Abscisic Acid Induces Rapid Reductions in Mesophyll Conductance to Carbon Dioxide

The rate of photosynthesis (A) of plants exposed to water deficit is a function of stomatal (gs) and mesophyll (gm) conductance determining the availability of CO2 at the site of carboxylation within the chloroplast. Mesophyll conductance often represents the greatest impediment to photosynthetic uptake of CO2, and a crucial determinant of the photosynthetic effects of drought. Abscisic acid (ABA) plays a fundamental role in signalling and co-ordination of plant responses to drought; however, the effect of ABA on gm is not well-defined. Rose, cherry, olive and poplar were exposed to exogenous ABA and their leaf gas exchange parameters recorded over a four hour period. Application with ABA induced reductions in values of A, gs and gm in all four species. Reduced gm occurred within one hour of ABA treatment in three of the four analysed species; indicating that the effect of ABA on gm occurs on a shorter timescale than previously considered. These declines in gm values associated with ABA were not the result of physical changes in leaf properties due to altered turgor affecting movement of CO2, or caused by a reduction in the sub-stomatal concentration of CO2 (Ci). Increased [ABA] likely induces biochemical changes in the properties of the interface between the sub-stomatal air-space and mesophyll layer through the actions of cooporins to regulate the transport of CO2. The results of this study provide further evidence that gm is highly responsive to fluctuations in the external environment, and stress signals such as ABA induce co-ordinated modifications of both gs and gm in the regulation of photosynthesis.


Introduction
The rate of photosynthesis (A) in drought stressed plants is frequently constrained by the availability of carbon dioxide at the site of carboxylation. The concentration of CO 2 within the chloroplast envelope (C c ) is determined by stomatal (g s ) and mesophyll (g m ) conductance [1][2][3]. During water deficit, levels of ABA within the leaf are often enhanced by the transport of ABA from the roots to the shoots [4] and the conversion of 'fixed' glycosylated ABA stored within the vacuole to 'free' ABA in the cytosol [5,6]. Stomatal conductance is negatively correlated to the concentration of ABA in the xylem [7] and leaf [8]. The effect of ABA on stomatal closure may be further enhanced as apoplastic ABA concentrates at the sites of evaporation close to the stomatal pores [9], and pH changes on a cellular level facilitate the movement of ABA to the guard cells [10]. Generally, stomatal and mesophyll conductance operate in tandem, showing marked reductions to drought [2,11,12]. It was shown that this tight coordination was controlled in an ultradian fashion in Eucalyptus citriodora plants subjected to different degrees of water deficit [13]. However, despite transport of CO 2 across the mesophyll layer representing the largest resistance step in the uptake of CO 2 for photosynthesis and a critical component in the drought stress response of plants, the response of g m to ABA is not clear.
Mesophyll conductance is determined by physical [14][15][16] and biochemical [17,18] factors. Any reduction in leaf turgor could alter physical resistances, such as apoplastic space and the porosity of the cell wall encountered by CO 2 during transport from the sub-stomatal air-space to the chloroplast [9,19,20]. It is therefore necessary to distinguish between the physical influence of changes in leaf turgor experienced during drought, and any direct biochemical effects of ABA on g m . Relatively few studies have investigated the effect of ABA on g m , and have produced contrasting results; possibly due to the experimental approaches utilised to modify [ABA], either through increased endogenous production of ABA or exogenous application e. g. [21,22], and the methods employed to measure g m e.g. [23,24,25]. Exogenous application of ABA to hydroponically grown soybean (Glycine max L.) and tobacco (Nicotiana tabaccum L.) resulted in reduced g s and g m values over the course of ten days [3]. Identical reductions of both g s and g m were observed in cut-leaves of drought stressed and well-watered wild-type and ABA deficient mutants of tobacco (Nicotiana plumbaginifolia) two hours after exposure to a solution containing ABA [9]. The addition of ABA to the nutrient solution of 'sand-grown' sunflowers (Helianthus annuus L.) also reduced canopy-level g m after three days [26]. In contrast, perlite grown sunflower plants supplied with an exogenous ABA solution showed reduced g s , but no alteration of g m after three days [25]. Furthermore, no difference was observed in g m values of wild-type and ABA insensitive mutants of Arabidopsis thaliana [27].
