Portal de Periódicos da CAPES

Glucose sensing and signaling by two glucose receptors in the yeast Saccharomyces cerevisiae

1998; Springer Nature; Volume: 17; Issue: 9 Linguagem: Inglês

10.1093/emboj/17.9.2566

1460-2075

Sabire Özcan,

Plant nutrient uptake and metabolism

Article1 May 1998free access Glucose sensing and signaling by two glucose receptors in the yeast Saccharomyces cerevisiae Sabire Özcan Sabire Özcan Present address: Department of Cell Biology and Physiology, Washington University, School of Medicine, St. Louis, MO, 63110 USA Search for more papers by this author Jim Dover Jim Dover Department of Genetics, Washington University School of Medicine, St Louis, MO, 63110 USA Search for more papers by this author Mark Johnston Corresponding Author Mark Johnston Department of Genetics, Washington University School of Medicine, St Louis, MO, 63110 USA Search for more papers by this author Sabire Özcan Sabire Özcan Present address: Department of Cell Biology and Physiology, Washington University, School of Medicine, St. Louis, MO, 63110 USA Search for more papers by this author Jim Dover Jim Dover Department of Genetics, Washington University School of Medicine, St Louis, MO, 63110 USA Search for more papers by this author Mark Johnston Corresponding Author Mark Johnston Department of Genetics, Washington University School of Medicine, St Louis, MO, 63110 USA Search for more papers by this author Author Information Sabire Özcan2, Jim Dover1 and Mark Johnston 1 1Department of Genetics, Washington University School of Medicine, St Louis, MO, 63110 USA 2Present address: Department of Cell Biology and Physiology, Washington University, School of Medicine, St. Louis, MO, 63110 USA *Corresponding author. E-mail: [email protected] The EMBO Journal (1998)17:2566-2573https://doi.org/10.1093/emboj/17.9.2566 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info How eukaryotic cells sense availability of glucose, their preferred carbon and energy source, is an important, unsolved problem. Bakers' yeast (Saccharomyces cerevisiae) uses two glucose transporter homologs, Snf3 and Rgt2, as glucose sensors that generate a signal for induction of expression of genes encoding hexose transporters (HXT genes). We present evidence that these proteins generate an intracellular glucose signal without transporting glucose. The Snf3 and Rgt2 glucose sensors contain unusually long C-terminal tails that are predicted to be in the cytoplasm. These tails appear to be the signaling domains of Snf3 and Rgt2 because they are necessary for glucose signaling by Snf3 and Rgt2, and transplantation of the C-terminal tail of Snf3 onto the Hxt1 and Hxt2 glucose transporters converts them into glucose sensors that can generate a signal for glucose-induced HXT gene expression. These results support the idea that yeast senses glucose using two modified glucose transporters that serve as glucose receptors. Introduction Glucose, the preferred carbon and energy source for most eukaryotic cells, has significant and varied effects on cell function. Consequently, maintaining glucose homeostasis is of great importance to many organisms. How cells perceive and respond to glucose is an important, unanswered question. A signal transduction pathway responsible for glucose-induced gene expression in baker's yeast (Saccharomyces cerevisiae) has come into focus recently. Transcription of several of the 20 genes that encode hexose transporters or highly related proteins (HXT genes) (Bisson et al., 1993; Kruckeberg, 1996; Boles and Hollenberg, 1997) is induced by glucose. Expression of HXT1 is induced only in response to high concentrations of glucose; transcription of HXT2 and HXT4 is induced only by low levels of glucose (Özcan and