Heteroduplex analysis of Salmonella virulence plasmids and their prevalence in isolates of defined sources

Microbial Pathogenesis 1991 ; 11 : 391-397

Heteroduplex analysis of Salmonella virulence plasmids and their prevalence in isolates of defined sources Maria A . Montenegro,' Giovanna MoreIIi 2 and Reiner Helmuth` 1 Robert von Ostertag Institute, Bundesgesundeitsamt, Thielallee 88-92, D-1000 Berlin 33, Germany and 2 Max Planck Institute for Molecular Genetics, lhnestr . 63, D- 1000 Berlin 33, Germany (Received May 27, 1991 ; accepted in revised form August 7, 1991)

Montenegro, M . A . (Robert von Ostertag Institute, Bundesgesundeitsamt, Thielallee 88-92, D1000 Berlin 33, Germany and Max Planck Institute for Molecular Genetics, lhnestr . 63, D-1000 Berlin 33, Germany), G . Morelli and R . Helmuth . Heteroduplex analysis of Salmonella virulence plasmids and their prevalence in isolates of defined sources . Microbial Pathogenesis 1991 ; 11 : 391-397 . Strains of the Salmonella serovars S . typhimurium, S . enteritidis, S. dublin, and S . choleraesuis harbour large plasmids which are required for extraintestinal colonization after oral infection of mice . Electron microscopic heteroduplex analysis showed that these virulence plasmids share large regions of homology . Nine hundred and eighty-six isolates of different origins were analysed for the presence of these plasmids by using a cloned fragment of a S. cho/eraesuis virulence plasmid as a gene probe . Virulence plasmids were detected in nearly 100% of strains isolated from animal organs or human blood . Frequencies of detection ranged from 48 to 87% in strains of faecal, food or environmental origin . These results suggest that Salmonella virulence plasmids are required for systemic infections in humans and livestock . Key words: Salmonella virulence plasmids ; heteroduplex analysis ; Salmonella gene probe ; plasmid epidemiology .

Introduction The virulence of certain Salmonella serotypes for mice is dependent on the possession -3 of virulence plasmids .' These serotypes include those most commonly isolated from human and animal diseases, namely S. typhimurium and S. enteritidis, as well as the host specific serotypes S. dublin, S . gallinarum pullorum and S . choleraesuis.1

4,5

It seems that the virulence plasmids are important for the efficient colonization of spleen, liver and mesenteric lymph nodes after oral inoculation of mice,"' and reduce the lethal dose required .' They are not involved in the ability of Salmonella strains to cross the gut wall .' The exact mechanism by which the virulence plasmids contribute to systemic infection remains to be elucidated . Genetic and physical analysis of these virulence plasmids has led to the identification of genes specifically involved in virulence . For those plasmids best studied from S . typhimurium, S . dublin and S . choleraesuis, a consensus region of 8 kb is essential for

Author to whom correspondence should be addressed . 0882-4010/91 /120391 +07 503 .00/0

© 1991 Academic Press Limited



Salmonella virulence plasmids

392

\_ 0

I V

A A

S . typhimurium pR028 pRQ29 S . enteritidis

I S . typhimurium pRQ28 pRQ30 S .dublin

S. typhimurium pRQ28 S . choleraesuis pRQ20

S . enteritidis S . dublin



I S.

enteritidis pRQ29 S. choleraesuis pRQ20

A

A '

pRQ29 pR030

pR030 LV l S .dublin S .choleroesuis pRQ20 10 kb

Fig . 1 . Schematic representation of all possible heteroduplices formed between the four Salmonella virulence plasmids pRQ20, pRQ28, pRQ29, and pRQ30 .' The arrows indicate the position of the internal repeat used to orientate the molecules . Parallel lines show regions of homology .

