Zbl. Bah. 277, 10-21 (1992) © Gustav Fischer Verlag , Stut tgart/New York
Characterization of the Virulence Regions in the Plasmids of Three Live Salmonella Vaccines
WOLFGANG B E Y ER a nd LUT Z GEUE Institut fiir Epizootiologie und T ierseuchen bekam pfung, Abteilung fiir Mo leku lare Epizootiologie, 0- 1903 Wusterhausen, Germany
With 4 Figures ' Received July 8, 1991 . Revision received Novem ber 12, 1991 . Accepted December 20, 1991
Summary Three live Salmonella vaccines, Zoosaloral'P " Dessau" , Bovisaloral 'f " Dessau" , and Suisaloralv "Dessau" successfully used in veterinary medicine in eastern Germany were analy sed for thei r plasmid DNA content. Plasmi ds of 60 MDa (5. typhimurium), 55 M Da and 26 MDa (S. dub lin) and 34 MDa (S. choleraesuis) were found . The three large plasmids con ta ined "a virulence region as shown by restriction enzyme ana lysis and hybridization studies with virulence-region-specific DNA probes. Restriction enzyme pattern ana lysis of th e who le plasmids revea led differences between vaccine strains and parental strains. We wanted to know if there were alte rations within the viru lence region of the vaccine strain plasmid s compared wi th parenta l strains cau sed by th e attenuation procedure. Therefore, the 6 kb-Cla I fragments of several plasmids were cloned and characterized by restriction enzyme fingerprinting. The exp ression of the CIa I frag men t-encoded proteins was ana lysed in the minicell-producing stra in DS 410. The experi ments revea led that the restriction pattern of the 6 kb-C la I fragments of Zoosaloral and Bovisaloral as well as the PAGE pattern of virulence-region-encoded proteins were in complete accordance with one ano ther, between vaccine strains and the parental strain of Zoosaloral, and with a wild-type strain of S. typ him urium.
Zusammenfassung Die drei Salmonella-Lebendvakzinen, Zoosaloralv "Dessau" (S. typhimurium), Bovisalora l® "Dessau" (S. dublin ) and Suisaloralf "Dessau" (S. cboleraesuis), die mit gutem Erfolg in der Veterinarmedizin auf dem Gebiet der neuen Bundeslander eingesetzt wurden, ent halten Plasmide von 60 MDa (S. typhimurium), 55 MDa und 26 MDa (S. dublin) bzw. 34 MDa (S. cho leraesuis). Auf den drei grolsen Plasmiden wu rde mittels Restriktionsenz ym-Analyse und Hybridisierung mit vir-spezifi schen DNA-Sonden eine Virulenzregion nachgewiesen. Die Restriktionsmusrer der vollstandigen Plasmide zeigen Unterschiede zwischen den Vakzine-Stamrnen und ihren Elrern stam men . Uns interessierte, ob es infolge der Attenuierung Vera nderungen inn erhalb der vir-Regionen der Vakzi ne-Plasmide im Vergleich zu den Eltern-Pl asmiden gibe. Dafiir w urden die ent sprechenden 6 kb -Cla I Fragmente miteinander
Virulence Plasmids in Live Salmonella Vaccines
11
verglichen. Die Expression der durch die CIa I Fragmente kodierten Proteine wurde im Minizellstamm DS 410 analysiert. Die Experimente ergaben, dag sowohl die Restriktionsmuster der 6 kb-Cla I Fragmente von Zoosaloral und Bovisaloral als auch die PAGE-Muster der durch die vir-Region kodierten Proteine untereinander, zwischen den Vakzine-Stamrnen und dem Eltern-Stamm von Zoosaloral sowie mit einem Wild-Stamm von S. typhimurium vollig iibereinstimmen.
Introduction Live vaccines are used as one of the most effective ways to induce a strong immunity against the action of pathogenic enteric bacteria including Salmonella. Being aware of the risks involved in the use of live bacteria, a list of requirements for strains used were layed down (1, 14, 26, 32). Keeping in mind our modern knowledge on virulence plasmids, the requirement of safety with regard to the presence of transmissible genetic virulence structures is of special interest. Although up to-day there has been no explantation of how a virulence plasmid of a Salmonella sp. acts to make its host virulent, the evidence strongly suggests that virulence plasmids are important in the pathogenesis of naturally occurring systemic Salmonella infections in animals. Roudier et al. (24) have shown that the plasmid is absolutely necessary but not sufficient to make wild-type Salmonella spp. virulent for mice. All the major Salmonella species that are highly hostadapted to animals carry a virulence plasmid including S. choleraesuis (13), S. dublin (2,6,31), S. gallinarum (4), S. pullorum (3), and S. abortus avis (22, 33). Moreover, virulence-encoding regions were proved to exist for the following serotypes: S. enteritidis, S. typhimurium, S. rostock, S. paratyphi c, S. moscow, S. blegdam (33), S. johannisburg, and S. kottbus (15). It is also likely that the plasmid contributes to the virulence of nontyphoid Salmonella species in human disease and may be responsible for the septicemic forms of salmonellosis (25). The absence of virulence plasmid sequences in certain extraintestinal isolates of Salmonella does not contradict the hypothesis that the plasmid contributes to virulence in human disease since other important factors influence the invasiveness of Salmonella including the LPS structure, motility, ability to express specific adherence/invasion proteins after attachment to epithelial cells, and probably specific virulence genes necessary for entry, multiplication and penetration into host cells. Otherwise, factors such as the inoculum size, the immunocompetence (10) and maybe the genetic resistance of the host are of importance for the fate of a Salmonella infection (for a review see ref. 8, 11). Over the last fifteen years, three live Salmonella vaccines were developed and applied in Veterinary Medicine with good results in eastern Germany (19 and communications presented at the meetings of the WHO Working Group on Salmonella Immunization in Animals held in Munich (1986), lena (1989), and lena (1991), respectively). The principle of these vaccines is the combination of two independently attenuating markers with no limitation on propagation (14,17,26). All three vaccine strains contain adenine as the first attenuating marker. Zoosaloral® "Dessau" (S. typhimurium) and Bovisaloralf "Dessau" (S. dublin) were made by using an S-form auxotrophic phenotype his- and thi" respectively, which may lead to a reduced survival within macrophages and/or cells of the reticulo-endothelial system (7). Suisaloralv "Dessau" (S. choleraesuis) was made from an R-form mutant. The auxotrophy mutations are accidental mutations induced by incubation of the parental strains with N-methyl-N'-
W. Beyer and L. Geue
12
nitro-nitrosoguanidine (17). The site(s) and the nature of these mutations have not been clarified beyond doubt. Therefore, alterations within the vir region of the virulence plasmids participating in the attenuation are not unlikely. We were interested in evaluating whether there were any alterations caused by the attenuation procedures within the virulence region of the virulence plasmids as compared to parental strains. We have been able to show that all three vaccine strains do contain a virulence plasmid with a restriction fragment pattern different from that of their parental strains. However, the restriction pattern of the 6 kb-Cla I fragments as well as the PAGE pattern of virulence-region-encoded proteins are in complete accordance with one another, between vaccine strains and the parental strain of Zoosaloral, and with a wild-type strain of S. typhimurium.
