INFECTION AND IMMUNITY, Apr. 1979, p. 90-93 0019-9567/79/04-0090/04$02.00/0

Vol. 24, No. 1

Mouse Protective Capabilities of Escherichia coli Hybrids Expressing Salmonella typhi Antigens B. B. DIENA,J* H. LIOR,' A. RYAN,' P. KROL,' E. M. JOHNSON,2 AND L. S. BARON2 Laboratory Center for Disease Control, Health Protection Branch, Health and Welfare Canada, Ottawa, Ontario KiA OL2, Canada,' and Department of Bacterial Immunology, Walter Reed Army Institute of Research, Washington, D.C. 200122 Received for publication 14 September 1978

An Escherichia coli hybrid, F1061, expressing Salmonella typhi somatic antigens 9 and 12, and a derivative of this hybrid, E. coli hybrid WR3078, expressing the S. typhi Vi antigen in addition to somatic antigens 9 and 12, were compared with S. typhi Ty2 in experiments to test their ability, as live vaccines, to protect Swiss white mice against death from challenge with a mouse-virulent Salmonella typhimurium hybrid expressing the S. typhi antigens 9, 12, Vi, and d. When the live, vaccinating organisms were administered intraperitoneally, 87.5% of the mice immunized with S. typhi Ty2 survived challenge, as compared with 62.5% of those immunized with E. coli hybrid F1061 and 55% of those inoculated with E. coli hybrid WR3078. When live organisms were administered orally at a dose of 109, 67.5% of the mice immunized with S. typhi Ty2 survived challenge as compared with 47.5% of those immunized with E. coli hybrid F1061 and 40% of those administered E. coli hybrid WR3078. Thus, the protection conferred by E. coli hybrid F1061 expressing only the S. typhi somatic antigens, although significant in this system, was inferior to that conferred by S. typhi Ty2 and the addition of the S. typhi Vi antigen to this hybrid (creating E. coli hybrid WR3078) did not enhance that protection.

We previously reported using mouse-virulent Salmonella typhimurium hybrids expressing Salmonella typhi antigens as challenge organisms in vaccinated Swiss-Webster white mice to test the protective abilities of typhoid vaccines (1-4). This assay system has proven capable of demonstrating differences in effectiveness among various kinds of typhoid vaccines with respect to their ability to confer protection against death of the animals (2-4) and has provided evidence that the Salmonella somatic antigens are important in conferring this protection (1). Although our earlier studies employed only nonliving vaccines administered intraperitoneally (i.p.), we have recently extended our use of this system to assay the protective capabilities of both living and nonliving vaccines administered either i.p. or orally (4). Of the various typhoid vaccines that we have examined, live S. typhi administered i.p. is the most effective in protecting the mice against death when challenged with an S. typhimurium hybrid expressing S. typhi antigens (4). However, since our primary concern is with evaluating those vaccines against S. typhi that are suitable for human use, we are particularly interested in examining, in this system, the protective capabilities of presently used as well as potential

human typhoid vaccines. We were, therefore, especially intrigued by the report of Kiefer et al. (7) that mice immunized i.p. with an Escherichia coli hybrid expressing the somatic antigens of S. typhi were significantly protected against death when challenged with an S. typhimurium hybrid expressing S. typhi antigens. Because an E. coli hybrid of this type would seem most promising for use as a safe, live, oral human typhoid vaccine, we were interested in comparing the protection it would confer in our assay system, administered orally as well as i.p., with that conferred by live S. typhi. Also, we wished to see whether the addition of the S. typhi Vi antigen to this E. coli hybrid (by conjugal gene transfer) would enhance that protection. In the present paper we report the results of these studies. MATERIALS AND METHODS Bacterial strains. The derivation and description of the S. typhi Hfr strain, WR4000, used to generate the E. coli hybrids, F1061 and WR3078, have been reported previously (6). The E. coli parent strain, F464 (serotype 08:K-:H-), and E. coli hybrid F1061 were provided by G. Schmidt. E. coli hybrid F1061 is the product of a mating between S. typhi Hfr WR4000 and E. coli F464, as described elsewhere (8); as the conse90

