|
Advertisement | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
J. Biol. Chem., Vol. 278, Issue 50, 50091-50100, December 12, 2003
Catalysis, Stereochemistry, and Inhibition of Ureidoglycolate Lyase*![]() ![]() From the School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, Georgia 30332-0400
Received for publication, April 11, 2003 , and in revised form, September 5, 2003.
Ureidoglycolate lyase (UGL, EC 4.3.2.3 [EC] ) catalyzes the breakdown of ureidoglycolate to glyoxylate and urea, which is the final step in the catabolic pathway leading from purines to urea. Although the sequence of enzymatic steps was worked out nearly 40 years ago, the stereochemistry of the uric acid degradation pathway and the catalytic properties of UGL have remained very poorly described. We now report the first direct investigation of the absolute stereochemistry of UGL catalysis. Using chiral chromatographic analyses with substrate enantiomers, we demonstrate that UGL catalysis is stereospecific for substrates with the (S)-hydroxyglycine configuration. The first potent competitive inhibitors for UGL are reported here. These inhibitors are compounds which contain a 2,4-dioxocarboxylate moiety, designed to mimic transient species produced during lyase catalysis. The most potent inhibitor, 2,4-dioxo-4-phenylbutanoic acid, exhibits a KI value of 2.2 nM and is therefore among the most potent competitive inhibitors ever reported for a lyase enzyme. New synthetic alternate substrates for UGL, which are acyl- -hydroxyglycine compounds, are described. Based on these alternate substrates, we introduce the first assay method for monitoring UGL activity directly. Finally, we report the first putative primary nucleotide and derived peptide sequence for UGL. This sequence exhibits a high level of similarity to the fumarylacetoacetate hydrolase family of proteins. Close mechanistic similarities can be visualized between the chemistries of ureidoglycolate lyase and fumarylacetoacetate hydrolase catalysis.
Ureidoglycolate lyase (UGL,1 EC 4.3.2.3 [EC] ) and allantoicase (Acase, EC 3.5.3.4 [EC] ) are the enzymes that catalyze the final steps in the catabolic pathway leading from purines to urea. In this pathway (Fig. 1), allantoic acid is formed from allantoin by allantoinase, Acase then converts allantoic acid to ureidoglycolate and urea, and, finally, UGL catalyzes the breakdown of ureidoglycolate to glyoxylate and urea. Acase was first found independently in frog liver by Krebs and Weil (1) and in the mycelium of Aspergillus niger by Brunel (2). UGL was first described by Valentine et al. (3, 4) in Streptococcus allantoicus and in a strain of Pseudomonas. It has since been established that UGL and Acase are both functional in many bacteria, yeast, fish, and rat (516), whereas in some species only one of these activities has been detected (2, 1723).
The stereochemistry of UGL catalysis, and indeed of the uric acid degradation pathway as a whole, has remained poorly described for many years. Conflicting results have been reported as to whether UGL reacts preferentially with ()-ureidoglycolate (5, 6, 9, 10, 18) or with (+)-ureidoglycolate (17). Moreover, allantoinase reacts with both enantiomers of allantoin, and Acase has been reported to produce ureidoglycolate exhibiting a negative rotation, but, paradoxically, to also convert ureidoglycolate exhibiting a positive rotation to glyoxylate and urea (11). Thus, the stereospecificity of UGL catalysis cannot be inferred from the stereochemistries of the preceding enzymes in the pathway. Furthermore, while the effects of pH, temperature, and bivalent cations have been investigated (5, 6, 9, 10, 2528), very little kinetic characterization of UGL has been carried out, and no structural information for UGL has ever been reported. In this paper we report the first direct evidence for the absolute stereochemistry of UGL catalysis. We also introduce the first potent competitive inhibitors for UGL, including what is one of the most potent inhibitors ever reported for a lyase enzyme. Novel substrates are reported here, which make possible the first assay method for monitoring UGL activity directly thus greatly facilitating kinetic and mechanistic investigations. Finally, we report the first primary structural data for UGL, we demonstrate that the derived peptide sequence for UGL exhibits a high level of similarity to that of the fumarylacetoacetate hydrolase family of proteins, and we point out the close mechanistic similarities that can be visualized for the catalytic chemistries of these enzymes.
