- Email: [email protected]
Structural and functional properties of aldose xylose reductase from the d -xylose-metabolizing yeast Candida tenuis
Chemico-Biological Interactions 130 – 132 (2001) 583 – 595
www.elsevier.com/locate/chembiont
Structural and functional properties of aldose xylose reductase from the D-xylose-metabolizing yeast Candida tenuis Bernd Nidetzky a,*, Peter Mayr a, Wilfried Neuhauser a, Michael Puchberger b a Di6ision of Biochemical Engineering, Institute of Food Technology, Uni6ersity of Agricultural Sciences, Muthgasse 18, A-1190 Vienna, Austria b Institute of Chemistry, Uni6ersity of Agricultural Sciences, Muthgasse 18, A-1190 Vienna, Austria
Abstract The primary structure of the aldose xylose reductase from Candida tenuis (CtAR) is shown to be 39% identical to that of human aldose reductase (hAR). The catalytic tetrad of hAR is completely conserved in CtAR (Tyr51, Lys80, Asp46, His113). The amino acid residues involved in binding of NADPH by hAR (D.K. Wilson, et al., Science 257 (1992) 81 – 84) are 64% identical in CtAR. Like hAR the yeast enzyme is specific for transferring the 4-pro-R hydrogen of the coenzyme. These properties suggest that CtAR is a member of the aldo/keto reductase superfamily. Unlike hAR the enzyme from C. tenuis has a dual coenzyme specificity and shows similar specificity constants for NADPH and NADH. It binds NADP+ approximately 250 times less tightly than hAR. Typical turnover numbers for aldehyde reduction by CtAR (15–20 s − 1) are up to 100-fold higher than corresponding values for hAR, probably reflecting an overall faster dissociation of NAD(P)+ in the reaction catalyzed by the yeast enzyme. © 2001 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Aldo/keto reductase; Stereochemistry; Specificity; Xylose metabolism
* Corresponding author. Tel.: + 43-1-360066274; fax: +43-1-360066251. E-mail address: [email protected] (B. Nidetzky). 0009-2797/01/$ - see front matter © 2001 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 0 0 9 - 2 7 9 7 ( 0 0 ) 0 0 2 8 5 - 4
584
B. Nidetzky et al. / Chemico-Biological Interactions 130–132 (2001) 583–595
1. Introduction The metabolism of D-xylose in yeasts and fungi occurs through a pathway in which the pentose is first reduced to xylitol. This step is catalyzed by aldose (xylose) reductase (EC 1.1.1.21; AR). Yeast ARs can be classified into two groups according to whether they are strictly specific for NADPH or show a dual coenzyme specificity [1]. Therefore, D-xylose reduction can occur via utilization of NADPH or less often, NADH. Xylitol is then oxidized in a strictly NAD+-dependent reaction by xylitol dehydrogenase (EC 1.1.1.14), a Zn2 + -dependent medium-chain alcohol/ sorbitol dehydrogenase [2,3], to yield D-xylulose. The resulting pentulose is phosphorolyzed and further metabolized via different pathways [1]. The utilization of D-xylose by yeasts has been of considerable interest pertaining to its physiology and potential use in the production of ethanol from renewable plant biomass in which D-xylose is a major component of the pentosan fraction [1]. Compared with the great number of reports dealing with the capabilities of different yeast strains to metabolize D-xylose and the development of strategies to improve pentose fermentation by natural and metabolically engineered microorganisms, relatively few studies have been published on the enzymology of the D-xylose pathway. Up to now primary structures of eight yeast ARs are known whose physiological role can be clearly assigned to the conversion of D-xylose [4], as opposed to a rather general function of aldehyde reduction [5] in, e.g. cellular detoxification processes. These enzymes have been purified and characterized in some detail, revealing that yeast ARs occur chiefly as homodimers with a molecular mass of the subunit of : 36 kDa whereas members of the aldo/keto reductase superfamily are usually monomers of the same size [6]. Lee has recently reviewed the relationships of structure and function of ARs from yeasts [4]. A major conclusion of his article has been that yeast ARs could be hybrids between aldo/keto reductases [7] and short-chain dehydrogenases [8]. Using the AR from Candida tenuis (CtAR) [9,10] as model of yeast AR, we show here on the basis of detailed sequence analysis and determination of the stereochemistry of reaction that the major structural and functional properties of yeast AR are those of the aldo/keto reductases. The physiological functions of human aldose reductase (hAR) are thought to include the detoxification of a wide range of reactive carbonyl-containing metabolites and the formation of intracellular polyol as a mechanism to counteract negative osmotic effects. Grimshaw [11] outlined elegantly that hAR has probably evolved to enzymic perfection as a detoxification catalyst by acting with low turnover numbers, but comparable and high catalytic efficiencies, with a wide variety of substrates containing an aldehyde as functional group. In contrast to hAR, CtAR is an inducible, key metabolic enzyme and required for growth and energy. It seems, therefore, interesting to compare the kinetic properties of hAR and CtAR since they are expected to mirror the differences in physiological function. We summarize here data, which lend support to the suggestion that (i) weak binding of coenzyme, and (ii) utilization of binding energy derived from noncovalent interaction with nonreacting portions of the aldehyde are properties of CtAR, which chiefly distinguish the yeast enzyme from hAR.
B. Nidetzky et al. / Chemico-Biological Interactions 130–132 (2001) 583–595
585
2. Results and discussion
2.1. Sequence comparison The alignment of the amino acid sequences of CtAR [10] and hAR [7] is shown in Fig. 1. The primary structures of the two enzymes are 39% identical. The relationships of structure and function of hAR have been established by a series of elegant crystallographic and mechanistic studies of the enzyme [12–15]. They provide an excellent basis for analyzing the amino acid sequence of CtAR which follows.
2.1.1. Catalytic residues The catalytic tetrad of hAR, Tyr48/Asp43/Lys77/His110 (for review, see [6]; [13, 15]) is completely conserved in the yeast enzyme. By analogy, Tyr51 is proposed to be the proton donor of aldehyde reduction by CtAR. The residue pair Asp46/Lys80 is expected to interact with the hydroxy group of Tyr51, thereby lowering its pKa value to approximately 7.5 – 8.0, as observed in the pH profile of kcat for NADH-dependent reduction of D-xylose. Apart from directing the stereoselectivity of the enzyme [13] His113 may facilitate (‘push’) proton donation by Tyr51, as proposed by Penning and co-workers for the mechanism of action of 3a-hydroxysteroid dehydrogenase [16]. The occurrence of a Tyr51-(X)3-Lys motif, where X is any amino acid, in CtAR and other yeast aldose reductases [4] is interesting since this motif presents the catalytic consensus sequence of short-chain dehydrogenases [8]. However, it must be noted that there is precedent of the occurrence of a Tyr-(X)3Lys motif in a member of the aldo/keto reductase superfamily, namely 3a-hydroxysteroid dehydrogenase. The tyrosine (Tyr205) did not have a role in catalysis and was found to be located remote from the active site [17,18]. However, the yeast enzymes appear to be unique in this respect among aldo/keto reductases because the proposed catalytic tyrosine is part of the Tyr-(X)3-Lys motif. Site-directed mutagenesis will clearly be needed to clarify a possible role of Lys55 in catalysis by CtAR. 2.1.2. Coenzyme binding The structure of the complex of hAR and NADPH has been solved at high resolution [12,19] so that residues involved in coenzyme binding are known for this enzyme. Eleven out of 17 residues are conserved in CtAR, including three out of five residues interacting with the 2%-phosphate group of AMP. Interestingly, Thr265 of hAR which hydrogen bonds with its hydroxyl side chain to the 2%-phosphate, is replaced by Leu276 in CtAR and it is tempting to speculate that this replacement could account partly for the much smaller preference of NADPH over NADH, which is observed for CtAR compared with hAR. Significantly, 3a-hydroxysteroid dehydrogenase, which has a dual coenzyme specificity but shows a larger preference of NADPH over NADH than CtAR, does not interact with the bound nucleotide via a residue corresponding to Thr265 in hAR [6,17,20]. The rate of aldehyde reduction by hAR is entirely limited by a slow conformational change, which
586
B. Nidetzky et al. / Chemico-Biological Interactions 130–132 (2001) 583–595
Fig. 1. Comparison of CtAR and hAR. The amino acid sequences were obtained under their GenBank accession numbers X15414 (hAR) and AF074484 (CtAR). CLUSTAL alignment was employed using the DNAStar program with standard settings. ALR – C. – ten is CtAR and ALR – human is hAR. The catalytic residues of hAR [13,15] and the residues involved in coenzyme binding by hAR [12] are marked by superscripts ‘c’ and ‘n’, respectively.
