Guard cell photosynthesis is critical for stomatal turgor production, yet does not directly mediate CO2 - and ABA-induced stomatal closing

Abstract

Stomata mediate gas exchange between the inter-cellular spaces of leaves and the atmosphere. CO2 levels in leaves (Ci) are determined by respiration, photosynthesis, stomatal conductance and atmospheric [CO2 ]. [CO2 ] in leaves mediates stomatal movements. The role of guard cell photosynthesis in stomatal conductance responses is a matter of debate, and genetic approaches are needed. We have generated transgenic Arabidopsis plants that are chlorophyll-deficient in guard cells only, expressing a constitutively active chlorophyllase in a guard cell specific enhancer trap line. Our data show that more than 90% of guard cells were chlorophyll-deficient. Interestingly, approximately 45% of stomata had an unusual, previously not-described, morphology of thin-shaped chlorophyll-less stomata. Nevertheless, stomatal size, stomatal index, plant morphology, and whole-leaf photosynthetic parameters (PSII, qP, qN, FV '/FM' ) were comparable with wild-type plants. Time-resolved intact leaf gas-exchange analyses showed a reduction in stomatal conductance and CO2 -assimilation rates of the transgenic plants. Normalization of CO2 responses showed that stomata of transgenic plants respond to [CO2 ] shifts. Detailed stomatal aperture measurements of normal kidney-shaped stomata, which lack chlorophyll, showed stomatal closing responses to [CO2 ] elevation and abscisic acid (ABA), while thin-shaped stomata were continuously closed. Our present findings show that stomatal movement responses to [CO2 ] and ABA are functional in guard cells that lack chlorophyll. These data suggest that guard cell CO2 and ABA signal transduction are not directly modulated by guard cell photosynthesis/electron transport. Moreover, the finding that chlorophyll-less stomata cause a 'deflated' thin-shaped phenotype, suggests that photosynthesis in guard cells is critical for energization and guard cell turgor production.

Keywords: Arabidopsis thaliana; CO 2; abscisic acid; chlorophyll; chlorophyllase; guard cell; photosynthesis; stomata; turgor.

Figure 1. Generation and phenotype of guard cell-targeted chlorophyll-deficient plants

Arabidopsis plants with reduced-chlorophyll levels in guard cells were generated using constitutive guard cell-specific expression of the highly active version of Citrus sinensis Chlorophyllase protein (Harpaz-Saad, et al. 2007). (a) The constitutively active N-terminally-truncated Chlorophyllase, (ChlaseΔN) (Harpaz-Saad, et al. 2007) was sub-cloned to the pHUASGW7 expression vector (pHUAS-ChlaseΔN) and then (b) transformed to an E1728, guard cell specific-expressing enhancer trap line (Gardner, et al. 2009) (GC-ChlaseΔN), which constitutively expresses GFP in guard cells. (c) Representative images of wild-type (WT) and three GC-ChlaseΔN lines (T3 generation) (#4,#5 and #8) grown under ambient CO₂ conditions for 5 weeks.

Figure 2. Chlorophyll fluorescence levels were reduced exclusively in GC-ChlaseΔN guard cells

(a–f) The abaxial epidermis layer of the fifth true rosette leaf was imaged using confocal microscopy. Representative images for wild-type (WT) control E1728 enhancer trap plants (a, c and e) and GC-ChlaseΔN (T3 generation) transgenic plants (b, d and f) are shown. Merged confocal laser microscopic images show Arabidopsis leaf guard cells among epidermal cells. (b, d) Thin-shaped stomata detected in GC-ChlaseΔN lines. An example is marked by white arrow. Panel (a, b): Confocal images of chlorophyll auto-fluorescence (red) overlaid on the corresponding bright-field images. Panel (c–f): confocal images recorded showing chlorophyll auto-fluorescence (red) and GFP fluorescence (green) corresponding to GFP expression specifically in guard cells of the E1728 enhancer trap plants. (g–h) mesophyll layer of the fifth rosette true leaf was imaged using confocal microscopy. Representative confocal images show chlorophyll auto-fluorescence (red) in mesophyll of the E1728 enhancer trap plants (g) and GC-ChlaseΔN (T3 generation) transgenic plants (h).

