Parasite Immunology 1990, 12, 199-21 I
Characterization of a family of monoclonal antibodies which bind Schistosoma japonicurn egg antigens and express an interstrain cross-reactive idiotype G.RICHARD OLDS & THOMAS F.KRESINA Brown University Program of Geographic Medicine, International Health Institute, Department of Medicine, Miriam Hospital, Providence, RI 02906, USA Accepted for publication 6 September 1989
Summary A family of monoclonal antibodies (MoAb) was derived from the somatic cell fusion of P,NSI myeloma cells and lymphoid cells from naturally infected mice or hyperimmunized mice. C57BL/6 mice were naturally infected with Schistosoma japonicum, and BALB/c mice were hyperimmunized with preparations of S. japonicum soluble egg antigens (SEA). The MoAbs which reflected the immune repertoire of naturally infected animals versus hyperimmune animals were characterized with regard to antibody isotype, antigen binding specificity, in-vitro immunosuppression of antigen-induced cell-mediated immune responses and the expression of SJ-CRIM, a major Cross-reactive idiotype which appears on polyclonal anti-SEA antibodies generated during murine S. japonicum infection. The data indicate that for MoAbs of the IgG isotype which bound SEA by ELISA, the most immunosuppressive anti-SEA MoAbs identified expressed SJ-CRIMand were derived from naturally infected mice. All anti-SEA MoAbs expressing SJ-CRIMshowed an identical banding pattern on immunoblot analysis which was abrogated by weak periodate treatment. The generation of expression of SJ-CRIM o n MoAbs using differing methodologies across an allotype barrier indicates that the expression of SJ-CRIM is encoded by a germline gene. These data indicate an association between expression of this germline interstrain cross-reactive idiotype and immunosuppressive capacity. In addition, the immunoregulatory network which develops during immune S . japonicum infection is initiated by a carbohydrate epitope(s) found on various SEA. These data have profound implications in the use of the cross-reactive idiotype as a serodiagnostic tool in schistosomiasis. Keywords: Schistosoma japonicum infection, monoclonal antibodies, interstrain cross-reactive idiotype
Correspondence: Thomas F.Kresina, PhD, Brown University Program of Geographic Medicine, Department of Medicine, Miriam Hospital, Providence, RI 02906, USA.
199
200
G.R.Olds & T.F.Kresina
Introduction An increasing number of studies have implicated the immune network, as first proposed by Jerne (1974), in the modulation of the immune responses to infectious agents. Antiidiotypic antibodies have been suggested to be candidates for vaccines in numerous infectious diseases including infection with Trypanosoma rhodesiense (Sacks & Sher 1983), hepatitis B virus (Dreesman & Kennedy 1985), Streptococcus pneumoniae (McNamara, Ward & Kohler 1984), reovirus (Sharpe et al. 1984), HTLV-111 virus (Dreesman 1986),and Schistosoma mansoni(Grzych et al. 1985, Kresina & Olds 1989). In addition, anti-idiotypic antibodies and the idiotypic repertoire have been implicated in the regulation of the protective immune response (Phillips et al. 1986, 1988) as well as the immunopathology observed in schistosomiasis (Powell & Colley 1985, 1987, Olds & Kresina 1985) by an increasing volume ofdata. These studies, as well as the data from this laboratory, have shown that idiotype-based pathway(s) function in an immunoregulatory role and are present during natural infection with schistosomiasis. Up to now, studies from this laboratory (Olds & Kresina 1985, Kresina & Olds 1986) have described the appearance of naturally occurring immunoregulatory idiotype-based immune network components (idiotypic and anti-idiotypic antibodies) during the course of chronic infection in schistosomiasis japonicum. In these studies, polyclonal auto-antiidiotypic antibodies were identified during chronic infection which describe a major crossreactive idiotype (SJ-CRIM)occurring on anti-soluble egg antigen (SEA) molecules. These auto-anti-idiotypic antibodies were profoundly immunosuppressive in aitro and in oiuo. On the other hand, polyclonal anti-SEA antibodies bearing SJ-CRIMwere observed during acute infection. These anti-antigen (SEA) molecules themselves were two log dilutions less than the corresponding anti-idiotype in immunosuppressive capacity. The idiotypes are, however, critical to the induction and characterization of these antiidiotypic molecules, and served as markers for the duration of infection. The expression of