INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Production of degrading enzymes (-lactamases) is the most common mechanism of bacterial resistance to -lactam antibiotics. The evolution of -lactamase determinants started long before the introduction of -lactams in clinical practice, presumably under the selective pressure of natural -lactam compounds produced in various microbial ecosystems, while the recent (on an evolutionary timescale) exploitation of -lactams for antimicrobial chemotherapy has strongly selected for spreading and evolution of similar resistance determinants among bacterial pathogens. In fact, it is well known that the continuous release of new -lactams in clinical practice has invariably been followed by the appearance of new -lactamases capable of degrading the newest compounds (7, 22).

Two different families of -lactam-degrading enzymes, which catalyze the same reaction, i.e., the opening of the -lactam ring by hydrolysis of the amide bond, but are structurally and mechanistically unrelated, have evolved in bacteria: (i) active-site serine--lactamases and (ii) metallo--lactamases (8, 13). The latter enzymes were identified some 25 years later than serine--lactamases and have remained less common among pathogenic bacteria (5, 8, 20). Nevertheless, they are potentially very dangerous as resistance effectors due to their efficient hydrolysis of carbapenem antibiotics, which are stable to hydrolysis by most -lactamases and often represent the "last-resort" agents for chemotherapy of multidrug-resistant pathogens, and due to their lack of susceptibility to the serine--lactamase inhibitors such as clavulanic acid and penicillanic acid sulfones (2, 3; reference 5 and references therein; 18, 19, 26, 29, 31). The recent emergence of mobile metallo--lactamase genes capable of horizontal spreading among nosocomial strains of Enterobacteriaceae, Pseudomonas aeruginosa, and other gram-negative nonfermenters (reference 6 and references therein; 26, 30, 36, 37; G. Cornaglia, M. L. Riccio, A. Mazzariol, P. Piccoli, L. Lauretti, R. Fontana, and G. M. Rossolini, Letter, Lancet 353:899-900, 1999) has considerably increased the attention to these enzymes, including them among the major threats for the 21st century in the field of microbial drug resistance (6).

Most known metallo--lactamases are encoded by chromosomal genes of some bacterial species that are primarily members of the environmental microbiota, such as Bacillus cereus, Stenotrophomonas maltophilia, Aeromonas spp., Myroides odoratus (formerly Flavobacterium odoratum), Legionella gormanii (reference 5 and references therein), and Chryseobacterium spp. (2, 3, 31), whereas some as yet unknown environmental species are the most likely sources of the mobile metallo--lactamase determinants that recently appeared among gram-negative pathogens. Therefore, environmental bacteria could be an important reservoir of similar resistance determinants.

In this work we carried out screening on various environmental samples for the presence of bacteria producing metallo--lactamase activity. In addition to several isolates belonging to various species that are known to produce similar enzymes, the screening also yielded a metallo--lactamase-producing isolate of Janthinobacterium lividum, a species in which production of metallo--lactamase activity has not been previously reported. Scanning for metallo--lactamase determinants carried by this isolate led to the identification of a new member of the highly divergent subclass B3 lineage of metallo--lactamases, named THIN-B, which appears to be resident in this species.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Media, reagents and reference strain. Nutrient broth (NB), Nutrient agar (NA), and Mueller-Hinton (MH) medium (Difco Laboratories, Detroit, Mich.) were used for bacterial cultures. Antibiotics and other reagents were from Sigma Chemical Co. (St. Louis, Mo.) unless otherwise specified. Imipenem was from Merck Research Laboratories (Rahway, N.J.), meropenem from Astra-Zeneca Pharmaceuticals (Macclesfield, Cheshire, United Kingdom), ceftazidime from Glaxo-Wellcome (Verona, Italy), aztreonam and cefepime from Bristol-Myers Squibb Co. (Wallingford, Conn.), and nitrocefin from Unipath (Milan, Italy). The J. lividum reference strain DSM1522^T was purchased from the German Collection of Microorganisms and Cell Cultures (DSMZ, Braunschweig, Germany).

