Vol. 53, No. 8APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Aug. 1987, p. 1721-17290099-2240/87/081721-09$02.00/0Copyright © 1987, American Society for Microbiology

Clinical Hypersensitivity to Specific Aerosol Challenge inParenterally Immunized Calves

FRANK G. R. TAYLOR,'* COLETTE D. R. JONES,2t AND JOHN BOURNE'Department of Veterinary Medicine' and Department ofAnimal Husbandry,2 University of Bristol, Langford, Bristol BS18

7DU, England

Received 30 December 1986/Accepted 11 May 1987

This paper reports an experiment designed to demonstrate that the calf lung can be sensitized to a specificrespirable challenge following parenteral immunization with a nonliving antigen (human serum albumin). Thepossibility that immune-mediated injury could subsequently interfere with nonspecific mucosal defenses wasalso investigated by infecting calves with Pasteurella haemolytica after the antigen challenge and assessingpulmonary clearance of the organism. The results indicated that specific aerosol challenge produces reversiblesigns of respiratory hypersensitivity and that persistence of incidental infection in the upper respiratory tractis potentiated. Since the calves were sensitized by an immunization regime which imitated conventionalvaccination, this study highlights the potential dangers of inactivated parenteral respiratory vaccines.

In common with other mucosal surfaces, the lung iscapable of developing an immune response following local orsystemic presentation of antigen. The subsequent responseto specific challenge serves to protect the mucosal surface,but under certain conditions of sensitization the outcomeappears to be harmful rather than beneficial. For example,Wilkie (16) demonstrated that hypersensitivity pneumonitiswas produced in calves following repeated exposure toaerosols of Micropolyspora faeni whether or not they hadbeen systemically primed. Furthermore, parenteral immuni-zation of calves with killed Pasteurella haemolytica resultedin an increased incidence of pneumonic pasteurellosis atchallenge (5). Subsequent comparison of vaccination routes,using a killed culture of P. haemolytica, demonstrated thatparenteral immunization was consistent with the develop-ment of pulmonary lesions at challenge, whereas local im-munization was not (17). In human medicine, pneumonitisresulting from occupational contact with airborne allergensis well documented and parenteral vaccination with inacti-vated organisms is known to have potentiated respiratorysyncytial virus and Mycoplasma pneumoniae infections (3).In the case of respiratory syncytial virus, it was proposedthat high levels of circulating immunoglobulin G (IgG) anti-body were induced in the absence of local humoral andcellular immunity, culminating in immune-mediated hyper-sensitivity at challenge (2). This type of reaction is consid-ered to be the consequence of complement activation andneutrophil influx following immune complex formation in thelower respiratory tract, which at its extreme is manifestedclinically as pneumonitis (15). In the case of vaccine-potentiated calf pasteurellosis, Wilkie et al. (17) proposed adifferent mechanism in which opsonization and phagocytosisof cytotoxic bacteria caused the death of macrophages, thuscompromising lung defenses. They suggested that this proc-ess was enhanced by the absence of a local IgA antibodyresponse.

* Corresponding author.t Present address: Dunn Clinical Nutrition Centre, Cambridge

CB2 1QL, England.

In addition to the local injury which results from immune-mediated disease, the disruption of mucosal defenses mayalso predispose unrelated opportunist infections. A greaterfrequency of viral respiratory infections has been recordedin asthmatic children compared with their nonasthmaticsiblings (12), and in the pig hypersensitivity to food antigensin the gut has been shown to allow the proliferation of entericpathogens (11). The possibility exists, therefore, that im-mune-mediated lung disease in calves may be potentiated byparenteral vaccination with nonliving antigens and that sub-sequent mucosal hypersensitivity may permit opportunistinfections.We have previously shown that aerosol challenge with

human serum albumin (HSA) in parenterally immunizedcalves caused an increase in the neutrophil population of freelung cells (14). However, we did not observe clinical signsconsistent with an immune-mediated inflammatory re-sponse. Since the duration of challenge was short (20 min)and only 4.2% of the aerosolized protein mass was concen-trated within respirable particles (aerodynamic size, 0 to 3,um), we supposed that the challenge to the lower respiratorytract may have been insufficient to provoke inflammation tothe point of hypersensitivity. This paper reports an experi-ment designed to investigate the following: (i) whethercalves immunized parenterally on two occasions with anonliving antigen will produce clinical signs of pneumonitisfollowing provocation with respirable antigen; (ii) whetherthe effect is antigen specific; and (iii) whether nonspecificmucosal defenses against opportunist infection are compro-mised by the event. Briefly, calves which had been immu-nized or sham immunized by parenteral injection with HSAwere challenged with an aerosol containing a high concen-tration of the antigen in respirable form. Each calf was thenclinically assessed, and after 24 h all were infected byintrapulmonary inoculation of nalidixic acid-resistant P.haemolytica (NARPH), the clearance of which was mnoni-tored by nasolaryngeal swabs. The aerosol challenge wassubsequently repeated at a later date to compare observa-tions with those of the first challenge. At slaughter, allmacroscopic lung lesions were scored and swabs were taken

1721

on Novem

ber 25, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

APPL. ENVIRON. MICROBIOL.

