INFECTION AND IMMUNrrY, July 1973, p. 83-90Copyright 0 1973 American Society for Microbiology

Vol. 8, No. 1Printed in U.S.A.

Immunoglobulin and Histamine-SensitivityResponse of Mice to Live Bordetella pertussis

BERNARD D. GELLER' AND MARGARET PITTMAN2

Laboratory of Bacterial Products, Division of Biologics Standards, National Institutes of Health, Bethesda,Maryland 20014.

Received for publication 8 February 1973

Mice injected intranasally (it.n.) and intraperitoneally (i.p.) with a nonlethaldose (2.5 x 105 colony-forming units) of live Bordetella pertussis were examinedfor 50 days for infection, respiratory tract immunoglobulins (Ig), changes in serumIg, and histamine sensitivity. With mice infected it.n., respiratory infectionmarkedly declined between day 20 and day 30. Ig classes (A, Gl, G25, G2b, but no

M), which had specificity for B. pertussis, were present in tracheobronchial wash(TBW) by day 15; by day 50, TBW immunodiffusion and immunoelectrophoreticprecipitin bands were more intense. A sharp rise in serum IgA after day 30 wasthe only significant change relative to controls among the five serum Ig examined.A high degree of histamine sensitivity developed by day 15 to 20 and persisted forthe 50 days. With mice inoculated i.p., no bacteria were recovered, no Ig or onlytraces were found in TBW and IgA only was specific, and no significant changesin the serum Ig relative to controls occurred. Histamine sensitivity developedsomewhat more slowly and to a lesser degree than in it.n.-injected mice but per-sisted for the 50 days. A similar small number of killed bacteria (pertussis vac-

cine) injected it.n. or i.p. likewise induced slowly developing histamine sensitivityin contrast to published reports of 4 to 5 day peak sensitivity and decline followingi.p. injection of 109 or more killed bacteria.

Whooping cough (pertussis), the diseasecaused by Bordetella pertussis, has been effec-tively controlled by intramuscular or subcu-taneous immunization with standardized per-tussis vaccine (14). The immunological mecha-nism involved in prophylaxis against this localrespiratory infection is not understood. Pittman(14) has postulated that, although the diseasedoes not develop, infection may not be pre-vented. If permeability of the respiratory tract,like in the experimental mouse intracerebral in-fection, is altered after infection, and if protec-tive antibodies gain access to and kill the bac-teria, then the characteristic symptoms of thedisease do not develop. Iida et al. (6) foundthat, with the use of immunofluorescence, anti-pertussis serum (equine) inoculated intraperi-toneally (i.p.) into intracerebrally infected micedid not contact the bacteria, largely localizedamong the cilia of the ependymal cells lining theventricles, until vascular permeability had beenincreased. Bacteria had to reach a certain

'Present address: Department of Pediatrics, Los AngelesCounty-USC Medical Center, Los Angeles, Calif. 90032.

2Present address: 3133 Connecticut Ave., N.W., Washing-ton, D.C. 20008.

concentration before effecting the change in thebrain barrier.As in prophylaxis, the nature of the mech-

anisms of recovery from the disease has notbeen fully explained. A number of papers onpassive protection in the mouse support the roleof immunoglobulins. Dolby and Dolby, (4) re-ported that 7S and 19S globulins mixed withchallenge culture were equally effective againstlung infection in the mouse, whereas 7S was 100times more effective than 19S against braininfection. Kendrick et al. (7) likewise found thatthe protective activity against the brain infec-tion was in the immunoglobulin class G (IgG)fraction of rabbit Ig. The effectiveness of thedifferent classes of Ig in man has not beenstudied adequately.

In an approach to a better understanding ofactive protection in pertussis, we studied, withthe use of the mouse respiratory infection model(1, cf. 14), the classes of respiratory and serumIg present during infection and recovery. Com-parisons were made with Ig that were presentafter i.p. injection of live B. pertussis. Thedevelopment and duration of histamine sensiti-zation was also followed. The mouse model

83

on June 30, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

GELLER AND PITTMAN

resembles the human infection in several as-pects. The duration of infection is similar, andthe onset and duration of histamine sensitivityin the mouse parallels paroxysmal coughing inthe child (13).

