OUP user menu

Galactose transporters discriminate steric anomers at the cell surface in yeast

Toshio Fukasawa, Hiroshi Sakurai, Yasuhisa Nogi, Enrico Baruffini
DOI: http://dx.doi.org/10.1111/j.1567-1364.2009.00517.x 723-731 First published online: 1 August 2009


Aldose-1-epimerase or mutarotase (EC catalyzes interconversion of α/β-anomers of aldoses, such as glucose and galactose, and is distributed in a wide variety of organisms from bacteria to humans. Nevertheless, the physiological role of this enzyme has been elusive in most cases, because the α-form of aldoses in the solid state spontaneously converts to the β-form in an aqueous solution until an equilibrium of α : β=36.5 : 63.5 is reached. A gene named GAL10 encodes this enzyme in yeast. Here, we show that the GAL10-encoded mutarotase is necessary for utilization of galactose in the milk yeast Kluyveromyces lactis, and that this condition is presumably created by the presence of the β-specific galactose transporter, which excludes the α-anomer from the α/β-mixture in the medium at the cell surface. Thus, we found that a mutarotase-deficient mutant of K. lactis failed to grow on medium, in which galactose was the sole carbon source, but, surprisingly, that the growth failure is suppressed by concomitant expression of the Saccharomyces cerevisiae-derived galactose transporter Gal2p, but not by that of the K. lactis galactose transporter Hgt1p. We also suggest the existence of another mutarotase in K. lactis, whose physiological role remains unknown, however.

  • GAL10
  • galactose metabolism
  • epimerase
  • mutarotase
  • sugar transporter
  • yeast


Galactose is important for most living organisms including humans, where this sugar is not only an essential carbon source in newborns as a constituent of lactose but also a building block of nerve tissue as galactolipid. Galactose is metabolized through the so-called ‘Leloir pathway’ as shown in Fig. 1, which is conserved among most organisms. The genes encoding the enzymes responsible form an operon in Escherichia coli K12 (Buttin, 1963). In 1994, Adhya and his colleagues identified the galM gene encoding mutarotase as a new member of the galactose operon in E. coli K12 (Bouffard, 1994). Introduction of a deletion in galM does not affect utilization of galactose at all, however. Instead, the defect leads to slow growth in the medium containing phenyl-β-d-galactopyranoside (hereafter phenyl-β-d-galactose) as the sole carbon source. These authors suggested that mutarotase is involved in efficient utilization of lactose due to the following reason: phenyl-β-d-galactose, when taken up by the cell through the function of the β-galactoside transporter (Kennedy, 1970), is split into a phenyl group and β-d-galactopyranoside (hereafter β-d-galactose) by the catalytic action of β-galactosidase. The latter compound, but not the former, suitable as a carbon source, has to be converted to α-galactose by the action of mutarotase before entering the Leloir pathway, because the first enzyme of the pathway, galactokinase, uses only α-galactose, but not its β-anomer, as the substrate (Howard & Heinrich, 1965, see Fig. 1). Therefore, the galM mutants normally fail to grow in the medium containing phenyl-β-d-galactose as the sole carbon source. As a corollary, they concluded that autocatalytic water-dependent conversion of β-galactose to its α-form is not efficient enough for the Leloir pathway to support normal cell growth. To determine whether mutarotase is really required for utilization of galactose or lactose, we used Kluyveromyces lactis, which is capable of utilizing both galactose and lactose as the carbon source, exceptionally among yeasts.

Figure 1

The role of mutarotase in the Leloir pathway. Numbers in parentheses represent reactions catalyzed by (1) aldose-1-epimerase (mutarotase), (2) galactokinase, (3) galactose-1-phosphate uridylyl transferase, and (4) UDP-glucose-4-epimerase (epimerase). Note that no. 1 carbon is the target of both mutarotase and galactokinase.

