In broad bean (Vicia faba L.), an apoplastic phloem loader, the sucrose concentration increases up to ∼ 2 mM in the leaf apoplast and up to ∼ 150 mM in the guard-cell apoplast during the photoperiod. This high concentration in the guard-cell apoplast results from transpiration and is sufficient osmotically to reduce stomatal aperture size by up to 3 µm or ∼ 25% of the maximum aperture size. In this paper, we investigated a parallel and required role for high bulk-leaf apoplastic sucrose concentration, which correlates with high photosynthesis rate. An empirically determined combination of lowered light intensity and lowered CO2 concentration reduced the photosynthesis rate to nominally one-fifth of the control value without a significant change in transpiration. This reduction in photosynthesis caused the sucrose concentration in the leaf apoplast – the immediate source pool for guard cells – to decrease by 70% (to 0.4 mM). In turn, sucrose concentration in the guard-cell apoplast decreased by ∼ 80% (to ∼ 40 mM). These results complete the required evidence for a non-exclusive, transpiration-linked, photosynthesis-dependent passive mechanism for the modulation of stomatal aperture size. In an ancillary investigation, hexoses in the bulk-leaf apoplast decreased when photosynthesis was lowered, but their concentrations in the guard-cell apoplast of control plants indicated that their osmotic contribution was negligible.
In their seminal work, Mott & Parkhurst (1991) found that transpiration rate is the correlative parameter when stomata respond to ambient relative humidity (RH) (for a meta-analysis, see Monteith 1995). As implied earlier, an increase in the guard-cell apoplastic sucrose concentration is a signal that effectively reflects transpiration rate. As guard-cell apoplastic sucrose is a recent product of photosynthesis in mesophyll cells (Lu et al. 1997), its concentration is hypothesized to correlate with the rate of photosynthesis as well as transpiration. We (Outlaw & De Vlieghere-He 2001) have thus proposed that guard-cell apoplastic sucrose concentration is an integrative signal that inhibits further opening of stomata when photosynthesis and transpiration are high. In this study, we have addressed the hypothesized effects of high bulk-leafapoplastic sucrose concentration on guard-cell apoplastic sucrose accumulation using broad bean, a plant often used for stomatal studies (e.g. Outlaw & De Vlieghere-He 2001; Taylor & Assmann 2001; Ritte & Raschke 2003; Frechilla, Talbott & Zeiger 2004; Roelfsema et al. 2006). Additional investigations assessed the importance of an osmotic role of reducing sugars in the guard-cell apoplast. Importantly, we note the complexity of stomatal regulation (see Discussion) in response to the environment and particularly mechanisms that sense perturbations in water balance [see Buckley (2005) for a recent comprehensive and critical review].
MATERIALS AND METHODS
Broad-bean (Vicia faba L. cv Longpod) seeds were purchased from Harris Seeds (Rochester, NY, USA). Plants were grown in Fafard No. 2 soil-less potting medium (Conrad Fafard, Inc., Agawam, MA, USA) in a growth cabinet (for details, see Ewert et al. 2000). In brief, the cabinet was programmed for a 16 h day [25/20 °C day/night temperature; photosynthetically active radiation (PAR), 500 µmol m−2 s−1 at leaf level] at a constant 60% RH. The first youngest fully expanded pair of leaflets of ∼ 3-week-old plants was used in all experiments.
Leaf photosynthesis rate, leaf temperature and conductance were measured with a CIRAS-1 Portable Photosynthesis System (PP Systems Co., Amesbury, MA, USA).
During the photosynthesis measurement, the control leaf temperature was 24.7 ± 0.2 °C and that of the low-CO2/shaded leaves, 25.2 ± 0.7 °C. Using plants grown similarly, leaf temperature under control conditions was monitored remotely, with complementary results (average = 24.8 °C). These small differences indicate a negligible average of 7% difference in water vapour driving forces between the leaf and the atmosphere under the two conditions.
Aperture sizes were determined microscopically from selected stomata on epidermal peels. The same criteria used for the selection of stomatal aperture size determinations were used for the selection of stomata for histochemical analysis (centrally located between the midrib and leaf edge, typical guard-cell dimensions and aperture size, not near veins). The uniform application of these criteria was to minimize interleaf variation in sucrose content (Outlaw & Manchester 1979) and achieve better concordance between samples for the two types of measurements than conductance would have provided. For control samples, the conductance-to-aperture size relationship was concordant with that reported for randomly sampled stomata (Zhang & Outlaw 2001); for treatments, the conductance was 37 ± 17% lower, compared with controls.
