Seand. J. Imtnunol. 9, 297-302, 1979
SHORT PAPER
Antigenic Analysis of Mycobacterium leprae O. CLOSS, R. N. MSHANA & M. HARBOE University of Oslo, Institute for Experimental Medical Research, UUeval Hospital, Oslo, Norway
Closs, O., Mshana, R.N. & Harboe, M. Antigenic Analysis of Mvcobmteritim leprae. Sca/id. J. Immunol. 9, 297-?.O2, 1979.
About twenty distinct antigenic components have been demonstrated in Mycobacterium leprae (M. leprae) by crossed immunoelectrophoresis against a rabbit antiserum produced by immunization wilh concentrated M. leprae antigen. This system allows a more detailed analysis of the antigenic relationship between M. leprae and olher mycobacteria and a better characterization of the antigenic content of various M. teprae preparations than wiih previously available antisera which reacted with tar fewer components. The anlibody activity in sera of patients with lepromatous leprosy was studied by incorporating the sera into the intermediate gel of the M. leprae reference system. Antibodies were found againsl only seven of the components. Since those compared are all known to be cross-reacting widely with antigens of oiher mycobacteria, it is speculated that cross-immunization may influence the antibody response in lepromatous leprosy. O. Closs, University «f Oslo, Institute for Experimental Medical Research. Ullevat Hospital, Oslo I, Norway.
The outcome and clinical manifestations after infection with M. leprae depend to a large extent on the host's immune response to the bacterium, but the major antigens involved in these reactions are at present unknown. A better characterization of the antigenic composition of M. leprae may be an important first step and may also help to clarify the antigenic relationship between this bacterium and other mycobacteria. With the supply of M. leprae from infected armadillos it has become possible to provide sufficient material for such antigenic studies. The antigenic structure of mycobacteria has been studied by various gel precipitation techniques [2, 9, 10], By crossed immunoelectrophoresis (CIE), between twenty and fifty different antigenic components (hereafter called components) are usually easily demonstrable in any mycobacteriutn. Therefore, it was disappointing when only seven components could be clearly defined in our first attempt to establish a CIE reference system for M. leprae [5]. 0300-9475/79/0300-0297 $02.00
Antisera against other mycobacteria frequently contained antibodies cross-reacting with these components, indicating that components with similar structure occur in many mycobacteria, and that they were not M. leprae specific. Moreover, antibodies against the same components were frequently found in the serum of patients with iepromatous leprosy [5]. During a mycobacterial infection cross-reactive antigens may be more immunogenic than the species specific ones [3]. This suggestion was supported by the results obtained during immunization of rabbits with different doses of M. bovis (BCG) [8]. Even with minute doses of BCG sonicate a relatively strong antibody response remained against a few highly crossreactive antigens. Comparison of antigenic extracts of armadillo-grown M. teprae and other mycobacteria by acrylamide gradient gel electrophoresis has revealed that the M. leprae preparation produced much fewer and weaker bands than preparations made in a similar way from cultivable mycobacteria [5]. This led us to
i: 1979 Blackwell Scientific Publications
297
298
O. CI0.SS, R. N. Mshana & M. Harhoe
assume that the antibody response in rabbits could be improved by increasing the dose of M. leprae antigen used for immunization.