Material and Methods Bacterial strains and growth conditions. Bacterial strains, descriptions and sources are listed in Table 1. Cells were grown in L broth containing pancreatic peptone 17 gil, yeast extract 3 gil, NaCI 5 gil, glucose 1 gil, or on L broth containing 10 gil agar at 37°C. Ampicillin was used at 100 ug/ml. Plasmid preparations for cloning procedures. The DNA isolation of native Salmonella plasmids was carried out as described by Kado and Liu (11) with the following modifications. Briefly, bacteria of a 200 ml culture were centrifuged in four 50 ml portions, the pellets lysed in 25 ml of 50mM Tris base, 3% SDS, 290 f!V100 ml of 11.2 N NaOH,
Table 1. Bacterial strains Strain
Characterization
Source
E. coli (K12) TG1
/lilac pro); thi; str A; sup E; rec A; end A; sbcB; hsdR-; proAB; lacIq ; Z/lM15
Central Inst. of Mol. BioI. of the Acad. of Sci., Berlin-Buch
E. coli (Kll) DS410 (M2141)
minicell producing strain /lilac pro); min A; min B; rps L; thi; thr
Max Planck Institut Berlin
W 3839
S. typhimurium, virulent field isolate
Inst. of Exp. Epidemiology, Wernigerode
Zoosaloral'"
S. typhimurium, live vaccine strain (ade, his)
Commercially available
"Dessau"
Bovisaloralf "Dessau"
S. dublin, live vaccine strain (ade, thi, nad)
Commercially available
Suisaloralf "Dessau"
S. choleraesuis, live vaccine strain (ade, R)
Commercially available
Moscow strain
S. typhimurium, parental strain of Zoosaloral
Inst. of Veterinary Ecology, Microbiology and Immunol., Dessau
AF2423246
S. dublin, parental strain of Bovisaloral (nad)
Inst. of Medical Microbiol. of the F.-Schiller-University, jena
Virulence Plasmids in Live Salmonella Vaccines
13
incubated at 55°C for 2 h. Then the probes were pooled and 50 ml precooled 3M NaOAc, pH 4.8, added dropwise. After centrifugation (15000 g), the DNA was precipitated with 0.8 volumes of isopropanol, further purified with phenol/chloroform (50% v/v) extractions and then precipitated with ethanol. For cloning procedures, the DNA was further purified by centrifugation through a caesium chloride-ethidiurn bromide density gradient as described (18). For plasmid profiles, a small-scale procedure was routinely used. About 2 colonies of fresh bacteria from an agar plate were lysed in 200 ul lysis solution, incubated at 55°C for 1 h and extracted with 800 [!l phenol/chloroform (50% v/v). After centrifugation (11000 g), the supernatant was used directly for electrophoresis. Plasmids and their restriction products were routinely analysed on horizontal 0.6% (w/v) agarose gels run in Tris-acerate buffer overnight. These gels were used for Southern transfers to nitrocellulose membranes. Recombinant DNA techniques, construction of recombinant plasmids. Plasmid DNA preparations, restriction analysis, ligations and transformations were performed by standard procedures essentially as described in (18). Enzymes were from Amersham Buchler GmbH & Co KG, Braunschweig, agarose for DNA electrophoresis was from Serva. The recombinant plasmids pWh 170, 180, 190 and 200 were obtained by cloning the 6 kb Cia I fragments known to harbour the virulence region from the virulence plasmids of W3839, Zoosaloral, Bovisaloral, and "Moscow strain", respectively, in Cia l-cleaved pBR322. The recombinant plasmids were propagated in E. coli TGl. Preparation of the 3.9 k b Hind III DNA probe p Wh 70 for analysis of plasmid profiles and restriction maps. DNA of the virulence plasmid of S. typhimurium W3839 was digested with Hind III and DNA fragments were separated by gel electrophoresis. The 3.9 kb fragment was recovered by clcctroclurion, ligated into Hind III-cleaved vector pUC19 and cloned into E. coli TG1. The clone was named pWh70. According to the maps of Baird et al. (2) and Michiels et al. (20), this fragment should contain a virulence-determining region. This was proved to be true by restriction enzyme fingerprinting with Eco RI, Bam HI, Cia I, and Xho I and by sequencing some hundreds of the nucleotides in end position. The purified insert of pWh 70 was labelled by randomprimed incorporation of digoxigenin-labelled deoxyuridine triphosphate according to the recommendations of Boehringer Mannheim or with a- J2-PdNTP using the DNA-labelling kit from Serva. Hybridizations. Hybridization and detection procedures with the digoxygenin DNAlabelling kit were carried out according to the recommendations of the manufacturer (Boehringer). For the hybridization with a- 32PdNTP-labelled probes (Amersham), we used a prehybridization procedure with" x sse (0.45 M NaCl, 0.045 M Na citrate), 0.5% (w/v) skim milk, 0.1 % (w/v) SDS and 100 [!g/ml sonicated and boiled unspecific DNA (4-5 h) and hybridization procedure with 3 x SSC, 0.2% (w/v) skim milk, 0.1 % (w/v) SDS and 10% (w/ v) PEG 6000 (Serva) and up to 50 ng/ml denatured labelled probe, 15 h, 65°C. The filters were washed briefly in 2 x sse, 0.1 '/'o 50S, RT, then once in 2 x SSC, 0.1 % SOS, 65°C, 30 min, and two times in 0.2 x sse, 0.1 0;;, SDS, 65 DC, 30 min and autoradiographed with xray film and intensifying screens (ORWO). Analysis of plasmid-directed protein synthesis in minicells. The E. coli mini cell-producing strain DS410 was used to identify plasmid-encoded proteins. DS410 was transformed with plasmids pBR322, pWh170, 180, 190, and 200 by the calcium chloride procedure (18). Minicells were isolated and labelled with 35S-methionine (> 37 TBq/mmol, Amersham) at a concentration of 1.8 MBq/run essentially as described (28). Labelled minicells were solubilized in 5DS-PAGE sample buffer and the protein content was electrophoretically separated in 10% SDS-PA gels (16). The gels were fixed in 10% (v/v) acetic acid (30 min), impregnated with Amplify" (Amersham) (30 min), dehydrated in 90% (v/v) ethanol, 10% (v/v) glycerol (3 h) air-dried overnight and exposed to X-ray films (ORWO) (3-6 d). As molecular mass standard, the rainbow-coloured markers from Amersham, radiolabelled after electrophoresis, were used.
14
W. Beyer and L. Geue Results
Plasmid profiles In order to characterize the virulence plasm ids, the following plasmid profiles were established. From the live Salmonella vaccines, Zoosaloral - 60 MDa ; Bovisaloral- 55 MD a and 26 MDa, Suisaloral - 34 MDa and from the parental strains S. typhimurium "Moscow strain" - 60 MDa and S. dublin AF2423246 - 55 MDa and 26 MDa. In comparison , the S. typhimurium field strain W3839 - 60 MDa was coprepared. All plasm ids, with the exception of the 26 MDa plasmid of Bovisaloral hybridized with the probe pWh70. The probes pWh 70 and pWh 170 have been explained under Material and Methods and Fig. 3.
Restriction enzyme analysis and hybridization All plasmids were cleaved with restriction enzymes Hind III, Bam HI, Eco RI and Cia I and the gelelectrophoretical cleavage patterns were compared. These enzymes have
Fig. 1a. Hind III-parrern of the plasmids of Zoosa loral and its parental strain. Lane 1 - Hind III marker; lane 2 - W 3839 ; lane 3 - "Moscow strain"; lane 4 and 5 - Zoosaloral; lane 6Pvu II marker. In accordance with Michiels er al. (20), the eight fragments of "Moscow strain" have been designated H I to H8 in the text. The calculated size of all fragments is indicated.
Virulence Plasmids in Live Salmonella Vaccines
15
recognit ion sites within the conserved virulence region (9, 20, 21, 29 ). While the cleavage patterns with Bam HI and Cia I were identical for Zoosaloral and the "Moscow strain ", there were striking differences in the Hind III pattern (Fig.1a). While the "Moscow strain" showed a cleavage pattern of 8 fragments (H 1- H 8) (20) frequently occurring in 60 MDa virulence plasmids, the highly conserved 3.9 kb fragment H3 (20, 22 ) as well as the 16 kb fragment H3 (20) were absent in the vaccine strain Zoosaloral. The following bands hybridized with the probe pWh 70 after Southern transfer of the DNA to nitrocellulose (nc) membr anes: W3839 - 3.9 kb, " Moscow strain" - 3.9 kb (H6), Zoosaloral- 19.8 kb (H2, Fig. 2). Identical digestion patterns with Bam HI and Cia I but striking differences with the other two enzymes were found in comparison of Bovisaloral and its parental strain AF2324236. Fig. 1b shows the cleavage pattern for Eco RI and HindIII. After transfer of the fragments to nc-membranes, in both strains the 3.9 kb band of the Hind III digestion and identical band s of the Eco RI digestion hybridized with the probe
2
3 4
Fig. 1b. Hind IIlIEco RI-pane rn of the plasmids of Bovisal and its parental strain. Lane 1 Bovisaloral; lane 2 - AF2324236 (both Hind III cleaved); lane 3 - Bovisaloral; lane 4 AF2324236 (both Eco RI cleaved).
