The Unique Hexokinase of Kluyveromyces lactis

The Crabtree-negative yeast Kluyveromyces lactis is capable of adjusting its glycolytic flux to the requirements of respiration by tightly regulating glucose uptake. RAG5 encoding the only glucose and fructose phosphorylating enzyme present in K. lactis is required for the up-regulation of glucose transport and also for glucose repression. To understand the significance of the molecular identity and specific function(s) of the corresponding kinase to glucose signaling, RAG5 was overexpressed and its gene product KlHxk1 (Rag5p) isolated and characterized. Stopped-flow kinetics and sedimentation analysis indicated a monomer-homodimer equilibrium of KlHxk1 in a condition of catalysis, i.e. in the presence of substrates and products. The kinetic constants of ATP-dependent glucose phosphorylation identified a 53-kDa monomer as the high affinity/high activity form of the novel enzyme for both glycolytic substrates suggesting a control of glucose phosphorylation at the level of dimer formation and dissociation. In contrast to the highly homologous hexokinase isoenzyme 2 of Saccharomyces cerevisiae (ScHxk2), KlHxk1 was not inhibited by free ATP in a physiological range of nucleotide concentration. Mass spectrometric sequencing of tryptic peptides of KlHxk1 identified unmodified serine at amino acid position 156. The corresponding amino acid in ScHxk2 is serine 157, which represents the autophosphorylation-inactivation site. KlHxk1 did not display, however, the typical pattern of inactivation under the respective in vitro conditions and maintained a high residual glucose phosphorylating activity. The biophysical and functional data are discussed with respect to a possible regulatory role of KlHxk1 in glucose metabolism and signaling in K. lactis.

The Crabtree-negative yeast Kluyveromyces lactis is capable of adjusting its glycolytic flux to the requirements of respiration by tightly regulating glucose uptake. RAG5 encoding the only glucose and fructose phosphorylating enzyme present in K. lactis is required for the up-regulation of glucose transport and also for glucose repression. To understand the significance of the molecular identity and specific function(s) of the corresponding kinase to glucose signaling, RAG5 was overexpressed and its gene product KlHxk1 (Rag5p) isolated and characterized. Stopped-flow kinetics and sedimentation analysis indicated a monomer-homodimer equilibrium of KlHxk1 in a condition of catalysis, i.e. in the presence of substrates and products. The kinetic constants of ATP-dependent glucose phosphorylation identified a 53-kDa monomer as the high affinity/high activity form of the novel enzyme for both glycolytic substrates suggesting a control of glucose phosphorylation at the level of dimer formation and dissociation. In contrast to the highly homologous hexokinase isoenzyme 2 of Saccharomyces cerevisiae (ScHxk2), KlHxk1 was not inhibited by free ATP in a physiological range of nucleotide concentration. Mass spectrometric sequencing of tryptic peptides of KlHxk1 identified unmodified serine at amino acid position 156. The corresponding amino acid in ScHxk2 is serine 157, which represents the autophosphorylation-inactivation site. KlHxk1 did not display, however, the typical pattern of inactivation under the respective in vitro conditions and maintained a high residual glucose phosphorylating activity. The biophysical and functional data are discussed with respect to a possible regulatory role of KlHxk1 in glucose metabolism and signaling in K. lactis.
Hexokinases catalyze the intracellular trapping and initiation of metabolism of glucose, fructose, and mannose and are essentially involved in glucose signaling in yeast, plants, and mammals (1)(2)(3)(4)(5)(6). In higher organisms, their glycolytic sugar substrate plays a dual role by acting as carbon source and hormone-like regulator (1). The yeast Saccharomyces cerevisiae has three genes encoding glucose kinases (ScHXK1, 1 ScHXK2, and ScGLK1), which are singly adequate to allow growth on glucose (7,8), however, only ScHXK1 and ScHXK2 can mediate repression by glucose of transcription of genes involved in the utilization of alternative carbon sources, in gluconeogenesis and respiration (9 -11). Transcriptional regulation in response to glucose in this yeast requires transport and phosphorylation of the sugar (1,3,9,12), whereas integral membrane proteins apparently function as sensors of the external glucose concentration (13,14). Signal transduction induced by high external glucose finally results in a predominance of ScHxk2, whereas expression of ScHXK1 and ScGLK1 is high on non-fermentable carbon sources or galactose (15)(16)(17).
