Connecting stomatal development and physiology


Stomatal regulation of leaf gas exchange is a complex optimization process (Cowan & Farquhar, 1977; Medlyn et al., 2011; Buckley & Schymanski, 2013), but the operational limits are set during leaf development through the unique pattern of stomata in the epidermis. It is this pattern that determines the maximum stomatal conductance to CO2 or water vapour, gsmax. With changing environmental conditions, such as the onset of a drier climate or increasing atmospheric CO2 concentration, gsmax must be reset in new leaves (or new species) by adjustment of the number and size of stomata. Much remains unknown about the mechanisms and constraints of this process. In this issue of New Phytologist, two articles by Dow et al. (2014a, pp. 1218–1226; 2014b, pp. 1205–1217) apply recent advances in the molecular genetics of stomatal development in combination with a physical diffusion model of gsmax to reveal a direct connection between the physiology of diffusion through stomatal pores and the role of genes that regulate the developmental steps leading to stomatal pattern.

‘This is the first direct evidence that the one-cell spacing rule is crucially important to the gas exchange capacity of leaves.’

More than a century ago Brown & Escombe (1900) devised the foundations of a physical model equating the geometry of pores in a perforated membrane with their conductance to the diffusion of gasses. With little modification this model remains ideally suited for directly calculating stomatal conductance from the dimensions of stomata (Franks & Farquhar, 2001; de Boer et al., 2012). A key element of this conductance model is the effect of concentration ‘shells’ at either end of the stomatal pore (Fig. 1). Represented visually as a series of ellipsoidal regions of equal gas concentration, the shells define a steep concentration gradient perpendicular to the flow of gas molecules along their path through the stomatal pore. They act as if to lengthen the diffusion pathway through the stomatal pores, reducing the rate of flow, and are accounted for in the model as an ‘end correction’ added to the actual stomatal pore depth. When stomata are spaced adequately then diffusion through each stoma behaves as if there were no neighbouring stomata and, using the model, the stomatal conductance on a per unit leaf area basis is obtained by multiplying the average conductance of a single stoma by the stomatal density (D, number of stomata per unit leaf area). If the spacing between stomata is less than several stomatal pore diameters then the concentration shells around each stoma begin to interfere with those of neighbouring stomata and the model cannot be applied in the usual way to calculate conductance. Inadequate spacing of stomata is suboptimal, but how is this quantified, and how do plants overcome this limitation under strong selection pressure for high stomatal density?

Figure 1.

Adequate spacing of stomata prevents interaction of gas concentration shells that form at either end of the stomatal pore. Shown in two-dimensional format is a diagrammatic representation of the concentration shells (blue lines, adapted from Brown & Escombe, 1900) superimposed on a cross-section of a single open stoma from the angiosperm Tradescantia virginiana. The ellipsoid concentration shells at either end of the stomatal pore are accounted for in the ‘end correction’ added to the length of the flow path through the stomatal pore.

