(PI3 kinase inhibitors) prevented the bicarbonate orABA-induced stomatal closure (Fig.4). These com-pounds however did not aVect the stomatal closurecaused by 10 M H2O2.
Patterns of H2O2 production in guard cells
The levels of H2O2 in guard cells were monitored byusing the cell permeable Xuorescent dye, H2DCF-DA.The presence of bicarbonate or ABA enhanced mark-edly H2O2 levels in guard cells (Fig.5). The amount of
Fig.2
Promotion of stomatal closure by H2O2 and its reversibil-ity upon H2O2 removal. Stomatal apertures in abaxial epidermisof Arabidopsis were determined at 1h intervals during 2h ofH2O2 application (+H2O2), followed by 2h after H2O2 removal(¡H2O2). Values are means of 100 measurements§SE
when H2O2 was removed from the medium. This indi-cates that at high concentrations, H2O2 may inducedamage. In contrast, up to a concentration of 10¡5M,the eVects of H2O2 on stomata were fully reversible.Therefore, in all further experiments 10¡5M, H2O2 wasused. The stomatal closure induced by bicarbonate wasslightly enhanced in the presence of H2O2 (Fig.3).
Catalase (CAT, an H2O2 scavenger) or DPI (aninhibitor of NAD(P)H oxidase and subsequently ofH2O2 production) are expected to reduce the levels ofH2O2. The presence of CAT or DPI prevented signiW-cantly, but only partially, the stomatal closure causedby both bicarbonate (Fig.3) and ABA (data notshown). The presence of wortmannin or LY 294002
Fig.3
The eVect of DPI (20 M), catalase
(100UmL¡1) and H2O2
(10¡5M) on stomatal closure induced by either 2mM bicar-bonate in abaxial epidermis of Columbia wild (a) or Atrboh D/F mutant (b) of Arabidop-sis
. Results are the averages (§SE) from at least three experiments
Fig.4The eVect of 10 M wortmannin (WM) or 100 M LY294002 or 10¡5M H2O2 on stomatal closure induced by 2mMHCO3 or 10 M ABA in the abaxial epidermis of Arabidopsis.Values are the means§SE from three experiments
123