SJ-CRIMon anti-SEA antibodies disappeared as the infection progressed from acute to chronic. This correlated temporarily with immune modulation; a beneficial downregulation which results in decreased granulomatosis. These results suggest that SJ-CRIM could represent a serum marker ofdisease modulation. These data concur with the results of studies in S. mansoni infection and together further indicate a central role for the expression of idiotype (SJ-CRIM)in the regulation of murine S.japonicum infection. In addition, these data suggest that the expression of SJ-CRIM,the naturally occurring cross-reactive idiotype expressed on anti-SEA antibodies in S.japonicum infection, could be serodiagnostic. In this regard, the loss of expression of polyclonal SJ-CRIMon antiSEA antibodies would reflect immune modulation of chronic schistosomiasis. Therefore, the present study was performed to determine the nature of SJ-CRIM,as expressed on polyclonal antibodies derived from natural infection, with monoclonal antibodies (MoAb) derived from either animals with natural infection or hyperimmune with parasite antigens. The identification of SJ-CRIMon MoAbs derived from natural infection would substantially facilitate the use of this reagent as an immunodiagnostic tool in S.japonicum infection. The results of the present study show that SJ-CRIMcan be expressed o n MoAbs derived from natural infection as well as MoAbs derived from egg immunized mice of another strain. Characterization of the expression of SJ-CRIMon these MoAbs indicates that ( I ) SJ-CRIM is encoded by a germline gene through its expression on MoAbs derived from mice of differing haplotypes; (2) SJ-CRIMrecognized
Monoclonal antibodies in S. japonicum
20 1
an epitope in soluble egg antigens which is comprised, at least in part, of carbohydrate moieties; and (3) the carbohydrate moieties are recurrent epitopes on numerous soluble egg antigen glycoproteins.
Materials and methods A N I M A L S A N D INFECTION
C57BL/6 mice (Jackson Laboratories, Bar Harbor, ME, USA) were infected with 25 cercariae of a Philippine strain of S.japonicum at Lowell University, Lowell, MA, USA (NIAID supply contract AI-02636). Animals with this intensity of infection will survive at least 30 weeks post-infection (Olds et al. 1982). BALB/c mice were immunized twice intraperitoneally with a purified preparation of S. japonicum SEA (100 pg) in Freund’s complete adjuvant (FCA) and rested for 4 weeks after the second immunization before being killed. B-CELLHYBRIDIZATION C57BL/6 mice were killed at 10-30 weeks of infection, their spleens were removed, and single cell suspensions prepared. BALB/c mice were killed after hyperimmunization with SEA/FCA, their spleens removed and single cell suspensions prepared. Splenocytes were fused with P3NS1 myeloma cells and cloned as described previously (Daniel & Olds 1985). Monoclonal antibodies derived from the differing time points during infection are noted in Table 1. Supernatants were screened for binding to SEA by ELISA and binding to polyclonal anti-SEA antibody by a radioimmunoassay (RIA) (Olds & Kresina 1985). Isotypes of MoAbs were determined by microimmunodiffusion (Olds et al. 1982). Supernatants from growing clones were also screened for their ability to suppress SEA-induced proliferative responses of splenocytes from 5-week infected animals (Garb, Stavitsky & Mahmoud 1981). In the SEA-induced proliferation assay, splenocytes from C57BL/6 5-week infected mice were stimulated with 10 pg of SEA as described previously (Garb et al. 1981). Cultures were then incubated for 3 days in 5 % CO2 at 37°C. Cultures were pulsed for the final 18 h with 3H-thymidine and proliferation determined by augmented incorporation of 3H compared to controls. For statistical comparison, data were expressed as mean ct/ min for quadruplicate samples and a Student’s t-test was used. Data are expressed in text and tables as per cent suppression and were calculated by the following formula: (Ctlmin control - ct/min background) - (ct/min experimental - ct/min background) x 100 (Ct/min control - ct/min background)
Background incorporation averaged 1026f87 ct/min. Monoclonal antibody preparations noted to suppress in-vitro SEA-induced proliferation were also tested for their ability to suppress concanavalin A, or PPD stimulated proliferation of splenocytes from BCG-treated mice as described previously (Garb et al. 1981, Collins & Watson 1979). These latter assays were used to determine the antigen specificity of the observed suppression.
