Stomatal responses to air pollutants are complex, varying among species and with concentration, environmental conditions and age. In general, short-term exposure to sulphur dioxide (SO2) promotes stomatal opening, whereas longer-term exposure can cause partial stomatal closure. By contrast, the effects of oxides of nitrogen (NOx) are often small or insignificant. The effects of ozone, and oxidative stress, are equally complex. Short-term exposure to ozone stimulates a rapid reduction in stomatal aperture, whilst longer-term exposure causes stomatal responses to become sluggish. The response of stomata to abscisic acid (ABA) has been shown to be slower in plants exposed to a combination of SO2 and NO2 suggesting an adverse effect on guard cell ABA signal transduction. In addition, ozone causes a reduction in stomatal closure under drought conditions. There is an increasing body of evidence to suggest that air pollutants and oxidative stresses can have a marked effect on the Ca2+ homeostasis of guard cells and the intracellular machinery responsible for stomatal movements. Here we discuss the effects of air pollutants on stomatal responses and their possible effects on Ca2+ based signalling in guard cells focusing on the effects of ozone and oxidative stress.
Most of the common anthropogenic air pollutants (e.g. SO2, oxides of nitrogen (NOx) and ozone) are natural components of the earth’s atmosphere. However, there is now abundant evidence that many of these atmospheric contaminants can profoundly affect stomatal control mechanisms even at very low concentrations. The impacts of air pollutants on stomata are complex and vary depending on a wide range of factors such as the period of exposure, concentration of the pollutant, species, age and interactions with additional environmental factors including other pollutants. The same pollutant can often cause stomatal opening or closure under different circumstances, with frequently unpredictable dose-response relationships. Stomatal responses to air pollutants have been extensively reviewed (Darrall, 1989; Saxe, 1990; Wellburn, 1994; Mansfield, 1998; Robinson et al., 1998). The effects of SO2, NOx (NO, NO2) and ozone are summarized in Table 1.
Table 1. A simplified summary of the effects of sulphur dioxide (SO2), oxides of nitrogen (NOx) and ozone on stomata on the basis of period of exposure and/or concentration
Short-term (low [SO2])
Long-term (high [SO2])
Small/insignificant effects (Reduced opening at high [NO])
Promotion of closure
Combinations of gases
Effects observed at [lower]
There is evidence that air pollutants affect stomata both indirectly and directly. Several studies suggest that low concentrations of air pollutants may influence stomata indirectly through damage to the subsidiary cells, or the epidermal cells that fulfill this role when there are no anatomically distinct subsidiary cells (Black & Black, 1979; Neighbour et al., 1988) or through changes in cell wall structure within the stomatal apparatus (Maier-Maercker & Koch, 1986; Maier-Maercker, 1989). The resultant reduction in the mechanical resistance towards the guard cells may result in wider stomatal apertures. This may play an important role in stomatal opening observed in response to short-term exposure low/concentrations of SO2.
There is also strong evidence to suggest that air pollutants influence stomata directly through their affect on stomatal guard cells. Atkinson et al. (1991) have shown that exposure of spring barley to low concentrations of SO2 and NO2 (24–35 ppb of each gas) in combination (see below for a discussion of the effects of combinations of air pollutants) had very little effect on stomatal opening. Nevertheless, the stomata of leaves that had been exposed to the pollutants closed more slowly, and also less completely than controls in response to ABA. Pearson & Mansfield (1993) have obtained comparable results to these in well-watered and droughted beech (Fagus sylvatica) exposed to realistic concentrations of ozone (up to a maximum of 120 ppb). In well-watered trees, leaves of the first flush displayed a decrease in stomatal conductance in response to ozone as expected. Withholding water resulted in partial stomatal closure in all trees. However, the extent of the closure was much smaller in trees that had been exposed to ozone. The results of these studies and others (Darrall, 1989; Saxe, 1990; Wellburn, 1994; Mansfield, 1998; Robinson et al., 1998) suggest that air pollutants have the potential to influence stomata directly by reducing the ability of guard cells to respond to stimuli such as ABA.
