 1 mm phenylmethyl sulfonyl urea (PMSF) added just before use and stirred for 15 min (based on Mayrhofer et al. 2005). Samples were centrifuged at 12 000 g for 20 min to pellet cell wall debris, and supernatants were assayed according to the method of Lehning et al. (1999) with modifications. Samples were not desalted prior to analysis; the pH 8.0 assay buffer consisted of 50 mm HEPES, 5 mm KCl, 50 mm MgCl2, 2 mm NaF and 10% (v/v) glycerol, and the isoprene produced was detected by gas chromatography as described by Wiberley et al. (2005). DMADP was synthesized according to the method of Davisson, Woodside & Poulter (1985).

DMADP assays

Tissue samples were submerged in 200 µL of 0.5 m H2SO4 in 2 mL crimp-seal vials (Alltech, Deerfield, IL, USA) and incubated for 90 min at 70 °C, in an adaptation of the protocols described by Fisher et al. (2001) and Brüggemann & Schnitzler (2002b). After incubation, 1 mL of headspace was withdrawn from each vial using water displacement, and the headspace was injected an air stream flowing through the Fast Isoprene Sensor. The Fast Isoprene Sensor was set up to record isoprene signals 10 times per second; these signals were summed for each sample, the background subtracted, and the net signal used to calculate the amount of isoprene produced from acid hydrolysis of DMADP. Standards were run using liquid isoprene serially diluted to 200 pmol mL–1 in N2. The range of this assay was checked using DMADP synthesized as previously described, and it was found to be linear from 0.5 to 32 nmol DMADP, with approximately 45% conversion of DMADP to isoprene (data not shown). Measured DMADP contents for leaves were divided by 0.45 to adjust for this percent conversion.

Statistics

Data were analysed with JMP (version 5, SAS Institute, Cary, NC, USA) using analysis of variance to determine differences.

RESULTS

Isoprene emission and photosynthesis

When plants were moved from 20 to 30 °C, or 30 to 20 °C, emission capacity rose significantly within 3 h but then decreased and stabilized over the following 3 d (Fig. 2a,b). Within 3 d of each move, emission capacities tended to stabilize around 30 nmol m–2 s–1, regardless of whether the trees had developed at 20 or 30 °C and regardless of whether they were growing at 20 or 30 °C when isoprene emission was assayed. The increases in emission capacities after moving were significant at the 0.05 level for the first and third moves. After plants were returned to their original growth temperatures and remained there for 2 weeks, emission capacities for the 20M trees were substantially higher than those of their 20U counterparts; this was not observed for the 30M trees as compared to their 30U counterparts (Fig. 3a). Photosynthesis rates for all groups of trees remained largely constant throughout the entire experiment (Fig. 2c,d), and were similar for moved and unmoved 20 and 30 °C trees (Fig. 3b). Variations in photosynthesis rates were generally not statistically significant.

Figure 2.

Isoprene emission (a,b) and photosynthesis (c,d) rates, IspS mRNA (e,f) and protein (g,h) quantities, extractable IspS activities (i,j) and whole-leaf dimethylallyl diphosphate (DMADP) contents (k,l) for trees that developed at 20 °C (light bars, left) or 30 °C (dark bars, right) and were moved from one temperature to the other. A light band at the top of the graph indicates that the trees were at 20 °C when the measurements below the band were collected; a dark band, that they were at 30 °C. Vertical lines within graphs indicate when trees were moved from one temperature to the other. Each value is the average of three leaves’ data ± SE. Isoprene emission and photosynthesis rates were measured at 30 °C and 1000 µmol m–2 s–1 light.

Figure 3.

