Veterinary Microbiology, 25 (1990) 199-207 Elsevier Science Publishers B.V., Amsterdam
199
Virulence and lienotoxicity of Bordetella bronchiseptica in mice T. Magyar Veterinary Medical Research Institute of the Hungarian Academy of Sciences, 1t-1581 Budapest, P.O. Box 18, Hungary (Accepted 4 April 1990)
ABSTRACT Magyar, T., 1990. Virulence and lienotoxicity of Bordetella bronchiseptica in mice. Vet. Microbiol., 25: 199-207. Whole-cell suspensions (WCSs) and cell-flee sonicated extracts (SEs) of seven Bordetella bronchiseptica strains were studied for lethality and lienotoxicity in mice. Lethality was assessed after intravenous and intracerebral inoculation, and lienotoxicity by splenic atrophy after intravenous inoculation. The strains represented phase I isolates with or without cytotoxin production, their phase II1 subcultures and a phase IV variant. The lethality and lienotoxicity of the SEs were in close positive correlation with cytotoxin production. The WCSs of all phase I strains were lethal, irrespective of their cytotoxin- and lienotoxin-producing ability. The only difference was that cytotoxic phase I strains caused splenic atrophy while the noncytotoxic phase I strain induced splenic hypertrophy in the surviving mice. The WCSs of phase III and IV variants were non-lethal and caused splenic hypertrophy even though all but one of them showed some cyto- and lienotoxic activity when their SEs were tested. The results indicate that B. bronchiseptica possesses two different mouse lethal factors: one seems to be identical with the cytotoxin, the other is associated with cell integrity and viability and, presumably, propagation in vivo. It also follows from the results that only the SEs are suitable for accurate determination of the lienotoxin-producing ability ofB. bronchiseptica.
INTRODUCTION
Bordetella bronchiseptica is a respiratory pathogen involved in the aetiology of atrophic rhinitis (AR) of swine. It produces a range of potential virulence determinants: haemagglutinins, an adhesin to nasal epithelial cells (Semj6n and Magyar, 1985), cytotoxin, mouse lethal factor (Collings and Rutter, 1985 ), adenylate cyclase (Weiss and Hewlett, 1986) and haemolysin (Goodnow, 1980). Although their role has not been elucidated satisfactorily yet, it seems that colonisation of the nasal cavity in large numbers and the cytotoxin-producing ability of the strains are the most important factors in 0378-1135/90/$03.50
© 1990 - - Elsevier Science Publishers B.V.
200
T.MAGYAR
the pathogenesis ofturbinate atrophy caused by B. bronchiseptica (Magyar et al., 1988). The toxigenicity of B. bronchiseptica can be tested by different methods. Depending on the method applied, various names are used to designate the toxin: dermonecrotoxin ( D N T ) in the guinea pig skin test (de Jong et al., 1980 ), cytotoxin in embryonic bovine lung (EBL) cell assay (Rutter and Luther, 1984 ) and mouse lethal factor in mice (Collings and Rutter, 1985 ). Kriiger and Horsch ( 1982 ) described a mouse lienotoxicity test using broth cultures for determining the toxin-producing ability of B. bronchiseptica. Strains that were lethal to mice and which at sublethal doses caused atrophy of the spleen were considered toxic. In contrast, atoxic strains did not cause death and induced significant hypertrophy of the spleen. A similar test was described earlier by Iida and Okonogi ( 1971 ) to demonstrate the DNT-producing ability of B. pertussis. However, they examined a partially purified extract of B. pertussis. The purpose of the work presented here was to study the relationship between virulence and toxin-producing ability of B. bronchiseptica in mice. For this, the effects of whole-cell suspensions (WCSs) and cell-free sonicated extracts (SEs) of B. bronchiseptica were compared in the so-called mouse lienotoxicity test using toxic and atoxic phase I isolates, their phase III variants and a phase IV (rough) strain. Another objective was to assess the value of WCSs or SEs in determining the lienotoxigenicity of B. bronchiseptica. MATERIALS AND METHODS
B. bronchiseptica strains B. bronchiseptica strains CF and B58 were isolated from two different pig herds with clinical AR. Strain PV6 was isolated from a herd in which clinical AR had never been observed. These strains agglutinated calf red blood cells (RBCs), were haemolytic on Bordet-Gengou (BG) agar medium supplemented with 15% sheep blood and did not grow in the presence of erythromycin. B. bronchiseptica strains CFp61, B65 and PV6p15 w e r e derived from the fresh isolates by repeated subculture on MacConkey agar. These strains did not agglutinate calf RBCs, were non-haemolytic on BG agar and grew in the presence of erythromycin. On the basis of these criteria, they were designated as phase III variants of the original phase I isolates (Lax, 1985 ). B. bronchiseptica ATCC 19 395 was obtained from the American Type Culture Collection and during maintenance in the laboratory it became a phase IV variant with rough colony morphology. It also was non-haemolytic, erythromycin resistant and failed to agglutinate calf RBCs.
Haemagglutination test This test was done as described previously (Magyar et al., 1988 ).
TOXICITY OF BORDETELLA BRONCHISEPTICA IN MICE
201
Preparation of whole-cell suspensions B. bronchiseptica strains grown for 24 h at 37 °C on BG agar supplemented with 15% (v/v) sheep blood were harvested and suspended in phosphate buffered saline (PBS), pH 7.2. The concentration of the bacterial suspensions was adjusted to approximately 5X 109 colony forming units (cfu) m1-1 by photometry.
