Expression and manipulation of PHOSPHOENOLPYRUVATE CARBOXYKINASE 1 identifies a role for malate metabolism in stomatal closure


  • Steven Penfield,

    1. Department of Biology, Centre for Novel Agricultural Products, University of York, Heslington, York YO10 5DD, UK
    Search for more papers by this author
    • Present address: College of Life and Environmental Sciences, University of Exeter, Geoffrey Pope Building, Stocker Road, Exeter EX4 4QD, UK.

  • Sarah Clements,

    1. Department of Molecular Biology and Biotechnology, University of Sheffield, Sheffield S10 2TN, UK
    Search for more papers by this author
  • Karen J. Bailey,

    1. Department of Molecular Biology and Biotechnology, University of Sheffield, Sheffield S10 2TN, UK
    Search for more papers by this author
  • Alison D. Gilday,

    1. Department of Biology, Centre for Novel Agricultural Products, University of York, Heslington, York YO10 5DD, UK
    Search for more papers by this author
  • Richard C. Leegood,

    1. Department of Animal and Plant Sciences, University of Sheffield, Sheffield S10 2TN, UK
    Search for more papers by this author
  • Julie E. Gray,

    1. Department of Molecular Biology and Biotechnology, University of Sheffield, Sheffield S10 2TN, UK
    Search for more papers by this author
  • Ian A. Graham

    Corresponding author
    1. Department of Biology, Centre for Novel Agricultural Products, University of York, Heslington, York YO10 5DD, UK
      (e-mail; e-mail; e-mail
    Search for more papers by this author

(e-mail; e-mail; e-mail


Malate, along with potassium and chloride ions, is an important solute for maintaining turgor pressure during stomatal opening. Although malate is exported from guard cells during stomatal closure, there is controversy as to whether malate is also metabolised. We provide evidence that phosphoenolpyruvate carboxykinase (PEPCK), an enzyme involved in malate metabolism and gluconeogenesis, is necessary for full stomatal closure in the dark. Analysis of the Arabidopsis PCK1 gene promoter indicated that this PEPCK isoform is specifically expressed in guard cells and trichomes of the leaf. Spatially distinct promoter elements were found to be required for post-germinative, vascular expression and guard cell/trichome expression of PCK1. We show that pck1 mutant plants have reduced drought tolerance, and show increased stomatal conductance and wider stomatal apertures compared with the wild type. During light–dark transients the PEPCK mutant plants show both increased overall stomatal conductance and less responsiveness of the stomata to darkness than the wild type, indicating that stomata get ‘jammed’ in the open position. These results show that malate metabolism is important during dark-induced stomatal closure and that PEPCK is involved in this process.


Stomatal guard cells are among the most studied and best understood of plant cell signalling systems, yet there remain considerable gaps in our knowledge of the processes of stomatal opening and closing. Stomatal opening is achieved by the accumulation of large amounts of solute in the vacuole. A large fraction of this solute is potassium salt (Cl and/or malate, depending on the plant and the conditions; Raschke and Schnabl, 1978), but sucrose is also involved, particularly later in the diurnal cycle (Talbott and Zeiger, 1996). Malate accumulates in guard cells of open stomata compared with guard cells of closed stomata. During opening, malate is synthesised from starch via phosphoenolpyruvate (PEP) carboxylase. During stomatal closure, in response to various environmental signals, the guard cell loses the ability to maintain its high vacuolar solute content and volume, with consequential loss of turgor and closure of the stomatal pore. The process of stomatal closure is the result of massive loss of the accumulated solutes, largely from the vacuole. This occurs both via transport across the tonoplast and plasmalemma, and probably via metabolism (Figure 1). We are particularly ignorant of the roles that malate metabolism plays during stomatal closure, as opposed to the export of osmolytes (such as K+, Cl, malate and sucrose) from the guard cells to surrounding epidermal cells. There is good evidence for malate efflux during stomatal closure (Van Kirk and Rashke, 1978; Negi et al., 2008; Vahisalu et al., 2008; Finkemeier and Sweetlove, 2009; Fernie and Martinoia, 2009; Gruber et al., 2010; Meyer et al., 2010; Sasaki et al., 2010), but the extent and importance of malate metabolism remains unclear. We do not know whether it is respired or converted back to carbohydrates, such as starch, by gluconeogenesis, or which enzymes are involved in these interconversions. Experiments to mis-express NADP malic enzyme in guard cells of maize plants resulted in decreased amounts of malate and increased stomatal conductance (Laporte et al., 2002), and experiments to down-regulate fumarase activity in tomato plants resulted in reduced flux through the tricarboxylic acid (TCA) cycle and defects in stomatal opening (Nunes-Nesi et al., 2007), underlining the correlation between amounts of malate and stomatal aperture control. Understanding the role of malate metabolism during stomatal closure is crucial to understanding how stomata regulate CO2 uptake and water loss by leaves.