MATERIALS AND METHODSMice. N:NIH(SW) strain female mice, approxi-

mately 1 month old and weighing 14 to 17 g, werereceived from the Rodent and Rabbit ProductionSection, Division of Research Services, National Insti-tutes of Health. They were housed 6, 10, or 16 per cageand given food and water ad libitum.

Bacterial techniques. The B. pertussis culture,strain no. 18323, was prepared by the method de-scribed for the challenge in the pertussis-vaccinepotency assay (17). A 17- to 20-h culture grown onBordet-Gengou (B-G) agar slants containing about15% defibrinated rabbit blood was suspended in asolution of 1.0% Protolysate (a pancreatic enzymedigest of casein, Mead Johnson & Co., Evansville,Ind.) and 0.6% NaCl, filtered through cotton andadjusted by spectrophotometry to 10 U.S. opacityunits (OPU) (17), then diluted to provide an inoculumof 10-3 OPU in a volume of 1 minim (approximately0.06 ml) for either intranasal (it.n.) or i.p. injection.(The U.S. Opacity Standard has an assigned value of10 units. It was originally adjusted to be equivalent inturbidity to 10 x 109 [direct count] bacteria in agedpertussis vaccine. Suspensions of freshly harvestedlive bacteria of equal turbidity have around 6 x 109bacteria, and colony counts are lower.) The inoculumcontained approximately 2.5 x 105 colony-formingunits (CFU; about 6 x 105 bacteria) and was nonle-thal.One drop of a tracheobronchial wash (TBW), a cut

surface of a lung lobe, and peritoneal fluid werecultured on plates of B-G agar containing no peptone.The lung fragment was rubbed on the surface and leftin situ. Peritoneal fluid was transferred with a cottonswab. Cultures were incubated for 3 to 5 days.Identity of B. pertussis was determined by typicalcolonial and bacterial morphology.

Sera. Blood obtained from the tail vein after a briefwarming of the mouse under an incandescent lampwas collected in polyethylene tubes (400 gliters) andspun for 5 min. Serum was withdrawn and storedindividually or in pools of equal volumes per mouse at-15 C.The reference normal mouse serum consisted of a

pool from N:NIH(SW) mice. It was distributed insmall samples before storage. A freshly thawed por-tion was used for each test. Residues were discarded.The levels of Ig classes M, A, G,, G2a, and G2b (1gM,IgA, IgG,I IgG2., IgG2b) in this reference were deter-mined by radial immunodiffusion by R. Asofsky, ofthe National Institute of Allergy and Infectious Dis-eases.

Goat antisera to mouse myeloma proteins, mademonospecific for IgM, IgA, IgG,, IgG,a, and IgG,b byadsorption, were used in the identification tests forspecific globulins. They were kindly supplied by R.Asofsky.TBW. After bleeding, the mouse was anesthetized

with sodium pentobarbital. The trachea was exposed,hemisected, and intubated with fine-gauge polyethyl-ene tubing (PE 10) which was tied in place withsuture. By use of a tuberculin syringe fitted with a26-gauge needle and filled with 1 ml of cold saline, thetracheobronchial tree was washed by gently injectingand withdrawing the saline. Slight tension exerted onthe trachea via the tubing insured a maximal returnof the injected fluid. In experiment I, one drop of theTBW from each animal was cultured for B. pertussisand the remaining TBW per group were pooled. Eachpool was concentrated in a collodion bag to approxi-mately one-tenth the original volume and examinedfor presence of the five subclasses of Ig. In experimentHI, individual unconcentrated TBW were tested forpresence of Ig.

Ig determinations. The presence of Ig in serumand TBW was determined by immunodiffusion andimmunoelectrophoresis. The quantitation of the im-munoglobulins was by radial immunodiffusion (RID)on cellulose acetate (18) or on agar (9), or both. Inaddition, individual TBW in experiment III wereexamined by immunofluorescent microscopy as fol-lows.

