Microscopy and Cell Counts

For both Arabidopsis and barley, stomatal and epidermal cell counts were taken from the abaxial surface of mature, fully expanded leaves or cotyledons. Cell counts were taken from the widest section of the first true leaf avoiding the mid vein. Dental resin (Coltene Whaledent) was applied in the region of maximum leaf width and left to set before removing the leaf and applying clear nail varnish to the resin. Stomatal counts were determined from nail varnish impressions by light microscopy (Olympus BX51). Five areas per leaf were sampled from four to eight plants of each genotype and treatment. For epidermal imaging (Fig. 2B–D), mature leaves were excised and the central vein of the leaf cut away. Leaf tissue was then serially dehydrated in ethanol. Samples were then placed into modified Clarke’s solution (4:1 ethanol to glacial acetic acid solution) then cleared in 50% bleach overnight.

For epidermal phenotyping, the second fully expanded mature leaf of seedlings was excised and a 3- to 5-cm strip midway along the proximodistal axis of these leaves were cut out. These leaf samples were then submerged in Clarke’s solution (3:1 ethanol to glacial acetic acid solution). Following 1 h of vacuum infiltration, the samples were left in Clarke’s solution for 24 h for fixation. Once fixed, the samples were transferred into 100% ethanol. Prior to imaging, the leaf samples were cleared in 50% bleach solution overnight. The midrib of each sample was then excised and the remaining leaf sections mounted in deionized water on microscope slides for imaging. Samples were viewed by light microscopy (Olympus BX51) using differential interference contrast functionality. For confocal microscopy (Figs. 4, A and B), barley samples were prepared as described (Wuyts et al., 2010) and viewed on an Olympus FV1000 using a 20× UPlan S-Apo N.A. 0.75 objective, 543-nm laser, 555- to 655-nm emission, and Fluorview software.