Plant shading and growth-cabinet CO2-concentration manipulation
Broad-bean leaves were shaded with a custom-fabricated shading device. Lattices of transparent plastic monofilament lines (spacing = 1 cm) were mounted in two plastic frames (20 × 10 cm2) that sandwiched the leaf, holding it horizontally. In treatments, neutral-density film (26 cm2) was mounted between the lights and leaves. The light intensity of control leaves was 450–500 µmol m−2 s−1 and that of shaded plants, 120–150 µmol m−2 s−1. Ambient temperature and humidity were maintained as described for growth conditions of all plants.
The growth cabinet CO2 concentration was lowered manually either by adding bottled CO2-free air (80/20, nitrogen/oxygen) or by pumping internal air through a 1 M NaOH solution inside the growth cabinet. The CO2 concentration was monitored with an infrared gas analyser (Model 225 MK3; the Analytical Development Co., Hoddesdon, UK). The growth cabinet CO2 concentration under the low CO2 condition was 350–360 µmol mol−1, 100 µmol mol−1 lower than that of control conditions. Plants were sampled at 1100 h. Control plants were maintained under growth cabinet conditions; at the onset of illumination, treated plants were shaded and the CO2 concentration was lowered gradually.
Quantitative histochemical assay of guard-cell sucrose content
Histochemical procedures for single-cell sucrose analysis were according to Lu et al. (1997). (For a general description of these methods, see Outlaw & Zhang 2001.) In brief, tissue was frozen in liquid-nitrogen slurry and stored at −80 °C until freeze-drying at −35 °C and < 10 µm Hg. Then, guard-cell pairs were individually dissected in a climate-controlled environment and the sucrose content of each pair was measured with oil-well (initial volume = 1 µL) and enzymatic-cycling techniques (0.5–2.5 pmol). Guard cells that were dissected from whole-leaf fragments contained both symplastic and apoplastic sucrose. Guard cells that were dissected from epidermis that had been washed before freezing contained only symplastic sucrose (for details, see Lu et al. 1997). Apoplastic contents were calculated by subtraction; SEs were calculated by an algorithm that provided a maximum apparent variance in the apoplastic pool (see Outlaw & De Vlieghere-He 2001). Unfortunately, the labour-intensive process of individually dissecting and transferring cells placed constraints on the breath of experimental design, permitting pairwise comparisons, but not the measurement of incremental changes.
Bulk-leaf apoplastic sap collection
Petioles of excised leaves were inserted immediately through the sealing grommet of a pressure chamber (Model 1000; PMS Instrument Co., Albany, OR, USA). The use of a special grommet (Protocol C, Ewert et al. 2000) was required because of the succulence and fragility of Vicia petioles. The first droplet to be extruded, ∼ 3 µL (sufficient to purge petiolar tracheary elements), was removed by blotting with tissue paper and discarded. The next aliquot, 8–10 µL, was collected with a restriction pipette and stored at −80 °C until analysis. The maximum pressure exerted was 1.1 MPa. Malic dehydrogenase activity (according to López-Millán et al. 2000) in the bulk-leaf apoplastic sap was < 0.5% of that in the leaf homogenate (mass basis), a determination made to ensure that cells were not ruptured at high pressures.
Guard-cell-sugar extraction for high-performance liquid chromatography (HPLC)
Fragments of broad-bean leaf or epidermis were sampled at 1100 h and were freeze-dried according to Lu et al. (1997). In brief, the tissue was frozen in liquid-nitrogen slurry and stored at −80 °C until freeze-drying at −35 °C and < 10 µm Hg. Then, tissues were stored under vacuum at −20 °C until dissection of guard cells. Guard cells were dissected from washed epidermis (containing symplastic sugars) or from leaf fragments (containing both symplastic and apoplastic sugars) for histochemical analysis. Individually dissected guard-cell pairs were pooled (180 pairs) for extraction in 30 µL of ice-cold water, which was immediately elevated to 95 °C, where it remained for 30 min. The extraction solution was then stored at −20 °C before being used for HPLC analysis of sugars.
