Lack of clonal relationship between non-typhi Salmonella straintypes from humans and those isolated from animals

living in close contact

Samuel Kariuki a;e;�, Gunturu Revathi b, Francis Gakuya c, Victor Yamo d,Jane Muyodi a, C. Anthony Hart e

a Centre for Microbiology Research, Kenya Medical Research Institute, P.O. Box 43640, Nairobi, Kenyab Department of Medical Microbiology, Kenyatta National Hospital, Nairobi, Kenya

c Department of Veterinary Public Health, Pharmacology and Toxicology, University of Nairobi, Nairobi, Kenyad Kenchic Limited, Nairobi, Kenya

e Department of Medical Microbiology and Genitourinary Medicine, University of Liverpool, Liverpool L69 3GA, UK

Received 4 December 2001; received in revised form 7 February 2002; accepted 14 March 2002

First published online 6 May 2002

Abstract

Antibiogram patterns and chromosomal DNA typing were used to compare 151 non-typhoidal Salmonella spp. (NTS) isolated frompatients and 78 from animals, environmental or food specimens obtained within or near the homes of patients with invasive salmonellosis.The majority of NTS from humans (137; 90.7%) were Salmonella enterica serotype Typhimurium (S. Typhimurium) and S. Enteritidis.Chicken specimens and feeds produced (24; 52.2%) S. Enteritidis, while S. Agona was the predominant (20; 77%) serovar among pigs anddairy cows. The majority (97; 64.2%) of NTS from humans were multidrug resistant, while NTS from cows, pigs, beef carcass swabs andsewers were fully susceptible to all antibiotics tested. Pulsed-field gel electrophoresis patterns of XbaI-digested genomic DNA of NTS fromthe humans and the chickens were different. However, S. Enteritidis from chickens, and S. Braenderup and S. Agona from cows and pigswere clustered together in one group. There was no significant relatedness between NTS isolates from humans and those from animals,food or the environment in close contact to humans. = 2002 Federation of European Microbiological Societies. Published by ElsevierScience B.V. All rights reserved.

Keywords: Non-typhoidal Salmonella ; Human and animal; Characterization

1. Introduction

Non-typhoidal Salmonella spp. (NTS) are an importantcause of infection in both humans and animals. Largeoutbreaks of infection have been associated with food-borne transmission including that from contaminatedpoultry and poultry products, meat and milk and otherdairy products [1^3]. Although NTS typically cause gas-troenteritis they are becoming increasingly important bac-terial pathogens in developing countries causing bacterae-mia and other invasive disease [4], and there is anincreasing prevalence of multidrug resistance [5,6]. In Ken-

ya, among immunocompromised individuals and the veryyoung, NTS frequently causes bacteraemic infections.Multidrug resistance, particularly to the commonly avail-able antibiotics, poses a major health concern, as alterna-tive therapeutic choices are either unavailable or too ex-pensive to be a¡ordable for most patients.Several studies have documented that farm animals are

the major reservoir for NTS in industrialised countries.For instance in the USA it is estimated that over 95% ofNTS infections are related to food-borne transmission [7].In most industrialised countries and some less industrial-ised countries Salmonella enterica serotype Enteritidis(S. Enteritidis) is transmitted through consumption offoods containing raw or incompletely cooked eggs andhome-cooked products containing eggs [1,8^10]. CertainS. Enteritidis clones are stable over long periods of timecausing several outbreaks in di¡erent geographical areas

0928-8244 / 02 / $22.00 = 2002 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved.PII: S 0 9 2 8 - 8 2 4 4 ( 0 2 ) 0 0 3 0 9 - 7

* Corresponding author, at address a. Tel. : +254 (2) 72 01 63;Fax: +254 (2) 71 16 73.

E-mail address: skariuki@wtnairobi.mimcom.net (S. Kariuki).

FEMSIM 1410 24-6-02

FEMS Immunology and Medical Microbiology 33 (2002) 165^171

www.fems-microbiology.org

[11]. The multidrug resistant (MDR) S. Typhimuriumphage type (DT) 104 strain, which has been responsiblefor epidemics particularly in Europe, the USA and Cana-da, has reservoirs in cattle and is transmitted mainlythrough consumption of contaminated meat, milk andmilk products [3,12^14]. However, there are no datafrom developing countries including Kenya on the likelysources of NTS that cause human infections. In thepresent study we used antimicrobial susceptibility testing,and plasmid and genomic DNA typing to investigate ifNTS from animals and environmental sources were relatedto NTS isolated from humans living in close contact.

