Bilirubin Oxidase Activity of Bacillus subtilis CotA

The spore coat protein CotA from Bacillus subtilis was previouslyidentified as a laccase. We have now found that CotA also showsstrong bilirubin oxidase activity and markedly higher affinityfor bilirubin than conventional bilirubin oxidase. This is thefirst characterization of bilirubin oxidase activity in a bacterialprotein.

INTRODUCTION

Top

Introduction

Bilirubin oxidase (BOD; EC 1.3.3.5) catalyzes the oxidationof bilirubin to biliverdin (2) and is used clinically to determinethe levels of total and conjugated bilirubin in serum (3, 12,14). It is distributed mainly in fungi, and BODs from Myrotheciumverrucaria (20) and Trachyderma tsunodae K-2593 (6) have beenpurified and characterized. BOD is thought to be a member ofthe multicopper oxidase family because it has been shown tobind a type I (blue) copper ion, a paramagnetic (electron paramagneticresonance) type II copper ion, and a pair of diamagnetic typeIII copper ions (6, 16). Moreover, strong sequence identitybetween BOD and the other multicopper oxidases, including laccase(EC 1.10.3.2), ascorbate oxidase (EC 1.10.3.3), and ceruloplasmin(EC 1.16.3.1), particularly at potential copper coordinationsites (8) (Fig. 1), has been widely reported.

View larger version (15K):
[in this window]
[in a new window]
FIG. 1. Comparison of the amino acid sequences of potential copper coordination sites in multicopper oxidases. CotA, B. subtilis CotA (NCBI entry NP_388511); BOD, M. verrucaria BOD (NCBI entry Q12737); CLA, Coprinus cinereus laccase (NCBI entry 1HFUA); TLA, Trametes villosa laccase (NCBI entry AAB47735); MLA, Melanocarpus albomyces laccase (NCBI entry CAE00180). The numbers 1, 2, and 3 indicate the potential coordination sites for type 1, 2, and 3 copper ions, respectively. The amino acid residues presumed to be involved in binding copper are boxed. The amino acid residues of CotA closely related to those of BOD are indicated by arrowheads.

On the other hand, BOD differs from other multicopper oxidasesin that it catalyzes reactions involving substances comprisedof tetrapyrroles, especially bilirubin (8, 21). It is knownthat BOD oxidizes some laccase substrates (20, 22); however,laccase, ascorbate oxidase, and ceruloplasmin show little orno BOD activity (1, 21), though one exception is the alkaliphiliclaccase from Myrothecium verrucaria 24G-4 (18). It has beenproposed that these differences in substrate specificity reflectthe heterogeneity of the amino acid sequences within the consensusdomains (8). X-ray crystallographic analysis of the structureof Coprinus cinereus laccase showed that the type I copper centerof the enzyme contains His396, Cys452, and His457 as the copperligands and a nonligating Leu462 at a position axial to theligand (4). In addition, Kumar et al. (11) reported that most,but not all, fungal laccases carry a Leu or a Phe at this axialposition 10 residues downstream from Cys452 (Fig. 1), thoughseveral other multicopper oxidases, including BOD, feature aMet at this position. Finally, Koikeda et al. (8) showed thatLeu442 and Asp504, within the C-terminal consensus domain ofM. verrucaria BOD, are not present in the other multicopperoxidases (Fig. 1).

Recently, the Bacillus subtilis endospore component CotA wasidentified as a bacterial laccase (7, 13), and the crystal structureof the protein was solved (5). This is the first bacterial laccasethat has been both functionally and structurally characterizedin detail. CotA shows relatively high (19.7 to 22.4%) aminoacid sequence identity with multicopper oxidases (13) and containsthe four consensus copper-binding sites (17). We also analyzedthe consensus domains of CotA and found that the amino acidsequences around the copper coordination sites are closely relatedto those of BOD in terms of the aforementioned features; thatis, CotA has a Met at the position corresponding to Leu462 inC. cinereus laccase and Leu/Asp at the positions correspondingto the Leu442/Asp504 in M. verrucaria BOD (Fig. 1). Notably,when we expressed the CotA gene in Escherichia coli and examinedcharacteristics of the product, we found that in addition tothe previously identified laccase activity, the enzyme alsoshows strong BOD activity. Here we describe the characteristicsof this first identified bacterial BOD. In addition, we describeour application of this enzyme for the determination of bilirubinlevels in human serum.

Production and purification of recombinant CotA.

Top

Production and Purification of...