202
G.R.Olds & TXKresina
ASSAY FOR
ANTI-SEAANTIBODY
Soluble egg antigen was prepared from the livers and intestines of infected CFI mice (Charles River, Cambridge, MA, USA) as described previously (Boros & Warren 1970). Binding of MoAbs to SEA was determined by a solid-phase ELISA in a microtitre assay (Olds & Kresina 1985) as described previously. For studies determining an association of carbohydrate moieties in the binding epitope, SEA was treated with 10 mM sodium metaperiodate (Woodward, Young & Boodgood 1985) after absorption to the ELISA plate for 30 min at room temperature in the absence of light. Sample wells not treated with sodium metaperiodate were incubated in an identical fashion with sodium acetate buffer. After incubation the MoAbs were tested for binding as noted previously (Olds & Kresina 1985). S E R O L O G I C A L A N A L Y S I S OF MONOCLONAL
ANTI-SEAA N T I B O D Y
The presence of SJ-CRIMon monoclonal anti-SEA antibodies was determined by an indirect method of precipitation as described for anti-idiotype activity (Olds & Kresina 1985). In the present case, autologous polyclonal anti-idiotypic antibody specifically purified from chronic (30 week) infected mouse serum was utilized as the labelled ligand. In previous studies (Olds & Kresina 1985), this polyclonal anti-idiotype preparation was utilized to define the appearance of SJ-CRIMon anti-SEA antibodies derived from acute infection. In the present study, serial dilutions (1 : 10 to 1 :320) of hybridoma culture supernatant were mixed with 10 pg of '251-labelledpolyclonal anti-idiotypic antibody in the presence of 10 pg of normal C57BL/6 or BALB/c immunoglobulin. The mixture was incubated for 1 h at 37°C and immune complexes precipitated by carrier precipitation (Olds & Kresina 1985). Data are presented as anti-idiotypic binding capacity of the MoAbs which was derived from the penultimate value on the linear portion of the binding curve as described previously (Olds & Kresina 1985). IMMUNOBLOT ANALYSIS OF M O N O C L O N A L ANTIBODIES
Elucidation of S.juponicurn antigen epitopes identified by murine MoAbs was determined by immunoblot analysis as described previously (Gershon & Palack 1983, King et ul. 1987). SEA was electrophoresed on 10% SDS polyacrylamide gels and Western blotted on to nitrocellulose paper (Jones 1980, Burnette 1981).
Results B-CELLHYBRIDIZATION-LYMPHOID
CELLS OF N A T U R A L L Y INFECTED M I C E
Three fusions were performed using splenocytes from lo-, 20- and 30-week infected mice. Monoclonal antibodies derived from these times of infection were characterized for their ability to bind SEA. Fifty-seven clones or 23% of total clones generated were observed to bind SEA. Of these, fewer than one in six were noted to suppress SEA-induced blastogenesis in vitro. Six immunosuppressive anti-SEA MoAbs were expanded and subcloned (Table 1). Monoclonal proteins isolated from culture supernatant of growing
203
Monoclonal antibodies in S . japonicum
Table 1. Characterization of monoclonal anti-SEA antibodies derived from na:urally infected C57BL/6 mice Suppression of Mice SEA blastogenesis* Anti-SEA (weeks of Monoclonal activity? Isotype Specificity (50 pg/ml) infection) antibody 12 30 30 10
10 10 10
106-1A9 100-3B10 100-3H 10 115-lA8 115-1BI 1 I15-1B4 115-1B9
IgG 1 IgG 1 IgM IgG 1 IgG 1 IgG I IgG I
Control Anti-SEA Anti-SEA Anti-SEA Anti-SEA Anti-SEA Anti-SEA
0Yo 43 f4% 37f3Y" 46+2% 0'?4 86 f 73 f4% 92+4%
Characterization of a family of monoclonal antibodies which bind Schistosoma japonicurn egg antigens and express an interstrain cross-reactive idiotype G.RICHARD OLDS & THOMAS F.KRESINA Brown University Program of Geographic Medicine, International Health Institute, Department of Medicine, Miriam Hospital, Providence, RI 02906, USA Accepted for publication 6 September 1989
