APMIS 100: 629-634. 1992

Tetracycline resistance genes in Kenyan hospital isolates of Salmonella typhimrium

SAMUEL KARIUKI'*2, NAZIR BEGUM MIRZA', YNGVILD WASTESON', DANIEL SENERWA', JOSEPH M. GATHUMA', and 0RJAN OLSVIK'

'Department of Microbiology and Immunology, Norwegian College of Veterinary Medicine, Oslo, Norway, 'Department of Public Health, Pharmacology and Toxicology, Faculty of Veterinary Medicine, and 'Depart-

ment of Medical Microbiology, Faculty of Medicine, University of Nairobi, Nairobi, Kenya

Kariuki, S., Mirza, N. B., Wasteson, Y., Senerwa, D., Gathuma, J. M. & Olsvik, 0. Tetracycline resistance genes in Kenyan hospital isolates of Salmonella typhimurium. APMIS 100: 629-634, 1992.

All 97 strains of Salmonella typhimurium isolated from patients at a hospital in Nairobi, Kenya, during 1988-90 were resistant to tetracycline. The minimum inhibitory concentration (MIC) showed a large distribution range from 1 pg/ml to 128 pg/ml. The strains were heterogeneous with respect to plasmid content, but initially all strains possessed, in addition to other plasmids, a large 60-, 63- or 65-MDa plasmid. The tetracycline resistance genes were characterized using oligonucleotide probes, and 20% of the resistant strains possessed tetracycline type A (tet' A), 6% tet' B, and 4% tet'C genes. Three strains possessed both type A and B tetracycline resistance determinants, which were shown to be located on the large 65-MDa plasmid. There was no correlation between strains isolated from stools, blood, cerebrospinal or epidural fluids, pus, or urine, with respect to the tetracycline genotypes, MIC values or plasmid content.

Key words: Salmonella typhimurium; tetracycline resistance genes; probes; plasmids; Kenya.

0. Olsvik, Enteric Diseases Branch, Division of Bacterial and Mycotic Diseases, National Center for Infectious Diseases, Centers for Disease Control, MS CO 3, Atlanta, GA. 30333, USA.

Salmonellosis is a public health problem in Kenya as in most other countries, and during the past 10 years there has been a marked increase in infection due to Salmonella typhimurium (1 4,29). Wumolu & Mirzu (29) furthermore noted the more alarming fact that strains of S. typhimurium isolated in Kenya were very resistant to the cur- rently available antibiotics, for instance com- pared to S. typhi which was still relatively sensi- tive.

Although tetracyclines are still the second most commonly used antibiotics worldwide, their efficacy has been curtailed by the high fre- quency of microbial resistance (17). The source

Received October 7, 1991. Accepted January 27, 1992.

of resistant bacteria is controversial. Administra- tion of antimicrobials to animal feed selects for antimicrobial-resistant bacteria, and has been suggested to increase the prevalence of such strains in the food chain (14). This might be an important cause of drug-resistant strains in humans. In addition, it is likely that multiple drug-resistant S. typhimurium strains result from therapeutic use, both in man and animals, with resistant strains found in hospitals possibly being of particular importance (1,2).

The aim of the present study was to character- ize the S. typhimurium strains isolated at the hos- pital in Nairobi. Special emphasis was placed on the presence of some tetracycline resistance genes in order to study the high frequency of tetracy- cline resistance observed in enteric strains in sev- eral developing countries (2, 10,25,29).

629

KARIUKI et a1

MATERIALS AND METHODS

Strains Ninety-seven Salmonella typhimurium strains ob-

tained from patients at a large hospital in Nairobi, Kenya were isolated from clinical samples during the period 1988-90. Fifty strains were from blood cultures, 32 from stool cultures, three from cerebro- spinal fluids, and a total of 12 strains from urine, pus or epidural fluids. The strains used in this study were identified according to the criteria listed in Bergey's Manual of Systematic Bacteriology (4).

Tetracycline MIC determination Minimum inhibitory concentration (MIC) of oxy-

tetracycline hydrochloride (Sigma Chemical Co., St. Louis, MO) (with a potency of 920 pg/ml) for the strains was examined by the two-step dilution method according to Ericsson & Sherris (9) using standardized inoculations of about lo4 CFU to a spot 5 mm in diameter. The results were interpreted according to the criteria set by the International Collaborative Study (ICS) in antibiotic sensitivity testing (9). E. coli strains possessing the tet'A (pTET A in F186), B (pKT007 in F175) and C (pBR322 in F176) genes, and tetracycline-sensitive strains, all from the collec- tion of the Department of Microbiology and Immu- nology, the Norwegian College of Veterinary Medi- cine, Oslo, Norway, and the Centers for Disease Con- trol, Atlanta, GA, were used as positive and negative controls throughout the study.

