Studies in volunteers to evaluate candidate Shigella vaccines: further experience with a bivalent Salmonella typhi-Shigella sonnei vaccine and protection conferred by previous Shigella sonnei disease Deirdre A. Herrington *§, Lillian Van De Verg*, Samuel B. Formal t, Thomas L. Hale t, Ben D. Tall*, Stanley J. Cryz +, Edmund C. Tramont t and Myron M. Levine* A bivalent vaccine consisting of Salmonella typhi strain Ty21a containing the 120MDa plasmid of Shigella sonnei and expressing both S. typhi and S. sonnei lipopolysaccharides ( LPS) on its surface was previously shown to protect significantly against S. sonnei disease in experimental challenge studies. However, protective efficacy could not be reconfirmed in volunteers with five subsequent lots of vaccine. One vaccine lot which resembled the initial protective lots of vaccine in biochemical and serological tests, and by electron microscopy, was administered to 16 volunteers who ingested three doses of 109 organisms each. Antibody secreting cells (ASC) specific for S. sonnei LPS were detected in the blood of 100% of vaccinees, but no protection of these vaccinees was demonstrated during a S. sonnei challenge study. To assess the ability of the volunteer model to detect infection-derived immunity, six volunteers who had had clinical shigellosis due to S. sonnei two months earlier were rechallenged with wild-type S. sonnei, together with 12 controls. Prior infection provided 100% protection against febrile illness (p=O.05) and diarrhoea (p=0.04), thereby validating the volunteer model for assessing Shigella vaccines. Keywords:Shigellasonnei; Salmonella typhi; bivalent vaccine Introduction The ultimate proof of safety and protective efficacy of a candidate Shigella vaccine depends upon studies in man. Experimental trials in small numbers of volunteers can help determine which candidate vaccines are worthy of further development and testing. For example, studies in volunteers have been used to demonstrate protection conferred by a streptomycin-dependent S. flexneri 2a oral vaccine ~. In these studies, a protective efficacy of 50% was seen in the volunteer studies in contrast to the field studies in which the vaccine conferred .~90% protection 2'3. In addition, rechallenge studies in volunteers have demonstrated that previous illness with S. flexneri 2a confers significant protection against rechallenge with the homologous strain ~. Based on the above experiences, we used volunteer trials to investigate the safety, immunogenicity and *Center for Vaccine Development, Division of Geographic Mi~dicine, Department of Medicine, University of Maryland at Baltimore, Baltimore, MD 21201, USA. t Department of Enteric Infections, Walter Reed Army Institute of Research, Walter Reed Army Medical Center, Washington DC, USA. tSwiss Serum and Vaccine Institute, Berne, Switzerland. §To whom correspondence should be addressed at Division of Geographic Medicine, University of Maryland, 10 South Pine Street, Baltimore, MD 21201, USA. (Received 26 October 1989; revised 16 January 1990; accepted 22 January 1990) 026"-410X/90/04035..~05 1990Butterworth-HeinemannLtd
efficacy of an oral attenuated bivalent S. typhi/S, sonnei vaccine (strain 5076-1C) which consists of S. typhi strain Ty21a containing the 120-megadalton plasmid of S. sonnei and expressing both S. typhi and S. sonnei lipopolysaccharides (LPS) on its surface ~'5. As reported previously, two initial lots of bivalent S. sonnei-S, typhi vaccine 5076-1C, but not a third lot, conferred significant protection against clinical shigellosis due to S. sonnei in volunteers 6. In previous studies with 5076-1C 6'7, as with earlier studies with other vaccines 2,8-1° and with re-challenges 2, neither measurement of serum antibody nor of intestinal fluid slgA to Shigella 0 antigens was sensitive in detecting immune responses. Recent reports have described the detection in the peripheral blood of trafficking immunoglobulin-secreting cells several days after oral immunization. Preliminary results from studies of oral typhoid 1L ~2 and cholera ~3 vaccines suggest that this may be a technique that markedly increases the ability to detect specific immune responses following oral immunization. In the present report, clinical studies with four additional lots of bivalent vaccine 5076-1C are presented. In one of these studies, the specific IgA-secreting lymphocyte response was measured and shown to be sensitive in detecting immune responses to the vaccine. Finally, the capacity of the volunteer challenge model to detect protective immunity in vaccinees or controls with recent S. sonnei illness is demonstrated, thereby validating
Vaccine, Vol. 8, August 1990 353
Bivalent S. typhi-S, sonnei vaccine: D.A. Herrington et al.
the use of this model for conducting efficacy studies of candidate S. sonnei vaccines. Materials and methods
Vaccine preparation The vaccine strain, 5076-1C, was prepared by conjugal transfer of the form I S. sonnei plasmid into the attenuated S. typhi strain Ty21a as reported 4. A seed lot of vaccine strain was prepared and used for the production of all lots for clinical studies. As previously described 6 for lots 2, 5, and 8, lot 14 was prepared at the Department of Biologic Research, Walter Reed Army Institute of Research (Washington, DC, USA) by harvesting after growth on solid agar, dispensing into vials, and lyophilizing. Lots SSVI 1 and SSVI 2, and lot 87-5-I, were formulated at the Swiss Serum and Vaccine Institute, Berne, Switzerland where they were harvested after growth in liquid medium in a fermenter. The procedure used for lot 87-5-I is summarized as follows. Agar-grown bacteria which agglutinated strongly with antisera specific for S. sonnei I O antigen and S. typhi d (flagella) were used to inoculate a fermenter containing Brain Heart Infusion broth-Hycase SF medium (BHI-HFS), and the culture was grown for 18 h at 30°C. The cells were collected by continuous flow centrifugation under aseptic conditions. The cell pellet was suspended in sterile sucrose Hycase-SF ascorbic acid cryoprotective medium, poured onto sterile metal plates, frozen at -70°C and lyophilized under aseptic conditions. The lyophilized vaccine bacteria were mixed with lactose and dispensed into individual sealed aluminium foil packets containing 3 x 109 viable cells per packet. Packets were stored at 4°C. Several random packets were reconstituted with sterile distilled water, and electron microscopy demonstrated that the ceils were sparsely piliated and possessed intact and attached flagella. Viable counts and agglutinations with specific antisera were confirmed for several random packets on the day prior to each vaccination. Volunteers Adult volunteers between the ages of 18 and 35 years from the Baltimore metropolitan area participated in the studies. The methods of medical screening, care of the volunteers, and obtaining informed consent have been described ~4.15. Protocols were reviewed by the Human Volunteer Research Committee of the University of Maryland at Baltimore and the Human Use Review Office of the Surgeon General's Human Subjects Research Review Board of the Department of the Army. Administration o f vaccine Vaccine (approximately 2x 10 9 viable cells) was administered orally with 2g sodium bicarbonate mixed in 150ml of water for all lots with the exception of lot 87-5-I. For lot 87-5-1, a commercially prepared packet of powdered buffer (2.65g sodium bicarbonate, 1.65g ascorbic acid, 0.2 g lactose) and a packet of vaccine were stirred into 100 ml sterile distilled water. The vaccine was ingested by each volunteer on three occasions separated by three or four days. The volunteers fasted 90 min before and after ingestion of the vaccine and returned 24 and 48h after each dose for questioning about adverse reactions and to provide a stool specimen for culture.
