INFECTION AND IMMUNITY, June 1994, p. 2562-2567 Vol. 62, No. 60019-9567/94/$04.00+0Copyright C) 1994, American Society for Microbiology

Identification of a Novel Antigen from Corynebacteriumpseudotuberculosis That Protects Sheep

against Caseous LymphadenitisJ. WALKER,I* H. J. JACKSON,' D. G. EGGLETON,2 E. N. T. MEEUSEN,' M. J. WILSON,'

AND M. R. BRANDON'Centre for Animal Biotechnology, School of Veterinary Science, The University of Melboume,'

and CSL Ltd.,2 Parkville, Victoria 3052, Australia

Received 13 December 1993/Returned for modification 29 January 1994/Accepted 1 April 1994

A 40-kDa protein antigen from Corynebacterium pseudotuberculosis has been identified by application of astrategy that employs locally derived antibody-secreting cells (ASC). ASC probes generated by culture ofASCobtained from the lymph node draining the site of infection showed a specificity restricted to a 40-kDa antigen.Analysis of immunoblots with sequential serum samples taken from sheep during the course of experimentalprimary infection with C. pseudotuberculosis also revealed the 40-kDa antigen as an early immunodominantantigen. Sheep vaccinated with two 100-,ug doses of a 40-kDa antigen preparation in aluminium hydroxideadjuvant were protected against infection with C. pseudotuberculosis, with an 82% reduction in the proportionof infected sheep and a 98% reduction in lung lesions. Sera from vaccinated sheep exhibited a strong responseonly to the 40-kDa antigen on immunoblots. These results strongly suggest that the 40-kDa antigen plays amajor role in immunity to caseous lymphadenitis.

One of the major difficulties limiting the development ofnew defined-subunit vaccines is the identification or selectionof protective antigens. Bacteria and parasites present theresearcher and the host with a complex array of immunogenicmolecules. To date, candidate vaccine antigens have beenselected on the basis of their known or putative role ininfection, e.g., bacterial toxins or essential enzymes of parasitessuch as glutathione-S-transferase (19), or on the basis of theiridentification with antibodies from immune sera. However, thecomplexity of the immune response to the pathogen is usuallyindicated by the very diverse antibody response in immunesera, and reliance on selection of molecules judged to beimmunodominant on the basis of their reactivity with serumantibody has not always been successful in identifying protec-tive antigens. For example, immunodominant membrane pro-teins of Mycoplasma hyopneumoniae (21) and Plasmodiumfalciparum (20) have not proved to be protective.

Caseous lymphadenitis (CLA) is a common and economi-cally important disease of sheep and goats, caused by infectionwith Corynebacterium pseudotuberculosis. The disease is char-acterized by the formation of encapsulated necrotic or caseouslesions that in sheep are located primarily in peripheral lymphnodes and the lungs. C. pseudotuberculosis is a facultativeintracellular pathogen that multiplies within macrophages. It isreadily phagocytosed by nonimmune macrophages, and follow-ing phagocytosis with ovine (11) and caprine (22) macro-phages, phagosomes fuse with lysosomes but bacteria continueto multiply within phagolysosomes, leading to cell death,release of bacteria, and ultimately a necrotic lesion. Theresponse of the host is to contain the pyogranuloma by theformation of a strong collagen capsule (15). The immuneresponse to natural infection has been shown to include astrong humoral component, with circulating antibodies de-tected to a large number of antigens (9). Cell-mediatedimmunity has also been implicated in control of CLA lesion

* Corresponding author. Phone: 61 3 344 7348. Fax: 61 3 347 4083.

development, with lymphocytes observed inside the collagencapsule in intimate contact with necrotic tissue (23). Theintracapsular lymphocytes have a defined architecture of threelayers, with high proportions of CD8 and of -yi T-cell subpopu-lations being striking features of the observed cellular responseto C. pseudotuberculosis infection.Two major virulence determinants of C. pseudotuberculosis

are currently recognized-a potent phospholipase D (PLD)exotoxin (14) of 31 kDa, believed important in facilitatingspread of the disease, and a toxic lipid associated with the cellwall (4, 12). Production of both of these factors by strains of C.pseudotuberculosis has been correlated with virulence (17).

