Potentiation of Antibiotic Bactericidal Activity by Normal Human Serum

Combinations of certain antibiotics and normal human serum at concentrations at which there was no killing by the agents when used alone were found to be bactericidal for Escherichia coli K-12 cells. This effect was observed with tetracycline, streptomycin (SM), trimethoprim, and ampicillin, but not with chloramphenicol or nalidixic acid. Synergy between SM and human serum was also observed against four of nine smooth strains of E. coli. A plasmid-bearing strain of E. coli K-12 was also killed by combinations of tetracycline or SM plus serum, even though the plasmid conferred resistance to tetracycline and SM. Evidence is presented that the synergy between antibiotics and serum is due to a complement-mediated effect on the bacterial cells that makes the cells more susceptible to the bactericidal effects of the antibiotics.     

Incidence of antibiotic-resistant Escherichia coli associated with frozen chicken carcasses and characterization of conjugative R plasmids derived from such strains.

Escherichia coli were isolated at concentrations of about 10(2)/ml from the fluid obtained after thawing each of five frozen chicken carcasses. Between 13 and 89% of the E. coli were resistant to mercury(II) or to at least one of eight antibiotics tested. Multiple resistance was more common than single resistance, and resistances to tetracycline, streptomycin, sulfathiazole, or chloramphenicol were more frequently encountered than was resistance to ampicillin or mercury(II). Resistance to kanamycin, gentamicin, or trimethoprim occurred rarely. Upwards of 30% of the E. coli apparently carried conjugative plasmids and could transfer at least one of their resistance determinants to E. coli K-12. Twenty-five conjugative R plasmids obtained in this way were characterized within strain K-12 by a variety of phenotypic criteria. Twenty could be assigned to the incompatibility groups Inc X, Inc N, Inc FI, or Inc I alpha.     

Multiresistance plasmid from commensal Neisseria strains.

Antibiotic-resistant commensal strains of Neisseria spp. and Branhamella catarrhalis were isolated from throat cultures, on the basis of their capacity to grow in the presence of penicillin, streptomycin, or sulfamethoxazole-trimethoprim. Several strains, which belonged to different species of Neisseria, were resistant to beta-lactams, streptomycin, sulfamethoxazole, and trimethoprim, harbored a 6.0-megadalton plasmid with identical HinfI restriction patterns, and produced beta-lactamase and streptomycin phosphotransferase. The resistance determinants for beta-lactams, streptomycin, and sulfamethoxazole, but not for trimethoprim, were transferred from all these strains to Escherichia coli by conjugation or transformation. The resulting transconjugants or transformants acquired the plasmid and the capacity to produce beta-lactamase and streptomycin phosphotransferase. The 6.0-megadalton plasmid complemented a mutation which determines production of thermosensitive dihydropteroate synthetase in E. coli. We conclude that an R plasmid coding for beta-lactamase, streptomycin phosphotransferase, and a sulfonamide-resistant dihydropteroate synthetase is common to these strains.     

Quinolone, fluoroquinolone and trimethoprim/sulfamethoxazole resistance in relation to virulence determinants and phylogenetic background among uropathogenic Escherichia coli

INTRODUCTION: The goal of this study was to assess how resistance to quinolones, fluoroquinolones and trimethoprim/sulfamethoxazole relates to the virulence potential and phylogenetic background of clinical Escherichia coli isolates. METHODS: Among 150 uropathogens (21% resistant to quinolones, 12% resistant to fluoroquinolones and 29.3% resistant to trimethoprim/sulfamethoxazole), E. coli phylogenetic group, 15 virulence-associated genes and 7 O antigens were analysed. Clonal group A (CGA) and genomic PCR profiles were studied among trimethoprim/sulfamethoxazole-resistant isolates. RESULTS: Isolates susceptible to the three antimicrobial agents were significantly associated with phylogenetic group B2, whereas resistant isolates exhibited shifts to non-B2 groups (quinolone and fluoroquinolone-resistant isolates to group A; trimethoprim/sulfamethoxazole-resistant isolates to group D). Diverse virulence traits, including UTI-associated O antigens, were significantly less frequent among resistant isolates, particularly those resistant to fluoroquinolones (median score, 3.9 virulence factors/strain) and also to quinolones (5.2) or trimethoprim/sulfamethoxazole (6.4), as compared with the corresponding drug-susceptible isolates (median scores of 7.9, 8.6 and 7.9, respectively). Among 44 trimethoprim/sulfamethoxazole-resistant isolates, 3 (6.8%) belonged to CGA. All these 3 CGA strains caused pyelonephritis (P=0.02) and exhibited the consensus virulence profile of previously described CGA strains from abroad. CONCLUSIONS: E. coli isolates resistant to quinolones, trimethoprim/sulfamethoxazole and especially fluoroquinolones were associated with reductions in virulence traits and shifts to non-B2 phylogenetic groups. Moreover, fluoroquinolone resistance usually occurred in low-virulence E. coli group A isolates rather than in isolates from groups B2 and D which had lost virulence traits. CGA accounted for 23% of trimethoprim/sulfamethoxazole-resistant E. coli producing pyelonephritis.     