In most organisms including bacteria (Poolman, 1990; Bouffard, 1994) and humans (Timson & Reece, 2003), mutarotase and UDP-glucose-4-epimerase (hereafter called epimerase), the third enzyme of the Leloir pathway, are encoded by two independent genes, galM and galE, respectively. In contrast, these enzymes are produced as a single peptide in yeasts, such as Saccharomyces cerevisiae, K. lactis, and Kluyveromyces fragilis. While epimerase was purified from yeasts including K. fragilis and S. cerevisiae since 1963 through 1980, it was noticed, surprisingly, that the molecular size of yeast enzymes was twice or more larger than that of epimerase from other sources, such as E. coli, wheat germ, porcine, or bovine (see Supporting Information, Table S1). This was solved by Poolman (1990), who found that the mutarotase gene cloned from Streptococcus thermophilus has strong similarity to the 3′-half of GAL10 from S. cerevisiae (Citron & Donelson, 1984). Majumdar (2004), Brahma & Bhattacharyya (2004), and Scott & Timson (2007), respectively, confirmed this fact using purified enzymes from S. cerevisiae and K. fragilis, showing that the epimerase and mutarotase activities reside separately in N- and C-terminal domains. The putative active sites have been located in the respective domains in a three-dimensional structure visualized by X-ray diffraction analysis (Thoden & Holden, 2005).

Materials and methods

Yeast genetic methods including media

Yeast genetic methods including media were those described by Sherman (2002). Yeast transformation was performed according to the ‘high efficient protocol’ described by Gietz & Woods (2002).

Yeast strains and plasmids

JA6gal10Δ was constructed in this work by insertion of the ‘popping-out’URA3 gene from S. cerevisiae between BglII and SalI sites in GAL10 from K. lactis JA6 (MATa, ade1-600 trp1-11 ura3-12), followed by selecting 5-fluoroorotic acid-resistant clones from Ura+ transformants (Alani, 1987). JA6gal10Δgal80Δ was also constructed in this work by replacing the GAL80 locus of JA6gal10Δ with the XhoI fragment of pD802, a plasmid carrying gal80URA3 (Zenke, 1993). The ORFs of ScGAL10 (P04397), KlGAL10 (P09609), and SpGAL10a (CAC21414) were PCR-cloned from S. cerevisiae W303-1a, K. lactis KA5-6C, and Schizosaccharomyces pombe JY742, respectively. Codes in parentheses indicate the accession numbers of NCBI. The yeast S. pombe possesses two GAL10 genes: a and b; the former encodes epimerase/mutarotase, whereas the latter encodes only epimerase. No evidence has been available to indicate whether these genes are transcribed in situ. The primers used for cloning are as follows:










Cloned GAL10 fragments were first inserted into pVT102U, an expression vector for S. cerevisiae (Vernet, 1987). The resulting plasmid then received a 70-bp-fragment-containing sequence for the histidine tag (His10) from pET16b (Novagen) at the 5′-end of GAL10. The region encompassing the His10-GAL10-ADH1 terminator was excised and inserted into p4XX, another expression vector of S. cerevisiae, which contains the promoter region of ScGAL1 and the selection marker of ScTRP1 (Mumberg, 1994). Finally, the automatically replicating sequence of S. cerevisiae in the resulting plasmid was replaced by that of K. lactis from pLF1 (Irene, 2004). A schematic representation of the plasmid construct is shown in Fig. S1.

Affinity purification of KlGal10p

Affinity purification of KlGal10p was carried out using TALON metal affinity resin (Clontech Laboratories Inc., Palo Alto, CA), essentially according to the supplier's protocol. In brief, cells of K. lactis strain JA6gal10Δgal80Δ carrying either pK394 or pK411, expressing the wild-type or mutarotase-domain-deleted Gal10p were grown to the late-log phase in a tryptophan-omitted synthetic medium at 30 °C. When the cell density reached an OD650 nm of 1.2–1.3, cells were harvested and washed with an equal volume of chilled water and then 0.1 M Tris/HCl (pH 8.0) containing 0.25 M NaCl. Cells were disrupted by vigorous agitation on a Vortex mixer in the presence of glass beads (d=0.5 mm), and the clear lysate obtained after high-speed centrifugation (100 000 g for 20 min) was gently mixed with TALON metal affinity resin for 30 min. The mixture was transferred into a column, and the resin was washed in a gravity-flow manner successively with buffer and buffer containing 5 mM imidazole. Gal10p or its deletion was eluted with buffer containing 50 mM imidazole. All the steps were carried out at a temperature <5 °C.