Experimental evidence indicating phosphoryl transfer by any glucose kinase to be required but not sufficient for glucose signaling in S. cerevisiae (9,18) was supported by the finding that glucose limitation stimulates phosphorylation of ScHxk2 in vivo (19) and promotes dissociation of the homodimeric enzyme in vitro (20 -22). Monomeric ScHxk2 displays increased substrate affinity and is inhibited by free ATP (whereas dimeric ScHxk2 is not) in a physiological range of nucleotide concentrations (23). The phosphorylated amino acid serine 14 is part of a protein kinase A consensus sequence, which is preceded by a nuclear localization motif mediating the observed transfer of the enzyme into the nucleus (10,24). ScHxk2 is phosphorylated in vitro at another site (serine 157) by autocatalysis when non-phosphorylateable 5-carbon analogs of glucose are present (25). The latter modification is accompanied by a complete, but fully reversible loss of glucose kinase activity (25). Remarkably, mutation of serine 157 to alanine caused abrogation of glucose-induced, but not fructose-induced, repression (18).
The Crabtree-negative yeast Kluyveromyces lactis is receiving increasing attention as a model organism in comparative studies on functional genomics in unicellular and higher eukaryotes (26 -28). In contrast to the Crabtree-positive yeast S. cerevisiae (29), in which the glycolytic flux is primarily correlated with the external glucose concentration, regulation of glucose uptake in K. lactis is apparently coupled to oxygen supply (26,30,31). The sensitivity of K. lactis to glucose repression is highly strain dependent (32,33) with glucose repressible strains being characterized by the presence of a glucose transporter gene that does not exist in strains that are insensitive to the sugar (34). Although the molecular mechanism of glucose repression in K. lactis is largely unknown, the expression of glucose transporters and the glucose transport capacity of the cell are likely to be involved (31,34,35).
The gene RAG5 of K. lactis encodes a protein that shares 70% identity with ScHxk1 and 73% identity with ScHxk2 (35). The observation that rag5 mutants were unable to utilize glucose and displayed neither glucokinase nor fructokinase activity led to the conclusion that RAG5 encodes the only hexokinase (EC 2.7.1.1) present in K. lactis (35). Remarkably, rag5 mutants were also affected in high and low affinity glucose transport and showed relief from glucose repression of several enzymes (35)(36)(37). In addition, K. lactis hexokinase (KlHxk1, Rag5p) turned out to be essential for glucose-induced transcription of the low affinity glucose transporter gene RAG1 and the KlPDCA gene encoding pyruvate decarboxylase (35,38) and was involved in the expression of the high affinity glucose permease Hgt1 (37). Strong evidence for a role of KlHxk1 in glucose signaling came from the in vivo complementation of a rag5 mutation by ScHXK2 and of hxk1 hxk2 mutations by RAG5. ScHXK2 and RAG5 turned out equivalent in complementing hexose phosphorylation, however, RAG5 failed to restore glucose repression in S. cerevisiae (35). The obvious conclusion that the individual molecular structure of the kinase might play a discriminating role in protein-protein interactions mediating the glucose effect was confirmed by a similar study on hexokinase of Hansenula polymorpha. 2 This paper presents basic molecular and functional properties of the unique hexokinase existing in K. lactis. The experimental approach was focused on the identification of general features enabling glucose kinases to participate in glucose-dependent signal transduction. Data are discussed with special consideration of the molecular mechanisms controlling glucose kinase activity of KlHxk1 including oligomer formation and dissociation and its significance to enzyme function and glucose signaling.