MaterialsAll buffers, -aminohexyl-agarose, allantoin, ureidoglycolate, allantoate (allantoic acid), phenylhydrazine, and acetopyruvate were from Sigma. Superose 12 HR 10/30 columns and DEAE-Sephacel were from Amersham Biosciences. Phenylacetamide was from Eastman Kodak Co. Benzamide, benzoyl- -hydroxyglycine, cinnamamide, and maleamic acid were from Aldrich. All inorganic salts, general acids, and bases were of reagent grade, and all HPLC grade solvents were of highest quality available. The -hydroxyglycine derivatives and their corresponding amides, along with 2,4-dioxo-4-phenylbutanoic acid and N-Ac-Phe-pyruvate were synthesized as described previously (4042).
MethodsReversed-phase HPLC analyses were performed on a Spherisorb C8 5-micron reversed-phase column (250 x 4.6 mm, Alltech, Deerfield, IL), using a LDC Constametric III system equipped with a LDC Spectromonitor 3100 variable wavelength detector and a Rheodyne 7125 injection valve (20-µl loop). Product quantitation was by standard curve based on peak height. The column was operated at 1.5 ml/min, and the
The continuous spectrophotometric glyoxylate assay (6) was carried out in 150 mM MES-Na buffer, pH 7.0, containing 11 mM phenylhydrazine, various concentrations of ureidoglycolate as substrate, various concentrations of inhibitor, and equal amounts of enzyme in a total volume of 600 µl at 30 °C. The formation of glyoxylate was followed as the change in absorbance at 324 nm for 500 s. The rate of reaction was calculated using the extinction coefficient of 1.7 x 104 l/mol/cm for phenylhydrazone (43). Assays for Acase activity were performed as described by Choi et al. (44). A direct assay for activity toward the
Synthesis of the oxamate affinity column was performed as described by O'Carra and Barry (46), except that the
Isolation of UGL and AcaseBurkholderia cepacia was grown under aerobic conditions on media with allantoin as the carbon source (97% allantoin, 3% yeast extract; adapted from Trijbels and Vogels (17)). UGL was purified from a cell-free extract obtained as the supernatant after sonication (3 by 2 min at 60% amplitude using a Branson 250 Sonifier (Danbury, CT)) of a bacteria cell suspension. The cell-free extract was applied to a DEAE-Sephacel column (2.6 x 30 cm) equilibrated with 50 mM Tris-HCl buffer, pH 6.5. Elution was achieved with a linear NaCl gradient (00.5 M) over 700 min. UGL was eluted from the column by
Acase was also isolated from B. cepacia. Cell-free extract was obtained as described above and loaded onto a DEAE-Sephacel column equilibrated with 50 mM Tris-HCl buffer, pH 6.5, as described for UGL. Acase was eluted from the column by 0.25 M NaCl. The concentrated material was loaded onto the Superose-Superose columns according to the procedure described above. A final pool of purified Acase was collected from the fractions that showed only Acase activity, and the buffer was exchanged to 50 mM HEPES buffer with 0.87 mM MnSO4, pH 7.5. Purified Acase typically exhibit a specific activity of 103.0 units/mg and 33% yield. SDS-PAGE revealed that Acase was not homogeneous but did exhibit a major band at 38 kDa. Kinetic assays confirm that the final pool of purified Acase contains less than 0.3% residual UGL activity. Sequence Analysis of UGLAutomated Edman degradation chemistry was used to determine the N-terminal protein sequence (47). PerkinElmer Life Sciences/Applied Biosystems models 492 and 494 Procise sequencers (Wellesley, MA) were employed for the degradations. The respective PTH-derivatives were identified by reversed-phase HPLC analysis in an on-line fashion employing either a PerkinElmer Life Sciences/Applied Biosystems model 140C PTH analyzer fitted with a PerkinElmer Life Sciences/Brownlee 2.1 mm inner diameter PTH-C18 column or a PerkinElmer Life Sciences/applied Biosystems