Collection and analysis of environmental samples. Environmental samples were collected in 50-ml sterile screw-cap polypropylene tubes, from the different sources listed in Table 1. Each sample was collected, transferred to the laboratory, and processed on the same day. Samples were carefully resuspended in an approximately equal volume of sterile saline and left to decant at room temperature for 4 h. A 100-µl volume of the supernatant from each sample was then inoculated in 5 ml of NB containing imipenem (5 µg/ml) and amphotericin B (20 µg/ml) and was incubated at 25°C aerobically until the development of turbidity (usually evident after 24 to 48 h). Dilutions of each culture (in order to obtain isolated colonies) were then plated on NA containing imipenem (5 µg/ml) and amphotericin B (20 µg/ml) (NACA medium) and were incubated at 25°C until the appearance of colonies. A representative for each different colony morphology was reisolated on NACA medium, subjected to gram staining, and tested for production of metallo--lactamase activity. Metallo--lactamase producers were identified at the species or genus level according to the Manual of Clinical Microbiology (23) or, in the case of J. lividum, according to Bergey's Manual of Systematic Bacteriology (35). The API 20NE (Bio-Mérieux, Marcy-L'Etoile, France), ATB 32GN (Bio-Mérieux), and Crystal E/NF (Becton Dickinson Microbiology Systems, Cockeysville, Md.) systems were used for the identification of some gram-negative isolates.

-Lactamase assays. Production of metallo--lactamase activity by the environmental isolates was assayed in crude cell extracts prepared from late-exponential-phase cells grown aerobically, at 25°C, in NB containing imipenem (5 µg/ml). Crude extracts were prepared as follows. Cells were collected by centrifugation, resuspended in 20 mM sodium phosphate buffer (PB) (pH 7.0) (1/10 of the original culture volume), and disrupted by sonication (six times for 15 s each time, at 50 W), and cell debris was removed by centrifugation at 10,000 × g for 10 min. -Lactamase activity was assayed spectrophotometrically by monitoring imipenem hydrolysis at 299 nm (change in extinction coefficient, 9,000 M¹ cm¹) at 25°C in PB. The initial substrate concentration was 150 µM. One unit of -lactamase acivity was defined as the amount of enzyme hydrolyzing 1 nmol of imipenem per min under the above assay conditions. Inhibition of enzymatic activity by EDTA was assayed by measuring the residual carbapenemase activity after incubation of the crude extract, for 20 min at 25°C, in the presence of 5 mM EDTA. A control without EDTA was always carried out in parallel. Induction experiments with JAC1 and DSM1522^T were carried out as follows. Cells were grown aerobically, at 25°C, in MH broth until the mid-exponential phase (A₆₀₀  0.3 to 0.4). The culture was then split into two flasks, of which one was added with the inducer at the desired concentration. Cells were collected after a further 3-h incubation. Preparation of crude extracts and -lactamase assays were carried out as described above. The protein concentration in solution was assayed with a commercial kit (Bio-Rad [Richmond, Calif.] protein assay), with bovine serum albumin as a standard.

Recombinant DNA methodology. Basic recombinant DNA procedures were performed as described by Sambrook et al. (33). Genomic DNA was extracted from J. lividum strains as described previously (16), with an additional extraction step with water-saturated ether before ethanol precipitation. For construction of the JAC1 genomic library, genomic DNA was partially digested with Sau3AI, and fragments in the 2- to 9-kb size range were purified by agarose gel electrophoresis using the Geneclean II kit (Bio 101, La Jolla, Calif.). The purified restriction fragments were ligated to BamHI-linearized and dephosphorylated pACYC184 (33), and the ligation mixture was transformed into Escherichia coli DH5 (GIBCO-BRL, Gaithersburg, Md.) by electroporation using a Gene Pulser apparatus (Bio-Rad) according to the manufacturer's instructions. The ratio of recombinant clones to those carrying an empty religated vector was >10, as shown by replica plating of transformants, selected on Luria-Bertani (LB) agar plates containing chloramphenicol (70 µg/ml), onto plates containing both chloramphenicol and tetracycline (20 µg/ml). Southern blot hybridization was performed as described previously (33) using a nylon membrane (Amersham Pharmacia Biotech, Milan, Italy) and a ³²P-labeled DNA probe.