Preparation of concentrated crude extracts and ammonium sulfate fractionation

Cells of JA6gal10Δgal80Δ were grown to the late-log phase in 500 mL of YPD fortified with 20 μg mL−1 of adenine sulfate. When the cell density reached an OD650 nm of 1.2–1.3, cells were collected, washed twice with an equal volume of water, and finally 250 mL of 20 mM Tris/HCl buffer (pH 8.0) containing 0.2 M NaCl and 1 mM each of EDTA and dithiothreitol. Cells were disrupted on a Vortex mixer in the presence of glass beads. Cell debris and unbroken cells were removed by centrifugation at 1100 g for 20 min in a clinical centrifuge, and the supernatant sample was subjected to high-speed (100 000 g) centrifugation for 20 min to yield 100 mL of a clear lysate. Solid ammonium sulfate was added to the resulting clear lysate to obtain precipitates at the indicated concentrations. All steps were carried out at a temperature <5 °C.


UDP-glucose-4-epimerase was assayed using the two-step method described previously (Fukasawa, 1980); UDP-galactose was incubated with a sample in the first reaction, and the UDP-glucose formed was determined by formation of NADH in the second reaction consisting of UDP-glucose-dehydrogenase and NAD+ in a spectrophotometer. Mutarotase was determined with a polarimeter based on the change in optical rotation of α-d-glucose according to the method described by Majumdar (2004), whose validity is shown in Fig. S3. Mutarotase is assayed enzymatically in some experiments with β-glucose-specific dehydrogenase (Sigma/Aldrich Co.) according to the method described by Bouffard (1994).


Construction of the gal10 deletion K. lactis strain

First, we introduced deletion covering both domains of GAL10 in K. lactis strain JA6 requiring uracil and tryptophan to yield JA6gal10Δ. Because JA6gal10Δ was unable to grow in the presence of galactose, it is difficult to obtain sufficient amount of induced cell mass for an enzyme assay. We therefore introduced deletion into the GAL80 gene that encodes the specific inhibitor of Gal4p to generate JA6gal10Δgal80Δ, using pD802, a plasmid containing gal80Δ (Zenke, 1993). The loss of Gal80p function leads to constitutive synthesis of the enzymes in the Leloir pathway, so that these activities are fully expressed, even when cells were grown on glucose. The regulatory proteins Gal4p and Gal80p exert the respective functions in the same manner both in S. cerevisiae and in K. lactis (Schaffrath & Breunig, 2000). As shown in Table 1, the activity of epimerase and mutarotase became undetectable in JA6gal10Δgal80Δ. By contrast, JA6gal80Δ grown on glucose exhibited both activities comparable to those in the original strain JA6 grown on galactose.

View this table:
Table 1

UDP-glucose-4-epimerase (epimerase) and aldose-1-epimerase (mutarotase) activity in various Kluyveromyces strains

Yeast strainGene carried on plasmidCarbon sourceEpimeraseMutarotase
JA6NoneGalactose41 ± 1.034.6 ± 1.0
JA6 gal80ΔNoneGlucose42 ± 1.025.6 ± 1.0
JA6gal10Δgal80ΔKlGAL10Glucose58 ± 535.5
JA6gal10Δgal80ΔKlgal10mutΔGlucose31 ± 3<2.0
JA6gal10ΔKlGAL10+vacant vectorGalactose125 ± 6110 ± 4.0
JA6gal10ΔKlgal10mutΔ+ScGAL2Galactose61 ± 5<2.0
  • Epimerase activity was expressed as micromoles of UDP-glucose formed per hour per milligram of protein. Mutarotase activity was calculated from a first-order rate constant for catalyzed mutarotation of α-glucose subtracted with the rate constant of spontaneous mutarotation of α-glucose determined with a polarimeter (JASCO DIP36). The rate constant of catalyzed mutarotation reaction thus obtained was multiplied 1000 times for clarity in presentation. The values with ± are the average of two independent samples with deviations.