Preparation of cell-free sonicated extracts B. bronchiseptica strains propagated for 24 h at 37°C on BG agar plates were suspended in distilled water at a concentration of ~ 10 l° cfu ml -l. The cell suspensions were disrupted ultrasonically (MSE Sonifier 1-66, Great Britain) at full power for 5 rain whilst cooling in an icy water bath. Viable counts were reduced by ~ 90%. Sonicates were then centrifuged at 12 000 × g for 60 min at 4°C and filtered through a 0.22/tm membrane filter (Millipore Corp. ). Filtrates were tested for toxicity or adenylate cyclase activity within 2 days while stored at 4 oC.
Treatment of sonicated extracts A portion of SEs of B. bronchiseptica strain CF and CFp61 was exposed to heat (56 ° C, 1 h) before it was tested for lienotoxicity.
Adenylate cyclase test Bacterial extracts were tested for adenylate cyclase activity as described in an earlier paper (Magyar et al., 1988 ).
Experimental procedure Lethality was assessed in intravenously or intracerebrally inoculated mice, and lienotoxicity by determining the spleen mass of surviving mice. For the lienotoxicity test, twofold dilutions of test materials were prepared. The diluent was PBS for WCSs and distilled water for SEs. Groups of five white CFLP (LATI, G6d6116, Hungary) mice weighing 16-18 g were inoculated intravenously with 0.2 ml of one of the serial dilutions. Deaths were recorded up to 7 days postinoculation when the survivors were killed and spleen masses were determined. Relative spleen masses were calculated for 10 g body mass (spleen mass (mg) per body mass (g) × 10). For the intracerebral test tenfold dilutions of WCSs were prepared and groups of 10 mice were inoculated with 0.03 ml of one of the serial dilutions. Deaths were recorded up to 10 days postinoculation. Cytotoxicity for EBL cells (Rutter and Luther, 1984) was tested using twofold dilutions of sonicated extracts.
202
T.MAGYAR
RESULTS
Intravenous inoculation of mice The WCSs and cell-free SEs of the B. bronchiseptica strains designated CF and B58 (Table 1 ) were lethal to mice and induced a reduction in spleen mass when given at sublethal doses. The WCSs of phase III variants ( C U p 6 1 and B65 ) did not cause death and induced a remarkable increase in spleen mass. In contrast, the SEs of these subcultures resulted in death and splenic atrophy but at much lower dilutions (about one tenth) than the SEs prepared from the original phase I isolates. Similar results were obtained with the phase IV B. bronchiseptica strain ATCC 19 395 (Table 2 ), with the difference that its WCS produced only a moderate increase in spleen mass. B. bronchiseptica PV6 inoculated in living form was similarly lethal as the WCSs of the other phase I strains (CF and B58) but the surviving mice showed considerable lienal hypertrophy (Table 2). The SEs of both PV6 and PV6pl5 were nonlethal and produced an increase in spleen mass, similarly to the WCS of PV6pl 5. Heat treatment completely abolished the lethal and lienotoxic activity of the SEs prepared from B. bronchiseptica strains CF and CFp61 (Table 3 ). All the mice survived and developed a dose-dependent increase in spleen mass.
Intracerebral inoculation of mice Table 4 shows the virulence of B. bronch&eptica strains as determined by intracerebral inoculation of mice. Phase I strains (CF, B58 and PV6) had TABLE 1 Mouse lienotoxicity test w i t h whole-cell suspensions (WCSs) and cell-free sonicated extracts (SEs) ofBordetella bronchiseptica strains CF, B58 and their phase III variants CFp61and B65
CFp61
D i l u t i o n CF
WCSs
2 4 8 16 32 64 128 256 512
SEs
B58
WCSs
SEs
B65
WCSs
SEs
WCSs
SEs
let a
smb let
sm
let
sm
let
sm
let
sm
let
sm
let
sm
let
5/5 5/5 5/5 3/5 2/5 0/5 ND c ND ND
37 42 59
15 21 31 43 57 54
0/5 0/5 0/5 0/5 0/5 0/5 ND ND ND
194 187 137 93 88 76
2/5 0/5 0/5 0/5 0/5 0/5 ND ND ND
18 26 34 46 55 58
5/5 3/5 3/5 0/5 0/5 0/5 ND ND ND
27 38 40 58 64
5/5 5/5 5/5 5/5 5/5 5/5 2/5 0/5 0/5
11 22 40
0/5 0/5 0/5 0/5 0/5 0/5 ND ND ND
126 110 114 93 82 70
5/5 5/5 1/5 0/5 0/5 0/5 ND ND ND
5/5 5/5 5/5 1/5 0/5 0/5 0/5 0/5 0/5
alet = lethality. bsm = relative spleen mass. CND ---not determined.
sm m
27 39 45 53
TOXICITY OF BORDETELLA BRONCHISEPTICAIN MICE
203
TABLE 2
Mouse lienotoxicity test with whole-cell suspensions (WCSs) and cell-free sonicated extracts (SEs) ofBordetella bronchiseptica strains ATCC 19 395, PV6 and PV6p~5 Dilution
ATCC 19 395 WCSs
2 4 8 16 32 64
PV6 SEs
PV6pl5
WCSs
SEs
WCSs
SEs
leta
sm b
let
sm
let
sm
let
sm
let
sm
let
sm
0/5 0/5 0/5 0/5 0/5 0/5
88 74 77 73 67 61
5/5 5/5 2/5 1/5 0/5 0/5
18 40 44 50
5/5 5/5 5/5 5/5 4/5 1/5
178 118
0/5 0/5 0/5 0/5 0/5 0/5
110 85 79 71 70 61
0/5 0/5 0/5 0/5 0/5 0/5
141 131 112 86 77 56
0/5 0/5 0/5 0/5 0/5 0/5
128 99 96 89 80 72
alet = lethality. bsm = relative spleen mass. TABLE 3
Mouse lienotoxicity test with heat-treated cell-free sonicated extracts ofBordetella bronchiseptica strains CF and
CFp61
Dilution
2 4 8 16 32 64
CFp61
CF let a
sm b
let
sm
0/5 0/5 0/5 0/5 0/5 0/5
96 80 77 75 60 52
0/5 0/5 0/5 0/5 0/5 0/5
83 73 68 61 58 55
~let = lethality. bsm = relative spleen mass.