FIG. 1. Aerosol delivery system. 1, Antigen solution reservoir; 2, DeVilbiss ultrasonic nebulizer; 3, large droplet trap; 4, high-efficiencyparticulate absorber (HEPA filter), with arrow indicating air inlet; 5, flow meter; 6, 4-in. (10-cm)-diameter flexible foil ducting; 7, aerodynamicparticle sizer; 8, calf with air-tight face mask; 9, roughing filter; 10, flow meter; 11, vacuum pump.

at all levels of the respiratory tract to determine the distri-bution of residual NARPH.

MATERIALS AND METHODSAnimals. Eight Friesian bull calves were purchased at 1 to

2 weeks of age from market and individually penned beyondbodily contact within the same housing unit. Alphabeticalidentities (A to H) were ascribed to the calves. The bodyweight of each calf was recorded weekly from the time ofimmunization to the point of slaughter.

Immunization procedure. At 3 to 4 weeks of age fourcalves were immunized by intragluteal injection of 40 mg ofHSA (Sigma Chemical Co.) in saline emulsified in Freundcomplete adjuvant (Difco Laboratories). The injections weregiven on two occasions at an interval of 2 weeks. Theremaining four were sham immunized on the same occasionswith equivalent volumes of saline emulsified in Freundcomplete adjuvant.Grouping of calves. Because of the numbers of observa-

tions required at each stage of the experiment, it was notpossible to study all eight calves at one time. Consequently,two smaller groups were formed (A to D and E to H), eachconsisting of two sham-immunized and two immunizedmembers. Comparisons were then made within each groupfollowing challenge and infection procedures.Serum antibody studies. Blood samples were taken from all

calves at regular intervals to assess the development ofanti-HSA IgG serum antibody by an indirect enzyme-linkedimmunosorbent assay system (13). Briefly, polyvinyl chlo-ride microtitration plates (Dynatech Laboratories, Inc.)were coated with HSA, washed, and incubated with serialtwofold dilutions of sera. Specific IgG antibody was thendetected by incubating the washed plate with a pigantibovine IgG produced at this laboratory. This was in turnlabeled by incubating the washed plate with a sheep antipigIgG conjugated to alkaline phosphatase (Sigma). Finally, theattached conjugate was quantified in the washed plate byincubation with phosphatase substrate (Sigma). Colorationwas read at 405 nm on an automatic reader (TitertekMultiscan MC; Flow Laboratories, Inc.). Endpoint titerswere expressed as the last detectable coloration in the seriesof serum dilutions which exceeded the background controlby a factor of at least 2.

Anti-P. haernolytica activity was evaluated by using thesame assay system. An antigen extract of the organism was

prepared by boiling a 48-h culture of the bacteria in brainheart infusion (Difco) and extracting the supernatant bycentrifugation (8). A 1:20 dilution of the supernatant wasthen used to coat the plates, and the assay was repeated asbefore.

Aerosol challenge. Calves inhaled a polydisperse liquidaerosol of HSA containing 248 mg of protein solution m3 ofair for a period of 25 min. The aerosol was generated from a0.5% solution of HSA in saline by an ultrasonic nebulizer(DeVilbiss model 65), and its characteristics were measuredin real time, using an aerodynamic particle sizer (TSI Incor-porated). Each calf inhaled an aerosol with an average massmedian aerodynamic diameter of 5.7 pm (geometric standarddeviation = 1.7). The antigen mass concentrated withinrespirable particles (<3.05 ,um) was on average 11.8% of thetotal generated by the nebulizer. Compared to our formerstudy (14), this represented an increase in the respirableantigen load by a factor of 2.8. To achieve this, turbulenceand particle deposition within the system were reduced byincreasing the diameter of the delivery ducting and reducingthe linear flow of aerosol from 200 to 28 cm/s. The aerosoldelivery system is illustrated in Fig. 1.The calves were challenged in two groups, A to D (group

I) and E to H (group II). Group I animals were challenged onday 32 following the start of vaccination, when serumantibody titers wre shown to be high, and group II waschallenged on day 39. At 24 h after challenge, each calfreceived an intrapulmonary inoculation of NARPH. Allcalves were rechallenged with HSA, using the same tech-nique, on days 47 (group II) and 54 (group I).On day 62 group I was subjected to a polydisperse liquid

aerosol of 0.5% bovine serum albumin (Sigma) in salinesolution to assess the effect of a nonspecific antigen chal-lenge. In this instance the aerosol had an average massmedian aerodynamic diameter of 6.25 p.m (geometric stan-dard deviation = 1.7), and the protein mass concentratedwithin respirable particles was on average 7.7% of thatgenerated by the nebulizer.

