THE JOURNAL OF INFECTIOUS DISEASES. VOL. 137. NO.6. JUNE 1978 © 1978 by the University of Chicago. 0022-1899178/3706-0003$01.09
Safety and Immunogenicity of a Neisseria meningitidis Type 2 Protein Vaccine in Animals and Humans W. D. Zollinger, R. E. Mandrell, P. Altieri, S. Berman, J. Lowenthal, and M. S. Artenstein*
From the Departments of Bacterial Diseases and Biologics Research, Walter Reed A rmy Institute of Research, Washington, D.C.
Two Neisseria meningitidis vaccines consisting principally of outer membrane protein (lot 1381-0) or outer membrane protein plus group C polysaccharide (lot 1381MI) were prepared from the group C type 2 strain 1381. Lipopolysaccharide and lipid were removed by gel filtration in the presence of sodium deoxycholate. The vaccines were found to be nontoxic and non pyrogenic in animals. They provided active protection in mice against mucin-enhanced killing by group B type 2 meningococci and induced good titers of type-specific bactericidal and hemagglutinating antibodies in rabbits. In five volunteers the vaccines were well tolerated and induced significant increases in serum bactericidal activity against both group C and group B strains. Three of five volunteers had a two- to fourfold increase in antibodies to the outer membrane proteins, but these antibodies did not appear to have bactericidal activity. The bactericidal antibodies to both group B and group C strains were directed against the capsular polysaccharides.
The success of the polysaccharide vaccines of Neisseria meningitidis groups A and C [1-3] leaves group B as the only major pathogenic serogroup of meningococci for which a vaccine is not available. Although the group B polysaccharide (2-'8-a-linked N-acetyl neuraminic acid) is chemically very similar to group C polysaccharide (2-'9-a-linked N-acetyl, O-acetyl neuraminic acid) [4], purified high-molecular-weight B polysaccharide has been found to be a poor immunogen in humans [5, 6]. Even natural infections with group B meningococci do not appear to result in a strong antibody response to the group B polysaccharide [7, 8]. In addition, Kasper et al. [9] found that antibodies to the B poly-
saccharide were apparently not bactericidal for all group B strains. These observations have led us to investigate the potential of the serotype protein antigens of the meningococcal outer membrane as possible immunizing agents against group B meningococcal disease. Humans have been shown to produce HA antibodies to the serotype proteins as a consequence of nasopharyngeal carriage and systemic infections with group Band C meningococci [7]. In addition, Kasper et al. [9] found that the specificity of bactericidal antibodies in convalescentphase sera from patients with group B disease correlated better with serotype than with serogroup. This finding suggested the presence of type-specific bactericidal antibodies in these sera. Although the outer membrane protein antigens have type specificity rather than group specificity, a single serotype (type 2 or II) has been associated with a high percentage of strains causing epidemics of group Band C meningococcal disease [10-12]. Thus, a vaccine consisting of the type 2 protein could conceivably provide protection against most group B strains with epidemic potential. In previous studies, serotype proteins extracted from the group B strain 99M using relatively harsh procedures were tested for safety and im-
Received for publication May 23. 1977. and in revised form December 27.1977. The authors thank MSG Adam Druzd, Ms. Brenda Brandt, Mr. Hans Hansel. and Mr. Calvin Powell for their assistance. The importance of prior unpublished studies of Dr. Dennis L. Kasper on a meningococcal protein vaccine is also acknowledged. In conducting the research described in this report. the investigators adhered to the Guide for Care and Use of Laboratory Animal Resources, National Academy of Sciences. National Research Council. Please address requests for reprints to Dr. W. D. Zollinger, Department of Bacterial Diseases, Walter Reed Army Institute of Research, Washington, D.C. 20012. .. Deceased. March 9.1976.
