APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Apr. 1995, p. 1580–1585 Vol. 61, No.

4
0099-2240/95/$04.0010
Copyright q 1995, American Society for Microbiology

Purification and Partial Characterization of an Aldo-Keto

Reductase from Saccharomyces cerevisiae
ANNERITA KUHN, CARINA VAN ZYL,† ANDRÉ VAN TONDER, AND BERNARD A. PRIOR*
Department of Microbiology and Biochemistry, University of the
Orange Free State, Bloemfontein 9300, South Africa
Received 11 October 1994/Accepted 6 February 1995

A cytosolic aldo-keto reductase was purified from Saccharomyces cerevisiae ATCC 26602 to homogeneity by
affinity chromatography, chromatofocusing, and hydroxylapatite chromatography. The relative molecular
weights of the aldo-keto reductase as determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis
and size exclusion chromatography were 36,800 and 35,000, respectively, indicating that the enzyme is mono-
meric. Amino acid composition and N-terminal sequence analysis revealed that the enzyme is closely related
to the aldose reductases of xylose-fermenting yeasts and mammalian tissues. The enzyme was apparently
immunologically unrelated to the aldose reductases of other xylose-fermenting yeasts. The aldo-keto reductase
is NADPH specific and catalyzes the reduction of a variety of aldehydes. The best substrate for the enzyme is
the aromatic aldehyde p-nitrobenzaldehyde (Km 5 46 mM; kcat/Km 5 52,100 s21 M21), whereas among the
aldoses, DL-glyceraldehyde was the preferred substrate (Km 5 1.44 mM; kcat/Km 5 1,790 s21 M21). The enzyme
failed to catalyze the reduction of menadione and p-benzoquinone, substrates for carbonyl reductase. The
enzyme was inhibited only slightly by 2 mM sodium valproate and was activated by pyridoxal 5*-phosphate. The
optimum pH of the enzyme is 5. These data indicate that the S. cerevisiae aldo-keto reductase is a monomeric
NADPH-specific reductase with strong similarities to the aldose reductases.

The aldo-keto reductases comprise a family of monomeric, in the liver and the brain (20, 56). Aldose reductase catalyzes
NADPH-dependent oxidoreductases with similar physical and the first reaction of the polyol pathway in which aldoses are
chemical properties which catalyze the reduction of alde- converted by an NADPH-dependent reaction to the corre-
hydes and ketones to the corresponding alcohol products. sponding polyalcohols (56). Aldose reductase appears to be
The catalyzed reaction favors alcohol formation, whereas the strongly involved in the etiology of several human pathologies
reverse reaction occurs only to a limited extent (56). These associated with hyperglycemia (32). Aldose reductase also
reductases have a preferential, but not mutually exclusive, occurs in xylose-fermenting yeasts such as Pichia stipitis, Can-
substrate specificity, and this overlapping specificity, together dida shehatae, and Pachysolen tannophilus (15, 22). It cata-
with indeterminate physiological roles, has precluded efforts lyzes the first step in the xylose metabolic pathway and has
to derive a more specific nomenclature for these enzymes
been studied because the production of ethanol from xylose
(20).
is of potential economic value (11). In spite of the fact that
Three distinct enzymes which exhibit the typical properties
of aldo-keto reductase are known: (i) aldehyde reductase (EC the yeast Saccharomyces cerevisiae is unable to grow on xy-
1.1.1.2), which catalyzes the reduction of a variety of alde- lose (3), a number of investigators have found that this
hydes, notably uronic acids and some ketones; (ii) aldose re- yeast can indeed metabolize xylose when another sugar is
ductase (EC 1.1.1.21), which catalyzes the reduction of alde- present (5, 41, 53, 54). However, the rate of xylose utiliza-
hydes, especially glycolaldehydes and polyolaldehydes, but is tion is slow, with only 74% of the available xylose being me-
less active with uronic acids; and (iii) carbonyl reductase (EC tabolized within 7 days (54). Both glucose- and ethanol-grown
1.1.1.184), which catalyzes the reduction of quinones, other cultures were found to possess similar levels of xylose re-
ketones, and aldehydes to their corresponding alcohols (56). ductase activity (ca. 5 mU/mg of protein), indicating that
The aldo-keto reductases consist of a single polypeptide chain the enzyme is constitutive and non-glucose repressible (53).
with a relative molecular mass of between 30,000 and 40,000 Furthermore, many S. cerevisiae strains grow on xylulose
for aldehyde and aldose reductases and approximately 30,000 (21), thereby indicating that the reasons for the failure of S.
for carbonyl reductases (56). They are found in a wide variety cerevisiae to grow on xylose may be found in the initial steps
of mammals, birds, reptiles, amphibia, fish, insects, and fungi from xylose to xylulose. The enzymes catalyzing these ini-
(18). tial steps in S. cerevisiae have not yet been characterized, and
Carbonyl as well as aldehyde reductases have been proposed an underst anding of the properties of these enzymes could
to be involved in the detoxification of reactive compounds assist in the development of a xylose-fermenting S. cerevisiae
strain.
