- Email: [email protected]
SEF17 fimbriae are essential for the convoluted colonial morphology of Salmonella enteritidis
FEMS Microbiology Letters 153 (1997) 33^42
SEF17 ¢mbriae are essential for the convoluted colonial morphology of Salmonella enteritidis Emma Allen-Vercoe, Mike Dibb-Fuller, Christopher J. Thorns, Martin J. Woodward* Department of Bacteriology, Central Veterinary Laboratory, Woodham Lane, New Haw, Addlestone, Surrey KT15 3NB, UK
Received 7 March 1997; revised 29 April 1997; accepted 2 May 1997
Abstract
Salmonella enteritidis isolated from poultry infections generated a convoluted colonial morphology after 48 h growth on colonisation factor antigen (CFA) agar at 25³C. A mutant S. enteritidis defective for the elaboration of the SEF17 fimbrial antigen, in which the agf gene cluster was inactivated by insertion of an ampicillin resistance gene cassette, and other wild-type 3 mutants at 25³C S. enteritidis transduced to this genotype failed to produce convoluted colonies. However, growth of SEF17 on CFA agar supplemented with 0.001% Congo red resulted in partial recovery of the phenotype. Immunoelectron microscopy demonstrated that copious amounts of the SEF17 fimbrial antigen were present in the extracellular matrix of convoluted colonies of wild-type virulent S. enteritidis isolates. Bacteria were often hyperflagellated also. Immunoelectron microscopy of SEF173 mutants grown on CFA agar+0.001% Congo red demonstrated the elaboration of an as yet undefined fimbrial structure. Isolates of S. enteritidis which were described previously as avirulent and sensitive to environmental stress failed to express SEF17 or produce convoluted colonies. These data indicate an essential role for SEF17, and possibly for another fimbria and flagella, in the generation of the convoluted colonial phenotype. The relationship between virulence and colonial phenotype is discussed.
Keywords : Salmonella enteritidis
; Salmonellosis; Fimbriae; Colony morphology; SEF17
1. Introduction
Poultry has been cited as an important reservoir of associated with human infection and food safety measures have been hampered by the ability of S. enteritidis to cause invasive, subclinical infection in farmed chicken £ocks that may be di¤cult to detect [1]. In particular, the apparent Salmonella enteritidis
* Corresponding author. Tel.: +44 (1932) 357582; Fax: +44 (1932) 347046; E-mail: [email protected]
tropism that S. enteritidis has for the reproductive tissues of laying hens results in vertical transfer [2]. Guard-Petter [3] observed that isolates of S. enteritidis di¡ered in their ability to infect laying hens and contaminate eggs and that this ability was associated with the production of `lacy' colonies, a phenotype ¢rst described by Jameson [4] as `convoluted' in reports of colonial morphology of S. typhimurium isolates. We will use the term convoluted throughout this paper. Humphrey et al. [5] showed that isolates of S. enteritidis of the same phage type di¡er not only in their pathogenic potential for mice and
0378-1097 / 97 / $17.00 ß 1997 Federation of European Microbiological Societies. Published by Elsevier Science B.V. PII S 0 3 7 8 - 1 0 9 7 ( 9 7 ) 0 0 2 1 9 - X
FEMSLE 7647 20-10-97
34
E. Allen-Vercoe et al. / FEMS Microbiology Letters 153 (1997) 33^42
chickens but also in their tolerance to certain environmental stresses. These authors noted that the tolerant/virulent phenotype was found to correlate with the elaboration of the convoluted phenotype which was found to be inherently stable, requiring no special culturing conditions for enhancement, and suggested that such traits may be genetically determined by genes controlling the expression of various extracellular components. This suggestion was supported by Guard-Petter et al. [6] who demonstrated the convoluted phenotype of S. enteritidis was associated with hyper£agellation and the formation of high-molecular-mass O-antigen although ¢mbriae were not detected. S. enteritidis has the potential to elaborate at least six di¡erent ¢mbrial types [7] and we have used a reversed genetics approach to insertionally inactivate genes encoding the elaboration of ¢mbriae and £agella in order to study their role in S. enteritidis pathogenesis. We report that S. enteritidis phage type 4 mutants unable to synthesise the thin, aggregative, hydrophobic ¢mbriae, SEF17 [8], through inactivation of the agfA gene as well as a naturally occurring avirulent S. enteritidis were unable to form convoluted colonies whereas virulent wild-types isolates did. We present evidence that SEF17 is expressed by bacteria elaborating convoluted colonies and that another hitherto undescribed ¢mbria may be associated with this phenotype also. 2. Materials and methods
2.1. Bacterial strains
phage type 4 strains S1400/94 and LA5 are wild-type ¢eld isolates from infected chickens. Strains EAV11 and EAV12 are SEF173 mutants of S1400/94 and LA5, respectively. S. enteritidis strains E, I [5] and C6B were kindly provided by Prof. T.J. Humphrey of the Food Microbiology Research Unit, Public Health Laboratory, Exeter. Strain C6B was a recently isolated strain that showed extreme tolerance of heat, acid and oxidative stress (T.J. Humphrey, personal communication). Strains EAV31, EAV32 and EAV33 are SEF173 mutants of E, I and C6B respectively. S. enteritidis strain 27655R [8] was a kind gift from Prof. T. Wadstrom of Lund University, Sweden, which displayed conS. enteritidis
stitutive expression of SEF17. Strain EAV30 is a SEF173 mutant of 27655R. 2.2. Growth conditions
All strains were maintained on Dorset's egg slopes at 20³C. The convoluted phenotype was promoted by growth of the strains on colonisation factor antigen (CFA) agar at 25³C for approximately 48 h, conditions previously suggested to achieve maximal expression of thin, aggregative ¢mbriae [9]. Congo red (Sigma) was added to this medium to a ¢nal concentration of 0.001% (v/v) and ampicillin (50 Wg ml31 ) was added when appropriate. 2.3. Genetic manipulation methods
Unless stated otherwise, all genetic manipulation methods were as described by Sambrook et al. [10]. 2.4. Polymerase chain reaction
PCR was performed essentially following the methods of Saiki et al. [11] with primers designed on the agf gene cluster using the sequence data of Collinson et al. [12] (GenBank accession number U43280) and were: AGF1F
3P-ATG ATG TTG ACA ATA CTG GGT-5P AGF1R 3P-TGA TAA ATG CAG TGA TTG TCC-5P
position 716^737 position 2009^1988.
