Vol. 58, No. 7
INFECTION AND IMMUNITY, JUlY 1990, p. 2367-2374
0019-9567/90/072367-08$02.00/0
Copyright X 1990, American Society for Microbiology
Synthesis and Immunologic Properties in Mice of Vaccines Composed of Staphylococcus aureus Type 5 and Type 8 Capsular Polysaccharides Conjugated to Pseudomonas aeruginosa Exotoxin A SCHNEERSON,l SHOUSUN C. SZU,l WILLIE F. VANN,2 JOSEPH SHILOACH,3 WALTER W. KARAKAWA,4 AND JOHN B. ROBBINS' National Institute of Child Health and Human Development,' Division of Bacterial Products, Office of Biologics Research and Review, Food and Drug Administration,2 and National Institute of Diabetes and Digestive and Kidney Diseases,3 Bethesda, Maryland 20892, and Department of Molecular and Cellular Biology, Pennsylvania State University, University Park, Pennsylvania 168024 Received 18 December 1989/Accepted 18 April 1990 ALI FATTOM,1* RACHEL
Epidemiological, serological and in vitro phagocytosis experiments provide evidence that the newly discovered type 5 and type 8 capsular polysaccharides (CPs) are both virulence factors and protective antigens for bacteremia caused by Staphylococcus aureus. Neither type 5 nor type 8 CP elicited serum antibodies when injected into mice. These two CPs were bound to Pseudomonas aeruginosa exotoxin A (ETA) to form conjugates by using the synthetic scheme devised for the CP (Vi) of SalmoneUa typhi and of pneumococcus type 12F (A. Fattom, W. F. Vann, S. C. Szu, A. Sutton, X. Li, D. Bryla, G. Schiffman, J. B. Robbins, and R. Schneerson, Infect. Immun. 56:2292-2298, 1988; S. C. Szu, A. L. Stone, J. D. Robbins, R. Schneerson, and J. B. Robbins, J. Exp. Med. 166:1510-1524, 1987). Both S. aureus CP-ETA conjugates elicited a rise in CP antibodies. As components of conjugates, both S. aureus CPs acquired T-cell-dependent properties, as shown by their ability to respond to carrier priming and to stimulate booster responses. The conjugate-induced antibodies facilitated type-specific opsonization of S. aureus by human polymorphonuclear leukocytes. The conjugates also induced ETA antibodies which neutralized the native toxin in vitro. Clinical studies of these two conjugates for active or passive immunization of patients at risk for S. aureus bacteremia are planned. akawa and Vann) (27, 36). (ii) Optimal synthesis of the type 5 and type 8 CPs requires cultivation of these strains in media low in phosphate (24, 25). (iii) S. aureus CPs contain an aminouronic acid which interferes with their colorimetric assay (16, 18, 20, 38, 52). (iv) Modified serologic techniques were required to identify their CPs (25). Surveillance in the United States, Israel, France, Switzerland, and Germany showed that strains of type 5 and 8 CPs (11 recognized types exist) compose -80% of the isolates from bacteremia (3, 6, 16, 17, 21, 25, 27, 39, 49). As is the case with pneumococci, meningococci, and Haemophilus influenzae, only some of the many CP types within S. aureus are associated with invasive infections (3, 27). In vitro studies showed that type 5 and 8 CPs shield S. aureus from agglutination by antibodies to cell wall structures (24). Capsulated strains of types 5 and 8 are poorly phagocytized by polymorphonuclear leukocytes (PMNs) (26). Addition of type-specific polyclonal antiserum or murine monoclonal CP antibodies facilitates type-specific opsonization by PMNs. Noncapsulated S. aureus strains, in contrast, are agglutinated by cell wall antiserum and phagocytized rapidly by PMNs alone. The structures of types 5 and 8 were elucidated (Fig. 1) (17, 18; M. Moreau, J. C. Richards, J.-M. Fournier, R. A. Byrd, W. W. Karakawa, and W. F. Vann, Carbohydr. Res., in press). The composition of S. aureus CP types 5 and 8 is identical: both have a ManNAcA in their repeat unit that can be used to introduce a sulfhydryl group. Slight cross-reactivity between these CPs is demonstrable with unabsorbed hyperimmune antisera (25, 27). Their comparatively low molecular sizes allow us to predict that these two CPs will be
Staphylococcus aureus causes several diseases by different pathogenic mechanisms. The most frequent and serious of these diseases are bacteremia and its complications in hospitalized patients. Despite the intensive use and development of new antibiotics, S. aureus remains a major cause of bacteremia with a high mortality; this pathogen accounts for about one-third of bacteremic isolates from hospitalized patients in the United States (11, 19, 35, 40, 48, 54; Editorial, Lancet i:953-954, 1986). Further, outbreaks of bacteremia due to S. aureus, some caused by organisms resistant to methicillin, continue to occur (7, 16). Curiously, the pathogenesis and mechanisms of protective immunity to S. aureus bacteremia remain obscure (29, 37, 45, 48; Editorial, Lancet). The discovery of S. aureus capsular polysaccharides (CPs) associated with bacteremia provided a new insight into the pathogenesis of and protective immunity to bacteremia caused by this pathogen (3, 16, 21, 25-28, 49). These CPs have unusual properties which may account for the fact that this information emerged only recently. (i) S. aureus isolates from the blood of patients do not have the mucoid appearance usually associated with capsulation; widely studied capsulated S. aureus strains, especially the mucoid strain Smith, were not associated with bacteremia in humans (20, 27, 31, 32, 37, 38, 41, 42, 52, 53). The amount of CP synthesized by bacteremic isolates is comparatively low; incubation of these strains with India ink does not demonstrate the CP as it does the large capsule surrounding the Smith strain (type 2 according to the classification of Kar*
Corresponding author. 2367
2368
FATTOM ET AL.