DNA sequencing and computer analysis of sequence data. DNA sequences of both strands were determined on plasmid templates by the dideoxy-chain termination method using an automatic DNA sequencer (model 4000; LI-COR Inc., Lincoln, Nebr.), the Thermosequenase DNA sequencing kit (Amersham), and IRD 800-labeled custom sequencing primers (MWG Biotech, Munich, Germany). Similarity searches against sequence databases were performed using an updated version of the BLAST program at the BLAST interface of the National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov/BLAST/). Computer analysis of sequence data was performed using an updated version (8.1) of the Wisconsin Package (version 8.1; Genetics Computer Group Inc., Madison, Wis.) at the Italian EMBL Node of Bari. The multiple sequence alignment was generated with the help of the PILEUP program of the Wisconsin package and was manually refined by considering the information available on the three-dimensional structure of metallo--lactamases (9-12, 38). It essentially corresponds to that proposed for the definition of a standard numbering scheme for class B -lactamases (14). The unrooted tree was constructed on the basis of the multiple sequence alignment with the help of the CLUSTAL W program at the Italian EMBL Node of Bari.

In vitro susceptibility testing. The vitro susceptibility of E. coli DH5 carrying the cloned bla_THIN-B gene was determined by a macrodilution broth method (24), using MH broth and a bacterial inoculum of 10⁵ CFU per tube. Results were recorded after incubation at 28°C for 24 h.

Nucleotide sequence accession number. The nucleotide sequence reported in this paper has been submitted to the EMBL/GenBank/DDBJ sequence databases and assigned accession number AJ250876.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Screening of environmental samples for metallo--lactamase-producing bacteria. Eleven environmental samples collected from different sources (Table 1) were screened for the presence of bacteria growing on a selective medium containing imipenem, a carbapenem antibiotic which is efficiently degraded by all known metallo--lactamases but resistant to hydrolysis by most active-site serine -lactamases (7). A broad-spectrum antifungal agent was also added to the medium to prevent fungal overgrowth. Bacterial isolates growing on the selective enrichment medium were obtained from each sample, and a total of 27 different isolates were collected using this strategy (Table 1).

View this table: [in this window] [in a new window]   TABLE 1.   Environmental samples analyzed for the presence of metallo--lactamase-producing bacteria, and results of the screening

Production of metallo--lactamase activity in J. lividum. Since production of metallo--lactamase activity has not been previously reported for J. lividum, the isolate of this species obtained from sample E, named JAC1, was subjected to further investigation.

View this table: [in this window] [in a new window]   TABLE 2.   Carbapenemase activity in crude extracts of J. lividum JAC1, J. lividum DSM1522T, and E. coli DH5(pCIRO)a

Cloning and characterization of a metallo--lactamase determinant from J. lividum JAC1. The genome of JAC1 was scanned for the presence of metallo--lactamase determinants by means of a shotgun cloning approach. For this purpose a genomic library of JAC1, constructed in the pACYC184 plasmid vector and transformed in E. coli DH5, was replica plated on MH medium containing imipenem at a concentration of 5 µg/ml. One clone growing on this medium was obtained out of approximately 3,000 screened recombinants. The presence of carbapenemase activity susceptible to inhibition by EDTA was detectable in the crude extract of this clone (Table 2).

View larger version (13K): [in this window] [in a new window]   FIG. 1.   Physical map of the insert of plasmid pCIRO, and subcloning strategy. Thick lines represent cloned DNA, while thin lines correspond to vector sequences. Production of metallo--lactamase activity (ML act.) was assayed as described in Materials and Methods on crude extracts prepared from late-exponential-phase cultures. The location of the blaTHIN-B ORF is indicated. B/S, BamHI/Sau3AI junction; H, HindIII; Hc, HincII; K, SacII; N, NotI; P, PstI; Sa, SalI; Sc, SacI; V, EcoRV. In plasmid pBCIRO-K the orientation of the blaTHIN-B ORF is the same as that of the Plac promoter flanking the polylinker.

