Adherence of wild-type and mutant strains of Vibrio cholerae to normal and immune intestinal tissue*

Bulletin of the World Health Organization, 57 (1): 123-128 (1979) Adherence of wild-type and mutant strains of Vibrio cholerae to normal and immune intestinal tissue* J. W. BHATTACHARJEEl & BRAHM S. SRIVASTAVA2 Adherence of some wild-type and mutant strains of Vibrio cholerae was studied and the relation between adherence, motility, multiplication, and virulence was examined. Vibrios adhered readily to freshly isolated tissue segments of rabbit intestine at 37°C. The number of vibrios adhering was dependent on the concentration of bacteria to which the tissue was exposed. Adherence was markedly reduced by pretreatment with chloramphenicol and also when intestinal tissue was taken from parenterally or orally immunized rabbits. Non-motile or feebly motile strains were unable to adhere. One motile strain was also found to adhere poorly. The adherence of non-motile or poorly adhering strains was not affected by chloramphenicol. Non-motile, poorly adhering, or slow-growing strains behaved as attenuated strains, suggesting the role of motility, adherence, and multiplication in the virulence of V. cholerae. Vibrio cholerae invades the intestine but the se- quence of events leading to the clinical manifestation of disease is rather poorly understood. Recent studies suggest that there is in V. cholerae a correlation between virulence, motility, and adherence of vibrios to intestinal mucosa (8, 9, 11, 13). Loss of motility was shown to result in a marked reduction in virulence (9) and fluorescent antibody studies and scanning and transmission electron microscopy demonstrated a close association of vibrios with the intestinal villi (5, 11, 13). Freter reported that the firm adherence of vibrios to the intestinal mucosa was inhibited by antibodies (6). Hence a virulent strain, besides elaborating toxin, is thought to be motile, adhere to the intestinal villi, and multiply in the gut. By extrapolation from this information it is inferred that an attenuated vibrio strain for the purpose of oral immunization must also be motile and adhere and multiply well in the intestine in order to induce immunity. The present study was undertaken to investigate this question. Using a variety of virulent vibrio strains, attenuated strains obtained by mutation or recombination, non-motile, and poorly adhering strains, the correla- tion between growth rate, motility, adherence, and virulence was examined. * From the Cholera Immunology Laboratory, Central Drug Research Institute, Lucknow, India. Research Fellow. Scientist. MATERIALS AND METHODS Bacterial strains These are listed in Table 1. Table 1. List of vibrio strains Vibrio Relevant features Source or referencestrains V. cholerae (classical) KB92 wild type K. Bhaskaran KB93 wild type K. Bhaskaran KB95 a rough variant of KB92 K. Bhaskaran 569B a feebly motile strain K. Bhaskaran KB378 a nonmotile strain K. Bhaskaran KB611 a streptomycin-dependent (2) strain KB599 a streptomycin-independent (14) revertant of KB61 1 Cl 4-S5 a dwarf colony mutant (1) CD1 a recombinant derived B. S. Srivastava from KB599 by mating with a virulent strain V. cholerae (biotype eltor) KB207 wild type K. Bhaskaran KB381 a poorly adhering strain new variant 3782 - 123 - 124 J. W. BHATTACHARJEE & B. S. SRIVASTAVA Media and buffer Cultures were grown in Difco brain-heart infusion broth (BHI) and viable counts were made on nutrient broth (Difco) supplemented with 1 % agar (Oxoid). Phosphate-buffered saline (PBS) was used for bacterial dilutions. Adherence of vibrios to rabbit intestine Bacteria grown overnight on nutrient agar slopes at 37°C were suspended in PBS and the required viable counts were diluted in the same buffer. A piece of intestine of an anaesthetized rabbit was removed after ligating the blood vessels. After gently washing in PBS, a 10-mm disc of the intestinal tissue was prepared. The tissue disc was dropped into 5 ml of bacterial suspension and gently shaken at 37°C for 30 min. The disc was then removed, washed twice with 10 ml of PBS and, after drying quickly on sterile filter paper was homogenized in 10 ml of PBS. The viable vibrios were counted in the homogenate. The index of adherence was calculated by dividing the number of vibrios adhering to the disc by the number of vibrios to which the disc had been exposed, and was expressed as a percentage. Since KB611 is a streptomycin dependent strain, media were supplemented