Transferable Antibiotic Resistance in

Transcription

1 ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, Oct. 1973, p American Society for Microbiology Vol. 4, No. 4 Printed in U.S.A. Transferable Antibiotic Resistance in Enteropathogenic Escherichia coli Between 1948 and 1968 B. SLOCOMBE AND R. SUTHERLAND Beecham Research Laboratories, Chemotherapeutic Research Centre, Betchworth, Surrey, England Received for publication 21 May 1973 Enteropathogenic strains of Escherichia coli which had been isolated in the United Kingdom during three periods between 1948 and 1968, namely 1948 to 1951, 1957 to 1960, and 1967 to 1968, were tested for susceptibility to ampicillin, streptomycin, tetracycline, chloramphenicol, and sulphonamides. Antibioticresistant strains were tested for their ability to transfer antibiotic resistance to an antibiotic-susceptible strain of E. coli K-12. A relatively high proportion of strains isolated between 1948 and 1951 was resistant to ampicillin, streptomycin, or sulphonamides. None of these strains transferred ampicillin or streptomycin resistance, but sulphonamide resistance was R-factor-mediated in three out of 14 sulphonamide-resistant strains. Resistance to tetracycline and chloramphenicol was rare before 1951 but had become common among enteropathogenic E. coli by Much of the antibiotic resistance of bacteria isolated between 1957 and 1960 was R-factor-mediated, and transferable resistance was about as prevalent among E. coli isolated between 1957 and 1960 as among strains isolated in 1967 and Nevertheless, there was no appreciable increase in the overall incidence of antibiotic resistance among these enteropathogenic strains of E. coli between 1957 and 1968, although transferable antibiotic resistance was common during this period. These results do not suggest that the emergence of transferable antibiotic resistance will inevitably lead to the rapid development of antibiotic resistance among this group of b acteria. The discovery in 1959 of transferable antibiotic resistance among enteric bacteria (1, 17) obtaining gram-negative bacilli isolated before sources in the United Kingdom with a view to resulted in an increased interest in the incidence of antibiotic resistance among these orga- antibiotic resistance to other bacteria. We were 1960 which could be tested for ability to transfer nisms and in speculation regarding the possibility of the rapid development of widespread Director of the Salmonella Reference Labora- fortunate in learning that Joan Taylor (former antibiotic resistance among gram-negative bacilli. It is evident that there has been an dale, London) had established a collection of tory, Central Public Health Laboratory, Colin- increase in the resistance to antibiotics among cultures of enteropathogenic strains of Escherichia coli isolated from 1948 onwards. Dr. certain groups of infective enteric bacteria in the past years, e.g., shigellae (6, 12, 17) and Taylor very kindly made cultures available to Salmonella typhimurium (2, 8), and that many us, and the results reported here describe the of these bacteria possess transferable resistance antibiotic susceptibility patterns of enteropathogenic E. coli isolated at three different periods factors. On the other hand, the evidence regarding other Enterobacteriaceae is much less clear, between 1948 and 1968, together with the incidence of transferable antibiotic resistance and although it has been established that bacteria capable of transferring antibiotic resistance among these organisms. are at present common among the intestinal flora of healthy subjects (4, 5, 11) as well as of MATERIALS AND METHODS hospital patients (4, 9, 10, 15), little is known Cultures. A total of 512 strains of E. coli was about the emergence of bacteria possessing obtained from the Salmonella Reference Laboratory, transferable R-factors and the development of Colindale, London, and these bacteria were tested for resistance. Accordingly, we approached various susceptibility to ampicillin, tetracycline, strep- 459