Smears of a suspension of a 20-h culture of B.pertussis were made on microscope slides, air-dried,and heat-fixed. The TBW from each mouse was addedto five slides. The slides were incubated for 20 min ina moist chamber, then washed twice in 200 ml of freshsaline containing 0.2% gelatin for 10 min per wash.Next, monospecific goat antiserum IgA, IgM, IgG,,IgG,a, or IgG,b was added to one of the five slides. Theslides were incubated for 20 min and washed twice.Normal goat serum was applied to a control culture-smear slide. Finally, diisothiocyanate-labeled anti-goat serum (rabbit; Cappel Laboratories, Inc.) wasadded to each slide. Slides were incubated for 20 minand washed three times. They were examined by useof a Zeiss ultraviolet microscope fitted for reflected-light fluorescent microscopy.Histamine sensitization. In experiment I, groups

of 16 mice per study group at the respective bleedingtimes were tested for histamine sensitivity by i.p.injection of 100 mg of histamine diphosphate per kg ofmouse weight (0.36 mg of histamine base/10 g ofmouse weight). In two subsequent experiments, thesensitizing effect of the small dose of living bacteriaand the same number of killed bacteria were com-pared. Pertussis vaccine lot 7b was used. Deathsusually occurred within 2 h. Final results were re-corded at 24 h and were reported as the percentage ofdeaths per group of 16 mice.Experinent I. Two groups of 124 mice each were

inoculated it.n. or i.p., respectively, with the dose ofculture described above. A control group of 156 micewas not inoculated. Before inoculation and 5, 10, 15,20, 30, and 50 days thereafter, four mice per groupwere bled. The five classes of Ig of the pooled sera pergroup were determined by RID, the TBW were cul-tured individually, and pools per group were tested forpresence of Ig classes by immunodiffusion and im-munoelectrophoresis. At day 5 and at each subse-quent period, 16 additional mice from each groupwere tested for sensitivity to histamine.Experiment II. Three groups of five mice each

84 INFECT. IMMUNMr

on June 30, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

MOUSE PERTUSSIS IMMUNOGLOBULINS

were bled. Then one group was inoculated with theculture it.n., another group was inoculated i.p., andthe third group was left untreated. Each animal was

bled weekly for 7 weeks. The Ig level of the serum poolper group per bleeding time were determined as inexperiment I. Since mice were not sacrificed afterbleeding, no cultures were made.Experiment III. Five mice per group were inocu-

lated either it.n. or i.p. with the culture. On day 15,TBW were obtained and examined individually with-out concentration by immunodiffusion, immunoelec-trophoresis, and immunofluoroscopy for presence ofIg classes.

RESULTS

B. pertussis infection. In experiment I, in-tranasal instillation of the nonlethal dose of B.pertussis caused a high incidence of respiratorytract infection which remained for 20 days, thendeclined (Table 1). On day 30 only a feworganisms were recovered from one of four mice.On day 50 all cultures were negative. At no timewas B. pertussis recovered from i.p.-inoculatedor control mice.Serum Ig. In experiment 1 the measured

levels of five serum Ig classes of the three studygroups of mice from 0 to 50 days are shown inFig. 1. Although there were rather wide varia-tions between the readings of each Ig class pergroup, there was an upward trend in theamounts of each class in the sera of the mice inall test groups. The differences between therespective Ig levels of it.n., i.p., and controlmice were not significant except for the IgAlevel of the it.n. mice. With the latter, the levelrose from 0.89 mg/ml on day 30 to 2.61 mg/ml on

day 50 (Fig. 1 and Table 2).In experiment II, instead of using separate

groups of mice per bleeding as in experiment I,the same five mice per group were bled weeklythroughout the experiment. The measurementsof Ig classes (Fig. 2) varied more widely thanthose in experiment I. In general, however, theresults of the two experiments were in agree-

ment. The rise of the IgA level of the it.n. mice

of experiment II from 0.34 mg/ml on day 28 to2.34 mg/ml on day 42 was the significant finding(Fig. 2 and Table 2).TBW Ig. In experiment I, the immunoelec-

trophoretic and immunodiffusion tests demon-strated the presence of Ig in the pooled concen-trates of TBW from it.n.-infected mice only.The results given in Fig. 3 illustrate the pres-ence and absence of Ig in the TBW from it.n.and i.p. mice, respectively. In both types oftests the bands for each of the four Ig classeswere more intense with day-50 than day-15TBW. At no time was IgM detected. Tests ofTBW from control mice (not shown) were nega-tive.