with streptomycin (1 mg/ ml) except for the 30 min period during which the vibrios were in contact with the disc. The effect of chloramphenicol on the adherence of vibrios was also examined, the procedure being exactly as described above, except that the vibrios were pretreated for 15 min with 100 ,ug of chlor- amphenicol per ml and that the antibiotic was present when the tissue disc was exposed to the bacterial suspension. This dose of chloramphenicol had no effect on the viable counts, nor on the motility of vibrios as observed in the hanging drops (see below). Motility Motility was examined microscopically in hanging drops. It was observed that the non-motile KB378 when stabbed into sloppy nutrient agar plates (0.5 %) grew only along the stab, whereas the motile strains formed a turbid zone of growth 20-22 mm in dia- meter in 9 h (9). Therefore, bacterial suspensions of identical optical density were taken. A straight inoculation needle was dipped into the suspension and then stabbed into nutrient agar plates. The plates were incubated at 37°C and the diameter of the turbid bacterial growth was measured after 9 h (for fast-growing strains) or 24 h (for slow-growing strains). Growth Growth was assessed by means of counts of viable vibrios growing in BHI. The growth of different strains was compared by comparing the generation times (t = 0.301/K). Virulence Bacterial cultures grown in BHI were serially diluted 10-fold and inoculated into the following animals models of cholera. Infant mice. Five-day-old infant Swiss mice were orally inoculated after starving for 5 hours (4). 0.1 ml of inoculum containing the required number of vibrios was introduced into the stomach using a round-tipped needle attached to a syringe; mortality was recorded up to 48 h. Adult mice. Adult Swiss mice of either sex weighing about 14-16 g were distributed into groups of 10 (1). The mice were injected intraperitoneally with or- ganisms suspended in 5% sterile mucin (pH 7.2). a Virulence of different strains was compared by their LD50 at 48 hours, calculated by the method of Reed & Muench (12). Adherence of vibrios to immune rabbit intestine The adherence was measured as described above except that the tissue discs were made from the intes- tine of immune rabbits. A minimum of four rabbits were included in each immunization experiment with equal numbers of controls and at least four discs were taken from each rabbit to examine bacterial adhe- rence. The rabbits weighing 1 kg were immunized either orally or parenterally. Parenteral immunization with heat-killed virulent KB92. Rabbits were immunized with six 1-ml doses of vaccine over 14 days. The vaccine was prepared by growing KB92 on agar slopes and 109 cells suspended in 1 ml of broth were killed by heating in a boiling water bath for 10 min. Four weeks after the last dose, intestinal tissue discs were prepared from these immunized rabbits and the adherence of V. cholerae was examined. Parenteral immunization with heat-killed or live attenuated CDJ. In contrast to the immunization schedule with KB92, rabbits were given two 1-ml sub- cutaneous doses separated by an interval of 21 days. Each dose contained 109 organisms either live or heat-killed. Ten days after the second dose of vaccine, the intestine was removed to study adherence. a Type 1701-W, purchased from Wilson Laboratories, Chicago, IL, USA. ADHERENCE OF VIBRIOS TO INTESTINAL TISSUE Oral immunization with live CDJ. Four daily doses of 1011 organisms were administered using a stomach tube after preliminary alkalinization of the stomach. Five days after the last dose of vaccine the intestine was taken for study. RESULTS Adherence of V. cholerae as a function of the con- centration of bacteria and time To obtain the optimum conditions for adherence, this property of vibrios was studied as a function of bacterial concentration and time. Keeping the area of the tissue disc constant, an increase in the number of bacteria in the 5 ml of bacterial suspension resulted in an increase in the number of bacteria adhering (Fig. 1). When the index of adherence was calculated a saturation phenomenon was seen, the index being at a maximum when the total number of vibrios was between 105 and 107. Incubation of a tissue disc for 20-30 min with vibrios was sufficient to yield optimum adherence (Fig. 2). Therefore, in the subsequent experiments, intestinal discs were exposed to 107 vibrios for 30 min at 37°C with gentle shaking. a I * NoO VIRISoDHRE w a:~~~~~~~~~~ U) Z 6 30 UJ08 11 theadeec ne o 0,0,ad18clswr 0 4 ~~~~~~~~~~~~~~~~~~~u LL c o Lu I0~~~~~~~~ 0 ADHERENC4 INDEX z * ~*No.OFVIBRIOSADHERED NUMBERS OF VIBRIOS Fig. 1. Adherence of vibrio strain KB207 to intestinal discs as a function of the number of vibrios per 5 ml of suspension. The means and standard deviations of the adherence index for 10Q6, 10Q7, and 108 cells were as follows: 4.2 + 0.5 (3), 2.2 + 0.7 (5) and 1.1 + 0.2 (3)-the values in brackets indicating the number of experiments. a w w 0 0 cn 0 6 z 10 410 I 10. 