model 140D PTH analyzer fitted with a PerkinElmer Life Sciences/Brownlee 0.8 mm inner diameter PTH-C18 column. The N-terminal sequence of the purified protein was obtained to 42 positions. Prior to enzymatic digestion, the UGL was desalted by applying it to an Aquapore butyl microbore (2.1 mm inner diameter) C8-RP300, reversed-phase column. Following reduction and alkylation with iodoacetic acid, UGL was digested overnight at 37 °C with trypsin (sequence grade, modified porcine; Promega, Madison, WI) at 1:10. The subsequent digestion mixture was purified on a PerkinElmer Life Sciences/Brownlee C8-RP300 microbore (1.0 mm inner diameter) reversed-phase column, and the chromatography was developed over 90 min using a gradient from 0 to 60% acetonitrile in water, 0.1% trifluoroacetic acid. The flow rate was maintained at 0.05 ml/min. Fragments were collected and characterized by sequence analysis.
Stereochemistry of UGL by Chiral ChromatographyComplete conversions of reactive enantiomers by UGL were performed in 150 mM MES-Na buffer, pH 7.0, containing 7 mM racemic cinnamoyl-
Pure (R)- and (S)-enantiomers were collected from the CHIRALPAK® ADTM column and recovered by evaporation of the mobile phase under vacuum at room temperature. Stereoselective conversion of purified enantiomers by UGL was analyzed by reversed-phase HPLC. Assays were performed in 150 mM MES-Na buffer, pH 7.0, containing purified (R)- or (S)-enantiomer of benzoyl-
A typical purification of UGL from B. cepacia is summarized in Table I. The enzyme obtained from the final gel filtration column is homogeneous as judged by SDS-PAGE (Fig. 2). Automated Edman analysis of the purified UGL provided the N-terminal sequence of the enzyme to 42 amino acids. Sequence information was also obtained for some internal fragments produced by enzymatic digestion using trypsin. Fig. 3, panel A, shows the N-terminal sequence and the sequences of six internal fragments. Fragments 28 and 17 are very short, and their sequences are contained within fragment 30. The ambiguity within fragment 35 precluded its use in protein data base searching.
A computer-based sequence alignment search (TblastN v2.0MP-WashU) of the ongoing genome sequencing project for B. cepacia2 identified significant similarity (E = 1.6 x 1015, 5.1 x 105, 0.0086, and 0.26) between the N-terminal sequence and three of the internal fragments (fragments 30, 22, and 24) with portions of one of the contigs (Bcep825a04.p2n33). As is evident from Fig. 3, panel B, there is a 90100% identity between the N-terminal sequence, the fragments, and portions of the contig. The open reading frame containing our N-terminal sequence and these fragments was identified and is referred to here as "hypothetical UGL" from B. cepacia. An expanded sequence alignment search (BLASTP v2.2.4 NCBI-psiblast (49)) of the GenBankTM non-redundant protein dataset (as of 12/5/02) using hypothetical UGL identified significant similarity (E = 10125 and 10124) with the sequence of a fumarylacetoacetate hydrolase family protein from Brucella suis 1330 (gi:23501151) (50) and a 2-hydroxyhepta-2,4-diene-1,7-dioateisomerase/5-carboxymethyl-2-oxohex-3-ene-1,7-dioatedecarboxylase from Brucella melitensis (gi:17987991) (51) (Fig. 3,Fig. 3, panel C). There is 62% identity between these three proteins. Because of this high level of sequence similarity, hypothetical UGL was used in a search for protein families that share this sequence (HMMER 2.2, Pfam data base as of 12/5/02 (52)). The result was a very good match (E = 6.2 x 1082)tothe fumarylacetoacetate hydrolase family of proteins (pfam0157). This family consists of fumarylacetoacetate hydrolase, and it also includes 2-hydroxyhepta-2,4-diene-1,7-dioateisomerase/5-carboxymethyl-2-oxohex-3-ene-1,7-dioatedecarboxylase from Escherichia coli strain W.