Purification of KlHxk1-K. lactis mutant strain JA6⌬rag5/pTSRAG5 carrying RAG5 in multicopy was grown in yeast nitrogen base medium lacking uracil and supplemented with adenine and amino acids (SC medium) according to Ref. 44 at 2% glucose in a Biostat ED type fermenter (B. Braun Diessel Biotech, Melsungen, Germany) at pH 5.5 and 30°C under permanent aeration until an A 600 of 5-6. Typically, 100 g of frozen yeast (Ϫ20°C) was resuspended in buffer A (50 mM potassium phosphate ϩ 1 mM EDTA ϩ 5 mM 2-mercaptoethanol ϩ 1 mM PMSF, pH 7.4) to give a cell density of 30 -35% (v/v). The cell suspension was vigorously shaken in a Vibrogen Cell Mill Vi4 (Buehler, Tuebingen, Germany) 2 ϫ 2 min at 4°C using glass beads of 0.35 mm diameter. After treatment with protamine sulfate (0.2% m/v) and pH readjust-ment, the supernatant was stirred into DEAE-Sepharose Fast Flow (70 g wet weight; Amersham Biosciences) equilibrated with buffer B (20 mM Tris/HCl ϩ 1 mM EDTA ϩ 5 mM 2-mercaptoethanol ϩ 1 mM PMSF, pH 7.4). The gel was washed with buffer B followed by buffer B ϩ 100 mM NaCl. Hexokinase was eluted by buffer B ϩ 250 mM NaCl and precipitated with solid ammonium sulfate at 90% saturation. Buffer B was added to the precipitate until a saturation of ammonium sulfate of 50% was reached. The supernatant was desalted (PD 10; Amersham Biosciences) and subjected to chromatography on a Resource Q column (Amersham Biosciences) equilibrated with buffer B. Hexokinase was desorbed at pH 7.4 using a linear gradient of NaCl from 0 to 250 mM and precipitated with ammonium sulfate as above. The precipitate was dissolved in buffer A containing ammonium sulfate at 50% saturation and loaded on a Resource ether column (Amersham Biosciences) equilibrated with the same buffer. The enzyme was eluted by linearly decreasing the salt concentration from 50 to 12.5% saturation, precipitated with ammonium sulfate at 90% saturation, and stored at 2-4°C.
Determination of Protein Concentration-The concentrations of purified KlHxk1 and ScHxk2 were determined spectrophotometrically at 280 nm using the absorption coefficients of 51,490 M Ϫ1 cm Ϫ1 and 43,240 M Ϫ1 cm Ϫ1 as calculated from DNA sequences (cf. Swiss-Prot accession numbers P33284 and P04807) according to Ref. 45.
N-terminal Sequencing, Tryptic Digestion, and Mass Spectrometry-30 g of KlHxk1 was bound to 10 mg of Poros 20 R1 reverse-phase material (Applied Biosystems, Weiterstadt, Germany) in a pipette tip with a plug of fine glass wool as a filter. After washing the gel with 3ϫ 100 l of 0.1% trifluoroacetic acid, the protein was eluted with 60 l of 60% acetonitrile containing 0.1% trifluoroacetic acid. One aliquot of this solution was sequenced in a Procise 494 protein sequencer (Applied Biosystems, Weiterstadt, Germany), another aliquot (50 l) was lyophilized, dissolved in 50 l of 100 mM Tris/HCl buffer, pH 8.2, and digested using 1 g of sequencing grade trypsin at 37°C for 15 h. The molecular mass of KlHxk1 was determined by electrospray ionization mass spectrometry employing a Q-TOF1 mass spectrometer (Micromass, Manchester, UK) with an accuracy of 0.01%. For mass measurements, aliquots of the desalted enzyme solution were taken to dryness in a SpeedVac centrifuge and dissolved in a mixture of methanol, water, and formic acid (50:49:1). The solution of tryptic peptides was introduced using the Micromass CapLC system comprising a low flow capillary LC pump and an autosampler. The mass spectrometer survey scan was acquired from m/z 450 to 1,000 with the switching criteria for MS to MS/MS including ion intensity and charge state. The Q-TOF1 instrument performed MS/MS on up to three co-eluting species. MS/MS spectra were processed to generate a peak list. The resulting spectra were searched against a non-redundant data base using MASCOT (46). 3 Equilibrium Sedimentation-Experiments were performed in a Beckman analytical ultracentrifuge XL-A using an An50Ti rotor equipped with six-channel cells. KlHxk1 was desalted in and equilibrated with 50 mM potassium phosphate buffer, pH 7.4, containing 1.0 mM EDTA, 1.0 mM PMSF, and 0.2 mM DTT. Sedimentation equilibrium was analyzed at 15,000 rpm and 15°C at scanning wavelengths of 230 and 280 nm. Calculation of the apparent molecular mass, M r(app) employed a partial specific volume of 0.73 ml/g (47). From the dependence of the M r(app) values on the enzyme concentration, dissociation constants were calculated for a monomer-homodimer equilibrium according to Equations 1-3. M and D represent the molecular mass of monomer and homodimer, respectively (Equation 1).