2 460 SLOCOMBE AND SUTHERLAND tomycin, chloramphenicol, and sulphamethoxazole. The cultures had been isolated from clinical sources throughout the United Kingdom at three different periods of time between 1948 and 1968: (i) , 112 strains; (ii) , 200 strains; and (iii) , 200 strains. These strains had been stored on Dorset egg agar slopes in the dark at room temperature. Between 1948 and 1951, benzylpenicillin, streptomycin, and sulphonamides were the most widely used antibiotics; the broad-spectrum antibiotics, chloramphenicol and the tetracyclines, became available shortly afterwards, and ampicillin was introduced in Most of the strains (90%) from the two latter periods had been isolated from babies with diarrhea or gastroenteritis and comprised a wide range of serotypes. The remainder (10%) were from other sources (urine, blood, cerebrospinal fluid). Records were not available for the strains isolated between 1948 and 1951, but it is known that they were of human origin and it is believed that they were mainly from infantile gastroenteritis, as was the case for most of the other strains. Minimal inhibitory concentrations. Twofold serial dilutions of the test antibiotics were prepared in 18-ml volumes of molten nutrient agar (Blood Agar Base, Oxoid) and poured in petri dishes. When set, the plates were dried at 37 C and inoculated with a drop (0.003 ml) of an undiluted overnight broth culture of the test organism by means of a replicating device delivering 20 cultures to each plate. Tests in liquid medium were done in 5-ml volumes of nutrient broth (Nutrient Broth no. 2, Oxoid) which were inoculated with a drop (0.03 ml) of an overnight broth culture of the test organism. The susceptibility of E. coli to sulphonamides was measured by serial dilution of sulphamethoxazole in antagonist-free agar (D.S.T. agar, Oxoid) containing 10% lysed horse blood, using as inoculum a drop of a 1: 1,000 dilution of an overnight broth culture. Minimal inhibitory concentrations (MICs) were determined after overnight incubation at 37 C. The f-lactamase activity of ampicillin-resistant strains of E. coli was determined by measurement of the destruction of ampicillin during the course of serial dilution tests by using Bacillus subtilis ATCC 6633 to assay residual antibiotic activity Ṫransfer of resistance. Antibiotic-resistant isolates of E. coli were grown overnight in Penassay ANTIMICROB. AG. CHEMOTHER. broth (Antibiotic Medium no. 3, Oxoid), and 0.2-ml volumes of these donor strains were added to 0.4-ml volumes of an ovemight culture of the recipient strain, E. coli K-12/711Nx, a non-lactose-fermenting, nalidixic acid-resistant mutant. The mixed cultures of donor and recipient bacteria were cultured in 9 ml of fresh, prewarmed Penassay broth, and after overnight incubation at 37 C they were plated on Mac- Conkey agar, incorporating 20,g of the appropriate test antibiotic per ml and 20 ug of nalidixic acid per ml; control plates containing either 20 Mg of antibiotic or 20 Ag of nalidixic acid per ml were also inoculated with the mating mixtures. A number of donor strains were resistant to nalidixic acid, and in these cases selection was made on agar containing 1,000 Mg of streptomycin per ml by using a streptomycin-resistant strain of E. coli K-12/711 as recipient. Inoculation of the plates was carried out with a multipoint inoculating device delivering approximately ml of culture. After overnight incubation at 37 C, the plates were examined for non-lactose-fermenting colonies growing in the presence of 20 Mg of the antibiotic and 20 Mg of nalidixic acid per ml. Lactose-negative colonies were subcultured into Penassay broth and retested for antibiotic resistance. The transfer of sulphonamide resistance was determined by using D.S.T. agar (Oxoid) containing 10% lysed horse blood, incorporating sulphamethoxazole (200 Mg/ml) and nalidixic acid (20 Mg/ml). Neither the donor nor recipient bacteria grew on the selective medium. Colonies which developed from the mixed cultures were purified and identified as E. coli K-12/711Nx and retested for sulphonamide and antibiotic resistance. In cases where the donor strain inhibited the growth of E. coli K-12/711Nx as a result