Functional Characterization of OXA-57, a Class D β-Lactamase from Burkholderia pseudomallei

Abstract

Class D β-lactamase OXA-57 was identified in a range of isolates of Burkholderia pseudomallei and Burkholderia thailandensis. Comparative kinetic analyses of wild-type and mutant forms of B. pseudomallei OXA-57 are reported. Implications of these data for β-lactam resistance and the proposed role of Ser-104 in β-lactam hydrolysis are discussed.

The gram-negative bacterium Burkholderia pseudomallei causes the fatal disease melioidosis. Resistance to ceftazidime correlates with β-lactamase production by B. pseudomallei (3), and the genome contains of a number of β-lactamases (6). In this study we provide some insights into β-lactam resistance and class D β-lactamase activity in B. pseudomallei.

An ∼730-bp product was obtained upon PCR amplification of B. pseudomallei strain 576 (Table 1) genomic DNA with primers (Table 2) specific for the class D β-lactamase gene. Nucleotide sequencing confirmed that this PCR product corresponds to the oxa57 open reading frame lacking the predicted signal sequence. The predicted amino acid sequence of OXA-57 is aligned with the OXA-59 homologue from B. pseudomallei strain K96243 (6) (Fig. 1). The OXA-57 and OXA-59 coding sequences are identical except for a single A-to-G base change, resulting in changing residue 170 from Asp to Asn. Sequence analysis (Fig. 1) reveals the three active-site elements common to all class D β-lactamases (11): S-X-X-K (residues 53 to 56), S-X-V (residues 104 to 106), and K-T/S-G (residues 201 to 203).

TABLE 1.

Bacterial strains screened in this study

Strain	Comments	Sourcea
B. pseudomallei
K96243	Clinical isolate, genome sequence strain	S. Songsivalai, Mahidol University, Nakhon Pathan, Thailand
576	Clinical isolate	Ty Pitt, HPA, London
4845	Clinical isolate	NCTC
E8	Environmental isolate	Ty Pitt, HPA, London
204	Clinical isolate	Ty Pitt, HPA, London
E25	Environmental isolate	Ty Pitt, HPA, London
708A	Gentamicin sensitive	David Dance, HPA, Plymouth
E38	Environmental isolate	David Dance, HPA, Plymouth
Mal6	Origin unknown	David Dance, HPA, Plymouth
B. thailandensis
E111	Environmental isolate	Ty Pitt, HPA, London
E125	Environmental isolate	Ty Pitt, HPA, London
E251	Environmental isolate	Ty Pitt, HPA, London
E255	Environmental isolate	Ty Pitt, HPA, London
E135	Environmental isolate	Ty Pitt, HPA, London
E254	Environmental isolate	Ty Pitt, HPA, London
E30	Environmental isolate	Ty Pitt, HPA, London
E132	Environmental isolate	Ty Pitt, HPA, London
E253	Environmental isolate	Ty Pitt, HPA, London
E264	Environmental isolate	Ty Pitt, HPA, London
E31	Environmental isolate	Ty Pitt, HPA, London
E202	Environmental isolate	Ty Pitt, HPA, London

^aNCTC, National Collection of Type Cultures; HPA, Health Protection Agency. London and Plymouth addresses are both United Kingdom.

TABLE 2.

PCR primers used in this study

Namec	Oligonucleotide sequencea (5′-3′)	Tmb (°C)
OXA57-1	AAAAGCCATATGAAGACGATCTGCACGGCG	76.1
OXA57-2	AAACTCGAGTAAAGAAGCCGGGGCAAGTCG	76.3
S104Pfor	CGCTGGCTCAAGTATCCGGTCGTGTGGTAT	78.7
S104Prev	ATACCACACGACCGGATACTTGAGCCAGCG	78.7
D170Nfor	CTCGGCAAGATGCTCAATCGCAAGCTGCCC	83.5
D170Nrev	GGGCAGCTTGCGATTGAGCATCTTGCCGAG	83.5
K232Nfor	ATCGTGCGTGGCAATCAGACGCTGGTGTTC	82.4
K232Nrev	GAACACCAGCGTCTGATTGCCACGCACGAT	82.4

^aUnderlined sequences are restriction enzyme sites (OXA57-1 and OXA57-2) or mutated codons (S104P for through K232Nrev).

