Vol. 53, No. 3INFECTION AND IMMUNITY, Sept. 1986, p. 474-4790019-9567/86/030474-06$02.00/0Copyright C) 1986, American Society for Microbiology

Resistance of C3H/HeJ Mice to the Effects ofHaemophilus pleuropneumoniae

B. W. FENWICK,*t B. I. OSBURN, AND H. J. OLANDER

Department of Veterinary Pathology, School of Veterinary Medicine, University of California, Davis, California 95616

Received 11 March 1986/Accepted 19 May 1986

Comparisons were made in the mortality associated with an inhaled dose of viable Haemophiluspleuropneumoniae type 5, strain J45, between adult C3H/HeN and C3H/HeJ mice. Mice of both strains werealso challenged with Escherichia coli strains O111:B4 and J5. The 50% lethal dose (LD50) of H.pleuropneumoniae in C3H/HeN mice was calculated to be 106.5 CFU. At a mean dose of 106.7 CFU a 46%mortality rate occurred in C3H/HeN mice, whereas only 10% of the C3H/HeJ mice died (P < 0.01). Deathsoccurred significantly earlier in C3H/HeN mice (P < 0.01). No deaths occurred later than 12 h postinfection ineither group. Pulmonary lesions in the mice that died were similar to those in pigs that die during the acutephase of H. pleuropneumoniae infection. In surviving mice of both strains, a mild resolving interstitial andbronchopneumonia was present which was not typical of subacute H. pleuropneumoniae infections in swine.Quantitative bacterial isolations from the lungs, liver, and spleen indicate that H. pleuropneumoniae did notmultiply in the lungs, was rapidly cleared, and did not become systemic. No deaths occurred in the miceinoculated with E. coli J5 or O111:B4 at mean doses of 106.3, 107.2, and 108.5 CFU, and 106.4, 0 and 108.2CFU, respectively. The difference in the mortality rate between the C3H/HeN and C3H/HeJ mice suggests thatendotoxin may be involved in acute deaths in pigs infected with H. pleuropneumonia. As indicated by the E. colichallenge, however, other factors are also likely to be involved. Because of the differences in the pathology andmicrobiology following H. pleuropneumoniae pulmonary infections in mice and pigs, mice do not appear to bean accurate model of the overall disease in swine.

Haemophilus (Actinobacillus) pleuropneumoniae is thecause of a severe and often fatal infectious pneumonia withpleuritis in pigs (20, 22, 23). Prominent features of thepulmonary lesions caused by H. pleuropneumoniae in pigsare venous thrombosis and necrosis. Results of previousstudies suggest that toxic factors associated with the bacteriaare important in the pathogenesis of the disease and arelikely to be the cause, either directly or indirectly, of thenecrosis and acute deaths (1, 4). The identity of the toxin ortoxins and their role in the disease have yet to be defined.Direct cytotoxic effects, endotoxins, and exotoxins havebeen suggested (1, 4, 8, 17).C3H/HeJ mice are relatively refractory to the toxic and

mitogenic effects of endotoxin due to a defect in the lps alleleon the fourth chromosome and other undefined geneticinfluences (15, 27, 28). On the other hand, the syngeneic andfully histocompatible C3H/HeN mice are fully responsive tothe effects of endotoxin (15). This difference (approximately1,000:1) has proven to be a powerful tool in the study of therole of endotoxin in the pathogenesis and immunopathologyof gram-negative infections.C3H/HeJ mice are hypersusceptible to Salmonella

typhimurium infections, while C3H/HeN mice are innatelyresistant (19). This difference has been interpreted as anindication that endotoxin is not a major factor in salmonel-losis (2). By using the same approach, the potential involve-ment of endotoxin in the pathogenesis of H. pleuro-pneumoniae was studied, and at the same time a proposedmouse model of H. pleuropneumoniae infections in swinewas evaluated (24, 25).

* Corresponding author.t Present address: Department of Experimental and Comparative

Pathology, University of Florida, Gainesville, FL 32610.