of bacteriocin activity, appropriate colicin-tolerant mutants were selected and used as recipient strains. A number of antibiotic-resistant cultures were filtered through membrane filters (0.22 ym pore size; Millipore Corp.), and the sterile filtrates were incubated with E. coli K-12/711Nx as described above. None of the cell-free filtrates conferred resistance on the recipient strain. Elimination of extrachromosomal antibiotic resistance. Antibiotic-resistant strains of E. coli were cultured on nutrient agar containing ethidium bromide (50 to 500 Mg/ml), and colonies growing on the plates containing the highest concentrations of ethidium bromide which supported normal growth were subcultured in drug-free medium and tested for antibiotic susceptibility (3). RESULTS Results in Table 1 show the distribution of the MICs of ampicillin, tetracycline, streptomycin, and chloramphenicol against E. coli isolated between 1948 and 1951, 1957 and 1960, and 1967 and Data expressed in Fig. 1 show the percentage inhibition of strains inhibited by serial concentrations of the antibiotics. Results in Table 2 show the incidence of transferable resistance among the antibiotic-resistant strains of E. coli, and the antibiotic resistance patterns transferred are described in Table 3. Resistance to ampicillin. Ampicillin inhibited 66% of the 112 strains of E. coli isolated between 1948 and 1951 at a concentration of 12.5 ug/ml or less; the remainder (34%) required inhibitory concentrations of from 25 to 500 gg of ampicillin per ml. Of the 200 strains isolated in the period 1957 to 1960, 95% were susceptible to 12.5 Ag of ampicillin per ml, and the remaining 5% of the strains were susceptible to concentrations of 25 to 250 Mg of ampicillin per ml. Most (85%) of the 200 strains isolated in 1967 and 1968 were susceptible to 12.5 Mg of ampicillin

3 VOL. 4, 1973 R-FACTOR TRANSFER IN E. COLI 461 TABLE 1. Antibiotic susceptibility of 512 strains of Escherichia coli isolated between 1948 and 1968 No. of Period 1 Antibiatic strains > No. of strains with MICa (,Ag/ml) of Ampicillin Chloramphenicol Streptomycin Tetracycline Ampicillin Chloramphenicol Streptomycin Tetracycline Ampicillin Chloramphenicol Streptomycin Tetracycline a I z CDz a Serial dilution in agar. AMPICIUIN TETRACYCLINE MINIMUM INHIBITORY CONCENTRATION ug/ml. STREPTOMYCIN CHLORAMPHENICOL , 112 Strains EZ , 200 Strains , 200 Strains FIG. 1. Susceptibility of 512 enteropathogenic strains of Escherichia coli isolated between 1948 and 1968 to four antibiotics. per ml, but two-thirds of the resistant strains insusceptible to ampicillin. Thus, the ampicilwere insusceptible to 500,ug of ampicillin per lin-resistant strains isolated between 1948 and ml. The incidence of ampicillin resistance was 1951 and between 1957 and 1960 were inhibited highest in the group of bacteria isolated be- by concentrations of 500 jig or less of ampicillin tween 1948 and 1951, but the resistant strains per ml, whereas most of the resistant strains isolated in 1967 and 1968 were generally more isolated in 1967 and 1968 were not susceptible to

4 462 SLOCOMBE AND SUTHERLAND TABLE 2. Transferable antibiotic resistance among Escherichia coli isolated between Resistant strainsa ANTIMICROB. AG. CHEMOTHER. Transferable resistance Antibiotic Time of No. of % of % of isolation strains No % of No. of resistant total total strains strains strains Ampicillin Chloramphenicol Streptomycin Tetracycline Sulphamethoxazole " b b a Not inhibited by 12.5 jag of test antibiotic per ml. Resistant to 200,g of sulphamethoxazole per ml. 500 gg of ampicillin per ml. Examination of a number of the ampicillin-resistant strains from the three periods showed that these strains were able to destroy ampicillin during the course of the antibacterial tests, indicating that the resistance to ampicillin was associated with,b-lactamase activity. None of the ampicillin-resistant strains isolated in the periods 1948 to 1951 or 1957 to 1960 transferred ampicillin resistance to the selected recipient strain, E. coli K-12/711Nx (Table 2). Similarly, those ampicillin-resistant strains isolated in 1967 