Microbial Pathogenesis 1991 ; 10 : 117-126
Comparison of the human immune response to live oral, killed oral or killed parenteral Salmonella typhi TY21 A vaccines* A . Kantele,' t H . Arvilommi, 2 J . M . Kantele,' L . Rintala' and P . H . Makela'
'National Public Health Institute, SF-00300 Helsinki, Finland and 2 National Public Health Institute, Turku, Finland (Received August 23, 1990 ; accepted in revised form October 16, 1990)
Kantele, A . (National Public Health Institute, SF-00300 Helsinki, Finland), H . Arvilommi, J . M . Kantele, L . Rintala and P . H . Makela . Comparison of the human immune response to live oral, killed oral or killed parenteral Salmonella typhi TY21 A vaccines . Microbial Pathogenesis 1991 ; 10 : 117-126 . The live oral typhoid vaccine Ty21 a has proved to confer protection against the disease at least as effectively as killed parenteral vaccines, whereas killed oral vaccines have not been protective in field trials . This prompted us to compare the immune response of subjects vaccinated either with live oral, killed oral or killed parenteral Salmonella typhi Ty21a vaccine . The immune responses were studied by analysis of peripheral blood antibody-secreting cells (ASC), believed to reflect the mucosal immune response . Live and killed bacteria administered by the oral route elicited immune responses of similar specificity and Ig class profile (IgA dominating), but the response to the live vaccine was significantly stronger and lasted longer . The administration route, on the other hand, influenced the antigenic specificity of the ASC response suggesting different processing of the antigen by the systemic and local immune systems . Thus, the response after oral vaccination was almost exclusively directed to the surface 0-antigen, whereas after parenteral vaccination an equally strong response was seen to the O-antigen, to lipopolysaccharide core and to flagella .
Key words : antibody-secreting cell ; typhoid fever; Salmonella typhi Ty2l a ; oral vaccination ; ELISPOT; secretory immunity .
Introduction We encounter most agents of infectious diseases at or through the large surface area of mucosal membranes and, therefore, protective immunity at these portals of entry would seem highly desirable . An oral vaccine is believed to stimulate the mucosal immune system more efficiently than a parenteral vaccine,' and live oral vaccines are believed to be more efficient than killed ones . 2 Thus, an oral live typhoid vaccine has proved to confer significant protection against the disease, 3-$ whereas oral killed typhoid vaccines have showed only little if any protective effect . 9-13 Specific antibodies have been demonstrated in secretions of orally vaccinated volunteers . 14-16 Specific, predominantly IgA antibody-secreting cells, believed to give information on the mucosal immune response, have been demonstrated in the peripheral blood after oral
'The work was carried out at NPHI, Helsinki and Jyvaskyla, Finland . tAuthor to whom correspondence should be addressed at : NPHI, Mannerheimintie 166, SF-00300 Helsinki, Finland . 0882-4010/91/0201171 10 $03 .00/0
~ci
1991 Academic Press Limited
118
A . Kantele
et a1
vaccination ." 21 These cells are believed to originate in the mucosal immune system : after antigen stimulation in Peyer's patches, activated antigen-specific cells migrate to the local lymph nodes and return from there via the lymphatics and blood to the .22 25 mucosal surfaces for antibody secretion We have recently confirmed the use of the peripheral blood antibody secreting cell (ASC) assay as a measure of the immunogenicity of an oral vaccine ." Circulating ASC have been demonstrated in humans after oral vaccination with both live 16,18 .19 5,17 microorganisms . Circulating ASC are also found after parenteral vacciand killed' 26-29 but comparison of the ASC responses after oral and parenteral immunization nation, suggests qualitative differences between the routes . However, these studies have been carried out using different antigens and somewhat different methods of investigation . To our knowledge, until now, no studies have been carried out that compare the immune responses in humans to the same antigen given both parenterally and orally as a live or a killed vaccine . Specifically, our aim was to discover how the quality of the vaccine (live or killed) and the route of its administration (oral or parenteral) each influenced the immune response, assessed in terms of the magnitude, the kinetics, the Ig class distribution and the specificity of the ASC response .
Results Comparison of the responses to live and killed oral vaccines An ASC response was found in all of the 20 volunteers who received the live oral vaccine (group LO), and in 12 of the 14 volunteers who received the killed oral vaccine (group KO) (Table 1, Fig . 1) . The specificity of the ASC responses was similar in both orally vaccinated groups : the responses were almost exclusively directed to the O-antigen (Fig . 2), in accordance with findings in previous studies ." In both groups, specific ASC were observed in the peripheral blood only after vaccination, showing a peak mostly on day 7 with a subsequent fall [Fig . 1 (a)-(b)] . A difference between the groups was found in the duration of the response : on day 14, specific ASC were still seen in 8 (62%) of the 13 volunteers studied in group LO (live vaccine) [Fig . 1 (a)] compared with 2 (14%) of the 14 volunteers in group KO
Table 1 The peak of the 0-9,12 specific ASC response in volunteers who received S. typhi Ty21 a either as a live oral (LO), killed oral (KO) or killed parenteral (KP) vaccine LO
KO
KP
20 20
14 12
8 8
178 137
24 58
18 28
Number of responders with 0-10 ASC/10 6 cells 11-100 ASC/106 cells 101-1000 ASC/10 6 cells 1001ASC/106 cells
2 3 13 2
4 7 3 0
3 4 1 0
Dominating Ig-classb IgA IgG IgM
61% 11% 28%
No . of individuals vaccinated No . of responders Mean number of ASC/106 cells' SEM
80% 0% 20%
20% 20% 60%
'Geometric mean of the sum of IgA-, IgG- and IgM-ASC/106 cells . ' Counted only of responses exceeding 10 ASC/10 6 cells.
119
Response to live/killed, oral/parenteral vaccine
T
4 T
10
12
10
14
12
14
T T Days after vaccination
Fig . 1 . 0-9,12 specific antibody-secreting cells (ASC) after oral administration of Salmonella typhi Ty2l a either as (a) a live (LO) or (b) a killed (KO) vaccine . Each curve represents the sum of IgA-, IgG- and IgMASC of one individual . The arrows indicate vaccine doses given .
1000
U V d
10
KO KP 0-4,5,12
LO KP H-d
Fig . 2 . The magnitude of the ASC responses (the sum of IgA-, IgG- and IgM-ASC) to the different antigens in the three vaccination groups (LO, KO, KP) calculated for all vaccinees in each group . Note that only IgA-ASC were determined for the LPS core in group LO (in other, similar experiments IgA-ASC has accounted for 55% of the total ASC), and only four subjects in group KP were studied for flagellar d antigen (H-d) .
120
A . Kantele et al .
Fig . 3 . 0-9,12 specific ASC in four volunteers after administration of two doses of killed parenteral
Salmonella typhi Ty21 a vaccine with one month interval (arrows) . Each curve is the sum of IgA-, IgG- and IgM-ASC of one individual labelled so as to make it possible to identify each response throughout the study .
(killed vaccine) [Fig . 1(b)] . The geometric mean of the peaks of the 0-9,12 specific responses was also higher in group LO than in group KO (Table 1, Fig . 2) ; this difference was statistically significant (P < 0 .001) . High responses (peak of more than 100 specific ASC) were seen in both groups, but there were several low responders and two non-responders in group KO (Table 1) . The immunoglobulin class distribution of the ASC responses was determined in those volunteers in whom the peak of the response exceeded 10 ASC/10 6 cells . Groups LO and KO proved similar : 61 and 80% of the responses, respectively, were IgA-dominated, and 28 and 20%, respectively, IgM-dominated (Table 1 ) .