728
Meningococcal Type 2 Protein Vaccine
munogenicity in animals and humans and were found to be safe and immunogenic in animals but not immunogenic in humans (D. L. Kasper, W. D. Zollinger, P. Altieri, and M. S. Artenstein, unpublished observations). Development of milder and more effective methods of isolating serotype proteins essentially free of lipopolysaccharide (LPS) and nucleic acid [7] provided the incentive for the present study. A group C type 2 strain of N. meningitidis, 1381, was used as the vaccine strain because the type 2 protein is common to both serogroups Band C [II], and it was of interest to investigate the possibility of using a combination of outer membrane protein and group C polysaccharide as a means of protecting against both group B and group C disease in recruits and of increasing the immunogenicity of the group C polysaccharide in children. In the present study two vaccine lots, 1381-0 (consisting principally of outer membrane proteins) and 1381-Ml (consisting of both outer membrane proteins and group C polysaccharide), were prepared and tested for safety and immunogenicity in animals and in humans. Materials and Methods
Bacterial strains and growth conditions. Meningococcal group B strains M986, M982, MI080, MIOII, M136, and Bl6B6 were provided by Dr. Carl Frasch (Bureau of Biologics, Food and Drug Administration, Bethesda, Md.). All remaining meningococcal strains were from the culture collection of the Department of Bacterial Diseases, Walter Reed Army Institute of Research, Washington, D.C. Escherichia coli strain 07:Kl(L):NM [6] was used for production of Kl capsular polysaccharide. Stock cultures were preserved in the lyophilized state. Serogroup and serotype information, when given, appears in parentheses following the strain designation and has the format (serogroup:. protein serotype), e.g., 99M(B: P2,3,6). The protein serotypes were determined by the solid-phase radioimmunoassay inhibition method [13]. The protein (P) serotypes refer to a comprehensive serotyping system which consists of antisera made against 23 different meningococcal strains from serogroups A, B, and C including the serotyping strains of Frasch and
729
Chapman [14] and of Gold and Wyle [IS], and additional strains characterized in our laboratory. Serotype P2 is the same as the type 2 of Frasch and Chapman and the factor II of Gold and Wyle. Media and growth conditions, except for preparation of the vaccines, were as previously described [7]. Preparation of vaccines. The meningococcal strain 1381 (C:P2) was used as the vaccine strain. Two IS-liter cultures were grown as previously described [16] for 16 hr at 37 C using the casamino acid medium of Watson and Sherp [17], modified to contain four times the prescribed concentration of sodium phosphate [18]. The cultures were killed by addition of 90% phenol to a final concentration of 0.5% after which the organisms were collected by continuous flow centrifugation and processed as previously described [7] to obtain the outer membrane complex from which vaccine lot 1381-0 was prepared. The culture medium supernatant was concentrated to 800 ml on an ultrafiltration apparatus equipped with PM-30 diaflow membranes (Amicon Corp., Lexington, Mass.) and then processed as previously described [19] to obtain the outer membrane complex from which vaccine lot 1381-Ml was prepared. The outer membrane complex obtained from the culture medium supernatant contained substantially more capsular polysaccharide than the outer membrane complex stripped off the organisms. The vaccine was prepared from the preparations of outer membrane complex by separation of the protein (lot 1381-0) or protein plus capsular polysaccharide (lot 1381-Ml) components from the complex. Lipid and LPS were removed from the outer membrane complex by gel filtration on Sephadex G-100 in 0.5% sodium deoxycholate (DOC), 1 mM EDTA, and 0.05 M glycine (pH 10). Nine parts of outer membrane complex suspended at -5 mg of proteinjrnl in distilled water was mixed with one part buffer concentrate containing 10% DOC, 0.01 M EDTA, and 0.5 M glycine (pH 10). The pH was then adjusted to II.5 with I N NaOH, and the solution was centrifuged at 100,000 g for 2 hr at 4 C. The small pellets were discarded, and the supernatant was immediately applied to the Sephadex G-IOO column (5 em X 90 em). Fractions were assayed for protein, 2-keto-3-
730
