Expression and activity of the Hxt7 high-affinity hexose transporter of Saccharomyces cerevisiae


  • Ling Ye,

    1. E.C. Slater Institute, The University of Amsterdam, Plantage Muidergracht 12, 1018 TV Amsterdam, The Netherlands
    Current affiliation:
    1. Institute of Geriatrics, Chinese PLA General Hospital, Beijing 100853, China.
    Search for more papers by this author
  • Jan A. Berden,

    1. E.C. Slater Institute, The University of Amsterdam, Plantage Muidergracht 12, 1018 TV Amsterdam, The Netherlands
    Search for more papers by this author
  • Karel van Dam,

    1. E.C. Slater Institute, The University of Amsterdam, Plantage Muidergracht 12, 1018 TV Amsterdam, The Netherlands
    Search for more papers by this author
  • Arthur L. Kruckeberg

    Corresponding author
    1. E.C. Slater Institute, The University of Amsterdam, Plantage Muidergracht 12, 1018 TV Amsterdam, The Netherlands
    2. Department of Molecular Cell Physiology, Free University of Amsterdam, De Boelelaan 1087, 1081 HV Amsterdam, The Netherlands
    • E.C. Slater Institute, The University of Amsterdam, Plantage Muidergracht 12, 1018 TV Amsterdam, The Netherlands.
    Search for more papers by this author


High-affinity hexose transport is required for efficient utilization of low hexose concentrations by the baker's yeast Saccharomyces cerevisiae. These low concentrations occur during the late exponential phase of batch growth on hexoses, during hexose-limited chemostat or fed-batch culture, or during growth on sugars such as sucrose and raffinose that are hydrolysed to hexoses outside the cell. The expression of the Hxt7 high-affinity glucose transporter of S. cerevisiae was examined during batch growth on glucose medium in a wild-type strain and a strain expressing only HXT7 (i.e. with null mutations in HXT1–HXT6). In the wild-type strain, HXT7 transcription was repressed at high glucose and was detected when the glucose in the culture approached depletion. In the HXT7-only strain, transcription of HXT7 was constitutive throughout the glucose growth phase and was increased further at low glucose concentrations. After glucose depletion, the levels of HXT7 mRNA declined rapidly in both strains. In contrast, the Hxt7 protein was relatively stable after glucose depletion. By monitoring the subcellular localization of an Hxt7::GFP fusion protein it was observed that Hxt7 was localized in the plasma membrane, even when expressed at high glucose concentrations in the HXT7-only strain. After glucose depletion Hxt7 was gradually endocytosed and targeted to the vacuole for degradation. The Hxt7::GFP fusion protein was a fully functional hexose transporter with a catalytic centre activity of approximately 200/sec. It is concluded that repression of HXT7 and degradation of Hxt7 at high glucose concentrations is dependent on a high glucose transport capacity. Copyright © 2001 John Wiley & Sons, Ltd.


The kinetics of glucose transport in the yeast Saccharomyces cerevisiae are subject to regulation in response to environmental conditions, especially the nature and concentration of the carbon source. At high glucose concentrations yeast exhibits low-affinity glucose transport, and at low glucose concentrations or in the absence of glucose (e.g. after the diauxic shift of a glucose-grown batch culture) the affinity displayed by yeast for glucose is high (Bisson and Fraenkel, 1984; Walsh et al., 1994). These differences in affinity presumably reflect the substrate affinities of the individual hexose transport proteins synthesized under these conditions (Reifenberger et al., 1997). Yeast expresses three high-affinity glucose transporter genes under low-glucose conditions, viz. HXT2, HXT6 and HXT7 (Diderich et al., 1999).

Among these three genes, HXT7 [or in some strains its close homologue, HXT6 (Liang and Gaber, 1996)], is the most highly transcribed. HXT7 is also transcribed under the widest variety of conditions; its mRNA is abundant in cells approaching the diauxic shift during batch culture on glucose medium (DeRisi et al., 1997; Diderich et al., 1999) and during batch culture on galactose and ethanol/glycerol media (Liang and Gaber, 1996). It is also transcribed under a wide variety of chemostat-cultivation conditions in both a wild-type laboratory strain and an industrial baker's yeast strain (Diderich et al., 1999; P. van Hoek, J. Pronk and A. Kruckeberg, unpublished results). The high codon adaptation index of the gene suggests that HXT7 mRNA can be translated efficiently when it is present (Kruckeberg, 1996).