Comparison of the responses to oral and parenteral vaccines The kinetics of the ASC response were similar after oral and parenteral administration of the vaccine, except that on day 14 no ASC were seen in any of the volunteers receiving killed parenteral vaccine (group KP), whereas after oral vaccination the response still persisted in some of the volunteers (see above) . Instead of the IgA-dominated response seen after oral vaccination, the response after parenteral vaccination was most often dominated by the IgM class ; however, this conclusion is based on only five responses exceeding 10 ASC/106 cells (Table 1 ) . Four of the volunteers were given two doses of the killed parenteral vaccine . The kinetics (Fig . 3), the magnitude, the Ig class distribution and the specificity of the responses were essentially similar after both of these doses in each individual . Specificity of the ASC responses was different after parenteral than after oral vaccination (Fig . 2) . All three vaccination groups responded to the 0-9,12 antigen, and also, although with somewhat lower numbers of ASC, to the 0-4,5,12 antigen (determined in groups KO and KP only) . However, only group KP, Le . those who had
Response to live/killed, oral/parenteral vaccine
1 21
been vaccinated parenterally, responded noticeably to the LPS core or to the flagella ; actually these responses in group KP were at least as high as that to the 0-9,12 antigen . No specific ASC were seen to the unrelated control antigen Haemophilus influenzae in any of the groups . The numbers of all ISC stayed essentially constant in all the volunteers in all groups throughout the study (data not shown), in accordance with previous findings after oral immunization . 3o
Discussion and conclusions The data reported here show differences in the magnitude, the duration, the Ig class distribution and the specificity of the specific ASC responses in subjects vaccinated either with live oral, killed oral or killed parenteral Salmonella typhi Ty21 a vaccine . We wanted to see the influence of the properties of the living organism and the administration route on the immune response . To exclude other sources of variation, we used the same antigen for all these vaccinations ; killed parenteral and killed oral vaccines were in fact the same killed vaccine preparation and the bacterial strain for it was obtained by culturing from the live vaccine preparation . It was more difficult to achieve a similar dose for all groups . It was obvious that in oral administration a dose of the live vaccine should contain lower numbers of bacteria than the killed vaccine, because the live vaccine strain is expected to multiply in the intestine . On the other hand, to induce an immune response, considerably higher numbers of bacteria are required for an oral than a parenteral administration of a vaccine; giving such high numbers of bacteria parenterally would cause fever and other adverse reactions . The live oral vaccine was the one used in several field trials in a dose at least 10 9 bacteria given on each of days 0, 2 and 4 . 3-6 The killed oral vaccine was given as three doses of 400 x 109 bacteria each ; this was the highest number of killed bacteria reported to have been used in field trials in India ." In killed parenteral vaccination the dose used (0 .5 x 10 9 bacteria) was the one currently recommended for parenteral typhoid vaccines .
Interestingly, the magnitude of the specific ASC response was significantly higher in group LO than in group KO, although the number of bacteria given per dose was lower in group LO than in group KO . Live S . typhi Ty2l a multiply in the intestine, and hence after an oral vaccination, the number of bacteria will increase . However, we did not expect this multiplication to be extensive, partly because of the slow multiplication rate typical of the Ty21 a mutant, and partly because of the limited nature of this multiplication in the intestine as described by Germanier and Fiarer . 31 Indeed, in volunteer studies, the vaccine strain could not be isolated from the vaccinated volunteer's faeces 1 or 2 days after vaccination .32, 33 Hence, the final number of bacteria used for vaccination in these groups may have been rather similar . Therefore, the significantly higher response in group LO as compared to group KO is likely to indicate a special advantage of the living bacteria in stimulating the immune system ; this could be due to their capacity to invade the mucosal epithelium and/or to persist longer in the mucosa or in the Peyer's patches . The lower responses in group KO compared with group LO parallel the results obtained for protection in field trials : only low or negligible protection has been found after vaccination with a killed oral typhoid vaccine, 9-13 whereas the live oral vaccine confers significant protection .` Some of the volunteers in group KO had high responses ; it is possible that the KO vaccine confers significant protection to a small proportion of vaccinated individuals, while the immune response in the majority remains inadequate, and therefore no or only low protection is found in field trials . Hence, the results show that even if the killed oral vaccine is significantly less effective in stimulating the immune system than a live vaccine, it is not ineffective . A
122
A . Kantele et a/
killed oral cholera vaccine has been shown to confer significant protection against cholera ." ," It might be that the difference between live and killed oral vaccines is more pronounced with Salmonella than with Vibrio cholerae, because the former are invasive, whereas the latter is not . Further, in contrast to cholera, cell-mediated immune response is believed to be essential for protection against typhoid fever .
36
The kinetics of the ASC responses were similar in the three vaccine groups, and were also similar to our earlier findings after oral live vaccine ." 19 A few volunteers
were also assessed after a second dose of the killed parenteral vaccine given 1 month after the first dose as recommended for parenteral typhoid vaccination ; surprisingly, the ASC response was qualitatively and quantitatively similar to those seen in the same individuals after the first dose . The response to the oral vaccines seemed to persist somewhat longer than that to the parenteral vaccine, possibly as a result of the prolonged exposure to the antigen, since the oral vaccines were given as three doses on days 0, 2 and 4 . Prolonged intestinal exposure to antigen has, indeed, been shown to prolong the peripheral blood ASC response . 37 Also the difference between groups
KO and LO (14% versus 63% responding on day 14) might, at least partly, be explained by a difference in the duration of antigen exposure : the live bacteria probably persist longer in the tissues than the killed bacteria . We were also interested in comparing the immunoglobulin class distribution of the ASC response in the three vaccination groups . Previously, IgA-dominated responses have been considered typical of oral vaccination, 17 whereas IgG-dominated responses have been reported after parenteral vaccination e .g . with tetanus 26 and cholera
toxoids ." However, parenterally administered pneumococcal polysaccharide vaccine has elicited an IgA-dominated response . 28 Our recent study (to be published)
with parenteral Hib-conjugate vaccine shows an IgA-dominated response to the polysaccharide part of the vaccine (polyribosylphosphate) and a simultaneous IgGdominated response to the protein part of the vaccine (diphtheria toxoid), which implies that the nature of the antigen influences the Ig class distribution of an ASC response . The present study, on the other hand, shows that also the administration route of the vaccine influences the Ig class distribution : the same antigen (0-9,12) elicited an IgA-dominated response when given orally, but an IgM-dominated response when given parenterally . An unexpected finding in the present study was the difference in the antigenic specificity of the ASC response after parenteral and oral vaccination . The response to oral vaccination was mainly directed to the 0 antigen, whereas responses to the LPS core and to the flagella were either absent or very low, as demonstrated also in our earlier studies ." Interestingly, killed parenteral vaccination with the same antigen not only elicited an 0-specific ASC response, but concomitantly also an equal response to the LPS core and to the flagella . As far as we know, this difference in the antigenic specificity after oral and parenteral vaccination has not been reported before . We can only speculate about the reasons for this difference . In intact bacteria, the 0 antigen is exposed on the bacterial surface and is therefore ready to react with antibodies e .g . on the surface of B cells . By contrast, the LPS core is not exposed, and would require processing of the antigen before presentation ; this processing might be more efficient after systemic than after oral immunization . The difference might also arise because vaccine bacteria have different fates in the two cases : after oral vaccination, whole bacteria would gain access from the gut lumen via the M cells directly to the B cells in Peyer's patches, whereas after parenteral vaccination the antigen would, in most cases, be ingested by the macrophages at the site of injection, and B cells would meet it only after digestion by macrophages . It seems that after oral immunization the response is almost exclusively directed to the surface of the bacteria ; such antibodies
Response to live/killed, oral/parenteral vaccine
123
would, in fact, be immediately useful in the gut . The response to the flagellar antigen, which is also an exposed structure, was low or absent after oral vaccination . One explanation for this might be that the protein would be rapidly destroyed in the intestinal milieu ; however, there is no direct evidence for this . The good immunogenicity of the 0 antigen is consistent with this hypothesis, since the LPS is very resistant to digestive enzymes . The low or absent response to flagellar antigen after oral administration of the vaccine should be confirmed by further studies, because of its potentially important consequences in the use of oral vaccines introducing foreign antigens inserted in the Salmonella flagellin ." In conclusion, the present study shows that the use of live instead of killed bacteria in an oral vaccine significantly enhances its immunogenicity : the ASC response after live oral vaccine was more vigorous and lasted longer . The antigenic specificity of the ASC response is strongly influenced by the route of administration : after oral vaccination the response was almost exclusively directed to the surface 0 antigen of the bacteria, whereas after parenteral vaccination an equally strong response was found to the 0 antigen, LPS core and flagella, suggesting different processing of the antigen by the systemic and local immune systems .
