Oliognucleotide primers

About

Primers for Use in Detecting Beta-Lactamases Abstract Oliognucleotide primers are provided that are specific for nucleic acid characteristic of certain beta-lactamases. The primers can be employed in methods to identify nucleic acid characteristic of family-specific beta-lactamase enzymes in samples, and particularly, in clinical isolates of Gram-negative bacteria. Background A disturbing consequence of the use, and over-use, of beta-lactam antibiotics (e.g., penicillins and cephalosporins) has been the development and spread of beta-lactamases. Beta-lactamases are enzymes that open the beta-lactam ring of penicillins, cephalosporins, and related compounds, to inactivate the antibiotic. The production of beta-lactamases is an important mechanism of resistance to beta-lactam antibiotics among Gram-negative bacteria. Expanded-spectrum cephalosporins have been specifically designed to resist degradation by the older broad-spectrum beta-lactamases such as TEM-1, 2, and SHV-1. Microbial response to the expanded-spectrum cephalosporins has been the production of mutant forms of the older beta-lactamases called extended-spectrum beta-lactamases (ESBLs). Although ESBL-producing Enterobacteriaceae were first reported in Europe in 1983 and 1984, ESBLs have now been found in organisms of diverse genera recovered from patients in all continents except Antarctica. The occurrence of ESBL-producing organisms varies widely with some types more prevalent in Europe (TEM-3), others more prevalent in the United States (TEM-10, TEM-12 and TEM-26), while others appear worldwide (SHV-2 and SHV-5). These enzymes are capable of hydrolyzing the newer cephalosporins and aztreonam. Studies with biochemical and molecular techniques indicate that many ESBLs are derivatives of older TEM-1, TEM-2, or SHV-1 beta-lactamases, some differing from the parent enzyme by one to four amino acid substitutions. In addition, resistance in Klebsiella pneumoniae and Escherichia coli to cephamycins and inhibitor compounds such as clavalante have also arisen via acquisition of plasmids containing the chromosomally derived AmpC beta-lactamase, most commonly encoded by Enterobacter cloacae, Pseudomonas aeruginosa, and Citrobacter freundii. It is of particular concern that genes encoding the beta-lactamases are often located on large plasmids that also contain genes for resistance to other antibiotic classes including aminoglycosides, tetracycline, sulfonamides, trimethoprim, and chloramphenicol. Furthermore there is an increasing tendency for pathogens to produce multiple beta-lactamases. These developments, which occur over a wide range of Gram-negative genera, represent a recent evolutionary development in which common Gram-negative pathogens are availing themselves of increasingly complex repertoires of antibiotic resistance mechanisms. Clinically, this increases the difficulty of identifying effective therapies for infected patients. Thus, there is a need for techniques that can quickly and accurately identify the types of beta-lactamases that may be present in a clinical isolate or sample, for example. This could have significant implications in the choice of antibiotic necessary to treat a bacterial infection. Summary of the invention The present invention is directed to the use of oligonucleotide primers specific to nucleic acids characteristic of (typically, genes encoding) certain beta-lactamases. More specifically, the present invention uses primers to identify family specific beta-lactamase nucleic acids (typically, genes) in samples, particularly, in clinical isolates of Gram-negative bacteria. Specific primers of the invention include the primer sequences set forth in SEQ ID NOs: 1-45. As used herein, a nucleic acid characteristic of a beta-lactamase enzyme includes a gene or a portion thereof. A “gene” as used herein, is a segment or fragment of nucleic acid (e.g., a DNA molecule) involved in producing a peptide (e.g., a polypeptide and/or protein). A gene can include regions preceding (upstream) and following (downstream) a coding region (i.e., regulatory elements) as well as intervening sequences (introns, e.g., non-coding regions) between individual coding segments (exons). The term “coding region” is used broadly herein to mean a region capable of being transcribed to form an RNA, the transcribed RNA can be, but need not necessarily be, further processed to yield an mRNA. Additionally, a method for identifying a beta-lactamase in a clinical sample is provided. Preferably, the clinical sample provided is characterized as a Gram-negative bacteria