2. Materials and methods

2.1. Patients

These were acute adult admissions to the medical wardsat the Kenyatta National Hospital and the Aga KhanHospital, Nairobi from 1998 to 2000. Blood cultures andstools were obtained from all febrile patients prior to anti-biotic treatment. Stools were obtained from cases of per-sistent diarrhoea.

2.2. Specimens from animals and environmental specimens

Cases from whom NTS were isolated were followed totheir home areas where more epidemiological data wereobtained and the following specimens were obtained: fae-ces and intestinal contents from rodents trapped aroundthe homes, water from nearby rivers and streams, and rawand cooked food from homes of patients and vendors inmarkets serving the areas. In addition, rectal swabs fromfarm animals including pigs, chickens, cows, and goatswere taken from around the homes of index cases orfrom neighbouring homes. Five abattoirs for beef andone for camel meat were sampled by means of carcassswabs, and e¥uent was sampled from various points in-side and outside the abattoirs. For epidemiological com-parison, faecal and liver specimens were obtained fromsick chickens identi¢ed from two large commercial chickenfarms 20 km from Nairobi. Stools were also obtained froma total of 16 workers from the abattoirs (6), pig and chick-en farms (10).

2.3. Laboratory procedures

Blood cultures and stool specimens were processed us-ing standard techniques. Brie£y, blood cultures were incu-bated in 5% CO2 at 37‡C for 18 h and if signs of bacterialgrowth were observed (air bubbles, turbidity or both) theywere subcultured on sheep blood agar and chocolate agar.The remaining blood cultures were reincubated for a fur-ther 7 days or until positive. Stools were processed bydirect plating onto selective media (XLD and brilliant

green agar) (Oxoid, Basingstoke, UK) and by overnightenrichment in selective Selenite F broth (Oxoid) followedby plating onto XLD and brilliant green agar (Oxoid), andincubated in air at 37‡C for 18 h. NTS were identi¢edusing agglutinating antisera (Murex Biotech, Dartford,UK) and their identi¢cation was con¢rmed biochemicallyusing API 20E strips (API System, Montalieu Vercieu,France).Environmental samples were initially cultured in Rappa-

port^Vassiliadis soya broth (Oxoid) for enrichment. Thebroth culture was then subcultured onto XLD and bril-liant green agars (Oxoid). NTS were identi¢ed as forblood and stool cultures. All NTS isolates were stored at370‡C on Protect Beads (Technical Service Consultants,Heywood, UK) until analysed.

2.4. Antimicrobial susceptibility testing

Susceptibility tests with commonly used antimicrobialagents were performed on Mueller^Hinton (Oxoid) agarby the disk di¡usion technique. The antibiotic disks (Ox-oid) used were ampicillin 10 Wg, co-amoxiclav 10:20 Wg,cefuroxime 30 Wg, ceftazidime 30 Wg, co-trimoxazole 25 Wg,chloramphenicol 30 Wg, cipro£oxacin 5 Wg, gentamicin10 Wg, nalidixic acid 10 Wg, streptomycin 10 Wg, and tetra-cycline 30 Wg. E-test strips (AB BioDisk, Solna, Sweden)were used to determine the minimum inhibitory concen-trations (MICs) of the same antimicrobial agents accord-ing to the manufacturer’s instructions. Escherichia coliATCC 25922 was used as control for bacterial growthand potency of antibiotics on disks and E-test strips.Disk zone sizes and E-test MICs were interpreted accord-ing to the National Committee for Clinical LaboratoryStandards guidelines [15].