Ozone- and oxidative stress-induced increases in [Ca2+]cyt
Several strands of evidence point towards the effects of ozone on the calcium homeostasis of plants and calcium-based signalling in guard cells. Clayton et al. (1999) have used Arabidopsis and tobacco seedlings expressing the calcium reporting protein aequorin to study the effects of ozone on [Ca2+]cyt in plants. They showed that realistic concentrations of ozone (35–200 ppb) induce an increase in whole-seedling [Ca2+]cyt (Fig. 1). The ozone-induced increase in [Ca2+]cyt, or calcium signature (McAinsh & Hetherington, 1998), was biphasic consisting of two ‘peaks’: an initial spike followed by a more prolonged elevation. The second peak was observed only at concentrations of ozone of 70 ppb or higher. The magnitude of the ozone-calcium signature was highly dose-dependent; the height and area of both the first and second peaks increased with increasing concentrations of ozone. Clayton et al. (1999) went on to perform a functional dissection of the ozone-calcium signature. They showed that the second of the two peaks could be abolished by either pretreating plants with LaCl3 or by reducing the period of the fumigation, and that this inhibited the induction of a gene for the antioxidant enzyme glutathione S-transferase (GST) by ozone. Therefore, they concluded that the second peak of the ozone-calcium signature was required for induction by ozone of GST expression. However, they were not able to assign a role to the first peak of the ozone-calcium signature although they speculated that each peak might be associated with different cell types (masked by the constitutive expression of the calcium reporter gene) and associated with different cell-specific responses.
It is tempting to suggest that the first peak of the ozone-calcium signature may be involved in the response of stomata to ozone. However, is there any support for this hypothesis? Several studies provide indirect evidence that ozone has the potential to cause an increase in guard cell [Ca2+]cyt resulting in stomatal closure. Fink (1991) has examined the impact of ozone on the distribution of calcium oxalate crystals in the needles of Norway spruce (Picea abies). Crystals were restricted to the epidermal walls of needles in control trees. By contrast, large deposits of calcium oxalate were observed growing in the vacuole of the epidermal, subsidiary cells and guard cells of needles of trees exposed to ozone. This implies a massive flux of calcium into the guard cells in response to ozone with the concomitant disruption of calcium homeostasis and hence calcium-based signalling. X-ray microanalysis studies provide further evidence of a link between ozone-induced increases in guard cell [Ca2+]cyt and ozone-induced stomatal closure. Le Thiec et al. (1994) have reported a decrease in stomatal conductance in Norway spruce (Picea abies) exposed to ozone which correlates with a marked increase in the calcium content of guard cells. In a more recent study, De Silva et al. (2001) have shown a similar decrease in stomatal conductance associated with an increase in the guard cell calcium content in the calcicole Leontodon hispidus in response to 100 ppb ozone.
More direct evidence for the effects of ozone on guard cell [Ca2+]cyt can be drawn from studies looking at the effect of oxidative stress on [Ca2+]cyt. Price et al. (1994) have examined the effects of H2O2, which is one of the reactive oxygen species generated when ozone enters solution (Heath, 1994; Wellburn, 1994), on [Ca2+]cyt in aequorin-transformed tobacco. They showed that H2O2 stimulates an increase in whole-seedling [Ca2+]cyt the magnitude of which was highly dose-dependent. We have subsequently shown that H2O2-calcium signature is biphasic with similar kinetics to the ozone-calcium signature (K. A. North et al., unpublished).