354
Vaccine, Vol. 8, August 1990
Challenge studies The methods of preparing the inoculum of wild type S. sonnei strain 53G, medical care of the volunteers, and bacteriology have been previously described in detail 6. Multiple vials of a seed lot of strain 53G were stored at - 7 0 ° C at the initiation of studies, and challenge inocula were prepared from this frozen stock.
Immunological studies Sera were obtained from the volunteers before, and 7, 14, 21 and 28 days after, the first dose of vaccine. Serum IgG and IgA antibodies to the O antigen of S. sonnei were measured by ELISA 6. Fourfold increases in titre were considered significant. Heparinized blood was collected for solid-phase enzyme-linked immunospot (ELISPOT) assays (see below) 0, 9, 14 and 21 days after the first dose of vaccine; lymphocytes were separated by FicoU Hypaque and stored in vapour nitrogen. The ELISPOT assay 16-1s was used for enumeration of antibody-secreting cells (ASC) (mononuclear cells which secreted Shigella anti-LPS antibody). In brief, thawed lymphocytes were resuspended in RPMI 1640 culture medium (Flow Laboratories) supplemented with 10% heat-inactivated fetal calf serum, gentamicin (I 5 ttg ml- 1), and L-glutamine (3 mg ml- ~) and adjusted to a concentration of 2.5 x 106 viable cells per ml. Cells were incubated for 3h at 37°C in the antigen-coated wells of flat-bottomed 96-well microtitre plates (Immunoplate I; Nunc). Specific IgA antibodies secreted by the cells were detected by addition of goat anti-human IgA (Kirkegaard Perry Laboratories) at a dilution of 1: 500 followed by a substrate-agarose overlay and microscopic counting of the coloured spots. Antigens used were Plesiomonas shigelloides LPS (kindly provided by Dr All Lindberg, Stockholm, Sweden), which is immunologically and biochemically identical to S. sonnei LPS t9 but technically easier to prepare since S. sonnei readily becomes rough in culture, and S. typhi LPS (Difco). Antigens were used at a concentration of 10/~g ml-1, and Vibrio cholerae Inaba LPS (Sigma) and buffer were used as negative controls.
Results
Clinical tolerance, vaccine shedding, immunologic response to vaccination and vaccine efficacy Table 1 summarizes clinical trials with seven separate lots of vaccine. Results for three of these vaccine lots (2, 5, and 8) were reported previously6. Separate cohorts of volunteers with no previous history of shigellosis or enrolment in previous Shigella vaccine studies were utilized for each trial. Vaccine excretion rates, serologic responses to S. sonnei LPS following ingestion of lots 14, SSVI 1 and SSVI 2 were similar to results seen following administration of the protective lots 2 and 5 (Table I). The vaccine strains recovered from coproculture were all found to express S. sonnei antigen as determined by agglutination with specific antiserum. Nevertheless, as summarized in Table 1, lots 14, SSVI 1 and SSVI 2 failed to protect against experimental challenge with pathogenic S. sonnei strain 53(3.
Bivalent S. t y p h i - S , sonnei vaccine: D.A. Herrington et al. Table 1 Bacteriology, immunological response to, and protective efficacy of various lots of bivalent Salmonella typhi-Shigella sonnei vaccine 5076-1C administered to volunteers
Seroconversions by ELISA to S. sonnei LPS
Vaccine lot" Lots 2 and 5 (n =43) ° Lots 8 and 14 (n = 31) Lots SSVl 1 and SSVl 2 (n = 30) Lot 87-5-1 (n= 16)
Vaccine excretion (%)
IgA (%)
IgG (%)
Attack rate after challenge Vaccinees
Controls
p valu&
23
25
23
10/28 (36%)
20/29 (69%)
0.01
23
16
19
12/25 (48%)
8/17 (47%)
0.60
33
43
27
3/20 (9%)
5/18 (28%)
0.29
75
56
19
5/13 (38%)
5/12 (42%)
0.60
• Results for vaccine lots 2, 5, and 8 were reported previously (ref. 6); "p value was determined by using Fisher's exact test, 1 tail; °n=number of volunteers receiving vaccine
Efforts to improve vaccine formulation: immunogenicity o f lot 87-5-1 The apparent loss of vaccine efficacy in volunteers studied led to an intensive examination of the biochemical and morphological characteristics of the protective versus non-protective lots. These examinations were undertaken in an effort to identify the critical components of the vaccine strain and the steps in the manufacturing process which may have rendered the strain less effective. The most extensive testing that was carried out compared lot 5 (protective) and lot 8 (non-protective) organisms directly reconstituted from lyophilized vaccine; no differences were found between these lots with respect to the presence of lactose-fermenting organisms; agglutination patterns with Salmonella group D, Shigella group D, or S. sonnei form I antisera were similar; plasmid DNA restriction endonuclease digestion patterns also failed to show a difference; neither lot exhibited mannose-sensitive haemagglutination (due to type 1 somatic pill). However, a difference was found between the lots with respect to flagella. The S. typhi d (flagellar) agglutination reaction was fourfold higher for lot 5 than for non-protective lot 8 organisms. This difference was corroborated by electron microscopy; lot 8 had no attached, intact flagella, whereas intact flagella were readily found on lot 5 organisms. Electron microscopy of samples of non-protective lots 14, SSVI 1 and SSVI 2 following gentle rehydration with water also showed that flagella either were not present, were detached from the bacterial cells, or were detached and fragmented. By systematic changes in the manufacturing process, a new vaccine lot (87-5-I) was prepared which closely resembled the original protective lots biochemically, serologically and by electron microscopic visualization. The vaccination study with lot 87-5-I revealed that vaccine organisms were recovered in small numbers in coprocultures from 12 out of 16 volunteers; no volunteers manifested adverse reactions. Nine out of 16 volunteers (56%) had a fourfold or greater rise of serum IgA anti-LPS as measured by ELISA, while 3 out of 16 (19%) had significant rises in serum IgG. Thus vaccine excretion and serum immune response to lot 87-5-I were similar to those elicited by protective lots 2 and 5 and by non-protective lots 14, SSVI 1, and SSVI 2 (Table I). Because a correlation between IgA production by peripheral blood lymphocytes several days after oral immunization and the detection of specific intestinal sIgA
to
o= o
I_ 10
O3 t_ t~ >. O .D cr~
..~....