2.5. Pulsed-¢eld gel electrophoresis (PFGE) ofmacrorestricted chromosomal DNA

Chromosomal DNA was prepared in agarose plugs asdescribed by Kariuki et al. [6] from an overnight bacterialculture in Luria broth (Oxoid). Agarose plugs were thendigested with XbaI (Life Technologies, Paisley, UK) ac-cording to the manufacturer’s instructions. PFGE of aga-rose plug inserts was then performed in a CHEF-DR IIsystem (Bio-Rad Laboratories, Hercules, CA, USA) on ahorizontal 1% agarose gel for 24 h at 120 V, pulse time of1^40 s, at 14‡C. A VDNA digest consisting of ca. 22 frag-ments of increasing size from 48 kb to about 1000 kb wasincluded as a DNA size standard. The gel was stained with0.05% ethidium bromide and photographed on an UVtransilluminator (UVP Inc., San Gabriel, CA, USA).The restriction endonuclease digest patterns were inter-preted by considering migration distance and intensity ofall visible bands, and by using guidelines described byTenover et al. [16]. Dendrograms of genetic similaritywere then constructed using data obtained by the Dice

FEMSIM 1410 24-6-02

S. Kariuki et al. / FEMS Immunology and Medical Microbiology 33 (2002) 165^171166

coe⁄cient method and clustered by the unweighted pairgroup arithmetic averaging method (Molecular Finger-printing Program version 1.4.1, Bio-Rad, UK) puttingthe isolates into PFGE types. Isolates within each PFGEtype produced indistinguishable PFGE fragment patternsin which case they were likely to be identical strains orthey showed one to two band di¡erences in their fragmentbanding patterns and were considered to be closely re-lated. Isolates from di¡erent PFGE were su⁄ciently di¡er-ent in their fragment banding patterns as to render themdi¡erent strain types.

2.6. Plasmid studies

Plasmid DNA extraction was performed using a Plas-mid Mini Prep Kit (Qiagen, West Sussex, UK) accordingto the manufacturer’s instructions. Plasmids were sepa-rated by electrophoresis on horizontal 0.8% agarose gelsat 100 V for 2 h. Plasmid sizes were determined by co-electrophoresis with plasmids of known sizes from E. colistrains V517 (NCTC 50193) (53.7, 7.2, 5.6, 3.9, 3.0, 2.7,2.1 kb) and 39R861 (NCTC 50192) (147, 63, 43.5, 6.9 kb).DNA bands were visualised with an ultraviolet transillu-minator (UVP Inc.) after staining with 0.05% ethidiumbromide.

3. Results

3.1. Bacteria from humans

A total of 592 specimens from 402 inpatients were pro-cessed between January 1998 and April 2000, giving atotal of 151 non-duplicate NTS. A total of 58 NTS eachcame from blood and stools of the same patients; 12 NTSwere isolated from blood and 49 NTS were obtained fromstools only. These isolates came from sporadic cases thatcame to seek medical attention at two tertiary hospitals inNairobi, Kenya. It was therefore di⁄cult to incriminateany particular common food source. The main Salmonellaserovars were S. Typhimurium (79; 52.3%) and S. Ente-

ritidis (58; 38.4%). Other serotypes including S. Agona,S. SaintPaul, S. Braenderup and S. Durban in small num-bers (1^3) were isolated from blood, stool cerebrospinal£uid, pus or liver biopsies.

3.2. Bacteria from animals and environmental specimens

A total of 220 swabs from beef carcasses and 47 abattoire¥uent samples taken at various points along the opensewer lines from ¢ve abattoirs were processed. From thesesamples only four NTS were isolated. In addition, faeceswere processed from 210 dairy cows, 122 pigs and 228rodents, and these gave a total of 28 NTS. A further 39NTS came from chickens and soya feed supplements fromtwo large commercial farms just outside Nairobi whilefaeces from chickens from small-scale farmers from thestudy area yielded seven NTS. S. Agona (20; 76.9%)was the main serovar from pigs and cows which camefrom the same farms, while S. Enteritidis (24; 52.2%)was the main serovar isolated from chicken specimens.Faeces from goats, swabs from camel carcasses, andfood and water samples from the homes of index casesdid not yield any NTS.

