Abscisic acid (ABA) modulates the activities of three major classes of ion channels-inward-and outwardrectifying K+ channels (IK,i and IK,out, respectively) and anion channels-at the guard-cell plasma membrane to achieve a net efflux of osmotica and stomatal closure. Disruption of ABA sensitivity in wilty abil-1 mutants ofArabidopsis and evidence that this gene encodes a protein phosphatase suggest that protein (de-)phosphorylation contributes to guard-cell transport control by ABA. To pinpoint the role of ABIl, the abil-1 dominant mutant allele was stably transformed into Nicotiana benthamiana and its influence on IK,ing, IK,out, and the anion channels was monitored in guard cells under voltage clamp. Compared with guard cells from wild-type and vectortransformed control plants, expression of the abil-l gene was associated with 2-to 6-fold reductions in IK,out and an insensitivity of both hc,in and IK,out to 20 ,uM ABA. In contrast, no differences between control and abil-1 transgenic plants were observed in the anion current or its response to ABA. Parallel measurements of intracellular pH (pHJ) using the fluorescent dye 2',7'-bis(2-carboxyethyl)-5-(and -6)-carboxyfluorescein (BCECF) in every case showed a 0.15-to 0.2-pH-unit alkalinization in ABA, demonstrating that the transgene was without effect on the pH; signal that mediates in ABA-evoked K+ channel control. In guard cells from the abil-l transformants, normal sensitivity ofboth K+ channels to and stomatal closure in ABA was recovered in the presence of 100 ,iM H7 and 0.5 ,iM staurosporine, both broad-range protein kinase antagonists. These results demonstrate an aberrant K+ channel behavior-including channel insensitivity to ABAdependent alkalinization of pHr-as a major consequence of abil-l action and implicate ABIl as part of a phosphatase/kinase pathway that modulates the sensitivity of guard-cell K+ channels to ABA-evoked signal cascades.Abscisic acid (ABA) participates in the growth and development of higher plants by controlling seed maturation and dormancy and signals conditions of water stress in vegetative tissues (1, 2). The hormone accumulates in the leaves during drought stress, evoking stomatal closure to reduce transpirational water loss (1). Stomatal closure is achieved through osmotic solute efflux from the guard cells, notably of K+ from the vacuole and cytoplasm across the plasma membrane, and the consequent decline in guard-cell turgor. ABA initiates this flux by modulating three classes of ion channels-inward-and outward-rectifying K+ channels (Iyi,n and I,.out, respectively) and anion channels-events coordinated by at least two independent signal cascades that lead to a rise in cytoplasmic free [Ca2+] ([Ca2+]) and intracellular pH (pHi) (1, 3, 4).Protein phosphorylation has also been implicated from studies of the ABA-insensitive 1 (abil) mutant that interferes with ABA responsiveness inArabidopsis thaliana. Mutations at this locus evoke a range of ABA-related phenotypes, notably aberrant control of stomatal aperture and ...

The eects of elevated CO 2 concentrations on stomatal movement, anion-and K + -channel activities were examined in guard cells from epidermal strips of Vicia faba. Membrane voltage was measured using intracellular, double-barrelled microelectrodes and ionchannel currents were recorded under voltage clamp during exposure to media equilibrated with ambient (350 ll á l A1 ), 1000 ll á l A1 and 10 000 ll á l A1 CO 2 in 20% O 2 and 80% N 2 . The addition of 1000 ll á l A1 CO 2 to the bathing solution caused stomata to close with a halftime of approx. 40 min, and with 10 000 ll á l A1 CO 2 closure occurred with a similar time course. Under voltage clamp, exposure to 1000 ll á l A1 and 10 000 ll á l A1 CO 2 resulted in a rapid increase (mean, 1.5 0.2-fold, n = 8; range 1.3-to 2.5-fold) in the magnitude of current carried by outward-rectifying K + channels (I K,out ). The eect of CO 2 on I K,out was essentially complete within 30 s and was independent of clamp voltage, but was associated with 25±40% (mean, 30 4%) decrease in the halftime for current activation. Exposure to CO 2 also resulted in a four-fold increase in background current near the free-running membrane voltage, recorded as the instantaneous current at the start of depolarising and hyperpolarising voltage steps, and a decrease in the magnitude of current carried by inward-rectifying K + channels (I K,in ). The eect of CO 2 on I K,in was generally slower than on I K,out ; it was allied with a transient acceleration of its activation kinetics during the ®rst 60±120 s of treatment; and it was associated with a negative shift in the voltage-sensitivity of gating over a period of 3±5 min. Measurements carried out to isolate the background currents attributable to anion channels (I Cl ), using tetraethylammonium chloride and CsCl, showed that CO 2 also stimulated I Cl and dramatically altered its relaxation kinetics. Within the timeframe of CO 2 action at the membrane, no signi®cant eect was observed on cytosolic pH, measured using the¯uorescent dye 2¢,7¢-bis-(2-carboxyethyl)-5,6-carboxy¯ourescein (BCECF) and ratio¯uorescence microphotometry. These results are broadly consistent with the pattern of guard-cell response to abscisic acid, and indicate that guard cells control both anion and K + channels to achieve net solute loss in CO 2 . By contrast with the eects of abscisic acid, however, the data indicate that CO 2 action is not mediated through changes in cytosolic pH and thereby implicate new and, as yet, unidenti®ed pathway(s) for channel regulation in the guard cells.Abbreviations: ABA = abscisic acid; BCECF = 2¢,7¢-bis-(2-carboxyethyl)-5,6-carboxy¯ourescein; I K,out , I K,in = outward-, inward-rectifying K + channel; pH i = cytosolic pH; TEA = tetraethylammonium chlo-ride; t 1/2 = activation halftime Correspondence to:

To explore the role of auxin-binding protein (ABP1) in planta, a number of transgenic tobacco (Nicotiana tabacum) lines were generated. The wild-type KDEL endoplasmic reticulum targeting signal was mutated to HDEL, another common retention sequence in plants, and to KEQL or KDELGL to compromise its activity. The auxin-binding kinetics of these forms of ABP1 were found to be similar to those of ABP1 purified from maize (Zea mays). To test for a physiological response mediated by auxin, intact guard cells of the transgenic plants were impaled with double-barreled microelectrodes, and auxin-dependent changes in K(+) currents were recorded under voltage clamp. Exogenous auxin affected inwardly and outwardly rectifying K(+) currents in a dose-dependent manner. Auxin sensitivity was markedly enhanced in all plants overexpressing ABP1, irrespective of the form present. Immunogold electron microscopy was used to investigate the localization of ABP1 in the transgenic plants. All forms were detected in the endoplasmic reticulum and the KEQL and KDELGL forms passed further across the Golgi stacks than KDEL and HDEL forms. However, neither electron microscopy nor silver-enhanced immunogold epipolarization microscopy revealed differences in cell surface ABP1 abundance for any of the plants, including control plants, which indicated that overexpression of ABP1 alone was sufficient to confer increased sensitivity to added auxin. Jones et al. ([1998] Science 282: 1114-1117) found increased cell expansion in transgenic plants overexpressing wild-type ABP1. Single cell recordings extend this observation, with the demonstration that the auxin sensitivity of guard cell K(+) currents is mediated, at least in part, by ABP1.