VOL. 24, 1979

MOUSE PROTECTION BY E. COLI-S. TYPHI HYBRIDS

of replacement of the rfb gene cluster of the E. coli recipient with the rfb cluster of the S. typhi donor, E. coli hybrid F1061 expresses the somatic (0) antigens 9 and 12 of S. typhi and lacks E. coli 08 specificity. E. coli hybrid WR3078 was generated by a mating between S. typhi Hfr WR4000 and a melibiose nonutilizing (mel) mutant of E. coli hybrid F1061; it was selected for receipt of the S. typhi mel genes, to which the structural genetic determinants of the Vi antigen (viaB) are closely linked (10), and it expresses the S. typhi Vi antigen in addition to the S. typhi somatic antigens 9 and 12. The derivation of S. typhimurium hybrid H42 has been described previously (1, 2); it expresses the 9, 12, Vi, and d antigens of S. typhi and has a 50% lethal dose of less than 50 organisms. The classical Ty2 strain of S. typhi is, of course, well known. Protection experiments. Organisms were harvested in phosphate-buffered saline, washed three times, appropriately diluted, and mixed 1:1 with a sodium bicarbonate solution (Travenol, 5% injection USP) just prior to inoculation. Swiss-Webster white mice (40 per group) HPB strain, random bred and weighing 16 to 18 g, were administered the live, vaccinating organisms either i.p. or orally. Mice immunized i.p. received a dose of 107 organisms suspended in 0.5 ml of solution; the groups immunized orally received either a dose of 107 organisms or a dose of 109 organisms, suspended in each case in 0.5 ml of solution. Orally administered doses were introduced into the stomach with a bent, 25-gauge, 2-inch needle with its point removed and its end smoothed off. All animals were challenged i.p., 5 weeks after immunization, with 2,500 organisms (0.5 ml) of S. typhimurium hybrid H42, as in our previous live immunization experiments (4). Survivors were counted after 21 days. Test for statistical significance. The results of assays were analyzed as RXC tables by using the usual chi-square criterion (9). quence

91

TABLE 1. Protective capabilities of i.p.administered live vaccines in mice challenged with S. typhimurium hybrid H42a Vaccinated with:

inoculated

% Survival

S. typhi Ty2 87.5 35/40 E. coli hybrid 62.5 25/40 F1061 E. coli hybrid 55.0 22/40 WR3078 E. coli parent F464 42.5 17/40 Control (not vacci2.5 1/40 nated) a The vaccinating dose was 107 organisms, and the animals were challenged i.p. after 5 weeks with 2,500 organisms (0.5 ml) of S. typhimurium hybrid H42. b Protection conferred by all of these strains is significantly better (P < 0.005) than the control.

coli parent strain, F464, from which the hybrids were derived. Although ranking lowest in the protection comparison, it surprisingly conferred significant protection with 42.5% of the i.p.-inoculated mice surviving challenge (Table 1). Nevertheless, in several repeated experiments with the parental and hybrid strains, the 20% difference in survival percentage seen here between E. coli F464 and the hybrid F1061 was

maintained consistently, confining the impor-

tance of the S. typhi somatic antigens in the protection observed in this system. Comparison of protection conferred by living organisms administered orally. Since the potential value of a hybrid such as E. coli F1061 would be its suitability for use in humans RESULTS as an oral vaccine, we were especially interested Comparison of protection conferred by in observing its performance in our assay system living organisms administered i.p. Our pre- when administered by the oral route. We had vious finding (4) that live S. typhi administered reported earlier (4) a 42% survival among mice i.p. conferred, in this assay system, the most orally immunized with 107 live S. typhi Ty2 and effective protection against death of the animals challenged with S. typhimurium hybrid H42. was confirmed. As shown in Table 1, i.p. adminThis figure is almost exactly what we obtained istration of live S. typhi Ty2 resulted in an 87.5% with 107 live S. typhi Ty2 administered orally in survival of the mice challenged with S. typhi- the present study (Table 2). However, at that murium hybrid H42, compared with a 62.5% dose level with live, orally administered F1061, survival among the group immunized i.p. with although protection was still significantly better the live E. coli hybrid F1061. The E. coli hybrid (P < 0.01) than in the control group, only 32.5% WR3078, expressing the Vi antigen in addition of the inoculated mice survived challenge. to its S. typhi 9 and 12 somatic antigens, showed Previously, we did not use oral dose levels of no improvement over E. coli hybrid F1061, with live S. typhi Ty2 higher than 107 because a only 55% of the immunized mice surviving the number of mice died at higher dosages, particuchallenge. The failure of Vi antigen addition to larly at 109. However, since the E. coli hybrids enhance the protection afforded by a hybrid can be used safely at 109, we used this dose of expressing the S. typhi somatic antigens has live S. typhi Ty2 as well to make the comparison. been observed previously in studies using S. As shown in Table 2, oral administration of 109 typhimurium hybrids (acetone killed) as vacci- live S. typhi Ty2 raised the level of protection nating strains (1). (percent survival) to 67.5%, and with an oral We examined also in this experiment the E. dose of 109 live E. coli hybrid F1061 the survival