B. Nidetzky et al. / Chemico-Biological Interactions 130–132 (2001) 583–595
587
precedes the actual release of NADP+ [14]. The motion of a loop of residues Gly213 to Leu227 in hAR is required for NADP+ release [12]. Specific interactions such as a salt link of Asp216 and Lys21, or interactions involving Cys298 anchor the loop and therefore, the coenzyme is held tightly in place explaining the extremely small dissociation constant of the complex of hAR and NADP+ [13,14,21]. Notably Asp216 and Cys298 of hAR are not conserved in CtAR, and the putative nucleotide-enfolding loop of CtAR has a four amino acid-long insertion at a position corresponding to Pro215 and Asp216 in hAR. Structural differences between hAR and CtAR in the loop, which mediates the so-called ‘conformational clamping’ in the reaction catalyzed by hAR, may determine the : 250-fold weaker binding of NADP+ by CtAR (see later). In addition, they could arguably be responsible for the observed 100-fold faster overall reduction of aldehyde substrates by the yeast enzyme, by allowing steps involved in coenzyme release to proceed at a higher rate. An interesting difference in the coenzyme-binding sites of mammalian AR and AR from yeast is brought to light by using chemical modification with pyridoxal 5%-phosphate. The aldehyde of pyridoxal 5%-phoshate has been shown to form a Schiff Base with the o-NH2 group of Lys262 of AR from human psoas muscle [22]. The lysine belongs to the coenzyme-interaction consensus sequence (-IPKS-) of aldo/keto reductases. Interestingly, CtAR modified with pyridoxal 5%-phosphate was found to be inactive [9] whereas the modified hAR showed an increase in kcat [22].
2.1.3. Substrate specificity The extremely broad spectrum of aldehyde substrates reduced by the different members of the aldo/keto reductase superfamily does not warrant a detailed discussion of possible differences in enzyme/aldehyde interactions and substrate specificities based on comparative sequence analysis alone. However, a few structural properties of CtAR are interesting. Molecular modeling of aldehyde binding at the active site of hAR [23] suggested that the correctly oriented 2-OH group of the substrate had hydrogen bonds with Ho1 of Trp111. Crystallographic studies suggest an involvement of Trp111 in inhibitor binding [24]. This residue is replaced by Phe114 in CtAR and the phenylalanine is conserved throughout yeast ARs [4] and most hydroxysteroid dehydrogenases [6,17]. Several apolar residues that line the substrate pocket of hAR are replaced by polar residues in CtAR, reflected by a 1.7-fold greater apparent hydrophobicity of the active site of hAR compared with that of the active site of CtAR [25]. Significantly, Val47 of hAR is substituted by an aspartate in CtAR and it seems possible that the replacement affects the microenvironment of the catalytic-acid tyrosine. The C-terminal region of hAR and CtAR show considerable differences in sequence, and it should be noted that this part of the protein has a role to determine the substrate specificity for hydroxylated, hydrophobic or charged aldehydes [26]. A final point is important to be mentioned. In his review on yeast ARs [4] Lee pointed out a seven amino acid-long peptide (Gly130 – 136 in CtAR), which occurs as an insertion at positions Asp125 and Glu126 of hAR whereby the two glycines of CtAR replace the carboxylic acids in hAR. The sequence of the peptide, GFYCGDG in CtAR, is overall reminiscent of
588
B. Nidetzky et al. / Chemico-Biological Interactions 130–132 (2001) 583–595
the Wierenga coenzyme-binding motif of other dehydrogenases e.g. [8], which led Lee to suggest that this part of the sequence could be somehow involved in yeast AR – coenzyme interaction. The structures of the ternary complex of the FR-1 protein, NADPH and the aldose-reductase inhibitor zopolrestat [27], and of the ternary complex of hAR and the same inhibitor [24] have shown that residues 121 – 135 of hAR are inolved in inhibitor binding by the FR-1 protein and hAR. Therefore, a role of the ‘Wierenga-type peptide’ for determining the substrate specificity of CtAR is well possible and must be considered. However, studies of AR from Pichia stipitis have shown that the cystein corresponding to Cys133 in CtAR is susceptible to modification by thiol reagents [4]. The modification was inhibited in the presence of NADPH, possibly suggesting that the cystein and NADPH come close in space in the binary complex.
2.2. Stereochemistry of hydride transfer When (4S)-[4-D]NADD was used as coenzyme for enzymatic reaction with the hydrogen was depleted upon CtAR-catalyzed oxidation of NADD and [4-D]NAD+ was obtained (Fig. 2 A). By contrast, when (4R)-[4-D]NADD was used, the deuterium was depleted at the C4 position of the nicotinamide ring (Fig. 2 B). Therefore, CtAR is specific for transferring the 4-pro-R hydrogen of NADH, which is typical of members of the aldo/keto reductase superfamily. By contrast, short-chain dehydrogenases/reductases show B-type stereospecificity [8]. The stereochemistry of the enzymatic reaction is expected to be a constant trait of members of an enzyme family. Therefore, and by considering the results of sequence comparison, we propose that CtAR is a true member of the aldo/keto reductase superfamily. D-xylose,
2.3. E6idence of direct nucleotide transfer to CtAR from a B-type dehydrogenase/NADH complex Srivastava and Bernhard [29] have shown that direct transfer of NADH from one dehydrogenase to another dehydrogenase can occur when the two enzymes differ in specificity pertaining to the C4H, which is transferred in the reaction. They suggested that transfer of NADH between two dehydrogenases of identical (A-type or B-type) specificity would have to take place via diffusion in aqueous solution. The basis of their experiments was to record the rate of enzymatic oxidation of NADH by a dehydrogenase in the presence of a large molar excess of a binary complex of NADH and another dehydrogenase of A- or B-type specificity, taking care that the concentration of free NADH available for reaction was exceedingly low. When the observed enzymatic rate was decreased relative to that of the control reaction lacking the dehydrogenase/NADH complex, nucleotide exchange between the two dehydrogenases was proposed not to occur by direct transfer. Results for CtAR-catalyzed reduction of D-xylose in the presence of two representative dehydrogenase/NADH complexes are shown in Table 1. The interpretation of the data is clear: direct transfer of NADH to CtAR takes place from a dehydrogenase of
B. Nidetzky et al. / Chemico-Biological Interactions 130–132 (2001) 583–595
589
B-type specificity but not one of A-type specificity. This finding is in excellent agreement with the generalization proposed by Srivastava and Bernhard [29]. Now, since the stereochemistry of yeast xylitol dehydrogenase is A (B.N., unpublished results), direct nucleotide transfer between xylitol dehydrogenase and AR will not occur, although one might intuitively expect it for two successive enzymes in a metabolic pathway. If yeast AR specifically recognized the XDH/NADH complex as coenzyme donor and therefore used NADH with large preference over NADPH in vivo, the first two steps of the xylose pathway would be ‘redox-neutral’, which has been discounted firmly by several physiological studies of D-xylose metabolism in yeast [1]. However, NADH has been shown to serve as in-vivo coenzyme of the AR-catalyzed reduction of D-xylose [30].
Fig. 2. Stereospecificity of hydride transfer catalyzed by CtAR, monitored by 1H-NMR. The products obtained on complete enzymatic oxidation of stereospecifically labeled [28] (A) (4S)-[4-D]NADD and (B) (4R)-[4-D]NADD are shown. The important signal is observed at a chemical shift of about 8.80 ppm. The spectrum in (A) corresponds to [4-D]NAD+, that in (B) corresponds to [4-H]NAD+. Methods: spectra were recorded at 27°C non-spinning in a 5-mm tube at 300.13 MHz with a Bruker Avance 300 spectrometer. All samples were dissolved in D2O (99.75%) and the resulting spectra were referenced to internal DSS (3-(trimethylsilyl)-1-propanesulfonic acid, sodium salt; d =0 ppm). A typical reaction mixture for enzymatic reduction contained 3.5 mg (4R)-[4-D]NADD or (4S)-[4-D]NADD with a deuterium content of ] 98%, 10 U CtAR and 300 mM D-xylose dissolved in 50 mM potassium phosphate buffer pD 7.0. When the oxidation of NADD was complete (monitored spectrophotometrically at 340 nm), the sample was used for NMR analysis without further treatment.
590
B. Nidetzky et al. / Chemico-Biological Interactions 130–132 (2001) 583–595
Table 1 Comparison of the rates of NADH-dependent reduction of absence of a second dehydrogenase in the assay Dehydrogenase used in assay
D-xylose
by CtAR in the presence or
Stereospecificity of the dehydrogenase in the binary enzyme NADH complex
None – Lactate A-type (4-pro-R) dehydrogenase Glutamate B-type (4-pro-S) dehydrogenase
Relative activity of CtAR (%)a 100 491 83 9 4
a A solution of 100 mM D-xylose and 44 mM NADH in 0.3 M phosphate buffer, pH 6.0, was preincubated at 25°C for 0.5 min with a more than 5000-fold excess (: 80 mM) of NADH binding sites of lactate dehydrogenase or glutamate dehydrogenase, relative to NADH-binding sites of CtAR. The reaction was then started by the addition of 0.015 mM CtAR, and the rate of NADH consumption was monitored spectrophotometrically. 9 are S.E., calculated from three independent determinations.
2.4. Steady-state kinetics The kinetic parameters for the reduction of D-xylose by CtAR and hAR are summarized in Table 2. Quite significant differences between the two enzymes are found when comparison is made on the basis of the values of the turnover number (: 85-fold difference), the Km for D-xylose (: 51-fold difference) and the binding of NADP+ ( :250-fold difference). The possible physiological significance of these differences has been discussed recently [11,31]. For CtAR, weak binding of the reactants and at the same time efficient substrate turnover with a high value of kcat makes perfect sense when the key role of the enzyme in the catabolic pathway of D-xylose is considered.
