Development of a DNA probe to detect Salmonella typhi

(1)

0095-1137/85/100600-06$02.00/0

Development

of

a

DNA Probe

to

Detect

Salmonella

typhi

FRAN A. RUBIN,* DENNIS J. KOPECKO, KERRY F. NOON, AND LOUIS S. BARON

Departmentof BacterialImmunology, Walter ReedArmyInstitute of Research, Washington,D.C. 20307-5100 Received 11 March 1985/Accepted 26 June 1985

This study was undertaken to identify a DNA sequence that could be used to facilitate the diagnostic identificationof Salmonella typhi, the causativeagentof typhoid fever. All virulentS. typhistrainsencodea

relativelyunique capsularantigentermedthe virulence (Vi)antigen. Two distinct genetic loci, viaAand viaB,

areinvolved inthesynthesis of this antigen. Thestructuralgenes,locatedatviaB,wereconsideredasapossible

specific DNAprobe. The viaB locus, contained in arecombinantcosmid, was subclonedto variousplasmid vectorsforthispurpose.Selected viaB-regionDNAfragmnentswerethenanalyzed for specificity inDNA colony hybridization reactionswithmorethan 170 strains representingavariety of entericbacteria. An8.6-kilobase EcoRI fragment washighlyspecificfortheviaBgene regionandwasconsidered agoodhybridization probe.

ThisDNA probe should prove useful in rapid diagnostic assays set up todetect S. typhiin mixedbacterial

samples(e.g.,stools) within afew hours of specimen collection.

AllstrainsofSalmonellatyphiand S.paratyphi C, as well

as a few atypical but genetically related Citrobacter and Salmonella strains, are capable of synthesizing a capsular

antigen termed Vi for virulence (6). This galactosamine

uronic acid polymer has been associated with the virulence

of the organism (8, 9). Two separate chromosomal loci necessary for Vi antigen expression, viaA and viaB, have

been identified in genetic studies of S. typhi (13, 14). The

viaBregion appears to encodethe structural genes forthis

antigen(13).Analogous andpresumablyallelic chromosomal

siteshave beenidentified inS.paratyphi C (26) and in some

strains ofCitrobacterfreundii(27).Althoughtheexpression

oftheVi antigen is relatively stablein S. typhi, Vi-positive

Citrobacter strains exhibit a rapid, reversible transition

between forms that express the Vi antigen and forms that

appearnot toexpressit,referredto asnon-ViorWforms(1,

27).

In previous studies, the viaB region of C.freundii was

cosmid cloned in an attempt to elucidate the phenomenon

involved in the reversible genetic expression of the Vi

antigen (L. S. Baron, D. J. Kopecko, E. M. Johnson, K.

Noon, N. Snellings, and C.A. Life, Abstr. Annu. Meet.

Am. Soc. Microbiol. 1983, H116, p. 125; manuscript in

preparation). Thepresentstudywasaimedatdetertnining if

a segment of the viaB gene region could be used as a

typhoid-specific DNA probe for

diagnostic

purposes. DNA

fromthe viaB region ofC.freundiiwas subclonedfromthe

previously isolated cosmid recombinantintoseveral

plasmid

vectors.An8.6-kilobase (kb) EcoRIfragment

recognized

the

viaB locus specifically when used in colony

hybridization

reactions with avariety of enteric strains.

MATERIALS ANDMETHODS

Bacterial strains and plasmids. Bacterial strains and

plasmidsarelisted inTables1and 2. S.typhiWR4201

(ViaA+

ViaB+) expresses Vi antigen; previously constructed

deriv-atives WR4205 (13) and WR4226 (27) were used as DNA

hybridization controls since they are ViaA-

ViaB+

and

ViaA'

ViaB-,

respectively.

Escherichia coli

WR2376,

a

Vi-positive E. coli C600recombinant

carrying

the viaB

lo-cus of C. freundii WR7004 (2), was also used as a DNA

hybridization control in some experiments. Salmonella

*

Corresponding

author.

strains from the Centers for Disease Control (CDC), Atlanta,

Ga.,used to determine probe specificity included groups A,

B,

C1,

C2, C3,

D1,

D2,

E1,

E2,

E3, E4,

F, G1,

G2,

H, I, J, K,

L,M, N,0, P, Q,R,S, T, U, V, W, X, Y, Z, 51, 52, 53, 54,

55,66,and 67.Additionalbacterial strains were obtained from

the collection at Walter Reed Army Institute of Research

(WRAIR), Washington,D.C.

Media andcultureconditions.Bacteria were grown at 37°C

onnutrient agarorin Penassay orbrainheartinfusionbroth

(Difco Laboratories, Detroit, Mich.).

