APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Jan. 1977, p. 10-14 Copyright ©) 1977 American Society for Microbiology
Vol. 33, No. 1 Printed in U.S.A.
Enteropathogenicity of Vibrio parahaemolyticus in the Ligated Rabbit Ileum DAVID F. BROWN,* PROCTER L. SPAULDING, AND ROBERT M. TWEDT U.S. Department ofHealth, Education, and Welfare, Public Health Service, Food and Drug Administration, Cincinnati, Ohio 45226 Received for publication 13 August 1976
The enteropathogenicity of Vibrio parahaemolyticus was investigated by contrasting the effects of whole cells, cell fragments, cell-free preparations, and media constituents injected into rtabbit ileal loops. Three of 20 cultures utilized were Kanagawa-negative strains from seawater and sea fish. The remaining 17 cultures included both Kanagawa-positive and -negative strains from Japanese victims of gastroenteritis. Broth culture filtrates concentrated 10-fold by dialysis against 30% Carbowax were unreactive, whereas lyophilized filtrates, regardless of Kanagawa type, as well as all sterile broth preparations containing .5% NaCl gave positive reactions in the rabbit gut. In contrast, crude lysates derived from broth cultures of Kanagawa-positive strains caused loop dilatation; lysate supernatants were unreactive. Lysates of cells washed from brain heart infusion agar were more reactive than lysates from Trypticase soy agar-grown cells. When agar-grown cell lysates prepared by disruption in saline were dialyzed against distilled water, they were devoid of gut reactivity. Reactivity was restored in dialysands resuspended in saline and in dialysates concentrated 10fold. The agar-grown cell lysates exhibited Kanagawa-type hemolysis. Our data support the conclusion that the rabbit loop reactivity observed with lyophilized, cell-free culture filtrates may result from excessively elevated NaCl concentrations, and that a toxic factor associated with large-cell particles may be dialyzable, depends on saline for expression, and resembles the Kanagawa hemolysin. tive V. parahaemolyticus strains (13, 14). By contrast, we described a toxic factor that was capable of causing enteric reactivity in the rabbit and that appeared to be associated with relatively large-cell particles from hemolytic strains (14). The present study was undertaken to further characterize the role of the particulate toxic factor in the enteropathogenicity of V. parahaemolyticus.
A close association between the source of Vibrio parahaemolyticus isolates and their hemolytic activity has been well established (6, 8, 10). According to this relationship, termed the Kanagawa phenomenon, most strains isolated from diarrheal stools are f3-hemolytic on a special blood agar (17); most marine isolates are not. We have shown that a strain's hemolytic ability is directly related to its capability to produce dilatation in the ligated rabbit ileal loop test (13, 14). Several workers have isolated and purified a thermostable direct hemolysin from culture filtrates of Kanagawa-positive strains (3, 9, 12, 18). The purified hemolysin has been shown to be a toxic protein of 45,000 daltons that exhibits a mean lethal dose for 20-g mice by intraperitoneal inoculation of 1.5 ,g, a minimal cutaneous response dose in guinea pig skin test of 2.5 ,ug, and a minimal ileal loop dilatation dose of 200 p.g (19, 20). Cardiotoxicity of the purified hemolysin has been demonstrated in rats and in cultured mouse heart cells (2). In previous investigations, we were unsuccessful in our attempts to demonstrate an enterotoxin in culture filtrates of Kanagawa-posi-
MATERIALS AND METHODS Cultures. The 20 cultures of V. parahaemolyticus used (Table 1) were, with three exceptions, isolated from human victims of gastroenteritis in Japan. They were obtained from H. Zen-Yoji and Y. Kudoh, Tokyo Metropolitan Research Laboratory of Public Health, Tokyo; Y. Miyamoto, Kanagawa Prefectural Public Health Laboratory, Yokohama; M. Sochard, Catholic University, Washington, D.C.; and R. Sakazaki, National Institute of Health, Tokyo, Japan. The media and culture conditions used for maintenance of stock cultures as well as for verification of strains have been previously described (15). All strains exhibited the well-established properties that identify V. parahaemolyticus (16). Preparation of inocula. Cultures were incubated in brain heart infusion (BHI; Difco) broth contain10
VOL. 33, 1977
ENTEROPATHOGENICITY OF V. PARAHAEMOLYTICUS
11
TABLE 1. Culture origin and designation of Vibrio parahaemolyticus strains Strain
Source
reaction
K serotype
Donor
Anzai Nakatsukawa 9337 9379 9382 11590 550 551 552 553 557 391 392 393 395 396 T3454-1 T6069 T6367 S-42-87 UT, Untypable.