Drug Resistance and R Plasmids in Salmonella typhi Isolated in Korea

A total of 949 strains of Salmonella typhi isolated in Korea from 1968 to 1975 were tested for drug resistance and distribution of R plasmids. Resistance was mostly restricted to streptomycin (SM) and sulfisomidine (SA), singly or in combination, at a low degree. A small number of strains (1.5%) were resistant to four or more drugs: chloramphenicol (CM), tetracycline (TC), SM, SA, ampicillin (AP), and kanamycin (KM). No strain was resistant to nalidixic acid or to a 1:20 mixture of trimethoprim and sulfamethoxazole. Nor was there any strain singly resistant to CM, TC, AP, or KM. Transfer experiments of multiple-drug resistance to Escherichia coli ML1410 showed that all the strains resistant to four or more drugs carried R plasmids, whereas those weakly resistant to three or less drugs did not. The quadruply resistant strains carried one R plasmid determining CM, TC, SM, and SA resistance, and sextuply resistant ones carried two plasmids, one determining CM, TC, SM, and SA resistance and the other determining AP and KM resistance. One strain carrying a plasmid determining AP and KM resistance was also found. The transfer frequency of CM, TC, SM, and SA resistance was much higher than that of AP and KM resistance. The resistance of S. typhi was more efficiently transferred to E. coli at 25°C than at 37°C.     

Molecular cloning and mechanism of trimethoprim resistance in Haemophilus influenzae.

We studied 10 trimethoprim-resistant (Tmpr) Haemophilus influenzae isolates for which agar dilution MICs were 10 to greater than 200 micrograms/ml. Trimethoprim resistance was transferred from two Tmpr H. influenzae isolates to a Tmps strain by conjugation or transformation. Wild-type Tmpr strains and Tmpr transcipients did not contain detectable plasmid DNA. The trimethoprim resistance gene was cloned into a cosmid vector, and recombinant plasmids were transduced into Escherichia coli. A 0.50-kilobase intragenic probe derived from a 12.9-kilobase fragment which encoded trimethoprim resistance hybridized with whole-cell DNA from Tmps and Tmpr strains. Southern blot analysis of restricted DNA from isogenic Tmps and Tmpr H. influenzae indicated that acquisition of trimethoprim resistance involved a rearrangement or change in nucleotide sequence. Hybridization was not seen with DNA derived from Tmpr E. coli containing dihydrofolate reductase I, II, and III genes or with Tmpr Neisseria meningitidis, Neisseria gonorrhoeae, and Pseudomonas cepacia. Southern hybridization with 12 multiply resistant encapsulated H. influenzae strains confirmed that the trimethoprim resistance gene was chromosomally mediated. Dihydrofolate reductase activity was significantly greater in cell sonicate supernatants of Tmpr strains in comparison with isogenic Tmps recipients. Differences were not found in the trimethoprim inhibition profile of dihydrofolate reductase activity in Tmps and Tmpr strains. We conclude that the mechanism of trimethoprim resistance in H. influenzae is overproduction of chromosomally located dihydrofolate reductase.     

Efficacy of trimethoprim in murine experimental infection with a thymidine kinase-deficient mutant of Escherichia coli.

