Frequency of tetracycline resistance and infectious resistance transfer in coliform bacteria.

TETRACYCLINE RESISTANCE AND INFECTIOUS RESISTANCE TRANSFER IN COLIFORMS 829 Frequency of Tetracycline Resistance and Infectious Resistance Transfer in Coliform Bacteria * by MESSELE GEDEBOU 1 & WARREN C. EVELAND 2 It is a well-known phenomenon that a micro- organism sensitive to a certain antibiotic or chemo- therapeutic agent may convert to a resistant form by spontaneous mutation or general transduction. A most remarkable cause of drug-resistance was not recognized until 1959 (cited by Watanabe, 1963): it is known as infectious transfer of drug-resistance, that is, transfer of resistance by conjugation among Enterobacteriaceae from a resistant donor to a sensitive recipient. The existence and spread of such resistance among enteric bacteria have been con- firmed wherever the search has been undertaken: Japan (Watanabe, 1963), England (Datta, 1962), the Federal Republic of Germany (Lebek, 1963), South Africa (Mare & Coetzee, 1965), the USA (Kabins & Cohen, 1966), Ethiopia (Mann & Gede- bou, 1966), Greece (Kontomichalou, 1967a), Czecho- slovakia (Macuch et al., 1967), and other parts of the world. This relatively new concept has recently been thoroughly reviewed by Anderson (1968), Datta (1965), and Mitsuhashi (1969). Several reports on infectious drug-resistance have been on outbreaks caused by drug-resistant enteric pathogenic bacteria and on their capacity to transfer their resistance in vitro to sensitive recipients. Considerable attention has also been given to the study of the mechanism of infectious transfer of drug-resistance and to the nature of the transfer factor involved (Anderson, 1968; Meynell et al., 1968; Mitsuhashi, 1969; Watanabe, 1967). Relative- ly few reports (Gunter & Feary, 1968; Mann & Gedebou, 1966; Mare, 1968; Mitsuhashi, 1969) have appeared on the frequency of the occurrence of transferable drug-resistance among intestinal bacteria which have not been involved in any out- break of disease. * The work was conducted during tenure of a WHO Fellowship by the first author. 1 Lecturer in Bacteriology, Faculty of Medicine, Haile Sellassie I University, Addis Ababa, Ethiopia. Present address: WHO Fellow, Department of Epidemiology, School of Public Health, University of Michigan, Ann Arbor, Mich., USA. ' Professor of Epidemiology, Department ofEpidemiology, School of Public Health, University of Michigan, Ann Arbor, Mich., USA. The present investigation was conducted to assay the frequency of the occurrence of tetracycline- resistant coliform bacteria isolated from in-patients and also the frequency of resistance transfer to a Salmonella recipient which is sensitive to tetra- cycline and to all the other drugs employed in this work. Surveys of this type are needed from different areas and hospitals in order to evaluate the public health threat posed by infectious drug-resistance. Materials and methods Stool specimens. Stool specimens were obtained from the Microbiology Laboratory of the University Hospital, Ann Arbor. They were primarily collected from in-patients for testing occult blood. A total of 300 specimens were randomly picked from the laboratory and were screened in this study. Tetra- cycline-resistant coliforms were isolated and tried as donors in the transfer experiments. One hundred strains of Escherichia coli collected from a tribe in Brazil, apparently unexposed to any therapeutic or prophylactic drug, were also included in the investi- gation. Recipient strain. Strain Salmonella typhimurium 74 from the National Collection of Type Cultures, Colindale, London, England, was the recipient bacterium throughout the investigation of infectious transfer of drug-resistance. It was sensitive to all the drugs employed in this study. Media. Difco MacConkey agar was employed for isolating coliforms from stools and brilliant- green agar was used for the drug-resistance transfer