MATERIALS AND METHODS Antigens. M. Icprac wn-i provided by the Worid Health Organization through its Immunology of Leprosy flMMLEP) programme. The bacilli, hatch AB19. were obtained as purified, lyophllized material, prepared from infected armadillo tissue by Dr P, Draper of the National Institute for Medical Research, Mill Hill. London [4]. Sonic extracts of the bacilli were prepared in the following way: 60 mg of bacilli (wet weight) or 6 mg of lyophilized .M. Icprac were added per millilitre of 0.9" o NaCI. and the suspension was sonified for 15 min using the Branson BU sonifier (Branson Sonic Power Co., Danbiiry. Conn.) with an effect of 90-100 W. If ihc extract was to be used in CIE, it was centrifuged al 20,000 g for 20 min to remove insoluble material. For immunization of rabbits the extract was mixed with an equal volume of Freund's incomplete adjuvant (FIA). In the present experiment the M. teprac sonicate was concentrated 10-fold in a Minicon Bcell(Amicon Corp.. Lexington. Mass., tJSA) before it was emulsified with FIA. The emulsion was kept at - 20 C between immunizations. Anii.sera. Antisera against a number of mycobacteria were prepared by hyperimmunizatlon of rabbits wiih sonified suspensions of Ihc bacteria in FIA as previously described |5). Anti-BCG immunoglobulin was kindly provided by Dakopatts A/S, Copenhagen (Code B124. lot 116). A rabbit antiserum against armadillo liver antigens was kindly provided by Mr Kesete Negassi. Imnuiniziiiion .schi-ilute. One rabbit (No. 042) was immunized with concentrated M. leprae antigen on days 0 and 14 and every 28th day thereafter. The antigen was injected intracutaneously at multiple sites in the neck region. Owing to technical difticullies some injections became located subcutaneously. To follow the immune response closely, the rabbit was bled on day 0 (pre-immune serum), day 14. and every week thereafter uniil day 160. The sera were slored with 15 mM NaN^aM C. Fructionaiion of ininiiinoghbutin. Immunoglobulins were isolated by (NHiJjSOi precipitation [6] of 20-25 ml of serum pooled from several bleedings. The purified immunoglobulin fractions were stored at 4 C with 15 niM NaN3 added. Cros.sed inwiunoclccfrophoresis. CIE with intermediate gel was carried out and the plates were washed, pressed and stained as described in detail elsewhere [I]. Briefly. CIE was performed on 5 v 5 cm glass plates in 1','-,, agarose (Litex type HSA with moderate electroendosmotic flow, Litex, Glostrup. Denmark) containing TRIS/barbital butTer, pH 8.6, and ionic strength 0.02, 10 [J1 of antigen were added in a circular well, 4 mm in diameter. The top gel contained 150 rj| antiserum and the intermediate gel contained 100 |jl buffer or antiserum.
RESULTS in the rabbit immunized with concentrated M. leprae antigen (No. 042) the first detectable antibodies appeared 2 weeks after the first immunization when three precipitin lines were seen in CIE, corresponding to M. leprae antigens 2, 5 and 7. Bolh the nutTiber of antibody specificities and the antibody concentrations increased with subsequent immunizations until a plateau was reached after the fifth immunization. Thereafter, the antibody concentrations remained fairly constant, although the activity against some components showed a tendency to decrease after the sixth immtmization. The highest number of lines was seen when concentrated immunoglobulins and 4- or 10fold concentrated antigens were used. Fig, 1 shows the same M. leprae antigen run in CIE against immunoglobulin prepared from rabbit 042 and against a pool of sera from rabbits immunized with regular doses; a marked diiTerence is seen. With the latter antibody preparation, only seven precipitin lines were visible, whereas more than twenty lines developed in the plate containing 042 immunoglobulins. The seven precipitin lines seen with the M. leprae serum pool corresponded to components 1. 2, 3, 4, 5, 7 and 8 of the previous reference system [5]. With the exception of component 1, these antigens all formed very distinct precipitin lines with the 042 immunoglobulin preparation. In addition, between eight and ten other distinct lines and a similar number of weaker lines were seen. Thus, altogether, the new precipitin pattern contained about twentyfour lines. Some of them were quite weak, however, and were sometimes lost during destaining. To test whether any of the components were of host origin, an antiserum against armadillo liver was incorporated in the intermediate gel. This antiserum produced twenty lines against armadillo liver in CIE but did not react with any of the antigens precipitated in the M. leprae system. When M. leprae antigen was run in CIE against antisera to other mycobacteria. notably anti-BCG, anti-M. avium and anti-M. Icpraemurium, a few more components were detected in addition to those precipitated by the 042 immunoglobulin preparation. Thus, the sonic extracts of M. /p/>r«£'probablycontainmorethan
Atitigens of Mycobacterium leprae
FIG. 1. Drawings of two CIE plalcs where the same Mycobacterium leprae sonicate was run against Iwo differeni antibody preparations. The intermediate gel beiween the two horizontal lines tlid not contain antibodies. (A) Antiserum from rabbits immunized witli standard doses of antigen. (B) Concentrated immunoglobulins from rabbits immunized with increased amounts of antigen. (C) Photograph of the plate drawn in B, stained with Coomassie brilliani blue. The numbers refer to the reference system previously described by Harboe er al. [5].