with resistance to beta-lactam antibiotics. The method includes, providing a pair of oligonucleotide primers, wherein one primer of the pair is complementary to at least a portion of the beta-lactamase nucleic acid in the sense strand and the other primer of each pair is complementary to a different portion of the beta-lactamase nucleic acid in the antisense strand; annealing the primers to the beta-lactamase nucleic acid; simultaneously extending the annealed primers from a 3′ terminus of each primer to synthesize an extension product complementary to the strands annealed to each primer wherein each extension product after separation from the beta-lactamase nucleic acid serves as a template for the synthesis of an extension product for the other primer of each pair; separating the amplified products; and analyzing the separated amplified products for a region characteristic of the beta-lactamase. The method, described above, can employ oligonucleotide primers that are specific for nucleic acid of the TEM family of beta-lactamases, the K1 beta-lactamases, the PSE family of beta-lactamases, and the SHV family of beta-lactamases. Additional primers that can be used include those that are specific for nucleic acid of the AmpC beta-lactamases found in Enterobacter cloacae, Citrobacter freundii, Serratia marcescens, Pseudomonas aeruginosa, and E. coli. Still other oligonucleotide primers that are suitable for use in the method of the present invention include primers that are specific for nucleic acid of the plasmid-mediated AmpC beta-lactamases designated as FOX-1, FOX-2, or MOX-1; primers specific for nucleic acid of the OXA-9 beta-lactamase; primers specific for nucleic acid of the OXA-12 beta-lactamase; primers specific for the nucleic acid group of OXA beta-lactamases representing OXA-5, 6, 7, 10, 11, 13, and 14 beta-lactamases; primers specific for the OXA-1 beta-lactamases; and primers specific for nucleic acid of the group of OXA beta-lactamases representing OXA-2, 3, and 15 beta-lactamases. EXAMPLE 1 Klebsiella pneumoniae with ESBLs and a Plasmid-mediated AmpC Beta-lactamase Materials and Methods Klebsiella pneumoniae 225 Klebsiella pneumoniae 225 was isolated from a 43-year-old white male patient who was working in a New York City sewage canal on Apr. 26, 1996. While in a pit containing a layer of sewage, a screen fell, knocking the patient down in the sewage. The patient struck his left forehead, sustaining a severe laceration approximately 20 centimeters (cm) long and down to the skull and was unconscious for approximately five minutes. The patient was taken to a local hospital where the wound was surgically debrided, irrigated, and closed. Intravenous cefazolin and gentamicin were administered pre-operatively, and cefazolin was continued 24 hours post-operatively. The patient was discharged and returned to Omaha, Nebr., four days later. On May 2, 1996, the patient was seen by an Omaha surgeon who diagnosed wound infection, ordered culture of the wound drainage, and initiated therapy with cephalexin and penicillin. The culture yielded growth of K. pneumoniae, Aeromonas hydrophila and Proteus penneri. By May 24, 1996, the patient was experiencing significant swelling and pain in the left scalp area, significant weakness and dizziness, and a low grade fever. An infectious disease consult was ordered, and the patient was hospitalized for further evaluation and management. Laboratory findings included a white blood cell count of 21,000 per cubic millimeter (cmm). Aspiration of a bulging left temporal mass from the patient yielded 7 milliliter (ml) of purulent fluid from which K. pneumoniae and A. hydrophila were cultured. The patient was empirically treated with piperacillin/tazobactam and ciprofloxacin. On May 25, 1996, the patient had incision drainage and debridement of the wound. Operative findings as well as preoperative CT scan of the patient's head did not reveal osteomyelitis. There was an abscess commencing in the region of the left zygoma and extending superior to the parietal region, with two small opaque foreign bodies in the caudal aspect of the collection. On May 27, 1996, following antibiotic susceptibility results, therapy was changed to imipenem/cilastatin. The drain was removed and the patient was discharged on May 29, 1996, and treated at home with intravenous imipenem/cilastatin via a peripheral inserted central catheter. After four weeks of therapy all signs of inflammation resolved. The patient remained free from infection at follow-up on Dec. 10, 1996. Susceptibility Tests Susceptibility tests were performed by