Potential Application of N-Carbamoyl-{beta}-Alanine Amidohydrolase from Agrobacterium tumefaciens C58 for {beta}-Amino Acid Production

An N-carbamoyl-β-alanine amidohydrolase of industrial interestfrom Agrobacterium tumefaciens C58 (βcar_At) has been characterized.βcar_At is most active at 30°C and pH 8.0 with N-carbamoyl-β-alanineas a substrate. The purified enzyme is completely inactivatedby the metal-chelating agent 8-hydroxyquinoline-5-sulfonic acid(HQSA), and activity is restored by the addition of divalentmetal ions, such as Mn²⁺, Ni²⁺, and Co²⁺. The native enzymeis a homodimer with a molecular mass of 90 kDa from pH 5.5 to9.0. The enzyme has a broad substrate spectrum and hydrolyzesnonsubstituted N-carbamoyl- ${alpha}$ -, -β-, - ${gamma}$ -, and - ${delta}$ -amino acids,with the greatest catalytic efficiency for N-carbamoyl-β-alanine.βcar_At also recognizes substrate analogues substitutedwith sulfonic and phosphonic acid groups to produce the β-aminoacids taurine and ciliatine, respectively. βcar_At is ableto produce monosubstituted β²- and β³-amino acids,showing better catalytic efficiency (k_cat/K_m) for the productionof the former. For both types of monosubstituted substrates,the enzyme hydrolyzes N-carbamoyl-β-amino acids with ashort aliphatic side chain better than those with aromatic rings.These properties make βcar_At an outstanding candidate forapplication in the biotechnology industry.

INTRODUCTION

Top

INTRODUCTION

N-Carbamoyl-β-alanine amidohydrolase (NCβAA) (EC 3.5.1.6),also known as β-alanine synthase or β-ureidopropionase,catalyzes the third and final step of reductive pyrimidine degradation.In this reaction, N-carbamoyl-β-alanine or N-carbamoyl-β-aminoisobutyricacid is irreversibly hydrolyzed to CO₂, NH₃, and β-alanineor β-aminoisobutyric acid, respectively (43). EukaryoticNCβAAs have been purified from several sources (10, 25,33, 39, 42, 44). Nevertheless, only two prokaryotic NCβAAs,belonging to the Clostridium and Pseudomonas genera (4, 29),have been purified to date, although this activity has beeninferred for several microorganisms due to the appearance ofthe reductive pathway of pyrimidine degradation (38, 45). PseudomonasNCβAA is also able to hydrolyze L-N-carbamoyl- ${alpha}$ -amino acids,and indeed, this activity is widespread in the bacterial kingdom(3, 23, 26, 46).

β-Amino acids have unique pharmacological properties, andtheir utility as building blocks of β-peptides, pharmaceuticalcompounds, and natural products is of growing interest (14).β-Alanine, a natural β-amino acid, is a precursorof coenzyme A and pantothenic acid in bacteria and fungi (vitaminB₅) (7). β-Alanine is widely distributed in the centralnervous systems of vertebrates and is a structural analogueof ${gamma}$ -amino-n-butyric acid and glycine, major inhibitory neurotransmitters,suggesting that it may be involved in synaptic transmissions(20). Another important natural β-amino acid is taurine(2-aminoethanesulfonic acid), which plays an important rolein several essential processes, such as membrane stabilization,osmoregulation, glucose metabolism, antioxidation, and developmentof the central nervous system and the retina (9, 28, 33). 2-Aminoethylphosphonate,the most common naturally occurring phosphonate, also knownas ciliatine, is an important precursor used in the biosynthesisof phosphonolipids, phosphonoproteins, and phosphonoglycans(5). β-Homoalanine (β-aminobutyric acid) has beenused successfully for the design of nonnatural ligands for therapeuticapplication against autoimmune diseases such as rheumatoid arthritis,multiple sclerosis, or autoimmune uveitis (30). Substitutedβ-amino acids can be denominated β², β³, andβ^2,3, depending on the position of the side chain(s) (R)on the amino acid skeleton (18). β²-Amino acids are notyet as readily available as their β³-counterparts, as theymust be prepared using multistep procedures (17).