MATERIALS AND METHODS

Bacterial strains and growth conditions. An encapsulatedstrain of H. pleuropneumoniae serotype 5 (strain J45) thathad been originally isolated from a naturally occurring caseof porcine pleuropneumonia was obtained from E. L.Biberstein, University of California, Davis. Bacteria weregrown on chocolate agar plates at 37°C in 10% CO2 for 18 h.The cells were harvested with sterile saline, washed twicewith cold saline, and stored on ice. Bacterial numbers werespectrophotometrically adjusted to the desired concentra-tion, as determined in previous experiments, with coldsterile saline. Escherichia coli 0111:B4 was purchased fromthe American Type Culture Collection, Rockville, Md., andE. coli J5 was provided by E. Ziegler, University of Califor-nia, San Diego. They were prepared in the same manner asH. pleuropneumoniae, except blood agar was used.

Mice. Male and female adult C3H/HeJ and C3H/HeN micewere purchased from Bantin and Kingman Inc., Fremont,Calif. Mice were separated by sex and housed in groups of 8or 10 and acclimatized for several weeks. They weighed 22to 28 g at the start of the experiment. Prior to the start of theexperiment, a few mice were randomly selected and killed,their lungs were cultured for mycoplasma and bacteria andexamined histologically. Serum collected from these micewas evaluated for antibodies to H. pleuropneumoniae J45and E. coli 0111:B4 and J5 by a microtube aggultinationassay (12). Positive control sera were produced in mice byrepeated subcutaneous injection of viable organisms.LD50 determination. Mice were anesthetized with halo-

thane to light narcosis. Replicate groups of 10 C3H/HeNmice were inoculated intranasally with various concentra-tions of bacteria diluted in 0.05 ml of sterile saline, asdescribed by Rushton (21). The concentrations of the cul-tures were periodically determined between inoculations by

474

on July 21, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

H. PLEUROPNEUMONIAE IN MICE 475

350

0co0h.0

20

20

Ecm

30

25

20

15

10

5

C3H/HeN

C3H/HeJ

0 100 200 300 400 500 600 700 800

Time (min)FIG 1. Total accumulated deaths by time and mouse strain caused by the intranasal inoculation of H. pleuropneumoniae (1067 CFU).

techniques described by Miles and Misra (11), and the meanconcentration of organisms per 0.05 ml of inoculum wascalculated. Clinical signs and mortality rates were recordedat 30-min intervals for the first 18 h and then twice daily for5 days. The 50% lethal doses (LD50s) were calculated basedon cumulative mortality at 5 days postinoculation.

Experimental infection. The calculated LD50 of H. pleuro-pneumoniae for C3H/HeN mice was simultaneously givenintranasally to both C3H/HeN and C3H/HeJ mice, as de-scribed previously (21). In a separate experiment, groups of16 mice from both strains were inoculated intranasally withvarious concentrations of E. coli 0111:B4 and J5. Tencontrol mice were inoculated with sterile saline.

Clinical signs and deaths were continuously recorded forthe first 18 h postinoculation and then twice daily for 5 days.Pathologic or microbiologic examination was alternatelyperformed on all animals that died. Six mice from bothstrains that were challenged with H. pleuropneumoniae wererandomly selected at 24 and 72 h postinoculation and exam-ined. Two mice from each strain that were given the E. colior saline were examined at 4, 24, and 72 h postinoculation.Five days after the inoculation all remaining mice weresacrificed and divided into equal groups for pathologic andmicrobiologic examinations. For pathologic examination,the lungs were insufflated with 10% Carlson-modified For-malin before they were removed from the thorax.

Bacterial recovery from tissues. The lungs, liver, andspleen were removed from each mouse and kept on ice. Theweight of the lungs and the combined weight of the liver andspleen were determined. The lungs or the liver and spleenwere placed into chilled sterile Ten Broeck tissue homoge-nizers, cold sterile saline was added to equal 10 ml, and thetissues were thoroughly homogenized. Undiluted homoge-nate or appropriate dilutions were plated onto chocolate agarand incubated at 37°C. Cultures were performed in triplicate.The plates with optimal numbers of organisms were counted24 h later, and the total number of organisms present in thelung or in the combined spleen and liver was calculated.

Statistics. The LD50 was calculated, as outlined by Reedand Muench (18). Mortality rates and time to death datawere analyzed by challenge experiment analysis techniques(9), using analysis of variance, the Student t test, andMann-Whitney techniques to verify population uniqueness.

Differences in bacterial recovery rates were evaluated by thechi-square test.

