INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Plants represent a complex mosaic of carbon sources (the mature leaves) and numerous sink tissues that all depend on the supply of organic carbon, mainly from the mature leaves via the long-distance transport system of the phloem (Williams et al., 2000). The transit of carbohydrates from the phloem into these various sink tissues, but also the post-phloem transport within a sink organ or from one sink to its adjacent sink, can occur symplastically via plasmodesmata or apoplastically across the plasma membrane (Büttner and Sauer, 2000; Williams et al., 2000). In most plants, Suc and hexoses are the main substrates for this carrier-mediated transmembrane transport: Suc, because it is delivered by the phloem, and monosaccharides, because Suc is frequently hydrolyzed by extracellular invertases (Ward et al., 1998; Büttner and Sauer, 2000; Williams et al., 2000).

Gene families have evolved (nine members in the Arabidopsis Suc transporter family [AtSUC genes] and 14 members in the Arabidopsis monosaccharide transporter family [AtSTP genes]) that enable the plant to regulate the necessary membrane transport processes. Regulation is needed with respect to the sink identity, developmental stage, environmental changes, and the metabolic needs of a given sink. In all plants, the reproductive organs represent the most important and the most complex sink organs. Numerous individual sinks, such as petals, stamina, and carpels with ovules, placenta, papillae, pollen, tapetum, anther wall, developing seed, etc., depend on a perfectly regulated carbon supply for timely correct development and reproductive success. In Vicia faba, it has been shown quite impressively that Suc transport on the one hand and invertase activity and monosaccharide transport on the other are strictly regulated and confined to discrete cell types and specific developmental stages during seed development (Weber et al., 1997). In Arabidopsis, it has been shown that different transporter genes are transcribed during the development of the male gametophyte, the pollen. One gene is transcribed very early in pollen development during the transition from the pollen tetrade to individual pollen grains (AtSTP2; Truernit et al., 1999). Three were shown to be transcribed at the very end of pollen maturation (AtSTP4 [Truernit et al., 1996; R. Stadler and N. Sauer, unpublished data], AtSUC1 [Stadler et al., 1999], and AtSTP9 [M. Büttner, unpublished data]). Translation of these three mRNAs occurs predominantly after the release of the pollen from the anther, during its germination, most likely to allow the rapid and competitive growth of the pollen tube toward the ovule. So far, no transporter gene has been shown to be expressed during the main period of pollen development, i.e. the formation of individual pollen grains after meiosis and the final maturation and exine formation. Also, in other plants, pollen-specific monosaccharide (petunia [Petunia hybrida]; Ylstra et al., 1998) or Suc transporters (tobacco [Nicotiana tabacum]; Lemoine et al., 1999) have been identified. Moreover, in petunia (Ylstra et al., 1998) and potato (Solanum tuberosum; Maddison et al., 1999) pollen-specific invertases were identified that might serve to produce substrates for the monosaccharide transporter(s).

Here, we report on the identification and characterization of the new Arabidopsis monosaccharide transporter gene AtSTP6 that is also expressed exclusively in pollen. The kinetic properties of the encoded protein were analyzed in yeast (Saccharomyces cerevisiae), where it complements the inability of hexose transport-deficient mutants to grow on Glc as sole carbon source. A transposon insertion in the AtSTP6 gene does not result in any detectable change in phenotype in homozygous knockout plants. The physiological role of this gene is discussed.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Cloning of the AtSTP6 cDNA

During a PCR-based search for potential new monosaccharide transporter sequences in the Arabidopsis genome, a small 204-bp fragment of AtSTP6 (GenBank accession no. AJ001659) had been isolated (Büttner et al., 2000). This fragment was used to screen a genomic Arabidopsis library (ecotype Columbia). Two fragments, pAS5 and pAS20, were identified that harbored the AtSTP6 promoter plus the entire coding sequence.

According to these sequence data, primers were designed for the PCR amplification of a full-length cDNA clone from cDNA libraries generated from mature plants only (plasmid library), from Arabidopsis plants harvested at different growth stages (phage library), or from Arabidopsis flowers (phage library). cDNAs could be amplified only from the flower-specific library that had generously been provided by Dr. Elliot M. Meyerowitz (California Institute of Technology, Pasadena). Sequencing of the obtained PCR products identified these PCR clones as AtSTP6 cDNA (GenBank accession no. AJ344337).