and 1968, which like earlier strains had ampicillin MIC values in the range 25 to 250,ug/ml, also failed to transfer ampicillin resistance. On the other hand, all strains isolated between 1967 and 1968 which were resistant to 500 gg of ampicillin per ml transferred ampicillin resistance associated with 6-lactamase activity. Treatment with ethidium bromide had no effect on the susceptibility to ampicillin of ampicillin-resistant strains of E. coli isolated before 1960, but resulted in the elimination or reduction of the resistance of strains with transferable ampicillin resistance isolated in 1967 and Resistance to streptomycin. In general, there was little difference in the streptomycin susceptibility patterns of the three groups of E. coli (75% susceptible to 12.5 g/iml in , 86% susceptible in , and 78% susceptible in ). However, all resistant organisms isolated between 1948 and 1951 were insusceptible to 500 Mug of streptomycin per ml, whereas most of the resistant strains isolated in the two more recent periods were less resistant and had MIC values in the range of 25 to 500 Mg of streptomycin per ml. None of the strains isolated between 1948 and 1951 transferred streptomycin resistance, but 68% of the streptomycin-resistant strains isolated between 1957 and 1960 and 66% of the most recently isolated strains were able to transfer streptomycin resistance to E. coli K-12/711Nx. R-factormediated streptomycin resistance was eliminated by treatment with ethidium bromide, but the streptomycin resistance of strains isolated before 1951 was unaffected by this treatment. Resistance to tetracycline. All strains of E. coli isolated between 1948 and 1951 were susceptible to tetracycline (MIC = 5,ug/ml or less), but resistance to tetracycline was common among bacteria isolated from 1957 onwards. Thus, 17.5% of the strains isolated between 1957 and 1960, and 22% of those isolated between 1967 and 1968, were not inhibited by 12.5,g of tetracycline per ml. The distribution of susceptibility to tetracycline of these two groups of E. coli was very similar (Fig. 1). Transfer of tetracycline resistance was readily demon-

5 VOL. 4, 1973 R-FACTOR TRANSFER IN E. COLI 463 TABLE 3. Antibiotic resistance patterns in Escherichia coli between 1948 and 1968 No. of strains Resistance patterna (112) (200) (200) Resistant Transferring Resistant Transferring Resistant Transferring resistance resistance resistance Su S SSu (S-1) T TSu 6 4(T-2) 1 0 TS (S-5) TSSu 4 3(TSu-1) 7 6(TS-1) TCS 1 0 CSSu 3 3(CS-1) 1 1 TSKSu 1 l(su-1) (TC-1) TCSSu (TCS-2) t(cssu-4) A AC AK 1 1 AS l(s-1) AT (A-1) ASu 6 3(Su-3) ACSu 1 0 ASK 1 0 ATS 2 1(T-1) 1 1(A-1) ASSu l(as-1) ATCSu 1 l(atc-1) ATSSu 2 2 (TT-1) ATCSSu 5 4(ACTS-3) ATSKSu 3 3(ATKSu-1) ATC KSSu 1 1 ATCKSuNx 1 1(ATCKSu-1) a Abbreviations: A, ampicillin; C, chloramphenicol; K, kanamycin; Nx, nalidixic acid; S, streptomycin; Su, sulphamethoxazole; T, tetracycline. strated with organisms isolated between 1957 and 1968, and 60% of the tetracycline-resistant strains isolated between 1957 and 1960 transferred tetracycline resistance, compared with 68% of the tetracycline-resistant strains isolated between 1967 and 1968 (Table 2). Resistance to chloramphenicol. Only one of the strains of E. coli isolated between 1948 and 1951 was resistant to chloramphenicol, and this strain failed to transfer its chloramphenicol resistance. A small proportion of strains isolated in the two later periods were resistant to chloramphenicol (7.5% in and 5% in ), and virtually all of these strains transferred their chloramphenicol resistance to E. coli K-12/711Nx (Table 2). Resistance to sulphonamides. Most (98 of 112) of the strains of E. coli isolated between 1948 and 1951 were susceptible to 20,ug of sulphamethoxazole per ml, but 14 strains were not inhibited by 200 Ag of sulphamethoxazole per ml. Three of these 14 resistant strains transferred sulphonamide resistance to E. coli K-12/71lNx. These three R-factor strains, which had been isolated in 1951, were resistant also to ampicillin, but ampicillin resistance was not transmissible. The incidence of sulphonamide-resistant strains was higher in the two later periods, and 40% of the sulphonamideresistant strains transferred sulphonamide resistance, usually in association with streptomycin or other resistance (Table 2). DISCUSSION An inherent weakness of a retrospective study of this kind is the possibility that the characteristics of the test organisms may have altered during the course of storage in the laboratory. It is possible that some of the strains examined