The equilibrium concentration of dimer, c D , can be expressed as a function of the equilibrium concentration of monomer, c M , and the dissociation constant, K D (Equation 2).
c M is related to the total (loading) enzyme concentration, c total , in a more complex manner (Equation 3).
Determination of ␤-Galactosidase Activity-The activity of crude ex-tracts prepared by glass-bead cell disruption was determined using o-nitrophenyl-␤-D-galactopyranoside as described in Ref. 48. One milliunit of ␤-galactosidase catalyzes the formation of 1 nmol of o-nitrophenolate/min at 30°C. Stopped-flow Analysis of Glucose Kinase Activity of KlHxk1-The dye indicator linked assay responding to proton liberation as catalyzed by hexose kinases has been described previously (23). Experiments were carried out at 25°C with an Applied Photophysics BioSequential DX.18 MV stopped-flow spectrometer (Leatherhead, UK). The slit widths of both monochromators were set to 7 mm. KlHxk1 was equilibrated on a Superdex 75 HR 10/30 column (Amersham Biosciences) with incomplete assay buffer (buffer I) consisting of 4.0 mM TEA/HCl, 10 mM MgCl 2 , and 1.0 mM DTT at pH 7.4. Kinetic studies were carried out in complete assay buffer (buffer C) prepared by supplementing buffer I with glucose, ATP, and p-nitrophenol(ate) and its subsequent mixing with equal volumes of KlHxk1 solution to give the final substrate and enzyme concentrations indicated. The buffering action of ATP was taken into account by stopped-flow titration of buffer C with HCl at different concentrations of the nucleotide substrate. The corresponding absorption coefficients of p-nitrophenol(ate) (Fig. 1, panel A) allow calculation of authentic enzyme activities. In addition, buffer I containing commercially available yeast hexokinase at the highest enzyme concentration employed in this study (1 mg/ml) was titrated with HCl. The close similarity of the corresponding two data sets (Fig. 1, panel B) indicates a negligible contribution of the protein to the buffering capacity of the assay system.
Autophosphorylation-Inactivation Assay-KlHxk1 and ScHxk2 were incubated in 50 mM Hepes buffer containing 12 mM MgCl 2 , 1.0 mM DTT, 4.0 mM ATP, and 100 mM D-xylose at pH 7.5 and 30°C as already described (25). The enzyme concentration was 167 g/ml, respectively. Glucose kinase activity was measured using glucose-6-phosphate dehydrogenase as auxiliary enzyme to monitor the time course of inactivation. ATP and D-xylose were omitted in the controls corresponding to a residual activity of 100%.
Biochemicals-Media constituents were from Difco, p-nitrophenol was a Merck (Darmstadt, Germany) product. PMSF and DTT were obtained from Sigma. ATP (special quality disodium salt), NADP ϩ (disodium salt), glucose 6-phosphate (disodium salt), glucose-6-phosphate dehydrogenase, phosphoglucose isomerase, yeast hexokinase, and SDS-PAGE standard were from Roche Diagnostics, sequencing grade trypsin was from Promega. ScHxk2 was prepared according to Ref. 23. All other substances used were of analytical grade.