At that point, the complete Arabidopsis genome was published (The Arabidopsis Genome Initiative, 2000), confirming the sequences obtained from pAS5 and pAS20. Vice versa, our cDNA sequences verified the computer-predicted intron/exon boundaries and confirmed that the AtSTP6 sequence (Munich Information Center for Protein Sequences no. At3g05960) is interrupted by two introns. The open reading frame in the AtSTP6 cDNA clone is 1,521 bp long and encodes a protein of 507 amino acids (sequence not shown) with a calculated molecular mass of 55.91 kD. AtSTP6 represents a typical member of the AtSTP protein family. It shares between 60% and 75% identical amino acids with already described family members (AtSTP1 [Sauer et al., 1990], AtSTP2 [Truernit et al., 1999], AtSTP3 [Büttner et al., 2000], AtSTP4 [Truernit et al., 1996], and AtSTP9 [M. Büttner, unpublished data]) and hydropathy analyses predict 12 transmembrane helices, a common feature to all AtSTPs (Büttner and Sauer, 2000). The calculated pI of AtSTP6 is 7.86, which is similar to the pIs of many plasma membrane transporters. The protein does not possess a consensus sequence for N-glycosylation.

Expression of AtSTP6 in Bakers' Yeast (Saccharomyces cerevisiae) and Analysis of Kinetic Parameters

Using specific primers, new NotI restriction sites were introduced at both ends of the AtSTP6 cDNA. These sites allowed the cloning of the PCR product into the yeast/Escherichia coli-shuttle vector pEX-Tag (Meyer et al., 2000) in sense (pACH61s) and antisense (pACH61a) orientation under the control of the yeast PMA1 promoter. Expression of AtSTP6 in this vector results in the C-terminal fusion to a biotinylation domain and a His tail (together 10 kD). Both domains can be used for purification of the protein and were shown previously not to interact with transport properties of plant transporters (Stolz et al., 1995). pACH61s and pACH61a were used to transform the yeast strains RE700A (Reifenberger et al., 1995) and EBY.VW4000 (Wieczorke et al., 1999). In strain RE700A, the genes of the endogenous hexose transporters Hxt1p to Hxt7p are deleted and the residual transport capacity for hexoses is negligible. Strain EBY.VW4000 lacks an additional 13 hexose transporter genes (Wieczorke et al., 1999) and has no detectable hexose transport activity.

Figure 1 shows that all transformed EBY.VW4000 lines, ScAS1E with AtSTP6 in sense orientation and ScAS2E with AtSTP6 in antisense orientation, can use maltose as an alternative carbon source. However, only ScAS1E lines have regained the capacity to grow on Glc as the sole carbon source. This shows that expression of recombinant AtSTP6 can complement multiple deletions of endogenous yeast hexose transporter genes and that AtSTP6 represents, in fact, a new Arabidopsis hexose transporter. This was supported by transport assays, where the transport capacity for ¹⁴C-labeled D-Glc was determined with transformed RE700A lines (ScAS1R with AtSTP6 in sense orientation and ScAS2R with AtSTP6 in antisense orientation). Clearly, the transport capacity for ¹⁴C-Glc of ScAS1R is 30- to 60-fold higher than the transport activity observed in ScAS2R cells (Fig. 2). The complementation of defects in plasma membrane-localized transport activities suggests that AtSTP6 may also be targeted to the plasma membrane of Arabidopsis cells.

View larger version (38K): [in this window] [in a new window]   Figure 1.   Growth complementation on D-Glc of hexose transport-deficient yeast cells by AtSTP6 expression. Three independent transformant lines of EBY.VW4000 expressing AtSTP6 in sense orientation (ScAS1E-1 to ScAS1E-3) and one transformant line expressing AtSTP6 in antisense orientation (ScAS2E-2) were grown on the alternative carbon source maltose for 2 d or on different concentrations of D-Glc for 4 d. Obviously, all lines can grow on maltose, but only the sense strains can grow on the different Glc media. Growth of sense strains was already seen after 2 d on all Glc concentrations, but was more prominent after 4 d.