HPLC analysis of sugars in bulk-leaf apoplastic sap and guard-cell extracts
Sugars were analysed with a Waters 2695 Alliance Separation Module with a temperature-controlled column chamber and autosampler (Waters Co., Milford, MA, USA). The column used was a 250 × 4.1 mm Hamilton RCX-10 anion exchange HPLC column (Hamilton Co., Reno, Nevada, USA). Samples were injected in a volume of 20 µL and the mobile phase was 150 mM NaOH running at a speed of 1 mL min−1. The detector was an ESA Coulochem II electrochemical detector with a gold electrode (ESA Biosciences, Inc., Chelmsford, MA, USA). The peaks of sugars were identified by comparing their retention times with those of standard sugars with Millenium32 data analysis software from Waters Co. Internal standards to insure recovery (> 90%) were added to guard-cell extracts before injection. The detection limit (P < 0.001) of sugars with this method was 4 pmol (2 × 10−7 M, 20 µL injection). Standard sucrose (Fisher Chemicals, Fairlawn, NJ, USA) and glucose, fructose, stachyose, raffinose and galactinol (Sigma-Aldrich Co., St. Louis, MO, USA) were obtained commercially.
Lowered photosynthesis rate without altering transpiration rate was effected by shading plus decreased ambient CO2 concentration
Bulk-leaf apoplastic sucrose is the intermediate pool between sucrose produced by recent photosynthesis in the mesophyll and that in the guard-cell apoplast in broad bean, a typical apoplastic phloem loader (Lu et al. 1997). We (Lu et al. 1997) hypothesized that a diminished bulk-leaf apoplastic sucrose pool by lowered photosynthesis would limit the size of the guard-cell apoplastic sucrose pool. Here, shading was used to decrease leaf photosynthesis, which did lower the bulk-leaf apoplastic sucrose pool, as expected (e.g. Walsh, Thorpe & Minchin 1998). In addition, to simplify interpretation, stomatal aperture size was maintained in order to distinguish the effects of lowering the leaf photosynthesis rate from lowering the leaf transpiration rate. Thus, after 5 h of shading [photon flux density (PFD) ∼ 30% of control] leaves on intact plants, the stomatal aperture size decreased from 10.2 ± 0.1 to 7.8 ± 0.1 µm (P < 0.001) (cf. columns 1 & 2, Fig. 1). Compensatorily, decreased CO2 concentration eliminated this effect (column 3, Fig. 1): with shading plus lowered ambient CO2 concentration, stomatal aperture size was essentially the same as that of the control (P = 0.49, Fig. 1), whereas leaf photosynthesis rate decreased (P < 0.001) more than sixfold (Fig. 2) from 9.8 ± 0.6 µmol CO2 m−2 s−1 to 1.5 ± 0.2 µmol CO2 m−2 s−1.
Bulk-leaf apoplastic sucrose, glucose and fructose concentrations were lower in low-CO2/shaded than in control plants
The effect of diminished leaf photosynthesis rate at the same stomatal aperture on bulk-leaf apoplastic sugar concentration was assessed as a prerequisite to testing the relationship between photosynthesis rate and sugar accumulation at the guard-cell apoplast. Glucose and fructose were characterized in addition to sucrose because broad-bean bulk-leaf apoplastic sap contains glucose and fructose in addition to the dominant sugar, sucrose (Delrot et al. 1983; Lohaus et al. 2001).
As shown in Fig. 3, the average bulk-leaf apoplastic sucrose concentration was ∼ 1.3 mM in control plants and ∼ 0.4 mM in shaded/low-CO2 plants (P = 0.002). Glucose and fructose concentrations in the bulk-leaf apoplast were less than half (P = 0.009 for glucose and P = 0.02 for fructose) that of sucrose in control plants, and shading/low-CO2 caused these sugars to decrease in concentration also (P = 0.03 for glucose and P = 0.02 for fructose, Fig. 3).
As shown in Fig. 3, the shading caused a significant decrease in bulk-leaf apoplastic sugar concentrations. However, the transpiration rate was not altered because the stomatal aperture size (Fig. 1), leaf temperature, ambient temperature and RH (Materials and Methods) were essentially unchanged. Here, guard-cell apoplastic sugar contents of control and treated plants were measured to assess the effects of lowered leaf apoplastic sugar concentration on sugar accumulation at the guard-cell apoplast.
Guard-cell apoplastic and guard-cell symplastic sucrose, glucose and fructose of control broad-bean plants were assayed by HPLC (Fig. 4). Guard-cell apoplastic glucose (0.11 ± 0.15 pmol pair−1, equivalent to ∼ 25 mM) and fructose (0.08 ± 0.17 pmol pair−1, equivalent to ∼ 20 mM) contents were much lower (P = 0.009 for glucose and P = 0.015 for fructose) than that of sucrose (0.63 ± 0.16 pmol pair−1) (Fig. 4, HPLC method) of control plants. The hexose concentrations in the guard-cell apoplast of control plants were negligible osmotically; in addition, the imprecision in their measurements would not have permitted the detection of a concentration decline or even a modest increase. Therefore, guard-cell apoplastic glucose and fructose contents were not determined in shaded/low-CO2 plants.