Figure 1.

 Simplified diagram of events that may occur in the guard cell during stomatal closure.
Loss of turgor during stomatal closure is correlated with loss of K+, Cl, sucrose and malate. The fate of sucrose during stomatal closure is unknown, but some is possibly converted to starch. Malate could either be lost by efflux across the plasmalemma, or metabolised via decarboxylation [via malic enzyme and/or the Krebs (tricarboxylic acid) cycle] or converted back to starch (or sucrose) via gluconeogenesis. mit, mitochondrion; chl, chloroplast; OAA, oxaloacetate; PPDK, pyruvate Pi dikinase; ME, malic enzyme; PEPCK, phosphoenolpyruvate (PEP) carboxykinase.

Phosponenolpyruvate carboxykinase (PEPCK; EC catalyses the conversion of oxaloacetate (OAA) to PEP, and is a key early step in the gluconeogenic pathway (Leegood and ap Rees, 1978) and as a decarboxylase in C4 photosynthesis (Wingler et al., 1999). Oxaloacetate can be derived directly from malate in the cytosol by the action of malate dehydrogenase. In plants, gluconeogenesis is important for the mobilisation of carbon derived from seed storage reserves, and we have previously shown that Arabidopsis PCK1 (At4g37870) is required for normal seedling growth and utilisation of reserves from the endosperm in Arabidopsis (Rylott et al., 2003; Penfield et al., 2004). PEPCKs are also known to be expressed in many plant tissues, where proposed roles include the regulation of cellular pH and the control of amino acid metabolism (Walker et al., 1999; Delgado-Alvarado et al., 2007). To investigate the role that the Arabidopsis PEPCK enzyme plays in the control of guard cell movements, pck1 mutant plants were characterised. The PCK1 gene was chosen because expression studies (Malone et al., 2007) and available microarray data show that PCK1 has a broad expression pattern which includes leaf tissue, in contrast to PCK2 expression which is absent in the leaf. Here we analyse the tissue- and cell-specific expression of Arabidopsis PCK1 in vegetative tissues and show that PCK1 is highly expressed in several cell types, including stomatal guard cells. Analysis of a promoter deletion series showed promoter elements responsible for post-germinative, vascular and guard cell-specific expression. pck-1 mutant plants were found to have increased stomatal apertures during light-induced stomatal opening and during dark-induced stomatal closure responses. Our results suggest that PEPCK activity is important for controlling stomatal apertures in response to changes in light intensity.


Identification of PCK1 promoter fragments controlling tissue-specific expression

We previously reported that a 2-kb PCK1 promoter fused to GUS was highly expressed in young seedlings (Penfield et al., 2004). Similar experiments have also shown PCK1 expression in leaf veins and hydathodes (Brown et al., 2010). In order to learn more about the regulation and function of PCK1 we constructed a 5′ promoter deletion series with the aim of identifying regions of the PCK1 promoter responsible for tissue-specific expression patterns (Figure 2a). Decreasing lengths of the PCK1 promoter upstream of the ATG were fused to the GUS transgene (see Experimental Procedures), and up to 30 independent transgenic events were analysed for each construct. In addition, six representative transgenic lines for each promoter fragment found to contain a single T-DNA integration event were propagated to the T3 generation and homozygous lines identified (a total of 66 independent lines) for quantitative analysis. We named these constructs after the length of the putative regulatory genomic sequence upstream of the translation start site, so that the original pPCK1:GUS lines we described (Penfield et al., 2004) were designated pPCK1-2076 (Figure 2).

Figure 2.

 Scheme describing the PCK1 promoter 5-prime deletion series.
(a) Diagram showing the deletion series constructs. Black indicates untranscribed sequences, grey indicates transcribed sequences other than introns (white box). Numbers indicate the start site of each promoter fragment measured in base pairs upstream of the predicted translation start site.
(b) Quantitative analysis of GUS activity from the deletion series. Values represent mean and standard error of six independent homozygous lines.
(c) Spatial analysis of GUS activity in seedlings 3 days after imbibition. One representative seedling from each of the six independent lines from (B) is shown.