Vaccination against CLA in sheep has been attempted witha variety of antigens, with variable success. Jolly (14) showedthe involvement of toxoid in protection. More recently, puri-fied toxoid has been shown to provide a similar level ofprotection to a complex array of culture supernatant-derivedantigens (5). Comparison of protection against CLA achievedwith whole-culture, culture supernatant, and whole-cell anti-gens has led others (2, 3) to conclude that cell-associatedantigens were required as well as toxoid for optimal protection.The ability of cell-associated antigens to protect lambs againstCLA has been recently confirmed by using bacterial cell wallswith muramyl dipeptide as an adjuvant (1). Hodgson et al. (13)have shown that infection of sheep with an avirulent toxin-minus strain of C. pseudotuberculosis with a deletion of thePLD toxin gene induces strong immunity to CLA, providingincontrovertible evidence that antigens other than PLD con-tribute to protection.

Despite many studies demonstrating the involvement of anantigen(s) other than toxin in protection, no such antigen hasbeen described. We report here the identification of an antigenfrom C. pseudotuberculosis by application of a strategy thatutilizes the local immune response. Others have used B cellsfrom a site of infection to derive monoclonal antibodies forantigen identification (24) but did not analyze directly thespectrum of specificities of antibodies produced by the B cells.Studies in our laboratories of the local immune response in

2562

on Novem

ber 29, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

SUBUNIT VACCINE AGAINST OVINE CASEOUS LYMPHADENITIS 2563

tissue at the time of infection, or rejection of infection, haveshown that antibody-secreting cells (ASC) obtained locallyproduce antibodies of highly restricted specificity, especiallycompared with the serum response (16). Such ASC probeshave been applied to the identification of a novel antigen, andfurthermore, we demonstrate that vaccination of sheep withthe native antigen provides a high level of protection againstexperimentally induced CLA.

MATERIALS AND METHODS

Bacterial strains and culture. C. pseudotuberculosis WA1030was grown at 37°C in brain heart infusion broth (Difco) or onthe same medium solidified with 1.5% Bacto Agar (Difco). Anumber of field isolates of C. pseudotuberculosis designatedA01 to A05 were also used in this study.

Sheep. All animals were from a CLA-free flock of Romney-Marsh sheep bred at the CSL Field-station, Woodend, Victo-ria, Australia. Adult sheep, used for immune response studies,were housed inside and fed a diet of chaff and pellets. Lambs,used for the vaccine trial, remained in paddocks.

Preparation of ASC probes. To study the local immuneresponse to infectious challenge with C. pseudotuberculosis,ASC probes were prepared by culture of ASC followinginfection. Naive adult sheep that had not previously beenvaccinated against CLA or infected with this bacterium wereused. Sheep were infected by subcutaneous injection of 1 ml ofdilute suspension of caseous material obtained from a CLAlesion; the suspension was prepared by homogenizing approx-imately 0.1 g of exudate with 1 ml of phosphate-buffered salineand contained 6.8 x 106 CFU of C. pseudotuberculosis per ml.The subcutaneous injection was given on the outer aspect ofthe right hock, resulting in a local infection, detected as aswelling within 7 days. Following infection, sequential samplesof blood were taken and sheep were killed by intravenousinjection of 30 ml of sodium pentabarbitone (Lethabarb)either 7 or 14 days after infection. The popliteal lymph nodehas been shown by previous tracer studies with Evans blue dyeto drain the area of infection. The left and right popliteallymph nodes were dissected out, and mononuclear leukocytesuspensions were prepared on Ficoll-Isopaque gradients. Themononuclear cells, containing ASC, were cultured at 2 x 106cells per ml in Dulbecco's modified essential medium (Difco)containing 10% (vol/vol) heat-inactivated fetal calf serum, 2 x10-' M 2-mercaptoethanol, and the antibiotics penicillin (100U/ml) and streptomycin (100 ,ug/ml). Mitogens were added toASC cultures at the following concentrations: pokeweed mito-gen (Gibco), 25 ,ug/ml, final concentration; lipopolysaccharide(LPS; Sigma), 50 ,ug/ml, final concentration; and dextransulfate (D.SO4; BDH) 20 ,ug/ml, final concentration. After 7days of culture at 37°C in a 5% CO2 incubator, cells wereremoved by centrifugation (1,000 x g, 10 min), and thesupernatants were retained. These supernatants obtained byculture of ASC are referred to as ASC probes.