View larger version (14K): [in this window] [in a new window]   Figure 2.   Transport of radiolabeled 14C-Glc by AtSTP6-expressing yeast cells and by controls. Transport of 14C-Glc in transformed lines of RE700A expressing AtSTP6 in sense orientation (ScAS1R) or in antisense orientation (ScAS2R) was determined at an initial concentration of 0.1 mM 14C-Glc. Only the sense strain can transport Glc at high rates. Glc transport in antisense controls is only marginal.

The K_m value of recombinant AtSTP6 was determined in the yeast strain ScAS1E. The Lineweaver-Burk plot of a typical analysis is shown in Figure 3. The average value of two analyses gave a K_m for D-Glc of 20.5 µM (individual values were 14 and 27 µM), characterizing AtSTP6 as a high-affinity Glc transporter of Arabidopsis.

View larger version (13K): [in this window] [in a new window]   Figure 3.   Determination of the Km value for D-Glc uptake of the AtSTP6-expressing yeast strain ScAS1E. The Lineweaver-Burk plot of a typical Km determination is presented. All values were determined in Na-phosphate buffer (pH 5.5).

The inhibitor sensitivity of AtSTP6 was studied at an initial outside concentration of 0.1 mM ¹⁴C-D-Glc and with inhibitor concentrations of 50 µM. Transport inhibition by the uncoupler dinitrophenol and the almost complete disappearance of Glc transport activity in the presence of the uncoupler carbonyl cyanide-m-chlorophenylhydrazone suggest that AtSTP6-driven transport is pmf dependent (Fig. 4A). As with the other plant monosaccharide transporters, no inhibition is observed with the SH group inhibitor p-(chloromercuri) benzene sulfonic acid.

View larger version (22K): [in this window] [in a new window]   Figure 4.   Sensitivity of AtSTP6 to inhibitors and substrate specificity. A, Transport of 0.1 mM 14C-Glc was determined in the presence of the uncouplers dinitrophenol or carbonyl cyanide-m-chlorophenylhydrazone or in the presence of the SH-group inhibitor p-(chloromercuri) benzene sulfonic acid. All inhibitors were added to a final concentration of 50 µM. B, Relative transport rates of radiolabeled D-Glc or other potential substrates at an initial outside concentration of 0.1 mM. All data represent average values of two independent transport tests.

Analyses of the transport rates of other potential substrates of AtSTP6 were performed with different ¹⁴C-labeled sugars at 0.1 mM (Fig. 4B). Uptake of Rib by AtSTP6 seems to be negligible and transport of Xyl is very low. However, significant transport rates were determined for Gal, Fru, Man, and the synthetic Glc analog 3-O-methyl-Glc. This characterizes AtSTP6 as a broad-spectrum monosaccharide transporter with a preference for hexoses.

Identification of the AtSTP6 Protein

Despite a high degree of sequence conservation within the membrane-spanning domains of the different AtSTPs, little sequence conservation is found in the C termini of these proteins. Therefore, an antiserum was raised in rabbits against a fusion of the E. coli maltose-binding protein and 35 amino acids from the very C terminus of AtSTP6. Figure 5 shows a western blot of total membrane fractions from bakers' yeast and fission yeast (Schizosaccharomyces pombe) strains expressing different Arabidopsis monosaccharide transporter cDNAs (AtSTP1-4 in fission yeast YGS-5 and AtSTP6, AtSTP9, AtSTP11, and AtSTP13 in yeast; this paper; Büttner and Sauer, 2000; M. Büttner and N. Sauer, unpublished data). Obviously, the affinity-purified anti-AtSTP6 antiserum is highly specific and recognizes a protein of about 50 kD only in the membrane protein fraction of ScAS1R (Fig. 5). Obviously, this apparent molecular mass is smaller than the calculated molecular mass of AtSTP6 (55.9 kD) plus its C-terminal fusion (10 kD; see above). However, lipophilic membrane proteins typically migrate at lower apparent molecular mass, a consequence of excess SDS binding (Sauer and Stadler, 1993; Meyer et al., 2000). No band is detected in the antisense control, ScAS2R, or in any of the other yeast membrane fractions. The observed apparent molecular mass is slightly below the predicted value, a feature common to most transport proteins.