Materials and methods Volunteers and vaccination . Forty-two healthy volunteers (28 women and 14 men, aged 2045 years) consented to participate in the study . None of them had a history of or contact with individuals having typhoid fever, nor had they been vaccinated against typhoid fever . Twenty subjects were vaccinated with a live oral, 14 with a killed oral and eight with a killed parenteral typhoid vaccine . Informed consent was obtained from each volunteer before the study . The study protocol was approved by the ethical committee of this Institute . Data on group LO have been reported in an earlier study ." All the vaccines were prepared of Salmonella typhi Ty2l a . The oral live vaccine was a commercially available oral S . typhi Ty2l a vaccine (Vivotif Berna, Swiss Serum and Vaccine Institute, Bern, Switzerland) in suspension formulation ; it contained at least 10 9 bacteria/dose, each lyophilized dose was suspended immediately before ingestion in the buffer provided by the manufacturer . The killed parenteral and oral vaccines were prepared at the Vaccine Laboratory of this Institute according to its routine methods, in brief : the strain Salmonella typhi Ty2la was cultured from the oral live vaccine and grown overnight in Brain Heart Infusion broth . The bacteria were killed by incubating the bacterial suspension for 45 min in a waterbath at 56°C and cooled ; phenol was added to a concentration of 0 .5% before overnight incubation at 21'C . Then, the suspension was washed once with saline (0 .9% NaCI) and suspended in a vaccine buffer (0 .05 M glycine, 0 .01 M succinate, 0 .085 M NaCl, pH 6 .7-7 .0, with 0 .3% phenol) . From this stock suspension, the oral preparation was made to contain 400x10 9 bacteria/dose . Each oral dose was dialysed against the vaccine buffer overnight before ingestion . The dialysed vaccine was tested for the absence of any pathogenic bacteria, and for the presence of other bacteria (less than 2000/dose) and molds (less than 20/dose) . From the same stock suspension, the parenteral vaccine was diluted to contain 10 9 bacterial/ml and phenol added to a final concentration of 0 .3% . The quality of the preparations was controlled at each step according to routine methods . Toxicity tests were carried out in guinea pigs and mice, and the antigenic nature tested by agglutination with specific antisera . To minimize the effect of gastric acid, a capsule containing 0 .8 g of sodium bicarbonate was to be ingested 2 min before the killed oral vaccine preparation . The oral vaccines were given in three doses at two day intervals . The total number of bacteria given orally was at least 3x10 9 (live, group LO) or 1 .2x 1012 (killed, group KO) . The killed parenteral vaccine was given intramuscularly in the left arm as one dose of 0 .5x10 9 bacteria (group KP) . Four subjects in group KP also received a second dose of the same vaccine one month later . Study design. Day 0 was the first vaccination day . Samples of heparinized venous blood were collected on days 0, 5, 7, 9 and 14 in all groups, and for those receiving two doses of killed parenteral vaccine, additional samples were collected on the corresponding days in
124
A . Kantele eta[
relation to the second vaccine dose . Responses to vaccination were followed by quantitation of specific ASC and total immunoglobulin-secreting cells (ISC) among peripheral blood lymphocytes in each blood sample . Isolation of lymphocytes. Mononuclear cells containing mainly lymphocytes were obtained by Ficoll-Paque (Pharmacia, Uppsala, Sweden) centrifugation of heparinized venous blood . Isolated cells washed three times in Hank's balanced salt solution (Flow Laboratories, Irvine, U .K .), were suspended in culture medium, and adjusted to the concentration of 2x10' cells/ml as described previously ." Assay of specific antibody-secreting cells . The number of specific ASC among newly isolated blood lymphocytes was determined by the ELISPOT-method . The isolated cells were allowed to secrete antibodies in antigen-coated 96-well microtiter plates (Immunoplate I, Nunc, Roskilde, Denmark) . Thereafter, the specific antibodies bound to the antigen were detected by application of enzyme-labelled antisera followed by substrate-agarose overlay ; the latter immobilizes the decaying substrate and allows the area of antibodies to be visualized as a coloured spot . Hence, each spot represents one cell secreting antibodies to the coating antigen . Coating was performed with three whole cell bacterial antigens, selected to share with S . typhi its total 0-antigen (0-9,12), only the 0-antigenic factor 12 (0-4,5,12), the lipopolysaccharide (LPS) core or the flagellar d-antigen . Haemophilus influenzae was used as an unrelated control antigen . All these antigens have been described by Kantele 19 except the S. typhimurium 04,5,12 strain ." The coating antigens were incubated at a concentration of 10 8 bacteria/ml PBS overnight at 20°C or for three hours at 37°C . After coating, the plates were masked with 1 bovine serum albumin in PBS for 30 min at 37°C . Next, 10 5 isolated mononuclear cells in culture medium were incubated in each well . Then, alkaline phosphatase-conjugated anti-IgA, G or M (Orion Diagnostica, Helsinki, Finland) diluted 1 :100 in 1 %BSA-PBS was allowed to react overnight at 4°C . The substrate (5-bromo-4-chloro-3-indolyl phosphate; Sigma) was applied in hot (50°C) agarose as described by Sedgwick and Holt ." Spots were counted under low magnification after storage at 4°C . Assay of a// immunoglobulin-secreting cells (c/asses IgA, IgG and IgM) . All immunoglobulin secreting cells (ISC) were quantitated with ELISPOT as described for specific ASC above with the following differences : coating was performed with human IgA, IgG or IgM-specific antisera, 19 the cells were added at a concentration of 5 x 104 cells/well and the second antibody used in the total IgG-ASC assay was as described by Kantele . 19 Statistical methods . The geometric means of the peak responses were calculated for all vaccinated subjects within each group (the peak value for each non-responder was assigned the value of 1) . Statistical comparison was carried out by Student's t-test using log transformed data .
We thank Mrs Anja Ratilainen for careful assistance in the laboratory work . This study was supported by the Emil Aaltonen Foundation, the Finnish found for the replacement of animal experimentation, the Finnish Medical Foundation and partly by the Diarrhoeal Diseases Control Programme, WHO .