Antibiotics

were used

at thefollowingfinal concentrations: kanamycin, 20

pug/ml;

tetracycline, 10

,ug/ml;

chloramphenicol,

20 ,ug/ml;

spec-tinomycin,25 ug/ml;andampicillin, 25 ,ug/ml.

Viantigen expression. Viantigen-expressing bacterial

col-onies on agar media were identified microscopically by

oblique

illumination (27). Vi antigen-expressing forms are seen as dense, bright, orange-tinted colonies which are

readily distinguishablefrom thedull, translucentcolonies of

non-Vi forms. Vi antigen expression was verified by slide

agglutination with rabbit antiserum prepared against

Vi-encapsulated C.freundii WR7004 cells.

An

additional test

for Vi antigen expression involved the sensitivity of Vi

antigen-expressingcells to

Vi-specific

typing phage. Adrop

ofVi phagewas spotted onan areaofanutrient agar

plate

that was heavily swabbed with a bacterial culture. After

overnight incubationat 37°C, cell

lysis

was observed

in

the

spotted area only in the case

of

cells

expressing

the Vi

antigen.

Isolation and manipulation of DNA. Bacterial cells were

grownat37°C for16to18 h inPenassay broth. PlasmidDNA

was isolated by a cleared

lysis

method with Triton X-100

detergent followedby

plasmid

purification

on cesium chlo-ride

density

gradients

(18).

Digestion

ofDNA with

restric-tion endonucleases was carried out under the

conditions

specified bythe vendor(New

England

BioLabs, Inc.,

Bev-erly, Mass.; International

Biotechnologies, Inc.,

New

Ha-ven, Conn.). Plasmids and restriction

endonuclease-generated DNA fragments were resolved and

analyzed

by

horizontal gel

electrophoresis

(International

Biotechnolo-gies)in 0.7to 2.0%agarose (SeaKem;FMC

Corp.,

Marine

ColloidsDiv.,

Rockland,

Maine;

International

Biotechnolo-gies)prepared in TBEbuffer(89mMTris

[pH 8.3],

2.5 mM

EDTA, 89 mM boric

acid).

We visualized DNA bands

by

stainingthegelin 0.5 ,ugof aqueous ethidiumbromideperml

600

on April 10, 2020 by guest

http://jcm.asm.org/

(2)

TABLE 1. Plasmidcloningvectors

Plasmid Size(kb) Relevant characteristics" Source

pHC79 6.5 Apr(Pstl), Tcr B. Hohn(12) pDPT429b 8.7 Cmr(EcoRI), Spr D. Taylor' pRK290 20.0 Tcr(EcoRl) D. Helinski(4) pACKC1 4.0 Cmr(EcoRI), Kmr V. Burdett" pBR325 6.0 Cmr(EcoRI), Apr, Tcr 3

aSingle restriction sitesthat inactivatedrugresistanceinthevectorsare included in parentheses next to the appropriate antibiotic resistance. Ap', ampicillin resistant;Tcr, tetracycline resistant; Cmr, chloramphenicol resis-tant;spr,spectinomycinresistant;andKmr, kanamycin resistant.

bSingle-copyvectorderived fromplasmidR100.

' SmithKline BeckmanCorp., Philadelphia.Pa. dDukeUniversity, Durham,N.C.

and then illuminating it with a 300-nm UV light source

(Fotodyne, New Berlin, Wis.).

Recombinant plasmids were constructed in vitro by

liga-tion, with T4 DNA ligase, ofendonuclease-linearized vector

DNA toendonuclease-generated DNAfragments (New En-gland BioLabs) at 17°C for 16 to 18 h with the buffer

described by Maniatis et al. (22). E. (oli HB101 cells were

prepared

for transformation with

plasmid

DNA

by

the method of Kushner

(19).

Preparation of32P-labeled DNA probes. To purify DNA

fragments

foruse as

probes

in

hybridization

experiments,we

digested plasmids

with the selected restriction endonu-cleases and resolved the

resulting fragments

by

agarose

(International Biotechnologies)

gel electrophoresis.

After ethidium bromide

staining

of the

gel,

the appropriate DNA bandwas cut out andthe DNA waselectroeluted with(i) a concentrator

(model 1750;

ISCO,

Lincoln, Nebr.)

or

(ii)

a

dialysis

membrane filled with theagaroseslice and TEbuffer

(0.01

MTris

[pH 8.0],

0.001 M

EDTA)

with 0.1x TBEbuffer

surrounding the membrane (100 V for 2 h followed

by

reversed current for 2

min)

(22). Fragments were further

purified

by

another round of agarose gel

electrophoresis

followed

by

electroelution. Vi

antigen

genelocus

fragments

wereradiolabeled in vitro

by

nick translation (25) witha kit

from New England Nuclear Corp., Boston, Mass.