View larger version (27K): [in this window] [in a new window]   Figure 5.   Identification of recombinant AtSTP6 protein in SDS extracts of total membranes from AtSTP6-expressing yeast cells. SDS-solubilized proteins (25 µg per lane) from yeast strains expressing the indicated AtSTP cDNA clones in sense (s) or antisense (as) orientation were separated on SDS gels, blotted to nitrocellulose filters, and incubated with anti-AtSTP6 antiserum at a dilution of 1:500 (w/v). Binding of the antiserum was detected with the Lumi-Light Kit (Boehringer, Mannheim, Germany). AtSTP1 to AtSTP4 were expressed in fission yeast, and AtSTP6 to AtSTP13 were expressed in bakers' yeast.

Analysis of AtSTP6-Promoter::GUS Plants

After the detailed characterization of AtSTP6 in yeast as a plasma membrane-localized, high-affinity, broad-spectrum monosaccharide transporter, we analyzed the site of AtSTP6 expression in planta. Therefore, we generated a construct containing 2,700 bp upstream of the start ATG plus the coding sequence for the first 125 amino acids, which is interrupted by the first intron. This genomic fragment was cloned in frame in front of the E. coli GUS gene in pBI101 (Jefferson et al., 1987) and used to perform an Agrobacterium tumefaciens-mediated transformation of Arabidopsis. Five of 21 independent transformants showed no -glucuronidase (GUS) staining in any tissue. In the other 16 lines, GUS histochemical staining was observed exclusively in anthers (Fig. 6, A and B) and not in any of the vegetative tissues (Fig. 6D). This expression pattern explained the already mentioned identification of AtSTP6 cDNAs exclusively in the flower-specific cDNA library. More detailed analyses revealed that the anther-specific GUS staining is confined to the pollen grains (Fig. 6C). Faint blue staining observed in anther walls results from color leaking out from heavily stained pollen. No staining was seen in the walls of depollinated anthers (data not shown). During flower development, GUS staining is first observed in stages 11 and 12 (Bowman and Smyth, 1993), right before the opening of floral buds, and becomes more intense during the later stages of floral development (Fig. 6A).

View larger version (54K): [in this window] [in a new window]   Figure 6.   Analyses of AtSTP6-promoter-GUS constructs in transgenic Arabidopsis plants. A and B, GUS histochemical staining was observed only in anthers. C, At higher magnification, only the pollen grains show blue GUS staining. D, All other tissues, such as roots, leaves, or the inflorescence stem (not shown) show absolutely no GUS activity. Scale bars = 1 mm in A, B, and D and 0.25 mm in C.

Identification of an En-1 Insertion Mutant of AtSTP6

For analyses of the physiological role of AtSTP6 in Arabidopsis pollen, a set of 3,000 (En-1)-tagged mutant lines (Baumann et al., 1998; Wisman et al., 1998) was screened for Atstp6 mutants via a PCR-based screening strategy (Baumann et al., 1998). One mutant line (En27) was isolated and shown to carry an insertion at position +536 within the AtSTP6 gene (Fig. 7). Seeds from this primary mutant were germinated and homozygous plants were identified by PCR (Fig. 7).

View larger version (10K): [in this window] [in a new window]   Figure 7.   Identification of the (En-1)-tagged Atstp6 mutant line En27. The AtSTP6 gene (introns are presented in white, exons in black, and nos. indicate the length of the respective fragment) in the mutant line En27 carries an (En-1)-insertion 535 bp after the start ATG. Arrows 1 to 3 indicate the positions of primers that were used to identify the mutant. An agarose gel with PCR products from different primer combinations in an Arabidopsis wild-type (WT) plant and in the homozygous En27 mutant is shown.