The purpose of this investigation was to purify the enzyme
* Corresponding author. Mailing address: Department of Microbi- catalyzing the conversion of xylose to xylitol in S. cerevisiae
ology and Biochemistry, University of the Orange Free State, P.O. Box
ATCC 26602 to homogeneity and to characterize its physical,
339, Bloemfontein 9300, South Africa. Phone: 0-27-51-4012396. Fax:
0-27-51-482004. Electronic mail address: PRIOR@WWG3.UOVS. immunological, and kinetic properties. The results indicate
AC.ZA. that this enzyme is an aldo-keto reductase with properties
† Present address: Research and Development Department, AECI, similar to those of a wide range of aldose reductases found in
Modderfontein 1645, South Africa. other microorganisms and animals.

1580
VOL. 61, 1995 ALDO-KETO REDUCTASE FROM S. CEREVISIAE 1581

MATERIALS AND METHODS TABLE 1. Purification of aldo-keto reductase from

S. cerevisiae ATCC 26602
Microorganism and growth conditions. Yeast cultures were maintained on
yeast peptone dextrose (YPD) agar slants (1% [wt/vol] yeast extract, 2% [wt/vol] Amt of
Total Specific Purifi-
peptone, 2% [wt/vol] glucose, 1.5% [wt/vol] agar). S. cerevisiae ATCC 26602 was total Yield
Step activity activity cation
cultured in 9 liters of the above-described medium without agar in a fermentor protein (%)
(U) (U/mg) (fold)
(VirTis, Gardiner, N.Y.) at 308C and an aeration rate of 7.5 liters/min. After 24 (mg)
h of growth, the cells were harvested and washed by centrifugation for 5 min at
5,000 3 g and 48C. Supernatant 1,274 4.15 0.0032 100 1
P. tannophilus NRRL Y-2460, C. shehatae CSIR Y492, P. stipitis CSIR Y633, Red Sepharose 12.6 5.39 0.428 130 134
and Candida utilis CSIR Y12 were grown in 400 ml of medium containing 1% Chromatofocusing 3.05 74
(wt/vol) yeast extract, 2% (wt/vol) peptone, and 2% (wt/vol) xylose in 3-liter Hydroxylapatite 0.8 3.11 3.90 75 1,218
flasks at 308C. S. cerevisiae strains CSIR Y191, 8282 ADE, ATCC 4126, and
NRRL Y132 were grown in 400 ml of YPD medium at 308C. All cells were
harvested in the late exponential phase.
Enzyme purification. Lyophilized cells were disrupted either by using a mortar
were then electrophoretically transferred onto a Hybond C nitrocellulose filter.
and pestle or by adding glass beads and disrupting the cells in a cell homogenizer
The blotted proteins were identified immunochemically by the sequential addi-
(MSK; Braun Melsungen AG, Melsungen, Germany). The extraction buffer
tion of anti-aldo-keto reductase serum followed by goat anti-rabbit immunoglob-
consisted of 100 to 150 ml of potassium phosphate (pH 7.4) containing freshly
ulin G antibodies conjugated to alkaline phosphatase, nitroblue tetrazolium salt,
prepared 0.5 mM phenylmethylsulfonyl fluoride (Boehringer Mannheim), 1 mM
and 5-bromo-4-chloro-3-indolyl-phosphate (Merck Darmstadt). Antiserum
pepstatin A (Merck Darmstadt), and 50 mM leupeptin (Merck Darmstadt) to
raised against purified aldose reductase from P. tannophilus (10) was kindly
inhibit protease activity. The homogenate was centrifuged for 30 min at 48,400 3
provided by P. L. Bolen.
g and 48C.
The purification was carried out according to a method described by Vander
Jagt et al. (52) with the exception that fractions containing aldo-keto reductase
activity were pooled and desalted by ultrafiltration with a 50-ml stirred Amicon RESULTS AND DISCUSSION
(Danvers, Mass.) ultrafiltration cell fitted with a YM-10 membrane instead of by
gel filtration. All steps were performed at 0 to 48C. The protocol for purification of aldo-keto reductase from S.
Protein determination. Protein concentrations were estimated either by mea-
suring the A280 or by the bicinchoninic acid method (45) with the Pierce bicin-
cerevisiae, summarized in Table 1, provided a 1,218-fold puri-
choninic acid protein assay kit (Rockford, Ill.). fication of enzyme with a 75% yield to a final specific activity
Enzyme assay. Aldo-keto reductase activity was determined routinely at 308C of 3.90 U/mg of protein with D-xylose as the substrate. This
by monitoring the decrease in A340 in an assay mixture containing 50 mM compared favorably with the purification of aldose reduc-
potassium phosphate buffer (pH 7.4), 115 mM xylose, 0.12 mM NADPH,
and 10 mM 2-mercaptoethanol (44). The determination of the effects of pH
tase from human tissues by a similar procedure (52). The
and sodium valproate (Merck Darmstadt) on enzyme activity was carried out purified enzyme was found to be homogeneous by SDS-PAGE
in a similar manner except that 10 mM DL-glyceraldehyde was used as a sub- (Fig. 1).