Growth of HXT6- or HXT7-only strains (i.e. with mutations in the other HXT1HXT7 genes) on solid media is not different from the wild-type when 0.1% or 1% glucose is present. However, growth on 5% glucose media is significantly inhibited (Ko et al., 1993; Reifenberger et al., 1995). This might be due to downregulation of HXT7 expression or function at the transcriptional or post-translational levels. Alternatively, it might be due to a kinetic imbalance, i.e. expression of high-affinity transport activity in the presence of high extracellular glucose might impair transporter function (‘substrate inhibition’; Bisson et al., 1993) or force the intracellular glucose concentration out of its normal range (Teusink et al., 1998).

HXT6 and HXT7 are almost identical in the sequences of their open reading frames and in the DNA sequence for 96 bp upstream; further upstream the sequences of the two genes diverge (Reifenberger et al., 1995). When the expression of HXT6 and HXT7 was assessed independently of one another in wild-type strains, they were shown to be differentially regulated; in these cases HXT7 expression was much higher than HXT6 expression (Diderich et al., 1999; Reifenberger et al., 1997; P. van Hoek, J. Pronk and A. Kruckeberg, unpublished results). The relative contributions of HXT6 and HXT7 to high-affinity glucose transport may be strain-dependent (Liang and Gaber, 1996).

Glucose transport can exert a high degree of control over the rates of growth and glucose metabolism. In a set of strains in which HXT7 (as the only functional glucose transporter) was expressed to different levels by altering the HXT7 promoter and the gene copy number, transport displayed a control coefficient of 0.54 for growth and 0.90 for glucose flux (the sum of the control coefficients of all steps in glucose metabolism is, by definition, 1; Ye, 1999; Ye et al., 1999). It is therefore not surprising that hexose transporters are themselves highly regulated. Regulation occurs at the steps of transcription (via both induction and repression mechanisms), transport activity (via modulation of Km and Vmax) and protein removal from the membrane (Boles and Hollenberg, 1997). In the case of Hxt7, the HXT7 gene appears to be under the control of both the general glucose repression pathway and the Snf3-dependent glucose sensing pathway. Moreover, the affinity of Hxt7 for glucose is modulated in response to both genetic and nutritional factors (Reifenberger et al., 1997; Walsh et al., 1996; this report). And finally, the Hxt7 protein is subject to endocytosis and proteolysis in the vacuole upon treatment of nitrogen-starved cells with high concentrations of glucose (Krampe et al., 1998).

In order to understand better the regulation and control of glucose metabolism, it is important to describe the life history and kinetic properties of the proteins that participate in the pathways involved. The large number and broad kinetic diversity of the hexose transporters of S. cerevisiae make such a description particularly challenging, since the transporters must be studied in isolation, yet must be described in an integrated model of transporter function and the impact of transport on cellular metabolism and regulation. In this paper the expression of the HXT7 gene and the behaviour of the Hxt7 protein were examined, and the glucose transport activity of Hxt7 was assessed quantitatively. These analyses were performed on native gene products in a wild-type strain or in a congenic strain expressing only HXT7. They were augmented by surveillance of the expression and cellular trafficking of an Hxt7::GFP fusion protein, which was found to have essentially wild-type expression and transport characteristics, enabling us to determine the catalytic centre activity of Hxt7. We provide direct evidence that this glucose transporter is normally expressed when the external glucose concentration is close to the substrate affinity of the protein.

Materials and methods

Strains, media and growth

The yeast strains used were the wild-type MC996A (MATα ura3-52 his3-11,15 leu2-3,112 MAL2 SUC2 GAL MEL) and the HXT7-only strain RE607B (MATα hxt1Δ::HIS3::Δhxt4 hxt5::LEU2 hxt2Δ::HIS3 hxt3Δ::LEU2::Δhxt6 ura3-52) (Reifenberger et al., 1995) and HXT7::GFP derivatives of these (KY101 and LYG7, respectively; see below).