full expression of virulence .3 .8 •s Four virulence genes have been mapped on this region . 10• 1 1 The different virulence plasmids share large regions of DNA homology, as based on DNA fingerprinting 12 and DNA hybridization studies . 13 "' Here we present quantitative measurements of the region of homology based on electron microscopic heteroduplex analyses . Despite the convincing data presented from mouse studies, the relevance of Salmonella virulence plasmids for livestock and humans remains undecided . We have used an epidemiological approach to investigate the importance of plasmids for human and animal disease . The occurrence of virulence plasmids in strains of S . typhimurium and S . enteritidis isolated from different animal, human or environmental sources was tested with the help of colony hybridization against a cloned fragment from within the virulence region . Results Electron microscopic heteroduplex analysis of Salmonella virulence plasmids Electron microscopic heteroduplex analyses were performed between the S . typhimurium pRQ28, S . enteritidis pRQ29, S . dublin pRQ30, and S . choleraesuis pRQ20 plasmids . The resolution of the method used was 50 bp . Figure 1 summarizes all heteroduplices observed . A typical stem loop structure caused by an internal 200 by repeat was common to all four virulence plasmids and allowed the alignment of the heteroduplices (arrow in Fig . 1 ) . Although the S . typhimurium plasmid pRQ28 is 23 MDa larger than the S . enteritidis plasmid, 99% of the smaller plasmid is homologous to the larger one with only 530 by of mismatch . 92% of the S . choleraesuis pRQ20 plasmid was homologous to pRQ28 . Furthermore, 84% of pRQ20 annealed to pRQ29 . The lowest degree of homology was



M . A . Montenegro et al.

393

found between the S . dublin plasmid pRQ30 and the other plasmids . Only 42% hybridized with pRQ28 plasmid and 35% with both other plasmids . Distribution of virulence plasmids Despite the importance of Salmonella virulence plasmids for systemic infection of mice,' 3,1 only a few reports on the infection of livestock animals have been published ."" We postulated that if these plasmids are important for infection of livestock and humans, then they might be more common in clinical than in environmental isolates . Strains were chosen for testing which were isolated from diverse animal or human organs including blood, from faeces of sick humans and asymptomatic Salmonella excretors, food and the environment . A gene probe specific for the virulence plasmids was obtained by cloning a 3 .6 kb Hind III fragment from the virulence region of the S . choleraesuis plasmid pRQ20 to generate plasmid pRQ51 . pRQ51 increased the frequency of liver colonization for orally infected mice of the plasmidless S . typhimurium strain 955/81 1 by 1000-fold . The cloned fragment shows the same size and restriction sites as probes derived from S. typhimurium plasmid . 161 It also hybridizes with the 3 .6 kb Hind III fragment of the recombinant plasmid pYA420 .' $ Reference strains were screened by colony hybridization with pRQ51 (Table 1 ) . Only strains belonging to serovars known to harbour virulence plasmids, e .g . S. typhimurium, S . enteritidis, S. dublin, S. choleraesuis and S. gallinarum pullorum, hybridized to the cloned fragment . Hybridization was not found against Salmonella of 29 other serovars (Table 1) nor against 200 isolates of eight other species of Enterobacteriaceae . The probe also did not hybridize with 40 Aeromonas spp ., 20 Campylobacter spp ., 6 Pseudomonas spp ., 22 Listeria spp ., and 15 Staphylococcus aureus isolates under relaxed conditions . On average, 76% of S . typhimurium and 86% of S . enteritidis strains analysed hybridized with the probe (Table 2) . Eighty-five selected strains were further analysed for the presence of plasmids by gel electrophoresis and subsequent Southern blot . The probe always hybridized to the virulence plasmids and not to the bacterial chromosome, confirming that a positive hybridization reaction correlates with the presence of an autonomous virulence plasmid . The data show that virtually every strain isolated from an extraintestinal source possesses a virulence plasmid . In contrast, significantly lower frequencies were found in S . typhimurium strains from faeces or from food and the environment . Similarly, significantly lower frequencies of plasmid carrying strains were detected in S . enteritidis strains from faeces than from extraintestinal sources . Discussion Several authors have demonstrated the close relationship between Salmonella virulence plasmids .' 131416 The results presented here allow the first quantitative estimate of this homology . Heteroduplex analysis was performed using all possible hybrids between virulence plasmids from S. typhimurium, S . enteritidis, S . dublin and S. cho/eraesuis . Ninety-nine per cent homology was found between the S . typhimurium and S . enteritidis plasmids, suggesting a common ancestor . The S . enteritidis plasmid could have been generated by deletion(s) of the S . typhimurium plasmid or the S. typhimurium plasmid could have evolved by DNA integration into the smaller S . enteritidis plasmid . The S. choleraesuis plasmid pRQ20 could also have been derived from the S . typhimurium plasmid, since 92% of its sequences were found to be homologous to the S. typhimurium plasmid .