3.3. Epidemiological data from homes of patients

All 151 patients that were followed back to their homeareas came from within 40 km of the two study hospitals.A total of 122 (80.8%) kept 10^15 chickens and 1^5 otheranimals (goats, pigs, cows or pigs) in their homesteads toprovide eggs, meat and milk, respectively. The mainsource of meat was the local butcher who got his meatfrom one of ¢ve abattoirs within the city. Occasionally, thechickens and goats were also slaughtered in homesteadsfor meat. These small-scale farmers also grew vegetablesfor sale and household use. The other 29 patients rentedhouses and depended wholly on local markets and butch-ers for vegetables and meat, respectively. Treated tapwater was available to all visited homesteads but on occa-sional water shortages during the drought period (Janu-ary^March) nearby rivers or streams became sources of

Table 1MIC using the E-test of 11 antimicrobial agents for 151 non-typhi Salmonella isolates from medical wards at two hospitals in Nairobi (1997^2000)

Antimicrobial agent MIC (Wg ml31) % resistant

Range Mode MIC50 MIC90

Ampicillin 0.75^s 256 s 256 s 256 s 256 65Co-amoxyclav 0.5^32 0.75 6 16 8Cefuroxime 2^128 3 8 12 18Ceftazidime 0.125^16 0.25 0.5 2 1Co-trimoxazole 0.032^s 32 s 32 s 32 s 32 60Chloramphenicol 1.5^s 256 s 256 32 s 256 40Cipro£oxacin 0.006^0.25 0.023 0.023 0.08 0Gentamicin 0.19^64 0.75 1 8 9Nalidixic acid 1^s 256 3 3 s 256 11Streptomycin 3^s 1024 32 s 1024 s 1024 90Tetracycline 0.75^192 1 16 64 48

FEMSIM 1410 24-6-02

S. Kariuki et al. / FEMS Immunology and Medical Microbiology 33 (2002) 165^171 167

water for over 50% of the homes. A survey on use ofantibiotics by farmers showed that tetracycline is widelyavailable for sale ‘over the counter’ and is used extensivelyin poultry rearing. It is added to commercial poultry feedsand in drinking water for birds of all ages. Tetracycline,penicillin and sulfonamides were also used extensively indairy animals for prophylaxis, but there was no indicationof their use in beef animals.

3.4. Drug susceptibility testing

There were two main antibiotic susceptibility patternsexhibited by the NTS from humans in the current study:fully susceptible isolates constituting 34% and MDR (re-sistant to two or more antibiotics) isolates accounting for64.2%. For all commonly available antibiotics includingampicillin, co-trimoxazole, streptomycin, chlorampheni-col, and tetracycline MIC values were high for the MDRisolates (Table 1). Three of the MDR S. Typhimuriumisolates showed resistance to ceftazidime (MIC=16 Wgml31). Few isolates (2%) were resistant to only one anti-biotic, mainly to ampicillin or co-trimoxazole. Althoughno NTS were resistant to cipro£oxacin, a number of them(4%) showed reduced susceptibility (MIC=0.125 Wg ml31)and 11% were resistant to nalidixic acid.All NTS isolated from cows, pigs and sewers were fully

susceptible to all 11 antimicrobials tested. In contrast 15(38.5%) of the NTS from chickens from the two largecommercial farms were multiply resistant to ampicillin,co-trimoxazole, tetracycline and streptomycin (all MICs 256 Wg ml31) and co-trimoxazole (MIC s 32 Wg ml31).All ¢ve S. Enteritidis from small-scale farms were resistantto tetracycline only (MIC=16^32 Wg ml31).

3.5. PFGE patterns of NTS isolates

For NTS from humans that gave indistinguishablePFGE banding patterns only representative isolates wereselected for the dendrogram analysis. Fragment sizes of110 kb or less were omitted from analysis as they may

6

Fig. 1. A: Dendrogram showing estimated relatedness among humanand animal S. Typhimurium isolates. Hu, number of human isolateswithin each PFGE subgroup; Cl, number of chicken isolates withineach PFGE subgroup from two large commercial farms. Isolates withPFGE patterns showing s 70% similarity (corresponding to 1^2 banddi¡erence) were likely to be related, hence within the same PFGE type.Isolates showing 100% similarity produced PFGE patterns that were in-distinguishable. B: Dendrogram showing estimated relatedness amonghuman and animal S. Enteritidis isolates. Hu, number of human isolateswithin PFGE subgroup; Cl, number of chicken isolates within eachPFGE subgroup from two large commercial farms; Cs, number ofchicken isolates within each PFGE subgroup from small-scale farms.Isolates with PFGE patterns showing s 70% similarity (correspondingto 1^2 band di¡erence) were likely to be related, hence within the samePFGE type. Isolates showing 100% similarity produced PFGE patternsthat were indistinguishable.