16
W.Beyer and L. Geue
1
2
3
4
19,8
3,9
Fig. 2. Hybridization pattern of the Hind III cleavedplasmids of: lane 1 - Bovisaloral;lane 2 - W 3839, lane 3 - Moscow strain; lane 4 - Zoosaloral
pWh70. The fragments of the CIa I digestion (not shown) became hybridized with the insert from clone pWh170. As after the CIa I digestion of W3839, Zoosaloral, and "Moscow strain", identical fragments of about 6 kb in size, hybridized. After cloning the 6 kb CIa I fragments from W3839 (pWh170), Zoosaloral (pWh180), Bovisaloral (pWh190), and "Moscow strain" (pWh200), the restriction sites for each fragment for the enzymes Eco RI, Pst I, Sac I, Kpn I, Hind III and Bam HI were investigated and compared with previously published restriction maps (2, 5, 9, 20, 21, 23, 29, 30). The restriction sites of all clones tested were identical with one another and with data given in literature. Fig. 3 shows the restriction map of pWh170 which is representative of all clones tested. After the digestion of Suisaloral with CIa I, the defined virulence region was contained within the largest about 16 kb fragment. This fragment was not further characterized. Expression of the virulence region
The recombinant plasmids pWh170, 180, 190, and 200 were introduced into the minicell-producing strain DS 410. Metabolic labelling with 35S-metwas used to analyse
Virulence Plasmids in Live Salmonella Vaccines
o
2
5
4
3
HB
7
6
E
C
E
C
9 kb
8
XB
17
H
I I
C
33 kO
EP
SP
I(
I(
II
mkaC
31 kO mkaB
II
KBSHHB
I( III I
pWh170
I
70 kO
1128 kol
mkaA
mkaO
Fig. 3. Restriction map of pWh70 and pWh170. Abbreviations used: C = Cia I, E = Eco R I, S = Sac I, P = Pst I, K = Kpn I, H = Hind III, B = Bam H I. The alignment of the four protein coding sequence regions was taken from Taira and Rhen (28).
6
54
3
2
97,4
6QO
46.0
30.0 Fig. 4. Fluorograms of 35S-methionine labelled E. coli minicells run in 10% SDS-PAGE. The plasmids contained are as follows: lane 1 - marker; lane 2 - pBR322; lane 3 - pWh170; lane 4 - pWh180; lane 5 - pWh200; Ian 6 - pWh190. Numbers on the right indicate the positions of molecular mass markers (in kDa). 2
Zbl. Bakt. 27711
18
W.Beyer and L. Geue
the expression of plasmid-encoded proteins. A common protein pattern of 70 kDa , 31 kDa , and 28 kDa was demonstrated for all clones tested in SDS-PAGE (Fig. 4).
Discussion Three live Salmonella vaccines which have been used successfully in veterinary medicine in eastern Germany pro ved to conta in virulence plasmids as shown by restriction enzyme analysis and hybridization with virulence-r egion-specific DNA probes. We were interested in evaluat ing whether there were any alter ation s caused by the attenuation pro cedures within the virulence region of these plasmids as comp ared to their parental stra ins. Therefore we comp ared the restriction maps of the whole vaccine strain plasmids and of the appropriate Cia I fragments of the vaccine str ain plasmids with one another, with their parental strains and with an S. typhimurium wild type strain as well as the PAGE pattern of virulence-region-encoded proteins. The Hind III pattern of Zoos aloral'P "De ssau " and its parent "Moscow stra in" were different (Fig. 1a). One of both Hind III sites of the otherwise conserved 3.9 kb fragment must have become mutated in Zoosaloral. It seems that the 19.8 kb band consists of two fragme nts, one is the same as in the " Moscow strain" and one results from the additi on of the 3.9 kb fragment and the missing 16 kb fragment (H3). Indeed, this band hybridized with the pro be pWh 70 (Fig. 2). According to the map of Gulig and Curtiss (9) the 3 '-Hind III site that means the site outside the CIa I vir-fragme nt must be the mut ated one (Fig. 3). Th is conclusion is supported by the concordance of the Hind III sites within the CIa I fragments of Zoosaloral and "Moscow strain" . There were differences in the whol e plasmid restri ction pattern of Bovisaloral'" "Dessau" and its parent AF2423246. However, when comp aring the 6 kb-Cla I fragments of Bovisaloral with those of Zo osaloral , "Moscow strain", and the S. typhimurium wild type plasmid, no differences occurred in the sites of all enzymes tested. So it was not necessar y to cha racterize the appro priate Cia I fragment of AF242324 6 in detail. M oreover, in all digestion s of Bovisaloral and AF2423246, ident ical fragments hybridized with our vir-region-specific DNA prob es. So we conclude th at the observed differences in the restriction sites of Eco RI and Hind III, one additional site in both cases in the AF2423246 strain (Fig. lb ), are localized outside the conserved Cia I virregion . In summary, there were no differen ces within the virulence regions defined by the CIa I fragments between the examined vaccine strains, the parental strain of Zoosaloral and the S. typhimurium wild type strain W3839 regarding the recognit ion sites of all enzymes tested. In order to compare the vir-regions on the protein level, the expre ssion of the Cia I fragment-encoded proteins was ana lysed in the minicell-producing strain DS 410. Protein pattern s completely ident ical with one another and in accordance with literature data were found with proteins of 70 kDa, 3 1 kDa, and 28 kDa, respectively. The fourth described prote in of 33 kDa (mka C, 30) could not be ident ified. Indeed, it seems to be difficult to identify all four proteins at the same time if the complete virregion is used. So Gulig and Curtiss (9) found the 3 1 kDa protein only in an in vitr o transcription/t ranslati on rest. They did not describe the 70 kDa protein. Nore! et al. (21) did not find the 31 kDa and the 33 kDa proteins using the same CIa I fragment in a
Virulence Plasmids in Live Salmonella Vaccines
19
maxicell expression system. The analysis of a subcloned Cia I - Eco RI fragment, coordinates 0 to 1.5 (Fig. 3), by Taira and Rhen (30) revealed two protein bands of 33 kDa and 30 kDa. Although the theoretical size of mka C should be 33.8 kDa (30), the 30 kDa band was more pronounced. Therefore it cannot be excluded that an identical protein band is masked by the 30 kDa protein ~-Iactamase of pBR322 in our experiments. Differences in the signal strength between lanes 3/5 and 4/6, respectively, do not reflect real differences in the expression level between the strains. There are rather problems with the minicell expression system in our hands, maybe in the procedure to generate viable minicells. We reached different values in several experiments for the same strain under equal conditions. In summary, also the analysis of the expression of the vir-region-encoded proteins by SDS-PAGE showed no differences between vaccine strains, the parental strain of Zoosaloral and the wild type strain W3839, respectively. Although it is not opportune to make an assertion considering the biological function of the vir-regions of the vaccine strains on the basis of our test criteria, the complete concordance of all features examined supports the hypothesis that no structural and with it, also no functional changes in the virulence-encoding region of the vaccine strains virulence plasm ids took place. So these plasmids might be fully intact with regard to their functions. Because of the unspecific action of N-methyl-N' -nitro-nitrosoguanidine on DNA, there might be several other mutations on the chromosomes and the plasmids, maybe also within other known virulence factors. This has not yet been investigated. At this time, the attenuation can only be ascribed to the auxotrophies and the alteration in LPS of Suisaloral, respectively. Remembering the requirements for live vaccines and the potential risks linked with the presence of virulence plasmids in Salmonella strains, future live Salmonella vaccines should not contain virulence plasmids unless they are essential for immunogenic properties. This should be tested in every case. Acknowledgements. The authors wish to thank the Institutes mentioned in Table 1 for providing the strains. We thank Mrs. Ch. Schnick and Mrs. M. Lembke for technical assistance and Dr. R. Hellmuth (Robert v. Ostertag-Institut, Berlin) for critical reading of the manuscript.