Construction of mutarotase-domain deleted gal10 gene on an expression vector

We PCR-cloned GAL10 from K. lactis in an expression vector as described in Materials and methods. Next, we introduced an internal deletion into the mutarotase domain of KlGAL10 in the plasmid (see Fig. S2). These plasmids pK394 or pK411 carrying either the wild-type GAL10 (KlGAL10) or the mutarotase-deleted gal10 (Klgal10mutΔ) were transferred into JAgal10Δgal80Δ. As is shown in Table 1, KlGAL10-carrying cells grown on glucose exhibited normal activities of both epimerase and mutarotase. By contrast, Klgal10mutΔ-carrying cells were specifically defective in mutarotase, but exhibited subnormal activity of epimerase. The lower than normal level may be due to the metabolic instability of the altered protein, because presumable degradation products were seen in polyacrylamide gel of metal-affinity purified protein (Fig. 2). The activity of epimerase and mutarotase in the partially purified enzyme is shown in Table 2.

Figure 2

Polyacrylamide (8%)/SDS (0.1%) gel electrophoresis of metal-affinity purified KlGal10p and its deletion. Lanes contained mutarotase domain-deleted Gal10mutΔp (3.0 μg), Gal10p (2.6 μg), and M, a set of prestained size markers (SeeBlue Pre-stained, Invitrogen Co.), respectively. Enzyme samples were those described in Table 2. Markers are myosin (188 kDa), phosphorylase B (98 kDa), bovine serum albumin (62 kDa), glutamic dehydrogenase (49 kDa), and alcohol dehydrogenase (38 kDa). The gel was stained with a silver staining kit (ATTO Co.).

View this table:
Table 2

UDP-glucose-4-epimerase activity of metal affinity-purified Kluyveromyces Gal10p or its mutarotase domain deleted protein

Epimerase (U mg−1 protein)Mutarotase (U mg−1protein)
  • One unit of epimerase or mutarotase is defined as 1.0 μmol of UDP-glucose converted to UDP-galactose or 1.0 μmol of α-glucose converted to β-glucose per hour, respectively. Mutarotase is assayed enzymatically with β-glucose-specific dehydrogenase (Sigma/Aldrich Co.) according to the method described by Bouffard (1994).

Growth behavior of mutarotase-deficient strains on various carbon sources

The transformants expressing either the wild-type or the mutarotase-deleted Gal10p were then tested for growth on galactose, lactose, and phenyl-β-d-galactose. As shown in Fig. 3, Klgal10mutΔ-carrying cells of JAgal10Δgal80Δ did not grow on synthetic minimal medium containing either galactose or phenyl-β-d-galactose as the sole carbon source, on which KlGAL10-carrying cells of the same strain grew normally. Both strains grew well when glucose or lactose was present in the medium as the sole carbon source.

Figure 3

Growth behavior of Kluyveromyces lactis strain JA6 gal10Δgal80Δ carrying TRP1-selectable plasmids that express the indicated Gal10p on various sugars. Each plate was streaked with one-loop-full fresh cultures (1–2 × 107 mL−1) of the indicated strain grown in tryptophan-dropout synthetic medium containing 0.5% each of glycerol and sodium lactate. Streaks are a tryptophan-nonrequiring revertant of JA6gal10Δgal80 without plasmid (none), JA6gal10Δgal80Δ bearing a plasmid expressing the wild-type Gal10p (GAL10), and two independent transformants with a plasmid expressing mutarotase-domain-deleted Gal10p (gal10mutΔ). Four plates of tryptophan-dropout synthetic agar contained 2% glucose, 0.25% galactose, 0.25% lactose, and 10 mM phenyl-β-d-galactose, which were incubated for 48 h at 30°C and stored in a refrigerator before being photographed. Similar results were observed when the concentration of galactose or lactose was increased to 1.0% (data not shown).