([cc-32P]dCTP [3,000

Ci/mmol]).

After 1 hat 14WC,6 plIof 0.3 M

EDTAwasaddedtoterminate the reaction.

Unincorporated

nucleotides were separated from labeled DNA by

centrifu-gation through a 1-ml Sephadex G-50 column

equilibrated

andrunwith0.2% sodium

dodecyl

sulfate

(SDS)-0.1

MNaCl

TABLE 2. Summaryof in situcolonyhybridization experiments

Responseto probe:"

Bacterialspecies Strains tested Source

EcoRI-A EcoRI-B

C.freundii

WR7004Vi+ WRAIR ++++

C.

freundii

4182-83 CDC - +

C.freundii Fivestrains CDC

C. div'ersus Five strains CDC - +

C. amalonaticus Fivestrains CDC - +

S.

typhi

WR4201 (ViaA+ WRAIR ++ ++ +

ViaB+)

S.

typhi

WR4205(ViaA-- WRAIR ++++ + ViaB+)

S. typhi WR4226 (ViaA+ WRAIR

ViaB )

S.

typhi

Ty2 WRAIR ++++

S. tvphi Six strains CDC + + ++ +

S.

paratiyphi

C Two strains CDC ++++ ++++

S.dublin

Vi+

17-59 andaCDCstrain 19 + + + + ++++

S.

typhimurium

C-5 17

TML WRAIR

-CDC strain CDC

Salmonella spp. 130CDC strains' CDC - +'

E. coliK-12 AB313 E. Adelberg

-HB101 H.Boyer

-52R137 (LT+) 23

-E. coli 218(018:K1) R. Silver - +

437 (04:K12) R. Silver - +

439(K92) R. Silver - +

440(086:K2) R.Silver - +

441 (015:K7) R. Silver - +

442 (K15) R.Silver - +

501 (075:K100) R. Silver - +

S.flexneri Serotype lb,M25-8A WRAIR

Serotype

2a, M4243 WRAIR

Serotype

3, J17B WRAIR

Serotype

4, Willis WRAIR

Serotype

5, M90T WRAIR

Serotype

6, CCH060 WRAIR

S. sonnei 53G form I WRAIR

53G form II WRAIR

+,Verystronghybridization; + ++,stronghybridization; +, weak hybridization: -,nohybridizationobserved. bHybridizationdatainthisrowexcludethefollowingCDCstrains:S.

t-lphi.

S.

paratyphi

C,and

S. (clublitn

Vi-.

Weakhybridizationwasdetectedin27%of

Salnoniella

strainsprobedwith the EcoRI-B.

on April 10, 2020 by guest

http://jcm.asm.org/

(3)

inTE buffer. Specific activity of the probe was usually 2 x

108 cpm/,ug of probe DNA.

Filters for in situ colony hybridization. Pure bacterial

cultures were grownovernightandtransferred by toothpick to an 82-mm diameternitrocellulosefilter (BA 85;Schleicher & Schuell, Inc., Keene, N.H.; HAHY 082 50; Millipore Corp., Bedford, Mass.) layered on MacConkey agar. Gen-erally, 15 to 50 cultures were inoculated on each nitrocellu-lose filter.After 3 to 6 h of incubation at 37°C, the filters were removed from the agar and the attached cells were lysed

with 0.5 MNaOH and prepared by the method described by Moseley et al. (24). These nitrocellulose filters were then transferred face upfor 1 mineach to a series of three paper filters eachsaturated with 1.0 M ammonium acetate and 0.02 M NaOH. After 10 min on a fourth change of the latter solution, nitrocellulose filters were air dried and the DNA wasfixed by incubation at 70°C for 2 h in vacuo.

In addition to nitrocellulose, 541 paper (Whatman, Inc., Clifton, N.J.) was used as a solid support for DNA

hybrid-izations. The Whatman 541 papers were prepared by the method described byMaas (21). An 82-mm circularpiece of

Whatman 541 paper was placed overcoloniesthat had been inoculated onto a nutrient agar plate and incubated at 37°C

overnight. After approximately 15 min, the Whatman 541 paper was peeled off and placed colony side up on a paper filter saturated with 0.5 M NaOH-1.5 M NaCl (lysing solu-tion), steamed for 3 min, immersedin fresh

lysing

buffer for

1min,immersed in 1 M Tris (pH 7)-2 MNaCl(neutralization

solution) for 4 min, and air dried. Prehybridization of

Whatman 541 paper is not necessary; the

hybridization

experiments were carried out in the same manner as that described for nitrocellulose. In addition, theprobe could be

removed from Whatman 541 paper by washing in 0.5 M NaOH for 30 min and then washing in 2x SSC

(lx

SSC is 0.15 M NaCl plus 0.015 M sodium

citrate)-0.1%

SDS for 30

min; after air drying, hybridization could be

repeated

as

described above.