References 1 . Mestecky J . The common mucosal immune system and current strategies for induction of immune responses in external secretions . J Clin Immunol 1987; 7 : 265-79 . 2 . Dougan G, Maskell D, O'Callaghan D, Chatfield S, Charles I, Hormaeche C . Oral vaccination . Antonie van Leeuwenhoek 1988 ; 54 : 447-51 . 3 . Wahdan MH, Serie C, Cerisier Y, Sallam S, Germanier R . A controlled field trial of live Salmonella typhi strain Ty2l a oral vaccine against typhoid : three year results . J Infect Dis 1982 ; 145: 292-5 . 4 . Levine MM, Black RE, Ferreccio C et al. The efficacy of attenuated Salmonella typhi oral vaccine strain Ty2l a evaluated in controlled field trials . In : Holmgren J, Lindberg A, Mollby R, eds . Development of vaccines and drugs against diarrhea . Lund : Studentlitteratur, 1986 : 90-101 . 5 . Ferreccio C, Levine MM, Rodriguez H . Contreras R, Chilean Typhoid Committee. Comparative efficacy
Response to live/killed, oral/parenteral vaccine
125
of two, three or four doses of TY21 a live oral typhoid vaccine in enteric-coated capsules : a field trial in an endemic area . J Infect Dis 1989 ; 159 : 766-9 . 6 . Levine MM, Ferreccio C, Black R, Tacket CO, Germanier R, Chilean Typhoid Committee . Progress in vaccines against typhoid fever . Rev Infect Dis 1989 ; 11 : S552-67 . 7 . Black RE, Levine MM, Ferreccio C et al. Efficacy of one or two doses of Ty21a Salmonella typhi vaccine in enteric-coated capsules in a controlled field trial . Vaccine 1990; 8: 81-4 . 8 . Black RE, Levine MM, Clements ML, Young CR, Svennerholm A-M, Holmgren J . Protective efficacy in humans of killed whole-vibrio oral cholera vaccine with and without the B subunit of cholera toxin . Infect Immun 1987 ; 55 :1116-20 . 9 . DuPont LH, Hornick RB, Snyder MJ, Dawkins AT, Heiner GG, Woodward TE . Studies of immunity in typhoid fever . Protection induced by killed oral antigens or by primary infection . Bull WHO 1971 ; 44 : 667-72 . 10 . Borgono JM, Corey G, Engelhardt H . Field trials with killed oral typhoid vaccines . Devel Biol Stand 1976 ; 33 : 80-4 . 11 . Chuttani CS, Prakash K, Gupta P, Groven V, Kumar A . Controlled field trial of a high dose oral killed typhoid vaccine in India . Bull WHO 1977; 55 : 643-4 . 12 . Chuttani CS, Prakash K, Vergese A et al. Ineffectiveness of an oral killed typhoid vaccine in a field trial . Bull WHO 1973; 48 :756-7 . 13 . Chuttani CS, Prakash K, Vergese A, Sharma U, Singha P, Ghosh Ray B . Effectiveness of oral killed typhoid vaccine . Bull WHO 1971 ; 45: 445-50 . 14 . Svennerholm A-M, Sack DA, Holmgren J, Bardhan PK . Intestinal antibody responses after immunisation with cholera B subunit. Lancet 1982 ; i : 305-7 . 15 . Czerkinsky C, Prince SJ, Michalek SM et al. IgA antibody-producing cells in peripheral blood after antigen ingestion : evidence for a common mucosal immune system in humans . Proc Natl Acad Sci USA 1987 ; 84 : 2449-53 . 16 . Forrest BD . Identification of an intestinal immune response using peripheral blood lymphocytes . Lancet 1988 ; ii : 81-3 . 17 . Lycke N, Lindholm L, Holmgren J . Cholera antibody production in vitro by peripheral blood lymphocytes following oral immunization of humans and mice . Clin Exp Immunol 1985 ; 62 : 39-47 . 18 . Kantele A, Arvilommi H, Jokinen I . Specific immunoglobulin-secreting human blood cells after peroral vaccination against Salmonella typhi. J Infect Dis 1986; 153 : 1126-31 . 19 . Kantele A . Antibody-secreting cells in the evaluation of the immunogenicity of an oral vaccine . Vaccine 1990;8 :321-6 . 20 . Kantele A, Makela PH . Different profiles of the human immune response to primary and secondary immunization with an oral vaccine . Vaccine : in press . 21 . Kantele A, Kantele JM, Arvilommi H, Makela PH . Active immunity is seen as a reduction in the cell response to oral live vaccine . Vaccine : in press . 22 . Bienenstock J, Befus AD . Mucosal immunology-review . Immunology 1980; 41 : 249-70 . 23 . Phillips-Quagliata JM, Roux ME, Arny M, Kelly-Hatfield P, McWilliams M, Lamm ME . Migration and regulation of B-cells in the mucosal immune system . Ann NY Acad Sci 1983 ; 409 : 194-203 . 24 . Tomasi TB Jr . Mechanisms of immune regulation at mucosal surfaces . Rev Infect Dis 1983 ; 5 [suppl] : S784-92 . 25 . Strober W, Jacobs D . Cellular differentiation, migration, and function in the mucosal immune system . In : Gallin JI, Fauci AS, eds . Mucosal immunity . New York : Raven Press, 1985 : 1-30 . 26 . Stevens RH, Macy E, Morrow C, Saxon A. Characterization of a circulating subpopulation of spontaneous antitetanus toxoid antibody producing B cells following in vivo booster immunization . J Immunol 1979 ; 122 : 2498-504 . 27 . Thiele CJ, Morrow CD, Stevens RH . Human IgA antibody production after in vivo tetanus toxoid immunization : size and surface membrane phenotype analysis . J Clin Immunol 1982 ; 2 : 327-34 . 28 . Lue C, Tarkowski A, Mestecky J . Systemic immunization with pneumococcal polysaccharide vaccine induces a predominant IgA2 response of peripheral blood lymphocytes and increases of both serum and secretory anti-pneumococcal antibodies . J Immunol 1988; 140 : 3793-800 . 29 . Thomson PD, Harris NS . Detection of plaque-forming cells in the peripheral blood of actively immunized humans . J Immunol 1977 ; 118 :1480-2 . 30 . Kantele A, Takanen R, Arvilommi H . Immune response to acute diarrhea seen as circulating antibodysecreting cells . J Infect Dis 1988 ; 158 : 1011-16 . 31 . Germanier R, Furer E . Isolation and characterization of galE mutant Ty21 a of Salmonella typhi : a candidate strain for a live oral typhoid vaccine . J Infect Dis 1975; 131 : 553-8 . 32 . Black R, Levine MM, Young C et al . Immunogenicity of Ty2la attenuated 'Salmonella typhi' given with sodium bicarbonate or in enteric-coated capsules . Develop Biol Standard 1983 ; 53 : 9-14 . 33 . Gilman RH, Hornick RB, Woodward WE et al. Evaluation of a UPD-glucose-4-epimeraseless mutant of Salmonella typhi as a live oral vaccine . J Infect Dis 1977 ; 136 : 717-23 . 34 . Clemens JD, Sack DA, Harris JR et al. Field trial of oral cholera vaccines in Bangladesh . Lancet 1986 ; ii : 124-7 . 35 . Clemens JD, Harris JR, Sack DA et al. Field trial of oral cholera vaccines in Bangladesh : results of one year of follow-up . J Infect Dis 1988 ; 158 : 60-9 . 36 . Germanier R . Typhoid fever . In : Germanier R, ed . Bacterial vaccines . New York : Academic Press, 1984 : 137-65 .
126
A . Kantele et al .
37 . Kantele A . Immune response to prolonged intestinal exposure to antigen . Scand J Immunol : in press . 38 . Newton SMC, Jacob CO, Stocker BAD . Immune response to cholera toxin epitope inserted in Salmonella flagellin . Science 1989 ; 244 : 70-2 . 39 . Hoiseth SK, Stocker BAD . Aromatic-dependent Salmonella typhimurium are non-virulent and effective as life vaccines . Nature 1981 ; 291 : 238-9 . 40 . Sedgwick JD, Holt PG . A solid-phase immunoenzymatic technique for the enumeration of specific antibody-secreting cells . J Immunol Methods 1983 ; 57 : 301-9 .