Isoprene emission (a) and photosynthesis (b) rates, IspS mRNA (c) and protein (d) quantities, extractable IspS activities (e), and whole-leaf dimethylallyl diphosphate (DMADP) contents (f) for trees grown at 20 °C (lighter bars) or 30 °C (darker bars) and then moved (M, intermediate-darkness bars) or not moved (U, lightest and darkest bars) between the two temperatures. Trees were at the temperatures at which they had developed when all measurements were taken. Each value is the average of three leaves’ data ± SE. Isoprene emission and photosynthesis rates were measured at 30 °C and 1000 µmol m–2 s–1 light.

IspS mRNA, protein and extractable activity

With the exception of one outlying time point, IspS mRNA pool size stayed fairly constant throughout the entire experiment (Fig. 2e,f). IspS protein quantities did change, being higher after the second and third temperature changes than after the first (Fig. 2g,h), but these changes did not correspond to the observed variations in isoprene emission capacities. For 2 weeks following the last move, the 20U and 20M trees and the 30U and 30M trees all had similar IspS mRNA pool sizes (Fig. 3c), and also had similar IspS protein quantities (Fig. 3d), even though the 30M and 30U trees, and the 20M trees had higher emission capacities than did the 20U trees. Extractable IspS activity stayed fairly constant throughout the course of the experiment, and was sufficient to account for only about 8% of observed emission capacity (Figs 2i,j & 3e). Changes in IspS mRNA and protein levels, and extractable IspS activity, did not correlate with changes in isoprene emission capacities (Fig. 4a–c). These changes were also generally not statistically significant.

Figure 4.

Correlations of IspS mRNA (a), IspS protein (b), extractable IspS activities (c) and whole-leaf dimethylallyl diphosphate (DMADP) (d) with isoprene emission rates. Modified Woolf–Hanes correlation with IspS protein, whole-leaf DMADP and isoprene emission (e). Units for (e): x-axis = (µmol m–2 leaf)4; y-axis = (mg IspS µmol4 DMADP s m–8 leaf nmol–1 isoprene). Isoprene emission rates were measured at 30 °C and 1000 µmol m–2 s–1 light.

Whole-leaf DMADP levels

Whole-leaf DMADP contents dropped in both 20M and 30M trees as the experiment progressed; they were highest shortly after the first move and then dropped after the second and third moves (Fig. 2k,l). These variations were significant at the 0.1 level for 30M trees, but they did not correlate with changes in isoprene emission (Fig. 4d). In the comparison of moved and unmoved plants, DMADP contents were not significantly different in one group relative to any other (Fig. 3f), though isoprene emission capacities did differ among groups.

A modified Woolf–Hanes equation was used to relate isoprene emission to IspS protein and DMADP levels simultaneously, to explore whether the interaction of these two factors correlated with emission. The Woolf–Hanes equation:

image

was multiplied through by Vmax (Vmax = E·kcat), so that (E·[S]) / vi could be plotted as a function of [S]. Because IspS can exhibit cooperativity (Schnitzler et al. 2005) with a Hill coefficient of 4 (Sharkey et al. 2005), [S] was raised to the fourth power:

image

Each data point corresponds to one leaf, with E being its IspS quantity as determined by Western blot, [S] being its assayed DMADP content multiplied by 0.68, and vi being its isoprene emission rate. DMADP contents were multiplied by 0.68 because previous work has shown 65–71% of whole-leaf DMADP to be in the chloroplast (Rosenstiel et al. 2002). This transformation of the data improved the correlation between isoprene emission and IspS enzyme and substrate quantities (Fig. 4e). From the equation of the line of best fit for this plot, values for Km and kcat could be determined. Km was found to be 0.32 mm and kcat, 1.9 s–1.