Johnston, 1995, 1996; Schulte and Ciriacy, 1995). Glucose induction of the HXT genes is mediated by a repression mechanism involving the zinc-finger-containing protein Rgt1: in the absence of glucose, Rgt1 binds to the HXT promoters and represses their expression; glucose inactivates Rgt1 repressor function, leading to derepression of HXT expression (Marshall-Carlson et al., 1991; Erickson and Johnston, 1994; Özcan and Johnston, 1995; Özcan et al., 1996a). Inhibition of Rgt1 by glucose requires Grr1, which may be part of a ubiquitin-conjugating enzyme complex (Flick and Johnston, 1991; Barrall et al., 1994; Vallier et al., 1994; Li and Johnston, 1997). The glucose signal that triggers inhibition of Rgt1 function appears to be generated by Snf3 and Rgt2 (Özcan et al., 1996b), which are similar to glucose transporters and are members of a large family of 20 known or predicted glucose transporters in yeast. Most of these Hxt proteins are very similar to each other, sharing between 50 and 100% amino acid sequence identity. Snf3 and Rgt2 are the most divergent family members, being only ∼25% similar to their relatives (Bisson et al., 1993; Kruckeberg, 1996; Boles and Hollenberg, 1997). A distinguishing characteristic of Snf3 and Rgt2 is their unusually long C–terminal extensions (>200 amino acids) that are predicted to reside in the cytoplasm (Marshall-Carlson et al., 1990). All other known or predicted hexose transporters (from any organism) have C-terminal cytoplasmic tails of only ∼50 amino acids. Work in several laboratories suggested that Snf3 plays a regulatory rather than a metabolic role in glucose transport (Marshall-Carlson et al., 1990, 1991; Bisson et al., 1993; Ko et al., 1993; Özcan and Johnston, 1995; Liang and Gaber, 1996; Coons et al., 1995, 1997). We have presented evidence that Snf3 and Rgt2 are sensors of extracellular glucose that are involved in generation of an intracellular glucose signal that triggers the induction of HXT gene expression (Özcan et al., 1996b). The key result that led to this idea is that a dominant mutation in RGT2 and SNF3 (Marshall-Carlson et al., 1990) causes constitutive induction of HXT gene expression, even in the absence of the inducer glucose (Özcan et al., 1996b). We believe this mutation converts the glucose sensors into their glucose-bound form. This result led us to conclude that glucose sensing and signaling is a receptor-mediated process which is independent of glucose metabolism. Here we provide further evidence that glucose signaling is not the result of glucose transport and that the C-termini of Rgt2 and Snf3 are the glucose signaling domains of these glucose sensors. Results Snf3 and Rgt2 have separate but overlapping functions Previous results indicated that Snf3 and Rgt2 probably sense different levels of glucose (Özcan et al., 1996b); Snf3 seems to function as a sensor of low levels of glucose because it is required for low glucose-induced expression of HXT2 (Table I, compare lines 9D and 10D) but not high glucose-induced expression of HXT1 (line 9B). Rgt2 appears to sense high concentrations of glucose because it is required for full induction of HXT1 expression by high levels of glucose (Table I, compare lines 6B–8B) but not for induction of HXT2 by low levels of glucose (line 6D). To support the idea that Snf3 and Rgt2 have different affinities for glucose, we tested whether the SNF3 or the RGT2 gene in single copy is sufficient to complement rgt2 or snf3 mutants, respectively. The decrease in high glucose-induced HXT1 expression caused by rgt2 mutations cannot be restored by SNF3 (line 7B), nor can the defect in HXT2 induction by low levels of glucose in a snf3 mutant be restored by RGT2 (line 11D). Thus, RGT2 and SNF3 have separate, non-redundant