Veterinary Microbiology, 29 ( 1991 ) ! 81 - 194 Elsevier Science Publishers B.V., Amsterdam

i8i

Generation of aromatic-dependent Salmonella havana and evaluation of its immunogenic

potential in mice and sheep

T.K.S. Mukkur a'*, K.H. W al ke r b a n d B.A.D. S tocke r c

aCSIRO Division of Animal Health, McMaster Laboratory Private Bag No. !. Glebe. NSW 2037, Australia

bDepartment of Agriculture and Fisheries, Eli:abeth Macarthur Agricultural Institute. Regional Veterinary Laboratory Menangle. NSW 2568, Australia

~Department of Medical Microbiology and hnmunology. Stanford UniversiO'. School of Medicine. Stanford, CA, 94305, USA

(Accepted 17 April 1991 )

ABSTRACT

Mukkur, T.K.S., Walker, K,H. and Stocker, B.A.D., 1991. Generation of aromatic-dependent Sal- monella havana and evaluation of its immunogenic potential in mice and sheep. IFt. Microbiol., 29: !81-194.

The generation of aromatic-dependent ( a r o - ) Sabnonella havana (Group G2. 01.13, 23 ) from a smooth wild-type parent strain by transduction with phage PI is reported. Mice immunized with this live aro- S. havana strain (CS234) by the intraperitoneal (i.p.) route were protected against chal- lenge with wild-type 5". havana, whereas those immunized by the oral route were not. Mice immunized with two doses of formalin-killed aro- S. havana by the i.p. route were also unprotected, in spite of high antibody titers. However, only those mice immunized with live aro- S. harana by the i.p. route developed significant delayed-type hypersensitivity. Following i.p. inoculation in mice, the aro- S. havana strain CS234 was detected in the liver, spleen and mesenteric lymph nodes on day 9 but not on day 15 post-inoculation (p.i.). On the other hand, when mice were inoculated with the parent wild-type strain (CS4) or the aro- derivative strain CS234 by the oral route, the organisms x~'ere recovered from the mesenteric lymph nodes and intestine only on day 3 but not on day 6 post-inocu- lation. In sheep inoculated with the aro- strain CS234 in the gastroc muscle, organisms were re- covered from the muscle, and popliteal and medial iliac lymph nodes for up to 21 but not 28 days p.i. However, no mutant organisms were recovered from liver, spleen, mescnteric lymph nodes or faeces. In orally-inoculated sheep, the mutant organisms were recovered from the mcsenteric Ivmnh , c 2 ~ . rumcn, intestinal contents, and faeces up to 14-21 days post-inoculation but not at 28 days. When sheep immunised with the aro- S. havana strain CS234 by the intramuscular or oral route were chal- lenged with the parent wild-type S. havana strain CS4 by the oral route, the latter strain was detectable in the mesenteric lymph nodes and faeces of immune sheep up to 14 days post-challenge in contrast with the non-immune sheep, where the challenge strain was detectable even at 28 days post-challenge. Only sheep immunized by the intramuscular route developed high antibody levels and delayed-type hypersensitivity.

*Author for all correspondence.

0378-1135/91/$03.50 © 199 ! Elsevier Science Publishers B.V. All rights reserved.

182 T.K.S. MUKKLIR ET AL.

I N T R O D U C T I O N

Salmoneliosis of ruminant animals causes significant economic loss to the livestock industry world wide (Wray and Sojka, 1977), in addition to being a public health hazard. Carrier animals become active excretors particularly when they are subjected to stressful conditions such as those encountered dur- ing pregnancy, crowding and shipping over long distances, and in intensive calf-rearing enterprises. Although a multiplicity of salmonella serotypes have been isolated from outbreaks of clinical salmonellosis in sheep, the most sig- nificant species in Australia include Sahnonella typhimurium (Group B ), S. bovismorb(ficans (Group C2 ), S. havana (Group G2 ) and S. anatum (Group E l ) ( Dixon, 1980; Murray, 1984 ). Previously we demonstrated that an aro- matic-dependent (aro-) S. typhtmurium vaccine in mice and sheep and an aro- S. bovismorb(,qcans vaccine in mice were highly protective against oral challenge with w,ld-type organisms (Mukkur et al., 1987: Begg et al., 1990; Mukkur et al., 1991 ). To further assist in the realization of our ultimate ob- jective of formulating an aro- composite Salmonella vaccine for use in sheep, we developed a method for the generation of aromatic-dependent (aro-) S. havana and evaluated its immunizing potential in mice and sheep. Results of this investigation are reported below.

