• ATP-binding cassette (ABC) transporter;
  • multidrug resistance;
  • drought susceptibility;
  • guard cell;
  • stomata


  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

ATP-binding cassette (ABC) transporters are membrane proteins responsible for cellular detoxification processes in plants and animals. Recent evidence shows that this class of transporters may also be involved in many other cellular processes. Because of their homology with human multidrug resistance-associated proteins (MRP), cystic fibrosis transmembrane conductance regulator (CFTR) and sulfonylurea receptor (SUR), some plant ABC transporters have been implicated in the regulation of ion channel activities. This paper describes an investigation of the AtMRP4 gene and its role in stomatal regulation. Reporter gene studies showed that AtMRP4 is highly expressed in stomata and that the protein is localized to the plasma membrane. Stomatal aperture in three independent atmrp4 mutant alleles was larger than in wild-type plants, both in the light and in the dark, resulting in increased water loss but no change in the photosynthetic rate. In baker's yeast, AtMRP4 shows ATP-dependent, vanadate-sensitive transport of methotrexate (MTX), an antifolate and a substrate of mammalian MRPs. Treatment with MTX reduced stomatal opening in wild-type plants, but had no effect in atmrp4 mutants. These results indicate the involvement of AtMRP4 in the complex regulation of stomatal aperture.


  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Members of the ATP-binding cassette (ABC) transporter superfamily are found in all organisms, from prokaryotes to eukaryotes, including both the plant and animal kingdoms (Henikoff et al., 1997). In animals, the discovery of ABC transporters was closely linked to their ability to export potentially toxic drugs. Ectopic expression of genes from this family confers multidrug resistance (MDR) on cancer cells (e.g. Gottesman and Pastan, 1993). Similarly, in plants, the existence of ABC transporters was first demonstrated by the energized vacuolar transport of modified xenobiotics (Martinoia et al., 1993). Apart from their transport function, some plant ABC transporters may mediate or control ion fluxes. In animals, the cystic fibrosis transmembrane conductance regulator (CFTR) and sulfonylurea receptor (SUR) exhibit or modulate ion channel activity (Anderson et al., 1991; Bryan and Aguilar-Bryan, 1999; Higgins, 1995) and are receptors for the sulfonylurea compound glibenclamide (Schmid-Antomarchi et al., 1987; Schultz et al., 1996). Interestingly, Leonhardt et al. (1997, 1999) showed that treatment with the sulfonylurea glibenclamide, and other drugs that interact with the SUR and the CFTR, induced stomatal opening in the guard cells and affected their electrophysiological properties.

Guard cells are highly specialized epidermal cells, located in pairs on the aerial organs of the plant. Each pair of cells forms the boundary of a small pore, or stoma, in the epidermis. Guard cells play a major role in controlling gas exchange (mainly photosynthetic carbon dioxide uptake and water release by transpiration) between the plant and the surrounding atmosphere and are regulated by environmental (light and CO2) and physiological (hormones) stimuli. These stimuli induce dynamic changes in the intracellular concentrations of inorganic and organic ions (including K+ and malate2−), and soluble sugars (e.g. sucrose). Precise control of stomatal aperture is critical for optimal plant performance, and requires a sophisticated, highly coordinated interplay of ion channels, metabolite conversions, and signal transduction pathways (Grabov and Blatt, 1998; Leckie et al., 1998; Schroeder et al., 2001a). Our current understanding of guard cell signal transduction and the role individual genes play in this process has undoubtedly benefited from the analysis of mutants impaired in stomatal physiology (e.g. Kinoshita et al., 2001; Mustilli et al., 2002; Pei et al., 1997; Szyroki et al., 2001; Wang et al., 2001). A detailed understanding of stomatal physiology will have a major impact on the genetic engineering of drought tolerance in higher plants, including crops (Schroeder et al., 2001b).

It was recently shown that Arabidopsis thaliana plants, carrying a T-DNA insertion in the AtMRP5 gene (atmrp5-1), which encodes a multidrug resistance-associated protein (MRP; this class of transporters has been renamed ABCC in humans; Dean et al., 2001), no longer open their stomata in response to glibenclamide (Gaedeke et al., 2001). Detailed analysis revealed that the stomata of atmrp5-1 plants exhibited reduced stomatal aperture in light conditions and did not (or only weakly) respond to abscisic acid (ABA), Ca2+, or auxin (Klein et al., 2003). In contrast, the application of fuscicoccin, which activates the plasma membrane proton pump, restored stomatal opening. Interestingly, atmrp5-1 plants take up less water, compared to wild-type plants, and exhibit a drought-tolerant phenotype. These results confirm the hypothesis of Leonhardt et al. (1997, 1999) that ABC transporters may be involved in the regulation of guard cells. However, research on the mechanism by which AtMRP5 interacts with the other players in guard cell regulation, either through direct interaction with ion channels or as part of a signaling pathway, is still in its infancy. To help elucidate the role of the ABC transporters in stomatal opening, we have conducted a screen for further AtMRP genes that are specifically expressed in guard cells. Here, we show the effect of knockout mutations in the closely related AtMRP4 gene. The mutant plants atmrp4 exhibit a phenotype opposite to atmrp5-1. Phylogenetically, AtMRP4 belongs to the clade II type of AtMRP genes, being most closely related to AtMRP14 with 82% amino acid (aa) homology (Kolukisaoglu et al., 2002). Analysis of the structure of the Arabidopsis genome indicates that both genes originated from chromosomal segmental duplication events.


  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

cDNA cloning and expression analysis of AtMRP4

The longest AtMRP4 cDNA insert isolated from a cDNA library (Tommasini et al., 1997) lacked the first 726 bp. Following gene predicitions by Sanchez-Fernandez et al. (1998), and our own structural alignments of the entire AtMRP gene family (Kolukisaoglu et al., 2002), we found no intron in the missing 726 bp at the 5′ end. The 5′ sequence was therefore obtained by PCR amplification of genomic DNA and inserted upstream of the AtMRP4 cDNA, such that the entire DNA sequence was identical to GenBank entry AF243509. The open-reading frame (ORF) of the AtMRP4 cDNA encodes a protein of 1516 aa residues with a predicted molecular weight of 169 kDa. Hydropathy profile consensus prediction using the ARAMEMNON membrane protein database (Schwacke et al., 2003) predicts 15 transmembrane helices, organized in three transmembrane domains (TMD0–TMD2, see Figure 1). The two ABCs of AtMRP4 resemble those typical for ABC proteins (Martinoia et al., 2002). Each consists of a c. 200 aa domain, comprising the Walker A (GxxGxG) and Walker B (T/IYLLD) motifs of ATP-binding sites (Walker et al., 1982) and the ABC signature (Higgins, 1992). As found for most AtMRPs, AtMRP4 also contains a perfect N-terminal ABC signature and a degenerated C-terminal ABC signature with matches in five of the eight conserved amino acid positions (Martinoia et al., 2002).


Figure 1. Structure of the AtMRP4 gene, position of insertion mutants (a) and RT-PCR analysis of AtMRP4 and S16 expression in wild-type and atmrp4 mutants (b).

(a) Genomic structure of AtMRP4 and insertion sites. Shown are the predicted AtMRP4 protein structure as well as the genomic structure of AtMRP4. Exons are shown as boxes, whereas introns are marked as interjacent black lines. Three predicted transmembrane-spanning domains (light gray boxes) are named TMD0–TMD2 (TMD0 = aa positions 101–227; TMD1 = aa positions 329–573; and TMD2 = aa positions 951–1237). Nucleotide-binding domain (NBD)1 (aa positions 574–950) and NBD2 (aa positions 1238–1516; dark gray boxes) are indicated. The promoter region including the 5′ UTR is shown as dashed line with an arrow indicating the direction of transcription. A dark gray box shows the 3′ UTR sequence. Start of ORF (+1) and stop of ORF (+5440) are indicated. The insertion sites and orientation of T-DNAs (open triangles) and Ds transposon (filled triangle) are indicated. The position of the insertions is shown as nucleotide distance from the A (position +1) of the start codon ATG. The border sequence of insertion atmrp4-A, facing away from AtMRP4, was not determined. Exons 1–11 are numbered in gray circles. The N-terminal deletion in atmrp4-2 (positions −588 to +1545) is indicated as gray box with dashed border.

(b) RT-PCR analysis verifies the absence of AtMRP4 transcript in atmrp4 mutants (atmrp4-1atmrp4-3) in contrast to wild-type controls Ws-2 and Ler. cDNAs were synthesized from total RNA of mutant and wild-type plants and AtMRP4-specific and 40S ribosomal protein S16-specific transcripts (Kolukisaoglu et al., 2002) were detected by RT-PCR. Absence of genomic contaminations was verified using intron-spanning primers for AtMRP4 and genomic Ws-2 DNA as control (control).

Download figure to PowerPoint

RT-PCR analysis of AtMRP4 expression in different tissues revealed that this ABC transporter is expressed in low amounts in all tissues tested (Figure 2a; Kolukisaoglu et al., 2002). In order to obtain a more detailed picture of AtMRP4 expression, promoter fusions between 2 kbp of the 5′ untranslated region (UTR) of AtMRP4 and the uidA gene encoding β-glucuronidase (GUS) were constructed and used to transform Arabidopsis plants (Figure 2b, i–vi) or for particle-gun-mediated transient expression in Vicia faba leaves (Figure 2b, vii). The promoter fragment corresponds to the entire UTR between At2g47790 and At2g47800. In seedlings grown in permanent light or darkness (Figure 2b, i and iv), GUS expression was detected in primary roots and the basal region of hypocotyls close to the root–shoot interphase. No GUS activity was detected in root hairs. Interestingly, GUS activity was generally absent from the root tips of light-grown seedlings, while high GUS activity was observed in root tips of etiolated seedlings. For GUS staining, seedlings were exposed to 5-bromo-4-chloro-3-indolyl-β-d-glucuronic acid (X-Gluc) for 3 h. When incubation was allowed to continue for 24 h, the entire seedling showed GUS-specific staining (data not shown) confirming the RT-PCR results. In different leaf types, GUS activity, driven by the AtMRP4 promoter, was variable. This was illustrated by different intensities of blue staining observed in cotyledons (Figure 2b, i) and three rosette leaves of independent transgenic lines (Figure 2b, ii), which are representative of the staining patterns found in our transformants. In rosette leaves, petioles were stained in several cases. In cotyledons and rosette leaves, GUS activity was often unevenly distributed. In some, but not all, cases, the vascular tissue of cotyledons and rosette leaves exhibited higher GUS activity than the surrounding tissue. Irrespective of these irregular staining patterns, high GUS activity was always detected in the guard cells of cotyledons, rosette leaves, cauline leaves, hypocotyls, stems, and the fused carpel walls. In flowers, high GUS activity was found in the sepals (but not in the petals) and in the filaments (but not in the stamens or pollen sacs; Figure 2b, v). Young carpels generally stained strongly at the style, while GUS activity in the carpel walls increased during ripening of the silique (Figure 2b, v and vi). In order to verify the specificity of the high GUS activity in guard cells, epidermal strips of V. faba leaves were bombarded with particles coated with promoter–GUS fusion constructs. GUS activity driven by the AtMRP4 promoter was unequivocally observed in the guard cells of V. faba leaves (Figure 2b, vii). In addition, transgenic Arabidopsis plants, transformed with a T-DNA carrying a fusion of the AtMRP4 promoter and the green fluorescent protein (GFP), exhibited high GFP fluorescence in guard cells (data not shown).


