FEMSMicrobiologyLetters94 (1992)89-94 © 1992Federation of European MicrobiologicalSocieties0378-1097/92/S05.00 Publishedby Elsevier
FEMSLE 04930
Production of monovalent antisera by induction of immunological tolerance for capsular typing of Streptococcuspneumoniae Jorgen Henrichsen " and John B. Robbins b a WHO Collaborating ('enlre for Reference and Research on Pneumtn'oecL Statens Senunm~t~tut. Copenhagen, Denmark, and i, National h~stimte of Child health and Human Del'dopment, NIH, Bethesda, MD, USA
Received21 February 1992 Revisionreceived27 March 1992 Accepted i April 1992 Key words: Pneumococcal typing; Epitope-specific immunological tolerance; Streptococcus pneumoniae
1. SUMMARY Hyperimmune and high-titered polyclonal pneumococcal antisera, specific for cross.reactive types within groups, were produced in adult rabbits. Purified capsular polysaccharide was injected intravenously into adult rabbits. One week later, these rabbits were given multiple intravenous injections of formalin-inactivated pneumococci of the cross-reactive type by an established method. Each of the resultant antise~a were specific for the cross-reactive type indicating that the previous injection of the polysaccharide had induced epitope-specific tolerance. This method was successful for production of antisera against pneumococcal types 6A, 6B, 9N, 9V, 19F and 19A. Polyelonal rabbit pneumococcal antisera have some advantages over murine monoclonal antibodies for serologic studies and this method should be applicable for producing type-
J, Henrichsen,Statens Seruminstitut, Artillerivej5, DK-2300CopenhagenS, Denmark.
Correspondence to:
specific antibodies to cross-reactive polysaccharides of clinical interest. Further, this method is simpler and generally produces higher titered monovalent (factor) reagents than absorbed antisera.
2, INTRODUCTION There is extensive cross-reactivity among certain pneumococcal types [1]. The most cross-reactive types have been classified into groups [2]. The identification of these cross-reactive types has necessitated the absorbtion of antiserum produced by multiple intravenous injections of formalin-inactivated pneumococci [1]. These monovalent reagents are called 'factor sera'. Their widespread use is limited because their preparation is laborious and their titre of antibodies is generally low. Murine hybridomas have been induced in order to produce monoclonal pneumococcal antibodies [3]. This approach has limitations because some pneumococcal types, such as those of group
6, fail to elicit serum antibody responses in many strains of mice [4,5]. Immunological tolerance to polysaccharide antigens has been induced in adult rabbits [6,7]. Although not reported for polysaccharides in rabbits, it is possible to elicit 'epitope-specifie' tolerance by intravenous injection of comparatively large amounts of protein in mice and in rabbits [8]. In this study, we report the induction of epitope-specific tolerance to cross-reactive types of pneumococcal polysaccharides by injection of adult rabbits with high doses of purified capsular polysaccharides. This experimental procedure was successful in producing large amounts of hightitered 'factor sera' for representative pneumocoecaI types.
3. MATERIALS AND METHODS
3.I. Antigen preparations Pneumococeal polysaccharides types 6A, 6B, 9N, 9V, 19F and 19A were kindly donated by Merck Sharp & Dohme, West Point, PA. Formalin-inactivated bac'~eria of reference strains have been described by Lt~nd and Henrichsen [1]. 3.2. Immunization White rabbits, weighing 2-3.0 kg, were injected intravenously with 15 mg of a polysaccharide in 1.0 mi of saline. One week later, the rabbits were injected repeatedly with whole formalin-treated cells of the cross-reactive type and bled monthly as described [1]. Three rabbits were immunized with each antigen. Factor sera, absorbed with the cross-reactive types, were also prepared as described [1]. 3.3. Animals Animals were from an inbred colony of rabbits of Statens Seruminstitut. They were kept according to the guidelines of the Danish Department of Justice and under close veterinary control. 3.4. Antibody titres Antibody titres were determined as capsular titres [1,9].