Clinical examination. Several clinical parameters wereassessed at the time of aerosol challenge, on occasionsduring the 8 h thereafter, and once again 24 h later: (i)incidence of coughing; (ii) respiratory rate per minute (RR);(iii) dyspnea, defined here as respiratory distress featuringmarked expiratory lift, head extension, and mouth breathingin addition to an increase in RR; (iv) rectal temperature; (v)

1722 TAYLOR ET AL.

on Novem

ber 25, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

HYPERSENSITIVITY TO AEROSOL IN IMMUNIZED CALVES

0

._-

._

00

(U

I

c

C.0

00

-J

4

3

2

E G~~EFdO d13 jd 39 Inf(d 40) d, 47 d6.63

FIG. 2. Development of serum IgG antibody titers following (open columns) parenteral immunization and subsequent challenge with HSAand (shaded columns) pulmonary infection with NARPH. All calves were immunized on days 0 and 13. Group I calves were challenged withHSA aerosol on day 32, infected with NARPH on day 33, and rechallenged with HSA on day 54. Group II calves received the same treatmentson days 39, 40, and 47, respectively. d., Day; Inf., infection; SI., slaughter. Note absence of antibody to HSA in sham-immunized calves A,D, E, and G at first challenge and absence of antibody to NARPH in all calves prior to infection.

demeanor, defined here as being alert or dull in nature (analert calf demonstrated interest by getting up if recumbentand approaching an attendant; a dull calf was disinterested inthe attendant and preferred recumbency even when encour-

aged to rise); (vi) appetite. All calves had been weaned at 6weeks of age prior to aerosol exposure and calf nuts were fedonly twice daily in addition to ad lib hay, so that a handful ofnuts would normally be readily accepted at any time duringthe day. This act was therefore used as a criterion of appetite5 to 6 h after challenge. Lung auscultation was attemptedfollowing aerosol challenge, but because of its subjectivenature it was deemed unreliable as an indicator of underlyingclinical pathology.

P. haemolytica infection and clearance studies. The P.haemolytica type Al used in the experiment had beenisolated from the lung of a pneumonic calf and was kindlydonated by N. J. L. Gilmour of the Moredun ResearchInstitute. Nalidixic acid resistance was induced after twosubcultures, and the final strain was grown up in nutrientbroth (Oxoid) and stored in 4-ml aliquots at -70°C. Thisculture was used to seed 200 ml of nutrient broth number II

(Lab M) the day before the calves were infected. P.haemolytica in the 20-h broth was centrifuged (700 x g) for12 min and suspended in 60 ml of sterile saline, giving a finalconcentration of 1.15 x 108 CFU/ml. Each calf received 10ml of the suspension into the lower respiratory tract followedby 20 ml of saline. The inoculum was delivered through a4-mm (outer diameter) sterile perspex tube introducedthrough the nasal cavity to a distance of 60 to 80 cm from theexternal nares.For the purpose of assessing bacterial clearance following

infection, it was assumed that reduced clearance within thelower lungs of compromised calves would result in an

increased bacterial load being carried up the mucociliaryescalator. This load was quantified by laryngeal swabbing atregular intervals for up to 56 h after infection. Swabs, 30 cm

(Medical Wire and Equipment Co. Ltd.), were passedthrough the ventral meatus to the back of the nasopharynxand positioned in the larynx as the calf attempted to swal-low, at which time the calf produced a faint roar on breath-ing. The swab was left in this position for 30 s.To quantify the bacteria sampled, each swab tip was

placed in a bijou containing 4 ml of 1:4-strength Ringersolution with glass beads and mixed on a Vortex mixer. Thesuspension was serially diluted and plated out onto 6% horseblood agar (Lab M blood agar base) with and withoutnalidixic acid (50 jig/ml). The plates were incubated aerobi-cally at 37°C for 28 h, and the colonies were counted by themethod of Miles et al. (10).Postmortem studies. All calves were slaughtered during the

period 67 to 69 days following the start of vaccination,equivalent to 28 to 35 days postinfection with NARPH.Immediately after death, each calf was bled out and thelarynx, trachea, and lungs were removed intact. The distri-bution of residual NARPH within the respiratory tract wasdetermined by culturing swabs taken at seven selected sitesbetween the upper and lower respiratory tracts in each calf.Gross lung pathology was compared between calves asreported previously (7). Briefly, macroscopic lesions in eachlung were scored out of 10 in each of twelve sections ofapproximately equal size, three to each diaphragmatic lobeand one to each other lobe.