FEMSIM 1410 24-6-02

S. Kariuki et al. / FEMS Immunology and Medical Microbiology 33 (2002) 165^171168

represent plasmids. Fig. 1A,B are simpli¢ed dendrogramanalyses of S. Typhimurium and S. Enteritidis, respec-tively, showing the relatedness of strains from both hu-mans and animals. From the PFGE typing, S. Typhimu-rium isolates produced three main patterns, one commonto most isolates (54; 68.4% in PFGE type A) and twoother patterns (PFGE types C and D) that di¡ered fromeach other and from the main pattern by four bands.Within PFGE type A, 42 (53.2%) of S. Typhimurium pro-duced PFGE patterns that were indistinguishable.For S. Enteritidis from humans there were two main

PFGE patterns, the commonest one contained 41(70.7%) isolates in PFGE type A of which 28 (48.3%)produced PFGE patterns that were indistinguishable. Atotal of 12 (15%) were contained in PFGE type B. Forboth S. Typhimurium and S. Enteritidis there was limiteddiversity in PFGE patterns for the strains from one yearto another during the study period. PFGE analysis ofS. SaintPaul (three isolates), S. Agona (two isolates) andS. Braenderup (four isolates) from the humans indicatedthat the isolates within each serovar were closely related.The 20 S. Agona isolates from domestic animals (12

isolates from pigs and eight isolates from cows) producedindistinguishable PFGE banding patterns. However, therewere two PFGE patterns in S. Choleraesuis, one for iso-lates from chickens and the second for isolates from cattlecarcass swabs. All nine S. Braenderup isolates from chick-ens produced indistinguishable PFGE banding patterns.Similarly, both S. Typhimurium and S. Enteritidis isolatesfrom chickens from the commercial farms gave indistin-guishable PFGE patterns (Fig. 1), indicating that singlecommon strain types caused infection on both farms.The PFGE patterns of S. Enteritidis and S. Typhimuri-

um from humans were distinct and di¡erent from corre-sponding serovars isolated from either animals or environ-mental sources. Only three S. Enteritidis from humans(PFGE type C) showed close relatedness to S. Enteritidisfrom chickens, while three S. Typhimurium isolates fromchickens were only distantly related to isolates from hu-mans. For S. Enteritidis XbaI digest fragments from hu-man and animal isolates di¡ered in four bands at 120^280kb, while for S. Typhimurium banding patterns di¡ered infour bands at 520^640 kb. However, two S. Agona isolatesfrom pig farm workers produced PFGE patterns that wereindistinguishable from the 20 S. Agona isolates from ani-mals. PFGE fragment digests from S. Braenderup isolatesfrom humans di¡ered from those isolated from chickensand chicken soya feed supplements in four bands at 200^280 kb, clearly indicating that these strains were di¡erent.Similarly, S. Saintpaul isolates from humans were signi¢-cantly di¡erent from those isolated from cows in fourbands at 150^220 kb.

3.6. Plasmid DNA analysis

All MDR S. Typhimurium and S. Enteritidis from hu-

mans and from chickens had a common 100^110-kb plas-mid that was found to encode multiple antibiotic resis-tance phenotype. In vitro conjugation tests [17] showedthat the 100^110-kb plasmids transferred resistance to am-picillin, co-trimoxazole, streptomycin and tetracycline enbloc to E. coli K12. In addition, S. Typhimurium isolatescontained smaller non-transferable plasmids, 3, 5 and 15kb in size, while S. Enteritidis isolates from chickens con-tained only the 3-kb additional plasmid. Plasmids couldnot be isolated from sensitive strains of NTS from hu-mans, animals or environmental sources.