deoxyoctonate (KDO), and sialic acid, and those fractions containing the protein and sialic acid were pooled. The sample was recovered from the DOC buffer by acid (pH 5) precipitation and washing with ethanol (lot 1381-0) or by dialysis followed by precipitation and washing with ethanol (lot l381-Ml) as described previously [7]. The washed precipitates were suspended in distilled water and dialyzed against three or four changes of distilled water. Lot l381-Ml was frozen, and lot 1381-0 was further processed to remove the small amount of capsular polysaccharide. The sample was mixed with an equal volume of 50% trichloroacetic acid and kept at room temperature (about 24 C) for 1 hr. The precipitate was then collected by centrifugation at 2,000 g for 10 min, washed twice with distilled water and once. with ethanol, and dissolved in distilled water by addition of N aO H to pH 11.5. The pH was then adjusted to 10.0 with HCl, and the solution was dialyzed against three changes of distilled water. For filter sterilization the preparations were adjusted to pH 11 with 2 N NaOH and filtered through a Millipore filter with 0.22-pm pores (Millipore Corp., Bedford, Mass.). After filtration the pH was readjusted to 7 by addition of a calculated amount of sterile 2 N HC1. The final sterile products were bottled, lyophilized, and stored at -20 C. Chemical and analytical methods. Assays for protein, sialic acid, KDO, and nucleic acid were performed as previously described [7]. Sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis was performed as previously described [19], with use of an acrylamide concentration of7.5%. Serologic investigations. Several serologic methods were used to study antibody responses in animals and volunteers. Indirect HA assays specific for capsular polysaccharide, outer membrane protein, or LPS, and HA inhibition assays were performed as previously described [7]. Both the radioactive bactericidal test described by Kasper and Wyle [20] and the standard bactericidal test [21] were used. In the radioactive test, approximately 107 organisms per tube were employed, and ~25% of maximal release of radioactivity was considered significant killing, whereas in the standard test about 5 X 103 or-
Zollinger et al,
ganisms per tube were used, and ~50% reduction in colonies was considered significant killing. The titers given refer to the dilution of antiserum added to the assay mixture. In both cases serum from rabbits younger than four weeks old without intrinsic bactericidal activity toward the test strain was used as a source of complement. Antibodies to the outer membrane serotype antigens were measured by a solid-phase radioimmunoassay (SPRIA) performed as previously described [22]. The radioactive antigen-binding assay (RABA) was performed as described by Brandt et al. [8]. Based on replicate tests of a single serum [22], the 95% confidence limits for the SPRIA were estimated to be ±20% of the reported IJ-g of antibody jml. Thus a ~40% increase in antibody was considered significant at the 95% confidence level. Similarly, the 95% confidence limits for titers determined in the HA or bactericidal assays were estimated to be ±15% of the titer when expressed as the number of twofold dilutions, i.e., 10g2 (reciprocal of serum dilution). Since additional nonrandom variations can occur from day to day, all sera were routinely tested against a given antigen or organism in a single assay. Preparation of antigens. Antigens used in the serologic investigations, except for the highmolecular-weight group B polysaccharides, were prepared as previously described [7]. The procedure used to prepare outer membrane protein was essentially the same as that used to prepare vaccine lot 1381-0. High-molecular-weight group B polysaccharide and E. coli Kl capsular polysaccharide were prepared by a modification of the method of Gotschlich et al. [23] from cultures grown on the casamino acid medium of Watson and Sherp [17] modified to contain 3% certified casamino acids (Difco, Detroit, Mich.) and 0.1% glucose. In this medium the final pH was >7.0, which prevented acid breakdown of the polysaccharide. The polysaccharide was precipitated from the culture medium supernatant with 0.1% cetyltrimethylammonium bromide, dissolved in 0.9 M CaC1 2 , and fractionated with ethanol to remove nucleic acid as described by Gotschlich et al. [23]. After ethanol fractionation and three ethanol washes, the polysaccharide was suspended in distilled water and centrifuged at 100,000 g for 2 hr
Meningococcal Type 2 Protein Vaccine
at 4 C. Ten percent Emulphogene BC 720 and I M potassium phosphate (pH 8.0) were added to the clear supernatant to a final concentration of 2% (vol rvol) and 0.01 M, respectively. The solution was then twice subjected to ion exchange chromatography on DEAE Sepharose CL-6B equilibrated with 110 Emulphogene BC 720 and 0.01 M potassium phosphate (pH 8.0). The polysaccharide was eluted with a gradient of 0-0.6 M NaCl in starting buffer and recovered by precipitation and washing with ethanol. The final products contained
Safety and Immunogenicity of a Neisseria meningitidis Type 2 Protein Vaccine in Animals and Humans W. D. Zollinger, R. E. Mandrell, P. Altieri, S. Berman, J. Lowenthal, and M. S. Artenstein*
From the Departments of Bacterial Diseases and Biologics Research, Walter Reed A rmy Institute of Research, Washington, D.C.