RESULTS

Susceptibility of mice. Antibody in serum against H.pleuropneumoniae or E. coli 0111:B4 was not detected bythe bacterial aggultination assay in control mice. A few miceof both strains had low bacterial agglutination titers (nevergreater than 1:8) against E. coli J5. The LD50 for an inhaledchallenge of H. pleuropneumoniae J45 in C3H/HeN micewas calculated to be 1065 CFU. At a mean dose of 106.7 CFUthe mortality rate for C3H/HeN mice was 46.7% (28 of 60),whereas only 10% of the endotoxin-hyporesponsiveC3H/HeJ mice died (6 of 60) (P < 0.01). In addition, themean time to death was shorter in the C3H/HeN than in theC3H/HeJ mice (2.43 versus 6.75 h, respectively) (P < 0.01)(Fig. 1). No deaths in either strain of mice occurred laterthan 12 h postinoculation. No deaths (0 of 10) occurred witheither E. coli J5 or O111:B4 at mean doses of 106.3, 107.2,108.5 CFU, and 106-4, 107-5 and 108.2 CFU, respectively.

Clinical disease. Most of the mice recovered rapidly fromthe anesthesia. A small number of mice from both strains(5.0% C3H/HeJ and 8.3% C3H/HeN) died because of anes-thetic overdose and were replaced. All mice that recoveredfrom anesthesia were clinically normal by 15 min postinoc-ulation. The mice from both strains that died during the firstfew hours after inoculation initially showed roughened haircoats, weakness, and reluctance to move. In a short timethey rapidly developed severe respiratory distress withgasping, and they became comatose and died. The time fromthe initial sign of illness to death in C3H/HeN mice averaged25 min but was as short as 5 min (Table 1). This progressionappeared to be slower in the C3H/HeJ mice, but because ofthe small number of animals that died the observation wasnot statistically significant.The mice that died between 7 and 12 h postinoculation

were slower to develop severe respiratory distress, and theprogression to death was delayed (mean of 48 min betweenfirst clinical sign and death). Approximately one-half of theanimals that survived the inoculation also showed rough haircoats and a tendency to huddle together, but only a fewshowed any signs of respiratory disease which was, for the

VOL. 53, 1986

on July 21, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

476 FENWICK ET AL.

FIG 2. Typical histologic lesions in the lung of a C3H/HeN mouse that died soon after it was inoculated intranasally with H.pleuropneumoniae. There is widespread pulmonary edema with flooding of alveoli, lymphytic dilation, and a focal area of acute necosis.Magnification, x 125.

most part, upper respiratory in nature. There was no appar-ent difference between the mouse strains in the severity ofthe clinical signs in the mice that survived. The miceinoculated with either of the E. coli strains did not showclinical disease for approximately 24 h. From 24 to 72 hpostinoculation approximately 25% of the mice from bothstrains at any one time had roughened hair coats, werereluctant to move, and huddled together. No evidence ofrespiratory disease was apparent and no difference betweenbacterial or mouse strain was noted. All mice were clinicallynormal by 5 days postinoculation. The control mice werenormal throughout the experiment.

Pathology. All control mice and those failing to recoverfrom anesthesia were free of histologic pulmonary lesions.The pulmonary lesions from both strains of mice that died asa result of the H. pleuropneumoniae inoculation were simi-lar, regardless of the time of death. Gross examination

TABLE 1. Clinical course after intranasal inoculation ofC3H/HeN and C3H/HeJ mice with H pleuropneumoniae

(106.7 CFU)

Mouse strain Mean time to death (min Clinical course (minMSD, range)a ± SD, range)

C3H/HeN 146 + 453 (55-230) 25 ± 4.5 (5-65)C3H/HeJ 404 + 161.6 (275-710) 72 + 18.7 (35-130)

a Mean time between intranasal inoculation and death.b Mean time between first recognized clinical illness and death.

showed that the lungs were uniformly consolidated, wet, andheavy and they failed to collapse. Histological examinationshowed that (Fig. 2) alveoli were flooded with sero-hemorrohagic exudate, marked vescular congestion waspresent, and lymphatics were dilated by fibrin and edemafluid. Inflammatory cell infiltration was minimal.The animals surviving the H. pleuropneumoniae and E.

coli challenges that were examined on days 1, 3, and 5postinoculation had normal lungs on gross examination.Histologically, a mild to focally moderate interstitial andbronchopneumonia, with neutrophils, macrophages, andsmall amounts of fibrin, was present. The inflammation wascentered on conducting airways, and only rarely were alveoliaffected and then only in close association with affectedairways. Neither vascular thrombosis nor parenchymal ne-crosis was evident. The lesions were most severe on day 1,with considerable improvement and mild focal inflammationby day 3 and with near complete resolution by day 5postinoculation (Fig. 3).