Type 5:
INFECT. IMMUN.
-->4)- -D-ManNAcAp(1-->4)-3)-f-D-FucNAcp(13 A UAc
Type 8:
3) -J-D-FucNAcp( 1-->3)-#-D-ManNAcAp(1-->3)4
I
OAc FIG. 1. Repeat units of the S. aureus type 5 and type 8 CPs (17, 18; Moreau et al., in press).
poor immunogens in humans, especially in patients with reduced resistance (15, 47). Conjugating these CPs to proteins, as has been done for other polysaccharides, would be expected to enhance their immunogenicity both for active immunization and for preparing high-titered antisera in volunteers for passive immunization (13, 14, 46, 47, 51). We used the synthetic scheme developed for binding the CPs of Salmonella typhi (Vi) and of the pneumococcus type 12F to proteins for preparing conjugates of the S. aureus type 5 and type 8 CPs (8, 14, 51). Both S. aureus CPs have an 0-acetyl moiety which is sensitive to hydrolysis at high pH. Accordingly, the reactions for preparing conjugates of these CPs were devised for neutral or slightly low pH. Pseudomonas aeruginosa exotoxin A (ETA) was chosen as a carrier because its structure and biological activities have been characterized and because it has been suggested as a protective antigen for an important, and serologically distinct, hospital pathogen (5, 43, 44). We now report the synthesis, characterization, and immunologic properties in mice of type 5 and type 8 ETA conjugates as a first step in evaluating this approach for immunologic control of S. aureus bacteremia.
MATERIALS AND METHODS Chemicals. lodoacetic acid, dithiothreitol (DTT), thimerosal, bovine serum albumin (BSA), 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDAC), lysostaphin, avidin, biotin, Brij 35, n-nitrophenyl phosphate, and heparin were from Sigma Chemical Co., St. Louis, Mo. Cystamine dihydrochloride and cysteamine were from Fluka AG, Buchs, Switzerland. N-Succinimidyl 3-(2-pyridyldithio)propionate (SPDP) and bicinchoninic acid were from Pierce Chemical Co., Rockford, Ill. Sephacryl S-300, DEAE-Sephacel, CL6B Sepharose, and dextran T-250 were from Pharmacia Fine Chemicals, Piscataway, N.J. Bio-Gel P-6DG was from BioRad Laboratories, Richmond, Calif. Columbia broth was from Difco Laboratories, Detroit, Mich. Pyrogen-free sterile water and pyrogen-free sterile saline from Travenol Laboratories, Morton Grove, Ill., were used throughout the study. Goat anti-mouse immunoglobulins (immunoglobulin M [IgM], IgG, and IgA)-alkaline phosphatase conjugates were purchased from Kirkegaard & Perry Laboratories, Inc., Gaithersburg, Md. P. aeruginosa 103 was kindly donated by Stephen Leppla, National Institute of Dental Research, Bethesda, Md. ETA and goat anti-ETA were from List Laboratories, San Francisco, Calif. ETA was from the Swiss Serum Institute, Bern, Switzerland. RPMI 1640 medium was from Microbiological Associates, Bethesda, Md. Fetal calf serum (FCS) was from GIBCO Laboratories, Grand Island, N.Y.
Bacterial growth. S. aureus Lowenstein (type 5) and Wright (type 8) were cultivated in Columbia broth supplemented with 2% NaCl in a 100-liter fermentor at 37°C (16, 17, 25). The cells were grown for 18 to 22 h to an A56 of 18 for type 8 or 12 for type 5. At this stage, the cells were maximally agglutinable with their specific antisera and yielded little or no detectable agglutination with cell wall antiserum (25). Phenol-ethanol (1:1, vol/vol) was added to the fermentor to a final concentration of 2% and mixed slowly for 4 h at room temperature (no viable cells were detected after this treatment). The cells were then harvested by centrifugation at 25,000 x g and stored at -25°C until use. P. aeruginosa 103 was cultivated as described elsewhere (5). Purification of type 5 and type 8 CP. The cells were suspended at 0.5 g (wet weight) per ml in 0.05 M Tris-2 mM MgSO4, pH 7.5 (17, 18). Lysostaphin (100 to 150 ,ug/ml) was added and incubated at 37°C for 4 h with mixing. Thereafter, DNase and RNase were added to final concentrations of 40 ,ug/ml each, and the incubation was continued for an additional 2 h. The mixture was pelleted by centrifugation at 25,000 x g for 30 min, the supematant was transferred to dialysis tubing to which 80 pug each of DNase and RNase were added per ml, and the reaction mixture was dialyzed against the Tris buffer for 3 h at 37°C with a drop of toluene. Protease (0.5 mg/ml) was added, the outer fluid was replaced with fresh buffer, and dialysis was continued at 37°C for 3 h. The reaction mixture was filtered through a 0.45-,um-poresize membrane (Nalge/Sybron Corp., Rochester, N.Y.) and precipitated sequentially with 25 and 75% ethanol in the presence of 5 mM CaCl2. The 75% ethanol precipitate was dialyzed extensively against water at 3 to 8°C and freezedried. The powder was dissolved in 0.05 M sodium acetate0.1 M NaCl (pH 6.0) for type 5 or 0.05 M sodium acetate0.05 M NaCI for type 8. The CP was applied to a DEAESephacel column (2.5 by 50 cm) equilibrated in the same buffer. After the column had been washed with -5 column volumes of the starting buffer, the CP was eluted with 0.15 M NaCl in the sodium acetate buffer. Fractions which contained CP, detected by capillary precipitation, were pooled, dialyzed, and freeze-dried. The powder was dissolved in 0.2 M NaCl and applied to a Sephacryl S-300 column (2.5 by 90 cm) in 0.2 M NaCl. Fractions which contained the CP and no detectable cell wall antigens by capillary precipitation were pooled, dialyzed against water, and freeze-dried. P. aeruginosa ETA. ETA was purified from the culture supematant of P. aeruginosa 103 as described elsewhere (5; J. Shiloach, M. van de Walle, J. B. Kaufman, T. R. Clem, and R. Fass, in P. L. Yu, ed., Fermentation Technologies: Industrial Applications, in press). Two preparations of ETA