The antimicrobial activity of trimethoprim is antagonized by thymidine in in vitro susceptibility tests. The purpose of this investigation was to determine whether this antagonism also occurred during experimental infection in mice, which have high serum thymidine concentrations. We derived a mutant strain of Escherichia coli, TT-48, incapable of utilizing exogenous thymidine from parent strain E. coli KC-14 and then investigated the in vitro and in vivo antimicrobial activities of trimethoprim, sulfamethoxazole, cefdinir, and ofloxacin against these strains. E. coli TT-48 lacked the activity of thymidine kinase, which catalyzes the conversion of thymidine to thymidylate, but its growth curve remained close to that of the parent strain. The MICs of all of the antimicrobial agents tested, except cefdinir, for the mutant strain were slightly inferior to those for the parent strain. The bactericidal effect of trimethoprim against the parent strain was antagonized by thymidine at concentrations of more than 1 microg/ml, while that against the mutant strain was not affected by thymidine even at the highest concentration (10 microg/ml). The therapeutic efficacy of trimethoprim in experimental murine infections was significantly higher when the mutant rather than the parent strain was used, whereas the therapeutic efficacy of cefdinir or ofloxacin, whose antimicrobial action is independent of folic acid synthesis, was the same with both strains. Unexpectedly, sulfamethoxazole also had similar efficacy against both strains. Thus, high thymidine concentrations antagonized the antimicrobial activity of trimethoprim in vitro and in vivo.     

Molecular analysis of multiple-resistance plasmids transferred from gram-negative bacteria isolated in a urological unit.

Forty-one isolates of multiply resistant gram-negative bacteria causing infection in a urological unit of a Dublin hospital were collected during a 6-month period. Twenty-one isolates transferred multiple resistance to an Escherichia coli K-12 recipient in liquid matings. Serratia marcescens, Proteus morganii, Proteus vulgaris, and E. coli isolates harbored similar 120-megadalton IncC plasmids, whereas Enterobacter cloacae strains transferred a 160-megadalton plasmid of a different Inc group. Southern hybridization experiments were performed with purified fragments cloned from one IncC plasmid as probes. They were hybridized to plasmid sequences in total cellular DNA extracts, showing that the IncC plasmids were very closely related. This suggests that the same plasmid has transferred to different bacterial species in the hospital environment.     

Genetic transfer of antimicrobial resistance and enterotoxigenicity among Escherichia coli strains.

To understand the role of enterotoxin (Ent) plasmids in epidemics of enterotoxigenic (ET) Escherichia coli diarrhea in the United States, we studied the genetics of Ent plasmids in relation to E. coli serotypes and R plasmids. Twenty-nine ET E. coli strains, including all epidemic isolates available at the Centers for Disease Control, Atlanta, Ga. (CDC), were assessed for the ability to transfer antimicrobial resistances (if present) by conjugation, to mobilize a nonconjugative R plasmid, and to cotransfer enterotoxigenicity with R determinants. Of the 12 ET E. coli strains isolated in the United States, 5 were able to transfer R plasmids; one strain cotransferred detectable enterotoxigenicity. Another four U.S. isolates were able to mobilize plasmid DNA, but no toxin production was detected in transconjugants. Of 17 resistant ET E. coli from South Asia, 13 were able to transfer R plasmids; 5 of those 13 cotransferred detectable Ent plasmids. In all, 22 ET E. coli strains (76%) were able to initiate conjugation and genetic transfers. Six of these strains (20%) were able to cotransfer enterotoxigenicity with a conjugative R plasmid at a detectable frequency. One of the six strains transferred R and Ent determinants on a single plasmid. These data are addressed in relation to the observed immobility of Ent and R during outbreaks of ET E. coli, the efficacy of prophylactic tetracycline, and the worldwide occurrence of a limited number of ET E. coli serotypes.     

Synergistic interaction of trimethoprim and sulfamethoxazole on Paracoccidioides brasiliensis.

The in vitro interaction of trimethoprim and sulfamethoxazole on clinical isolates of Paracoccidioides brasiliensis was studied. With complete inhibition and a visual endpoint used as the criteria, three of four strains had minimal inhibitory concentrations that indicated resistance to sulfamethoxazole, and all four strains were resistant to trimethoprim. A marked synergism in inhibition was noted with the combination of these drugs against sulfa-resistant strains. A sulfamethoxazole/trimethoprim ratio of 5:1 was the most synergistic. Fifty percent inhibition, determined spectrophotometrically, of the strains could be achieved with sulfamethoxazole alone. In summary, the striking synergy observed suggests that combination chemotherapy with these drugs deserves further study.     

Occurrence of transposable trimethoprim resistance in clinical isolates of Escherichia coli devoid of self-transmissible resistance plasmids.

Fifty trimethoprim-resistant clinical isolates of Escherichia coli, devoid of self transmissible trimethoprim resistance plasmids, were examined for the presence of trimethoprim resistance transposons. Trimethoprim resistance was mobilized from 12 strains by transposition onto plasmid RP4. The trimethoprim resistance transposons isolated comprised two groups: those with and without linked streptomycin resistance.     

