Biochemical Characterization of TEM-92 Extended-Spectrum β-Lactamase, a Protein Differing from TEM-52 in the Signal Peptide

Abstract

A bla_TEM-92 gene was cloned from a Proteus mirabilis isolate and expressed in Escherichia coli. Production of the enzyme caused reduction of susceptibility to penicillins and narrow- to expanded-spectrum cephalosporins but not to moxalactam and cephamycins. Determination of kinetic parameters with the purified enzyme revealed hydrolysis of expanded-spectrum cephalosporins, while cephamycins, moxalactam, and aztreonam were very poorly or not hydrolyzed. Clavulanate and penicillanic acid sulfones acylated TEM-92 slowly, and deacylation occurred at measurable rates.

The introduction of expanded-spectrum cephalosporins in clinical practice represented a major breakthrough for antimicrobial chemotherapy of nosocomial infections caused by Enterobacteriaceae producing broad-spectrum β-lactamases such as the widespread TEM-1, TEM-2, and SHV-1 enzymes. However, the extensive use of these drugs eventually selected for resistant strains that produce enzymes capable of degrading them, designated extended-spectrum β-lactamases (ESBLs). A number of these ESBLs are variants of TEM-1 or TEM-2 which have modified substrate specificity due to the presence of one or more amino acid substitutions. The TEM-type variants are among the most common acquired ESBLs in nosocomial isolates of Enterobacteriaceae worldwide, and their diffusion constitutes a relevant emerging problem for antimicrobial chemotherapy (1, 2, 6, 9).

TEM-92 is a TEM-type ESBL recently detected in clinical isolates of Proteus mirabilis and Providencia stuartii from France (3). It contains the same amino acid substitutions as TEM-52 (E104K, M182T, and G238S) (14) plus a Q6K substitution in the signal peptide region (3). TEM-52 was originally detected in a Klebsiella pneumoniae isolate from France (14) and was subsequently found to be widespread in other countries, including Italy (10, 12). It was reported to be associated to a decreased susceptibility to moxalactam (14), an unusual feature for most ESBLs (1). Kinetic properties of this enzyme, however, have not been analyzed in detail with a purified enzyme preparation. In this paper we report on the purification and biochemical characterization of TEM-92.

Cloning and expression of the bla_TEM-92 gene in Escherichia coli.

The bla_TEM-92 gene was amplified by PCR from a P. mirabilis isolate (isolate 3Va) obtained at the “Circolo” Hospital of Varese (northern Italy) in 1998. Two primers, cgccggagctcATAAAATTCTTGAAGAC and cgcgggtaccTTACCAATGCTTAATCA, designed based on bla_TEM-flanking regions (the regions that hybridize with the template are in uppercase letters) and containing a SacI and a KpnI linker (underlined) to be inserted upstream and downstream of the coding sequence, respectively, were used for PCR, which was carried out essentially as described previously (11). The resulting amplicon, digested with the two enzymes, was cloned in the pBC-SK plasmid vector (Stratagene Inc., La Jolla, Calif.) to obtain recombinant plasmid pMGP92. Sequencing of the cloned amplicon was performed on both strands as described previously (12) and confirmed the presence of a bla_TEM allele encoding an enzyme identical to TEM-92 (3). In vitro susceptibility testing was carried out as recommended by the NCCLS (8).

Compared to E. coli DH5α(pBC-SK), DH5α(pMGP92) exhibited a significant reduction of susceptibility to various β-lactams, including penicillins, cephalothin, cefotaxime, and ceftazidime, but not to cefoxitin, cefotetan, or moxalactam (Table 1). The resistance pattern was overall similar to that of P. mirabilis 3Va (Table 1). The presence of inhibitors in association with penicillins significantly lowered MICs for both strains.

TABLE 1.

In vitro susceptibility to various β-lactams of E. coli DH5α(pMGP52), producing the ESBL TEM-92

Antibiotic	MIC (μg/ml)
	P. mirabilis 3Vaa	E. coli DH5α(pMGP92)	E. coli DH5α(pBC-SK)b
Amoxicillin	>128	>128	2
Amoxicillin-clavulanatec	4	4	1
Piperacillin	>256	256	1
Piperacillin-tazobactamd	2	2	0.5
Cephalothin	>128	>128	2
Cefotaxime	>128	>128	≤0.06
Ceftazidime	16	16	0.12
Aztreonam	8	1	0.12
Cefoxitin	4	2	1
Cefotetan	0.12	0.25	0.12
Moxalactam	≤0.06	0.12	0.12

^aOriginal host of TEM-92.