VOL. 58, 1990
used. The ADP ribosyltransferase activity, mouse lethality, and CHO cell cytotoxicity assays were performed as described previously (22). Thiolation of type 5 and type 8 CP. CPs were dissolved in water at 5 to 10 mg/ml, and cystamine was added to a final concentration of 0.25 M. The pH was adjusted to 4.8 with 0.1 N HCl, and the temperature was maintained at 37°C. EDAC were
was added to 50 mM, and the pH was maintained at 4.8 with 0.1 N HCI until the solution was stable (-2 h). The mixture was dialyzed against water at 3 to 8°C, freeze-dried, and stored at -20°C. A sample of the CP-cystamine was reduced with 100 mM DTT for 1 h at room temperature and desalted on a Bio-Gel P-6DG column. The material eluted in the void volume was then assayed for its SH content (51). Derivatization of ETA with SPDP. Free SH groups were modified by treatment with 0.01 M iodoacetic acid (in PBS) for 1 h and dialysis against phosphate-buffered saline (PBS) at 3 to 8°C for 3 days. To a 13.9-mg/ml solution of ETA, SPDP was added (40 mM in ethanol) to a final concentration of 4 mM, and the mixture was stirred for 1 h. The reaction mixture was passed through a Bio-Gel P-6DG column (2.5 by 40 cm) in PBS. The void volume fractions containing ETA were pooled, concentrated by vacuum dialysis, and stored at 3 to 8°C. The bound SPDP was estimated after reduction of
a
sample with 40 mM DTT and measurement of the A343,
using an extinction coefficient of 8.08 x 103 M1 cm-1 for pyridine-2-thione (8). Synthesis of type 5 and type 8 ETA conjugates. Cystaminederivatized type 5 or type 8 CP was dissolved in saline at 5 to 10 mg/ml. DTT was added to a final concentration of 100 mM, and the mixture was stirred for 1 h at room temperature. After dialysis against saline for 1 h at room temperature, the reaction mixture was passed through a Bio-Gel P-6DG column (2.5 by 40 cm) in PBS-0.002 M EDTA. The CP was collected in the void volume and concentrated on a membrane (Amicon Corp., Lexington, Mass.) under argon to -10 mg/ml. SPDP-derivatized ETA was added to the CP-cystamine at an estimated ratio of 1:1 (SPDP:SH groups) under argon and incubated at 3 to 8°C for 48 h. The reaction mixture was then passed through a Sephacryl S-300 column (2.5 by 90 cm) in 0.2 M NaCl. Void volume fractions, which contained both CP and ETA as shown by capillary precipitation, were pooled, adjusted to a concentration of 0.01% thimerosal, and stored at 3 to 8°C. Analysis. SH groups were assayed by the Ellman method with cysteamine as a standard (51). Nucleic acids were measured by A260, and protein was assayed with bicinchoninic acid with BSA as a standard (14). Sodium dodecyl sulfate-polyacrylamide gel electrophoresis was performed as described elsewhere with commercially obtained ETA as a standard (30). The effluent was monitored with a refractometer. S. aureus type 5 and 8 CPs cannot be assayed by colorimetric methods used for other polysaccharides. A Bio-Rad TSK 400 column and a high-performance liquid chromatography apparatus (Waters Associates, Inc., Milford, Mass.) were used to measure the CP in the type 5 CP-ETA conjugate. A sample of the conjugate was reduced with 200 mM DTT for 1 h and injected into the TSK 400 column. The area under the peak of the CP curve was measured, and the concentration was determined with the type 5 CP as a reference standard. Reduction of the type 8 CP-ETA conjugate was incomplete. Accordingly, it was dialyzed against water and freeze-dried, and its dry weight was measured. The type 8 CP content was calculated by subtracting its protein content from the total weight. Gel filtration, through either CL-6B Sepharose or Sephacryl
S. AUREUS TYPE 5 AND 8 CP CONJUGATES
2369
S-300, and the calculation of the Kd values were performed as described previously (14). In vitro phagocytosis. The phagocytosis assay was performed as described elsewhere (26). Heparinized blood from a donor was sedimented for 45 min at room temperature in 3% dextran T-250. The PMNs were centrifuged at 160 x g for 5 min, washed with saline containing 100 [l1 of heparin per ml, and suspended to a concentration 1.0 x 107 cells per ml in RPMI 1640 medium containing 5% heat-inactivated FCS. Cell viability was assayed by trypan blue staining. All assays were performed in sterile screw-cap siliconized glass tubes (13 by 100 mm). S. aureus was grown in Columbia broth supplemented with 2% NaCl for 18 h at 37°C, harvested by centrifugation, washed with saline, and adjusted to 2.0 x 107 CFU/ml of PBS. The reaction mixture contained in 1.0 ml 106 S. aureus cells, 106 PMNs, and 0.1 ml of the test serum. Controls included sera from mice immunized with the heterologous S. aureus CP conjugate or from unimmunized mice. The tubes were tumbled slowly on a rotating rack (Scientific Industries, Inc., Springfield, Mass.) at 37°C. At 0, 60, and 120 min, 100-pd samples were removed and added to 9.9 ml of water. Twofold dilutions of this mixture were made in water, and 100-pul samples were spread onto Columbia salt agar plates in triplicate. Colonies were counted after overnight incubation at 37°C. Immunization. Groups of 10 female, 18- to 20-g general purpose mice, were injected subcutaneously one, two, or three times, 2 weeks apart with 2.5 ,ug of CP alone or CP conjugated to ETA. The groups were exsanguinated 2 weeks after the first injection or 1 week after the second and third injections. For priming, mice were injected with 2 ,ug of ETA-SPDP derivative (equal to the amount of ETA