2.5. Construction of S. enteritidis S1400/94 agf
The agf region was ampli¢ed by PCR using DNA extracted from S1400/94 as a template with primer pair AGF1F and AGF1R. The product of the predicted size (1.3 kb) was obtained and cloned directly into pCRscript (Stratagene) to generate recombinant plasmid pVW680. A unique EheI site within agfA was con¢rmed into which an ampicillin resistance gene cassette (Stratagene) was cloned to generate recombinant plasmid pVW680A. A 2.2 kb agfA ::ampr fragment was ampli¢ed with primer pair AGF1F and AGF1R with puri¢ed pVW680A as template and was ligated with suicide vector pERFORM C (Allen-Vercoe and Woodward, in preparation), a
FEMSLE 7647 20-10-97
E. Allen-Vercoe et al. / FEMS Microbiology Letters 153 (1997) 33^42
chloramphenicol-resistant suicide vector derived from pJP5603 [13]. The ligation mixture was used to transform E. coli CC118 V-pir with selection made jointly for chloramphenicol- and ampicillin-resistant recombinants. The restriction map of plasmid DNA from one recombinant was con¢rmed and the construct was designated pVW681AC. In order to mobilise pVW681AC into S. enteritidis S1400/94 by conjugation, pVW681AC was transformed into E. coli K12 strain S-17 V-pir, an auxotrophic mobilising strain. Conjugal pro¢ciency was supplied in trans by the chromosomally inserted incP mini-plasmid, pRP4, and a mating of the form E. coli K12 S-17 V-pir pVW681ACUS. enteritidis S1400/94 was set up with selection made on glucose minimal medium supplemented with ampicillin and chloramphenicol. Exconjugants were obtained at a frequency of 1U1038 and one was subcultured in Luria-Bertani broth without antibiotic selective pressure. An ampicillin-resistant, chloramphenicol-sensitive derivative was selected in which the native chromosomal agf gene cluster had been replaced by homologous recombination with the agf ::ampr construct with consequent loss of the chloramphenicol-marked suicide vector and designated EAV11. 2.6. Hybridisation
DNA probes were prepared by PCR as described above and radioisotopically labelled with 32 P using `Ready to Go' dCTP labelling kit (Pharmacia) following the manufacturer's recommendations. Colony dot blot and Southern hybridisation, hybridisation conditions and post-hybridisation washes were as described previously [14].
35
mately 0.1 Wg of pUC19 DNA (New England Biolabs) and pulsed in a BioRad Gene Pulser electroporator using the settings 25 WF capacitance, 400 6 resistance, and electrical potential of 2 kV, conditions known to give maximal e¤ciency for transformation of this strain. Cells were allowed to recover for 1 h in SOC medium (Gibco-BRL) before spreading onto selection plates. A single isolate was tested for the presence of the pUC19 plasmid using the `Perfect Prep' plasmid DNA extraction kit (5PC3P, supplied by CP Labs) and was designated S1400/94/ pUC19. 2.9. Electron microscopy
Cultures were grown onto Formvar-coated grids (CVL Addlestone) placed on uncharged Hybond-N membrane (Amersham) overlaid on agar plates. After incubation, the grids were allowed to soak in ¢lter-sterile 1UPBS for 5 min in order to loosen the cells so that excess could be easily removed. The grids were stained with 2% phosphotungstic acid for 15 s prior to viewing with a Philips TM10 transmission electron microscope. 2.10. Immunoelectron microscopy
Immunoelectron microscopy was carried out on grids, onto which cells had been grown as described above, using a monoclonal antibody raised against SEF17 (Dibb-Fuller et al., submitted) and Auroprobe (Amersham) gold conjugate. The procedure described by Sojka et al. [17] was used. 3. Results
2.7. P22 transduction
The preparation of a P22 lysate of strain EAV11 and its use in the transduction of the agfA ::ampr genotype was carried out essentially using the method of Anderson et al. [15]. 2.8. pUC19 transformation
Wild-type S. enteritidis S1400/1994 was prepared for electroporation using the method of Boëttger [16]. Electrocompetent cells were mixed with approxi-
3.1. Characterisation of an agf mutant of S. enteritidis strain S1400/1994 S. enteritidis strain S1400/94 was chosen in the ¢rst instance to generate a SEF173 mutant. To do this, the agf gene cluster, encoding the elaboration of SEF17 ¢mbriae at the cell surface, was disrupted by the insertion of an ampicillin resistance gene cassette (see Section 2) and one ampicillin-resistant agf mutant designated EAV11 was selected for further analysis. Southern hybridisation experiments were
FEMSLE 7647 20-10-97
E. Allen-Vercoe et al. / FEMS Microbiology Letters 153 (1997) 33^42
36
r
done to con¢rm the genetic lesion was located within
luted
agf operon. Total genomic DNA was extracted from S. enteritidis strain S1400/94 and the isogenic agf3 derivative, digested to completion in separate experiments with SspI, SalI and HpaI, resolved by
introduced into wild-type S1400/94 and LA5 cells
agarose
which supported the hypothesis that the formation
the
gel
electrophoresis,
transferred
membranes and hybridised with the
agf
to
nylon
gene probe.
Hybridisation data con¢rmed insertion of the 0.9 kb
by
colonies.
To
electroporation.
do
All
this,
was
pUC19amp
transformants
produced
convoluted colonies when grown at 26³C on CFA agar
with
and
without
ampicillin
supplement
of the convoluted colonies was una¡ected by
L-lac-
tamase.
ampicillin resistance cassette within the targeted region (data not shown). Furthermore, colony dot blot hybridisation con¢rmed the absence of sequences de-
3.4. SEF17 was elaborated by cells from convoluted colonies
rived from the pERFORM C suicide vector used to create the
agfA
mutation.
It
was
di¤cult
to
remove
convoluted
colonies
from the surface of agar plates and disaggregate
3.2. S. enteritidis agf did not generate the convoluted colony phenotype
them in aqueous solutions, a property associated with
the
hydrophobic
nature
of
SEF17
¢mbriae
probably. Thus, to visualise the ¢mbriae by electron Guard-Petter
et
al.
[6]
reported
S. enteritidis
microscopy, cultures were grown directly onto the
freshly isolated from clinical samples produced an
surfaces
unusual,
our
(Hybond-N, Amersham) overlaid onto CFA agar.
hands, the convoluted colony morphology was ob-
Each wild-type strain, except strain I, produced co-
convoluted
colony
served for many strains of
morphology.
S. enteritidis,
In
including
pious
of
Formvar-coated
¢mbriae
and
grids
on
hyper£agellation
membranes
was
noted
S1400/94, when grown at 26³C for 48 h on CFA
in many cases (Fig. 2). Immunogold-labelled mono-
agar ; conditions which Sjo ë bring et al. [9] suggested
clonal antibody raised against SEF17 bound speci¢-
S. enter-
cally to the ¢mbriae, con¢rming their identity (Dibb-
promoted optimal expression of SEF17 in
itidis. Interestingly, EAV11 cultured under these con-
Fuller et al., submitted ; data not shown). Strain I
ditions gave a smooth colony morphology. To test
and all of the SEF17
whether the colony
agf
mutation nulli¢ed the convoluted
phenotype,
transduction
experiments
were
3
transductants produced no
detectable ¢mbriae when grown in an identical manner.
done to establish co-transduction between ampicillin resistance and smooth colony formation phenotypes. A P22 lysate was prepared using the EAV11 strain as a donor and used to transduce the
S. enteritidis
3.5. Congo red promoted the expression of a novel ¢mbrial structure
strains LA5, E, I and C6B and 27655R. Selection
Collinson et al. [18] demonstrated that SEF17 ex-
was made for ampicillin-resistant transductants all
pression conferred a distinctive red coloration on
of which were phenotypically smooth when grown
wild-type
on CFA agar without ampicillin at 26³C. With the
containing the hydrophobic diazo dye Congo red
exception of strain I, all of the wild-type strains pro-
whereas
duced the convoluted phenotype under these condi-
grew
tions (Fig. 1).
wild-type strains and their isogenic mutants were
a
27655 grown on CFA agar
transposon-generated
pale-pink.