Patient's stool Patient's stool Patient's stool Patient's stool Patient's stool Patient's stool Patient's stool Patient's stool Patient's stool Patient's stool Patient's stool Patient's stool Patient's stool Patient's stool Patient's stool Patient's stool Patient's stool Sea fish Sea fish Seawater
+ + + + + + + + + +
11 3 38 7 11 11 8 8 8 4 UTa 54 54 29 13 8 12 6 54 3
H. Zen-Yoji H. Zen-Yoji Y. Miyamoto Y. Miyamoto Y. Miyamoto M. Sochard R. Sakazaki R. Sakazaki R. Sakazaki R. Sakazaki R. Sakazaki R. Sakazaki R. Sakazaki R. Sakazaki R. Sakazaki R. Sakazaki Y. Kudoh Y. Kudoh Y. Kudoh Y. Miyamoto
ing 3% NaCl for 24 h at 35°C. The resulting cell density was .-109 cells/ml. Culture supernatants were prepared by centrifuging broth cultures at 25,000 x g for 10 min. Checks at intervals proved the sterility of the preparation. Culture filtrates were made by passing supernatants through a 0.45-Am membrane filter (Millipore Corp., Bedford, Mass.). Culture filtrates were concentrated 10-fold by lyophilization and/or by dialysis overnight at 4°C in 30% Carbowax (Union Carbide Corp., New York). Suspensions of 108 cells/ml were prepared from cultures grown on BHI agar and Trypticase soy agar (TSA; BBL) containing 3% NaCl for 18 h at 35°C. Cells were washed three times and resuspended in phosphate-buffered saline (PBS; 0.425% NaCl, 1.5275% Na2HPO4, 0.244% KH2PO4, pH 7.2). Crude lysates were prepared from broth cultures or washed cell suspensions of selected strains by shaking with glass beads according to Howe et al. (4). Broth culture lysates contained soluble extracellular material, cytoplasm, and growth medium in addition to cell fragments. A portion of each broth culture lysate was centrifuged at 25,000 x g, and the sediment was resuspended in an equal volume of sterile broth. The resulting lysate supernatants, resuspended lysate sediments, and original crude lysates were tested concomitantly. In certain cases, washed cell lysates were concentrated 10-fold by dialysis overnight at 4°C in 30% Carbowax. Two diluents for disrupting washed cell suspensions were compared in a preliminary series of experiments. After final washing, the cells of certain strains were resuspended either in distilled water (7) or in PBS containing glass beads before lysis (4). In a later study of the effects of dialysis on the ileal loop activity of the toxic particulate factor, lysates obtained by disrupting agar-grown cells in PBS were dialyzed overnight at 4°C in five volumes
-
-
-
of distilled water. A portion of each dialysand was centrifuged at 25,000 x g and the sediment was resuspended in PBS, whereas a portion of each dialysate was concentrated 10-fold against 30% Carbowax. The lysate, dialysand, resusp2nded dialysand, dialysate, and concentrated dialysate from each strain were tested in parallel. Ileal loop procedure. The reactivity of various test preparations was tested by a modification of the rabbit ileal loop technique (13, 14). One-milliliter volumes of control or test preparations were injected into individual loops. An index of fluid accumulation was derived (1) from the ratio of loop fluid volume-to-loop length. A test preparation was considered positive if the index was threefold greater than that of the negative control. Results were discarded if either positive or negative controls were inappropriate.