Suppression of growth failure of mutarotase-deficient strains by ScGal2p

Why, then, did the mutarotase-deficient E. coli K12 cells grow on galactose in the previous work (Bouffard, 1994)? In E. coli, galactose enters the cell by the galactose-specific transporter encoded by galP (Rotman, 1968) as well as by the β-galactoside-specific transporter encoded by lacY (Kennedy, 1970). We hypothesized that the galP-encoded transporter could mediate the transport of α-galactose, so that K12 cells were able to grow on galactose even in the absence of mutarotase. By contrast, only β-galactose could enter K. lactis JA6 cells, because of the lack of the α-galactose transporter. In K. lactis, the transport of galactose is mediated solely by the function of the β-galactoside transporter encoded by KlLAC12 in some strains including the present strain JA6, because lac12 mutants derived thereof are unable to grow either on lactose or on galactose (Riley, 1987). In another strain like 2359, galactose transport is mediated by Lac12p, a low-affinity transporter, as well as by Hgt1p, a high-affinity transporter (Baruffini, 2006). On the other hand, galactose is transported into the cell solely by the function of Gal2p in S. cerevisiae (Tschoppe, 1986). Based on an the assumption that both GAL2 and galP encode the α-galactose transporter, expression of Gal2p in the mutarotase-deficient JA6 cells could lead to successful growth on galactose. This assumption was in conformity with the recent finding that partially purified E. coli galP protein prefers the α-anomer of 13C-glucose to its β-anomer in in vitro binding judged by the solid-state nuclear magnetic resonance analysis (Patching, 2008). To test this hypothesis, we introduced an additional plasmid that expresses ScGal2p under the control of Gal4p into JA6gal10Δ bearing KlGAL10 or Klgal10mutΔ. The plasmid consists of one of the K. lactis centromeres, an autonomously replicating sequence, and URA3 as a selection marker in the backbone of an E. coli plasmid (Betina, 2001). In accordance with our hypothesis, the Klgal10mutΔ-carrying transformant was able to grow if ScGal2p was expressed concomitantly (Fig. 4). As the control, glucose was present in place of galactose in the medium, on which strains bearing either KlGAL10 or Klgal10mutΔ grew normally. In addition, we also tested KlHGT1, which is present only in some strains of K. lactis, but not in JA6 (Baruffini, 2006), for the ability to suppress the growth failure of mutarotase-deficient cells on galactose; the gene was previously identified as encoding a high-affinity glucose transporter (Billard, 1996), but has recently been demonstrated to encode the galactose transporter (Baruffini, 2006). Keep in mind that ScGal2p is also a high-affinity glucose transporter (Boles & Hollenberg, 1997), and that KlHgt1p allows a galactose transport-deficient K. lactis strain to grow on galactose to the same extent as ScGal2p (Baruffini, 2006). As can be clearly seen in the same figure, introduction of KlHGT1 did not restore the growth failure of the mutarotase-deficient yeast on galactose. Alternatively, one might argue that the difference observed between ScGal2p and KlHgt1p was due to the difference in the capacity of galactose uptake of the respective transporter. Such a possibility seems unlikely from the clear difference seen in Fig. 4 and also from the profile of C14-galactose uptake by a strain of K. lactis expressing ScGal2p and KlHgt1p (fig. 2b of Baruffini, 2006), which demonstrates that the efficiency of galactose transport by these transporters is similar to each other. To eliminate such a possibility, the kinetics of galactose uptake was studied, which indicated that the kinetic parameters of Km and Vmax are approximately the same between these transporters (see Fig. S4).

Figure 4

Suppression of growth failure of mutarotase-deficient Kluyveromyces lactis cells by ScGal2p. Cells of JA6gal10Δ carrying a set of the indicated plasmids were grown to the late-log phase (1–2 × 107 cells mL−1) in tryptophan- and uracil-dropout synthetic minimal medium containing 0.5% each of glycerol and sodium lactate. The cells were washed once with an equal volume of water, and resuspended in one-tenth the volume of the same medium without a carbon source. One-loop full of the indicated cells was streaked on the synthetic minimal medium containing 1% each of galactose or glucose. The plates were incubated at 30°C for 72 h. The left and the right halves of each plate were streaked with cells expressing the wild-type Gal10p (GAL10) and those expressing mutarotase-domain-deleted Gal10p (gal10mutΔ), respectively. Each streak contained the cells expressing the additional plasmid as indicated outside the plate. Vector represents KCp491, a vacant vector.