To
substantiate
3
SEF17 this
mutant
¢nding,
the
grown on Congo red CFA agar at 26³C for 48 h.
3.3. L-Lactamase expression has no e¡ect on the formation of convoluted colonies It was necessary to test whether expression of
S. enteritidis
Both wild-type and mutant strains produced bright red colonies although the wild-type strains seemed
L-
more adept at leaching the dye from the agar, as shown by the formation of a pale halo around the
lactamase, the gene product of the ampicillin resist-
colony. This was not apparent for colonies of mutant
ance gene cassette used to generate the mutant, af-
strains. The colony morphologies observed on Con-
fected the ability of
S. enteritidis
to produce convo-
go red CFA agar for wild-type strains were similar to
FEMSLE 7647 20-10-97
E. Allen-Vercoe et al. / FEMS Microbiology Letters 153 (1997) 33^42
37
Fig. 1. Colonial morphologies of wild-type and mutant strains grown on CFA agar for 5 days at 26³C. Position A1, strain S1400/94 ; A2, LA5 ; A3, 27655R ; B1, EAV11 ; B2, EAV12 ; B3, EAV30 ; C1, E ; C2, I ; C3, C6B ; D1, EAV31 ; D2, EAV32 ; D3, EAV33 ; E1, S1400/94 pUC19 ; E2, LA5 pUC19.
those observed when the strains were grown on CFA
consisted of bacteria connected by networks of very
agar alone. Interestingly, mutants EAV 11, 12, 31
¢ne, long ¢brous strands (Fig. 4). These ¢laments
and 33 grew partially convoluted on the dye-contain-
resemble ¢mbriae but are morphologically di¡erent
ing agar whereas mutants EAV 30 and 32 did not
from any
(Fig. 3). When prepared for electron microscopy as
ously
before, samples of the partially convoluted colonies
against known ¢mbriae.
FEMSLE 7647 20-10-97
S. enteritidis
and
did
not
¢mbrial type described previ-
react
with
antibodies
raised
E. Allen-Vercoe et al. / FEMS Microbiology Letters 153 (1997) 33^42
38
Fig. 2. Electron micrograph of S1400/94 bacteria grown on CFA agar at 26³C for 5 days demonstrating profuse complex ¢mbriation and hyper£agellation.
4. Discussion
therefore, the phenotype was a direct result of the
3
genetic lesion. We suggest that the biological data mu-
obtained were a true comparison between isogenic
tant phenotype was the result of the insertion of an
strain variants only in the production of SEF17 ¢m-
The genetic evidence con¢rmed the SEF17
ampicillin resistance gene cassette into the
agf
region
briae and our evidence suggests that SEF17 and one
only. It may be argued that unobserved mutational
other as yet unidenti¢ed ¢mbrial antigen are at least
agf con-
partially responsible for the formation of convoluted
events distant from but co-transducible with ferred
the
phenotype.
which harboured a plasmid,
pUC19,
That
the
wild-type
strain
multicopy ampicillin resistance
remained
sharing
sequence
S. enteritidis
identity
with
harbours genes
those
encoding
the
suggested
bundle-forming pili, BFP, has been gained [19] but
gene product
their elaboration has not been observed. Whether the
did not in£uence the behaviour of the mutant and,
novel ¢mbriae described here are BFP homologues
that the presence of the
convoluted
colonies. Evidence that
L-lactamase
FEMSLE 7647 20-10-97
E. Allen-Vercoe et al. / FEMS Microbiology Letters 153 (1997) 33^42
39
Fig. 3. Colonial morphologies of wild-type and mutant strains grown at 26³C for 5 days on CFA agar with 0.001% Congo red. Strains and strain positions as for Fig. 1.
remains to be tested. Guard-Petter and co-workers [3,6] identi¢ed £agella and carbohydrates as major constituents in the extracellular matrix of convoluted colonies but did not detect SEF17 or any other ¢mbriae. Collinson et al. [8] showed the SEF17 antigen subunit was highly hydrophobic and required ex-
treme treatments such as 90% formic acid for the protein to be resolved by PAGE. This may explain why Guard-Petter and colleagues were unable to demonstrate the presence of SEF17 within the extracellular matrix. Collinson et al. [18] showed that S. enteritidis
FEMSLE 7647 20-10-97
40
E. Allen-Vercoe et al. / FEMS Microbiology Letters 153 (1997) 33^42
Fig. 4. Electron micrograph of EAV12 bacteria grown on CFA agar supplemented with 0.001% Congro red at 26³C for 5 days.
27655R grown on T medium at 37³C produced a characteristic `rough' colony morphology whereas transposon-generated agf mutants were typically smooth. We have observed that 27655R expressed SEF17 over a broad temperature range and on various media whereas other wild-type S. enteritidis strains required speci¢c temperature and media for SEF17 expression which suggested strain 27655R was probably constitutive for SEF17 expression. Analysis of the S. enteritidis 27655 genome revealed an 815 kb inversion relative to S. typhimurium [12] and that the agf region was positioned close to the break point of the inversion [20]. It is unknown
whether the chromosomes of other wild-type strains of S. enteritidis share a similar inversion relative to S. typhimurium, or to what extent, if any, this inversion a¡ected the regulation of the agf gene cluster in strain 27655. Strain I was a natural isolate which failed to express SEF17 and the genetic basis may be associated with genomic rearrangements similar to those described for 27655 or other genetic lesions a¡ecting regulation. This requires investigation. The di¡erences observed by Collinson and colleagues [18] in the coloration of SEF173 and wildtype colonies grown on Congo red-containing media could not be repeated in this laboratory and may
FEMSLE 7647 20-10-97
E. Allen-Vercoe et al. / FEMS Microbiology Letters 153 (1997) 33^42 have been associated with the speci¢c mutation we made or the presence of a compensating ¢mbrial antigen. We observed partial restoration of the con-
3
voluted phenotype by SEF17
mutants grown in the
presence of Congo red associated with the elaboration of a novel ¢mbrial structure. The role of Congo red in induction of ¢mbrial expression, the nature and biological function of the novel ¢mbriae and the interrelationship between the various structures in the extracellular matrix of convoluted colonies are unknown. Whether extracellular matrix is produced by
S. enteritidis
during
infection
is
unknown,
although Humphrey et al. [5] showed the convoluted
[2] St Louis, M.E., Morse, D.L., Potter, M.E., DeMel¢, T.M., Guzewich, J.J., Tauxe, R.V. and Blake, P.A. (1988) The emergence of grade A eggs as a major source of
itidis
Salmonella enter-
infections. New implications for the control of salmonel-
losis. J. Am. Med. Assoc. 259, 2103^2107. [3] Guard-Petter, J. (1993) Detection of two smooth colony phenotypes in a
Salmonella enteritidis
isolate which vary in their
ability to contaminate eggs. Appl. Environ. Microbiol. 59, 2884^2890. [4] Jameson, J.E. (1966) Di¡erentiation of
Salmonella
strains by
colonial morphology. J. Pathol. Bacteriol. 91, 141^148. [5] Humphrey, T.J., Williams, A., McAlpine, K., Lever, M.S., Guard-Petter, J. and Cox, J.M. (1996) Isolates of
enterica
Salmonella
enteritidis PT4 with enhanced heat and acid tolerance
are more virulent in mice and more invasive in chickens. Epidem. Infect. 117, 79^88.