References 1. Anonymus: Oral enteric bacterial vaccines. WHO Techn. Rep. Ser. Nr. 500 (1972) 2. Baird, G. 0., E.]. Manning, and W. Jones: Evidencefor related virulence sequences in plasmids of Salmonella dublin and Salmonella typhimurium. J. Gen. Microbiol. 131 (1985) 1815-1823 3. Barrow, P. A. and M. A. Lovell: The association between a large molecular mass plasmid and virulence in a strain of Salmonella pullorum. J. Gen. Microbial. 134 (1988) 2307-2316 4. Barrow, P. A., ]. M. Simpson. M. A. Lovell. and M. M. Binns: Contribution of Salmonella gallinarum large plasmid toward virulence in fowl typhoid. Infect. Immun. 55 (1987) 388-392 5. Beninger, P. R., G. Chieami, K. Tanabe, C. Roudier, ]. Pierer, and D. G. Guiney: Physical and genetic mapping of the Salmonella dublin virulence plasmid pSDL2. Relationship to plasmids from other Salmonella strains. J. Clin. Invest. 81 (1988) 1341-1347
20
W. Beyer and L. Geue
6. Chikami, G. K., J. Fierer, and D. G. Guiney: Plasmid mediated virulence in Salmonella dublin demonstrated by use of a 5-oriT construct. Infect. Immun. 50 (1985) 420-424 7. Fields, P. I., R. U. Swanson, C. G. Haidaris, and F. Heffron: Mutants of Salmonella typhimurium that cannot survive within the macrophage are avirulent. Proc. Nat!. Acad. Sci. USA 83 (1986) 5189-5193 8. Finlay, B. B. and S. Falkow: Salmonella as an intracellular parasite. Malec. Microbial. 3 (1989) 1833-1841 9. Gulig, P. A. and R. Curtiss III: Cloning and transposon insertion mutagenesis of virulence genes of the 100-kilobase plasmid of Salmonella typhimurium. Infect. Immun. 56 (1988) 3262-3271 10. Hook, E. W.: Salmonellosis: certain factors influencing the interaction of Salmonella and the human host. Bull. N.Y. Acad. Med. 37 (1961) 499-512 11. Hsu, H. S.: Pathogenesis and immunity in murine salmonellosis. Microbiol. Rev. 53 (1989) 390-409 12. Kado, C. I. and S. T. Liu: Rapid procedure for detection and isolation of large and small plasmids. J. Bact. 145 (1981) 1365-1373 13. Kawahara, K., Y. Haraguchi, M. Tsuchimoto, N. Terakado, and H. Danbara: Evidence of correlation between 50-kilobase plasmid of Salmonella choleraesuis and its virulence. Microb. Pathogen. 4 (1988) 155-163 14. Koch, H.: Untersuchungen zur Prufung von bakteriellen Mutanten als potentielle Impfstarnme. Arch. expo Vet.-Med., Leipzig 34 (1980) 43-50 15. Korpela, K., M. Ranki, s. Sukupolvi, P. H. Miikelii, and M. Rhen: Occurrence of Salmonella typhimurium virulence plasmid-specific sequences in different serovars of Salmonella. FEMS Microbiol. Lett. 58 (1989) 49-54 16. Laemmli, U. K.: Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227 (1970) 680-685 17. Linde, K.: Herstellung von stabilen Salmonella-Impfstammen durch Kopplung von zwei unabhangig voneinander nichtvermehrungsbegrenzenden attenuierenden Markern. Arch. expo Vet.-Med. (Leipzig) 34 (1980) 19-32 18. Maniatis, T., E. F. Fritsch, and]. Sambrook: Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring HarbourlNY (1982) 19. Meyer, H., K. Beer, L. Fischer, G. Gareiss, W. Grab, R. Korber, B. Lortzing, and G. Steinbach: Prevention and control of salmonellosis in farm animals in East Germany. Mh. Vet.-Med. 12 (1990) 403-406 20. Michiels, T., M. Y. Popoff, S. Durviaux, C. Coynault, and G. Cornelis: A new method for the physical and genetic mapping of large plasmids: application to the localization of the virulence determinants on the 90 kb plasmid of Salmonella typhimurium. Microb. Pathogen 3 (1987) 109-116 21. Norel, F., C. Coynault, I. Miras, D. Hermant, and M. Y. Popoff: Cloning and expression of plasmid DNA sequences involved in Salmonella serotype typhimurium virulence. Molec. Microbiol. 3 (1989) 733-743 22. Popoff, M. Y., I. Miras, C. Coynault, C. Lasselin, and P. Pardon: Molecular relationships between virulence plasmids of Salmonella serotypes typhimurium and dublin and large plasmids of other Salmonella serotypes. Ann. Microbiol. (Inst. Pasteur) 135 A (1984) 389-398 23. Rhen, M., M. Virtanen, and P. H. Miikelii: Localization by insertion mutagenesis of a virulence associated region on the Salmonella typhimurium 96 kilobase pair plasmid. Microb. Pathogen. 6 (1989) 153-158 24. Roudier, c., M. Krause, J. Fierer, and D. Guiney: Correlation between the presence of sequences homologous to the vir region of Salmonella dublin plasmid pSDL2 and the virulence of twenty-two Salmonella serotypes in mice. Infect. Immun. 58 (1990) 1180-1185 25. Sapbra, I. and]. W. Wenter: Clinical manifestations of salmonellosis in man. N. Engl. J. Med. 256 (1957) 128-134
Virulence Plasmids in Live Salmonella Vaccines
21
26. Scholl, W., G. Grunert und K. Lind e: Anforderungen an potentielle bakterielle Lebendimpfstoffe zum Einsatz in der Veterinarrnedizin. Z. ges. Hyg. 23 (1977) 678-679 27. Scholl, W. und G. Greinert: Suisaloral'" "Dessau" - ein Salmonella-choleraesuis-i.ebendimpfstoff zur oralen, parenteralen und kombinierten Anwendung. Arch. expo Vet.Med. (Leipzig) 34 (1980) 91-97 28. Stok er, N. G., J. M. Pratt, and I. B. Holland : In vivo Gene Expression Systems in Procaryotes. In: B. D. Ham es and S. J. Higgins (eds.), Tran scription and Translation - a Practical Approach, pp. 153-] 77. IRL Press, Oxford -Washington (1984) 29. Taira, S. and M. Rhen : Identificat ion and genetic analysis of mkaA - a gene of the Salmon ella typhimurium virulence plasmid necessary for intracellular growth . Microb . Pathogen. 7 (1989) 165-1 73 30. Taira, S. and M. Rhen: Molecular organization of genes constituting the virulence determinant on the Salmo nella typh imurium 96 kilobase pair plasmid. FEBS Lett. 257 (1989) 274-278 31. Terakado, N., T. Sekizak i, K. Hashimoto, and S. Naitoh: Correlation between the presence of a 50-megadalton plasmid in Salmonella dublin and virulence for mice. Infect. Immun. 41 (1983) 443-444 32. Urbaneck, D. and W. Scholl: Anforderungen an Lebendimpfstoffe fur den Einsatz in der Veterinarrnedizin. Arch. expo Vet.-Med. (Leipzig) 34 (1980) 1-7 33. Williamson, C. M., G. D. Baird, and E. J. Manning: A common virulence region on plasmids from eleven serotypes of Salmonella. J. Gen. Microbiol. 134 (1988) 975-982
Dr. Wolfgang Beyer, Universitar Hohenheim, Insr. fur Tiermedizin und Tierhygiene mit Tierklinik, Postfach 700 562, D-7000 Stuttgart 70
Characterization of the Virulence Regions in the Plasmids of Three Live Salmonella Vaccines
WOLFGANG B E Y ER a nd LUT Z GEUE Institut fiir Epizootiologie und T ierseuchen bekam pfung, Abteilung fiir Mo leku lare Epizootiologie, 0- 1903 Wusterhausen, Germany
With 4 Figures ' Received July 8, 1991 . Revision received Novem ber 12, 1991 . Accepted December 20, 1991
Summary Three live Salmonella vaccines, Zoosaloral'P " Dessau" , Bovisaloral 'f " Dessau" , and Suisaloralv "Dessau" successfully used in veterinary medicine in eastern Germany were analy sed for thei r plasmid DNA content. Plasmi ds of 60 MDa (5. typhimurium), 55 M Da and 26 MDa (S. dub lin) and 34 MDa (S. choleraesuis) were found . The three large plasmids con ta ined "a virulence region as shown by restriction enzyme ana lysis and hybridization studies with virulence-region-specific DNA probes. Restriction enzyme pattern ana lysis of th e who le plasmids revea led differences between vaccine strains and parental strains. We wanted to know if there were alte rations within the viru lence region of the vaccine strain plasmid s compared wi th parenta l strains cau sed by th e attenuation procedure. Therefore, the 6 kb-Cla I fragments of several plasmids were cloned and characterized by restriction enzyme fingerprinting. The exp ression of the CIa I frag men t-encoded proteins was ana lysed in the minicell-producing stra in DS 410. The experi ments revea led that the restriction pattern of the 6 kb-C la I fragments of Zoosaloral and Bovisaloral as well as the PAGE pattern of virulence-region-encoded proteins were in complete accordance with one ano ther, between vaccine strains and the parental strain of Zoosaloral, and with a wild-type strain of S. typ him urium.