It was previously reported that recombinant CotA protein wasmainly produced as a precipitate when pET21a(+) and E. coliAH3517 were used for the gene expression and that attempts torecover the soluble protein from the precipitate were unsuccessful(13). This means that the gene product was produced as an inclusionbody in the cell extract. To improve the protein's solubility,we evaluated its production by use of several combinations ofexpression vectors, host cells, and cultivation conditions.We found that E. coli JM109 transformed with pUC119 carryingthe CotA gene produces large amounts of the soluble proteinunder low-temperature conditions. For construction of the expressionvector, the oligonucleotide primers used to amplify the CotAgene fragment by PCR were 5'-ATAAAAGCTTATGAAATGACACTTGAAAAATT-3',which contains a unique HindIII restriction site overlappingthe 5' initial codon (italics), and 5'-TCATGTGGATCCTGTGTGAGCATAAAAGCTCC-3',which contains a unique BamHI restriction site proximal to the3' end of the termination codon (italics). Chromosomal B. subtilisDNA isolated as described previously (15) served as the template.The amplified 1.5 kbp fragment was digested with HindIII andBamHI and ligated with pUC119 vector linearized with the samerestriction enzymes to generate pUC119BOX8. E. coli JM109 wasthen transformed with pUC119BOX8 and cultivated in Luria-Bertanimedium containing 0.25 mM CuSO₄ and 50 µg/ml ampicillin.With this protocol, the yield of soluble CotA was dependenton the cultivation temperature (Fig. 2). At 21°C there wasvery little production, and beyond 31°C soluble proteinaccounted for less than 5% of the total CotA protein. But at24°C approximately 50% of the CotA protein was detectedin the soluble fraction.

View larger version (87K):
[in this window]
[in a new window]
FIG. 2. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of B. subtilis CotA heterologously produced in E. coli. Insoluble fractions of crude extract from pUC119BOX8 cultivated at 21, 24, and 31°C were loaded onto lanes 1, 3, and 5, respectively; the soluble fractions were loaded onto lanes 2, 4, and 6, respectively. The arrow indicates the position of CotA.

So cultivation of the recombinant E. coli was carried out at24°C for 20 h, after which the cells were harvested by centrifugation,suspended in 10 mM potassium phosphate buffer (pH 7.0) containing1 mM CuSO₄, and disrupted by ultrasonication. The entire operationwas done at room temperature ( $~$ 25°C), and the fractions containingCotA were checked by sodium dodecyl sulfate-polyacrylamide gelelectrophoresis at each purification step. The crude extractwas heated at 65°C for 40 min, and the denatured proteinwas then removed by centrifugation. The remaining protein solutionwas then loaded onto a Q Sepharose Big Beads column (AmershamBiosciences) (50 by 300 mm) equilibrated with 10 mM potassiumphosphate buffer (pH 7.0), and after the column was washed withthe same buffer, the protein was eluted with a linear gradientof 0 to 0.5 M KCl in the same buffer. The fractions containingCotA were collected, and solid (NH₄)₂SO₄ was added to achievea 25% (wt/vol) concentration. The protein solution was thenloaded onto a phenyl Sepharose 6 Fast Flow column (AmershamBiosciences) (50 by 300 mm) equilibrated with 10 mM potassiumphosphate buffer (pH 7.0) containing 25% (NH₄)₂SO₄, and afterthe column was washed with the same buffer, the protein waseluted with a linear gradient of 25% to 0% (NH₄)₂SO₄ in thesame buffer. The CotA-containing fractions were collected anddialyzed against 10 mM potassium phosphate buffer (pH 7.0),after which the protein solution was loaded onto a DEAE SepharoseFast Flow column (Amersham Biosciences) (50 by 300 mm) equilibratedwith 10 mM potassium phosphate buffer (pH 7.0). After the columnwas washed with the same buffer, the protein was eluted witha linear gradient of 0 to 0.5 M NaCl in the same buffer. Thefractions containing CotA were pooled, concentrated by ultrafiltration, and used as the purified preparation. Because theconcentrated enzyme (19 mg/ml) was easily precipitated whenNaCl was removed from the buffer by dialysis, we ordinarilyused buffer containing 0.6 M NaCl as the enzyme solution. About2 g of purified enzyme was prepared from the cells obtainedfrom a 38-liter culture.

Activities and enzymatic properties.

Top

Activities and Enzymatic...

Conjugated bilirubin is known to be a better indicator of liverfunction than total bilirubin (3, 12, 14), and ditaurobilirubinis widely utilized as the standard for conjugated bilirubin.For that reason, we assayed the enzyme activity of the isolatedCotA by use of both bilirubin and ditaurobilirubin (solubilizedmaterials; Interference Check A) (Sysmex, Kobe, Japan) as substrates.The oxidation of bilirubin or ditaurobilirubin at 37°C wasmonitored spectrophotometrically at 450 nm ( ${varepsilon}$ ₄₅₀ = 32 and 74mM^–1 cm^–1 for bilirubin and ditaurobilirubin, respectively).Unless otherwise specified, the reaction mixture contained 100mM potassium phosphate (pH 6.0), 20 µg of substrate, and1 to 3 mU of enzyme in a total volume of 1 ml. The reactionwas started by the addition of the enzyme. The optimum pH ofthe enzyme was determined by running the standard assay at 37°Cin the following 40 mM buffer systems: glycine-HCl (pH 2.6 to3.4), acetate-NaOH (pH 3.4 to 5.9), citrate-NaOH (pH 5.9 to6.8), potassium phosphate (pH 6.3 to 7.4), Tris-HCl (pH 7.4to 8.7), and glycine-NaOH (pH 8.4 to 11.0). The optimum pH forbilirubin oxidation was found to be about pH 7.0 whereas thatfor ditaurobilirubin oxidation was about pH 4.0 (Fig. 3A). M.verrucaria BOD also reportedly oxidizes ditaurobilirubin atacidic pH, which proved useful for selective measurement ofconjugated bilirubin at about pH 5.5 (10). Thus, the large differencein the optimum pHs for CotA-catalyzed oxidation of bilirubinand ditaurobilirubin may be advantageous for selective determinationof conjugated bilirubin levels in serum.