2.5. Substrate specificity 2.5.1. Hydrophobicity By studying the NADH-dependent enzymatic reduction of a series of straightchain aldehydes ranging from propanal to hexanal we have shown recently [31] that Table 2 Comparison of the kinetic parameters of CtAR and hAR for the reduction of
D-xylose
Parameter
CtARa
hARb
kcat (s−1) Km,xylose (mM) KiNAD(P)+ (mM) Ki NAD(P)H (mM)
17.0 76 195 (NAD+), 1.5 (NADP+) 16 (NADH)
0.20 1.5 0.006 0.010
a From [9], experiments carried out at 25°C in 50 mM phosphate buffer, pH 7.0, using NADH as coenzyme. b From [14], experiments carried out at 25°C in 33 mM phosphate buffer, pH 8.0, using NADPH as coenzyme.
B. Nidetzky et al. / Chemico-Biological Interactions 130–132 (2001) 583–595 Table 3 Comparison of the kinetic parameters for reaction of CtARa and hARb with D-glucose, respectively, and derivatives of the parent sugar Aldehyde
CtAR kcat (s−1)
Parent 14.7 2-Deoxy 2-Deoxy-2-fluo 1.0 ro
D-galactose
591
and
hAR Km (mM)
kcat/Km (M−1 s−1)
kcat (min−1)
Km (mM)
kcat/Km (M−1 s−1)
228
66 B0.06 11
15.1 11.1 3.00
12.9 10.4 12.5
19.5 17.8 4
92
From [31], reactions done at 25°C in 50 mM potassium phosphate buffer, pH 7.0, and using 220 mM NADH. b From [32], reactions done at 27°C in 0.1 M sodium phosphate buffer, pH 7.2, using 160 mM NADPH. a
CtAR is able to utilize hydrophobic bonding with nonreacting portions of the substrate to bring about increases in the specificity constant ( : 150-fold) and the turnover number (:2.6-fold). The significant increase in kcat with increasing hydrophobicity of the alkyl chain attached to the reacting aldehyde indicates that the rate of dissociation of NAD+ cannot be fully rate-limiting for the overall reaction catalyzed by the yeast enzyme. If it were, the value of kcat would be expected to be constant across the series of the straight-chain aldehydes. In contrast to CtAR, the turnover numbers for reduction of the same series of aldehydes by hAR showed no variation when the alkyl chain increased in length from two to five carbon atoms [25], in agreement with a kinetic mechanism in which coenzyme dissociation is rate-limiting [14]. Further analyses of the kinetic data have shown that the substrate binding pockets of hAR and CtAR appear to be 2.5 times [25] and 1.4 times [31] more hydrophobic than n-octanol, respectively.
2.5.2. Hydrogen bonding 6ersus inducti6e/field effect in the reaction of CtAR and hAR with sugars Table 3 summarizes kinetic parameters for the enzymatic reduction of aldose sugars and derivatives of these sugars in which the nonreacting hydroxy group at C-2 has been replaced by hydrogen or fluorine. The substitution OH H or OH F has a relatively small effect on the overall structural properties of the sugar such as the proportion of free aldehyde present in aqueous solution, for example (see [32]). The ability of F to weakly accept a hydrogen for bonding which, of course, H cannot (for discussion, see [33]), makes deoxy and deoxyfluoro analogues potentially useful probes to study hydrogen bonding interactions in enzymatic reactions. However, the inductive/field effect of the substituent must be considered in each case and particularly that of the fluorine would seem important, its
B. Nidetzky et al. / Chemico-Biological Interactions 130–132 (2001) 583–595
592
substituent constant (sI) in aliphatic systems being 0.52 [34]! The electron-withdrawing fluorine is expected to affect a stronger polarization of the neighboring carbonyl group, which could arguably facilitate hydride transfer. In addition, the rate of reactions in which there is development of negative charge on an atom in the transition state — the alcoholate oxygen, for example — are known to be increased inductively by a nearby electron-withdrawing group. In summary, therefore, when hydrogen bonding at the 2-OH is most important in the enzymatic reaction, the expected order of reactivity is -OH \ F]H. In contrast, when electronic effects are predominant, the fluoro substrate will be more reactive than the deoxy while its reactivity relative to the OH is difficult to predict. The results shown for hAR are consistent with neither of the two cases. The data for CtAR probably reflect a combination of hydrogen-bonding effects and electronic effects, especially for the reaction with the fluoro substrate. In any case, the use of deoxy and deoxyfluoro sugars reveals marked differences between hAR and CtAR, which it is worth to examine in better detail by using more sophisticated kinetic techniques, e.g. rapid-kinetic and isotope-effect studies [14].
2.5.3. Reaction with three-carbon aldehydes In Table 4 we compare kinetic parameters for hAR- and CtAR-catalyzed reductions of three-carbon aldehydes, which differ mainly in the oxidation state at the a-carbon atom. The results show clearly that the a-carbon atom has an important role for determining the catalytic efficiencies of both enzymes. In the case of hAR the increases in kcat/Km for reactions with acrolein, methylglyoxal, and DL-glyceraldehyde, relative to the catalytic efficiency for reduction of propanal, reflect decreases in Km by about the same factor. Consequently, kcat of hAR is essentially constant across the series of three-carbon aldehydes in Table 4. In the case of CtAR, the observed variations in kcat/Km are distributed between changes in kcat (11-fold) and Km (18-fold). The variation in kcat seen for the yeast enzyme corroborates the above suggestion that noncovalent interactions with the aldehyde can be utilized by CtAR to decrease the activation free energy of the rate-limiting Table 4 Comparison of the kinetic parameters for reaction of CtARa and hARb with three-carbon aldehydes Aldehyde
Propanal Acrolein Methylglyoxal DL-glyceraldehyde a b
CtAR
hAR
kcat (s−1)
Km (mM)
kcat/Km (M−1 s−1)
kcat (min−1)
Km (mM)
kcat/Km (M−1 min−1)
13.0 2.4 26.3 15.5
365 13 6.3 2
36 185 4175 7750
112 87 142 120
1700 80 8 16
1.1×103 1.8×104 3.0×105 1.2×105
Reactions at 25°C in 50 mM potassium phosphate buffer, pH 7.0, and using 220 mM NADH. From [35] at 25°C in 0.1 M sodium phosphate buffer, pH 7.0, using 100 mM NADPH.
B. Nidetzky et al. / Chemico-Biological Interactions 130–132 (2001) 583–595
593
step(s) reflected by the value of kcat. A comparison of the kinetic parameters of CtAR for NADH-dependent reduction of acrolein (which has an electron-withdrawing double bond between C-2 and C-3) and propanal are interesting. The value of kcat for reaction with acrolein is 1/5.5 times that for the reaction with propanal. Yet, the corresponding catalytic efficiency for reaction with propanal is only 1/5 times that for reaction with acrolein. Binding interactions between CtAR and aldehydes may be productive in which case they will be expressed both at the Michaelis complex and the transition state, and lead to an increase in kcat/Km while kcat remains unaffected. Nonproductive binding is expressed at the Michaelis complex but not at the transition state, and its effect will be a decrease in the observed value of kcat by the fraction of productively bound substrate, with kcat/Km remaining unaffected [36]. The results shown in Table 4 could suggest that nonproductive binding occurs to large extent in the reaction of CtAR with acrolein, and reaction to give products takes place only from a relatively small fraction of productive complexes. In the series of aldehydes in Table 4, DL-glyceraldehyde is the best substrate for NADH-dependent reduction by CtAR. Methylglyoxal, which is the most reactive compound among the four aldehydes tested, is the preferred substrate of hAR. Overall, hAR shows catalytic efficiencies for reactions with three-carbon aldehydes that are at least ten times those of CtAR for the same reactions. However, the differences between the two enzymes are especially large (nearly 100-fold) when comparison is made on the basis of the second-order rate constants for the reduction of chemically reactive aldehydes, acrolein and methylglyoxal. Therefore, the kinetic data shown in Table 4 are in good agreement with the suggestion that (i) catalytic efficiency of aldehyde reduction by CtAR is determined to a significant extent by noncovalent bonding with parts of the substrate (e.g. hydroxyls in DL-glyceraldehyde or other sugars) remote from the site of chemical transformation; and (ii) by forming an extremely tight, thus reactive complex with NADPH, hAR is able to reduce a very broad spectrum of aldehyde substrates with catalytic efficiencies that are largely independent on such bonding [11] but seem to mirror the intrinsic chemical reactivities of the aldehydes.
3. Conclusions Comparison of CtAR with the well characterized hAR shows that the yeast enzyme belongs to the aldo/keto reductase superfamily. In the nomenclature of aldo-keto reductases [37], CtAR has been assigned AKR 285. CtAR shares many structural and functional properties with its mammalian counterpart enzyme. Clear differences between the two enzymes pertain to the magnitude of the turnover number and the stabilities of binary enzyme/coenzyme complexes. By comparison with hAR, which is thought to be well adapted for the reduction of structurally diverse and potentially toxic aldehydes [11], it will be interesting to pinpoint the structural features of CtAR which are the key to reductase function with fluxional efficiency in mainstream catabolism of pentose sugars and determine the major ‘kinetic’ differences to hAR.