Over the last two decades the extensive molecular and genetic resources available for the model plant Arabidopsis thaliana have enabled researchers to identify a number of the genes and signalling pathways that regulate stomatal development and pattern. One of the earliest and most significant findings was the isolation through a mutant screen of the too many mouths (tmm) mutant, which showed abnormal clustering of stomata (increased occurrence of two or more stomata in contact with each other). Cloning of TMM revealed it encoded a receptor-like protein (Nadeau & Sack, 2002), providing some of the first evidence that a signalling mechanism was in place to inhibit the development of adjacent stomata. Further work since has revealed that TMM acts together with members of the ERECTA family of Leucine Rich Repeat Receptor Like Kinases (LRR-RLK), which includes ERECTA, ERECTA-LIKE1 and ERECTA-LIKE2 (Shpak et al., 2005). These LRR-RLK are receptors for members of the EPIDERMAL-PATTERNING FACTOR LIKE (EPFL) family of peptides (Lee et al., 2012; Torii, 2012) and together these receptors and peptide ligands act to regulate both the entry into the stomatal lineage and the spacing of stomata. The former process occurs via an asymmetric division that requires the activity of both the transcription factor SPEECHLESS (SPCH) and the plant specific protein BREAKING OF AYSMMETRY IN THE STOMATAL LINEAGE (BASL; MacAlister et al., 2007; Dong et al., 2009). Reductions in SPCH lead to fewer cells entering the stomatal lineage and subsequent reductions in stomatal density (complete knockout results in an epidermis devoid of stomata). In wild-type plants, the signalling mechanism acts to both correctly orient cell divisions and, following the formation of a stomatal lineage cell, inhibit neighbouring cells from entering the stomatal lineage. Therefore, individual or combinatorial mutations in TMM, BASL, EPF1, EPF2 and the ERECTA family result in stomatal clustering defects of varying severity whereby two and sometimes several stomata develop adjacent to one another with no intervening epidermal cells. Environmental variables such as light intensity, atmospheric CO2 concentration and water availability are known to influence stomatal density (Casson & Gray, 2008). However, some of the genes involved in these responses exhibit significant pleiotropic effects (Masle et al., 2005), so it is difficult to isolate model developmental systems and genetic mutants for studying stomatal function while minimizing nonstomatal side effects. As a result of recent breakthroughs in stomatal developmental genetics a comprehensive and versatile molecular tool kit is now available to alter stomatal patterning whilst maintaining other leaf traits, enabling researchers to explore the role of stomatal development in the optimization of leaf gas exchange.

The ‘one-cell spacing’ rule, whereby stomata are usually separated by at least one epidermal cell, is a fundamental characteristic of stomatal development (Sachs, 1991). This has long intrigued developmental biologists as an example of cellular ‘spacing pattern’ control, but until now there has been no quantitative assessment of its physiological importance. In a novel approach, Dow et al. (2014b) exploit the manipulation of stomatal patterning genes to violate the spacing rule and produce genotypes with increased clustering over a range of stomatal densities. Stomatal functioning in this high-clustering group was compared with a second set of genotypes spanning a similar range of stomatal densities but with very low clustering. Amongst the low-clustering genotypes, under conditions known to induce the widest possible stomatal apertures, the maximum stomatal conductance measured with an infrared gas analyser (Diffusive gsmax) closely approximates the theoretical maximum calculated from stomatal anatomy with the conductance model (Anatomical gsmax). However, in genotypes with high clustering the Diffusive gsmax was substantially lower than Anatomical gsmax, despite similar stomatal densities, indicating that clustering severely compromised stomatal conductance. This is the first direct evidence that the one-cell spacing rule is crucially important to the gas exchange capacity of leaves.

The full extent of the disruption caused by stomatal clustering is unknown. In addition to its effect on diffusion shells there are obvious impediments to the functioning of the stomatal guard cells themselves. In many species, particularly angiosperms, the stomatal pore is created by the guard cells bending apart from each other and displacing the walls of adjoining epidermal cells (Fig. 1). With high clustering, stomata are pushing against one another and neither can open properly. Similarly, to generate their high osmotic pressures guard cells take up considerable quantities of potassium from adjoining epidermal cells, so high clustering would diminish access to this resource and impede opening. Dow et al. discuss these and other potential limitations that clustering places on the normal operation of stomata but a full understanding awaits further investigation.

Across tmm, basl, tmm;basl double and tmm;erl1;erl2 triple mutants, clustering appears to be more prevalent with increasing stomatal density (Fig. 2a). However, higher stomatal density is required to increase gsmax, so unless clustering is avoided the loss of diffusive capacity will be greater at higher Anatomical gsmax (Fig. 2b). It therefore seems that developmental plasticity in stomatal density coevolved with a suite of genes and signalling mechanisms that control against stomatal clustering. Another solution to reduce clustering at higher densities is to produce smaller stomata, and indeed part of the reason that wild A. thaliana ecotypes exhibit relatively high gas exchange capacities and high stomatal densities is that they have small stomata. But smaller stomata requires smaller guard cell nuclei and therefore smaller plant genome size (Franks et al., 2012), so even with developmental measures in place to prevent clustering, leaf gas exchange capacity is again ultimately constrained by cellular dimensions.