View larger version (45K): [in this window] [in a new window]   FIG. 2.   Nucleotide sequence of the DNA insert of plasmid pBCIRO-K containing the blaTHIN-B ORF and flanking regions. The putative ribosome-binding site is overlined. Protein translation is reported below the sequence, and the putative signal peptide for protein secretion is underlined.

Comparison of the THIN-B enzyme with other metallo--lactamases. A multiple sequence alignment analysis with other class B -lactamases confirmed the closest similarity of THIN-B with the L1 enzyme of S. maltophilia and with the other enzymes of subclass B3. THIN-B could be aligned over the entire sequence with these enzymes without introducing major gaps (Fig. 3), and the percent identities among them (24.2 to 35.6%) were considerably higher than those between THIN-B and the metallo--lactamases of subclasses B1 (9.6 to 17.8%) and B2 (15.7 to 16.5%) (Table 3).

View larger version (71K): [in this window] [in a new window]   FIG. 3.   Comparison of the THIN-B amino acid sequence (boldfaced) with those of other molecular class B -lactamases of subclasses B1, B2, and B3. IMP-1, IMP-1 enzyme encoded by the blaIMP-1 gene cassette found in Serratia marcescens TN9106 (25) and in other gram-negative strains (1, 17); CcrA, CcrA enzyme from Bacteroides fragilis TAL3636 (28); Bc-II, -lactamase II from B. cereus 569/H (15); VIM-1, VIM-1 enzyme encoded by the blaVIM-1 gene cassette found in P. aeruginosa VR-143/97 (19); IND-1, IND-1 enzyme from C. indologenes 001 (3); BlaB, BlaB enzyme from C. meningosepticum CCUG4310 (31); CphA, CphA enzyme from A. hydrophila AE036 (21); SfhI, SfhI enzyme from Serratia fonticola UTAD54 (EMBL/GenBank accession number AF197943); L1, L1 enzyme from S. maltophilia IID 1275 (39); FEZ-1, FEZ-1 enzyme from L. gormanii ATCC 33297T (4); GOB-1, GOB-1 enzyme from C. meningosepticum PINT (2). The BBL numbering scheme (14) is indicated below the sequences; the numbering of the L1 enzyme (38) is also indicated, in italics. Identical residues are marked with an asterisk. Residues of the L1 enzyme involved in binding of Zn2+ are indicated by the letter z. Secondary structure elements of L1 (38) are also indicated above the sequences: 3, 310 helix; b, extended strand participating in -ladder; t, hydrogen-bonded turn; a, -helix. The invariant residues in all proteins, or in those of subclass B3, are shaded.

View this table: [in this window] [in a new window]   TABLE 3.   Pairwise percent amino acid sequence identity between class B -lactamases

View larger version (16K): [in this window] [in a new window]   FIG. 4.   Unrooted tree showing phylogenetic relationships among metallo--lactamases. The names of the enzymes are the same as in Fig. 3. Numbers at branching points indicate the number per 1,000 bootstrap trials returned for that point.

Influence of THIN-B production on -lactam susceptibility. The substrate specificity of THIN-B was investigated by testing the susceptibility to various -lactams of E. coli DH5(pBCIRO-K), which carries a cloned copy of the bla_THIN-B gene and produces the THIN-B enzyme (Fig. 1), in comparison with that of DH5(pBC-SK), carrying the empty plasmid vector. THIN-B production was associated with a decrease in the in vitro susceptibility of the bacterial host to ampicillin, carbenicillin, piperacillin, cefotaxime, cephalothin, cefuroxime, cefoxitin, ceftazidime, cefepime, imipenem, and meropenem, while susceptibility to aztreonam was not affected (Table 4). The relative MIC increases were higher overall with cephalosporins (except for cefepime) and meropenem (Table 4).