Summary A family of monoclonal antibodies (MoAb) was derived from the somatic cell fusion of P,NSI myeloma cells and lymphoid cells from naturally infected mice or hyperimmunized mice. C57BL/6 mice were naturally infected with Schistosoma japonicum, and BALB/c mice were hyperimmunized with preparations of S. japonicum soluble egg antigens (SEA). The MoAbs which reflected the immune repertoire of naturally infected animals versus hyperimmune animals were characterized with regard to antibody isotype, antigen binding specificity, in-vitro immunosuppression of antigen-induced cell-mediated immune responses and the expression of SJ-CRIM, a major Cross-reactive idiotype which appears on polyclonal anti-SEA antibodies generated during murine S. japonicum infection. The data indicate that for MoAbs of the IgG isotype which bound SEA by ELISA, the most immunosuppressive anti-SEA MoAbs identified expressed SJ-CRIMand were derived from naturally infected mice. All anti-SEA MoAbs expressing SJ-CRIMshowed an identical banding pattern on immunoblot analysis which was abrogated by weak periodate treatment. The generation of expression of SJ-CRIM o n MoAbs using differing methodologies across an allotype barrier indicates that the expression of SJ-CRIM is encoded by a germline gene. These data indicate an association between expression of this germline interstrain cross-reactive idiotype and immunosuppressive capacity. In addition, the immunoregulatory network which develops during immune S . japonicum infection is initiated by a carbohydrate epitope(s) found on various SEA. These data have profound implications in the use of the cross-reactive idiotype as a serodiagnostic tool in schistosomiasis. Keywords: Schistosoma japonicum infection, monoclonal antibodies, interstrain cross-reactive idiotype
Correspondence: Thomas F.Kresina, PhD, Brown University Program of Geographic Medicine, Department of Medicine, Miriam Hospital, Providence, RI 02906, USA.
199
200
G.R.Olds & T.F.Kresina
Introduction An increasing number of studies have implicated the immune network, as first proposed by Jerne (1974), in the modulation of the immune responses to infectious agents. Antiidiotypic antibodies have been suggested to be candidates for vaccines in numerous infectious diseases including infection with Trypanosoma rhodesiense (Sacks & Sher 1983), hepatitis B virus (Dreesman & Kennedy 1985), Streptococcus pneumoniae (McNamara, Ward & Kohler 1984), reovirus (Sharpe et al. 1984), HTLV-111 virus (Dreesman 1986),and Schistosoma mansoni(Grzych et al. 1985, Kresina & Olds 1989). In addition, anti-idiotypic antibodies and the idiotypic repertoire have been implicated in the regulation of the protective immune response (Phillips et al. 1986, 1988) as well as the immunopathology observed in schistosomiasis (Powell & Colley 1985, 1987, Olds & Kresina 1985) by an increasing volume ofdata. These studies, as well as the data from this laboratory, have shown that idiotype-based pathway(s) function in an immunoregulatory role and are present during natural infection with schistosomiasis. Up to now, studies from this laboratory (Olds & Kresina 1985, Kresina & Olds 1986) have described the appearance of naturally occurring immunoregulatory idiotype-based immune network components (idiotypic and anti-idiotypic antibodies) during the course of chronic infection in schistosomiasis japonicum. In these studies, polyclonal auto-antiidiotypic antibodies were identified during chronic infection which describe a major crossreactive idiotype (SJ-CRIM)occurring on anti-soluble egg antigen (SEA) molecules. These auto-anti-idiotypic antibodies were profoundly immunosuppressive in aitro and in oiuo. On the other hand, polyclonal anti-SEA antibodies bearing SJ-CRIMwere observed during acute infection. These anti-antigen (SEA) molecules themselves were two log dilutions less than the corresponding anti-idiotype in immunosuppressive capacity. The idiotypes are, however, critical to the induction and characterization of these antiidiotypic molecules, and served as markers for the duration of infection. The expression of SJ-CRIMon anti-SEA antibodies disappeared as the infection progressed from acute to chronic. This correlated temporarily with immune modulation; a beneficial downregulation which results in decreased granulomatosis. These results suggest that SJ-CRIM could represent a serum marker ofdisease modulation. These data concur with the results of studies in S. mansoni infection and together