In 3-day-old seedlings we found that reducing the length of the promoter fragment to 923 bp upstream of the ATG caused a quantitative reduction in GUS expression so that the pPCK1-923 lines showed a level of expression just less than half of that for the pPCK1-2076 constructs (Figure 2b). However, the tissue-specific expression pattern of GUS in seedlings remained unchanged (Figure 2c). Further deletion of the PCK1 promoter to pPCK-870 caused a large reduction in expression level and also a striking loss of shoot expression, with the exception of the base of the hypocotyl (Figure 2b,c). These data showed that this 53-bp promoter region was required for up-regulation of PCK transcription in the shoot, but also repression of PCK1 expression in root tissues. This same promoter region was also required for PCK1 expression in endosperm tissues (data not shown). The pPCK1-807, pPCK1-750 and pPCK1-674 deletion lines all exhibited an expression pattern resembling the pPCK1-870 lines (Figure 2c) but shorter promoter fragments did not promote GUS expression in seedlings. Therefore we concluded that sequences between −923 and −870 of the PCK1 promoter were important for the post-germinative expression pattern of PCK1, and that these work in tandem with unknown further cis-elements lying further upstream of the transcription start site. Interestingly, the sequence of this region is conserved in the PCK promoters of Brassica species (Figure S1 in Supporting Information).

PCK1 is expressed in stomatal guard cells

Phosphoenolpyruvate carboxykinase has previously been shown to be expressed in many plant tissues, most notably in the vasculature (Walker et al., 2001; Chen et al., 2004; Brown et al., 2010) and reproductive tissues (Famiani et al., 2000; Bahrami et al., 2001; Malone et al., 2007), in which it may function in nitrogen and organic acid metabolism (Famiani et al., 2005; Delgado-Alvarado et al., 2007). To extend our analysis we stained non-seed tissues of pPCK1:GUS expressing plants with the aim of identifying further sites of PCK1 activity (Figure 3). A principal site of PCK1 promoter activity in leaves was the vascular tissue, in agreement with previous PCK1 promoter fusion experiments and studies showing immunolocalisation of PEPCK to both phloem and phloem companion cells (Walker et al., 2001; Chen et al., 2004; Brown et al., 2010). In addition to the vasculature, the Arabidopsis PCK1 promoter was also found to be active in the hydathodes, trichomes (see also Chen et al., 2000; Brown et al., 2010) and stomatal guard cells (Figure 3a–c), as well as in emerging lateral root primordia (Figure 3d). PCK1 is therefore likely to be expressed in a complex tissue-specific manner in young plants. In reproductive tissues the expression of PCK1:GUS was limited to the vasculature (Figure 3e).

Figure 3.

 The expression of phosphoenolpyruvate carboxykinase in non-seed tissues.
(a) whole 14-day-old seedling stained to show expression in the vascular tissues, hydathodes, trichomes and young expanding leaves Scale bar 4 mm.
(b, c) Close-up of leaf illustrating trichome-specific expression Scale bar 1 mm. The small speckles of GUSdf are guard cells, seen at higher magnification in (c). Scale bar = 50 μm.
(d) PCK1:GUS expression in lateral root primordia. Scale bar 0.5 mm.
(e) Close-up of an Arabidopsis flower. The visible GUS expression is limited to the vascular tissue. Scale bar 0.5 mm.

To further investigate the activity of the PCK1 promoter, we stained leaves of our promoter deletion series and analysed the expression in different vegetative tissues and cell types. Due to the complex expression pattern we simply scored independent T1 lines visually and indicated the presence or absence of GUS activity in each tissue in each line (Table 1). In contrast to the situation in seedlings, GUS expression in vascular tissues was found absolutely to require the promoter region between pPCK1-316 and pPCK1-230, although further increasing the promoter length promoted GUS expression in a greater number of T1 lines. Interestingly, hydathode expression also required similar regions of the PCK1 promoter, suggesting that the PCK1 expression in these tissues is controlled by the same regulatory region. The pPCK1-230 to pPCK1-316 region was also important for expression in lateral root primordia.

Table 1.   GUS expression in Arabidopsis vegetative tissues in the PCK1 promoter deletion series. Each line represents an independent transformation event. The frequency of GUS expression is calculated as the number of lines expressing in that tissue divided by the total number of lines tested, expressed as a percentage to one decimal place
PromoterFrequency of GUS expression (%)Number of lines
VasculatureTrichomeLRPGuard cellHydathode
  1. LRP, lateral root primordia.