Homozygous En27 plants were able to self-pollinate, were fertile, and produced normal seeds, pods, and seed numbers, suggesting that the Atstp6 mutation does not interfere significantly with pollen viability, with pollen tube growth or with the pollen tubes' ability to fertilize ovules. This interpretation was supported by light microscopic analyses of En27 anthers showing no difference in pollen development, pollen size, or pollen number, when compared with WT anthers (data not shown).

Finally, we tested the ability of En27 pollen and of pollen from the corresponding ecotype Columbia to germinate and grow on two different synthetic pollen growth media (medium1, Stadler et al., 1999; and medium 2, Fan et al., 2001) to detect possible Atstp6-dependent growth differences. On both media, the germination rate of pollen from Columbia WT was quite low (6.5%-7%) compared with pollen from another wild-type (WT) line (Arabidopsis ecotype C24; 70%-85%). However, in several independent analyses, no significant difference between the germination rates of En27 pollen (medium 1, 5.95% ± 0.85%; and medium 2, 7.0% ± 1.3%) and of pollen from the corresponding ecotype Columbia (medium 1, 4.0% ± 1.6%; and medium 2, 6.6% ± 2.1%) could be observed.

Localization of AtSTP6 mRNA by in Situ Hybridization

The En27 mutant line was also used to confirm the localization data obtained with AtSTP6-promoter::GUS plants by in situ hybridization. Ideally, only sequences from the 3'-untranslated region should be used for this technique, to avoid potential cross-reactions with the highly conserved coding regions of mRNAs from related genes. However, frequently these 3' sequences are too short to obtain sufficiently strong labeling of the probe. A knockout line is the perfect control for this type of analysis. Figure 8 shows that in sections of mature anthers from Arabidopsis WT, the ³⁵S-labeled AtSTP6-antisense probe reacts specifically only with the pollen grains (Fig. 8A). No signals were obtained with the same probe on sections from developmentally younger anthers (e.g. the tetrade state in Fig. 8D), on sections from mature anthers from En27 knockout plants (Fig. 8C), or on sections of mature WT anthers treated with a ³⁵S-labeled AtSTP6-sense probe (Fig. 8B). These data confirm the GUS data presented in Figure 6.

View larger version (125K): [in this window] [in a new window]   Figure 8.   Localization of AtSTP6 transcripts in pollen grains by in situ hybridizations. A, Section from an Arabidopsis Columbia WT anther shows pollen-specific signals, when treated with a radiolabeled AtSTP6-antisense probe. B, No signals were detected in similar sections after incubation with an AtSTP6-sense probe. C, Section from an En27 anther was treated with an AtSTP6-antisense probe. D, Section from an Arabidopsis Columbia WT anther from the tetrade state was treated with an AtSTP6-antisense probe. No accumulation of signals is seen in C and D. Scale bar = 20 µm in A through D.

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

This paper describes the molecular, kinetic, and physiological characteristics of the new Arabidopsis monosaccharide transporter gene AtSTP6 (At3g05960). The presented data characterize the encoded protein as an energy-dependent, pmf-driven, high-affinity monosaccharide transporter that can accept Glc, Man, Fru, and Gal as potential substrates. These kinetic properties and the sequence homology to other, previously characterized transporters describe AtSTP6 as a plasma membrane-type monosaccharide transporter and as a member of the AtSTP transporter family (Büttner and Sauer, 2000). The localization in the plasma membrane is also supported by the observation that a yeast strain lacking 20 endogenous transporter genes (Wieczorke et al., 1999) regains its ability to grow on Glc as sole carbon source upon expression of the Arabidopsis AtSTP6 gene (Fig. 1).