S. enter-
[6] Guard-Petter, J., Keller, L.H., Mahbubur Rahman, M., Carl-
isolates to invade the reproductive organs of
son, R.W. and Silvers, S. (1996) A novel relationship between
phenotype was associated with the ability of
itidis
41
laying hens. Roles for SEF17 in cell aggregation and tissue invasion as well as enhanced survival in the presence of stomach acids have been suggested [8]. Hyper£agellation is known to be a virulence trait
O-antigen variation, matrix formation and invasiveness of
monella enteritidis.
Sal-
Epidem. Infect. 117, 219^231.
[7] Thorns, C.J. (1995) Salmonella ¢mbriae : novel antigens in the detection and control of Salmonella infections. Br. Vet. J. 151, 643^655.
Proteus mirabilis [21] and O-antigen variation may
[8] Collinson, S.K., Emo ë dy, L., Muller, K.-H., Trust, T.J. and
represent an evasive mechanism by which the bacte-
Kay, W.W. (1991) Puri¢cation and characterisation of thin,
of
ria escape non-speci¢c host immune defences [6]. Congo red binding has been associated with the vir-
aggregative ¢mbriae from
Salmonella enteritidis.
J. Bacteriol.
173, 4773^4781. è n, A. (1994) Plasminogen, [9] Sjo ë bring, U., Pohl, G. and Olse
ulence of various bacterial pathogens and the novel,
absorbed by
surface-exposed structural antigen reported in this
enteritidis
paper may represent a further Congo red-binding
vated by simultaneously captured tissue-type plasminogen ac-
protein associated with the virulence of
S. enteritidis.
Whether co-expression of these various virulence determinants is subject to global regulation remains to
Escherichia coli
expressing curli or by
Salmonella
expressing thin, aggregative ¢mbriae, can be acti-
tivator (t-PA). Mol. Microbiol. 14, 443^452. [10] Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning : A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.
be answered and it is interesting to speculate whether
[11] Saiki, R.K., Scharf, S., Faloona, F., Mullis, K.B., Horn, G.T.,
the convoluted phenotype is a correlate of virulence.
Erlich, H.A. and Arnheim, N. (1985) Enzymic ampli¢cation of
L-globin sequences and restriction site analysis of sickle cell
anaemia. Science 230, 1350^1354.
Acknowledgments
[12] Collinson, S.K., Liu, S.-L., Clouthier, S.C., Banser, P.A., Doran, J.L., Sanderson, K.E. and Kay, W.W. (1996) The
agfA, ¢mA, sefA sefD on the Salmonella enteritidis and/or Salmonella typhimurium XbaI-BlnI genomic restriction maps. Gene 169, 75^ location of four ¢mbrin-encoding genes
The authors acknowledge support from the Department of Health and gratefully acknowledge the electron microscopy skills of Bill Cooley. The authors thank Prof. T.J. Humphrey for helpful discussions.
and
80. [13] Penfold, R.J. and Pemberton, J.M. (1992) An improved suicide vector for construction of chromosomal insertion mutants in bacteria. Gene 118, 145^146. [14] Woodward, M.J., Allen-Vercoe, E. and Redstone, J. (1996) Distribution, gene sequence and expression in vivo of the
References
plasmid encoded ¢mbrial antigen of
Salmonella
serotype en-
teritidis. Epidem. Infect. 117, 17^28. [15] Anderson, R.J., Patricia London ì o, L., Pickard, D.J. and Dou-
[1] Hoop, R.K. and Popischil, A. (1993) Bacteriological, serological, histological and immunohistochemical ¢ndings in laying hens naturally infected with 4. Vet. Rec. 133, 391^393.
Salmonella enteritidis
phage-type
gan, G. (1996) Development of attenuated
Salmonella
strains
that express heterologous antigens. In : Methods in Molecular Medicine : Vaccine Protocols (Robinson, A., Farrar, G. and Wiblin, C., Eds.), pp. 47^62. Humana Press, Totowa, NJ.
FEMSLE 7647 20-10-97
E. Allen-Vercoe et al. / FEMS Microbiology Letters 153 (1997) 33^42
42
[16] Bo ë ttger, E.C. (1988) High-e¤ciency generation of plasmid
Schoolnik, G.K. (1993) Cloning and characterisation of the
cDNA libraries using electro-transformation. BioTechniques
bundle-forming pilin gene of enteropathogenic
6, 878^880.
distribution in
[17] Sojka, M.G., Dibb-Fuller, M. and Thorns, C.J. (1996) Characterisation of monoclonal antibodies speci¢c to SEF21 ¢m-
Salmonella enteritidis and their reactivity with other Salmonellae and Enterobacteria. Vet. Microbiol. 48, 207^221. briae of
[18] Collinson, S.K., Doig, P.C., Doran, J.L., Clouthier, S., Trust, T.J. and Kay, W.W. (1993) Thin, aggregative ¢mbriae mediate binding of
Salmonella enteritidis
to ¢bronectin. J. Bacteriol.
Salmonella
[19] Sohel, I., Puente, J.L., Murray, W.J., Vuopio-Varkila, J. and
and its
575.
XbaISalmonella enteritidis Salmonella typhimurium LT2.
[20] Liu, S.L., Hessel, A. and Sanderson, K.E. (1993) The
BlnI-CeuI
genomic
cleavage map of
shows an inversion relative to Mol. Microbiol. 10, 655^664.
[21] Allison, C., Lai, H.C. and Hughes, C. (1992) Coordinate expression of virulence genes during swarm cell di¡erentiation and population migration of
175, 12^18.
E. coli
serotypes. Mol. Microbiol. 7, 563^
biol. 6, 1583^1591.