3.5. Determination of antigen Determination of antigen concentration was done by ELISA. Wells were coated with a purified polyclonal rabbit anti-group 6 antiserum diluted l: 10 000 in PBS (0.1 M NaCl in 0.05 M phosphate buffer pH 7.4). The sample to be tested (in casu: various pneumoeoccal antisera) was then added in two-fold dilutions in PBS witl-:Tween 20 (0.1%) starting with 1:10. Washing was performed three times with PBS with 0.1% Tween 20. A murine monoclonal anti-type 6A antibody (produced by immunization with a pneumococcal type 6A-tetanus toxoid conjugate [10], diluted 1:500 with PBS/Tween was added. Washing three times was repeated and a rabbit-anti-mouse antibody, diluted 1:500 in PBS/Tween conjugated with peroxidase (Dakopatts, Glostrup, Denmark) was added and allowed to react with its substrate (OPD). Absorption was measured photometrically at 490 nm Negative controls were two-fold d;!utions with PBS of normal rabbit serum and PBS alone. For comparative purposes we included two-fold dilutions of both types 6A and 6B polysaceharides starting with a concentration of 1 /.tg/ml and ending with about 5 ng. This kind of assay does not determine the concentration of antigen, but allows an interpretation of the result (arbitrary ELISA units) into a range of concentrations corresponding to at most two steps of dilution.
4. RESULTS
4.1. Group 6 The straetures and serological properties of the group 6 ~.-Ivsaccharides, types 6A and 6B, are closely relatt, q~ 'tl 1-13] Type 6A: -~ 2)-a-D-Oalp-(1 --* 3)-a-D-Glup(1 ~ 3)-Ct-L-Rhap-(1 -~ 3)-o-ribitol-5PO4 -~ Type 6B: ~ 2)-a-D-Galp-(1 ~ 3)-a-D-Glup(1 --* 3)-a-L-Rhap-(1 --* 4)-D-ribitol-5PO4
Table I Capsular reaction titres of types within group 6 in rabbits injected with either types 6A or 6B polysaccharides (ps) prior to immunization with whole pneumococci lmmunogen
Titre Type 6A
Type 6B
Type 6A cells ~' Type 6B ps, type 6A cells ~'
512 256 (8)
256 4 ( < 2)
Type 6B cells a Type 6A ps, type 6B cells b
128 < 2 (4)
512 256 (6a)
a Bleeding of one representative rabbit immunized according to the routine schedule [1]. ~' Best bleeding of one rabbit in each of the [our groups ~third and fourth bleeding from types 6A and 6B animals, respectively). The best responses of the least responsive rabbil,, are given in parentheses (third and second bleeding, respectively)~
Prior immunization with the polysaccharide of one type did not reduce the titre elicited by the formalin-inactivated bacteria of the same type. The titre to the cross-reactive type was, however, almost eliminated by the tolerizing dose of polysaccharide (Table 1). The anti-type 6A serum obtained from the tolerized rabbits contained trace amounts of type 6B polysaecharide ( ~ 5 ng) and this amount was clearly less in the serum obtained after the third bleeding than that obtained after the first. The mono-specificity of the antisera became less pronounced when the tolerizing polysaccharides were no longer measurable. In contrast, factor sera, obtained by absorption with whole bacteria, contained between 10 2 and 10 3 times as much
polysaccharide (results not shown but obtained as described in MATERIAt.S AND ME'rHODS).
.~.2. Group 9 There are 4 types within this group: most isolates from patients are either type 9N or type 9V [14]. These two types are only slightly related structurally to each other and are not "very crossreactive [15]. The structures of pneumococcal types 9N [16] and 9V [17] are as follows: Type 9N: ~ 4)-a-o-GlcpA-(l --, 3)-a-D-Ghip-(I ~3)-~+oManpNAc-( 1 -'. 4)-a-D-GIco-(I -'. 4)-~-DGlcpNAc-( I Type 9V: --) 4)-a-D-GIcpA-(! ---)3)~-r,-Galp-( I --) 3)-fl-oI OAc ManpNAc-( i --, 4)-/3-D-GIcp(I ~ 4)~t-D-GIcp-(I I OAc Injection of the type 9V polysaccharide eliminated the titres to types 9V and to type 9A without affecting the homologous antibody responses to type 9N induced by the formalin-inactivated ceils. The homologous titre elicited by pneumococcal type 9V cells was similarly unaffected by injection of type 9N polysaccharide. The titre to the cross-reactive type 9N, when immunizing with type 9V cells, is almost negligible (Table 2).