299
300
O. Closs, R. N. Mshana & M. Harboe
twenty-four antigens, but some of them produce only weak lines and appear to be present in very low concentrations. When strong rabbit antisera against M. avium, M. duvalii, M. lepraemurium. M. noncliromogenictim, M. .smeffmalis or Nocardia a.steroides were added to the intermediate gel of the M. leprae CIE system, many of the lines were pulled downwards, showing that the antiserum contained cross-reacting antibodies, and indicating that the corresponding mycobacterium contained a structurally related component. Most of the components of M. leprae seemed to react with one or more of the heterologous antisera. Only a few lines, which were weak, remained unaffected by all of the antisera. These weak lines may represent M. leprae specific components. To add another serum to the intermediate gel of a CIE reference system is a sensitive way of detecting antibodies in that serum. The
antibodies in sera of patients with lepromatous leprosy were examined in this manner, adding a few selected sera and two preparations of concentrated immunoglobulin from pooled lepromatous sera to the intermediate gel of the M. leprae reference system. In the lepromatous serum pools, antibody activity was demonstrated against a very restricted number of antigens (Fig. 2). Some of the individual sera reacted with as many as ten to twelve components, but a number of these antibody specificities were obviously not frequent enough to avoid being diluted out in a serum pool. No additional lines were seen in the intermediate gel. The components reacting with antibodies in lepromatous sera were among those most frequently involved in reactions with antisera to other mycobacteria, which indicated that these components are not M. leprae specific but that structurally related components occur in many mycobacteria.
FIG. 2. CIE of Mycobacterium leprae sonicate. In the top gel. concentrated immunoglobulins from rabbit hyperimmunized with M. leprae, and in the intermediate gel, concentrated immunoglobulins from a pool of sera from patients with lepromatous leprosy. The lepromatous serum pool contained strong antibody activity againsl the components marked wilh two arrows (Nos. 1, 5 and 7) and weak activity against components marked with one arrow (Nos. 3, 4, 6 and one of the additional components). The numbering is ihe same as in Fig. 1.
Antigens of Mycobacterium leprae 301 DISCUSSION The difficulties in developing a reference system for M. leprae which defines a large number of antigens could indicate that the available preparations of leprosy bacilli contain fewer antigenic components than other mycobacteria. Both experimental rabbit antisera and sera from lepromatous patients have failed to show antibody activity against more than seven or eight different components in CIE [5, 7] and eleven to twelve lines in immunodiffusion [11]. Why more components were detected by simple immunodiffusion than by CIE is difficult to explain, since the latter technique is usually more sensitive. However, the rabbit antisera used in the double diffusion study also reacted with human serum [11], and it was not clearly shown that all eleven precipitin lines corresponded to M. teprae derived components. Based on the present results, it is evident that M. leprae is not very different from other mycobacteria concerning the number of components which elicit antibody formation in rabbits. The antiserum produced by our rabbit contained antibodies against twenty-four different components, but the M. leprae preparationclearly contained more components, since several additional lines developed when other mycobacteria! antisera were used. It was shown that neither armadillo liver extract, nor antiserum against armadillo liver antigens, affected any of the lines when incorporated into the intermediate gel of our M. teptac reference system. Therefore, none of the components in the system seem to be of armadillo origin. Different lines of evidence indicate that M. leprae. isolated from human biopsies or from infected armadillo tissues, contains relatively few intact bacilli. A high percentage of the bacilli may be cell walls with little of the cell constituents left inside. Most of the lines in our mycobacterial reference systems appear to represent components of the cytosol fraction, while only few components have any clear association with the cell wall. To develop a multicomponent reference system for M. teprae antigens, we shall have to work with concentrated antigens both for immunization and in CIE, because many components are probably present in very low concentration in the extracts we make. The establishment of an improved reference system for M. leprae will allow a more detailed
analysis of its relationship to other mycobacteria. It will also create an importatit tool with which to characterize the antigenic composition of various preparations of M. leprae. Possibly the system can be improved and the number of antigens defined by it further increased. Numbering of more components should wait until a new, larger batch of polyvalent antibodies has been produced. The antibody activity in sera from patients with lepromatous leprosy has been studied in a system permitting detection of antibodies against more than twenty immunologically defined components of M. teprae. It is striking that antibodies were found only against a very restricted number of these components. The components that induce antibody formation are widely cross-reacting; that is, components with similar structure are found in many mycobacterial species. The humoral immune response in lepromatous leprosy may therefore represent an anamnestic response to antigens against which other mycobacteria have primed the immune system. This hypothesis is supported by the finding of antibodies against some of the same components in rabbits irnmunized with very low doses of BCG antigen [8] and in mice infected with M. Icpracmttrium [3]. Lepromatous patients are thought to lack specific cell mediated itiimunity but to have a strong humoral immune response to M. teprae [12]. The present results should perhaps make us consider the possibility that, except for a few antigens against which they produce an anamnestic antibody response, lepromatous patients lack both cell mediated and humoral immune responses to the bacillus.