Microbiology. The concentrations of viable H. pleuropneu-moniae or E. coli with which the mice were inoculated wasdetermined by averaging the concentration before, during,and at the end of the inoculation period. The total timeinvolved was approximately 1.75 h. No organisms werecultured from the mice that were examined prior to the startof the experiment or from the controls.

Isolations were not made from the livers or spleens of anyof the mice, irrespective of the organism or time of exami-

INFECT. IMMUN.

on July 21, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

H. PLEUROPNEUMONIAE IN MICE 477

*u. J __s _ Y_!.r_ww sJ;P-S..w-wh ug -.FIG 3. Typical histologic lesions in the lung of a C3H/HeN mouse 3 days after intranasal inoculation of H. pleuropneumoniae. There is

a mild to focally moderate bronchopneumonia and mild interstitial inflammatory cell infiltration. Magnification, x 125.

nation. The mean number of organisms present in the lungsof mice from the various treatment groups by time ispresented in Table 2. Pulmonary clearance of H. pleuro-pneumoniae and both E. coli strains occurred rapidly and

TABLE 2. Mean pulmonary isolation rates per gram oftissue (loglo)

Isolation rate/g of tissueDose Mouse (loglo) at the following

Organism (CFU) strain times postinoculation":

Day 0 Day 1 Day 3 Day 5

H. pleuropneumoniae 1067 C3H/HeN 4.63 2.02 0 0J45 C3H/HeJ 4.54 2.80 1.42 0

E. coli J5 106.3 C3H/HeN 3.10 2.16 0 0C3H/HeJ 3.59 2.04 0 0

E. coli J5 107 C3H/HeN 3.66 2.43 0 0C3H/HeJ 3.59 2.44 0 0

E. coli J5 10.5 C3H/HeN 3.92 2.38 0 0C3H/HeJ 4.00 2.12 1.14 0

E. coli 0111:B4 106.4 C3H/HeN 3.07 2.08 0 0C3H/HeJ 2.98 2.40 0 0

E. coli 0111:B4 107-5 C3H/HeN 3.52 2.80 1.20 0C3H/HeJ 3.96 2.01 1.08 0

E. coli 111:B4 1082 C3H/HeN 3.98 2.45 1.13 0C3H/HeJ 4.10 2.39 1.06 1.04

Controls (saline) C3H/HeN 0 0 0 0C3H/HeJ 0 0 0 0

a At the time of death for H. pleuropneumoniae infected mice and at 4 hpostinoculation for E. ccoli infected mice.

was-complete by postinoculation day 5 in all but one group;C3H/HeJ give E. coli 0111:84 at 108.5 CFU. There was nodetectable difference in the rate of pulmonary clearancebetween the mouse strains, regardless of the organism. Theisolations that did occur were on day 3 from one infectedmouse and on day 5 from two infected mice.

DISCUSSION

In these studies we used the inherent difference inendotoxin responsiveness between C3H/HeJ and C3H/HeNmice to evaluate the involvement of endotoxin in the patho-genesis of H. pleuropneumoniae infections and at the sametime to test the validity of a mouse model for the disease inswine.The results demonstrate that the endotoxin-hypore-

sponsive C3H/HeJ mice are significantly more resistant tothe acute effects of inhaled H. pleuropneumoniae, as mea-sured by mortality and mean time to death, than the nor-mally endotoxin-responsive C3H/HeN mice. This observa-tion implicitly suggests a role of endotoxin in the early stagesof H. pleuropneumoniae infections; but as indicated by thedifference between the effects of H. pleuropneumoniae andE. coli, other factors are likely to be involved. Moreover,mice do not appear to be an accurate model of H.pleuropneumoniae infections in pigs in that inhaled H.pleuropneumoniae acts more as a toxin than a true infection,and both the pathologic and microbiologic findings differfrom those seen in the later stage of H. pleuropneumoniae