4.3. Group 19 The structure of type~ i9F and 19A, responsible for almost all diseases caused by group 19
Table 2 Capsular reaction titres of types within group 9 in rabbits injected with t:ither types 9N or 9V potysaccharides prior to immunization with whole pneumococci [mmunogen
Titre Type 9N
Type 9~.
Type 9A
Type 9L
Type 9N cells ~ Type 9V ps, type 9N cells ~'
256 1024 (512)
64 < 2 ( < 2)
128 < 2 ( < 2)
256 128 (128)
Type 9V cells ~ Type 9N ps, type 9V cells b
4 2 ( < 2)
64 2.56(128)
128 512 (64)
64 256 (32)
a Bleeding of one representative rabbit immunized according to the routine schedule [1]. b Best bleeding of one rabbit in each of the four groups (fifth b!eeding from both types 9N and 9V animals). The best responses of the least responsive rabbits are given in parentheses (third ,arough fifth and second bleeding, respectively).
pneumococci, are very similar as shown by Jennings et al. [18] and Katzenellenbogen and Jennings [19]: Type 19F: -~ 4)-fl-D-ManNAcp-(1 ~ 4)-Ot-DGlcp-(l -, 2)-a-L-Rhap-(PO4) Type 19A: ---}4~-B-o-ManNAcp-(l -~ 4)-a-DGIcp-(1 -, 3)-a-L-Rhap-(PO4) -} As shown in Table 3 injection of type 19A polysaccharide prior to immunization with type 19F whole cells almost eliminates the type 19A reactivity of the serum obtained and vice versa.
5, DISCUSSION Immunological tolerance (or maybe rather paralysis) has been induced in adult rabbits by high doses (1-30 rag) of pneumococcal polysaccharides [6,20]. Tolerance has also been elicited in adult rabbits to a part of a protein by neonatal injection of high doses of a fragment of that molecule [21]. Our results show, for the first time, that epitope-specific tolerance can also be induced by potysaccharides. The results shown in the tables were representative of the experiments and demonstrate that the paralysis as well as the immune response vary, but not very much. Different rabbits mount different responses to the different pneumococcal capsular polysaccharide antigens and the most useful antisera were often obtained at differcnt periods of the immunization
(Tables 1-3). It is, however, important that in all three pneumococcal groups very useful monovalent antisera were obtained. The best explanation for the epitope-specific tolerance .achieved in this study is as follows: the reactive polysaccharides have at least several shared epitopes; B-cells secrete antibodies of a single specificity; the high dosages of polysaccharide interacted with B-cells containing membrane-bound IgM antibodies of a single specificity and, thereby, paralyscd them. The B-cells stimulated by the whole bacteria were those producing antibodies to the epitopes not shared by the cross-reactive bacteria; this refractory period lasted about several months. Antibody levels achieved by this technique are generally higher than those achieved by absorbed antisera and the sera contain much less free polysaccharide. This method can probably be used for other systems including Klebsiella and E. coil-capsulated bacterial pathogens. Immunological paralysis has been shown also to occur in man by Pichichero [22], who described a 9-month old boy with meningitis caused by Streptococcus pneumoniae type 18C. Capsular polysaceharide was present in his urine until he was 36 months old. At the age of 24 months he responded well to immunization with tetanus toxold, Haemophilus influenzae type b and to some pneumococcai capsular polysaccharides, but not to type 18C At the age of 48 months, i.e. 1 year after antigenuria was last demonstrated, he had a normal response to type 18C after pneumococcal vaccination.