ACKNOWLEDGMENTS This work was supported by grants frotn Anders Jahre's Fund for the Promotion of Science, The Norwegian Research Council for Science and the Humanities and by the World Health Organization through its Immunology of Leprosy Programme (IMMLEP). We thank R. J. W. Rees and P. Draper of the National Institute for Medical Research, London, for purified M. teprae. K. Negassi for supplying anti-armadillo serum and Grethe S. Fuglestad for excellent technical assistance.
302
O. CIOS.S, R. N. Mshana & M. Harboe
REFERENCES 1 Axelsen, N.H., Kroli, J. & Weeke. B. (eds.) Quantilative immunoelectrophoresis. Methods and applications. Scand. J. Immunol, 2, Suppl. 1. 1973. 2 Closs. O., Harboe, M. & Wassum, A.M. Cross reactions between mycohacteria. F. Crossed immunoeleetrophoresis of soluble antigens of Mycobacterium lepraemurium and comparison with BCG. Scand. J. Immimol. 4. Suppi. 1. 173, 1975. 3 Closs. 0 . & Kronvall. G. Experimental murine leprosy IX. Antibodies against Mycobacterium lepraemurium in C.IH and C57B1 mice with murine leprosy and in patients with lepromatous leprosy. Scand. J. Immunot. 4, 735, 1975. 4 Draper, P. Cell walls of Myeobacterium leprae. Int.J. Lepr. 44,^)5. 1976. 5 Harboe, M.. Closs. O., Bjorvatn. B.. Kronvall, G. & Axelsen, N.H. The antibody response in rabbiis to immunization with Mycobacteriittu leprae. Infect. Immunity. 18,792, 1977. 6 Harboe, N. & Ingild, A. Immunization, isolation of immunoglobulins and estimation of antibody titre. Scand. J. Immunol. 1, Suppl. 1. 161. 197.1.
7 Kronvall, G., Bjiine, G., Stanford. J.L., Menzel, S, & Samuel, D. Mycobacterial antigens in antibody responses of leprosy patients. Int. J. Lepr. 43, 299, 1975. 8 Mshana, R.N., Closs, O. & Harboe, M. Antibody response in rabbits to Mycobacterium bovis BCG. Scand. J. Immunol. 9, 175. 1979. 9 Roberts, D.B.. Wright. G.L.. Jr, Affronti, L.F. & Reich, M. Characterization and comparison of mycobacterial antigen by two dimensional immunoelectrophoresis. Infect. Immunity. 6, 567, 1972. 10 Stanford, J.L. Immunodiffusion analysis. A rational basis for the taxonomy of mycobacteria. Ann. Soc. beige Med. trop. 53, 321.1973. 11 Stanford, J L . . Rook. G.A.W., Convit, J.. Godal, T., Kronvall, G., Rees. R.J.W. & Walsh, G.P. Preliminary taxonomic studies on ihe leprosy bacillus. Br. J. exp. Path. 56, 579, 1976. 12 Turk, J.L. & Bryceson, A.D.M. Immunological phenomena in leprosy and related diseases. Adv. Immunol, 13, 209, 1971. Received 16 October 1978 Received in revised form 4 January 1979
SHORT PAPER
Antigenic Analysis of Mycobacterium leprae O. CLOSS, R. N. MSHANA & M. HARBOE University of Oslo, Institute for Experimental Medical Research, UUeval Hospital, Oslo, Norway
Closs, O., Mshana, R.N. & Harboe, M. Antigenic Analysis of Mvcobmteritim leprae. Sca/id. J. Immunol. 9, 297-?.O2, 1979.