FEMSLE 7647 20-10-97
Proteus mirabilis.
Mol. Micro-
SEF17 ¢mbriae are essential for the convoluted colonial morphology of Salmonella enteritidis Emma Allen-Vercoe, Mike Dibb-Fuller, Christopher J. Thorns, Martin J. Woodward* Department of Bacteriology, Central Veterinary Laboratory, Woodham Lane, New Haw, Addlestone, Surrey KT15 3NB, UK
Received 7 March 1997; revised 29 April 1997; accepted 2 May 1997
Abstract
Salmonella enteritidis isolated from poultry infections generated a convoluted colonial morphology after 48 h growth on colonisation factor antigen (CFA) agar at 25³C. A mutant S. enteritidis defective for the elaboration of the SEF17 fimbrial antigen, in which the agf gene cluster was inactivated by insertion of an ampicillin resistance gene cassette, and other wild-type 3 mutants at 25³C S. enteritidis transduced to this genotype failed to produce convoluted colonies. However, growth of SEF17 on CFA agar supplemented with 0.001% Congo red resulted in partial recovery of the phenotype. Immunoelectron microscopy demonstrated that copious amounts of the SEF17 fimbrial antigen were present in the extracellular matrix of convoluted colonies of wild-type virulent S. enteritidis isolates. Bacteria were often hyperflagellated also. Immunoelectron microscopy of SEF173 mutants grown on CFA agar+0.001% Congo red demonstrated the elaboration of an as yet undefined fimbrial structure. Isolates of S. enteritidis which were described previously as avirulent and sensitive to environmental stress failed to express SEF17 or produce convoluted colonies. These data indicate an essential role for SEF17, and possibly for another fimbria and flagella, in the generation of the convoluted colonial phenotype. The relationship between virulence and colonial phenotype is discussed.
Keywords : Salmonella enteritidis
; Salmonellosis; Fimbriae; Colony morphology; SEF17
1. Introduction
Poultry has been cited as an important reservoir of associated with human infection and food safety measures have been hampered by the ability of S. enteritidis to cause invasive, subclinical infection in farmed chicken £ocks that may be di¤cult to detect [1]. In particular, the apparent Salmonella enteritidis
* Corresponding author. Tel.: +44 (1932) 357582; Fax: +44 (1932) 347046; E-mail: [email protected]
tropism that S. enteritidis has for the reproductive tissues of laying hens results in vertical transfer [2]. Guard-Petter [3] observed that isolates of S. enteritidis di¡ered in their ability to infect laying hens and contaminate eggs and that this ability was associated with the production of `lacy' colonies, a phenotype ¢rst described by Jameson [4] as `convoluted' in reports of colonial morphology of S. typhimurium isolates. We will use the term convoluted throughout this paper. Humphrey et al. [5] showed that isolates of S. enteritidis of the same phage type di¡er not only in their pathogenic potential for mice and
0378-1097 / 97 / $17.00 ß 1997 Federation of European Microbiological Societies. Published by Elsevier Science B.V. PII S 0 3 7 8 - 1 0 9 7 ( 9 7 ) 0 0 2 1 9 - X
FEMSLE 7647 20-10-97
34
E. Allen-Vercoe et al. / FEMS Microbiology Letters 153 (1997) 33^42
chickens but also in their tolerance to certain environmental stresses. These authors noted that the tolerant/virulent phenotype was found to correlate with the elaboration of the convoluted phenotype which was found to be inherently stable, requiring no special culturing conditions for enhancement, and suggested that such traits may be genetically determined by genes controlling the expression of various extracellular components. This suggestion was supported by Guard-Petter et al. [6] who demonstrated the convoluted phenotype of S. enteritidis was associated with hyper£agellation and the formation of high-molecular-mass O-antigen although ¢mbriae were not detected. S. enteritidis has the potential to elaborate at least six di¡erent ¢mbrial types [7] and we have used a reversed genetics approach to insertionally inactivate genes encoding the elaboration of ¢mbriae and £agella in order to study their role in S. enteritidis pathogenesis. We report that S. enteritidis phage type 4 mutants unable to synthesise the thin, aggregative, hydrophobic ¢mbriae, SEF17 [8], through inactivation of the agfA gene as well as a naturally occurring avirulent S. enteritidis were unable to form convoluted colonies whereas virulent wild-types isolates did. We present evidence that SEF17 is expressed by bacteria elaborating convoluted colonies and that another hitherto undescribed ¢mbria may be associated with this phenotype also. 2. Materials and methods
2.1. Bacterial strains
phage type 4 strains S1400/94 and LA5 are wild-type ¢eld isolates from infected chickens. Strains EAV11 and EAV12 are SEF173 mutants of S1400/94 and LA5, respectively. S. enteritidis strains E, I [5] and C6B were kindly provided by Prof. T.J. Humphrey of the Food Microbiology Research Unit, Public Health Laboratory, Exeter. Strain C6B was a recently isolated strain that showed extreme tolerance of heat, acid and oxidative stress (T.J. Humphrey, personal communication). Strains EAV31, EAV32 and EAV33 are SEF173 mutants of E, I and C6B respectively. S. enteritidis strain 27655R [8] was a kind gift from Prof. T. Wadstrom of Lund University, Sweden, which displayed conS. enteritidis
stitutive expression of SEF17. Strain EAV30 is a SEF173 mutant of 27655R. 2.2. Growth conditions
All strains were maintained on Dorset's egg slopes at 20³C. The convoluted phenotype was promoted by growth of the strains on colonisation factor antigen (CFA) agar at 25³C for approximately 48 h, conditions previously suggested to achieve maximal expression of thin, aggregative ¢mbriae [9]. Congo red (Sigma) was added to this medium to a ¢nal concentration of 0.001% (v/v) and ampicillin (50 Wg ml31 ) was added when appropriate. 2.3. Genetic manipulation methods
Unless stated otherwise, all genetic manipulation methods were as described by Sambrook et al. [10]. 2.4. Polymerase chain reaction
PCR was performed essentially following the methods of Saiki et al. [11] with primers designed on the agf gene cluster using the sequence data of Collinson et al. [12] (GenBank accession number U43280) and were: AGF1F
3P-ATG ATG TTG ACA ATA CTG GGT-5P AGF1R 3P-TGA TAA ATG CAG TGA TTG TCC-5P
position 716^737 position 2009^1988.