microdilution methodology using the MicroScan Walkaway system (Dade MicroScan Inc., Sacramento, Calif.) and by NCCLS microdilution methodology in Mueller-Hinton broth (CM 405, Oxoid, Basingstoke, England) using an inoculum of approximately 5×105 CFU/ml (National Committee for Clinical Laboratory Standards, 1997, Approved Standard M7-A4) and also by NCCLS disk diffusion methodology (National Committee for Clinical Laboratory Standards, 1997, Approved Standard M2-A6). Clavulanate Double-Disk Potentiation Test Using the procedure of Brun-Buisson et al., Lancet., ii 302-306 (1987), a Mueller-Hinton agar plate (CM 337, Oxoid, Basingstoke, England) was inoculated with K. pneumoniae 255 as for a standard disk diffusion test. Disks (BBL, Cockeysville, Md.) containing aztreonam, cefotaxime, ceftriaxone, and ceftazidime were strategically placed around an amoxicillin-clavulanate disk prior to incubation at 35° C. ESBL production was inferred by the presence of characteristic distortions of the inhibition zone indicative of clavulanate potentiation of the test drug. Three-Dimensional Test Using a modification of the procedure of Thomson and Sanders, Antimicrob. Agents Chemother., 36:1877-1882 (1992), the surface of a Mueller-Hinton agar plate was inoculated with E. coli ATCC 25922 as for a standard disk diffusion test. A slit made in the agar with a sterile no. 11 scalpel blade was then inoculated with a heavy suspension of cells of K. pneumoniae 225 that had been grown to logarithmic phase in 10 ml tryptone soy broth (CM 129, Oxoid), centrifuged, and resuspended in 100 microliter (μl) TRIS EDTA buffer (T-9285 Sigma Chemical Co., St. Louis, Mo.) for 40 minutes. Disks containing aztreonam, cefotaxime, ceftriaxone, ceftazidime and cefoxitin were placed on the agar 3 millimeters (mm) away from the inoculated slit, and the plate was incubated in the usual manner. Enzymatic inactivation of the antibiotics was inferred if the margin of the inhibition zone was distorted in the vicinity of the slit in a manner that indicated loss of drug activity (hydrolysis) as the drug diffused through the inoculated slit. Isoelectric Focusing, Cefotaxime Hydrolysis, and Inhibitor Determinations Using a modification of the methods of Sanders et al., Antimicrob. Agents Chemother., 30:951-952 (1986), Bauernfeind et al., Infection, 18 ;294-298 (1990), and Thomson et al., Antimicrob. Agents Chemother., 35;1001-1003 (1991), sonic extracts of K. pneumoniae 225 and strains of E. coli that produced reference beta-lactamases, were characterized by determining the isoelectric focusing point (pI) of each beta-lactamase, inhibitor profile in the presence and absence of 1,000 micromolar (μM) clavulanate and 1,000 μM cloxacillin, and ability to hydrolyze 0.75 μg/ml cefotaxime solution. Plasmid Isolations Plasmid DNA isolated using alkaline lysis (Manniatis et al., Molecular Cloning: A Laboratory Manual, Cold Spring Harbor, N.Y., 1982) was performed with the following modifications. Cell pellets were washed twice with 3% TRITON-X 100 dissolved in Tris-ethylene diaminetetraacetate (Tris-EDTA) (pH 8). After neutralization, supernatant was extracted with phenol plus 1/10 volume 10% sodium dodecyl sulfate (SDS) followed by one extraction using phenol:chloroform:isoamyl alcohol (25:24:1) followed by one or more chloroform:isoamyl (24:1) extractions until supernatant was clear. When designated, some samples were treated with plasmid-safe DNase (Epicentre Technologies, Madison, Wis.) as described by the manufacturer. Plasmid DNA was electrophoresed on the day of preparation to decrease the possibility of DNA damage (nicking) during storage. Plasmids were separated by agarose (0.8 %) gel electrophoresis using 1× Tris-acetate-EDTA (TAE) as the buffer system. In some cases, plasmids were visualized without previous isolation by lysing the bacterial cells within the well of the agarose gel. One colony of Klebsiella pneumoniae 225 was suspended into 5 μl of protoplasting buffer (30 mM Tris-HCl (pH 8), 5 mM EDTA, 50 mM NaCl, 20% weight by volume (w/v) sucrose, 50 μg/ml RNase A and 50 μg/ml lysozyme; the RNase A and lysozyme added just prior to use) and incubated 30 minutes at 37° C. Into each well of the agarose gel, 2 μl of room temperature lysis buffer (89 mM Tris (pH 8.3), 89 mM boric acid, 25 mM EDTA, 2% w/v SDS, 5% w/v sucrose and 0.04% Bramaphenol blue) was loaded just prior to the addition of protoplast suspension. The protoplast suspension was loaded and the gel was run for 15 minutes at 30 volts to lyse the protoplasts. After 15 minutes the voltage was increased to 120 volts and the gel was run for 