The situation in guard cells and trichomes was found to be more complex. In both cases we found that reducing the promoter length from −2076 to −1527 all but eliminated expression. However, further reductions in promoter length allowed an increase in the number of T1 lines expressing GUS in guard cells and eventually also trichomes, demonstrating that one or more promoter elements located between −1072 and −760 mediate the repression of PCK1 expression in these cells. Continuing reduction of promoter length to −520 resulted in few if any lines showing guard cell or trichome expression. Therefore, we concluded that expression in trichomes and stomatal guard cells is regulated by multiple sites in the promoter, with two key positive determinants (−1527 to −2076, and −760 to −674), and an area mediating repression, the most important of which was (−1527 to −923). We also found that at least the promoter region upstream of −230 was necessary for expression in any Arabidopsis tissue. This corresponds approximately to the transcription start site used by the majority of full-length cDNAs encoding PCK1 present in the database (Figure 2).

pck1-2 plants have increased susceptibility to drought

We previously reported two T-DNA insertion alleles of PCK1, pck1-1 and pck1-2. Both have reduced PCK activity and exhibit phenotypes consistent with reduced gluconeogenesis during seedling establishment. pck1-2 which is disrupted in the first exon of the PCK1 gene, is the stronger allele. It has a more severe seedling phenotype and extremely low PEPCK activity (Penfield et al., 2004; Brown et al., 2010). In this report we focus on the stomatal responses of the pck1-2 mutant but we also investigate the phenotype of the less severe pck1-1 mutant, which has a disruption in the eighth intron of PCK1, and retains approximately 5% PEPCK activity (Penfield et al., 2004). As might be expected, pck1-1 had a similar but less severe stomatal phenotype than pck1-2 in most of our experiments (Figure S2). Under normal watering conditions, pck1 plants did not appear any different from their background ecotype controls, but after water had been withheld for 1 week, the pck1-2 plants were visibly more affected and less turgid than Col-0 background controls, suggesting that their leaves lose water at an enhanced rate (Figure 4). After 1 week of drought the rosette leaves were harvested and the relative water content of the plants was calculated and found to be significantly lower in pck1-2 than Col-0 controls. There were no significant differences in stomatal densities between pck1 and control plants, indicating that the drought-susceptible phenotype was not due to an increase in stomatal number (Table S1). Thus pck1-2 plants may transpire more water than the wild type, resulting in the quicker onset of drought symptoms.

Figure 4.

pck1-2 plants are sensitive to drought.
(a) Photographs of typical 9-week-old pck1-2 and control (Col-0) plants after 1 week of withholding water.
(b) Histogram showing the relative water content of pck1-2 and Col-0 control plants after 1 week of drought. Values are means of 10 measurements of 10 leaves from three different plants. Error bars represent the standard error. Asterisk indicates a value that is significantly different from the control value (P < 0.05).

pck1-2 stomata have enhanced opening in light

Because PCK1 is expressed in guard cells, and carbon metabolism has been implicated in the control of stomatal closure, we tested whether the pck1 mutants showed normal stomatal responses to light, dark and the drought stress hormone abscisic acid (ABA). To examine stomatal opening in response to light, gas exchange of mature leaves (attached to plants) was measured. The pck1-2 dark-adapted plants had a higher mean stomatal conductance than controls at the start of the experiment (Figure 5a). For up to 25 min after exposure to saturating light, mean stomatal conductances of pck1-2 were consistently higher than controls. After 25 min, the increase in pck1-2 stomatal conductance appeared to cease but the stomatal conductance of control plants continued to increase. These results suggested that the stomata of pck1-2 plants were unable to close as fully as control stomata in the absence of light but were able to open in the presence of light. The defect in the ability to maintain stomatal closure in the dark was also strongly evident in a second allele, pck1-1, demonstrating that control of stomatal aperture is disrupted in multiple pck1 alleles (Figure S2A).

Figure 5.

pck1-2 mutants show abnormal stomatal responses to changes in light intensity.
(a) Stomatal conductances of leaves of dark-adapted pck1-2 and Col-0 control plants measured by infrared gas analysis following exposure to saturating light for the times indicated.
(b) Stomatal conductances of leaves of plants adapted to growth light intensity, following exposure to increasing light intensities indicated for 5 min each.
(c) Rates of photosynthesis measured by infrared gas analysis during exposure to increasing light intensities (as in B). Values in (a–c) are means of four measurements of four different plants.
(d) Stomatal aperture measurements in the dark and after exposure to light for the times indicated. Each value is a mean of 120 aperture measurements over three separate experiments.
(e) Stomatal conductances of leaves of plants adapted to saturating light following dark treatment in the infrared gas analyser chamber for the times indicated. Values are mean of four measurements of four different plants.
(f) Stomatal conductances of leaves of plants subjected to alternating periods of light and darkness. Plants were adapted to a growth light intensity of 140 μm photons m2 sec−1 (0 time-point) and alternately exposed to 5 min in the absence of light (black bars) and 5 min saturating light intensity (white bar) in the infrared gas analyser chamber. Values are means of three measurements on three different plants. Error bars represent the standard error. Asterisks indicate values that are significantly different from control value (P < 0.05).