Probable existence of another mutarotase gene in K. lactis

During the course of the present work, we found that JA6gal10Δ cells exhibit severe galactose sensitivity. The growth of these cells on nonfermentable carbon sources was completely arrested on addition of galactose to 0.1% (data not shown). The galactose sensitivity was suppressed if any epimerase-expressing plasmid was introduced into JA6gal10Δ cells (Fig. 5). It is known that deficiency of epimerase causes severe galactose sensitivity in many organisms including bacteria (Fukasawa & Nikaido, 1959a, b), yeast (Douglas & Hawthorne, 1964; Wasilenko & Fridovich-Keil, 2006), and humans (Holton, 1981). Epimerase-deficient mutants of gram-negative bacteria, such as E. coli or Salmonella, were killed due to cell lysis due to the presence of galactose in the medium (Fukasawa & Nikaido, 1959a). The galactose sensitivity of epimeraseless mutants is suppressed by an additional defect in galactokinase or any mutations that shut off other galactose-inducible genes, indicating that galactose must be metabolized to generate UDP-galactose in order for the galactose sensitivity to be seen (Fukasawa & Nikaido, 1959b; Nogi, 1977; Mumma, 2008). Thus, as can be seen in Fig. 5, galactose-resistant colonies outgrew epimeraseless colonies during the prolonged incubation due to second mutations, which would include any mutations that prohibit formation of UDP-galactose in the cell. The observed galactose sensitivity of gal10-deletion mutants of JA6 strongly suggests the existence of another mutarotase, which escaped our polarimetric assay due to the low sensitivity. In fact, we detected a low activity of mutarotase activity in a concentrated crude extract from JA6gal10Δgal80Δ cells as well as in its precipitates with ammonium sulfate by the enzymatic assay (Table 3 and Fig. S5). Trials to purify the mutarotase in question have so far been unsuccessful, not only due to the low abundance in the starting materials but also due to the instability of the enzyme. We conclude, therefore, that the level of the second mutarotase is not high enough to support the growth of the gal10 deletion on galactose, but to cause the galactose sensitivity.

Figure 5

Galactose sensitivity of Kluyveromyces lactis strains expressing Gal10p from various yeasts. Cells of strain JA6gal10Δgal80Δ carrying various plasmids were grown to 4–5 × 108 mL−1 in tryptophan-dropout synthetic medium containing 1% each of sodium lactate and glycerol. The cells were streaked on a pair of agar plates of the same medium with or without galactose (0.1%). The plasmids were expressing Schizosaccharomyces pombe Gal10p (SpGAL10), Saccharomyces cerevisiae Gal10p (ScGAL10), K. lactis Gal10p (KlGAL10), and mutarotase-domain deletion of K. lactis Gal10p (Klgal10mutΔ), respectively. None represents a tryptophan-nonrequiring revertant from JA6gal10Δgal80Δ without a plasmid. Note that galactose-resistant colonies are overgrowing on the streak.

View this table:
Table 3

Ammonium sulfate fractionation of aldose-1-epimerase (mutarotase) from crude extracts of JA6gal10Δgal80Δ

FractionTotal protein (mg)Total activity (U)Specific activity (U mg−1)Recovery (%)
Crude extract71010 80015100
0–55% precipitates47558011952
55–80% precipitates252108.42
  • One unit of enzyme was defined as the activity of 1.0 μmol of α-glucose converted to β-glucose per hour. The assay was carried out using β-glucose dehydrogenase (Sigma/Aldrich Co.) according to the method described by Bouffard (1994). The experiment on which these values are based is shown in Fig. S5.