1-1.5 hours. Before staining in ethidium bromide (0.5 μg/ml), the gel was washed in large volumes of water with at least two changes to remove the SDS. The plasmid bands were visualized with a UV transilluminator. The gel consisted of 4.8% agarose, 1× Tris-borate-EDTA (TBE) (89 mM Tris, 89 mM boric acid, and 2.5 mM EDTA) (pH 8.3) and 10% SDS. The running buffer was 1× TBE plus 10% SDS. Southern Analysis Plasmid DNA was prepared by alkaline lysis separated as described above. To achieve high resolution separation, gels were electrophoresed for 17-18 hours at 35 volts. DNA was transferred using 0.4 M NaOH to Zeta-Probe blotting membranes using a vacuum blotter (Bio-RAD) as described by the manufacturer. TEM specific probes (5′-TGCTTAATCAGTGAGGCACC-3′ (SEQ ID NO:1) nucleotides 1062-1042; numbering of Sutcliff, Proc. Nat. Aca. Sci. USA, 75:3737-3741 (1978)) and SHV (5′-TTAGCGTTGCCAGTGCTCG-3′ (SEQ ID NO:11 nucleotides 988-970; numbering of Mercier et al., Antimicrob. Agents Chemother., 34:1577-1583 (1990)) were labeled using the Genius System Oligonucleotide 3′-End labeling kit (Boehringer Mannheim, Indianapolis, Ind.). Prehybridization and hybridization followed the recommendation of the manufacturer using 1% SDS at 37° C. Initially, blots were washed with 5× SSC (twice at room temperature for 5 minutes and twice at room temperature for 30 minutes followed by washings using tetramethylammonium chloride (TMAC); once at 37° C. for 15 minutes and twice at 48° C. for 20 minutes. Labeled probe hybridized to plasmid DNA was detected using the Genius Luminescent detection kit (Boehringer Mannheim) as described by manufacturer. PCR amplifications were carried out as described below in Example 3 with the following modifications. The composition of the reaction mixture was 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 1.5 mM MgCl2, 0.2 mM (each) of the four deoxynucleoside triphosphates, and 1.2 U of Taq polymerase (GIBCO, Gaithersburg, Md.) in a total volume of 48 μl. A total of 2 μl of sample lysate containing the DNA template was added to the reaction mixture. The PCR parameters consisted of initial denaturation step at 95° C. for 5 minutes followed by 24 amplification cycles consisting of a denaturation step of 96° C. for 15 seconds; primer annealing at 55° C. for 15 seconds and extension at 72° C. for 2 minutes. Amplified product was detected by agarose (2%) gel electrophoresis using a 1× TAE buffering system. Some of the PCR products were sequenced by automated PCR cycle-sequencing with dye-terminator chemistry using a DNA stretch sequencer from Applied Biosystems (Foster City, Calif.). Restriction Fragment Length Polymorphism (RFLP) SHV-specific PCR products (1/5 volume) were used directly in a restriction endonuclease assay (Niesch-Inderbinen et al., Eur. J. Clin. Microbiol. Infect. Dis., 15:398-402 (1996)) using the restriction endonuclease, NheI (New England Biolabs, Beverly, Mass.). Enzyme reactions were carried out as directed by the manufacturer. To ensure that each sample received the same amount of enzyme, an enzyme mix containing the buffer system and enzyme was aliquoted to each sample. Samples were resolved using 2% agarose and a 1× TAE buffer system. Transformation and Conjugation Transformations were done using a modified Hanahan method (TSS) described by CLONTECH (CLONTECH Laboratories, Inc., Palo Alto, Calif., Transformer site. Directed Mutagenesis Kit—2nd version). Plasmids were separated as described above, excised from the gel, and electroeluted from the gel slice. The DNA was transformed into E. coli HB101. Conjugation experiments were carried out by filter mating using E. coli, strain C600, as the recipient. Transconjugants were selected on Luria-Bertani agar plates containing 30 μg/ml of naladixic acid. An endol test was performed on the transconjugant (E. coli C600) to further differentiate it from the donor (K. pneumoniae 225). Results Susceptibility Tests The results of the microdilution tests performed with K. pneumoniae 225 using the NCCLS microdilution methodology were as follows: MIC>64 μg/ml: ticarcillin, ticarcillin/clavulanate, piperacillin, piperacillin/tazobactam, ceftazidime, cefixime, loracarbef, cephalothin, cefazolin, cefoxitin, aztreonam, ampicillin/sulbactam MIC 64 μg/ml: amoxicillin/clavulanate, cefpodoxime MIC 16 μg/ml: ceftriaxone, cefotaxime MIC 1 μg/ml: imipenem, cefepime MIC 0.5 μg/ml: ciprofloxacin MIC 0.06 μg/ml: meropenem Other susceptibility results obtained in MicroScan tests were (MicroScan MICs shown in parentheses): Resistant: cefuroxime (>16 μg/ml), gentamicin (>8 μg/ml), tobramycin (>8 