^bpBC-SK is the cloning vector.

^cAmoxicillin-clavulanate at a 2:1 ratio.

^dTazobactam at fixed concentration of 4 μg/ml.

Purification of the TEM-92 β-lactamase.

TEM-92 was purified from a stationary-phase culture of E. coli DH5α(pMGP92) grown aerobically, at 37°C, in 6 liters of Luria-Bertani broth containing chloramphenicol (30 μg/ml). Cells were harvested by centrifugation (13,000 × g for 10 min at 4°C), washed twice with 50 mM Tris HCl buffer (pH 8.0), resuspended in 100 ml of the same buffer, and disrupted by sonication (five times for 30 s each time at 60 W). Cell debris was removed by centrifugation (105,000 × g for 30 min at 4°C), and the cleared supernatant was loaded onto a Q-Sepharose FF column (2 by 20 cm; Amersham Biosciences, Milan, Italy) equilibrated with 50 mM Tris HCl buffer (pH 8.0). The β-lactamase was eluted with a linear gradient of NaCl (0 to 1 M) in the same buffer. The fractions containing β-lactamase activity were pooled, dialyzed at 4°C against 25 mM bis-Tris buffer (pH 7.0), and loaded onto a Mono P HR 5/20 column (Amersham Biosciences) equilibrated with the same buffer. The protein was eluted with 10-fold-diluted Polybuffer 74 (Amersham Biosciences). The fractions containing β-lactamase activity were pooled, dialyzed at 4°C against 50 mM sodium phosphate buffer (pH 7.0), and stored at −80°C. During purification the presence of β-lactamase activity was assayed by monitoring hydrolysis of 100 μM cefotaxime as a substrate, at 260 nm, at 30°C in 50 mM sodium phosphate buffer (pH 7.0). Protein concentration was determined by the method of Bradford using a commercial protein assay kit (Bio-Rad, Richmond, Calif.) with bovine serum albumin as the standard.

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis and analytical isoelectric focusing were carried out as described previously (11). After electrophoresis, the purified protein preparation appeared as a single 28-kDa band, and the purity of the preparation was estimated to be higher than 95% (data not shown). In isoelectric focusing, the purified protein exhibited a pI of 6.0 (data not shown).

Kinetic parameters of TEM-92 and interaction with inhibitors.

Determination of kinetic parameters of the purified TEM-92 β-lactamase was carried out essentially as described previously (13), using a Lambda 2 spectrophotometer (Perkin-Elmer, Rahway, N.J.). Enzyme assays were always performed at 30°C in 50 mM sodium phosphate buffer (pH 7.0), containing 0.2 M KCl to prevent enzyme instability. Inhibition by clavulanic acid, tazobactam, and sulbactam was monitored with 100 μM nitrocefin as the reporter substrate. Inhibitor parameters with clavulanate, sulbactam, and tazobactam were determined as described previously (4). For determination of the 50% inhibitory concentration, the enzyme was preincubated with the inhibitor for 5 min at 30°C.

Under the experimental conditions used, TEM-92 was able to efficiently hydrolyze (k_cat/K_m >10⁵ M⁻¹ · s⁻¹) several β-lactams, including penicillin G, narrow-spectrum cephalosporins, cefotaxime, and cefepime, while ceftazidime was hydrolyzed somewhat less efficiently. A very slow turnover was observed with cefotetan, and no hydrolysis was detectable with moxalactam, cefoxitin, and aztreonam (Table 2). Although not hydrolyzed, cefoxitin and aztreonam were recognized by the enzyme (the latter with high affinity), while interaction with moxalactam was poor or nonexistent (Table 2). In comparison with TEM-3, TEM-92 showed an overall better catalytic activity against cefotaxime, suggesting that the M182T global suppressor mutation (16) present in the latter enzyme might provide a contribution in this sense.

TABLE 2.