92

DIENA ET AL.

INFECT. IMMUN.

figure rose to 47.5%. The E. coli hybrid expressing the Vi antigen, WR3078, again showed no improvement over F1061, with only 40% of the orally immunized (109 dose) mice surviving. Once again, at the 109 dose, the E. coli parent strain exhibited a significant (P < 0.01) level of protection, with 32.5% of the mice in that group surviving. DISCUSSION from these experiments that both It is evident oral and i.p. immunization of Swiss-Webster white mice with live E. coli hybrid F1061 confers significant protection against death from challenge with S. typhimurium hybrid H42. It is also evident, however, that the protection conferred by live E. coli F1061 is inferior to that conferred by live S. typhi Ty2, 62.5% versus 87.5% in the i.p.-immunized mice, and 47.5% versus 67.5% in the mice immunized orally. Furthermore, addition of the S. typhi Vi antigen to E. coli hybrid F1061 (creating E. coli hybrid WR3078) did not produce a hybrid with better protective capability than F1061, which expresses only the S. typhi somatic antigens 9 and 12. The failure of Vi antigen addition to enhance the protective capability of F1061 is not only similar to our earlier observations with acetone-killed S. typhimurium hybrids as vaccinating strains (1), but is also consistent with our previous findings with regard to the Vi antigen in that we have yet to discover TABLE 2. Protective capabilities of orally administered live vaccines in mice challenged with S. typhimurium hybrid H42' Dose

Survivors/no.

% Survival

109

27/40b

67.5

17/40b

42.5

109 107

19/40b 13/40c

47.5 32.5

WR3078

109 107

16/40b 12/40c

40.0 30.0

E. coli parent

109

13/40C

32.5

107

8/40

20.0

2/40

5.0

Vaccinated with:

S. typhi Ty2

107 E. coli hybrid F1061

E. coli hybrid

F464

Control (not vaccinated) a

Animals

inoculated

were challenged i.p. after 5 weeks with

2,500 organisms (0.5 ml) of S. typhimurium hybrid H42.

b Protection is significantly better (P < 0.005) than the control. Protection is significantly better (P < 0.01) than the control. c