0 5 10 20 30 TIME (MIN) Fig. 2. Adherence of vibrio strain KB207 to intestinal discs as a function of time. Adherence of the wild-type and mutant strains of V. cholerae Calculation of the adherence index for different strains reveals that a few mutant strains adhered poorly to the intestine, in particular the non-motile or poorly motile strains like KB378, 569B, and C14-S5. KB381, which is a motile strain, also adhered poorly (Table 2). KB207, KB92, KB93, KB95, KB611, KB599, and CD1 appear to. be normally adhering strains. Roughness of a strain had no effect on adherence. Chloramphenicol-sensitive adherence Chloramphenicol treatment significantly reduced the number of KB207 vibrios adhering to intestinal tissue discs, whereas it had no effect on a non-motile strain KB378 (Table 3). The antibiotic had little or no effect on the adherence of KB381, which had been passaged twice through rabbits. Animal passage did not enhance the adherence of KB381. It should be noted that chloramphenicol did not affect the motility of motile strains. 125 J. W. BHATTACHARJEE & B. S. SRIVASTAVA Table 2. Comparative data on adherence, motility, growth, and virulence of some wild-type and mutant vibrios Virulence Vibrio strains index dherenenexof Motility Generation Mortality of infant mice aAndherence erange ofde time (min) _____________ LD50 Of 107 108 adult mice KB207 8.4 5-11 motile 22 10/10 - < 100 KB92 8.3 motile 21 10/10 - < 50 KB93 11.6 motile 21 ND ND ND KB95 11.3 5-13.3 ND ND ND ND ND KB611 9.0 5.5-12 motile 37 ND ND 108 KB599 9.3 9-11.3 motile 37 0/10 0/10 108 C14-S5 0.08 0.05 -0.11 feebly 42 ND ND 107 motile KB381 0.55 0.3 -0.6 motile 21 0/10 3/10 2 x 10s 569B 0.037 0.025-0.05 feebly 21 5/10 10/10 105 motile KB378 0.015 0.008-0.025 non-motile 22 0/10 0/10 2 x 105 CD1 10.0 6-15 motile 22 0/10 ND 106 a Mortality is expressed as the number of dead mice over the total number of mice, for mice fed 107 or 108 vibrios: ND = not done. Table 3. Chloramphenicol sensitivity of adherence of V. cholerae to intestinal discs a Adherence index Vibrio strain Control ChloramphenicolControl ~~treated KB207 7.4 0.072 KB378 0.022 0.017 KB381 0.55 0.51 a Adherence in the presence of chloramphenicol was examined in duplicate in 3 experiments. The data represent the results from one such reproducible experiment. Adherence of vibrios to immunized intestine KB207 and CD1, both normally-adhering strains, were used as challenge strains. The index of adherence was significantly reduced in discs of intestine from immunized rabbits compared with the controls (Table 4). With rabbits immunized parenterally with live or dead CD1, similar results were obtained. The adhe- rence index was 0.52. It is assumed that washing the intestine from immunized animals removed only the faecal material and that the mucus layer and any antibodies present on its surface are unlikely to have been disturbed. Table 4. Adherence of V. cholerae to discs from immunized rabbits Rabbits VjChallange Adherence Rahnge ofRabbits vibrio strain index indhexec Control group KB207 7.0 4.1-10.5 CD1 8.2 6.0-11.0 Group immunized par- KB207 0.09 0.015-0.25 enterally with V. cholerae (classical) CD1 0.56 0.49-0.9 Group immunized orally with live CD1 KB207 0.5 0.09-1.4 To rule out any possibility that the intestinal tissue homogenate itself had agglutinating activity in these experiments, vibrios were incubated with tissue homogenate for 30-60 min. There was no decrease in the viable counts in comparison with controls held in PBS. Motility The motility of the various wild-type and mutant strains was compared (Table 2). When examined microscopically in a hanging drop, all except 569B, C14-S5, and KB378 exhibited motility typical of V. cholerae. KB378 was non-motile whereas C14-S5 and 569B were feebly motile. 