RESULTS
Glucose Sensitivity of KlHxk1 Donor Strain-Crude extracts of multicopy strain JA6⌬rag5/pTSRAG5 prepared by French Press cell disruption following growth at 2% glucose contained 4 -5-fold higher glucose kinase activity than extracts of the corresponding wild type strain JA6 expressing RAG5 in single copy. Whereas glucose repression of ␤-galactosidase was clearly relieved in the JA6⌬rag5 deletion mutant (Table I, upper line), plasmid pTSRAG5 fully restored the glucose effect (bottom line). The activity of ␤-galactosidase was higher in the rag5 deletion mutant both in the presence and absence of the inductor galactose (Table I, 977 versus 6.4 and 16.8 versus 4.1 milliunits/mg). Wild type cells grown in the absence of glucose exhibited activity levels of ␤-galactosidase that were similar to those of the rag5 mutants (1,250 and 20 milliunits/mg for growth at 2% galactose or 3% glycerol, respectively) indicating that RAG5 was essential for mediating glucose repression.
Molecular Characterization of KlHxk1-SDS-PAGE following 25-fold purification of KlHxk1 revealed a single protein band that migrated almost indistinguishably from the 54-kDa subunit of ScHxk2 (Fig. 2, lanes 5 and 6).  Since strain JA6⌬rag5 does not grow on glucose, it was pre-grown in glycerol medium, kept in glucose medium overnight, and subsequently shifted to (glucose ϩ galactose) medium.
b Value from single experiment. (25). Modification of the corresponding amino acid serine 156 of KlHxk1 and its functional consequences are considered below.
Oligomeric Structure of KlHxk1-The radial distribution of KlHxk1 at sedimentation equilibrium was analyzed in the absence of effectors and in the presence of hexose substrate and product. The resulting sets of apparent molecular mass data could be described quantitatively by assuming a monomerhomodimer equilibrium using a subunit molecular mass of 53,476 Da as determined by mass spectrometry (Fig. 3, Table  II). In the absence of substrate or product, KlHxk1 is largely dimeric at high enzyme levels, whereas millimolar glucose causes significant dissociation (Fig. 3, panel A). The dissociative effect of glucose 6-phosphate is less pronounced than that of the sugar substrate. At low enzyme concentrations, KlHxk1 is almost completely dissociated in all three conditions. The dependence of the equilibrium concentrations of monomer and homodimer on the enzyme concentration is illustrated for the effector-free condition in Fig. 3, panel B. The hydrodynamic data indicate a control of oligomer formation and dissociation by metabolic effectors and by the enzyme concentration itself under experimental conditions that might well correspond to a physiological situation (cf. "Discussion").
Stopped-flow Analysis of Glucose Kinase Activity of KlHxk1-The interrelation between oligomeric state and catalytic function of KlHxk1 was analyzed by monitoring the kinetics of ATP-dependent glucose phosphorylation at two protein concentrations (1 g/ml and 1 mg/ml) where, according to equilibrium sedimentation, the enzyme should predominantly exist in either the monomeric or homodimeric state. Fig. 4 illustrates the dependence of enzyme activity on the concentrations of glucose (panels A and C) and ATP (panels B and D) at low (panels A and B) and high (panels C and D) concentration of KlHxk1. The corresponding kinetic parameters indicate that glucose affinity of KlHxk1 when acting at low enzyme concentration is more than twice as high as compared with the high enzyme level (Table III). This observation is paralleled by a significant increase of V max reaching a specific catalytic activity at low enzyme concentration that is very similar to the corresponding value reported for ScHxk2 (23). Experiments employing variable ATP at a saturating concentration of glucose gave numerical values of K M(ATP) and V max(ATP) that are almost identical with those describing the dependence of enzyme activity on the concentration of the sugar (Table III).