About twenty distinct antigenic components have been demonstrated in Mycobacterium leprae (M. leprae) by crossed immunoelectrophoresis against a rabbit antiserum produced by immunization wilh concentrated M. leprae antigen. This system allows a more detailed analysis of the antigenic relationship between M. leprae and olher mycobacteria and a better characterization of the antigenic content of various M. teprae preparations than wiih previously available antisera which reacted with tar fewer components. The anlibody activity in sera of patients with lepromatous leprosy was studied by incorporating the sera into the intermediate gel of the M. leprae reference system. Antibodies were found againsl only seven of the components. Since those compared are all known to be cross-reacting widely with antigens of oiher mycobacteria, it is speculated that cross-immunization may influence the antibody response in lepromatous leprosy. O. Closs, University «f Oslo, Institute for Experimental Medical Research. Ullevat Hospital, Oslo I, Norway.
The outcome and clinical manifestations after infection with M. leprae depend to a large extent on the host's immune response to the bacterium, but the major antigens involved in these reactions are at present unknown. A better characterization of the antigenic composition of M. leprae may be an important first step and may also help to clarify the antigenic relationship between this bacterium and other mycobacteria. With the supply of M. leprae from infected armadillos it has become possible to provide sufficient material for such antigenic studies. The antigenic structure of mycobacteria has been studied by various gel precipitation techniques [2, 9, 10], By crossed immunoelectrophoresis (CIE), between twenty and fifty different antigenic components (hereafter called components) are usually easily demonstrable in any mycobacteriutn. Therefore, it was disappointing when only seven components could be clearly defined in our first attempt to establish a CIE reference system for M. leprae [5]. 0300-9475/79/0300-0297 $02.00
Antisera against other mycobacteria frequently contained antibodies cross-reacting with these components, indicating that components with similar structure occur in many mycobacteria, and that they were not M. leprae specific. Moreover, antibodies against the same components were frequently found in the serum of patients with iepromatous leprosy [5]. During a mycobacterial infection cross-reactive antigens may be more immunogenic than the species specific ones [3]. This suggestion was supported by the results obtained during immunization of rabbits with different doses of M. bovis (BCG) [8]. Even with minute doses of BCG sonicate a relatively strong antibody response remained against a few highly crossreactive antigens. Comparison of antigenic extracts of armadillo-grown M. teprae and other mycobacteria by acrylamide gradient gel electrophoresis has revealed that the M. leprae preparation produced much fewer and weaker bands than preparations made in a similar way from cultivable mycobacteria [5]. This led us to
i: 1979 Blackwell Scientific Publications
297
298
O. CI0.SS, R. N. Mshana & M. Harhoe
assume that the antibody response in rabbits could be improved by increasing the dose of M. leprae antigen used for immunization.