Hybridization. The solution for

prehybridization

and

hy-bridization consisted of50% formamide, 5x SSC

(0.75

M

NaCl,0.075 M sodiumcitrate), 0.1%SDS,1mM

EDTA,

and

lx Denhardt solution(0.02% Ficoll

[Pharmacia

Fine

Chem-icals, Piscataway, N.J.],

0.02%

polyvinylpyrrolidone, 0.02%

bovine serum albumin). Nitrocellulose

filters,

prepared

as

described above, wereincubated for2 to 4 h in

prehybridi-zation solution containing 50 ,ig of

heat-denatured,

soni-cated salmon sperm DNA per ml. The filters were then transferred tofreshhybridization solution

containing

labeled

probe DNA (106 cpm) and 50 ,ug of

heat-denatured,

soni-cated salmon sperm DNA per ml. The

probe

DNA was

denatured with alkali as describedbyHill and

Payne

(10)

or

by boiling for 10 min. The filters were

hybridized

overnight

at37°C. Excesshybridization mixturewas

removed,

and the filters were washed once in 5x SSC-0.1% SDS at room

temperature for 15 min, then three times in 2x SSC-0.1%

SDS at 65°C for 15 min each,and

finally

threetimes in 0.1x SSC-0.1% SDS at 65°C for 15 min each.

Hybridized

filters were air dried, andautoradiogramswere

exposed

for4or18 h at -80°C with Kodak XAR film and

regular

intensifying

screens.

As a part of the plasmid mapping

studies,

the Southern

blothybridization technique (28) wasusedtotransferDNA

from an agarose gel onto a nitrocellulose filter in 6x SSC.

Probe hybridization to theSouthern blotswascarriedoutas described byManiatis etal. (22).

Studies of probe hybridization

sensitivity.

A minifold II

apparatus (Schleicher &Schuell)wasusedto

deposit

10-fold

dilutions of

overnight

bacterial cultures onto nitrocellulose

filters. We

prepared

dilutionsin0.9% salineand

plated

them

on nutrient agarto obtain viable counts. These filterswere

processed

and

hybridized

in the samemanner asfor

colony

hybridization

studies,

as described above. Inan attempt to

increase

sensitivity,

10% dextran sulfate or increased

amountsof

probe

DNA(107to

108

total

cpm)

orbothwere added to the

hybridization

mixture in some

experiments.

Kodak XAR film was

exposed

at

-80°C

for various times from 18 to 72 h.

RESULTS

Construction ofDNA

probes

specific

forVi

antigen

struc-tural genes. Ina

study,

chromosomalDNAfromE.

coli

WR2376,

whichcontained the

chromosomally

integrated

C.

freundii

WR7004 genes

encoding

Mel'

(melibiose

utiliza-tion)

and the

adjacent

Vi

antigen

structuralgenes

(i.e.,

the viaB

locus),

was

partially

digested

with endonuclease PstI and the

resulting

materialwascosmid cloned intothevector

plasmid

pHC79.

One recombinant

cosmid,

pWR75,

con-tained a 31-kb insert and

expressed

both

tetracycline

resist-ance and the Vi

antigen

in E.

coli

HB101,

which

normally

contains functional viaA sequences

(1).

In further studies

aimed at

investigating

the reversible nature of Vi

antigen

expression,

wesubclonedVi

antigen

genesfrom

pWR75

into

the

single-copy

plasmid

pDPT429

by

using

a

partial

EcoRI

digest

of both

plasmids.

One

resulting

recombinant

plasmid,

pWR80,

was

isolated,

which has a 29-kb

fragment

from

pWR75

inserted into

pDPT429 (Baron

et

al.,

Abstr. Annu. Meet. Am. Soc. Microbiol.

1983).

We used Vi

antigen-expressing plasmid

pWR80

as our

beginning

material to

identify

and

study

potential

Vi

gene-specific

DNA

probes.

Initially,

wereduced the insert

by

partially digesting pWR80

with EcoRI and

inserting

an 18-kb

fragment

intothe EcoRI

site of the

broad-host-range

vector

pRK290

to

generate

pWR122,

aVi

antigen-expressing

recombinant

plasmid.

This

18-kb viaB DNA insert in

pWR122

consists oftwo EcoRI

fragments,

whichwe

designated

EcoRI-Aand

EcoRI-B

(8.6

and 9.4

kb,

respectively).

Sincevector

pRK290

wasderived

from Pseudomonas sp.

(4),

it was

hoped

that this vector

wouldnot share

homology

with enteric bacteria.