Plasmid DNA isolation Plasmid DNA was isolated both as described by

Birnboim & Doly (3), and as described by Kado & Liu (16), and separated by electrophoresis under water cooled conditions in a 1% vertical standard low en- dosmotic agarose (Bio Rad Laboratories, Richmond, CA) in Tris phosphate-EDTA buffer (0.8 M Tris, 0.008 M sodium EDTA) (pH 8.0) for four h at 80 V, 60 mA. The gels were then stained with ethidium bromide and pictures taken on an ultraviolet transil- luminator. Plasmid DNA from E. coli strains V517 and pDkg was used as molecular weight control. The different plasmid profiles were grouped by the molecular weights of the plasmids into six plasmid profile groups, each consisting of a core profile of the large plasmids.

Colony and Southern blot hybridization with y '2P-labelled synthetic probes

The colony blots on nylon membranes (Biotech- nology Systems, Boston, Mass) were prepared and single-stranded DNA was fixed to the nylon mem- branes by exposure to UV light for five min (11). Membrane-bound DNA was hybridized with oligo- nucleotide probes for tetracycline resistance gene types A (tet' A: 5'GCC TCC TGC GCG ATC TGG3', Tm 62"C), B (tet' B: 5'CAG TGC TGT TGT TGT

CAT TAA3', Tm 62°C) and C (tet' C: 5'TTG CAT GCA CCA TTC CTT GCG3', Tm 66°C). These oligonucleotide sequences were generated from data obtained from DePaola et al. (8), Hillen & Schollmeier (13), and Waters et al. (31). The probes were labelled in the 5'-end by y 32P ATP using T4 kinase (18), and hybridized to DNA on the membranes, probes A and B at 60"C, and C at 64°C (18). The plasmid DNA in the gels were transferred to nylon membranes (Gene Screen, NEN Research Products, Boston, Mass) by a Southern blot procedure as described by Wasteson & OIsvik (30), and then hybridized with the probes as described above. The signals were developed by ex- posure to X-ray diagnostic film with intensifying screens at -70°C for 48 h.

RESULTS

MIC for tetracycline All the 97 strains tested were initially resistant

to oxytetracycline, but nine strains became sen- sitive during laboratory handling. The distri- bution of MIC for tetracycline among the strains is shown in Fig. 1. Two clusters of resist- ance¶ at low MIC (4-8 pg per ml) and high MIC (128 pg per ml or higher) seem to be present. There was, however, no correlation between low and high MIC and genotype of resistance or other characteristics examined.

Plasmid content The 97 strains were in general different with

respect to plasmid content, and no clone seems to dominate during the period the strains were collected. The plasmids could be grouped by

50 I 1

.B 40

(r ' 30 0

i i m 8

10

0 0.5 1 2 4 8 16 32 64 128256

MIC (pglml) oxytetracycline Fig. 1. Distribution of minimum inhibitory concen- tration of tetracycline among clinical isolates of Salmonella typhimurium from a hospital in Nairobi, Kenya.

630

TETRACYCLINE RESISTANCE GENES IN SALMONELLA TYPHIMURIUM

their molecular weights into six plasmid profile groups, each consisting of a core profile of large plasmids. The plasmid profile groups and the molecular weights of the plasmids constituting them are listed in Table 1. Seventy-one of the strains tested possessed a 65-MDa plasmid, 16 strains a 63-MDa plasmid, and one strain a 60- MDa plasmid in addition to other plasmids (Fig. 2). During laboratory handling, a total of nine strains lost the large (65-60-MDa) plasmid. These plasmids were seen in the first plasmid profiling performed locally. A pattern of minor plasmids was present in most of the strains.

Tetracycline resistance genes Twenty-seven out of 97 strains tested pos-

sessed one of the tetracycline resistance determi- nants A, B or C (Table 2). Twenty of the strains possessed the tet' A, six the tet' B, and four the tet' C determinant. Three strains carried both tet' A and tet' B genes. Fig. 3 shows the tet' C probe hybridizing with the 65-MDa plasmid. Both the tet' A and B genes were also found located on the large plasmid

Clinical origin of strains There was no correlation between MIC, plas-

mid content or tetracycline genotype and the clinical isolation site of the strains.