Figure 2. Expression and localization of AtMRP4.

(a) Results of RT–PCR analysis suggest that AtMRP4 is expressed in low amounts in all tissues tested. cDNAs from different Arabidopsis tissues were synthesized from total RNA and transcripts specific for AtMRP4 and 40S ribosomal protein S16 were detected by PCR. Five micrograms of total RNA and equal volumes of PCR products were separated on 1% formaldehyde (RNA; lowest panel) and 2.5% native agarose gels (PCR products; upper two panels), respectively.

(b) Histochemical localization of GUS activity (i–vii) and subcellular localization of AtMRP–GFP (viii–xiii). (i-vi) GUS activity in A. thaliana plants stably transformed with approximately 2 kbp of the 5′ UTR of AtMRP4 fused to the uidA gene. (i) Eight-day-old seedling grown in 8 h light/16 h dark showing GUS activity in the cotyledons mainly in punctuate structures, parts of the hypocotyls, and the primary root. The arrow points to the root tip, which is not stained. (ii) Rosette leaves taken from independent 21-day-old seedlings showing different GUS staining patterns mainly with respect to the petiole and vascular tissue, which are not always stained. In all cases, punctuate structures representing stomata are strongly stained. (iii) Surface view of the abaxial side of a rosette leaf viewed with differential interference contrast optics exhibiting a strong GUS activity in the stomata. (iv) Eight-day-old etiolated seedling. Arrow, in comparison to (i), the root tip exhibits strong GUS activity. (v,vi) Flowers and siliques of 5-week-old transgenic A. thaliana plants. Stomata of the carpel wall are strongly stained. (vii) GUS staining in guard cells of a V. faba leaf 2 days after particle gun bombardment of a leaf with the promoter–GUS fusion construct. (viii–xiii) Laser scanning confocal microscope analysis of plant material expressing a GFP-tagged version of AtMRP4. Images represent internal optical sections generated by CLSM from onion epidermis cells expressing transiently (ix, x) and Arabidopsis seedlings expressing ectopically AtMRP4–EGFP (ix–xiii). AtMRP4-specific GFP fluorescence was found mainly in the periphery of cells, suggesting that AtMRP4 protein is located at the plasma membrane. (viii) Onion epidermis cell expressing the vacuolar marker KCO1–GFP labeling the tonoplast surrounding the central vacuole. The position of the nucleus is indicated by an arrow. (ix) Onion epidermis cell expressing AtMRP4–EGFP shows fluorescence at the periphery of cells. In some onion cells bombarded with AtMRP4–EGFP, dense vesicles, filled with GFP fluorescence (indicated by an arrow), were observed. These may have been caused by overloading the cells' secretory pathway with excessive quantities of fusion protein. (x) Same cells in a bright field. (xi) Close up of root epidermis cells showing fluorescence at the periphery of cells. (xii) Close up of an Arabidopsis guard cell pair showing fluorescence at the periphery of cells. Note that circular green fluorescence of chloroplasts (marked by an asterisk) is not because of GFP fluorescence but because of autofluorescence of bleaching chlorophyll. (xiii) Same cell in a bright field. Scale bars are 2 mm (i, ii, iv), 50 µm (iii, vii), 20 µm (viii, ix), and 5 µm (xi, xii).

(c) Arabidopsis microsomal fractions overexpressing MRP4–GFP were separated by linear sucrose gradient centrifugation, and Western blots were imunoprobed with anti-GFP antisera. Origin of immunopositive fractions was ascertained using antisera against the marker proteins vacuolar V-type H+-ATPase, ER-localized BIP and the plasma membrane-bound P-type H+-ATPase (Geisler et al., 2000).

Subcellular localization of AtMRP4

Analysis of AtMRP4 using the ARAMEMNON database (Schwacke et al., 2003) predicts the presence of a chloroplast-targeting signal with average to high probability. To determine precisely the intracellular localization of AtMRP4, an enhanced version of the GFP (EGFP) was inserted into the AtMRP4 cDNA within an intracellular loop separating the predicted transmembrane helices 14 and 15. This particular location of EGFP was chosen to minimize interference with the catalytic active cytoplasmic nucleotide-binding folds of AtMRP4.

Confocal microscope analysis of onion epidermal cells transiently expressing AtMRP4–EGFP showed fluorescence at the periphery of cells, suggesting that AtMRP4 protein is located at the plasma membrane. The vacuolar GFP marker KCO1, a component of the slow-vacuolar K+ channel (Czempinski et al., 2002; Schönknecht et al., 2002) was used as a control (Figure 2b, viii). KCO1–GFP labeled the tonoplast surrounding the central vacuole, but not the cytoplasm, which is largely restricted to a region surrounding the nucleus (arrow in Figure 2b, viii). These labels clearly distinguish the tonoplast and the plasma membranes. This also confirms the observation that AtMRP4 is localized on the plasma membrane (Figure 2b, ix). In some onion cells bombarded with AtMRP4–EGFP, dense vesicles, filled with GFP fluorescence (arrows in Figure 2b, ix), were observed. These may have been caused by the high activity of the CaMV 35S promoter used to drive the GFP fusion, overloading the cells' secretory pathway with excessive quantities of fusion protein. Bombardment of onion epidermis with atTOC159–GFP, a component of the outer envelope protein import machinery of chloroplasts (Bauer et al., 2002; kindly provided by F. Kessler, University of Neuchatel, CH), assured the presence of plastids in these cells. Fluorescence signals were visible on a large amount of intracellular structures mainly close to the plasma membrane and appeared clearly different from AtMRP4–GFP signals (M. Jasinski and E. Martinoia, unpublished observation; data not shown). We concluded that AtMRP4–GFP is not associated with plastids.

Seedlings transformed with AtMRP4–EGFP showed high levels of fluorescence in both the root and the shoot. As observed in bombarded onion tissue, fluorescence in the root epidermis (xi) and leaf stomata (xii) was restricted to the cells' periphery. Vector-transformed control plants revealed no GFP fluorescence (data not shown).

To verify these data immunologically, Arabidopsis microsomes were separated by linear sucrose gradient density centrifugation and probed with monoclonal anti-GFP antibodies. Results showed a single band of the expected size (c. 200 kDa) in gradient fractions 9–12 (sucrose concentrations between 42 and 50%). AtMRP4 co-localized with the plasma membrane-bound P-type H+-ATPase (Figure 2c), while markers for other membranes, such as the vacuolar V-type H+-ATPase or BIP, an ER-specific marker, were found in other fractions (Figure 2c).

Isolation of allelic knockout mutants in AtMRP4

T-DNA insertion mutants were identified for the AtMRP4 gene (At2g47800) in collections of T-DNA-transformed Arabidopsis lines obtained from Arabidopsis Stock Centers (ABRC at Ohio State, USA, and NASC, University of Nottingham; Kolukisaoglu, Möller, Sieber, Zeidler, Schulz, unpublished). Sequence analysis of PCR fragments identified T-DNA insertions at positions +5609 (atmrp4-1), +1545 (atmrp4-2), and +6067 (atmrp4-A; Figure 1a). Only one T-DNA insertion was located within the coding region of AtMRP4 (atmrp4-2). Insertions in atmrp4-1 and atmrp4-A were found 57 and 662 bp downstream of the stop codon, respectively. Both T-DNA insertion sites were defined by sequence analysis of the appropriate PCR products. Surprisingly, the insertion in atmrp4-2 caused a deletion of AtMRP4 sequence from −588 to +1545, removing a significant portion of the putative promoter region, as well as the entire TMD0, and most of TMD1 domains. Database searches of Cold Spring Harbor Laboratory Genetrap DB (Martienssen, 1998) resulted in identification of mutant line atmrp4-3 (ET 1399), which bears a dissociation (Ds) element in the 10th exon of AtMRP4. Upon integration of the Ds element, a duplication of AtMRP4 sequences +5311 to +5378 occurred at the insertion site. The atmrp4-1 and atmrp4-2 insertion mutants were backcrossed to the wild type. The phenotype of atmrp4 co-segregated with insertions in AtMRP4 (as verified by PCR and selection on media containing appropriate antibiotics). The number and structure of T-DNA insertions, or Ds elements, in single lines was confirmed by Southern blot analysis. atmrp4-1 was found to contain only a single T-DNA insert, whereas atmrp4-2 lines showed two independent T-DNA insertions. These T-DNAs could not be separated by backcrossing with the wild type, thus revealing their close proximity. PCR and Southern blot analysis, as well as growth on media containing kanamycin as selection marker for the Ds element inserted in atmrp4-3, allowed the isolation and establishment of lines homozygous for the disrupted AtMRP4 gene. As the T-DNA insertion of atmrp4-A did not show any phenotypic alteration, possibly because of its location of more than 600 bp downstream of the AtMRP4 stop codon, this line was not subjected to further analysis.

RT-PCR experiments performed on total RNA extracted from aerial parts of soil-grown Arabidopsis showed that the AtMRP4 transcript did not accumulate in homozygous atmrp4-1, atmrp4-2, and atmrp4-3 lines. This result suggests that the atmrp4-1 and atmrp4-3 mRNAs are unstable when the insertion is either close to the stop codon (atmrp4-3) or right in front of the polyadenylation signal (atmrp4-1). These findings concur with the observation that all atmrp4 mutant lines display identical mutant phenotypes.