However,

when

32P-labeled

pWR122

wasusedas a

probe,

it

hybridized

weakly

to DNAs ofsome E.

coli

and

Shigella

strains,

and further

cloning

ofthe insertwas necessary. The 18-kb viaB

insert of

pWR122

was then cloned into

pACKC1

(a small,

amplifiable ColEl

derivative

vector)

by

ligation

of

EcoRI-digested pWR122

and

pACKC1, resulting

in the

construction

ofa Vi

antigen-expressing

recombinant

plasmid,

pWR127.

We

separately

subcloned EcoRI-A and

EcoRI-B,

the two

fragments

ofthe viaB

region,

into the vectors

pBR325

and

pACKC1,

respectively,

toconstruct

pWR141

and

pWR137.

Cells

harboring

plasmids pWR137

or

pWR141

donotexpress

the Vi

antigen.

In all ofthe

cloning

studies,

we assessedVi

antigen

expression by using

the three methods described

above. Before further

subdividing

these two

viaB

gene

fragments,

we

attempted

to assess their

hybridization

spec-ificity.

Probe

specificity.

We conducted

colony

hybridization

ex-periments

todetermine ifeither the EcoRI-Aor

EcoRI-B

of

the viaB

region

could be used as

hybridization probes

for

detecting

the presence of the Vi gene locus in the test

bacterial strains. BacterialDNA wasfixedon nitrocellulose

filters and

probed

with

32P-labeled

DNA as described.

Sev-eral

positive

and

negative

controls were included on each filter. C.

freundii

WR7004,

from which the

probe

Vi

antigen

geneswere

originally

cloned,

servedas one

positive

control.

on April 10, 2020 by guest

http://jcm.asm.org/

(4)

a

b

I

c

d

e

2 p

I

11 E

A

0

3S-FIG. 1. Autoradiogram of colony hybridization on Whatman filterpaperafter incubationwith32P-labeled EcoRI-A. KodakXAR film was exposed for 4 h. Strains: la, C. freundii WR7004; lb,S.

typhi WR4201; lc, S. typhi WR4205; ld, S. typhi WR4226; le, E. coli K-12AB313; 2a,E.coli218; 2b,E.coli437; 2c,S.typhi Ty 2; 2d, E. coli 439; 2e, E. coli 440; 3a, E. coliK-12HB101(pWR127);3b, E. coli441; 3c, E. coli 442; 3d,E.coli501;and3e,notinoculated.

S. typhi WR4201 was always included as a typical Vi

antigen-expressing typhoid strain. S. typhi WR4205 contains

amutation in theviaAregion but hasanintact viaB locus.It

thereforewas usedasapositive control for thepresence of Vi structuralgenes in S. typhi. However, S. typhi WR4226

contains anintact viaA region, but the viaB locus has been

replaced by S. typhimurium chromosomal DNA; this created

a ViaB- phenotype, and thus, strain WR4226 served as a

negative S. typhi control. The Vi-positive E. coli WR2376 (1)

wasused asanotherpositive control. Strain 17-59 isarare,

Vi-positive isolate of S. dublin andwasusedas anadditional

positive control (20).

A representative sample of various enteric bacterial strains (e.g., E. coli, S. typhimurium, S. sonnei, and S. flexneri) was used to test the specificity of the various probes. Table 2 summarizes the results of these studies. An example ofanautoradiogram ofoneof these filters is shown

in Fig. 1.

Both EcoRI-Aand EcoRI-B were tested ashybridization

probes with nitrocellulose filters as well as Whatman 541 paper. The EcoRI-A probe only hybridized with DNA samples containing the viaB locus, whether or not the Vi antigenwas expressed(Table 2). Of 140 variousSalmonella strains(Table 2) obtained from the CDC, theEcoRI-Aprobe hybridized only to DNA from colonies of S. typhi, S.

paratyphi C, and Vi-positiveS. dublin,as onewouldexpect ofahighly specific probe. No hybridization ofEcoRI-Awas

detected against DNA from 16Citrobacter strains obtained from the CDC.

The EcoRI-Bprobe was less specific. Although EcoRI-B

hybridized strongly to DNA samples containing the viaB

locus, a weak hybridization signal was detected against

many Salmonella andCitrobacter strains. Furthermore, an

unexpected strong hybridization ofEcoRI-B toCitrobacter strain4182-83 wasobserved.