Yeast cells were grown in a rotary shaker at 30°C in YP medium (2% peptone, 1% yeast extract) containing glucose or maltose at the concentrations indicated. Transformation of yeast cells was carried out by a lithium acetate method (Gietz and Schiestl, 1995). Cell samples for Northern and Western analysis were harvested by brief centrifugation at 4°C and frozen in liquid nitrogen prior to extraction.

Northern blot analysis

Total RNA was isolated from the yeast cells by acid-phenol extraction as described (Diderich et al., 1999). RNA samples were separated by electrophoresis in 1% agarose formaldehyde gels (Sambrook et al., 1989). Transfer to nylon membranes, prehybridization, hybridization with HXT7- and PDA1-specific oligonucleotide probes, and washing were performed as described (Diderich et al., 1999).

HXT7::GFP chimeric DNA fragment construction and transformation

Plasmid pFA6a–GFPS65T–kanMX6A was amplified using primers LY1 (GACTCACGATGACAAGCCATTGTACAAGAGAATGTTCAGCACCAAAAGTAAAGGAGAAGAACTTTTC) and LY2 (CTTTGCATTTCTCTTTAAAGTTTCTTTGTCTCCGTCCCACTCAACGGATGGCGGCGTTAGTATC) using the Amplify kit (Roche) as described by the manufacturer. The product contains the S65T allele of green fluorescent protein (GFP) and a kanMX selectable marker, flanked by DNA corresponding to the sequences at the 3′ end of the HXT7 open reading frame (primer LY1, underlined) and to the DNA 3′ of the HXT7 open reading frame (primer LY2, underlined). The amplification product was transformed into yeast strains MC996A and RE607B (Reifenberger et al., 1995) and transformants were selected for resistance to 200 µg/ml G418 on YP 2% glucose plates. Transformants that were stably G418-resistant were screened for proper integration of the GFP–kanMX cassette by analytical PCR.

Measurement of growth and glucose consumption

Growth was monitored by measurement of the optical density (OD600) of batch cultures at various time points. The residual glucose in the medium at these time points was determined by quenching an aliquot of the cultures in an equal volume of 5% trichloroacetic acid and measuring the glucose concentration enzymatically with hexokinase and glucose 6-phosphate dehydrogenase; the absorbance change of NADH at 340 nm was measured with a COBAS Auto-analyzer (Roche).

Glucose transport assay

Glucose uptake was assayed as described (Walsh et al., 1994) at five concentrations of D-U-(14C)-glucose [1, 2.5, 5, 10 and 25 mM (371, 297, 148, 74 and 30 MBq/mol, respectively)]. The data were fit to the Michaelis–Menten equation using ENZFITTER software (Elsevier-Biosoft). Total cell protein was measured by the method of Lowry (Lowry et al., 1951) using a COBAS Auto-analyzer (Roche) after digestion of cells overnight in 1 N NaOH. Bovine serum albumin was used as a standard. Cell number was determined by counting at least 1000 cells with a haemocytometer.


Living cells were mixed with an equal amount of 1% (w/v) low-melt agarose at 37°C and 6–8 µl was applied to a microscope slide. Cells were examined with a Leitz Aristoplan epifluorescence microscope using filter cube 1001 HQ-FITC (Chroma) for GFP excitation. Micrographs were recorded using an Apogee CCD camera and processed for display using Image-Pro Plus and Adobe Photoshop.

Fluorescence spectroscopy

A Hitachi RF-5001PC fluorimeter was used to scan the excitation and emission spectra of whole cell suspensions in 0.1 M potassium phosphate buffer, pH 6.5. Emission spectra were collected between 500 and 550 nm, with excitation at 489 nm and excitation and emission slit widths of 3 nm. Spectra of cells expressing HXT7::GFP and HXT7 were normalized to cell density, and background correction was done by subtracting the spectra of cells expressing HXT7 from the spectra of cells expressing HXT7::GFP.