126 ADHERENCE OF VIBRIOS TO INTESTINAL TISSUE The behaviour of these strains was studied in soft agar stabs. All motile strains except KB611 and KB599 gave 22-24 mm zones of growth at 9 hours. KB611 and KB599 gave zones of 7-10 mm at 24 hours which may be attributed to their poor rates of growth. The non-motile strain KB378 did not spread into the agar. Feebly motile strains 569B and C14-S5 gave zones of growth of 2-4 mm at 24 hours. It may be emphasized that strain 569B is a fast-growing strain. Multiplication and growth Growth of the vibrio strains was measured in vitro and the generation times were compared (Table 2). Multiplication of some of the vibrio strains was measured in animal models, either closed ileal loops of adult rabbits or orally fed suckling mice. The data showed that the fast-growing strains multiplied in the animal models, the titre increasing by 1-2 log factors, whereas the slow- growing strains did not increase in number (1, 14). Virulence of V. cholerae in relation to motility, adherence, and multiplication The virulent strains KB207 and KB92, as judged by the behaviour in animals models, were normal with respect to their growth, motility, and adherence to intestine. The strains showing attenuation of virulence (except CD1) were deficient in one or more of these three characteristics. KB378 and 569B grew as fast as the virulent strains but were non-motile or feebly motile and adhered poorly. KB381 grew rapidly and was motile but adhered poorly. KB611 and KB599 adhered normally and were motile but grew very slowly. C14-S5 was poor in all respects. CD1 is. a recombinant isolated in this laboratory as a potential oral immunizing agent and was found to grow well, and was motile and adhered normally. DISCUSSION Several virulent and attenuated strains of V. cholerae adhere to intestinal tissue, whereas some mutant strains do not. Earlier work suggested a relationship between motility and adherence and we confirmed that adherence of vibrios was impaired in the non-motile strains we observed. One motile strain (KB381) was also found to be poorly adhering. Adherence of KB207 was sensitive to chlor- amphenicol, which is an inhibitor of protein synthesis; however, the index of KB381 was only slightly affected and that of a non-motile strain did not fall further. It should be noted that chloramphe- nicol had no effect on the motility of a motile strain. The drug sensitive nature of adherence suggests that adherence of vibrios required a de novo synthesis of an essential protein (7, 10) and that the sysnthesis or activity of this protein is impaired in BK381. Adherence of the vibrios to intestine appears to be an important step in pathogenesis, an idea that is supported by the observation that adherence of V. cholerae to immunized intestine is significantly reduced. A high level of bacterial adherence to the intestine might be one of the factors promoting colonization and delivery of the enterotoxin to the correct site. Failure of the vibrios to adhere may be one of the significant components in successful immunity to cholera. The motility and rate of growth have also been recognized as factors contributing to virulence of V. cholerae (9). Strains growing slowly, adhering poorly, or which are non-motile or feebly motile tend to behave as attenuated strains. There is no doubt that enterotoxin production is essential for the secretion of fluid but this is affected by several virulence factors which appear to exert their influence independently of each other. All the virulent strains of V. cholerae used in this investigation were pro- ficient in motility, adherence, and multiplication. The foregoing discussion suggests that adherence, together with other factors, may be a desirable feature in any attenuated strain considered as a potential immunizing agent. C14-S5 was found to lack these desirable features and indeed this strain did not give promising results in preliminary human trials (3). Further studies would be desirable to determine whether adherence is a useful parameter for assessing the potential activity of a particular immunizing agent. Increasing the concentration of vibrios exposed to a fixed area of intestine resulted in an increase in the number of vibrios adhering. However, the index of adherence declined above a certain level. This may have been due to the saturation of specific receptor sites, or to the fact that a large proportion of vibrios undergo non-specific adherence, which is probably irrelevant, and are eluted gradually. Jones et al. (10) observed 90% elution of adhered vibrios. Adherence may not be random attachment of V. cholerae to intestine; it has recently been shown that extracts of gut mucosa can interfere with the adherence of vibrios to the intestine (8) and that intestinal antitoxins prevented the binding of cholera toxin to intestinal receptors (15). 