Bacterial strains that produce capsular antigensmay

con-tain regions of DNA that code for common biosynthetic

enzymes. In recently reported hybridization experiments,

cloned Kl capsular antigengenes exhibited homology with

DNA from strains ofE. colicapsular types K92, K7, and

K100 (5). Therefore, we probed several E. colistrains that

produce capsular antigens withEcoRI-A and EcoRI-Btosee

Pv I

PP P

''71

I I I

A

E

I I Kb

5

10

FIG. 2. Restriction endonucleasedigestionmapof theEcoRI-A of the viaBgeneregion. Restriction endonuclease digestion sitesare designatedasfollows:E, EcoRI; A, AvaI; B, BglII; P,Pstl;andPv, PvuII.

ifany hybridization could be detected. Afterhybridization with Whatman 541 paper, autoradiograms resulting from a 4-hexposurewereidentical whenEcoRI-AorEcoRI-Bwas used as a probe. Strong hybridization was observed with

DNA frompositive control strains with viaBsequences,but hybridization was not detected against the other strains

tested, which included negative controls and strains ofE. coli that produce common capsular antigens. Although hy-bridization withEcoRI-Awasnotdetectedevenwithlonger exposure, weak hybridization between EcoRI-B and these E. coli strainswasobserved when the filterswere autoradio-graphed overnight.

Restrictionmapping of EcoRI-A. Since the EcoRI-A of the viaBgeneregion appearedtoserve as ahighly specific DNA

probe,wedecidedtomapitssites for endonucleasecleavage with severalrestrictionenzymes.PlasmidpWR141 contains the EcoRI-A ofpWR127 incorporated into pBR325. Single and doublerestriction endonucleasedigests of pWR141were

resolved by electrophoresis on agarose gels, and we

ana-lyzed the DNA fragments by size and Southern blot hybrid-ization toconstructarestriction map(Fig. 2).

Study of probe sensitivity. To determine the fewest number ofbacteria thatcould be detected with the EcoRI-A probe and theradiolabeling procedure,weperformed the following study. The DNA from 101 to105 cells of each of three test

bacterial strains was fixed on a nitrocellulose filter and

probed with EcoRI-A. When dextran sulfate and additional probe

(107

to 108 cpm) were used in the hybridization

mixture,

103

Vi-positive cells could be clearly detected (Table 3). In some experiments, as few as 100 to 500 Vi-positive cells were detected, but detection was made

difficultbecause of increased nonspecific reactivity.

TABLE 3. Sensitivity of EcoRl-A probe

No.of Hybridization reaction with": bacterial cells C. freundii E.coli S.tvphi

WR7004 HB101 WR4201

105 ++++ --;++ +

104 ++ - ++

103 + _ +

102 - (+)b _ (+)b

101 _ _ _

aHybridization observed with addition ofdextran sulfate andprobeDNA (107cpm)tothehybridization mixture.

bIn someexperiments with probeDNA(108cpm), asfewas100to 500cells could be detected.

IF a X - n

a I

I

on April 10, 2020 by guest

http://jcm.asm.org/

(5)

DISCUSSION

Rapid identification tests for microbial pathogens are

currently being developed by recombinant DNAtechnology combined with radio- or enzyme-linked immunoassay tech-niques. DNA probe detection systems have been reported

for the following enteric bacteria: enterotoxigenic E.coli (10,

24), Vibrio spp. (15, 16), Yersinia enterocolitica(11),

Salmo-nella spp. (7), and Shigella spp. (2).

Typhoid fever remains a serious public health problemin

developing countries and continues to be endemic in many

areas of the world. Currently, microbiological identification

of S. typhi from clinical specimens generally requires 36 to 48 h. To simplify identification,we assessedthe Vi capsular antigen structural genes for use in the developmentof arapid detection DNA probe system for S. typhi.

Our approach for development of the DNA probe for the

detection of S. typhi involved cloning the viaB region ofC. freundii. The smallest recombinant clone that expresses Vi antigen contained an 18-kb DNA insert. Digestion of this 18-kb cloned insert with EcoRI restriction endonuclease

produced two fragments, which weredesignatedasEcoRI-A

(8.6 kb) andEcoRI-B(9.4 kb). Each of these fragmentswas tested as a possible probe for detectingS.typhi. When used

to probe a variety of enteric strains, includinghighly related Salmonella and Citrobacter strains, EcoRI-B was not spe-cific. Weak hybridization of EcoRI-B was observed with many Salmonella strains, some Citrobacterstrains, and E. coli strains producing capsular antigens.EcoRI-A,however, was specific for strains containingviaBgene sequences and should have considerable potential as a DNA probe in a

rapid diagnostic system for identifying S. typhi. Further study may reveal a largerprobe composed ofEcoRI-A and part ofEcoRI-B that combines the specificity ofEcoRI-A

with increased hybridization sensitivity.