further indicate a central role for the expression of idiotype (SJ-CRIM)in the regulation of murine S.japonicum infection. In addition, these data suggest that the expression of SJ-CRIM,the naturally occurring cross-reactive idiotype expressed on anti-SEA antibodies in S.japonicum infection, could be serodiagnostic. In this regard, the loss of expression of polyclonal SJ-CRIMon antiSEA antibodies would reflect immune modulation of chronic schistosomiasis. Therefore, the present study was performed to determine the nature of SJ-CRIM,as expressed on polyclonal antibodies derived from natural infection, with monoclonal antibodies (MoAb) derived from either animals with natural infection or hyperimmune with parasite antigens. The identification of SJ-CRIMon MoAbs derived from natural infection would substantially facilitate the use of this reagent as an immunodiagnostic tool in S.japonicum infection. The results of the present study show that SJ-CRIMcan be expressed o n MoAbs derived from natural infection as well as MoAbs derived from egg immunized mice of another strain. Characterization of the expression of SJ-CRIMon these MoAbs indicates that ( I ) SJ-CRIM is encoded by a germline gene through its expression on MoAbs derived from mice of differing haplotypes; (2) SJ-CRIMrecognized
Monoclonal antibodies in S. japonicum
20 1
an epitope in soluble egg antigens which is comprised, at least in part, of carbohydrate moieties; and (3) the carbohydrate moieties are recurrent epitopes on numerous soluble egg antigen glycoproteins.
Materials and methods A N I M A L S A N D INFECTION
C57BL/6 mice (Jackson Laboratories, Bar Harbor, ME, USA) were infected with 25 cercariae of a Philippine strain of S.japonicum at Lowell University, Lowell, MA, USA (NIAID supply contract AI-02636). Animals with this intensity of infection will survive at least 30 weeks post-infection (Olds et al. 1982). BALB/c mice were immunized twice intraperitoneally with a purified preparation of S. japonicum SEA (100 pg) in Freund’s complete adjuvant (FCA) and rested for 4 weeks after the second immunization before being killed. B-CELLHYBRIDIZATION C57BL/6 mice were killed at 10-30 weeks of infection, their spleens were removed, and single cell suspensions prepared. BALB/c mice were killed after hyperimmunization with SEA/FCA, their spleens removed and single cell suspensions prepared. Splenocytes were fused with P3NS1 myeloma cells and cloned as described previously (Daniel & Olds 1985). Monoclonal antibodies derived from the differing time points during infection are noted in Table 1. Supernatants were screened for binding to SEA by ELISA and binding to polyclonal anti-SEA antibody by a radioimmunoassay (RIA) (Olds & Kresina 1985). Isotypes of MoAbs were determined by microimmunodiffusion (Olds et al. 1982). Supernatants from growing clones were also screened for their ability to suppress SEA-induced proliferative responses of splenocytes from 5-week infected animals (Garb, Stavitsky & Mahmoud 1981). In the SEA-induced proliferation assay, splenocytes from C57BL/6 5-week infected mice were stimulated with 10 pg of SEA as described previously (Garb et al. 1981). Cultures were then incubated for 3 days in 5 % CO2 at 37°C. Cultures were pulsed for the final 18 h with 3H-thymidine and proliferation determined by augmented incorporation of 3H compared to controls. For statistical comparison, data were expressed as mean ct/ min for quadruplicate samples and a Student’s t-test was used. Data are expressed in text and tables as per cent suppression and were calculated by the following formula: (Ctlmin control - ct/min background) - (ct/min experimental - ct/min background) x 100 (Ct/min control - ct/min background)
Background incorporation averaged 1026f87 ct/min. Monoclonal antibody preparations noted to suppress in-vitro SEA-induced proliferation were also tested for their ability to suppress concanavalin A, or PPD stimulated proliferation of splenocytes from BCG-treated mice as described previously (Garb et al. 1981, Collins & Watson 1979). These latter assays were used to determine the antigen specificity of the observed suppression.
