Functional mimicry: elicitation of a monoclonal anti-idiotypic antibody hydrolizing ß-lactams

^a Laboratoire de Technologie Enzymatique-UPRES-A CNRS 6022, Université de Technologie de Compiègne-BP 20529, 60205 Compiègne Cedex, France

	ABSTRACT

TOP ABSTRACT INTRODUCTION ANTI-IDIOTYPES AND CATALYTIC... THE ROLE OF ANTI-IDIOTYPIC... CONCLUSIONS REFERENCES

 
Antigen mimicry by anti-idiotypic antibodies is investigated as a reliable strategy to achieve molecular imprinting of an enzymatic activity. A monoclonal anti-idiotypic antibody (Ab2-9G4H9) was elicited by using a monoclonal antibody (Ab1-7AF9) specific for the ß-lactamase active site. Catalytic features of Ab2 were characterized with ß-lactamase substrates. The antibody combining site appeared to have retained a part of the catalytic specificity. The relevance of the idiotypic mimicry concept for the generation of catalytic antibodies was further demonstrated by eliciting a third generation antibody (Ab3), which was shown to recognize ß-lactamase: the complete internal image properties of Ab2 9G4H9, including binding and catalytic properties, were thus checked.—Avalle, B., Thomas, D., Friboulet, A. Functional mimicry: elicitation of a monoclonal anti-idiotypic antibody hydrolizing ß-lactams. FASEB J. 12, 1055–1060 (1998)

Key Words: mimicry • catalytic antibody • abzyme • anti-idiotype • ß-lactamase

	INTRODUCTION

TOP ABSTRACT INTRODUCTION ANTI-IDIOTYPES AND CATALYTIC... THE ROLE OF ANTI-IDIOTYPIC... CONCLUSIONS REFERENCES

 
DESIGN OF PROTEINS by molecular bioimprinting procedures is currently attracting attention because of its tremendous potential for eliciting catalysts endowed with tailor-made properties. Over the last decade, the most intriguing results were obtained with antibodies catalyzing the hydrolysis of predesigned substrates. Since Jencks' proposal (1) that antibodies directed against haptens mimicking features of the rate-determining transition state could increase the rate of transformation of the corresponding compounds, a wide range of reactions have been tackled (2, 3), from the less complex (4, 5) to the most disfavored ones (6, 7). Elicited antibodies generally display exquisite specificity and weak catalytic capabilities (for recent reviews, see refs 8–10). Other applications may encounter restrictions in the rationality of the hapten design, since structural data and efficient synthetic strategy have to be available.

An alternative strategy has been developed in the present work that aims to elicit antibodies that mimic the active site of a model enzyme. This approach relies on the fact that besides the traditional scheme of antibody–antigen interaction, the immune system has proved to be able to generate antibodies specific for idiotypic (or antigen-specific) determinants expressed within the variable region of other antibodies. This property led Jerne (11) to regard the immune system as a network of interacting idiotypes. According to Jerne's proposal, for each immunoglobulin (Ab1) generated against an antigenic determinant, there exists a complementary antibody (anti-idiotypic, Ab2) directed against the idiotypic determinants of Ab1. Among the Ab2 molecules, some (Ab2ß) recognize idiotopes in the combining site of Ab1 and mimic the antigen. Such antibodies are designed as internal images of the antigen. Although the physiological role of anti-idiotypic antibodies as a mechanism for regulating the immune response is sometimes tainted with doubt, the molecular mimicry phenomenon is widely documented by structural studies (12, 13). Since structural and electrostatic features of the model antigen can be exhibited by Ab2, it may be possible to mimic an enzymatic active site as long as information related to catalytic residues is retained through the idiotypic cascade. This paper describes the elicitation of a monoclonal antibody displaying ß-lactamase-like activity. This approach calls on knowledge at the crossroads of various fields such as general protein chemistry, enzymology, and immunology, as well as protein–protein interactions, for which the scope of applications is wide. Mimicry of enzymatic sites by anti-idiotypic antibodies resulting from immune regulation dysfunction has been proposed to explain the appearance of abzymes in autoimmune disorders (14–17).