any role for it in the protection conferred in this system (1-4). Although our comparisons of the protective capabilities of E. coli hybrid F1061 and its parent F464 consistently indicated the importance of the S. typhi somatic antigens expressed by the hybrid, it would appear, nevertheless, that a substantial percentage of the overall ability of F1061 to confer protection in this system derives from the parent strain, F464, itself. We do not have sufficient data on the protective capabilities of different E. coli strains in this system to make valid comparisons, but with at least one other strain tested (E. coli WR3991, expressing O antigen 102, no K antigen), we have not observed the protection levels seen with F464. Unfortunately, we have not been successful in attempts to transfer the S. typhi somatic antigens to WR3991 and thus have no data to indicate how other E. coli hybrids would compare with F1061. However, we would suspect that factors other than the specificity of the S. typhi somatic antigens would affect the protective capabilities of hybrids in this system. Although there are, as yet, no data to establish that comparative protective capabilities of live typhoid vaccines against S. typhimurium hybrid H42 in our system would be reflected similarly in their effectiveness against S. typhi in humans, we are, of course, hopeful that such will eventually prove to be the case. Therefore, it seems reasonable to consider what our present results might possibly imply regarding the potential of E. coli hybrids such as F1061 for use as live, oral, human typhoid vaccines. Initially, our data suggest that an attenuated S. typhi strain, such as the galE mutant Ty2la proposed by Germanier and Furer (5), might be a more effective human immunizing agent than an E. coli hybrid. Frankly, we had hoped that the protective capability exhibited by F1061 would more closely approximate that of S. typhi than was the case in our present experiments. Furthermore, the sizable contribution of its E. coli parent strain, F464, to that protective capability raises the question as to whether a similar contribution could be expected in the human situation. On the other hand, the E. coli hybrid F1061 did confer a significant degree of protection in our system, and it does not have the undesirable potential for producing a systemic infection in vaccinated humans that an S. typhi mutant has, should back-mutation occur. Also, it must be acknowledged that we did not test, in the present experiments, an attenuated S. typhi strain to establish how closely it would approximate, in our system, the protective capabilities of the nonattenuated S. typhi Ty2 strain. In the final

MOUSE PROTECTION BY E. COLI-S. TYPHI HYBRIDS

VOL. 24, 1979

analysis, however, it is only with human subjects in actual field trials that a realistic evaluation of the merits of these vaccines can be accomplished to determine whether the protection conferred by them in our test system reflects their effectiveness as human immunizing agents. LITERATURE CITED

4. Diena, B. B., A. Ryan, R. Wallace, E. M. Johnson, L. S. Baron, and F. E. Ashton. 1977. Effectiveness of parenteral and oral typhoid vaccination in mice chal5.

6.

1. Diena, B. B., L. S. Baron, E. M. Johnson, R. Wallace, and F. E. Ashton. 1974. Role of typhoid antigens in

protection and pathogenicity for mice. Infect. Immun. 9:1102-1104. 2. Diena, B. B., E. M. Johnson, L. S. Baron, R. Wallace, and L Greenberg. 1973. Assay of typhoid vaccines with Salmonella typhosa-Salmonella typhimurium hybrids. Infect. Immun. 7:5-8. 3. Diena, B. B., A. Ryan, R. Wallace, F. E. Ashton, E. M. Johnson, and L S. Baron. 1975. Ineffectiveness of Vi and chemically treated endotoxins as typhoid vaccines in mice challenged with a Salmonella typhosaSalmonella typhimurium hybrid. Infect. Immun. 12: 1470-1471.

93

7.

8. 9.

10.

lenged with a Salnonella typhi-Salmonella typhimurium hybrid. Infect. Immun. 15:997-998. Germanier, R., and E. Furer. 1975. Isolation and characterization of Gal E mutant Ty 21a of Salmonella typhi: a candidate for a live, oral typhoid vaccine. J. Infect. Dis. 131:553-558. Johnson, E. M., S. Falkow, and L. S. Baron. 1964. Chromosome transfer kinetics of Salmonella Hfr strains. J. Bacteriol. 88:395-400. Kiefer, W., P. Gransow, G. Schmidt, and 0. Westphal. 1976. Salmonellosis in mice: immunization experiments with Salmonella-Escherichia coli hybrids. Infect. Immun. 13:1517-1518. Kiefer, W., G. Schmidt, B. Jann, and K. Jann. 1976. Genetic transfer of Salmonella 0 antigens to Escherichia coli 08. J. Gen. Microbiol. 92:311-324. Snedecor, G. W. 1957. Statistical methods, 5th ed., p. 212-236. Iowa State College Press, Ames. Snellings, N. J., E. M. Johnson, and L. S. Baron. 1977. Genetic basis of Vi antigen expression in Salmonella paratyphi C. J. Bacteriol. 131:57-62.

Mouse protective capabilities of Escherichia coli hybrids expressing Salmonella typhi antigens.

INFECTION AND IMMUNITY, Apr. 1979, p. 90-93 0019-9567/79/04-0090/04$02.00/0 Vol. 24, No. 1 Mouse Protective Capabilities of Escherichia coli Hybrids...
584KB Sizes 0 Downloads 0 Views