In experiment Ill, five individual unconcen-

trated TBW collected on day 15 from it.n. micehad the same four classes of Ig as did the pooledTBW in experiment I. In contrast, in experi-ment III the TBW from three of five i.p. miceshowed faint bands for IgG2a and IgG2b and one

showed a band for IgG, and for IgA. Althoughthe same technique was used in each experi-ment, serum source of the small amounts of Igin TBW from the i.p. mice in experiment HIcannot be excluded. There were marked differ-ences in the numbers and intensities of theprecipitin bands in the immunoelectrophoreticpattems of the TBW from individual it.n. andi.p. mice (Fig. 4).The immunofluorescent test of the individual

TBW from it.n. mice confirmed the presence ofthe four classes of Ig and also their specificityfor B. pertussis and the lack of IgM (Table 3).With the TBW from i.p. mice, IgA only was

bound to B. pertussis and in trace amount.Histamine sensitization. The it.n. mice of

experiment I gradually developed histaminesusceptibility which was at a peak of 94 to 88%between days 15 and 30 and then declined to69% by day 50 (Fig. 5). Sensitization of the i.p.mice developed more slowly and reached a lowerpeak of 56% on day 20. Thereafter sensitivityremained almost constant through day 50. This

TABLE 1. Recovery of organisms from mice inoculated with B. pertussis, experiment I

Route of inocu- Cultures positive/4 mice culturedalation Source of culturelationOb 5 10 15 20 30 50

Intranasal Tracheobronchial wash 0 3 4 (2)c 4 4 1 0Lung smear 0 3 3 4 (3)c 3 ld 0

Intraperitoneal Swab of peritoneal cavity 0 0 0 0 00 0 0

a Cultures of control mice caged in close proximity to inoculated mice were always negative.b Day after inoculation.c Number in parentheses is number of individual cultures that showed a mixture of B. pertussis and other

bacteria.d Only two colonies were present in the culture.

85VOL. 8, 1973

on June 30, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

GELLER AND PWTMAN

- W.%-0.6

0.4F0.3

0 110 5 10 15 20 30 50

DAYFIG. 1. Experiment I: Serum Ig levels of mice from

0 to 50 days after intranasal or intraperitoneal injec-tion of a nonlethal dose of live B. pertussis. Each pointrepresents a pool of sera from four sacrificed mice pertreatment group.

sensitization of the i.p. mice was unexpected.Induction of histamine sensitivity in mice

usually has been demonstrated by i.p. injectionof about 109 or more killed bacteria. Sensitivityreaches a peak in 4 to 5 days and then declines(8). Sensitization with a later peak by a smallnumber of live bacteria given it.n. has beenattributed to multiplication of the bacteria to a

sensitizing concentration (13). Because there is

little or no multiplication of B. pertussis in theperitoneal cavity, the relatively small dose of2.5 x 105 CFU given i.p. had not been expectedto induce sensitization, much less sensitizationof prolonged duration.Two experiments were performed to compare

the sensitizing action of the same number of liveand dead bacteria (Fig. 6). The live bacteriainjected it.n. induced a response that wassimilar to the one shown in Fig. 5, with peak onday 20, whereas sensitization induced by livebacteria injected i.p. was lower than shown inFig. 5: the peak of 31% was on day 28 to 32 whenthe test was terminated. A similar pattern wasinduced by the vaccine injected it.n. However,the vaccine given i.p. induced a level of sensitiv-ity (53%) similar to the level (56%) induced bylive bacteria injected i.p. (Fig. 5), but thedecline appeared to be more rapid. Furthertesting would be required to determine whetherthe difference observed were due to chanceselection of mice of different susceptibility tosensitization. It is definite, however, that asmall number of live or killed B. pertussis cellsare capable of inducing late histamine sensitiv-ity by either i.p. or it.n. route of injection.

DISCUSSIONThe B. pertussis infection of the respiratory

tract of mice followed the course describedpreviously for this model (e.g., 13, 15). The bac-teria were recovered in cultures of the TBW andlungs through 20 days, then the bacteria de-clined markedly by day 30. The duration ofinfection resembled that of human pertussis.

TABLE 2. Serum immunoglobulins of mice 49 to 50days after B. pertussis inoculated intrarnasally (it.n.)

and intraperitoneally (i.p.)

Immuno- Expt I Expt IIglobulin Con- j ~~Con-class it.n. i.p. Col it.n. -C ol

IgG2b 1.7a 2.0 1.1 0.47 0.55 0.65IgG23 2.2 2.0 2.5 1.30 1.00 1.60IgGi 1.1 1.1 0.6 1.80 1.80 1.90Total IgG 5.00 5.10 4.20 3.57 3.45 4.15

IgA 2.61b 0.95 1.05 2.34b c 0.61 0.67IgM 0.55 0.55 0.57 0.15 0.16 0.09Total Ig 8.16 6.60 5.82 6.06 4.22 4.91

IgA (%) 32.0 14.4 18.0 38.6 14.4 13.6a Measured in milligrams per milliliter.b Significantly different from IgA levels of i.p. and

control mice (P < 0.01).c Day 42 serum.