It has long been known that there exist disaccharide transporters capable of distinguishing the α/β linkage in the sugar, for example the β-galactoside transporter (permease) for lactose, that is, β-d-galactopyranosyl β-d-glucopyranose, or the α-galactoside transporter for melibiose, that is, α-d-galactopyranosyl α-d-glucopyranose (Kennedy, 1970). For monosaccharide transporters, however, little attention has been paid to the recognition specificity of the α/β-anomer of substrates. The present results strongly suggest that Gal2p of S. cerevisiae or Hgt1p of K. lactis are specific for α-galactose or β-galactose, respectively. Kluyveromyces lactis possesses the ability to use lactose, for which the necessary enzymes are under the control of the Gal4p/Gal80p system as well (see Schaffrath & Breunig, 2000). By contrast, some strains of S. cerevisiae are able to use melibiose by the function of the α-galactoside transporter and α-galactosidase encoded by MEL1, which is also controlled by the combination of Gal4p/Gal80 (see Bhat & Murthy, 2001). Sugar-metabolizing systems would have evolved depending on the available sugars in a given ecological niche at an evolutionary stage. Original habitats of S. cerevisiae are bark and sap exudates of oak trees, which are rich in saccharose and less in raffinose (Phaff, 1986). By contrast, many laboratory strains of K. lactis were originally isolated from milk-derived products, in which lactose is the major carbon source (see Schaffrath & Breunig, 2000). A question then arises: is the β-galactoside specificity of KlHgt1p and KlLac12p, or the α-galactoside specificity of ScGal2p and the melibiose transporter merely a coincidence? Structural as well as evolutionary analyses of these transporters may lead to the answer in the future.

We have suggested that there exists a mutarotase other than Gal10p in K. lactis, whose physiological role remains unknown. We speculate that the mutarotase in question originally uses an as yet unknown compound as the substrate, but exerts its function on galactose as well, because of fuzziness in the substrate specificity. Majumdar (2004) and Brahma & Bhattacharyya (2004) have also identified constitutively synthesized mutarotase in strains of S. cerevisiae and K. fragilis, respectively. There exist two ORFs similar to the mutarotase domain of GAL10 in S. cerevisiae genome, namely YHR210c and YNR071C (Goffeau, 1996), whose functions are still unknown. Another mutarotase-like ORF, YMR099c, is known in S. cerevisiae to code for hexose-6-phosphate mutarotase, which does not rotate hexoses at all, however (Graille, 2006). A highly conserved homolog of YMR099c is identified in K. lactis, which is referred to as KLLA0C14300p without biochemical information (http://cbi.labri.fr/Genolevures/). It remains to be determined whether or not this ORF is responsible for the observed activity of mutarotase in JAgal10D.

Supporting Information

Additional Supporting Information may be found in the online version of this article:

Fig. S1. Construction of Gal4pÔÇÉdependent expression plasmid for Gal10p in Kluyveromyces lactis. KlÔÇÉars stands for automatically replicating sequence of K. lactis and is derived from plasmid KARS101 (Irene et al., 2004). TRP1 is derived from an S. cerevisiae plasmid pJJ246 (Jones & Prakash, 1990). The promoter region including UASG and GAL1p (upstream activating sequence and the transcription initiation site, respectively) is derived from p4XX (Mumberg et al., 1994). ADH1t is derived from pVT102U (Vernet et al., 1987), which encompasses the transcription termination site. Shaded region at the NÔÇÉterminus of GAL10 represents histidineÔÇÉtag containing the target peptide of protease Factor Xa (in red), which comes from the cloning/expression region of pET16b (Novagen Co.). The amino acid sequence is shown in the uppermost area.