The cell-specific expression of AtSTP6 was analyzed in AtSTP6-promoter::GUS plants and with in situ hybridization analyses (Figs. 6 and 8). Both data sets clearly show that AtSTP6 expression is confined to the latest stage in pollen development (stages 11 and 12 according to Bowman and Smyth, 1993). Thus, AtSTP6 is the fifth gene encoding a pollen plasma membrane localized sugar carrier and the fourth monosaccharide transporter gene that is expressed at this very stage of pollen maturation. In previous analyses, the Suc transporter gene AtSUC1 (Stadler et al., 1999) and the monosaccharide transporter genes AtSTP2 (Truernit et al., 1999), AtSTP4 (Truernit et al., 1996; R. Stadler and N. Sauer, unpublished data), and AtSTP9 (M. Büttner, unpublished data) were shown to be expressed at the very same stage of pollen development. All of these genes are transcribed, but not or only weakly translated before the release of the pollen grains from the anthers. However, large amounts of proteins could be detected after germination of the pollen grains on the papillae of Arabidopsis stigmas (Stadler et al., 1999).

Similar analyses could not be performed to discriminate between the timing of transcription and translation of the AtSTP6 gene and its mRNA because antisera from different rabbits (Oryctolagus cuniculus) were useful only for the specific identification of the AtSTP6 protein on western blots (Fig. 5). In contrast, all attempts to identify the AtSTP6 protein in thin sections of Arabidopsis anthers or in sections of AtSTP6-expressing yeast cells failed (data not shown). This is a clear indication that our anti-AtSTP6 antisera were directed against a primary structure resulting from SDS solubilization (i.e. a linear antigenic region in AtSTP6), but not against a secondary or tertiary structure resulting from fixation for light microscopy (i.e. a fixed, folded structure in the AtSTP6 protein). Nevertheless, the fact that AtSTP6 expression is only found in stages 11 and 12 of pollen grain development (Bowman and Smyth, 1993), a time when the exine is formed and the grain is fully developed, makes a role of AtSTP6 in sugar uptake at this stage more than unlikely.

This suggests a function of AtSTP6 for the sugar supply of the germinating pollen or the growing pollen tube, a possibility that has already been discussed for other pollen-specific transporters. It is impossible to speculate which of the potential hexose substrates (Fig. 4B) is transported by AtSTP6 at that point. However, in vitro germination experiments demonstrated anyway that in tobacco (Lemoine at al., 1999) and Arabidopsis (N. Sauer and R. Stadler, unpublished data), pollen germination is much better on Suc in the medium. In Arabidopsis, this could be imported by AtSUC1 (Stadler et al., 1999); in tobacco, this could be imported by NtSUT3 (Lemoine et al., 1999). Of course, part of this Suc might also be cleaved by a pollen-specific invertase and the resulting hexoses might be imported. This was observed in tobacco pollen and in pollen (Lemoine et al., 1999) of petunia (Ylstra et al., 1998). A pollen-specific invertase that might catalyze this hydrolysis has been identified only recently in tomato (Lycopersicon esculentum; Goetz et al., 2001). However, addition of Glc to in vitro-germinated pollen frequently leads to an immediate burst of the pollen tubes (Lemoine et al., 1999; N. Sauer and R. Stadler, unpublished data), a reaction that normally occurs highly regulated at the very end of the pollen tube to release the nuclei for fertilization. Thus, disaccharides and monosaccharides may have different functions during germination, with one serving as a signal (e.g. in pollen tube guidance and/or in the release of the nuclei from the tip) and the other as a metabolite. In this respect, the presence of several monosaccharide transporters (AtSTP4, AtSTP6, and AtSTP9) in Arabidopsis pollen tubes underlines that growth and development of the male gametophyte is a highly regulated process. Future experiments involving analyses of double and triple knockout mutants will be necessary to unravel the distinct roles of these monosaccharide transporters and to elucidate their physiological role during pollen development, germination, and pollen tube growth.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Strains and Growth Conditions

Arabidopsis plants were grown in potting soil in the greenhouse under ambient conditions. For cloning in Escherichia coli, we used strain DH5 (Hanahan, 1983). Basic molecular biology techniques were applied according to Sambrook et al. (1989).