39 4

Salmonella virulence plasmids

Table ll Colony hybridization of various Salmonella species to a virulence plasmid derived gene probe Hybridization Species

Serogroup

(Number)

S . agona S . derby S . heidelberg S . typhimurium variatio copenhagen

B B B B B

16 12 13 61 6

S . blockley S . bovis morbificans S . braenderup S . chester S . choleraesuis S . hadar S . infantis S . Livingstone S . munchen S . newport S . oranienburg S . potsdam S . reading S . tennessee S . thompson

C C C C C C C C C C C C C C C

6 6 6 2 16 10 26 6 6 10 6 1 2 6 6

S . dublin S . enteritidis S . gallinarum pullorum S . panama

D D

54 45 14 58

S. S. S. S.

E

anatum give newington Orion

D

D

+

46 6

16 12 13 15" 6 6 6 2

16 10 26 6 6 10 6 1 2 6 6 53 42 14

a

1 3' 58

E

12 2 2 2

S . london S . senftenberg

E E

12 14

12 14

S . cerro S. cubana S. worthington

G G G

6 6 5

6 6 5

S . java

J

10

10

12

12

Arizonae a

E E

12 2 2 2

DNA isolation confirmed the absence of plasmids .

Part of the homologous regions seems to represent common replication and partition functions, because all four plasmids are known to belong to the same incompatibility group .' 1,21 In this respect it is interesting to note that these virulence plasmids are extremely stable; they are not lost by storage of more than 30 years or by frequent subculture in different environments (Helmuth et al. unpublished data) . We also used an epidemiological approach to study the role of virulence plasmids for Salmonella infections of humans and livestock . We found a strong correlation between the ability of strains to cause extraintestinal infections and plasmid carriage ; virtually all strains from extraintestinal sources possessed a virulence plasmid . This observation indicates that the presence of the virulence plasmids is a prerequisite, not only in mice but also in livestock and humans, for efficient colonization of deeper tissues . The few strains (2%) from organs that failed to hybridize and did not carry virulence plasmids, all stem from animals and might represent secondary contamination of the organs after slaughter . Alternatively, they might carry different virulence genes



M . A . Montenegro

395

et al.

Table 2 Frequencies of virulence plasmid sequences according to the source of the isolates Serotype S. typhimurium

S . enteritidis

Sources all sources Faeces Excretors Organs Food Environment all sources Faeces Excretors Organs

Total No . of strains

derived

No . of strains that hybridized (%)

610 275 33 210 55 37

462 187 16 206 31 22

(76) (68) (48) b (98)° (56) (59)

376 230 44 102

324 (86) 199(87) 25 (57)` 100 (98)`

For definition of sources see Materials and methods . as compared to faeces . `P < 0 .001 as compared with the other sources . a

bP < 0 .05

not homologous with the probe we used . However, none of these strains was invasive in the mouse model . Plasmid carrying strains were also found in specimens from faeces of humans and animals, food and from the environmental strains but at significantly lower frequencies . Thus, the virulence plasmid is not a prerequisite for efficient colonization of these habitats . The lowest frequencies of plasmid carriage were found in strains isolated from asymptomatic carriers . Nevertheless, some plasmid carrying strains were found among the isolates from this source . It is possible that some asymptomatic carriers represent hosts infected with strains carrying virulence plasmids and that these hosts excrete bacteria without clinical symptoms due to a balance between the infecting strain and the immune system . Other healthy excretors might represent colonization by bacteria lacking the virulence plasmid or which are of reduced virulence for other reasons . These healthy excretors should not represent as large a danger for infection of other humans . We conclude that a differentiation of Salmonella strains according to virulence plasmid content is desirable in order to assess the risks and gravity of an infection .