strate. For assays at different pH values, 50 mM sodium citrate (pH 3 to 5.4) Analysis of the enzyme by gel electrophoresis in the pres-
and 50 mM potassium phosphate (pH 6 to 8) buffers were used. The effect of
sodium valproate on enzyme activity was determined in 50 mM sodium citrate
ence of SDS (Fig. 1) revealed one band with an Mr of 36,800 6
buffer (pH 5), whereas the effect of pyridoxal 59-phosphate (Merck Darmstadt) 1,000 (n 5 6). Size exclusion chromatography on Toyopearl
(0.25 or 0.5 mM) on aldo-keto reductase activity was determined for 60 min at HW-55F resulted in the elution of the enzyme activity as a
room temperature in the dark as described by Morjana et al. (39). The appro- symmetrical peak corresponding to an Mr of approximately
priate blanks to correct for nonspecific oxidation of NADPH were prepared for
each assay. The Michaelis kinetic parameters, Km and kcat, were determined by
35,000. This result, taken together with that of SDS-PAGE,
fitting data directly to the rate equation with SAS software (SAS Institute Inc.). indicates that the S. cerevisiae aldo-keto reductase is a mono-
One unit of enzyme activity was defined as the amount of enzyme catalyzing the meric enzyme. Xylose reductases have been reported to be
oxidation of 1 mmol of NADPH per min under the above-described assay con- either monomeric (24) or dimeric (10, 42, 55) in a number of
ditions.
PAGE. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-
yeasts. The evaluated Mr for this enzyme is, however, in the
PAGE) was performed on a 10 to 20% gradient gel by the method of Laemmli same range (30,000 to 40,000) as those reported for several
(35). Protein bands were visualized with Coomassie brilliant blue G (Sigma other monomeric aldo-keto reductases (56).
Chemical Co., St. Louis, Mo.) or by silver staining (46). Standard proteins The amino acid composition of aldo-keto reductase from S.
(Boehringer Mannheim) used for estimation of molecular mass were a2-macro-
globulin (Mr 5 170,000), b-galactosidase (Mr 5 116,400), fructose-6-phosphate
cerevisiae (data not shown) was found to contain a high proline
kinase (Mr 5 85,200), glutamate dehydrogenase (Mr 5 55,600), aldolase (Mr 5 content (6%), typical for aldo-keto reductases (14, 38). The
39,200), triose phosphate isomerase (Mr 5 26,600), trypsin inhibitor (Mr 5 results of compositional relatedness analysis (to determine the
20,100), and lysozyme (Mr 5 14,300). difference index [DI], where 0 represents two identical proteins
Molecular mass determination by size exclusion chromatography. The mo-
lecular mass of the aldo-keto reductase was also determined under nonreducing
and 100 represents two proteins with no identical amino acids
conditions by size exclusion chromatography (1) using a Toyopearl HW-55F [37]) of aldo-keto reductase from S. cerevisiae and aldo-keto
(Tosohaas, Stuttgart, Germany) column (645 by 8 mm) attached to a fast protein reductases from various other sources indicated that the aldo-
liquid chromatography system. A 50 mM phosphate buffer (pH 7.5) containing keto reductase from S. cerevisiae is most closely related to the
0.3 M NaCl and 5 mM 2-mercaptoethanol was used as the running buffer at a
flow rate of 0.25 ml/min. The column was calibrated with the following standard
aldose reductase from human kidney (DI 5 7.38 [2]). It is also
proteins (Boehringer Mannheim): ferritin (Mr 5 450,000), aldolase (Mr 5 closely related to the xylose reductase from P. stipitis (DI 5
240,000), catalase (Mr 5 158,000), hen egg albumin (Mr 5 45,000), chymot- 8.00 [24]) and the aldose reductases in rabbit muscle (DI 5
rypsinogen A (Mr 5 25,000), and cytochrome c (Mr 5 12,500). 8.16 [17]), C. shehatae (DI 5 8.33 [26]), P. tannophilus (DI 5
Amino acid analysis. Approximately 500 mg of enzyme was precipitated with
30% (wt/vol) trichloroacetic acid, and the precipitate was washed with ice-cold
8.42 [10]) and human muscle (DI 5 9.55 [38]). DI analysis
acid acetone (0.1% [vol/vol] concentrated HCl in acetone). The amino acid indicated that the aldo-keto reductase of S. cerevisiae was less
composition of the dried acetone powder was analyzed by ion-exchange chro- closely related to aldehyde (DIs ranging from 10.33 to 21.11 [2,
matography (7) on an amino acid high-performance liquid chromatography 30, 49]) and carbonyl (DI . 16 [27]) reductases.
system (Waters, Milford, Mass.). The compositional relatedness between aldo-
keto reductase from S. cerevisiae and aldo-keto reductases from various other
Homology of the first 13 residues of the N-terminal se-
sources was assessed by the method of Metzger et al. (37). quence apparently exists among aldose reductases from vari-
N-terminal sequence analysis. The N-terminal sequence of the purified pro- ous sources (Fig. 2). Each N-terminal sequence of aldose re-
tein was determined by analyzing the dried acetone powder by a procedure ductase from various sources contains a highly conserved
described by Hewick et al. (25) and modified by Brandt et al. (13).