M A T E R I A L S A N D M E T H O D S

The wild-type smooth strain of S. havana (CS4) used in this investigation was isolated from the faeces of a profusely scouring sheep held in a feedlot in Perth, Western Australia prior to being loaded aboard ships destined for the Midd!e East. Other bacterial strains used in this investigation are listed in Table i.

Determination of LDso The LDso of wild-type and aromatic-dependent ( a ro - ) S. havana was es-

timated by inoculating Balb/c mice with 104-10 II colony forming units (CFU) by the intraperitoneal ( i.p. ) route and calculated according to Litch- fiela and Wilcoxon ( 1949 ). Where possible, mice were euthanased when they were observed to be moribund. The pathogenicity of S. havana for mice was also examined by administration of the organisms by the oral route.

Generation ~!faromaticdependent ( a r o ) S. havana The process stages involved in the generation of atx~- S. havana from the

wild-type strain of Australian origin (CS4/SL5749) are shown in Table i. Phage P22, used to convert virulent strains of S, typhimurium, S. dublic,,~*

S. o'phi into non-virulent, candidate live-vaccine strains by transducing in mutant aro- alleles (Hoiseth and Stoeker, 1981; Smith et al., 1984) adsorbs

G E N E R A T I O N ( )F A R O M A T I C D E P E N D E N T .~,L.IIO.~,'I:I.LI I l l I i I,%. I

T A B L E i

Bacterial strains

183

Strain Number Description* Origin o r Reference

SL1479

SL2838

CS4/SL5749 CS233/SL5770

CS234/SL5775

S. o'phimurium UCD live-vaccine strain: CRR426 S. typhimurium M7471 CRR426 [aroA554::Tn 10 (Tc-s, non-roy.) ] =tg- 903::Tn I 0 galE S. harana wild-type S. harana =hj-903::Tn I 0 CRR426 [ arc.4554::Tn I 0 (Tc-s, non-roy. ) ] S. harana CRR. =bj-903::TnlO (Tc-s) CRR426 [amA554::TnlO (Tc-s. non-roy. ) l

Smith et al., 1984

Nnalue and Stocker. 1987

From profusely scouring sheep From CS4/SL5749 by transduction From CS233.SL5770 by selection for tetracycline sensitivity

*Additional mutant characters of strain SLI479 and SL2838, not here relevant, omitted. CRR. com- plex rearrangemenl mutation, here Tn 10-generated deletion or deletion-inversion causing loss of te- tracycline-resistance. Tc-s, tetracycline-sensitive, non-roy., unable to revert by "'clean excisioa'" of transposon.

only to smooth strains of O groups B, D or A and so could not be used to transduce a mutant aro allele to S. havana of O group G2 (01, 13, 23 ). Phage Pl does not act on smooth Salmonellae of O group B, but is adsorbed by mu- tants making type Rc (galactose-deficient) iipopolysaccharide (LPS), in- c:uding those with mutations at ga! E, which specifies the enzyme UDP-gal- actose epimerase; this phage has been used to transduce genes from gal E strains of S. typhimurlum (O group B) to ga! E strains of S. choleroesuis ( ~ o u p C !; 069 7 ) and the reverse (Nna!ue and gtncker~ ! 987 ). We therefore planned to isolate a gal E mutant from the wild-type {smooth) S. harana strain° for use as a recipient in P l-mediated transduction. However, contrary to expectation, S. havana CS4/SL5749 could adsorb phage P 1. This was shown by the appearance of colonies ( ~ 50 per 0.0 ! ml drop of undiluted lysate) after overnight incubation at 30°C of plates of blood agar base with chlor- amphenicol ( 12 gg/mi ) flood-inoculated with S. havana wild-type ~train and spotted with phage PI Cm Ct-s {which when it lysogenizes converts its host to chloramphenicol-resistance and is temperature sensitive for induction). The yield ofchloramphenico!-resistant colonies from S. havana was only about 1/200 of that from a gal E strain orS. o'phimurium.