Antigen identification. ASC probes were used to screenwhole-cell antigens of C. pseudotuberculosis by immunoblots.Washed whole cells (10 ,ul of cells per lane at an optical densityat 600 nm [OD600] of 1.0) were solubilized with loading buffer,run on 12.5% (wt/vol) sodium dodecyl sulfate (SDS)-polyacryl-amide gels (Mini-Gels; Bio-Rad) under reducing conditionsusing a discontinuous buffer system, and then electroblottedonto nitrocellulose (BA35; Schleicher & Schuell). After block-ing overnight in Tris-buffered saline (pH 7.4) with 0.05%(vol/vol) Tween 20 (TBS), blots were incubated with eitherserum (diluted 1/100 in TBS) or ASC probe (undiluted) for 1h at room temperature. After four washes in TBS, bound sheep

antibodies were detected by incubation with an anti-sheephorseradish peroxidase conjugate (Dako) diluted 1/600 in TBSfollowed by washing and incubation in diaminobenzidine sub-strate (Sigma; 2 mg/ml) and H202 (20 mM) in citrate buffer,pH 5.5.

Antigen purification. The 40-kDa antigen identified by ASCprobes was prepared from 50 liters of culture supernatant thathad been concentrated by pressure dialysis. Ammonium sul-fate was added to 50% saturation, and precipitated proteinswere removed by centrifugation. The supernatant was appliedto a phenyl-Sepharose hydrophobic chromatography column(Pharmacia) and eluted with a decreasing (NH4)2SO4 gradient.ASC probes were used to monitor fractionation of the 40-kDaantigen preparation.

Vaccination trial. Sixty-four wether lambs, approximately 6months old, were randomly allocated to one of three groupsand were identified by ear tags. Sheep in group 1 received 100jig of 40-kDa antigen preparation in aluminum hydroxide,group 2 animals received 5 ,ug of 40-kDa antigen preparationper dose given in aluminum hydroxide, and group 3 sheepreceived a placebo injection of aluminum hydroxide alone.Each animal was given two subcutaneous injections 4 weeksapart. Three weeks following the second vaccination, sheepwere bled and challenged by subcutaneous injection of awashed suspension of 2 x 106 in vitro-grown C. pseudotuber-culosis cells in the lower right back leg. Three months afterinfectious challenge, the sheep were slaughtered at a commer-cial abattoir. As ears and tags were removed before inspection,sheep were identified by consecutive numbering of the gam-brels or hangers, and these numbers were coded to ear tagnumbers for later analysis. Carcasses were examined by De-partment of Primary Industry inspectors. Lungs were removedand placed in plastic bags for subsequent scoring of lesions.Lesions were detected by palpation, and any enlarged lymphnode or swelling in tissue was examined by serial incisions.After all tissues had been examined and results had beenrecorded, gambrel numbers were decoded and group lesionscores were collated.Assay of serum antibodies. Antibody responses in sera were

assayed by enzyme-linked immunosorbent assay (ELISA).Whole cells diluted to an OD600 of 0.01 or culture supernatantor 40-kDa antigen preparation diluted to approximately 5 ,ugof protein per ml in bicarbonate buffer, pH 9.6, were dispensedin 100-,lI aliquots into 96-well microtiter plates (Nunc Immu-noplate). After coating overnight at 4°C, wells were washedand blocked with TBS for 1 h. Dilutions of sera were incubatedfor 1 h at 37°C, wells were washed, and 100-tLI aliquots ofanti-sheep peroxidase conjugate (Dako) diluted 1/5,000 in TBSwere added. Conjugate was incubated for 1 h, wells werewashed with TBS, and bound conjugate was detected with100-,u aliquots of tetramethylbenzidine (100 ,ug/ml; Aldrich)substrate with 20 mM H202. ELISA reactions were stoppedwith 50 p,l of 2 M H2SO4.

Statistical methods. Results of the vaccination trial were

analyzed by standard statistical methods (10).

RESULTS

Antigen identification. Sequential serum samples takenfrom naive (CLA-free) sheep during the course of a primaryinfection were screened by immunoblots against whole-cellantigens. Serum antibody responses are shown in Fig. la. Littleantibody response was detected during the first 4 to 5 days(Fig. la, panels i and ii). A band at approximately 28 kDaobserved with preinfection sera from naive sheep increased inintensity with sera from day 4 of infection onwards. This

VOL. 62, 1994

on Novem

ber 29, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

2564 WALKER ET AL.

a(i)

97kD -67kD -59kD -

37kD -

K'1'A

0 2 4 7Days

b(i)

00

k,

97kD-

67kD-59kD-

37kD-

a(ii)

4.-

a(iii)

IY&

(a)97kD67kD

59kD-

37kD-I4-

0 2 4 7 11 14Days

b(ii)

_k, _:^

8'^ *

DA._.N_ 4_#3#s g

{

{, K

i .-s

s CT s V);> X > 4

gk R

b(iii)

I.'kE

cI

C6

1 97kD-

67kD-59kD-

37kD-

b(iv). lI

I.w. -D

M.