We decided to characterize NCβAA (β-carbamoylase)from Agrobacterium tumefaciens C58 (βcar_At) after showingthat some dihydropyrimidinases belonging to the Arthrobacterand Sinorhizobium genera are able to hydrolyze different 5-or 6-substituted dihydrouracils to the corresponding N-carbamoyl-β-aminoacids (18, 22). If βcar_At could decarbamoylate the reactionproducts of dihydrouracils, different β-amino acids wouldbe obtained enzymatically in the same way that ${alpha}$ -amino acidsare produced via the hydantoinase process (6, 21). We thereforedescribe the physical, biochemical, kinetic, and substrate specificityproperties of recombinant βcar_At.

MATERIALS AND METHODS

Top

MATERIALS AND METHODS

General protocols and reagents.
Standard methods were used for cloning and expression (1, 31).Restriction enzymes, T4 DNA ligase, and thermostable Pwo polymerasefor PCR were purchased from Roche Diagnostic S.L. (Barcelona,Spain). β-Leucine, β-homoalanine, taurine, ciliatine, ${alpha}$ -amino-β-alanine, and glycine were purchased from Sigma-Aldrich(Madrid, Spain), 5-aminopentanoic acid and 6-aminohexanoic acidwere purchased from Alfa Aesar (Labortecnic, Granada, Spain),β-phenylalanine, 3-aminoisobutyric acid, 4-aminobutyricacid, and β-alanine were purchased from Acros (Labortecnic,Granada, Spain), and 2-amino-3-ureidopropionic acid (Albizziin)was purchased from Bachem (Cymitquimica S.L., Barcelona, Spain).2-Phenyl-3-ureidopropionic acid was synthesized as previouslydescribed elsewhere (35). The rest of the N-carbamoyl-aminoacids used in this work were synthesized according to methodsin the literature (2).

Microbes and culture conditions.
Agrobacterium tumefaciens ATCC 33970 (strain C58) was used asa possible donor of the β-carbamoylase gene. It was cultivatedat 30°C for 20 h in Luria-Bertani medium (LB medium; 1%tryptone, 0.5% yeast extract, 0.5% NaCl, pH 7.2). Escherichiacoli BL21 (36) was used to clone and express the β-carbamoylasegene.

Cloning and sequence analysis of the βcar_At gene.
A DNA region from the Agrobacterium tumefaciens C58 circularchromosome (NC_003062) which was highly similar to the NCβAAgene from Pseudomonas aeruginosa PAO1 (AE004091) was amplifiedby PCR. The primers used for PCR amplification were Atβcar5(5'-AATCTAGAACGGCGGGTAAAAACTTGACGGT-3') and Atβcar3 (5'-TTCTGCAGTCAATGATGATGATGATGATGTTGCACGATCTCCGCAGTC-3').The latter included a polyhistidine tag (His₆ tag) before thestop codon. The XbaI- and PstI-digested fragment was purifiedfrom agarose gel by use of a Qiaquick gel extraction kit (Qiagen)and then ligated into the pBluescript II SK (+) plasmid (StratageneCloning Systems) to create plasmid pAMG4.

Once the fragment had been cloned, it was sequenced using thedye termination dideoxy nucleotide sequencing method in an ABI377 DNA sequencer (Applied Biosystems). Sequencing was carriedout twice from both strands, using standard T3 and T7 primers,edited, and assembled. The assembled sequences were alignedand compared with different NCβAAs of proven activity (10,15, 27, 39).

Site-directed mutagenesis.
Three different substitution mutants of arginine 291 were developed(R291E, R291K, and R291Q mutants). Mutagenesis of the pAMG4plasmid was performed using a QuikChange II site-directed mutagenesiskit from Stratagene following the manufacturer's protocol. Mutationswere confirmed by using the dye termination dideoxy nucleotidesequencing method as described above.