Western blot analysis

Cells were harvested by centrifugation, washed once in 1% KCl, and extracted by abrasion with glass beads in 50 mM Tris–HCl, pH 8, 10 mM EDTA, 5% glycerol plus protease inhibitors (1 µg/ml leupeptin, 1 µg/ml pepstatin A, 0.2 mM AEBSF in DMSO). The lysates were cleared by centrifugation at 1300×g for 3 min. Supernatants containing 15 µg protein were heated at 40°C for 15 min in sample buffer and electrophoresed in 10% SDS–PAGE minigels. The proteins were transferred to nitrocellulose membranes at 25 V for 30 min, 50 V for 30 min and 100 V for 30 min at 4°C. The blots were sequentially incubated with anti-GFP rabbit antiserum diluted 1:10 000 (Molecular Probes) or anti-Hxt7 antiserum diluted 1:250 (Krampe et al., 1998), and with horseradish peroxidase-conjugated goat anti-rabbit secondary antibody diluted 1:3000 (Bio-Rad). Detection was carried out by chemiluminescence using the SuperSignal chemiluminescent substrate kit (Pierce).

Results and discussion

HXT7 is highly expressed at low concentrations of glucose

The expression patterns of HXT7 in the wild-type strain MC996A and the HXT7-only strain RE607B (bearing null mutations in HXT1HXT6) were compared during batch cultivations on 1% (about 55 mM) glucose medium. The growth curve, glucose consumption profile, and HXT7 mRNA and protein levels were measured simultaneously. The behaviour of the strains was reproducible under standard conditions; results of representative experiments are described below.

The wild-type strain MC996A had a higher growth rate than the HXT7-only strain (μ=0.38/h and 0.31/h, respectively; Figure 1A). HXT7 mRNA was expressed at a high level in the wild-type strain only after 8 h (residual glucose 18 mM) and reached a maximum level at 9 h (residual glucose 3 mM). After 10 h, glucose was depleted from the medium and the strain underwent diauxic shift. During the diauxic shift the level of HXT7 mRNA declined rapidly (Figure 1B) and was undetectable after 1 h. As expected, the appearance of Hxt7 protein was delayed compared to the appearance of the HXT7 mRNA; the protein level remained high for at least 2 h after glucose depletion (Figure 1C).

Figure 1.

Expression of HXT7 during batch cultivation on glucose. Single colonies of strains MC996A and RE607B were inoculated to 2% maltose pre-cultures and grown to early stationary phase. Cultures of 1% glucose medium were inoculated to an optical density (OD600) of approximately 0.2 and sampled over a 12 h period. (A) Time course of growth (▪) and residual glucose concentration (•). (B) Time course of HXT7 mRNA levels detected by Northern blot analysis. Filled arrow, HXT7 mRNA; open arrow, PDA1 mRNA used as a loading control. (C) Time course of Hxt7 protein levels detected by Western blot analysis

The HXT7-only strain had a prolonged glucose consumption profile (Figure 1A); this is explained by the lower glucose transport capacity of this strain than the wild-type (Ye et al., 1999). The level of HXT7 mRNA in this strain was high throughout the exponential growth phase (measurements began at 6 h, at a residual glucose concentration of 45 mM). The mRNA level increased further as the glucose concentration approached zero (Figure 1B). Within 1 h after glucose depletion the level of HXT7 mRNA had decreased below the threshold of detection. As with HXT7 mRNA, the Hxt7 protein was observed at all stages of exponential growth of the HXT7-only strain and its level remained high for several hours after glucose exhaustion (Figure 1C).

Reifenberger et al. (1995) described unpublished results showing that (in the strains used also in this study) mRNA levels from HXT7 and β-galactosidase activity from HXT7 promoter–lacZ fusions are very high under derepressing conditions; expression of HXT6 is much lower. Our results on expression of HXT genes under a variety of growth conditions confirm those observations (Diderich et al., 1999). When Reifenberger et al. compared growth and glycolytic flux between the wild-type strain and the HXT7-only strain, they found that the rates were diminished by about 10% in the mutant (Reifenberger et al., 1995), and that the glucose transport capacity (Vmax) of the mutant was 65–70% of the wild-type capacity (Reifenberger et al., 1997). In contrast, these rates in the HXT6-only strain ranged from 15% to 35% of the wild-type rates. Moreover, growth of the HXT6-only strain was slower than that of the HXT7-only strain. This effect was more pronounced in the presence of antimycin A, which blocks mitochondrial electron transport (Reifenberger et al., 1995). The growth inhibition by this respiratory inhibitor is consistent with a relatively low level of HXT6 expression, since one response of yeast to severe reduction in glucose transport capacity is to derepress its respiratory capacity (and thus increase the ATP yield per mole of glucose (Ye et al., 1999). Indeed, it was observed that the HXT6-only strain of Reifenberger et al. (1995) fermented only 40% of the glucose it consumed, compared to 80% for the wild-type and HXT7-only strain. Thus, for the strains examined here, HXT7 encodes the majority of the high affinity glucose transport activity. The stronger growth and transport phenotype for HXT6 and the HXT6/7 chimera reported by others (Ko et al., 1993; Liang and Gaber, 1996) may represent strain differences. However, the growth rate of this HXT6/7-only strain (CY290) was only one-third of that of the wild-type strain (0.13/h vs. 0.36/h; see Walsh et al., 1996); in contrast, the growth rate of the HXT7-only strain used here (RE607B) was 80% of the wild-type rate.