FIG. 1. Analysis of sera (a) and ASC probes (b) from naive sheepwith a primary infection of C. pseudotuberculosis against whole-cellantigens. (a) Sequential serum samples from three sheep between days0 and 7 of infection (i) or days 0 and 14 of infection or serum taken atday 14 of infection (iii). Arrows indicate 40-kDa antigen. (b) ASCprobes derived from four sheep 7 days (i) and 14 days (ii, iii, and iv)after a primary infection. Arrows indicate 40-kDa antigen. ASC probeswere prepared by culturing mononuclear leukocytes obtained frompopliteal lymph nodes draining infected or contralateral uninfectedsites in the presence of either pokeweed mitogen (PWM) or LPS-D.SO4 (LPS).

antigen was identified as a major cell protein of 28 kDa on

SDS-polyacrylamide gel electrophoresis (PAGE) of wholecells (Fig. 2a, lane 1) that gave a background reaction on

immunoblots. Other minor bands did not increase in intensityfrom days 2 to 7 of infection (Fig. la, panel i). Infections of C.pseudotuberculosis that were allowed to proceed for 14 daysstimulated antibodies to a wide range of antigens, the immu-noblot patterns of sera becoming more complex from day 7onwards, as the infection progressed.

(b)

H](C)

m

FIG. 2. (a) SDS-polyacrylamide gel protein profiles of antigenpreparations of C. pseudotuberculosis. Lanes: 1, whole cells; 2, concen-trated culture supernatant (9 ,ug of protein loaded); 3, 40-kDa antigenpreparation (3 p.g of protein loaded). (b) Immunoblot of supernatantantigens (as shown in panel a, lane 2), reacted with ASC probesderived from a challenged node. ASC probe cultures were stimulatedwith either phytohemagglutinin (PHA) or LPS-D.SO4 (LPS). (c)Immunoblot of 40-kDa antigen probed with representative pre- andpostimmunization sera from a sheep vaccinated with the 40-kDaantigen preparation.

ASC probes prepared from lymph nodes taken from sheepkilled at day 7 of infection showed little reactivity on immu-noblots (Fig. lb, panel i). The background band at 28 kDaobserved on immunoblots with preinfection sera was also seenwith ASC probes prepared from lymph nodes draining infectedsites of sheep killed at day 7 (Fig. lb, panel i) and from lymphnodes draining uninfected sites (Fig. lb, panel ii). ASC probesfrom sheep killed 14 days after a primary infection showedstrong and restricted reactivity on immunoblots to whole-cellantigens (Fig. lb), especially compared with serum antibody.Probes from ASC draining the infected site were stronglyreactive, while those from the contralateral uninfected siteshowed background reactivity to the 28-kDa protein (Fig. lb).The spectrum of reactivity of ASC probes at day 14 to

antigens of C. pseudotuberculosis showed some variation be-tween sheep, with a number of bands highlighted by somesheep (Fig. lb, panels iii and iv). However, a 40-kDa antigenwas consistently and strongly recognized on immunoblots andin two sheep was the major antigen recognized (Fig. lb, panelsii and iii). Analysis of immunoblots of sera showed thatinfected sheep also had circulating antibody to the 40-kDaantigen, and sequential serum samples showed this was one ofthe earliest antigens recognized during the course of infection(Fig. la, panel ii). ASC probes also strongly recognized the40-kDa antigen in concentrated culture supernatant (Fig. 2b),with a minor band at 38 kDa also visible, which we believe tobe a smaller form of the same antigen. Polyclonal sera raised tothe 40-kDa antigen preparation also recognized the 38-kDaband, seen in Fig. 2c as the lower band of a doublet.ELISA of antibodies in scrum taken at day 14 and in ASC

probes from the same sheep to whole-cell and supernatantantigens of C. pseudotuberculosis confirmed the results ob-tained from the immunoblots (Tables 1 and 2). Antibody

4

INFECT. IMMUN.

