Stomatal Opening Involves Polar, Not Radial, Stiffening Of Guard Cells

Abstract

It has long been accepted that differential radial thickening of guard cells plays an important role in the turgor-driven shape changes required for stomatal pore opening to occur [1-4]. This textbook description derives from an original interpretation of structure rather than measurement of mechanical properties. Here we show, using atomic force microscopy, that although mature guard cells display a radial gradient of stiffness, this is not present in immature guard cells, yet young stomata show a normal opening response. Finite element modeling supports the experimental observation that radial stiffening plays a very limited role in stomatal opening. In addition, our analysis reveals an unexpected stiffening of the polar regions of the stomata complexes, both in Arabidopsis and other plants, suggesting a widespread occurrence. Combined experimental data (analysis of guard cell wall epitopes and treatment of tissue with cell wall digesting enzymes, coupled with bioassay of guard cell function) plus modeling lead us to propose that polar stiffening reflects a mechanical, pectin-based pinning down of the guard cell ends, which restricts increase of stomatal complex length during opening. This is predicted to lead to an improved response sensitivity of stomatal aperture movement with respect to change of turgor pressure. Our results provide new insight into the mechanics of stomatal function, both negating an established view of the importance of radial thickening and providing evidence for a significant role for polar stiffening. Improved stomatal performance via altered cell-wall-mediated mechanics is likely to be of evolutionary and agronomic significance.

Keywords: Arabidopsis; atomic force microscopy; cell wall; computational modelling; mechanics; stomata.

Graphical abstract

Figure 1

Stomata Show Stage-Dependent Patterns of Modulus (A) Force map of a leaf epidermis showing the spatial pattern of E_a. Stomata (indicated by asterisks) at different stages of differentiation are distributed across the epidermis and show different patterns of E_a, indicated by relative signal value (yellow, high; red/black, low). (B) Bioassays of young and mature stomata indicate that they both respond to low CO₂ by increasing pore area and to high CO₂ by decreasing pore area. Single asterisk indicates significant difference p < 0.01, n > 23; double asterisk indicates significant difference p < 0.001, n > 23 (ANOVA was performed on “young” or “mature” datasets, followed by a Tukey test). Error bars indicate SEM. (C) Force map of a guard mother cell (GMC) showing the symmetrical cross wall separating the two daughter guard cells. (D) Distribution of E_a across the diameter (as shown in schematic) of the GMC shown in (C). Three peaks of E_a of similar value are detected, corresponding to the three walls of the GMC. (E) Distribution of E_a around the circumference of the GMC shown in (C), with the start point at the equator (as shown in the schematic). A series of peaks of E_a are observed. (F) Force map of a young stomata consisting of two separated guard cells. (G) Distribution of E_a across the diameter of the stomatal complex shown in (F). Four peaks are detected, corresponding to the pairs of walls defining the guard cells. The maximum peak value is similar for all four walls. (H) Distribution of E_a around the circumference of the stomatal complex shown in (F). Two main peaks of E_a are observed at the poles of the stomatal complex. The shoulder on the second peak corresponds to the junction with the epidermal cell on the right-hand guard cell. (I) Force map of a mature stomata consisting of two guard cells. (J) Distribution of E_a across the diameter of the stomatal complex shown in (I). Four peaks are detected, corresponding to the pairs of walls defining the guard cells. The maximum E_a value for the inner radial walls is higher than the peak E_a for the outer radial walls. (K) Distribution of E_a around the circumference of the stomatal complex shown in (I). Two main peaks of E_a are observed at the poles of the stomatal complex. The minor third peak corresponds to the junction with the epidermal cell on the right-hand guard cell. Representative images and analyses of young (F–H) and mature (I–K) guard cells are shown. Force maps were obtained from a total of 14 young and 18 mature guard cells. Scale bars in (A), (C), (F), and (I), 10 μm. See also Figures S1 and S2.

Figure 2

Finite Element Modeling Indicates Only a Minor Role for Radial Stiffening in Stomatal Function but Demonstrates that Fixing Stomatal Poles Has a Major Influence on Aperture Response to Change of Turgor Pressure (A) Cross-sections through guard cells modeled using the baseline parameters (circular cross-section and uniform wall thickness) or the variable wall thickness (VWT) model in which a rounded triangular geometry leads to differential inner wall thickness. The cell wall is modeled as an anisotropic material, parameterized by cellulose micro-fibrils embedded in an isotropic matrix. The micro-fibrils are oriented circumferentially in all models. (B) Modeled relationship of stomatal aperture to guard cell turgor pressure. In the baseline model (purple), aperture increases as pressure increases above about 1.3 MPa, reaching a maximum value as pressure exceeds 5 MPa. Both epidermal and guard cell turgor are increased initially (gray area) after which only guard cell turgor increases. Modification of the model to include a VWT (shown in A) leads to a slight alteration in curve shape (green) so that opening occurs at a slightly lower turgor pressure and the maximal aperture attained is slightly lower. (C) Exploration of the VWT model by increasing or decreasing the inner (ventral) wall thickness by 10% indicates essentially no outcome on the aperture/pressure response curve (lines superimposed). (D) Modification of the baseline model (purple) so that the poles of the guard cells are fixed to prevent stomatal complex elongation leads to a modified output curve (blue) in which pore opening occurs at a lower turgor pressure and the final aperture attained is larger than the baseline model. (E) Effective Lagrange strain for the inside of a guard cell modeled using the baseline parameters. The colored scale indicates the range of strain calculated in different regions of the cell, with a gradient of strain occurring across the cell radius with the inner radial wall having a high strain. (F) Effective stress pattern in the guard cell modeled in (E). A radial stress pattern is generated with high stress at points along the inner radial wall. (G) As in (E) but with VWT parameters used in the model. A decreased strain gradient occurs across the cell. (H) Effective stress pattern in the guard cell modeled in (G). The stress pattern observed in (F) is dissipated so that less extreme gradients are formed, with maximal stress occurring in a medial region. (I) As in (E) but with the stomatal poles fixed (as in D). The pattern is modified from (E) so that high strain gradients form in localized regions toward the guard cell poles. (J) Effective stress pattern in the guard cell modeled in (I). Steeper stress gradients now form toward the guard cell poles compared to (F). In (E), (G), and (I) the strain is dimensionless and is capped at 1 for consistency across the figures. Only the regions immediately neighboring the point at which the pore adjoins the polar wall exceed this value. Strain is a dimensionless tensor describing the deformation of the material, which in simple cases is defined as length change per length. Stress is a tensor which characterizes the internal forces within a material as force per area. In (F), (H), and (J) the unit of stress is MPa, where 1 Pa = 1 Nm⁻².