Overexpression and purification of wild-type and mutated βcar_At.
The transformant BL21 strain was grown in LB medium supplementedwith 100 µg ml^–1 of ampicillin. A single colonywas transferred into 10 ml of LB medium with ampicillin at theabove-mentioned concentration in a 100-ml flask. This culturewas incubated overnight at 37°C with shaking. Five hundredmilliliters of LB medium with the appropriate concentrationof ampicillin was inoculated with 5 ml of the overnight culturein a 2-liter flask. After 3 h of incubation at 37°C withvigorous shaking, the optical density at 600 nm (OD₆₀₀) of theresulting culture was 0.3 to 0.5. For induction of β-carbamoylasegene expression, isopropyl-β-D-thiogalactopyranoside (IPTG)was added to a final concentration of 0.2 mM, and incubationwas continued at 34°C for an additional 6 h. The cells werecollected by centrifugation (Beckman JA2-21 rotor; 7,000 x g,4°C, 10 min), washed twice, and resuspended in 50 ml washbuffer (300 mM NaCl, 0.02% NaN₃, 50 mM sodium phosphate, pH7.0). The cell walls were disrupted by sonication, using a UP200 S ultrasonic processor (Hielscher GmbH, Germany), on icefor six periods of 60 s at pulse mode 0.5 and 60% sonic power.The cell debris was precipitated by centrifugation (BeckmanJA2-21 rotor; 10,000 x g, 4°C, 20 min), and the supernatantwas applied to a column with 3 ml of Talon metal-affinity resin(Clontech Laboratories, Inc.). After being washed four timeswith wash buffer, β-carbamoylase enzyme was eluted withelution buffer (100 mM NaCl, 0.02% NaN₃, 300 mM imidazole, 2mM Tris, pH 8.0). The purified enzyme was dialyzed against 100mM sodium phosphate buffer, pH 8.0, and stored at 4°C untiluse.

To ensure that only a single class of metal ions is presentat the active site, an apoenzyme form of βcar_At was preparedby incubating 45 to 50 µM of purified enzyme with 10 mMof 8-hydroxyquinoline-5-sulfonic acid (HQSA) at 4°C overnight.HQSA was removed by dialysis in four stages at 12-hour intervals,all at 4°C, using 100 mM sodium phosphate buffer, pH 8.0.

Enzyme assays.
The standard enzymatic reaction was carried out with purifiedβcar_At (at a final concentration of 0.03 µM) alongwith N-carbamoyl-β-alanine substrate (25 mM) dissolvedin 100 mM sodium phosphate buffer (pH 8.0) in a 500-µlreaction volume. The reaction mixture was incubated at 30°Cfor 20 min, and 25-µl aliquots were stopped by the additionof 475 µl of 1% H₃PO₄. After centrifugation, the resultingsupernatants were analyzed by high-performance liquid chromatography(HPLC). N-Carbamoyl-β-alanine and β-alanine were detectedby an HPLC instrument (Finnigan SpectraSystem HPLC system; Thermo,Madrid, Spain) equipped with a Hypersil-amino acid C₁₈ column(4.6 x 250 mm; Thermo). The mobile phase was NaH₂PO₄ (20 mM;pH 4.5), pumped at a flow rate of 0.75 ml min^–1 and measuredat 200 nm. The specific activity of the enzyme was defined asthe amount of enzyme that catalyzed the formation of 1 µmolof β-amino acid at 30°C min^–1 and mg^–1of protein.

Substrate specificity studies were performed with each differentN-carbamoyl-β-amino acid dissolved in 100 mM sodium phosphatebuffer (pH 8.0) along with the purified enzyme at differentconcentrations (0.03 to 5 µM). Reactions were carriedout at 30°C, with the apoenzyme preincubated (1 h) at 4°Cwith NiCl₂, and stopped by the addition of 1% H₃PO₄. The mobilephase of the different substrates and their corresponding β-aminoacids was methanol-phosphoric acid (20 mM) (5:95 to 30:70 [vol/vol],depending on the compound) at pH 3.2, pumped at a flow rateof 0.75 ml min^–1. Compounds were detected with a UV detectorat a wavelength of 200 nm.

Molecular mass analysis.
Size-exclusion chromatography-HPLC analysis was performed toestimate the molecular mass of the native enzyme, using a nondenaturedprotein molecular weight marker kit (Bio-Rad, Madrid, Spain).The enzyme was eluted in 100 mM citrate, sodium phosphate, Tris,and borate buffers (pH 5.5 and 6.0, pH 6.0 to 8.0, pH 7.0 to8.5, and pH 8.0 to 9.0, respectively) at a flow rate of 0.4ml/min and measured at 280 nm in an HPLC system with a Biosep-SEC-S2000column (Phenomenex, Madrid, Spain). The molecular mass of themonomeric form was estimated by sodium dodecyl sulfate-polyacrylamidegel electrophoresis, using a low-molecular-weight marker kit(GE Healthcare, Barcelona, Spain).