lethal effects o f similar degree: a single dose o f ~ 3 cfu was sufficient to kill more than 50% o f the mice. In contrast, phase III and IV strains (CFp61, B65, PV6pl 5 and ATCC ) were non-lethal to mice even at doses as large as ~ 3 × 106 cfu.
Properties ofB. bronchiseptica strains Table 5 summarizes the results o f tests aimed at determining the properties o f the B. bronchiseptica strains studied. All phase I strains showed similar adenylate cyclase production, while phase III variants had no adenylate cyclase activity. Strains CF and B58 produced cytotoxin in high titres as measured in EBL assay, whereas B. bronchiseptica strains CFp61, B65 and ATCC 19 395 exhib-
204
T.MAGYAR
TABLE 4 Lethality caused by
Bordetella bronchiseptica
Dilution (cfu ml -~)
CF
CFp61
B58
-
10 a
-
9
-
10 10 10 10 10 10
-
10 10 10 10 10 10
108 107 106 10 5 10 4 103 102
2 3
s t r a i n s in i n t r a c e r e b r a l l y i n o c u l a t e d m i c e
2 2 3
B65
ATCC 19 3 9 5
PV6
PV6pl5
9
-
10 10 10 10 10 10
-
10 10 10 10 10 10 10
1 3 4
a N u m b e r o f s u r v i v o r s o u t o f 10 m i c e p e r g r o u p . TABLE 5 Characterisation of Strain
Phase
Bordetella bronchiseptica s t r a i n s
Haemolysis HA b Adenylate on B G a g a r~ cyclasec
Cyto-
Lienotoxicity
of the
L e t h a l i t y in m i c e
inoculated with
toxicity titre
Suspension a Extract a Suspension
CF
CUp61 B58 B65 ATCC 19 3 9 5 PV6 PV6pt5
I II1 I 1II IV
+ + -
+ + -
1042 0 955 0 ND h
2048 8-64 g 2048 8-256 4-64
I III
+ -
+ -
1103 0
0-8 0-2
Extract
i.v5
i.c. f
i.v.
32 0 16 0 0
128 16 512 32 16
+ + -
+ + -
+ + + + +
0 0
0 0
+ -
+ -
-
aBG = Bordet-Gengou. bHaemagglutination of calf erythrocytes, cCyclic AMP per min per assay (pmol). d H i g h e s t d i l u t i o n t h a t c a u s e d r e d u c t i o n in s p l e e n m a s s . e l n t r a v e n o u s l y , q n t r a c e r e b r a l l y , g R a n g e s r e p r e sent values obtained for several different batches of crude extract, hNot determined.
ited much lower cytotoxicity, with cytotoxin titres varying by batch. However, the values clearly indicated that these strains had not completely lost their cytotoxin-producing ability. B. bronchiseptica PV6 and PV6pl5 gave negligible reactions in this test. Similar differences were observed between the strains in lienotoxic titres of the different cell-free SEs. DISCUSSION
The results presented above suggest that the SEs and the WCSs of B. bronchiseptica react in a dissimilar way in the mouse lienotoxicity test. When examined in intravenously inoculated mice, lethality and lienotoxicity of SEs of
TOXICITY OF BORDETELLA BRONCH1SEPTICA IN MICE
205
B. bronchiseptica showed a close correlation with cytotoxicity for EBL cells. The extracts of highly cytotoxic strains (CF, B58) caused death and were lienotoxic in higher dilutions than those having low cytotoxic titres (CFp61, B65 and ATCC 19 395 ). The extracts of the barely cytotoxic strains PV6 and PV6pl 5 had no lethal and lienotoxic effect at all: on the contrary, they induced an increase in spleen mass. On the other hand, all WCSs were lethal both in intravenously and intracerebrally inoculated mice when they were prepared from phase I strains, irrespective of their lieno- or cytotoxin-producing ability. The WCS of strain PV6, a phase I, non-lienotoxic strain not producing cytotoxin and D N T (Magyar et al., 1988), proved to be lethal; however, unlike lienotoxic phase I strains, it caused a marked increase in the survivors' spleen mass. Thus, B. bronchiseptica strain PV6 seems to produce a factor lethal but not lienotoxic to mice. All these facts indicate that probably two mouse lethal factors (MLFs) can be distinguished: one is associated with cell integrity, viability and, probably, propagation in vivo, while the other can be found in SEs possessing cytotoxic and lienotoxic activity. K u m e et al. ( 1986 ) purified a dermonecrotic toxin from B. bronchiseptica which was lethal for mice and cytotoxic for EBL cells, indicating that these three activities are those of one toxin. Therefore, in this work the lethal factor, which seems to be identical with the cytotoxin, is termed mouse lethal toxin (MLT) to distinguish it from the other MLF associated with live phase I B. bronchiseptica. Whether lienotoxicity is an additional activity of the MLTDNT-cytotoxin or of a separate toxic substance could only be elucidated by testing the purified cytotoxin in the mouse lienotoxicity test. Cytotoxic phase I B. bronchiseptica strains produce both MLT and MLF. Thus, an additive effect of the two would be expected. B. bronchiseptica PV6, however, which only produces MLF but no MLT, was at least as lethal as strains possessing both activities. Furthermore, the WCS of B. bronchiseptica CF was more lethal in intravenously inoculated mice than that of strain B58 though the latter showed higher MLT activity than strain CF when their SEs were compared in the mouse lienotoxicity test. The role of MLT in the lethality of WCSs of phase I strains remains to be elucidated. For this, data should be obtained on the mechanism of release of MLT from B. bronchiseptica cells during propagation in vivo. In any case, lienotoxin seems to be released in sufficient quantity to cause splenic atrophy in mice intravenously inoculated with lienotoxic phase I strains. The WCSs of MLT-positive phase III and IV strains were non-lethal and non-lienotoxic, similarly to that of the MLT-negative phase III strain PV6pl 5. This suggests that phase III and IV strains do not produce MLF and the quantity of MLT and lienotoxin released from the disrupted bacterial cells is not sufficient to exert any adverse effect.