The disparity in responses of g m to ABA may be related to interspecific differences in drought stress physiology. For example, concomitant reductions in both g s and g m are frequently observed in response to water-deficit e.g. [2]. However, in drought tolerant plants, persistent long-term water deficit may result in an enhancement of g m in comparison to the early stages of drought, whilst g s remains low [28,29]. One other potential explanation is the methodology employed to gauge g m . It is not possible to measure g m directly, and all approaches involve a number of assumptions and potential uncertainties [23]. Mesophyll conductance can be calculated through simultaneous measurement of leaf gas exchange and chlorophyll fluorescence parameters (the variable J method) [30,31], determination of leaf gas exchange and carbon isotope discrimination [32,33] or analysis of the curve of the photosynthetic response to increased [CO 2 ] in the internal sub-stomatal air-space (C i ) [34]. It has been suggested that stomatal closure during drought stress may alter rates of photosynthetic 'recapture' of CO 2 released during photorespiration, affecting the measurement of levels of respiration in the light (R d ) and thus the determination of g m using the variable J method [23,35]. However, observations of g m sensitivity and insensitivity to ABA have been recorded using both the variable J method [3,9,25] and the carbon isotopic discrimination approach [9,25,26]. This may suggest that it is not possible to fully account for the disparity in g m responses to ABA due to uncertainties with the methodological approach utilised to characterise g m . Furthermore, analysis of the effects of drought on g m levels in rice (Oryza sativa) using both the variable J method and analysis of carbon isotope discrimination in recently synthesised sugars method indicated that the variable J technique was equally effective as the carbon isotopic discrimination approach in gauging g m at low levels of water availability, but crucially produced more robust results under well-watered control conditions [2].
Abscisic acid plays a fundamental role in the drought stress response of plants. Given the wide-range of g m responses to ABA reported in the literature over different timescales, we investigated g m sensitivity to exogenous ABA application in four commercially important woody species. To characterise the direct biochemical response of g m to ABA, measurements were taken over a short time period (four hours) to minimise any physical effects that may be associated with leaf water status that could affect the movement of CO 2 from the sub-stomatal air-space to the chloroplast envelope. This study aimed to: i) investigate the effect of exogenous ABA on stomatal and mesophyll conductance over a four hour time-course; ii) characterise any interspecific variations in the g m response of the four species to ABA; iii) assess diffusive constraints imposed by g s and g m following ABA treatment that determine rates of photosynthesis.

Materials and Methods
Plant material, growing conditions and ABA treatment Cherry (Prunus avium L.), black poplar (Populus nigra L.), olive (Olea europaea L.) and rose (Rosoideae rosa, hybrid tea rose "Camp David") were grown in a greenhouse at the National Research Council, Monterotondo, Rome, Italy. The plants were approximately one year-old and grown in comparatively large 6 dm 3 pots, where they did not experience root-restriction, containing a sand-perlite mixture (1:3) under natural sunlight and photoperiod from June to August. The respective daily maximum and minimum air temperatures were 38 and 20°C. To avoid any water and nutrient limitation, the saplings were watered every other day to pot water capacity and fertilised once a week, with Hoagland nutrient solution to supply nutrients at free access rates [36].
The evening prior to measurement the plants were watered to pot water capacity. The next morning branches were cut under distilled water; control treatment branches remained in distilled water, and cis-trans ABA (99% purity, Sigma) was added to the water of branches subject to ABA treatment. An ABA solution of 10 −4 M concentration was used. Simultaneous measurement of gas exchange and chlorophyll fluorescence was then performed every hour over a four hour period between 08:00 and 12:00 in a well ventilated air-conditioned room at 25°C with control and ABA-treated shoots placed under a metal halide light emitting 800 μmol m -2 s -1 PPFD.

Gas exchange and fluorescence measurements
Leaf gas exchange and fluorescence parameters of the central leaf section were simultaneously measured using a LI-6400-40 leaf chamber fluorometer (Li-Cor, Inc., Nebraska, USA) equipped with a 2 cm 2 cuvette. One branch for the ABA and control treatments was taken from each of four plants, with the youngest fully expanded leaf from four branches measured for each treatment. The measurements were made at a saturating photon flux density (PPFD) of 1600 μmol m -2 s -1 measured using the internal quantum sensor within the leaf chamber and leaf temperature of 25°C for all four species. Leaves were exposed to a contaminant and pollutant free flux of synthetic air, composed of a mixture of nitrogen (80%), O 2 (20%) and CO 2 (385 ppm). The relative humidity of the air flow (500 μmol s-1) was maintained at 40-50%. To reduce diffusion leaks through the chamber gasket [37], a supplementary external chamber gasket composed of the same polymer foam was added to create an interspace between the two gaskets (i.e. a double-gasket design with a 5 mm space separating the internal and external gaskets). Then the CO 2 and H 2 O gradients between the in-chamber air and pre-chamber air were minimized by feeding the IRGA exhaust air into the interspace between the chamber and the pre-chamber gaskets [38].