Figure 3. Expression of chlorophyllase in guard cells did not alter whole leaf chlorophyll levels in GC-ChlaseΔN plants, but produced a large fraction of chlorophyll-less stomata that are thin-shaped

(a) Total chlorophyll per leaf area (leaf disc) was quantified from the fifth rosette leaf of 4 week-old transgenic GC-ChlaseΔN (T3 generation) lines (#4, #5, and #8) and the wild-type (WT) control E1728 enhancer trap line. Data are means ± s.e.m. (n=6; 4 discs from the fifth leaf from 6 different plants). (b–e) The abaxial epidermis layer of the fifth rosette leaf was imaged using confocal (b, c) and DIC (d, e) microscopy. Representative images for wild-type (WT) control E1728 enhancer trap plants (b, d) and GC-ChlaseΔN (T3 generation) transgenic plants (c, e) are shown. Two morphological types of stomata are detected 1) normal kidney-shaped stomata (asterisk) in WT (b, d) and GC-ChlaseΔN plants (c, e) and 2) thin-shaped stomata (arrow) exclusively in GC-ChlaseΔN plants (c, e). (f) Percentage of thin-shaped stomata from the total number of stomata (kidney-shaped and thin-shaped stomata in true leaf). Data are the mean ± s.e.m. (n=4 leaves from independent plants; in each leaf 4 images [x40 magnification] were taken and data were averaged). Unpaired student’s t test between wild-type and GC-ChlaseΔN presented as “*” P < 0.01. Note that the same data shown in panel (f) are presented in figure S2a.

Figure 4. More than 90% of GC-ChlaseΔN guard cells show reduced chlorophyll levels

Total chlorophyll auto-fluorescence of individual guard cells was analyzed from confocal guard cell Z-stack images. (a) The average chlorophyll auto-fluorescence measured for the wild-type (WT) control E1728 plants was designated as 100%. Average chlorophyll auto-fluorescence of GC-ChlaseΔN stomata is depicted as percentages relative to wild-type (WT) controls. Data are the mean ± s.e.m. WT [E1728] n=8 plants, total ~ 400 stomata; line #4 n=5 plants, total 122 stomata; line #5 n=5 plants, total 400 stomata; line #8 n=5 plants, total 168 stomata. Unpaired student’s t test between wild-type and GC-ChlaseΔN presented as “*” P<0.01. (b) Scatter plot of the relative chlorophyll auto-fluorescence per stoma of representative plants, where each circle represents the chlorophyll fluorescence of a guard cell relative to the average of all WT guard cells measured. Gray circles represent kidney-shaped stomata and white circles represent thin-shaped stomata. Horizontal lines depict the mean and standard deviation for each plant. Note that the same data shown in (b) are included in figure S1b.

Figure 5. Guard cell chlorophyll-deficient plants display reduced stomatal conductance and respond to [CO₂] shifts

Time-resolved stomatal conductance responses and net carbon assimilation rates at the imposed [CO₂] shifts (bottom in p.p.m.) in wild-type (WT) and in three independent GC-ChlaseΔN (T3 generation) lines (#4, #5, and #8) were analyzed using intact whole leaf gas exchange. (a, f) Stomatal conductance in mol H₂O m⁻²s⁻¹. (b, g) Data shown in (a, f) were normalized to the last 30 sec time-point of 450 ppm [CO₂] exposure. (c, h) Net carbon assimilation rates (μmol CO₂m⁻²s⁻¹). (d, i) Intrinsic water use efficiency (iWUE). No difference was observed between wild-type and GC-ChlaseΔN transgenic plants with respect to chlorophyll fluorescent parameters, in leaves that were pre-adapted at 150 μmol m⁻² s⁻¹ photosynthetically active radiation (PAR), under ambient CO₂ levels. (e, j) The corresponding inter-cellular [CO₂] (Ci) levels are shown. (k) Maximum efficiency of photosystem II (PSII) photochemistry at 150μE, F_v′/F_m′. (l) PSII operating efficiency (ΦPSII). (m) non-photochemical quenching (qN). (n) Photochemical quenching (qP). Data in (a–n) are the mean of n = 3 plants ± s.e.m. (k–n) Unpaired student’s t test between wild-type and GC-ChlaseΔN show no significant differences.