Materials and methods Bacterial strains. S . typhimurium and S. enteritidis isolates originated from the institutes given in the acknowledgment . The majority of the animal strains were of cattle or porcine origin . Depending on the isolation specimen and the symptoms provoked, they were grouped into five categories : faeces-isolates originating from faecal samples of animals or people showing any symptoms of an infection ; excretors-isolates originating from faecal samples of animals or humans showing no symptoms of the infection and which were solely detected on the basis of routine surveillance ; organsisolates originating from internal organs, including blood of animals or people with or without symptoms of the infection ; food-isolates originating from ingredients or food prepared for human consumption ; environment-isolates originating from any other location than already stated, the majority of them originating from water . Plasmid DNA isolation . The plasmid content of selected strains was analysed by agarose gel electrophoresis after isolating plasmid DNA according to Kado and Liu ." Purification of plasmids for electron microscopy was accomplished by a modification of the method of Portnoy et al."' Electron microscopic heteroduplex analysis . The electron microscopic heteroduplex analysis was performed according to Morelli et al. 23



396

Salmonella virulence plasmids

Preparation of a virulence plasmid specific gene probe . A 3 .6 kb Hind I I I fragment common to the virulence plasmids from S. typhimurium, S. enteritidis, S. dublin and S . cho/eraesuis was cloned from the S . choleraesuis plasmid pRQ20' into the plasmid vector pKT231 24 by standard techniques . 25 The resulting recombinant plasmid was named pRQ51 . The Hind III insert was purified from agarose gels by electroelution and used as a gene probe in hybridization experiments, after labelling it with 32 P using a nick translation kit from Amersham Corporation . Preparation of filters for colony hybridization . Strains to be analysed were grown in L broth' in microdilution plates at 37°C overnight and transferred to L agar plates with a multipoint inoculator . After overnight incubation at 37°C, grown colonies were further transferred to Schwarzband filters (No . 586, Schleicher & Schull, Inc .) . The filters were processed by the method of Maas ." Each filter always included a negative and a positive control strain . Hybridization conditions . Al hybridizations against Salmonella isolates were performed under stringent conditions . Filters were prehybridized at 42°C for at least 2 h and hybridized overnight at the same temperature in solutions containing 50% formamide, as described previously ." When relaxed hybridization conditions were required, the hybridization solution contained 25% formamide . After washing the filters, they were exposed to X-ray films on intensifying screens . Infection experiments. The virulence properties of selected strains were assayed after oral infection of mice as described by Helmuth et al .' Statistical analysis . The significance of differences between the frequencies of plasmidcarriers detected among the strains of various sources was calculated using one-sided Fisher's exact test . P < 0 .05 was regarded as statistically significant .

We thank 0 . Pietzsch and R . Stephan from the Bundesgesundeitsamt and T . Richter from the Landesmedizinaluntersuchungsamt for providing Salmonella isolates . The expert technical assistance of Cornelia Bunge and Antje Steinbeck is gratefully appreciated . We also thank Ella Gi ndel for her support of this work and Mark Achtman for critical reading of the manuscript .

References 1 . Helmuth R, Stephan R, Bunge C, Hoog B, Steinbeck A, Bulling E . Epidemiology of virulence-associated plasmids and outer membrane protein patterns within seven common Salmonella serotypes . Infect Immun 1985 ; 48 :175-82 . 2 . Terakado N, Sekizaki T, Hashimoto K, Naitoh S . Correlation between the presence of a fifty-megadalton plasmid in Salmonella dub/in and virulence for mice . Infect Immun 1983 ; 41 : 443-4 . 3 . Gulig PA . Virulence plasmids of Salmonella typhimurium and other Salmonellae . Microb Pathogen 1990; 8 : 3-11 . 4 . Barrow PA, Simpson JM, Lovell MA, Binns MM . Contribution of Salmonella ga/linarum large plasmid toward virulence in fowl typhoid . Infect Immun 1987 ; 55 : 388-92 . 5 . Barrow PA, Lovell MA . The association between a large molecular mass plasmid and virulence in a strain of Salmonella pullorum . J Gen Microbiol 1988 ; 134 : 2307-16 . 6 . Pardon P, Popoff MY, Coynault C, Marly J, Miras I . Virulence-associated plasmids of Salmonella serotype typhimurium in experimental murine infection . Ann Inst Pasteur Microbiol 1986 ; 137(B) : 4760. 7 . Manning EJ, Baird GD, Jones PW . The role of plasmid genes in the pathogenicity of Salmonella dublin . J Med Microbiol 1986; 21 : 239-43 . 8 . Williamson CM, Pullinger GD, Lax AJ . Identification of an essential virulence region on Salmonella plasmids. Microb Pathogen 1988 ; 5 : 469-73 . 9 . Kawahara K, Tsuchimoto M, Sudo K, Terakado N, Danbara H . Identification and mapping of mba regions of the Salmonella choleraesuis virulence plasmid pKDSC50 responsible for mouse bacteremia . Microb Pathogen 1990 ; 8 : 13-21 . 10 . Taira S, Rhen M . Molecular organization of genes constituting the virulence determinant on the Salmonella typhimurium 96 kilobase pair plasmid . FEBS Lett 1989 ; 257 : 274-8 . 11 . Matsui H, Abe A, Kawahara K, Terakado N, Danbara H . Positive regulator for the expression of Mba protein of the virulence plasmid, pKDSC50, of Salmonella choleraesuis . Microb Pathogen 1991 ; 10 : 459-64 . 12 . Williamson CM, Baird GD, Manning EJ . A common virulence region on plasmids from eleven serotypes of Salmonella . J Gen Microbiol 1988 ; 134 : 975-82 . 13 . Popoff MY, Miras I, Coynault C, Lasselin C, Pardon P . Molecular relationships between virulence