μg/ml), amikacin (>32 μg/ml), trimethoprim/sulfamethoxazole (>2/38), tetracycline (>8 μg/ml), nitrofurantoin (>64 μg/ml), chloramphenicol (>16 μg/ml) Susceptible: ofloxacin and levofloxacin (both 2 μg/ml), cefotetan (16 μg/ml) Discrepancies between MicroScan and conventional NCCLS results were obtained with ciprofloxacin (0.5 μg/ml in conventional microdilution test, susceptible by disk test, 2 μg/ml in MicroScan test) and cefotetan (resistant in disk test with 12 mm zone diameter, susceptible by MicroScan, 16 μg/ml). The susceptibility results for the Aeromonas hydrophila isolate were not considered unusual and are not reported. Double Disk and Three Dimensional Tests The clavulanate double-disk potentiation test was positive with each of the antibiotics tested, indicating that K. pneumoniae possessed one or more clavulanate-sensitive beta-lactamases capable of hydrolyzing aztreonam, cefotaxime, ceftriaxone, and ceftazidime. This result was consistent with beta-lactamase activity of Bush group 2be or, possibly, high level activity of Bush group 2b. The three dimensional test was positive for each of the antibiotics tested, aztreonam, cefotaxime, ceftriaxone, ceftazidime, and cefoxitin, indicating β-lactamase-mediated hydrolysis of each drug. The positive result with cefoxitin was notable, being consistent with production of a Bush group 1 beta-lactamase. Isoelectric focusing-Based Tests Isoelectric focusing yielded five beta-lactamase bands with pI values of 5.4, 6.8, 7.6, 8.2, and 39.0, values consistent with TEM-1 (pI 5.4), PSE-3, OXA-9 or unknown enzyme (pI 6.8), SHV-1, SHV-2, or SHV-8 (pI 7.6), SHV-5 (pI 8.2) and AmpC (pI 39.0) (Table 2, below). Only the pI 39.0 enzyme was resistant to clavulanate, confirming that this was a Bush group 1 (AmpC) beta-lactamase. The beta-lactamase bands which hydrolyzed cefotaxime, as detected by microbiological assay, were pI 7.6, pI 8.2, and pI 39.0. These results suggested the presence of clavulanate-sensitive ESBLs of pI values 7.6 and 8.2, and added support to the identification of an AmpC enzyme with a pI value 39.0. These results are supported and/or confirmed by PCR. Polymerase Chain Reaction (PCR) PCR analysis was used initially to confirm and/or identify the beta-lactamases observed during isoelectric focusing (Table 2, above). Primer sets specific for the TEM or SHV gene families, Enterobacter AmpC, OXA-9 and integron sequences were used in a PCR (Table 1, above). PCR identified the presence of TEM and SHV-like genes, an Enterobacter AmpC-like gene, the OXA-9 gene and integron sequences (data not shown). Plasmids Multiple plasmid isolations from K. pneumoniae 225 revealed the organism carried only two plasmids. Using a supercoiled DNA ladder, the estimated sizes of these plasmids were approximately 17 kb and approximately 90 kb. Three different isolation procedures were used to extract plasmid DNA; alkaline lysis (Manniatis et al., Molecular Cloning: A Laboratory Manual, Cold Spring Harbor, N.Y., 1982), lysozymes/SDS (Crosa et al., “Plasmids” In Gerhardt et al., Manual of Methods for General Bacteriology, American Society for Microbiology, Washington, DC, pages 266-282, 1981), and cell lysis and extraction within the well of the gel. All procedures yielded the same two plasmids indicating that plasmids were not being lost during any one type of isolation procedure. Alkaline lysis yielded the purest preparation of plasmid DNA and was therefore used for southern blot analysis. It was possible that residual chromosomal DNA comigrated and therefore masked a possible plasmid. To address this, an enzyme, plasmid-safe DNase (Epicentre Technologies, Madison, Wis.), which does not cleave supercoiled DNA, was used to treat the DNA plasmid preparations before electrophoresis. Treatment with plasmid-safe DNase degraded the chromosomal DNA band while having no effect on the 17 kb and 90 kb plasmids, indicating no other plasmids were present in K. pneumoniae 225 (data not shown). Southern Analysis It was surprising that an organism expressing 5 and possibly 6 beta-lactamases would have only two extrachromosomal pieces of DNA. Therefore, whether beta-lactamase genes were encoded on one or both plasmids was evaluated. Southern analysis revealed that both the 90 kb and 17 kb plasmid encoded TEM-like genes, however, only the 90 kb plasmid encoded the SHV-like genes (data not shown). Transformation and Conjugation It was possible that the plasmid-mediated AmpC gene encoded by K. pneumoniae 225 cross-hybridized with the TEM-specific probe and that one or both plasmids encoded the AmpC enzyme observed during isoelectric