Figure 2.

Breakdown of the one-cell spacing rule leads to increased clustering and severely compromised gas exchange capacity. (a) Arabidopsis genotypes comprising individual mutations in the stomatal patterning genes TMM or BASL, or combinatorial mutations in TMM and BASL, or TMM, ERL1 and ERL2, all exhibit substantial stomatal clustering. Amongst this group, clustering appears to increase with stomatal density (data taken from table 1 in Dow et al., 2014b, in this issue of New Phytologist pp. 1205–1217). (b) The impact of stomatal clustering on the maximum measured stomatal conductance, Diffusive gsmax. For pairs of mutants with similar stomatal density but either high or low clustering, Diffusive gsmax of the high clustering mutant is plotted against that of the low clustering mutant (black squares with fitted red curve). The gap between Diffusive gsmax in high clustering vs low clustering mutants (vertical distance between the black 1:1 line and the red curve) widens with increasing Diffusive gsmax.

Dow et al. (2014b) have uncovered the physiological significance of the one-cell spacing rule, but what is the significance of gsmax itself? What is gained by an elaborate developmental mechanism for altering stomatal pattern and gsmax? Given the sophisticated mechanism of stomatal aperture control, why not a ‘one size fits all’ model, whereby gsmax is fixed and optimal gas exchange is achieved entirely through aperture control? In their companion paper Dow et al. (2014a) again apply the powerful combination of molecular genetics and the diffusive conductance model, this time with a suite of Arabidopsis genotypes that have mutations or transgene activity resulting in altered stomatal density (without clustering). Remarkably, in response to short-term changes in CO2 concentration, all genotypes displayed the same change in stomatal conductance relative to their respective Anatomical gsmax despite a five-fold range of stomatal density. From this Dow and colleagues conclude that the short-term response to CO2 is specific to the individual stoma and simply scales with stomatal density. We now know also that this beautifully efficient and versatile system is crucially dependent on the genetic mechanisms that prevent clustering as stomatal density changes (Fig. 3). With the right developmental controls in place, plasticity in leaf gas exchange capacity is more effectively achieved through altered stomatal density. This allows individual stomata to operate within a narrow range of optimal guard cell turgor sensitivity across a wide range of gsmax (Franks et al., 2012). In a sense this is a one size fits all model, except that the stoma as a functional unit is conserved while gsmax and the absolute range of stomatal control are scaled via stomatal density.

Figure 3.

Conservation of the stoma as the functional unit across a range of maximum stomatal conductance (gsmax) relies upon the genetic infrastructure that prevents clustering during leaf development. The top two panels represent the gradation from low gsmax to high gsmax via increased stomatal density in plants with intact developmental control against clustering. The bottom two panels represent the same gradation in genotypes with diminished control over stomatal clustering. Increased clustering decreases gsmax. Drawings of stomata (dark green shapes) and epidermal cells (light green with blue outline) are adapted from fig. 1 in Dow et al. (2014b, in this issue of New Phytologist pp. 1205–1217).

Based on the observation that stomata tend to operate at a fixed proportion of gsmax, Dow et al. propose that this operating point should be predictable from basic information about stomatal anatomy and their environment. To demonstrate this they apply a simple modification of the widely successful Ball–Woodrow–Berry equation, substituting the CO2 assimilation term for gsmax. This gives an expression for the operating stomatal conductance in terms gsmax multiplied by an environmental sensitivity factor, the ratio of relative humidity to CO2 concentration (see eqn 3 in Dow et al., 2014a). Dow et al. are quick to emphasize that this new empirical model, which reliably simulates their gas exchange data, does not capture all the conditions a plant experiences in nature. However, it does capture the integration of stomatal development, physiology and evolution in an elegantly simple and powerfully applicable form. This, in combination with new molecular tools for manipulation of stomatal properties, lays the foundations for a broader understanding of optimal plant adaptation to environmental change across all timescales.