Fig. S2. Restriction map of Kluyveromyces lactis GAL10 and its deletions. The restriction sites of GAL10 are deduced from the published sequence in the database ( http://cbi.labri.fr/Genolevures/). The introduction of deletion covering both epimerase and mutarotase domains were carried out by three steps: 1) The GAL10 gene, PCRÔÇÉcloned from K. lactis KA5ÔÇÉ6C (MATa ade his leu) DNA, was inserted to pBluescript SK(+) between XhoI and HindIII (pBlueÔÇÉKlGAL10). 2) The hisGÔÇÉURA3ÔÇÉhisG region excised from pNK51 (Alani et al., 1987) with BglII and BamHI was inserted in pBluescript SK(+) at the BamHI site (pBluePOÔÇÉURA3). 3) The BglIIÔÇÉSalI region of pBlueÔÇÉKlGAL10 was replaced with the XhoI and BamHI fragment from pBluePOÔÇÉURA3. 4) The XhoIÔÇÉPvuII fragment of the resultant plasmid, pBlueKlÔÇÉgal10+ö::POÔÇÉURA3, was used for transformation of JA6 (MATa, ade1ÔÇÉ600 trp1ÔÇÉ11 ura3ÔÇÉ12) to select Ura+ colonies. Successful disruption of GAL10 was confirmed by galactoseÔÇÉnegative phenotype of transformants, from which 5ÔÇÉfluoroorotic acidÔÇÉresistant clones were selected to create JA6gal10+ö. Introduction of deletion to the mutarotase domain was carried out by removing the PstIÔÇÉHpaI fragment from GAL10; precisely, the cloned GAL10 was cleaved first with PstI, endÔÇÉblunted with T4 DNA polymerase, and with HpaI. The resultant ends in GAL10 were ligated to yield pK411.

Fig. S3. Linearity of polarometric assay of mutarotase. a) The enzyme sample used was the same crude extracts described in Table 1 in the text, which was prepared from Kluyveromyces lactis strain JA6gal80+ö (gal80D) and JA6gal80+ögal10+ö (gal80Dgal10D) grown on glucose. b) The rate constant was calculated from the slope of log (+-0ÔÇÉ+-e)/(+-tÔÇÉ+-e) for the crude extract of JA6gal80+ö in a), where +-0, +-e and +-t are rotation at 0ÔÇÉtime, equilibrium and time t, respectively. mcL stands for microliter.

Fig. S4. LineweaverÔÇôBurk plot showing galactose uptake in the strain 2359lac12hgt1+ö (Baruffini et al., 2006) transformed with either KCp491ÔÇÉHGT1 (black line and squares) or KCp491ÔÇÉGAL2 (grey line and squares). or KCp491ÔÇÉGAL2 (grey line and squares). Galactose uptake was measured as previously reported (Baruffini et al., 2006), except that uptake activity was measured after 10 s of incubation with [1ÔÇÉ14C]ÔÇÉDÔÇÉgalactose at final concentrations ranging from 0.5 to 10 mM. For each concentration, the value comes from the subtraction of the uptake activity of the involved strain with the background uptake activity of strain 2359lac12+öhgt1+ö transformed with KCp491. Values are means of three independent uptake experiments.

Fig. S5. GAL10ÔÇÉindependent mutarotase activity in concentrated crude extracts and ammonium sulfate precipitates from JAgal10gal80+ö. The assay was carried out according to Bouffard et al. (1994). Reaction was started by the addition of +-ÔÇÉglucose solution, which is prepared immediately before each assay. The absorbance of the no enzyme control was subtracted automatically by setting a reference cuvette without sample in a dual beam spectrophotometer (Hitchi). The protein concentration of crude extracts, ammonium sulfate precipitates at 0% to 55% (55%ppt), and 55% to 80% (80% ppt) was 7.1 mg/ml, 4.7 mg/ml, and 2.5 mg/ml, respectively. The amount of enzyme used is 10 l each of crude extract, 10ÔÇÉtimes diluted 55% ppt, and 10ÔÇÉtimes diluted 80% ppt.

Table S1. Molecular size and characteristics of uridine diphosphate glucose 4ÔÇÉepimerse purified from various sources

Please note: Wiley-Blackwell is not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.


Many of the present experiments were performed by T.F. at Dr M. Kasahara's laboratory of Teikyo University and also at Dr A. Abe's laboratory of Kitasato University, to whom we are grateful. We thank Dr T. Lodi of Parma University for encouraging collaboration between Italian and Japanese scientists. We also thank Dr N. Gunge, Dr K. Breunig, Dr L. Fabiani, Dr M. Yamamoto, and Dr R. Dickson for the yeast strains, plasmids, and information on handling K. lactis.


  • Editor: André Goffeau


View Abstract