Molecular Cloning of AtSTP6 Genomic and cDNA Sequences

A genomic library of Arabidopsis (Stanford/Columbia in GEM11) has generously been provided by Dr. Sabine Schäfer (Max-Planck-Institut für Züchtungsforschung, Köln, Germany). This library was screened with a radiolabeled probe corresponding to a previously described, 204-bp fragment of AtSTP6 (GenBank accession no. AJ001659) that had been identified during a PCR search for potential monosaccharide transporter sequences in the Arabidopsis genome (Büttner et al., 2000). The 14-kb insert of one positive lambda clone (lambda Co4/2) was digested with EcoRI, the resulting fragments were cloned into pUC19, and the terminal sequences of all fragments were determined. Two fragments, pAS5 and pAS20, were shown to harbor the promoter and the entire coding sequence of the AtSTP6 gene. The complete sequences of these fragments were determined.

A flower-specific Arabidopsis cDNA library (ecotype Columbia) has been obtained from Dr. Elliot M. Meyerowitz and was used as template for PCR-analyses with the primers STP6-start (5'-TGC AGC GGC CGC GTG GTG GTG TCT GCT CG-3') and STP6-33 (5'-CTA GGC GGC CGC AGC ATT TCT CTT CTC TAT GTC ATG GTG G-3'). These primers were designed to amplify the complete AtSTP6 cDNA, to remove the TGA stop codon, and to introduce NotI restriction sites (underlined) on both ends of the PCR product (33 bp upstream from the start ATG and right after the last coding triplet). The resulting fragment was cloned into pGEM-T Easy (Promega, Madison, WI), and the insert of the resulting plasmid, pACH60, was sequenced.

Expression of AtSTP6 cDNA Sequences in Yeast (Saccharomyces cerevisiae) and Transport Analyses

The NotI insert of pACH60 was excised and cloned into the yeast/E. coli shuttle vector pEX-Tag (Meyer et al., 2000) that drives expression of inserted cDNAs under the control of the promoter of the yeast plasma membrane H⁺-ATPase (PMA1). The yeast/E. coli shuttle vector pEX-Tag is based on the previously described NEV-E vector (Sauer and Stolz, 1994). It harbors additional sequences in the EcoRI cloning site of NEV-E that allow the in-frame cloning of cDNAs to the N terminus of a sequence coding for the biotinylation domain of the oxaloacetate decarboxylase from Klebsiella pneumoniae (Schwarz et al., 1988) followed by a tail of six His residues. Both fragments can be used for the affinity purification of the cloned cDNA fragment (Stolz et al., 1995). The resulting plasmids, pACH61s (insert in sense orientation) and pACH61a (insert in antisense orientation), were used to transform (Gietz et al., 1992) yeast strains RE700A (Reifenberger et al., 1995; yielding strains ScAS1R [pACH61s] and ScAS2R [pACH61a]) and EBY.VW4000 (Wieczorke et al., 1999) yielding strains ScAS1E [pACH61s] and ScAS2E [pACH61a]).

Transport of ¹⁴C-labeled compounds was analyzed in Na-phosphate buffer (pH 5.5) in the presence of 10 mM ethanol as described (Sauer et al., 1990).

Complementation of Glc Transport-Deficient Yeast Cells by AtSTP6

Yeast cells were grown overnight in maltose-casamino acids (CAA) medium (0.67% [w/v] yeast nitrogen base, 1% [w/v] casamino acids, and 2% [w/v] maltose) to an OD₆₀₀ of 1, harvested, and washed with and resuspended in Na-phosphate buffer (pH 5.5). Different dilutions of this cell suspension were spotted onto petri plates containing maltose-CAA medium, or Glc-CAA medium with 0.2%, 2%, or 5% (w/v) D-Glc as sole carbon source.

In Situ Hybridization and Section of Plant Material

An AtSTP6 probe was generated by PCR from the pAS20 genomic clone using the oligonucleotides pas20r1 (5'-CGG AAT GCT TCT CCA GC-3') and STP6-32 (5'-TTA AGC GGC CGC GCG TTT GGT AGA AGA CAT TGA GC-3'). The resulting PCR fragment (775-bp coding region and 268-bp 3'-flanking sequence) was cloned into pGEM-T Easy (Promega) yielding the plasmids pAS100s (cDNA in sense orientation with respect to the T7 promoter) and pAS100a (cDNA in antisense orientation with respect to the T7 promoter). Linearized plasmids were in vitro transcribed with T7 polymerase in the presence of ³⁵S-UTP.

