Biochemical Characterization of β-Lactamases Bla1 and Bla2 from Bacillus anthracis

Abstract

The Sterne and Ames strains of Bacillus anthracis carry chromosomal genes bla1 and bla2, which confer β-lactam resistance when expressed in Escherichia coli. MIC measurements and steady-state kinetic analyses indicate that Bla1 possesses penicillinase activity while Bla2 possesses penicillinase, cephalosporinase, and carbapenem-hydrolyzing activities.

Anthrax, caused by Bacillus anthracis, can be treated with antibiotics such as tetracyclines, macrolides, quinolones, and β-lactams. However, recent studies have shown that naturally occurring B. anthracis isolates can show variable sensitivity to multiple classes of antibiotics (5, 7, 20, 23; C. H. Choe, S. S. Bouhaouala, I. Brook, T. B. Elliot, and G. B. Knudson, Letter, Antimicrob. Agents Chemother. 44:1766, 2000). The Sterne and Ames strains of B. anthracis carry chromosomal genes bla1 and bla2, which are predicted to encode a group 2a penicillinase and a functional group 3 metalloenzyme (6, 8). Bla1 shares 94% amino acid identity with the Bacillus cereus group 2a β-lactamase (19) and contains three functional motifs, S₇₀XXK, S₁₃₀DN, and K₂₃₄SG (ABL numbering) (1, 14), present in group 2 β-lactamases. Bla2 shares 92% amino acid identity with the group 3 B. cereus 569H enzyme (17), shows conservation of the zinc-chelating motif common to group 3 enzymes (15), and is EDTA susceptible (data not shown).

The bla1 and bla2 genes were previously cloned from the B. anthracis Sterne strain and expressed in Escherichia coli (8). However, the β-lactamases were poorly expressed in the original constructs. In this work, a high-level expression system was modified to overexpress these enzymes as ompA-bla1 and ompA-bla2 gene fusion products (25). The gene for mature Bla1 or Bla2 enzyme was fused to the leader sequence of the E. coli OmpA protein, cloned into pET24a(+), and transformed into E. coli BL21Star(DE3). pET24a(+) containsan IPTG (isopropyl-β-d-thiogalactopyranoside)-inducibleT7lacUV5 promoter for large-scale protein expression and a Kan^r marker. Overlap extension PCR was employed as previously described (12) to construct pCMOB1 and pCMOB2. Briefly, four primers, ompA-bla1-bot, ompA-bla1-top, bla1-BamHI, and pET-SphI (Table 1), were used to PCR amplify two DNA fragments with overlapping ends from vectors pET24a(+)-TEM1 (containing the ompA signal sequence) and pUTE523 (containing bla1) (25). The two overlapping PCR products were combined in a single PCR and amplified by using the external primers bla1-BamHI and pET-SphI to generate the fusion product, ompA-bla1. The fusion product was digested with BamHI and XbaI and ligated into pET-24a(+). Ligation products were transformed into E. coli BL21Star(DE3), and plasmid DNA was isolated from a set of transformants. The sequence of the ompA-bla1 fusion gene was determined and confirmed by ABI3100 automated sequencing. The same procedure was used to construct pCMOB2. The four primers, ompA-bla2-bot, ompA-bla2-top, bla2-BamHI, and pET-SphI (Table 1), were used to PCR amplify two DNA fragments with overlapping ends from vectors pET24a(+)-TEM1 and pUTE490 (containing bla2) (25).

TABLE 1.

Primers used in this study

Primer	Nucleotide sequencea
ompA-bla1-bot	5′-CACTTGTCCAGTCTTTTCTTTGGCCTGCGCTACGGTAGCGAAACCAGCCAG-3′
ompA-bla1-top	5′-CGCTACCGTAGCGCAGGCCAAAGAAAAGACTGGACAAGTGAAACATAAAAATC-3′
bla1-BamHI	5′-GGGCGCGGGATCCGAGATTATCACCTAAAAGCATTAACTATAACTTCAGCCG-3′
pET-SphI	5′-GAGCACCGCCGCCGCAAGGAATGGTGCATG-3′
ompA-bla2-bot	5′-TACTTTATGCTCTACCTTTCGTTCGGCCTGCGCTACGGTAGCGAAACCAGCCAG-3′
ompA-bla2-top	5′-CGCTACCGTAGCGCAGGCCGAACGAAAGGTAGAGCATAAAGTAATAAAAAATGAGACAG-3′
bla2-BamHI	5′-GGGCGCGGGATCCTCCTCGCTTTTTTATTTTAACAAATCCAATGTATGTAAAAG-3′

^aThe underlined sequence indicates the BamHI restriction site present in the primer sequence.