McAinsh et al. (1996) have examined the effects of H2O2 on stomata. They showed H2O2 reversibly inhibits stomatal opening and promotes stomatal closure in Commelina communis (Fig. 2). This effect is highly dose-dependent and increases with increasing concentrations of H2O2. It was also demonstrated that H2O2 induces increases in guard cell [Ca2+]cyt using the fluorescent calcium-reporting indicator fura-2 microinjected into individual guard cells. Importantly, the H2O2-induced increase in guard cell [Ca2+]cyt and the H2O2-induced inhibition of stomatal opening were both inhibited by the calcium chelator EGTA suggesting a causal relationship between the two. H2O2-induced stomatal closure and increases in guard cell [Ca2+]cyt have also been observed in Arabidopsis (Pei et al., 2000). Taken together, these data and the results of the indirect studies described previously suggest that ozone can cause an increase in guard cell [Ca2+]cyt. However, the location of ozone-induced increases in [Ca2+]cyt together with the mechanism by which they affect stomata, possibly through calcium-signalling and changes in guard cell calcium homeostasis, remains to be confirmed. We are currently addressing these questions using targeted aequorin (Kiegle et al., 2000) and cameleon techniques (Allen et al., 1999).
The impact of ozone and oxidative stress on guard cell ion channels
It is clear that ozone has the potential to cause an increase in guard cell [Ca2+]cyt which may affect stomata either directly via calcium-based signalling processes or through the disruption of guard cell calcium homeostasis and hence calcium-signalling, compromising the ability of stomata to respond to environmental stimuli. Ozone and oxidative stress may also affect other components of the signal transduction pathways in guard cells, including the ion channels that mediate the fluxes of anions and cations that drive changes in guard cell turgor and hence the size of the stomatal pore. Early evidence for the affect of ozone on ion channel activity came from studies of calcium transport in plasma membrane vesicles from pinto bean (Phaseolus vulgaris) (Castillo & Heath, 1990). Recently, however, Torsethaugen et al. (1999) have been able to demonstrate an effect of ozone on guard cell ion channels in broad bean (Vicia faba) using the patch clamp technique. Treatment of guard cell protoplasts with ozone resulted in inhibition of the plasma membrane inward-rectifying K+ channel in guard cells (K+in) that mediates the K+ uptake required for stomatal opening. This was associated with an ozone-induced inhibition of stomatal opening and the promotion of stomatal closure in both intact leaves and isolated epidermis. The authors suggest that the action of ozone on K+in may be a consequence of ozone-induced increases in guard cell [Ca2+]cyt, since this channel is known to be inhibited by calcium (Schroeder & Hagiwara, 1989), or direct oxidation of the channel protein.
Pei et al. (2000) have also used patch clamping to examine the effects of H2O2 on guard cell ion channels. Using this technique they have identified a calcium-permeable channel in the plasma membrane of Arabidopsis guard cells that is activated by H2O2 (Fig. 3). They have shown that activation of this channel precedes the H2O2-induced increase in guard cell [Ca2+]cyt suggesting a role for this channel in mediating calcium influx into guard cells in response to exposure to H2O2. Interestingly, Pei et al. (2000) have also shown that ABA stimulates H2O2 production in guard cells, monitored using 2,7-dichlorofluorescin diacetate, and activates plasma membrane calcium-permeable channels. Furthermore, blocking H2O2 production using diphenylene iodonium chloride inhibited ABA-induced stomatal closure. In addition, activation of calcium-permeable channels by H2O2 together with H2O2- and ABA-induced stomatal closure were all disrupted by the recessive ABA-insensitive mutant gca2. These data indicate that ABA-induced H2O2 production and H2O2-activated calcium-permeable channels are important components of the signalling pathway by which ABA stimulates stomatal closure. The work of Pei et al. (2000) raises the intriguing possibility that ozone may affect stomata through the action of H2O2, generated when ozone enters solution, on H2O2-activated calcium-permeable channels in the guard cell plasma membrane. This has the potential to activate these channels resulting in an increase in guard cell [Ca2+]cyt. Since H2O2-activated calcium-permeable channels, together with increases in guard cell [Ca2+]cyt (McAinsh et al., 1990), both form a part of the ABA signalling pathway leading to stomatal closure, the activation of these channels by ozone/H2O2 has the potential to disrupt the guard cell ABA calcium signal which would have a marked affect on stomatal responses to ABA.