86 INFECT. IMMUNITY

on June 30, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

MOUSE PERTUSSIS IMMUNOGLOBULINS

2.0 IgG2b antibodies were specifically bound to B.-i__ pertussis. IgA gave the most intense reaction. In

1.0 contrast, no Ig was demonstrated in pooled0.8 __ TBW concentrates from the i.p.-inoculated0.6 mice in experiment I, but in experiment m0.4 some individual TBW had traces of one or two

IgG subclasses and IgA. Only IgA was specifi-0.2~~~~~... A# cally bound to B. pertussis.0.2 -i - con/rol Both culture-inoculated and control mice

0o1 showed an upward trend in the amounts of five30 classes of Ig in the serum. Maturation was one

IgG20 influential factor. Mice were about 4 weeks old\y Lat the beginning of the experiments. At 50 days

-.... the measured quantities of the respective Ig080 " classes did not differ significantly between the08 / it.n., i.p., and control mice except for IgA

(Table 2). Differences in the levels between3.0 igGi experiment I and experiment II may have been2.0 I influenced by the repeated weekly bleedings. In

/i\,. both experiments the most striking and signifi-cant observation was the rise in serum IgA in1I.0

0.8 / .. . the it.n. mice after day 30; it constituted moreE 0.6 - \ than 30% of the total Ig. Although antibody

04 - 'xV.l' activity of the serum globulins was not deter-mined, the high postinfection rise in serum IgA

0.2 appears to be correlated with the increase in2.0 - IgA .?. serum IgA antibody in cholera patients. North-

rup and Hossain (12) observed more than a

.1.030-fold increase in serum IgA antibody against.08 - /\ / 'Vibrio cholerae after disappearance of the0.6 - /t < j vibrios from the intestines of the patients.0°4 S lx \ r Increase in total serum IgA was minimum0.4 /.../ A.~\ X:/ (Northrup, personal communication). Charac-

V teristic of both diseases is the localization of the0.2 infectious bacteria on a membrane surface0.4 Igm which is capable of producing IgA. In each

infection the stimulus for the marked rise in0.2 serum IgA was no doubt related to the site of the

infection.0.1 1 Information on the antipertussis specificity of006 / serum globulins of both mouse and man is0.06 needed. Wiedermannova and Wiedermann (19)0.04r) 14 21 2 3 42 49 measured by immunoelectrophoresis the serum

DAY IgG, IgM, and IgA levels in children 3, 5, and 60FIG. 2. Experiment II: Serum Ig levels of mice bled weeks after beginning of infection. No signifi-

weekly from 0 to 49 days after intranasal or in- cant differences relative to time or to agglutinintraperitoneal injection of a nonlethal dose of B. titers were found. However, B. pertussis anti-pertussis. Each point represents a pool of sera from body activity of the classes of Ig was notfive mice which were bled consecutively throughout determined.the experiment.

In our study the time sequence of the disap-pearance of B. pertussis and the presence of

The study of the Ig classes provided new apparent increasing concentrations of respira-information. IgG,, IgG2,, IgG2b, and IgA, but no tory immunoglobulins suggest that Ig played anIgM, appeared in the respiratory tract of the antibacterial role. North (11) found that serumit.n.-inoculated mice by day 15 and increased from it.n.-infected mice bled on days 30 to 50with time as evidenced by the increased inten- passively protected mice, whereas 14-day serumsity of the immunoprecipitin bands. Immuno- was ineffective. IgG antibody is antibacterial andfluorescence showed that the four classes of passively protective for mice against B.

VOL. 8, 1973 87

on June 30, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

GELLER AND PITTMAN

day 50

/ /bG½a 1g6i

\itn

IgG2b IgA

day 15 I...,-.

itn 'k

I \FIG. 3. Experiment I: Immunoelectrophoretic precipitin bands of four classes of Ig in pooled tracheobron-

chial wash (TBW) collected 15 days (bottom) and 50 days (top) after intranasal (itn) injection of live B.pertussis. No Ig precipitin bands developed with the TBW from the intraperitoneally(ip)-injected mice.Troughs contained anti-mouse serum (rabbit).