. ......"
...~ 5
Z
10 Time a f t e r
Figure 1
15
20
25
vaccination (days}
Mean numbers of mononuclear cells secreting IgA against
S. sonnei (P. shigelloides LPS, see text) or S. typhi LPS as detected in peripheral blood on days 0, 9, 14, and 21 after vaccination with S. typhUS. sonnei bivalent vaccine 5076-1C. O, P. shigelloides L P S ; / k , S. typhi LPS
has been reported x1,18, circulating lymphocytes were collected from the vaccinees on several occasions after vaccination with lot 87-5-I. Lymphocytes were available for analysis from 13 vaccinees. Antibody secreting cells (ASC) specific for P. shigelloides polysaccharide antigen were detected on day 9 in all 13 vaccinees and were no longer found on day 21 (Figure I ). Ten out of 13 vaccinees (77%) also had ASC specific for S. typhi LPS (Figure I), while none of the volunteers produced ASC against the control LPS from V. cholerae serotype Inaba (data not shown). These results indicate that the S. sonnei/S, typhi bivalent vaccine triggered an immune response to O antigen in 100% of the vaccinees tested, even though no protection was apparent in the volunteer model (see below).
Challenge study to assess vaccine efficacy
(lot 87-5-1)
One month after receiving the third dose of vaccine, 13 vaccinees and 12 unvaccinated controls were challenged with approximately 4 x 10 z smooth, form I colonies of S. sonnei strain 53G in an experimental study to assess
Vaccine, Vol. 8, A u g u s t 1990
355
Bivalent S. typhi-S, sonnei vaccine: D.A. Herrington et al. Table 2 Protective efficacy conferred by prior shigellosis due to S. sonnei as determined by experimental challenge in volunteers Protection against
Veterans
Controls
p value"
Protective efficacy (%)
Febrile illness (T> 37.8°C) Dysentery Diarrhoea b
0/6 1/6 0/6
6/12 8/12 7/12
0.05 0.13 0.04
100 78 100
'p value determined by Fisher's exact test, two tail; ~Diarrhoea defined as 1 stool of >300ml or 2 or more stools totalling >200ml within a 48 h period
the efficacy of the vaccine. Clinical shigellosis developed in 5 out of 13 vaccinees and 5 out of 12 controls (Table I). In this challenge study, there was no evidence of protection or of amelioration of disease due to the vaccine.
Challenge study to assess infection-derived immunity to S. sonnei In view of the variable efficacy of 5076-1C vaccine, we considered it critical to validate the ability of the volunteer model of experimental S. sonnei shigellosis to detect protective efficacy. To accomplish this goal we used the model to measure infection-derived immunity against S. sonnei. Accordingly, six of the ~,".unteers who experienced clinical shigellosis while participating in the vaccine challenge study (four of whom had previously received vaccine) were rechallenged two months later along with 12 naive controls. The data are summarized in Table 2. Only one of the six rechallenged volunteers met the definition of illness; this volunteer had a single dysenteric stool with no other symptoms. In contrast, 8 out of 12 of the control volunteers developed illness (diarrhoea 7 out of 12; fever 6 out of 12; dysentery 8 out of 12). While the possibility that previous vaccination of four of the rechallenged volunteers contributed to their subsequent protection cannot be excluded, the results nevertheless demonstrate the utility of the challenge model for detecting protective immunity.
not be optimal. Indeed, in this hybrid, the S. sonnei O-disaccharide repeat unit is not complexed with the S. typhi LPS coreS; there is no direct evidence, however, to suggest that this would alter the immunogenicity of the polysaccharide antigen. It is conceivable that Shigella antigens other than LPS may contribute in a critical way to vaccine efficacy. For example, the S. sonnei plasmid harboured by 5076-1C does not express the 'invasion plasmid antigens' (outer membrane proteins) which induce antibody responses after natural Shigella infections23'24. Finally, although lyophilization is necessary to address economic and practical considerations in vaccine delivery, it is a harsh treatment that can adversely affect the vaccine strain 25. Unfortunately, a trial assessing the protective efficacy of vaccine prepared by direct harvest from culture material (without lyophilization) was not conducted because of regulatory agency restrictions. Analysis of parameters other than protection against oral challenge suggested that the bivalent vaccine successfully delivered Shigella antigen to the GALT. Most interesting was the demonstration of circulating IgA ASC in all vaccinees (Figure I). This finding is indicative of stimulation of the intestinal immune system and perhaps of a local secretory IgA response against both S. sonnei and S. typhi somatic antigen xx. The lack of correlation of these ASC and protection against shigellosis may be a function of the magnitude of the ASC response. Preliminary data indicate that active Shigella infection in individuals subsequently demonstrated to have protective immunity (Table 2) stimulated significantly higher mean levels of IgA ASC than vaccination with non-protective lot 87-5-1 of bivalent vaccine (L. Van De Verg et al., in press). The immunizing capacity of natural illness due to S. sonnei was confirmed in the homologous strain rechallenge study, supporting the idea that the volunteer challenge model can provide a valid measure of vaccine efficacy. The difficulty of developing and testing S. typhi carrier based vaccines and the need for characterized markers of efficacy for Shigella vaccines is evident from these studies.