Stereospecificity of UGL CatalysisThe stereospecificity of UGL catalysis was investigated using chiral chromatography. In these experiments we utilized two novel synthetic substrates for UGL-benzoyl- -hydroxyglycine and cinnamoyl- -hydroxyglycine; each of these substrates is enzymatically converted by UGL to the corresponding amide (see below). Fig. 4, panel A, shows the elution profile of racemic benzoyl- -hydroxyglycine on a CHIRALPAK® ADTM column, resolved using 92.5% hexane, 7.5% ethanol, 0.15% trifluoroacetic acid. Baseline separation of the two enantiomers was obtained under these elution conditions. Enzymatic conversion with the enzyme peptidylamidoglycolate lyase (PGL, EC 4.3.2.5
[EC]
), which we have previously shown to react only with the (S)- -hydroxyglycine moiety (40, 5355), was used to establish that the peak eluting at 37.2 min represents the (S)-enantiomer of benzoyl- -hydroxyglycine. After reaction of racemic benzoyl- -hydroxyglycine with UGL until no further enzymatic conversion could be detected, the chromatogram shown in Fig. 4, panel B, was obtained. It is evident from the disappearance of the peak at 37.2 min that UGL reacts preferentially with the (S)-enantiomer. A similar experiment was performed using cinnamoyl- -hydroxyglycine and the results are shown in panels C and D of Fig. 4. Panel C shows base-line separation of the enantiomers of racemic cinnamoyl- -hydroxyglycine obtained on the CHIRALPAK® ADTM column, using 90% acetonitrile, 10% isopropanol, 0.1% trifluoroacetic acid as the eluent. Once again, enzymatic conversion with PGL was used to establish that the peak eluting at 12.0 min represents the (S)-enantiomer. After complete reaction of racemic cinnamoyl- -hydroxyglycine with UGL (panel D), the disappearance of the peak for the (S)-enantiomer again verifies preferential UGL reactivity toward the (S)-hydroxyglycine moiety.
To unequivocally demonstrate the stereospecificity of UGL for the (S)-enantiomers of benzoyl- -hydroxyglycine and cinnamoyl- -hydroxyglycine, and to confirm that the expected corresponding amide products are actually being formed from UGL turnover, the experiments shown in Fig. 5 were carried out. Pure enantiomers of benzoyl- -hydroxyglycine and cinnamoyl- -hydroxyglycine were individually collected from the CHIRALPAK® ADTM column. Fig. 5, panel A, shows chromatograms obtained by reversed-phase HPLC after UGL was incubated with the purified (R)- or (S)-enantiomer of benzoyl- -hydroxyglycine, respectively. Benzoyl- -hydroxyglycine elutes at 3.4 min, and the benzamide product elutes at 4.8 min. Stereospecific catalysis by UGL is evident from the 98% conversion of the (S)-enantiomer, while less than 3% of the (R)-enantiomer is converted to product. A similar experiment was performed using cinnamoyl- -hydroxyglycine, and the results are shown in panel B of Fig. 5; in the chromatograms in panel B cinnamoyl- -hydroxyglycine elutes at 5.4 min and the cinnamamide product elutes at 8.2 min. Stereospecific catalysis by UGL is again evident from the 97% conversion of the (S)-enantiomer, while less than 8% of the (R)-enantiomer is converted.
New Alternate Substrates for UGLThe natural substrates for UGL and Acase are ureidoglycolate and allantoate, respectively. We investigated the reactivities of a number of acyl- -hydroxyglycine compounds as potential alternate substrates for these enzymes. We anticipated that appropriately designed alternate substrates would enable facile direct assay of these enzymes by HPLC, thus obviating the need for the coupled assays used by previous investigators (2939). In addition, reactive alternate substrates are very useful as templates for the design of potent new competitive inhibitors.