SO2, NO2 and combinations of air pollutants
As discussed previously, many factors influence the impacts of air pollutants on stomata including interactions between pollutants (Darrall, 1989; Saxe, 1990; Wellburn, 1994; Mansfield, 1998; Robinson et al., 1998). The effects of air pollutants in combination can be more than additive, less than additive or antagonistic (Table 1). However, in general combinations of pollutants act to reduce the concentration at which an effect on stomata can be observed. The interaction of SO2 and NO2 is a good example of this phenomenon. For example, Carlson (1983) has shown that in soybean (Glycine max) exposure to SO2 and NO2 alone (0–600 ppb) resulted in only small reductions in stomatal conductance. However, when plants were exposed to SO2 and NO2 in combination (0–600 ppb of each gas) there was a significant reduction in stomatal conductance. We have examined the cellular basis for the interactive effects of SO2 and NO2 on stomata focusing on plant calcium homeostasis and whole-seedling [Ca2+]cyt using aequorin-transformed Arabidopsis. Exposure to SO2 or NO2 alone (300 ppb) resulted in only small changes in [Ca2+]cyt (L. T. Montgomery et al., unpublished). These ranged from no change (approx. 50%) to transient increases of 100 nM above resting [Ca2+]cyt lasting 2–3 min However, when plants were exposed to SO2 and NO2 in combination (300 ppb of each gas) large increases in [Ca2+]cyt were always observed (Fig. 4). The SO2/NO2-calcium signature was biphasic with similar kinetics to those of the ozone-calcium signature (Clayton et al., 1999) consisting of two ‘peaks’: an initial spike followed by a more prolonged elevation. The first peak ranged from 100 to 400 nM above resting [Ca2+]cyt lasting 1–2 min, whereas the magnitude and duration of the second peak was more variable although it was always present. This suggests that the concentrations of SO2/NO2 used may have been near the threshold concentration required for the appearance of the second peak.
Based on the strong correlation between the magnitude of stomatal responses to SO2 and NO2, alone and in combination, and the kinetics of the SO2-, NO2- and SO2/NO2-calcium signatures, small changes in stomatal conductance reflecting small changes in [Ca2+]cyt and large changes in stomatal conductance reflecting large changes in [Ca2+]cyt, it is tempting to suggest that the observed changes in whole-seedling [Ca2+]cyt are indicative of changes in guard cell [Ca2+]cyt. This would imply a causal relationship between the effect of these two pollutants on guard cell [Ca2+]cyt and stomatal conductance similar to that proposed for ozone. However, direct measurements of the effects of SO2 and NO2 on guard cell [Ca2+]cyt, using targeted aequorin (Kiegle et al., 2000) or cameleons techniques (Allen et al., 1999) together with studies of the effects of SO2 and NO2, alone and in combination, on guard cell ion channels will be required in order to confirm this hypothesis.
The impacts of air pollutants on stomata are complex and reflect both indirect and direct effects of pollutants on guard cells. There is an increasing body of evidence to suggest that pollutants have the potential to affect the signal transduction pathways by which guard cells respond to environmental stimuli and, in particular, calcium-based signalling through alterations in the calcium homeostasis of guard cells. This may reflect alterations in the activity of guard cell ion channels. Nevertheless, there are many questions that remain to be addressed. Do SO2-, NO2- and ozone-induced changes in whole-seedling [Ca2+]cyt reflect changes in guard cell [Ca2+]cyt as has been hypothesized? What are the mechanisms by which pollutant-calcium signatures are generated and do they encode stimulus (pollutant)-specific signalling information? How are the signalling elements both upstream and downstream of calcium affected by air pollutants? Are changes in calcium-signalling in guard cells a key factor in the response of stomata to air pollutants or a secondary effect? The combination of whole-plant, cellular and molecular approaches that are now available for studying stomata (Schroeder et al., 2001) will be a powerful tool in answering many of these outstanding questions.
N.H.E. and L.T.M. were funded by grants from the Biotechnology and Biological Sciences Research Council (U.K.). K.A.N. is in receipt of a studentship from the Natural Environmental Research Council (U.K.). M.R.M. is grateful to the Royal Society for the award of a University Research Fellowship.