Discussion
Acknowledgements
The bivalent Salmonella typhi-Shigella sonnei 5076-1C hybrid is a prototype of vaccines using attenuated Salmonella strains to deliver heterologous antigens to the gut-associated lymphoid tissue (GALT) 2°-22. Initial tests with lots 2 and 5 of 5076-1C vaccine were promising (ref. 6, Table I), but subsequent lots prepared by using either a similar protocol (lots 8 and 14) or a modified procedure (lots SSVI 1 and SSVI 2) lacked efficacy against S. sonnei challenge (Table 1). Systematic comparison of protective and non-protective lots revealed a preponderance of attached flagella in the protective preparations. Nonetheless, a manufacturing protocol specifically designed to preserve flagellar structure did not improve the efficacy of the bivalent vaccine (lot 87-5-I, Table I). Although the initial success of 5076-1C validates the concept of S. typhi as an antigen carrier for oral immunization of humans, the subsequent failures documented in this communication show that more research needs to be carried out to make this approach practical and reliable. It is possible that the presentation of the heterologous somatic antigen in the 5076-1C vaccine may
The authors are grateful to the volunteers who participated in these studies and also to Dr Carol Tacket and Dr J. Glenn Morris for assistance during the clinical trials, Catherine Black and Ronald Grochowski for skilled nursing care, and Brenda Blodgett for help in recruiting the volunteers. This work was supported by the Department of the Army and the US Army Medical Research and Development Command research contract DAMD 17-88.C-8039 to MML.
356
Vaccine, Vol. 8, A u g u s t 1990
References 1
2 3
DuPont, H.L, Hornick, R.B., Snyder, M.J., Libonati, J.P., Formal, S.B. and Gangarosa, E.J. Immunity in shigellosis. I1. Protection induced by oral live vaccine or primary infection. J. Infect. Dis. 1972,125, 12 Mel, D.M., Terun, A.L. and Vuksic, L. Studies on vaccination against bacillary dysentery. 3. Effective oral immunization against Shigella flexneri 2a in a field trial. Buff. WHO 1965, 32, 647 Mel, D.M., Gangarosa, E.J. and Radaovanovic, M.D. Studies on vaccination against bacillary dysentery. 6. Protection of children by oral immunization with streptomycin-dependent Shigella strains. Bull. WHO 1971, 45, 457
B i v a l e n t S. t y p h i - S , sonnei vaccine: D.A. H e r r i n g t o n et al. 4
5
6
7
8
9
10
11 12
13
14
Formal, S.B., Baron, L.S., Kopecko, D.J., Washington, O., Powell, C. and Life, C.A. Construction of a potential bivalent vaccine strain: Introduction of Shigella sonnei Form I antigen genes into the gal E Salmonella typhi Ty 21a typhoid vaccine strain. Infect. Immun. 1981, 34, 746 Seid, R.C. Jr., Kopecko, D.J., Sadoff, J.C., Schneider, H., Baron, L.S. and Formal, S.B. Unusual lipopolysaccharide antigens of a Salmonella typhi oral vaccine expressing the Shigella sonnei Form I antigen. J. Biol. Chem. 1984, 259, 9028 Black, R.E., Levine, M.M., Clements, M.L., Losonsky, G., Herrington, D., Berman, S. and Formal, S~B. Prevention of shigellosis by a Salmonella typhi-Shigella sonnei bivalent vaccine. J. Infect. Dis. 1987, 155, 1260 Tramont, E.C., Chung, R., Berman, S., Keren, D., Kapfner, C. and Formal, S.B. Safety and antigenicity of typhoid-Shigella sonnei vaccine (strain 5076-1C). J. Infect. Dis. 1984, 149, 133 Formal, S.B. and Levine, M.M. Shigellosis. In: Bacterial Vaccines (Ed. Germanier, R.) Academic Press, Orlando, 1983, pp. 167-186 DuPont, H.L., Hornick, R.B., Snyder, M.J., Libonati, J.P., Formal, S.B. and Gangarosa, E.J. Immunity in shigellosis. I. Response of man to attenuated strains of Shigella. J. Infect. Dis. 1972, 125, 5 Levine, M.M., DuPont, H.L., Gangarosa, E.J., Hornick, R.B., Snyder, M.J., Libonati, J.P. et al. Shigellosis in custodial institutions. II. Clinica!, immunologic and bacteriologic response of institutionalized children to oral attenuated shigella vaccines. Am. J. Epidemiol. 1972, 96, 40 Forrest, B.D. Identification of an intestinal immune response using peripheral blood lymphocytes. Lancet 1988, I, 81 Kantele, A., Arvilommi, H. and Jokinen, I. Specific immunoglobulinsecreting human blood cells after peroral vaccination against Salmonella typhi. J. Infect. Dis. 1986, 153, 1126 Lycke, N., Hellstrom, U. and Holmgren, J. Circulating cholera antitoxin memory cells in the blood one year after oral cholera vaccinations in humans. Scand. J. Immunol. 1987, 26, 207 Levine, M.M., Black, R.E., Clements, M.L., Lanata, C., Sears, S., Honda, T. et al. Evaluation in humans of attenuated Vibrio cholerae El Tor Ogawa strain Texas Star-SR as a live oral vaccine. Infect. Immun. 1984, 43, 515
15
16
17
18
19
20
21
22 23
24
25
Levine, M.M., Kaper, J.B., Herrington, D.A., Losonsky, G., Morris, J.G., Ciements, ML. et al. Volunteer studies of deletion mutants of Vibrio cholerae O1 prepared by recombinant techniques. Infect. Immun. 1988, 56, 161 Czerkinsky, C.C., Nilsson, L.A., Nygren, H., Ouchterlony, O. and Tarkowski, A. A solid-phase enzyme-linked immunospot (ELISPOT) assay for enumeration of specific antibody-secreting cells. J. Immunol. Methods 1983, 65, 109 Sedgwick, J.D. and Holt, P.G. A solid phase immunoenzymatic technique for the enumeration of specific antibody-secreting cells. J. Immunol. Methods 1983, 57, 301 Kantele, A.M., Takanen, R. and Arvilommi, H. Immune response to acute diarrhea.seen as circulating antibody-secreting cells. J. Infect. Dis. 1988, 158, 1011 Basu, S., Tharanathan, R.N., Kontrohr, T. and Mayer, H. Chemical structure of the lipid A component of Plesiomonas shigelloides and its taxonomical significance. FEMS Microbiol. Lett. 1985, 28, 7 Maskell, D., Liew, F.Y., Sweeney, K., Dougan, G. and Hormaeche, C. Attenuated Salmonella typhimurium as live oral vaccines and carriers for delivering antigens to the secretory immune system. In: Vaccines 86. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, 1986, pp. 213-217 Curtiss, R. and Kelly, S.M. Salmonella typhimurium deletion mutants lacking adenylate cyclase and cyclic AMP receptor protein are avirulent and immunogenic. Infect. Immun. 1987, 55, 3035 Stccker, B.A.D. Auxotrophic Salmonella typhi as a live vaccine. Vaccine 1988, 6, 141 Hale, T.L., Sansonetti, P.J., Schad, P.A., Austin, S. and Formal, S.B. Characterization of virulence plasmids and plasmid-associated outer membrane proteins in Shigella flexneri, Shigella sonnei, and Escherichia coli. Infect. Immun. 1983, 40, 340 Hale, T.L., Oaks, E.V., Formal, S.B., Dinari, G. and Echeverria, P. Immune response to Shigella infections and to Shigella vaccines. In: Vaccines 86. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, 1986, pp. 81~35 Levine, M.M., DuPont, H.L., Hornick, R.B., Synder, M.J., Woodward, W., Gilman, R.H. and Libonati, J.P. Attenuated, streptomycindependent Salmonella typhi oral vaccine: Potential deleterious effects of lyophilization. J. Infect. Dis. 1976, 133, 424