202
G.R.Olds & TXKresina
ASSAY FOR
ANTI-SEAANTIBODY
Soluble egg antigen was prepared from the livers and intestines of infected CFI mice (Charles River, Cambridge, MA, USA) as described previously (Boros & Warren 1970). Binding of MoAbs to SEA was determined by a solid-phase ELISA in a microtitre assay (Olds & Kresina 1985) as described previously. For studies determining an association of carbohydrate moieties in the binding epitope, SEA was treated with 10 mM sodium metaperiodate (Woodward, Young & Boodgood 1985) after absorption to the ELISA plate for 30 min at room temperature in the absence of light. Sample wells not treated with sodium metaperiodate were incubated in an identical fashion with sodium acetate buffer. After incubation the MoAbs were tested for binding as noted previously (Olds & Kresina 1985). S E R O L O G I C A L A N A L Y S I S OF MONOCLONAL
ANTI-SEAA N T I B O D Y
The presence of SJ-CRIMon monoclonal anti-SEA antibodies was determined by an indirect method of precipitation as described for anti-idiotype activity (Olds & Kresina 1985). In the present case, autologous polyclonal anti-idiotypic antibody specifically purified from chronic (30 week) infected mouse serum was utilized as the labelled ligand. In previous studies (Olds & Kresina 1985), this polyclonal anti-idiotype preparation was utilized to define the appearance of SJ-CRIMon anti-SEA antibodies derived from acute infection. In the present study, serial dilutions (1 : 10 to 1 :320) of hybridoma culture supernatant were mixed with 10 pg of '251-labelledpolyclonal anti-idiotypic antibody in the presence of 10 pg of normal C57BL/6 or BALB/c immunoglobulin. The mixture was incubated for 1 h at 37°C and immune complexes precipitated by carrier precipitation (Olds & Kresina 1985). Data are presented as anti-idiotypic binding capacity of the MoAbs which was derived from the penultimate value on the linear portion of the binding curve as described previously (Olds & Kresina 1985). IMMUNOBLOT ANALYSIS OF M O N O C L O N A L ANTIBODIES
Elucidation of S.juponicurn antigen epitopes identified by murine MoAbs was determined by immunoblot analysis as described previously (Gershon & Palack 1983, King et ul. 1987). SEA was electrophoresed on 10% SDS polyacrylamide gels and Western blotted on to nitrocellulose paper (Jones 1980, Burnette 1981).
Results B-CELLHYBRIDIZATION-LYMPHOID
CELLS OF N A T U R A L L Y INFECTED M I C E
Three fusions were performed using splenocytes from lo-, 20- and 30-week infected mice. Monoclonal antibodies derived from these times of infection were characterized for their ability to bind SEA. Fifty-seven clones or 23% of total clones generated were observed to bind SEA. Of these, fewer than one in six were noted to suppress SEA-induced blastogenesis in vitro. Six immunosuppressive anti-SEA MoAbs were expanded and subcloned (Table 1). Monoclonal proteins isolated from culture supernatant of growing
203
Monoclonal antibodies in S . japonicum
Table 1. Characterization of monoclonal anti-SEA antibodies derived from na:urally infected C57BL/6 mice Suppression of Mice SEA blastogenesis* Anti-SEA (weeks of Monoclonal activity? Isotype Specificity (50 pg/ml) infection) antibody 12 30 30 10
10 10 10
106-1A9 100-3B10 100-3H 10 115-lA8 115-1BI 1 I15-1B4 115-1B9
IgG 1 IgG 1 IgM IgG 1 IgG 1 IgG I IgG I
Control Anti-SEA Anti-SEA Anti-SEA Anti-SEA Anti-SEA Anti-SEA
0Yo 43 f4% 37f3Y" 46+2% 0'?4 86 f 73 f4% 92+4%