Table 3 Capsular reaction titres of types within group 19 in rabbits injected with either types 19F or 19A polysaccharides (ps) prior to immunization with whole pneumococci Immnogen
Titrc Type 19F
Type 19A
Type 19F cells" Type 19A ps, type 19F cells b
) 512 > 512 (256)
256 4 (8)
Type 19A cells ' Type 19F ps, tpe 19A cells b
512 4 (4)
512 128 (64)
Type 19B
Type 19C
8 2 ( < 2)
2 < 2 ( < 2)
FEMSLE 04930
Production of monovalent antisera by induction of immunological tolerance for capsular typing of Streptococcuspneumoniae Jorgen Henrichsen " and John B. Robbins b a WHO Collaborating ('enlre for Reference and Research on Pneumtn'oecL Statens Senunm~t~tut. Copenhagen, Denmark, and i, National h~stimte of Child health and Human Del'dopment, NIH, Bethesda, MD, USA
Received21 February 1992 Revisionreceived27 March 1992 Accepted i April 1992 Key words: Pneumococcal typing; Epitope-specific immunological tolerance; Streptococcus pneumoniae
1. SUMMARY Hyperimmune and high-titered polyclonal pneumococcal antisera, specific for cross.reactive types within groups, were produced in adult rabbits. Purified capsular polysaccharide was injected intravenously into adult rabbits. One week later, these rabbits were given multiple intravenous injections of formalin-inactivated pneumococci of the cross-reactive type by an established method. Each of the resultant antise~a were specific for the cross-reactive type indicating that the previous injection of the polysaccharide had induced epitope-specific tolerance. This method was successful for production of antisera against pneumococcal types 6A, 6B, 9N, 9V, 19F and 19A. Polyelonal rabbit pneumococcal antisera have some advantages over murine monoclonal antibodies for serologic studies and this method should be applicable for producing type-
J, Henrichsen,Statens Seruminstitut, Artillerivej5, DK-2300CopenhagenS, Denmark.
Correspondence to:
specific antibodies to cross-reactive polysaccharides of clinical interest. Further, this method is simpler and generally produces higher titered monovalent (factor) reagents than absorbed antisera.
2, INTRODUCTION There is extensive cross-reactivity among certain pneumococcal types [1]. The most cross-reactive types have been classified into groups [2]. The identification of these cross-reactive types has necessitated the absorbtion of antiserum produced by multiple intravenous injections of formalin-inactivated pneumococci [1]. These monovalent reagents are called 'factor sera'. Their widespread use is limited because their preparation is laborious and their titre of antibodies is generally low. Murine hybridomas have been induced in order to produce monoclonal pneumococcal antibodies [3]. This approach has limitations because some pneumococcal types, such as those of group
6, fail to elicit serum antibody responses in many strains of mice [4,5]. Immunological tolerance to polysaccharide antigens has been induced in adult rabbits [6,7]. Although not reported for polysaccharides in rabbits, it is possible to elicit 'epitope-specifie' tolerance by intravenous injection of comparatively large amounts of protein in mice and in rabbits [8]. In this study, we report the induction of epitope-specific tolerance to cross-reactive types of pneumococcal polysaccharides by injection of adult rabbits with high doses of purified capsular polysaccharides. This experimental procedure was successful in producing large amounts of hightitered 'factor sera' for representative pneumocoecaI types.
3. MATERIALS AND METHODS
3.I. Antigen preparations Pneumococeal polysaccharides types 6A, 6B, 9N, 9V, 19F and 19A were kindly donated by Merck Sharp & Dohme, West Point, PA. Formalin-inactivated bac'~eria of reference strains have been described by Lt~nd and Henrichsen [1]. 3.2. Immunization White rabbits, weighing 2-3.0 kg, were injected intravenously with 15 mg of a polysaccharide in 1.0 mi of saline. One week later, the rabbits were injected repeatedly with whole formalin-treated cells of the cross-reactive type and bled monthly as described [1]. Three rabbits were immunized with each antigen. Factor sera, absorbed with the cross-reactive types, were also prepared as described [1]. 3.3. Animals Animals were from an inbred colony of rabbits of Statens Seruminstitut. They were kept according to the guidelines of the Danish Department of Justice and under close veterinary control. 3.4. Antibody titres Antibody titres were determined as capsular titres [1,9].