Guard-cell apoplastic and guard-cell symplastic sucrose contents of both control and shaded/low-CO2 plants were assayed with quantitative histochemical methods. The guard-cell apoplastic sucrose content (Fig. 4) decreased (P < 0.001) from 0.75 ± 0.1 pmol pair−1 in control plants to 0.16 ± 0.08 pmol pair−1 in treated plants. The decrease correlated with the decrease from 1.3 to 0.4 mM (P = 0.002, Fig. 3) in the bulk-leaf apoplastic sucrose concentration. In contrast, guard-cell symplastic sucrose content was the same in control and shaded/low-CO2 plants (P = 0.16) (histochemical method, Fig. 4). The results from quantitative histochemistry and HPLC agreed. With histochemistry, the sucrose content of the guard-cell symplast was 1.09 ± 0.07 pmol guard-cell pair−1 (n = 43) and that of the apoplast was 0.75 ± 0.1 pmol guard-cell pair−1 (n = 47) of control plants. These values are almost identical to values obtained with HPLC from extracts of pooled guard cells (1.26 ± 0.2 and 0.63 ± 0.2 pmol guard-cell pair−1 for the symplastic and apoplastic compartments, respectively).
A linear relationship (Fig. 5, R2 = 0.80) between bulk-leaf apoplastic sucrose concentrations and corresponding guard-cell apoplastic sucrose contents exists. The highest bulk-leaf apoplastic sucrose concentration was ∼ 2 mM, similar to that of Outlaw & De Vlieghere-He (2001) and Lu et al. (1997). The highest guard-cell apoplastic sucrose content (control plants) was ∼ 0.9 pmol guard-cell pair−1, which is also in the range of previous reports (Lu et al. 1995, 1997; Outlaw & De Vlieghere-He 2001). The guard-cell apoplastic sucrose contents of shaded/low-CO2 plants (0.07–0.25 pmol guard-cell pair−1) are similar to those of broad-bean plants with closed stomata (0.22 pmol guard-cell pair−1, Lu et al. 1995), or with a low leaf transpiration rate (0.2 pmol guard-cell pair−1, Outlaw & De Vlieghere-He 2001).
The evolution of adjustable stomata (Raven 2002) was required in the adaptation of erect plants to a terrestrial environment because they permit the ascent of solute-laden sap and atmospheric CO2 acquisition while isolating the plant's water status from that of the surroundings. Water is typically the limiting resource, and unsurprisingly, the crucial and momentary regulation of stomatal aperture size results from the integration of complex parallel, redundant, overlapping and antagonistic pathways that respond to endogenous and environmental signals. As an example of the complexity, when roots sense dry soil, they convey that information to the shoot via an elevated ABA concentration in the transpiration stream (Davies, Kudoyarova & Hartung 2005), but other means of increasing the ABA concentration in guard cells are via foliar redistribution (Pei & Kuchitsu 2005) and synthesis of ABA by guard cells themselves (Cornish & Zeevaart 1986). Although the molecular identity of ABA receptors is a work in progress (Shen et al. 2006), long-standing evidence for internal and external perception sites exists (Outlaw 2003). Once ABA is perceived, a very specific pattern of cytosolic Ca2+ concentration oscillations is an obligate part of the signalling pathway (Allen et al. 2000) in Arabidopsis under certain conditions, but the requirement for elevation of cytosolic Ca2+ concentration is not absolute (Allan et al. 1994). Further complexity resides in the means by which the Ca2+ signal is generated (Schroeder et al. 2001; Hetherington & Brownlee 2004), how this Ca2+ signal might integrate with the ABA-independent CAS-mediated Ca2+ signal (Pei & Kuchitsu 2005) or the interplay with Ca2+ signals that arise from distinct stimuli, notably, sucrose (Furuichi, Cunningham & Muto 2001) and CO2 (Vavasseur & Raghavendra 2005). Finally, recent work (Xie et al. 2006 ) demonstrates shared components between the ABA network and the stomatal response to humidity. In summary, this sketch emphasizes the plethora of known means by which plants metabolically respond to water insufficiency or perturbation. However, these metabolic responses do not preclude modulation by passive means.