594
B. Nidetzky et al. / Chemico-Biological Interactions 130–132 (2001) 583–595
Acknowledgements Financial support from the Austrian Science Fund is gratefully acknowledged (grant P-12569-MOB to B.N.).
References [1] P. Mishra, A. Singh, Microbial pentose utilization, in: Advances in Applied Microbiology, vol. 39, Academic Press, San Diego, CA, 1993, pp. 91 – 185. [2] B. Persson, J. Hallborn, M. Walfridsson, B. Hahn-Ha¨gerdal, S. Kera¨nen, M. Penttila¨, H. Jo¨rnvall, Dual relationships of xylitol and alcohol dehydrogenases in families of two protein types, FEBS Lett. 324 (1993) 9–14. [3] R. Lunzer, Y. Mamnun, D. Haltrich, K.D. Kulbe, B. Nidetzky, Structural and functional properties of a yeast xylitol dehydrogenase, a zinc-containing metalloenzyme similar to mediumchain sorbitol dehydrogenases, Biochem. J. 336 (1998) 91 – 99. [4] H. Lee, The structure and function of yeast xylose (aldose) reductases, Yeast 14 (1998) 977 – 984. [5] A. Kuhn, C. van Zyl, A. van Tonder, B. Prior, Purification and partial characterization of an aldo–keto reductase from Saccharomyces cere6isiae, Appl. Environ. Microbiol. 61 (1995) 1580 – 1585. [6] J.M. Jez, M.J. Bennett, B.P. Schlegel, M. Lewis, T.M. Penning, Comparative anatomy of the aldo–keto reductase superfamily, Biochem. J. 326 (1997) 625 – 636. [7] K.M. Bohren, B. Bullock, B. Wermuth, K.H. Gabbay, The aldo – keto reductase superfamily. cDNAs and deduced amino acid sequences of human aldehyde and aldose reductases, J. Biol. Chem. 264 (1989) 9547–9551. [8] H. Jo¨rnvall, B. Persson, M. Krook, S. Atrian, R. Gonzalez-Duarte, J. Jeffery, D. Ghosh, Short-chain dehydrogenases/reductases (SDR), Biochemistry 34 (1995) 6003 – 6013. [9] W. Neuhauser, D. Haltrich, K.D. Kulbe, B. Nidetzky, NAD(P)H-dependent aldose reductase from the xylose-assimilating yeast Candida tenuis, Biochem. J. 326 (1997) 683 – 692. [10] B. Ha¨cker, A. Habenicht, M. Kiess, R. Mattes, Xylose utilisation: cloning and characterisation of the xylose reductase from Candida tenuis, Biol. Chem. 380 (1999) 1395 – 1403. [11] C.E. Grimshaw, Aldose reductase: model for a new paradigm of enzymic perfection in detoxification catalysts, Biochemistry 31 (1992) 10139 – 10145. [12] D.K. Wilson, K.M. Bohren, K.H. Gabbay, F.A. Quiocho, An unlikely sugar substrate site in the 1.65 A structure of the human aldose reductase holoenzyme implicated in diabetic complications, Science 257 (1992) 81–84. [13] K.M. Bohren, C.E. Grimshaw, C.J. Lai, D.H. Harrison, D. Ringe, G.A. Petsko, K.H. Gabbay, Tyrosine-48 is the proton donor and histidine-110 directs substrate stereochemical selectivity in the reduction reaction of human aldose reductase: enzyme kinetics and crystal structure of the Y48H mutant enzyme, Biochemistry 33 (1994) 2021 – 2032. [14] C.E. Grimshaw, K.M. Bohren, C.J. Lai, K.H. Gabbay, Human aldose reductase: rate constants for a mechanism including interconversion of ternary complexes by recombinant wild-type enzyme, Biochemistry 34 (1995) 14356–14365. [15] I. Tarle, D.W. Borhani, D.K. Wilson, F.A. Quiocho, J.M. Petrash, Probing the active site of human aldose reductase. Site-directed mutagenesis of Asp-43, Tyr-48, Lys-77, and His-110, J. Biol. Chem. 268 (1993) 25687–25693. [16] B.P. Schlegel, J.M. Jez, T.M. Penning, Mutagenesis of 3a-hydroxysteroid dehydrogenase reveals a ‘push–pull’ mechanism for proton transfer in aldo – keto reductases, Biochemistry 37 (1998) 3538–3548. [17] T.M. Penning, Molecular endocrinology of hydroxysteroid dehydrogenases, Endocr. Rev. 18 (1997) 281–305.
B. Nidetzky et al. / Chemico-Biological Interactions 130–132 (2001) 583–595
595
[18] J.E. Pawlowski, T.M. Penning, Overexpression and mutagenesis of the cDNA for rat liver 3a-hydroxysteroid/dihydrodiol dehydrogenase. Role of cysteines and tyrosines in catalysis, J. Biol. Chem. 269 (1994) 13502–13510. [19] D.W. Borhani, T.M. Harter, J.M. Petrash, The crystal structure of the aldose reductase.NADPH binary complex, J. Biol. Chem. 267 (1992) 24841 – 24847. [20] S.S. Hoog, J.E. Pawlowski, P.M. Alzari, T.M. Penning, M. Lewis, Three-dimensional structure of rat liver 3a-hydroxysteroid/dihydrodiol dehydrogenase: a member of the aldo – keto reductase superfamily, Proc. Natl. Acad. Sci. 91 (1994) 2517 – 2521. [21] C.E. Grimshaw, K.M. Bohren, C.J. Lai, K.H. Gabbay, Human aldose reductase: subtle effects revealed by rapid kinetic studies of the C298A mutant enzyme, Biochemistry 34 (1995) 14366 – 14373. [22] N.A. Morjana, C. Lyons, T.G. Flynn, Aldose reductase from human psoas muscle. Affinity labeling of an active site lysine by pyridoxal 5%-phosphate and pyridoxal 5%-diphospho-5%-adenosine, J. Biol. Chem. 264 (1989) 2912–2919. [23] H.L. De Winter, M. von Itzstein, Aldose reductase as a target for drug design: molecular modeling calculations on the binding of acyclic sugar substrates to the enzyme, Biochemistry 34 (1995) 8299–8308. [24] D.K. Wilson, I. Tarle, J.M. Petrash, F.A. Quiocho, Refined 1.8 A, structure of human aldose reductase complexed with the potent inhibitor zopolrestat, Proc. Natl. Acad. Sci. USA 90 (1993) 9847–9851. [25] S. Srivastava, S.J. Watowich, J.M. Petrash, S.K. Srivastava, A. Bhatnagar, Structural and kinetic determinants of aldehyde reduction by aldose reductase, Biochemistry 38 (1999) 42 – 54. [26] K.M. Bohren, C.E. Grimshaw, K.H. Gabbay, Catalytic effectiveness of human aldose reductase. Critical role of C-terminal domain, J. Biol. Chem. 267 (1992) 20965 – 20970. [27] D.K. Wilson, T. Nakano, J.M. Petrash, F.A. Quiocho, 1.7 A, structure of FR-1, a fibroblast growth factor-induced member of the aldo–keto reductase family, complexed with coenzyme and inhibitor, Biochemistry 34 (1995) 14323–14330. [28] S.B. Mostad, A. Glasfeld, Using high field NMR to determine dehydrogenase stereospecificity with respect to NADH, J. Chem. Educ. 70 (1993) 504 – 506. [29] D.K. Srivastava, S.A. Bernhard, Mechanism of transfer of reduced nicotinamide adenine dinucleotide among dehydrogenases. Transfer rates and equilibria with enzyme – enzyme complexes, Biochemistry 26 (1987) 1240–1246. [30] N. Meinander, G. Zacchi, B. Hahn-Ha¨gerdal, A heterologous reductase affects the redox balance of recombinant Saccharomyces cere6isiae, Microbiology 142 (1996) 165 – 172. [31] W. Neuhauser, D. Haltrich, K.D. Kulbe, B. Nidetzky, Noncovalent enzyme/substrate interactions in the catalytic mechanism of yeast aldose reductase, Biochemistry 37 (1998) 1116 – 1123. [32] M.E. Scott, R.E. Viola, The use of fluoro- and deoxy-substrate analogs to examine binding specificity and catalysis in the enzymes of the sorbitol pathway, Carbohydr. Res. 313 (1998) 247–253. [33] I.P. Street, C.R. Armstrong, S.G. Withers, Hydrogen bonding and specificity. Fluorodeoxy sugars as probes of hydrogen bonding in the glycogen phosphorylase– glucose complex, Biochemistry 25 (1986) 6021–6027. [34] A.R. Fersht, in: W.H. Freeman (Ed.), Enzyme Structure and Mechanism, second ed., Freeman, New York, 1985, p. 81. [35] D.L. Vander Jagt, N.S. Kolb, T.J. Vander Jagt, J. Chino, F.J. Martinez, L.A. Hunsaker, R.E. Royer, Substrate specificity of human aldose reductase: identification of 4-hydroxynonenal as an endogenous substrate, Biochim. Biophys. Acta 1249 (1995) 117 – 126. [36] W.P. Jencks, Binding energy, specificity, and enzymic catalysis, Adv. Enzymol. 43 (1975) 219 – 410. [37] J.M. Jez, T.G. Flynn, T.M. Penning, A new nomenclature for the aldo-keto reductase superfamily, Biochem. Pharmacol. 54 (1997) 639– 647.