In view of this result, we nex~ tried to transfer a previously-characterized non-reverting mutation at aro,4 to our wild-type S. havana strain by co-trans- duction with an adjacent silent insertion oftransposon Tn 10, designated Zbj- 903::Tn 10 (Hoiseth and Stocker, 1981; Nnalue and Stocker, 1987), with selection for tetracycline-resistance. The strain used as donor, SL2838, is a S. typhimurium line with a gal E mutation which was given by transduction,

! 8 4 T.K.S. MUKKUR ETAL.

mutation Zbj-903::Tn 10 and the linked Tn 10-generated non-reverting aro A mutation of an S. typhimurium live-vaccine strain, SL1479 (Nnalue and Stocker, 1987; Smith et al., 1984). Drops of a PI Ct-s lysate of strain SL2838, obtained by thermal induction of SL2838 converted to chloramphenicol-re- sistance by lysogenization with this phage, were applied to plates selective for tetracycline resistance which had been flood-inoculated with the F havana recipient. After 3 d incubation at 37°C, nine tetracycline-resistant colonies were observed. Eight of these gave prototrophic growth, like the parent strain; they are presumed to have originated either from incorporation of donor Zbj- 903:'Tn 10 without the adjacent aro A mutation or from transposition of Tn 10 from a transduced chromosomal fragment into some unrelated site in the recipient. One tetracycline-resistant clone designated CS233/SL5770 was ar- omatic-dependent but unaltered in serological character and phage sensitivity pattern. This aromatic-dependent transductant was plated on the medium de- vised by Bochner et al. (1980) which inhibits growth oftetracycline-resistant strains. All of the 14 clones tested were tetracycline-sensitive but still aro- matic-dependent and unaltered in other characters tested; one such isolate, numbered CS234/SL5775, constitutes our candidate S. havana live-vaccine strain. No reversion of this strain to aromatic-independence was detected; this was to be expected, since it has the aro A mutation ofthe S. typhimurium live vaccine strain, SL1479, which gave no revertants in extensive trials (Smith et al. 1984).

The LDso of the parent wild-type strain orS. havana when administered by the i.p. route was 1 × l0 T CFU whereas that ofthe aro- derivative was I X 109 CFU per mouse. However, the wild-type organisms were apparently non-

* 1 . . . . . . . . . . _ * r L I I paUlU$¢mt: wncn up tu lu-- C F U per mouse was administered by the oral route.

hnmunizatic, n-challenge experiments Survival (protection from death) was used as the measure of immunizing

efficacy as no published information on alternative definitive predictive cri- teria for assessment of vaccine protection against acute salmonellosis was available. Any mice that were observed to be moribund were euthanased and recorded as a mortality.

(a) Mice. Two groups of 12-week-old Balb/c mice (27 per group) were im- munized once by either the i.p. (Group i) or oral routes (Graup !i) with i x 107 or 10'°CFU ofaro- S. havana (CS234/SL5775) respectively. A third group (Group !!I) of 27 mice was immunized with two doses of formalin- killed aro- X havana (CS234/SL5775) by the i.p. route, administered one week apart ( l0 s CFU and 5 X l0 s CFU respectively). A fourth group (Group IV) of 27 mice served as unimmunized controls. Four weeks post-immuni- zation, 10 mice from each of the groups were challenged with 2 LDso or 10

GENERATION OF AROMATIC DEPENDENT S,4LMONELL4 tlAI~4NA 185

LDso of wild-type S. havana by the i.p. route and mortality recorded for 15 days post-challenge. At the same time, blood was collected from three heavily anaesthetized mice in each group, by excision of the axilla with a scalpel blade, for determination of antibody titers. Mice remained anaesthetized during the procedure and died due to exsanguination. The remaining four mice were used for assessing the development of delayed-type hypersensitivity (method described below).

(b) Sheep. A total of 36 sheep were used in this experiment. One group of ! 2 sheep was vaccinated intramuscularly with one dose containing 10 I° CFU of aro- S. havana strain CS234 while another group of 12 sheep was similarly immunized by the oral route. A third group of 12 sheep was used as non- immune controls. Blood samples were collected at weekly intervals up to 3 weeks post-immunization, when they were challenged orally with 10 Im CFU per sheep ofthe parent wild-type S. havana strain CS4. Three sheep from each group were killed (by stunning with a captive bolt followed by cervical cor- dotomy) at days 7, 14, 21 and 28 post-challenge and the bacterial load in the mesenteric lymph nodes and presence/absence in the faeces determined. The sera were used for determination of antibody titers using agglutinating and enzyme-linked immunosorbent assays (ELISA).