Common Plasmid Specifying Tobramycin Resistance Found in Two Enteric Bacteria Isolated from Burn Patients

Tobramycin-resistant burn wound isolates of Klebsiella pneumoniae and Enterobacter cloacae, together with Escherichia coli K-12 transconjugants from these two strains, were examined for plasmid deoxyribonucleic acid (DNA). All the resistant strains contained a common, high-molecular-weight, covalently closed circular DNA plasmid that was absent in the tobramycin-susceptible E. coli recipient strain. The common plasmid residing in E. cloacae was designated pIE098, and that residing in K. pneumoniae was designated pIE099. Both plasmid species were found to have a molecular mass of approximately 60 x 10(6) daltons and a guanine-plus-cytosine content of 50 mol%. The DNA that was extracted from all of the tobramycin-resistant strains tested was able to hybridize to 86 to 100% with pIE098 and pIE099 [(3)H]DNA generated by EcoRI to produce fragments of a size similar to those generated by BamHI. This study illustrates the usefulness of simple screening methods for antibiotic resistance plasmids in a hospital epidemiological situation.     

Synergistic Activity of Trimethoprim and Amikacin Against Gram-Negative Bacilli

The in vitro effect of trimethoprim on the inhibitory and bactericidal activity of amikacin against 20 strains each of Klebsiella pneumoniae and Serratia marcescens, 15 strains of Escherichia coli, and 10 strains of Pseudomonas aeruginosa was examined by the checkerboard technique in microtiter plates. Trimethoprim had a synergistic effect on the inhibitory and bactericidal activity of amikacin against the majority of non-pseudomonas strains tested. The mean +/- standard deviation fractional inhibitory concentration indexes were 0.59 +/- 0.19 for the Klebsiella strains, 0.48 +/- 0.18 for the Serratia strains, and 0.60 +/- 0.22 for the E. coli strains tested. Respective mean +/- standard deviation fractional bactericidal concentration indexes for these organisms were 0.55 +/- 0.17, 0.54 +/- 0.29, and 0.61 +/- 0.22. A total of 40% of the Klebsiella strains, 80% of the Serratia strains, and 46% of the E. coli strains had a fractional inhibitory concentration equal to or less than 0.25 for both of these antimicrobial agents and were considered to be synergistically inhibited by the combination. By applying this criterion to bactericidal activity, synergy was demonstrated against 50, 65, and 46% of these strains, respectively. All of the Enterobacteriaceae tested were inhibited by clinically achievable concentrations of trimethoprim and amikacin. Antagonism was not demonstrated with any of the organisms tested. Trimethoprim had no antibacterial effect on the Pseudomonas strains and did not alter amikacin's activity against these bacteria.     

Bactericidal activity of orally available agents against methicillin-resistant Staphylococcus aureus

BACKGROUND: The recent proliferation of community-acquired (CA) methicillin-resistant Staphylococcus aureus (MRSA) has led to a marked increase in the need for outpatient treatment of MRSA infections. Many oral agents active against MRSA have been available for years, and a paucity of literature compares them, leaving physicians with little guidance for choosing among them. The purpose of the present study was to compare the bactericidal effects of orally available antibiotics against MRSA and to determine whether there were differences in antimicrobial killing activity against CA-MRSA and hospital-acquired (HA) MRSA isolates. METHODS: A total of 12 unique patient MRSA isolates were studied. Six strains were CA, carrying the staphylococcal chromosomal cassette (SCCmec) type IVa, while six were HA and carried SCCmec type II. Time-kill methods were used to study the bactericidal activity of the orally available antimicrobials linezolid, rifampicin, trimethoprim/sulfamethoxazole, clindamycin, minocycline, and moxifloxacin alone and in combination in vitro. RESULTS: Trimethoprim/sulfamethoxazole was rapidly bactericidal resulting in >2 log(10) cfu/mL decrease at 8 h and >3 log(10) cfu/mL decrease at 24 h in vitro. No antibiotic combination exhibited better killing than trimethoprim/sulfamethoxazole alone. Adding rifampicin to trimethoprim/sulfamethoxazole showed a trend towards antagonism in vitro. There were no differences in the bactericidal activity of any antimicrobial or antimicrobial combination against MRSA isolates carrying SCCmec type IVa versus those carrying SCCmec type II. CONCLUSION: Trimethoprim/sulfamethoxazole is rapidly bactericidal against MRSA in vitro when compared with most other orally available antimicrobials. No differences in bactericidal activity were detected when activities against CA-MRSA and HA-MRSA were compared.     