VOL. 53, 1986

on July 21, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

478 FENWICK ET AL.

infections in pigs. However, the clinical disease and pulmo-nary lesions during the first few hours after H.pleuropneumoniae exposure in mice are similar to acutelethal infections in swine.C3H/HeJ mice are hypersusceptible to S. typhimurium

septicemia and to E. coli urinary tract infections (2, 3, 15,19). The reason for their unusually high susceptibility togram-negative infections is uncertain but has been attributedto defective autosomal codominant Ips genes and is associ-ated with their hyporesponsiveness to the biologic andimmunologic effects of endotoxin (27-29). In contrast,C3H/HeJ mice were more resistant to the lethal effects of aninhaled dose of H. pleuropneumoniae than were C3H/HeNmice. However, infections evidenced by significant bacterialmultiplication and dissemination did not take place.The effect of an inhaled dose of H. pleuropneumoniae in

C3H/HeJ mice is strongly dose dependent and results in oneof three events: rapid acute death within a few hours, mildclinical disease, or no observable effect. A similar H.pleuropneumoniae dose-related effect has been experimen-tally demonstrated in swine (26) and in other strains of mice(24, 25). However, in mice, as opposed to in swine, H.pleuropneumoniae did not multiply significantly, did notbecome systemic, and was rapidly cleared from the lungs.Interestingly, at the same or higher dose the two strains of E.coli were cleared as fast as H. pleuropneumoniae withoutnotable toxicity.The acute toxicity of H. pleuropneumoniae in mice, as

well as the pathology and dose-related nature of the effect,suggests that endotoxin may be involved in acute H.pleuropneumoniae infections. If these effects were relatedonly to the concentration of lipid A, it would be expectedthat E. coli at the same or higher dose would also be toxic.The fact that a similar toxic reaction commonly occurs whenswine are immunized with heat-killed (and other) H.pleuropneumoniae vaccines (S. Henry, Abilene, Kans., per-sonal communication, 1985) further supports the inherenttoxic nature of H. pleuropneumoniae. The lesions similar tothose which occur in pigs that die during the acute stages ofpleuropneumonia can be reproduced in pigs by theintratracheal administration of culture supernatants (S.Rosendal, W. R. Mitchell, M. Weber, M. R. Wilson, andM. R. Zamen, Proc. Int. Pig Vet. Soc. Congr., p. 221, 1980)or the purified lipopolysaccharide (LPS) of H. pleuro-pneumoniae J45 (2a). Inhalation of H. pleuropneumoniaeLPS by mice also induces lesions similar to those seen in themice that die after being inocculated intranasally with H.pleuropneumoniae (data not shown). In addition, ultrastruc-tural examination of the lesions induced by H. pleuropneu-moniae in pigs showed endothelial changes similar to thoseof the Shwartzman reaction preceding vascular thrombosis(16). The content and composition of the LPS among gram-negative bacteria vary widely (6, 10), as do their toxogenicand thrombogenic properties (13). Haemophilus sp. andrelated organisms are known to have vesicular structures, or"blebs," on their outer cell walls which are morphologicallysimilar to LPS vesicles (7). These structures are implicatedin tissue damage and inflammation associated with gram-negative bacterial infections.

In the mice, all deaths occurred during the first few hoursafter inoculation with H. pleuropneumoniae. In the acuteform of H. pleuropneumoniae infections in swine, death alsocommonly occurs within the first 24 h, often with a veryrapid clinical course (4, 5). In these pigs, as in the mice thatdied, the lungs are characterized by widespread alveolarflooding, vascular congestion, dilated fibrin-filled lymphat-

ics, focal necrosis, and minimal inflammatory cell infiltra-tion. The principle pathologic difference between the acutelesions in swine and mice is the greater amount of fibrinpresent in the lungs of swine.Swine that survive the acute stages of an H. pleuro-

pneumoniae infection commonly develop focal areas ofnecrosis in the lung associated with venous thrombosiswithout any recognizable airway association. These pulmo-nary infarcts commonly contain H. pleuropneumoniae, thusresulting in the occurrence of chronic carriers (14). Thelesions in the mice surviving the H. pleuropneumoniaeinoculation and those given E. coli were distinctively airwayoriented and lacked areas of necrosis, and the presence of H.pleuropneumoniae or E. coli could not be demonstrated.