Immunochemical methods. Antiserum to purified aldo-keto reductase was
LNNG region (positions 6 to 9, according to S. cerevisiae aldo-
raised in Californian rabbits (29), and immunoblotting was performed as de- keto reductase numbering). The asparagine residue at position
scribed by Towbin et al. (50). Briefly, proteins were resolved by SDS-PAGE and 8 (S. cerevisiae aldo-keto reductase numbering) in this motif is
1582 KUHN ET AL. APPL. ENVIRON. MICROBIOL.

FIG. 2. Alignment of N-terminal amino acid sequences of aldo-keto reduc-

tases. a, S. cerevisiae aldo-keto reductase (this paper); b, human placenta aldose
reductase (9); c, C. shehatae aldose reductase (26); d, P. tannophilus aldose
reductase (10); e, rat lens aldose reductase (16); f, human liver aldehyde reduc-
tase (8); g, S. salmonicolor aldehyde reductase (30); h, P. stipitis xylose reductase
FIG. 1. (A) SDS-PAGE (10% acrylamide) shows a homogeneous enzyme (24, 47); (i) GCY (yeast gene product) (40); (j) human placenta carbonyl reduc-
preparation after the final purification step, namely hydroxylapatite (lanes 6 and tase (57). Boldface type is used to indicate identical residues. The numbering
7). Lanes 1 and 2, calibration proteins; lane 3, proteins which were present after used is that of S. cerevisiae aldo-keto reductase.
the red Sepharose purification step; lane 4, proteins present after the desalting
step (more protein was loaded in lane 4 than in lane 3, resulting in more and
thicker bands); lane 5, proteins present after the chromatofocusing step. The
protein bands were visualized by silver staining (46). (B) Western blotting (im- cultures was observed (data not shown), in spite of the fact that
munoblotting) of aldo-keto reductase illustrates the specificity of the antiserum
which was raised against the purified aldo-keto reductase. Lane 1, crude extract aldo-keto reductase specific activity was found in P. stipitis, C.
of S. cerevisiae; lane 2, purified aldo-keto reductase. shehatae, and P. tannophilus (between 20 and 165 mU/mg) and
C. utilis (2.1 mU/mg). Similarly, antiserum raised against P.
tannophilus NRRL Y-2460 failed to react with a crude extract
of S. cerevisiae ATCC 26602 on immunoblots (data not shown),
replaced by a threonine in the aldehyde reductases from hu- thereby indicating that the aldo-keto reductase of S. cerevisiae
man liver and Sporobolomyces salmonicolor. The proline resi- is immunologically unrelated to that found in xylose-ferment-
due in the 13th position is also well conserved, and lysine and ing yeasts. Poor interspecies cross-reactivity of anti-aldehyde
methionine residues in the 11th and 12th positions, respec- reductase antibody has, however, been reported (56). This
tively, are well conserved in some reductases. The conserved feature might also be characteristic of anti-aldose reductase
glycine in position 9 and the proline in position 13 are thought antibodies.
to compose part of the consensus sequence for the ‘‘Rossmann Optimal reduction of DL-glyceraldehyde by purified S. cer-
fold’’ involved in nucleotide binding (16). evisiae aldo-keto reductase was observed at pH 5 (Fig. 3). At
A gene encoding a protein (GCY) homologous to mamma- pH 3, 87% of the activity remained, whereas at pH 8 the
lian aldo-keto reductases has been isolated from S. cerevisiae enzyme activity was only 10% of the value at pH 5. The pref-
(36, 40). Its N-terminal sequence and its kinetic properties, erence for a slightly acidic pH optimum is typical of aldose
however, indicate that this enzyme is probably a carbonyl reductases (16), although the optimum pH of this aldo-keto
reductase not related to the aldo-keto reductase described reductase is apparently 1 to 2 pH units lower than has previ-
here. ously been observed for other yeast (10, 12, 33, 55) and some
Analysis of purified S. cerevisiae aldo-keto reductase on im- mammalian (38, 49) aldose reductases. The optimum pH re-
munoblots developed with anti-S. cerevisiae aldo-keto reduc- ported here for the S. cerevisiae enzyme is, however, similar to
tase revealed a single protein (Fig. 1B). No bands were de- the values of 4.8 and 5.6 reported for the pig and rabbit lens
tected when preimmune serum was used (data not shown). aldose reductases, respectively (14, 48).
The cross-reactivity of the antiserum with crude extracts of The broad substrate specificity of purified S. cerevisiae aldo-
S. cerevisiae CSIR Y 191, S. cerevisiae 8282 ADE, S. cerevisiae keto reductase is apparent from Table 2. The aldo-keto reduc-
ATCC 4126, and S. cerevisiae NRRL Y 132 grown in YPD tase discriminates among substrates almost entirely in the
broth was examined (data not shown). Aldo-keto reductase binding step, since kcat values are essentially independent of
activity in these strains was found to be insignificant (#0.9 mU/ the substrate. The high affinity of the enzyme for p-nitroben-
mg). Weak bands indicated the presence of an aldo-keto re- zaldehyde is typical for all aldo-keto reductases (51). In gen-
ductase in these S. cerevisiae strains. However, a number of eral, the enzyme preferred aromatic or aliphatic aldehydes to
other bands were also observed, indicating nonspecific cross- aldose sugars. Its substrate specificity, as shown by increasing
reactivity possibly as a result of the use of polyclonal anti- Km values, was as follows: p-nitrobenzaldehyde, DL-glyceralde-
serum. hyde, D-glyceraldehyde, L-glyceraldehyde, D-glucose, D-xylose,
When S. cerevisiae ATCC 26602 was grown in YPD broth and L-arabinose. This order of substrate affinity is in agreement
containing 3% NaCl to test whether aldo-keto reductase activ- with observations reported for aldose reductases from various
ity could be induced by NaCl as was found for renal medullary mammalian sources (2, 6, 14) and the yeasts P. stipitis (55) and
cells (6), no increase in the specific activity of the aldo-keto P. tannophilus (19). On the one hand, the S. cerevisiae aldo-
reductase relative to the activity found in S. cerevisiae which keto reductase had a lower affinity for the aromatic and ali-
was grown in YPD broth without NaCl (3 to 8 mU/mg) was phatic aldehydes compared with the values observed for mam-
found. This finding was confirmed by immunoblotting, which malian aldose reductases, whereas on the other hand, the yeast
showed no marked increase in the intensity of the aldo-keto aldo-keto reductase had a greater affinity for glucose. Argu-
reductase band (data not shown) and which indicates that ments that the open-chain aldehyde form of glucose is the true
aldo-keto reductase activity is not osmotically regulated in S. physiological substrate have been made, and a Km of 0.66 mM
cerevisiae. has been calculated for aldose reductase when reacted with
No cross-reactivity between antiserum raised against S. cer- this form of glucose (28). Accounting for the aldehyde form of
evisiae aldo-keto reductase and crude extracts of xylose-grown glucose and galactose leads to very low corrected Km values
VOL. 61, 1995 ALDO-KETO REDUCTASE FROM S. CEREVISIAE 1583

was, however, observed with 2 mM sodium valproate. No ac-

tivity was observed with the quinones menadione and p-ben-
zoquinone, which are good substrates for carbonyl reductase
(51).