Vaccine, Vol. 8, A u g u s t 1990
357
efficacy of an oral attenuated bivalent S. typhi/S, sonnei vaccine (strain 5076-1C) which consists of S. typhi strain Ty21a containing the 120-megadalton plasmid of S. sonnei and expressing both S. typhi and S. sonnei lipopolysaccharides (LPS) on its surface ~'5. As reported previously, two initial lots of bivalent S. sonnei-S, typhi vaccine 5076-1C, but not a third lot, conferred significant protection against clinical shigellosis due to S. sonnei in volunteers 6. In previous studies with 5076-1C 6'7, as with earlier studies with other vaccines 2,8-1° and with re-challenges 2, neither measurement of serum antibody nor of intestinal fluid slgA to Shigella 0 antigens was sensitive in detecting immune responses. Recent reports have described the detection in the peripheral blood of trafficking immunoglobulin-secreting cells several days after oral immunization. Preliminary results from studies of oral typhoid 1L ~2 and cholera ~3 vaccines suggest that this may be a technique that markedly increases the ability to detect specific immune responses following oral immunization. In the present report, clinical studies with four additional lots of bivalent vaccine 5076-1C are presented. In one of these studies, the specific IgA-secreting lymphocyte response was measured and shown to be sensitive in detecting immune responses to the vaccine. Finally, the capacity of the volunteer challenge model to detect protective immunity in vaccinees or controls with recent S. sonnei illness is demonstrated, thereby validating
Vaccine, Vol. 8, August 1990 353
Bivalent S. typhi-S, sonnei vaccine: D.A. Herrington et al.
the use of this model for conducting efficacy studies of candidate S. sonnei vaccines. Materials and methods
Vaccine preparation The vaccine strain, 5076-1C, was prepared by conjugal transfer of the form I S. sonnei plasmid into the attenuated S. typhi strain Ty21a as reported 4. A seed lot of vaccine strain was prepared and used for the production of all lots for clinical studies. As previously described 6 for lots 2, 5, and 8, lot 14 was prepared at the Department of Biologic Research, Walter Reed Army Institute of Research (Washington, DC, USA) by harvesting after growth on solid agar, dispensing into vials, and lyophilizing. Lots SSVI 1 and SSVI 2, and lot 87-5-I, were formulated at the Swiss Serum and Vaccine Institute, Berne, Switzerland where they were harvested after growth in liquid medium in a fermenter. The procedure used for lot 87-5-I is summarized as follows. Agar-grown bacteria which agglutinated strongly with antisera specific for S. sonnei I O antigen and S. typhi d (flagella) were used to inoculate a fermenter containing Brain Heart Infusion broth-Hycase SF medium (BHI-HFS), and the culture was grown for 18 h at 30°C. The cells were collected by continuous flow centrifugation under aseptic conditions. The cell pellet was suspended in sterile sucrose Hycase-SF ascorbic acid cryoprotective medium, poured onto sterile metal plates, frozen at -70°C and lyophilized under aseptic conditions. The lyophilized vaccine bacteria were mixed with lactose and dispensed into individual sealed aluminium foil packets containing 3 x 109 viable cells per packet. Packets were stored at 4°C. Several random packets were reconstituted with sterile distilled water, and electron microscopy demonstrated that the ceils were sparsely piliated and possessed intact and attached flagella. Viable counts and agglutinations with specific antisera were confirmed for several random packets on the day prior to each vaccination. Volunteers Adult volunteers between the ages of 18 and 35 years from the Baltimore metropolitan area participated in the studies. The methods of medical screening, care of the volunteers, and obtaining informed consent have been described ~4.15. Protocols were reviewed by the Human Volunteer Research Committee of the University of Maryland at Baltimore and the Human Use Review Office of the Surgeon General's Human Subjects Research Review Board of the Department of the Army. Administration o f vaccine Vaccine (approximately 2x 10 9 viable cells) was administered orally with 2g sodium bicarbonate mixed in 150ml of water for all lots with the exception of lot 87-5-I. For lot 87-5-1, a commercially prepared packet of powdered buffer (2.65g sodium bicarbonate, 1.65g ascorbic acid, 0.2 g lactose) and a packet of vaccine were stirred into 100 ml sterile distilled water. The vaccine was ingested by each volunteer on three occasions separated by three or four days. The volunteers fasted 90 min before and after ingestion of the vaccine and returned 24 and 48h after each dose for questioning about adverse reactions and to provide a stool specimen for culture.
354
Vaccine, Vol. 8, August 1990
Challenge studies The methods of preparing the inoculum of wild type S. sonnei strain 53G, medical care of the volunteers, and bacteriology have been previously described in detail 6. Multiple vials of a seed lot of strain 53G were stored at - 7 0 ° C at the initiation of studies, and challenge inocula were prepared from this frozen stock.