In vivo multiplication o f aromatic-dependent (aro- ) S. havana

(a) Mice. Two groups of 12 i~alb/c mice were inoculated with 5 × 10 7 CFU of the wild-type parent strain oz" of its aro- derivative by the i.p. route, and a third group of i2 mice were inoculated with i0 :° CFU by the oral route. On days 3, 6, 9 and 15 post-inoculation, three mice from each group were euthan- ased and the liver, spleen and mesenteric lymph nodes crushed in 2.0 ml of sterile physiological saline using a Colworth Stomacher (400). A viable bac- terial count was then obtained by spread-plating serial 10-fold dilutions on xylose-lysine deoxycholate (XLD) agar plates.

(b) Sheep. One group of 12 sheep was immunized with 10 I° CFU ofaro- S. havana strain CS234 suspended in i.0 ml of phosphate-buffered saline by the intramuscular (i.m.) route in the gastroc muscle whereas another group of 12 sheep were immunized by the oral route with l 0 I° CFU, delivered in gelatin capsules (Mukkur et ai., 1987). A third group of 12 sheep served as an un- immunized control group. At 7, 14, 21 and 28 days post-immunization, three sheep from each group were sacrificed and the distribution of the aro- S. havana s'-ain CS234 in various organs/tissues of sheep wa~ determined.

The organs/tissues sampled in the intramuscularly immunized group in- cluded liver, spleen, muscle, lymph nodes (popliteal, medial, lilac, j,junal, ileal and caecal ), heart blood, and faeces whereas those sampled in the orally

18b T.K.S. MUKKUR ET AL.

immunized group included liver, spleen, rumen and intestinal (jejunal, ileal, caecal) contents, lymph nodes (anterior mediastinal, jejunal, ileal, caecal), bile, heart blood and faeces. Isolation of salmonella from bile, heart blood and faeces was accomplished by enrichment in mannitol-cystin~ selenite broth. Other organs/tissues were weighed, crushed using a Colworth Stomacher 400, 10 ml of sterile physiological saline added, serial ten-fold dilutions spread- plated on XLD plates for direct bacterial counts, and CFU per gram calcu- lated. A portion of this homogenate was also inoculated into marmitoi-cystine selenite broth.

hmmmological parameters

(a) Agglulination liters The somatic (O) or flagella (H) antibody liters were determined only for

sheep sera, using procedures described previously (Mukkur et al., 1987 ).

(b) ELISA S. havana.specific antibody tilers in the sera of individual mice or sheep

were measured using ELISA as described previously (Engvall and Perlmann, 1972). Antibody levels were calculated as endpoint titers using the cut-off point determined by an average of the optical densities of quadruplicate an- tigen-free wells in each plate. The antigen used for coating the ELISA plates (M 129B Dynatech ) was an ultrasonicate of the parent S. havana strain CS4 ( 100 lig protein per ml ).

(c) Delayed-o'pe h.|,persensitivio' (DTH)

Mice. Delayed-type hypersensitivity (DTH) was measured by injecting a sonicated S. havana preparation ( 10/lg protein per 5 gl saline) into the foot- pads of four mice at four weeks post-immunization and measuring their thickness with a vernier caliper at 48 h post-inoculation.

Sheep. For determination of DTH, groups of 12 sheep each were each im- munized with the aro- S. havana strain CS234 by either the i.m. or oral route, and a third group of 12 sheep served as unimmunized controls. The DTH was determined in three sheep from each group at days 7, 14 and 21 post-immu- nization as described previously ( Mukkur, et al. 1987 ). The antigen used was the ultrasonicate of the parent S. havana strain CS4 (25/~g protein per 25/LI saline ).