The Km value of S. cerevisiae aldo-keto reductase with
NADPH as a cofactor was similar to values reported for mam-
malian aldose reductases (2, 22) and aldose reductases from
the yeasts P. tannophilus (19) and P. stipitis (55). No activity,
however, was detected with NADH as a cofactor, which is
somewhat surprising, since aldose reductases often are able to
utilize both cofactors (51), although the aldose reductases gen-
erally have a greater affinity for NADPH than NADH in mam-
malian tissues (58) and in yeasts (42, 55). This finding is not
that unusual, however, in the case of yeasts. Bruinenberg et al.
(15) have reported the presence of NADPH-dependent xylose
reductases exhibiting no activity with NADH in C. utilis and
certain strains of Candida tenuis. Insignificant reverse reaction
rates with 200 mM xylitol (,5% of the forward reaction rates
with 50 mM xylose), typical of an aldo-keto reductase, were
found.
The aldo-keto reductase was activated 40% by 250 mM pyr-
FIG. 3. The effect of pH on the activity of purified aldo-keto reductase of S. idoxal 59-phosphate after 30 min and 35% by 125 mM pyridoxal
cerevisiae ATCC 26602. The activity of the purified enzyme was measured as 59-phosphate after the same amount of time. During this pe-
described in Materials and Methods.
riod, the control activity decreased by 10%. This activation by
pyridoxal 59-phosphate, possibly as a result of modification of
an essential lysine in the coenzyme binding site, is in agree-
and high kcat/Km values in the case of the rat testis aldose ment with previous findings for aldose reductase from human
reductase (31). Indeed, Grimshaw (23) has used aldose re- muscle (39).
ductase to measure directly the rate of ring opening of D- The physiological function of this aldo-keto reductase is
glucose, and his results further show that aldose reductase does uncertain. The diversity of aldehyde structures reduced by the
not itself catalyze the ring-opening reaction. A correlation aldose reductases has indicated that they are involved in the
between the local concentration of glucose in a tissue, the rate detoxification of both endogenous and exogenous aldehydes
of ring opening, and the accumulation of sorbitol has not been (22). In pentose-fermenting yeasts, aldose reductase catalyzes
made, and there is no direct evidence to date that aldose the first step in the metabolism of D-xylose (43). D-Xylose is
reductase in vivo uses the acyclic form of D-glucose as a sub- converted to xylitol by aldose reductase, and the xylitol is
strate. oxidized further via D-xylulose through the pentose phosphate
The catalytic efficiency (kcat/Km) of the S. cerevisiae aldo- and Embden-Meyerhof pathways. The inability of S. cerevisiae
keto reductase was lower than those of the mammalian aldose to grow on xylose (54) indicates that the aldo-keto reductase
reductases on aromatic and aliphatic aldehydes, whereas this activity is unlikely to function in the catabolism of xylose (53).
rate was higher on the aldoses (17, 22, 52). The S. cerevisiae Furthermore, a role in osmoregulation, as found for aldose
aldo-keto reductase failed to exhibit activity on D-glucuronate, reductases in PAP-HT25 cells (rabbit renal medullary cells in
an anionic substrate used to characterize aldehyde reductase tissue culture) (6) and in the barley embryo (4), is unlikely, as
(22). Aldose reductases are generally less active with uronic the aldo-keto reductase levels in S. cerevisiae were not changed
acids than are aldehyde reductases (56). In addition, sodium by osmotic stress.
valproate, which has been shown to inhibit aldehyde reductase The construction of a recombinant S. cerevisiae strain able to
(17, 38, 51), is ineffective against the purified S. cerevisiae grow and metabolize xylose to xylitol (24, 34, 47) has been
enzyme at a concentration of 1 mM. Slight inhibition (16%) attempted by cloning the aldose reductase of P. stipitis. The
presence of an aldo-keto reductase in S. cerevisiae could
have potential application in the conversion of xylose to xylitol
TABLE 2. Michaelis constants and maximal reaction velocities (a potentially important nonnutritive sweetener [24]). How-
for substrates of S. cerevisiae aldo-keto reductasea ever the NADPH specificity of the enzyme would make it
unsuitable for the development of a recombinant S. cerevisiae
kcat (SEM)b Km (SEM) kcat/Km strain able to produce ethanol, as only yeasts with NADH-
Substrate
(s21) (mM) (mM21zs21)
specific xylose reductase activity are able to ferment xylose
p-Nitrobenzaldehyde 2.37 (0.05) 0.046 (0.004) 52.1 (15).
