Beta-lactamases represent the main mechanism of bacterial resistance to beta-lactam antibiotics. The recent emergence of bacterial strains producing inhibitor-resistant TEM (IRT) enzymes could be related to the frequent use of beta-lactamase inhibitors such as clavulanic acid, sulbactam and tazobactam in hospitals and in general practice. The IRT beta-lactamases differ from the parental enzymes TEM-1 or TEM-2 by one, two or three amino acid substitutions at different locations. This paper reviews the phenotypic, genetic and biochemical characteristics of IRT beta-lactamases in an attempt to shed light on the pressures that have contributed to their emergence.

Sera of 52 Lyme borreliosis patients classified according to their clinical features were analysed by Western blot using as antigens Borrelia strains belonging to three recently described genomic species. The antibody response was demonstrated to be homologous within each genospecies. Serum reactivity was studied for each of the type strains Borrelia burgdorferi sensu stricto (strain B31T), Borrelia garinii (strain 20047T) and group VS461. Seven of 15 sera (46.6%) of patients with meningoradiculitis showed preferential reactivity with Borrelia garinii (strain 20047T), all of 8 sera (100%) of patients with acrodermatitis chronica atrophicans showed preferential reactivity with group VS461 (strain VS461) and 8 of 16 sera (50%) of patients with arthritis showed preferential reactivity with Borrelia burgdorferi sensu stricto (strain B31T). The presence of a strong response to OspA and OspB proteins of Borrelia burgdorferi sensu stricto was found only in this last group of patients. These results suggest that there are clinical implications of the recently described modifications in the taxonomy of Borrelia burgdorferi.

Two different strains of Escherichia coli exhibiting unusual patterns of resistance to beta-lactam antibiotics were isolated from patients at Cochin Hospital. Both isolates showed a low level of resistance to amoxycillin, ticarcillin and ureidopenicillins but were susceptible to cephalosporins, aztreonam and imipenem; beta-lactamase inhibitors potentiated the activities of the beta-lactams to only a limited extent. All resistance characteristics of the strains were transferable by conjugation to E. coli K12. Resistance was shown to be due to beta-lactamases of pI 5.20 and relative molecular masses of 24,000. The hydrolytic and inhibition profiles of these enzymes were similar to each other but differed from those of broad-spectrum beta-lactamases (TEM-1). The rates of hydrolysis (Vmax) of amoxycillin (c. 200%) were higher than that for TEM-1 (84%). Ticarcillin, ureidopenicillins and cephaloridine were hydrolyzed slowly. However, as for TEM-1, no hydrolysis was observed with cefoxitin, third generation cephalosporins, aztreonam and imipenem. The high Km values demonstrated the poor affinity of these enzymes for their substrates. Unlike TEM-1, they were poorly inhibited by beta-lactamase inhibitors. These two enzymes differed from each other as follows: (i) the concentrations of clavulanic acid required for 50% beta-lactamase inhibition were 31 mumol/L for one enzyme (E-SAL) and 9.4 mumol/L for the other (E-GUER); (ii) p-chloromercuribenzoate was a more active inhibitor of E-SAL then E-GUER. The titration curve method and DNA-DNA hybridization studies demonstrated that both enzymes were structurally related to TEM-1. The novel plasmid-encoded enzymes produced by the two isolates of E. coli appeared to be almost identical and to be derived from TEM-enzymes. On the basis of their presumed phylogeny and their biological properties, we propose that these beta-lactamases be given the generic name TRI (TEM Resistant to beta-lactamase Inhibitors).

The properties of PSE-2 beta-lactamase have been examined by using two new PSE-2-producing plasmids, pMG33 and pMG74, as well as plasmid R151, found in Pseudomonas aeruginosa. PSE-2 beta-lactamase resembled other PSE enzymes in activity against carbenicillin, but it also resembled OXA enzymes, such as OXA-1, in rapid hydrolysis of oxacillin, cloxacillin, and methicillin and in inhibition by sodium chloride but not by cloxacillin. Antisera that inactivated TEM-1, TEM-2, OXA-1, or PSE-1 and PSE-4 beta-lactamase failed to cross-react with PSE-2, which thus appears to be immunologically distinct. The plasmids determining PSE-2 varied in geographical origin, size, transfer proficiency, and incompatibility specificity, but all determined resistance to carbenicillin, gentamicin, kanamycin, streptomycin, spectinomycin, sulfonamide, and tobramycin. From a pUZ8-R151 recombinant plasmid in Escherichia coli, the PSE-2 beta-lactamase gene could be transposed to a second plasmid in a 6.4-megadalton unit together with resistance to gentamicin, kanamycin, streptomycin, spectinomycin, sulfonamide, and tobramycin. Transposition was recA independent. We propose the designation Tn1404 for this unit, which, like transposons carrying OXA-1, PSE-1, PSE-4, and some transposons determining TEM-1, includes genes for beta-lactam, aminoglycoside, and sulfonamide resistance.

The molecular diversity of inhibitor-resistant TEM (IRT) enzymes was explored using a strategy which involved DNA amplification by polymerase chain reaction (PCR), analysis of restriction fragment length polymorphism (RFLP), and direct nucleotide sequencing. The study of plasmid-borne genes from 27 strains, resistant to amoxicillin and beta-lactamase-inhibitor combinations, identified mutations resulting in amino acid change at positions 69, 244, 275, and 276 known to be associated with the IRT phenotype and a mutation at nucleotide position 162 in the promoter region. These mutations were found to lie on two different gene sequences, described here as "TEM-1B like" and "TEM-2 like" restriction linkage groups. Further analysis, of nucleotide sequences of promoter and coding regions of the beta-lactamases, confirmed that a given mutation causing IRT phenotype could be associated with two different gene sequence frameworks and two different causal mutations could lie on identical gene sequence framework. These data argue in favor of convergent phenotypic evolution of IRT enzymes under the selective pressure imposed by the intensive clinical use of beta-lactam-beta-lactamase inhibitor combinations.