Stat i.~t ica1.4 nal rses Student's t test was used to determine the significance of differences be-

tween the log,. mean values for agglutination and ELISA titres and Fisher's

GENERATION OF AROMATIC D E P E N D E N T X.ILM, O X E L L . ! !!,1 ! ~I,Y.-I 187

TABLE 2

The isolation of virulent (CS4) or aro- S. havana (CS234) in Balb/c mice following inoculation by the i.p. route

Strain Day p.i. C F U a ( ----- S E M ) in o r g a n s

Liver Spleen MLN b

× I0 ~

Parent 3 6.0 ___0.76 !.2 +0.52 5.32+_ 1.08 S. havana (CS4/SL5749) 6 700 +42 210 +-88 6.2+-2.29

x 10 ~

an~- S. havana (CS234/ 3 !.5 +_ I.O2 4.5 +__2.8 4.83+_2.62 SL5775) 6 41.3 +_29.3 52.0 _+17.1 6.9 _+0.27

9 0.40+0.06 0.45+_0.16 0.4 +_0.2 i 5 N D ~ N D N D

"~CFU = colony forming units/organ. hMNL=mesenleric lymph nodes. "N D = not detectable.

TABLE 3

The isolation ofaro- S. havana (CS234) in sheep inoculated by the intramuscular route b

Organ/tissue Mean CFU a per gram of organs/tissue on day post-inoculation

7 14 21 28

Liver 0 0 0 Spleen 0 0 0 Muscle (650) (3/3) E' 0

200-800 d Faeces 0 0 0 Lymph nodes

Popliteal (133) (3/3) E 0 100-200 E

Medial iliac ( 1 1 7 ) 0 { 2 / 3 ) E 0 150-200

Jejunal 0 0 0 Ileal 0 0 O ('aecal 0 0 0

(2/3)

0 0 0

0

0

0

0 0 0

• ~CFU =colony forming t.nit~-. "One ml of the inoc:zlum containing I × I0 ~" CFU/ml g, as injected into the gastroc muscle. ' E denotes isolation of salmonella organisms using an enrichment procedure, whereas the proportion shown within parenthesis represents number of sheep in v.'hich the isolation of salmone!la organisms from organs by enrichment culture divided by the !oral number of sheep used on various da.~s post inoculation. dDenotes range of CFU observed.

188 T.K.S. MUKKUR ETAL.

exact test was used to determine the significance of protection observed in mice.

RESULTS

In vivo distribution ofaro- S. havana strain (CS234)

(a) Mice. On day three post-inoculation, the number of live wild-type S. ha- rana (CS4/SL5749) recovered from liver, spleen and mesenteric lymph nodes was significantly greater than that observed for its aro - derivative (CS234/ SL5775 ) (Table 2 ). Further, there was a substantial increase in the live wild- type and to a lesser degree the aro - S. havana recovered from the liver and spleen on day 6 post-inoculation. By day 9, all mice inoculated with wild-type

TABLE 4

The isolation ofaro- S. harana (CS234) in sheep inoculated by the oral route

Organ/tissues Mean CFU a per gram of organs/tissue on day post-inoculation

7 14 21 28

Hearl blood 0 0 0 0 Bile 0 0 0 0 Liver 0 0 0 0 Spleen 0 0 0 0 Rumcn contents ( I / 3 ) E" 0,0 ( I / 3 ) E 0 0

(9OO) Jejunum ( I00(~ 0 0 0

0-2~..~3 ~ Ileum (27 300) (90 000) C" 0

2900-52 000 0-270 000 ( I / 3 ) E

Caecum (72 000) (183) 0 0 8600-200 000 0-5:~0

( I / 3 ) E Faeces (3/3) E 0 0.0 ( I /3 ) E 0 Lymph nodes

Anterior mediasfinal 0 Jejunal 0 Ileal 0.0 ( I /3 ) E

Caccal

0 0 0 0 0 0 ( I /3 ) E 0 0 0-183 (150 0,0 (1/3) E 0 0-450

"~('FU. colony forming units. qX'noles range of CFU observed. °E denotes isolation of salmonella b~ enrichment culture. The numbers within parentheses denote the number of sheep from ~,hich mutant salmonella were isolated, divided by the total number of sheep vaccinated.

GENERATION OF AROMATIC DEPENDENT S.4L,~IONIfLLA 11.41:4,VA ! 89

S. havana had died. In contrast, there was a significant decline in the numbers of live, aro- S. havana recovered from the liver, spleen and mesenteric lymph nodes on day 9 post-inoculation. However, by day 15, aro- organisms could not be recovered even by enrichment cultures (Table 2 ).

Neither wild-type nor aro- S. havana could be detected in the liver and spleen of mice inoculated by the oral route, although they were detectable in the intestine and mesenteric lymph nodes by enrichment cultures on day 3 but not day 6 p.i. (data not shown ).