Listed in Table II are the results obtained for the reactions of several acyl-
The ability of benzoyl- -hydroxyglycine to act as a competitive inhibitor for UGL-catalyzed conversion of ureidoglycolate was also investigated, since the latter is a much more reactive substrate. If the Km for benzoyl- -hydroxyglycine is largely a measure of binding affinity, its numerical value should be similar to measured KI for benzoyl- -hydroxyglycine inhibition of ureidoglycolate turnover. Benzoyl- -hydroxyglycine inhibition of ureidoglycolate turnover was found to be purely competitive by (Dixon plots; EnzymeKinetics, Trinity Software), and as shown in Table III, the measured KI value is quite similar to the Km for benzoyl- -hydroxyglycine itself.
N-Ac-L-Phe-
New Inhibitors for UGL and AcaseOur laboratory recently reported that 2,4-diketo-5-acetamido-alkanoic acids are the first potent competitive inhibitors for peptidylamidoglycolate lyase (41). The 2,4-dioxocarboxylate moiety in these compounds was designed to mimic the transient species produced during lyase catalysis in which ketonization at the
Kinetic experiments established that benzoylpyruvate is indeed a very potent inhibitor of UGL, and Dixon plots confirmed that inhibition is purely competitive. As shown in Table IV, the KI value for benzoylpyruvate inhibition is 2.2 nM. Thus, benzoylpyruvate is one of the most potent inhibitor ever reported for a lyase enzyme (see "Discussion"). It is also evident from Table IV that acetopyruvate and N-Ac-Phe-pyruvate, both of which possess a 2,4-dioxocarboxylate moiety, are also potent inhibitors for UGL. In acetopyruvate, a simple methyl substituent is appended to the 2,4-dioxocarboxylate moiety, whereas a branched
The pyruvate compounds are much less potent inhibitors for Acase (Table IV), although acetoacetate (KI = 16 µM) is certainly sufficiently potent to be potentially useful in future mechanistic and structural studies on this enzyme. A two-step reaction mechanism for hydrolysis of allantoate, in which allantuate (the imino analog of ureidoglycolate) is an intermediate, has been proposed (24, 58). Maleamic acid, the carbon analog of allantuate, can obviously not undergo Acase-catalyzed breakdown, and we therefore reasoned that it might act as a potent competitive inhibitor for Acase. As listed in Table IV, maleamic acid exhibited competitive inhibition toward Acase with a KI in the millimolar range and is thus an extremely poor inhibitor.
Several aspects of the stereochemistry of the uric acid degradation pathway have remained unresolved for many years. Allantoin is the first metabolite formed in the pathway, and this molecule possesses a chiral carbon atom. However, the enzyme allantoinase, which then converts allantoin to allantoic acid, is not stereospecific, and reacts with both enantiomers of allantoin. A second complication is that the next enzyme in the pathway, allantoicase, has been reported to react with allantoic acid to produce ureidoglycolate exhibiting a negative rotation but, paradoxically, has also been reported to convert ureidoglycolate exhibiting a positive rotation to glyoxylate and urea (11). Third, conflicting results have been reported as to whether the next enzyme, UGL, reacts preferentially with ()-ureidoglycolate (5, 6, 9, 10, 18) or with (+)-ureidoglycolate (17) to produce the final products, glyoxylate and urea. Clearly, definitive stereochemical studies that are based on determinations of absolute configurations have long been needed for these enzymes to confidently establish the stereochemistry of the uric acid degradation pathway. The results presented here demonstrate clearly that UGL reacts only with substrates possessing an (S)-hydroxyglycine moiety. This constitutes the first direct evidence for the absolute stereochemistry and stereospecificity of UGL catalysis. Previous investigations of the stereochemistry of UGL and the other enzymes in the uric acid degradation pathway have been based solely on optical rotations, and these studies did not establish stereospecificity as opposed to preferential reactivity of one of the enantiomers (stereoselectivity). We note that in addition to elucidating the stereospecificity of an important step in uric acid degradation, our stereochemical results may be of interest from a biotechnological perspective, since lyases have considerable potential for the biocatalytic processing of chiral compounds for industrial or pharmaceutical applications (59).