206
T.MAGYAR
It also follows from the results that lienotoxicity test with WCSs may give misleading results as regards the lienotoxin-producing ability of B. bronchiseptica. Inoculation of mice with B. bronchiseptica WCS, either by the intravenous or the intracerebral route, shows whether or not a strain is lethal for mice; however, only cell-free SEs can be used for accurate determination of the lienotoxin-producing ability of B. bronchiseptica in mice. Heat treatment abolished D N T activity (Kume et al., 1986). Lienotoxin and MLT were also destroyed by heating the SEs of B. bronchiseptica, suggesting that both activities are heat labile. The cell wall of Gram-negative bacteria contains heat-stable endotoxin. Spleen hypertrophy obtained by intravenous inoculation of heat-treated SEs indicates that a heat-stable component, presumably the endotoxin, is responsible for the increase in spleen mass. This is the dominant effect when MLT is absent or its quantity is insufficient. On the other hand, when MLT is present in sufficient quantity, a reduction in spleen mass will take place. In an earlier work B. bronchiseptica PV6 was shown to colonise pigs as well as B58 but not to cause AR (Magyar et al., 1988), since it produced an adhesin to nasal epithelial cells but not cytotoxin. When inoculating gnotobiotic piglets with strain PV6 followed by toxigenic P. multocida, however, strain PV6 was able to assist colonisation by P. multocida but to a lesser extent than the cytotoxic phase I strain B58 (Chanter et al., 1989). This suggests that some other virulence factors produced by phase I B. bronchiseptica strains may also be important in the pathogenicity ofB. bronchiseptica, as supported by the lethality observed in mice intravenously inoculated with strain PV6. During the phase shift, strain PV6 loses a range of different factors like adenylate cyclase, haemolysin and ability to agglutinate calf RBCs, and it becomes non-lethal (MLF-negative) at the same time. It is yet to be elucidated whether MLF is a new virulence determinant or is identical with any of the known factors or a complex of effects. The study of mutants defective in one or more virulence factors would be a possible way to get a closer insight into this problem. ACKNOWLEDGEMENTS
I thank Mrs. E. Laczi for her skilful technical assistance and Dr. A. Sz6kely for help with the English translation.
REFERENCES Chanter, N., Magyar, T. and Rutter, J.M., 1989. Interactions between Bordetella bronchiseptica and toxigenic Pasteurella multocida in atrophic rhinitis of pigs. Res. Vet. Sci., 47: 48-53. Collings, L.A. and Rutter, J.M., 1985. Virulence ofBordetella bronchiseptica in the porcine respiratory tract. J. Med. Microbiol., 19: 247-255.
TOXICITY OF BORDETELLA BRONCHISEPTICA 1N MICE
207
De Jong, M.F., Oei, H.L. and Tetenburg, G.J., 1980. AR-pathogenicity-test for Pasteurella multocida isolates. Proc. 5tb Int. Pig Vet. Soc. Congress, Copenhagen, Denmark, p. 211. Goodnow, R.A., 1980. Biology ofBordetella bronchiseptica. Microbiol. Rev., 44: 722-738. Iida, T. and Okonogi, T., 1971. Lienotoxicity ofBordetella pertussis in mice. J. Med. Microbiol., 4:51-60. KriJger, M. and Horsch, F., 1982. Untersuchungen zur Differenzierung von Bordetella bronchiseptica St~immen. 1. Mitteilung: Nachweis des hitzenlabilen Exotoxins und des H~imagglutinins. Arch. Exp. Vet. Med., Leipzig, 36:691-698. Kume, K., Nakai, T., Samejima, Y. and Sugimoto, C., 1986. Properties ofdermonecrotic toxin prepared from sonic extracts ofBordetella bronchiseptica. Infect. Immun., 52: 370-377. Lax, A.J., 1985. Is phase variation in Bordetella caused by mutation and selection? J. Gen. Microbiol., 131: 913-917. Magyar, T., Chanter, N., Lax, A.J., Rutter, J.M. and Hall, G.A., 1988. The pathogenesis of turbinate atrophy in pigs caused by Bordetella bronchiseptica. Vet. Microbiol., 18:135-146. Rutter, J.M. and Luther, P.D., 1984. A cell culture assay for toxigenic Pasteurella multocida from atrophic rhinitis of pigs. Vet. Rec., 108:451-454. Semj6n, G. and Magyar, T., 1985. A bovine haemagglutinin ofBordetella bronchiseptica responsible for adherence. Acta Vet. Hung., 33:129-136. Weiss, A.A. and Hewlett, E.L., 1986. Virulence factors of Bordetella pertussis. Annu. Rev. Microbiol., 40:661-686.