DL-Glyceraldehyde 2.58 (0.13) 1.44 (0.26) 1.79
D-Glyceraldehyde 3.32 (0.18) 1.57 (0.29) 2.12
ACKNOWLEDGMENTS
L-Glyceraldehyde 2.50 (0.14) 6.38 (1.05) 0.39
D-Xylose 3.37 (0.18) 27.90 (4.51) 0.12 This work was supported by the Foundation for Research Develop-
L-Arabinose 1.61 (0.05) 32.63 (2.81) 0.049 ment, South Africa.
D-Glucose 0.71 (0.02) 9.34 (0.79) 0.076 We thank Lodewyk Kock and Stefaans Kilian for bringing to our
NADPH 0.013 (0.002) attention the observation that S. cerevisiae could utilize D-xylose and
a for their helpful suggestions. The N-terminal sequence and amino acid
Enzyme activities were assayed in triplicate with 50 mm potassium phosphate
buffer (pH 5.0) containing 0.12 mM NADPH and 10 mM 2-mercaptoethanol, analyses by Wolf Brandt are gratefully acknowledged. Paul Bolen is
and Km and kcat values were determined by using a direct fit of the Michaelis- thanked for providing antiserum against P. tannophilus. Robert Schall
Menten rate equation. and Alta Stassen are thanked for their assistance with nonlinear re-
b
SEM, standard error of the mean. gression analysis of the kinetic data.
1584 KUHN ET AL. APPL. ENVIRON. MICROBIOL.

REFERENCES 28. Inagaki, K., I. Miwa, and J. Okuda. 1982. Affinity, purification, and glucose
1. Andrews, P. 1965. The gel filtration of proteins related to their molecular specificity of aldose reductase from bovine lens. Arch. Biochem. Biophys.
weights over a wide range. Biochem. J. 96:595–606. 216:337–344.
2. Ansari, N. H., A. Bhatnagar, S. Lui, and S. K. Srivastava. 1991. Purification 29. Johnstone, A., and R. Thorpe. 1987. Immunochemistry in practice, 2nd ed.
and characterization of aldose reductase and aldehyde reductase from hu- Blackwell Scientific Publications, Oxford.
man kidney. Biochem. Int. 25:755–765. 30. Kataoka, M., H. Sakai, T. Morikawa, M. Katoh, T. Miyoshi, S. Shimizu, and
3. Barnett, J. A., R. W. Payne, and D. Yarrow. 1983. Yeasts: characteristics and H. Yamada. 1992. Characterization of aldehyde reductase of Sporobolomyces
identification. Cambridge University Press, Cambridge. salmonicolor. Biochim. Biophys. Acta. 1122:57–62.
4. Bartels, D., K. Engelhardt, R. Roncarati, K. Schneider, M. Rotter, and F. 31. Kawasaki, N., T. Tanimoto, and A. Tanaka. 1989. Characterization of aldose
Salamini. 1991. An ABA and GA modulated gene expressed in the barley reductase and aldehyde reductase from rat testis. Biochim. Biophys. Acta
embryo encodes an aldose reductase related protein. EMBO J. 10:1037– 996:30–36.
1043. 32. Kinoshita, J. H., P. F. Kador, W. G. Robison, M. B. Datilis, L. M. Cobo, and
5. Batt, C. A., S. Carvallo, D. D. Easson, M. Akedo, and A. J. Sinskey. 1986. C. Kupfer. 1984. Aldose reductase and complications of diabetes. Ann.
Direct evidence for a xylose metabolic pathway in Saccharomyces cerevisiae. Intern. Med. 101:82–91.
Biotechnol. Bioeng. 28:549–553. 33. Kise, S., N. Koizumi, and H. Maeda. 1988. Properties of NAD(P)H-linked
6. Bedford, J. J., S. M. Bagnasco, P. F. Kador, H. W. Harris, and M. B. Burg. aldose reductase from Cryptococcus lactativorus. J. Ferment. Technol. 66:
1987. Characterization and purification of a mammalian osmoregulatory 615–623.
protein, aldose reductase, induced in renal medullary cells by high extracel- 34. Kötter, P., and M. Ciriacy. 1993. Xylose fermentation by Saccharomyces
lular NaCl. J. Biol. Chem. 262:14255–14259. cerevisiae. Appl. Microbiol. Biotechnol. 38:776–783.
7. Bidlingmeyr, B. A., S. A. Cohen, and T. L. Tarvic. 1984. Rapid analysis 35. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of
of amino acids using pre-column derivatization. J. Chromatogr. 336:93– the head of bacteriophage T4. Nature (London) 227:680–685.