.
www.elsevier.com/locate/chembiont
Structural and functional properties of aldose xylose reductase from the D-xylose-metabolizing yeast Candida tenuis Bernd Nidetzky a,*, Peter Mayr a, Wilfried Neuhauser a, Michael Puchberger b a Di6ision of Biochemical Engineering, Institute of Food Technology, Uni6ersity of Agricultural Sciences, Muthgasse 18, A-1190 Vienna, Austria b Institute of Chemistry, Uni6ersity of Agricultural Sciences, Muthgasse 18, A-1190 Vienna, Austria
Abstract The primary structure of the aldose xylose reductase from Candida tenuis (CtAR) is shown to be 39% identical to that of human aldose reductase (hAR). The catalytic tetrad of hAR is completely conserved in CtAR (Tyr51, Lys80, Asp46, His113). The amino acid residues involved in binding of NADPH by hAR (D.K. Wilson, et al., Science 257 (1992) 81 – 84) are 64% identical in CtAR. Like hAR the yeast enzyme is specific for transferring the 4-pro-R hydrogen of the coenzyme. These properties suggest that CtAR is a member of the aldo/keto reductase superfamily. Unlike hAR the enzyme from C. tenuis has a dual coenzyme specificity and shows similar specificity constants for NADPH and NADH. It binds NADP+ approximately 250 times less tightly than hAR. Typical turnover numbers for aldehyde reduction by CtAR (15–20 s − 1) are up to 100-fold higher than corresponding values for hAR, probably reflecting an overall faster dissociation of NAD(P)+ in the reaction catalyzed by the yeast enzyme. © 2001 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Aldo/keto reductase; Stereochemistry; Specificity; Xylose metabolism
* Corresponding author. Tel.: + 43-1-360066274; fax: +43-1-360066251. E-mail address: [email protected] (B. Nidetzky). 0009-2797/01/$ - see front matter © 2001 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 0 0 9 - 2 7 9 7 ( 0 0 ) 0 0 2 8 5 - 4
584
B. Nidetzky et al. / Chemico-Biological Interactions 130–132 (2001) 583–595
1. Introduction The metabolism of D-xylose in yeasts and fungi occurs through a pathway in which the pentose is first reduced to xylitol. This step is catalyzed by aldose (xylose) reductase (EC 1.1.1.21; AR). Yeast ARs can be classified into two groups according to whether they are strictly specific for NADPH or show a dual coenzyme specificity [1]. Therefore, D-xylose reduction can occur via utilization of NADPH or less often, NADH. Xylitol is then oxidized in a strictly NAD+-dependent reaction by xylitol dehydrogenase (EC 1.1.1.14), a Zn2 + -dependent medium-chain alcohol/ sorbitol dehydrogenase [2,3], to yield D-xylulose. The resulting pentulose is phosphorolyzed and further metabolized via different pathways [1]. The utilization of D-xylose by yeasts has been of considerable interest pertaining to its physiology and potential use in the production of ethanol from renewable plant biomass in which D-xylose is a major component of the pentosan fraction [1]. Compared with the great number of reports dealing with the capabilities of different yeast strains to metabolize D-xylose and the development of strategies to improve pentose fermentation by natural and metabolically engineered microorganisms, relatively few studies have been published on the enzymology of the D-xylose pathway. Up to now primary structures of eight yeast ARs are known whose physiological role can be clearly assigned to the conversion of D-xylose [4], as opposed to a rather general function of aldehyde reduction [5] in, e.g. cellular detoxification processes. These enzymes have been purified and characterized in some detail, revealing that yeast ARs occur chiefly as homodimers with a molecular mass of the subunit of : 36 kDa whereas members of the aldo/keto reductase superfamily are usually monomers of the same size [6]. Lee has recently reviewed the relationships of structure and function of ARs from yeasts [4]. A major conclusion of his article has been that yeast ARs could be hybrids between aldo/keto reductases [7] and short-chain dehydrogenases [8]. Using the AR from Candida tenuis (CtAR) [9,10] as model of yeast AR, we show here on the basis of detailed sequence analysis and determination of the stereochemistry of reaction that the major structural and functional properties of yeast AR are those of the aldo/keto reductases. The physiological functions of human aldose reductase (hAR) are thought to include the detoxification of a wide range of reactive carbonyl-containing metabolites and the formation of intracellular polyol as a mechanism to counteract negative osmotic effects. Grimshaw [11] outlined elegantly that hAR has probably evolved to enzymic perfection as a detoxification catalyst by acting with low turnover numbers, but comparable and high catalytic efficiencies, with a wide variety of substrates containing an aldehyde as functional group. In contrast to hAR, CtAR is an inducible, key metabolic enzyme and required for growth and energy. It seems, therefore, interesting to compare the kinetic properties of hAR and CtAR since they are expected to mirror the differences in physiological function. We summarize here data, which lend support to the suggestion that (i) weak binding of coenzyme, and (ii) utilization of binding energy derived from noncovalent interaction with nonreacting portions of the aldehyde are properties of CtAR, which chiefly distinguish the yeast enzyme from hAR.
B. Nidetzky et al. / Chemico-Biological Interactions 130–132 (2001) 583–595
585
2. Results and discussion
2.1. Sequence comparison The alignment of the amino acid sequences of CtAR [10] and hAR [7] is shown in Fig. 1. The primary structures of the two enzymes are 39% identical. The relationships of structure and function of hAR have been established by a series of elegant crystallographic and mechanistic studies of the enzyme [12–15]. They provide an excellent basis for analyzing the amino acid sequence of CtAR which follows.
2.1.1. Catalytic residues The catalytic tetrad of hAR, Tyr48/Asp43/Lys77/His110 (for review, see [6]; [13, 15]) is completely conserved in the yeast enzyme. By analogy, Tyr51 is proposed to be the proton donor of aldehyde reduction by CtAR. The residue pair Asp46/Lys80 is expected to interact with the hydroxy group of Tyr51, thereby lowering its pKa value to approximately 7.5 – 8.0, as observed in the pH profile of kcat for NADH-dependent reduction of D-xylose. Apart from directing the stereoselectivity of the enzyme [13] His113 may facilitate (‘push’) proton donation by Tyr51, as proposed by Penning and co-workers for the mechanism of action of 3a-hydroxysteroid dehydrogenase [16]. The occurrence of a Tyr51-(X)3-Lys motif, where X is any amino acid, in CtAR and other yeast aldose reductases [4] is interesting since this motif presents the catalytic consensus sequence of short-chain dehydrogenases [8]. However, it must be noted that there is precedent of the occurrence of a Tyr-(X)3Lys motif in a member of the aldo/keto reductase superfamily, namely 3a-hydroxysteroid dehydrogenase. The tyrosine (Tyr205) did not have a role in catalysis and was found to be located remote from the active site [17,18]. However, the yeast enzymes appear to be unique in this respect among aldo/keto reductases because the proposed catalytic tyrosine is part of the Tyr-(X)3-Lys motif. Site-directed mutagenesis will clearly be needed to clarify a possible role of Lys55 in catalysis by CtAR. 2.1.2. Coenzyme binding The structure of the complex of hAR and NADPH has been solved at high resolution [12,19] so that residues involved in coenzyme binding are known for this enzyme. Eleven out of 17 residues are conserved in CtAR, including three out of five residues interacting with the 2%-phosphate group of AMP. Interestingly, Thr265 of hAR which hydrogen bonds with its hydroxyl side chain to the 2%-phosphate, is replaced by Leu276 in CtAR and it is tempting to speculate that this replacement could account partly for the much smaller preference of NADPH over NADH, which is observed for CtAR compared with hAR. Significantly, 3a-hydroxysteroid dehydrogenase, which has a dual coenzyme specificity but shows a larger preference of NADPH over NADH than CtAR, does not interact with the bound nucleotide via a residue corresponding to Thr265 in hAR [6,17,20]. The rate of aldehyde reduction by hAR is entirely limited by a slow conformational change, which
586
B. Nidetzky et al. / Chemico-Biological Interactions 130–132 (2001) 583–595
Fig. 1. Comparison of CtAR and hAR. The amino acid sequences were obtained under their GenBank accession numbers X15414 (hAR) and AF074484 (CtAR). CLUSTAL alignment was employed using the DNAStar program with standard settings. ALR – C. – ten is CtAR and ALR – human is hAR. The catalytic residues of hAR [13,15] and the residues involved in coenzyme binding by hAR [12] are marked by superscripts ‘c’ and ‘n’, respectively.