3.5. Determination of antigen Determination of antigen concentration was done by ELISA. Wells were coated with a purified polyclonal rabbit anti-group 6 antiserum diluted l: 10 000 in PBS (0.1 M NaCl in 0.05 M phosphate buffer pH 7.4). The sample to be tested (in casu: various pneumoeoccal antisera) was then added in two-fold dilutions in PBS witl-:Tween 20 (0.1%) starting with 1:10. Washing was performed three times with PBS with 0.1% Tween 20. A murine monoclonal anti-type 6A antibody (produced by immunization with a pneumococcal type 6A-tetanus toxoid conjugate [10], diluted 1:500 with PBS/Tween was added. Washing three times was repeated and a rabbit-anti-mouse antibody, diluted 1:500 in PBS/Tween conjugated with peroxidase (Dakopatts, Glostrup, Denmark) was added and allowed to react with its substrate (OPD). Absorption was measured photometrically at 490 nm Negative controls were two-fold d;!utions with PBS of normal rabbit serum and PBS alone. For comparative purposes we included two-fold dilutions of both types 6A and 6B polysaceharides starting with a concentration of 1 /.tg/ml and ending with about 5 ng. This kind of assay does not determine the concentration of antigen, but allows an interpretation of the result (arbitrary ELISA units) into a range of concentrations corresponding to at most two steps of dilution.
4. RESULTS
4.1. Group 6 The straetures and serological properties of the group 6 ~.-Ivsaccharides, types 6A and 6B, are closely relatt, q~ 'tl 1-13] Type 6A: -~ 2)-a-D-Oalp-(1 --* 3)-a-D-Glup(1 ~ 3)-Ct-L-Rhap-(1 -~ 3)-o-ribitol-5PO4 -~ Type 6B: ~ 2)-a-D-Galp-(1 ~ 3)-a-D-Glup(1 --* 3)-a-L-Rhap-(1 --* 4)-D-ribitol-5PO4
Table I Capsular reaction titres of types within group 6 in rabbits injected with either types 6A or 6B polysaccharides (ps) prior to immunization with whole pneumococci lmmunogen
Titre Type 6A
Type 6B
Type 6A cells ~' Type 6B ps, type 6A cells ~'
512 256 (8)
256 4 ( < 2)
Type 6B cells a Type 6A ps, type 6B cells b
128 < 2 (4)
512 256 (6a)
a Bleeding of one representative rabbit immunized according to the routine schedule [1]. ~' Best bleeding of one rabbit in each of the [our groups ~third and fourth bleeding from types 6A and 6B animals, respectively). The best responses of the least responsive rabbil,, are given in parentheses (third and second bleeding, respectively)~
Prior immunization with the polysaccharide of one type did not reduce the titre elicited by the formalin-inactivated bacteria of the same type. The titre to the cross-reactive type was, however, almost eliminated by the tolerizing dose of polysaccharide (Table 1). The anti-type 6A serum obtained from the tolerized rabbits contained trace amounts of type 6B polysaecharide ( ~ 5 ng) and this amount was clearly less in the serum obtained after the third bleeding than that obtained after the first. The mono-specificity of the antisera became less pronounced when the tolerizing polysaccharides were no longer measurable. In contrast, factor sera, obtained by absorption with whole bacteria, contained between 10 2 and 10 3 times as much
polysaccharide (results not shown but obtained as described in MATERIAt.S AND ME'rHODS).
.~.2. Group 9 There are 4 types within this group: most isolates from patients are either type 9N or type 9V [14]. These two types are only slightly related structurally to each other and are not "very crossreactive [15]. The structures of pneumococcal types 9N [16] and 9V [17] are as follows: Type 9N: ~ 4)-a-o-GlcpA-(l --, 3)-a-D-Ghip-(I ~3)-~+oManpNAc-( 1 -'. 4)-a-D-GIco-(I -'. 4)-~-DGlcpNAc-( I Type 9V: --) 4)-a-D-GIcpA-(! ---)3)~-r,-Galp-( I --) 3)-fl-oI OAc ManpNAc-( i --, 4)-/3-D-GIcp(I ~ 4)~t-D-GIcp-(I I OAc Injection of the type 9V polysaccharide eliminated the titres to types 9V and to type 9A without affecting the homologous antibody responses to type 9N induced by the formalin-inactivated ceils. The homologous titre elicited by pneumococcal type 9V cells was similarly unaffected by injection of type 9N polysaccharide. The titre to the cross-reactive type 9N, when immunizing with type 9V cells, is almost negligible (Table 2).