104. 36. Magdolen, V., U. Oechsner, P. Trommler, and W. Bandlow. 1990. Transcrip-
8. Bohren, K. M., B. Bullock, B. Wermuth, and K. H. Gabbay. 1989. The tional control by galactose of a yeast gene encoding a protein homologous to
aldo-keto reductase superfamily. cDNAs and deduced amino acid sequences mammalian aldo/keto reductases. Gene 90:105–114.
of human aldehyde and aldose reductases. J. Biol. Chem. 264:9547–9551. 37. Metzger, H., M. B. Shapiro, J. E. Mosimann, and J. E. Vinton. 1968. As-
9. Bohren, K. M., J. L. Page, R. Shankar, S. P. Henry, and K. H. Gabbay. 1991. sessment of compositional relatedness between proteins. Nature (London)
Expression of human aldose and aldehyde reductases, site-directed mutagen- 219:1166–1168.
esis of a critical lysine 262. J. Biol. Chem. 266:24031–24037. 38. Morjana, N. A., and T. G. Flynn. 1989. Aldose reductase from human psoas
10. Bolen, P. L., J. A. Bietz, and R. W. Detroy. 1985. Aldose reductase in the muscle. Purification, substrate specificity, immunological characterization,
yeast Pachysolen tannophilus: purification, characterization and N-terminal and effect of drugs and inhibitors. J. Biol. Chem. 264:2906–2911.
sequence. Biotechnol. Bioeng. Symp. 15:129–148. 39. Morjana, N. A., C. Lyons, and T. G. Flynn. 1989. Aldose reductase from
11. Bolen, P. L., and R. W. Detroy. 1985. Induction of NADPH-linked D-xylose human psoas muscle. Affinity labeling of an active site lysine by pyridoxal
reductase and NAD-linked xylitol dehydrogenase activities in Pachysolen 59-phosphate and pyridoxal 59-diphospho-59-adenosine. J. Biol. Chem. 264:
tannophilus by D-xylose, L-arabinose, or D-galactose. Biotechnol. Bioeng. 2912–2919.
27:302–307. 40. Oechsner, U., V. Magdolen, and W. Bandlow. 1988. A nuclear yeast gene
12. Bolen, P. L., and D. A. McCracken. 1990. Properties of aldose reductase (GCY) encodes a polypeptide with high homology to a vertebrate eye lens
from the methanol yeast Candida boidinii. J. Ferment. Bioeng. 69:211–214. protein. FEBS Lett. 238:123–128.
13. Brandt, W. F., H. Alk, M. Chauhan, and C. Von Holt. 1984. A simple 41. Pearson, T., J. L. F. Kock, and S. G. Kilian. 1990. The value of carbon
modification converts the spinning cup sequencer into a vapour phase se- compound mixtures in the identification of the yeast genus Saccharomyces.
quencer. FEBS Lett. 174:228–232. Acta Varia 5:135–146.
14. Branlant, G. 1982. Properties of an aldose reductase from pig lens. Com- 42. Rizzi, M., P. Erlemann, N. A. Bui-Thanh, and H. Dellweg. 1988. Xylose
parative studies of an aldehyde reductase from pig lens. Eur. J. Biochem. fermentation by yeasts. 4. Purification and kinetic studies of xylose reductase
129:99–104. from Pichia stipitis. Appl. Microbiol. Biotechnol. 29:148–154.
15. Bruinenberg, P. M., P. H. M. de Bot, J. P. van Dijken, and W. A. Scheffers. 43. Skoog, K., and B. Hahn-Hägerdal. 1988. Xylose fermentation. Enzyme Mi-
1984. NADH-linked aldose reductase: the key to anaerobic alcoholic crob. Technol. 10:66–80.
fermentation of xylose by yeasts. Appl. Microbiol. Biotechnol. 19:256– 44. Smiley, K. L., and P. L. Bolen. 1982. Demonstration of D-xylose reductase
260. and D-xylitol dehydrogenase in Pachysolen tannophilus. Biotechnol. Lett.
16. Carper, D. A., G. Wistow, C. Nishimura, C. Graham, K. Watanabe, Y. Fujii, 4:607–610.
H. Hayashi, and O. Hayaishi. 1989. A superfamily of NADPH-dependent 45. Smith, P. K., R. I. Krohn, G. T. Hermanson, A. K. Mallia, F. H. Gartner,
reductases in eukaryotes and prokaryotes. Exp. Eye Res. 49:377–388. M. D. Provenzano, E. K. Fujimoto, N. M. Goeke, B. J. Olsen, and D. C.
17. Cromlish, J. A., and T. G. Flynn. 1983. Purification and characterization of Klenk. 1985. Measurement of protein using bicinchoninic acid. Anal. Bio-
two aldose reductase isoenzymes from rabbit muscle. J. Biol. Chem. 258: chem. 150:76–85.
3416–3424. 46. Switzer, R. C., C. R. Merril, and S. Shifrin. 1979. A highly sensitive silver
18. Davidson, W. S., D. J. Walton, and T. G. Flynn. 1987. A comparative study stain for detecting proteins and peptides in polyacrylamide gels. Anal. Bio-
of the tissue and species distribution of NADPH-dependent aldehyde re- chem. 98:231–237.
ductase. Comp. Biochem. Physiol. 60B:309–315. 47. Takuma, S., N. Nakashima, M. Tantirungkij, S. Kinoshita, H. Okada, T.