Disruption of AtMRP4 affects stomatal functioning

The strong expression of AtMRP4 in guard cells prompted the question whether the absence of AtMRP4 leads to changes in stomatal movement. In a first experiment, leaves were excised from dark-adapted plants of all mutants and the corresponding wild types. The leaves were placed on 20 mm KCl and 0.5 mm CaCl2 in the light (to promote stomatal opening) for 2.5 h and subsequently incubated in the dark on the same medium for a further 4 h. Microscopic measurements of epidermal peels revealed that, irrespective of the light conditions, the stomatal aperture (data not shown), as well as the calculated width to length ratio of the stomatal pore, was larger in atmrp4-1, atmrp4-2, and atmrp4-3 mutants than in the corresponding wild types (Figure 3a).


Figure 3. Defects in stomatal opening and closing of wild-type and atmrp4 mutant plants in response to light and darkness lead to changes in carbon isotope signatures of mature leaves.

(a) Detached rosette leaves of equal developmental stage from dark-adapted 6-week-old wild-type and atmrp4 plants were incubated in the light for 2.5 h, and subsequently transferred into darkness for another 4 h. Stomatal apertures (pore width and length) were measured on abaxial epidermal peels within 7 min after peeling. Depicted is the ratio of width to length of mean values and the standard error to the means (SEM) of three independent experiments; n ≥ 60 per experiment. Stomata of all atmrp4 alleles display pronounced opening in the light and dark when compared to the corresponding wild-type ecotypes (Ws-2, atmrp4-1, atmrp4-2; and Ler, atmrp4-3). Asterisks indicate statistically significant differences between mutants and the wild type for plants grown in light or darkness (t-test; P < 0.01). Inset: inhibition of light-induced stomatal opening in the presence of 10 µm ABA. Mutants still respond to ABA.

(b) Stomatal apertures in intact plants shortly before and soon after the onset of the light period in the growth chamber. Leaves were detached from plants 30 min before (dark) or 90 min after the light switched on in the growth chamber, epidermal strips were prepared, and stomatal pore sizes were scored within 10 min. Mean values ± SEM were averaged from three independent experiments; n ≥ 80 measurements per experiment. Asterisks indicate that the differences of stomatal apertures between Ws-2 wild-type (empty bar) and atmrp4-2 plants (filled bar) are significant (t-test; P < 0.05).

(c) The 10 youngest (young) and the following 10 fully expanded rosette leaves (mature) of 6-week-old Ws-2 (open bar) and atmrp4-2 (gray bar) were separately detached from single plants and dried. The stable carbon isotope signature of bulk leaf material was determined by isotope ratio mass spectrometry. Depicted are the mean δ13C values ± SD of a representative experiment of two with similar tendencies (n = 4 single plants per genotype and experiment).

Download figure to PowerPoint

In order to verify that the knockout in AtMRP4 does not affect the density or spacing of stomata in the abaxial epidermal layer, the stomatal frequency, based on more than 800 stomata per genotype (four plants per genotype, six different leaves per plant), was calculated. The mutants atmrp4-1 (178 ± 24 stomata mm−2), atmrp4-2 (176 ± 29 stomata mm−2), and atmrp4-3 (128 ± 26 stomata mm−2) were compared with the respective wild types Wassilewskija (Ws-2) 174 ± 45 stomata mm−2 and Landsberg erecta (Ler; 129 ± 27 stomata mm−2). Results showed no significant difference between the stomatal density in the mutants and their corresponding wild types. Interestingly, the Ler ecotype shows fewer stomata per unit surface area than Ws-2. This is consistent with our observation that, in drying experiments, Ler loses fresh weight (FW) more slowly than Ws-2 (data not shown).

Stomatal pore width is controlled by a large number of endogenous and ecological parameters. One of the most prominent effectors of stomatal aperture is the water stress hormone ABA, which normally triggers stomatal closure and inhibits opening in the light. In order to investigate whether the increased stomatal aperture in atmrp4 was caused by insensitivity to ABA, leaves from dark-adapted atmrp4-2 and atmrp4-3 were incubated for 2.5 h, in the light, in the presence of 10 µm ABA. Mutant stomatal pore widths were measured and compared to the corresponding wild type. ABA inhibited the opening of stomata in the light in wild-type and atmrp4 mutant plants (Figure 3a, inset). In addition, stomata of atmrp4-2 mutants closed after application of ABA in the light (width/length ratios of stomata before ABA application: Ws-2, 0.43 ± 0.02; and atmrp4-2, 0.63 ± 0.01; after 2 h, ABA-application: Ws-2, 0.24 ± 0.01; and atmrp4-2, 0.14 ± 0.01). We concluded that the increased aperture of atmrp4 mutants was not the result of changes in sensitivity to ABA. In this respect, atmrp4 mutants behaved very differently to atmrp5 mutants, which are insensitive to ABA (Klein et al., 2003).

Stomatal pore size was also investigated on the leaves of intact plants in a growth chamber. The width of Ws-2 and atmrp4-2 stomatal apertures was compared before and after exposure to light under controlled conditions (Figure 3b). Leaves were detached from intact 6-week-old Ws-2 and atmrp4-2 plants, 30 min before and 90 min after exposure to light. Leaves were examined and stomatal apertures measured immediately upon detachment. At the end of the dark period (30 min before exposure to light), stomatal apertures in dark-adapted atmrp4-2 plants were clearly larger than in Ws-2. Likewise, stomata of atmrp4-2 were larger than those of Ws-2 after 90 min of light irradiation (Figure 3b). Thus, differences in stomatal pore width, resulting from changes in light conditions, occurred regardless of whether the leaves remained on the plant, or were excised and exposed to KCl. It was therefore hypothesized that atmrp4 mutants would experience increased water loss, and that atmrp4 mutants should exhibit lower δ13C values compared to Ws-2 plants. It is well established that ribulose bisphosphate carboxylase/oxygenase (Rubisco) effectively discriminates between 13CO2 and 12CO2 (Farquhar et al., 1989). When stomata are open, gas exchange between the leaf intercellular space and the atmosphere is optimal, and Rubisco preferentially fixes 12CO2, leading to a stronger discrimination (more negative δ13C values in plant dry matter). In contrast, when stomata are closed, diffusion of atmospheric CO2 (Ca) into the intercellular leaf space is restricted, leading to a decrease in the CO2 concentration in the leave (Ci), and the probability that Rubisco fixes 13CO2 increases. Stable carbon isotope ratios were analyzed in dry material collected from rosette leaves of single Ws-2 and atmrp4-2 plants. Leaves were separated into fractions of 10 ‘younger’ leaves and 10 ‘mature’, fully expanded leaves by counting from the top of the rosette. Fully expanded leaves of atmrp4-2 indeed exhibited a statistically not significant tendency to lower δ13C values, when compared to the wild type, while no difference was observed in ‘younger’ rosette leaves (Figure 3c). Younger leaves are less likely to exhibit significant differences in δ13C values than mature leaves, as the carbon isotope signature of bulk leaf material represents a time-integrated signal, and only a small fraction of the carbon of young leaves is derived by their own photosynthesis. Therefore, the observation that atmrp4 exhibited increased stomatal pore sizes under growth chamber conditions was reflected by a decrease in the corresponding δ13C/12C values. Taken together, these observations suggest that atmrp4 has an increased rate of stomatal transpiration, and that atmrp4 plants would be consequently less tolerant to drought stress conditions than wild type.

atmrp4 mutants show increased water loss and wilt faster during desiccation

As mutations in AtMRP4 affect the regulation of stomatal aperture, one might expect the mutant plants to display a different rate of transpiration and water loss. To investigate whether AtMRP4 plays an important role during drought stress response, water loss was measured using two methods. Firstly, a time-course of FW loss was measured in excised rosettes of soil-grown plants. As expected for leaves with greater pore apertures, atmrp4-1 and atmrp4-2 plants displayed increased water loss compared to the wild type (Figure 4a).


Figure 4. In the absence of water supply atmrp4 plants loose water faster and wilt earlier than wild-type plants.

(a) Time course of water loss of excised rosettes of 6-week-old wild-type Ws-2 (squares), atmrp4-1 and atmrp4-2 (black and gray circles, respectively) plants. Data show percentage of initial FW lost from entire rosettes given as the mean of four individual plants. A representative of three independent experiments is shown. Error bars represent SDs.

(b) Seven-week-old plants of the wild-type ecotypes (Ws-2 and Ler), atmrp4 and abi1 grown in soil under short-day condition and with normal water supply (twice a week) were drought stressed by terminating irrigation. Shown is the phenotype after 13 days without water.

(c) The loss of water in pots was determined by determining the weight of saran-wrapped pots with individual plants at the indicated time points during desiccation as detailed in Experimental procedures. Values (means of eight individual pots with SDs) of a representative experiment are illustrated as percentage of the initial water content. Left panel: Squares, Ws-2 wild type; black circles, atmrp4-1; and gray circles, atmrp4-2. Right panel: squares, Ler wild type; and black circles, atmrp4-3.

Download figure to PowerPoint

Secondly, water loss and wilting was analyzed over several days in intact mature plants, as they desiccated after cessation of watering. Figure 4(b) shows plants, after 13 days of drought. All mutant plants (alleles atmrp4-1–atmrp4-3) wilted earlier than the corresponding wild types. However, the wilting phenotype in all atmrp4 individuals was less severe than in the classical ABA-insensitive (abi)1 mutant, (compare abi1 and atmrp4-3, Ler background). Single atmrp4 plants grown in soil also displayed earlier wilting than the wild type (data not shown). Loss of water from individual plants in saran-wrapped pots leaving only the foliage exposed to the atmosphere was also accelerated for atmrp4 plants when compared to wild-type plants (Figure 4c).

Increased transpiration rates of atmrp4-2 in response to light and decreased partial pressure of CO2

Transpiration rates in intact wild-type and atmrp4-2 plants were measured using gas exchange analysis of individual leaves, under various environmental conditions (Figure 5). In the first set of experiments, changes in the transpiration rate were investigated while shifting plants from darkness to light (400 µmol photosynthetically active radiation (PAR) m−2 sec−1) and back to darkness. The experiment began with dark-adapted plants at 400 p.p.m. CO2. Switching light on and off induced an increase and a decrease in transpiration rate, respectively (Figure 5a), reflecting stomatal opening and closing, respectively. Analysis of steady-state levels of transpiration revealed that atmrp4-2 had significantly higher (1.4-fold) transpiration rates in the light, as compared to the wild type. In the dark, a slight, though statistically insignificant, increase in transpiration rate was observed in the mutant compared to the wild type. Linear (light on) and exponential fitting (‘light on’ and ‘light off’), with first-order exponential functions, was used to analyze the kinetics of transitions between steady states for each curve (Figure 5b,c). The analysis showed that switching on the light induced a faster increase in transpiration rate in atmrp4-2, than in the wild type, suggesting that stomatal opening, in response to light, occurs more quickly in the mutant. However, no significant differences between the wild type and mutant were observed when the time constant for the closing reaction (light off) was calculated. This result suggested that the effect of the mutation in AtMRP4 is not restricted to absolute changes in stomatal pore size, but also effects the kinetics of stomatal opening.