Time course of HXT7::GFP expression

Detection of the Hxt7 protein by immunoblot analysis does not demonstrate that the protein was functionally expressed. In order for this hexose transporter to be functional it must reside in the plasma membrane. Particularly in the case of the HXT7-only strain in early exponential growth (at high residual glucose concentrations), the location of the protein is an important question, since under these conditions the protein is not expressed in wild-type strains. Moreover, in non-growing cells Hxt7 undergoes endocytosis and vacuolar degradation in response to high glucose concentrations (Krampe et al., 1998). In order to monitor the expression andtargeting of the Hxt7 protein, a chimeric Hxt7::GFP protein was engineered and examined by fluorimetry and epifluorescence microscopy.

The HXT7 open reading frames of the wild-type and HXT7-only strains were fused in frame at their 3′ ends with GFP by targeted integration of a DNA cassette containing the GFP open reading frame, the kanMX selectable marker, and flanking DNA corresponding to the sequences around the HXT7 stop codon. The resulting strains were named KY101 (wild-type HXT7::GFP) and LYG7 (HXT7::GFP-only).

LYG7 and KY101, and their parental strains, RE607B and MC996A, were grown to mid-exponential phase in 5% glucose medium and then shifted to 1% glucose medium. After the medium shift the HXT7::GFP fluorescence, the expression and localization of the HXT7::GFP fusion protein in the living cells, the optical density (OD600) and the residual glucose concentration in the medium were measured simultaneously during 12 h.

Relative fluorescence due to HXT7::GFP was high in the HXT7::GFP-only strain LYG7 immediately after shifting the culture to 1% glucose medium. This relative fluorescence increased further as the glucose concentration declined below 20 mM. At 5 h glucose was almost depleted (0.25 mM) and the relative fluorescence reached a maximum (Figure 2A). Growth continued after the diauxic shift but the relative fluorescence decreased gradually. The cellular localization of Hxt7::GFP was examined after shifting LYG7 to 1% glucose medium, as shown in Figure 3A. The Hxt7::GFP fluorescence was localized in the plasma membrane until glucose exhaustion at about 5 h, at which time the fluorescent staining of the membrane reached its maximum intensity. After glucose exhaustion, the Hxt7::GFP fluorescence appeared in punctate structures internal to the plasma membrane. These gradually coalesced and accumulated deep within the cell. The punctate structures were presumably endocytic vesicles and the site of GFP accumulation was presumably the vacuole (Kruckeberg et al., 1999).

Figure 2.

Time course of relative fluorescence of Hxt7::GFP protein expression in wild-type and HXT7-only strains. Cells were pre-cultured to mid-exponential phase in 5% glucose medium and then shifted to 1% glucose medium. At the indicated times Hxt7::GFP relative fluorescence (arbitrary units) (•), OD600 (⧫) and residual glucose in the medium (□) were measured simultaneously. (A) Strain LYG7 (HXT7-only). (B) Strain KY101 (wild-type)

Figure 3.

Time course of expression of Hxt7::GFP protein in single living HXT7-only and wild-type cells. Cells harvested at the indicated times from the cultures described in Figure 2 were examined by epifluorescence microscopy. Additional samples were examined at 13 and 24 h. (A) Strain LYG7 (HXT7-only). (B) Strain KY101 (wild-type)

When KY101 cells were shifted from 5% to 1% glucose medium, Hxt7::GFP fluorescence by this wild-type strain was undetectable for the first several hours (Figure 2B). When the glucose was nearly exhausted (8–0.16 mM, at 6–7 h) and the expression of high-affinity glucose transporters was induced by the low glucose concentration, fluorescence was detected, although the level was still quite low in comparison to that of the LYG7 strain. After the diauxic shift the fluorescence continued to increase gradually for some time.