We investigated stomatal conductance responses to increasing light intensity and found that mean stomatal conductance of pck1-2 plants was higher than controls at all light intensities from 0 to 2000 μmol m−2 sec−1 (Figure 5b). The mean conductance of pck1-2 and control plants decreased between 0 and 100 μmol m−2 sec−1 and then increased at similar rates, as the light intensity increased above the growth light intensity of 140 μmol m−2 sec−1. However, pck1-2 stomatal conductances were higher than controls at the start and remained higher throughout the experiment, suggesting that pck1-2 stomata shut less readily than control stomata at low light intensities but open fully in response to high light intensities. Rates of photosynthesis were monitored at the same time as stomatal conductance during this experiment. The rate of photosynthesis of pck1-2 was slightly higher than controls throughout the experiment, which is consistent with the observed increases in stomatal conductance, but for pck1-2 the rate of photosynthesis was not significantly higher than Col-0 at any time-point (Figure 5c). For pck1-1 however, the photosynthetic rate was significantly higher at 1000–1500 μmol m−2 sec−1 (Figure S2C).

To further investigate the light-induced stomatal opening response we carried out direct measurements of stomatal apertures in epidermal peels from dark-adapted pck1 plants. Mean stomatal apertures increased for both control and mutant stomata during exposure of epidermal peels to light, but pck1-2 stomata had larger mean apertures than control stomata at the start of the experiment and at all time-points measured for up to 4 h after the onset of exposure to light (Figure 5d). In the same assay pck1-1 stomata also exhibited a much larger mean stomatal aperture in the dark than controls, and significantly increased stomatal opening was also observed 30 and 90 min after exposure to light (Figure S2D). Although pck1-1 and pck1-2 stomata were significantly more open, we were not able to determine from this experiment whether they had enhanced sensitivity to light-induced opening as they were also more open at the start of the experiment in the absence of light. Thus, stomata with reduced levels of PCK1 may lack the ability to close completely at low light intensities, and this phenotype is consistent across more than one allele.

Stomatal responses to darkness

To investigate the pck1 dark-induced stomatal closure response further, leaves from light-adapted plants were monitored during exposure to darkness in the gas-exchange chamber (Figure 5e). At the start of the experiment, mean stomatal conductances for control and pck1-2 plants were similar, and this remained the case following 5- and 10-min exposure to darkness. After 15 min in the dark, the decline in pck1-2 stomatal conductance ceased, and pck1-2 conductance remained constant until the end of the experiment, whilst control stomatal conductance continued to decrease suggesting that stomata lacking PCK1 were unable to completely close in the absence of light over the 30-min experiment.

Stomatal response to rapid changes in light

As in the above experiments pck1 plants appeared to exhibit either enhanced stomatal opening in the light or reduced stomatal closure in the absence of light, we investigated their response to fluctuating light and dark periods of 5 min each. At growth-room light intensity, pck1-2 mean stomatal conductances were similar to controls (Figure 5f). In the absence of light, control stomatal conductances decreased and in saturating light they increased. An overall increase in control stomatal conductance was observed for all plants by the end of the experiment, suggesting that saturating light caused the stomata to open more in 5 min of light than they were able to close in 5 min in the absence of light. This overall opening effect was significantly enhanced in the pck1-2 plants from 5 to 35 min of fluctuating light (Figure 5f). These results suggest that pck1-2 stomata cannot close as rapidly as control stomata in the absence of light and are able to open more rapidly in the presence of light.

Stomatal response to ABA

We investigated the ability of pck1 stomata to respond to ABA by measurement of stomatal apertures in epidermal strip bioassays. pck1-2 stomata showed a robust stomatal closure response to applied ABA in promotion of closure bioassays (Figure 6; pck1-1 was not included in this experiment). Although pck1-2 stomatal apertures were significantly wider in the absence of ABA and in the solvent control incubation, they reduced to have similar mean apertures as control stomata following incubation with 0.1, 1 and 10 μm ABA. Thus our results suggest that pck1-2 stomata close normally in response to ABA. We also monitored changes in stomatal conductance in response to changes in humidity and CO2 concentration by infra-red gas analyser (IRGA). We observed no significant differences to controls in these experiments (data not shown).