Zusammenfassung Die drei Salmonella-Lebendvakzinen, Zoosaloralv "Dessau" (S. typhimurium), Bovisalora l® "Dessau" (S. dublin ) and Suisaloralf "Dessau" (S. cboleraesuis), die mit gutem Erfolg in der Veterinarmedizin auf dem Gebiet der neuen Bundeslander eingesetzt wurden, ent halten Plasmide von 60 MDa (S. typhimurium), 55 MDa und 26 MDa (S. dublin) bzw. 34 MDa (S. cho leraesuis). Auf den drei grolsen Plasmiden wu rde mittels Restriktionsenz ym-Analyse und Hybridisierung mit vir-spezifi schen DNA-Sonden eine Virulenzregion nachgewiesen. Die Restriktionsmusrer der vollstandigen Plasmide zeigen Unterschiede zwischen den Vakzine-Stamrnen und ihren Elrern stam men . Uns interessierte, ob es infolge der Attenuierung Vera nderungen inn erhalb der vir-Regionen der Vakzi ne-Plasmide im Vergleich zu den Eltern-Pl asmiden gibe. Dafiir w urden die ent sprechenden 6 kb -Cla I Fragmente miteinander
Virulence Plasmids in Live Salmonella Vaccines
11
verglichen. Die Expression der durch die CIa I Fragmente kodierten Proteine wurde im Minizellstamm DS 410 analysiert. Die Experimente ergaben, dag sowohl die Restriktionsmuster der 6 kb-Cla I Fragmente von Zoosaloral und Bovisaloral als auch die PAGE-Muster der durch die vir-Region kodierten Proteine untereinander, zwischen den Vakzine-Stamrnen und dem Eltern-Stamm von Zoosaloral sowie mit einem Wild-Stamm von S. typhimurium vollig iibereinstimmen.
Introduction Live vaccines are used as one of the most effective ways to induce a strong immunity against the action of pathogenic enteric bacteria including Salmonella. Being aware of the risks involved in the use of live bacteria, a list of requirements for strains used were layed down (1, 14, 26, 32). Keeping in mind our modern knowledge on virulence plasmids, the requirement of safety with regard to the presence of transmissible genetic virulence structures is of special interest. Although up to-day there has been no explantation of how a virulence plasmid of a Salmonella sp. acts to make its host virulent, the evidence strongly suggests that virulence plasmids are important in the pathogenesis of naturally occurring systemic Salmonella infections in animals. Roudier et al. (24) have shown that the plasmid is absolutely necessary but not sufficient to make wild-type Salmonella spp. virulent for mice. All the major Salmonella species that are highly hostadapted to animals carry a virulence plasmid including S. choleraesuis (13), S. dublin (2,6,31), S. gallinarum (4), S. pullorum (3), and S. abortus avis (22, 33). Moreover, virulence-encoding regions were proved to exist for the following serotypes: S. enteritidis, S. typhimurium, S. rostock, S. paratyphi c, S. moscow, S. blegdam (33), S. johannisburg, and S. kottbus (15). It is also likely that the plasmid contributes to the virulence of nontyphoid Salmonella species in human disease and may be responsible for the septicemic forms of salmonellosis (25). The absence of virulence plasmid sequences in certain extraintestinal isolates of Salmonella does not contradict the hypothesis that the plasmid contributes to virulence in human disease since other important factors influence the invasiveness of Salmonella including the LPS structure, motility, ability to express specific adherence/invasion proteins after attachment to epithelial cells, and probably specific virulence genes necessary for entry, multiplication and penetration into host cells. Otherwise, factors such as the inoculum size, the immunocompetence (10) and maybe the genetic resistance of the host are of importance for the fate of a Salmonella infection (for a review see ref. 8, 11). Over the last fifteen years, three live Salmonella vaccines were developed and applied in Veterinary Medicine with good results in eastern Germany (19 and communications presented at the meetings of the WHO Working Group on Salmonella Immunization in Animals held in Munich (1986), lena (1989), and lena (1991), respectively). The principle of these vaccines is the combination of two independently attenuating markers with no limitation on propagation (14,17,26). All three vaccine strains contain adenine as the first attenuating marker. Zoosaloral® "Dessau" (S. typhimurium) and Bovisaloralf "Dessau" (S. dublin) were made by using an S-form auxotrophic phenotype his- and thi" respectively, which may lead to a reduced survival within macrophages and/or cells of the reticulo-endothelial system (7). Suisaloralv "Dessau" (S. choleraesuis) was made from an R-form mutant. The auxotrophy mutations are accidental mutations induced by incubation of the parental strains with N-methyl-N'-
W. Beyer and L. Geue
12
nitro-nitrosoguanidine (17). The site(s) and the nature of these mutations have not been clarified beyond doubt. Therefore, alterations within the vir region of the virulence plasmids participating in the attenuation are not unlikely. We were interested in evaluating whether there were any alterations caused by the attenuation procedures within the virulence region of the virulence plasmids as compared to parental strains. We have been able to show that all three vaccine strains do contain a virulence plasmid with a restriction fragment pattern different from that of their parental strains. However, the restriction pattern of the 6 kb-Cla I fragments as well as the PAGE pattern of virulence-region-encoded proteins are in complete accordance with one another, between vaccine strains and the parental strain of Zoosaloral, and with a wild-type strain of S. typhimurium.