focusing. In an attempt to isolate each plasmid from the other, transformation experiments were carried out. Each plasmid was extracted and gel purified. The approximately 17 kb plasmid was transformed into E. coli HB101, and selected using ampicillin. After confirming that only a 17 kb plasmid was present in the HB 101 transformants, a disk diffusion assay was performed. The transformant did not exhibit diminished susceptibility to any of the drugs in Table 3 except ampicillin and amikacin, indicating that the 17 kb plasmid did not encode AmpC or extended spectrum beta-lactamase genes. Several attempts to transform the large plasmid into E. coli, (strains HB101 and MV1190) failed. Transformation using the 90 kb plasmid produced transformants that were resistant only to ampicillin. When plasmid DNA was isolated from these transformants many sized plasmids, all less than 90 kb, were present (data not shown). The data obtained from the transformation of the 17 kb plasmid suggested that the plasmid-mediated AmpC gene was encoded on the 90 kb plasmid. Therefore, conjugation experiments were performed. Conjugation between K. pneumoniae 225 and E. coli C600 resulted in the transfer of both the 90 kb and 17 kb plasmids. Cefoxitin resistance of the transconjugant indicated transfer of the AmpC gene (Table 3). Taken together, these data strongly suggest that the AmpC gene is located on the 90 kb plasmid. Restriction Fragment Length Polymorphism (RFLP) Some ESBL-SHV enzymes with a pI of 7.6 (SHV-2, SHV-7) contain a glycine to serine amino acid substitution at position 238. In the structural SHV-gene the nucleotide mutation resulting in the amino acid substitution creates a new endonuclease restriction site, NheI. This restriction site is not present in the structural gene of SHV-1, SHV-6, SHV-8, or SHV-11, but these enzymes also have a pI of 7.6. Therefore, RFLP analysis using NheI can help distinguish between these two groups of enzymes (Nüesch-Inderbinen et al., Eur. J. Clin. Microbiol. Infect. Dis., 39:185-191 (1996)). Isoelectric focusing data suggested that the identity of the pI 7.6 beta-lactamase could be SHV-2, SHV-6, SHV-8 or a hyperproducer of SHV-1. To help distinguish between SHV-2 and SHV-6, SHV-8 or a hyperproducer of SHV-1, RFLP analysis on SHV-specific PCR products from K. pneumoniae 225 were performed using NheI. The presence of the NheI site in the SHV-specific PCR product will result in 2 bands: 219 bp and 164 bp. The absence of the NheI site will result in no cleavage and a full length fragment: 383 bp. SHV-specific PCR products amplified from template prepared from K. pneumoniae 225 show both full length and cleaved products. These data suggest that SHV-1, SHV-6, or SHV-8 as well as an SHV ESBL is encoded by K. pneumoniae 225 DNA. EXAMPLE 2 Beta-lactamases Responsible for Resistance to Expanded-Spectrum Cephalosporins among Klebsiella pneumoniae, Escherichia coli and Proteus mirabilis Isolates Recovered in South Africa Materials and Methods Bacterial Strains During a period of three months in 1993, 37 strains of Klebsiella pneumoniae (13 blood, 5 burn, 7 wound, 11 tracheal isolates), 4 strains of Proteus mirabilis (all wound isolates) and 4 strains of Escherichia coli (1 blood, 1 burn, 2 wound isolates) were collected from patients at the following medical centers in South Africa: Tygerberg Hospital near Cape Town, King Edward VIII Hospital in Durban, Chris Hani Baragwanath Hospital in Soweto and Pretoria Academic Hospital in Pretoria. The strains were provided in response to a request for all strains of Enterobacteriaceae, lacking inducible beta actamases, that were intermediate or resistant to cefotaxime or ceftazidime. The total number of strains screened is unknown, and at this time the referring hospitals did not perform more sensitive screening tests for ESBL detection. Therefore accurate prevalence data were not obtained. Thirty-four of the 43 patients involved (including all from whom blood isolates were obtained) had received a third generation cephalosporin during the four weeks prior to isolation of the above organisms. Fifteen patients (including 8 blood isolate patients) were receiving either cefotaxime or ceftazidime at the time the isolates were cultured and were considered not to be responding to these agents. Susceptibility Testing and Antibiotics Antibiotic susceptibility was determined by standard disk diffusion (NCCLS Standard M2-T4, 1994) and agar dilution (NCCLS Standard M7-T2, 1994) procedures. Standard powders of antimicrobial agents were kindly provided by the following companies: piperacillin and tazobactam (Lederle Laboratories, Wayne, N.J.); cefoxitin and imipenem, (Merck, Rathway, N.J.); cefotaxime, (Hoechst-Roussel Pharmaceuticals Inc., Somerville, N.J.); ceftazidime, (Glaxo Group Research Ltd., Greenford, England); aztreonam and cefepime, (Bristol-Myers Squibb, Princeton, N.J.). Disks for agar diffusion were obtained from Becton Dickinson Microbiology Systems (Cockeysville, Md.). For quality control purposes, the following quality control strains were run simultaneously with the test organisms E. coli ATCC 25922, Pseudomonas aeruginosa ATCC 27853, E. coli ATCC 35218, and Staphylococcus aureus ATCC 29213. Throughout this study, results were interpreted using NCCLS criteria for disk diffusion (NCCLS Standard M2-T4, 1994) and broth dilution (NCCLS Standard M7-T2, 1994). Double-Disk Test All the strains were screened for the production of extended-spectrum beta-lactamases by using the double-disk test as described by Jarlier et al., Rev. Infect. Dis. 10:867-878 (1988). A potentiation of the zones of cefotaxime, ceftriaxone, ceftazidime or aztreonarn by clavulanic acid represented a positive test and was indicative of possible presence of an extended-spectrum beta-lactamase. Beta-lactamase Characterization Overnight cultures in 5 ml trypticase soy broth were diluted with 45 ml fresh broth and incubated with shaking for 4 hours at 37° C. Cells were harvested by centrifugation at 4° C., washed with 1 M potassium-phosphate buffer (pH 7.0), suspended and sonicated. After sonication, crude extracts were obtained by centrifugation at 5,858×g for 1 hour. One strain, K. pneumoniae Pit 68, with a suspected AmpC beta-lactamase, was induced with cefoxitin as described below in Example 3. The rate of hydrolysis of 100 μM solutions of nitrocephin, cephalothin, cefotaxime, ceftazidime and aztreonam was performed by spectrophotometric assays on crude beta-lactamase extracts (Naumovski et al., Antimicrob. Agents Chemother., 36:1991-1996 (1992)). The beta-lactamases in the sonic extracts were assessed for isoelectric points (pIs), general substrate and inhibitor characteristics in polyacrylamide gels. As controls, crude beta-lactamase preparations from the following organisms possessing different TEM and SHV enzymes were examined simultaneously with the K. pneumoniae, E. coli and P. mirabilis strains: TEM-1 [from E. coli RTEM (R6K)], TEM-2 [from E. coli 1752E (RP1)], TEM-10 [from E. coli C600 (pK2)], TEM-26 [from E. coli HB101 (PJPQ101), SHV-1 [from E. coli J53 (R1010)], SHV-2 [from Klebsiella ozaenae 2180], SHV-3 [from E. coli J53 (pUD18)], SHV-4 [from E. coil J53-2 (pUD21)] and SHV-5 [from E. coli ClaNal (pAFF2)]. DNA Amplification Using Polymerase Chain Reaction (PCR) The organisms were inoculated into 5 ml of Luria Bertani (LB) broth (Difco, Detroit, Mich.) and incubated for 20 hours at 37° C. with shaking. Cells from 1.5 ml of overnight culture were harvested by centrifugation at 17,310×g in an Hermle centrifuge for 5 minutes. After the supernatant was decanted, the pellet was resuspended in 500 μl of distilled water. The cells were lysed by heating at 95° C. for 10 minutes and cellular debris was removed by centrifugation for 5 minutes at 17,310×g. The supernatant was used as source of template for amplification. The following oligonucleotide primers specific for the SHV and TEM genes were designed by using MacVector version 4.5 (Kodak/IBI): SHV genes: A [5′-(CACTCAAGGATGTATTGTG) -3′] (SEQ ID NO:10) and B [5′-(TTAGCGTTGCCAGTGCTCG)-3′] (SEQ ID NO:11) corresponding to nucleotide numbers 103 to 121 and 988 to 970, respectively, of Mercier et al., Antibicrob. Agents Chemother. 34:1577-1583 (1990)); TEM genes: C [5′-(TCGGGGAAATGTGCGCG)-3′ (SEQ ID NO:5) and D [5′-(TGCTTAATCAGTGAGGCACC)-3′ (SEQ ID NO:1) corresponding to nucleotide numbers 90 to 105 and 1062 to 1042, respectively, of Sutcliff et al., Proc. Nat. Aca. Sci., USA, 75:3737-3741 (1978). Primers A and B amplified a 885 base pair fragment while primers C and D amplified a 971 base pair fragment. The specificity of the SHV and TEM primers for amplification of SHV and TEM genes respectively was tested by using the following beta-lactamase controls; TEM-1 (pACYC177), MIR-1 (from K. pneumoniae 96D) and SHV-7 (pCLL3410). PCR amplifications were carried out on a DNA Thermal Cycler 480 instrument (Perkin-Elmer, Cetus, Norwalk, Conn.) using the Gene Amp DNA amplification kit containing AmpliTaq polymerase (Perkin Elmer, Roche Molecular Systems, Inc., Branchburg, N.J.). The composition of the reaction mixture was as follows: 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 0.1% TRITON X-100, 1.5 mM MgCl2, 