B. Nidetzky et al. / Chemico-Biological Interactions 130–132 (2001) 583–595
587
precedes the actual release of NADP+ [14]. The motion of a loop of residues Gly213 to Leu227 in hAR is required for NADP+ release [12]. Specific interactions such as a salt link of Asp216 and Lys21, or interactions involving Cys298 anchor the loop and therefore, the coenzyme is held tightly in place explaining the extremely small dissociation constant of the complex of hAR and NADP+ [13,14,21]. Notably Asp216 and Cys298 of hAR are not conserved in CtAR, and the putative nucleotide-enfolding loop of CtAR has a four amino acid-long insertion at a position corresponding to Pro215 and Asp216 in hAR. Structural differences between hAR and CtAR in the loop, which mediates the so-called ‘conformational clamping’ in the reaction catalyzed by hAR, may determine the : 250-fold weaker binding of NADP+ by CtAR (see later). In addition, they could arguably be responsible for the observed 100-fold faster overall reduction of aldehyde substrates by the yeast enzyme, by allowing steps involved in coenzyme release to proceed at a higher rate. An interesting difference in the coenzyme-binding sites of mammalian AR and AR from yeast is brought to light by using chemical modification with pyridoxal 5%-phosphate. The aldehyde of pyridoxal 5%-phoshate has been shown to form a Schiff Base with the o-NH2 group of Lys262 of AR from human psoas muscle [22]. The lysine belongs to the coenzyme-interaction consensus sequence (-IPKS-) of aldo/keto reductases. Interestingly, CtAR modified with pyridoxal 5%-phosphate was found to be inactive [9] whereas the modified hAR showed an increase in kcat [22].
2.1.3. Substrate specificity The extremely broad spectrum of aldehyde substrates reduced by the different members of the aldo/keto reductase superfamily does not warrant a detailed discussion of possible differences in enzyme/aldehyde interactions and substrate specificities based on comparative sequence analysis alone. However, a few structural properties of CtAR are interesting. Molecular modeling of aldehyde binding at the active site of hAR [23] suggested that the correctly oriented 2-OH group of the substrate had hydrogen bonds with Ho1 of Trp111. Crystallographic studies suggest an involvement of Trp111 in inhibitor binding [24]. This residue is replaced by Phe114 in CtAR and the phenylalanine is conserved throughout yeast ARs [4] and most hydroxysteroid dehydrogenases [6,17]. Several apolar residues that line the substrate pocket of hAR are replaced by polar residues in CtAR, reflected by a 1.7-fold greater apparent hydrophobicity of the active site of hAR compared with that of the active site of CtAR [25]. Significantly, Val47 of hAR is substituted by an aspartate in CtAR and it seems possible that the replacement affects the microenvironment of the catalytic-acid tyrosine. The C-terminal region of hAR and CtAR show considerable differences in sequence, and it should be noted that this part of the protein has a role to determine the substrate specificity for hydroxylated, hydrophobic or charged aldehydes [26]. A final point is important to be mentioned. In his review on yeast ARs [4] Lee pointed out a seven amino acid-long peptide (Gly130 – 136 in CtAR), which occurs as an insertion at positions Asp125 and Glu126 of hAR whereby the two glycines of CtAR replace the carboxylic acids in hAR. The sequence of the peptide, GFYCGDG in CtAR, is overall reminiscent of
588
B. Nidetzky et al. / Chemico-Biological Interactions 130–132 (2001) 583–595
the Wierenga coenzyme-binding motif of other dehydrogenases e.g. [8], which led Lee to suggest that this part of the sequence could be somehow involved in yeast AR – coenzyme interaction. The structures of the ternary complex of the FR-1 protein, NADPH and the aldose-reductase inhibitor zopolrestat [27], and of the ternary complex of hAR and the same inhibitor [24] have shown that residues 121 – 135 of hAR are inolved in inhibitor binding by the FR-1 protein and hAR. Therefore, a role of the ‘Wierenga-type peptide’ for determining the substrate specificity of CtAR is well possible and must be considered. However, studies of AR from Pichia stipitis have shown that the cystein corresponding to Cys133 in CtAR is susceptible to modification by thiol reagents [4]. The modification was inhibited in the presence of NADPH, possibly suggesting that the cystein and NADPH come close in space in the binary complex.
2.2. Stereochemistry of hydride transfer When (4S)-[4-D]NADD was used as coenzyme for enzymatic reaction with the hydrogen was depleted upon CtAR-catalyzed oxidation of NADD and [4-D]NAD+ was obtained (Fig. 2 A). By contrast, when (4R)-[4-D]NADD was used, the deuterium was depleted at the C4 position of the nicotinamide ring (Fig. 2 B). Therefore, CtAR is specific for transferring the 4-pro-R hydrogen of NADH, which is typical of members of the aldo/keto reductase superfamily. By contrast, short-chain dehydrogenases/reductases show B-type stereospecificity [8]. The stereochemistry of the enzymatic reaction is expected to be a constant trait of members of an enzyme family. Therefore, and by considering the results of sequence comparison, we propose that CtAR is a true member of the aldo/keto reductase superfamily. D-xylose,
2.3. E6idence of direct nucleotide transfer to CtAR from a B-type dehydrogenase/NADH complex Srivastava and Bernhard [29] have shown that direct transfer of NADH from one dehydrogenase to another dehydrogenase can occur when the two enzymes differ in specificity pertaining to the C4H, which is transferred in the reaction. They suggested that transfer of NADH between two dehydrogenases of identical (A-type or B-type) specificity would have to take place via diffusion in aqueous solution. The basis of their experiments was to record the rate of enzymatic oxidation of NADH by a dehydrogenase in the presence of a large molar excess of a binary complex of NADH and another dehydrogenase of A- or B-type specificity, taking care that the concentration of free NADH available for reaction was exceedingly low. When the observed enzymatic rate was decreased relative to that of the control reaction lacking the dehydrogenase/NADH complex, nucleotide exchange between the two dehydrogenases was proposed not to occur by direct transfer. Results for CtAR-catalyzed reduction of D-xylose in the presence of two representative dehydrogenase/NADH complexes are shown in Table 1. The interpretation of the data is clear: direct transfer of NADH to CtAR takes place from a dehydrogenase of
B. Nidetzky et al. / Chemico-Biological Interactions 130–132 (2001) 583–595
589
B-type specificity but not one of A-type specificity. This finding is in excellent agreement with the generalization proposed by Srivastava and Bernhard [29]. Now, since the stereochemistry of yeast xylitol dehydrogenase is A (B.N., unpublished results), direct nucleotide transfer between xylitol dehydrogenase and AR will not occur, although one might intuitively expect it for two successive enzymes in a metabolic pathway. If yeast AR specifically recognized the XDH/NADH complex as coenzyme donor and therefore used NADH with large preference over NADPH in vivo, the first two steps of the xylose pathway would be ‘redox-neutral’, which has been discounted firmly by several physiological studies of D-xylose metabolism in yeast [1]. However, NADH has been shown to serve as in-vivo coenzyme of the AR-catalyzed reduction of D-xylose [30].
Fig. 2. Stereospecificity of hydride transfer catalyzed by CtAR, monitored by 1H-NMR. The products obtained on complete enzymatic oxidation of stereospecifically labeled [28] (A) (4S)-[4-D]NADD and (B) (4R)-[4-D]NADD are shown. The important signal is observed at a chemical shift of about 8.80 ppm. The spectrum in (A) corresponds to [4-D]NAD+, that in (B) corresponds to [4-H]NAD+. Methods: spectra were recorded at 27°C non-spinning in a 5-mm tube at 300.13 MHz with a Bruker Avance 300 spectrometer. All samples were dissolved in D2O (99.75%) and the resulting spectra were referenced to internal DSS (3-(trimethylsilyl)-1-propanesulfonic acid, sodium salt; d =0 ppm). A typical reaction mixture for enzymatic reduction contained 3.5 mg (4R)-[4-D]NADD or (4S)-[4-D]NADD with a deuterium content of ] 98%, 10 U CtAR and 300 mM D-xylose dissolved in 50 mM potassium phosphate buffer pD 7.0. When the oxidation of NADD was complete (monitored spectrophotometrically at 340 nm), the sample was used for NMR analysis without further treatment.
590
B. Nidetzky et al. / Chemico-Biological Interactions 130–132 (2001) 583–595
Table 1 Comparison of the rates of NADH-dependent reduction of absence of a second dehydrogenase in the assay Dehydrogenase used in assay
D-xylose
by CtAR in the presence or
Stereospecificity of the dehydrogenase in the binary enzyme NADH complex
None – Lactate A-type (4-pro-R) dehydrogenase Glutamate B-type (4-pro-S) dehydrogenase
Relative activity of CtAR (%)a 100 491 83 9 4
a A solution of 100 mM D-xylose and 44 mM NADH in 0.3 M phosphate buffer, pH 6.0, was preincubated at 25°C for 0.5 min with a more than 5000-fold excess (: 80 mM) of NADH binding sites of lactate dehydrogenase or glutamate dehydrogenase, relative to NADH-binding sites of CtAR. The reaction was then started by the addition of 0.015 mM CtAR, and the rate of NADH consumption was monitored spectrophotometrically. 9 are S.E., calculated from three independent determinations.
2.4. Steady-state kinetics The kinetic parameters for the reduction of D-xylose by CtAR and hAR are summarized in Table 2. Quite significant differences between the two enzymes are found when comparison is made on the basis of the values of the turnover number (: 85-fold difference), the Km for D-xylose (: 51-fold difference) and the binding of NADP+ ( :250-fold difference). The possible physiological significance of these differences has been discussed recently [11,31]. For CtAR, weak binding of the reactants and at the same time efficient substrate turnover with a high value of kcat makes perfect sense when the key role of the enzyme in the catabolic pathway of D-xylose is considered.