DISCUSSION

Initially all the 97 strains of S. typhimurium investigated were resistant to tetracycline, and

TABLE I , Plasmid profile groups of 97 tetracycline- resistant Salmonella typhimurium strains isolated at

a hospital in Nairobi, Kenya Plasmid No. of Molecular weight (MDa)" of plas- profile strains mids groups Majorb Minor' 1 12 65 5.2 4.0 2 37 65 36 5.2 4.0 2.4 2.0 3 12 65 46 36 5.2 4.0 2.4 2.0 4 10 65 46 4.0 5 17 63 (60)d 4.0

\ ,

6 9 5.2 4.0 2.4 a MDa = Megadaltons

Major = Plasmid always present Minor =Plasmid not always present A 60-MDa plasmid was present in one strain only (see also Fig. 2).

Fig. 2. Plasmid profiles of strains possessing the large virulence plasmids of different sizes Lane A: S. typhimurium strain B: S. typhimurium strain C: S. typhimurium strain

DPHPT 30 (65 MDa) DPHPT 26 (63 MDa) DPHFT 24 (60 MDa)

similar results have been observed both for nor- mal cohabiting enteric Escherichia coli and EPEC strains isolated from neonates born at the same hospital (25, 26). S. typhimurium strains from other parts of the world indicate a much lower percentage of tetracycline resistance (10, 14, 15).

Genetic characterization of bacteria using plasmid profiles has proved a useful tool in the investigation of outbreaks and epidemics (20, 23), and Holmberg et al. (15) found plasmid pattern analysis rapid, relatively simple to per- form, as specific as phage typing, and superior to biotyping and antimicrobial susceptibility

63 I

KARIUKI e/ al.

TABLE 2. Distribution of tetracycline-resistant gene types among tetracycline-resistant Salmonella typhi-

murium straias from a homital in Nairobi. Kenva Clinical Total No. of strains origin tet A tet B tet C Non-ABC& Blood 50 12 5 0 33 Stool 32 5 1 2 24 CSFb 3 2 0 1 0 Othersc 12 1 0 1 10 a Non-ABC=Strains negative for tet' A, B, and C

resistance determinants CSF = Cerebrospinal fluid. Others = Urine, pus and aspirated pleura and knee- joint fluids.

typing. The plasmid profile grouping of the strains at the hospital revealed that several of these strains were quite heterogeneous with re- spect to plasmid content, and that apparently no single clone was the cause. This would also have been somewhat surprising as most of the patients were probably infected before arriving at the hospital. A similar heterogenicity in plas-

Fig. 3. Localization of tet' C genes on the 65-MDa plasmid in two S. typhimurium strains by hybridiza- tion on a Southern blot of the plasmid separation gel A: S. typhimurium strains DPHPT 19 B: S. typhimurium strains DPHPT 18 A.

632

mid content was observed during an outbreak of enteropathogenic E. coli of serotype 01 1l:HNT at the same hospital (25). Initially all the 97 strains possessed a large plasmid of 60-65 MDa; a plasmid of this size is assumed to carry genes involved in the virulence of the bacterium (12). However, strains containing only one plas- mid of considerably lower weight have been iso- lated from cattle and humans affected by S. typhimuriurn (23). The strains that lost the large plasmid also became sensitive to tetracycline.

Several authors have lately questioned the res- ervoirs of tetracycline-resistant salmonella or the transmittable tetracycline resistance genes (6, 24). The strains examined in our study pos- sessing genes hybridizing with the oligonuc- leotides used had all these genes located on plas- mids. Extensive use of tetracyclines as a feed additive for growth promotion in cattle in sev- eral countries, among them the USA, has raised the question whether this might be one of the reasons for the increase in the isolation fre- quency of tetracycline- (and other antibiotic-) resistant salmonella in humans (6, 14, 24). This seems not to be the case in Kenya where tetracy- cline is only used on a small scale as an additive in chicken feed.