Identification of antimicrobial resistance and class 1 integrons in Shiga toxin-producing Escherichia coli recovered from humans and food animals

OBJECTIVES: The objective of this study was to identify antimicrobial resistance and class 1 integrons among Shiga toxin-producing Escherichia coli (STEC). METHODS: Two-hundred and seventy-four STEC recovered from poultry, cattle, swine and humans were characterized by antimicrobial susceptibility testing, screened for the presence of class 1 integrons by PCR, and assayed for integron transfer by conjugation. RESULTS: Ninety-three (34%) of the isolates were resistant to streptomycin, followed by 89 (32%) to sulfamethoxazole, 83 (30%) to tetracycline, 48 (18%) to ampicillin, 29 (11%) to cefalothin, 22 (8%) to trimethoprim/sulfamethoxazole, 18 (7%) to gentamicin, 13 (5%) to chloramphenicol and 10 (4%) to cefoxitin. Class 1 integrons were detected in 43 (16%) of the 274 isolates. The adenyl acetyltransferase gene, aadA, which confers resistance to streptomycin, was identified in integrons from 41 (95%) of these 43 isolates, and the dfrA12 gene, which confers resistance to trimethoprim, was identified in integrons from eight (19%) of the isolates. The sat1 gene, which confers resistance to streptothricin, an antimicrobial that has never been approved for use in the United States, was identified in integrons from three (7%) of the isolates. Transfer of integrons by conjugation between strains of E. coli resulted in transfer of antimicrobial-resistant phenotypes for ampicillin, chloramphenicol, cefalothin, gentamicin, tetracycline, trimethoprim, sulfamethoxazole and streptomycin. CONCLUSIONS: Antimicrobial resistance is common in STEC. Class 1 integrons located on mobile plasmids have facilitated the emergence and dissemination of antimicrobial resistance among STEC in humans and food animals.     

R-mediated beta-lactamases and episomal resistance to the beta-lactam drugs in different bacterial hosts

Ten penicillinase plasmids of varying taxonomic origin were studied after transfer to a variety of bacterial hosts. Nine of the ten plasmids specified enzymes with the following identical, or very similar, properties: substrate profile, molecular weight, susceptibility to heat and inhibitors, and electrophoretic mobility, i.e., TEM-like enzymes. The tenth R-mediated beta-lactamase was a cephalosporinase. Plasmids with TEM-like enzymes mediated resistance patterns identical towards the beta-lactam drugs, whereas the resistance pattern of the cephalosporinase plasmid was distinctly different. Expression of enzyme and resistance had a dual R-factor and host specificity. Escherichia coli K-12 and Salmonella typhi constituted one group of the same R-factor phenotype expressions. Most, but not all, penicillinase plasmids exhibited in Proteus PM1 a considerably lower order of beta-lactamase activity and an even lower order of resistance to the beta-lactam drugs than the previous two hosts. This difference was most pronounced for the resistance to carbenicillin, which was mediated by the plasmids specifying the synthesis of TEM-like enzymes. Release by osmotic shock was complete in the host E. coli K-12 for the TEM-like enzymes, but was lower for the cephalosporinase and minimal or negative in the PM1 host. Crypticity factor for benzylpenicillin, ampicillin, and carbenicillin was not related to the increase in resistance mediated by the penicillinase plasmids in both K-12 and PM1 hosts. Inoculum size effects for the penicillins and 6-aminopenicillanic acid were higher in PM1 than in K-12 R(+) cultures. The expression of penicillinase plasmids in wild-type bacteria was strain specific and not species specific. For two plasmids of different phenotypes for beta-lactamase activity (and resistance) in K-12 and PM1 hosts, a positive correlation was found between their phenotype and the relative amount of episomal deoxyribonucleic acid, as detected by ethidium bromide density gradient centrifugation. This is interpreted as indicating differences in the mode of replication of the plasmids in the two hosts.     

Persistence of an antibiotic resistance plasmid in intestinal Escherichia coli of chickens in the absence of selective pressure.