Figure 6.

 Stomatal apertures of pck1-2 respond normally to ABA treatment.
Mean stomatal apertures of pck1-2 and Col-0 stomata following exposure of epidermal strips to ABA concentrations indicated for 2 h. MeOH is methanol solvent control incubation. Values are means of 120 aperture measurements over three separate experiments. Error bars represent the standard error. Asterisked values are statistically significantly different from the control value (P < 0.05).


Our promoter deletion analysis indicates that multiple cis-elements present in the PCK1 promoter are required for the normal expression of this gene. In particular, we found that a 53-bp region is important for post-germinative expression in Arabidopsis. Blast searches reveal that 30 bp of this is highly conserved in putative PCK promoters of Brassica rapa and Brassica oleracea (Figure S1), suggesting functional conservation. Interestingly, we showed previously that PCK1 expression was induced in a wave-like manner in post-germinative embryos starting with the root/hypocotyl junction. Our results here show that the expression in the root/hypocotyl junction is under the control of a distinct promoter region from the rest of the seedling, suggesting that PCK1 is induced in a two-step manner after germination, first in the collet and then in the rest of the seedling by a distinct and propagating signal. Our analysis also revealed that distinct promoter regions were found to be important for vascular expression and for guard cell expression. It has previously been suggested that gene expression in guard cells may be mediated by Dof domain transcription factors via 5′-TAAAG-3′ target site motifs (Plesch et al., 2001). The PCK1 promoter region contains nine of these motifs between −1617 and −141 bp. Our analysis suggests that trichome and guard cell expression is coordinately regulated by at least three distinct promoter regions, but that guard cell expression was less subject to transgene position effects than expression in trichomes. This may indicate a similar physiological role for PCK1 in these tissues that is distinct from that in young seedlings or that proposed for the phloem (Rylott et al., 2003; Penfield et al., 2004; Wingler et al., 1999; Chen et al., 2004; Brown et al., 2010), but this remains to be tested.

There was much controversy in the late 1970s and early 1980s over the photosynthetic and metabolic capacities of guard cells, and some of these questions have never been satisfactorily resolved. This partly arises from the fact that studies were made in different systems (in different species and using leaves, isolated epidermis, guard cells and guard cell protoplasts). Experiments with radio-labelled malate indicated that malate decarboxylation and gluconeogenesis could occur in guard cells of epidermal strips of Commelina communis (Dittrich and Rashke, 1977) but other studies reported that PEPCK activity could not be detected in guard cells. Schnabl (1981) reported activity for NADP-malic enzyme and low amounts of pyruvate orthophosphate dikinase, but not PEPCK in Vicia faba guard cell protoplasts, and Outlaw et al. (1981) reported activity for NAD- and NADP-malic enzyme, but not PEPCK or pyruvate orthophosphate dikinase in guard cells dissected from V. faba. NADP-malic enzyme has also been shown to be expressed in guard cells of Arabidopsis (Wheeler et al., 2005). These results have been interpreted to indicate either that gluconeogenesis is unlikely to occur in guard cells (Outlaw, 2003), or that gluconeogenesis may occur in guard cells of some species, to convert malate back to starch, but that the decarboxylation step is most probably catalysed by NAD-or NADP-malic enzyme and not PEPCK. We were therefore surprised to find evidence for relatively high expression of the PCK1 gene in Arabidopsis guard cells. However, the expression of PCK1 in guard cells is also suggested by transcriptome and proteome analyses of guard cell protoplasts (Leonhardt et al., 2004; Zhao et al., 2008).