in the conjugate). Two weeks later, they were injected with the conjugate, and they were bled after 1 more week (14, 46). Reference antisera against each CP type and ETA were prepared by injecting 10 BALB/c mice subcutaneously three times 2 weeks apart with 5 ,ug of the appropriate conjugate. The mice were bled 1 week after the last injection, and their sera were pooled and assigned a value of 100 enzyme-linked immunosorbent assay (ELISA) units for the CP and the ETA. Serologic studies. Rabbit type 5, type 8, and cell wall antisera were prepared as described elsewhere (25). Murine serum type 5 and type 8 antibodies were measured by ELISA. Immulon 1 plates (Dynatech Laboratories, Inc., Chantilly, Va.) were coated with 4 ,ug of avidin per ml in PBS, incubated overnight at room temperature, and washed with 0.1% Brij 35-saline. Biotinylated CPs were prepared as described previously (50). The biotin-CPs (4.0 ,ug/ml in PBS) were added to the plates and incubated at room temperature for 6 to 8 h. The plates were washed with 0.1% Brij 35-saline and then blocked with 0.5% BSA in PBS for 1 h at room temperature. Twofold dilutions of the sera were made in 1% BSA-0.1% Brij 35 in PBS and incubated at room temperature overnight. The plates were washed 10 times, goat anti-mouse immunoglobulins conjugated to alkaline phosphatase were added, and the plates were incubated at room temperature for 4 h. n-Nitrophenylphosphate (1.0 mg/ml in 1 M Tris hydrochloride buffer [pH 9.8] containing 0.3 mM MgSO4) was added, and the A407 was read after 30 min in an MR600 microplate reader (Dynatech). ETA antibodies were measured by ELISA; Immulon 1 plates were coated with 4.0 ,ug of ETA per ml in sodium carbonate buffer (pH 9.2) overnight at room temperature. Thereafter, the procedure was the same as for the CP antibodies.
2370
FATTOM ET AL.
INFECT. IMMUN.
TABLE 1. Composition of S. aureus type 5 and type 8 CPs and their conjugates with P. aeruginosa ETA CP
type
Nucleic acid'
5
p,
200 compared with the unconjugated ETA preparation 2. No deaths occurred in mice injected with 25 ,ug of type 8 CP-ETA per mouse (-20 ,ug of ETA per dose of conjugate). Serum antibodies elicited by type 5 or 8 CP alone or conjugated to ETA. Three injections of either CP alone did not elicit a rise in CP antibodies (Tables 3 and 4). Identical results were obtained with doses of 0.05, 0.5, and 25 ,ug of S. aureus CP injected according to the same schedule (data not shown). The first injection of the conjugates elicited low levels of CP antibodies for both type 5 and type 8 (P = 0.001). Both conjugates elicited a CP booster response after the second injection (P = 0.001); the type 5 CP-ETA elicited a -30-fold rise, and the type 8 CP-ETA stimulated a -15fold rise. The third injection of type 8 CP-ETA induced a further increase in type 8 CP antibodies (P = 0.001); type 5 CP-ETA did not elicit a significant rise after the third injection.
Carrier priming. One injection of 2 ,ug of ETA per mouse (the amount in a dose of each conjugate) did not elicit a rise in levels of CP antibodies (data not shown). Injection of the ETA-primed mice with either conjugate showed an enhanced CP antibody response compared with injection of
unprimed animals (P = 0.001 for type 5 and P = 0.001 for type 8) (Tables 3 and 4). The priming effect was carrier specific, since BSA, a nonrelated protein, had no effect on the CP antibody response to the type 5 CP-ETA. ETA antibodies. One injection of the type 5 CP-ETA elicited a rise in ETA antibodies (P < 0.001); a second injection of type 5 CP-ETA elicited a booster response (P < 0.001) (Table 5). The rise in ETA antibodies which occurred after the third injection of this conjugate was not significant. Type 8 CP-ETA failed to elicit a significant rise in ETA antibodies after the first injection. The second and third injections of type 8 CP-ETA each elicited significant rises (P < 0.001). Mice primed with ETA responded to one injection of either conjugate with antibody levels comparable to those obtained after two injections. As expected, injection of either CP alone did not elicit ETA antibodies. Phagocytic activities of conjugate-induced antibodies. No phagocytosis of S. aureus type 5 strains by human PMNs occurred with added normal rabbit or mouse serum (Fig. 4). Both rabbit antisera (see Materials and Methods) and sera from mice immunized with the conjugates facilitated the internalization of the bacteria and their subsequent death, as indicated by the decline in viable-cell count within 120 min. The degree of phagocytosis facilitated by the rabbit and murine antisera was similar for both CP types (Fig. 4; data shown for type 5 only). Neither mouse nor rabbit type 5 antiserum facilitated phagocytosis of S. aureus type 8 organisms, and vice versa (data not shown). These results demonstrate that antibodies to the type 5 and type 8 CPs, derived from immunization with ETA conjugates of these antigens, are opsonic and type specific. Neutralization of ETA by conjugate-induced antibodies. The reference antisera (see Materials and Methods) prepared in BALB/c mice and assigned a value of 100 ELISA units were assayed for their neutralization of the cytotoxicity of ETA in CHO cells. The two commercial toxins failed to demonstrate cytotoxic activity against this cell line. Accord-
TABLE 4. Serum S. aureus type 8 CP antibodies elicited in mice by type 8 CP alone or conjugated to P. aeruginosa ETA' Type 8 CP antibody concn after injection no.b:
Immunogen
Saline Type 8 CP Type 8 CP-ETA ETA primed
1
INFECTION AND IMMUNITY, JUlY 1990, p. 2367-2374
0019-9567/90/072367-08$02.00/0