2.5. Construction of S. enteritidis S1400/94 agf
The agf region was ampli¢ed by PCR using DNA extracted from S1400/94 as a template with primer pair AGF1F and AGF1R. The product of the predicted size (1.3 kb) was obtained and cloned directly into pCRscript (Stratagene) to generate recombinant plasmid pVW680. A unique EheI site within agfA was con¢rmed into which an ampicillin resistance gene cassette (Stratagene) was cloned to generate recombinant plasmid pVW680A. A 2.2 kb agfA ::ampr fragment was ampli¢ed with primer pair AGF1F and AGF1R with puri¢ed pVW680A as template and was ligated with suicide vector pERFORM C (Allen-Vercoe and Woodward, in preparation), a
FEMSLE 7647 20-10-97
E. Allen-Vercoe et al. / FEMS Microbiology Letters 153 (1997) 33^42
chloramphenicol-resistant suicide vector derived from pJP5603 [13]. The ligation mixture was used to transform E. coli CC118 V-pir with selection made jointly for chloramphenicol- and ampicillin-resistant recombinants. The restriction map of plasmid DNA from one recombinant was con¢rmed and the construct was designated pVW681AC. In order to mobilise pVW681AC into S. enteritidis S1400/94 by conjugation, pVW681AC was transformed into E. coli K12 strain S-17 V-pir, an auxotrophic mobilising strain. Conjugal pro¢ciency was supplied in trans by the chromosomally inserted incP mini-plasmid, pRP4, and a mating of the form E. coli K12 S-17 V-pir pVW681ACUS. enteritidis S1400/94 was set up with selection made on glucose minimal medium supplemented with ampicillin and chloramphenicol. Exconjugants were obtained at a frequency of 1U1038 and one was subcultured in Luria-Bertani broth without antibiotic selective pressure. An ampicillin-resistant, chloramphenicol-sensitive derivative was selected in which the native chromosomal agf gene cluster had been replaced by homologous recombination with the agf ::ampr construct with consequent loss of the chloramphenicol-marked suicide vector and designated EAV11. 2.6. Hybridisation
DNA probes were prepared by PCR as described above and radioisotopically labelled with 32 P using `Ready to Go' dCTP labelling kit (Pharmacia) following the manufacturer's recommendations. Colony dot blot and Southern hybridisation, hybridisation conditions and post-hybridisation washes were as described previously [14].
35
mately 0.1 Wg of pUC19 DNA (New England Biolabs) and pulsed in a BioRad Gene Pulser electroporator using the settings 25 WF capacitance, 400 6 resistance, and electrical potential of 2 kV, conditions known to give maximal e¤ciency for transformation of this strain. Cells were allowed to recover for 1 h in SOC medium (Gibco-BRL) before spreading onto selection plates. A single isolate was tested for the presence of the pUC19 plasmid using the `Perfect Prep' plasmid DNA extraction kit (5PC3P, supplied by CP Labs) and was designated S1400/94/ pUC19. 2.9. Electron microscopy
Cultures were grown onto Formvar-coated grids (CVL Addlestone) placed on uncharged Hybond-N membrane (Amersham) overlaid on agar plates. After incubation, the grids were allowed to soak in ¢lter-sterile 1UPBS for 5 min in order to loosen the cells so that excess could be easily removed. The grids were stained with 2% phosphotungstic acid for 15 s prior to viewing with a Philips TM10 transmission electron microscope. 2.10. Immunoelectron microscopy
Immunoelectron microscopy was carried out on grids, onto which cells had been grown as described above, using a monoclonal antibody raised against SEF17 (Dibb-Fuller et al., submitted) and Auroprobe (Amersham) gold conjugate. The procedure described by Sojka et al. [17] was used. 3. Results
2.7. P22 transduction
The preparation of a P22 lysate of strain EAV11 and its use in the transduction of the agfA ::ampr genotype was carried out essentially using the method of Anderson et al. [15]. 2.8. pUC19 transformation
Wild-type S. enteritidis S1400/1994 was prepared for electroporation using the method of Boëttger [16]. Electrocompetent cells were mixed with approxi-
3.1. Characterisation of an agf mutant of S. enteritidis strain S1400/1994 S. enteritidis strain S1400/94 was chosen in the ¢rst instance to generate a SEF173 mutant. To do this, the agf gene cluster, encoding the elaboration of SEF17 ¢mbriae at the cell surface, was disrupted by the insertion of an ampicillin resistance gene cassette (see Section 2) and one ampicillin-resistant agf mutant designated EAV11 was selected for further analysis. Southern hybridisation experiments were
FEMSLE 7647 20-10-97
E. Allen-Vercoe et al. / FEMS Microbiology Letters 153 (1997) 33^42
36
r
done to con¢rm the genetic lesion was located within
luted
agf operon. Total genomic DNA was extracted from S. enteritidis strain S1400/94 and the isogenic agf3 derivative, digested to completion in separate experiments with SspI, SalI and HpaI, resolved by
introduced into wild-type S1400/94 and LA5 cells
agarose
which supported the hypothesis that the formation
the
gel
electrophoresis,
transferred
membranes and hybridised with the
agf
to
nylon
gene probe.
Hybridisation data con¢rmed insertion of the 0.9 kb
by
colonies.
To
electroporation.
do
All
this,
was
pUC19amp
transformants
produced
convoluted colonies when grown at 26³C on CFA agar
with
and
without
ampicillin
supplement
of the convoluted colonies was una¡ected by
L-lac-
tamase.
ampicillin resistance cassette within the targeted region (data not shown). Furthermore, colony dot blot hybridisation con¢rmed the absence of sequences de-
3.4. SEF17 was elaborated by cells from convoluted colonies
rived from the pERFORM C suicide vector used to create the
agfA
mutation.
It
was
di¤cult
to
remove
convoluted
colonies
from the surface of agar plates and disaggregate
3.2. S. enteritidis agf did not generate the convoluted colony phenotype
them in aqueous solutions, a property associated with
the
hydrophobic
nature
of
SEF17
¢mbriae
probably. Thus, to visualise the ¢mbriae by electron Guard-Petter
et
al.
[6]
reported
S. enteritidis
microscopy, cultures were grown directly onto the
freshly isolated from clinical samples produced an
surfaces
unusual,
our
(Hybond-N, Amersham) overlaid onto CFA agar.
hands, the convoluted colony morphology was ob-
Each wild-type strain, except strain I, produced co-
convoluted
colony
served for many strains of
morphology.
S. enteritidis,
In
including
pious
of
Formvar-coated
¢mbriae
and
grids
on
hyper£agellation
membranes
was
noted
S1400/94, when grown at 26³C for 48 h on CFA
in many cases (Fig. 2). Immunogold-labelled mono-
agar ; conditions which Sjo ë bring et al. [9] suggested
clonal antibody raised against SEF17 bound speci¢-
S. enter-
cally to the ¢mbriae, con¢rming their identity (Dibb-
promoted optimal expression of SEF17 in
itidis. Interestingly, EAV11 cultured under these con-
Fuller et al., submitted ; data not shown). Strain I
ditions gave a smooth colony morphology. To test
and all of the SEF17
whether the colony
agf
mutation nulli¢ed the convoluted
phenotype,
transduction
experiments
were
3
transductants produced no
detectable ¢mbriae when grown in an identical manner.
done to establish co-transduction between ampicillin resistance and smooth colony formation phenotypes. A P22 lysate was prepared using the EAV11 strain as a donor and used to transduce the
S. enteritidis
3.5. Congo red promoted the expression of a novel ¢mbrial structure
strains LA5, E, I and C6B and 27655R. Selection
Collinson et al. [18] demonstrated that SEF17 ex-
was made for ampicillin-resistant transductants all
pression conferred a distinctive red coloration on
of which were phenotypically smooth when grown
wild-type
on CFA agar without ampicillin at 26³C. With the
containing the hydrophobic diazo dye Congo red
exception of strain I, all of the wild-type strains pro-
whereas
duced the convoluted phenotype under these condi-
grew
tions (Fig. 1).
wild-type strains and their isogenic mutants were
a
27655 grown on CFA agar
transposon-generated
pale-pink.