Comparison of the human immune response to live oral, killed oral or killed parenteral Salmonella typhi TY21 A vaccines* A . Kantele,' t H . Arvilommi, 2 J . M . Kantele,' L . Rintala' and P . H . Makela'
'National Public Health Institute, SF-00300 Helsinki, Finland and 2 National Public Health Institute, Turku, Finland (Received August 23, 1990 ; accepted in revised form October 16, 1990)
Kantele, A . (National Public Health Institute, SF-00300 Helsinki, Finland), H . Arvilommi, J . M . Kantele, L . Rintala and P . H . Makela . Comparison of the human immune response to live oral, killed oral or killed parenteral Salmonella typhi TY21 A vaccines . Microbial Pathogenesis 1991 ; 10 : 117-126 . The live oral typhoid vaccine Ty21 a has proved to confer protection against the disease at least as effectively as killed parenteral vaccines, whereas killed oral vaccines have not been protective in field trials . This prompted us to compare the immune response of subjects vaccinated either with live oral, killed oral or killed parenteral Salmonella typhi Ty21a vaccine . The immune responses were studied by analysis of peripheral blood antibody-secreting cells (ASC), believed to reflect the mucosal immune response . Live and killed bacteria administered by the oral route elicited immune responses of similar specificity and Ig class profile (IgA dominating), but the response to the live vaccine was significantly stronger and lasted longer . The administration route, on the other hand, influenced the antigenic specificity of the ASC response suggesting different processing of the antigen by the systemic and local immune systems . Thus, the response after oral vaccination was almost exclusively directed to the surface 0-antigen, whereas after parenteral vaccination an equally strong response was seen to the O-antigen, to lipopolysaccharide core and to flagella .
Key words : antibody-secreting cell ; typhoid fever; Salmonella typhi Ty2l a ; oral vaccination ; ELISPOT; secretory immunity .
Introduction We encounter most agents of infectious diseases at or through the large surface area of mucosal membranes and, therefore, protective immunity at these portals of entry would seem highly desirable . An oral vaccine is believed to stimulate the mucosal immune system more efficiently than a parenteral vaccine,' and live oral vaccines are believed to be more efficient than killed ones . 2 Thus, an oral live typhoid vaccine has proved to confer significant protection against the disease, 3-$ whereas oral killed typhoid vaccines have showed only little if any protective effect . 9-13 Specific antibodies have been demonstrated in secretions of orally vaccinated volunteers . 14-16 Specific, predominantly IgA antibody-secreting cells, believed to give information on the mucosal immune response, have been demonstrated in the peripheral blood after oral
'The work was carried out at NPHI, Helsinki and Jyvaskyla, Finland . tAuthor to whom correspondence should be addressed at : NPHI, Mannerheimintie 166, SF-00300 Helsinki, Finland . 0882-4010/91/0201171 10 $03 .00/0
~ci
1991 Academic Press Limited
118
A . Kantele
et a1
vaccination ." 21 These cells are believed to originate in the mucosal immune system : after antigen stimulation in Peyer's patches, activated antigen-specific cells migrate to the local lymph nodes and return from there via the lymphatics and blood to the .22 25 mucosal surfaces for antibody secretion We have recently confirmed the use of the peripheral blood antibody secreting cell (ASC) assay as a measure of the immunogenicity of an oral vaccine ." Circulating ASC have been demonstrated in humans after oral vaccination with both live 16,18 .19 5,17 microorganisms . Circulating ASC are also found after parenteral vacciand killed' 26-29 but comparison of the ASC responses after oral and parenteral immunization nation, suggests qualitative differences between the routes . However, these studies have been carried out using different antigens and somewhat different methods of investigation . To our knowledge, until now, no studies have been carried out that compare the immune responses in humans to the same antigen given both parenterally and orally as a live or a killed vaccine . Specifically, our aim was to discover how the quality of the vaccine (live or killed) and the route of its administration (oral or parenteral) each influenced the immune response, assessed in terms of the magnitude, the kinetics, the Ig class distribution and the specificity of the ASC response .
Results Comparison of the responses to live and killed oral vaccines An ASC response was found in all of the 20 volunteers who received the live oral vaccine (group LO), and in 12 of the 14 volunteers who received the killed oral vaccine (group KO) (Table 1, Fig . 1) . The specificity of the ASC responses was similar in both orally vaccinated groups : the responses were almost exclusively directed to the O-antigen (Fig . 2), in accordance with findings in previous studies ." In both groups, specific ASC were observed in the peripheral blood only after vaccination, showing a peak mostly on day 7 with a subsequent fall [Fig . 1 (a)-(b)] . A difference between the groups was found in the duration of the response : on day 14, specific ASC were still seen in 8 (62%) of the 13 volunteers studied in group LO (live vaccine) [Fig . 1 (a)] compared with 2 (14%) of the 14 volunteers in group KO
Table 1 The peak of the 0-9,12 specific ASC response in volunteers who received S. typhi Ty21 a either as a live oral (LO), killed oral (KO) or killed parenteral (KP) vaccine LO
KO
KP
20 20
14 12
8 8
178 137
24 58
18 28
Number of responders with 0-10 ASC/10 6 cells 11-100 ASC/106 cells 101-1000 ASC/10 6 cells 1001ASC/106 cells
2 3 13 2
4 7 3 0
3 4 1 0
Dominating Ig-classb IgA IgG IgM
61% 11% 28%
No . of individuals vaccinated No . of responders Mean number of ASC/106 cells' SEM
80% 0% 20%
20% 20% 60%
'Geometric mean of the sum of IgA-, IgG- and IgM-ASC/106 cells . ' Counted only of responses exceeding 10 ASC/10 6 cells.
119
Response to live/killed, oral/parenteral vaccine
T
4 T
10
12
10
14
12
14
T T Days after vaccination
Fig . 1 . 0-9,12 specific antibody-secreting cells (ASC) after oral administration of Salmonella typhi Ty2l a either as (a) a live (LO) or (b) a killed (KO) vaccine . Each curve represents the sum of IgA-, IgG- and IgMASC of one individual . The arrows indicate vaccine doses given .
1000
U V d
10
KO KP 0-4,5,12
LO KP H-d
Fig . 2 . The magnitude of the ASC responses (the sum of IgA-, IgG- and IgM-ASC) to the different antigens in the three vaccination groups (LO, KO, KP) calculated for all vaccinees in each group . Note that only IgA-ASC were determined for the LPS core in group LO (in other, similar experiments IgA-ASC has accounted for 55% of the total ASC), and only four subjects in group KP were studied for flagellar d antigen (H-d) .
120
A . Kantele et al .
Fig . 3 . 0-9,12 specific ASC in four volunteers after administration of two doses of killed parenteral
Salmonella typhi Ty21 a vaccine with one month interval (arrows) . Each curve is the sum of IgA-, IgG- and IgM-ASC of one individual labelled so as to make it possible to identify each response throughout the study .
(killed vaccine) [Fig . 1(b)] . The geometric mean of the peaks of the 0-9,12 specific responses was also higher in group LO than in group KO (Table 1, Fig . 2) ; this difference was statistically significant (P < 0 .001) . High responses (peak of more than 100 specific ASC) were seen in both groups, but there were several low responders and two non-responders in group KO (Table 1) . The immunoglobulin class distribution of the ASC responses was determined in those volunteers in whom the peak of the response exceeded 10 ASC/10 6 cells . Groups LO and KO proved similar : 61 and 80% of the responses, respectively, were IgA-dominated, and 28 and 20%, respectively, IgM-dominated (Table 1 ) .