Material and Methods Bacterial strains and growth conditions. Bacterial strains, descriptions and sources are listed in Table 1. Cells were grown in L broth containing pancreatic peptone 17 gil, yeast extract 3 gil, NaCI 5 gil, glucose 1 gil, or on L broth containing 10 gil agar at 37°C. Ampicillin was used at 100 ug/ml. Plasmid preparations for cloning procedures. The DNA isolation of native Salmonella plasmids was carried out as described by Kado and Liu (11) with the following modifications. Briefly, bacteria of a 200 ml culture were centrifuged in four 50 ml portions, the pellets lysed in 25 ml of 50mM Tris base, 3% SDS, 290 f!V100 ml of 11.2 N NaOH,
Table 1. Bacterial strains Strain
Characterization
Source
E. coli (K12) TG1
/lilac pro); thi; str A; sup E; rec A; end A; sbcB; hsdR-; proAB; lacIq ; Z/lM15
Central Inst. of Mol. BioI. of the Acad. of Sci., Berlin-Buch
E. coli (Kll) DS410 (M2141)
minicell producing strain /lilac pro); min A; min B; rps L; thi; thr
Max Planck Institut Berlin
W 3839
S. typhimurium, virulent field isolate
Inst. of Exp. Epidemiology, Wernigerode
Zoosaloral'"
S. typhimurium, live vaccine strain (ade, his)
Commercially available
"Dessau"
Bovisaloralf "Dessau"
S. dublin, live vaccine strain (ade, thi, nad)
Commercially available
Suisaloralf "Dessau"
S. choleraesuis, live vaccine strain (ade, R)
Commercially available
Moscow strain
S. typhimurium, parental strain of Zoosaloral
Inst. of Veterinary Ecology, Microbiology and Immunol., Dessau
AF2423246
S. dublin, parental strain of Bovisaloral (nad)
Inst. of Medical Microbiol. of the F.-Schiller-University, jena
Virulence Plasmids in Live Salmonella Vaccines
13
incubated at 55°C for 2 h. Then the probes were pooled and 50 ml precooled 3M NaOAc, pH 4.8, added dropwise. After centrifugation (15000 g), the DNA was precipitated with 0.8 volumes of isopropanol, further purified with phenol/chloroform (50% v/v) extractions and then precipitated with ethanol. For cloning procedures, the DNA was further purified by centrifugation through a caesium chloride-ethidiurn bromide density gradient as described (18). For plasmid profiles, a small-scale procedure was routinely used. About 2 colonies of fresh bacteria from an agar plate were lysed in 200 ul lysis solution, incubated at 55°C for 1 h and extracted with 800 [!l phenol/chloroform (50% v/v). After centrifugation (11000 g), the supernatant was used directly for electrophoresis. Plasmids and their restriction products were routinely analysed on horizontal 0.6% (w/v) agarose gels run in Tris-acerate buffer overnight. These gels were used for Southern transfers to nitrocellulose membranes. Recombinant DNA techniques, construction of recombinant plasmids. Plasmid DNA preparations, restriction analysis, ligations and transformations were performed by standard procedures essentially as described in (18). Enzymes were from Amersham Buchler GmbH & Co KG, Braunschweig, agarose for DNA electrophoresis was from Serva. The recombinant plasmids pWh 170, 180, 190 and 200 were obtained by cloning the 6 kb Cia I fragments known to harbour the virulence region from the virulence plasmids of W3839, Zoosaloral, Bovisaloral, and "Moscow strain", respectively, in Cia l-cleaved pBR322. The recombinant plasmids were propagated in E. coli TGl. Preparation of the 3.9 k b Hind III DNA probe p Wh 70 for analysis of plasmid profiles and restriction maps. DNA of the virulence plasmid of S. typhimurium W3839 was digested with Hind III and DNA fragments were separated by gel electrophoresis. The 3.9 kb fragment was recovered by clcctroclurion, ligated into Hind III-cleaved vector pUC19 and cloned into E. coli TG1. The clone was named pWh70. According to the maps of Baird et al. (2) and Michiels et al. (20), this fragment should contain a virulence-determining region. This was proved to be true by restriction enzyme fingerprinting with Eco RI, Bam HI, Cia I, and Xho I and by sequencing some hundreds of the nucleotides in end position. The purified insert of pWh 70 was labelled by randomprimed incorporation of digoxigenin-labelled deoxyuridine triphosphate according to the recommendations of Boehringer Mannheim or with a- J2-PdNTP using the DNA-labelling kit from Serva. Hybridizations. Hybridization and detection procedures with the digoxygenin DNAlabelling kit were carried out according to the recommendations of the manufacturer (Boehringer). For the hybridization with a- 32PdNTP-labelled probes (Amersham), we used a prehybridization procedure with" x sse (0.45 M NaCl, 0.045 M Na citrate), 0.5% (w/v) skim milk, 0.1 % (w/v) SDS and 100 [!g/ml sonicated and boiled unspecific DNA (4-5 h) and hybridization procedure with 3 x SSC, 0.2% (w/v) skim milk, 0.1 % (w/v) SDS and 10% (w/ v) PEG 6000 (Serva) and up to 50 ng/ml denatured labelled probe, 15 h, 65°C. The filters were washed briefly in 2 x sse, 0.1 '/'o 50S, RT, then once in 2 x SSC, 0.1 % SOS, 65°C, 30 min, and two times in 0.2 x sse, 0.1 0;;, SDS, 65 DC, 30 min and autoradiographed with xray film and intensifying screens (ORWO). Analysis of plasmid-directed protein synthesis in minicells. The E. coli mini cell-producing strain DS410 was used to identify plasmid-encoded proteins. DS410 was transformed with plasmids pBR322, pWh170, 180, 190, and 200 by the calcium chloride procedure (18). Minicells were isolated and labelled with 35S-methionine (> 37 TBq/mmol, Amersham) at a concentration of 1.8 MBq/run essentially as described (28). Labelled minicells were solubilized in 5DS-PAGE sample buffer and the protein content was electrophoretically separated in 10% SDS-PA gels (16). The gels were fixed in 10% (v/v) acetic acid (30 min), impregnated with Amplify" (Amersham) (30 min), dehydrated in 90% (v/v) ethanol, 10% (v/v) glycerol (3 h) air-dried overnight and exposed to X-ray films (ORWO) (3-6 d). As molecular mass standard, the rainbow-coloured markers from Amersham, radiolabelled after electrophoresis, were used.
14
W. Beyer and L. Geue Results
Plasmid profiles In order to characterize the virulence plasm ids, the following plasmid profiles were established. From the live Salmonella vaccines, Zoosaloral - 60 MDa ; Bovisaloral- 55 MD a and 26 MDa, Suisaloral - 34 MDa and from the parental strains S. typhimurium "Moscow strain" - 60 MDa and S. dublin AF2423246 - 55 MDa and 26 MDa. In comparison , the S. typhimurium field strain W3839 - 60 MDa was coprepared. All plasm ids, with the exception of the 26 MDa plasmid of Bovisaloral hybridized with the probe pWh70. The probes pWh 70 and pWh 170 have been explained under Material and Methods and Fig. 3.
Restriction enzyme analysis and hybridization All plasmids were cleaved with restriction enzymes Hind III, Bam HI, Eco RI and Cia I and the gelelectrophoretical cleavage patterns were compared. These enzymes have
Fig. 1a. Hind III-parrern of the plasmids of Zoosa loral and its parental strain. Lane 1 - Hind III marker; lane 2 - W 3839 ; lane 3 - "Moscow strain"; lane 4 and 5 - Zoosaloral; lane 6Pvu II marker. In accordance with Michiels er al. (20), the eight fragments of "Moscow strain" have been designated H I to H8 in the text. The calculated size of all fragments is indicated.
Virulence Plasmids in Live Salmonella Vaccines
15
recognit ion sites within the conserved virulence region (9, 20, 21, 29 ). While the cleavage patterns with Bam HI and Cia I were identical for Zoosaloral and the "Moscow strain ", there were striking differences in the Hind III pattern (Fig.1a). While the "Moscow strain" showed a cleavage pattern of 8 fragments (H 1- H 8) (20) frequently occurring in 60 MDa virulence plasmids, the highly conserved 3.9 kb fragment H3 (20, 22 ) as well as the 16 kb fragment H3 (20) were absent in the vaccine strain Zoosaloral. The following bands hybridized with the probe pWh 70 after Southern transfer of the DNA to nitrocellulose (nc) membr anes: W3839 - 3.9 kb, " Moscow strain" - 3.9 kb (H6), Zoosaloral- 19.8 kb (H2, Fig. 2). Identical digestion patterns with Bam HI and Cia I but striking differences with the other two enzymes were found in comparison of Bovisaloral and its parental strain AF2324236. Fig. 1b shows the cleavage pattern for Eco RI and HindIII. After transfer of the fragments to nc-membranes, in both strains the 3.9 kb band of the Hind III digestion and identical band s of the Eco RI digestion hybridized with the probe
2
3 4
Fig. 1b. Hind IIlIEco RI-pane rn of the plasmids of Bovisal and its parental strain. Lane 1 Bovisaloral; lane 2 - AF2324236 (both Hind III cleaved); lane 3 - Bovisaloral; lane 4 AF2324236 (both Eco RI cleaved).