199
Virulence and lienotoxicity of Bordetella bronchiseptica in mice T. Magyar Veterinary Medical Research Institute of the Hungarian Academy of Sciences, 1t-1581 Budapest, P.O. Box 18, Hungary (Accepted 4 April 1990)
ABSTRACT Magyar, T., 1990. Virulence and lienotoxicity of Bordetella bronchiseptica in mice. Vet. Microbiol., 25: 199-207. Whole-cell suspensions (WCSs) and cell-flee sonicated extracts (SEs) of seven Bordetella bronchiseptica strains were studied for lethality and lienotoxicity in mice. Lethality was assessed after intravenous and intracerebral inoculation, and lienotoxicity by splenic atrophy after intravenous inoculation. The strains represented phase I isolates with or without cytotoxin production, their phase II1 subcultures and a phase IV variant. The lethality and lienotoxicity of the SEs were in close positive correlation with cytotoxin production. The WCSs of all phase I strains were lethal, irrespective of their cytotoxin- and lienotoxin-producing ability. The only difference was that cytotoxic phase I strains caused splenic atrophy while the noncytotoxic phase I strain induced splenic hypertrophy in the surviving mice. The WCSs of phase III and IV variants were non-lethal and caused splenic hypertrophy even though all but one of them showed some cyto- and lienotoxic activity when their SEs were tested. The results indicate that B. bronchiseptica possesses two different mouse lethal factors: one seems to be identical with the cytotoxin, the other is associated with cell integrity and viability and, presumably, propagation in vivo. It also follows from the results that only the SEs are suitable for accurate determination of the lienotoxin-producing ability ofB. bronchiseptica.
INTRODUCTION
Bordetella bronchiseptica is a respiratory pathogen involved in the aetiology of atrophic rhinitis (AR) of swine. It produces a range of potential virulence determinants: haemagglutinins, an adhesin to nasal epithelial cells (Semj6n and Magyar, 1985), cytotoxin, mouse lethal factor (Collings and Rutter, 1985 ), adenylate cyclase (Weiss and Hewlett, 1986) and haemolysin (Goodnow, 1980). Although their role has not been elucidated satisfactorily yet, it seems that colonisation of the nasal cavity in large numbers and the cytotoxin-producing ability of the strains are the most important factors in 0378-1135/90/$03.50
© 1990 - - Elsevier Science Publishers B.V.
200
T.MAGYAR
the pathogenesis ofturbinate atrophy caused by B. bronchiseptica (Magyar et al., 1988). The toxigenicity of B. bronchiseptica can be tested by different methods. Depending on the method applied, various names are used to designate the toxin: dermonecrotoxin ( D N T ) in the guinea pig skin test (de Jong et al., 1980 ), cytotoxin in embryonic bovine lung (EBL) cell assay (Rutter and Luther, 1984 ) and mouse lethal factor in mice (Collings and Rutter, 1985 ). Kriiger and Horsch ( 1982 ) described a mouse lienotoxicity test using broth cultures for determining the toxin-producing ability of B. bronchiseptica. Strains that were lethal to mice and which at sublethal doses caused atrophy of the spleen were considered toxic. In contrast, atoxic strains did not cause death and induced significant hypertrophy of the spleen. A similar test was described earlier by Iida and Okonogi ( 1971 ) to demonstrate the DNT-producing ability of B. pertussis. However, they examined a partially purified extract of B. pertussis. The purpose of the work presented here was to study the relationship between virulence and toxin-producing ability of B. bronchiseptica in mice. For this, the effects of whole-cell suspensions (WCSs) and cell-free sonicated extracts (SEs) of B. bronchiseptica were compared in the so-called mouse lienotoxicity test using toxic and atoxic phase I isolates, their phase III variants and a phase IV (rough) strain. Another objective was to assess the value of WCSs or SEs in determining the lienotoxigenicity of B. bronchiseptica. MATERIALS AND METHODS
B. bronchiseptica strains B. bronchiseptica strains CF and B58 were isolated from two different pig herds with clinical AR. Strain PV6 was isolated from a herd in which clinical AR had never been observed. These strains agglutinated calf red blood cells (RBCs), were haemolytic on Bordet-Gengou (BG) agar medium supplemented with 15% sheep blood and did not grow in the presence of erythromycin. B. bronchiseptica strains CFp61, B65 and PV6p15 w e r e derived from the fresh isolates by repeated subculture on MacConkey agar. These strains did not agglutinate calf RBCs, were non-haemolytic on BG agar and grew in the presence of erythromycin. On the basis of these criteria, they were designated as phase III variants of the original phase I isolates (Lax, 1985 ). B. bronchiseptica ATCC 19 395 was obtained from the American Type Culture Collection and during maintenance in the laboratory it became a phase IV variant with rough colony morphology. It also was non-haemolytic, erythromycin resistant and failed to agglutinate calf RBCs.
Haemagglutination test This test was done as described previously (Magyar et al., 1988 ).
TOXICITY OF BORDETELLA BRONCHISEPTICA IN MICE
201
Preparation of whole-cell suspensions B. bronchiseptica strains grown for 24 h at 37 °C on BG agar supplemented with 15% (v/v) sheep blood were harvested and suspended in phosphate buffered saline (PBS), pH 7.2. The concentration of the bacterial suspensions was adjusted to approximately 5X 109 colony forming units (cfu) m1-1 by photometry.