The localization of Hxt7::GFP in KY101 was also examined (Figure 3B). Before 6 h, no Hxt7::GFP fluorescence was detectable. After 6 h, when the glucose concentration in the medium decreased below 8 mM, Hxt7::GFP appeared in the plasma membrane but the signal remained weak. After 7 h, the glucose was depleted; the level of Hxt7::GFP in the membrane continued to increase gradually. At 9 h and later, although the glucose concentration was zero, the fluorescent signal was stronger than at 7 h. After 10 h the Hxt7::GFP fusion protein was gradually removed from the membrane by endocytosis and targeted to the vacuole.

In order to examine the fate of the Hxt7 protein in HXT7-only strains, a Western blot analysis was performed (Figure 4). LYG7 and RE607B cells from the same cultures as described above were harvested at 1, 5, 8 and 24 h. Hxt7::GFP protein from LYG7 lysates was detected with an anti-GFP antibody. The Hxt7::GFP fusion protein was detectable at 1 h (residual glucose 36 mM) in accordance with the results of fluorimetry (Figure 2A) and epifluorescence microscopy (Figure 3A). At 5 h (0.25 mM residual glucose) Hxt7::GFP expression reached a maximum and some Hxt7::GFP fusion protein started to be degraded to breakdown products. At 8 h the Hxt7::GFP level decreased and more degradation products were detected. At 24 h only some degradation product was observed. No cross-reacting signal was observed in the RE607B strain with this anti-GFP antibody. These results demonstrate that the Hxt7::GFP protein was stable in the HXT7::GFP-only strain at high glucose concentrations, and that the residual GFP fluorescence in this strain after the diauxic shift was due solely to residual GFP moieties.

Figure 4.

Immunodetection of Hxt7::GFP with anti-GFP antibody. LYG7 and RE607B cells from the cultures analysed by fluorimetry (Figure 2) and fluorescence microscopy (Figure 3) were harvested at 1, 5, 8 and 24 h. Cell lysates were separated by SDS–polyacrylamide gel electrophoresis, transferred to nitrocellulose, and Hxt7::GFP was detected with anti-GFP antibody

It has been demonstrated previously that Hxt7 is rapidly endocytosed and degraded in the vacuole when induced (raffinose-grown) cells are starved for nitrogen and treated with high glucose concentrations (Krampe et al., 1998). However, our results demonstrate that high glucose was not sufficient to stimulate endocytosis of Hxt7 when the protein was inappropriately expressed at high glucose concentrations during mid-exponential growth in the HXT7-only strain. Perhaps the activity of the transporter in growing, glucose-metabolizing cells prevented it from being endocytosed.

Glucose transport and catalytic-centre activity of the Hxt7::GFP glucose transporter

LYG7 and its parental strain RE607B were grown to mid-exponential phase in 5% glucose medium and then shifted to 1% glucose medium. Referring to the time course depicted in Figure 2A, after 3 h (mid-exponential phase) and 5 h (glucose exhaustion) cells were harvested and assayed for zero-trans glucose uptake. Data from a representative experiment are shown as Eadie–Hofstee plots in Figure 5. The data fit to a single-component transport system obeying Michaelis–Menten kinetics. The Km and Vmax values are also presented in Figure 5. The kinetics of glucose transport conferred by the HXT7 and HXT7::GFP gene products were not markedly different, although the Hxt7::GFP fusion protein reproducibly showed slightly higher Km values than those of the native Hxt7 protein at the same time points. The affinity for glucose of the Hxt7 and Hxt7::GFP proteins increased by 40% between mid-exponential phase and glucose depletion. For LYG7, the Vmax at diauxic shift was almost twice as high as in mid-exponential phase. This result correlates well with HXT7::GFP expression levels obtained by fluorimetry and epifluorescence microscopy.

Figure 5.