DISCUSSION

When mature P. trichocarpa leaves were subjected to repeated changes in growth temperature, isoprene emission capacities varied somewhat as temperatures changed, but not in a manner that correlated with IspS mRNA, protein levels or extractable activity, or with whole-leaf DMADP contents. Isoprene emission capacities increased significantly within 3 h of most of the moves, whether they were from lower to higher or higher to lower temperature. These changes were too large to be accounted for by the change in time of day alone, as it had previously been observed that in these growing conditions, emission is fairly constant from 0900 until 1600 h. Circadian regulation of isoprene emission has been demonstrated (Mayrhofer et al. 2005; Wilkinson et al. 2006; Loivamäki et al. 2007) and this may have interacted with the temperature changes in some way even though emission had been found to be constant throughout the times when samples were collected here. It therefore seems that either the act of physically moving the plants or their being subjected to a change in temperature, regardless of its direction, caused their isoprene emission capacities to rise.

The effects of various physical disturbances on isoprene emission have been studied previously. Loreto & Sharkey (1993) found that when Mucuna deeringeniana and Pueraria montana leaves were cut or burned, their isoprene emission decreased within minutes and stayed low for at least an hour. Loreto et al. (2006) wounded Phragmites leaves and observed an increase in isoprene emission rates, but this lasted only about 10 min; the same has been observed for Quercus leaves (Sharkey, unpublished observation). Hanson & Sharkey (2001) moved Sphagnum and Quercus specimens from lower to higher and higher to lower temperatures, with different light levels, and found that when plants were moved to higher temperatures, their emission rates increased within 1 d; when moved to lower temperatures, emission rates decreased within 1 d, and plants moved to the same temperature but higher light levels had no change in emission. The cause of the increase in isoprene emission 3 h after moving poplars from one temperature to the other remains unclear, but it could indicate a connection between isoprene metabolism and cellular signalling systems.

After the initial increase that followed each move, isoprene emission capacities stabilized at about 30 nmol m–2 s–1; this rate was the same regardless of whether plants developed at 20 or 30 °C or were growing at 20 or 30 °C when isoprene emission was assayed. This represented a substantial increase in emission capacity for the 20M plants but not for the 30M plants. As long as 2 weeks after the last move, the 20M plants still had higher emission capacities than their 20U counterparts, while the 30M and 30U plants had very similar emission capacities. Therefore, once a plant is induced to make larger amounts of isoprene by being transferred to a higher temperature, it continues to be capable of emitting at high levels, even when it has been growing at a lower temperature for at least 2 weeks.

IspS mRNA and protein quantities varied throughout this experiment, but these variations did not correlate with changes in isoprene emission capacity. In mature leaves, particularly those that developed at 30 °C, IspS protein levels were highest after the second move, even though emission capacities were not. The control of isoprene emission in mature leaves subjected to changing temperatures is therefore shared between IspS activity and supply of its substrate, because protein and product quantities for IspS are largely unrelated. This is at odds with the previously observed correlation in developing leaves (Wiberley et al. 2005; Sharkey et al. 2007). It seems likely that while isoprene synthase expression and isoprene emission are closely related in developing leaves that are grown at constant temperature, once the leaves are mature and growth temperatures change, the control of emission becomes more complex.

IspS activity assays were performed but did not contribute to the understanding of the regulation of isoprene emission capacity in these conditions. No IspS activities measured in poplar to date have been sufficient to account for observed emission capacities, and extractable activity can be measured even when leaves are not emitting isoprene (Mayrhofer et al. 2005; Magel et al. 2006; Loivamäki et al. 2007). While extractable IspS activity can account for observed emission rates in Quercus (Lehning et al. 1999; Brüggemann & Schnitzler 2002a) and Phragmites (Scholefield et al. 2004), the same has not been found in poplar, and it may be that the regulation in poplar is more complex or existing extraction techniques are insufficient. IspS activity may well be important in controlling isoprene emission in response to temperature changes, but present techniques with poplar do not allow it to be measured in a valuable way.