2.5. Substrate specificity 2.5.1. Hydrophobicity By studying the NADH-dependent enzymatic reduction of a series of straightchain aldehydes ranging from propanal to hexanal we have shown recently [31] that Table 2 Comparison of the kinetic parameters of CtAR and hAR for the reduction of
D-xylose
Parameter
CtARa
hARb
kcat (s−1) Km,xylose (mM) KiNAD(P)+ (mM) Ki NAD(P)H (mM)
17.0 76 195 (NAD+), 1.5 (NADP+) 16 (NADH)
0.20 1.5 0.006 0.010
a From [9], experiments carried out at 25°C in 50 mM phosphate buffer, pH 7.0, using NADH as coenzyme. b From [14], experiments carried out at 25°C in 33 mM phosphate buffer, pH 8.0, using NADPH as coenzyme.
B. Nidetzky et al. / Chemico-Biological Interactions 130–132 (2001) 583–595 Table 3 Comparison of the kinetic parameters for reaction of CtARa and hARb with D-glucose, respectively, and derivatives of the parent sugar Aldehyde
CtAR kcat (s−1)
Parent 14.7 2-Deoxy 2-Deoxy-2-fluo 1.0 ro
D-galactose
591
and
hAR Km (mM)
kcat/Km (M−1 s−1)
kcat (min−1)
Km (mM)
kcat/Km (M−1 s−1)
228
66 B0.06 11
15.1 11.1 3.00
12.9 10.4 12.5
19.5 17.8 4
92
From [31], reactions done at 25°C in 50 mM potassium phosphate buffer, pH 7.0, and using 220 mM NADH. b From [32], reactions done at 27°C in 0.1 M sodium phosphate buffer, pH 7.2, using 160 mM NADPH. a
CtAR is able to utilize hydrophobic bonding with nonreacting portions of the substrate to bring about increases in the specificity constant ( : 150-fold) and the turnover number (:2.6-fold). The significant increase in kcat with increasing hydrophobicity of the alkyl chain attached to the reacting aldehyde indicates that the rate of dissociation of NAD+ cannot be fully rate-limiting for the overall reaction catalyzed by the yeast enzyme. If it were, the value of kcat would be expected to be constant across the series of the straight-chain aldehydes. In contrast to CtAR, the turnover numbers for reduction of the same series of aldehydes by hAR showed no variation when the alkyl chain increased in length from two to five carbon atoms [25], in agreement with a kinetic mechanism in which coenzyme dissociation is rate-limiting [14]. Further analyses of the kinetic data have shown that the substrate binding pockets of hAR and CtAR appear to be 2.5 times [25] and 1.4 times [31] more hydrophobic than n-octanol, respectively.
2.5.2. Hydrogen bonding 6ersus inducti6e/field effect in the reaction of CtAR and hAR with sugars Table 3 summarizes kinetic parameters for the enzymatic reduction of aldose sugars and derivatives of these sugars in which the nonreacting hydroxy group at C-2 has been replaced by hydrogen or fluorine. The substitution OH H or OH F has a relatively small effect on the overall structural properties of the sugar such as the proportion of free aldehyde present in aqueous solution, for example (see [32]). The ability of F to weakly accept a hydrogen for bonding which, of course, H cannot (for discussion, see [33]), makes deoxy and deoxyfluoro analogues potentially useful probes to study hydrogen bonding interactions in enzymatic reactions. However, the inductive/field effect of the substituent must be considered in each case and particularly that of the fluorine would seem important, its
B. Nidetzky et al. / Chemico-Biological Interactions 130–132 (2001) 583–595
592
substituent constant (sI) in aliphatic systems being 0.52 [34]! The electron-withdrawing fluorine is expected to affect a stronger polarization of the neighboring carbonyl group, which could arguably facilitate hydride transfer. In addition, the rate of reactions in which there is development of negative charge on an atom in the transition state — the alcoholate oxygen, for example — are known to be increased inductively by a nearby electron-withdrawing group. In summary, therefore, when hydrogen bonding at the 2-OH is most important in the enzymatic reaction, the expected order of reactivity is -OH \ F]H. In contrast, when electronic effects are predominant, the fluoro substrate will be more reactive than the deoxy while its reactivity relative to the OH is difficult to predict. The results shown for hAR are consistent with neither of the two cases. The data for CtAR probably reflect a combination of hydrogen-bonding effects and electronic effects, especially for the reaction with the fluoro substrate. In any case, the use of deoxy and deoxyfluoro sugars reveals marked differences between hAR and CtAR, which it is worth to examine in better detail by using more sophisticated kinetic techniques, e.g. rapid-kinetic and isotope-effect studies [14].
2.5.3. Reaction with three-carbon aldehydes In Table 4 we compare kinetic parameters for hAR- and CtAR-catalyzed reductions of three-carbon aldehydes, which differ mainly in the oxidation state at the a-carbon atom. The results show clearly that the a-carbon atom has an important role for determining the catalytic efficiencies of both enzymes. In the case of hAR the increases in kcat/Km for reactions with acrolein, methylglyoxal, and DL-glyceraldehyde, relative to the catalytic efficiency for reduction of propanal, reflect decreases in Km by about the same factor. Consequently, kcat of hAR is essentially constant across the series of three-carbon aldehydes in Table 4. In the case of CtAR, the observed variations in kcat/Km are distributed between changes in kcat (11-fold) and Km (18-fold). The variation in kcat seen for the yeast enzyme corroborates the above suggestion that noncovalent interactions with the aldehyde can be utilized by CtAR to decrease the activation free energy of the rate-limiting Table 4 Comparison of the kinetic parameters for reaction of CtARa and hARb with three-carbon aldehydes Aldehyde
Propanal Acrolein Methylglyoxal DL-glyceraldehyde a b
CtAR
hAR
kcat (s−1)
Km (mM)
kcat/Km (M−1 s−1)
kcat (min−1)
Km (mM)
kcat/Km (M−1 min−1)
13.0 2.4 26.3 15.5
365 13 6.3 2
36 185 4175 7750
112 87 142 120
1700 80 8 16
1.1×103 1.8×104 3.0×105 1.2×105
Reactions at 25°C in 50 mM potassium phosphate buffer, pH 7.0, and using 220 mM NADH. From [35] at 25°C in 0.1 M sodium phosphate buffer, pH 7.0, using 100 mM NADPH.
B. Nidetzky et al. / Chemico-Biological Interactions 130–132 (2001) 583–595
593
step(s) reflected by the value of kcat. A comparison of the kinetic parameters of CtAR for NADH-dependent reduction of acrolein (which has an electron-withdrawing double bond between C-2 and C-3) and propanal are interesting. The value of kcat for reaction with acrolein is 1/5.5 times that for the reaction with propanal. Yet, the corresponding catalytic efficiency for reaction with propanal is only 1/5 times that for reaction with acrolein. Binding interactions between CtAR and aldehydes may be productive in which case they will be expressed both at the Michaelis complex and the transition state, and lead to an increase in kcat/Km while kcat remains unaffected. Nonproductive binding is expressed at the Michaelis complex but not at the transition state, and its effect will be a decrease in the observed value of kcat by the fraction of productively bound substrate, with kcat/Km remaining unaffected [36]. The results shown in Table 4 could suggest that nonproductive binding occurs to large extent in the reaction of CtAR with acrolein, and reaction to give products takes place only from a relatively small fraction of productive complexes. In the series of aldehydes in Table 4, DL-glyceraldehyde is the best substrate for NADH-dependent reduction by CtAR. Methylglyoxal, which is the most reactive compound among the four aldehydes tested, is the preferred substrate of hAR. Overall, hAR shows catalytic efficiencies for reactions with three-carbon aldehydes that are at least ten times those of CtAR for the same reactions. However, the differences between the two enzymes are especially large (nearly 100-fold) when comparison is made on the basis of the second-order rate constants for the reduction of chemically reactive aldehydes, acrolein and methylglyoxal. Therefore, the kinetic data shown in Table 4 are in good agreement with the suggestion that (i) catalytic efficiency of aldehyde reduction by CtAR is determined to a significant extent by noncovalent bonding with parts of the substrate (e.g. hydroxyls in DL-glyceraldehyde or other sugars) remote from the site of chemical transformation; and (ii) by forming an extremely tight, thus reactive complex with NADPH, hAR is able to reduce a very broad spectrum of aldehyde substrates with catalytic efficiencies that are largely independent on such bonding [11] but seem to mirror the intrinsic chemical reactivities of the aldehydes.
3. Conclusions Comparison of CtAR with the well characterized hAR shows that the yeast enzyme belongs to the aldo/keto reductase superfamily. In the nomenclature of aldo-keto reductases [37], CtAR has been assigned AKR 285. CtAR shares many structural and functional properties with its mammalian counterpart enzyme. Clear differences between the two enzymes pertain to the magnitude of the turnover number and the stabilities of binary enzyme/coenzyme complexes. By comparison with hAR, which is thought to be well adapted for the reduction of structurally diverse and potentially toxic aldehydes [11], it will be interesting to pinpoint the structural features of CtAR which are the key to reductase function with fluxional efficiency in mainstream catabolism of pentose sugars and determine the major ‘kinetic’ differences to hAR.
594
B. Nidetzky et al. / Chemico-Biological Interactions 130–132 (2001) 583–595
Acknowledgements Financial support from the Austrian Science Fund is gratefully acknowledged (grant P-12569-MOB to B.N.).