(b) Sheep. Aromatic-dependent ( a r o - ) S. havana strain CS234 was not de- tectable in any tissue or organ tested on day 28 p.i., regardless of the route of immunization (Tables 3 and 4).

Immunization-challenge experiments and immunological analyses

(a) Mice. Mice immunized with live aro- S. havana by the i.p. route were protected against an i.p. challenge with 2 x 107 (2 LDso) or ! X I 0 s ( 10 LDso) CFU of the wild-type organisms. On the other hand, mice immunized orally with one dose ( IO '°) CFU of the live mutant organism or two doses of for- malin-killed S. havana, administered one week apart ( 10 s and 5× IO s CFU respectively) were not protected (Table 5). Mice immunized with live or killed aro- S. havana by the i.p. route developed high ELISA antibody tilers,

TABLE 5

Antibody tilers, delayed-type hypersensitivity ( D T H ) and protection in mice immunized with live or killed aro- S. harana (CS234)

Vaccine preparation Total vaccine dose ELISA tiler DTH and route of ( C F U ' ) ( -+SEM) (ram) administration administered/ ( -+ SEM )

m o u s e

No. of mice surviving/ total no. of mice challenged by intraperitoneal route

2 LDso 10 LDso

Live, a r o - S. havana. 10 '° 1381"c+-691 0.153-+0.02 I 0 / I O~ 10/lOs oral Live. aro - S. havana I × I0 ' i 0358 ' _ + 1667 0.793_+0.11 I 0 / ! 0 ~ 10/10 r intraperitoneal Formalin-killed aro- IO s. 5X !0 a {one 8338d+ 1936 0.198-+0.02 3/10 h 1/10 h

h.rano, week apart ) intraperitoncal Unimmunized - ND h 0.183+0.02 0 /10 0 /10

ICFU =colony forming units. hND = not detectable, ": Statistically significant differences obs~n'ed in vaccinaled mice: c an# d versus e (P< 0.01 ): f versus n (P<O.OI): fand gversus i (P<O.O01).

] qO T.K.S. MUKKUR ET AL.

TABLE 6

The isolation o f £ harana strain CS4 after challeng~ of sheep b either vaccinated intramuscularly with the a r o - strain CS234 or unimmunized controls

Organ/tissue Immunisation status Mean CFU ~ per gram of organs/tissues on day post challenge

7 14 21 28

Jejunall.vmph node Immune 0 0 0 0 Non-immune 0 0 0 0

Ileal lymph node Immune ( 667 ) 0,0 ( I/3 ) 0 0 200-1100 E ~

Non-immune (750 (700) (900) 100-1500 500-1200 700-1300

Caecal lymph node Immune (400) 0 (2/3) E 0 0-I000)

Non-immune (700) (600) { 266) 600-800 200-1400 i 00-500

Faeces Immune (3 /3) E 0 0 Non-immune (3 /3) E O, (2 /3) E O, (2/3) E

(700) 400-I000 (I/3) E 0

(166) 0-300 0 0,0 ( I / 3 ) E

~CFI! = colony forming units. hThree sheep per Ireal.ment at specified intervals were used. "E denotes isolation of salmonella by enrichment culture. Numbers within parentheses denote the number of sheep from which the challenge organisms were isolated, divided by the total number of sheep challenged.

TABLE 7

The isolation of X havana strain CS4 after challenge of sheep b either vaccinated orally with the aro -

Organz/tissues lmmup.i~a:ion ;talus Mean CFU ~ per gram of organs/tissues on day post- challenge

7 14 21 28

Jejunal lymph node Immune 0 0 0 0 Non-immune 0 0 0 0

Ileal bmph node Immune (3 /3) E ~ 0 0 0 Non-immune (826) ( ! I00) ( I00) (316)

500-1200 1000-2000 0-300 150-400 ('aecall.vmph node Immune (3 /3) E (3/3~ F O 0

Non-immune (366) (1333) ( i00) (433) 200-600 1200-1500 0-200 150-600

Faeces Immune (3 /3) E ~ O O 0 Non-immune (3 /3) E (3/3) E O. (2/3) E ( I / 3 ) E

"('FI. T = colon.~ form ing units. hThree sheep per treatment at specified intervals were used. ' Denotes isolation of salmonella by enrichment culture. Numbers within parentheses denote the num- ber of sheep from which the challenge organisms were isolated divided by the tolal number of sheep challenged.