When whole-leaf DMADP was assayed, no clear relationship between emission and DMADP content was found. DMADP levels were higher at the start of the experiment and then dropped towards the end, but there was no corresponding change in emission capacity. The assay included DMADP in the chloroplast, which is available to isoprene synthesis, and cytosolic DMADP, which is not, so the chloroplastic fraction of DMADP could have varied in a manner related to the changes in emission without being detected. Between 65 and 71% of whole-leaf DMADP has been found to be in the chloroplast (Rosenstiel et al. 2002), but the consistency of this distribution through different species and conditions is not known. Some crosstalk between the plastidic and cytosolic DMADP synthesis pathways has been observed (Bick & Lange 2003; Laule et al. 2003; Dudareva et al. 2005), but this involves much lower concentrations of DMADP than are needed by IspS, and also seems to occur mainly as export from the chloroplast to the cytosol. On a given day, DMADP content can be inversely proportional to isoprene emission rate (Wolfertz et al. 2003), but the present study deals with a longer time span, which may alter the relationship between isoprene emission and DMADP content. The modified Woolf–Hanes analysis greatly improved the correlations between DMADP levels, IspS quantity and isoprene emission. The Km and kcat values found in this analysis are similar to those found in in vitro assays for other Populus species (Sasaki, Ohara & Yazaki 2005; Silver & Fall 1995; Miller, Oschinski & Zimmer 2001; Schnitzler et al. 2005). This supports the idea that an interaction of IspS protein level and DMADP supply controls isoprene emission rates in mature leaves.

Regulation of the MEP pathway has been demonstrated at several steps in the production of other isoprenoids: deoxyxylulose 5-phosphate synthase (DXS), the first enzyme of the pathway, has an apparent regulatory role in isoprenoid synthesis in developing tomato fruits (Lois et al. 2000), in lavender (Munoz-Bertomeu et al. 2006) and in Arabidopsis (Estévez et al. 2001). Feedback regulation of DXS has been suggested to regulate isoprene emission as well (Wolfertz et al. 2003, 2004). Deoxyxylulose 5-phosphate reductoisomerase, the pathway's committing enzyme, had a regulatory role indicated by Carretero-Paulet et al. (2002, 2006); and 1-hydroxy-2-methyl-butenyl 4-diphosphate reductase (HDR), which catalyses the penultimate step of the pathway, producing DMADP and its isomer isopentenyl diphosphate, has been shown to regulate other enzymes of the pathway post-transcriptionally (Guevara-García et al. 2005). In addition, isopentenyl diphosphate isomerase, which interconverts the two products of HDR, has been shown to correlate with isoprene emission (Brüggemann & Schnitzler 2002c) and to regulate isoprenoid accumulation in Zea mays (Albrecht & Sandmann 1994) and in a unicellular green alga (Sun, Cunningham & Gantt 1998). It is therefore reasonable that isoprene emission from mature leaves should be regulated by enzymes upstream of IspS in the isoprene biosynthetic pathway, with control being at the level of one of the upstream enzymes or shared among several.

It has been well demonstrated that isoprene emission varies over the course of a growing season, and is highest in summer and decreases in the autumn (Pier & McDuffie 1997; Schnitzler et al. 1997; Goldstein et al. 1998; Fuentes & Wang 1999; Fuentes et al. 1999; Geron et al. 2000; Xiaoshan et al. 2000; Lehning et al. 2001; Petrón et al. 2001; Rinne et al. 2002). Here, the intermediate-term effects of increasing and decreasing growth temperatures on emission capacities were examined, comparing emission capacities from trees grown at different temperatures and subjected or not subjected to temperature changes, and investigating molecular explanations for the observed emissions. This study shows that the control of emission in these conditions is complex, and likely shared among IspS and upstream enzymes in the MEP pathway. Development of tools for the study of regulation of DMADP supply for isoprene synthesis is currently underway.

ACKNOWLEDGMENTS

The authors thank Carlo Calfapietra and Tanya Falbel for assistance with preparation and acquisition of IspS polyclonal antibodies. This research was supported by the National Science Foundation Grants IBN-0212204 and IOB-0640853.

Ancillary