MATERIALS AND METHODS Antigens. M. Icprac wn-i provided by the Worid Health Organization through its Immunology of Leprosy flMMLEP) programme. The bacilli, hatch AB19. were obtained as purified, lyophllized material, prepared from infected armadillo tissue by Dr P, Draper of the National Institute for Medical Research, Mill Hill. London [4]. Sonic extracts of the bacilli were prepared in the following way: 60 mg of bacilli (wet weight) or 6 mg of lyophilized .M. Icprac were added per millilitre of 0.9" o NaCI. and the suspension was sonified for 15 min using the Branson BU sonifier (Branson Sonic Power Co., Danbiiry. Conn.) with an effect of 90-100 W. If ihc extract was to be used in CIE, it was centrifuged al 20,000 g for 20 min to remove insoluble material. For immunization of rabbits the extract was mixed with an equal volume of Freund's incomplete adjuvant (FIA). In the present experiment the M. teprac sonicate was concentrated 10-fold in a Minicon Bcell(Amicon Corp.. Lexington. Mass., tJSA) before it was emulsified with FIA. The emulsion was kept at - 20 C between immunizations. Anii.sera. Antisera against a number of mycobacteria were prepared by hyperimmunizatlon of rabbits wiih sonified suspensions of Ihc bacteria in FIA as previously described |5). Anti-BCG immunoglobulin was kindly provided by Dakopatts A/S, Copenhagen (Code B124. lot 116). A rabbit antiserum against armadillo liver antigens was kindly provided by Mr Kesete Negassi. Imnuiniziiiion .schi-ilute. One rabbit (No. 042) was immunized with concentrated M. leprae antigen on days 0 and 14 and every 28th day thereafter. The antigen was injected intracutaneously at multiple sites in the neck region. Owing to technical difticullies some injections became located subcutaneously. To follow the immune response closely, the rabbit was bled on day 0 (pre-immune serum), day 14. and every week thereafter uniil day 160. The sera were slored with 15 mM NaN^aM C. Fructionaiion of ininiiinoghbutin. Immunoglobulins were isolated by (NHiJjSOi precipitation [6] of 20-25 ml of serum pooled from several bleedings. The purified immunoglobulin fractions were stored at 4 C with 15 niM NaN3 added. Cros.sed inwiunoclccfrophoresis. CIE with intermediate gel was carried out and the plates were washed, pressed and stained as described in detail elsewhere [I]. Briefly. CIE was performed on 5 v 5 cm glass plates in 1','-,, agarose (Litex type HSA with moderate electroendosmotic flow, Litex, Glostrup. Denmark) containing TRIS/barbital butTer, pH 8.6, and ionic strength 0.02, 10 [J1 of antigen were added in a circular well, 4 mm in diameter. The top gel contained 150 rj| antiserum and the intermediate gel contained 100 |jl buffer or antiserum.
RESULTS in the rabbit immunized with concentrated M. leprae antigen (No. 042) the first detectable antibodies appeared 2 weeks after the first immunization when three precipitin lines were seen in CIE, corresponding to M. leprae antigens 2, 5 and 7. Bolh the nutTiber of antibody specificities and the antibody concentrations increased with subsequent immunizations until a plateau was reached after the fifth immunization. Thereafter, the antibody concentrations remained fairly constant, although the activity against some components showed a tendency to decrease after the sixth immtmization. The highest number of lines was seen when concentrated immunoglobulins and 4- or 10fold concentrated antigens were used. Fig, 1 shows the same M. leprae antigen run in CIE against immunoglobulin prepared from rabbit 042 and against a pool of sera from rabbits immunized with regular doses; a marked diiTerence is seen. With the latter antibody preparation, only seven precipitin lines were visible, whereas more than twenty lines developed in the plate containing 042 immunoglobulins. The seven precipitin lines seen with the M. leprae serum pool corresponded to components 1. 2, 3, 4, 5, 7 and 8 of the previous reference system [5]. With the exception of component 1, these antigens all formed very distinct precipitin lines with the 042 immunoglobulin preparation. In addition, between eight and ten other distinct lines and a similar number of weaker lines were seen. Thus, altogether, the new precipitin pattern contained about twentyfour lines. Some of them were quite weak, however, and were sometimes lost during destaining. To test whether any of the components were of host origin, an antiserum against armadillo liver was incorporated in the intermediate gel. This antiserum produced twenty lines against armadillo liver in CIE but did not react with any of the antigens precipitated in the M. leprae system. When M. leprae antigen was run in CIE against antisera to other mycobacteria. notably anti-BCG, anti-M. avium and anti-M. Icpraemurium, a few more components were detected in addition to those precipitated by the 042 immunoglobulin preparation. Thus, the sonic extracts of M. /p/>r«£'probablycontainmorethan
Atitigens of Mycobacterium leprae
FIG. 1. Drawings of two CIE plalcs where the same Mycobacterium leprae sonicate was run against Iwo differeni antibody preparations. The intermediate gel beiween the two horizontal lines tlid not contain antibodies. (A) Antiserum from rabbits immunized witli standard doses of antigen. (B) Concentrated immunoglobulins from rabbits immunized with increased amounts of antigen. (C) Photograph of the plate drawn in B, stained with Coomassie brilliani blue. The numbers refer to the reference system previously described by Harboe er al. [5].