functions. Table 1. Rgt2p and Snf3p have distinct, but overlapping functions in glucose signaling Genotype Construct Mean β-galactosidase activity (U) ± SD HXT1::lacZ HXT2::lacZ A Gly B 4% Glu C Gly D Gly + 0.1% Glu 1 WT vector 1.1 ± 0.15 353 ± 32 14 ± 0.5 348 ± 30 2 CEN–SNF3 1.4 ± 0.2 364 ± 22 21 ± 4 335 ± 18 3 CEN–RGT2 1.2 ± 0.3 394 ± 80 19 ± 7 314 ± 40 4 ADH–SNF3 30 ± 7 329 ± 21 274 ± 23 327 ± 57 5 ADH–RGT2 37 ± 9 396 ± 74 298 ± 46 308 ± 26 6 rgt2Δ vector 0.6 ± 0.14 68 ± 7 19 ± 4 397 ± 83 7 CEN–SNF3 0.9 ± 0.2 72 ± 8 8 CEN–RGT2 1.6 ± 0.1 403 ± 69 9 snf3Δ vector 0.7 ± 0.09 227 ± 30 10 ± 1.8 24 ± 2 10 CEN–SNF3 21 ± 3 332 ± 77 11 CEN–RGT2 11 ± 1.5 27 ± 4 12 rgt2Δ snf3Δ vector 0.42 ± 0.2 0.88 ± 0.2 6 ± 1.4 14 ± 3 13 ADH–HXT1 0.7 ± 0.2 0.7 ± 0.1 4 ± 0.9 12 ± 1 Abbreviations: Gly, 5% glycerol + 0.5% galactose; Gly + 0.1% glu, 5% glycerol + 0.1% glucose; 4% Glu, 4% glucose. Vector = pRS316; CEN–SNF3 = pBM3111; CEN–RGT2 = pBM3272; ADH–SNF3 = pBM3135; ADH–RGT2 = pBM3333; ADH–HXT1 = pBM3362. The data shown in lanes 6C/D and 9A/B are from Özcan et al. (1996b). Induction of HXT1 expression by high concentrations of glucose is completely abolished in the snf3 rgt2 double mutant (Table I, line 12B). Consequently, the snf3 rgt2 double mutant grows poorly on high concentrations of glucose (Figure 3; see below). This is in contrast to snf3 mutants, which exhibit no reduction of high glucose-induced HXT1 expression (line 9B), and rgt2 mutants, which have only ∼5- to 6-fold reduction in HXT1 transcription at high concentrations of glucose (line 6B) (Özcan et al., 1996b). Thus, Snf3 appears to contribute to induction of HXT1 transcription by high levels of glucose. We believe that these results reflect the different relative affinities of these sensors for glucose: Rgt2 is probably a sensor of high levels of glucose (a low-affinity receptor); Snf3 is probably a sensor of low levels of glucose (a high-affinity receptor; see Discussion). Figure 1.Analysis of glucose transport activity of overexpressed SNF3, RGT2 and HXT1 genes in hxt1Δ–hxt7Δ (hxt) strain. The hxt mutant transformed with the ADH1 vector alone, ADH1–SNF3, ADH1–RGT2 and ADH1–HXT1 was scored for growth on YNB medium containing either 2% galactose or 2% glucose with antimycin A (1 μg/ml). The cells were grown first on YNB medium lacking uracil with 2% galactose as carbon source and then replica plated on YNB medium containing 2% glucose with antimycin A. Download figure Download PowerPoint Snf3 and Rgt2 are limiting components of the glucose signaling pathway Previous data obtained with dominant SNF3-1 and RGT2–1 mutants suggested that these two proteins are the limiting components of the glucose signaling pathway (Özcan et al., 1996b). This is supported by the observation that overexpression of either SNF3 or RGT2 causes constitutive expression of HXT1 and HXT2 (i.e. even in the absence of glucose) (Table I, lines 4A and 4C, and lines 5A and 5C). We believe this is because the glucose receptors, like all receptors, are in equilibrium between the unliganded and ligand-bound form. Higher levels of the receptors necessarily increase the amount of receptor in the ligand-bound form, leading to constitutive signaling. Note that HXT1 expression in the absence of glucose is only partially constitutive (lines 4A and 5A); this is because induction of HXT1 expression at high levels of glucose requires a second, Rgt2-independent pathway (Özcan et al., 1996b). Glucose transport is neither necessary nor sufficient for signaling To test whether Snf3 and Rgt2 can transport glucose, we expressed them in a strain unable to grow on glucose because it is deleted for seven HXT genes (hxt1Δ–hxt7Δ, called the hxt null mutant) (Reifenberger et al., 1995, 1997) (Figure 