To directly correlate structural and functional data, KlHxk1 was subjected to stopped-flow kinetic analysis in a range of enzyme concentration (1-1,000 g/ml) where significant changes of its oligomeric state were likely to occur (Fig. 5). Semilogarithmic presentation of these data gave an inversely sigmoidal curve indicating two plateau regions (Fig. 5, panel  B). Although a final value of glucose kinase activity could not be observed at the highest enzyme concentrations, application of the fitting procedure used to describe a one-step associationdissociation equilibrium gave a dissociation constant of K D ϭ (6.6 Ϯ 0.9) ϫ 10 Ϫ7 M. This value is most similar to the dissociation constant determined by sedimentation analysis in the presence of glucose 6-phosphate (Table II).
Preliminary kinetic investigation of KlHxk1 at low enzyme concentration employing glucose-6-phosphate dehydrogenase (and phosphoglucose isomerase, when fructose was the sugar   Table II for the effectorfree condition. Partial concentrations of monomers and dimers correspond to the respective total concentration of KlHxk1. substrate) in a coupled assay revealed that glucose and fructose allow the same affinity for ATP. The affinity for glucose is 6-fold higher than that for fructose with ATP as the second substrate. KlHxk1 also accepts nucleotide substrates other than ATP, however, replacement of ATP by ITP, UTP, CTP, and GTP results in a significant decrease of V max to ϳ40, 15, 5, and 5%, respectively (primary data not shown).
Inactivation of ScHxk2 and KlHxk1-The identification of unmodified serine 156 in KlHxk1 led to the investigation of the inactivation behavior of the enzyme because autophosphorylation of ScHxk2 at the corresponding serine 157 is known to impair glucose kinase activity (25). Incubation of the latter enzyme with MgATP and D-xylose, a non-phosphorylateable 5-carbon analog of glucose, caused the typical time course of inactivation resulting in a rapid and complete loss of glucose kinase activity (Fig. 6). In contrast, inactivation of KlHxk1 followed a first-order kinetics revealing an extrapolated residual glucose phosphorylating activity of ϳ60%.

DISCUSSION
The yeast S. cerevisiae employs three glucose kinases to fully accomplish catalysis and regulation of glucose metabolism (3,49,50). In contrast, a unique hexokinase is apparently adequate for efficient glucose sensing, transport, and phosphoryl-ation in K. lactis (35). The corresponding gene, RAG5, was cloned and sequenced, but information on the gene product KlHxk1 (Rag5p) was largely missing (35). This work demonstrates that purified KlHxk1 is capable of forming a homodimeric structure (Figs. 2 and 3). The subunit molecular mass of 53,476 Da as determined by electrospray ionization-MS confirms the DNA sequence recently published. 4 Comparison of the oligomeric stability of KlHxk1 and ScHxk2 (Table II) reveals a remarkable similarity for the effector-free condition and the effect of glucose 6-phosphate. Moreover, the dissociation constants of both enzymes determined in the presence of sugar phosphate are identical within the limits of error with the K D value calculated to fit the dependence of glucose kinase activity of KlHxk1 on the enzyme concentration (Table II and Fig. 5). This finding may reflect a predominating role of glucose 6-phosphate in the metabolic control of the oligomeric state of KlHxk1 under the conditions of catalysis. Evaluation of the dissociating effect exerted by glucose requires further investigation because the presence of the sugar alone seems to be without physiological significance. Because of the strong UV absorbance of the purine moiety, however, the influence of the nucleotide substrate and/or product on the oligomerization equilibrium has not been studied by sedimentation analysis.  The concentrations of glucose and ATP employed in this work correspond to a physiological condition in S. cerevisiae (51,52), but neither their respective in situ concentrations nor that of KlHxk1 in K. lactis are known. Based on the purification data of the enzyme (not shown) and its specific catalytic activity (Table III), estimation of the average cytosolic concentration of KlHxk1 in the glucose repressible strains indicated in Table  I gave values from ϳ0.5 to 1.5 mg/ml. According to sedimentation and stopped-flow analysis (Figs. [3][4][5], KlHxk1 should predominantly exist in the homodimeric state in the above range of enzyme concentration. The latter conclusion implies the possibility of monomer formation via covalent modification of KlHxk1 despite a high local enzyme concentration in the cell as already discussed for ScHxk2 (23). Indeed, KlHxk1 contains the same N-terminal stretch of amino acids, which carries in ScHxk2 the in vivo phosphorylation site serine 14 (20). Phosphorylation of this residue in ScHxk2 is stimulated by external glucose limitation (19) and promotes monomer formation (21). If hexokinase phosphorylation also takes place in K. lactis and depends on glucose availability in a similar manner, the identification of unmodified serine 14 in KlHxk1 described above is the expected result of the suppression of the modification by the high glucose concentration employed during cell growth (cf. "Experimental Procedures").