Sebunya and Saunders (24, 25) proposed a mouse model ofH. pleuropneumoniae infection in which they suggested itsusefulness in studying the pathogenesis of the disease inswine as well as in evaluating potential vaccines. The resultsof the present experiment are similar, with the differences inLD50, mortality rates, and time to death easily explained bythe use of different bacterial and mouse strains. Both studiesshow that in the strains of mice examined, H. pleuro-pneumoniae in general fails to multiply significantly; deathsmainly occur shortly after inoculation; very large doses of H.pleuropneumoniae are required for infection; deaths areclosely associated with the inoculating dosage; and exceptfor the mice that died acutely, the lesions are not similar tothose in H. pleuropneumoniae-infected pigs. In a previousstudy (25), when H. pleuropneumoniae was inocculatedintranasally into mice at low dosages it multipled for a shorttime (24 h) and then was rapidly cleared. We found that atreasonable challenge doses (equal to the infective dose forpigs) mice can rapidly clear H. pleuropneumoniae J45 fromthe lungs and show few clinical signs with minimalpathologic changes in the lungs.The findings in this report suggest that the acute toxic

effects of H. pleuropneumoniae infection in mice, and pos-sibly in swine, are due to endotoxin and that H. pleuro-pneumoniae is inherently more toxic than E. coli, for an asyet to be determined reason. Utilization of a mouse model ofH. pleuropneumoniae infections may be helpful in elucidat-ing the pathogenesis of rapid and apparently toxic deathsthat occasionally occur in the early stages of H. pleuro-pneumoniae infections in pigs. The murine model may alsobe helpful in the development of less toxic H. pleuropneu-moniae vaccines. It is apparent, however, that because of aninnate resistance to H. pleuropneumoniae infections in thestrains of mice that have been studied, the mouse appearsnot to be an appropriate model for the overall study of thepathogenesis and immunology of H. pleuropneumoniae in-fections in swine.

ACKNOWLEDGMENTS

We thank E. L. Biberstein for reviewing this manuscript.This work was supported in part by special grant 83-CRSR-2-2175

from the U.S. Department of Agriculture.

LITERATURE CITED1. Bendixen, P. H., P. E. Shewen, S. Rosendal, and B. N. Wilkie.

1981. Toxicity of Haemophilus pleuropneumoniae for porcinelung macrophages, peripheral blood monocytes, and testicularcells. Infect. Immun. 33:673-676.

2. Eisenstein, T. K., W. Deakins, L. Killan, P. H. Saluk, and B. M.Sultzer. 1982. Dissociation of innate susceptibility to salmonellainfection and endotoxin responsiveness in C3HeB/FeJ mice andother strains in the C3H lineage. Infect. Immun. 36:696-703.

INFECT. IMMUN.

on July 21, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

H. PLEUROPNEUMONIAE IN MICE 479

2a.Fenwick, B. W., B. I. Osburn, and H. J. Olander. 1986. Isolationand biological characterization of two lipopoly saccharides anda capsular-enriched polysaccharide preparation from Haemoph-ilus pleuropneumoniae. Am. J. Vet. Res. 47:1433-1441.

3. Hagberg, L., R. Hull, S. Hull, J. R. McGhee, S. M. Michalek,and C. S. Eden. 1984. Difference in susceptibility to gram-negative urinary tract infections between C3H/HeJ andC3H/HeN mice. Infect. Immun. 46:839-844.

4. Hani, H., F. Kong, J. Nicolet, and E. Scholl. 1973. PorcineHemophilus parahaemolyticus pleuropneumoniae VI. Patho-genesis. Schweizer Arch. Tierheik. 115:196-203.

5. Hani, H., H. Konig, J. Nicolet, and E. Scholl. 1973. Haemophiluspleuropneumoniae in swine. V. Pathomorphology. SchweizArch. Tierheilk. 115:191-1%.

6. Hitchcock, P. J., and T. M. Brown. 1983. Morphological heter-ogeneity among Salmonella lipopolysaccharide chemotypes insilver-stained polyacrylamide gels. J. Bacteriol. 154:269-277.

7. Holt, S. C., A. C. R. Tanner, and S. S. Socransky. 1980.Morphology and ultrastructure of oral strains of Actinobacillusactinomycecomitans and Haemophilus aphrophilus. Infect. Im-mun. 30:588-600.

8. Katsumi, K., T. Nakai, and A. Sawata. 1986. Interaction betwenheat-stable hemolytic substance from Haemophilus pleuro-pneumoniae and porcine pulmonary macrophage in vitro. In-fect. Immun. 51:563-570.