Copyright X 1990, American Society for Microbiology
Synthesis and Immunologic Properties in Mice of Vaccines Composed of Staphylococcus aureus Type 5 and Type 8 Capsular Polysaccharides Conjugated to Pseudomonas aeruginosa Exotoxin A SCHNEERSON,l SHOUSUN C. SZU,l WILLIE F. VANN,2 JOSEPH SHILOACH,3 WALTER W. KARAKAWA,4 AND JOHN B. ROBBINS' National Institute of Child Health and Human Development,' Division of Bacterial Products, Office of Biologics Research and Review, Food and Drug Administration,2 and National Institute of Diabetes and Digestive and Kidney Diseases,3 Bethesda, Maryland 20892, and Department of Molecular and Cellular Biology, Pennsylvania State University, University Park, Pennsylvania 168024 Received 18 December 1989/Accepted 18 April 1990 ALI FATTOM,1* RACHEL
Epidemiological, serological and in vitro phagocytosis experiments provide evidence that the newly discovered type 5 and type 8 capsular polysaccharides (CPs) are both virulence factors and protective antigens for bacteremia caused by Staphylococcus aureus. Neither type 5 nor type 8 CP elicited serum antibodies when injected into mice. These two CPs were bound to Pseudomonas aeruginosa exotoxin A (ETA) to form conjugates by using the synthetic scheme devised for the CP (Vi) of SalmoneUa typhi and of pneumococcus type 12F (A. Fattom, W. F. Vann, S. C. Szu, A. Sutton, X. Li, D. Bryla, G. Schiffman, J. B. Robbins, and R. Schneerson, Infect. Immun. 56:2292-2298, 1988; S. C. Szu, A. L. Stone, J. D. Robbins, R. Schneerson, and J. B. Robbins, J. Exp. Med. 166:1510-1524, 1987). Both S. aureus CP-ETA conjugates elicited a rise in CP antibodies. As components of conjugates, both S. aureus CPs acquired T-cell-dependent properties, as shown by their ability to respond to carrier priming and to stimulate booster responses. The conjugate-induced antibodies facilitated type-specific opsonization of S. aureus by human polymorphonuclear leukocytes. The conjugates also induced ETA antibodies which neutralized the native toxin in vitro. Clinical studies of these two conjugates for active or passive immunization of patients at risk for S. aureus bacteremia are planned. akawa and Vann) (27, 36). (ii) Optimal synthesis of the type 5 and type 8 CPs requires cultivation of these strains in media low in phosphate (24, 25). (iii) S. aureus CPs contain an aminouronic acid which interferes with their colorimetric assay (16, 18, 20, 38, 52). (iv) Modified serologic techniques were required to identify their CPs (25). Surveillance in the United States, Israel, France, Switzerland, and Germany showed that strains of type 5 and 8 CPs (11 recognized types exist) compose -80% of the isolates from bacteremia (3, 6, 16, 17, 21, 25, 27, 39, 49). As is the case with pneumococci, meningococci, and Haemophilus influenzae, only some of the many CP types within S. aureus are associated with invasive infections (3, 27). In vitro studies showed that type 5 and 8 CPs shield S. aureus from agglutination by antibodies to cell wall structures (24). Capsulated strains of types 5 and 8 are poorly phagocytized by polymorphonuclear leukocytes (PMNs) (26). Addition of type-specific polyclonal antiserum or murine monoclonal CP antibodies facilitates type-specific opsonization by PMNs. Noncapsulated S. aureus strains, in contrast, are agglutinated by cell wall antiserum and phagocytized rapidly by PMNs alone. The structures of types 5 and 8 were elucidated (Fig. 1) (17, 18; M. Moreau, J. C. Richards, J.-M. Fournier, R. A. Byrd, W. W. Karakawa, and W. F. Vann, Carbohydr. Res., in press). The composition of S. aureus CP types 5 and 8 is identical: both have a ManNAcA in their repeat unit that can be used to introduce a sulfhydryl group. Slight cross-reactivity between these CPs is demonstrable with unabsorbed hyperimmune antisera (25, 27). Their comparatively low molecular sizes allow us to predict that these two CPs will be
Staphylococcus aureus causes several diseases by different pathogenic mechanisms. The most frequent and serious of these diseases are bacteremia and its complications in hospitalized patients. Despite the intensive use and development of new antibiotics, S. aureus remains a major cause of bacteremia with a high mortality; this pathogen accounts for about one-third of bacteremic isolates from hospitalized patients in the United States (11, 19, 35, 40, 48, 54; Editorial, Lancet i:953-954, 1986). Further, outbreaks of bacteremia due to S. aureus, some caused by organisms resistant to methicillin, continue to occur (7, 16). Curiously, the pathogenesis and mechanisms of protective immunity to S. aureus bacteremia remain obscure (29, 37, 45, 48; Editorial, Lancet). The discovery of S. aureus capsular polysaccharides (CPs) associated with bacteremia provided a new insight into the pathogenesis of and protective immunity to bacteremia caused by this pathogen (3, 16, 21, 25-28, 49). These CPs have unusual properties which may account for the fact that this information emerged only recently. (i) S. aureus isolates from the blood of patients do not have the mucoid appearance usually associated with capsulation; widely studied capsulated S. aureus strains, especially the mucoid strain Smith, were not associated with bacteremia in humans (20, 27, 31, 32, 37, 38, 41, 42, 52, 53). The amount of CP synthesized by bacteremic isolates is comparatively low; incubation of these strains with India ink does not demonstrate the CP as it does the large capsule surrounding the Smith strain (type 2 according to the classification of Kar*
Corresponding author. 2367
2368
FATTOM ET AL.