The MICs of β-lactam antibiotics were determined by using twofold dilutions in Luria-Bertani-kanamycin broth (LB-Kan; 25 μg/ml) (Table 2). Kanamycin was added to maintain the plasmid. In the absence of IPTG, expression of the cloned genes still occurs; therefore, IPTG was not added to the LB-Kan broth. An inoculum of 10⁵ E. coli BL21Star(DE3) cells/ml containing pCMOB1, pCMOB2, or the pET24a(+) religated vector was used. The MICs were read after overnight incubation at 37°C. Bla1-producing E. coli possessed significant resistance to penicillins but remained susceptible to cephalosporins, imipenem, and aztreonam. Bla2-producing E. coli showed resistance to penicillins and some cephalosporins, reduced susceptibility to imipenem, and susceptibility to aztreonam.

TABLE 2.

Antibiotic susceptibilities of E. coli BL21Star(DE3) expressing B. anthracis Bla1, Bla2, or the pET24a(+) vector

Antibioticb	MIC (μg/ml)a
	pET24a(+)	Bla1	Bla2
Amoxicillin	0.25	≥256	16
Ampicillin	0.125	≥256	8
Benzylpenicillin	2	≥256	16
Carbenicillin	0.25	≥256	64
Cefotaxime	<0.125	<0.125	16
Ceftazidime	<0.125	<0.125	0.5
Cephalothin	0.25	0.5	4
Cefoxitin	1	2	1
Imipenem	<0.125	<0.125	2
Aztreonam	<0.125	<0.125	<0.125

^aThe MIC determinations were performed in duplicate by using the twofold dilution method in LB-Kan broth (25 μg/ml).

^bAntibiotic concentrations ranged from 0.125 to 256 μg/ml.

The Bla1 and Bla2 enzymes were purified to >95% homogeneity by using previously described methods (18) with the following changes: (i) cultures were inoculated into LB-Kan broth (25 μg/ml), induced at a final concentration of 0.1 mM IPTG, and allowed to grow overnight at 25°C; (ii) 50 μM ZnSO₄ was not used for Bla1 protein purification; (iii) enzymes were eluted using a 0 to 1 M NaCl linear gradient; and (iv) Bla2, after collection from cation exchange chromatography, was loaded onto an S-200 Sepharose high-performance liquid chromatography gel permeation column at a flow rate of 1 ml/min, while Bla1 was loaded onto the G-75 Sephadex gel filtration column. Using amino acid analysis, we determined the concentration of Bla2 to be 0.4 mg/ml. The concentration of Bla1 protein was determined by use of the Bradford assay to be 0.5 mg/ml (4). Bla1 migrated at about 31,000 Da and Bla2 migrated at approximately 28,000 Da by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (data not shown). The SignalP version 1.1 program (http://www.cbs.dtu.dk/services/SignalP/) was used to predict the cleavage sites for Bla1 (VEA₁₅-K₁₆EK) and Bla2 (VQA₂₆-E₂₇RK) signal sequences, which were positively confirmed by N-terminal protein sequencing. β-Lactam hydrolysis by Bla1, performed at Johnson & Johnson, Raritan, N.J., was assayed in 1-ml reaction mixtures at 25°C by using a Shimadzu UV-1601 spectrophotometer (26). The concentrated Bla1 enzyme was diluted into 50 mM phosphate buffer (pH 7.0) with 10 μg of bovine serum albumin/ml when enzyme concentrations of ≤0.05 mg/ml were used. Substrates were prepared fresh in 50 mM phosphate buffer (pH 7.0). Substrates were assayed at least twice, with benzylpenicillin included as a daily reference. K_m and V_max values were obtained by averaging the results of a least-squares fit to the Michaelis-Menten equation and Hanes, Eadie-Hofstee, and Cornish-Bowden direct linear plots. Inhibition data for clavulanic acid, tazobactam, and EDTA were obtained by preincubating 0.81 pM Bla1 with various concentrations of inhibitor in 946 μl of phosphate buffer at 25°C for 5 min, adding nitrocefin to a final concentration of 100 μM in 1 ml, and measuring initial hydrolysis rates at 495 nm. The 50% inhibitory concentrations (IC₅₀s) were determined from graphs plotting the percentage of control activity versus inhibitor concentration. Hydrolysis of phenoxymethylpenicillin (16), nitrocefin, imipenem, and cefotaxime (15) by Bla2, performed at the Baylor College of Medicine, Houston, Tex., was measured by using a DU 640 Beckman Coulter spectrophotometer. The reactions were performed in a total volume of 500 μl. The substrate and buffer (50 mM HEPES, 50 μM ZnSO₄ [pH 7.5], 20 μg of bovine serum albumin/ml) were preincubated at 30°C for 5 min. Substrates were prepared fresh in assay buffer. K_m and k_cat values were derived from at least four independent initial-velocity measurements by nonlinear regression using EnzymeKinetics Pro version 2.34. Inhibition data for EDTA were obtained by preincubating 15 nM Bla2 with or without 50 μM EDTA in 110 μl of assay buffer at 25°C for 5 min, and 20 μl of either of these solutions was then added to 81 μM nitrocefin in 120 μl and incubated at 25°C for 5 min or overnight (≥16 h).