It is important not only to demonstrate the presence of phenotypical tetracycline resistance; a further characterization of the tetracycline re- sistance genes will present additional infor- mation, which will assist in the investigation of the epidemiology of the tetracycline-resistant S. typhimurium strains, and the molecular epidemi- ology of the tetracycline resistance genes (5, 7, 21,27,28). The tet' A gene was observed in 20% of these hospital isolates; tet' B and tet' C in 6 and 4%, respectively. Three strains carried both tet' A and tet' B determinants. One cannot draw strict conclusions based on the present results concerning the reservoirs of tetracycline-resist- ant genes in the S. typhimurium strain collections examined, as a majority of the strains were non- typeable by the oligonucleotides employed. The tetracycline genes might be of types other than A, B or C , or, due to the narrow specificity oligonucleotides give, very limited mutations of one or two base pairs can result in no hybridiza- tion (13, 18, 21, 22, 31).

Using polynucleotides, Marshall et ul. (19) found class A determinants in 21.3%, class B (7" 10 type) in 73.3%, class C in 8%, and multideter-

TETRACYCLINE RESISTANCE GENES IN SALMONELLA TYPHIMURIUM

minants in 3.5% of a collection of enteric micro- organisms. It is interesting to observe that tet‘ B was not as predominant in our materials as had been observed by Marshall et al. (19). These and other data suggest that the gram- negative enteric flora of humans and animals might serve as reservoirs for these resistance determinants (6, 24).

The use of probes for characterization and classification of antimicrobial resistance genes is an important diagnostic feature assisting in the demonstration of antimicrobial resistance pools and the epidemiology of these genes in different bacterial populations (20, 27, 28). Such information might help develop policies for a safer and more efficient use of antimicrobials.

REFERENCES

1.

2.

3.

4.

5.

6.

7.

8.

Anderson, E. S.: Drug resistance in Salmonella typhimurium and its implications. Br. Med. J. 3: 333-339, 1968. Bhatiu, R., Svaze, S. & Agarwul, D. S.: Transfer- able multiple drug resistance in Salmonella typhi- muriurn. Indian J. Med. Res. 74: 642-647, 1981. Birnboim, H. C. & Doly, J.: A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acids Res. 7: 15 13-1 523, 1979. Brenner, D. G. : Enterobucteriucea. In: Krieg, N. R. & Holt, J. G. (Eds.): Bergey’s manual of sys- tematic bacteriology, Vol. I . Williams and Wilk- ins, Baltimore, 1984, pp. 408-5 16. Bryan, R. N . , Ruth, J. L . , Smith, R. D. & Le Bon, J. M.: Diagnosis of clinical samples with synthetic oligonucleotide hybridization probes. In: Levin, L. (Ed.): Microbiology. American Society for Microbiology, Washington D.C.,

Cohen, M . : Epidemiological studies of the emerg- ence and transmission of drug-resistant bacteria. In: Olsvik, 0. & Bukholm, G. (Eds.): Application of molecular biology in diagnosis of infectious diseases. Norwegian College of Veterinary Medi- cine, Oslo, 1990, pp. 33-38. Davies, J. & Gray, G.: Evolutionary relationships among genes for antibiotic resistance. CIBA Foundation Symp. 1102: 219-232, 1984. DePuola, A, , Hill, W E. & Hurrell, E: Determi- nation of tetracycline resistance genotypes in pond raised catfish microflora using DNA probes. Abstract no. Q145, American Society for Microbiology’s Annual Meeting 1990, American Society for Microbiology, Washington D.C., 1990, p. 312.

1986, pp. 113-116.

9. Ericsson, H . M. & Sherris, J. C.: Antibiotic sensi- tivity testing. Report of an international colla- borative study. Acta path. microbiol. scand. Sect. B, Suppl. No. 217, 1971.

10. Farrar, W E . : Antibiotic resistance in developing countries. J. Infect. Dis. 152: 1103-1 106, 1985.

1 1. Griinstein, M . & Hogness, D. S.: Colony hybridi- zation; a method of isolation of cloned DNAs that contain a specific gene. Proc. Natl. Acad. Sci. USA, 72: 3961-3965, 1975.

12. Gulig, II A. & Curtiss III, R.: Plasmid-associated virulence of Salmonella typhimurium. Infect. Im- mun. 55: 2891-2901, 1987.

13. Hillen, W & Schollmeier, K.: Nucleotide sequence of the Tn 10 encoded tetracycline RT resistance gene. Nucleic Acids Res. 11: 525-539, 1983.

14. Holmberg, S. D., Osterholm, M. T , Senger, R. A. & Cohen, M. L.: Drug-resistant Salmonella from animals fed antimicrobials. N. Engl. J. Med.