Comparison of the responses to oral and parenteral vaccines The kinetics of the ASC response were similar after oral and parenteral administration of the vaccine, except that on day 14 no ASC were seen in any of the volunteers receiving killed parenteral vaccine (group KP), whereas after oral vaccination the response still persisted in some of the volunteers (see above) . Instead of the IgA-dominated response seen after oral vaccination, the response after parenteral vaccination was most often dominated by the IgM class ; however, this conclusion is based on only five responses exceeding 10 ASC/106 cells (Table 1 ) . Four of the volunteers were given two doses of the killed parenteral vaccine . The kinetics (Fig . 3), the magnitude, the Ig class distribution and the specificity of the responses were essentially similar after both of these doses in each individual . Specificity of the ASC responses was different after parenteral than after oral vaccination (Fig . 2) . All three vaccination groups responded to the 0-9,12 antigen, and also, although with somewhat lower numbers of ASC, to the 0-4,5,12 antigen (determined in groups KO and KP only) . However, only group KP, Le . those who had
Response to live/killed, oral/parenteral vaccine
1 21
been vaccinated parenterally, responded noticeably to the LPS core or to the flagella ; actually these responses in group KP were at least as high as that to the 0-9,12 antigen . No specific ASC were seen to the unrelated control antigen Haemophilus influenzae in any of the groups . The numbers of all ISC stayed essentially constant in all the volunteers in all groups throughout the study (data not shown), in accordance with previous findings after oral immunization . 3o
Discussion and conclusions The data reported here show differences in the magnitude, the duration, the Ig class distribution and the specificity of the specific ASC responses in subjects vaccinated either with live oral, killed oral or killed parenteral Salmonella typhi Ty21 a vaccine . We wanted to see the influence of the properties of the living organism and the administration route on the immune response . To exclude other sources of variation, we used the same antigen for all these vaccinations ; killed parenteral and killed oral vaccines were in fact the same killed vaccine preparation and the bacterial strain for it was obtained by culturing from the live vaccine preparation . It was more difficult to achieve a similar dose for all groups . It was obvious that in oral administration a dose of the live vaccine should contain lower numbers of bacteria than the killed vaccine, because the live vaccine strain is expected to multiply in the intestine . On the other hand, to induce an immune response, considerably higher numbers of bacteria are required for an oral than a parenteral administration of a vaccine; giving such high numbers of bacteria parenterally would cause fever and other adverse reactions . The live oral vaccine was the one used in several field trials in a dose at least 10 9 bacteria given on each of days 0, 2 and 4 . 3-6 The killed oral vaccine was given as three doses of 400 x 109 bacteria each ; this was the highest number of killed bacteria reported to have been used in field trials in India ." In killed parenteral vaccination the dose used (0 .5 x 10 9 bacteria) was the one currently recommended for parenteral typhoid vaccines .
Interestingly, the magnitude of the specific ASC response was significantly higher in group LO than in group KO, although the number of bacteria given per dose was lower in group LO than in group KO . Live S . typhi Ty2l a multiply in the intestine, and hence after an oral vaccination, the number of bacteria will increase . However, we did not expect this multiplication to be extensive, partly because of the slow multiplication rate typical of the Ty21 a mutant, and partly because of the limited nature of this multiplication in the intestine as described by Germanier and Fiarer . 31 Indeed, in volunteer studies, the vaccine strain could not be isolated from the vaccinated volunteer's faeces 1 or 2 days after vaccination .32, 33 Hence, the final number of bacteria used for vaccination in these groups may have been rather similar . Therefore, the significantly higher response in group LO as compared to group KO is likely to indicate a special advantage of the living bacteria in stimulating the immune system ; this could be due to their capacity to invade the mucosal epithelium and/or to persist longer in the mucosa or in the Peyer's patches . The lower responses in group KO compared with group LO parallel the results obtained for protection in field trials : only low or negligible protection has been found after vaccination with a killed oral typhoid vaccine, 9-13 whereas the live oral vaccine confers significant protection .` Some of the volunteers in group KO had high responses ; it is possible that the KO vaccine confers significant protection to a small proportion of vaccinated individuals, while the immune response in the majority remains inadequate, and therefore no or only low protection is found in field trials . Hence, the results show that even if the killed oral vaccine is significantly less effective in stimulating the immune system than a live vaccine, it is not ineffective . A
122
A . Kantele et a/
killed oral cholera vaccine has been shown to confer significant protection against cholera ." ," It might be that the difference between live and killed oral vaccines is more pronounced with Salmonella than with Vibrio cholerae, because the former are invasive, whereas the latter is not . Further, in contrast to cholera, cell-mediated immune response is believed to be essential for protection against typhoid fever .
36
The kinetics of the ASC responses were similar in the three vaccine groups, and were also similar to our earlier findings after oral live vaccine ." 19 A few volunteers
were also assessed after a second dose of the killed parenteral vaccine given 1 month after the first dose as recommended for parenteral typhoid vaccination ; surprisingly, the ASC response was qualitatively and quantitatively similar to those seen in the same individuals after the first dose . The response to the oral vaccines seemed to persist somewhat longer than that to the parenteral vaccine, possibly as a result of the prolonged exposure to the antigen, since the oral vaccines were given as three doses on days 0, 2 and 4 . Prolonged intestinal exposure to antigen has, indeed, been shown to prolong the peripheral blood ASC response . 37 Also the difference between groups
KO and LO (14% versus 63% responding on day 14) might, at least partly, be explained by a difference in the duration of antigen exposure : the live bacteria probably persist longer in the tissues than the killed bacteria . We were also interested in comparing the immunoglobulin class distribution of the ASC response in the three vaccination groups . Previously, IgA-dominated responses have been considered typical of oral vaccination, 17 whereas IgG-dominated responses have been reported after parenteral vaccination e .g . with tetanus 26 and cholera
toxoids ." However, parenterally administered pneumococcal polysaccharide vaccine has elicited an IgA-dominated response . 28 Our recent study (to be published)
with parenteral Hib-conjugate vaccine shows an IgA-dominated response to the polysaccharide part of the vaccine (polyribosylphosphate) and a simultaneous IgGdominated response to the protein part of the vaccine (diphtheria toxoid), which implies that the nature of the antigen influences the Ig class distribution of an ASC response . The present study, on the other hand, shows that also the administration route of the vaccine influences the Ig class distribution : the same antigen (0-9,12) elicited an IgA-dominated response when given orally, but an IgM-dominated response when given parenterally . An unexpected finding in the present study was the difference in the antigenic specificity of the ASC response after parenteral and oral vaccination . The response to oral vaccination was mainly directed to the 0 antigen, whereas responses to the LPS core and to the flagella were either absent or very low, as demonstrated also in our earlier studies ." Interestingly, killed parenteral vaccination with the same antigen not only elicited an 0-specific ASC response, but concomitantly also an equal response to the LPS core and to the flagella . As far as we know, this difference in the antigenic specificity after oral and parenteral vaccination has not been reported before . We can only speculate about the reasons for this difference . In intact bacteria, the 0 antigen is exposed on the bacterial surface and is therefore ready to react with antibodies e .g . on the surface of B cells . By contrast, the LPS core is not exposed, and would require processing of the antigen before presentation ; this processing might be more efficient after systemic than after oral immunization . The difference might also arise because vaccine bacteria have different fates in the two cases : after oral vaccination, whole bacteria would gain access from the gut lumen via the M cells directly to the B cells in Peyer's patches, whereas after parenteral vaccination the antigen would, in most cases, be ingested by the macrophages at the site of injection, and B cells would meet it only after digestion by macrophages . It seems that after oral immunization the response is almost exclusively directed to the surface of the bacteria ; such antibodies
Response to live/killed, oral/parenteral vaccine
123
would, in fact, be immediately useful in the gut . The response to the flagellar antigen, which is also an exposed structure, was low or absent after oral vaccination . One explanation for this might be that the protein would be rapidly destroyed in the intestinal milieu ; however, there is no direct evidence for this . The good immunogenicity of the 0 antigen is consistent with this hypothesis, since the LPS is very resistant to digestive enzymes . The low or absent response to flagellar antigen after oral administration of the vaccine should be confirmed by further studies, because of its potentially important consequences in the use of oral vaccines introducing foreign antigens inserted in the Salmonella flagellin ." In conclusion, the present study shows that the use of live instead of killed bacteria in an oral vaccine significantly enhances its immunogenicity : the ASC response after live oral vaccine was more vigorous and lasted longer . The antigenic specificity of the ASC response is strongly influenced by the route of administration : after oral vaccination the response was almost exclusively directed to the surface 0 antigen of the bacteria, whereas after parenteral vaccination an equally strong response was found to the 0 antigen, LPS core and flagella, suggesting different processing of the antigen by the systemic and local immune systems .