Two Neisseria meningitidis vaccines consisting principally of outer membrane protein (lot 1381-0) or outer membrane protein plus group C polysaccharide (lot 1381MI) were prepared from the group C type 2 strain 1381. Lipopolysaccharide and lipid were removed by gel filtration in the presence of sodium deoxycholate. The vaccines were found to be nontoxic and non pyrogenic in animals. They provided active protection in mice against mucin-enhanced killing by group B type 2 meningococci and induced good titers of type-specific bactericidal and hemagglutinating antibodies in rabbits. In five volunteers the vaccines were well tolerated and induced significant increases in serum bactericidal activity against both group C and group B strains. Three of five volunteers had a two- to fourfold increase in antibodies to the outer membrane proteins, but these antibodies did not appear to have bactericidal activity. The bactericidal antibodies to both group B and group C strains were directed against the capsular polysaccharides.
The success of the polysaccharide vaccines of Neisseria meningitidis groups A and C [1-3] leaves group B as the only major pathogenic serogroup of meningococci for which a vaccine is not available. Although the group B polysaccharide (2-'8-a-linked N-acetyl neuraminic acid) is chemically very similar to group C polysaccharide (2-'9-a-linked N-acetyl, O-acetyl neuraminic acid) [4], purified high-molecular-weight B polysaccharide has been found to be a poor immunogen in humans [5, 6]. Even natural infections with group B meningococci do not appear to result in a strong antibody response to the group B polysaccharide [7, 8]. In addition, Kasper et al. [9] found that antibodies to the B poly-
saccharide were apparently not bactericidal for all group B strains. These observations have led us to investigate the potential of the serotype protein antigens of the meningococcal outer membrane as possible immunizing agents against group B meningococcal disease. Humans have been shown to produce HA antibodies to the serotype proteins as a consequence of nasopharyngeal carriage and systemic infections with group Band C meningococci [7]. In addition, Kasper et al. [9] found that the specificity of bactericidal antibodies in convalescentphase sera from patients with group B disease correlated better with serotype than with serogroup. This finding suggested the presence of type-specific bactericidal antibodies in these sera. Although the outer membrane protein antigens have type specificity rather than group specificity, a single serotype (type 2 or II) has been associated with a high percentage of strains causing epidemics of group Band C meningococcal disease [10-12]. Thus, a vaccine consisting of the type 2 protein could conceivably provide protection against most group B strains with epidemic potential. In previous studies, serotype proteins extracted from the group B strain 99M using relatively harsh procedures were tested for safety and im-
Received for publication May 23. 1977. and in revised form December 27.1977. The authors thank MSG Adam Druzd, Ms. Brenda Brandt, Mr. Hans Hansel. and Mr. Calvin Powell for their assistance. The importance of prior unpublished studies of Dr. Dennis L. Kasper on a meningococcal protein vaccine is also acknowledged. In conducting the research described in this report. the investigators adhered to the Guide for Care and Use of Laboratory Animal Resources, National Academy of Sciences. National Research Council. Please address requests for reprints to Dr. W. D. Zollinger, Department of Bacterial Diseases, Walter Reed Army Institute of Research, Washington, D.C. 20012. .. Deceased. March 9.1976.