Figure 5. Disruption of AtMRP4 results in changes of stomatal transpiration.

(a) Dark-adapted single leaves attached to single soil-grown, 6-week-old Ws-2 (open symbols) and atmrp4-2 (closed symbols) plants were analyzed by gas exchange measurement in the presence of 400 p.p.m. CO2. After equilibration in the dark for 30 min, PAR was set to 400 µE m−2 sec−1 for 1 h (light) and then turned off again (dark). Changes in transpiration were recorded at 1-min intervals. Illustrated are the mean values and SDs of 10 data series for each genotype. Non-parametric statistical analysis revealed that the difference between wild-type and atmrp4-2 transpiration rates is significant for the entire period 2–3 min after light is switched on until 15 min after light was switched off again (Mann–Whitney test, P < 0.05).

(b) The change in the transpiration after light is switched on is significantly larger in the atmrp4-2 mutant. Values were calculated by linear regression of transpiration rates within the first 10 min (means ± SEM; n = 10 for each genotype). Comparable differences were obtained when time constants of exponential fits were calculated as in (c).

(c) Time constants (means ± SEM; n = 10 for each genotype) describing the changes in transpiration rate when light was switched off. Constants were calculated by fitting the transpiration values of each data series to exponential equations (single exponential decay, three parameters). There is no significant difference in the velocity of decrease in transpiration between wild-type and atmrp4-2 plants.

(d) Transpiration (circles; left axis) and CO2 fixation rates (a; squares; right axis) in the light (400 µE m−2 sec−1 PAR) in response to changes in the partial pressure of CO2 in the air stream. Measurements were started in the morning at 9 am. After each change in [CO2], plants were equilibrated for 45 min before data points were recorded. Ws-2, open symbols; and atmrp4-2, closed symbols. n = 8 for each genotype.

Asterisks indicate that transpiration rates of wild-type and atmrp4-2 plants are significantly different (non-parametric analysis; Mann–Whitney test, P < 0.05).

Download figure to PowerPoint

As the ABA response, as discussed previously, was not affected in atmrp4, the capacity to close and open stomata in the light, in response to increases or decreases in Ca (ambient CO2 partial pressure), was investigated. Experiments were designed to include quick, drastic changes in Ca (e.g. from 1200 p.p.m. CO2 to nominally 0 p.p.m.) as well as step-by-step changes (Figure 5d). At high Ca(>800 p.p.m. CO2), no significant difference in the transpiration rate between the wild type and mutant was observed, indicating that the capacity to close the stomata in response to high CO2 was not affected by the mutation. In contrast, at Ca values of 400 p.p.m. CO2 or less, atmrp4-2 exhibited a clearly higher transpiration rate (1.6-fold higher at nominally 0 p.p.m. CO2) than the wild type. Therefore, the absence of AtMRP4 did not alter stomatal closure in response to high Ca, but did enhance opening at low Ca.

The antifolate methotrexate (MTX) inhibits stomatal opening in the light

Using the patch-clamp technique, changes in stomatal aperture and transpiration rates were investigated in relation to alterations in the activities of known guard cell ion channels, located at the plasma membrane. However, although the inwardly and outwardly rectifying potassium channels, as well as the slowly activating ABA-induced anion channel, were detected, no changes in their activities were observed between atmrp4 and wild-type guard cells (data not shown). It was therefore concluded that AtMRP4 does not act independently as an ion channel nor does it directly control channel activities, e.g. by physical interaction. However, we cannot exclude that the effects of AtMRP4 disruption on ion channel activities are too weak to be detected.

MRPs are known to be ATP-dependent membrane transporters for diverse amphiphilic organic anions. In our hands, heterologous expression of AtMRP4 under control of various promoters in the Δycf1Δbpt1 strain of Saccharomyces cerevisiae, which lacks detectable vacuolar glutathione conjugate transport activity (Klein et al., 2002; Sharma et al., 2002), failed to produce measurable ATP-dependent transport of the model organic anions 2,4-dinitrobenzene-glutathione and estradiol-17-(β-d-glucuronide). Recently, it was proposed that AtMRP4, like AtMRP1, could function as a high-capacity pump for folates and the antifolate drug MTX (Rea et al., 2003). Indeed, ATP-dependent uptake of tritiated MTX into microsomes isolated from Δycf1Δbpt1 yeasts transformed with full-length AtMRP4 is 2.1-fold higher than that into microsomes of empty vector-transformed yeasts (Figure 6a). Vanadate is the classical inhibitor of ABC-type transport processes. In the presence of vanadate, both the AtMRP4 transport and control transport activities were reduced to the same level arguing for the presence of additional transport mechanisms in yeasts, which cannot be attributed to ABC proteins. The difference of the vanadate-sensitive MTX-transport activity between microsomes from AtMRP4 transformants and empty vector transformants is consequently even higher (factor 3.6). We conclude that heterologously expressed AtMRP4 acts as a MTX transporter.


Figure 6. The antifolate MTX is actively transported by AtMRP4 (a) and affects stomatal opening in the light (b).

(a) ATP-dependent transport of [3H]-MTX into microsomal vesicles isolated from Δycf1Δbpt1 S. cerevisiae strains transformed with empty vector (pNEV) or AtMRP4 in pNEV vector (MRP4), respectively, in the absence or presence of 1 mm sodium vanadate.

(b) Differential sensitivity of stomatal opening in the light to MTX in Ws-2 wild-type (empty bars) and atmrp4-2 plants (filled bars). Dark-adapted detached leaves of Ws-2 and atmrp4-2 were incubated in the light in the presence and absence of the indicated concentrations of MTX. Data represent the means of stomatal pore width to length ratios and SEM. The asterisk indicates that the difference in the apertures of Ws-2 values in the absence and presence of 10 µm MTX is significant (t-test, P < 0.05).

Download figure to PowerPoint

MTX inhibits folate synthesis at the dihydrofolate reductase (DHFR) step (Cossins, 1987; Schweitzer et al., 1990). Although no information is available as to why plants should pump folates across the plasma membrane into the extracellular space, the mutation in AtMRP4 was investigated to determine whether it could perturb folate homeostasis in guard cells, leading ultimately to the stomatal phenotypes observed in atmrp4. The effects of increased concentrations of MTX on stomatal opening in the light were therefore investigated (Figure 6b).

Light-induced stomatal opening in wild-type plants was significantly reduced in the presence of 10 µm (to approximately 80% of opening in the absence of MTX). In contrast, light-induced stomatal opening in atmrp4-2 was not affected by MTX. Furthermore, the application of up to 100 µm MTX in the dark did not alter stomatal pore size in Ws-2 or atmrp4-2. In contrast, folic and folinic acid had no effect on light-induced opening in wild-type and mutant plants (data not shown). It was therefore concluded that light-induced stomatal opening was partially inhibited by the antifolate drug MTX, and that this inhibition was dependent on the presence of AtMRP4.


  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

The plasma membrane localization of AtMRP4 suggests novel functions for plant ABCC/MRP-type ABC transporters

Because of the transport activities found in isolated vacuoles and tonoplast vesicles with typical MRP substrates such as organic anions, it was suggested that this class of transporters is mainly involved in vacuolar storage of potentially toxic compounds, including certain secondary metabolites and heavy metals (Martinoia et al., 2000). However, experimental evidence for the tonoplast localization of plant MRP-type ABC transporters has been presented only for AtMRP2 and the wheat MRP TaMRP1, using tonoplast-enriched vesicle fractions (Liu et al., 2001; Theodoulou et al., 2003).

In the human genome, 13 genes of the ABCC/MRP subclass have been identified (Dean et al., 2001). Nearly all human and rodent MRPs catalyze the ATP-dependent vectorial transport of different organic anions, with varying affinities toward different substrates (König et al., 1999). Notable exceptions include the chloride channel CFTR (Anderson et al., 1991) and SUR, which together with Kir6.2, forms the functional K-ATP channel (Bryan and Aguilar-Bryan, 1999). Interestingly, a major factor establishing the functional diversity of the MRPs may be their location on different regions of the plasma membrane, especially in polarized mammalian cells. In A. thaliana and rice genomes, 14 and at least 17 MRP/ABCC-type genes can be found, respectively (Kolukisaoglu et al., 2002; H.U. Kolukisaoglu, unpublished). As the range of functions of these MRPs in planta remains to be elucidated, it is of interest to establish subcellular and tissue localizations of these proteins.

Using GFP-tagging and membrane fractionation techniques, we demonstrated that AtMRP4 is located at the plasma membrane (Figure 2c). This is, to our knowledge, the first plant ABCC/MRP localized to the plasma membrane, suggesting that vacuolar sequestration of toxins is not the only function of the MRPs in plants. While physiological functions for MRP-mediated export of organic anions are found for the delivery of catabolites and toxic substances across the canalicular membrane of liver hepatocytes in mammals, the contribution of plasma membrane MRPs in plants is unclear. Furthermore, the high expression of AtMRP4 in guard cells suggested that AtMRP4 may have specific roles in stomata. AtMRP4 is the second gene, after AtMRP5, which has been found to exhibit preferential expression in this highly differentiated cell type. Interestingly, the plasma membrane localization and the demonstration that the entire AtMRP4 protein functions as a methotrexate transporter after heterologous expression suggest that the predicted plastid transit peptide is not functional.

AtMRP4 and AtMRP5 have an opposite influence on stomatal regulation

Our recent analysis of a T-DNA knockout mutant in AtMRP5 revealed that the regulation of stomatal movements may be a novel function for plant MRPs (Gaedeke et al., 2001; Klein et al., 2003). However, the localization of AtMRP5 is unclear, and the mode of action of this protein in the context of guard cell regulation remains to be elucidated. Briefly, stomata of atmrp5-1 plants have a tendency to remain more closed than the wild type in the light and were insensitive to classical hormones regulating the aperture of the pore, most notably ABA (Klein et al., 2003). In accordance with these findings, atmrp5-1 plants have a higher water use efficiency (WUE) their leaves loose water more slowly when decapitated and are less sensitive to wilting.

Surprisingly, our data suggest that the guard cell ABC transporter AtMRP4 plays an opposite role in the regulation of stomata. Stomata of all atmrp4 mutants are more open in the light and dark. Bioassays using epidermal fragments immediately analyzed after detaching the leaves from the plants also show this effect (Figure 3). However, atmrp4 mutants, in contrast to abi1 and abi2 mutants (Pei et al., 1997), still respond to ABA and are therefore not affected by the ABA signal transduction pathway.