4. Discussion

Although in the USA and Europe domestic animals arethe major reservoir and foods of animal origin are themajor vehicles of NTS infection in humans [1,3,12], our¢ndings indicate that this may not be the case for the NTSwe studied. We found that the distribution of NTS sero-vars among animals from home environments of indexcases was important. No signi¢cant numbers of NTSwere isolated from rodents, carcass swabs or e¥uentfrom abattoirs, and the serovars S. Agona and S. Chole-raesuis that were isolated from pigs and dairy cows didnot appear among humans. Human isolates were predomi-nantly S. Typhimurium and S. Enteritidis. In contrast,serovar S. Enteritidis was the predominant isolate fromchicken specimens and feed from two large-scale commer-cial farms, in addition to other serovars.Antimicrobial susceptibility testing showed that NTS

from humans were multiply resistant to antibiotics com-monly available in Kenya whereas the NTS from animalsand environmental sources close to the homes of indexcases were fully susceptible to all antibiotics tested. Acommon transferable 100^110-kb plasmid was found toencode the multidrug resistance phenotype amongS. Typhimurium and S. Enteritidis from humans andfrom the chickens from the large-scale commercial farms.As has been previously documented [5], the 100^110-kbplasmids were common among MDR NTS serovarsfrom di¡erent ecological sources in Kenya. However, itappears that this MDR-encoding plasmid has not beentransmitted to NTS isolated from cattle, pigs, or chickensfrom the small-scale farms that we studied despite the factthat index cases from whom MDR NTS were isolatedlived within the same homes or homes close to these ani-mals. The MDR NTS from chickens were all from the twolarge-scale chicken farms that used soya feed supplementsimported from a source in Europe. These feed supple-ments had the same NTS serovars isolated from them,thus giving credence to the assumption that the importedsoya feed was the source of the MDR NTS. Indeed feedshave been the main source of MDR NTS infection out-breaks among poultry from other industrialised countriesas well [18]. In addition, several studies from industrialised

FEMSIM 1410 24-6-02

S. Kariuki et al. / FEMS Immunology and Medical Microbiology 33 (2002) 165^171 169

countries have also shown that NTS from farm animalsare multiply resistant to commonly available drugs andlately to both quinolones and the extended spectrum L-lac-tams [12,19^21].Using dendrogram analysis of XbaI-digested genomic

DNA separated by PFGE we observed that there werecertain dominant PFGE groups among the NTS from hu-mans: three types for S. Typhimurium and two types forS. Enteritidis, indicating that a few dominant strain typeswere largely responsible for infection in humans. However,even for NTS serovars such as S. Typhimurium andS. Enteritidis that were common to humans and animals,there were major di¡erences in the PFGE patterns as torender them separate genotypes. These NTS serovars fromhumans were clearly unrelated to corresponding serovarsfrom animal sources. We also observed that a single cloneof S. Agona was in circulation among pig herds, chickensand zero-grazed cows. In addition, S. Braenderup isolatesfrom both chickens and cows produced indistinguishablePFGE patterns, indicating that these strains were likely tobe from common sources. Comparing S. Braenderup andS. Saintpaul strains from humans and those from animalsand environmental sources, there was no signi¢cant rela-tionship between them, again suggesting that the sourcesof infection for these strains were di¡erent. However,S. Agona isolated from two farm workers and S. Agonafrom animals had indistinguishable PFGE patterns, indi-cating that these two workers had possibly acquired theinfection from handling the animals on the farm. Apartfrom this case of S. Agona, the majority of NTS, includingS. Typhimurium and S. Enteritidis from animals, wereclearly di¡erent from NTS strains isolated from humans.In agreement with our ¢ndings, other recent studies that

were based on a mathematical model for serovar distribu-tion [22] and PFGE genomic DNA typing [20] also ob-served that there was no signi¢cant evidence that NTSfrom pigs and cattle, and raw animal products includingmeat and poultry were sources of infection for NTS foundin humans in parts of the USA, suggesting that animalreservoirs may not always represent a source for humanNTS infection outbreaks. In addition, using several geno-typing methods including genomic DNA, plasmid pro¢l-ing and ribotyping, a study of S. Enteritidis from eightpoultry farms from across England found that a total of54 di¡erent strain types may be in circulation on thefarms. In contrast, using PFGE-RFLP typing, Murphyet al. [14] found in Ireland that S. Typhimurium DT104isolates from both humans and animal sources were pre-dominantly belonging to one strain.From the results of antimicrobial susceptibility testing,

plasmid and genomic DNA typing, it appears that NTSfrom animal and environmental sources are not closelyrelated to NTS isolated from humans living close to theseanimals, and therefore are unlikely to originate from com-mon sources. We further observed that NTS serovars fromanimals and environmental sources from homes of index