Rou

te o

f im

mun

izat

ion

TAB

LE 8

Sero

logi

cal

resp

onse

of

shee

p im

mun

ized

wit

h th

e a

ro

- S

. h

~n

'an

a s

trai

n C

S234

Agg

luti

nati

on l

iter

( +

SE

M )

on d

ays

post

-im

mun

izat

ion

O a

ntig

en-s

peci

fic

H a

ntig

en-s

peci

fic

0 7

14

21

0 7

14

21

EL

ISA

til

er (

+_ S

EM

) on

day

s po

st-i

mm

uniz

atio

n

0 7

14

21

Intr

amus

cula

r N

D ~

33

+7

"*

93-+

35~

26±

7d

ND

23

+_9 ~

66

+-13

b

30 +

-I0d

N

D

639+

325

f 43

9"1z

194

1~

Ora

l N

D

ND

N

D

ND

N

D

ND

N

D

ND

N

D

ND

N

D

Uni

mm

uniz

ed

ND

N

D

ND

N

D

ND

N

D

ND

N

D

ND

N

D

ND

~ND

den

ote~

"no

te d

etec

tabl

e'.

*Sig

nifi

cant

dif

fere

nces

bet

wee

n an

tibo

dy l

iter

s: b

ver

sus

c an

~d d

( P<O

.05

L e

ver

sus

fand

g (

P<O

.O!

).

481

-+ 2

65 g

ND

N

D

Z f: "C

z f.p ,...

m

192 T.K.S. MUKKUR ET AL.

TABLE 9

The development o f cutaneous hypersensitivity in sheep immunized with the aro-S, havana strain CS234

Route of immunization Double skin-fold thickness ( m m ) at 48 h post-inoculation (mean + SEM ) on days post-immunization

7 14 21

Intramuscular 3.3 a +_ J.3 3.(P + 0.2 2.5 + 0.2 Oral i .4b+0.21 1.3b+0.4 1.6+0.5 Unimmunized I. I c + 0.5 IA b + 0.2 !.4 + 0.5

*: Significant differences between treatment groups: a versus b and c ( P < 0 . 0 5 ) .

whereas those immunized with live aro- organisms by the oral route devel- oped low antibody levels. However, only those mice immunized with the live organisms by the i.p. route showed significant development of delayed-type hypersensitivity (Table 5).

(b) Sheep. Regardless of the route of immunization, the parent S. havana strain CS4 was detectable in the ileal and caecal lymph nodes and from faeces from 7-14 days post-challenge but no organisms were detectable on days 21 and 28 (Tables 6 and 7). In contrast the above strain persisted in these tissues in non-immune sheep and was detectable even at day 28 post-challenge (Tables 6 and 7).

Serological analysis revealed that sheep immunized by the i.m. route de- veloped high O- and H-agglutination titers compared to those immunized by the oral route (Table 8 ), maximum titers being observed at 14 days p.i. Sim- ilar results were observed when the antibody levels were determined using ELISA. Sheep immunized by the i.m. route also developed significant DTH response (P<0.01) in comparison with those immunized by the oral route where no such development was noted (Table 9). Significant DTH response in sheep immunized by the i.m. route was recorded on days 7-14 post- immunization.

DISCUSSION

The finding in this investigation that S. havana could be transduced with phage P I without being made gal E suggested a rough type of lipopolysac- charide structure, confirmation of which must await analysis by gas liquid chromatography-mass spectrometry.

The immunogenic potential of the aro- S. havana strain, CS234, in sheep was evaluated by comparing the in vivo persistence of the parent fiel~ i~olate of S. havana strain CS4 in the mesemeric lymph nodes in immunized versus

GENERATION OF AROMATIC DEPENDENT ,$:,ILMONI='LL4 ll,41:4NA 193

non-immune sheep challenged orally with the above strain. While the field strain CS4 was detected even at four weeks post-challenge in nonimmune sheep, the organism was not detected at week 3 in the immunized group of animals, thus demonstrating its immunizing efficacy. A similar system of evaluating the immunizing efficacy of ~almonella vaccines in the mouse model has been a standard practice for the past several years (Collins and Carter, 1972).