4

Li=

C*lj cn cnCL4 0.4 a40-. 04 04

on Novem

ber 29, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

SUBUNIT VACCINE AGAINST OVINE CASEOUS LYMPHADENITIS 2565

TABLE 1. Titers in serum' assayed by ELISA using whole-celland supernatant antigens

ELISA OD459Day 7-Day infection 14-Day infection

postinfectionWhole Supernatant Whole Supernatantcells antigens cells antigens

0 0.580 0.207 0.400 0.1622 0.545 0.189 0.361 0.1604 0.544 0.204 0.362 0.1657 0.620 0.225 0.380 0.5171 l - 0.593 1.35514 0.744 1.523

All sera were assayed at a 1/100 dilution.

responses developed in sera from day 7 onwards. Antibodyreactivity of the probe from the challenged site reactedstrongly with antigens at day 14, but the ASC probe from 7-dayinfections was negative, as were probes prepared from lymphnodes draining uninfected sites. Other experiments haveshown that inclusion of mitogens in cultures leads to a higherlevel of antibody, as measured by ELISA, but does not affectthe limited spectrum of antigens recognized by ASC probes on

immunoblots.Antigen preparation. The 40-kDa antigen was fractionated

from 50 liters of culture supernatant by hydrophobic chroma-tography with a decreasing gradient of 50 to 0% saturatedammonium sulfate. The 40-kDa antigen eluted between 25 and15% saturated (NH4)2SO4 and was approximately 50% pure as

judged by scanning densitometry of Coomassie-stained SDS-PAGE gels (Fig. 2a, lane 3); this is defined as the 40-kDaantigen preparation. The major contaminating moieties were

peptides of less than 20 kDa derived from the culture medium.The large scale precluded some techniques, such as gel chro-matography, for further purification. Silver staining highlighteda red-to-rust brown diffuse band at the top of the gel. The31-kDa PLD toxin was the major protein present in bacterialculture supernatants (Fig. 2a, lane 2) but could not be detectedeither on stained gels of the 40-kDa antigen preparation or on

immunoblots of sera from sheep immunized with the 40-kDaantigen preparation; such blots showed that only the 40-kDaantigen was immunogenic (Fig. 2c). The contaminating bandsof <20 kDa visible in Fig. 2a, lane 3, were not immunogenicwhen reacted with ASC probes (Fig. 2b) or with the sheep sera

(Fig. 2c). Six field isolates of C. pseudotuberculosis were shownby immunoblotting experiments using ASC probes to alsocontain the 40-kDa antigen (not shown).Immune response to vaccination. Antibody titers of vacci-

TABLE 2. Titers in ASC probes" assayed by ELISA using whole-cell and supernatant antigens

ELISA OD450

Origin of prohe Mitogen" 7-Day infection 14-Day infection

Whole Supernatant Whole Supernatantcells antigens cells antigens

Challenged site PWM 0.079 0.034 0.092 0.408LPS-D.SO4 0.151 0.065 t).105 0.333

Unchallenged PWM 0.140 0.048 0.067 0.040site LPS-D.SO4 0.225 0.105 0.092 0.072

aASC probes were assayed at a 1/2 dilution."PWM, pokeweed mitogen.

TABLE 3. Field trial lesion scores of sheep 3 months afterinfectious challenge with C. pseuidotuibercuilosis

No. of infected No. of sheep with Mean no. of lungTreatment group sheep/no. in popliteal lymph lesions + SD

group node lesions (range)

40-kDa antigen100 ,ug 2/21" 0 0.1 0.5 (1-2)"5 ,ug 9/22 0 1.4 2.6 (1-12)"

Unvaccinated 12/22 5 6.7 + 12.6 (1-46)controls

" Significantly different from control group, see text for details.

nated sheep were measured by ELISA prior to infectiouschallenge. Following primary and secondary vaccinations,there was no difference in the group mean titers to the 40-kDaantigen of sheep vaccinated with 5 ,ug of antigen (mean OD450+ standard deviation = 0.318 + 0.166) compared with those ofsheep vaccinated with 100 ,ug of antigen (0.294 + 0.177).Unvaccinated control sheep had significantly lower titers(OD450 ± standard deviation = 0.075 + 0.122). Immunoblotsof sera from vaccinated sheep showed a strong response onlyto the 40-kDa antigen (Fig. 2c).

Lesion scores. Infectious challenge with C. pseudotuberculo-sis in the right hock 3 months prior to slaughter resulted in asatisfactory level of infection, with 55% of control animalsdeveloping lesions (Table 3).

Vaccination with 100-[wg doses of the 40-kDa antigen prep-aration had a major effect on development of disease andsignificantly reduced the infection rate from 55 to 5% (z testvalue = 3.58, P < 0.002). Vaccination with 5-pLg doses ofantigen resulted in a small and insignificant reduction in theinfection rate to 41% (z test value = 1.25, P < 0.20).Lung lesion scores provided the best measure of protection

achieved by vaccination with the 40-kDa antigen preparation.More control sheep developed lung lesions than developedlesions in lymph nodes, and those infected often had multiplelesions in the lung (Table 3). Of the nine control sheep withlung lesions, five did not have lesions in popliteal or otherlymph nodes.