Using gas exchange analysis, it was demonstrated that the closing reaction in response to high CO2 partial pressure was not affected in atmrp4 mutants (Figure 5d). Like the ABA data, this suggests that the capacity to close stomata completely remains unchanged under normal watering conditions. By contrast, transpiration at atmospheric and lower CO2 doses was increased in atmrp4 mutants. Furthermore, transpiration rates of atmrp4-2 mutants increased faster in response to a dark–light transition but not when light was turned off again (Figure 5a–c). Although atmrp4-2 mutants finally reached a higher transpiration steady state in the light when compared to the Ws-2 wild-type plants (Figure 5a), a difference in the velocity of transpiration rate change during the ‘light on’ phase but not the ‘closing’ (light off) reaction (Figure 5b,c) suggests that disruption of AtMRP4 causes primarily defects in the regulation of stomatal opening. Interestingly, the mutation in AtMRP4 did not alter the photosynthetic CO2 fixation and the value of the CO2 compensation point (Figure 5d). Thus, the differences in the reaction of the stomata was not because of the fact that atmrp4 mutants consume more CO2, thereby lowering Ci, which could in turn be the signal to induce stomatal opening. As for the light response (Figure 5a), these results indicated that in mutants, the capacity to open stomata in response to low CO2 could be more drastically affected than the closing capacity. Alternatively or in addition, the atmrp4 mutant could be less sensitive to a process induced by low CO2 doses or high light that normally controls stomatal opening and restricts transpiration. Our data clearly show that at atmospheric CO2 partial pressure, the transpiration rate of atmrp4-2 mutants is increased in contrast to the wild type in the light. Thus, as disruption of AtMRP4 does not change photosynthesis, the atmrp4 shows a decreased WUE under ‘natural’ growth conditions, ultimately leading to faster water loss from rosettes and consequently earlier wilting mutant phenotypes (Figure 4). The notion that atmrp4 mutants have a decreased WUE is further supported by our results on carbon isotope discrimination, where mature leaves showed a more negative δ13C value (Figure 3c). δ13C represents a time-integrated value of the ratio of the intercellular CO2 (Ci) to ambient CO2 concentration (Ca), and thus the capacity to exchange gases via the stomata (Farquhar et al., 1989).

In the absence of data on the epistatic relationship or physical contact between AtMRP4 and AtMRP5, one can only speculate as to whether both or possibly more ABC transporters interact in stomatal regulation. The difference in the ABA sensitivity argues against a common regulatory pathway for AtMRP4 and AtMRP5 in guard cells. By analogy to the proposed functions of HsCFTR and HsSUR, we tested whether the disruption in AtMRP4 directly alters ion channel activities in A. thaliana guard cells. We were unable to detect any differences between the atmrp4 mutant plants when compared with wild-type plants for outward and inward rectifying K+ channel activities or the slowly activating ABA-induced anion channel. The AtMRP4 does not therefore represent this anion channel in spite of its homology to the chloride channel CFTR nor does it support the hypothesis that our phenotypic data could at least partially be explained by the absence or downregulation of this channel. However, it should be noted that for guard cell ion channels to our knowledge, only the guard cell outward rectifying K+ (GORK) channel, shows a stomatal phenotype that was identified by a mutant approach (Hosy et al., 2003). Interestingly, disruption of GORK, like mutations in AtMRP4 and AtMRP5, does not cause a severe stomatal phenotype. Absence of GORK seems to gain importance under certain stress conditions, although the corresponding channel activity is clearly absent in gork-1 plants.

AtMRP4 and folate transport: do folates play a role in guard cell regulation?

One possibility of explaining the phenotypes of atmrp4 mutants could be that AtMRP4 is not directly involved in ion fluxes across the guard cell plasma membrane but alters metabolic properties of the guard cells, leading to a larger stomatal pore size. Classical examples of metabolic steps involved in osmoregulation of guard cell turgor are the accumulation of sucrose as a solute balancing potassium ion efflux late during the day. Likewise, potassium is counterbalanced by the synthesis of malate in addition to chloride fluxes (Allaway, 1973; Talbott and Zeiger, 1996). In order to find a potential metabolic pathway that might be affected by the disruption of AtMRP4, we looked at potential substrates for this ABC transporter. It is well established that different mammalian MRP (ABCC) transporters are able to pump physiological folates and MTX, and to a lesser extent polyglutamylate out of the cell (Chen et al., 2002; Zeng et al., 2001). Here, we demonstrate that AtMRP4 is a functional ATP-dependent transporter for MTX in plants (Figure 6a).

Tetrahydrofolate polyglutamate coenzymes are essential co-factors for many one-carbon transfer reactions, including the synthesis of methionine, serine, purine, or thymidylate (McGuire and Bertino, 1981). In plants, tetrahydrofolate is made exclusively in mitochondria, but its derivatives also occur in other compartments (Ravanel et al., 2001). In pea leaves, folates are distributed in different ratios between a mitochondrial fraction, a chloroplast fraction, and a fraction including cytosol, vacuole, and the nucleus depending on the light condition (Jabrin et al., 2003; Neuburger et al., 1996). This implies the presence of folate transporters on different subcellular membranes. However, folate transport has not been studied in plants, except for the observation that selected MTX-resistant Datura innoxia cell lines accumulated less MTX within the cell (Wu et al., 1993). Despite no obvious role of folates in guard cells, or any information as to why plants should pump folates across the plasma membrane into the extracellular space (Figure 2b), we have investigated the possibility of a physiological link between folate transport via AtMRP4 and stomatal regulation. Only MTX but not folic and folinic acid reproducibly inhibited the opening of stomata in the light in wild-type plants, while atmrp4-2 mutants were insensitive to MTX (Figure 6b). In the dark, no differences in stomatal apertures were observed between wild-type and mutant plants.

The MTX inhibition of stomatal aperture can be interpreted in different ways: (i) AtMRP4 can not only transport physiological folates and/or MTX, but may also serve as a binding protein. One could assume that the AtMRP4-folate complex is able to interact with ion channels at the plasma membrane and that the ratio between folate-bound and folate-free AtMRP4 causes a ‘fine tuning’ of stomatal regulation. Addition of MTX could saturate the folate-binding site of AtMRP4 and could consequently maximize the inhibitory effect on stomatal opening. In this case, atmrp4 lacks the binding protein and the interacting channels would not be up- or downregulated by the AtMRP4–folate complex, representing a negative regulator of stomatal opening. However, the contribution of stomatal regulation via AtMRP4 may be low as suggested by the observation that the atmrp4 mutants do not show significant differences in ion channel activities, and that the wilting phenotype is less severe than in abi1 (Figure 4). (ii) A complex metabolic regulation might also take place. In C3 plants, folates support huge metabolic fluxes, mainly participating in the mitochondrial glycine-to-serine conversion by the tetrahydrofolate-mediated glycine decarboxylase/serine hydromethyltransferase complex during photorespiration, which is considered to be the major pathway for the generation of single-carbon units via serine (Bauwe and Kolukisaoglu, 2003; Hanson and Gregory, 2002). However, we have no evidence of the activity of this pathway in guard cells and how changes in metabolite levels implicated in this pathway would affect stomatal regulation. Furthermore, the analysis of amino acid levels in whole leaves of wild-type and atmrp4 mutants did not show differences in the levels of glycine and serine, giving no direct evidence of the participation of AtMRP4 (data not shown).

In summary, our results show that the regulation of stomatal opening and therefore the interaction with the gaseous environment of plants implicate the coordinated action of several MRP-type ABC transporters (AtMRP4 and AtMRP5) that are involved in a complex process that integrates fast-changing environmental signals into hormonal and metabolic pathways. For ABC transporters, it is well known that single proteins can fulfill multiple functions. Our analysis of AtMRP4 demonstrates that biochemical assays of transport activities, examination of expression and mutants reveal different aspects of the same protein. We have no information on the physiological function of AtMRP4 in tissues other than guard cells where AtMRP4 is still expressed (Figure 2). In this respect, it could be argued that AtMRP4 is a multipurpose membrane protein. It is, however, likely that one function of AtMRP4 is the transport of folates. It will be interesting to investigate folate transport on the whole-plant level and to examine whether a folate transport function is indeed a metabolic control step in guard cell regulation.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Plant culture

Arabidopsis thaliana plants were grown on soil in growth chambers with either 8- or 16-h light period (21°C; relative humidity 70%; 200 µmol m−2 sec−1 PAR). If not stated otherwise, plants were watered twice a week. Plant material for all physiological experiments was grown on soil for 3 weeks, and then single plants were transferred into new pots. Single plants were grown in the phytotron with 8 h of light and were used 3 weeks after transfer. V. faba plants were grown for 3 weeks in a greenhouse.

Cloning of AtMRP4 cDNA

The cDNA of AtMRP4 was isolated by screening a size-selected Arabidopsis cDNA library (Tommasini et al., 1997). Based on the sequence information of expressed sequence tag (EST) 147I22T7 (Accession number U96399) and the genomic AtMRP4 sequence deposited by Sanchez-Fernandez et al. (1998; Accession number AJ002584), an intron-less, 397-bp, AtMRP4-specific, PCR-amplified fragment starting from position +1581 was used as a radiolabeled probe. Clone pBSC-MRP4-2 was found to contain the longest ORF of AtMRP4 but lacking the first 726 bp. The 5′ end of AtMRP4 was cloned by PCR on genomic DNA using the Expand High Fidelity PCR System (Roche, Rotkreuz, CH). Using a NotI restriction site that was located in the upper primer (UP; 5′-gct gcggccgc a gtcgaccccggg atg tgg ttg ctt tcg tct tct cca tg; underlined are NotI, SalI, and XmaI sites, respectively; lower primer (LP) 4A-a) and an AtMRP4 endogenous BamHI site (bp 1.548 of the amplified 5′ end), the first 822 bp of the incomplete AtMRP4 ORF in pBSC-MRP4-2 were exchanged bp 1–1.548 of AtMRP4, resulting in vector pBSC-MRP4-full.

