Production and manipulation of 11CO2 for rapid pulse labelling
11CO2 was produced on the BNL 41 inch (Japan Steel Works, Ltd, Tokyo, Japan) cyclotron in a 212-mL volume gas target pressurized to 200 psi with scientific grade N2 (99.9995% purity; MG Industries, Malvern, PA, USA). Irradiation using 17 MeV H+ generated 11C via the 14N(p,α)11C nuclear transformation as 11CO2 (Ferrieri & Wolf 1983). The radioactive gas was immediately released from the target after bombardment (Fig. 1), and the tracer concentrated through selective adsorption onto a bed of Ni(0) catalyst (Fig. 2). The catalyst bed was comprised of 135 mg of silica support Ni catalyst (Shimadzu, Inc., Kyoto, Japan) mixed with molecular sieve 4 Å (100 mesh; Alltech, Inc., Lancaster, PA, USA). The material was contained within a stainless steel tube using standard high performance liquid chromatography column frits (Alltech, Inc.). Prior to use, the catalyst was heated to 400 °C under H2 flow to ensure complete reduction of the Ni metal. Adsorption of 11CO2 from the target N2 gas stream was quantitative for a target gas flow rate set at 1.5 L min−1 and a bed temperature ranging between 20 and 30 °C. Once the tracer was concentrated on the bed, the module was resealed and rapidly heated to 400 °C using resistance cartridge heaters. In the absence of extraneous H2, 11CO2 rapidly desorbed from the Ni(0) surface without effect on its chemical integrity. The desorbed 11C was passed through an oxidation furnace to ensure all 11C was oxidized to CO2 and swept to the experimental platform in a controlled flow of air mixed with a desired level of unlabelled CO2 and a portion diverted to the leaf chamber. The catalyst within the module was restored to its reduced state by flushing it with H2 for a few minutes prior to reducing temperature back to ambient.
Figure 1. Schematic representation of the experimental platform used for pulse administration of 11CO2. The platform included a complete interface with the cyclotron target, a tracer pulse generator, leaf chamber and gas exchange instrumentation. The leaf chamber had its own red/blue LED light source providing a maximum irradiance of 800 µmol m−2 s−1. This chamber resided within a larger environmental chamber that maintained ambient light (average 400 µmol m−2 s−1) and temperature.
Download figure to PowerPoint
Figure 2. Schematic representation of the 11CO2 pulse generator module. This module consisted of a Ni(0) catalyst bed that quantitatively removed 11CO2 from the target gas, and allowed it to be introduced as a discrete pulse in the plant gas mixture.
Download figure to PowerPoint
We used a source gas of zero air (0% CO2 in air; Praxair, Inc., Danburg, CT, USA) blending it with 800 p.p.m. of CO2 in air to a level of 420 p.p.m. CO2 using electronic mass flow controllers on the gas feed stream (MKS Industries, Wilmington, MA, USA), with a total flow rate of 400 mL min−1. The gas flow was split on exiting the pulse module (Fig. 1) yielding a flow rate of 150 mL min−1 that first passed through a dew point generator (Li-Cor model LI-610; Li-Cor Inc., Lincoln, NE, USA), then through the leaf chamber, and exiting through the sample cell of an infrared gas analyser (IRGA, model LI-6262; Li-Cor, Inc.). The other split component, 250 mL min−1, was routed directly through the reference cell of the IRGA to provide gas exchange information. Outflow was also temporarily diverted to a gas chromatograph immediately prior to 11CO2 administration to measure total isoprene emissions.