RESULTS In our earlier studies (13, 14), we were unable to demonstrate an enterotoxic factor present in cell-free filtrates of Kanagawa-positive V. parahaemolyticus broth cultures even when the latter were concentrated 10-fold by dialysis. However, Sakazaki et al. (11) produced rabbit ileal loop dilatation with all V. parahaemolyticus strains when the test filtrates were concentrated 10-fold by lyophilization; filtrates concentrated by dialysis were unreactive. To investigate this discrepancy, we compared BHI broth cultures, culture filtrates, and concentrated culture filtrates of four hemolytic strains of V. parahaemolyticus (Table 2). It is apparent that filtrates concentrated 10-fold by
APPL. ENVIRON. MICROBIOL.
BROWN, SPAULDING, AND TWEDT
12
lyophilization were reactive in the rabbit gut, whereas filtrates concentrated by dialysis were not. Furthermore, most culture supernatants concentrated by lyophilization dilated the ligated rabbit gut regardless of the strain's Kanagawa reaction. In contrast, broth cultures of only Kanagawa-positive strains were capable of inducing the gut response (Table 3). Suspecting that a medium constituent was responsible for the observations, we tested sterile BHI broth concentrated 10-fold by lyophilization or by dialysis as well as single-strength broth prepared without salt. Whereas the dialyzed and single-strength broths were unreactive, the lyophilized broth produced loop dilatation in all tests. Furthermore, BHI broth prepared at 10-fold normal concentration but without NaCl was unreactive. We then tested sterile broth containing NaCl in the range of 0 to 30%. Concentrations of 5% and above were all loop positive. To investigate the possible existence of a toxic factor associated with the cell structure, we compared rabbit ileal loop reactivity of various preparations of four Kanagawa-positive strains of V. parahaemolyticus grown in BHI broth (Table 4). As expected, broth cultures produced severe dilatation 31 out of 40 times tested, or 78% positive. The percent positive reactivity of culture supernatants was 20%; of TABLE 2. Rabbit ileal loop reactivity of broth cultures and culture filtrates from Kanagawapositive Vibrio parahaemolyticus strains
Strain
Broth cultures
Culture filtrates
Concentrated filtrates
Dialyzed 550 551 552 553 Total positive % Positive a
4/4a 4/4 3/3 4/4 15/15 100
0/4
0/3
0/4 0/3 0/4 0/15 0
0/3 0/3 0/3 0/12 0
Lyophilized 4/4
4/4 3/3 4/4
15/15 100
Number of times positive/number of times tested.
lysates, 63%; of resuspended lysates, 50%; and of lysate supernatants, 0%. To determine the effect of cell nutrition on the formation of the toxic factor, we compared cell suspensions and lysates of four hemolytic strains of V. parahaemolyticus grown on TSA and BHI agar (Table 5). Although whole cell preparations from both media were active, those from TSA were significantly less so. In addition, the lysates derived from TSA-grown cells were only minimally active. In preliminary experiments, washed cell lysates of hemolytic strains were concentrated by dialysis against 30% Carbowax. Whereas the lysates were rabbit loop reactive, they were devoid of that ability after concentration. Furthermore, in another early study comparing the efficacy of disrupting washed cell suspensions in PBS or distilled water, preparations in the latter proved to be inactive. Because these studies suggested that the ionic content of the suspending media affected the reactivity of the toxic factor associated with lysate particles, we studied more thoroughly the effect of dialysis of lysates. Lysates of cells from three Kanagawa-positive strains grown on BHI agar were reactive in the ileal loop test, whereas the same lysates TABLE 3. Rabbit ileal loop reactivity of broth cultures and culture supernatants from Vibrio parahaemolyticus strains Strain
Source
550 551 552 553 391 392 393 395 396 T-3454-1 T-6069 T-6367 S-42-87
Stool Stool Stool Stool Stool Stool Sto(l Stool Stool Stool Sea fish Sea fish Seawater
Kanagawa reaction
Broth culture
+ + + +
+ + + +
-
-
Lyophilized supernatant + + + + + + + + + + +
TABLE 4. Rabbit ileal loop reactivity of various preparations from broth cultures of Kanagawa-positive Vibrio parahaemolyticus strains Strain
550 551 552 553 Total positive % Positive a
Broth cultures
9/10a 8/10 6/10
8/10 31/40 78
9/10 5/10 5/10 7/10
Resuspended lysates 6/10 5/10 4/10 5/10
Lysate supernatants 0/10 0/10 0/10 0/10
26/40 63
20/40 50
0/40 0
Culture supernatants 3/10 2/10 2/10 1/10
Lysates
8/40 20
Number of times positive/number of times tested.