In addition to evaluatingthediagnosticpotential of

EcoRI-A and EcoRI-B, we have been using these fragments to detect recombinant clones carrying similar S. typhi DNA sequences so that a comparison can be made between the viaB regionsof C.freundii and S. typhi. Thesehybridization

probes will also be helpful in studying the structural and

regulatory components controlling expression of the Vi

antigen.

Using various enteric bacterial strains, we tested probe

specificity with nitrocellulose and Whatman 541 paper and found Whatman 541paper tohave severaladvantageswhen usedin thecolony hybridization protocol. As apaper with high wet strength,it iseasierto handle thannitrocellulose. In addition, the papers do not have to be baked in a vacuum oven to fix DNAto the solidsupport. Anotheradvantage of Whatman 541paper is thatprehybridization isnotnecessary

(S. Moseley, personal communication). Finally,

hybridized

probe can beremovedeasily and the samples can be tested sequentially withdifferentprobes.

In sensitivity studies, the EcoRI-A probe detected

i04

Vi-expressing cellswith the standard hybridization solution

asdescribedabove. Since dextransulfatehasbeen shownto

increasesensitivity (29,30), we reexamined the

sensitivity

of

our probe with this reagent included in the

hybridization

solution. A 10-fold increase in sensitivity was observed

when dextran sulfate was added (Table 3). We expect that the EcoRI-A probe can be placed into a nonradioactive

labeled system in which detector signals can be

amplified,

resultingin afurther increase in sensitivityas well as

rapid

identification of S. typhi.

To assess the specificity of theEcoRI-A probe,

testing

of

selected human stool samples obtained in

Lima, Peru,

has

already begun. Also, efforts are under way to determine

whetherthis probe canbeused to aidin the detection ofS.

typhi carriers in Chile.

ACKNOWLEDGMENTS

WethankDeanTaylor and Vickers Burdett forplasmidvectors, RichardSilver,RandallHolmes,KayeWachsmuth,andAlma Murlin for theirhelpfulassistance insupplyingstrains,PatriciaGuerryfor expert scientific advice, and Eamestine Durham for typing

assistance.

LITERATURE CITED

1. Baron, L.S., D.J. Kopecko, S. M. McCowen,N. J. Snellings,

E.M.Johnson, W. C. Reid,and C. A. Life. 1982. Genetic and molecularstudiesontheregulationofatypicalcitrateutilization and variable Vi antigen expression in enteric bacteria, p. 175-194. In Hollaender(ed.), Genetic engineeringof

microor-ganismsfor chemicals. PlenumPress, NewYork.

2. Boileau, C.R., H. M. d'Hauteville, and P.J. Sansonetti. 1984. DNA hybridization technique to detect Shigella species and enteroinvasive Escherichiacoli.J. Clin. Microbiol. 20:959-961. 3. Bolivar, F. 1978. Construction and characterization of new

cloning vehicles. III. Derivatives ofplasmid pBR322

carrying

unique EcoRI generated recombinant molecules. Gene 4: 121-136.

4. Ditta, G., S. Stanfield, D. Corbin, and D. R. Helinski. 1980. Broad host range DNAcloningsystem for

gram-negative

bac-teria: construction ofagenebank ofRhizobiummeliloti. Proc. Natl. Acad. Sci. U.S.A. 77:7347-7351.

5. Echarti, C., B. Hirschel, G.J. Boulnois, J. M. Varley, F. Waldvogel,and K. N.Timmis.1983.

Cloning

and

analysis

ofthe Kl capsule biosynthesis genes of Escherichia coli: lack of homology withNeisseriameningitidisgroup B DNAsequences.

Infect. Immun. 41:54-60.

6. Edwards, P.R., and W. H. Ewing. 1972. Identification of

En-terobacteriaceae, 3rded., p. 146-207. Burgess

Publishing

Co.,

Minneapolis.

7. Fitts, R.,M.Diamond,C. Hamilton,and M. Neri. 1983. DNA-DNA hybridization assay for detection ofSalmonella spp. in foods. Appl. Environ. Microbiol. 46:1146-1151.

8. Flexi, A., S. S.Bhatnagar, and R. M. Pitt. 1934. Observations

onthe properties ofthe Vi

antigen

ofB.

typhosus.

Br. J. Exp.

Pathol. 15:346-354.

9. Heyns, K.,G.Kiessling,W. Lindenberg, H.Paulsen,and M. E. Webster. 1959. D-Galaktosaminuronsaure

(2-Amino-2-desoxy-D-galakturonsaure) als Baustein des

Vi-Antigens.

Chem. Ber. 92:2435-2437.

10. Hill, W. E., and W. L. Payne. 1984. Genetic methods forthe detection of microbial pathogens. Identification of

enterotoxi-genic Escherichia coliby DNA

colony

hybridization:

collabo-rativestudy. J. Assoc. Off. Anal. Chem. 67:801-807.