To
substantiate
3
SEF17 this
mutant
¢nding,
the
grown on Congo red CFA agar at 26³C for 48 h.
3.3. L-Lactamase expression has no e¡ect on the formation of convoluted colonies It was necessary to test whether expression of
S. enteritidis
Both wild-type and mutant strains produced bright red colonies although the wild-type strains seemed
L-
more adept at leaching the dye from the agar, as shown by the formation of a pale halo around the
lactamase, the gene product of the ampicillin resist-
colony. This was not apparent for colonies of mutant
ance gene cassette used to generate the mutant, af-
strains. The colony morphologies observed on Con-
fected the ability of
S. enteritidis
to produce convo-
go red CFA agar for wild-type strains were similar to
FEMSLE 7647 20-10-97
E. Allen-Vercoe et al. / FEMS Microbiology Letters 153 (1997) 33^42
37
Fig. 1. Colonial morphologies of wild-type and mutant strains grown on CFA agar for 5 days at 26³C. Position A1, strain S1400/94 ; A2, LA5 ; A3, 27655R ; B1, EAV11 ; B2, EAV12 ; B3, EAV30 ; C1, E ; C2, I ; C3, C6B ; D1, EAV31 ; D2, EAV32 ; D3, EAV33 ; E1, S1400/94 pUC19 ; E2, LA5 pUC19.
those observed when the strains were grown on CFA
consisted of bacteria connected by networks of very
agar alone. Interestingly, mutants EAV 11, 12, 31
¢ne, long ¢brous strands (Fig. 4). These ¢laments
and 33 grew partially convoluted on the dye-contain-
resemble ¢mbriae but are morphologically di¡erent
ing agar whereas mutants EAV 30 and 32 did not
from any
(Fig. 3). When prepared for electron microscopy as
ously
before, samples of the partially convoluted colonies
against known ¢mbriae.
FEMSLE 7647 20-10-97
S. enteritidis
and
did
not
¢mbrial type described previ-
react
with
antibodies
raised
E. Allen-Vercoe et al. / FEMS Microbiology Letters 153 (1997) 33^42
38
Fig. 2. Electron micrograph of S1400/94 bacteria grown on CFA agar at 26³C for 5 days demonstrating profuse complex ¢mbriation and hyper£agellation.
4. Discussion
therefore, the phenotype was a direct result of the
3
genetic lesion. We suggest that the biological data mu-
obtained were a true comparison between isogenic
tant phenotype was the result of the insertion of an
strain variants only in the production of SEF17 ¢m-
The genetic evidence con¢rmed the SEF17
ampicillin resistance gene cassette into the
agf
region
briae and our evidence suggests that SEF17 and one
only. It may be argued that unobserved mutational
other as yet unidenti¢ed ¢mbrial antigen are at least
agf con-
partially responsible for the formation of convoluted
events distant from but co-transducible with ferred
the
phenotype.
which harboured a plasmid,
pUC19,
That
the
wild-type
strain
multicopy ampicillin resistance
remained
sharing
sequence
S. enteritidis
identity
with
harbours genes
those
encoding
the
suggested
bundle-forming pili, BFP, has been gained [19] but
gene product
their elaboration has not been observed. Whether the
did not in£uence the behaviour of the mutant and,
novel ¢mbriae described here are BFP homologues
that the presence of the
convoluted
colonies. Evidence that
L-lactamase
FEMSLE 7647 20-10-97
E. Allen-Vercoe et al. / FEMS Microbiology Letters 153 (1997) 33^42
39
Fig. 3. Colonial morphologies of wild-type and mutant strains grown at 26³C for 5 days on CFA agar with 0.001% Congo red. Strains and strain positions as for Fig. 1.
remains to be tested. Guard-Petter and co-workers [3,6] identi¢ed £agella and carbohydrates as major constituents in the extracellular matrix of convoluted colonies but did not detect SEF17 or any other ¢mbriae. Collinson et al. [8] showed the SEF17 antigen subunit was highly hydrophobic and required ex-
treme treatments such as 90% formic acid for the protein to be resolved by PAGE. This may explain why Guard-Petter and colleagues were unable to demonstrate the presence of SEF17 within the extracellular matrix. Collinson et al. [18] showed that S. enteritidis
FEMSLE 7647 20-10-97
40
E. Allen-Vercoe et al. / FEMS Microbiology Letters 153 (1997) 33^42
Fig. 4. Electron micrograph of EAV12 bacteria grown on CFA agar supplemented with 0.001% Congro red at 26³C for 5 days.
27655R grown on T medium at 37³C produced a characteristic `rough' colony morphology whereas transposon-generated agf mutants were typically smooth. We have observed that 27655R expressed SEF17 over a broad temperature range and on various media whereas other wild-type S. enteritidis strains required speci¢c temperature and media for SEF17 expression which suggested strain 27655R was probably constitutive for SEF17 expression. Analysis of the S. enteritidis 27655 genome revealed an 815 kb inversion relative to S. typhimurium [12] and that the agf region was positioned close to the break point of the inversion [20]. It is unknown
whether the chromosomes of other wild-type strains of S. enteritidis share a similar inversion relative to S. typhimurium, or to what extent, if any, this inversion a¡ected the regulation of the agf gene cluster in strain 27655. Strain I was a natural isolate which failed to express SEF17 and the genetic basis may be associated with genomic rearrangements similar to those described for 27655 or other genetic lesions a¡ecting regulation. This requires investigation. The di¡erences observed by Collinson and colleagues [18] in the coloration of SEF173 and wildtype colonies grown on Congo red-containing media could not be repeated in this laboratory and may
FEMSLE 7647 20-10-97
E. Allen-Vercoe et al. / FEMS Microbiology Letters 153 (1997) 33^42 have been associated with the speci¢c mutation we made or the presence of a compensating ¢mbrial antigen. We observed partial restoration of the con-
3
voluted phenotype by SEF17
mutants grown in the
presence of Congo red associated with the elaboration of a novel ¢mbrial structure. The role of Congo red in induction of ¢mbrial expression, the nature and biological function of the novel ¢mbriae and the interrelationship between the various structures in the extracellular matrix of convoluted colonies are unknown. Whether extracellular matrix is produced by
S. enteritidis
during
infection
is
unknown,
although Humphrey et al. [5] showed the convoluted
[2] St Louis, M.E., Morse, D.L., Potter, M.E., DeMel¢, T.M., Guzewich, J.J., Tauxe, R.V. and Blake, P.A. (1988) The emergence of grade A eggs as a major source of
itidis
Salmonella enter-
infections. New implications for the control of salmonel-
losis. J. Am. Med. Assoc. 259, 2103^2107. [3] Guard-Petter, J. (1993) Detection of two smooth colony phenotypes in a
Salmonella enteritidis
isolate which vary in their
ability to contaminate eggs. Appl. Environ. Microbiol. 59, 2884^2890. [4] Jameson, J.E. (1966) Di¡erentiation of
Salmonella
strains by
colonial morphology. J. Pathol. Bacteriol. 91, 141^148. [5] Humphrey, T.J., Williams, A., McAlpine, K., Lever, M.S., Guard-Petter, J. and Cox, J.M. (1996) Isolates of
enterica
Salmonella
enteritidis PT4 with enhanced heat and acid tolerance
are more virulent in mice and more invasive in chickens. Epidem. Infect. 117, 79^88.