experiments. BBL Kligler's iron-sugar agar was employed to characterize the coliforms and also to identify the Salmonella derivatives after transfer experiments. Drugs: Drug-sensitivity tests were done on BBL trypticase-soy agar using BBL Sensidisks of tetra- cycline, chloramphenicol, streptomycin, kanamycin, neomycin and ampicillin, all in a concentration of 5 ,g/ml except for streptomycin; streptomycin was used in a concentration of 10 ,ug/ml because the recipient Salm. typhimurium was resistant 2523E to 2 ,ug/ml, the lower available concentration. Tetracyn from J. B. Roering & Company, New York, was used as a powdered tetracycline for incorporating into MacConkey and brilliant-green agar media. It consisted of tetracycline hydro- chloride buffered with ascorbic acid and was pre- pared primarily for intravenous therapeutic use. Each vial contained a mixture of 500 mg tetra- cycline hydrochloride and 1.5 g ascorbic acid. Isolation of tetracycline-resistant coliforms. The procedure in this investigation was similar to that of Mann & Gedebou (1966). A stool specimen was streaked on both MacConkey agar containing 25 jig/ml tetracycline hydrochloride and plain Mac- Conkey agar without the drug and incubated over- night at 37°C. The plain medium was used to check the presence of viable coliforms in the stool speci- mens. A lactose-fermenting colony appearing on tetracycline-containing MacConkey agar was picked and subcultured on a trypticase-soy agar slant for use as a possible donor in the resistance transfer experiment. The tetracycline-resistant coliform was then tested for its drug-resistance pattern on a trypticase-soy agar plate with Sensidisks of tetra- cycline, chloramphenicol, streptomycin, ampicillin, kanamycin and neomycin. The plate was read after incubation at 37°C for 24 hours. A zone of inhibition of 10 mm or greater was considered to indicate sensitivity. The drug-sensitive Salm. typhi- murium was simultaneously tested as a control. Drug-resistance transfer. A tetracycline-resistant coliform and the sensitive Salm. typhimurium were grown separately overnight at 37°C in nutrient broth. Then a tube of 5 ml of fresh nutrient broth was inoculated with 0.5 ml of each of the two cultures. After overnight incubation, a drop of the mixed culture was serially transferred with a Pasteur pipette in three 5-ml broth tubes. From each dilu- tion, a drop was spread with a glass spreader on brilliant-green agar containing 25 ,ug/ml tetra- cycline hydrochloride and also on plain brilliant- green agar and incubated overnight. The Salmonella and coliform colonies were distinguished by their characteristic colours on the lactose-containing medium. If a colony typical of Salmonella appeared on the medium containing tetracycline, it was picked and streaked on another tetracycline-con- taining agar plate and incubated for further con- firmation of the acquired resistance to tetracycline. It was further identified by its reactions on Kligler's iron-sugar agar and slide agglutination with serum prepared against the somatic antigen of Salm. typhi- murium. The derivative tetracycline-resistant Salmo- nella was then tested for its resistance pattern by the disc method using the aforementioned six Sensidisks. This was done to determine whether the Salmonella received resistance to tetracycline only or to other drugs to which the donor coliform was resistant. Results Tetracycline-resistant coliforms were isolated from 131 of the 300 stool specimens (43.7%Y.). Resistance to the drug was demonstrated by growth on Mac- Conkey agar containing 25 ,ug/ml tetracycline. Only 28 of the 131 strains (21.4%) were singly resistant to tetracycline alone (Table 1) as determined by the sensitivity testing with the Sensidisks; the remaining 78.6% were doubly or multiply resistant. Table 1 shows all the resistance patterns of the isolated coliforms. None of the 100 E. coli strains from Brazil grew on MacConkey agar containing