1). Expression in this strain of any one of the seven HXT genes restores growth on glucose (Reifenberger et al., 1995). We overexpressed in this strain SNF3, RGT2 and HXT1 from the ADH1 promoter on a multicopy plasmid and assayed growth on glucose media. Overexpression of HXT1 fully restored growth of the hxt null strain. By contrast, neither SNF3 nor RGT2, when overexpressed, were able to restore growth of the hxt null mutant (Figure 1). While we cannot be certain that increased levels of Rgt2 and Snf3 are expressed and reach the membranes in these cells, the fact that expression of SNF3 and RGT2 from the ADH1 promoter leads to constitutive HXT gene expression (Table I, lines 4A, 4C, 5A and 5C) suggests that this is the case. Thus, even though they are similar to glucose transporters, Snf3 and Rgt2 appear unable to transport sufficient amounts of glucose to correct the growth defect of the hxt null mutant. Figure 2.(A) Schematic representation of the Snf3, Rgt2 and Hxt1 protein structure. The repeats of the C-terminal tails of Snf3 and Rgt2 are indicated by boxes. In addition, the C-terminal tail deletions from Table III are shown. The numbers indicate the amino acid position. (B) Alignment of the repeated sequences in the C-terminal tail of Snf3 and Rgt2. Snf3 has two repeats [amino acids 678–702 (2. repeat) and 774 to 798 (1. repeat)], Rgt2 has only one (666–690). The amino acids that are identical within the repeats are indicated. Download figure Download PowerPoint To test if bona fide glucose transporters can provide for glucose signaling, we expressed the HXT1 and HXT2 genes from the ADH1 promoter on a multicopy plasmid in snf3 and rgt2 mutant cells and tested for restoration of HXT gene expression (Table II). Both of these plasmids express functional glucose transporters because they enable a mutant defective in glucose transport to grow on glucose (Figure 1, and data not shown), but neither is able to restore the glucose signaling defect of snf3 or rgt2 mutants, indicating that the HXT induction defect in these mutants is not due simply to impaired glucose transport. Table 2. The glucose signaling defect of rgt2Δ and snf3Δ mutants is not simply caused by impaired transporta Genotype Plasmid Mean β-galactosidase activity (U) ± SD HXT1::lacZ HXT2::lacZ AGly B4% Glu CGly DGly + 0.1% Glu 1 WT vector 1.1 ± 0.15 353 ± 32 14 ± 0.5 348 ± 30 2 ADH–HXT1 1.3 ± 0.2 328 ± 31 20 ± 4 340 ± 16 3 ADH–HXT2 0.8 ± 0.16 336 ± 51 18 ± 4 334 ± 59 4 rgt2Δ vector 0.66 ± 0.14 68 ± 7 5 ADH–HXT1 0.5 ± 0.1 75 ± 11 6 ADH–HXT2 0.6 ± 0.12 61 ± 4 7 snf3Δ vector 10 ± 1.8 24 ± 2 8 ADH–HXT1 11 ± 2 29 ± 4 9 ADH–HXT2 9 ± 1 37 ± 6 a Abbreviations: see Table I; vector = pRS426; ADH–HXT1 = pBM3362; ADH–HXT2 = pBM3138. The C-terminal tails of Snf3 and Rgt2 are necessary for glucose signaling An unusual feature of Snf3 and Rgt2 that distinguishes them from all other known glucose transporters is their long C-terminal tails, which are predicted to reside in the cytoplasm. The sequences of the Snf3 and Rgt2 tails are dissimilar, except for a stretch of 25 amino acids, 16 of which are identical among the repeats. Snf3 has two of these sequences; Rgt2 has only one (Figure 2). Deletion of the Rgt2 C-terminal tail (RGT2Δ2) abolishes its ability to sense high levels of glucose and induce HXT1 expression (Table III, line 3B), and deletion of the Snf3 tail (SNF3Δ2) abolishes its ability to sense low levels of glucose and induce HXT2 expression (line 7D). Furthermore, the dominant mutations SNF3-1 and RGT2-1 (Arg231 and Arg229, respectively, changed to lysine), which cause constitutive (glucose-independent) expression of the HXT genes (Table III, lines 10A and C, and 12A and C; see also Özcan et al., 1996b), do not manifest their effect