S. enter-
[6] Guard-Petter, J., Keller, L.H., Mahbubur Rahman, M., Carl-
isolates to invade the reproductive organs of
son, R.W. and Silvers, S. (1996) A novel relationship between
phenotype was associated with the ability of
itidis
41
laying hens. Roles for SEF17 in cell aggregation and tissue invasion as well as enhanced survival in the presence of stomach acids have been suggested [8]. Hyper£agellation is known to be a virulence trait
O-antigen variation, matrix formation and invasiveness of
monella enteritidis.
Sal-
Epidem. Infect. 117, 219^231.
[7] Thorns, C.J. (1995) Salmonella ¢mbriae : novel antigens in the detection and control of Salmonella infections. Br. Vet. J. 151, 643^655.
Proteus mirabilis [21] and O-antigen variation may
[8] Collinson, S.K., Emo ë dy, L., Muller, K.-H., Trust, T.J. and
represent an evasive mechanism by which the bacte-
Kay, W.W. (1991) Puri¢cation and characterisation of thin,
of
ria escape non-speci¢c host immune defences [6]. Congo red binding has been associated with the vir-
aggregative ¢mbriae from
Salmonella enteritidis.
J. Bacteriol.
173, 4773^4781. è n, A. (1994) Plasminogen, [9] Sjo ë bring, U., Pohl, G. and Olse
ulence of various bacterial pathogens and the novel,
absorbed by
surface-exposed structural antigen reported in this
enteritidis
paper may represent a further Congo red-binding
vated by simultaneously captured tissue-type plasminogen ac-
protein associated with the virulence of
S. enteritidis.
Whether co-expression of these various virulence determinants is subject to global regulation remains to
Escherichia coli
expressing curli or by
Salmonella
expressing thin, aggregative ¢mbriae, can be acti-
tivator (t-PA). Mol. Microbiol. 14, 443^452. [10] Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning : A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.
be answered and it is interesting to speculate whether
[11] Saiki, R.K., Scharf, S., Faloona, F., Mullis, K.B., Horn, G.T.,
the convoluted phenotype is a correlate of virulence.
Erlich, H.A. and Arnheim, N. (1985) Enzymic ampli¢cation of
L-globin sequences and restriction site analysis of sickle cell
anaemia. Science 230, 1350^1354.
Acknowledgments
[12] Collinson, S.K., Liu, S.-L., Clouthier, S.C., Banser, P.A., Doran, J.L., Sanderson, K.E. and Kay, W.W. (1996) The
agfA, ¢mA, sefA sefD on the Salmonella enteritidis and/or Salmonella typhimurium XbaI-BlnI genomic restriction maps. Gene 169, 75^ location of four ¢mbrin-encoding genes
The authors acknowledge support from the Department of Health and gratefully acknowledge the electron microscopy skills of Bill Cooley. The authors thank Prof. T.J. Humphrey for helpful discussions.
and
80. [13] Penfold, R.J. and Pemberton, J.M. (1992) An improved suicide vector for construction of chromosomal insertion mutants in bacteria. Gene 118, 145^146. [14] Woodward, M.J., Allen-Vercoe, E. and Redstone, J. (1996) Distribution, gene sequence and expression in vivo of the
References
plasmid encoded ¢mbrial antigen of
Salmonella
serotype en-
teritidis. Epidem. Infect. 117, 17^28. [15] Anderson, R.J., Patricia London ì o, L., Pickard, D.J. and Dou-
[1] Hoop, R.K. and Popischil, A. (1993) Bacteriological, serological, histological and immunohistochemical ¢ndings in laying hens naturally infected with 4. Vet. Rec. 133, 391^393.
Salmonella enteritidis
phage-type
gan, G. (1996) Development of attenuated
Salmonella
strains
that express heterologous antigens. In : Methods in Molecular Medicine : Vaccine Protocols (Robinson, A., Farrar, G. and Wiblin, C., Eds.), pp. 47^62. Humana Press, Totowa, NJ.
FEMSLE 7647 20-10-97
E. Allen-Vercoe et al. / FEMS Microbiology Letters 153 (1997) 33^42
42
[16] Bo ë ttger, E.C. (1988) High-e¤ciency generation of plasmid
Schoolnik, G.K. (1993) Cloning and characterisation of the
cDNA libraries using electro-transformation. BioTechniques
bundle-forming pilin gene of enteropathogenic
6, 878^880.
distribution in
[17] Sojka, M.G., Dibb-Fuller, M. and Thorns, C.J. (1996) Characterisation of monoclonal antibodies speci¢c to SEF21 ¢m-
Salmonella enteritidis and their reactivity with other Salmonellae and Enterobacteria. Vet. Microbiol. 48, 207^221. briae of
[18] Collinson, S.K., Doig, P.C., Doran, J.L., Clouthier, S., Trust, T.J. and Kay, W.W. (1993) Thin, aggregative ¢mbriae mediate binding of
Salmonella enteritidis
to ¢bronectin. J. Bacteriol.
Salmonella
[19] Sohel, I., Puente, J.L., Murray, W.J., Vuopio-Varkila, J. and
and its
575.
XbaISalmonella enteritidis Salmonella typhimurium LT2.
[20] Liu, S.L., Hessel, A. and Sanderson, K.E. (1993) The
BlnI-CeuI
genomic
cleavage map of
shows an inversion relative to Mol. Microbiol. 10, 655^664.
[21] Allison, C., Lai, H.C. and Hughes, C. (1992) Coordinate expression of virulence genes during swarm cell di¡erentiation and population migration of
175, 12^18.
E. coli
serotypes. Mol. Microbiol. 7, 563^
biol. 6, 1583^1591.
FEMSLE 7647 20-10-97
Proteus mirabilis.
Mol. Micro-