Bla1 is a penicillinase with most efficient hydrolysis for benzylpenicillin, ampicillin, amoxicillin, and piperacillin (Table 3). The Bla1 k_cat value of 1,900 s⁻¹ for benzylpenicillin is similar to published values of 1,920 and 2,000 s⁻¹ for the B. cereus β-lactamase I (9, 11). Cephaloridine and nitrocefin demonstrated measurable hydrolysis rates. Bla1 had K_m values of <70 μM for most penicillins, cephaloridine, and nitrocefin. Hydrolysis values of <0.2 s⁻¹ were observed for cefepime, cefotaxime, cefoxitin, cefpodoxime, ceftazidime and ceftriaxone, and imipenem (data not shown). The tazobactam IC₅₀ value for Bla1, 0.086 ± 0.03 μM, was 80-fold lower than the value obtained for clavulanic acid, 6.9 ± 0.8 μM. The IC₅₀ value obtained for EDTA for Bla1 was >250 mM.

TABLE 3.

Hydrolysis parameters of B. anthracis Bla1 β-lactamase

Antibiotic	kcat (s−1)	Relative kcat valuea	Km (μM)	kcat/Km (μM−1 s−1)	Relative kcat/Km valuea
Benzylpenicillin	1,900 ± 120	100	31 ± 2	61	100
Amoxicillin	2,700 ± 330	140	68 ± 10	40	65
Ampicillin	2,600 ± 140	140	50 ± 1	52	85
Carbenicillin	340 ± 15	18	38 ± 2	8.9	15
Cloxacillin	20 ± 5	1.1	NAb	NA	NA
Oxacillin	230 ± 2	12	120 ± 6	1.9	3.1
Piperacillin	660 ± 50	35	13 ± 0.6	51	83
Methicillin	58 ± 1	3.1	19 ± 4	3.1	5.0
Cephaloridine	36 ± 1	1.9	32 ± 4	1.1	1.8
Nitrocefin	42 ± 1	2.2	19 ± 2	2.2	3.6
Aztreonam	≤11	≤0.58	NA	NA	NA

^aRelative k_cat and k_cat/K_m values were calculated with benzylpenicillin as the reference compound.

^bNA, not available. Rates were too low to determine a K_m value.

Table 4 lists the kinetic parameters for β-lactam hydrolysis by Bla2. Bla2 hydrolyzed nitrocefin with values similar to those for the B. cereus metallo-β-lactamase (10, 21, 24), and it efficiently hydrolyzed phenoxymethylpenicillin. The catalytic efficiency (k_cat/K_m) for cefotaxime hydrolysis by Bla2 was about sixfold greater than that of β-lactamase II and B. cereus 569/H/9 metallo-β-lactamases and was similar to that of IMP-1, -3, and -6 metalloenzymes (13, 18, 27). The catalytic efficiency of imipenem hydrolysis by Bla2 was approximately eightfold higher than that of the B. cereus metallo-β-lactamases (10, 21, 24). Bla2 possessed an imipenem catalytic efficiency similar to that of the group 3 enzymes BlaB, FEZ-1, and VIM-2 (2, 3, 22). Some differences in catalytic efficiencies may be due to experimental conditions such as pH and zinc ion concentration. In the absence of EDTA, Bla2 exhibited hydrolysis of chromogenic nitrocefin within 5 min, while in the presence of EDTA, Bla2 showed slight hydrolysis of nitrocefin only after overnight incubation (data not shown); therefore, Bla2 is inhibited by EDTA.

TABLE 4.

Hydrolysis parameters of the purified B. anthracis Bla2

Antibiotic	kcat (s−1)	Km (μM)	kcat/Km (μM−1 s−1)a	Relative kcat/Km valueb
Nitrocefin	313 ± 21	75 ± 5	4.18 ± 0.32	100
Phenoxymethylpenicillin	NAc	>830	4.41 ± 0.12	106
Cefotaxime	NA	>300	6.59 ± 0.62	158
Imipenem	NA	>500	1.51 ± 0.09	36

^aThe k_cat/K_m ratio was calculated, except for nitrocefin, from the initial slope by using the equation v = k_cat/K_m ([E]₀[S]), where [E]₀ is total enzyme concentration and [S] is substrate concentration.

^bRelative k_cat and k_cat/K_m values were calculated with nitrocefin as the reference compound.

^cNA, not available. K_m was too high.

In this work, we characterized the enzymes encoded by the silent chromosomal bla1 and bla2 genes from the nonvirulent B. anthracis Sterne strain. Bla1 preferentially hydrolyzed penicillins and was inhibited by tazobactam and clavulanic acid. Bla2 exhibited carbapenem-, penicillin-, and cephalosporin-hydrolyzing activities and was inhibited by EDTA. Further investigations of bla gene expression in prototypical β-lactam-susceptible B. anthracis strains and in rare β-lactam-resistant isolates will reveal whether increased expression of these genes is associated with resistance to these antibiotics.