728
Meningococcal Type 2 Protein Vaccine
munogenicity in animals and humans and were found to be safe and immunogenic in animals but not immunogenic in humans (D. L. Kasper, W. D. Zollinger, P. Altieri, and M. S. Artenstein, unpublished observations). Development of milder and more effective methods of isolating serotype proteins essentially free of lipopolysaccharide (LPS) and nucleic acid [7] provided the incentive for the present study. A group C type 2 strain of N. meningitidis, 1381, was used as the vaccine strain because the type 2 protein is common to both serogroups Band C [II], and it was of interest to investigate the possibility of using a combination of outer membrane protein and group C polysaccharide as a means of protecting against both group B and group C disease in recruits and of increasing the immunogenicity of the group C polysaccharide in children. In the present study two vaccine lots, 1381-0 (consisting principally of outer membrane proteins) and 1381-Ml (consisting of both outer membrane proteins and group C polysaccharide), were prepared and tested for safety and immunogenicity in animals and in humans. Materials and Methods
Bacterial strains and growth conditions. Meningococcal group B strains M986, M982, MI080, MIOII, M136, and Bl6B6 were provided by Dr. Carl Frasch (Bureau of Biologics, Food and Drug Administration, Bethesda, Md.). All remaining meningococcal strains were from the culture collection of the Department of Bacterial Diseases, Walter Reed Army Institute of Research, Washington, D.C. Escherichia coli strain 07:Kl(L):NM [6] was used for production of Kl capsular polysaccharide. Stock cultures were preserved in the lyophilized state. Serogroup and serotype information, when given, appears in parentheses following the strain designation and has the format (serogroup:. protein serotype), e.g., 99M(B: P2,3,6). The protein serotypes were determined by the solid-phase radioimmunoassay inhibition method [13]. The protein (P) serotypes refer to a comprehensive serotyping system which consists of antisera made against 23 different meningococcal strains from serogroups A, B, and C including the serotyping strains of Frasch and
729
Chapman [14] and of Gold and Wyle [IS], and additional strains characterized in our laboratory. Serotype P2 is the same as the type 2 of Frasch and Chapman and the factor II of Gold and Wyle. Media and growth conditions, except for preparation of the vaccines, were as previously described [7]. Preparation of vaccines. The meningococcal strain 1381 (C:P2) was used as the vaccine strain. Two IS-liter cultures were grown as previously described [16] for 16 hr at 37 C using the casamino acid medium of Watson and Sherp [17], modified to contain four times the prescribed concentration of sodium phosphate [18]. The cultures were killed by addition of 90% phenol to a final concentration of 0.5% after which the organisms were collected by continuous flow centrifugation and processed as previously described [7] to obtain the outer membrane complex from which vaccine lot 1381-0 was prepared. The culture medium supernatant was concentrated to 800 ml on an ultrafiltration apparatus equipped with PM-30 diaflow membranes (Amicon Corp., Lexington, Mass.) and then processed as previously described [19] to obtain the outer membrane complex from which vaccine lot 1381-Ml was prepared. The outer membrane complex obtained from the culture medium supernatant contained substantially more capsular polysaccharide than the outer membrane complex stripped off the organisms. The vaccine was prepared from the preparations of outer membrane complex by separation of the protein (lot 1381-0) or protein plus capsular polysaccharide (lot 1381-Ml) components from the complex. Lipid and LPS were removed from the outer membrane complex by gel filtration on Sephadex G-100 in 0.5% sodium deoxycholate (DOC), 1 mM EDTA, and 0.05 M glycine (pH 10). Nine parts of outer membrane complex suspended at -5 mg of proteinjrnl in distilled water was mixed with one part buffer concentrate containing 10% DOC, 0.01 M EDTA, and 0.5 M glycine (pH 10). The pH was then adjusted to II.5 with I N NaOH, and the solution was centrifuged at 100,000 g for 2 hr at 4 C. The small pellets were discarded, and the supernatant was immediately applied to the Sephadex G-IOO column (5 em X 90 em). Fractions were assayed for protein, 2-keto-3-
730