Preparation of cell-free sonicated extracts B. bronchiseptica strains propagated for 24 h at 37°C on BG agar plates were suspended in distilled water at a concentration of ~ 10 l° cfu ml -l. The cell suspensions were disrupted ultrasonically (MSE Sonifier 1-66, Great Britain) at full power for 5 rain whilst cooling in an icy water bath. Viable counts were reduced by ~ 90%. Sonicates were then centrifuged at 12 000 × g for 60 min at 4°C and filtered through a 0.22/tm membrane filter (Millipore Corp. ). Filtrates were tested for toxicity or adenylate cyclase activity within 2 days while stored at 4 oC.
Treatment of sonicated extracts A portion of SEs of B. bronchiseptica strain CF and CFp61 was exposed to heat (56 ° C, 1 h) before it was tested for lienotoxicity.
Adenylate cyclase test Bacterial extracts were tested for adenylate cyclase activity as described in an earlier paper (Magyar et al., 1988 ).
Experimental procedure Lethality was assessed in intravenously or intracerebrally inoculated mice, and lienotoxicity by determining the spleen mass of surviving mice. For the lienotoxicity test, twofold dilutions of test materials were prepared. The diluent was PBS for WCSs and distilled water for SEs. Groups of five white CFLP (LATI, G6d6116, Hungary) mice weighing 16-18 g were inoculated intravenously with 0.2 ml of one of the serial dilutions. Deaths were recorded up to 7 days postinoculation when the survivors were killed and spleen masses were determined. Relative spleen masses were calculated for 10 g body mass (spleen mass (mg) per body mass (g) × 10). For the intracerebral test tenfold dilutions of WCSs were prepared and groups of 10 mice were inoculated with 0.03 ml of one of the serial dilutions. Deaths were recorded up to 10 days postinoculation. Cytotoxicity for EBL cells (Rutter and Luther, 1984) was tested using twofold dilutions of sonicated extracts.
202
T.MAGYAR
RESULTS
Intravenous inoculation of mice The WCSs and cell-free SEs of the B. bronchiseptica strains designated CF and B58 (Table 1 ) were lethal to mice and induced a reduction in spleen mass when given at sublethal doses. The WCSs of phase III variants ( C U p 6 1 and B65 ) did not cause death and induced a remarkable increase in spleen mass. In contrast, the SEs of these subcultures resulted in death and splenic atrophy but at much lower dilutions (about one tenth) than the SEs prepared from the original phase I isolates. Similar results were obtained with the phase IV B. bronchiseptica strain ATCC 19 395 (Table 2 ), with the difference that its WCS produced only a moderate increase in spleen mass. B. bronchiseptica PV6 inoculated in living form was similarly lethal as the WCSs of the other phase I strains (CF and B58) but the surviving mice showed considerable lienal hypertrophy (Table 2). The SEs of both PV6 and PV6pl5 were nonlethal and produced an increase in spleen mass, similarly to the WCS of PV6pl 5. Heat treatment completely abolished the lethal and lienotoxic activity of the SEs prepared from B. bronchiseptica strains CF and CFp61 (Table 3 ). All the mice survived and developed a dose-dependent increase in spleen mass.
Intracerebral inoculation of mice Table 4 shows the virulence of B. bronch&eptica strains as determined by intracerebral inoculation of mice. Phase I strains (CF, B58 and PV6) had TABLE 1 Mouse lienotoxicity test w i t h whole-cell suspensions (WCSs) and cell-free sonicated extracts (SEs) ofBordetella bronchiseptica strains CF, B58 and their phase III variants CFp61and B65
CFp61
D i l u t i o n CF
WCSs
2 4 8 16 32 64 128 256 512
SEs
B58
WCSs
SEs
B65
WCSs
SEs
WCSs
SEs
let a
smb let
sm
let
sm
let
sm
let
sm
let
sm
let
sm
let
5/5 5/5 5/5 3/5 2/5 0/5 ND c ND ND
37 42 59
15 21 31 43 57 54
0/5 0/5 0/5 0/5 0/5 0/5 ND ND ND
194 187 137 93 88 76
2/5 0/5 0/5 0/5 0/5 0/5 ND ND ND
18 26 34 46 55 58
5/5 3/5 3/5 0/5 0/5 0/5 ND ND ND
27 38 40 58 64
5/5 5/5 5/5 5/5 5/5 5/5 2/5 0/5 0/5
11 22 40
0/5 0/5 0/5 0/5 0/5 0/5 ND ND ND
126 110 114 93 82 70
5/5 5/5 1/5 0/5 0/5 0/5 ND ND ND
5/5 5/5 5/5 1/5 0/5 0/5 0/5 0/5 0/5
alet = lethality. bsm = relative spleen mass. CND ---not determined.
sm m
27 39 45 53
TOXICITY OF BORDETELLA BRONCHISEPTICAIN MICE
203
TABLE 2
Mouse lienotoxicity test with whole-cell suspensions (WCSs) and cell-free sonicated extracts (SEs) ofBordetella bronchiseptica strains ATCC 19 395, PV6 and PV6p~5 Dilution
ATCC 19 395 WCSs
2 4 8 16 32 64
PV6 SEs
PV6pl5
WCSs
SEs
WCSs
SEs
leta
sm b
let
sm
let
sm
let
sm
let
sm
let
sm
0/5 0/5 0/5 0/5 0/5 0/5
88 74 77 73 67 61
5/5 5/5 2/5 1/5 0/5 0/5
18 40 44 50
5/5 5/5 5/5 5/5 4/5 1/5
178 118
0/5 0/5 0/5 0/5 0/5 0/5
110 85 79 71 70 61
0/5 0/5 0/5 0/5 0/5 0/5
141 131 112 86 77 56
0/5 0/5 0/5 0/5 0/5 0/5
128 99 96 89 80 72
alet = lethality. bsm = relative spleen mass. TABLE 3
Mouse lienotoxicity test with heat-treated cell-free sonicated extracts ofBordetella bronchiseptica strains CF and
CFp61
Dilution
2 4 8 16 32 64
CFp61
CF let a
sm b
let
sm
0/5 0/5 0/5 0/5 0/5 0/5
96 80 77 75 60 52
0/5 0/5 0/5 0/5 0/5 0/5
83 73 68 61 58 55
~let = lethality. bsm = relative spleen mass.
lethal effects o f similar degree: a single dose o f ~ 3 cfu was sufficient to kill more than 50% o f the mice. In contrast, phase III and IV strains (CFp61, B65, PV6pl 5 and ATCC ) were non-lethal to mice even at doses as large as ~ 3 × 106 cfu.