Tryptic peptide analysis has shown the presence of unmodified serine 156 in KlHxk1. This residue is part of the conserved sequence PLGFTFS(F/Y)PA existing in several yeast hexokinases, among them the enzymes of S. cerevisiae and H. polymorpha. 2 Autophosphorylation of ScHxk2 at the corresponding serine 157 is stimulated by non-phosphorylatable 5-carbon analogs of glucose and accompanied by a loss of glucose kinase activity (25). In contrast, functional investigation of the effect of D-xylose on KlHxk1 revealed a significantly different pattern of inactivation and a high residual glucose kinase activity (Fig.  6). Because serine 157 in ScHxk2 is located in the immediate vicinity of the glucose binding site (53,54), the inactivation pattern of the homologous KlHxk1 enzyme suggests functionally important differences in the architecture of its active center. The obvious question whether the resistance of KlHxk1 against inactivation in the presence of D-xylose is related on a molecular basis to the observed insensitivity of the enzyme to ATP inhibition (Fig. 4) cannot be answered at present. It should be noted, however, that KlHxk1 was not inhibited up to a calculated concentration of the free nucleotide of 0.1 mM at 5 mM total ATP (K A ϭ 10 4 M Ϫ1 for the formation of the Mg-ATP complex (55)). This finding is in contrast to the observed inhibition of ScHxk2 (23) and other yeast glycolytic enzymes (56) by ATP and contradicts a regulation of glycolysis in K. lactis via ATP inhibition of KlHxk1. In addition, it rules out a control of glucose kinase activity of KlHxk1 by the intracellular concentration of free magnesium ions determining the concentration of uncomplexed ATP as hypothesized for ScHxk2 (23).
The kinetic data portray monomeric KlHxk1 as the high affinity/high activity form of the enzyme for the two glycolytic substrates ( Fig. 4 and Table III). This finding raises the question whether and under which circumstances glucose uptake and phosphorylation may be improved in vivo through stimulation of KlHxk1 dissociation. Although the differences in substrate affinity and catalytic activity of monomeric and dimeric KlHxk1 did not exceed a factor of 2.5 (Table III), data may reflect a tendency and do not necessarily need to describe the in situ situation in quantitative terms. Likewise, it is still unknown whether the influence of metabolites on the oligomeric state of KlHxk1 observed in vitro (Fig. 3) has physiological consequences. Answers to these questions will have to consider that even moderate changes of metabolite and/or enzyme concentrations may allow an effective modulation of the glucosedependent signal transduction cascade. In the case of K. lactis galactokinase (57), even small changes in the concentration of the enzyme may have profound influence on signaling (48,58). Upon substrate binding but independent of catalysis, galactokinase binds to the Gal4 inhibitor Gal80 and relieves Gal80mediated repression of galactose metabolic genes. Thus, signaling also depends on the intracellular concentration of galactose and on the rate of galactose metabolism because both factors affect the concentration of the galactokinase-galactose complex. In contrast, no protein component transmitting the external glucose signal to the transcription machinery of K. lactis has been identified. If it exists, its regulatory function may (q) were 167 g/ml. Glucose kinase activity was determined using glucose-6-phosphate dehydrogenase as auxiliary enzyme. Controls lacking ATP and D-xylose correspond to a residual activity of 100%. specifically depend on the monomer-homodimer equilibrium of KlHxk1 that has been evolutionarily conserved in K. lactis and S. cerevisiae.