deoxyoctonate (KDO), and sialic acid, and those fractions containing the protein and sialic acid were pooled. The sample was recovered from the DOC buffer by acid (pH 5) precipitation and washing with ethanol (lot 1381-0) or by dialysis followed by precipitation and washing with ethanol (lot l381-Ml) as described previously [7]. The washed precipitates were suspended in distilled water and dialyzed against three or four changes of distilled water. Lot l381-Ml was frozen, and lot 1381-0 was further processed to remove the small amount of capsular polysaccharide. The sample was mixed with an equal volume of 50% trichloroacetic acid and kept at room temperature (about 24 C) for 1 hr. The precipitate was then collected by centrifugation at 2,000 g for 10 min, washed twice with distilled water and once. with ethanol, and dissolved in distilled water by addition of N aO H to pH 11.5. The pH was then adjusted to 10.0 with HCl, and the solution was dialyzed against three changes of distilled water. For filter sterilization the preparations were adjusted to pH 11 with 2 N NaOH and filtered through a Millipore filter with 0.22-pm pores (Millipore Corp., Bedford, Mass.). After filtration the pH was readjusted to 7 by addition of a calculated amount of sterile 2 N HC1. The final sterile products were bottled, lyophilized, and stored at -20 C. Chemical and analytical methods. Assays for protein, sialic acid, KDO, and nucleic acid were performed as previously described [7]. Sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis was performed as previously described [19], with use of an acrylamide concentration of7.5%. Serologic investigations. Several serologic methods were used to study antibody responses in animals and volunteers. Indirect HA assays specific for capsular polysaccharide, outer membrane protein, or LPS, and HA inhibition assays were performed as previously described [7]. Both the radioactive bactericidal test described by Kasper and Wyle [20] and the standard bactericidal test [21] were used. In the radioactive test, approximately 107 organisms per tube were employed, and ~25% of maximal release of radioactivity was considered significant killing, whereas in the standard test about 5 X 103 or-
Zollinger et al,
ganisms per tube were used, and ~50% reduction in colonies was considered significant killing. The titers given refer to the dilution of antiserum added to the assay mixture. In both cases serum from rabbits younger than four weeks old without intrinsic bactericidal activity toward the test strain was used as a source of complement. Antibodies to the outer membrane serotype antigens were measured by a solid-phase radioimmunoassay (SPRIA) performed as previously described [22]. The radioactive antigen-binding assay (RABA) was performed as described by Brandt et al. [8]. Based on replicate tests of a single serum [22], the 95% confidence limits for the SPRIA were estimated to be ±20% of the reported IJ-g of antibody jml. Thus a ~40% increase in antibody was considered significant at the 95% confidence level. Similarly, the 95% confidence limits for titers determined in the HA or bactericidal assays were estimated to be ±15% of the titer when expressed as the number of twofold dilutions, i.e., 10g2 (reciprocal of serum dilution). Since additional nonrandom variations can occur from day to day, all sera were routinely tested against a given antigen or organism in a single assay. Preparation of antigens. Antigens used in the serologic investigations, except for the highmolecular-weight group B polysaccharides, were prepared as previously described [7]. The procedure used to prepare outer membrane protein was essentially the same as that used to prepare vaccine lot 1381-0. High-molecular-weight group B polysaccharide and E. coli Kl capsular polysaccharide were prepared by a modification of the method of Gotschlich et al. [23] from cultures grown on the casamino acid medium of Watson and Sherp [17] modified to contain 3% certified casamino acids (Difco, Detroit, Mich.) and 0.1% glucose. In this medium the final pH was >7.0, which prevented acid breakdown of the polysaccharide. The polysaccharide was precipitated from the culture medium supernatant with 0.1% cetyltrimethylammonium bromide, dissolved in 0.9 M CaC1 2 , and fractionated with ethanol to remove nucleic acid as described by Gotschlich et al. [23]. After ethanol fractionation and three ethanol washes, the polysaccharide was suspended in distilled water and centrifuged at 100,000 g for 2 hr
Meningococcal Type 2 Protein Vaccine
at 4 C. Ten percent Emulphogene BC 720 and I M potassium phosphate (pH 8.0) were added to the clear supernatant to a final concentration of 2% (vol rvol) and 0.01 M, respectively. The solution was then twice subjected to ion exchange chromatography on DEAE Sepharose CL-6B equilibrated with 110 Emulphogene BC 720 and 0.01 M potassium phosphate (pH 8.0). The polysaccharide was eluted with a gradient of 0-0.6 M NaCl in starting buffer and recovered by precipitation and washing with ethanol. The final products contained