^bT_m, calculated melting temperature.

^cfor and rev, forward and reverse primers, respectively.

FIG. 1.

Amino acid sequence alignment of OXA-57, OXA-59 from strain K96243 (6), OXA-42 and OXA-43 (10) from B. pseudomallei, and OXA-1 (12) from P. aeruginosa (accession numbers AJ631966, AJ632249, CAD32564, CAD32565, and 1M6KA, respectively). Active-site residues are underlined, differences between the B. pseudomallei strains are in boldface, and identical residues are represented by dots.

To determine the distribution of class D β-lactamases, 9 strains of B. pseudomallei and 12 strains of Burkholderia thailandensis (1) (Table 1) were screened by PCR (Table 2). A 534-bp product, encoding residues 50 to 227 of OXA-57 and including all three conserved structural elements (Fig. 1), was observed in all strains tested here.

The oxa-57 gene, lacking the predicted signal sequence, was cloned into the pET28a expression vector (Novagen, Madison, Wis.) by kanamycin selection. Three site-directed mutant proteins, D170N (OXA-59 equivalent), K232N, and S104P, were produced with the QuikChange site-directed mutagenesis kit (Stratagene Ltd., La Jolla, Calif.). Oligonucleotide primers are shown in Table 2. Purification of recombinant wild-type and mutant proteins was accomplished with a Co²⁺ affinity Talon column as a primary step. Size exclusion chromatography showed two peaks consistent with the presence of a monomer (26.7 kDa; ca. 90%) and dimer (53.4 kDa; ca. 10%). Circular dichroism spectra of wild-type and mutant proteins show minima at 208 and 221 nm, indicating that all forms of the enzyme are in a folded state.

Hydrolytic activities of wild-type OXA-57 and mutant enzymes were determined for nitrocefin and a number of β-lactam antibiotics (Table 3). The wild-type enzyme and the D170N and K232N mutant enzymes hydrolyzed the same compounds with similar kinetic parameters. S104P failed to hydrolyze any of the compounds tested here. Hydrolytic profiles of wild-type OXA-57 and of the class A β-lactamase from B. pseudomallei (2, 5, 13) are, taken together, consistent with most observed patterns of β-lactam drug resistance and sensitivities demonstrated by the organism. The only inconsistency is that OXA-57 demonstrates high hydrolytic activity for piperacillin (Table 3), although low MICs for selected B. pseudomallei strains have been reported (4, 7). This suggests that piperacillin sensitivity may not be a common feature of B. pseudomallei strains and that natural resistance to this compound would limit its value if it were used clinically for the treatment of melioidosis.

TABLE 3.

Steady-state kinetic parameters of wild-type OXA-57 and the D170N and K232N mutant enzymes

Substrate	Enzyme	Km (μM)a	kcat (s−1)a	kcat/Km (s−1 M−1)
Nitrocefin	WTc	17.9 ± 2.1	63.0 ± 7.5	3.51 × 106
	D170N	50.4 ± 13	79.5 ± 21	1.58 × 106
	K232N	13.2 ± 3.3	29.7 ± 7.4	2.24 × 106
Ampicillin	WT	71.9 ± 9.6	0.54 ± 0.1	7.54 × 103
	D170N	38.9 ± 2.7	2.12 ± 0.2	5.44 × 104
	K232N	157 ± 26	1.42 ± 0.2	8.98 × 103
Penicillin G	WT	29.5 ± 5.1	0.40 ± 0.1	1.36 × 104
	D170N	56.3 ± 8.5	0.61 ± 0.1	1.07 × 104
	K232N	46.6 ± 11	0.75 ± 0.2	1.60 × 104
Cloxacillin	WT	28.7 ± 4.2	0.51 ± 0.1	1.79 × 104
	D170N	166 ± 16	1.12 ± 0.1	6.69 × 103
	K232N	180 ± 29	1.04 ± 0.2	5.77 × 103
Oxacillin	WT	40.7 ± 6.1	1.23 ± 0.2	3.02 × 104
	D170N	37.2 ± 2.5	2.37 ± 0.2	6.37 × 104
	K232N	194 ± 47	3.07 ± 0.8	1.57 × 104
Piperacillin	WT	21.3 ± 3.8	2.11 ± 0.4	9.90 × 104
	D170N	22.1 ± 1.1	5.7 ± 0.3	2.58 × 105
	K232N	9.48 ± 1.2	4.31 ± 0.5	4.55 × 105
Cephalothin	WT	163 ± 11	0.28 ± 0.1	1.73 × 103
	D170N	100 ± 7.5	0.86 ± 0.1	8.59 × 103
	K232N	55.4 ± 3.9	0.65 ± 0.1	1.17 × 104
Cephaloridineb	WT	97.9 ± 15	13.0 ± 1.9	1.33 × 105