The leaf chamber was fabricated from Plexiglas on a split clamshell hinge that allowed us to seal targeted leaves between neoprene rubber gaskets. The chamber housed a bank of red/blue LED lights providing an irradiance of 800 µmol m−2 s−1. It also had four inlet ports mounted across the top half and four exhaust ports mounted across the bottom to provide adequate gas mixing. The 150 mL min−1 flow rate of air was sufficient to void a 30-s radiotracer pulse from the chamber volume within 3 min. Leaf surface temperatures were constantly monitored using fine wire thermocouples (Omega Engineering, Inc., Stamford, CT, USA) The leaf chamber was housed within a larger environmental chamber that maintained air temperatures between 20 and 30 °C with 1 °C accuracy, and an average light irradiance of 450 µmol m−2 s−1, using a water-filtered 300-W greenhouse lamp.
The leaf chamber was also equipped with a PIN diode gamma radiation detector (Bioscan, Inc., Washington, DC, USA) to provide direct measurement of 11CO2 in the leaf chamber. Typically, time-activity data from this detector (Fig. 3) reveals information about the pulse magnitude, tissue fixation and subsequent loss of tracer from the leaf reflecting a combination of phloem transport of mobile [11C]photosynthate, emission of [11C]BVOC, and photorespiration of 11CO2. The amount of carbon fixation was used to normalize measurements from the same plant when re-labelled. We typically administered 20 mCi pulses of 11CO2. Where it was necessary to vary the amount of carbon-11 in the pulse, the gas target was irradiated for different times. PIN diode detector response was periodically calibrated by replacing the leaf with a silica thin layer chromatography strip (Fisher Scientific, Inc., Fairlawn, NJ, USA) spotted with known amounts of aqueous [11C]carbonate.
Figure 3. Decay corrected time-activity data (measured by the PIN diode detector) illustrating the shape of the pulse passing through the leaf chamber. Features of this curve include measurement of pulse size and leaf fixation of 11C for data normalization, and leaf loss of tracer reflecting a combination of sugar export and volatile emissions (BVOCs and CO2).
Download figure to PowerPoint
Measurement of leaf [11C] isoprene emission
Administration of discrete pulses of 11CO2 allowed us to assess three parameters regarding [11C]isoprene emission. First, we measured transit times reflecting speed of tracer movement through the isoprenoid pathway. Second, we measured integrated flux and related it to amount of recently fixed carbon. Third, we measured changes in the source of carbon for isoprene biosynthesis over time providing insight into whether treatments changed the relative proportions of recently fixed carbon and/or old carbon sources.
Air from the leaf chamber was exhausted through a well-shielded soda-lime CO2 trap (Alltech Inc.; 80–100 mesh) to capture 11CO2 from the pulse and from any subsequent leaf 11CO2 emissions, as well as any radiolabelled volatile monoterpenes. We verified that isoprene was not retained by the soda-lime packing, but that monoterpenes were. This was accomplished using a short column (6 in. × 1/4 in o.d) packed with soda lime particles (20–40 mesh) coupled to a thermal conductivity detector of a gas chromatograph. We observed no loss of isoprene when passed through this packed column compared against standard samples passed through the empty column. A similar test for monoterpene retention by soda-lime was performed using β-pinene. Unlike isoprene, this higher molecular weight substance was completely retained.
The exhaust from this first soda-lime trap passed to an auxiliary platform that housed an auto-sampling capillary gas chromatograph with a photometric detector to quantify total isoprene emissions. Retention time and detector response were calibrated using isoprene standards. A by-pass on the GC sampling loop allowed gas to pass to a quartz furnace filled with 20 g of copper oxide (Aldrich, Milwaukee, WI, USA; 30–60 mesh) and maintained at 800 °C to oxidize [11C]isoprene to 11CO2. The 11CO2 from this process was captured on a second soda-lime trap placed near a well-shielded sodium iodide scintillation detector (Canberra, Inc., Meriden, CT, USA), configured for single 511 keV photon counting and sensitive to less than 1 µCi of radioactivity. All data were decay-corrected for display and analysis.
