The discovery that the plant hormone abscisic acid (ABA) had a pronounced antitranspirant effect when supplied to detached leaves (Mittelheuser & van Stevenick, 1969) initiated extensive research activity aimed at discovering its mode of action, which continues to this day. Successive generations of researchers have aimed to determine the biochemical and genetic mechanism(s) regulating its biosynthesis, transport and metabolism, its binding to receptors and their location, and its downstream effects on intracellular signal transduction, gene expression and cellular ionic regulation in diverse cell types. Much of this work has used stomatal guard cells as a model, in view of the over-arching importance of ABA in regulating plant water loss, as exemplified by the higher transpiration rates and lower water status of ABA-deficient mutants compared with wild-type (WT) plants. An enduring controversy has been the relative importance of the roles of tissue water relations (hydraulic effects) and ABA status (chemical signalling) in regulating transpiration (Christmann et al., 2007). Thus various experimental techniques have attempted to separate these effects (e.g. by growing ABA-deficient and WT plants at different relative humidities), even though stomatal response to ABA depends on tissue water status (Tardieu & Davies, 1992). Recent work of Pantin et al., in this issue of New Phytologist (pp. 65–72), further indicates the futility and artificiality of this paradigmatic separation. Their work apparently indicates that ABA not only directly mediates stomatal closure (chemical signalling), but also systemically decreases leaf hydraulic conductance (Kleaf) upstream of the stomata.
Whole plants and detached leaves of various ABA-insensitive Arabidopsis mutants (abi1-1, abi2-1, ost2-1, ost2-2, slac1-1), when either sprayed with an ABA solution (400 μM) or supplied with ABA (50 μM) via the transpiration stream, showed partial stomatal closure (Pantin et al.). Detached leaves of these mutants showed a 36–85% decrease in stomatal conductance (in some cases equivalent to the WT response of an 83% decrease) in response to xylem-supplied ABA. Furthermore, transpiration of intact abi1-1, abi2-1, ost2-2 plants was decreased by 22–36% following a foliar ABA spray (400 μM), a similar response to WT plants. Previous work (discussed in Pantin et al.) indicates that supplying ABA to epidermal strips of the same mutants was unable to elicit stomatal closure, indicating a disparity in the cellular and tissue responses to ABA. Measurements in detached leaves indicated that xylem-supplied ABA (50 μM) decreased leaf hydraulic conductance (Kleaf – calculated as the ratio of transpiration rate and the water potential gradient between solution and leaf, which in detached leaves simplifies to the ratio of transpiration rate to leaf water potential) of both WT and ost2-2 plants by c. 35% (Pantin et al.), suggesting dual hydraulic and chemical mechanisms by which ABA could elicit stomatal closure.
Another recent report corroborates this decrease in Kleaf in response to feeding ABA to detached leaves. When WT Arabidopsis leaves were xylem-supplied with 10 μM ABA, Kleaf decreased by c. 50%, yet foliar application of the same ABA concentration (while decreasing transpiration rate) had no statistically significant effect on detached leaf Kleaf (Shatil-Cohen et al., 2011). This disparity highlighted the role of bundle sheath cells (which enclose dead, conductive xylem vessels) in sensing long-distance ABA signals. In contrast to previous reports where incubation of WT leaf protoplasts in 1 μM ABA for 3 h had no significant effect on osmotic water permeability (Morillon & Chrispeels, 2001), incubating bundle sheath protoplasts in 1 μM ABA for 1 h decreased osmotic water permeability by c. 40% (Shatil-Cohen et al., 2011). Reconciliation of these results in the same study would confirm the intriguing possibility that ionic regulation of different cell types is differentially responsive to ABA, which would interact with spatial differences in ABA concentration observed in different cell types (Christmann et al., 2007) in regulating hydraulic responses to ABA.
‘Do root and shoot tissues show completely opposing responses to ABA?’
However, reports of ABA-induced declines in hydraulic conductance apparently contradict the stimulation of root cell (Hose et al., 2000) and root system (Beaudette et al., 2007) hydraulic conductivity (Lp) by exogenously supplied ABA and in ABA overproducing transgenics (Thompson et al., 2007). Such work indicated that ABA acted at the plasmalemma to stimulate symplastic water transport across the root cylinder (Hose et al., 2000), which has more recently been correlated with enhanced activity and expression of aquaporin proteins. Do root and shoot tissues show completely opposing responses to ABA? In trying to reconcile these apparently opposing effects of exogenous ABA on root and shoot hydraulic conductance, the percentage change in response was plotted against ABA concentration for several literature reports (Fig. 1a), and compared with the dose–response curve for the effects of xylem-supplied ABA on detached leaf transpiration (Fig. 1b). Interestingly, early reports of ABA suppression of Lp (Markhart et al., 1979) are consistent with more recent reports of ABA suppression of Kleaf, with such effects noted when at least 10 μM ABA is supplied.
While limitations of space prevent a more complete consideration of the sensitivity of transpiration to xylem-supplied ABA in detached leaves, it seems that stomatal closure may be more sensitive than decreases in hydraulic conductance (cf. Fig. 1a,b). However, Kleaf and transpiration are directly related as defined earlier. Furthermore, in all calculations of Kleaf that rely on measuring leaf water potential (Ψleaf) and transpiration rate of detached leaves, ABA feeding has not statistically changed Ψleaf (Shatil-Cohen et al., 2011; Pantin et al.) irrespective of how Ψleaf was measured. In such cases, effects of ABA on Kleaf (hydraulic) and transpiration rate (chemical) are confounded. Instead, measurements of Kleaf using techniques that do not rely on quantifying detached leaf transpiration rate (Sack et al., 2002), in plants of contrasting ABA status (Thompson et al., 2007), would provide direct evidence that xylem-supplied ABA does indeed act as a regulator of Kleaf.
Historically, the role of xylem-supplied ABA in mediating stomatal closure in response to soil water deficit has been extensively questioned, since in vivo xylem ABA concentrations did not always elicit stomatal closure when fed to detached leaves (see discussion in Zhang & Davies, 1991). Observations that xylem sap pH altered the distribution of ABA between apoplastic and symplastic compartments such that xylem alkalization enhanced apoplastic ABA concentrations (Wilkinson & Davies, 1997) apparently accounted for drought-induced stomatal closure before any increase in xylem ABA concentration. An alternative view arising from the work of Pantin et al. is that synergistic effects of xylem-supplied ABA (in the nanomolar range) on both Kleaf and transpiration create a feedforward response eliciting stomatal closure, perhaps via transient changes in leaf water status sensitizing guard cells to existing ABA concentrations (Tardieu & Davies, 1992).
By slowing the rate of water loss, these changes are likely to be of real adaptive significance when plants are exposed to prolonged periods of soil drying inducing massive ABA accumulation. To what extent these changes are readily reversible is also a key issue for plants growing in environments with fluctuating soil water availability, since xylem ABA concentrations can remain elevated (and stomata remain partially closed) for several days after soil water status returns to normal (Correia & Pereira, 1994). Whether the phenomenon described by Pantin et al. is amenable to genetic or environmental regulation, via crop breeding or agronomic manipulations, is a key unanswered question, which may assist in improving crop water use efficiency.