tetra- cycline; they grew only on plain MacConkey agar without tetracycline. TABLE 1 DRUG-RESISTANCE PATTERNS OF ISOLATED COLIFORMS Coliforms transferring Resistant to: a i No. any resistanceisolated No. | % tet chl sm amp neo kan 1 0 0 tet sm amp neo kan 1 1 100 tet chi sm amp 11 2 18.1 tet sm neo kan 1 0 0 tet chi neo kan 1 0 0 tet chl sm neo 1 1 100 tet chl sm 2 1 50 tet chl amp 7 2 28.6 tet sm neo 3 1 33.3 tet sm amp 23 11 47.8 tet chl 7 3 42.8 tet amp 9 5 55.5 tet sm 36 12 33.3 tet 28 10 35.7 Total 131 49 37.4 a tet = tetracycline; chl = chloramphenicol; amp = ampi- cillin; neo = neomycin; sm = streptomycin; kan = kanamycln. 830 NOTES TETRACYCLINE RESISTANCE AND INFECTIOUS RESISTANCE TRANSFER IN COLIFORMS Among the 131 resistant coliforms, 49 (37.4%) were able to transfer resistance to tetracycline alone or to either one or more of the other drugs to Salm. typhimurium which was sensitive prior to mating with the resistant donors. Only 15 (11.4%) transferred, at the same time, resistance to tetracycline and to one or more of the other drugs (Table 2). Only 3 triply and 7 doubly resistant donors transferred wholly their resistance pattern to the recipient Salm. typhimurium (Table 2). The other resistant donors transferred only part of their resistance spectra. TABLE 2 DRUG-RESISTANCE PATTERNS OF DONOR COLIFORMS AND DERIVATIVE SALMONELLA TYPHIMURIUM Donor Derivative Resistance Resistance No. pattern aN. pattern a tet sm amp neo kan 1 tet 1 tet chl sm amp 2 tet amp 1 tet I tet chi sm neo 1 tet chl sm 1 tet sm amp 11 tet sm amp 2 tet sm 2 tet 7 tet sm neo I tet 1 tet chl sm 1 tet chl sm 1 tet chl amp 2 tet amp 1 tet _ tet sm 13 tet sm 7 tet 6 tet chl 3 tet 3 tet amp 5 tet 5 tet 9 tet 9 Total 49 49 a tet= tetracycline; chl = chloramphenicol; amp = ampi- cillin; neo - neomycin; sm = streptomycin; kan = kanamycin. Discussion The frequency of the occurrence of tetracycline- resistant coliforms reported in this work is signifi- cantly high (43.7 %). Another study elsewhere (Mann & Gedebou, 1966) gave a slightly lower frequency (36 %). In the latter, specimens were obtained from out-patients who were, very likely, less exposed to drugs than the in-patients in this work. These studies, however, excluded all tetra- cycline-sensitive strains, some of which could be resistant to any other drug. The actual frequency of the occurrence of coliforms resistant to any one or more drugs must therefore be considered to be greater than 43.7%, which is the frequency using only tetracycline as a selective marker in isolation. Gunter & Feary (1968) found 38.1 % of their clinically isolated E. coli strains resistant to one or more drugs. However, their isolation was not restricted to any one drug as ours was. Thus their value for the frequency of drug-resistance appears rather low. On the other hand, a frequency as high as 84.2% has been reported (Mitsuhashi et al., 1967). Although 19 of the coliform donors were resistant to 3 or more drugs, only 4 donor strains transferred triple resistance (Table 2). Simultaneous transfer of resistance to up to 8 drugs has been reported (Kontomichalou, 1967b). Transfer of multiple resistance depends on various factors (Anderson, 1968; D. H. Smith, 1966). The efficiency of transfer is influenced by the types of both the recipient and the donor strains (Ander- son, 1968). Certain species or even certain strains are better recipients than others (Watanabe, 1963; Watanabe & Fukasawa, 1961) and some donor species or strains are more efficient than others; a strain could be a good recipient for a particular donor and a poor one for another (Anderson, 1968). The value for the frequency of the occurrence of the resistance transfer factor among bacteria is thus dependent on these and other factors. This could account for the low frequency of resistance transfer factors in this study (37.4 %), when compared with reports of 57.6% (Gunter & Feary, 1968) and 84% (Mitsuhashi et al., 1967). The results in this and other reports (Gunter & Feary, 1968; Macuch et al., 1967; Mann & Gedebou, 1966; Mare, 1968; W. H. Smith, 1966) are in line with the consensus that if naturally occurring coliforms acquire infectious drug-resistance, they could act as the source of transferable resistance for pathogens that share the same intestinal habitat. 