Statistical analysis. Student's t test was used for compari-son of results; P values of <0.05 were considered significant.

RESULTS

Serum antibody studies. The development of serum IgGantibody is shown in Fig. 2, which also illustrates theexperimental protocol. By the time of their first HSA aerosolchallenge the immunized calves (B, C, F, and H) haddeveloped high levels of circulating antibody. Subsequent to

VOL. 53, 1987 1723

on Novem

ber 25, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

1724 TAYLOR ET AL.

SHAM-IMMUNISED IMMUNISED

B

959 I I of r---" Ir**

0 2 4 6 8 24 0 2 4 6

FIG. 3. Respiratory rates before, during,group II.

Time (hours)and after the first HSA aerosol challenge of noninfected calves. Symbols: A, group I; *,

aerosol challenges, all of the nonimmunized calves (A, D, E,and G) produced low levels of circulating antibody. None ofthe calves demonstrated circulating antibody to the antigenextract of P. haemolytica type Al prior to infection, but alldeveloped titers thereafter. No difference was apparentbetween sham-immunized and immunized calves in ability tomount a systemic IgG antibody response to infection.

First aerosol challenge. Clinical observations on the firstHSA aerosol challenge of noninfected calves are shown inFig. 3 and 4 and Table 1. Despite a lapse of 7 days between

SHAM-IMMUNISED

the first challenges of groups I and II (days 32 and 39,respectively), the clinical findings were essentially the sameand are recorded together here. The number of observationsof RR and temperature were increased after aerosolization ofgroup II as a result of the trends indicated by group I.

Before challenge, there was no significant difference be-tween the RRs recorded in sham-immunized and immunizedcalves (Fig. 3). During the aerosolization period the RR wasassessed at 5, 10, and 20 min (mean plots are shown in Fig.3 at time zero), and these values indicated significant in-

IMMUNISED

A

G

0

2 4 6 8 240 2 4 6 8 24

Time (hours)FIG. 4. Rectal temperatures before and after the first HSA aerosol challenge of noninfected calves. Symbols: A, group I; *, group Il.

120 -

a)

c

E0._

a)t-

co

a0..

00'O.

cr

80 -

40 -

-41-8 24

105 -

0-

.-

C1)

-

Q

C-

a)

103 -

101 -

APPL. ENVIRON. MICRO1310L.

on Novem

ber 25, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

HYPERSENSITIVITY TO AEROSOL IN IMMUNIZED CALVES

TABLE 1. Clinical observations made at the first HSA aerosol challenge of noninfected calves and during the 0- to 8-h period thereafter"

Animal group/ Challenge Postchallengedesignation Mean RR No. of coughs RR (h)" Demeanor Appetite

Sham immunizedIA 39 32 (5) Alert NormalD 43 42 (4) Alert Normal

IIE 49 2 46 (8) Alert NormalG 46 52 (4 & 6) Alert Normal

ImmunizedIB 60 80 (6) Dull; preferred recumbency AnorexiaC 52 90 (4)' Dull; persistent recumbency Anorexia

IIF 61 7 116 (4) Dull; preferred recumbency PoorH 38 64 (7) Alert Normal

aRespiratory rates per minute (RR) at challenge are given as the mean of three observations. Postchallenge respiratory rates are given as the maximum recordedfor each calf. Demeanor and appetite were recorded at 5 h postchallenge.bHours in parentheses are number of hours postchallenge that maximum RR were recorded.' Plus dyspnoea.

creases from prechallenge levels in both sham-immunizedand immunized calves (P < 0.01; df = 14 in each case).However, in immunized calves the response was greaterthan in sham-immunized animals (P < 0.05; df = 22). In the0- to 8-h period following challenge, all immunized calvesdemonstrated RRs which exceeded at some time any of theobservations for sham-immunized calves. Unfortunately,there were insufficient observations to establish the statisti-

10

8

CD

460

-J

10

cal significance of this finding at specific times. By 24 h theRRs of sham-immunized calves had returned to pre-challenge values, but in immunized calves the rates were stillgreater than pre-challenge values (P < 0.02; df = 6) despitea marked decline from immediate postchallenge levels.Changes in rectal temperatures from prechallenge values

were less consistent (Fig. 4). Unfortunately, too few record-ings were available to analyze temperature changes between

* * HSA immunisedSham-immunised

30 50 60

Time (hours)FIG. 5. Numbers of CFU of NARPH isolated from nasolaryngeal swabs after intrapulmonary infection.