4.3. Group 19 The structure of type~ i9F and 19A, responsible for almost all diseases caused by group 19
Table 2 Capsular reaction titres of types within group 9 in rabbits injected with t:ither types 9N or 9V potysaccharides prior to immunization with whole pneumococci [mmunogen
Titre Type 9N
Type 9~.
Type 9A
Type 9L
Type 9N cells ~ Type 9V ps, type 9N cells ~'
256 1024 (512)
64 < 2 ( < 2)
128 < 2 ( < 2)
256 128 (128)
Type 9V cells ~ Type 9N ps, type 9V cells b
4 2 ( < 2)
64 2.56(128)
128 512 (64)
64 256 (32)
a Bleeding of one representative rabbit immunized according to the routine schedule [1]. b Best bleeding of one rabbit in each of the four groups (fifth b!eeding from both types 9N and 9V animals). The best responses of the least responsive rabbits are given in parentheses (third ,arough fifth and second bleeding, respectively).
pneumococci, are very similar as shown by Jennings et al. [18] and Katzenellenbogen and Jennings [19]: Type 19F: -~ 4)-fl-D-ManNAcp-(1 ~ 4)-Ot-DGlcp-(l -, 2)-a-L-Rhap-(PO4) Type 19A: ---}4~-B-o-ManNAcp-(l -~ 4)-a-DGIcp-(1 -, 3)-a-L-Rhap-(PO4) -} As shown in Table 3 injection of type 19A polysaccharide prior to immunization with type 19F whole cells almost eliminates the type 19A reactivity of the serum obtained and vice versa.
5, DISCUSSION Immunological tolerance (or maybe rather paralysis) has been induced in adult rabbits by high doses (1-30 rag) of pneumococcal polysaccharides [6,20]. Tolerance has also been elicited in adult rabbits to a part of a protein by neonatal injection of high doses of a fragment of that molecule [21]. Our results show, for the first time, that epitope-specific tolerance can also be induced by potysaccharides. The results shown in the tables were representative of the experiments and demonstrate that the paralysis as well as the immune response vary, but not very much. Different rabbits mount different responses to the different pneumococcal capsular polysaccharide antigens and the most useful antisera were often obtained at differcnt periods of the immunization
(Tables 1-3). It is, however, important that in all three pneumococcal groups very useful monovalent antisera were obtained. The best explanation for the epitope-specific tolerance .achieved in this study is as follows: the reactive polysaccharides have at least several shared epitopes; B-cells secrete antibodies of a single specificity; the high dosages of polysaccharide interacted with B-cells containing membrane-bound IgM antibodies of a single specificity and, thereby, paralyscd them. The B-cells stimulated by the whole bacteria were those producing antibodies to the epitopes not shared by the cross-reactive bacteria; this refractory period lasted about several months. Antibody levels achieved by this technique are generally higher than those achieved by absorbed antisera and the sera contain much less free polysaccharide. This method can probably be used for other systems including Klebsiella and E. coil-capsulated bacterial pathogens. Immunological paralysis has been shown also to occur in man by Pichichero [22], who described a 9-month old boy with meningitis caused by Streptococcus pneumoniae type 18C. Capsular polysaceharide was present in his urine until he was 36 months old. At the age of 24 months he responded well to immunization with tetanus toxold, Haemophilus influenzae type b and to some pneumococcai capsular polysaccharides, but not to type 18C At the age of 48 months, i.e. 1 year after antigenuria was last demonstrated, he had a normal response to type 18C after pneumococcal vaccination.
Table 3 Capsular reaction titres of types within group 19 in rabbits injected with either types 19F or 19A polysaccharides (ps) prior to immunization with whole pneumococci Immnogen
Titrc Type 19F
Type 19A
Type 19F cells" Type 19A ps, type 19F cells b
) 512 > 512 (256)
256 4 (8)
Type 19A cells ' Type 19F ps, tpe 19A cells b
512 4 (4)
512 128 (64)
Type 19B
Type 19C
8 2 ( < 2)
2 < 2 ( < 2)