299
300
O. Closs, R. N. Mshana & M. Harboe
twenty-four antigens, but some of them produce only weak lines and appear to be present in very low concentrations. When strong rabbit antisera against M. avium, M. duvalii, M. lepraemurium. M. noncliromogenictim, M. .smeffmalis or Nocardia a.steroides were added to the intermediate gel of the M. leprae CIE system, many of the lines were pulled downwards, showing that the antiserum contained cross-reacting antibodies, and indicating that the corresponding mycobacterium contained a structurally related component. Most of the components of M. leprae seemed to react with one or more of the heterologous antisera. Only a few lines, which were weak, remained unaffected by all of the antisera. These weak lines may represent M. leprae specific components. To add another serum to the intermediate gel of a CIE reference system is a sensitive way of detecting antibodies in that serum. The
antibodies in sera of patients with lepromatous leprosy were examined in this manner, adding a few selected sera and two preparations of concentrated immunoglobulin from pooled lepromatous sera to the intermediate gel of the M. leprae reference system. In the lepromatous serum pools, antibody activity was demonstrated against a very restricted number of antigens (Fig. 2). Some of the individual sera reacted with as many as ten to twelve components, but a number of these antibody specificities were obviously not frequent enough to avoid being diluted out in a serum pool. No additional lines were seen in the intermediate gel. The components reacting with antibodies in lepromatous sera were among those most frequently involved in reactions with antisera to other mycobacteria, which indicated that these components are not M. leprae specific but that structurally related components occur in many mycobacteria.
FIG. 2. CIE of Mycobacterium leprae sonicate. In the top gel. concentrated immunoglobulins from rabbit hyperimmunized with M. leprae, and in the intermediate gel, concentrated immunoglobulins from a pool of sera from patients with lepromatous leprosy. The lepromatous serum pool contained strong antibody activity againsl the components marked wilh two arrows (Nos. 1, 5 and 7) and weak activity against components marked with one arrow (Nos. 3, 4, 6 and one of the additional components). The numbering is ihe same as in Fig. 1.
Antigens of Mycobacterium leprae 301 DISCUSSION The difficulties in developing a reference system for M. leprae which defines a large number of antigens could indicate that the available preparations of leprosy bacilli contain fewer antigenic components than other mycobacteria. Both experimental rabbit antisera and sera from lepromatous patients have failed to show antibody activity against more than seven or eight different components in CIE [5, 7] and eleven to twelve lines in immunodiffusion [11]. Why more components were detected by simple immunodiffusion than by CIE is difficult to explain, since the latter technique is usually more sensitive. However, the rabbit antisera used in the double diffusion study also reacted with human serum [11], and it was not clearly shown that all eleven precipitin lines corresponded to M. teprae derived components. Based on the present results, it is evident that M. leprae is not very different from other mycobacteria concerning the number of components which elicit antibody formation in rabbits. The antiserum produced by our rabbit contained antibodies against twenty-four different components, but the M. leprae preparationclearly contained more components, since several additional lines developed when other mycobacteria! antisera were used. It was shown that neither armadillo liver extract, nor antiserum against armadillo liver antigens, affected any of the lines when incorporated into the intermediate gel of our M. teptac reference system. Therefore, none of the components in the system seem to be of armadillo origin. Different lines of evidence indicate that M. leprae. isolated from human biopsies or from infected armadillo tissues, contains relatively few intact bacilli. A high percentage of the bacilli may be cell walls with little of the cell constituents left inside. Most of the lines in our mycobacterial reference systems appear to represent components of the cytosol fraction, while only few components have any clear association with the cell wall. To develop a multicomponent reference system for M. teprae antigens, we shall have to work with concentrated antigens both for immunization and in CIE, because many components are probably present in very low concentration in the extracts we make. The establishment of an improved reference system for M. leprae will allow a more detailed
analysis of its relationship to other mycobacteria. It will also create an importatit tool with which to characterize the antigenic composition of various preparations of M. leprae. Possibly the system can be improved and the number of antigens defined by it further increased. Numbering of more components should wait until a new, larger batch of polyvalent antibodies has been produced. The antibody activity in sera from patients with lepromatous leprosy has been studied in a system permitting detection of antibodies against more than twenty immunologically defined components of M. teprae. It is striking that antibodies were found only against a very restricted number of these components. The components that induce antibody formation are widely cross-reacting; that is, components with similar structure are found in many mycobacterial species. The humoral immune response in lepromatous leprosy may therefore represent an anamnestic response to antigens against which other mycobacteria have primed the immune system. This hypothesis is supported by the finding of antibodies against some of the same components in rabbits irnmunized with very low doses of BCG antigen [8] and in mice infected with M. Icpracmttrium [3]. Lepromatous patients are thought to lack specific cell mediated itiimunity but to have a strong humoral immune response to M. teprae [12]. The present results should perhaps make us consider the possibility that, except for a few antigens against which they produce an anamnestic antibody response, lepromatous patients lack both cell mediated and humoral immune responses to the bacillus.