View this table: [in this window] [in a new window]   TABLE 4.   -Lactam susceptibility of E. coli DH5(pBCIRO-K), carrying the cloned blaTHIN-B gene and producing the THIN-B enzymea compared to that of the E. coli host containing the plasmid vector pBC-SK

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Results of the screening performed in this work indicated that metallo--lactamase-producing bacteria are widespread in the environmental microbiota and that such an approach can be successful for their detection. In the screening procedure imipenem was used as a selective agent in consideration of the fact that carbapenemase activity is a constant feature of metallo--lactamases (2, 3; reference 5 and references therein; 18, 19, 26, 29, 31), but its concentration was kept relatively low in consideration of the low carbapenem MICs exhibited by some metallo--lactamase producers (32, 34). Nevertheless, the possibility that the selective conditions used in the medium might have biased the screening in favor of certain species cannot be excluded. On the other hand, the screening also yielded a consistent number of isolates that were able to grow on the selective enrichment medium while not producing detectable carbapenemase activity. In these cases resistance to the antibiotic present in the medium could be related to one or more of the following mechanisms: (i) low affinity of the -lactam targets; (ii) production of enzymes with very low rates of turnover against carbapenems (undetectable under the assay conditions adopted in this study), and (iii) presence of permeability barriers or efflux systems. It would be interesting to evaluate whether modification of various parameters (nature and concentration of the selective agent, nature of the medium, temperature and atmosphere used for incubation) could increase the screening sensitivity, and also to analyze samples from different sources (such as various types of aquatic environments).

Of the metallo--lactamase producers isolated in this study, most belonged to species that were already known for this trait, of which S. maltophilia was the most common. This suggests an overall broader diffusion and higher prevelance of this species, compared to the others, in the environmental microbiota, although an influence of the screening procedure on this result cannot be ruled out at this stage. The screening also yielded a metallo--lactamase-producing isolate of J. lividum, a species in which no similar activity has been reported previously. In J. lividum, production of metallo--lactamase activity is likely a species-related trait, since it was also detected in the type strain, and it is apparently regulated, since it is detectable at higher levels upon exposure to -lactam compounds. Scanning the genome of the J. lividum isolate for carbapenemase determinants led to the isolation and identification of a chromosomal gene, named bla_THIN-B, encoding a new molecular class B enzyme. The chromosomal origin and base composition of this gene, together with the presence of closely related sequences in the J. lividum type strain, strongly suggest that bla_THIN-B is a resident gene of this species.

Sequence analysis showed that the THIN-B enzyme belongs to the highly divergent subclass B3 lineage of metallo--lactamases and is most closely related to the L1 enzyme of S. maltophilia. With the addition of THIN-B, subclass B3, which until recently included only one member (29), now has four different enzymes, all from environmental species, clustered in two different evolutionary sublineages, one including the L. gormanii FEZ-1 (4) and Chryseobacterium meningosepticum GOB-1 (2) metallo--lactamases, and the other including L1 and THIN-B (Fig. 4). The almost 36% sequence identity to L1 and the complete conservation of the residues that in L1 are known to be directly or indirectly involved in metal coordination (38) suggest that the three-dimensional fold of THIN-B is likely very similar to that of L1 and that the geometry of zinc coordination in the active site of THIN-B is the same as that of L1 and different from those of the enzymes of subclass B1 (38; reference 40 and references therein). The conservation of the two cysteine residues that in L1 (at positions 256 and 290, in the BBL numbering) form an intramolecular disulfide bridge, which constrains the loop between -strand 12 and the C-terminal -helix 5 (38), suggests that a similar disulfide bridge is likely present also in THIN-B.

Production of the THIN-B enzyme in E. coli caused a decrease in host susceptibility to a broad array of -lactams, including penicillins, narrow-spectrum cephalosporins, cephamycins, oxyimino cephalosporins, and carbapenems. Only aztreonam was apparently unaffected, in agreement with the properties of other metallo--lactamases (2, 3; reference 5 and references therein; 18, 19, 26, 29, 31). THIN-B, therefore, appears to be a class B enzyme of broad substrate specificity which can be assigned to group 3a in the functional classification of -lactamases (8). A biochemical and structural analysis of THIN-B, in comparison with the other members of subclass B3, would be useful for acquiring a better understanding of the structure-function relationships of this emerging lineage of metallo--lactamases.