The availability of mutants in the Arabidopsis gene PCK1 with substantially reduced amounts of PEPCK activity (Penfield et al., 2004) allowed us to test whether this enzyme is involved in stomatal aperture control. Our results indicate that PEPCK is involved in the control of Arabidopsis stomatal responses. In particular PEPCK, and by implication malate metabolism, are required for normal stomatal closure in response to darkness. Furthermore, pck1 stomata maintained larger apertures for several hours after exposure to light. SLAC1, and ALMT families of S- and R-type anion channels which have malate transport activity have been shown to be necessary for stomatal closure in response to stimuli including dark, ABA, increases in CO2 concentration and reductions in relative humidity (Negi et al., 2008; Vahisalu et al., 2008; Gruber et al., 2010; Meyer et al., 2010; Sasaki et al., 2010). We therefore suggest that, in Arabidopsis, both malate metabolism and malate transport may be involved in controlling stomatal responses to light and dark, but that malate export may be more important than malate metabolism when stomata close in response to ABA, CO2 and humidity stimuli. Our results suggest that malate metabolism appears to be important when stomata close in response to darkness. It should be noted that, like crassulacean acid metabolism (which may have derived from stomatal metabolism; Cockburn, 1983), decarboxylation of malate may be achieved by different routes (dehydrogenation to OAA by malate dehydrogenase followed by decarboxylation to PEP by PEPCK and/or direct decarboxylation to pyruvate by malic enzyme) and that the balance between transport and metabolism is likely to differ in different species and under different environmental conditions. Our results suggest that lack of PEPCK activity does not affect guard cell ABA signalling. The most likely explanation for the pck1 drought stress phenotype is that the stomata open more widely in the light and close less fully in the dark.

Although our results indicated a relatively high level of PCK1 expression in guard cells, our experiments did not directly probe whether the increase in stomatal gas exchange that we observed in pck1 mutants was due to an alteration in guard cell metabolism. Tomato plants with constitutively reduced succinate dehydrogenase activity and reduced flux through the TCA cycle have recently been shown to have phenotypic effects similar to those we observed for PCK1 mutants (increased stomatal conductance, normal stomatal closure in response to ABA, and no change in stomatal density; Araújo et al., 2011). These plants also have reduced levels of organic acids, including malate, in their leaves and in mesophyll and guard cell protoplasts. However, down-regulation of succinate dehydrogenase specifically in guard cells had no effect on their stomatal aperture control suggesting that changes in apoplastic metabolite concentrations induced the observed changes in stomatal aperture. Indeed, apoplastic malate concentrations are well known to affect stomatal aperture control (Hedrich and Marten, 1993), and thus it will require further experiments to determine whether PCK1 activity affects stomatal conductance via a direct effect on guard cell metabolism.

Experimental Procedures

Plant material

Two Arabidopsis thaliana pck1 T-DNA disruption lines were used in this study. pck1-1 from the Wisconsin T-DNA collection in the Ws background and pck1-2 corresponding to SALK line 032133 as described previously (Penfield et al., 2004). For phenotypic analysis plants were grown on compost in a controlled environment room with a 10-h photoperiod (140 μmol m−2), 20°C day temperature and 16°C night temperature, 60% humidity with irrigation every 3 days, following germination on 0.5× MS medium supplemented with 1% sucrose.

Promoter GUS deletion analysis

PCK1 promoter fragments were amplified by PCR from Arabidopsis genomic DNA as described (Penfield et al., 2004), using PEPCK reverse primer and one of the following forward primers which amplified the promoter lengths described: PCK1-230 5′-CTCGAGAACCACTTTCGCTTCTCTTCACATTCGC-3′; PCK1-316 5′-CTCGAGCTCTTCCTAGGAAGATAGATC-3′; PCK1 -520 5′-CTCGAGCATTATACCACTACGTAGAGGTA-3′; PCK1-674 5′-CTCGAGGCAAGTTGGTCGCTTATAACACGTT-3′ PCK1-780 5′-CTCGAGATGGAGATCACGAATATTAGACCGTAA-3′; PCK1-807 5′-CTCGAGCATCATCAAACAACTTGATAAAATAAGTG-3′; PCK1-870 5′-CTCGAGCAATCGGAGCTATGTCTATCGTTTTC-3′; PCK1-923 5′-CTCGAGTGTCCTCTCACGAATCTACTTAC-3′; PCK1-1072 5′-CTCGAGTGACGAAGGTTCTTGTAATTGTG-3′; PCK1-1527 5′-CTCGAGATTGGGTCCAGACTATTAAAGTTG-3′. Exact primer locations were determined by the availability of GC-rich regions suitable for primer annealing. Promoter fragments were purified by gel electrophoresis and cloned into pGEM-T (Promega, After sequencing, the promoters were sub-cloned into pGREENGUS by HindIII/XhoI digestion as described for the full-length PCK1 promoter (Penfield et al., 2004). T2 lines (20–50 individual plants) were stained for GUS expression in seedlings and for each construct six independent representative T3 homozygous transgenic lines were isolated which segregated kanamycin resistance and GUS expression in a 3:1 ratio, indicating a single T-DNA integration event. GUS expression was assayed as described (Pritchard et al., 2002; Penfield et al., 2004). Briefly, 20 seedling were extracted in 4-methylumbelliferyl β-d-glucuronide (MUG) extraction buffer (50 mm NaPO4 pH 7.0, 10 mm EDTA pH 8.0, 0.1% sarcosyl, 0.1% Triton) and 20 μl extract was reacted with 1 mm MUG for 1 and 2 h before stopping the reaction with sodium carbonate. The yield of 4-methylumbelliferone (4MU) was assessed by comparison with a standard curve which was confirmed to be linear in the range used. Plant tissues were photographed using a Leica SP6 stereomicroscope fitted with a SPOT RT image capture system (Diagnostic Instruments,