6 464 SLOCOMBE AND SUTHERLAND which now appear to be susceptible to antibiotics may have possessed R-determinants at the time of isolation, or that antibiotic-resistant strains which now fail to transfer resistance may once have possessed transfer factors. In practice, the data obtained suggest that the enteropathogenic strains of E. coli described here may not have altered greatly on storage, and it seems possible that the antibiotic resistance patterns reported may well be a true reflection of those prevailing at the time of isolation of the cultures. This supposition is based upon the following facts. (i) The strains of E. coli isolated in 1967 and 1968 were examined relatively soon after isolation when it was unlikely that there would have been any significant change in the antibiotic resistance patterns of this group. (ii) The overall incidence of resistance and the frequency of bacteria possessing transmissible R- factors were similar for strains isolated between 1957 and 1960 and for strains isolated in 1967 and 1968, with the exception of resistance to ampicillin. It seems unlikely, therefore, that there was any great loss of resistance from the former group of bacteria as a result of storage, with the possible exception of ampicillin. (iii) There was a relatively high incidence of resistance to ampicillin, streptomycin, and sulphonamides among E. coli isolated between 1948 and 1951, and so it is less likely that there was any significant loss of resistance on storage. These bacteria were isolated before chloramphenicol and the tetracyclines became freely available in the United Kingdom, and so the lack of resistant organisms is not unexpected. (iv) The data suggest that the antibiotic resistance of E. coli isolated between 1948 and 1951 is not extrachromosomally mediated, and therefore it seems reasonable to assume that failure to demonstrate transferable antibiotic resistance among bacteria from this period is due to the fact that these bacteria never possessed transfer factors, rather than to loss during storage. Another potential failing of this type of study is that the strains examined may not represent a true sample of the bacterial populations predominant at the time of isolation. This could apply particularly to the enteropathogenic bacteria isolated before 1951, for which precise records are unavailable, but less so to the strains isolated between 1957 and 196.8, since they were isolated from various areas of the United Kingdom and were of numerous different serotypes. Nevertheless, despite these reservations, the results presented'here appear to be meaningful in tracing the development of trans- ANTIMICROB. AG. CHEMOTHER. ferable antibiotic resistance among enteropathogenic E. coli isolated in the United Kingdom between 1948 and The relatively high incidence of ampicillinresistant strains of E. coli isolated before 1951 may be associated with the general use of benzylpenicillin during this period, which could have resulted in the selection of ampicillinresistant mutants. The resistance of ampicillinresistant strains isolated between 1948 and 1951 was associated with,b-lactamase activity, but the f-lactamase produced by these strains differed from that produced by the majority of strains isolated 20 years later, in 1967 and The former is probably chromosomally mediated, since it is not possible to demonstrate transfer of the,8-lactamase, it is not eliminated by treatment with ethidium bromide, and the substrate and inhibition profiles of the cell-free enzyme are different from those of the extrachromosomally mediated,8-lactamase possessed by the majority of the ampicillin-resistant strains isolated in 1967 and 