Immunological studies Sera were obtained from the volunteers before, and 7, 14, 21 and 28 days after, the first dose of vaccine. Serum IgG and IgA antibodies to the O antigen of S. sonnei were measured by ELISA 6. Fourfold increases in titre were considered significant. Heparinized blood was collected for solid-phase enzyme-linked immunospot (ELISPOT) assays (see below) 0, 9, 14 and 21 days after the first dose of vaccine; lymphocytes were separated by FicoU Hypaque and stored in vapour nitrogen. The ELISPOT assay 16-1s was used for enumeration of antibody-secreting cells (ASC) (mononuclear cells which secreted Shigella anti-LPS antibody). In brief, thawed lymphocytes were resuspended in RPMI 1640 culture medium (Flow Laboratories) supplemented with 10% heat-inactivated fetal calf serum, gentamicin (I 5 ttg ml- 1), and L-glutamine (3 mg ml- ~) and adjusted to a concentration of 2.5 x 106 viable cells per ml. Cells were incubated for 3h at 37°C in the antigen-coated wells of flat-bottomed 96-well microtitre plates (Immunoplate I; Nunc). Specific IgA antibodies secreted by the cells were detected by addition of goat anti-human IgA (Kirkegaard Perry Laboratories) at a dilution of 1: 500 followed by a substrate-agarose overlay and microscopic counting of the coloured spots. Antigens used were Plesiomonas shigelloides LPS (kindly provided by Dr All Lindberg, Stockholm, Sweden), which is immunologically and biochemically identical to S. sonnei LPS t9 but technically easier to prepare since S. sonnei readily becomes rough in culture, and S. typhi LPS (Difco). Antigens were used at a concentration of 10/~g ml-1, and Vibrio cholerae Inaba LPS (Sigma) and buffer were used as negative controls.
Results
Clinical tolerance, vaccine shedding, immunologic response to vaccination and vaccine efficacy Table 1 summarizes clinical trials with seven separate lots of vaccine. Results for three of these vaccine lots (2, 5, and 8) were reported previously6. Separate cohorts of volunteers with no previous history of shigellosis or enrolment in previous Shigella vaccine studies were utilized for each trial. Vaccine excretion rates, serologic responses to S. sonnei LPS following ingestion of lots 14, SSVI 1 and SSVI 2 were similar to results seen following administration of the protective lots 2 and 5 (Table I). The vaccine strains recovered from coproculture were all found to express S. sonnei antigen as determined by agglutination with specific antiserum. Nevertheless, as summarized in Table 1, lots 14, SSVI 1 and SSVI 2 failed to protect against experimental challenge with pathogenic S. sonnei strain 53(3.
Bivalent S. t y p h i - S , sonnei vaccine: D.A. Herrington et al. Table 1 Bacteriology, immunological response to, and protective efficacy of various lots of bivalent Salmonella typhi-Shigella sonnei vaccine 5076-1C administered to volunteers
Seroconversions by ELISA to S. sonnei LPS
Vaccine lot" Lots 2 and 5 (n =43) ° Lots 8 and 14 (n = 31) Lots SSVl 1 and SSVl 2 (n = 30) Lot 87-5-1 (n= 16)
Vaccine excretion (%)
IgA (%)
IgG (%)
Attack rate after challenge Vaccinees
Controls
p valu&
23
25
23
10/28 (36%)
20/29 (69%)
0.01
23
16
19
12/25 (48%)
8/17 (47%)
0.60
33
43
27
3/20 (9%)
5/18 (28%)
0.29
75
56
19
5/13 (38%)
5/12 (42%)
0.60
• Results for vaccine lots 2, 5, and 8 were reported previously (ref. 6); "p value was determined by using Fisher's exact test, 1 tail; °n=number of volunteers receiving vaccine
Efforts to improve vaccine formulation: immunogenicity o f lot 87-5-1 The apparent loss of vaccine efficacy in volunteers studied led to an intensive examination of the biochemical and morphological characteristics of the protective versus non-protective lots. These examinations were undertaken in an effort to identify the critical components of the vaccine strain and the steps in the manufacturing process which may have rendered the strain less effective. The most extensive testing that was carried out compared lot 5 (protective) and lot 8 (non-protective) organisms directly reconstituted from lyophilized vaccine; no differences were found between these lots with respect to the presence of lactose-fermenting organisms; agglutination patterns with Salmonella group D, Shigella group D, or S. sonnei form I antisera were similar; plasmid DNA restriction endonuclease digestion patterns also failed to show a difference; neither lot exhibited mannose-sensitive haemagglutination (due to type 1 somatic pill). However, a difference was found between the lots with respect to flagella. The S. typhi d (flagellar) agglutination reaction was fourfold higher for lot 5 than for non-protective lot 8 organisms. This difference was corroborated by electron microscopy; lot 8 had no attached, intact flagella, whereas intact flagella were readily found on lot 5 organisms. Electron microscopy of samples of non-protective lots 14, SSVI 1 and SSVI 2 following gentle rehydration with water also showed that flagella either were not present, were detached from the bacterial cells, or were detached and fragmented. By systematic changes in the manufacturing process, a new vaccine lot (87-5-I) was prepared which closely resembled the original protective lots biochemically, serologically and by electron microscopic visualization. The vaccination study with lot 87-5-I revealed that vaccine organisms were recovered in small numbers in coprocultures from 12 out of 16 volunteers; no volunteers manifested adverse reactions. Nine out of 16 volunteers (56%) had a fourfold or greater rise of serum IgA anti-LPS as measured by ELISA, while 3 out of 16 (19%) had significant rises in serum IgG. Thus vaccine excretion and serum immune response to lot 87-5-I were similar to those elicited by protective lots 2 and 5 and by non-protective lots 14, SSVI 1, and SSVI 2 (Table I). Because a correlation between IgA production by peripheral blood lymphocytes several days after oral immunization and the detection of specific intestinal sIgA
to
o= o
I_ 10
O3 t_ t~ >. O .D cr~
..~....
. ......"