Plant material was fixed, embedded into methacrylate, and rehydrated as described (Stadler and Sauer, 1996). After equilibration with 0.2 M Na-phosphate buffer (5 min at ambient temperature), sections were treated with Pronase (1 mL of 10 mM Tris/HCl [pH 7.5], 1 mM EDTA plus 250 µL of Pronase stock solution for 10 min at ambient temperature; Pronase stock: 40 mg mL¹ Pronase in 50 mM Tris/HCl [pH 7.5] and 5 mM EDTA, pre-incubated for 4 h at 37°C) to enhance the accessibility of the sections for the RNA probes. Washes, prehybridization, hybridization, and detection were performed as described (Stadler et al., 1999).

Preparation of Yeast Total Membranes, SDS-PAGE, and Western Blots

Total membranes from yeast strains expressing different Arabidopsis monosaccharide transporter cDNAs were prepared as described (Sauer and Stolz, 2000). Membrane proteins were solubilized and separated on denaturing SDS-polyacrylamide gels according to the protocol of Laemmli (1970). Western blots were performed as described (Dunn, 1986).

Identification of a Transposon (En-1)-Tagged AtSTP6 Knockout Line of Arabidopsis

A population of 3,000 Arabidopsis lines that carry approximately 15,000 independent insertions of the autonomous maize (Zea mays) element En-1 was screened for En-1 insertions in the AtSTP6 gene as described (Baumann et al., 1998). In brief, following a three-dimensional pooling strategy, PCR reactions were performed with all combinations of the AtSTP6-specific primers AtSTP6t5' (5'-CGA GAA CAA CTA CTG CAA GTA CGA TAA CC-3') and AtSTP6t3' (5'-TCT CAA TGA TAA GTA GTG AAC CGA AGA GG-3') and the En-1 transposon-specific primers En205 (5'-GAA GAA GCA CGA CGG CTG TAG AAT AGG-3') and En8130 (5'-GAG CGT CGG TCC CCA CAC TTC TAT AC-3'). PCR reactions were tested for AtSTP6-specific products by Southern analysis using a 460-bp probe (region +327 to +786 of the AtSTP6 gene). Hybridizing PCR products were sequenced with the corresponding primers to reconfirm the En-1 insertions.

Construction of AtSTP6-Promoter::GUS Fusions and Plant Transformation

A 3,196-bp genomic HindIII/PstI-fragment from pAS5 (2,698 bp of 5'-flanking and promoter sequences plus 498 bp of coding sequence, including the first intron) was cloned into the HindIII/PstI-digested vector pUC19-derived plasmid (pUC-GUS-0b) that harbors the E. coli GUS gene and the terminator of the nopaline synthase gene. From this plasmid, an AtSTP6-promoter/GUS/nopaline-synthase terminator box was excised with HindIII/EcoRI and inserted into the respective sites of pBI101 (Jefferson et al., 1987), yielding the plasmid pBI-STP6-GUS. After transformation of the Agrobacterium tumefaciens strain GV3101 (Holsters et al., 1980), the construct was used for Arabidopsis transformation by dipping (Clough and Bent, 1998).

Preparation of Anti-AtSTP6 Antiserum

A 142-bp fragment corresponding to the last 35 amino acids of the AtSTP6 coding region was amplified using the oligonucleotides STP6-Cter5E (5'-GAC AGA ATT CAT CGC CAT TGA TGA CAT GAG-3') and STP6-Cter3H (5'-CAG GAA GCT TGA CAC AAA TCA ACA GAC TCG-3') and cloned into EcoRI/HindIII-digested pMAL-c2 (New England Biolabs, Schwalbach, Germany).

After expression of the fusion construct in E. coli BL21(DE) (Novagen, Madison, WI), extracts were separated on SDS-polyacrylamide gels (Laemmli, 1970) and fusion proteins were extracted (Sauer and Stadler, 1993). Antisera against these extracts were raised in rabbits (Oryctolagus cuniculus) (Dr. Pineda, Antikörper Service, Berlin). Crude anti-AtSTP6 antisera were affinity purified as described (Sauer and Stadler, 1993).