VOL. 53, 1987 1725

on Novem

ber 25, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

1726 TAYLOR ET AL.

SHAM-IMMUNISED IMMUNISED

A ~IG H

0 2 4 6 8 24 0 2 4 6 8 24

Time (hours)FIG. 6. Respiratory rates before, during, and after the second HS

group II.

calves at the various times, and although immunized calvesF and H demonstrated high temperatures between 7 and 8 hpost-challenge, no overall trend was apparent.Of the remaining clinical evaluations, only two of the eight

calves produced coughs during the first aerosol challenge,and all sham-immunized calves displayed an alert demeanorand normal appetite at 5 h postchallenge, whereas three ofthe four immunized calves were dull with depressed appe-tites (Table 1). These signs disappeared within 24 h. Nochanges were noted in consistency of feces or frequency ofdefecation.

A challenge of NARPH-infected calves. Symbols: A, group I, 0,

P. haemolytica infection and clearance. At 24 h after thefirst HSA aerosol challenge, all calves received a lunginoculum of NARPH, the clearance of which was monitoredby nasolaryngeal swabs. Despite a lapse of 7 days betweeninfection of groups I and II (days 33 and 40, respectively),the bacterial loads recovered were essentially the same andare recorded together here. There was no statistical differ-ence between the numbers of NARPH isolated from sham-immunized and immunized calves at any point over 56 h(Fig. 5). Following infection the rectal temperatures in allcalves rose to a peak of 104.8 to 106.7°F (40.4 to 41.5°C)

TABLE 2. Clinical observations made at the second HSA aerosol challenge of NARPH-infected calves and during the 0- to 8-hperiod thereafter'

Animal group/ Challenge Postchallengedesignation Mean RR No. of coughs RR (h)b Demeanor Appetite

Sham immunizedIA 42 1 48 (3, 5, 7) Alert NormalDc 81 4 82 (2) Alert Poor

IIE 44 10 46 (8) Alert NormalG 32 1 46 (4) Alert Normal

ImmunizedIB 64 37 74 (3) Alert NormalC 61 42 83 (3) Dull; persistent recumbency Poor

IIF 61 14 110 (3) Alert NormalH 44 16 64 (3) Alert Normal

Respiratory rates per minute (RR) at challenge are given as the mean of three observations. Postchallenge respiratory rates are given as the maximum recordedfor each calf. Demeanor and appetite were recorded at 5 h postchallenge.bHours in parentheses are number of hours postchallenge that maximum RR were recorded.C Pneumonic calf: prechallenge RR = 78.

120 -

0

c

._E

L-

o

0..0-

I._

0CU

0.

C)

80 -

40 -

APPL. ENVIRON. MICROBIOL.

on Novem

ber 25, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

HYPERSENSITIVITY TO AEROSOL IN IMMUNIZED CALVES

SHAM-IMMUNISED

A

0 2I I I, Ir--"4 6 8 24

IMMUNISED

B

c

0 24

Time (hours)FIG. 7. Respiratory rates before, during, and after bovine serum albumin aerosol challenge of group I NARPH-infected calves.

between 8 and 10 h and returned to resting levels by 70 h.The only calf to show overt signs of pneumonia followinginfection was calf D, which developed a chronically raisedrespiratory rate. This persisted through to slaughter, but therectal temperature was never greater than that recorded inother calves.Second aerosol challenge. Clinical observations on the

second HSA aerosol challenge of calves infected withNARPH are shown in Fig. 6 and Table 2. Again, despite alapse between the second challenges of groups I and II (days54 and 47, respectively), the clinical findings were essentiallythe same and are recorded together.

Before aerosolization, there was no significant differencebetween the RRs recorded in sham-immunized and immu-nized calves (Fig. 6). During challenge the RRs recorded at5, 10, and 20 min indicated significant increases aboveprechallenge levels in immunized calves only (mean plots areshown in Fig. 6 at time zero: P < 0.01; df = 14). In the 0- to8-h period following challenge, all immunized calves dem-onstrated RRs which exceeded those observed at each time

point for sham-immunized calves A, G, and E. During theperiod 2 to 5 h these differences were significant (P < 0.05;df = 5 at each of four time points). A fourth sham-immunizedcalf (D) began with a prechallenge rate of 78 per min, havingdeveloped a chronically high RR, presumably as a result ofP. haemolytica infection. This rate showed no tendency toincrease during the postchallenge period, but it was excludedfrom the statistical comparison with immunized calves. By24 h the RRs of all sham-immunized calves were no greaterthan either prechallenge or challenge levels, whereas thoseof immunized calves still exceeded their prechallenge levels(P < 0.05; df = 6).