VOL. 33, 1977
13
ENTEROPATHOGENICITY OF V. PARAHAEMOLYTICUS
were completely inactive after overnight di- of their recent study of false-positive ileal loop alysis against distilled water (Table 6). When reactions exhibited by filtrates of V. parahaethe dialysands were centrifuged and resus- molyticus broth cultures (5). In contrast to this pended in PBS to their orignal volume, most of suggestion, the enterotoxic factor reported by the reactivity of the particles was restored in Sakazaki et al. (11) to be present in similar each case. The dialysates, in contrast, were lyophilized preparations of most strains of V. inactive at their original concentration. How- parahaemolyticus was heat labile. The results of our present study demonstrate ever, when they were concentrated 10-fold, the dialysates of two of three strains produced ileal (Table 4) that a toxic factor present in broth loop dilatation. The maximum NaCl concentra- cultures of live cells is present only in minimal tion obtainable in the concentrated dialysates concentration in the culture supernatant. That by our procedure was 1.5%. the factor is associated with cell particles is To determine the relationship of the toxic borne out by the reactivity of lysates resusparticulate factor to the Kanagawa hemol- pended in sterile broth as well as the lack of ysin, we tested various preparations from five reactivity of unconcentrated lysate supernastrains of V. parahaemolyticus on Wagatsuma tant. Available nutrients influence the expresagar. The results (Table 7) demonstrate that ly- sion of the toxic factor (Table 5). The BHI agar sates derived from Kanagawa-positive cultures appeared to support formation of a particulate yield Kanagawa-type hemolysis. Furthermore, toxic principle that was qualitatively or quantiwhereas lysate supernatants also show hemoly- tatively different from that formed on TSA. The activity of the toxic particulate factor is sis, filtration through 0.45-,m membranes removes this reactivity. dependent on the presence of a minimal saline concentration in the suspending medium. DISCUSSION When lysates were dialyzed overnight against Our data suggest that the positive ileal loop distilled water, the ability of the dialysand to dilatations observed (Tables 2 and 3) when cul- produce rabbit ileal loop dilatation was abolture filtrates concentrated 10-fold by lyophiliza- ished (Table 6). The toxic factor itself was not tion were introduced into rabbit ileal loops re- lost completely from the dialysand particles besulted from the elevated salt concentrations in cause resuspending them in fresh saline resuch preparations. Similar conclusions were stored toxicity. Nevertheless, the toxic factor reached by Johnson and Calia as a consequence may be dialyzable because the dialysate, concentrated 10-fold, elicited gut reactivity. The toxic reactivity of concentrated dialysates canTABLE 5. Rabbit ileal loop reactivity of various not be attributed to salt effects since the maxipreparations from Kanagawa-positive Vibrio mum NaCl concentration obtainable in these parahaemolyticus strains grown on two different preparations by our procedure was only 1.5%, media well below the critical level. Brain heart infusion agar
Trypticase soy agar
Cell sus-
Cell sus-
Strain
pensions 550 551 552 553
Total positive % Positive a
3/4a 3/4 2/4 3/4 11/16 69
Lysates 2/4 2/4 1/4 2/4 7/16 44
pensions 2/4 1/4 2/4 2/4 7/16 44
Lysates 0/4 1/4 1/4 2/4 4/16 25
Number of times positive/number of times tested.