831 832 NOTES The spread of such resistant coliforms, at least among hospital patients, is indeed high. The potential danger of carriers of drug-resistant in- testinal flora was considered as early as the first comprehensive report (Watanabe, 1963) on this relatively newly discovered phenomenon. The possible sources of resistance factors for the intest- inal flora have been discussed by Anderson (1968). The finding in this and other studies (Gunter & Feary, 1968; Mare, 1968), that the frequency of the occurrence of coliforms with transferable drug- resistance is high among persons exposed to one or more drugs and nil or extremely low among persons apparently unexposed to any therapeutic or pro- phylactic agent, is direct and strong evidence to warrant strict control of the indiscriminate use of drugs. Such unnecessary use of drugs certainly provides the selection pressure in favour of the increase of drug-resistant strains. How a particular bacterial strain acquires the infectious resistance factor-that is, without transfer from another bac- terium-is still a matter for speculation. However, the evidence from this and other studies (Gunter & Feary, 1968; Harada et al., 1960; Mare, 1968), that such resistant strains are absent or occur with extremely low frequency among persons apparently healthy and not under treatment, strongly indicates a significant role of drug use in the emergence of infectious resistance factors. Minimal use of drugs in human as well as in veterinary medicine has been suggested as the only practicable measure, at present, to control the spread of resistance factors (Cohen, 1969). Among the immediate and essential steps toward the control of this threatening public health problem would be a thorough understanding of the exact molecular nature of the episome involved and the mechanism of the resistance-mediating process. ACKNOWLEDGEMENT We thank Dr P. G. Mann of the Public Health Labor- atory Service, Manor Hospital, Bath, England, who arranged for the National Collection of Type Cultures, Colindale, London, England, to send us Salmonella typhimurium strain 74. REFERENCES Anderson, E. S. (1968) Ann. Rev. Microbiol., 22, 131- 180 Cohen, S. (1969) J. infect. Dis., 119, 104-106 Datta, N. (1962) J. Hyg. (Lond.), 60, 301-310 Datta, N. (1965) Brit. med. Bull., 21, 254-259 Gunter, C. A. & Feary, T. W. (1968) J. Bact., 96, 1556- 1561 Harada, K., Suzuki, M., Kameda, M. & Mitsuhashi, S. (1960) Jap. J. exp. Med., 30, 289-299 Kabins, A. S. & Cohen, S. (1966) New Engl. J. Med., 275, 248-252 Kontomichalou, P. (1967a) Path. et Microbiol. (Basel), 30, 71-93 Kontamichalou, P. (1967b) Path. et Microbiol. (Basel), 30, 185-200 Lebek, G. (1963) Zbl. Bakt., L Abt. Orig., 189, 213- 223 Macuch, P., Seckarova, A., Parrakova,E., Krcmery, B. & Bymola, F. (1967) Z. allg. Mikrobiol., 1, 159-162 Mann, P. G. & Gedebou, M. (1966) Ethiop. med. J., 4, 181-188 Mare, I. J. (1968) Nature (Lond.), 220, 1046-1047 Mare, I. J. & Coetzee, J. N. (1965) S. Afr. med. J., 39, 864-865 Meynell, E., Meynell, G. G. & Datta, N. (1968) Bact. Rev., 32, 55-83 Mitsuhashi, S. (1969) J. infect. Dis., 119, 89-100 Mitsuhashi, S., Hashimoto, H., Egawa, R., Tanaka, T. & Nagai, Y. (1967) J. Bact., 93, 1242-1245 Smith, D. H. (1966) Drug resistance of enteric bacteria mediated by R factors. In: Hobby, G. L., ed., Anti- microbial agent and chemotherapy-1966, Baltimore, American Society for Microbiology. pp. 274-280 Smith, W. H. (1966) J. Hyg. (Lond.), 64, 465-474 Watanabe, T. (1963) Bact. Rev., 27, 87-115 Watanabe, T. (1967) Fed. Proc., 26, 23-28 Watanabe, R., & Fukasawa, T. (1961) J. Bact., 81, 669-678