An RGSH6 motif and a NotI site (underlined) were added to the C-terminus of AtMRP4 by exchanging the very last 71 bp of its 3′ end with the insert of clone pGEM-71 using an endogenous SnoI site (bp 4.377). Therefore, the 3′ end of pBSC-MRP4-2 was amplified by PCR (UP, 5′-agg aag cag gag ttg ttt aga caa gag; and LP, 5′-cat ctcgag gcggccgc tctaga tca gtg atg gtg atg gtg atg aga acc acg tat tcg ggc aga tcg gag) and cloned into pGEM-T Easy (Promega, Wisconsin, USA), resulting in pGEM-71. NotI–SnoI and SnoI–NotI fragments were excised from pBSC-MRP-full and pGEM-71, respectively, and a three-point ligation into pYES-2 (Invitrogen, Basel, CH) resulted in pYES-MRP4. The entire insert was sequenced to verify the absence of PCR errors.

Detection of AtMRP4-specific transcripts by RT-PCR

Total RNA from Arabidopsis soil-grown plants was prepared using the RNeasy Plant Kit (Qiagen, Hilden, Germany). DNaseI (Qiagen) treatment was performed with column-bound RNA. Oligo-dT-primed cDNA from 1 µg of total RNA was synthesized using the Reverse Transcription System (Promega). Transcripts specific for AtMRP4 and 40S ribosomal protein S16 (Accession number F19995) were detected by PCR for 25–30 cycles at 52°C annealing temperature. RT-PCR primers used were: 4B-s, 4B-a (see below); S16-UP, 5′-ggc gac tca acc agc tac tga; and S16-LP, 5′-cgg taa ctc ttc tgg taa cga.

Five micrograms of total RNA and equal volumes of PCR products were separated on 1% formaldehyde and 2.5% native agarose gels, respectively, and fluorescence of ethidium bromide-stained bands was quantified using the scion image software, version 3.62a ( Negative controls in the absence of enzyme in the RT reaction yielded no products. Absence of genomic DNA contaminations was further verified by fragment size determination as the 4B-s/4B-a amplification product spanned an intron.

Promoter–GUS fusion and GUS expression analysis

The 5′ UTR of AtMRP4 (positions −1996 to −1 relative to the start codon) was amplified by PCR using DNA from bacterial artificial chromosome (BAC) clone F17A22 (Accession number AC005309) as a template and the High Fidelity Expand PCR Kit. Primers used were: UP, 5′-cac ctt gcc gtt tca gcg aaa ttg; and LP, 5′-ttt tcc cag tga aaa aat at. The PCR product was cloned into the pENTRY D-TOPO vector according to the manufacturer's instructions (Invitrogen), resulting in pENTR-MRP4Pr in which the sequence of the promoter fragment was verified. This construct served as the entry vector to transfer the promoter fragment into promoterless-GUS and -GFP reporter vectors via the GATEWAY™ system using the LR reaction according to the manufacturer's instructions. The binary destination vectors used were pMDC102 (GFP6) and pMDC108 (GUS; Curtis and Grossniklaus, 2003), resulting in pM102-MRP4Pr and pM108-MRP4Pr, respectively. Wild-type A. thaliana Ws-2 plants were transformed by floral dipping using Agrobacterium strain GV3101 (Clough and Bent, 1998). More than 40 T1 transformants for each construct were selected on half-strength MS medium agar plates containing hygromycin for 7 days and were subsequently transferred to fresh sterile agar plates without antibiotic until the reporter gene was visualized (8-h light period or complete darkness). Biolistic bombardment into V. faba guard cells with pM108-MRP4Pr was performed as described by Jung et al. (2002). GUS activity was assayed as described by Rashotte et al. (2001) using 2 mm X-Gluc (Biosynth, Stadd, Switzerland) as substrate. Incubation at 37°C was stopped after 3 h. GUS staining of whole seedlings was analyzed using a Nikon SMZ1500 binocular, and pictures were taken with a Nikon Coolpix camera. Detailed GUS analysis was performed with a Leica DMR microscope equipped with a Leica DC300 F charge coupled device (CCD) camera and controlled by Leica im1000 software (Leica, Heerbrugg, CH). Digital pictures were processed using adobe photoshop 7.0 software (Adobe Systems) without changing intensities or color properties. GFP transformants inspected by epifluorescence microscopy with a fluorescein isothiocyanate filter set exhibited GFP fluorescence in seedlings which was indistinguishable from the results of the GUS staining (data not shown).

Localization of AtMRP4–GFP

The AtMRP4 cDNA was subcloned NotI–NotI from vector pYES-MRP4 into vector pRTΩ (Überlacker and Werr, 1996), resulting in pRTΩ–MRP4. The entire gene of an enhanced GFP version (EGFP) was PCR amplified (UP, 5′-act ggtacc gga gta aag gag aag aac ttt tca c; and LP 5′-agt ggtacc att tgt ata gtt cat cca tgc cat g) and inserted into the KpnI (bp 3.323) site of AtMRP4. The EGFP fusion was sequenced to verify the absence of PCR errors. The cassette containing the CaMV 35S promoter, AtMRP4–EGFP and the polyadenylation signal was excised with AscI and inserted into vectors pGPTV-bar (Überlacker and Werr, 1996), and the resulting binary construct pGPTV–MRP4–GFP was used to transform Arabidopsis wild-type plants. Resistant transformants were selected on soil by BASTA watering (1 : 20 000).

Homozygous transgenic Arabidopsis plants were grown for 5–7 days on MS plates under continuous light. Onion epidermis cell layers were transfected with construct pGPTV–MRP4–GFP using a low-pressure partical inflow gun. Transfected onion epidermis cells and young Arabidopsis seedlings were analyzed by confocal laser scanning microscopy (CLSM). FITC fluorescence using the corresponding filter sets was recorded, and then stored images were coloured green using adobe photoshop. Arabidopsis microsomes were prepared from light-grown liquid-shaking cultures and separated by continuous sucrose gradient centrifugation as described by Geisler et al. (2000). The anti-GFP antibody used was Living Colors a.V. (JL-8) Monoclonal Antibody (Clontech, Palo Alto, CA, USA).

Yeast strains, preparation of yeast membrane vesicles, and in vitro transport studies

The full-length sequence of AtMRP4 was cloned NotI–NotI into the yeast expression vector pNEV (Sauer and Stolz, 1994), resulting in pN-AtMRP4. pNEV, and pN-AtMRP4 were transformed into the Saccharomyces cerevisiae YMK2 strain (MATα ura3–52 leu2–3112 his6Δycf1::HIS6Δbpt1::LEU2; Klein et al., 2002) by standard procedures. Isolation of yeast microsomal vesicles from yeasts grown in synthetic dropout (SD) selective medium and uptake experiments with [3′,5′,7-3H]methotrexate (Na salt; 292 GBq mmol−1, Amersham; final activity 1 µCi ml−1) using the rapid filtration technique (Durapore, Millipore) followed published protocols (Klein et al., 2002).

Isolation of atmrp4 mutants

A collection of 32 880 T-DNA insertion lines available through the Arabidopsis Stock Centers at Ohio State (ABRC) and the University of Nottingham (NASC) were screened for an AtMRP4 knockout allele. To screen the entire gene region of AtMRP4 and non-translated 5′ and 3′ sequences of AtMRP4, a set of primers was designed (mrp4A-s, mrp4A-a, mrp4B-s, and mrp4B-a; Table 1). These primers were used in combination with T-DNA-derived border primers in PCRs with DNAs from T-DNA-transformed plants arranged in pools of 1000 plants. Subsequent rounds of re-amplification of PCR products and generation of hybridization probes were performed with nested PCR primers (mrp4An-s, mrp4An-a, mrp4Bn-s, and mrp4Bn-a) to avoid cross-reactions with the primary PCR primer sequences. T-DNA insertions in AtMRP4 were identified in Arabidopsis lines generated by Bechthold et al. (1993; atmrp4-2) and Forsthoefel et al. (1992; atmrp4-1 and atmrp4-A) and confirmed by sequencing of isolated PCR products, amplified with combinations of gene-specific and T-DNA border primers as described previously by Gaedeke et al. (2001). Individual lines, homozygous for the T-DNA insertions, were subsequently isolated from plant populations of 400 plants corresponding to positively scored DNA pools, which were derived from 20 plants each. Homozygous plants were identified by PCR amplification with a combination of gene-specific and T-DNA primers, heterozygous plants yielded additional PCR products with a combination of two gene-specific primers encompassing the respective T-DNA insertion (Table 1). Southern blot analyses were performed with T-DNA-specific as well as AtMRP4-specific DNA probes. Co-segregation of T-DNA insertion and mutant phenotype was ascertained by crossing homozygous atmrp4 mutant plants with the corresponding wild types. Segregation analysis of F2 progeny revealed 3 : 1 ratio for the dominant selection markers BASTA or kanamycin. Co-segregation was shown by proving PCR fragment amplification with combinations of gene-specific and T-DNA border primers only with DNAs from T-DNA-containing plants. Offspring of homozygous lines segregated in the F3 generation as 100% resistant plants, while hemizygous plants segregated again in a 3 : 1 ratio. An additional line (atmrp4-3) that contained a Ds insertion in the AtMRP4 gene was found in the Cold Spring Harbor Laboratory Genetrap DB (Martienssen, 1998) by database searches (ET 1399). This insertion was verified in single lines by PCR using primer combinations lmrp4A-s and Ds5-4. The wild-type allele was detected by PCR with the primer combination lmrp4A-s and lmrp4A-a. Primers mrp4A-s and mrp4An-s in combination with Ds5-4 were used for generating probes and sequencing. Expression of AtMRP4 and the S16 gene as control in atmrp4 mutant as well as wild-type plants were performed with first-strand cDNA (30 cycles: 1 min at 94°C, 30 sec at 58°C, and 1 min at 72°C with end elongation after the last cycle) in an MJ Research PTC200 cycler.

Table 1.  PCR primers used for identification and verification of atmrp4 insertion mutants, for generating hybridization probes and RT-PCR analysis
Primer namePrimer sequenceMutation screened
mrp4Bn-s5′-AGGAAGCAGGAGTTGTTTAGACAAGAGatmrp4-1, atmrp4-3

Stomatal aperture measurements

All measurements were started in the morning after 15–16 h of darkness. If not stated otherwise, leaves from a minimum of three independent single plants per genotype and experiment were detached and floated on 20 mm KCl and 0.5 mm CaCl2 in Petri dishes. Light-induced stomatal opening was analyzed after 2.5 h in the light (200 µmol m−2 sec−1) followed by 4 h in complete darkness in order to measure dark-induced closure. Parts of the abaxial epidermis were peeled off with sharp forceps, transferred into a drop of floating solution on a glass slide, and immediately processed for microscopical analysis of stomatal apertures. Bright-field pictures of stomata were taken with the Leica DMR/DC 300F setup. The width and length of the stomatal pores was determined using the measuring tool integrated in the im1000 software. MTX and ABA experiments were performed by adding the substance concentrations indicated at the beginning of the treatment in the light. In the case of MTX, the floating solution was buffered with 10 mm 2-morpholinoethanesulphonic acid (MES)–KOH, pH 5.8. A 10 mm MTX stock was prepared by dissolving the drug in the floating buffer. ABA (mixed isomers, Fluka) was dissolved as a 10 mm stock in MeOH. Stomata measurements of leaves from intact plants were performed by detaching the leaves, peeling off epidermal fragments, and microscopic image acquisition within less than 10 min. Samples were collected 30 min before the end of the dark night period and 90 min after the beginning of illumination.