TABLE 7. Kanagawa-type hemolysis by cultures and various preparations from lysates of Vibrio parahaemolyticus grown on BHI agar
pernatants
Lysate su-
Lysate filtrates
-
-
-
+
_
_
+ + +
+ + +
-
Strain
Cultulres
Lysates
393 557 553 554 9337
-
+ + +
-
_
TABLE 6. Rabbit ileal loop reactivity of various preparations from lysates of Kanagawa-positive Vibrio parahaemolyticus strains grown on BHI agar Strain
Lysates
Lysate dialysands
3/3a 0/3 3/3 0/3 3/3 0/3 0 Number of times positive/number of times tested. 553 9337 11590
Concentrated dialysates
Resuspended
Lysate dialy-
2/3 3/3 2/3
0/3
2/3
0/3 0/3
0/3
dialysands
sates
3/3
14
BROWN, SPAULDING, AND TWEDT
These data suggest that the toxic principle associated with lysate particles is dissociable and active in the absence of the particles themselves, though not in the absence of a minimal saline concentration. The direct relationship of the toxic particulate factor with the purified hemolysin is indicated by two principal lines of evidence. (i) There is a close association between the ability of broth cultures or cell suspensions of a V. parahaemolyticus isolate to dilate the ligated rabbit ileum and that strain's ability to produce Kanagawa-type hemolysin (13, 14); and (ii) nonfilterable lysate particles of Kanagawa-positive cells produce rabbit loop dilatation and also hemolyze Wagatsuma agar. Lysates of Kanagawa-negative cells are unreactive in both tests (Table 7). We cannot determine, from our results, whether the hemolysin has associated with the lysate particles as a consequence of preparation or whether that association indicates substantial physiological importance. It is possible that the lysate particles contain the source of hemolysin synthesis. ACKNOWLEDGMENTS We thank H. Zen-Yoji, Y. Kudoh, Y. Miyamoto, M. Sochard, and R. Sakazaki for providing cultures.
LITERATURE CITED 1. Drucker, M., R. Yeiven, and T. G. Sacks. 1967. Pathogenesis of Escherichia coli enteritis in the ligated rabbit gut. Isr. J. Med. Sci. 3:445-452. 2. Honda, T., K. Gohsima, Y. Takeda, Y. Sugino, and T. Miwatani. 1976. Demonstration of the cardiotoxicity
of the thermostable direct hemolysin (lethal toxin) produced by Vibrio parahaemolyticus. Infect. Immun. 13:163-171. 3. Honda, T., S. Taga, T. Takeda, M. A. Hasibuan, Y. Takeda, and T. Miwatani. 1976. Identification of lethal toxin with the thermostable direct hemolysin produced by Vibrio parahaemolyticus, and some physiochemical properties of the purified toxin. Infect. Immun. 13:133-139. 4. Howe, J. M., W. R. Featherstone, W. J. Stadelman, and G. J. Banwart. 1965. Amino acid composition of certain bacterial cell wall proteins. Appl. Microbiol. 13:650-652. 5. Johnson, D. E., and F. M. Calia. 1976. False-positive rabbit ileal loop reactions attributed to Vibrio parahaemolyticus broth filtrates. J. Infect. Dis. 133:436440. 6. Kato, T., Y. Obara, H. Ichinoe, K. Nagashima, S. Akiyama, K. Takizawa, A. Matsushima, S. Yamai, and Y. Miyamoto. 1965. Grouping of V. parahaemoly-