Protein characterization.
The thermal stability of the β-carbamoylase enzyme wasmeasured after 60 min of preincubation at temperatures from25 to 50°C in 100 mM sodium phosphate buffer at pH 8.0.The enzyme assay was then carried out at 30°C for 20 minwith the purified β-carbamoylase enzyme and N-carbamoyl-β-alaninesubstrate. The effects of different compounds on enzyme activitywere evaluated. A 2 mM concentration of each metal (HgCl₂, MgCl₂,NiCl₂, MnCl₂, CoCl₂, CuCl₂, ZnCl₂, CaCl₂, PbCl₂, FeCl₂, FeCl₃,RbCl, CsCl, LiCl, NaCl, and KCl) and 5,5'-dithiobis-(2-nitrobenzoicacid) (DTNB) and a 10 mM concentration of β-mercaptoethanol,dithiothreitol (DTT), EDTA, HQSA, 2-iodoacetamide, bicine, borate,and glycine buffers were incubated with the β-carbamoylase(0.03 µM) in 100 mM sodium phosphate buffer, pH 8.0 (finalvolume, 500 µl), at 4°C for 60 min. The effect ofeach compound on specific activity was determined by standardenzyme assay.

Modeling studies.
The βcar_At model was obtained by the Swiss-Model server(34), using the substrate-bound structure of Saccharomyces kluyveriNCβAA (Skβas) (PDB accession no. 2V8H_A), which hasbeen solved at 2.0 Å, as a template (19). The stereochemicalgeometry of the final model was validated by PROCHECK (16) andWHATCHECK (13), included in the model assessment tools of theSwiss-Model server. Manual model building of the structureswas performed with the Swiss PDB viewer (12) and pymol (8).

Nucleotide sequence accession number.
The nucleotide sequence of the βcar_At gene has been depositedin the GenBank database under accession number EF507843.

RESULTS AND DISCUSSION

Top

RESULTS AND DISCUSSION

Sequence analysis and structure prediction of βcar_At.
The βcar_At amino acid sequence was compared to those ofother NCβAAs of proven activity (Saccharomyces kluyveri,GenBank accession no. AAK60518; Dictyostelium discoideum, GenBankaccession no. AAK60519; Drosophila melanogaster, GenBank accessionno. AAK60520; Homo sapiens, GenBank accession no. NP_057411;Rattus norvegicus, GenBank accession no. Q03248; Arabidopsisthaliana, GenBank accession no. BAB09868) (10, 15, 27, 39).The highest sequence identity was found with that from Saccharomyceskluyveri (Skβas; 36.70%); the amino acid sequence identitywith all other enzymes was <10%. Surprisingly, a BLAST analysisshowed that βcar_At presented higher sequence similarityto several L-N- ${alpha}$ -carbamoylases, among them the previously describedSinorhizobium meliloti L-N- ${alpha}$ -carbamoylase (SmLcar; 79.89%) (23).The same results were previously reported for Skβas (10),which shows a higher sequence identity with bacterial L-N- ${alpha}$ -carbamoylasesthan with mammalian NCβAAs.

Divergent evolution from a common ancestral gene was proposedfor different N-carbamoyl amidohydrolases acting on N-carbamoyl-substitutedcompounds (10). This hypothesis has been supported further bycomputational methods which have shown that several enzymesthat have distantly related primary structures share the samestructural scaffold (20, 24). βcar_At would be includedin the subfamily composed of bacterial L-N-carbamoylases andSkβas. The structural model of βcar_At, determinedusing the known 2.0-Å resolution structure of substrate-boundSkβas (2V8H), clearly shows that the catalytic center iscompletely conserved (Fig. 1). An overall average G factor of–0.07 was recorded (G factor scores should be above –0.5).Ramachandran plot statistics indicated that 97.8% of the main-chaindihedral angles are found in the most favorable regions.

View larger version (38K):
[in this window]
[in a new window]

FIG. 1. Superposition of active sites of closed Skβas (green) bound to its natural substrate (PDB accession no. 2V8H_A) and of a βcar_At model (blue). N-Carbamoyl-β-alanine atoms are shown as follows: red, oxygen; yellow, carbon; and blue, nitrogen.