The fact that mice immunized i.p. with either the live or formalin-killed aro- S. havana had high antibody titers but only those immunized with the live mutant organisms by the i.p. route developed significant delayed-type hypersensitivity and were protected against challenge with the wild-type or- ganisms, clearly suggested that development of delayed-type hypersensitivity to salmonella antigens following vaccination was an indicator of resistance to salmonellosis. This conclusion was also supported by the development of high antibody titers and DTH in intramuscularly-immunized sheep which were considered to be protected against oral challenge with the parent S. havana strain CS4. However, since orally-immunized mice or sheep were also pro- tected but developed either low or no antibody titers and DTH as reported previously in the case of aro- S. typhimurium and S. bovismorbificans vac- cines (Mukkur et al., 1987, ! 991; Begget al., 1990) the protection observed was clearly due to stimulation of the local protective immune mechanism (s) the nature of which is currently being investigated. Regardless, the results re- ported here clearly indicate that immunization with aro- S. havana can pro- mote an effective systemic and/or local protective immune response. On this basis, it should be included in any future composite vaccine for sheep where exposure to S. havana is likely.

ACKNOWLEDGEMENTS

The authors are deeply grateful for the competent technical assistance pro- vided by Ms M. Jamieson during the execution of this project.

This work was supported in part by the U.So/Australia Bilateral Science and Technology Agreement Award made to the first author, and the Rural Credits Development Fund, Reserve Bank of Australia.

REFERENCES

Bochner, B.R.. Huang, Hui-Chuan, Schieven, G.L. and Ames, B.N., 1980. Positive selection for loss of tetracycline resistance. J. Bacteriol., 143: 926-933.

Begg, A.P., Walker, K.H., Love, D.N. and Mukkur, T.K.S., 1990. Evaluation of protection against experimental salmonellosis in sheep immunised with i or 2 doses of live aromatic-depen- dent Sa#nonella o,phimurium. Aust. Vet. J., 67: 294-298.

194 T.K.S. MUKKUR ET AL.

Collins, F.M. and Carter, P.B., 1972. Comparative immunogenicity of heat-killed and living oral Sabnonella vaccines. Infect. lmmun., 6:45 !-458.

Dixon, S.F., 1980. Isolates typed by the Salmonella Reference Laboratory and the Escherichia coli Laboratory during i 979. Aust. Vet. J., 56: 457.

Engvall, E. and Perlmann, P., 1972. Enzyme-linked immunosorbent assay, ELISA. III. Quanti- ration of specific antibodies by enzyme-labeled anti-immunoglobulin in antigen-coated tubes. J. Immunol., 109: 129-135.

Hoiseth, S.R. and Stocker, B.A.D., 1981. Aromatic-dependent Sahnonella o,phimurium are non- virulent and effective as live vaccines. N,~:ure ( Lond. ), 291: 238-239.

Litchfield, J.T. and Wilcoxon, F., 1949. A simplified method of evaluating dose-effect experi- ments. J. Pharmacol. Exp. Ther., 96: 99-113.

Mukkur, T.K.S., McDowell, G.H., Stocker, B.A.D, and Lascelles, A.K., 1987. Protection against experimental salmonellosis in mice and sheep by immunisation with aromatic-dependent Sahnone!la o'phimurium. J, Med. Microbiol., 24: I 1-19.

Mukkur, T.K.S., Stocker, B.A.D. and Walker, K.H., 1991. Genetic manipulation of Sahnonella serotype Bovismorbificans to aromatic-dependence and evaluation of its vaccine potential in mice. J. Med. Microbiol. 34: 57-62.

Murray, C.J., 1984. Isolates of Salmonella and Eschelqchia coli serotyped at the Salmonella Reference Laboratory. in 1982 and 1983 from veterinary and human sources. Aust. Vet. J., 61: 273-274.

Nnalue, N.A. and Stocker. B.A.D., ! 987. The effects of O antigen character and enterobacterial common antigen content on the in vivo persistence of aromatic-dependent Salmonella sp. live-vaccine strains. Microb. Pathog., 3:3 !-44.

Smith, B.P.. Reina-Guerra, M., Hoiseth. S.K., Stocker, B.A.D.. Habasha, F., Johnson, E. and Merritt, F., 1984. Aromatic-dependent Sabnonella o'phimurium as modified live vaccines lbr cah'es. Am. J. Vet. Res., 45: 59-66.

Wray, C. and Sojka, W.J., 1977. Reviews of the progress of dairy science: bovine salmonellosis. J. Dairy Res., 44: 383-425.