^aValues are means±standard errors of the means (n = 3).

^bAssay was not performed for the D170N and K232N mutant enzymes.

^cWT, wild-type OXA-57.

Inhibition of wild-type OXA-57 and the D170N and K232N mutant enzymes, by clavulanic acid or NaCl, was studied with nitrocefin as a reporter substrate. Dixon and/or Cornish-Bowden plots were consistent with clavulanic acid acting as a competitive inhibitor with respect to the nitrocefin substrate. K_i values of 3.4, 4.6, and 3.9 μM were obtained for the wild-type, D170N, and K232N forms of the enzyme, respectively. None of these three enzymes were inhibited by NaCl, even at a concentration of 1 M.

Although ceftazidime is the front-line drug used to treat acute melioidosis, resistance has been observed in clinical and laboratory-generated strains which contain mutations in either the class A or class D β-lactamases (5, 10, 13). Transcription of class D β-lactamases is increased in ceftazidime-resistant mutants, and two forms of the enzyme, OXA-42 and OXA-43, which contain the K232N or S104P mutation, respectively, were identified (10). Kinetic studies of the K232N mutant enzyme showed little difference in the hydrolytic profile of β-lactams compared to that for OXA-57 and no evidence of ceftazidime hydrolysis (Table 3). The S104P mutant enzyme displayed no measurable β-lactam hydrolytic activity, although this enzyme appears folded and conformationally identical to wild-type OXA-57. The absence of hydrolytic activity for ceftazidime in K232N and S104P is consistent with results obtained by Niumsup and Wuthiekanun (10), who also reported no detectable ceftazidimase activity for OXA-42 or OXA-43 and concluded that overexpression of these enzymes was unlikely to be the reason for ceftazidime resistance. However, Escherichia coli expressing OXA-43 (containing the S104P mutation) displayed some oxacillinase activity (10), although no such activity could be detected in the recombinant S104P enzyme produced here. The complete lack of activity in this mutant enzyme suggests that the B. pseudomallei E15 strain harboring OXA-43 (10) may indeed be more susceptible to a wider range of antibiotics, including oxacillins, compared to the E15 strain harboring the wild-type enzyme.

A proposed enzyme-catalyzed reaction mechanism for the B. pseudomallei class D β-lactamase is proposed (Fig. 2). This mechanism is based on results obtained here for the S104P mutant enzyme and NaCl inhibition studies of wild-type OXA-57, along with published reaction schemes for class D β-lactamases (9, 12). The S104P mutant enzyme is the first active-site mutant class D β-lactamase constructed by site-directed mutagenesis and subsequently characterized. Complete lack of activity in S104P supports the proposed role of this conserved serine as a proton relay group (Fig. 2) (8) and, for the B. pseudomallei enzyme, indicates that Ser-104 is essential for catalysis. It has been proposed that, in the Pseudomonas aeruginosa class D β-lactamase OXA-10, a water molecule, Wat-2 in Fig. 2, is involved in the deacylation step of the reaction. This proposal is based on the observation that OXA-10 is sensitive to inhibition by NaCl and that a chloride ion has been observed to displace Wat-2 in an OXA-10 crystal structure (11). Wild-type OXA-57 from B. pseudomallei is not inhibited by NaCl, suggesting that the local environment of the OXA-57 active site may either not support an inhibitory chloride ion binding site or that deacylation in OXA-57 may not necessarily require a water molecule equivalent to Wat-2 (Fig. 2).

FIG. 2.

Enzyme-catalyzed mechanism for β-lactam hydrolysis in the class D β-lactamase from B. pseudomallei (9, 12). The question mark indicates that Wat-2 may not be required for deacylation in this enzyme.