APPL. ENVIRON. MICROBIOL. ticus with a hemolysis reaction. Shokuhin Eisei Kenkyu 15:83-86. 7. Lee, J. S. 1972. Inactivation of Vibrio parahaemolyticus in distilled water. Appl. Microbiol. 23:166-167. 8. Miyamoto, Y., T. Kato, Y. Obara, S. Akiyama, K. Takizawa, and S. Yamai. 1969. In vitro hemolytic characteristic of Vibrio parahaemolyticus: its close correlation with human pathogenicity. J. Bacteriol. 100:1147-1149. 9. Obara, Y. 1971. Studies on hemolytic factors of Vibrio parahaemolyticus. II. The extraction of hemolysin and its properties. (In Japanese) J. Jpn. Assoc. Infect. Dis. 45:392-398. 10. Sakazaki, R., K. Tamura, T. Kato, Y. Obara, S. Yamai, and K. Hobo. 1968. Studies on the enteropathogenic facultatively halophilic bacteria, Vibrio parahaemolyticus. III. Enteropathogenicity. Jpn. J. Med. Sci. Biol. 21:325-331. 11. Sakazaki, R., K. Tamura, A. Nakamura, T. Kurata, A. Kurata, A. Gohda, and Y. Kazuno. 1974. Studies on enteropathogenic activity of Vibrio parahaemolyticus using ligated gut loop model in rabbits. Jpn. J. Med. Sci. Biol. 27:35-43. 12. Sakurai, J., A. Matsuzaki, and T. Miwatani. 1973. Purification and characterization of thermostable direct hemolysin of Vibrio parahaemolyticus. Infect. Immun. 8:775-780. 13. Twedt, R. M., and D. F. Brown. 1973. Vibrio parahaemolyticus: infection or toxicosis? J. Milk Food Technol. 36:129-134. 14. Twedt, R. M., and D. F. Brown. 1974. Studies on the enteropathogenicity of Vibrio parahaemolyticus in the ligated rabbit ileum, p. 211-217. In T. Fujino, G. Sakaguchi, R. Sakazaki, and Y. Takeda (ed.), International Symposium on Vibrio parahaemolyticus. Saikon Publishing Co., Tokyo. 15. Twedt, R. M., P. L. Spaulding, and H. E. Hall. 1969. Morphological, cultural, biochemical, and serological comparison of Japanese strains of Vibrio parahaemolyticus with related cultures isolated in the United States. J. Bacteriol. 98:511-518. 16. U.S. Department of Health, Education, and Welfare. 1972. Bacteriological analytical manual for foods, 3rd ed. Public Health Service, Food and Drug Administration, U.S. Department of Health, Education, and Welfare, Washington, D.C. 17. Wagatsuma, S. 1968. A medium for the test of the hemolytic activity of Vibrio parahaemolyticus. (In Japanese) Media Circle 13:159-161. 18. Zen-Yoji, H., H. Hitakoto, S. Morozumi, and R. A. LeClair. 1971. Purification and characterization of a hemolysin produced by Vibrio parahaemolyticus. J. Infect. Dis. 123:665-667. 19. Zen-Yoji, H., Y. Kodoh, H. Igarashi, K. Ohta, and K. Fukai. 1974. Purification and identification on enteropathogenic toxins "a" and "a"' produced by Vibrio parahaemolyticus and their biological and pathological activities, p. 237-243. In T. Fujino, G. Sakaguchi, R. Sakazaki, and Y. Takeda (ed.), International Symposium on Vibrio parahaemolyticus. Saikon Publishing Co., Tokyo. 20. Zen-Yoji, H., Y. Kudoh, H. Igarashi, K. Ohta, K. Fukai, and T. Hoshino. 1975. Further studies on characterization and biological activities of an enteropathogenic toxin of Vibrio parahaemolyticus. Toxicon 13:134-135.