Despite more frequent recording of rectal temperaturesduring 0 to 8 h postchallenge, there was no significantdifference between sham-immunized and immunized calvesat any time point. The data are not presented.

During the second aerosol challenge a greater incidence ofcoughing was recorded in all calves (Table 2). In immunizedcalves the incidence was significantly greater than in sham-immunized controls (P < 0.05; df = 6). Unlike the first

TABLE 3. Distribution of NARPH recovered from the respiratory tract of calves at slaughter together with gross lung lesion scoresa

Sham-immunized animals Immunized animalsSite

A (34) D (35) E (28) G (29) B (35) C (35) F (29) H (29)

Lung tissue - - + b - - - - -

Bronchi - - - - + + +Trachea - - - + - + +Larynx N' - - - - + + +Tonsillar sinus N - + + + + + +Pharynx N - + + - - + +Ventral meatus N - - + - + - +

Total lung score 20 23 6 5 8 27 26 14

a Number in parentheses is number of days between NARPH infection and slaughter.bIn abscess.C N, Samples not taken.

120-

0C..

0

0

0

0

h.

CL

0

con

C:

80-

40

VOL. 53, 1987 1727

on Novem

ber 25, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

1728 TAYLOR ET AL.

challenge, only one of the four immunized calves (C) dis-played a dull demeanor and poor appetite at 5 h despiteincreases in RR.The isolation of NARPH by nasolaryngeal swabbing was

resumed in all calves following the second HSA challenge.All demonstrated a transient rise in the numbers isolated,which peaked between 6 and 12 h and returned to prechal-lenge levels by 18 h. Again, no significant difference wasdiscerned between sham-immunized and immunized calves.Rectal temperatures did not rise significantly nor differbetween the groups during the period of swabbing. The dataare not presented.

Challenge with bovine serum albumin aerosol. Since theimmunized calves of group I (B and C) had shown such ahigh incidence of coughing during the second HSA challenge(Table 2), this group was subjected to an aerosol challenge of0.5% bovine serum albumin for 25 min 8 days later (day 62),to ensure that the effect was not due to nonspecific aerosolirritation. A total of three coughs were then recorded duringaerosolization of sham-immunized calves A and D, but nonewere recorded in the immunized pair. In addition, the 0- to8-h RRs in the immunized pair showed no tendency toexceed those of the controls (Fig. 7), in contrast to previouschallenges with specific antigen (Fig. 3 and 6).

Live weight gains. No significant difference in individualweight gains was demonstrable between sham-immunizedand immunized calves during the experimental period. Thedata are not presented.Postmortem findings. Immediately after death NARPH

was isolated with greater frequency from tissues of the upperrespiratory tracts of immunized calves (Table 3). However,lung scoring indicated no difference in the extent of grosslung lesions between the two populations. A marked featureof these gross lesions was discrete abscessation in one or theother of the diaphragmatic lobes in seven of the eight calves.

DISCUSSION

The short-term reversible responses of immunized calvesto respirable aerosol challenge as indicated by increases inRR (Fig. 3 and 6) and changes in demeanor and appetite(Table 1) suggest immune-mediated respiratory disease. Thelack of such responses in sham-immunized calves demon-strates that the phenomenon is antigen specific. Experimen-tal lung hypersensitivity has been reported in cattle onprevious occasions, but the sensitizing regimes involvedeither multiple parenteral immunization (4) or multiple res-piratory exposures (16). In the current study the immuniza-tion procedure mimics conventional vaccination regimesinasmuch as parenteral injections of a nonliving antigen inadjuvant were given on two occasions at an interval of 2weeks. The findings are therefore more relevant to vaccine-induced disease than previous reports of experimentalhypersensitivity.

Using an identical immunization regime, we have alreadydemonstrated that serum IgG antibody activity correlateswith titers found in lung wash and that aerosol challenge withHSA causes an increase in the neutrophil population of freelung cells, but we were unable to demonstrate clinical signsof pneumonitis (14). In the current study, we increased our

respirable antigen load by a factor of 2.8 and subsequentlyprovoked a clinical response. This indicates that hypersen-sitivity is as much a function of the respirable antigen load atchallenge as the regime used for sensitization. One explana-tion for the difference in our results is that in this studyhigher levels of respirable antigen combined with IgG anti-

body in the lower respiratory tract to produce more immunecomplexes and in consequence a greater inflammatory re-sponse. However, we have not characterized the nature ofthe hypersensitivity response here; indeed, there seems tobe more than one type involved.The respiratory rates of immunized calves were signifi-