9. Liddle, F. D. 1978. Evaluation of survival in challenge experi-ments. Microbiol. Rev. 42:237-249.

10. Logan, S. M., and T. J. Trust. 1984. Structural and antigenicheterogeneity of lipopolysaccharides of Campylobacter jejuniand Campylobacter coli. Infect. Immun. 45:210-216.

11. Miles, A. A., and S. S. Misra. 1938. The estimation of thebactericidal power of blood. J. Hyg. 38:732-748.

12. Mittle, K. R., R. Higgins, S. Lariviere, and D. Leblanc. 1984. A2-mercaptoethanol tube agglutination test for diagnosis of Hae-mophilus pleuropneumoniae infection in pigs. Am. J. Vet. Res.45:715-719.

13. Morrison, D. C., and C. G. Cochran. 1973. Direct evidence forHageman factor (factor XIII) activation by bacterial lipopoly-saccharides. J. Exp. Med. 140:787-811.

14. Nielson, R., and M. Mandrup. 1977. Pleuropneumonia in swinecaused by Haemophilus parahaemoliticus, a study of the epi-demiology of the infection. Nord. Vet. Med. 29:465-473.

15. O'Brien, A. D., D. L. Rosenstreich, I. Scher, G. Campbell, R. P.

MacDermott, and S. B. Formal. 1980. Genetic control of sus-ceptibility to Salmonella typhimurium in mice: role of the Lpsgene. J. Immunol. 124:20-24.

16. Perfumo, C. J., C. Rehbinder, and K. Karlsson. 1983. Swinepleuropneumonia produced by Haemophilus pleuropneumoniaeIII. An electron microscopic study. Zentralbl. Vet. Med.30:678-684.

17. Pijoan, C. 1982. Hemophilus pleuropneumoniae. Mod. Vet.Pract. 63:653-654.

18. Reed, L. J., and H. Muench. 1938. A simple method of estimat-ing fifty percent endpoints. Am. J. Hyg. 27:493-497.

19. Robson, H. G., and S. I. Vas. 1972. Resistance of inbred mice toSalmonella typhimurium. J. Infect. Dis. 126:378-386.

20. Rosendal, S., and W. R. Mitchell. 1983. Epidemiology of Hae-mophilus pleuropneumoniae infection in pigs: a survey ofOntario pork producers, 1981. Can. J. Comp. Med. 47:1-5.

21. Rushton, B. 1978. Induction of pneumonia in mice with Pasteu-rella haemolytica. J. Comp. Pathol. 88:477-480.

22. Saunders, J. R., A. D. Osburn, and T. K. Sebunya. 1981.Pneumonia in Saskatchewan swine: Abattoir incidence ofintrathoracic lesions in pigs from a herd infected with Haemo-philus pleuropneumoniae and from other herds. Can. Vet. J.22:244-247.

23. Schultz, R. A., T. F. Young, R. F. Ross, and D. R. Jeske. 1982.Prevalence of antibody to Haemophilus pleuropneumoniae inIowa swine. Am. J. Vet. Res. 43:1848-1851.

24. Sebunya, T. N. K., and J. R. Saunders. 1982. Studies onimmunity to Haemophilus pleuropneumoniae infections inmice. Am. J. Vet. Res. 43:1793-1798.

25. Sebunya, T. N. K., and J. R. Saunders. 1982. Pulmonaryclearance of Haemophilus pleuropneumoniae and Seratiamarcesens in mice. Am. J. Vet. Res. 43:1799-1801.

26. Sebunya, T. N. K., J. R. Saunders, and A. D. Osburne. 1983.Dose response relationship of Haemophilus pleuropneumoniaaerosols in pigs. Can. J. Comp. Med. 47:54-56.

27. Sultzer, B. M. 1968. Genetic control of leukocyte responses toendotoxin. Nature (London) 219:1253-1254.

28. Watson, J., M. Largen, and K. P. McAdam. 1978. Geneticcontrol of endotoxin responses in mice. J. Exp. Med. 147:39-49.

29. Watson, J., R. Riblet, and B. A. Taylor. 1977. The response ofrecombinant inbred strains of mice to bacterial lipopolysaccha-rides. J. Immunol. 118:2088-2093.

VOL. 53, 1986

on July 21, 2020 by guesthttp://iai.asm

.org/D

ownloaded from