Relative water content (RWC)

Leaves from the rosette were excised, and their fresh weight was scored immediately. After floating them in deionised water at 4°C overnight, their rehydrated weight was determined. Finally, they were dried in an oven at 70°C overnight and weighed. The RWC was calculated using the equation RWC = (fresh weight−dry weight)/(rehydrated weight−dry weight). Values were statistically tested using unpaired t-tests.

Gas exchange analysis

To record responses to saturating light intensity, control and mutant plants that had been grown alongside each other were taken before the start of the photoperiod and dark adapted for 1 h. A fully expanded leaf, attached to the plant, was placed in the IRGA chamber at ambient CO2 and humidity in the absence of light for 10 min followed by exposure to saturating light (1500 μmol m−2sec−1) and stomatal conductances were measured every 5 min. To monitor responses to increasing light intensity plants were taken during the first half of the photoperiod. The light intensity in the IRGA leaf chamber was set to the growth light intensity (140 μmol m−2 sec−1) for 10 min and was then decreased to zero and increased every 5 min, up to 2000 μmol m−2 sec−1. Responses to darkness were recorded from plants in the first half of their photoperiod. Leaves were exposed to saturating light in the IRGA chamber for 10 min. The light was then switched off and stomatal conductance measured every 5 min in the absence of light. To monitor responses to fluctuating light leaves were placed in the IRGA chamber at the growth light intensity for the first 10 min. The light intensity was then alternated between 1500 and 0 μmol m−2 sec−1 every 5 min and stomatal conductance was recorded at the end of each 5-min period. Values were statistically tested using unpaired t-tests. Leaf gas-exchange measurements were made using an IR gas analyser (model 6400; Li-Cor, External air was scrubbed of CO2 and mixed with a supply of pure CO2 to result in a reference concentration of 400 μl L−1. Flow rate was 500 μmol sec−1 and external humidity was 55–65%. The upper half of the leaf chamber (2 × 3 cm) contained a light-emitting diode light source, and the bottom half held the leaf-temperature thermocouple which maintained a temperature of 20°C. Two GaAsP PAR sensors were fitted, one located inside the upper half of the leaf chamber and the other located externally, beside the leaf chamber.

Stomatal aperture measurements

Epidermal bioassays were adapted from Webb and Hetherington (1997). Epidermal peels were taken from the abaxial surface of Arabidopsis leaves and floated on resting buffer (10 mm MES, pH 6.2). For the rate of stomatal opening in light experiment, peels were taken from plants towards the end of the night period in a dark room illuminated with a low-intensity red filtered light and then incubated in the dark in resting buffer for 1 h. Peels were then transferred to opening buffer (10 mm MES, 50 mm KCl, pH 6.2) perfused with CO2-free air, maintained at 20°C and illuminated with 300 μmol m−2 white light for up to 4 h. For ABA-induced stomatal closure assays peels were maintained in opening buffer in 300 μmol m−2 light for 2 h before ABA was added to final concentrations of 0–10 μmol and incubated for a further 2 h before stomatal apertures were measured. As ABA is dissolved in methanol as a stock solution, a methanol control containing equivalent solvent to the 10 μmol ABA incubation was included. Stomatal aperture measurements were performed using a light microscope and graticule. Forty apertures were measured for each value and each experiment was repeated three times. For each experiment, peels were taken from three different plants. Values were statistically tested using unpaired t-tests.


We acknowledge financial support for this project from The Garfield Weston Foundation to the Centre for Novel Agricultural Products and a BBSRC studentship to SC.

Author Contributions

SDP and IAG designed and performed the expression analysis of PCK1, with assistance from ADG. Phenotypic analyses of pck1 mutants in guard cells were initiated by KB and RCL, and designed and completed by SC, JEG and RCL. SDP and JEG wrote the manuscript with assistance from IAG and RCL.