Kinetics of zero-trans glucose uptake by Hxt7- and Hxt7::GFP-only cells grown to mid-exponential phase and the diauxic shift. Uptake of 14C-glucose by strains RE607B and LYG7 from mid-exponential (E) and diauxic shift (D) phases of batch cultivation on glucose. Cells were harvested in mid-exponential phase at OD600 0.85 (RE607B, □) and 0.86 (LYG7, ▪), and at diauxic shift at OD600 2.9 (RE607B, ▵) and 3.0 (LYG7,□). The lines represent the fits of the data to the Michaelis–Menten equation. Dotted lines, strain RE607B; solid lines, strain LYG7. The residual glucose concentrations in the cultures at the first time point were approximately 25 mM, and at the second time point were less than 1 mM and undetectable, respectively. Uptake of each glucose concentration for every sample was determined in triplicate

The fluorescent signal from Hxt7::GFP was used to determine the cellular concentration and catalytic-centre activity of this fusion protein in the cells used for the glucose uptake assay, using purified GFP as a fluorescent standard. The concentrations of pure GFP required to match the fluorescent signal of cells expressing Hxt7::GFP in mid-exponential phase and diauxic shift were 2.7 and 3.9 nM, respectively. The density of the Hxt7::GFP cell suspension in the fluorimetric assay was 1.12×108 cells/ml (mid-exponential phase) and 1.31×108 cells/ml (diauxic shift). Therefore, each cell contained 1.4×104 (mid-exponential phase) or 1.8×104 (diauxic shift) Hxt7::GFP molecules. The ratio of Hxt7::GFP molecules per cell between diauxic shift and mid-exponential phase is 1.3. This ratio agrees with the ratio of relative fluorescence of 1.4 between diauxic shift (5 h) and mid-exponential phase (3 h) Hxt7::GFP cells in culture (Figure 2A).

The cellular abundance of the Hxt7::GFP chimera and the glucose transport activity mediated by that protein were used to calculate the catalytic-centre activity of Hxt7 for glucose. For the first time point (mid-exponential phase), the Hxt7::GFP cell suspension contained 0.61 mg total cell protein/ml, giving a relationship of 4.5 pmol Hxt7::GFP/mg total cell protein. The Vmax for glucose transport of these cells, 53 nM/min/mg total cell protein, was divided by this value, resulting in an estimate of 197/s for the catalytic-centre activity of Hxt7::GFP. For the second time point (diauxic shift), the Hxt7::GFP cell suspension contained 0.57 mg total cell protein/ml, giving a relationship of 7.0 pmol Hxt7::GFP/mg total cell protein. The Vmax for glucose transport of these cells, 99 nmoles/min/mg total cell protein, was divided by this value, resulting in an estimate of 237/s for the catalytic-centre activity of Hxt7::GFP. All of the Hxt7::GFP in both cell types appeared to be at the plasma membrane when examined by fluorescence microscopy (Figure 6).

Figure 6.

Hxt7::GFP was located in the plasma membrane in cells assayed for glucose uptake. Cells from the suspensions used in the glucose transport experiment were subsequently examined by epifluorescence and phase-contrast microscopy. (A). Cells from mid-exponential phase. (B) Cells from the diauxic shift

HXT7 is one of the most abundantly expressed hexose transporters at low glucose concentrations and accounts for a significant proportion of the high-affinity glucose transport capacity expressed by yeast. The catalytic centre activity for Hxt7 is about four-fold higher than the catalytic centre activity previously determined for an Hxt2::GFP fusion protein (Kruckeberg et al., 1999). Therefore, the contribution of Hxt7 to total high-affinity glucose transport is not simply due to its high level of expression, but also reflects the intrinsically high activity of this transporter molecule.

The Km of Hxt7 for glucose transport was found by Reifenberger et al. (1997) to be 2 mM in cells grown on high glucose concentrations, and 1 mM in HXT7-only cells grown on low glucose concentrations. We found similar values for the Km for glucose, and noticed a similar two-fold increase in affinity during the transition from high glucose to low glucose conditions. It has previously been observed that the transport kinetics of the Hxt6/7 chimera are altered by mutation of the SNF3 glucose sensor gene; the Km is increased 1.5–4-fold as a result of a snf3Δ mutation (Walsh et al., 1996). The increases in the catalytic centre activity and substrate affinity of Hxt7 (and Hxt7::GFP) between high and low glucose conditions suggest that the transport kinetics of Hxt7 are modulated post-transcriptionally by the sensed glucose concentration. The uptake kinetics of Hxt2 (Reifenberger et al., 1997) and Gal2 (Diderich et al., 1999; Liang and Gaber, 1996; Ramos et al., 1989; Reifenberger et al., 1997) have also been shown to be modulated according to the carbon source and its concentration. It would appear that modulation of transporter kinetics is a common phenomenon amongst yeast hexose transporters.