We verified in one experiment that radioactivity retained on the second trap reflected only [11C]isoprene emission by passing the outflow from the first soda-lime trap through a small glass cartridge packed with Tenex GR (graphitized carbon molecular sieve; Alltech). This material has a high retention for VOCs and was preloaded with unlabelled isoprene. The trapped radioactivity was quantified, and then the contents desorbed into a packed column radio gas chromatograph (6 ft. × 1/8 in. o.d. stainless steel column packed with 80–100 mesh Porapak T; Analabs, Inc., North Haven, CT, USA) equipped with thermal conductivity and radiation detectors connected in series. We observed one major radioactive peak that co-eluted with unlabelled isoprene. The integral of the radioactivity peak matched the starting amount of radioactivity on the cartridge. A trace amount of [11C]ethylene was also observed (< 0.01% of total trapped [11C]isoprene) which is ignored in this work.
Quantification of leaf isoprene emissions
We defined the transit time as the time for tracer to diffuse to the chloroplast, via the initial stages of the Calvin cycle, enter isoprenoid pathway, diffuse in mesophyll cells as [11C]isoprene for final release through the stomata, and subsequently be transported to the analytical system's isoprene trap. To infer a transit time, the time of half-maximum activity in the trap was estimated from the time of the maximum of a transform of the data using Eqn 1 (see inset in Fig. 4):
Figure 4. Integrated [11C]isoprene emissions from birch (a low isoprene emitter) and aspen (a high isoprene emitter). The data was decay corrected and normalized to 11C fixation. The insert figure illustrates the half maximum transformation of data using 1, and can be used to measure the transit time (to half maximum) for tracer to move through the leaf.
Download figure to PowerPoint
- Atransformed = Ahalfmax − ABS(Ahalfmax − At)(1)
Isoprene activity was presented as percent of fixed 11C relying on the PIN diode detector to measure tissue fixation, and calibrated response of the NaI detector. The half-maximum activity (Ahalfmax) was determined by dividing the maximum activity by two. The maximum activity was an average of several values at the emission plateau. Activity at specific time points (At) were the sum of [11C]isoprene emissions that accumulated within the trap over time. The transit time was measured as the time from the detector's indication of the initial tracer pulse (seen through the shielding from the leaf chamber and gas lines) to the half-maximum time in [11C]isoprene emission, and reflected a combination of leaf physico-chemical processes as well as the delay due to gas flow (which was less than 1 min for the system's plumbing). We made no attempt to disentangle these aspects, but assumed that the system's features were constant since the same flow rates were used throughout all the studies.
Integrated [11C]isoprene emission after treatment was expressed as a fraction of the emission from that plant before treatment (baseline). Similarly, total carbon isoprene emission after treatment was expressed as a fraction of the emission rate before treatment for each plant, averaging 10 independent GC measurements of isoprene.
There were four experiments. In experiment 1 we compared isoprene emissions from both aspen and birch seedlings representing a high isoprene emitter and a low isoprene emitter, respectively. In experiment 2 we measured isoprene emissions from a single aspen plant in a series of consecutive tracer runs spanning 4.5 h with slightly increasing leaf temperature. In experiment 3 we measured ‘local’ isoprene emissions from intact aspen leaves of several plants, each treated with JA. In experiment 4 we measured ‘systemic’ emissions from several intact hybrid poplar leaves, to determine the response to JA treatment of another leaf. Seeds of aspen (Populus tremuloides) and of birch (Betula papyrifera) were sown 2 years prior to experiments. The seedlings were coppiced in the winter prior to experiments, and the stems that sprouted were used in experiments 1, 2 and 3. For experiment 4, hybrid poplar (Populus nigra clone NC5271) cuttings were dipped in 0.1% indole-3-butyric acid (TakeRoot; Schultz Co., Bridgeton, MO, USA) and planted in fafards ♯2 potting soil in 2.83 L ‘Tall-one’ tree-pots (Stuewe & Sons, Corvallis, OR, USA). NC5271 clones were used for experiments approximately 45–60 d after sprouting, when stems were approximately 0.4 m in height.