cantly raised above those of controls during the first HSAchallenge (P < 0.05; Fig. 3). In the next 2 to 8 h theydemonstrated a trend to further increases, but there wereinsufficient observations for analysis at each time point.During the second challenge RRs rose significantly aboveprechallenge values in immunized calves only (P < 0.01; Fig.6) and thereafter exceeded those of control calves A, G, andE at each of four time points between 2 and 5 h (P < 0.05 foreach time point). Despite reductions of RR in immunizedcalves 24 h after each challenge, RR remained higher thantheir prechallenge levels (P < 0.02 [Fig. 3]; P < 0.05 [Fig.6]). These time intervals suggest hypersensitivity responseswhich contain immediate, immune complex, and delayed-type components as defined by Gel and Coombs (6). Inaddition, the second HSA challenge evoked coughing in allcalves during aerosolization, although the incidence wassignificantly greater in immunized calves (P < 0.05; Table 2).Not only is this evidence of an immediate-type response, butit also indicates a change in response to challenge in allcalves following the preliminary aerosol exposure.The results of our investigation into the effect of hyper-

sensitivity on mucosal defenses are equivocal. In our exper-iment there is no difference in the clearance of NARPHbetween calves as judged by nasolaryngeal swabbing (Fig.5). However, this criterion of infection is open to criticismsince it remains uncertain whether inadequate clearance inthe lower respiratory tract would increase the level ofbacteria appearing on the mucociliary escalator. Moreover,both nasal and laryngeal surfaces are sampled by this tech-nique, and quantification of bacteria arriving from the lowerrespiratory tract could be adversely influenced by contami-nants originating from points of colonization in the upperrespiratory tract. However, there was no significant differ-ence between sham-immunized and immunized calves interms of bacterial disease as judged by clinical assessment,live-weight gain, or gross lung lesions at postmortemexamination.

At postmortem examination seven of the eight calvesdisplayed diaphragmatic lobe abscesses, indicating that ourtechnique of infecting the lungs produced highly localizedreactions. Such lesions would not mimic a natural diffuseinfection and could therefore affect the criteria used here toassess nonspecific defenses. Nevertheless, at slaugther therewas a difference between calves in terms of the persistenceof infection in that NARPH was more consistently isolatedfrom the upper respiratory tract of those which had previ-ously demonstrated hypersensitivity. Furthermore, the pres-ence or absence of infection could not be correlated toNARPH antibody activity (Fig. 2). It is also interesting thatcalves C and F, which produced the highest recorded RRsafter HSA challenge (Fig. 3 and 6), produced the mostconsistent upper respiratory tract isolates and the highestlung scores (Table 3). In a small study such as this theevidence is necessarily scant, but it does suggest that, if thelung becomes sensitized, suitable conditions of challengemay render tissues more susceptible to colonization by thechallenge organism or incidental opportunists. Against this,the only animal to display overt pneumonia following infec-tion was sham-immunized calf D. This calf developed apersistently raised respiratory rate (Fig. 6 and 7) and pro-

APPL. ENVIRON. MICROBIOL.

on Novem

ber 25, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

HYPERSENSITIVITY TO AEROSOL IN IMMUNIZED CALVES

duced the lowest weight gain of all between infection andslaughter (data not shown). Even so, despite a high terminallung score, no NARPH could be isolated from the respira-tory tract (Table 3).

Wilkie and colleagues (17) ascribed the adverse effects ofkilled parenteral P. haemolytica vaccines to alveolar macro-phage cytotoxicity following specific opsonization and en-hanced ingestion of toxigenic bacteria. While the importanceof macrophage cytotoxicity in the pathogenesis of pasteurel-losis is undeniable, the experiment reported here indicatesthat other adverse immune-mediated mechanisms may alsobe potentiated following parenteral vaccination. Further-more, these changes were apparent after a challenge dura-tion of only 25 min, and it is interesting to speculate whateffect prolonged exposure would have had on the resultingpneumonitis. In a large field study, Martin et al. (9) found apositive correlation between increased mortality in cattleand the use of various types of respiratory vaccine; how-ever, these were not exclusively inactivated vaccines norwere they all administered parenterally.

In conclusion, our results indicate the potential dangers ofusing nonliving parenteral vaccines in calves which will beexposed to high concentrations of respirable antigen chal-lenge. Such conditions are likely to be a feature of over-crowding and poor environmental hygiene. Furthermore, ifhypersensitivity to challenge is potentiated by the absence ofa local response to parenteral vaccination (2), it is clearlydesirable that respiratory vaccines promote local as well assystemic immunity. Local immunity is best achieved bymucosal presentation of antigen in a replicating form whichis capable of colonizing the epithelium (1). Commercialmucosal vaccines are invariably attenuated for field use, butthis is not permissible when an organism may revert tovirulence. In such cases the protective immunity afforded bylocal administration of the killed organism or its subunitsshould be investigated as an alternative to parenteral admin-istration alone.