.ITNIP

|~~~ ~~~~~~~~~~. ..... ......-

|~~~~~~T_N Ii,

FIG. 4. Experiment III: Immunoelectrophoretic patterns of individual tracheobronchial washes from fiveintranasally (ITN) and five intraperitoneally (IP) injected mice collected on day 15. Troughs were filled withanti-mouse serum (rabbit).

INFw=. ImuNrryr88

on June 30, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

MOUSE PERTUSSIS IMMUNOGLOBULINS

TABLE 3. Fluorescence of Ig bound to B. pertussis insmears treated with TBWfrom mice 15 days after B.

pertussis inoculationa

Anti-mouse serum it.n. ap.(goat)

IgG, 4

IgG2. +IgG2b + _IgA + + 4

IgM -

a Diisothiocyanate-labeled antigoat serum (rabbit)was used as indicator.

C.,

<.:, w 1000-

Fo cn

SCJ 0-

=Ei

con

IMJ

,...-1/1xon___

I.......4 .......

......

'l- - , .

,_____

305 10 15 20DAYS

FIG. 5. Development and duration of histaminesensitivity of mice following intranasal (it.n.) infec-tion or intraperitoneal (i.p.) injection (no infection) oflive B. pertussis (about 2.5 x 106 colony-formingunits), 16 mice per point.

Live Spertussis5 pertussis vaccine

it.n. 0

100 _ i P. o

DAYS

FIG. 6. Sensitivity of mice to 100 mg of histaminediphosphate following intrarnasal (it.n.) or intra-peritoneal (i.p.) injection of the same small numberof live and killed B. pertussis (about 2.5 x 106colony-forming units or 6 x 1O6 bacteria). Results oftwo tests are shown, 32 mice per point. No controlmice died.

pertussis (4, 7). The mechanism of IgA actionhas not been fully explained. IgA inhibitsgrowth of viruses (16). Less is known about itsantibacterial function (21). Very recently Wil-liams and Gibbons (20) presented evidence of

action through inhibition of bacterial adherenceto the mucosal surface.Our study has not answered the question of

how parenterally administered pertussis vac-cine prevents the development of whoopingcough. However, trace amounts of Ig in therespiratory tract of some i.p. mice suggests thatserum Ig may pass the lung barrier. The samequestion has been raised in regard to the pre-vention of cholera by parenterally administeredcholera vaccine (10). The explanation in eachcase may be analogous. Recently Fubara andFreter (5) observed the entry of serum anti-bodies in the intestinal lumen of animals immu-nized with V. cholerae. Earlier Crabbe et al. (3)reported the development of IgA antibodies inthe intestinal mucosa following parenteral aswell as oral immunization with ferritin. A studyof the antibody immunoglobulin response inmice challenged it.n. after recovery from respi-ratory infection and after i.p. vaccination of liveand dead bacteria should provide applicableinformation. Although not considered here, thedirect intervention of cells by phagocytosis ordelayed hypersensitivity must not be ignored.As observed previously (13), the gradual de-

velopment and persistence of histamine suscep-tibility after subsidence of infection in themouse paralleled the onset and duration ofparoxysmal coughing relative to infection in thechild. A new observation was the late peakdevelopment (20 days) and persistent suscepti-bility induced by a small number of bacteria.This was in contrast to 4- to 5-day peaking anddecline following the customary use of 109 ormore bacteria for induction of histamine sus-ceptibility. This difference suggests that themechanisms of action of large and small num-bers of bacteria may be different. One aspect toinvestigate would be the relative amounts of IgEinduced by the different amounts of bacteria(2).

LITERATURE CITED1. Bumet, F. M., and C. Timmins. 1937. Experimental

infection with Haemophilus pertussis in the mouse byintranasal inoculation. Brit. J. Exp. Pathol. 18:83-90.

2. Clausen, C. R., J. Munoz, and R. K. Bergman. 1970. Areaginic type of antibody stimulated by extracts ofBordetella pertussis in inbred strains of mice. J.Immunol. 104:312-319.

3. Crabbe, P. A., D. R. Nash, H. Bazin, H. Eyssen, and J. F.Heremans. 1969. Antibodies of the IgA type in intesti-nal plasma cells of germfree mice after oral or parent-eral immunization with ferritin. J. Exp. Med.130:723-744.