As mentioned, stomata are also responsible for CO2 acquisition and often open at decreased CO2 concentrations (Vavasseur & Raghavendra 2005). Light provides additional opening responses via redundant mechanisms, mediated by phototropins, guard-cell chlorophyll and possibly phytochrome or zeaxanthin (Fan, Zhao & Assmann 2004). Although these particular responses are guard-cell delimited, CO2 and light are the substrates for leaf photosynthesis, implying indirect effects on stomata in situ and opening the possibility of unknown signals that tie photosynthesis to aperture size (Messinger, Buckley & Mott 2006). In the present study, decreased light intensity was used to decrease the leaf apoplastic sucrose concentration (via its effects on leaf photosynthesis). Simultaneously, the CO2 concentration was lowered in order to maintain stomatal aperture size (and, hence, transpiration). Thus, both these stimuli have the same effect on leaf photosynthesis, but opposite effects on stomata. Implicit in the design is the recognition that these stimuli under the experimental conditions override the effects of sucrose accumulation in the guard-cell apoplast on stomatal aperture size.
Essentially, aperture changes result from changes in the osmotic potential difference between the symplast and the apoplast of the subtending guard cells and interactions with other cells (Franks, Cowan & Farquhar 1998). Potassium salts are the predominant fluctuating osmotica, but in the guard-cell symplast, sucrose is also an important osmoticum that supplements or replaces potassium salts and is thus involved in sustaining stomatal opening (e.g. Tallman & Zeiger 1988; Lu et al. 1995; Talbott & Zeiger 1998). In the guard-cell apoplast, enough sucrose may accumulate (Lu et al. 1995) during the photoperiod to diminish stomatal aperture size osmotically by ∼ 3 µm (Lu et al. 1997; Outlaw & De Vlieghere-He 2001) if considered alone. The hypothetical mechanism for guard-cell apoplastic sucrose accumulation in apoplastic phloem loaders (Outlaw 2003) is outlined in the following. Photoassimilated sucrose is released to the bulk-leaf apoplast in the phloem. The transpiration stream sweeps some of this sucrose towards stomata. Water evaporation from or near the guard-cell apoplast deposits the sucrose there, a mechanism modelled with petiolar-fed plasma membrane impermeant 14C-mannitol (Ewert et al. 2000). High concentration of sucrose in the guard-cell apoplast is hypothesized to be one signal that integrates transpiration rate, leaf photosynthesis rate and phloem translocation rate (Lu et al. 1997) in the fine regulation of gas exchange. Of these three hypothetical factors, the role of transpiration rate was proven earlier (Outlaw & De Vlieghere-He 2001). Here, we report a direct robust correlation (Fig. 5) between the concentration of sucrose in the bulk-leaf apoplast and the concentration of sucrose in the guard-cell apoplast at the same nominal transpiration rate, as discussed later. Lowering the photosynthesis rate lowered the bulk-leaf apoplastic sucrose concentration (Fig. 3), and diminishing translocation increases the bulk-leaf apoplastic sucrose concentration (e.g. Krapp, Quick & Stitt 1991; Gottwald et al. 2000). Therefore, the qualitative roles of all three elements of the integrative theory have now been established empirically.
The relationship of bulk-leaf apoplastic sucrose concentration to guard-cell apoplastic sucrose concentration was established by ‘clamping’ transpiration, which is the product of the driving force for water effusion and conductance. Based on leaf and ambient temperatures and constant ambient RH, the driving force was maintained within 7% (see Materials and Methods). Under control conditions, the photosynthesis rate (Fig. 2) was typical for broad-bean plants (e.g. Hariadi & Shabala 2004). Lowering the light intensity and the CO2 concentration lowered the photosynthesis rate (Fig. 2) [and, hence, bulk-leaf apoplastic sucrose concentration (Fig. 3)] and were, in addition, used to maintain stomatal aperture size (Fig. 1). Overall, this strategy was successful: in control plants, the average guard-cell apoplastic sucrose content was 0.75 ± 0.1 pmol guard-cell pair−1[similar to our earlier work (Lu et al. 1995, 1997; Outlaw & De Vlieghere-He 2001)], whereas the correlate value in shaded/low-CO2-treated plants was 0.16 ± 0.08 pmol guard-cell pair−1[similar to that at the onset of illumination (Lu et al. 1995) and that of low-transpiration plants (Outlaw & De Vlieghere-He 2001)]. However, a clear limitation of the present work, imposed by the difficulty of quantitative histochemistry, was the comparison of data from only two conditions. On the other hand, the experimental design (collection of bulk-leaf apoplastic sap from one leaflet and guard cells from the other leaflet of the pair) permitted pairwise comparisons, providing for a strong conclusion under these conditions. Thus, over a nominal fourfold range of leaf apoplastic sucrose concentrations, there was a corresponding (R2 = 0.80) nominal 3.6-fold range in guard-cell apoplastic sucrose content (Fig. 5).