Materials and methods Volunteers and vaccination . Forty-two healthy volunteers (28 women and 14 men, aged 2045 years) consented to participate in the study . None of them had a history of or contact with individuals having typhoid fever, nor had they been vaccinated against typhoid fever . Twenty subjects were vaccinated with a live oral, 14 with a killed oral and eight with a killed parenteral typhoid vaccine . Informed consent was obtained from each volunteer before the study . The study protocol was approved by the ethical committee of this Institute . Data on group LO have been reported in an earlier study ." All the vaccines were prepared of Salmonella typhi Ty2l a . The oral live vaccine was a commercially available oral S . typhi Ty2l a vaccine (Vivotif Berna, Swiss Serum and Vaccine Institute, Bern, Switzerland) in suspension formulation ; it contained at least 10 9 bacteria/dose, each lyophilized dose was suspended immediately before ingestion in the buffer provided by the manufacturer . The killed parenteral and oral vaccines were prepared at the Vaccine Laboratory of this Institute according to its routine methods, in brief : the strain Salmonella typhi Ty2la was cultured from the oral live vaccine and grown overnight in Brain Heart Infusion broth . The bacteria were killed by incubating the bacterial suspension for 45 min in a waterbath at 56°C and cooled ; phenol was added to a concentration of 0 .5% before overnight incubation at 21'C . Then, the suspension was washed once with saline (0 .9% NaCI) and suspended in a vaccine buffer (0 .05 M glycine, 0 .01 M succinate, 0 .085 M NaCl, pH 6 .7-7 .0, with 0 .3% phenol) . From this stock suspension, the oral preparation was made to contain 400x10 9 bacteria/dose . Each oral dose was dialysed against the vaccine buffer overnight before ingestion . The dialysed vaccine was tested for the absence of any pathogenic bacteria, and for the presence of other bacteria (less than 2000/dose) and molds (less than 20/dose) . From the same stock suspension, the parenteral vaccine was diluted to contain 10 9 bacterial/ml and phenol added to a final concentration of 0 .3% . The quality of the preparations was controlled at each step according to routine methods . Toxicity tests were carried out in guinea pigs and mice, and the antigenic nature tested by agglutination with specific antisera . To minimize the effect of gastric acid, a capsule containing 0 .8 g of sodium bicarbonate was to be ingested 2 min before the killed oral vaccine preparation . The oral vaccines were given in three doses at two day intervals . The total number of bacteria given orally was at least 3x10 9 (live, group LO) or 1 .2x 1012 (killed, group KO) . The killed parenteral vaccine was given intramuscularly in the left arm as one dose of 0 .5x10 9 bacteria (group KP) . Four subjects in group KP also received a second dose of the same vaccine one month later . Study design. Day 0 was the first vaccination day . Samples of heparinized venous blood were collected on days 0, 5, 7, 9 and 14 in all groups, and for those receiving two doses of killed parenteral vaccine, additional samples were collected on the corresponding days in
124
A . Kantele eta[
relation to the second vaccine dose . Responses to vaccination were followed by quantitation of specific ASC and total immunoglobulin-secreting cells (ISC) among peripheral blood lymphocytes in each blood sample . Isolation of lymphocytes. Mononuclear cells containing mainly lymphocytes were obtained by Ficoll-Paque (Pharmacia, Uppsala, Sweden) centrifugation of heparinized venous blood . Isolated cells washed three times in Hank's balanced salt solution (Flow Laboratories, Irvine, U .K .), were suspended in culture medium, and adjusted to the concentration of 2x10' cells/ml as described previously ." Assay of specific antibody-secreting cells . The number of specific ASC among newly isolated blood lymphocytes was determined by the ELISPOT-method . The isolated cells were allowed to secrete antibodies in antigen-coated 96-well microtiter plates (Immunoplate I, Nunc, Roskilde, Denmark) . Thereafter, the specific antibodies bound to the antigen were detected by application of enzyme-labelled antisera followed by substrate-agarose overlay ; the latter immobilizes the decaying substrate and allows the area of antibodies to be visualized as a coloured spot . Hence, each spot represents one cell secreting antibodies to the coating antigen . Coating was performed with three whole cell bacterial antigens, selected to share with S . typhi its total 0-antigen (0-9,12), only the 0-antigenic factor 12 (0-4,5,12), the lipopolysaccharide (LPS) core or the flagellar d-antigen . Haemophilus influenzae was used as an unrelated control antigen . All these antigens have been described by Kantele 19 except the S. typhimurium 04,5,12 strain ." The coating antigens were incubated at a concentration of 10 8 bacteria/ml PBS overnight at 20°C or for three hours at 37°C . After coating, the plates were masked with 1 bovine serum albumin in PBS for 30 min at 37°C . Next, 10 5 isolated mononuclear cells in culture medium were incubated in each well . Then, alkaline phosphatase-conjugated anti-IgA, G or M (Orion Diagnostica, Helsinki, Finland) diluted 1 :100 in 1 %BSA-PBS was allowed to react overnight at 4°C . The substrate (5-bromo-4-chloro-3-indolyl phosphate; Sigma) was applied in hot (50°C) agarose as described by Sedgwick and Holt ." Spots were counted under low magnification after storage at 4°C . Assay of a// immunoglobulin-secreting cells (c/asses IgA, IgG and IgM) . All immunoglobulin secreting cells (ISC) were quantitated with ELISPOT as described for specific ASC above with the following differences : coating was performed with human IgA, IgG or IgM-specific antisera, 19 the cells were added at a concentration of 5 x 104 cells/well and the second antibody used in the total IgG-ASC assay was as described by Kantele . 19 Statistical methods . The geometric means of the peak responses were calculated for all vaccinated subjects within each group (the peak value for each non-responder was assigned the value of 1) . Statistical comparison was carried out by Student's t-test using log transformed data .
We thank Mrs Anja Ratilainen for careful assistance in the laboratory work . This study was supported by the Emil Aaltonen Foundation, the Finnish found for the replacement of animal experimentation, the Finnish Medical Foundation and partly by the Diarrhoeal Diseases Control Programme, WHO .