when the C-terminal tails of Snf3 and Rgt2 are deleted (lines 11A and C, and 13A and C). Thus, the C-terminal tails of both Snf3 and Rgt2 are essential for glucose signaling. Figure 3.The growth defect of the snf3 rgt2 double mutant on 2% glucose with antimycin A (1 μg/ml) is complemented by overexpression of the HXT1 gene. Cells were grown first on YNB–2% galactose plates and then transferred (replica plated) to YNB–2% glucose plates with antimycin A. Download figure Download PowerPoint Table 3. The C-terminal tail of Rgt2p and Snf3p is essential for glucose signalinga Genotype Plasmid Mean β-galactosidase activity (U) ± SD HXT1::lacZ HXT2::lacZ AGly B4% Glu CGly DGly + 0.1% Glu 1 rgt2Δ vector 0.6 ± 0.14 68 ± 7 2 CEN–RGT2 1.6 ± 0.1 403 ± 69 (5.9×) 3 RGT2Δ2 1.0 ± 0.3 71 ± 10 (1.0×) 4 RGT2Δ1 1.2 ± 0.2 134 ± 24 (2.0×) 5 snf3Δ vector 10 ± 1.8 24 ± 2 6 CEN–SNF3 21 ± 3 332 ± 77 (13.8×) 7 SNF3Δ2 11 ± 2 21 ± 4 (0.88×) 8 SNF3Δ1 13 ± 2 69 ± 15 (2.9×) 9 WT vector 1.1 ± 0.15 353 ± 32 14 ± 0.5 348 ± 30 10 RGT2-1 33 ± 4 384 ± 13 272 ± 36 367 ± 74 11 RGT2-1ΔT 1.3 ± 0.3 360 ± 56 13 ± 3 336 ± 38 12 SNF3-1 29 ± 5 349 ± 37 301 ± 31 327 ± 57 13 SNF3-1ΔT 0.9 ± 0.1 355 ± 48 16 ± 3 359 ± 32 a Abbreviations: see Table I; vector = pRS316; CEN–RGT2 = pBM3272; RGT2Δ1 = pBM3312; RGT2Δ2 = pBM3279; CEN–SNF3 = pBM3111; SNF3Δ1 = pBM3319; SNF3Δ2 = pBM3363; RGT2-1 = pBM3270; RGT2-1ΔT = pBM3277; SNF3-1 = pBM3259; SNF3-1ΔT = pBM3335. The 25 amino acid repeats seem to be the functional units of the C-terminal tails of Snf3 and Rgt2 because an Rgt2 protein that retains its 25 amino acid repeat but is missing all distal sequences (RGT2Δ1) is still partially functional, mediating 2-fold induction of HXT1 expression (Table III, compare lines 1B, 2B and 4B). Similarly, Snf3 retaining one of its two repeats (SNF3Δ1) is partially functional, providing ∼3-fold induction of HXT2 expression (compare lines 5D, 6D and 8D). However, overexpression of the Snf3 or Rgt2 tails by themselves does not restore the glucose signaling defect of snf3 or rgt2 mutants (Table IV, lines 3B, 4B, 7D and 8D). Thus, the C-terminal tails of Rgt2 and Snf3 are necessary, but are not by themselves sufficient, for signaling. Table 4. Overexpression of the C-terminal tail of Rgt2p and Snf3p is not sufficient for glucose signalinga Genotype Plasmid Mean β-galactosidase activity (U) ± SD HXT1::lacZ HXT2::lacZ AGly B4% Glu CGly DGly + 0.1% Glu 1 rgt2Δ vector 0.6 ± 0.14 68 ± 7 2 CEN–RGT2 1.6 ± 0.1 403 ± 69 3 RGT2-T 0.6 ± 0.04 64 ± 5 4 SNF3-T 0.5 ± 0.07 67 ± 11 5 snf3Δ vector 10 ± 1.8 24 ± 2 6 CEN–SNF3 21 ± 3 332 ± 77 7 RGT2-T 13 ± 3 28 ± 7 8 SNF3-T 12 ± 1 22 ± 3 a Abbreviations: see Table I; vector = pRS426; CEN–RGT2 = pBM3272; CEN–SNF3 = pBM3111; SNF3-T = pBM3578; RGT2-T = pBM3576. Attachment of the Snf3 C-terminus to Hxt1 and Hxt2 converts these glucose transporters into glucose sensors To test whether the C-terminal tail of Snf3 allows glucose signaling when attached to other glucose transporters, we attached it to the Hxt1 and Hxt2. Indeed, both HXT1–SNF3 and HXT2–SNF3 chimeras partially complement the signaling defect of snf3 and rgt2 mutants (Table V, lines 3B, 4B, 8D and 9D). Since neither wild-type HXT1 nor HXT2 could repair the signaling defect of snf3 and rgt2 mutants (even when overexpressed) (Table II), we conclude that the C-terminal tail of Snf3 confers upon these glucose transporters the ability to signal glucose availability. However, these chimeric proteins do not signal as well as Snf3 or Rgt2, indicating that other residues of Snf3 and Rgt2 (probably in the transmembrane domain) are important for optimal function of the glucose sensors. Table 5. Attachment of Snf3p C-terminus to Hxt1p and Hxt2p converts them into glucose sensorsa Genotype Plasmid Mean