Vaccination with 100-pLg doses of the 40-kDa antigen prep-aration gave a significant 98% reduction in the incidence oflung lesions (unpaired t test, t = 2.046, 41 df, P < 0.025; Table3). The 5-pLg dose sheep showed a 79% reduction (t = 1.932, 42df, P < 0.05). Vaccination also tended to reduce the lesion size,with control group lesions having an average diameter of 67mm and lesions in the 5- and 100-p.g dose groups havingaverage diameters of 37 and 26 mm, respectively. These datashow that in addition to preventing actual infection as reflectedin the number of lung lesions, vaccination with the 40-kDaantigen preparation also limited the development of estab-lished foci of infection.The right popliteal lymph node drains the site of infectious

challenge; however, only 5 of the 22 right popliteal lymphnodes from control sheep showed lesion development. Vacci-nation of sheep with the 40-kDa antigen preparation at 5- and100-,ug dose levels prevented infection in the right popliteallymph node (Table 3). The only other lymph nodes with lesiondevelopment were the prescapular nodes. Three control sheepand one sheep in the 5-p.g dose group had lesions in theprescapular lymph nodes. No sheep in the 100-,ug dose grouphad detectable lesions in peripheral or visceral lymph nodes.

VOL. 62, 1994

on Novem

ber 29, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

2566 WALKER ET AL.

DISCUSSION

Antibody probes produced by ASC taken from the lymphnodes draining the site of infection with C. pseudotuberculosisshowed restricted antigenic reactivity on immunoblots, partic-ularly compared with the diverse spectrum of antigens identi-fied by antibodies from serum present in the sheep at the timeof sampling of lymph nodes. This restricted response identifieda novel 40-kDa antigen from C. pseudotuberculosis. The kinet-ics of the serum immune response of naive sheep also high-lighted the 40-kDa antigen as one of the predominant antigensrecognized (Fig. la) at day 7 postinfection. By day 14, whenASC from lymph nodes were obtained, the serum response washeterogeneous and the 40-kDa antigen response was no longerobvious over other antigens on immunoblots. Taken together,the restricted response of ASC probes and the more hetero-geneous response in serum show that the greater part of serumantibody to an infectious challenge is not being synthesized inthe draining lymph nodes. The pool of antibodies in serumspecific for antigens other than the 40-kDa antigen are pre-sumably being synthesized in other lymphoid tissues, like thespleen.

Vaccination of sheep with 100 ,ug of the 40-kDa antigenpreparation gave up to 82% reduction in the proportion ofinfected sheep. This level of protection compares favorablywith the levels of protection achieved with currently availablecommercial vaccines (5, 7). The lower dose level of 5 ,ug of40-kDa antigen preparation did not provide a significant levelof protection in terms of infected sheep but did significantlyreduce the incidence of lung lesions. This lower level ofprotection was not reflected in lower group mean titers ofantibody, with group means of the 5- and 100-,ug dose groupsnot being significantly different. In addition, there was nodiscernible relationship between lung lesion scores and anti-body titers. These data strongly implicate cellular immunity asthe primary mechanism of immune rejection of infection.Immunoblot analysis of sera from sheep vaccinated with the

40-kDa antigen preparation showed a strong antibody re-sponse restricted to the 40-kDa protein (Fig. 2c). The lower-molecular-mass components of approximately 20 and less than15 kDa in the 40-kDa antigen preparation were not immuno-genic. These were assumed to be derived from mediumcomponents of bovine origin. There was no response detectedto the 31-kDa PLD toxin. We have not analyzed the cellularimmune response to vaccination. The significant protectionachieved by vaccination suggests that the 40-kDa antigen playsa major role in protection against CLA. To date, the only otherrecognized protective antigen has been PLD, which is also amajor determinant of virulence (5, 13, 14, 17).