M . A . Montenegro et al.

397

plasmids of Salmonella serotypes typhimurium and dublin and large plasmids of other Salmonella serotypes . Ann Inst Pasteur Microbiol 1984 ; 135(A) : 389-98 . 14 . Korpela K, Ranki M, Sukupolvi S, Makela PH, Rhen M . Occurrence of Salmonella typhimurium virulence plasmid-specific sequences in different serovars of Salmonella . FEMS Microbiol Lett 1989; 58 : 49-54 . 15 . Jones PW, Collins P, Lax A . The role of large plasmids in the pathogenesis of Salmonellosis in cattle . J Med Microbiol 1988; 27 : X . 16 . Woodward MJ, McLaren l, Wray C . Distribution of virulence plasmids within Salmonellae . J Gen Microbiol 1989 ; 135 : 503-11 . 17 . Poppe C, Curtiss III R, Gulig PA, Gyles CL . Hybridization studies with a DNA probe derived from the virulence region of the 60 Mdal plasmid of Salmonella typhimurium. Can J Vet Res 1989; 53 : 378-84 . 18 . Gulig PA, Curtiss III R . Cloning and transposon insertion mutagenesis of virulence genes of the 100kilobase plasmid of Salmonella typhimurium . Infect Immun 1988; 56 : 3262-71 . 19 . Ou JT, Baron LS, Dai X, Life CA . The virulence plasmids of Salmonella serovars typhimurium, choleraesuis, dublin, and enteritidis, and the cryptic plasmids of Salmonella serovars Copenhagen and sendai belong to the same incompatibility group, but not those of Salmonella serovars durban, gallinarum, give, infantis and pullorum . Microb Pathogen 1990 ; 8 :101-7 . 20 . Cerin H, Hackett J . Molecular cloning and analysis of the incompatibility and partition functions of the virulence plasmid of Salmonella typhimurium . Microb Pathogen 1989 ; 7 : 85-99 . 21 . Kado Cl, Liu ST. Rapid procedure for detection and isolation of large and small plasmids . J Bact 1981 ; 145 :1365-73 . 22 . Portnoy DA, Moseley SL, Falkow S . Characterization of plasmids and plasmid-associated determinants of Yersinia enterocolitica pathogenesis . Infect Immun 1981 ; 31 : 775-82 . 23 . Morelli G, Montenegro MA, Hillenbrandt G, Scherzinger E, Trautner TA . The genome of B . subtilis phage SPP1 : Assignment of 5'-3'-polarity to the complementary strands of SPP1, DNA . Mol Gen Genet 1978 ; 164 : 93-7 . 24 . Bagdasarian M, Lurz R, Ruckert B et al. Specific plasmid cloning vectors . II . Broad host range, high copy number, RSF 1010-derived vectors, and a host-vector system for gene cloning in Pseudomonas . Gene 1981 ; 16 : 237-47 . 25 . Maniatis T, Fritsch EF, Sambrook J . Molecular cloning : a laboratory manual, Cold Spring Harbor : Cold Spring Harbor Laboratory Press, 1982; 11-15 . 26 . Maas R . An improved colony hybridization method with significantly increased sensitivity for detection of single copy genes . Plasmid 1983 ; 10 : 296-8 . 27 . Montenegro MA, Boulnois GJ, Timmis KN . Molecular epidemiology by colony hybridization using cloned genes . In : Pi hler A, Timmis KN, eds . Advanced molecular genetics . Berlin : Springer-Verlag, 1984 ; 92-103 .