View larger version (18K):
[in this window]
[in a new window]
FIG. 3. Optimum pH and thermostability. (A) Effects of pH on CotA-catalyzed oxidation of bilirubin (open circles) and ditaurobilirubin (closed circles). (B) Residual BOD activity after incubation at the indicated temperatures for 30 min. (C) Effect of incubation time at 65°C on BOD activity. In panels B and C, open circles and closed circles represent B. subtilis CotA and T. tsunodae K-2593 BOD, respectively.

We next compared the thermostability of CotA with that of T.tsunodae K-2593 BOD (Takara Shuzo Co., Ltd.) (Fig. 3B and C)and, as reported previously (13), found CotA to be highly thermostable.To determine its thermostability, the enzyme (10 mg/ml) wasincubated in 50 mM potassium phosphate buffer (pH 7.0) containing0.1% Triton X-100 at various temperatures, and the residualactivity was determined using the standard assay described above.We found that CotA had lost none of its BOD activity after beingheated at 84°C for 30 min, whereas the activity of the T.tsunodae enzyme was completely lost (Fig. 3B). CotA also retainedits full activity after heating at 65°C for at least 60min, whereas T. tsunodae BOD showed a half-life of inactivationof about 40 min at the same temperature (Fig. 3C). Thermostabilityhas so far been reported only for M. verrucaria BOD, which rapidlyinactivates at 70°C (20). CotA thus appears to be the mostthermostable BOD identified to date.

The Michaelis constants were determined from double-reciprocalplots of the initial oxidation rates and concentrations of bilirubinor ditaurobilirubin at 37°C and pH 7.0. CotA showed typicalMichaelis-Menten kinetics for both bilirubin and ditaurobilirubin.With bilirubin as the substrate, the apparent K_m and V_max valueswere 0.0080 mM and 28 µmol/min/mg, respectively. Withditaurobilirubin as the substrate, these constants were 0.015mM and 10 µmol/min/mg, respectively. This observed K_mvalue of CotA for bilirubin is extremely low, and the V_max/K_mratio is more than 20 times higher than that obtained with M.verrucaria BOD (Table 1), which is frequently utilized for determiningserum bilirubin levels.

View this table:
[in this window]
[in a new window]
TABLE 1. Kinetic constants for bilirubin and ditaurobilirubin

Serum bilirubin assay.

Top

Serum bilirubin assay.

We assayed total bilirubin in serum by use of CotA accordingto the method described by Otsuji et al. (14) and compared theresults to those obtained using the commercially available IatroT-Bil kit (Mitsubishi Kagaku Iatron, Inc., Tokyo, Japan). Serawere obtained from 51 healthy individuals and from 19 patientswith hepatobiliary disease. Figure 4 shows the correlation betweenthe results obtained with the two assays. The correlation coefficientwas 0.994 over a wide range of measurements, and the slope ofthe linear regression equation was approximately 1.08, showingthat CotA could be used for accurate determination of serumbilirubin levels.

View larger version (14K):
[in this window]
[in a new window]
FIG. 4. Correlation of total bilirubin levels measured in 70 samples by use of CotA (x axis) and an Iatro T-Bil kit (y axis). The regression line was defined by the equation Y = 1.08X + 0.27; R² = 0.994.

The genes encoding laccases have been cloned for a number ofspecies, but in most cases the proteins are produced at levelstoo low to be used for commercial purposes (9, 19). In the presentstudy, however, we were able to develop an effective protocolfor the high-level production of soluble CotA in E. coli. Theisolated enzyme showed markedly higher affinity for bilirubinthan conventional BOD and also exhibited greater stability.CotA thus shows great potential for use as a diagnostic reagentand may serve as the basis for a useful and practical methodof bilirubin determination.

A possible link between laccase activity of CotA and productionof a melanin-like pigment which appears to protect spores againstUV light or hydrogen peroxide has been suggested (5, 13). Althoughthe physiological significance of the BOD activity is currentlyunknown, the significantly high affinity for bilirubin showsthat CotA may have additional biological function beyond thelaccase activity.

FOOTNOTES

* Corresponding author. Mailing address: Asahi Kasei Pharma Corporation, Shizuoka 410-2321, Japan. Phone: 81 (0) 558 76 8564. Fax: 81 (0) 558 76 7149. E-mail: sakasegawa.sb{at}om.asahi-kasei.co.jp

REFERENCES

Top