Currently, vaccines against CLA are based on culture super-natants containing inactivated PLD toxin (5). Protection isattributed to responses to the PLD toxoid, and vaccinationwith purified inactivated PLD has achieved approximately 79%reduction in infection, similar to the 75 to 80% levels ofprotection achieved with the crude complex commercial vac-cines (5-7). In other studies, we have found that antibodyresponses of sheep vaccinated with one particular commer-cially available vaccine, Glanvac (CSL Ltd), showed low andinconsistent titers to the 40-kDa antigen. Variability in cellularor humoral responses to the 40-kDa antigen may provide animmunological basis for the intermediate levels of protectionachieved with commercial and experimental vaccines (1, 2, 5,7). As discussed previously, cell-mediated immunity may play arole in protection against intracellular infection with C.pseudotuberculosis. The absence of a direct correlation be-tween antibody titers to the 40-kDa antigen and disease status

indicates either that antibody to the 40-kDa antigen is involvedin protection through more indirect mechanisms, for example,antibody-dependent cell-mediated cytotoxicity, or that anti-body is not involved in protection. Cytotoxic T-cell responsesto the 40-kDa antigen may mediate protection against CLA, ashas been shown in other intracellular bacterial infections.There is also the possibility that T-cell responses to antigensthat do not elicit an antibody response mediate protection.Further studies are required to elucidate the contributions ofhumoral and cellular immunity in protection against ovineCLA.The function and location of the 40-kDa antigen are not yet

known. The antigen is present at significant levels in culturesupernatant and in detergent extracts of whole cells and isprobably noncovalently attached to the cell or is activelysecreted. The higher reactivity in ELISAs of ASC probesspecific for the 40-kDa antigen with supernatant antigens thanwith cellular antigens (Table 2) supports the concept that the40-kDa antigen is a secreted protein. Studies are continuing toclone the gene for and express the 40-kDa antigen anddetermine its role in immunity and pathogenesis. The identi-fication of a major protective antigen of C. pseudotuberculosisother than PLD should contribute to the development of anefficacious defined vaccine against CLA.

Studies of serum antibody responses following natural infec-tion and vaccination have highlighted a wide range of antigensfrom C. pseudotuberculosis (8, 9). These previous studiesshowed a number of antigens in the 40-kDa region on immu-noblots, but the strong response to a very wide range ofantigens prevented the selection of any particular antigen asbeing important in immunity. Identification of protective anti-gens by immunoscreening using serum antibodies has yieldedimportant candidate vaccine antigens, for example, paramyo-sin of Schistosoma mansoni (18). In previous studies of CLA(9) and in mycobacterial infections, which like CLA are causedby facultative intracellular pathogens, effective vaccine anti-gens have not yet been identified. We have shown here thatASC probes generated by ASC taken from lymph nodesdraining the site of a challenge infection with C. pseudotuber-culosis identified a highly protective antigen. This study dem-onstrates that generation of ASC probes from a site ofinfection provides a rational approach to the identification ofcandidate vaccine antigens which may have application to awide range of diseases of both humans and animals.

ACKNOWLEDGMENTS

This work was supported by the Meat Research Corp. (Australia).We thank Vern Bowles for critically reviewing the manuscript,

Christine Kerr for typing, and Ken Snibson for photography.

REFERENCES

1. Brogden, K. A., L. Chedid, R. C. Cutlip, H. D. Lehmkuhl, and J.Sacks. 1990. Effect of MDP on immunogenicity of Corynebacte-rium pseudotuberculosis whole-cell vaccines in mice and lambs.Am. J. Vet. Res. 51:200-202.

2. Burrell, D. H. 1983. Caseous lymphadenitis vaccine. N. S. W. Vet.Proc. 1983:53-57.

3. Cameron, C. M., J. L. Minnaar, and M. R. Purdon. 1969.Immunising properties of Corynebacterium pseudotuberculosis cellwalls. Onderstepoort J. Vet. Res. 36:211-216.

4. Carne, H. R., N. Wickham, and J. C. Kater. 1956. A toxic lipidfrom the surface of Corynebacterium ovis. Nature (London) 178:701-702.

5. Eggleton, D. G., C. V. Doidge, H. D. Middleton, and D. W. Minty.1991. Immunisation against ovine caseous lymphadenitis: efficacy

INFEc-r. IMMUN.

on Novem

ber 29, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

SUBUNIT VACCINE AGAINST OVINE CASEOUS LYMPHADENITIS 2567

of monocomponent Corynebacterium pseudotuberculosis toxoidvaccine and combined clostridial-corynebacterium vaccine. Aust.Vet. J. 68:320-321.

6. Eggleton, D. G., J. A. Haynes, H. D. Middleton, and J. C. Cox.1991. Immunisation against ovine caseous lymphadenitis: correla-tion between Corynebacterium pseudotuberculosis toxoid contentand protective efficacy in combined clostridial-corynebacteriumvaccines. Aust. Vet. J. 68:322-325.