16
W.Beyer and L. Geue
1
2
3
4
19,8
3,9
Fig. 2. Hybridization pattern of the Hind III cleavedplasmids of: lane 1 - Bovisaloral;lane 2 - W 3839, lane 3 - Moscow strain; lane 4 - Zoosaloral
pWh70. The fragments of the CIa I digestion (not shown) became hybridized with the insert from clone pWh170. As after the CIa I digestion of W3839, Zoosaloral, and "Moscow strain", identical fragments of about 6 kb in size, hybridized. After cloning the 6 kb CIa I fragments from W3839 (pWh170), Zoosaloral (pWh180), Bovisaloral (pWh190), and "Moscow strain" (pWh200), the restriction sites for each fragment for the enzymes Eco RI, Pst I, Sac I, Kpn I, Hind III and Bam HI were investigated and compared with previously published restriction maps (2, 5, 9, 20, 21, 23, 29, 30). The restriction sites of all clones tested were identical with one another and with data given in literature. Fig. 3 shows the restriction map of pWh170 which is representative of all clones tested. After the digestion of Suisaloral with CIa I, the defined virulence region was contained within the largest about 16 kb fragment. This fragment was not further characterized. Expression of the virulence region
The recombinant plasmids pWh170, 180, 190, and 200 were introduced into the minicell-producing strain DS 410. Metabolic labelling with 35S-metwas used to analyse
Virulence Plasmids in Live Salmonella Vaccines
o
2
5
4
3
HB
7
6
E
C
E
C
9 kb
8
XB
17
H
I I
C
33 kO
EP
SP
I(
I(
II
mkaC
31 kO mkaB
II
KBSHHB
I( III I
pWh170
I
70 kO
1128 kol
mkaA
mkaO
Fig. 3. Restriction map of pWh70 and pWh170. Abbreviations used: C = Cia I, E = Eco R I, S = Sac I, P = Pst I, K = Kpn I, H = Hind III, B = Bam H I. The alignment of the four protein coding sequence regions was taken from Taira and Rhen (28).
6
54
3
2
97,4
6QO
46.0
30.0 Fig. 4. Fluorograms of 35S-methionine labelled E. coli minicells run in 10% SDS-PAGE. The plasmids contained are as follows: lane 1 - marker; lane 2 - pBR322; lane 3 - pWh170; lane 4 - pWh180; lane 5 - pWh200; Ian 6 - pWh190. Numbers on the right indicate the positions of molecular mass markers (in kDa). 2
Zbl. Bakt. 27711
18
W.Beyer and L. Geue
the expression of plasmid-encoded proteins. A common protein pattern of 70 kDa , 31 kDa , and 28 kDa was demonstrated for all clones tested in SDS-PAGE (Fig. 4).
Discussion Three live Salmonella vaccines which have been used successfully in veterinary medicine in eastern Germany pro ved to conta in virulence plasmids as shown by restriction enzyme analysis and hybridization with virulence-r egion-specific DNA probes. We were interested in evaluat ing whether there were any alter ation s caused by the attenuation pro cedures within the virulence region of these plasmids as comp ared to their parental stra ins. Therefore we comp ared the restriction maps of the whole vaccine strain plasmids and of the appropriate Cia I fragments of the vaccine str ain plasmids with one another, with their parental strains and with an S. typhimurium wild type strain as well as the PAGE pattern of virulence-region-encoded proteins. The Hind III pattern of Zoos aloral'P "De ssau " and its parent "Moscow stra in" were different (Fig. 1a). One of both Hind III sites of the otherwise conserved 3.9 kb fragment must have become mutated in Zoosaloral. It seems that the 19.8 kb band consists of two fragme nts, one is the same as in the " Moscow strain" and one results from the additi on of the 3.9 kb fragment and the missing 16 kb fragment (H3). Indeed, this band hybridized with the pro be pWh 70 (Fig. 2). According to the map of Gulig and Curtiss (9) the 3 '-Hind III site that means the site outside the CIa I vir-fragme nt must be the mut ated one (Fig. 3). Th is conclusion is supported by the concordance of the Hind III sites within the CIa I fragments of Zoosaloral and "Moscow strain" . There were differences in the whol e plasmid restri ction pattern of Bovisaloral'" "Dessau" and its parent AF2423246. However, when comp aring the 6 kb-Cla I fragments of Bovisaloral with those of Zo osaloral , "Moscow strain", and the S. typhimurium wild type plasmid, no differences occurred in the sites of all enzymes tested. So it was not necessar y to cha racterize the appro priate Cia I fragment of AF242324 6 in detail. M oreover, in all digestion s of Bovisaloral and AF2423246, ident ical fragments hybridized with our vir-region-specific DNA prob es. So we conclude th at the observed differences in the restriction sites of Eco RI and Hind III, one additional site in both cases in the AF2423246 strain (Fig. lb ), are localized outside the conserved Cia I virregion . In summary, there were no differen ces within the virulence regions defined by the CIa I fragments between the examined vaccine strains, the parental strain of Zoosaloral and the S. typhimurium wild type strain W3839 regarding the recognit ion sites of all enzymes tested. In order to compare the vir-regions on the protein level, the expre ssion of the Cia I fragment-encoded proteins was ana lysed in the minicell-producing strain DS 410. Protein pattern s completely ident ical with one another and in accordance with literature data were found with proteins of 70 kDa, 3 1 kDa, and 28 kDa, respectively. The fourth described prote in of 33 kDa (mka C, 30) could not be ident ified. Indeed, it seems to be difficult to identify all four proteins at the same time if the complete virregion is used. So Gulig and Curtiss (9) found the 3 1 kDa protein only in an in vitr o transcription/t ranslati on rest. They did not describe the 70 kDa protein. Nore! et al. (21) did not find the 31 kDa and the 33 kDa proteins using the same CIa I fragment in a
Virulence Plasmids in Live Salmonella Vaccines
19
maxicell expression system. The analysis of a subcloned Cia I - Eco RI fragment, coordinates 0 to 1.5 (Fig. 3), by Taira and Rhen (30) revealed two protein bands of 33 kDa and 30 kDa. Although the theoretical size of mka C should be 33.8 kDa (30), the 30 kDa band was more pronounced. Therefore it cannot be excluded that an identical protein band is masked by the 30 kDa protein ~-Iactamase of pBR322 in our experiments. Differences in the signal strength between lanes 3/5 and 4/6, respectively, do not reflect real differences in the expression level between the strains. There are rather problems with the minicell expression system in our hands, maybe in the procedure to generate viable minicells. We reached different values in several experiments for the same strain under equal conditions. In summary, also the analysis of the expression of the vir-region-encoded proteins by SDS-PAGE showed no differences between vaccine strains, the parental strain of Zoosaloral and the wild type strain W3839, respectively. Although it is not opportune to make an assertion considering the biological function of the vir-regions of the vaccine strains on the basis of our test criteria, the complete concordance of all features examined supports the hypothesis that no structural and with it, also no functional changes in the virulence-encoding region of the vaccine strains virulence plasm ids took place. So these plasmids might be fully intact with regard to their functions. Because of the unspecific action of N-methyl-N' -nitro-nitrosoguanidine on DNA, there might be several other mutations on the chromosomes and the plasmids, maybe also within other known virulence factors. This has not yet been investigated. At this time, the attenuation can only be ascribed to the auxotrophies and the alteration in LPS of Suisaloral, respectively. Remembering the requirements for live vaccines and the potential risks linked with the presence of virulence plasmids in Salmonella strains, future live Salmonella vaccines should not contain virulence plasmids unless they are essential for immunogenic properties. This should be tested in every case. Acknowledgements. The authors wish to thank the Institutes mentioned in Table 1 for providing the strains. We thank Mrs. Ch. Schnick and Mrs. M. Lembke for technical assistance and Dr. R. Hellmuth (Robert v. Ostertag-Institut, Berlin) for critical reading of the manuscript.