Glucose repression and glucose transport capacity

We found that HXT7 mRNA and Hxt7 protein were derepressed on high glucose medium in the HXT7-only strain, compared to the wild-type strain. This was accompanied by a high level of high-affinity transport activity in the HXT7-only strain under these conditions. Reifenberger et al. (1995) also found that HXT7-only cells grown on 2% glucose medium express high-affinity glucose uptake. Under the same conditions the wild-type strain displays only low-affinity uptake (Reifenberger et al., 1995; unpublished results). Our data suggest that the high-affinity kinetics of the HXT7-only strain on high glucose is due to derepression of HXT7. Our observations that Hxt2 (Kruckeberg et al., 1999) and Hxt7 (this work) are upregulated in strains that only express those glucose transporters support earlier studies, which suggest that yeast cells are able to maintain relatively high glucose transport capacities despite null mutations in most of the metabolically active transporter genes (Reifenberger et al., 1997; Walsh et al., 1996).

The canonical glucose repression pathway in yeast requires functional hexokinase II (encoded by the HXK2 gene; Entian, 1980; Ma et al., 1989; Rose, et al., 1991). Glucose repression of HXT6 and HXT7 is diminished by mutation of HXK2 (Liang and Gaber, 1996; Petit et al., 2000), directly demonstrating that they are targets of the canonical pathway. We have previously argued that the canonical glucose repression pathway is fully functional only when cells express sufficient glucose transport capacity (Teusink et al., 1998; Ye et al., 1999) . We propose that the upregulation of HXT7 observed in the HXT7-only strain, compared to the wild-type, is also due to relief of glucose repression resulting from reduced transport capacity and intracellular concentrations of glucose, and cannot be accounted for by alterations in glucose signalling that are independent of glucose metabolism (Liang and Gaber, 1996). A number of laboratories have found that wild-type levels of glucose transport capacity are required for normal glucose repression of various genes, including those involved in respiration, gluconeogenesis and utilization of alternative carbon sources such as sucrose and maltose (Gamo et al., 1994; Özcan et al., 1993; Reifenberger et al., 1997; Weirich et al., 1997; Ye et al., 1999). For example, the degree of glucose repression of maltase activity during growth on glucose plus maltose is correlated with the extent to which glucose transport capacity is reduced in HXT-only strains and the hxt1–hxt7 null strain. Thus, maltase is fully derepressed in the HXT6-only strain with a very low transport capacity (to the same extent as the hxt null strain) but is partially repressed in the HXT7-only strain (Reifenberger et al., 1997). By controlled expression of HXT7 in a strain with no other hexose transport activity, we have shown that the degree of repression of invertase and respiratory capacity is correlated with the level of HXT7-dependent glucose transport capacity; at very low levels of HXT7 expression, invertase activity is largely derepressed, while at maximal levels of HXT7 expression, invertase is fully repressed (Ye et al., 1999).

We found that HXT7 expression was induced at low glucose concentrations. This induction was clear in the HXT7-only strain, in which the level of expression at low glucose was significantly higher than the derepressed levels associated with high glucose concentrations (Figures 1 and 2). Induction (as well as derepression) may account for the level of HXT7 expressed by the wild-type strain as well. Reifenberger et al. (1997) noticed that the high-affinity transport capacity measured in the HXT7-only strain at high glucose was only two-thirds of the capacity measured in low glucose-grown cells. They suggested that this was due to glucose repression of HXT7 at high glucose concentrations. Our results suggest that they were observing induction of HXT7-dependent high-affinity transport at low glucose concentrations.


This work was financially supported by the Association of Biotechnology Centres in The Netherlands (ABON) and EC Project BIO4-CT98-0562 (DG 12-SSM1).