As a means of providing perspective, here and elsewhere (Outlaw & De Vlieghere-He 2001), sucrose concentrations have been calculated from the sucrose contents and a conversion factor, the aqueous volume of the guard-cell apoplast. Note that these calculations provide only imprecise estimates. Although the guard-cell wall volume of embedded tissue of broad bean has been meticulously quantified (Ewert et al. 2000), how much of the wall volume is available to bulk solution and how much is occupied by structural elements are uncertain. Our conversion factor apportions equal volumes to these two elements (see Fry 1988), implying that the calculated concentrations are potentially overestimated by up to twofold or underestimated. Neither of these potentials for error undermines the overall conclusions (Lu et al. 1995, 1997; Outlaw & De Vlieghere-He 2001).
The results reported here (Figs 4 & 5) and elsewhere (Lu et al. 1995, 1997; Outlaw & De Vlieghere-He 2001) provide the basis for passive roles of transpiration and photosynthesis in the modulation of stomatal aperture size in apoplastic phloem loaders. However, other observations require further study for explanation. Firstly, the minimum average calculated concentration of sucrose in the guard-cell apoplast (Fig. 4) was 38 ± 19 mM (cf. ∼ 1.3 mM in the bulk-leaf apoplast, Fig. 3). This minimum guard-cell apoplastic sucrose concentration holds for plants at the end of the normal dark period (Lu et al. 1995) and for plants under minimum transpiration conditions (Outlaw & De Vlieghere-He 2001). Secondly, a large increase in the guard-cell symplastic sucrose concentration occurs in response to an abrupt decrease in transpiration (Outlaw & De Vlieghere-He 2001). Thirdly, an osmotically relevant increase in guard-cell symplastic sucrose concentration occurs towards the end of the photoperiod (Talbott & Zeiger 1996, 1998). So far, as mentioned, there is limited foundation for an explanation of these observations. Kinetic (Reddy & Rama Das 1986; Outlaw 1995; Lu et al. 1997) and molecular bases (Kopka et al. 1997) for guard-cell sucrose transport are known, but signalling and regulation have only begun to be studied in earnest (Kopka et al. 1997; Stadler et al. 2003; Liang et al. 2005). Regardless, the physical model that we proposed (Lu et al. 1997) apparently overrides other underlying mechanisms under the conditions that we studied, which undoubtedly serve other functions, such as homeostasis in sucrose-regulated gene expression (reviewed in Smeekens 2000; Rolland et al. 2002; Rook et al. 2006), and merit study. We note again, for emphasis, that the mechanism discussed here is not exclusive.
Albeit at lower concentrations than that of sucrose, glucose and fructose are major solutes in the bulk-leaf apoplast of broad bean (Fig. 3, Delrot et al. 1983; Lohaus et al. 2001) and both decreased there, though proportionately somewhat less than sucrose, in response to the shading/low-CO2 treatment (Fig. 3). However, glucose and fructose were not detected in the guard-cell apoplast and were not major osmolytes (each < 30 mM) there under control conditions. Therefore, no attempt was made to measure these hexoses under the lowered photosynthesis rate. The low hexose concentration in the guard-cell apoplast is consistent with the absence of invertase there (Hite et al. 1993). Whereas sucrose in the guard-cell apoplast is a product of recent photosynthesis in the mesophyll (Lu et al. 1997), the data are consistent, with the hexoses in the bulk-leaf apoplast, particularly fructose, having a different origin (cf. Figs 3 & 4) such as the transpiration stream (Minchin & McNaughton 1987; Chikov & Bakirova 2004; Chikov et al. 2005), perhaps as a result of phloem leakage (e.g. Ayre, Keller & Turgeon 2003). However, sound conclusions require further study, which is merited because hexoses, like sucrose, regulate gene expression (reviewed in Koch 1996; Pego et al. 2000; Couee et al. 2006; Francis & Halford 2006).
Tianran Jiang, Fanxia Meng, Danielle Sherdan and Guorong Zhang are thanked for their help in conducting the experiments and during the manuscript revision.