19. Ditzelmüller, G., C. P. Kubicek, W. Wöhrer, and M. Röhr. 1984. Xylose Seki, and T. Yoshida. 1991. Isolation of xylose reductase gene of Pichia
metabolism in Pachysolen tannophilus: purification and properties of xylose stipitis and its expression in Saccharomyces cerevisiae. Appl. Biochem. Bio-
reductase. Can. J. Microbiol. 30:1330–1336. technol. 28/29:327–340.
20. Flynn, T. G. 1982. Aldehyde reductases: monomeric NADPH-dependent 48. Tanimoto, T., H. Fukuda, and J. Kawamura. 1983. Purification and some
oxidoreductases with multifunctional potential. Biochem. Pharmacol. 31: properties of aldose reductase from rabbit lens. Chem. Pharm. Bull. 31:
2705–2712. 2395–2403.
21. Gong, C.-S., L.-F. Chen, M. C. Flickinger, L.-C. Chiang, and G. T. Tsao. 49. Tanimoto, T., S. Sato, and P. F. Kador. 1990. Purification and properties of
1981. Production of ethanol from D-xylose by using D-xylose isomerase and aldose reductase and aldehyde reductase from EHS tumor cells. Biochem.
yeasts. Appl. Environ. Microbiol. 41:430–436. Pharmacol. 39:445–453.
22. Griffin, B. W. 1992. Functional and structural relationships among aldose 50. Towbin, H., T. Staehelin, and J. Gordon. 1979. Electrophoretic transfer of
reductase, L-hexonate dehydrogenase (aldehyde reductase), and recently proteins from polyacrylamide gels to nitrocellulose sheets: procedure and
identified homologous proteins. Enzyme Microb. Technol. 14:690–695. some applications. Proc. Natl. Acad. Sci. USA 76:4350–4354.
23. Grimshaw, C. E. 1986. Direct measurement of the rate of ring opening of 51. Turner, A. J., and T. G. Flynn. 1982. The nomenclature of aldehyde reduc-
D-glucose by enzyme-catalyzed reduction. Carbohydr. Res. 148:345–348. tases, p. 401–402. In H. Weiner and B. Wermuth (ed.), Enzymology of
24. Hallborn, J., M. Walfridsson, U. Airaksinen, H. Ojamo, B. Hahn-Hägerdahl, carbonyl metabolism: aldehyde dehydrogenase and aldo-keto reductase.
M. Pentillä, and S. Keränen. 1991. Xylitol production by recombinant Sac- Alan R. Liss, Inc., New York.
charomyces cerevisiae. Bio/Technology 9:1090–1095. 52. Vander Jagt, D. L., L. A. Hunsaker, B. Robinson, L. A. Stangebye, and L. M.
25. Hewick, R. M., M. W. Hunkapiller, L. E. Hood, and W. J. Dreyer. 1981. A Deck. 1990. Aldehyde and aldose reductases from human placenta. J. Biol.
gas-liquid solid phase peptide and protein sequenator. J. Biol. Chem. 256: Chem. 265:10912–10918.
7990–7997. 53. Van Zyl, C., B. A. Prior, S. G. Kilian, and E. V. Brandt. 1993. Role of
26. Ho, N. W. Y., F. P. Lin, S. Huang, P. C. Andrews, and G. T. Tsao. 1990. D-ribose as a cometabolite in D-xylose metabolism by Saccharomyces cerevi-
Purification, characterization, and amino terminal sequence of xylose reduc- siae. Appl. Environ. Microbiol. 59:1487–1494.
tase from Candida shehatae. Enzyme Microb. Technol. 12:33–39. 54. Van Zyl, C., B. A. Prior, S. G. Kilian, and J. L. F. Kock. 1989. D-xylose
27. Imamura, Y., T. Higuchi, Y. Nozaki, E. Sugino, S. Hibino, and M. Otagiri. utilization by Saccharomyces cerevisiae. J. Gen. Microbiol. 135:2791–
1993. Purification and properties of carbonyl reductase from rabbit kidney. 2798.
Arch. Biochem. Biophys. 300:570–576. 55. Verduyn, C., R. van Kleef, J. Frank, H. Schreuder, J. P. van Dijken, and
VOL. 61, 1995 ALDO-KETO REDUCTASE FROM S. CEREVISIAE 1585

W. A. Scheffers. 1985. Properties of the NAD(P)H-dependent xylose reduc- 57. Wermuth, B., K. Bohren, G. Heinemann, J.-P. von Wartburg, and K. H.
tase from the xylose-fermenting yeast Pichia stipitis. Biochem. J. 226:669– Gabbay. 1988. Human carbonyl reductase. Nucleotide sequence analysis of a
677. cDNA and amino acid sequence of the encoded protein. J. Biol. Chem.
56. Wermuth, B. 1985. Aldo-keto reductases, p. 209–230. In T. G. Flynn and H. 263:16185–16188.
Weiner (ed.), Enzymology of carbonyl metabolism 2: aldehyde dehydroge- 58. Wermuth, B., H. Burgisser, K. Bohren, and J.-P. Von Wartburg. 1982.
nase, aldo-keto reductase, and alcohol dehydrogenase. Alan R. Liss, Inc., Purification and characterization of human-brain aldose reductase. Eur. J.
New York. Biochem. 127:279–284.