Kinetic parameters determined with the purified TEM-92 β-lactamase^a

Substrate	TEM-92			TEM-1			TEM-3
	Km (μM)	kcat (s−1)	kcat/Km (M−1 · s−1)	Km (μM)	kcat (s−1)	kcat/Km (M−1 · s−1)	Km (μM)	kcat (s−1)	kcat/Km (M−1 · s−1)
Penicillin G	10	150	1.5 × 107	19	1,600	8.4 × 107	4	38	9.5 × 106
Nitrocefin	49	69	1.4 × 106	55	930	1.7 × 107	6	44	7.3 × 106
Cephalothin	36	6	1.7 × 105	246	160	6.5 × 105	7	14	1.9 × 106
Cefazolin	27	9	3.3 × 105	513	200	3.9 × 105	96	52	5.4 × 105
Cefotaxime	15	170	1.1 × 107	6,000	9	1.5 × 103	51	51	1.0 × 106
Ceftazidime	166	10	6.0 × 104	4,300	0.3	7 × 101	175	7	4.0 × 104
Cefepime	115	26	2.3 × 105	ND	ND	ND	ND	ND	ND
Loracarbef	19	51	2.7 × 106	ND	ND	ND	ND	ND	ND
Cefoxitin	137b	<0.01c	NC	ND	ND	ND	ND	ND	ND
Cefotetan	80	0.06	7.5 × 102	ND	ND	ND	ND	ND	ND
Moxalactam	>1,000c	<0.01c	NC	ND	ND	ND	ND	ND	ND
Aztreonam	8b	<0.01c	NC	1,430	1	7 × 102	1	0.2	2 × 105

^aK_m and k_cat values are the means of three different measurements. The standard deviation was always lower than 10%. Kinetic parameters of TEM-1 and TEM-3 (15) are also shown for comparison. NC, not calculated; ND, not determined.

^bDetermined as K_i.

^cTurnover rates were lower than 0.01 s⁻¹ after exposure of the enzyme (at a concentration of 25 nM) to saturating substrate concentrations (1 mM for moxalactam).

TEM-92 was inhibited by clavulanate, sulbactam, and tazobactam. The 50% inhibitory concentrations were 0.15, 0.1 and 0.02 μM, respectively. Inactivation experiments performed with the above compounds showed that interaction was always characterized by low acylation efficiencies (k₊₂/K ranging from 1.4 × 10⁻² to 4.2 × 10⁻³ μM⁻¹ · s⁻¹) and that the hydrolysis of the inhibitor was not negligible (k₊₃ ≠ 0), since the dissociation of the enzyme-inhibitor complex was measurable (Table 3). Both of these parameters could account for the remarkable increase of MICs of amoxicillin-clavulanate and of piperacillin-tazobactam if a larger bacterial inoculum was used (data not shown).

TABLE 3.

Kinetic parameters of interaction of TEM-92 with β-lactamase inhibitors^a

Inhibitor	k+3 (s−1)	k+2/K (μM−1 · s−1)
Clavulanic acid	5.6 × 10−3	4.2 × 10−3
Sulbactam	4.1 × 10−4	3.3 × 10−3
Tazobactam	3.1 × 10−3	1.4 × 10−2

^aData are the means of three different measurements. The standard deviation was always lower than 10%.

Concluding remarks.

The genes encoding TEM-1 and TEM-2 have undergone a very high selective pressure because of a number of broad-spectrum β-lactam antibiotics which have been synthesized and introduced into clinical practice since the 1980s. The TEM-type ESBLs variably show a reduced catalytic efficiency toward penicillins and a broadening of the spectrum toward expanded-spectrum cephalosporins and aztreonam (1). Cephamycins and moxalactam normally remain stable to these enzymes, due to the presence of a substitution in the 7α position of the β-lactam nucleus which is known to displace the hydrolytic water molecule present in the active site of class A β-lactamases (7). However, production of TEM-52 was reported to be associated with reduced susceptibility to moxalactam (14). In this study we were unable to confirm that result with TEM-92, which differs from TEM-52 only in the presence of an additional mutation in the signal peptide. Moreover, kinetic analysis carried out with the purified enzyme showed that it is actually unable to efficiently hydrolyze moxalactam and cephamycins, although some of these molecules could interact with the enzyme with a moderate affinity. Concerning moxalactam, we were also unable to confirm the high-affinity interaction previously reported for TEM-52 (14).

In a recent nationwide survey we demonstrated that TEM-52 was the most prevalent TEM-type ESBL in Italian hospitals (12). Since the sequence of the leader peptide was not investigated in that survey, the enzyme could also have been TEM-92, and this was actually found to be the case in a number of isolates that were subsequently investigated (M. Perilli and G. Amicosante, unpublished results). TEM-92 differs from TEM-52 in only one amino acid substitution in the signal peptide sequence (Q6K). Since the positively charged amino-terminal region of the signal peptide plays an important role in efficient protein secretion across the membrane (5), the substitution could have functional significance in expression of the resistance phenotype.