cases were unique antibiotic-sensitive single strain typesthat remained stable over the study period. Only the twolarge commercial farms had the same strain types of NTSfrom chickens that were MDR and these may have origi-nated from imported feed supplements. However, as im-ports of protein feed supplements for animals, particularlychickens, are becoming more common, surveillance studiesneed to be implemented so as to detect NTS that may be asource of infection and spread of antimicrobial resistanceamong the animals from small-scale farmers. In addition,the MDR phenotype in humans needs to be monitoredwith a view to ¢nding ways of implementing a prudentantibiotic usage policy in order to reduce resistance, par-ticularly to commonly available drugs. Further studies willbe required to investigate whether patient-to-patient trans-mission of NTS plays any role in the epidemiology of NTSinfections in Kenya.

Acknowledgements

We thank the Director of the Kenya Medical ResearchInstitute for permission to publish this work. S.K. wassupported by the Wellcome Trust Research DevelopmentAward in Tropical Medicine.

References

[1] Ahmed, R., Soule, G., Demczuk, W.H., Clark, C., Khakhria, R.,Ratnam, S., Marshall, S., Ng, L., Woodward, D.L., Johnson,W.M. and Rodgers, F.G. (2000) Epidemiologic typing of Salmonellaenterica serotype Enteritidis in a Canada-wide outbreak of gastro-enteritis due to contaminated cheese. J. Clin. Microbiol. 38, 2403^2406.

[2] Cody, S.H., Abbott, S.L., Mar¢n, A.A., Schulz, B., Wagner, P.,Robbins, K., Mohle-Boetani, J.C. and Vugia, D.J. (1999) Two out-breaks of multidrug-resistant Salmonella serotype typhimuriumDT104 infections linked to raw-milk cheese in Northern California.J. Am. Med. Assoc. 281, 1805^1810.

[3] Threlfall, E.J. (2000) Epidemic Salmonella typhimurium DT 104 ^ atruly international multiresistant clone. J. Antimicrob. Chemother.46, 7^10.

[4] Graham, S.M., Molyneux, E.M., Walsh, A.L., Cheesbrough, J.S.,Molyneux, M.E. and Hart, C.A. (2000) Nontyphoidal Salmonellainfections of children in tropical Africa. Pediatr. Infect. Dis. J. 19,1189^1196.

[5] Kariuki, S., Gilks, C., Corkill, J., Kimari, J., Benea, A., Waiyaki, P.and Hart, C.A. (1996) Multi-drug resistant non-typhi Salmonella inKenya. J. Antimicrob. Chemother. 38, 425^434.

[6] Kariuki, S., Oundo, J.O., Muyodi, J., Lowe, B., Threlfall, E.J. andHart, C.A. (2000) Genotypes of multidrug-resistant Salmonella ente-rica serotype Typhimurium from two regions of Kenya. FEMS Im-mun. Med. Microbiol. 29, 9^13.

[7] Mead, P., Slutsker, S.L., Dietz, V., McCaig, L.F., Bresee, J.S., Sha-piro, C., Gri⁄n, P.M. and Tauxe, R.V. (1999) Food-related illnessand death in the United States. Emerg. Infect. Dis. 5, 607^625.

[8] Indar-Harrinauth, L., Daniels, N., Prabhakar, P., Brown, C., Baccus-Taylor, G., Comissiong, E. and Hospedales, J. (2001) Emergence ofSalmonella enteritidis phage type 4 in the Caribbean: case-controlstudy in Trinidad and Tobago, West Indies. Clin. Infect. Dis. 32,890^896.

FEMSIM 1410 24-6-02

S. Kariuki et al. / FEMS Immunology and Medical Microbiology 33 (2002) 165^171170

[9] Liebena, E., Garcia-Migura, L., Breslin, M.F., Davies, R.H. andWoodward, M.J. (2001) Diversity of strains of Salmonella entericaserotype Enteritidis from English poultry farms assessed by multiplegenetic ¢ngerprinting. J. Clin. Microbiol. 39, 154^561.