15. Holmberg, S. D., Wachsmuth, I. K . , Hickman- Brenner, E W & Cohen, M. L.: Comparison of plasmid profile analysis, phage typing and anti- microbial susceptibility testing in characteriza- tion of Salmonella typhimurium isolates from out- breaks. J. Clin. Microbiol. 1156: 175-182, 1984.

16. Kudo, C. I. & Liu, S. T : Rapid procedure for detection and isolation of small and large plas- mids. J. Bact. 145: 1365-1373, 1981.

17. Levy, S. B., McMurry, L. M. , Burdett. V , Courva- lin, II, Hillen, W, Roberts, M. C . & Taylor, D. E.: Nomenclature for tetracycline resistance determi- nants. Antimicrob. Agents Chemother. 33:

18. Maniatis, T , Fritsch, E. E & Sanbrook, J.: Mol- ecular cloning: A laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, 1981.

19. Marshall, B., Tuchibuna, C. & Levy, S. B.: Fre- quency of tetracycline resistance determinants classes among lactose fermenting coliforms. Anti- microb. Agents Chemother. 24: 835-840, 1983.

20. Mayer, L. W: Use of plasmid profiles in epide- miological surveillance of disease outbreaks and in tracing of antibiotic resistance. Clin. Micro- biol. Rev. 1: 228-243, 1988.

21. Mendez, B., Tachibana, C. & Levy, S. B.: Hetero- geneity of tetracycline resistance determinant. Plasmid 3: 99-108, 1980.

22. Olsvik, O., Hornes, E., Wasteson, Y & Lund, A.: Detection of virulence determinants in enteric Escherichiu coli using nucleic acid probes and polymerase chain reaction. In: Wadstrom, T , Ma- lela, P H., Svennerholm, A-M. & Wolf- Watz, H. (Eds.): Molecular pathogenesis of gastrointesti- nal infections. Plenum Press, New York, 1991,

23. Olsvik, 0., Sorum, H. , Birkness, K . , Wachsmuth, K., f idstad, M. , Lassen, J . , Fossum, K. & Feeley, J . C.: Plasmid characterization of Salmonella

311: 617-622, 1984.

1373-1374, 1989.

pp. 267-272.

63 3

KARIUKI et al.

typhimurium transmitted from animals to humans. J. Clin. Microbiol. 22: 336-338, 1985.

24. Roberts, M. C.: Characterization and epidemi- ology of tetracycline resistance genes. In: Olsvik, 0. & Bukholm, G. (Eds.): Application of molecu- lar biology in diagnosis of infectious diseases. Norwegian College of Veterinary Medicine, Oslo, 1990, pp. 39-44.

25. Senerwa, D., Mutanda, L. N. , Gathuma, J. M. & Olsvik, 0. : Antimicrobial resistance of enteropa- thogenic Escherichia coli strains from a noso- comial outbreak in Kenya. APMIS 99: 728-734, 1991.

26. Senerwa, D., Olsvik, O., Mutanda, L. N., Lindqvist, K. J,, Gathuma, J. M., Fossum, K. & Wachsmuth, K.: Enteropathogenic Escherichia coli serotype 01 11:HNT isolated from preterm neonates in Nairobi, Kenya. J. Clin. Microbiol. 27: 1307-131 1, 1989.

27. Tenover, E C.: Studies of antimicrobial resistance genes using DNA probes. Antimicrob. Agents Chemother. 29: 721-725, 1986.

28. Tenover, E C.: Diagnostic deoxyribonucleic acid probes for infectious diseases. J. Clin. Microbiol. Rev. 1: 82-101, 1988.

29. Wamola, I. A . & Mirza, N. B.: Problems of Sal- monella infections in a hospital in Kenya. East Afric. Med. J. 58: 677-683, 1981.

30. Wasteson, I: & Olsvik, 0.: Specific DNA frag- ments coding for STI and LT I toxins, and K88 (F4) adhesin in enterotoxigenic Escherichia coli. J. Vet. Med. B, 38: 445452, 1991.

31. Waters, S. H. , Rogowsky, I?, Grinsted, J.. Alten- buchner, J. & Schmitt, R.: The tetracycline resist- ance determinants of RP1 and Tn 1721: nucleo- tide sequence analysis. Nucleic Acids Res. 11: 6089-6105, 1983.

634