Type 5:
INFECT. IMMUN.
-->4)- -D-ManNAcAp(1-->4)-3)-f-D-FucNAcp(13 A UAc
Type 8:
3) -J-D-FucNAcp( 1-->3)-#-D-ManNAcAp(1-->3)4
I
OAc FIG. 1. Repeat units of the S. aureus type 5 and type 8 CPs (17, 18; Moreau et al., in press).
poor immunogens in humans, especially in patients with reduced resistance (15, 47). Conjugating these CPs to proteins, as has been done for other polysaccharides, would be expected to enhance their immunogenicity both for active immunization and for preparing high-titered antisera in volunteers for passive immunization (13, 14, 46, 47, 51). We used the synthetic scheme developed for binding the CPs of Salmonella typhi (Vi) and of the pneumococcus type 12F to proteins for preparing conjugates of the S. aureus type 5 and type 8 CPs (8, 14, 51). Both S. aureus CPs have an 0-acetyl moiety which is sensitive to hydrolysis at high pH. Accordingly, the reactions for preparing conjugates of these CPs were devised for neutral or slightly low pH. Pseudomonas aeruginosa exotoxin A (ETA) was chosen as a carrier because its structure and biological activities have been characterized and because it has been suggested as a protective antigen for an important, and serologically distinct, hospital pathogen (5, 43, 44). We now report the synthesis, characterization, and immunologic properties in mice of type 5 and type 8 ETA conjugates as a first step in evaluating this approach for immunologic control of S. aureus bacteremia.
MATERIALS AND METHODS Chemicals. lodoacetic acid, dithiothreitol (DTT), thimerosal, bovine serum albumin (BSA), 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDAC), lysostaphin, avidin, biotin, Brij 35, n-nitrophenyl phosphate, and heparin were from Sigma Chemical Co., St. Louis, Mo. Cystamine dihydrochloride and cysteamine were from Fluka AG, Buchs, Switzerland. N-Succinimidyl 3-(2-pyridyldithio)propionate (SPDP) and bicinchoninic acid were from Pierce Chemical Co., Rockford, Ill. Sephacryl S-300, DEAE-Sephacel, CL6B Sepharose, and dextran T-250 were from Pharmacia Fine Chemicals, Piscataway, N.J. Bio-Gel P-6DG was from BioRad Laboratories, Richmond, Calif. Columbia broth was from Difco Laboratories, Detroit, Mich. Pyrogen-free sterile water and pyrogen-free sterile saline from Travenol Laboratories, Morton Grove, Ill., were used throughout the study. Goat anti-mouse immunoglobulins (immunoglobulin M [IgM], IgG, and IgA)-alkaline phosphatase conjugates were purchased from Kirkegaard & Perry Laboratories, Inc., Gaithersburg, Md. P. aeruginosa 103 was kindly donated by Stephen Leppla, National Institute of Dental Research, Bethesda, Md. ETA and goat anti-ETA were from List Laboratories, San Francisco, Calif. ETA was from the Swiss Serum Institute, Bern, Switzerland. RPMI 1640 medium was from Microbiological Associates, Bethesda, Md. Fetal calf serum (FCS) was from GIBCO Laboratories, Grand Island, N.Y.
Bacterial growth. S. aureus Lowenstein (type 5) and Wright (type 8) were cultivated in Columbia broth supplemented with 2% NaCl in a 100-liter fermentor at 37°C (16, 17, 25). The cells were grown for 18 to 22 h to an A56 of 18 for type 8 or 12 for type 5. At this stage, the cells were maximally agglutinable with their specific antisera and yielded little or no detectable agglutination with cell wall antiserum (25). Phenol-ethanol (1:1, vol/vol) was added to the fermentor to a final concentration of 2% and mixed slowly for 4 h at room temperature (no viable cells were detected after this treatment). The cells were then harvested by centrifugation at 25,000 x g and stored at -25°C until use. P. aeruginosa 103 was cultivated as described elsewhere (5). Purification of type 5 and type 8 CP. The cells were suspended at 0.5 g (wet weight) per ml in 0.05 M Tris-2 mM MgSO4, pH 7.5 (17, 18). Lysostaphin (100 to 150 ,ug/ml) was added and incubated at 37°C for 4 h with mixing. Thereafter, DNase and RNase were added to final concentrations of 40 ,ug/ml each, and the incubation was continued for an additional 2 h. The mixture was pelleted by centrifugation at 25,000 x g for 30 min, the supematant was transferred to dialysis tubing to which 80 pug each of DNase and RNase were added per ml, and the reaction mixture was dialyzed against the Tris buffer for 3 h at 37°C with a drop of toluene. Protease (0.5 mg/ml) was added, the outer fluid was replaced with fresh buffer, and dialysis was continued at 37°C for 3 h. The reaction mixture was filtered through a 0.45-,um-poresize membrane (Nalge/Sybron Corp., Rochester, N.Y.) and precipitated sequentially with 25 and 75% ethanol in the presence of 5 mM CaCl2. The 75% ethanol precipitate was dialyzed extensively against water at 3 to 8°C and freezedried. The powder was dissolved in 0.05 M sodium acetate0.1 M NaCl (pH 6.0) for type 5 or 0.05 M sodium acetate0.05 M NaCI for type 8. The CP was applied to a DEAESephacel column (2.5 by 50 cm) equilibrated in the same buffer. After the column had been washed with -5 column volumes of the starting buffer, the CP was eluted with 0.15 M NaCl in the sodium acetate buffer. Fractions which contained CP, detected by capillary precipitation, were pooled, dialyzed, and freeze-dried. The powder was dissolved in 0.2 M NaCl and applied to a Sephacryl S-300 column (2.5 by 90 cm) in 0.2 M NaCl. Fractions which contained the CP and no detectable cell wall antigens by capillary precipitation were pooled, dialyzed against water, and freeze-dried. P. aeruginosa ETA. ETA was purified from the culture supematant of P. aeruginosa 103 as described elsewhere (5; J. Shiloach, M. van de Walle, J. B. Kaufman, T. R. Clem, and R. Fass, in P. L. Yu, ed., Fermentation Technologies: Industrial Applications, in press). Two preparations of ETA