ACKNOWLEDGMENTS This work was supported by grants frotn Anders Jahre's Fund for the Promotion of Science, The Norwegian Research Council for Science and the Humanities and by the World Health Organization through its Immunology of Leprosy Programme (IMMLEP). We thank R. J. W. Rees and P. Draper of the National Institute for Medical Research, London, for purified M. teprae. K. Negassi for supplying anti-armadillo serum and Grethe S. Fuglestad for excellent technical assistance.
302
O. CIOS.S, R. N. Mshana & M. Harboe
REFERENCES 1 Axelsen, N.H., Kroli, J. & Weeke. B. (eds.) Quantilative immunoelectrophoresis. Methods and applications. Scand. J. Immunol, 2, Suppl. 1. 1973. 2 Closs. O., Harboe, M. & Wassum, A.M. Cross reactions between mycohacteria. F. Crossed immunoeleetrophoresis of soluble antigens of Mycobacterium lepraemurium and comparison with BCG. Scand. J. Immimol. 4. Suppi. 1. 173, 1975. 3 Closs. 0 . & Kronvall. G. Experimental murine leprosy IX. Antibodies against Mycobacterium lepraemurium in C.IH and C57B1 mice with murine leprosy and in patients with lepromatous leprosy. Scand. J. Immunot. 4, 735, 1975. 4 Draper, P. Cell walls of Myeobacterium leprae. Int.J. Lepr. 44,^)5. 1976. 5 Harboe, M.. Closs. O., Bjorvatn. B.. Kronvall, G. & Axelsen, N.H. The antibody response in rabbiis to immunization with Mycobacteriittu leprae. Infect. Immunity. 18,792, 1977. 6 Harboe, N. & Ingild, A. Immunization, isolation of immunoglobulins and estimation of antibody titre. Scand. J. Immunol. 1, Suppl. 1. 161. 197.1.
7 Kronvall, G., Bjiine, G., Stanford. J.L., Menzel, S, & Samuel, D. Mycobacterial antigens in antibody responses of leprosy patients. Int. J. Lepr. 43, 299, 1975. 8 Mshana, R.N., Closs, O. & Harboe, M. Antibody response in rabbits to Mycobacterium bovis BCG. Scand. J. Immunol. 9, 175. 1979. 9 Roberts, D.B.. Wright. G.L.. Jr, Affronti, L.F. & Reich, M. Characterization and comparison of mycobacterial antigen by two dimensional immunoelectrophoresis. Infect. Immunity. 6, 567, 1972. 10 Stanford, J.L. Immunodiffusion analysis. A rational basis for the taxonomy of mycobacteria. Ann. Soc. beige Med. trop. 53, 321.1973. 11 Stanford, J L . . Rook. G.A.W., Convit, J.. Godal, T., Kronvall, G., Rees. R.J.W. & Walsh, G.P. Preliminary taxonomic studies on ihe leprosy bacillus. Br. J. exp. Path. 56, 579, 1976. 12 Turk, J.L. & Bryceson, A.D.M. Immunological phenomena in leprosy and related diseases. Adv. Immunol, 13, 209, 1971. Received 16 October 1978 Received in revised form 4 January 1979