...~ 5
Z
10 Time a f t e r
Figure 1
15
20
25
vaccination (days}
Mean numbers of mononuclear cells secreting IgA against
S. sonnei (P. shigelloides LPS, see text) or S. typhi LPS as detected in peripheral blood on days 0, 9, 14, and 21 after vaccination with S. typhUS. sonnei bivalent vaccine 5076-1C. O, P. shigelloides L P S ; / k , S. typhi LPS
has been reported x1,18, circulating lymphocytes were collected from the vaccinees on several occasions after vaccination with lot 87-5-I. Lymphocytes were available for analysis from 13 vaccinees. Antibody secreting cells (ASC) specific for P. shigelloides polysaccharide antigen were detected on day 9 in all 13 vaccinees and were no longer found on day 21 (Figure I ). Ten out of 13 vaccinees (77%) also had ASC specific for S. typhi LPS (Figure I), while none of the volunteers produced ASC against the control LPS from V. cholerae serotype Inaba (data not shown). These results indicate that the S. sonnei/S, typhi bivalent vaccine triggered an immune response to O antigen in 100% of the vaccinees tested, even though no protection was apparent in the volunteer model (see below).
Challenge study to assess vaccine efficacy
(lot 87-5-1)
One month after receiving the third dose of vaccine, 13 vaccinees and 12 unvaccinated controls were challenged with approximately 4 x 10 z smooth, form I colonies of S. sonnei strain 53G in an experimental study to assess
Vaccine, Vol. 8, A u g u s t 1990
355
Bivalent S. typhi-S, sonnei vaccine: D.A. Herrington et al. Table 2 Protective efficacy conferred by prior shigellosis due to S. sonnei as determined by experimental challenge in volunteers Protection against
Veterans
Controls
p value"
Protective efficacy (%)
Febrile illness (T> 37.8°C) Dysentery Diarrhoea b
0/6 1/6 0/6
6/12 8/12 7/12
0.05 0.13 0.04
100 78 100
'p value determined by Fisher's exact test, two tail; ~Diarrhoea defined as 1 stool of >300ml or 2 or more stools totalling >200ml within a 48 h period
the efficacy of the vaccine. Clinical shigellosis developed in 5 out of 13 vaccinees and 5 out of 12 controls (Table I). In this challenge study, there was no evidence of protection or of amelioration of disease due to the vaccine.
Challenge study to assess infection-derived immunity to S. sonnei In view of the variable efficacy of 5076-1C vaccine, we considered it critical to validate the ability of the volunteer model of experimental S. sonnei shigellosis to detect protective efficacy. To accomplish this goal we used the model to measure infection-derived immunity against S. sonnei. Accordingly, six of the ~,".unteers who experienced clinical shigellosis while participating in the vaccine challenge study (four of whom had previously received vaccine) were rechallenged two months later along with 12 naive controls. The data are summarized in Table 2. Only one of the six rechallenged volunteers met the definition of illness; this volunteer had a single dysenteric stool with no other symptoms. In contrast, 8 out of 12 of the control volunteers developed illness (diarrhoea 7 out of 12; fever 6 out of 12; dysentery 8 out of 12). While the possibility that previous vaccination of four of the rechallenged volunteers contributed to their subsequent protection cannot be excluded, the results nevertheless demonstrate the utility of the challenge model for detecting protective immunity.
not be optimal. Indeed, in this hybrid, the S. sonnei O-disaccharide repeat unit is not complexed with the S. typhi LPS coreS; there is no direct evidence, however, to suggest that this would alter the immunogenicity of the polysaccharide antigen. It is conceivable that Shigella antigens other than LPS may contribute in a critical way to vaccine efficacy. For example, the S. sonnei plasmid harboured by 5076-1C does not express the 'invasion plasmid antigens' (outer membrane proteins) which induce antibody responses after natural Shigella infections23'24. Finally, although lyophilization is necessary to address economic and practical considerations in vaccine delivery, it is a harsh treatment that can adversely affect the vaccine strain 25. Unfortunately, a trial assessing the protective efficacy of vaccine prepared by direct harvest from culture material (without lyophilization) was not conducted because of regulatory agency restrictions. Analysis of parameters other than protection against oral challenge suggested that the bivalent vaccine successfully delivered Shigella antigen to the GALT. Most interesting was the demonstration of circulating IgA ASC in all vaccinees (Figure I). This finding is indicative of stimulation of the intestinal immune system and perhaps of a local secretory IgA response against both S. sonnei and S. typhi somatic antigen xx. The lack of correlation of these ASC and protection against shigellosis may be a function of the magnitude of the ASC response. Preliminary data indicate that active Shigella infection in individuals subsequently demonstrated to have protective immunity (Table 2) stimulated significantly higher mean levels of IgA ASC than vaccination with non-protective lot 87-5-1 of bivalent vaccine (L. Van De Verg et al., in press). The immunizing capacity of natural illness due to S. sonnei was confirmed in the homologous strain rechallenge study, supporting the idea that the volunteer challenge model can provide a valid measure of vaccine efficacy. The difficulty of developing and testing S. typhi carrier based vaccines and the need for characterized markers of efficacy for Shigella vaccines is evident from these studies.
Discussion
Acknowledgements
The bivalent Salmonella typhi-Shigella sonnei 5076-1C hybrid is a prototype of vaccines using attenuated Salmonella strains to deliver heterologous antigens to the gut-associated lymphoid tissue (GALT) 2°-22. Initial tests with lots 2 and 5 of 5076-1C vaccine were promising (ref. 6, Table I), but subsequent lots prepared by using either a similar protocol (lots 8 and 14) or a modified procedure (lots SSVI 1 and SSVI 2) lacked efficacy against S. sonnei challenge (Table 1). Systematic comparison of protective and non-protective lots revealed a preponderance of attached flagella in the protective preparations. Nonetheless, a manufacturing protocol specifically designed to preserve flagellar structure did not improve the efficacy of the bivalent vaccine (lot 87-5-I, Table I). Although the initial success of 5076-1C validates the concept of S. typhi as an antigen carrier for oral immunization of humans, the subsequent failures documented in this communication show that more research needs to be carried out to make this approach practical and reliable. It is possible that the presentation of the heterologous somatic antigen in the 5076-1C vaccine may
The authors are grateful to the volunteers who participated in these studies and also to Dr Carol Tacket and Dr J. Glenn Morris for assistance during the clinical trials, Catherine Black and Ronald Grochowski for skilled nursing care, and Brenda Blodgett for help in recruiting the volunteers. This work was supported by the Department of the Army and the US Army Medical Research and Development Command research contract DAMD 17-88.C-8039 to MML.