References 1 . Mestecky J . The common mucosal immune system and current strategies for induction of immune responses in external secretions . J Clin Immunol 1987; 7 : 265-79 . 2 . Dougan G, Maskell D, O'Callaghan D, Chatfield S, Charles I, Hormaeche C . Oral vaccination . Antonie van Leeuwenhoek 1988 ; 54 : 447-51 . 3 . Wahdan MH, Serie C, Cerisier Y, Sallam S, Germanier R . A controlled field trial of live Salmonella typhi strain Ty2l a oral vaccine against typhoid : three year results . J Infect Dis 1982 ; 145: 292-5 . 4 . Levine MM, Black RE, Ferreccio C et al. The efficacy of attenuated Salmonella typhi oral vaccine strain Ty2l a evaluated in controlled field trials . In : Holmgren J, Lindberg A, Mollby R, eds . Development of vaccines and drugs against diarrhea . Lund : Studentlitteratur, 1986 : 90-101 . 5 . Ferreccio C, Levine MM, Rodriguez H . Contreras R, Chilean Typhoid Committee. Comparative efficacy
Response to live/killed, oral/parenteral vaccine
125
of two, three or four doses of TY21 a live oral typhoid vaccine in enteric-coated capsules : a field trial in an endemic area . J Infect Dis 1989 ; 159 : 766-9 . 6 . Levine MM, Ferreccio C, Black R, Tacket CO, Germanier R, Chilean Typhoid Committee . Progress in vaccines against typhoid fever . Rev Infect Dis 1989 ; 11 : S552-67 . 7 . Black RE, Levine MM, Ferreccio C et al. Efficacy of one or two doses of Ty21a Salmonella typhi vaccine in enteric-coated capsules in a controlled field trial . Vaccine 1990; 8: 81-4 . 8 . Black RE, Levine MM, Clements ML, Young CR, Svennerholm A-M, Holmgren J . Protective efficacy in humans of killed whole-vibrio oral cholera vaccine with and without the B subunit of cholera toxin . Infect Immun 1987 ; 55 :1116-20 . 9 . DuPont LH, Hornick RB, Snyder MJ, Dawkins AT, Heiner GG, Woodward TE . Studies of immunity in typhoid fever . Protection induced by killed oral antigens or by primary infection . Bull WHO 1971 ; 44 : 667-72 . 10 . Borgono JM, Corey G, Engelhardt H . Field trials with killed oral typhoid vaccines . Devel Biol Stand 1976 ; 33 : 80-4 . 11 . Chuttani CS, Prakash K, Gupta P, Groven V, Kumar A . Controlled field trial of a high dose oral killed typhoid vaccine in India . Bull WHO 1977; 55 : 643-4 . 12 . Chuttani CS, Prakash K, Vergese A et al. Ineffectiveness of an oral killed typhoid vaccine in a field trial . Bull WHO 1973; 48 :756-7 . 13 . Chuttani CS, Prakash K, Vergese A, Sharma U, Singha P, Ghosh Ray B . Effectiveness of oral killed typhoid vaccine . Bull WHO 1971 ; 45: 445-50 . 14 . Svennerholm A-M, Sack DA, Holmgren J, Bardhan PK . Intestinal antibody responses after immunisation with cholera B subunit. Lancet 1982 ; i : 305-7 . 15 . Czerkinsky C, Prince SJ, Michalek SM et al. IgA antibody-producing cells in peripheral blood after antigen ingestion : evidence for a common mucosal immune system in humans . Proc Natl Acad Sci USA 1987 ; 84 : 2449-53 . 16 . Forrest BD . Identification of an intestinal immune response using peripheral blood lymphocytes . Lancet 1988 ; ii : 81-3 . 17 . Lycke N, Lindholm L, Holmgren J . Cholera antibody production in vitro by peripheral blood lymphocytes following oral immunization of humans and mice . Clin Exp Immunol 1985 ; 62 : 39-47 . 18 . Kantele A, Arvilommi H, Jokinen I . Specific immunoglobulin-secreting human blood cells after peroral vaccination against Salmonella typhi. J Infect Dis 1986; 153 : 1126-31 . 19 . Kantele A . Antibody-secreting cells in the evaluation of the immunogenicity of an oral vaccine . Vaccine 1990;8 :321-6 . 20 . Kantele A, Makela PH . Different profiles of the human immune response to primary and secondary immunization with an oral vaccine . Vaccine : in press . 21 . Kantele A, Kantele JM, Arvilommi H, Makela PH . Active immunity is seen as a reduction in the cell response to oral live vaccine . Vaccine : in press . 22 . Bienenstock J, Befus AD . Mucosal immunology-review . Immunology 1980; 41 : 249-70 . 23 . Phillips-Quagliata JM, Roux ME, Arny M, Kelly-Hatfield P, McWilliams M, Lamm ME . Migration and regulation of B-cells in the mucosal immune system . Ann NY Acad Sci 1983 ; 409 : 194-203 . 24 . Tomasi TB Jr . Mechanisms of immune regulation at mucosal surfaces . Rev Infect Dis 1983 ; 5 [suppl] : S784-92 . 25 . Strober W, Jacobs D . Cellular differentiation, migration, and function in the mucosal immune system . In : Gallin JI, Fauci AS, eds . Mucosal immunity . New York : Raven Press, 1985 : 1-30 . 26 . Stevens RH, Macy E, Morrow C, Saxon A. Characterization of a circulating subpopulation of spontaneous antitetanus toxoid antibody producing B cells following in vivo booster immunization . J Immunol 1979 ; 122 : 2498-504 . 27 . Thiele CJ, Morrow CD, Stevens RH . Human IgA antibody production after in vivo tetanus toxoid immunization : size and surface membrane phenotype analysis . J Clin Immunol 1982 ; 2 : 327-34 . 28 . Lue C, Tarkowski A, Mestecky J . Systemic immunization with pneumococcal polysaccharide vaccine induces a predominant IgA2 response of peripheral blood lymphocytes and increases of both serum and secretory anti-pneumococcal antibodies . J Immunol 1988; 140 : 3793-800 . 29 . Thomson PD, Harris NS . Detection of plaque-forming cells in the peripheral blood of actively immunized humans . J Immunol 1977 ; 118 :1480-2 . 30 . Kantele A, Takanen R, Arvilommi H . Immune response to acute diarrhea seen as circulating antibodysecreting cells . J Infect Dis 1988 ; 158 : 1011-16 . 31 . Germanier R, Furer E . Isolation and characterization of galE mutant Ty21 a of Salmonella typhi : a candidate strain for a live oral typhoid vaccine . J Infect Dis 1975; 131 : 553-8 . 32 . Black R, Levine MM, Young C et al . Immunogenicity of Ty2la attenuated 'Salmonella typhi' given with sodium bicarbonate or in enteric-coated capsules . Develop Biol Standard 1983 ; 53 : 9-14 . 33 . Gilman RH, Hornick RB, Woodward WE et al. Evaluation of a UPD-glucose-4-epimeraseless mutant of Salmonella typhi as a live oral vaccine . J Infect Dis 1977 ; 136 : 717-23 . 34 . Clemens JD, Sack DA, Harris JR et al. Field trial of oral cholera vaccines in Bangladesh . Lancet 1986 ; ii : 124-7 . 35 . Clemens JD, Harris JR, Sack DA et al. Field trial of oral cholera vaccines in Bangladesh : results of one year of follow-up . J Infect Dis 1988 ; 158 : 60-9 . 36 . Germanier R . Typhoid fever . In : Germanier R, ed . Bacterial vaccines . New York : Academic Press, 1984 : 137-65 .
126
A . Kantele et al .
37 . Kantele A . Immune response to prolonged intestinal exposure to antigen . Scand J Immunol : in press . 38 . Newton SMC, Jacob CO, Stocker BAD . Immune response to cholera toxin epitope inserted in Salmonella flagellin . Science 1989 ; 244 : 70-2 . 39 . Hoiseth SK, Stocker BAD . Aromatic-dependent Salmonella typhimurium are non-virulent and effective as life vaccines . Nature 1981 ; 291 : 238-9 . 40 . Sedgwick JD, Holt PG . A solid-phase immunoenzymatic technique for the enumeration of specific antibody-secreting cells . J Immunol Methods 1983 ; 57 : 301-9 .