VOL. 58, 1990
used. The ADP ribosyltransferase activity, mouse lethality, and CHO cell cytotoxicity assays were performed as described previously (22). Thiolation of type 5 and type 8 CP. CPs were dissolved in water at 5 to 10 mg/ml, and cystamine was added to a final concentration of 0.25 M. The pH was adjusted to 4.8 with 0.1 N HCl, and the temperature was maintained at 37°C. EDAC were
was added to 50 mM, and the pH was maintained at 4.8 with 0.1 N HCI until the solution was stable (-2 h). The mixture was dialyzed against water at 3 to 8°C, freeze-dried, and stored at -20°C. A sample of the CP-cystamine was reduced with 100 mM DTT for 1 h at room temperature and desalted on a Bio-Gel P-6DG column. The material eluted in the void volume was then assayed for its SH content (51). Derivatization of ETA with SPDP. Free SH groups were modified by treatment with 0.01 M iodoacetic acid (in PBS) for 1 h and dialysis against phosphate-buffered saline (PBS) at 3 to 8°C for 3 days. To a 13.9-mg/ml solution of ETA, SPDP was added (40 mM in ethanol) to a final concentration of 4 mM, and the mixture was stirred for 1 h. The reaction mixture was passed through a Bio-Gel P-6DG column (2.5 by 40 cm) in PBS. The void volume fractions containing ETA were pooled, concentrated by vacuum dialysis, and stored at 3 to 8°C. The bound SPDP was estimated after reduction of
a
sample with 40 mM DTT and measurement of the A343,
using an extinction coefficient of 8.08 x 103 M1 cm-1 for pyridine-2-thione (8). Synthesis of type 5 and type 8 ETA conjugates. Cystaminederivatized type 5 or type 8 CP was dissolved in saline at 5 to 10 mg/ml. DTT was added to a final concentration of 100 mM, and the mixture was stirred for 1 h at room temperature. After dialysis against saline for 1 h at room temperature, the reaction mixture was passed through a Bio-Gel P-6DG column (2.5 by 40 cm) in PBS-0.002 M EDTA. The CP was collected in the void volume and concentrated on a membrane (Amicon Corp., Lexington, Mass.) under argon to -10 mg/ml. SPDP-derivatized ETA was added to the CP-cystamine at an estimated ratio of 1:1 (SPDP:SH groups) under argon and incubated at 3 to 8°C for 48 h. The reaction mixture was then passed through a Sephacryl S-300 column (2.5 by 90 cm) in 0.2 M NaCl. Void volume fractions, which contained both CP and ETA as shown by capillary precipitation, were pooled, adjusted to a concentration of 0.01% thimerosal, and stored at 3 to 8°C. Analysis. SH groups were assayed by the Ellman method with cysteamine as a standard (51). Nucleic acids were measured by A260, and protein was assayed with bicinchoninic acid with BSA as a standard (14). Sodium dodecyl sulfate-polyacrylamide gel electrophoresis was performed as described elsewhere with commercially obtained ETA as a standard (30). The effluent was monitored with a refractometer. S. aureus type 5 and 8 CPs cannot be assayed by colorimetric methods used for other polysaccharides. A Bio-Rad TSK 400 column and a high-performance liquid chromatography apparatus (Waters Associates, Inc., Milford, Mass.) were used to measure the CP in the type 5 CP-ETA conjugate. A sample of the conjugate was reduced with 200 mM DTT for 1 h and injected into the TSK 400 column. The area under the peak of the CP curve was measured, and the concentration was determined with the type 5 CP as a reference standard. Reduction of the type 8 CP-ETA conjugate was incomplete. Accordingly, it was dialyzed against water and freeze-dried, and its dry weight was measured. The type 8 CP content was calculated by subtracting its protein content from the total weight. Gel filtration, through either CL-6B Sepharose or Sephacryl
S. AUREUS TYPE 5 AND 8 CP CONJUGATES
2369
S-300, and the calculation of the Kd values were performed as described previously (14). In vitro phagocytosis. The phagocytosis assay was performed as described elsewhere (26). Heparinized blood from a donor was sedimented for 45 min at room temperature in 3% dextran T-250. The PMNs were centrifuged at 160 x g for 5 min, washed with saline containing 100 [l1 of heparin per ml, and suspended to a concentration 1.0 x 107 cells per ml in RPMI 1640 medium containing 5% heat-inactivated FCS. Cell viability was assayed by trypan blue staining. All assays were performed in sterile screw-cap siliconized glass tubes (13 by 100 mm). S. aureus was grown in Columbia broth supplemented with 2% NaCl for 18 h at 37°C, harvested by centrifugation, washed with saline, and adjusted to 2.0 x 107 CFU/ml of PBS. The reaction mixture contained in 1.0 ml 106 S. aureus cells, 106 PMNs, and 0.1 ml of the test serum. Controls included sera from mice immunized with the heterologous S. aureus CP conjugate or from unimmunized mice. The tubes were tumbled slowly on a rotating rack (Scientific Industries, Inc., Springfield, Mass.) at 37°C. At 0, 60, and 120 min, 100-pd samples were removed and added to 9.9 ml of water. Twofold dilutions of this mixture were made in water, and 100-pul samples were spread onto Columbia salt agar plates in triplicate. Colonies were counted after overnight incubation at 37°C. Immunization. Groups of 10 female, 18- to 20-g general purpose mice, were injected subcutaneously one, two, or three times, 2 weeks apart with 2.5 ,ug of CP alone or CP conjugated to ETA. The groups were exsanguinated 2 weeks after the first injection or 1 week after the second and third injections. For priming, mice were injected with 2 ,ug of