0.2 mM (each) of the four deoxynucleoside triphosphates, and 1.2 U of AmpliTaq in a total volume of 49 μl. A total of 1 μl of sample lysate was added to the reaction mixture, and was centrifuged briefly before 50 μl of mineral oil was layered on the surface. The PCR program consisted of an initial denaturation step at 96° C. for 15 seconds; followed by 24 cycles of DNA denaturation at 96° C. for 15 seconds, primer annealing at 50° C. for 15 seconds and primer extension at 72° C. for 2 minutes. After the last cycle the products were stored at 4° C. The PCR products (1/10 volume) were analyzed by electrophoresis using 1.4% agarose gels in TAE buffer (0.04 M Tris-acetate, 0.002 M EDTA [pH 8.5]). The gels were stained with ethidium bromide and the PCR products were visualized with ultra-violet light. A single band was observed for TEM amplified products using a single primer set. Two amplified products were observed with the SHV primer set. The larger product which corresponded to the expected size of the SHV specific product was gel purified using a 1.4% agarose in TAE gel and the purified PCR product was used for sequence analysis. PCR products were sequenced by automated PCR cycle-sequencing with dye-terminator chemistry using a DNA stretch sequencer from Applied Biosystems. Results Resistance Phenotypes All the strains except K. pneumoniae Pit 68, gave a positive disk potentiation when using cefotaxime, ceftriaxone, aztreonam and/or ceftazidime disks. Minimum inhibitory concentrations of piperacillin, piperacillin/tazobactam, cefotaxime, ceftazidime, aztreonam and cefoxitin revealed 3 different resistance phenotypes (Kpn1, 2 and 3) in the K. pneumoniae strains, and 2 (Ecl and 2) in E. coli strains. The phenotypes Kpn1 and Ec1 involved high level resistance to ceftazidime (MIC>128 μg/ml) but susceptibility to cefotaxime (MIC range 0.25-1 μg/ml), while Kpn2 and Ec2 involved decreased susceptibility to both cefotaxine (MIC range 4-64 μg/ml) and ceftazidime (MIC range 4-128 μg/ml). Kpn 3, represented by K. pneumoniae Pit 68, involved resistance to cefoxitin (MIC>128 μg/ml) and decreased susceptibility to cefotaxime, ceftazidime and aztreonam (MIC>2 μg/ml) (Table 4). The P. mirabilis isolates showed decreased susceptibility to ceftazidime (MIC range 16-64 μg/ml) and susceptibility to cefotaxime (MIC range 0.25-0.5 μg/ml). Beta-lactamases Strains representing the Kpnl and Ecl phenotypes produced beta-lactamases with pI values of 5.6 and 7.6 respectively, while phenotypes Kpn2 and Ec2 involved enzymes with pI's of 5.4, 7.6, and 8.2 (Table 5). K. pneumoniae Pit 68, representing phenotype Kpn3, produced two beta-lactamases with pls of 5.4 and 8.0. The P. mirabilis strains showed a single enzyme with a pI value of 5.6 (Table 5). The enzymes of pI 5.4, 5.6, 7.6 and 8.2 aligned with TEM-1 (pI 5.4), TEM-10 or 26 (pI 5.57), SHV-1, 2 or 8 (pI 7.6) and SHV-5 (pI 8.2) respectively (Table 5). It was therefore necessary to investigate these enzymes further. On isoelectric focusing gels, all of the beta-lactamases except for the enzymes with a pI of 8.0 were inhibited by clavulanate, a characteristic of Bush group 2 enzymes. The enzyme with a pI of 8.0 was inhibited by cloxacillin which correlates with Bush group 1 cephalosporinases. The substrate-based technique showed hydrolysis of 0.75 Fg/ml cefotaxime at the bands focusing at: 5.6, 8.0, 8.2 and some enzymes with a pI of 7.6 (Table 5). Control enzymes of TEM-10, TEM-26, SHV-2 and SHV-5 showed hydrolysis of cefotaxime in this assay. DNA Ampilification and Sequencing The DNA from organisms producing single beta-lactamases were amplified and sequenced. Strains producing ESBLs with pIs of 5.6, which aligned with TEM-10 and TEM-26, were amplified with the TEM primers (Table 7). Amino acids at positions 104, 164 and 240 (Ambler numbering (1)) were utilized to determine that this enzyme was more similar to TEM-26. Amino acids deduced from amplicon sequences included lysine at position 104, serine at position 164 and glutamine at position 240 (Table 7). Strains producing ESBLs with pI values 7.6 and 8.2, which aligned with SHV-2 and SHV-5 respectively, were amplified with SHV primers (Table 7). Amino acids at positions 205, 238 and 240 (Labia numbering (2)) were used to identify the ESBL involved. Arginine at position 205, serine at position 238 and glutamic acid at position 240 of the deduced amino acid sequence of strains producing an ESBL with a pI of 7.6 indicated the presence of SHV-2 (Table 7). K. pneumoniae Pit 82, producing an ESBL with a pI 8.2, had a lysine at position 240 indicating the presence of SHV-5 .