Isotope discrimination measurements

For the determination of stable carbon isotope ratio, rosette leaves of 6-week-old single plants were collected in two fractions: the ‘young leaves’ fraction consisted of the 10 youngest leaves counting from the last rosette leaf formed. Leaves were counted when they reached a length of 5 mm. The ‘mature leaves’ fraction contained the remaining, fully expanded leaves. All determinations were repeated twice, each with four independent replicates (plants) per genotype. Leaf material was dried in an oven at 70°C for 36 h and ground to a fine powder. The powdered material was then weighed into tin capsules and subjected to isotope ration mass spectrometry (IRMS) as described by Wanek et al. (2001). The continuous-flow IRMS system consisted of an elemental analyser (EA 1110, CE Instruments, Milan, Italy) connected to a gas isotope ratio mass spectrometer (DeltaPlus, Finnigan MAT, Bremen, Germany). Reference gases were calibrated to the Vienna Pee Dee Belemnite (V-PBD) standard using IAEA-CH-6 and IAEA-CH-7 as reference material (International Atomic Energy Agency, Vienna). The natural abundance of 13C in the samples is reported as follows:

  • image

where R is the ratio of the mass 13/mass 12 for the sample and standard material, respectively. The standard deviation of repeated measurements of a laboratory standard was 0.10‰ for δ13C.

Gas exchange measurements

Single-leaf gas exchange measurements were performed with LI-COR LI-6400 photosynthesis systems (DMP, Fehraltorf, Switzerland) equipped with light emitting diode (LED) light source 6400-02 fixed on top of the 6 cm2 standard cuvette/infrared gas analyzer (IRGA). Fully expanded leaves of comparable developmental status were chosen by counting the leaf number from the center of the rosette (in general leaves 8–10). Gas flow was adjusted to a constant rate of 0.25 mmol sec−1. CO2 partial pressure and light intensities were varied as indicated. Relative humidity was 60%, and the chamber temperature was set to 22°C. Measurements were started at fixed times using plants that were adapted to darkness for at least 16 h (night period).

Kinetics of water loss from excised rosettes and drought tolerance measurements

In this set of experiments, plants were grown in pots with equal amounts of soil. Water loss determinations from excised rosettes were carried out under standard conditions as described by Klein et al. (2003). Soil water loss was determined by wrapping pots and the soil surface with saran wrap, leaving only the rosette leaves in contact with the atmosphere to reduce soil evaporation. Before irrigation was terminated, the plants were watered once more by drenching for 2 h. At the indicated times, the weight of the pots was determined. Water content was calculated as the ratio m(t)/DW where m(t) is the pot mass at time t, DW is the dry weight and were expressed as percentage of the initial value m(t0)/DW. Dry weights were obtained after incubation of the wrapped pots including the plant to complete dryness at 70°C for 7 days at the end of the experiment.


Statistical analysis was performed using spss 11.5, and linear and exponential curve fitting was performed using sigmaplot version 8.02 (SPSS Inc., Chicago, USA).


  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Arabidopsis T-DNA lines were obtained through Randy Scholl and Sean May from the ABRC and NASC Arabidopsis Stock Centers at Ohio State University and University of Nottingham, respectively. Ds insertion line ET1399 was supplied by Joe Simorowski from the Cold Spring Harbor Laboratory and abi1 was donated by Erwin Grill, University of Munich. We thank Bernhard Schmid, University of Zurich, for lending us a second LI-COR setup. The authors thank Alain Vavasseur (CEA Cadarache, France) Marjolaine Girin, Aurélie Pédezert, Nadège Fahrni, and Lucien Bovet (all University of Neuchatel) for discussion and technical help, Christoph Ringli (University of Zurich) for help with Southern blots, Katrin Czempinski and Bernd Müller-Röber (University of Potsdam) for the KCO1–GFP construct, Felix Mauch (University of Fribourg) for providing the partical inflow gun, Bruno Stieger and Bruno Hagenbuch (University Hospital Zurich) for 3H-MTX. B.S. likes to thank Ulf-Ingo Flügge for continuous support. This work was supported by the Swiss National Foundation, Deutsche Forschungsgemeinschaft (Schu/821-2), the European Community (LATIN, BIOTEC 4), the Ministerium für Schule, Wissenschaft und Forschung des Landes NRW, the Post-doctoral Fellowship Program of Korea Science and Engineering Foundation (KOSEF; S.J.S.), Novartis and the Alexander von Humboldt-Foundation (Feodor-Lynen fellowship to M.K. and M.G. and Novartis fellowship to M.G.).