7. Eggleton, D. G., H. D. Middleton, C. V. Doidge, and D. W. Minty.1991. Immunisation against ovine caseous lymphadenitis: compar-ison of Corynebacterium pseudotuberculosis vaccines with andwithout bacterial cells. Aust. Vet. J. 68:317-319.

8. Ellis, J. A., D. A. Hawk, K. W. Mills, and D. L. Pratt. 1991. Antigenspecificity and activity of ovine antibodies induced by immunisa-tion with Corynebacterium pseudotuberculosis culture filtrate. Vet.Immunol. Immunopathol. 28:303-316.

9. Ellis, J. A., D. A. Hawk, K. W. Mills, and D. L. Pratt. 1991. Antigenspecificity of antibody responses to Corynebacterium pseudotuber-culosis in naturally infected sheep with caseous lymphadenitis.Vet. Immunol. Immunopath. 28:289-301.

10. Glantz, S. A. 1987. Primer of biostatistics. McGraw-Hill Book Co.,New York.

11. Hard, G. C. 1972. Examination by electron microscopy of theinteraction between peritoneal phagocytes and Corynebacteriumovis. J. Med. Microbiol. 5:483-491.

12. Hard, G. C. 1975. Comparative toxic effect of the surface lipid ofCorynebacterium ovis on peritoneal macrophages. Infect. Immun.12:1439-1449.

13. Hodgson, A. L. M., J. Krywult, L. A. Corner, J. S. Rothel, and A. J.Radford. 1992. Rational attenuation of Corynebacterium pseudo-tuberculosis: potential cheesy gland vaccine and live deliveryvehicle. Infect. Immun. 60:2900-2905.

14. Jolly, R. D. 1965. The pathogenic action of the exotoxin ofCorynebacterium ovis. J. Comp. Pathol. 75:417-431.

15. Jubb, K. V. F. 1985. Vol. 3, chapter 2, p. 200-201. In K. V. F. Jubb,P. C. Kennedy, and N. Palmer (ed.), Pathology of domestic

animals. Academic Press Ltd., London.16. Meeusen, E. N. T., and M. R. Brandon. 1994. The use of antibody

secreting cell probes to reveal tissue restricted immune responsesduring infection. Eur. J. Immunol. 24:469-475.

17. Muckle, C. A., and C. L. Gyles. 1983. Relation of lipid content andexotoxin production to virulence of Corynebacterium pseudotuber-culosis in mice. Am. J. Vet. Res. 44:1149-1153.

18. Pearce, E. J., S. L. James, J. P. Dalton, A. Barral, C. Ramos, M.Strand, and A. Sher. 1986. Immunochemical characterisation andpurification of Sm-97, a Schistosoma mansoni antigen monospe-cifically recognised by antibodies from mice protectively immu-nised with a non-living vaccine. J. Immunol. 137:3593-3600.

19. Smith, D. B., K. M. Davern, P. G. Board, W. V. Tiu, E. G. Garcia,and G. F. Mitchell. 1986. The Mr 26,000 antigen of Schistosomajaponicum recognised by resistant WEHI 129/J mice is a parasiteglutathione S-transferase. Proc. Natl. Acad. Sci. USA 83:8703-8707.

20. Smythe, J. A., R. L. Coppel, G. V. Brown, R. Ramasamy, D. J.Kemp, and R. F. Anders. 1988. Identification of two integralmembrane proteins of Plasmodium falciparum. Proc. Natl. Acad.Sci. USA 85:5195-5199.

21. Strasser, M., J. Frey, G. Besetti, M. Kobisch, and J. Nicolet. 1991.Cloning and expression of a species-specific early immunogenic36-kilodalton protein of Mycoplasma hyopneumoniae in Esche-richia coli. Infect. Immun. 59:1217-1222.

22. Tashjian, J. J., and S. G. Campbell. 1983. Interaction betweencaprine macrophages and Corynebacterium pseudotuberculosis: anelectron microscopic study. Am. J. Vet. Res. 44:690-693.

23. Walker, J., H. Jackson, M. R. Brandon, and E. Meeusen. 1991.Lymphocyte subpopulations in pyogranulomas of caseous lymph-adenitis. Clin. Exp. Immunol. 86:13-18.

24. Wallis, R. S., S. L. Aide, D. V. Havlir, M. H. Tahmasseb, T. M.Daniel, and J. J. ElIner. 1989. Identification of antigens ofMycobacterium tuberculosis using human monoclonal antibodies. J.Clin. Invest. 84:214-219.

VOL. 62, 1994

on Novem

ber 29, 2020 by guesthttp://iai.asm

.org/D

ownloaded from