Figure 6. Kidney-shaped stomata of the GC-ChlaseΔN lines respond to [CO₂] shifts, while thin-shaped stomata are continuously closed

Stomatal apertures in response to [CO₂] changes were measured in wild-type (WT) and GC-ChlaseΔN plants. Intact plants were grown for 4 weeks in a growth chamber under ambient CO₂ levels. Plants were than exposed to 200 ppm CO₂, in a CO₂-regulated growth chamber, for 2 hours and then stomatal apertures in the fifth leaf were directly imaged and analyzed. Plants that were pre-incubated for 2h in 200 ppm CO₂ were subsequently transferred to an 800-ppm CO₂ chamber for an additional 45 min and then stomatal apertures of the fifth leaf were directly analyzed. (a–c) Stomatal aperture measurements of WT and GC-ChlaseΔN (line #4, #5 and #8) in response to CO₂ changes. In each experiment (a–c) n=4 plants, total ~ 120 stomata were measured for each genotype and treatment (200/800 ppm CO₂). Data represent means ± s.e.m. ([CO₂] treatment blind experiments). Pairwise student’s t test between the different CO₂ treatments is presented above columns in each graph (n.s. = non-significant).

Figure 7. Kidney-shaped chlorophyll-less stomata respond to CO₂ shifts

Stomatal apertures in response to [CO₂] changes were measured in WT from kidney-shaped stomata, while in the transgenic (GC-ChlaseΔN) plants stomatal apertures were measured solely from kidney-shaped stomata that did not have a measurable chlorophyll-fluorescence signal. Plants were exposed to 200 ppm CO₂, in a CO₂-regulated growth chamber, for 2 hours and then stomatal apertures in the fifth leaf were sampled and analyzed. Plants that were pre-incubated for 2h in 200 ppm CO₂ were subsequently transferred to an 800 ppm-CO₂ chamber for an additional 90 min and then stomatal apertures in the fifth leaf were analyzed. Stomata in epidermal peels were imaged using both bright field and a red chlorophyll-fluorescence filter, in order to detect the chlorophyll-less stomata in the transgenic lines. (a–c) The average and the corresponding scatter plot of the data are presented for each line, where each circle represents one stomatal aperture measured. Black circles represent stomatal apertures after 200 ppm and gray circles after 800 ppm CO₂ induction. In each experiment (a–c) n=3–4 plants, total ~ 45 stomata were measured for each treatment (200/800 ppm CO₂). Column and scatter plot horizontal lines represent means ± s.e.m. (CO₂ treatment blind analyses). Pairwise student’s t tests between the same line under different CO₂ treatments and between the different lines under the same treatment are presented in each graph. Note that the same data are presented in figure S4.

Figure 8. Kidney-shaped chlorophyll-less stomata respond to ABA

Stomatal apertures in response to ABA were measured in WT from kidney-shaped stomata, while in the transgenic (GC-ChlaseΔN) plants stomatal apertures were measured solely from kidney-shaped stomata that did not have any measurable chlorophyll-fluorescence signal. (a, b) Average stomatal apertures and (c, d) the corresponding scatter plot of the data are presented for each line, where each circle represents one stomatal aperture measured. Circles represent stomatal apertures measured after 1 h exposure to 0.1% ETOH (black circles) or 10 uM ABA (gray circles). (a, c) Line 8, n = 8 plants, total ~ 240 stomata were measured for each treatment and genotype. (b, d) Line 4, n = 3 plants, total ~ 110 stomata were measured for each treatment and genotype. Error bars represent means ± s.e.m. (ABA treatment blind analyses). Pairwise student’s t tests between the same line under different treatments are presented in each graph. 5^th true leaves were floated on opening buffer for 2 h under 120 μE white light to pre-open stomata. Leaves were then exposed to 10 uM ABA or 0.1% ETOH for 1 h. Stomata in epidermal peels were imaged using both bright field and a red chlorophyll-fluorescence filter, in order to detect the chlorophyll-less stomata in the transgenic lines.