  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  • Allaway, W.G. (1973) Accumulation of malate in guard cells of Vicia faba during stomatal opening. Planta, 110, 6370.
  • Anderson, M.P., Gregory, R.J., Thompson, S., Souza, D.W., Paul, S., Mulligan, R.C., Smith, A.E. and Welsh, M.J. (1991) Demonstration that CFTR is a chloride channel by alteration of its anion selectivity. Science, 253, 202205.
  • Bauer, J., Hiltbrunner, A., Weibel, P., Vidi, P.A., Alvarez-Huerta, M., Smith, M.D., Schnell, D.J. and Kessler, F. (2002) Essential role of the G-domain in targeting of the protein import receptor atToc159 to the chloroplast outer membrane. J. Cell. Biol. 159, 845854.
  • Bauwe, H. and Kolukisaoglu, H.Ü. (2003) Genetic manipulation of glycine decarboxylation. J. Exp. Bot. 54, 15231535.
  • Bechthold, N., Ellis, J. and Pelletier, G. (1993) In planta Agrobacterium mediated gene transfer by infiltration of adult Arabidopsis thaliana plants. C. R. Acad. Sci. 316, 11941199.
  • Bryan, J. and Aguilar-Bryan, L. (1999) Sulfonylurea receptors: ABC transporters that regulate ATP-sensitive K+ channels. Biochim. Biophys. Acta, 1461, 285303.
  • Chen, Z.S., Lee, K., Walther, S., Raftogianis, R.B., Kuwano, M., Zeng, H. and Kruh, G.D. (2002) Analysis of methotrexate and folate transport by multidrug resistance protein 4 (ABCC4): MRP4 is a component of the methotrexate efflux system. Cancer Res. 62, 31443150.
  • Clough, S.J. and Bent, A.F. (1998) Floral dip: a simplified method for Agrobacterium-mediated transformation of Arabidopsis thaliana. Plant J. 16, 735743.
  • Cossins, E.A. (1987) Folate biochemistry and the metabolism of one-carbon units . In The Biochemistry of Plants, Vol. 11 (Davies, D.D., ed.). New York: Academic Press, pp. 317353.
  • Curtis, M.D. and Grossniklaus, U. (2003) A gateway cloning vector set for high-throughput functional analysis of genes in planta. Plant Physiol., 133, 462469.
  • Czempinski, K., Frachisse, J.M., Maurel, C., Barbier-Brygoo, H. and Mueller-Roeber, B. (2002) Vacuolar membrane localization of the Arabidopsis‘two-pore’ K+ channel KCO1. Plant J. 29, 809820.
  • Dean, M., Rzhetsky, A. and Allikmets, R. (2001) The human ATP-binding cassette (ABC) transporter superfamily. Genome Res. 11, 11561166.
  • Farquhar, G.D., Ehleringer, J.R. and Hubick, K.T. (1989) Carbon isotope discrimination and photosynthesis. Annu. Rev. Plant Physiol. Plant Mol. Biol. 40, 503537.
  • Forsthoefel, N.R., Wu, Y., Schulz, B., Bennett, M.J. and Feldmann, K.A. (1992) T-DNA insertion mutagenesis in Arabidopsis: prospects and perspectives. Aust. J. Plant Physiol. 19, 353366.
  • Gaedeke, N., Klein, M., Kolukisaoglu, H.Ü. et al. (2001) The Arabidopsis thaliana ABC transporter AtMRP5 controls root development and stomata movement. EMBO J. 20, 18751887.
  • Geisler, M., Frangne, N., Gomès, E., Martinoia, E. and Palmgren, M.G. (2000) The ACA4 gene of Arabidopsis encodes a vacuolar membrane calcium pump that improves salt tolerance in yeast. Plant Physiol. 124, 18141827.
  • Gottesman, M.M. and Pastan, I. (1993) Biochemistry of multidrug resistance mediated by the multidrug transporter. Annu. Rev. Biochem. 62, 385427.
  • Grabov, A. and Blatt, M. (1998) Co-ordination of signalling elements in guard cell ion channel control. J. Exp. Bot. 49, 351360.
  • Hanson, A.D. and Gregory, J.F., III (2002) Synthesis and turnover of folates in plants. Curr. Opin. Plant Biol. 5, 244249.
  • Henikoff, S., Greene, E.A., Pietrokovski, S., Bork, P., Attwood, T.K. and Hood, L. (1997) Gene families: the taxonomy of protein paralogs and chimeras. Science, 278, 609614.
  • Higgins, C.F. (1992) ABC transporter: from microorganisms to man. Annu. Rev. Cell Biol. 8, 67113.
  • Higgins, C.F. (1995) The ABC of channel regulation. Cell, 82, 693696.
  • Hosy, E., Vavasseur, A., Mouline, K. et al. (2003) The Arabidopsis outward K+ channel GORK is involved in regulation of stomatal movements and plant transpiration. Proc. Natl. Acad. Sci. USA, 100, 55495554.
  • Jabrin, S., Ravanel, S., Gambonnet, B., Douce, R. and Rebeille, F. (2003) One-carbon metabolism in plants. Regulation of tetrahydrofolate synthesis during germination and seedling development. Plant Physiol. 131, 14311439.
  • Jung, J.-Y., Kim, Y.-W., Kwak, J.M., Hwang, J.-U., Young, J., Schroeder, J.I., Hwang, I. and Lee, Y. (2002) Phosphatidylinositol 3- and 4-phosphate are required for normal stomatal movements. Plant Cell, 14, 23992412.
  • Kinoshita, T., Doi, M., Suetsugu, N., Kagawa, T., Wada, M. and Shimazaki, K.I. (2001) Phot1 and phot2 mediate blue light regulation of stomatal opening. Nature, 414, 656660.
  • Klein, M., Mamnun, Y.M., Eggmann, T., Schuller, C., Wolfger, H., Martinoia, E. and Kuchler, K. (2002) The ATP-binding cassette (ABC) transporter Bpt1p mediates vacuolar sequestration of glutathione conjugates in yeast. FEBS Lett. 520, 6367.
  • Klein, M., Perfus-Barbeoch, L., Frelet, A., Gaedeke, N., Reinhardt, D., Mueller-Roeber, B., Martinoia, E. and Forestier, C. (2003) The plant multidrug resistance ABC transporter AtMRP5 is involved in guard cell hormonal signalling and water use. Plant J. 33, 119129.
  • Kolukisaoglu, H.Ü., Bovet, L., Klein, M., Eggmann, T., Geisler, M., Wanke, D., Martinoia, E. and Schulz, B. (2002) Family business: the multidrug-resistance related protein (MRP) ABC transporter genes in Arabidopsis thaliana. Planta, 216, 107119.
  • König, J., Nies, A.T., Cui, Y., Leier, I. and Keppler, D. (1999) Conjugate export pumps of the multidrug resistance protein (MRP) family: localization, substrate specificity and MRP2-mediated drug resistance. Biochim. Biophys. Acta, 1461, 377394.
  • Leckie, C., McAinsh, M., Montgomery, L., Priestley, A., Staxen, I., Webb, A. and Hetherington, A. (1998) Abscisic acid induced stomatal closure mediated by cyclic ADP-ribose. J. Exp. Bot. 49, 339349.
  • Leonhardt, N., Marin, E., Vavasseur, A. and Forestier, C. (1997) Evidence for the existence of a sulfonylurea-receptor-like protein in plants: modulation of stomatal movements and guard cell potassium channels by sulfonylureas and potassium channel openers. Proc. Natl. Acad. Sci. USA, 94, 1415614161.
  • Leonhardt, N., Vavasseur, A. and Forestier, C. (1999) ATP binding cassette modulators control abscisic acid-regulated slow anion channels in guard cells. Plant Cell, 11, 11411151.
  • Liu, G., Sanchez-Fernandez, R., Li, Z.-S. and Rea, P.A. (2001) Enhanced multispecifity of Arabidopsis multidrug resistance-associated protein-type ATP-binding cassette transporter, AtMRP2. J. Biol. Chem. 267, 86488656.
  • Martienssen, R. (1998) Functional genomics: probing plant gene function and expression with transposons. Proc. Natl. Acad. Sci. USA, 95, 20212026.
  • Martinoia, E., Grill, E., Tommasini, R., Kreuz, K. and Amrheim, N. (1993) ATP-dependent glutathione S-conjugate ‘export’ pump in the vacuolar membrane of plants. Nature, 364, 247249.
  • Martinoia, E., Klein, M., Geisler, M., Sánchez-Fernández, R. and Rea, P.A. (2000) Vacuolar transport of secondary metabolites and xenobiotics. In Vacuolar Compartments. Annual Plant Reviews, Vol. 5 (Robinson, D.G. and Rogers, J.C., eds). Sheffield: Academic Press, pp. 221253.
  • Martinoia, E., Klein, M., Geisler, M., Bovet, L., Forestier, C., Kolukisaoglu, H.U., Mueller-Roeber, B. and Schulz, B. (2002) Multifunctionality of plant ABC transporters: more than just detoxifiers. Planta, 214, 345355.
  • McGuire, J.J. and Bertino, J.R. (1981) Enzymatic synthesis and function of folylpolyglutamates. Mol. Cell. Biochem. 38, 1948.
  • Mustilli, A., Merlot, S., Vavasseur, A., Fenzi, F. and Giraudat, J. (2002) Arabidopsis OST1 protein kinase mediates the regulation of stomatal aperture by abscisic acid and acts upstream of reactive oxygen species production. Plant Cell, 14, 30893099.
  • Neuburger, M., Rébeillé, F., Jourdain, A., Nakamura, S. and Douce, R. (1996) Mitochondria are the major site for folate and thymidylate synthesis in plants. J. Biol. Chem. 271, 94669472.
  • Pei, Z.M., Kuchitsu, K., Ward, J.M., Schwarz, M. and Schroeder, J.I. (1997) Differential abscisic acid regulation of guard cell slow anion channel in Arabidopsis wild-type and abi1 and abi2 mutants. Plant Cell, 9, 409423.
  • Rashotte, A.M., DeLong, A. and Muday, G.K. (2001) Genetic and chemical reductions in protein phosphatase activity alter auxin transport, gravity response, and lateral root growth. Plant Cell, 13, 16831697.
  • Ravanel, S., Cherest, H., Jabrin, S., Grunwald, D., Surdin-Kerjan, Y., Douce, R. and Rebeille, F. (2001) Tetrahydrofolate biosynthesis in plants: molecular and functional characterization of dihydrofolate synthetase and three isoforms of folylpolyglutamate synthetase in Arabidopsis thaliana. Proc. Natl. Acad. Sci. USA, 98, 1536015365.
  • Rea, P.A., Sánchez-Fernández, R., Chen, S., Peng, M., Klein, M., Geisler, M. and Martinoia, E. (2003) The plant ABC transporter superfamily: the functions of a few and identities of many. In ABC Proteins – from Bacteria to Man (Holland, I.B., Cole, S.P.C., Kuchler, K. and Higgins, C.F. eds). New York: Academic Press, pp. 335355.
  • Sanchez-Fernandez, R., Ardiles-Diaz, W., Van Montagu, M., Inze, D. and May, M.J. (1998) Cloning and expression analyses of AtMRP4, a novel MRP-like gene from Arabidopsis thaliana. Mol. Gen. Genet. 258, 655662.
  • Sauer, N. and Stolz, J. (1994) SUC1 and SUC2: two sucrose transporters from Arabidopsis thaliana; expression and characterization in baker's yeast and identification of the histidine-tagged protein. Plant J. 6, 6777.
  • Schmid-Antomarchi, H., De Weille, J., Fosset, M. and Lazdunski, M. (1987) The antidiabetic sulfonylurea glibenclamide is a potent blocker of the ATP-modulated K+ channel in insulin secreting cells. Biochem. Biophys. Res. Commun. 146, 2125.
  • Schönknecht, G., Spoormaker, P., Steinmeyer, R., Bruggeman, L., Ache, P., Dutta, R., Reintanz, B., Godde, M., Hedrich, R. and Palme, K. (2002) KCO1 is a component of the slow-vacuolar (SV) ion channel. FEBS Lett. 511, 2832.
  • Schroeder, J.I., Allen, G.J., Hugouvieux, V., Kwak, J.M. and Waner, D. (2001a) Guard cell signal transduction. Annu. Rev. Plant Physiol. Plant Mol. Biol. 52, 627658.
  • Schroeder, J.I., Kwak, J.M. and Allen, G.J. (2001b) Guard cell abscisic acid signal transduction network and engineering of plant drought hardiness. Nature, 410, 327330.
  • Schultz, B.D., Deroos, A.D.G., Venglarik, C.J., Singh, A.K., Frizzell, R.A. and Bridges, R.J. (1996) Glibenclamide blockade of CFTR chloride channels. Am. J. Physiol. 15, L192L200.
  • Schwacke, R., Schneider, A., Van Der Graaff, E., Fischer, K., Catoni, E., Desimone, M., Frommer, W.B., Flügge, U.I. and Kunze, R. (2003) ARAMEMNON, a novel database for Arabidopsis integral membrane proteins. Plant Physiol. 131, 1626.
  • Schweitzer, B.I., Dicker, A.P. and Bertino, J.R. (1990) Dihydrofolate reductase as a therapeutic target. FASEB J. 4, 24412452.
  • Sharma, K.G., Mason, D.L., Liu, G., Rea, P.A., Bachhawat, A.K. and Michaelis, S. (2002) Localization, regulation, and substrate transport properties of Bpt1p, a Saccharomyces cerevisiae MRP-type ABC transporter. Eukaryot Cell, 1, 391400.
  • Szyroki, A., Ivashikina, N., Dietrich, P. et al. (2001) KAT1 is not essential for stomatal opening. Proc. Natl. Acad. Sci. USA, 98, 29172921.
  • Talbott, L.D. and Zeiger, E. (1996) Central roles for potassium and sucrose in guard-cell osmoregulation. Plant Physiol. 111, 10511057.
  • Theodoulou, F.L., Clark, I.M., He, X.L., Pallett, K.E., Cole. D.J. and Hallahan, D.L. (2003) Co-induction of glutathione-S-transferases and multidrug resistance associated protein by xenobiotics in wheat. Pest Manag. Sci. 59, 202214.
  • Tommasini, R., Vogt, E., Schmid, J., Fromentau, M., Amrhein, N. and Martinoia, E. (1997) Differential expression of genes coding for ABC transporters after treatment of Arabidopsis thaliana with xenobiotics. FEBS Lett. 411, 206210.
  • Überlacker, B. and Werr, W. (1996) Vectors with rare cutter restriction enzyme sites for expression of open reading frames in transgenic plants. Mol. Breed. 2, 293295.
  • Walker, J.E., Saraste, M., Runswick, M.J. and Gay, N.J. (1982) Distantly related sequences in the α- and β-subunits of ATP synthase, myosin, kinases and other ATP-requiring enzymes and a common nucleotide binding fold. EMBO J. 1, 945951.
  • Wanek, W., Heintel, S. and Richter, A. (2001) Preparation of starch and other carbon fractions from higher plant leaves for stable carbon isotope analysis. Rapid Commun. Mass Spectrom. 15, 11361140.
  • Wang, X.-Q., Ullah, H., Jones, A.M. and Assmann, S.M. (2001) G protein regulation of ion channels and abscisic acid signaling in Arabidopsis guard cells. Science, 292, 20702072.
  • Wu, K., Atkinson, I.J., Cossins, E.A. and King, J. (1993) Methotrexate resistance in Datura innoxia– uptake and metabolism of methotrexate in wild-type and resistant cell lines. Plant Physiol. 101, 477483.
  • Zeng, H., Chen, Z.S., Belinsky, M.G., Rea, P.A. and Kruh, G.D. (2001) Transport of methotrexate (MTX) and folates by multidrug resistance protein (MRP) 3 and MRP1: effect of polyglutamylation on MTX transport. Cancer Res. 61, 72257232.