ACKNOWLEDGMENTS

We thank A. J. F. Webster for use of aerobiology facilities in theDepartment of Animal Husbandry at Bristol University and M. I.Gilmour for preparing the antigen extract of P. haemolytica.

This work was part of a study financed by the Agricultural andFood Research Council (AG7/104).

LITERATURE CITED1. Bourne, F, J., and T. J. Newby. 1981. Mucosal immunity. In

Practice 3(5):5-11.2. Chanock, R. M., A. Z. Kapikian, J. Mills, H. W. Kim, and R. H.

Parrott. 1970. Influence of immunological factors in respiratory

syncytial virus disease. Arch. Environ. Health 21:347-355.3. Craighead, J. E. 1975. Report of a workshop: disease accentu-

ation after immunization with inactivated microbial vaccines. J.Infect. Dis. 131:749-754.

4. Dungworth, D. L. 1965. The pulmonary response of sensitizedcattle to aerosol administration of antigen, p. N1-N1S. InProceedings of the Symposium on Acute Bovine PulmonaryEmphysema, Laramie, Wyoming, 1965. University of Wyo-ming, Laramie.

5. Friend, S. C. E., B. N. Wilkie, R. G. Thomson, and D. A.Barnum. 1977. Bovine pneumonic pasteurellosis: experimentalinduction in vaccinated and nonvaccinated calves. Can. J.Comp. Med 41:77-83.

6. Gell, P. G. H., and R. R. A. Coombs. 1968. Classification ofallergic reactions responsible for clinical hypersensitivity anddisease, p. 575-596. In Clinical aspects of immunology, 2nd ed.Blackwell Scientific Publications, Ltd., Oxford.

7. Jones, C. D. R., and A. J, F. Webster. 1984. Relationshipsbetween counts of nasopharyngeal bacteria, temperature, hu-midity and lung lesions in veal calves. Res. Vet. Sci. 37:132-137.

8. Klaassen, J. M., B. L. Bernard, and R. F. DiGiacomo. 1985.Enzyme-linked immunosorbent assay for immunoglobulin Gantibody to Pasteurella multocida in rabbits. J. Clin. Microbiol.21:617-621.

9. Martin, S. W., A. H. Meek, D. G. Davis, J. A. Johnson, andR. A. Curtis. 1981. Factors associated with morbidity andmortality of feedlot calves: the Bruce County beef project, yeartwo. Can. J. Comp. Med. 45:103-112.

10. Miles, A. A., S. A. Misra, and J. 0. Irwin. 1938. The estimationof the bactericidal power of the blood. J. Hyg. 38:732-749.

11. Miller, B. G., T. J. Newby, C. R. Stokes, D. J. Hampson, P. J.Brown, and F. J. Bourne. 1984. The importance of dietaryantigen in the cause of postweaning diarrhoea in pigs. Am. J.Vet. Res. 45:1730-1733.

12. Minor, T. E., J. W. Baker, E. C. Dick, A. N. DeMeo, J. J.Ouellette, M. Cohen, and C. E. Reed. 1974. Greater frequency ofviral respiratory infections in asthmatic children as comparedwith their nonasthmatic siblings. J. Pediatr. 85:472-477.

13. Taylor, F. G. R., D. Patel, and F. J. Bourne. 1983. Comparisonof sensitivities of ELISA and radioimmunoassay for detectionof class-specific antibody in mouse serum. J. Immunol. Methods65:65-73.

14. Taylor, F. G. R., M. Singleton, and F. J. Bourne. 1985. Effect ofaerosol challenge on a population of free lung cells in parenter-ally immunized calves. Res. Vet. Sci. 38:243-245.

15. Turner-Warwick, M. 1978. Immunology of the lung. Curr. Top.Immunol. 10:30-32.

16. Wilkie, B. N. 1976. Humoral and cell-mediated immune re-sponse of cattle to Micropolysporafaeni and clinical response toaerosol challenge. Int. Arch. Allergy Appl. Immunol. 50:359-373.

17. Wilkie, B. N., R. J. F. Markham, and P. E. Shewen. 1980.Response of calves to lung challenge exposure with Pasteurellahaemolytica after parenteral or pulmonary immunization. Am.J. Vet. Res. 41:1773-1778.

VOL, 53, 1987 1729

on Novem

ber 25, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from