The acyl-
We introduce here the first potent competitive inhibitors for UGL, including what is one of the most potent inhibitors ever reported for a lyase enzyme. Our design rationale was based on our present finding that benzoyl-
Although a number of reversible inhibitors for other lyases have been reported in the literature, very few of these exhibit KI values in the low nM range. The hydroxylamine analog of Phe,
An ever increasing body of sequence and structural information is surfacing, and when carrying out comparisons between protein families it is often found that catalytic groups can have distinctive patterns of variations. Furthermore, structural relationships are often found between protein families that have very different catalytic functions (65). The results reported here indicate that UGL has very high sequence similarity to the fumarylacetoacetate hydrolase family of enzymes. As illustrated in Fig. 6, close mechanistic similarities can be visualized between the chemistries of UGL and fumarylacetoacetate hydrolase catalysis. Based on their crystal structures of fumarylacetoacetate hydrolase and of its complexes with products and with a competitive inhibitor, Timm and co-workers (48, 66) have proposed that the enzymatic cleavage of the CC bond in fumarylacetoacetate catalysis proceeds via initial nucleophilic attack on the carbonyl carbon to form a tetrahedral alkoxy intermediate, which then breaks down via ketonization to yield the products fumarate and the enolate of acetoacetate. This mechanism is conceptually very analogous to the mechanism which we have proposed for amidoglycolate lyases (41), which, as illustrated in Fig. 6, entails ketonization at the
As described in detail by Timm and co-workers (48, 66), the active site of fumarylacetoacetate hydrolase is composed of a catalytic triad consisting of a glutamate carboxylate that stabilizes a histidine base that directly activates a nucleophilic water molecule. Also present is an oxyanion hole, which is positioned to stabilize the proposed tetrahedral alkoxy intermediate, a Ca2+ ion that likely interacts with substrate and product, and a Mg2+ ion. As discussed by Timm et al. (66), several features of this active site topography are quite distinct from those of other "catalytic triad-containing" hydrolases, and this is likely a reflection of the fact that fumarylacetoacetate hydrolase carries out cleavage of a relatively unreactive carbon-carbon bond. Whether or not the sequence and mechanistic similarities between UGL and fumarylacetoacetate hydrolase also extends to the active site structure of UGL remains to be determined.
* This work was supported by National Institutes of Health Grant GM40540. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 The abbreviations used are: UGL, ureidoglycolate lyase (EC 4.3.2.3
[EC]
); Acase, allantoicase (EC 3.5.3.4
[EC]
); PGL, peptidylamidoglycolate lyase (EC 4.3.2.5
[EC]
); HPLC, high performance liquid chromatography; MES-Na, sodium 2-(N-morpholino)ethanesulfonate; contig, group of overlapping clones.
2 B. cepacia Sequencing Project (as of 12/3/02), Sanger Centre, Hinxton, UK, www.sanger.ac.uk/Projects/B_cepacia/
We thank Dr. Greta Olson for initial work with bacterial growth conditions and enzyme assays and Kis Robertson for measuring the relative activity of Acase. We thank Drs. Michael G. Jennings and Christine Smith of Monsanto (St. Louis, MO) for their help in obtaining the N-terminal and internal protein sequence information for UGL. We thank the B. cepacia Sequencing Group, a project funded by a grant from Beowulf Genomics, at the Sanger Institute for allowing access to the genomic sequence data as it became available. We thank Drs. Charlie D. Oldham and Allison B. Moore for many technical discussions and recommendations.
This article has been cited by other articles:
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Advertisement | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||