Overexpression, purification, and subunit structure of βcar_At.
The His₆-tagged enzyme was purified in a one-step procedureusing immobilized cobalt affinity chromatography, with an apparentmolecular mass of approximately 45 kDa under denaturing conditions(estimated purity, >90%) (Fig. 2). Size-exclusion chromatographyshowed that the relative molecular mass of the native enzymefrom pH 5.5 to 9.0 varied from 88 to 100 kDa (data not shown),proving that βcar_At is a homodimer, as reported for otherNCβAAs and L-N-carbamoylases (Skβas [20], NCβAAfrom Pseudomonas putida [29] and SmLcar [23]). However, it variedconsiderably from those from rat (25) and calf liver (41), whichhave a hexameric structure (about 240 kDa), and those from Arabidopsisthaliana and Zea mays, which are decameric (about 440 kDa) (42).

View larger version (120K):
[in this window]
[in a new window]

FIG. 2. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis of each purification step of βcar_At from E. coli BL21 harboring the pAMG4 plasmid. Lane M, low-molecular-weight marker; lane 1, whole recombinant cells before rupture; lanes 2 and 3, supernatant and pellet of the resuspended crude extract after cell sonication; lane 4, eluate after adding the sonicated supernatant to the metal affinity column; lane 5, flowthrough after washing the metal affinity column with buffer; lane 6, purified enzyme.

Influence of pH and temperature on enzyme activity.
βcar_At activity was assayed in several buffers from pH6.0 to 10.5 (bicine, sodium phosphate, Tris, borate-HCl, borate-NaOH,and glycine-NaOH) at a concentration of 100 mM. The enzyme wasactive in sodium phosphate buffer (pH 6.5 to 8.0) and Tris buffer(pH 8.0 to 9.0), with an optimum pH for N-carbamoyl-β-alaninehydrolysis of 8.0. This is similar to those described for P.putida (29) and human liver (34) but slightly higher than thosefor other NCβAAs, which have an optimum pH of 7.5 (10)or 7.0 (25, 41, 42). Thermal stability studies performed onβcar_At have shown that residual activity gradually decreasesat temperatures over 30°C for 60 min and that after incubationat 45°C for 60 min, only 20% of the activity remains (Fig.3). Similar assays have been done only with the P. putida enzyme,which showed high thermostability, retaining 80% of the initialactivity after incubation at 65°C for 30 min. The optimaltemperature of βcar_At was evaluated at temperatures from25 to 60°C, with maximum activity at 30°C (Fig. 3).In view of the above data, it should be borne in mind that themaximum activity may be achieved at higher temperatures than30°C, but exact determination could be hampered by the durationof the activity assays. With the exception of NCβAA fromP. putida (29), which has an optimal temperature of 60°C,the optimal temperature range for hydrolysis of all previouslystudied NCβAA enzymes is 25 to 37°C.

View larger version (8K):
[in this window]
[in a new window]

FIG. 3. Effect of temperature on βcar_At activity. βcar_At activity was measured at different temperatures by a standard enzyme assay (black circles). βcar_At activities after incubation for 60 min at the indicated temperatures are shown (white triangles). The remaining βcar_At activity was measured at 30°C for 20 min, using N-carbamoyl-β-alanine substrate.

Effects of chemical agents and metal ions.
Buffers in which the enzyme had no detectable activity werestudied as possible inhibitors. After preincubation in 10 mMborate, bicine, and glycine, remaining βcar_At activitywas 6, 13, and 22%, respectively, compared to that of the untreatedenzyme. The enzymatic function was not affected by reducingcompounds such as DTT and β-mercaptoethanol or by the sulfhydrylreagents DTNB and iodoacetamide, all at 10 mM, thus showingthat no cysteine residue is crucial for enzyme activity or proteinstability. NCβAAs are described as metalloenzymes (15,20, 42). This also proved true for βcar_At, since the chelatingagents EDTA and HQSA, both at 10 mM, decreased its activitydrastically, to 16% and 0%, respectively. To evaluate the activationor inhibition of βcar_At by different metal ions, its activitywas assayed in the presence of a 2 mM concentration of thesecompounds, using the non-HQSA-treated and apoenzyme forms (Table1). Incubation with Cu²⁺ resulted in strong inhibition, whileZn²⁺ and Hg²⁺ caused only slight inhibition. However, metalions such as Ni²⁺, Mn²⁺, and Co²⁺ greatly enhanced activity.These metals may bind to His86, His193, Asp97, Glu132, and His385,as inferred from sequence comparison and modeling studies. Enzymeinactivation by HQSA was reversible after removal of the chelatingagent by dialysis and addition of Ni²⁺, Mn²⁺, and Co²⁺ at 2mM, but not Zn²⁺ (Table 1). Whether one of these three cationsis the natural cofactor of the enzyme remains unclear. It isworth noting that rat, maize, and S. kluyveri NCβAAs havebeen described as Zn²⁺ dependent (although no other metals havebeen checked) (15, 20, 39). On the other hand, prokaryotic NCβAA(βcar_At and P. putida NCβAA) and L-N-carbamoylaseactivities are optimum with cofactors other than Zn²⁺ (23, 29).The question remains as to whether eukaryotic NCβAAs wouldbe more active with Mn²⁺, Co²⁺, or Ni²⁺ catalytic cofactorsthan with Zn²⁺. Of the three best cofactors detected for βcar_Atunder experimental conditions, Co²⁺, Mn²⁺, and Ni²⁺, we continuedour studies using Ni²⁺. The best metal ion/protein ratio wasfound to be 25:1 when Ni²⁺ was present for at least 20 min at4°C. Under these conditions, maximum activity was maintainedand no inhibition was detected (Fig. 4).