The variable J method has proven to be effective in determining g m in both well-watered and drought stressed rice varieties [2], we therefore chose this approach to assess the effect of ABA on g m in the four plant species. Mesophyll conductance was calculated using the variable J method involving simultaneous measurements of gas-exchange and chlorophyll fluorescence parameters as described by Harley et al. (30) and Loreto et al. (31) (Eqs 1 and 2): where the electron transport rate (J F ) is calculated from fluorescence [39]: where F m is the fluorescence maximum and the partitioning factor (β) between photosystems I and II was considered to be 0.5 and leaf absorbance (α) (0.85) [40]. The CO 2 compensation point to photorespiration (Γ Ã ) was measured on individual attached leaves of intact plants by increasing C i at four different levels of photosynthetically active radiation [41,42]. Levels of respiration in the light (R d ) were analysed using the Kok method [43,44]; and respiration in the dark (R n ) was measured by switching off the light in the cuvette, when CO 2 release from the leaf had become stable for approximately five to ten minutes this was recorded and considered to represent R n [45]. Values of Γ Ã , R d and R n used in the calculation of g m utilising the variable J method are given in Table 1. Total conductance to CO 2 (g tot ) was calculated as: Leaf water status Immediately following the gas exchange measurements, each leaf was detached and weighed to determine leaf fresh mass (F M ). The leaves were then placed in a plastic bag and with the cutend submerged in distilled water and allowed to rehydrate in darkness at 5°C for 18 hours. After rehydration the leaves were dried using paper towels to remove any water on their surfaces, and then the leaves were weighed to determine the saturated mass (S M ). Leaves were then dried at 80°C for 48 hours to measure dry mass (D M ). The relative water content (RWC) of each leaf was then calculated as follows [46]: Table 1. Values of the CO 2 compensation point to photorespiration (Γ*), respiration in the light (R d ) and respiration in the dark (R n ) used to calculate g m levels of the four plant species using the variable J method [30,31]. ± indicates one standard deviation.

Statistical analyses
Statistical analyses were performed using SPSS 20 (IBM, New York, USA). A one-way ANOVA with LSD post-hoc test was used to assess differences in variance between samples subjected to control conditions and ABA treatment. A significant difference between treatments was assumed to occur at a P-value <0.05. Linear regression was used to investigate potential relationships between conductance to CO 2 and A in the four plant species under control conditions and ABA treatment.

Results
Following abscission of the cuttings, the RWC of the leaves was recorded at each hourly measurement interval to assess any alteration in leaf water content that may have influenced g m via a change in turgor. The RWC of control cherry leaves remained within the range of 88.7-94.1% throughout the measurement period (Fig 1a). However, after two hours the mean RWC of leaves exposed to exogenous ABA increased from 90.9% to 95.7%; a significant difference between the RWC values of the control and ABA treated leaves only became apparent after four hours of exposure (one-way ANOVA LSD post-hoc test, F 1,6 = 25.954; P = 0.00223). The RWC of rose leaves exhibited a similar pattern, with the mean RWC of ABA treated leaves exhibiting an increase from 94.6% to 97.2% after two hours. However, RWC parameters of control leaves also showed a slight increase during the measurement period over a range of 92.2 to 96.7%. Statistically significant differences were observed in the RWC values of control and ABA treated rose leaves after two hours (one-way ANOVA LSD post-hoc test, F 1,6 = 9.416; P = 0.0220) and this persisted for the remainder of the measurement period (Fig 1d). The RWC of poplar leaves showed no significant decrease between the control and ABA treatments, although RWC rose slightly by 1.6-3.4% over the four hour measurement period (Fig 1c). The RWC values of olive leaves displayed no significant change during the four hour measurement period, and no significant treatment effect associated with the application of ABA (Fig 1b). Exposure to exogenous ABA induced consistent declines in A in all four of the species studied. The control plants showed no declines in A that may have been associated with branch abscission (Fig 2). These reductions in A following exposure to ABA corresponded to identical declines in g s and g m . However, the rate of reduction in g s and g m values differed between the four species. Cherry exhibited a consistent decline in A, g s and g m throughout the measurement period following ABA treatment (Fig 2a-2c). Olive and poplar displayed rapid reductions in g s (olive: -66.7%; poplar-54.8%) and g m (olive: -47.3%; poplar-40.9%) in the first hour after exposure to exogenous ABA, before maintaining relatively stable g s and g m values for the remainder of the measurement period. In contrast, rose does not show a decline in levels of A, g s and g m in the first hour after ABA treatment, before exhibiting a rapid decline in the second hour (A: -44.2%; g s : -53.5%; g m : -44.5%) then stabilising in the third and fourth hour (Fig 2j-2l). The lower levels of conductance to CO 2 and A following four hours of ABA treatment led to decreases in C i of 8.4 to 28.0% in all four plant species. However, this reduced C i was statistically significant only in poplar and rose (Fig 3).