Figure 3

Measured Change in Stomatal Dimensions during Opening and Closing Supports a Fixed Position of the Stomatal Poles (A) Modeled change in stomatal complex length with increase in guard cell pressure predicts a gradual increase in length at pressures above 1 MPa, both for the baseline (purple) and the VWT model (green), whereas the fixed pole model imposes a constant complex length (blue). (B) Measured complex length in mature stomata triggered to close by elevated CO₂ (red), open by depleted CO₂ (green), or incubated under ambient CO₂ levels (blue). Complex length does not overtly change relative to pore width. Regression analysis was used to calculate the line indicated but is supported with only a low confidence value (p = 0.354, n = 360), suggesting a very limited relationship of complex length and pore width. (C) Measured pore length in mature stomata triggered to close by elevated CO₂ (red), open by depleted CO₂ (green), or incubated under ambient CO₂ levels (blue). Pore length increases with pore width. Regression analysis was used to calculate the line indicated, which is supported with p < 0.0001 (n = 360), suggesting a close relationship of pore length and pore width. Note that the size parameters used for the model are based on those from the literature for Vicia faba, thus the absolute magnitudes of stomatal complex length are greater in (A) than in (B).

Figure 4

Polar Cell Wall Structure Plays a Role in Stiffening and Stomatal Function (A) Labeling of stomata with the COS⁴⁸⁸ probe reveals a high level of signal (green) at the stomatal poles (left). Treatment of tissue with polygalacturonase (4 hr) leads to loss of COS⁴⁸⁸ binding (right). (B) Bioassays after pre-treatment with buffer (control), cellulose, or polygalacturonase (PGase) indicate that stomata retain the ability to open in response to low CO₂ after all treatments, but the stomatal aperture attained after PGase treatment is significantly smaller, both at ambient and low CO₂, relative to the control. ANOVA was performed across all samples with post hoc Tukey. Columns indicated with the same letter cannot be distinguished from each other at the 0.05 confidence limit (n = 40). Error bars indicate SEM. (C) Force map of epidermis from a control sample showing the spatial pattern of E_a after 4 hr incubation of tissue in buffer. Relative signal value is indicated by high (yellow) to low (red/black). (D) Distribution of E_a across the diameter (as shown in schematic in Figure 1) of the stomata indicated by asterisk in (C). Four peaks are detected, corresponding to the pairs of walls defining the guard cells. The maximum peak value for the inner radial walls (peaks 2 and 3) is higher than the peak value for the outer radial walls (peaks 1 and 4). (E) Distribution of E_a around the circumference (as shown in schematic in Figure 1) of the stomatal complex shown in (C). Two main peaks of E_a are observed at the poles of the stomatal complex. (F) Force map of epidermis showing the spatial pattern of E_a after 4 hr incubation of tissue in cellulase. Relative signal value is indicated by high (yellow) to low (red/black). (G) Distribution of E_a across the diameter of the stomata indicated by asterisk in (F). Four peaks are detected, corresponding to the pairs of walls defining the guard cells. The maximum E_a for the inner radial walls is higher than the peak value for the outer radial walls. (H) Distribution of E_a around the circumference of the stomatal complex shown in (F). Two main peaks of E_a are observed at the poles of the stomatal complex. (I) Force map of epidermis showing the spatial pattern of E_a after 4 hr incubation of tissue in polygalacturonase. Relative signal value is indicated by high (yellow) to low (red/black). (J) Distribution of E_a across the diameter of the stomata indicated by asterisk in (I). Two broad, asymmetric peaks are detected, with the highest values at the inner radial walls of the two guard cells. The peaks corresponding to the outer radial wall (peaks 1 and 4) are only barely detectable. (K) Distribution of E_a around the circumference of the stomatal complex shown in (I). Two main peaks of E_a are observed at the poles of the stomatal complex. The E_a value of these peaks is lower than those observed in (E) and (H). Representative images and analysis are shown for control (C–E), cellulase (F–H), and polygalacturonase-treated tissue (I–K). The analyses were repeated at least three times with similar results (data shown in Figure S4). Scale bars in (A), (C), (F), and (I), 10 μm. See also Figures S3 and S4.