11. Hill,W.E.,W. L.Payne,andC. C.G. Aulisio. 1983. Detection andenumeration ofvirulent Yersinia enterocolitica infood

by

DNA colony hybridization.

Appl.

Environ. Microbiol. 46: 636-641.

12. Hohn, B., and J. Collins. 1980. A small cosmid for efficient

cloningoflargeDNA

fragments.

Gene 11:291-298.

13. Johnson, E.M.,B.Krauskopf,and L. S. Baron. 1965.Genetic

mapping of Vi and somatic

antigenic

determinants in Salmo-nella. J.Bacteriol. 90:302-308.

14. Johnson, E.M.,B. Krauskopf, and L. S. Baron. 1966. Genetic

analysis ofthe viaA-his chromosomal

region

in Salmonella. J. Bacteriol. 92:1457-1463.

15. Kaper, J.B., R. K.Campen, R.J. Seidler, M. M.Baldini,and S. Falkow. 1984.

Cloning

of the thermostable direct or

Kanagawa

phenomenon-associated

hemolysin

of Vibrio

parahaemolyticus. Infect. Immun.45:290-292.

16. Kaper, J. B., and M. M. Levine. 1981. Cloned cholera enterotoxin genesin study and

prevention

of cholera. Lancet ii:1162-1164.

on April 10, 2020 by guest

http://jcm.asm.org/

(6)

17. Krishnapillai, V., and L. S. Baron. 1964. Alterations in the

mousevirulence of Salmonella typhimurium by genetic recom-bination. J. Bacteriol. 87:598-605.

18. Kupersztoch, Y. M., and D. R. Helinski. 1973. A catenated DNA molecule as an intermediate in the replication of the resistance transfer factor R6K in Escherichia coli. Biochem. Biophys. Res. Commun. 54:1451-1459.

19. Kushner, S.R.1978. Animproved method for transformation of Escherichia coli with ColEl derived plasmids, p. 17-23. In H. W. Boyer and S. Nicosia (ed.), Genetic engineering. Elsevier/North-Holland Biomedical Press, Amsterdam. 20. LeMinor,L., andP. Nicolle. 1964. Sur deux souches de

Salmo-nelladublin possedantl'antigene Vi. Ann. Inst. Pasteur (Paris) 107:550-556.

21. Maas, R. 1983. An improved colony hybridization method with significantly increasedsensitivity for detection of singlegenes.

Plasmid 10:296-298.

22. Maniatis, T., E. F. Fritsch, and J.Sambrook. 1982. Molecular cloning. A laboratorymanual.ColdSpring Harbor Laboratory,

Cold Spring Harbor, N.Y.

23. McConnell, M., H. R. Smith, G. A. Willshaw,S.M. Scotland, and B. Rowe.1980. Plasmids coding for heat-labile enterotoxin

production isolated from Escherichia coli 078: comparison of properties. J. Bacteriol. 143:158-167.

24. Moseley, S. L., P. Echeverria, J. Seriwatana, C. Tirapat, W.

Chaicumpa, T.Sakuldaipeara,andS. Falkow. 1982. Identifica-tion ofenterotoxigenic Escherichia coli by colony hybridization usingthreeenterotoxingeneprobes.J. Infect. Dis. 145:863-869. 25. Rigby, P. W.J., M. Dieckmann, C. Rhodes, and P. Berg. 1977. Labelingdeoxyribonucleic acidtohigh specific activityin vitro bynick translation withDNApolymeraseI. J. Mol.Biol. 113: 237-251.

26. Snellings, N. J., E. M. Johnson, and L. S. Baron. 1977.Genetic basis of Vi antigen expression in Salmonella paratyphi C. J. Bacteriol. 131:57-62.

27. Snellings, N. J., E. M. Johnson, D. J. Kopecko, H. H. Collins, and L.S. Baron.1981.Geneticregulation of variable Vi antigen expression in a strain of Citrobacter freundii. J. Bacteriol. 145:1010-1017.

28. Southern, E. M. 1975. Detection of specific sequences among DNAfragments separated bygel electrophoresis. J. Mol. Biol. 98:503-517.

29. Totten, P.A., K. K. Holmes, H. H. Handsfield, J. S. Knapp, P. L.Perine,and S. Falkow. 1983. DNAhybridization technique for the detection of Neisseria gonorrhoeae in men with urethri-tis. J. Infect.Dis. 148:462-471.

30. Wahl,G.M., M.Stern,andG. R.Stark.1979. Efficienttransfer of largeDNAfragments fromagarosegelsto diazonbenzyloxy-methyl-paper andrapid hybridization by using dextran sulfate. Proc. Natl. Acad. Sci. U.S.A. 76:3683-3687.