View this table:
[in this window]
[in a new window]

TABLE 1. Effects of metal ions and chemical agents on the activity of βcar_At^a

View larger version (7K):
[in this window]
[in a new window]

FIG. 4. Effect of Ni²⁺ metal ion on βcar_At activity. The reaction mixture contained 100 mM sodium phosphate buffer (pH 8.0), 25 mM N-carbamoyl-β-alanine substrate, 0.03 µM of βcar_At, and several different Ni²⁺ concentrations. Data are the means of three independent experiments and are shown as percentages of the enzyme activity at different Ni²⁺/βcar_At ratios.

Substrate specificity and kinetic characterization.
The ability of purified βcar_At to hydrolyze different substrateswas examined. To this end, the kinetic parameters K_m, k_cat,and k_cat/K_m were obtained from hyperbolic saturation curvesby least-squares fitting of the data to the Michaelis-Mentenequation. Reactions were carried out with different concentrationsof substrates at 30°C after preincubation of the enzymewith NiCl₂.

The involvement and roles of several residues comprising thecatalytic center of L-N-carbamoylases and Skβas have beenproved (19, 24), and they are completely conserved in βcar_At(Fig. 1), in which Arg291 would be the key residue for recognitionof the substrate carboxyl group. To reaffirm this hypothesis,three mutants were developed (R291E, R291K, and R291Q mutants).None of the mutants showed activity using the standard assay,although only the lysine mutant enzyme secondary structure elementswere unaltered when analyzed by far-UV circular dichroism (datanot shown). The R291K mutation suggests how essential this residueis in substrate binding, supporting previous results found forother enzymes (19, 24).

βcar_At is also able to hydrolyze substrate analogues inwhich the carboxyl group is replaced by a sulfonic or phosphonicgroup, although its affinity is somewhat lower for these compounds(Table 2). This is probably due to the different polarizabilityof the carbonyl, phosphonyl, and sulfonyl groups, which wouldaffect their interaction with the Arg291 guanidinium group.

View this table:
[in this window]
[in a new window]

TABLE 2. Kinetic parameters of βcar_At enzyme

By analogy with Skβas, βcar_At catalytic activity comesfrom large-scale movement of the catalytic domain (19, 20).We chose this feature to find out whether βcar_At mightaccommodate longer or shorter N-carbamoyl-amino acids betweenthe lid and the catalytic domains. βcar_At hydrolyzed compoundsfrom N-carbamoyl-glycine (ureidoethanoic acid) to N-carbamoyl- ${delta}$ -aminopentanoicacid (5-ureidopentanoic acid), but not longer substrates, showingthe highest affinity for N-carbamoyl-β-alanine (3-ureidopropanoicacid) (Fig. 5). These results confirm that βcar_At is aureidohydrolase and mainly a β-ureidopropionase or NCβAA.P. putida NCβAA was the first member shown to hydrolyzeN-carbamoyl- ${alpha}$ -, -β-, and - ${gamma}$ -amino acids (29), but to thebest of our knowledge, βcar_At is the first enzyme to hydrolyzeN-carbamoyl- ${delta}$ -amino acids as well.