ETA-SPDP derivative (equal to the amount of ETA in the conjugate). Two weeks later, they were injected with the conjugate, and they were bled after 1 more week (14, 46). Reference antisera against each CP type and ETA were prepared by injecting 10 BALB/c mice subcutaneously three times 2 weeks apart with 5 ,ug of the appropriate conjugate. The mice were bled 1 week after the last injection, and their sera were pooled and assigned a value of 100 enzyme-linked immunosorbent assay (ELISA) units for the CP and the ETA. Serologic studies. Rabbit type 5, type 8, and cell wall antisera were prepared as described elsewhere (25). Murine serum type 5 and type 8 antibodies were measured by ELISA. Immulon 1 plates (Dynatech Laboratories, Inc., Chantilly, Va.) were coated with 4 ,ug of avidin per ml in PBS, incubated overnight at room temperature, and washed with 0.1% Brij 35-saline. Biotinylated CPs were prepared as described previously (50). The biotin-CPs (4.0 ,ug/ml in PBS) were added to the plates and incubated at room temperature for 6 to 8 h. The plates were washed with 0.1% Brij 35-saline and then blocked with 0.5% BSA in PBS for 1 h at room temperature. Twofold dilutions of the sera were made in 1% BSA-0.1% Brij 35 in PBS and incubated at room temperature overnight. The plates were washed 10 times, goat anti-mouse immunoglobulins conjugated to alkaline phosphatase were added, and the plates were incubated at room temperature for 4 h. n-Nitrophenylphosphate (1.0 mg/ml in 1 M Tris hydrochloride buffer [pH 9.8] containing 0.3 mM MgSO4) was added, and the A407 was read after 30 min in an MR600 microplate reader (Dynatech). ETA antibodies were measured by ELISA; Immulon 1 plates were coated with 4.0 ,ug of ETA per ml in sodium carbonate buffer (pH 9.2) overnight at room temperature. Thereafter, the procedure was the same as for the CP antibodies.
2370
FATTOM ET AL.
INFECT. IMMUN.
TABLE 1. Composition of S. aureus type 5 and type 8 CPs and their conjugates with P. aeruginosa ETA CP
type
Nucleic acid'
5
p,
200 compared with the unconjugated ETA preparation 2. No deaths occurred in mice injected with 25 ,ug of type 8 CP-ETA per mouse (-20 ,ug of ETA per dose of conjugate). Serum antibodies elicited by type 5 or 8 CP alone or conjugated to ETA. Three injections of either CP alone did not elicit a rise in CP antibodies (Tables 3 and 4). Identical results were obtained with doses of 0.05, 0.5, and 25 ,ug of S. aureus CP injected according to the same schedule (data not shown). The first injection of the conjugates elicited low levels of CP antibodies for both type 5 and type 8 (P = 0.001). Both conjugates elicited a CP booster response after the second injection (P = 0.001); the type 5 CP-ETA elicited a -30-fold rise, and the type 8 CP-ETA stimulated a -15fold rise. The third injection of type 8 CP-ETA induced a further increase in type 8 CP antibodies (P = 0.001); type 5 CP-ETA did not elicit a significant rise after the third injection.
Carrier priming. One injection of 2 ,ug of ETA per mouse (the amount in a dose of each conjugate) did not elicit a rise in levels of CP antibodies (data not shown). Injection of the ETA-primed mice with either conjugate showed an enhanced CP antibody response compared with injection of
unprimed animals (P = 0.001 for type 5 and P = 0.001 for type 8) (Tables 3 and 4). The priming effect was carrier specific, since BSA, a nonrelated protein, had no effect on the CP antibody response to the type 5 CP-ETA. ETA antibodies. One injection of the type 5 CP-ETA elicited a rise in ETA antibodies (P < 0.001); a second injection of type 5 CP-ETA elicited a booster response (P < 0.001) (Table 5). The rise in ETA antibodies which occurred after the third injection of this conjugate was not significant. Type 8 CP-ETA failed to elicit a significant rise in ETA antibodies after the first injection. The second and third injections of type 8 CP-ETA each elicited significant rises (P < 0.001). Mice primed with ETA responded to one injection of either conjugate with antibody levels comparable to those obtained after two injections. As expected, injection of either CP alone did not elicit ETA antibodies. Phagocytic activities of conjugate-induced antibodies. No phagocytosis of S. aureus type 5 strains by human PMNs occurred with added normal rabbit or mouse serum (Fig. 4). Both rabbit antisera (see Materials and Methods) and sera from mice immunized with the conjugates facilitated the internalization of the bacteria and their subsequent death, as indicated by the decline in viable-cell count within 120 min. The degree of phagocytosis facilitated by the rabbit and murine antisera was similar for both CP types (Fig. 4; data shown for type 5 only). Neither mouse nor rabbit type 5 antiserum facilitated phagocytosis of S. aureus type 8 organisms, and vice versa (data not shown). These results demonstrate that antibodies to the type 5 and type 8 CPs, derived from immunization with ETA conjugates of these antigens, are opsonic and type specific. Neutralization of ETA by conjugate-induced antibodies. The reference antisera (see Materials and Methods) prepared in BALB/c mice and assigned a value of 100 ELISA units were assayed for their neutralization of the cytotoxicity of ETA in CHO cells. The two commercial toxins failed to demonstrate cytotoxic activity against this cell line. Accord-
TABLE 4. Serum S. aureus type 8 CP antibodies elicited in mice by type 8 CP alone or conjugated to P. aeruginosa ETA' Type 8 CP antibody concn after injection no.b:
Immunogen
Saline Type 8 CP Type 8 CP-ETA ETA primed
1