1968 (S. Elson, personal communication). These facts suggest that R-factor-mediated ampicillin resistance was not common among enteropathog'enic E. coli between 1948 and However, it is possible that some of the strains which are now susceptible to ampicillin- may have possessed ampicillin R-factors which were subsequently lost during storage. For example, R-factor-mediated ampicillin resistance has been described among E. coli isolated in Mexico in 1956 (7) and among E. coli isolated in London in 1961 (13). In the latter study, three out of seven ampicillinresistant strains isolated in 1961 had lost their resistance to ampicillin when retested in 1968; the remaining four strains possessed transferable ampicillin resistance. Moreover, since transferable streptomycin, and sulphonamide resistance became relatively common among enteropathogenic E. coli between 1951 and 1960, it would not be unexpected to find transferable penicillin resistance appearing during this period. One third of the ampicillin-resistant strains of E. coli isolated in 1967 and 1968 had a nontransferable,b-lactamase, determinant similar to that produced by strains isolated in the earlier periods, whereas the remainder of the ampicillin-resistant strains possessed a transferable extrachromosomally mediated fb-lactamase. The pattern of streptomycin resistance of the E. coli examined is in some ways analogous to that of the ampicillin resistance of these strains. Thus, there was a relatively high incidence of streptomycin-resistant strains of E. coli isolated

7 VOL. 4, 1973 before 1951, which is not entirely unexpected since streptomycin was freely available at this time. It seems probable too that the resistance of these particular strains is chromosomally mediated since, apart from the fact it is nontransferable and is not "cured" by ethidium bromide, the level of resistance of strains isolated before 1951 is very high, a characteristic of streptomycin-resistant mutants, in contrast to the lower levels of resistance found in the R-factor-mediated streptomycin-resistant strains isolated from 1957 onwards. In addition, most of the streptomycin-resistant strains isolated before 1951 were susceptible to sulphonamides, whereas most of the streptomycin-resistant strains isolated from 1957 onwards which transferred streptomycin resistance were also resistant to sulphonamides. It is unlikely, therefore, that the streptomycin-resistant strains isolated before 1951 are R-factor bacteria lacking transfer factors. These facts suggest that transferable streptomycin resistance was not a feature of the streptomycin resistance of E. coli isolated before 1951 but became common relatively soon afterwards, so that by 1957 the proportion of strains carrying streptomycin R- factors was as high as that observed in strains isolated in It is perhaps not unexpected that three of' the 14 sulphonamide-resistant strains isolated between 1948 and 1951 were able to transfer sulphonamide resistance, since this. class of compounds had been widely used for more than 10 years before this, although, in fact, the incidence of sulphonamide resistance of the strains isolated before 1951-was relatively low. It is probable that the three strains isolated in 1967 capable of vtransferring sulphonamide resistance came from the same source, and it is of interest that the sulphonamide resistance of these strains was not linked with transferable streptomycin resistance as was generally the case for strains isolated in later periods. By 1951, sulphonamide resistance had become more common and was often R-factor mediated, associated with streptomycin or multiple resistance. It seems likely, therefore, that there was a marked increase in transferable sulphonamide resistance between 1951 and The data relating to' chloramphenicol and tetracycline resistance show that resistance to these two antibiotics, largely R-factor mediated, appeared sometime after 1951, coincident with the increasing usage of these antibiotics. By 1957 most of the chloramphenicol- and tetracycline-resistant strains were capable of transferring antibiotic resistance, and as with