356
Vaccine, Vol. 8, A u g u s t 1990
References 1
2 3
DuPont, H.L, Hornick, R.B., Snyder, M.J., Libonati, J.P., Formal, S.B. and Gangarosa, E.J. Immunity in shigellosis. I1. Protection induced by oral live vaccine or primary infection. J. Infect. Dis. 1972,125, 12 Mel, D.M., Terun, A.L. and Vuksic, L. Studies on vaccination against bacillary dysentery. 3. Effective oral immunization against Shigella flexneri 2a in a field trial. Buff. WHO 1965, 32, 647 Mel, D.M., Gangarosa, E.J. and Radaovanovic, M.D. Studies on vaccination against bacillary dysentery. 6. Protection of children by oral immunization with streptomycin-dependent Shigella strains. Bull. WHO 1971, 45, 457
B i v a l e n t S. t y p h i - S , sonnei vaccine: D.A. H e r r i n g t o n et al. 4
5
6
7
8
9
10
11 12
13
14
Formal, S.B., Baron, L.S., Kopecko, D.J., Washington, O., Powell, C. and Life, C.A. Construction of a potential bivalent vaccine strain: Introduction of Shigella sonnei Form I antigen genes into the gal E Salmonella typhi Ty 21a typhoid vaccine strain. Infect. Immun. 1981, 34, 746 Seid, R.C. Jr., Kopecko, D.J., Sadoff, J.C., Schneider, H., Baron, L.S. and Formal, S.B. Unusual lipopolysaccharide antigens of a Salmonella typhi oral vaccine expressing the Shigella sonnei Form I antigen. J. Biol. Chem. 1984, 259, 9028 Black, R.E., Levine, M.M., Clements, M.L., Losonsky, G., Herrington, D., Berman, S. and Formal, S~B. Prevention of shigellosis by a Salmonella typhi-Shigella sonnei bivalent vaccine. J. Infect. Dis. 1987, 155, 1260 Tramont, E.C., Chung, R., Berman, S., Keren, D., Kapfner, C. and Formal, S.B. Safety and antigenicity of typhoid-Shigella sonnei vaccine (strain 5076-1C). J. Infect. Dis. 1984, 149, 133 Formal, S.B. and Levine, M.M. Shigellosis. In: Bacterial Vaccines (Ed. Germanier, R.) Academic Press, Orlando, 1983, pp. 167-186 DuPont, H.L., Hornick, R.B., Snyder, M.J., Libonati, J.P., Formal, S.B. and Gangarosa, E.J. Immunity in shigellosis. I. Response of man to attenuated strains of Shigella. J. Infect. Dis. 1972, 125, 5 Levine, M.M., DuPont, H.L., Gangarosa, E.J., Hornick, R.B., Snyder, M.J., Libonati, J.P. et al. Shigellosis in custodial institutions. II. Clinica!, immunologic and bacteriologic response of institutionalized children to oral attenuated shigella vaccines. Am. J. Epidemiol. 1972, 96, 40 Forrest, B.D. Identification of an intestinal immune response using peripheral blood lymphocytes. Lancet 1988, I, 81 Kantele, A., Arvilommi, H. and Jokinen, I. Specific immunoglobulinsecreting human blood cells after peroral vaccination against Salmonella typhi. J. Infect. Dis. 1986, 153, 1126 Lycke, N., Hellstrom, U. and Holmgren, J. Circulating cholera antitoxin memory cells in the blood one year after oral cholera vaccinations in humans. Scand. J. Immunol. 1987, 26, 207 Levine, M.M., Black, R.E., Clements, M.L., Lanata, C., Sears, S., Honda, T. et al. Evaluation in humans of attenuated Vibrio cholerae El Tor Ogawa strain Texas Star-SR as a live oral vaccine. Infect. Immun. 1984, 43, 515
15
16
17
18
19
20
21
22 23
24
25
Levine, M.M., Kaper, J.B., Herrington, D.A., Losonsky, G., Morris, J.G., Ciements, ML. et al. Volunteer studies of deletion mutants of Vibrio cholerae O1 prepared by recombinant techniques. Infect. Immun. 1988, 56, 161 Czerkinsky, C.C., Nilsson, L.A., Nygren, H., Ouchterlony, O. and Tarkowski, A. A solid-phase enzyme-linked immunospot (ELISPOT) assay for enumeration of specific antibody-secreting cells. J. Immunol. Methods 1983, 65, 109 Sedgwick, J.D. and Holt, P.G. A solid phase immunoenzymatic technique for the enumeration of specific antibody-secreting cells. J. Immunol. Methods 1983, 57, 301 Kantele, A.M., Takanen, R. and Arvilommi, H. Immune response to acute diarrhea.seen as circulating antibody-secreting cells. J. Infect. Dis. 1988, 158, 1011 Basu, S., Tharanathan, R.N., Kontrohr, T. and Mayer, H. Chemical structure of the lipid A component of Plesiomonas shigelloides and its taxonomical significance. FEMS Microbiol. Lett. 1985, 28, 7 Maskell, D., Liew, F.Y., Sweeney, K., Dougan, G. and Hormaeche, C. Attenuated Salmonella typhimurium as live oral vaccines and carriers for delivering antigens to the secretory immune system. In: Vaccines 86. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, 1986, pp. 213-217 Curtiss, R. and Kelly, S.M. Salmonella typhimurium deletion mutants lacking adenylate cyclase and cyclic AMP receptor protein are avirulent and immunogenic. Infect. Immun. 1987, 55, 3035 Stccker, B.A.D. Auxotrophic Salmonella typhi as a live vaccine. Vaccine 1988, 6, 141 Hale, T.L., Sansonetti, P.J., Schad, P.A., Austin, S. and Formal, S.B. Characterization of virulence plasmids and plasmid-associated outer membrane proteins in Shigella flexneri, Shigella sonnei, and Escherichia coli. Infect. Immun. 1983, 40, 340 Hale, T.L., Oaks, E.V., Formal, S.B., Dinari, G. and Echeverria, P. Immune response to Shigella infections and to Shigella vaccines. In: Vaccines 86. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, 1986, pp. 81~35 Levine, M.M., DuPont, H.L., Hornick, R.B., Synder, M.J., Woodward, W., Gilman, R.H. and Libonati, J.P. Attenuated, streptomycindependent Salmonella typhi oral vaccine: Potential deleterious effects of lyophilization. J. Infect. Dis. 1976, 133, 424
Vaccine, Vol. 8, A u g u s t 1990
357