4. Dolby, J. M., and D. E. Dolby. 1969. The antibodyactivities of 19S and 7S fractions from rabbit antiserato Bordetella pertussis. Immunology 16:737-747.

5. Fuhara, E. S., and R. Freter. 1972. Availability of locallysynthesized and systemic antibodies in the intestines.Infect. Immunity 6:965-981.

VOL. 8, 1973 89

on June 30, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

90 GELLER AND PITTMAN

6. Iida, T., N. Kusano, A. Yamamoto, and M. Konosu. 1966.An immunofluorescence study of the action of antibodyin experimental intracerebral infection of mice withBordetella pertussis. J. Pathol. Bacteriol. 92:359-367.

7. Kendrick, P. L., G. Eldering, and N.-S. Ling. 1970.Globulin fractions separated from antipertussis rabbitserum. Arch. Environ. Health 21:388-396.

8. Maitland, H. B., R. Kohn, and A. D. MacDonald. 1955.The histamine-sensitizing property of Haemophiluspertussis. J. Hyg. 53:196-211.

9. Mancini, G., A. 0. Carbonara, and J. F. Heremans. 1965.Immunochemical quantitation of antigens by singleradial immunodiffusion. Immunochemistry 2:235-254.

10. Mosley, W. H., J. C. Feeley, and M. Pittman. 1971. Theinterrelationships of serological responses in humans,and the active mouse protection test to cholera vaccineeffectiveness. In International Symposium on En-terobacterial Vaccines, Berne, 1968. Symp. Ser. Im-munobiological Standardization. 15:185-196. Karger,Basel and New York.

11. North, E. A. 1946. Passive immunization by the intrana-sal route in experimental pertussis. Aust. J. Exp. Biol.Med. Sci. 24:253-259.

12. Northrup, R. S., and S. A. Hossain. 1970. Immunoglobu-lins and antibody activity in the intestine and serum incholera. II. Measurement of antibody activity in jejunalaspirates and sera of cholera patients by radioim-munodiffusion. J. Infect. Dis. 121:S142-146.

13. Pittman, M. 1951. Sensitivity of mice to histamine duringrespiratory infection by Hemophilus pertussis. Proc.Soc. Exp. Biol. Med. 77:70-74.

14. Pittman, M. 1970. Bordetella pertussis. Bacterial and

INFECT. IMMUNITY

host factors in the pathogensis and prevention ofwhooping cough. In S. Mudd (ed.), Infectious agentsand host reactions. W. B. Saunders Co., Philadelphia.

15. Proom, H. 1947. The immunological aspects of experi-mental Haemophilus pertussis infection. J. Pathol.Bacteriol. 59:165-180.

16. Tomasi, T. B., Jr. 1969. The concept of local immunityand the secretory system, p. 3-10. In D. H. Dayton, Jr.,P. A. Small, Jr., R. M. Chanock, H. E. Kaufman, andT. B. Tomasi, Jr. (ed.), The secretory immunologicsystem. Superintendent of Documents, U.S. Govern-ment Printing Office, Washington, D.C.

17. U.S. Public Health Service Regulations (42 CRF, Part73.4004) Revised June 1, 1971.

18. Vergani, C., R. Stabilini, and A. Agostoni. 1967. Quanti-tative determination of serum immunoglobulins bysingle radial immunodiffusion on cellulose acetate.Immunochemistry 4:233-237.

19. Wiedermannova, D., and D. Wiedermann. 1969. Diequantitativen Serumverhaltnisse der ImmunoglobulineG, A und M im Vergleich mit den serologisch erfass-baren Antikorpern wiihrend der Scharlach- und Keu-chustenerkrankung. Z. Immunitaetsforschung AllergieKlin. Immunol. 138:1-12.

20. Williams, R. C., and R. J. Gibbons. 1972. Inhibition ofbacterial adherence by secretory immunoglobulin A: amechanism of antigen disposal. Science 177:697-699.

21. Various participants. 1969. In D. H. Dayton, Jr., P. A.Small, Jr., R. M. Chanock, H. E. Kaufman, and T. B.Tomasi, Jr; (ed.), The secretory immunologic system.Superintendent of Documents, U.S. GovernmentPrinting Office, Washington, D.C.

on June 30, 2020 by guesthttp://iai.asm

.org/D

ownloaded from