R-FACTOR TRANSFER IN E. COLI 465 streptomycin, the incidence of transferable chloramphenicol and tetracycline resistance apparently remained unchanged between 1957 and Although it is known that R-factors were present in bacteria prior to 1959 (7, 14, 16), when the phenomenon of transferable antibiotic resistance was first reported, data on the frequency of such strains are lacking. Our results with a relatively large number of homogenous bacteria isolated between 1948 and 1968 show that R-factor-mediated sulphonamide resistance was present in E. coli as long ago as 1951, and that transferable antibiotic resistance became established among enteropathogenic strains of E. coli in the United Kingdom at some time between 1951 and It was unexpected, however, that the incidence of R-factormediated resistance was about as high among antibiotic-resistant strains isolated between 1957 and 1960 as in strains collected in 1967 and Moreover, although strains capable of transferring antibiotic resistance to susceptible bacteria were widespread between 1957 and 1968, there was apparently no marked increase in the overall incidence of resistance of these strains of E. coli to the four antibiotics tested, or to sulphonamides, during this period. These results show that the phenomenon of transferable antibiotic resistance has been common among E. coli for a longer period of time than has hitherto been supposed and do not support the view that the emergence of transferable antibiotic resistance among enteric bacteria will inevitably lead to a rapid or general increase in the antibiotic resistance of this group of bacteria. ACKNOWLEDGMENTS We are very grateful to Joan Taylor, former Director of the Salmonella Reference Laboratory, Colindale, London, for allowing us access to her collection of enteropathogenic E. coli and for supplying the cultures examined here. We thank also B. Rowe, Director of the Salmonella Reference Laboratory, and R. Gross for providing additional cultures. LITERATURE CITED 1. Akiba, T., K. Koyama, Y. Ishiki, S. Kimura, and T. Fukushima On the mechanism of the development of multiple drug-resistant clones of Shigella. Japan J. Microbiol. 4: *2. Anderson, E. S Drug resistance in Salmonella typhimurium and its implications. Brit. Med. J. 3: Bouarichaud, D. H., M. R. Scavizzi, and Y. A. Chabbert Elimination by ethidium bromide of antibiotic resistance in enterobacteria and staphylococci. J. Gen. Microbiol. 54: Datta, N Drug resistance and R-factors in the bowel bacteria of London patients before and after admission to hospital. Brit. Med. J. 2:

8 466 SLOCOMBE AND SUTHERLAND 5. Datta, N., M. C. Faiers, D. S. Reeves, W. Brumfitt, and I. Orskov R-factors in Escherichia coli in faeces after oral chemotherapy in general practice. Lancet 1: Davies, J. R., W. N. Farrant, and A. H. C. Uttley Antibiotic resistance of Shigella sonnei. Lancet 2: Evans, J., E. Galindo, J. Olarte, and S. Falkow fb-lactamase of R-factors. J. Bacteriol. 96: Guinee, P. A. M Bacterial drug resistance in animals. Ann. N.Y. Acad. Sci. 182: Kabins, S. A., and S. Cohen Resistance-transfer factor in Enterobacteriaceae. N. Engl. J. Med. 275: Lewis, M. J Transferable drug resistance and other transferable agents in strains of Escherichia coli from two human populations. Lancet 1: Moorhouse, E. C Transferable drug resistance in ANTIMICROB. AG. CHEMOTHER. enterobacteria isolated from urban infants. Brit. Med. J. 2: Ross, S., G. Controni, and W. Khan Resistance of shigellae to ampicillin and other antibiotics. J. Amer. Med. Ass. 221: Slocombe, B., and R. Sutherland Sensitivity of gram-negative bacilli to ampicillin after six years' clinical use. J. Clin. Pathol. 22: Smith, D. H R-factor infection of Escherichia coli lyophilized in J. Bacteriol. 94: Smith, D. H., and S. E. Armour, Transferable R-factors in enteric bacteria causing infection of the genitourinary tract. Lancet 2: Smith, H. W., and S. Halls Observations on infective drug resistance in Britain. Brit. Med. J. 1: Watanabe, T Infective heredity of multiple drug resistance in bacteria. Bacteriol. Rev. 27: Downloaded from on June 14, 2018 by guest