Photosynthetic rates of the four species were largely determined by conductance to CO 2 (Fig 4). Under control conditions, rose generally exhibited the highest levels of A, g s and g m , while cherry displayed the lowest. However, after four hours of ABA treatment, rose and cherry showed identical photosynthetic rates and levels of conductance to CO 2 . The leaves treated with exogenous ABA exhibited lower levels of g s , g m and g tot that correspond to lower A. The relationship between A and conductance to CO 2 was most robust when g s and g m were combined to determine g tot (Fig 4c).

Discussion
Abscisic acid plays a fundamental role in plant responses to water-deficit. The effect of ABA on g m is poorly characterised in comparison to the influence of ABA on stomatal behaviour e.g. [4,  47,48]. This study demonstrated that exogenous ABA produced a rapid reduction in g m values within two hours of application in all of the four species studied. These declines in g m corresponded to reductions in g s , analogous to co-ordination of g s and g m observed during drought [2,12,49] and as circadian pattern [13]. The reductions in g s following ABA application are indicative of stomatal sensitivity to ABA and active physiological stomatal behaviour in all four of the plant species studied [50,51]. Photosynthesis was closely related to conductance to CO 2 in the control and ABA treated cuttings (Fig 4), suggesting that ABA determined A through its action upon g s and g m . The speed of response of g m , g s and A to ABA application varied between the four species, possibly due to differences in water transport e.g. [52] affecting the uptake and movement of ABA, or biochemical differences in the effect of ABA signalling between the plants e.g. [53,54]. Mesophyll conductance is composed of physical and chemical components that determine the movement of CO 2 from the internal sub-stomatal air-space to inside the chloroplast envelope [55]. Increased RWC in ABA treated cherry and rose (Fig 1), possibly due to stomatal closure reducing transpirative water-loss from the leaf, may have altered the physical properties of the leaf and thus affected g m indirectly. However, in the evening prior to abscission the plants were watered to pot capacity, and all of the leaves exhibited high RWC values that correspond to full leaf turgor [56], likely suggesting that the proportionally small percentage increases of 2-3% in RWC following ABA application did not induce significant alteration of leaf structural properties. It is noteworthy that significant differences in the RWC of control and ABA treated leaves of cherry only became apparent after two hours, whereas g m values in ABA treated leaves declined significantly within one hour (Fig 2c); suggesting that the initial declines in following ABA application were solely the result of biochemical changes, and not due to alteration of the physical properties of leaves associated with a change in foliar water status.
The effect of ABA on g m observed in this study was rapid, occurring within one hour of exposure in three of the four species studied. Short-term fluctuations in g m have also been recorded in response to [CO 2 ] [37], salinity [1,57], light quality [58,59], light intensity [37] and temperature [18], and have been ascribed to the action of carbonic anhydrase and cooporins that transport CO 2 across the plasma membrane [17,[60][61][62][63][64][65]. The conversion of gaseous CO 2 to aqueous carbonic acid (H 2 CO 3 ) represents one of the largest resistance steps encountered in photosynthetic CO 2 -uptake [66]. The results of this study may suggest that ABA acts to reduce the activity of cooporins involved in the transport of CO 2 . Cooporins belong to a group of proteins known as aquaporins, that also facilitate the movement of water across membranes [67]. As foliar [ABA] increased, the activity of aquaporins have been shown to decline during drought stress [68,69], and induce stomatal closure through increased guard cell permeability and reduced water movement across the membrane [70]. Exogenous application of ABA to fully hydrated leaves achieved similar results [71]. However, increased [ABA] may also enhance the expression of certain aquaporins to enhance drought tolerance through increased water transport [72].