127 128 J. W. BHATTACHARJEE & B. S. SRIVASTAVA ACKNOWLEDGEMENTS We are grateful to Dr Nitya Nand, Director of the Institute, for his interest and support. Thanks are due to Mr A. K. Sarkar and Mr M. U. Khan for their generous technical and laboratory help. One of us (J.W.B.) was a Research Fellow of the Indian Council of Scientific and Industrial Research when this work was done. This is communication No. 2299 of this Institute. RtSUMt ADHARENCE DE SOUCHES SAUVAGES ET MUTANTES DE VIBRIO CHOLERAE AU TISSU INTESTINAL PRtLEVE SUR UN ANIMAL IMMUNISE OU NON L'adh6rence de Vibrio cholerae a l'intestin du lapin a e 6tudi6e en vue de rechercher une eventuelle correlation entre cette adherence, la motilite des bacteries et leur rapidit6 de croissance, d'une part, et la virulence d'autre part. Divers types de souches bacteriennes - sauvage, rugueuse (< rough #), de motilite nulle, de faible motilite, a mauvaise adh6rence, streptomycinod6pendante, a croissance lente, att6nuee et recombin6e (tableau 1)- ont ete mises en pr6sence de segments de tissu fraichement pr6leves, a 37°C. La plupart des souches de vibrions ont adhere rapidement au tissu, et le maximum d'adherence a et6 observe lorsque le tissu etait expose pendant 30 minutes aux bacteries contenues dans 5 ml de suspension a une concentration se situant entre 105 et 107. Les souches de motilit6 nulle et de faible motilit6 se sont revelees incapables d'adherence, mais l'une des souches a motilite normale (KB381) n'a elle-meme manifest6 qu'une capacite d'adh6rence limitee. Dans ce dernier cas, la capacite d'adherence n'a d'ailleurs pas ete modifi6e par le traitement des vibrions au chloramphenicol, alors que ce traitement a notablement reduit l'adherence des vibrions d'une souche presentant normalement un indice d'adhe- rence 6leve (KB207). L'antibiotique n'a pas eu d'effet sur une souche presentant une motilite nulle (KB378) (tableau 3). L'effet du chloramphenicol (qui est un inhibiteur de la synthese des prot6ines) sur I'adherence amene a penser que celle-ci exige la synthese de novo d'une proteine essentielle. Si les bacteries sont en presence d'intestin de lapin immunise, leur adherence est r6duite de maniere significative, ce qui suggere que l'adh6rence constitue un element important dans la pathogenese. On n'a pas constate d'effet sur l'adh6rence de la rapidit6 de croissance des bacteries, dont une souche streptomyci- nodependante (KB61 1) a manifeste une aussi bonne adherence que des vibrions de souches sauvages (KB 207, KB92 et 93) (tableau 2). D'une maniere generale, les souches de motilite nulle, a mauvaise adherence ou a croissance lente se sont comportees comme des souches attenu6es, alors que les souches virulentes presentaient des caracteristiques normales sur ces trois plans. Ceci permet d'avancer que I'adherence, jointe A des facteurs comme la motilit6 et une vitesse de croissance normale, pourrait etre consideree comme un caractere souhaitable pour toute souche att6nuee dont 1'emploi pour l'immu- nisation orale serait envisage. REFERENCES 1. BHASKARAN, K. & SINHA, V. B. Journal of hygiene, 65: 135-148 (1967). 2. BHATrACHARJEE, J. W. & SINHA, V. B. Indian journal of medical research, 65: 29-34 (1977). 3. CASH, R. A. ET AL. Infection and immunity, 19: 763-764 (1974). 4. CHAICUMMP, W. & ROWLEY, D. Journal of infectious diseases, 125: 480-485 (1972). 5. LABREC, E. H. ET AL. In: Proceedings of the Cholera Research Symposium, 24-29 January 1965. Honolulu, Hawaii, Washington, DC, US Government Printing Office, 1965, p. 272 (US Public Health Service Publication No. 1328). 6. FRETER, R. Texas reports on biology and medicine, 27 (Suppl. 1): 299-316 (1969). 7. FRETER, R. Infection and immunity, 6: 134-141 (1972). 8. FRETER, R. & JONES, G. W. Infection and immunity, 14: 247-256 (1976). 9. GUENTZEL, M. N., & BERRY, L. J. Infection and immunity, 11: 890-897 (1975). 10. JoNES, G. W. ET AL. Infection and immunity, 14: 232-239 (1976). 11. NELSON, E. T. ET AL. Infection and immunity, 14: 527-547 (1976). 12. REED, L. J. & MUENCH, H. American journal of hygiene, 27: 493-497 (1938). 13. SCHRANK, G. D. & VERWEY, W. F. Infection and immunity, 13: 195-203 (1976). 14. SINHA, V. B. & BHATrACHARJEE, J. W. Bulletin of the World Health Organization, 53: 482-484 (1976). 15. Wu, A. L. & WALKER, W. A. Infection and infection, 14: 1034-1042 (1976).