β-galactosidase activity (U) ± SD HXT1::lacZ HXT2::lacZ AGly B4% Glu CGly DGly + 0.1% Glu 1 rgt2Δ vector 0.6 ± 0.14 68 ± 7 2 CEN–RGT2 1.6 ± 0.1 403 ± 69 (5.9×) 3 pHXT1-HXT1/SNF3 0.8 ± 0.06 182 ± 41 (2.7×) 4 pHXT2-HXT2/SNF3 0.7 ± 0.14 153 ± 37 (2.3×) 5 pADH-HXT1/SNF3 5 ± 0.8 263 ± 20 (3.9×) 6 snf3Δ vector 10 ± 1.8 24 ± 2 7 CEN–SNF3 21 ± 3 322 ± 77 (13×) 8 pHXT1-HXT1/SNF3 9.2 ± 1.2 89 ± 11 (3.7×) 9 pHXT2-HXT2/SNF3 11 ± 1.5 117 ± 21 (4.9×) 10 pADH-HXT1/SNF3 56 ± 6 94 ± 11 (3.9×) a Abbreviations: see Table I; vector = pRS316; CEN–RGT2 = pBM3272; CEN–SNF3 = pBM3111; pHXT1-HXT1/SNF3 = pBM3436; pHXT2-HXT2/SNF3 = pBM3654; pADH-HXT1/SNF3 = pBM3273. Overexpression of the HXT1–SNF3 chimera from the ADH1 promoter in the snf3 and rgt2 mutant strain also causes a low level of constitutive HXT expression (i.e. in the absence of glucose) (Table V, lines 5A and 10C). This is similar to results obtained when SNF3 and RGT2 were overexpressed (Table I), and further supports the view that the concentration of glucose sensors is the limiting factor for signaling. Snf3 and Rgt2 are required for glucose repression of GAL1 and SUC2 expression The snf3 rgt2 double mutant is severely defective in induction of HXT expression (like grr1 mutants), and therefore grows poorly on glucose-containing media. Overexpression of HXT1 (from the ADH1 promoter on a multicopy plasmid) in the snf3 rgt2 double mutant restores growth of this mutant on glucose (Figure 3), suggesting that its poor growth on glucose is due to a defect in glucose transport. Overexpression of HXT1 does not, however, repair the glucose induction defect in HXT expression (Table I, lines 13B and D), supporting the idea that the glucose induction signal is generated independently of glucose metabolism. Because the snf3 rgt2 double mutant has severely reduced glucose transport, we expected it to be defective in glucose repression of GAL1 and SUC2 expression (like grr1 mutants) (Flick and Johnston, 1991; Özcan et al., 1994). Indeed, GAL1 and SUC2 expression is not repressed by 4% glucose in the snf3 rgt2 double mutant, in contrast to snf3 or rgt2 single mutants (Table VI). The snf3 rgt2 double mutant displays a 3-fold decrease in SUC2 expression at low concentrations of glucose because low levels of glucose induce SUC2 expression, and this requires SNF3 (Özcan et al., 1997). Table 6. The snf3Δ rgt2Δ double mutant is defective in glucose repressiona Genotype Mean β-galactosidase activity (U) ± SD GAL1::lacZ SUC2::lacZ Glu/gal Gal Glu Gly + 0.1% Glu WT <0.1 128 ± 19 <0.1 75 ± 12 snf3Δ rgt2Δ 33 ± 7 137 ± 26 16 ± 3 26 ± 5 rgt2Δ <0.1 112 ± 15 60% identical to Snf3 and Rgt2 from S.cerevisiae. Interestingly, the K.lactis Rag4 protein also has a long C-terminal tail of ∼251 amino acids that contains one copy of the same repeated sequence found in the C-terminal tails of Snf3 and Rgt2. This finding is consistent with the idea that the 25 amino acid C-terminal repeat plays an important role in signaling by Snf3 and Rgt2. It is possible that Rag4 also functions as a glucose sensor in the yeast K.lactis, and might regulate the high glucose-induced expression of the K.lactis hexose transporter RAG1. Other transporters of small molecules also function as sensors. Sensing of certain sugars by bacteria is mediated by sugar transporters (Postma et al., 1993; Saier et al., 1996). However, this cannot be viewed as a receptor-mediated event, because signal generation is coupled to transport and metabolism (phosphorylation) of the sugar. The glucose transporter Rco3 of Neurospora crassa may function as a nutrient sensor; like Snf3 and Rgt2, it is required for expression of glucose transporter activity, glucose regulation of gene expression and glucose repression (Madi et al., 1997). Mep2 app