We studied eight strains of Escherichia coli resistant to high levels of trimethoprim that were isolated over a 6-week period in a commercial breeding flock of broilers. The strains originated from fecal samples and from a carcass immediately after slaughter. Seven of eight strains belonged to the same infrequent biotype. They were also resistant to ampicillin and streptomycin, and some were resistant to tetracycline and potassium tellurite. All the strains transferred trimethoprim and ampicillin resistance to E. coli. Analysis of the donors and of the transconjugants by agarose gel electrophoresis after digestion by restriction endonucleases and by nucleic acid hybridization indicated that resistance to trimethoprim (dfrI) and to ampicillin (bla TEM-1) was mediated by a 65-kilobase plasmid, pIP1531. Persistence of resistance to trimethoprim and ampicillin in this flock was therefore due to two cumulative factors, both occurring in the absence of selective pressure, namely the dissemination of a particular plasmid between strains and the ability of an atypical E. coli strain to stably colonize many animals.     

Emergence of aminoglycoside 3-N-acetyltransferase IV in Escherichia coli and Salmonella typhimurium isolated from animals in France.

We studied two outbreaks of calf salmonellosis caused by apramycin and gentamicin-resistant Salmonella typhimurium strains. In both cases, the responsible strains were resistant to ampicillin, chloramphenicol, kanamycin, streptomycin, tetracycline, and trimethoprim; one strain was also resistant to nalidixic acid in one outbreak. A systematic survey of the intestinal Escherichia coli strains of calves from the two affected flocks showed that 11 of 24 animals sampled were also colonized by apramycin- and gentamicin-resistant E. coli strains. These isolates belonged to four biotypes and were resistant to ampicillin, chloramphenicol, kanamycin, streptomycin, tetracycline, trimethoprim, and nalidixic acid. All of the strains were resistant to high levels of apramycin (MICs, 512 to 1,024 micrograms/ml) and to gentamicin (MICs, 8 to 32 micrograms/ml), and these resistances were always transferred en bloc. In S. typhimurium, this coresistance was borne by plasmids that were approximately 39 kilobases long (outbreak 1) or 90 kilobases long (outbreak 2), whereas in E. coli, the coresistance was due to plasmids that were approximately 110 kilobases long in both outbreaks. The two plasmids of Salmonella and four plasmids of E. coli encoded type IV aminoglycoside 3-N-acetyltransferases. The intensive use of curative and preventive treatments in calf production could be responsible for the emergence of enzymic resistance to apramycin and gentamicin.     

Antibiotic resistance and small R plasmids among Escherichia coli isolates from outpatient urinary tract infections in northern Norway.

Escherichia coli strains from outpatient urinary tract infections in northern Norway over a period of 1 year were examined for resistance to nine commonly used antibiotics. Strains collected during 4.5 months were examined for R plasmid content by using conjugation and in vitro transformation. Of the E. coli strains, 42% were resistant to one or more antibiotics. Resistance was highest to sulfonamide (20.8% of all strains), nitrofurantoin (14.5%), and tetracycline (10.1%), whereas less than 6% of the strains were resistant to ampicillin, carbenicillin, cephalothin, nalidixic acid, or trimethoprim-sulfamethoxazole. No strain was resistant to gentamicin. Tetracycline resistance was more common in men than in women. Resistance to cephalothin, nalidixic acid, and sulfonamide was higher in strains from older people. Resistance to sulfonamide was more frequent in the urban community. These was no seasonal variation in antibiotic resistance, although the incidence of urinary tract infection varied with seasons. Plasmid-determined resistance to ampicillin, streptomycin, sulfonamide, and tetracycline was found. About 18% of the resistant strains from the urban municipality carried R plasmids, most of which were small plasmids mediating resistance to sulfonamide and streptomycin. The overall frequency of resistance in strains collected from rural areas was similar to the urban frequency, but in the rural strains, R plasmids were found in only 5% of the resistant strains.     

Incidence and mechanisms of resistance to trimethoprim in clinically isolated gram-negative bacteria.

Over an 18-month period (October 1973 to April 1975), 133 strains of gram-negative bacteria with acquired resistance to trimethoprim (TM) were isolated from infected urines cultured at the Royal Free Hospital. The overall frequency of resistance was 3.2%. A disproportionately high number of resistant strains (63.1%) were Kebsiella aerogenes. Resistance to TM mediated by R plasmids occurs infrequently (9% of all resistant strains); the majority of TMR plasmids isolated belonged to one incompatability group (W). Chromosomally mediated resistance to TM in most Escherichia coli and K. aerogenes strains appears to be due mainly to production of a dihydrofolate reductase with a reduced susceptibility to TM. In some strains, increased activity of the DHFR was also a contributing factor. Increase in enzyme level alone was only great enough to account for the level of resistance to TM in a small number of cases.