View larger version (12K):
[in this window]
[in a new window]

FIG. 5. Comparison of nonsubstituted carbamoyl compounds, precursors of ${alpha}$ -, β-, ${gamma}$ -, ${delta}$ -, and ${varepsilon}$ -amino acids, as substrates. Bar ${alpha}$ , N-carbamoyl- ${alpha}$ -glycine (ureidoethanoic acid); bar β, N-carbamoyl-β-alanine (3-ureidopropanoic acid); bar ${gamma}$ , N-carbamoyl- ${gamma}$ -aminobutyric acid (4-ureidobutanoic acid); bar ${delta}$ , N-carbamoyl- ${delta}$ -aminopentanoic acid (5-ureidopentanoic acid); and bar ${varepsilon}$ , N-carbamoyl- ${varepsilon}$ -aminohexanoic acid (6-ureidohexanoic acid). Data, shown as k_cat/K_m, are the means of three independent experiments.

βcar_At was able to produce several monosubstituted β²-and β³-amino acids, showing better catalytic efficiency(k_cat/K_m) for the production of the first class (Table 2). Forexample, the k_cat/K_m ratio for 2-methyl-3-ureidopropionic acidwas 60 times better than that for 3-methyl-3-ureidopropionicacid, decreased drastically with a larger substituent (phenylgroup) for β²-carbamoyl, and was not detected at all forthe β³-counterpart (Table 2). Up to now, β²-aminoacids have not been as readily available as their β³-counterpartsand must be prepared using multistep procedures (17). However,the βcar_At enzyme will simplify their synthesis.

Bearing in mind the flaws that any model has, the relative positionof the substrate in the model provides an idea of the specificityof βcar_At for N-carbamoyl-β²- and -β³-amino acids.The bulkiness of the N-carbamoyl-β²-amino acid side chainscauses a lower affinity for larger groups, and therefore stericclashes might make the largest contribution to affinity loss.A closer inspection of the homology model identifies the sidechains of Trp218 (Trp251 in Skβas) and Ala139 (Ser167 inSkβas) located where the substrate substituent would fit,thus explaining the loss of affinity due to the steric effect(Fig. 6). It is worth noting that the presence of an amino groupinstead of a methyl group in the substrate increased the affinityapproximately 100 times, thus showing that some interactionscaused by the presence of the amino group favor binding. However,this fact cannot be explained by the homology model, as no hydrogenbonding or van der Waals contacts can be inferred.

View larger version (39K):
[in this window]
[in a new window]

FIG. 6. Modeled βcar_At showing the most probable residues involved in steric impediments of 2- and 3-substituted N-carbamoyl-β-amino acids.

Hydrolysis of N-carbamoyl-β³-amino acids was achieved witha methyl group in this position, but not with phenyl or isopropylmoieties. Visual inspection of the βcar_At model suggeststhat it would be impossible to allocate a substituent largerthan a single atom group because of steric clashes, probablywith the Glu132 carbonyl group and/or the catalytic Glu131 sidechain (Fig. 6). Interestingly, although the affinity for the3-methyl-substituted substrate was almost the same as that forN-carbamoyl-β-alanine (3.44 and 2.14 mM, respectively),the k_cat of the enzyme showed a 100-fold decrease for the substitutedcompound. The presence of the methyl substituent might perturbthe catalytic Glu131 carboxyl group environment, thus hinderingits proton-shuttling task for activation of the polarized watermolecule which bridges the enzyme's binuclear metal center.

Our laboratory is currently evaluating the application of βcar_Atin the same context as the "hydantoinase process," along withdifferent dihydropyrimidinases. The broad substrate specificityof this enzyme makes it an attractive tool for the productionof substituted and nonsubstituted β-amino acids.

ACKNOWLEDGMENTS

This work was supported by the Spanish Ministry of Educationand Science through grant BIO2007-67009, by the Spanish Ministryof Education and Science and the European Social Fund throughresearch grant BES-2008-003751, and by Andalusian Regional Councilof Innovation, Science and Technology project P07-CVI-2651.

We thank Andy Taylor and Deborah Fuldauer for critical discussionsof the manuscript and Pedro Madrid-Romero for technical assistance.

FOOTNOTES

* Corresponding author. Mailing address: Departamento de Química-Física, Bioquímica y Química Inorgánica, Edificio CITE I, Universidad de Almería, La Cañada de San Urbano, E-04120 Almería, Spain. Phone: 34 950 015055. Fax: 34 950 015008. E-mail: fjheras{at}ual.es

Published ahead of print on 14 November 2008.

REFERENCES

Top