References [1] P. Mishra, A. Singh, Microbial pentose utilization, in: Advances in Applied Microbiology, vol. 39, Academic Press, San Diego, CA, 1993, pp. 91 – 185. [2] B. Persson, J. Hallborn, M. Walfridsson, B. Hahn-Ha¨gerdal, S. Kera¨nen, M. Penttila¨, H. Jo¨rnvall, Dual relationships of xylitol and alcohol dehydrogenases in families of two protein types, FEBS Lett. 324 (1993) 9–14. [3] R. Lunzer, Y. Mamnun, D. Haltrich, K.D. Kulbe, B. Nidetzky, Structural and functional properties of a yeast xylitol dehydrogenase, a zinc-containing metalloenzyme similar to mediumchain sorbitol dehydrogenases, Biochem. J. 336 (1998) 91 – 99. [4] H. Lee, The structure and function of yeast xylose (aldose) reductases, Yeast 14 (1998) 977 – 984. [5] A. Kuhn, C. van Zyl, A. van Tonder, B. Prior, Purification and partial characterization of an aldo–keto reductase from Saccharomyces cere6isiae, Appl. Environ. Microbiol. 61 (1995) 1580 – 1585. [6] J.M. Jez, M.J. Bennett, B.P. Schlegel, M. Lewis, T.M. Penning, Comparative anatomy of the aldo–keto reductase superfamily, Biochem. J. 326 (1997) 625 – 636. [7] K.M. Bohren, B. Bullock, B. Wermuth, K.H. Gabbay, The aldo – keto reductase superfamily. cDNAs and deduced amino acid sequences of human aldehyde and aldose reductases, J. Biol. Chem. 264 (1989) 9547–9551. [8] H. Jo¨rnvall, B. Persson, M. Krook, S. Atrian, R. Gonzalez-Duarte, J. Jeffery, D. Ghosh, Short-chain dehydrogenases/reductases (SDR), Biochemistry 34 (1995) 6003 – 6013. [9] W. Neuhauser, D. Haltrich, K.D. Kulbe, B. Nidetzky, NAD(P)H-dependent aldose reductase from the xylose-assimilating yeast Candida tenuis, Biochem. J. 326 (1997) 683 – 692. [10] B. Ha¨cker, A. Habenicht, M. Kiess, R. Mattes, Xylose utilisation: cloning and characterisation of the xylose reductase from Candida tenuis, Biol. Chem. 380 (1999) 1395 – 1403. [11] C.E. Grimshaw, Aldose reductase: model for a new paradigm of enzymic perfection in detoxification catalysts, Biochemistry 31 (1992) 10139 – 10145. [12] D.K. Wilson, K.M. Bohren, K.H. Gabbay, F.A. Quiocho, An unlikely sugar substrate site in the 1.65 A structure of the human aldose reductase holoenzyme implicated in diabetic complications, Science 257 (1992) 81–84. [13] K.M. Bohren, C.E. Grimshaw, C.J. Lai, D.H. Harrison, D. Ringe, G.A. Petsko, K.H. Gabbay, Tyrosine-48 is the proton donor and histidine-110 directs substrate stereochemical selectivity in the reduction reaction of human aldose reductase: enzyme kinetics and crystal structure of the Y48H mutant enzyme, Biochemistry 33 (1994) 2021 – 2032. [14] C.E. Grimshaw, K.M. Bohren, C.J. Lai, K.H. Gabbay, Human aldose reductase: rate constants for a mechanism including interconversion of ternary complexes by recombinant wild-type enzyme, Biochemistry 34 (1995) 14356–14365. [15] I. Tarle, D.W. Borhani, D.K. Wilson, F.A. Quiocho, J.M. Petrash, Probing the active site of human aldose reductase. Site-directed mutagenesis of Asp-43, Tyr-48, Lys-77, and His-110, J. Biol. Chem. 268 (1993) 25687–25693. [16] B.P. Schlegel, J.M. Jez, T.M. Penning, Mutagenesis of 3a-hydroxysteroid dehydrogenase reveals a ‘push–pull’ mechanism for proton transfer in aldo – keto reductases, Biochemistry 37 (1998) 3538–3548. [17] T.M. Penning, Molecular endocrinology of hydroxysteroid dehydrogenases, Endocr. Rev. 18 (1997) 281–305.
B. Nidetzky et al. / Chemico-Biological Interactions 130–132 (2001) 583–595
595
[18] J.E. Pawlowski, T.M. Penning, Overexpression and mutagenesis of the cDNA for rat liver 3a-hydroxysteroid/dihydrodiol dehydrogenase. Role of cysteines and tyrosines in catalysis, J. Biol. Chem. 269 (1994) 13502–13510. [19] D.W. Borhani, T.M. Harter, J.M. Petrash, The crystal structure of the aldose reductase.NADPH binary complex, J. Biol. Chem. 267 (1992) 24841 – 24847. [20] S.S. Hoog, J.E. Pawlowski, P.M. Alzari, T.M. Penning, M. Lewis, Three-dimensional structure of rat liver 3a-hydroxysteroid/dihydrodiol dehydrogenase: a member of the aldo – keto reductase superfamily, Proc. Natl. Acad. Sci. 91 (1994) 2517 – 2521. [21] C.E. Grimshaw, K.M. Bohren, C.J. Lai, K.H. Gabbay, Human aldose reductase: subtle effects revealed by rapid kinetic studies of the C298A mutant enzyme, Biochemistry 34 (1995) 14366 – 14373. [22] N.A. Morjana, C. Lyons, T.G. Flynn, Aldose reductase from human psoas muscle. Affinity labeling of an active site lysine by pyridoxal 5%-phosphate and pyridoxal 5%-diphospho-5%-adenosine, J. Biol. Chem. 264 (1989) 2912–2919. [23] H.L. De Winter, M. von Itzstein, Aldose reductase as a target for drug design: molecular modeling calculations on the binding of acyclic sugar substrates to the enzyme, Biochemistry 34 (1995) 8299–8308. [24] D.K. Wilson, I. Tarle, J.M. Petrash, F.A. Quiocho, Refined 1.8 A, structure of human aldose reductase complexed with the potent inhibitor zopolrestat, Proc. Natl. Acad. Sci. USA 90 (1993) 9847–9851. [25] S. Srivastava, S.J. Watowich, J.M. Petrash, S.K. Srivastava, A. Bhatnagar, Structural and kinetic determinants of aldehyde reduction by aldose reductase, Biochemistry 38 (1999) 42 – 54. [26] K.M. Bohren, C.E. Grimshaw, K.H. Gabbay, Catalytic effectiveness of human aldose reductase. Critical role of C-terminal domain, J. Biol. Chem. 267 (1992) 20965 – 20970. [27] D.K. Wilson, T. Nakano, J.M. Petrash, F.A. Quiocho, 1.7 A, structure of FR-1, a fibroblast growth factor-induced member of the aldo–keto reductase family, complexed with coenzyme and inhibitor, Biochemistry 34 (1995) 14323–14330. [28] S.B. Mostad, A. Glasfeld, Using high field NMR to determine dehydrogenase stereospecificity with respect to NADH, J. Chem. Educ. 70 (1993) 504 – 506. [29] D.K. Srivastava, S.A. Bernhard, Mechanism of transfer of reduced nicotinamide adenine dinucleotide among dehydrogenases. Transfer rates and equilibria with enzyme – enzyme complexes, Biochemistry 26 (1987) 1240–1246. [30] N. Meinander, G. Zacchi, B. Hahn-Ha¨gerdal, A heterologous reductase affects the redox balance of recombinant Saccharomyces cere6isiae, Microbiology 142 (1996) 165 – 172. [31] W. Neuhauser, D. Haltrich, K.D. Kulbe, B. Nidetzky, Noncovalent enzyme/substrate interactions in the catalytic mechanism of yeast aldose reductase, Biochemistry 37 (1998) 1116 – 1123. [32] M.E. Scott, R.E. Viola, The use of fluoro- and deoxy-substrate analogs to examine binding specificity and catalysis in the enzymes of the sorbitol pathway, Carbohydr. Res. 313 (1998) 247–253. [33] I.P. Street, C.R. Armstrong, S.G. Withers, Hydrogen bonding and specificity. Fluorodeoxy sugars as probes of hydrogen bonding in the glycogen phosphorylase– glucose complex, Biochemistry 25 (1986) 6021–6027. [34] A.R. Fersht, in: W.H. Freeman (Ed.), Enzyme Structure and Mechanism, second ed., Freeman, New York, 1985, p. 81. [35] D.L. Vander Jagt, N.S. Kolb, T.J. Vander Jagt, J. Chino, F.J. Martinez, L.A. Hunsaker, R.E. Royer, Substrate specificity of human aldose reductase: identification of 4-hydroxynonenal as an endogenous substrate, Biochim. Biophys. Acta 1249 (1995) 117 – 126. [36] W.P. Jencks, Binding energy, specificity, and enzymic catalysis, Adv. Enzymol. 43 (1975) 219 – 410. [37] J.M. Jez, T.G. Flynn, T.M. Penning, A new nomenclature for the aldo-keto reductase superfamily, Biochem. Pharmacol. 54 (1997) 639– 647.
.