It is unclear whether alterations in g m are a by-product of the g s response, or whether the two processes are linked by co-ordinated physiological signalling [73]. Instantaneous increases in C i applied to plants grown under optimal conditions induce reductions in g m , as the limiting effect of CO 2 availability on A declines at higher C i [37]. This would suggest that if the g m responses observed following ABA treatment in the present study were solely the result of reduced C i due to stomatal closure, an increase in g m might be expected. However, reduced g m was recorded in all four species after application of exogenous ABA; furthermore, only two of the species exhibited significant declines in C i after ABA treatment (Fig 3). Application of the same concentration of ABA used in this study to the roots of sunflower did not affect g m but did reduce g s . The relationship between g m and C i in ABA treated sunflower was also identical to control plants; showing a positive correlation between g m and C i at concentrations above 200 ppm [25]. However, detached leaves of wild-type tobacco plants when treated with exogenous ABA exhibited a positive g m -C i relationship at sub-ambient C i , whilst the control counterparts showed a negative g m -C i relationship [73]. Furthermore, g m response to C i has been shown to occur on a shorter timescale than g s response to C i in tobacco, wheat and in both wild-type and mutant Arabidopsis thaliana that lacked the capacity for stomatal closure [74]. This may suggest that the concomitant declines in g s and g m observed in this study following ABA treatment are the result of a shared signalling mechanism regulating rates of stomatal and mesophyll conductance to CO 2 [73] in addition to any subsequent effects on g m associated with stomatal closure [73][74][75], respiration [23] or photorespiration [24].
It is noteworthy that the effect of ABA on g m observed in the present study occurs more rapidly than previously reported e.g. [3,9,25,26]. Increased foliar [ABA] following drought stress induces stomatal closure to restrict water-loss from the internal leaf [76]. It would initially appear incongruous to reduce transport of CO 2 across the mesophyll layer in concert with decreased g s to minimise water-loss from the internal leaf [73]; as a high g m to g s ratio would permit the maintenance of a degree of CO 2 uptake and enhanced water use efficiency during drought [27]. However, the rapid declines in g m after application with exogenous ABA may indicate a selective advantage of reduced conductance to CO 2 across the mesophyll during episodes of water-deficit. A decline in the activity of cooporins responsible for the transport of CO 2 may reduce energy consumption [77]. Higher cellular [78] and apoplastic [79] ABA may also be associated with changes in pH that alter membrane properties through the action of proteins such as aquaporins [80], thus affecting g m [17]. Nonetheless, it is presently unclear as to the nature of the mechanisms responsible for such rapid alterations in g m following ABA treatment or their functional significance. Further analysis of the expression of cooporin protein RNA may elucidate the biochemistry underlying this response e.g. [81]. Application with exogenous ABA induced declines in g m within one hour in three of the species studied (and within two hours in the remaining species, rose); this would suggest that ABA has a clear effect on g m , and the interaction of ABA and g m plays an important role in plant drought stress response.

Conclusion
The results of this study show consistent reductions in A, g s and g m values of all four plants following exposure to exogenous ABA. Photosynthesis in the plant species was positively related to the availability of CO 2 within the chloroplast envelope (Fig 4). The observed declines in g m occurred on a shorter timescale than those reported in previous studies; suggesting that ABA serves to induce rapid reductions in g m following exposure to water deficit. These falls in g m values occurred prior to any significant alteration in RWC; suggesting that they are not associated with physical effects of increased turgor affecting the permeability of cell membranes, and are instead the result of alteration to the biochemical properties of the interface between the mesophyll and the internal sub-stomatal air-space. The reductions in g m recorded in the present study are unlikely to be the by-product of stomatal closure causing a decrease in C i , as plants exposed to ABA exhibited reduced g m rather than increased conductance to counter lower availability of CO 2 in the sub-stomatal air-space. The effect of ABA on g m is likely through the diminished activity of cooporins and carbonic anhydrase, whether these responses act on a dose-dependent basis or occur over different timescales is currently unclear. However, the findings of the present study indicate that ABA functions by inducing rapid reductions in g m that are associated with concomitant declines in g s as part of a co-ordinated gas exchange response to water deficit.