[10] Palmer, S., Parry, S., Perry, D., Smith, R., Evans, M., Nehaul, L.,Roberts, R., Walapu, M. and Wright, D. (2000) The role of out-breaks in developing food safety policy: population based surveil-lance of salmonella outbreaks in Wales 1986^98. Epidemiol. Infect.125, 467^472.

[11] Murakami, K., Horikawa, K., Ito, T. and Otsuki, K. (2001) Envi-ronmental survey of salmonella and comparison of genotypic char-acter with human isolates in Western Japan. Epidemiol. Infect. 126,159^171.

[12] Fey, P.D., Safranek, T.J., Rupp, M.E., Dunne, E.F., Ribot, E., Iwen,P.C., Bradford, P.A., Angulo, F.J. and Hinrichs, S.H. (2000) Cef-triaxone-resistant salmonella infection acquired by a child from cat-tle. New Engl. J. Med. 342, 11242^11249.

[13] Glynn, M.K., Bopp, C., Dewitt, M.S.W., Dabney, P., Mokhtar, M.and Angulo, F.J. (1998) Emergence of multidrug-resistant Salmonellaenterica serotype Typhimurium DT 104 infections in the UnitedStates. New Engl. J. Med. 338, 1333^1338.

[14] Murphy, T.M., McNamara, E., Hill, M., Roony, N., Barry, J., Egan,J., O’Connell, A.J., O’Loughlin, J. and McFaddyen, S. (2001) Epi-demiological studies of human and animal Salmonella typhimuriumDT104 and DT104b isolates in Ireland. Epidemiol. Infect. 126, 3^9.

[15] National Committee for Clinical Laboratory Standards (1997) Meth-ods for Dilution Antimicrobial Susceptibility Tests for Bacteria thatGrow Aerobically: Approved Standard, 4th edn. National Commit-tee for Clinical Laboratory Standards document M7-A4, Wayne, PA.

[16] Tenover, F.C., Arbeit, R.D., Goering, R.V., Mickelsen, P.A., Mur-ray, B.E., Persing, D.H. and Swaminathan, B. (1995) Interpretingchromosomal DNA restriction patterns produced by pulsed-¢eld gelelectrophoresis : criteria for bacterial strain typing. J. Clin. Microbiol.33, 2233^2239.

[17] Walia, S.K., Madhavan, T., Chagh, T.D. and Sharma, K.B. (1987)Characterization of self-transmissible plasmids determining lactosefermentation and multiple antibiotic resistance in clinical strains ofKlebsiella pneumoniae. Eur. J. Clin. Microbiol. Infect. Dis. 7, 279^284.

[18] Shirota, K., Katoh, H., Murase, T., Ito, T. and Otsuki, K. (2001)Monitoring of layer feed and eggs for Salmonella in eastern Japanbetween 1993 and 1998. J. Food Prot. 64, 734^737.

[19] Cruchaga, S., Echeita, A., Aladuena, A., Garcia-Pena, J., Frias, N.and Usera, M.A. (2001) Antimicrobial resistance in salmonellae fromhumans, food and animals in Spain in 1998. J. Antimicrob. Chemo-ther. 47, 315^321.

[20] Winokur, P.L., Brueggemann, A., DeSalvo, D.L., Ho¡mann, L., Ap-ley, M.D., Uhlenhopp, E.K., Pfaller, M.A. and Doern, G.V. (2000)Animal and human multidrug-resistant, cephalosporin-resistant Sal-monella isolates expressing a plasmid-mediated CMY-2 AmpC beta-lactamase. Antmicrob. Agents Chemother. 44, 2777^2783.

[21] Wiu¡, C., Madsen, M., Baggesen, D.L. and Aaestrup, F.M. (2000)Quinolone resistance among Salmonella enterica from cattle, broilers,and swine in Denmark. Microb. Drug Resist. 6, 11^17.

[22] Sarwari, A.R., Magder, L.S., Levine, P., McNamara, A.M., Knower,S., Armstrong, G.L., Etzel, R., Hollingsworth, J. and Morris Jr., J.G.(2001) Serotype distribution of Salmonella isolates from food animalsafter slaughter di¡ers from that of isolates found in humans. J. Infect.Dis. 183, 1295^1299.

FEMSIM 1410 24-6-02

S. Kariuki et al. / FEMS Immunology and Medical Microbiology 33 (2002) 165^171 171