References 1. Anonymus: Oral enteric bacterial vaccines. WHO Techn. Rep. Ser. Nr. 500 (1972) 2. Baird, G. 0., E.]. Manning, and W. Jones: Evidencefor related virulence sequences in plasmids of Salmonella dublin and Salmonella typhimurium. J. Gen. Microbiol. 131 (1985) 1815-1823 3. Barrow, P. A. and M. A. Lovell: The association between a large molecular mass plasmid and virulence in a strain of Salmonella pullorum. J. Gen. Microbial. 134 (1988) 2307-2316 4. Barrow, P. A., ]. M. Simpson. M. A. Lovell. and M. M. Binns: Contribution of Salmonella gallinarum large plasmid toward virulence in fowl typhoid. Infect. Immun. 55 (1987) 388-392 5. Beninger, P. R., G. Chieami, K. Tanabe, C. Roudier, ]. Pierer, and D. G. Guiney: Physical and genetic mapping of the Salmonella dublin virulence plasmid pSDL2. Relationship to plasmids from other Salmonella strains. J. Clin. Invest. 81 (1988) 1341-1347
20
W. Beyer and L. Geue
6. Chikami, G. K., J. Fierer, and D. G. Guiney: Plasmid mediated virulence in Salmonella dublin demonstrated by use of a 5-oriT construct. Infect. Immun. 50 (1985) 420-424 7. Fields, P. I., R. U. Swanson, C. G. Haidaris, and F. Heffron: Mutants of Salmonella typhimurium that cannot survive within the macrophage are avirulent. Proc. Nat!. Acad. Sci. USA 83 (1986) 5189-5193 8. Finlay, B. B. and S. Falkow: Salmonella as an intracellular parasite. Malec. Microbial. 3 (1989) 1833-1841 9. Gulig, P. A. and R. Curtiss III: Cloning and transposon insertion mutagenesis of virulence genes of the 100-kilobase plasmid of Salmonella typhimurium. Infect. Immun. 56 (1988) 3262-3271 10. Hook, E. W.: Salmonellosis: certain factors influencing the interaction of Salmonella and the human host. Bull. N.Y. Acad. Med. 37 (1961) 499-512 11. Hsu, H. S.: Pathogenesis and immunity in murine salmonellosis. Microbiol. Rev. 53 (1989) 390-409 12. Kado, C. I. and S. T. Liu: Rapid procedure for detection and isolation of large and small plasmids. J. Bact. 145 (1981) 1365-1373 13. Kawahara, K., Y. Haraguchi, M. Tsuchimoto, N. Terakado, and H. Danbara: Evidence of correlation between 50-kilobase plasmid of Salmonella choleraesuis and its virulence. Microb. Pathogen. 4 (1988) 155-163 14. Koch, H.: Untersuchungen zur Prufung von bakteriellen Mutanten als potentielle Impfstarnme. Arch. expo Vet.-Med., Leipzig 34 (1980) 43-50 15. Korpela, K., M. Ranki, s. Sukupolvi, P. H. Miikelii, and M. Rhen: Occurrence of Salmonella typhimurium virulence plasmid-specific sequences in different serovars of Salmonella. FEMS Microbiol. Lett. 58 (1989) 49-54 16. Laemmli, U. K.: Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227 (1970) 680-685 17. Linde, K.: Herstellung von stabilen Salmonella-Impfstammen durch Kopplung von zwei unabhangig voneinander nichtvermehrungsbegrenzenden attenuierenden Markern. Arch. expo Vet.-Med. (Leipzig) 34 (1980) 19-32 18. Maniatis, T., E. F. Fritsch, and]. Sambrook: Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring HarbourlNY (1982) 19. Meyer, H., K. Beer, L. Fischer, G. Gareiss, W. Grab, R. Korber, B. Lortzing, and G. Steinbach: Prevention and control of salmonellosis in farm animals in East Germany. Mh. Vet.-Med. 12 (1990) 403-406 20. Michiels, T., M. Y. Popoff, S. Durviaux, C. Coynault, and G. Cornelis: A new method for the physical and genetic mapping of large plasmids: application to the localization of the virulence determinants on the 90 kb plasmid of Salmonella typhimurium. Microb. Pathogen 3 (1987) 109-116 21. Norel, F., C. Coynault, I. Miras, D. Hermant, and M. Y. Popoff: Cloning and expression of plasmid DNA sequences involved in Salmonella serotype typhimurium virulence. Molec. Microbiol. 3 (1989) 733-743 22. Popoff, M. Y., I. Miras, C. Coynault, C. Lasselin, and P. Pardon: Molecular relationships between virulence plasmids of Salmonella serotypes typhimurium and dublin and large plasmids of other Salmonella serotypes. Ann. Microbiol. (Inst. Pasteur) 135 A (1984) 389-398 23. Rhen, M., M. Virtanen, and P. H. Miikelii: Localization by insertion mutagenesis of a virulence associated region on the Salmonella typhimurium 96 kilobase pair plasmid. Microb. Pathogen. 6 (1989) 153-158 24. Roudier, c., M. Krause, J. Fierer, and D. Guiney: Correlation between the presence of sequences homologous to the vir region of Salmonella dublin plasmid pSDL2 and the virulence of twenty-two Salmonella serotypes in mice. Infect. Immun. 58 (1990) 1180-1185 25. Sapbra, I. and]. W. Wenter: Clinical manifestations of salmonellosis in man. N. Engl. J. Med. 256 (1957) 128-134
Virulence Plasmids in Live Salmonella Vaccines
21
26. Scholl, W., G. Grunert und K. Lind e: Anforderungen an potentielle bakterielle Lebendimpfstoffe zum Einsatz in der Veterinarrnedizin. Z. ges. Hyg. 23 (1977) 678-679 27. Scholl, W. und G. Greinert: Suisaloral'" "Dessau" - ein Salmonella-choleraesuis-i.ebendimpfstoff zur oralen, parenteralen und kombinierten Anwendung. Arch. expo Vet.Med. (Leipzig) 34 (1980) 91-97 28. Stok er, N. G., J. M. Pratt, and I. B. Holland : In vivo Gene Expression Systems in Procaryotes. In: B. D. Ham es and S. J. Higgins (eds.), Tran scription and Translation - a Practical Approach, pp. 153-] 77. IRL Press, Oxford -Washington (1984) 29. Taira, S. and M. Rhen : Identificat ion and genetic analysis of mkaA - a gene of the Salmon ella typhimurium virulence plasmid necessary for intracellular growth . Microb . Pathogen. 7 (1989) 165-1 73 30. Taira, S. and M. Rhen: Molecular organization of genes constituting the virulence determinant on the Salmo nella typh imurium 96 kilobase pair plasmid. FEBS Lett. 257 (1989) 274-278 31. Terakado, N., T. Sekizak i, K. Hashimoto, and S. Naitoh: Correlation between the presence of a 50-megadalton plasmid in Salmonella dublin and virulence for mice. Infect. Immun. 41 (1983) 443-444 32. Urbaneck, D. and W. Scholl: Anforderungen an Lebendimpfstoffe fur den Einsatz in der Veterinarrnedizin. Arch. expo Vet.-Med. (Leipzig) 34 (1980) 1-7 33. Williamson, C. M., G. D. Baird, and E. J. Manning: A common virulence region on plasmids from eleven serotypes of Salmonella. J. Gen. Microbiol. 134 (1988) 975-982
Dr. Wolfgang Beyer, Universitar Hohenheim, Insr. fur Tiermedizin und Tierhygiene mit Tierklinik, Postfach 700 562, D-7000 Stuttgart 70