Properties ofB. bronchiseptica strains Table 5 summarizes the results o f tests aimed at determining the properties o f the B. bronchiseptica strains studied. All phase I strains showed similar adenylate cyclase production, while phase III variants had no adenylate cyclase activity. Strains CF and B58 produced cytotoxin in high titres as measured in EBL assay, whereas B. bronchiseptica strains CFp61, B65 and ATCC 19 395 exhib-
204
T.MAGYAR
TABLE 4 Lethality caused by
Bordetella bronchiseptica
Dilution (cfu ml -~)
CF
CFp61
B58
-
10 a
-
9
-
10 10 10 10 10 10
-
10 10 10 10 10 10
108 107 106 10 5 10 4 103 102
2 3
s t r a i n s in i n t r a c e r e b r a l l y i n o c u l a t e d m i c e
2 2 3
B65
ATCC 19 3 9 5
PV6
PV6pl5
9
-
10 10 10 10 10 10
-
10 10 10 10 10 10 10
1 3 4
a N u m b e r o f s u r v i v o r s o u t o f 10 m i c e p e r g r o u p . TABLE 5 Characterisation of Strain
Phase
Bordetella bronchiseptica s t r a i n s
Haemolysis HA b Adenylate on B G a g a r~ cyclasec
Cyto-
Lienotoxicity
of the
L e t h a l i t y in m i c e
inoculated with
toxicity titre
Suspension a Extract a Suspension
CF
CUp61 B58 B65 ATCC 19 3 9 5 PV6 PV6pt5
I II1 I 1II IV
+ + -
+ + -
1042 0 955 0 ND h
2048 8-64 g 2048 8-256 4-64
I III
+ -
+ -
1103 0
0-8 0-2
Extract
i.v5
i.c. f
i.v.
32 0 16 0 0
128 16 512 32 16
+ + -
+ + -
+ + + + +
0 0
0 0
+ -
+ -
-
aBG = Bordet-Gengou. bHaemagglutination of calf erythrocytes, cCyclic AMP per min per assay (pmol). d H i g h e s t d i l u t i o n t h a t c a u s e d r e d u c t i o n in s p l e e n m a s s . e l n t r a v e n o u s l y , q n t r a c e r e b r a l l y , g R a n g e s r e p r e sent values obtained for several different batches of crude extract, hNot determined.
ited much lower cytotoxicity, with cytotoxin titres varying by batch. However, the values clearly indicated that these strains had not completely lost their cytotoxin-producing ability. B. bronchiseptica PV6 and PV6pl5 gave negligible reactions in this test. Similar differences were observed between the strains in lienotoxic titres of the different cell-free SEs. DISCUSSION
The results presented above suggest that the SEs and the WCSs of B. bronchiseptica react in a dissimilar way in the mouse lienotoxicity test. When examined in intravenously inoculated mice, lethality and lienotoxicity of SEs of
TOXICITY OF BORDETELLA BRONCH1SEPTICA IN MICE
205
B. bronchiseptica showed a close correlation with cytotoxicity for EBL cells. The extracts of highly cytotoxic strains (CF, B58) caused death and were lienotoxic in higher dilutions than those having low cytotoxic titres (CFp61, B65 and ATCC 19 395 ). The extracts of the barely cytotoxic strains PV6 and PV6pl 5 had no lethal and lienotoxic effect at all: on the contrary, they induced an increase in spleen mass. On the other hand, all WCSs were lethal both in intravenously and intracerebrally inoculated mice when they were prepared from phase I strains, irrespective of their lieno- or cytotoxin-producing ability. The WCS of strain PV6, a phase I, non-lienotoxic strain not producing cytotoxin and D N T (Magyar et al., 1988), proved to be lethal; however, unlike lienotoxic phase I strains, it caused a marked increase in the survivors' spleen mass. Thus, B. bronchiseptica strain PV6 seems to produce a factor lethal but not lienotoxic to mice. All these facts indicate that probably two mouse lethal factors (MLFs) can be distinguished: one is associated with cell integrity, viability and, probably, propagation in vivo, while the other can be found in SEs possessing cytotoxic and lienotoxic activity. K u m e et al. ( 1986 ) purified a dermonecrotic toxin from B. bronchiseptica which was lethal for mice and cytotoxic for EBL cells, indicating that these three activities are those of one toxin. Therefore, in this work the lethal factor, which seems to be identical with the cytotoxin, is termed mouse lethal toxin (MLT) to distinguish it from the other MLF associated with live phase I B. bronchiseptica. Whether lienotoxicity is an additional activity of the MLTDNT-cytotoxin or of a separate toxic substance could only be elucidated by testing the purified cytotoxin in the mouse lienotoxicity test. Cytotoxic phase I B. bronchiseptica strains produce both MLT and MLF. Thus, an additive effect of the two would be expected. B. bronchiseptica PV6, however, which only produces MLF but no MLT, was at least as lethal as strains possessing both activities. Furthermore, the WCS of B. bronchiseptica CF was more lethal in intravenously inoculated mice than that of strain B58 though the latter showed higher MLT activity than strain CF when their SEs were compared in the mouse lienotoxicity test. The role of MLT in the lethality of WCSs of phase I strains remains to be elucidated. For this, data should be obtained on the mechanism of release of MLT from B. bronchiseptica cells during propagation in vivo. In any case, lienotoxin seems to be released in sufficient quantity to cause splenic atrophy in mice intravenously inoculated with lienotoxic phase I strains. The WCSs of MLT-positive phase III and IV strains were non-lethal and non-lienotoxic, similarly to that of the MLT-negative phase III strain PV6pl 5. This suggests that phase III and IV strains do not produce MLF and the quantity of MLT and lienotoxin released from the disrupted bacterial cells is not sufficient to exert any adverse effect.
206
T.MAGYAR
It also follows from the results that lienotoxicity test with WCSs may give misleading results as regards the lienotoxin-producing ability of B. bronchiseptica. Inoculation of mice with B. bronchiseptica WCS, either by the intravenous or the intracerebral route, shows whether or not a strain is lethal for mice; however, only cell-free SEs can be used for accurate determination of the lienotoxin-producing ability of B. bronchiseptica in mice. Heat treatment abolished D N T activity (Kume et al., 1986). Lienotoxin and MLT were also destroyed by heating the SEs of B. bronchiseptica, suggesting that both activities are heat labile. The cell wall of Gram-negative bacteria contains heat-stable endotoxin. Spleen hypertrophy obtained by intravenous inoculation of heat-treated SEs indicates that a heat-stable component, presumably the endotoxin, is responsible for the increase in spleen mass. This is the dominant effect when MLT is absent or its quantity is insufficient. On the other hand, when MLT is present in sufficient quantity, a reduction in spleen mass will take place. In an earlier work B. bronchiseptica PV6 was shown to colonise pigs as well as B58 but not to cause AR (Magyar et al., 1988), since it produced an adhesin to nasal epithelial cells but not cytotoxin. When inoculating gnotobiotic piglets with strain PV6 followed by toxigenic P. multocida, however, strain PV6 was able to assist colonisation by P. multocida but to a lesser extent than the cytotoxic phase I strain B58 (Chanter et al., 1989). This suggests that some other virulence factors produced by phase I B. bronchiseptica strains may also be important in the pathogenicity ofB. bronchiseptica, as supported by the lethality observed in mice intravenously inoculated with strain PV6. During the phase shift, strain PV6 loses a range of different factors like adenylate cyclase, haemolysin and ability to agglutinate calf RBCs, and it becomes non-lethal (MLF-negative) at the same time. It is yet to be elucidated whether MLF is a new virulence determinant or is identical with any of the known factors or a complex of effects. The study of mutants defective in one or more virulence factors would be a possible way to get a closer insight into this problem. ACKNOWLEDGEMENTS
I thank Mrs. E. Laczi for her skilful technical assistance and Dr. A. Sz6kely for help with the English translation.
REFERENCES Chanter, N., Magyar, T. and Rutter, J.M., 1989. Interactions between Bordetella bronchiseptica and toxigenic Pasteurella multocida in atrophic rhinitis of pigs. Res. Vet. Sci., 47: 48-53. Collings, L.A. and Rutter, J.M., 1985. Virulence ofBordetella bronchiseptica in the porcine respiratory tract. J. Med. Microbiol., 19: 247-255.
TOXICITY OF BORDETELLA BRONCHISEPTICA 1N MICE
207
De Jong, M.F., Oei, H.L. and Tetenburg, G.J., 1980. AR-pathogenicity-test for Pasteurella multocida isolates. Proc. 5tb Int. Pig Vet. Soc. Congress, Copenhagen, Denmark, p. 211. Goodnow, R.A., 1980. Biology ofBordetella bronchiseptica. Microbiol. Rev., 44: 722-738. Iida, T. and Okonogi, T., 1971. Lienotoxicity ofBordetella pertussis in mice. J. Med. Microbiol., 4:51-60. KriJger, M. and Horsch, F., 1982. Untersuchungen zur Differenzierung von Bordetella bronchiseptica St~immen. 1. Mitteilung: Nachweis des hitzenlabilen Exotoxins und des H~imagglutinins. Arch. Exp. Vet. Med., Leipzig, 36:691-698. Kume, K., Nakai, T., Samejima, Y. and Sugimoto, C., 1986. Properties ofdermonecrotic toxin prepared from sonic extracts ofBordetella bronchiseptica. Infect. Immun., 52: 370-377. Lax, A.J., 1985. Is phase variation in Bordetella caused by mutation and selection? J. Gen. Microbiol., 131: 913-917. Magyar, T., Chanter, N., Lax, A.J., Rutter, J.M. and Hall, G.A., 1988. The pathogenesis of turbinate atrophy in pigs caused by Bordetella bronchiseptica. Vet. Microbiol., 18:135-146. Rutter, J.M. and Luther, P.D., 1984. A cell culture assay for toxigenic Pasteurella multocida from atrophic rhinitis of pigs. Vet. Rec., 108:451-454. Semj6n, G. and Magyar, T., 1985. A bovine haemagglutinin ofBordetella bronchiseptica responsible for adherence. Acta Vet. Hung., 33:129-136. Weiss, A.A. and Hewlett, E.L., 1986. Virulence factors of Bordetella pertussis. Annu. Rev. Microbiol., 40:661-686.