	ANTI-IDIOTYPES AND CATALYTIC ACTIVITY

TOP ABSTRACT INTRODUCTION ANTI-IDIOTYPES AND CATALYTIC... THE ROLE OF ANTI-IDIOTYPIC... CONCLUSIONS REFERENCES

 
A preliminary paper attempted to elicit an anti-idiotypic antibody that exhibits cholinesterase-like activity (18). A monoclonal immunoglobulin M (IgM)² was obtained by immunizing mice with an IgG binding to a part of the acetylcholinesterase active site (19). This IgM was characterized as one of the most efficient abzymes described so far (k_cat=81 s^-1). However, this raised many questions. Since cholinesterase activity is widespread in eukaryotic expression systems used to produce monoclonal antibodies, part of the detected activity may be due to a catalyst contaminating the preparation. Moreover, the eliciting Ab1 was purchased from the American Type Cell Collection (cell line HB73), which prevented our approach from being fully validated. The focus of this preliminary work on esterase activity rests mainly on the demonstration of the potential to mimic a catalytic activity by this method. The extent and reproducibility of the anti-idiotypic approach for generating catalytic antibodies had to be evaluated, avoiding, as much as possible, drawbacks encountered with the esterase activity. Finally, another step can be reached by considering the elicitation of a catalytic activity that does not exist in the systems used to express antibodies to warrant the relationship between antibody and activity and to consider potential therapeutic applications.

The generation of antibodies that would catalyze the opening of ß-lactam ring antibiotics, an activity naturally catalyzed by microbial ß-lactamases, appeared to be an appropriate model for demonstrating the relevance of the anti-idiotypic mimicry scheme. ß-Lactamase A from Bacillus cereus (20) catalyzes, by formation of an acyl-serine intermediate, the hydrolysis of penicillin and cephalosporin substrates, as ampicillin (1) and PADAC (2), with catalytic constants of 2.8 x 10⁵ min^-1 and 7.53 x 10² min^-1, respectively, in phosphate buffer 0.1 M pH 7.4 (see Scheme ). The enzyme displays a 10-fold lower K_M value for PADAC than for ampicillin (22 vs. 200 µM). Immunization of Balb/C mice was performed using 40 µg ß-lactamase for each of the four 2-wk interval injections. One hundred fifty monoclonal antibodies specific for B. cereus ß-lactamase A were obtained by standard protocols (21) and selected by enzyme-linked immunosorbent assay (ELISA), in which 2 µg of ß-lactamase was coated at solid phase.

View larger version (13K): [in this window] [in a new window]   Figure S1.

The influence of Ab1 on enzymatic activity was determined as the most direct and valuable criterion for active site recognition. Experimental data were confirmed by statistical Student's t test calculations, based on quantitative results of experiments repeated three times. Nine monoclonal antibodies were selected for their putative inhibitory effect on ß-lactamase activity. The influence of Ab1 purified on a protein A-agarose column was assessed by determining k_cat and K_M values of ß-lactamase-Ab1 complexes. One monoclonal Ab1, 7AF9, appeared to induce a significant change of kinetic parameters: k_cat values of PADAC hydrolysis were reduced by 50% in its presence, whereas K_M increased from 20 to 73 µM. k_cat/K_M was significantly lowered by 7AF9 binding, presumably because of a disturbing acylation step. A plot of ß-lactamase residual activity in the presence of 7AF9 ( Fig. 1) indicates that the inhibitory effect depends on the antibody concentration. Lineweaver-Burk linear regression ( Fig. 1, inset) of experimental data indicates that inhibition follows a mixed-type mechanism, with a K_i value of about 1 x 10^-5 M. The addition of 0.75 mM benzylpenicillin in antigen recognition ELISA experiments was shown to alter 7AF9 binding by 37%. From these results, a rough epitopic pattern can be drawn: 7AF9 appears to recognize some residues of the ß-lactamase cavity that participate in catalysis.

View larger version (21K): [in this window] [in a new window]   Figure 1. Inhibition of ß-lactamase activity by monoclonal antibody 7AF9. Measurements of PADAC hydrolysis (15 µM) were performed in 0.1 M phosphate buffer (pH 7.4) containing 0.5% gelatin. ß-Lactamase concentration was fixed at 5 nM. Effect of 7AF9 on ß-lactamase activity () was compared to that of another anti ß-lac~tamase antibody, 110D9 (). Inset: Lineweaver-Burk linear regression. 7AF9 () and anti ß-lactamase monoclonal antibody ().

Biozzi mice, a highly responsive mice strain, were injected with 150 µg 7AF9. After each immunization, the appearance of Ab2 in blood samples was tested for Ab2 binding to 7AF9 and for potential catalytic activity. Measurements of immunological activity of antiserum specific for 7AF9 were performed on 200-fold diluted antiserum, using a test in a microtitration plate where 2.5 µg 7AF9 F(ab')₂ pepsin-digested fragments were adsorbed in solid phase. The resulting signal, indicating the anti-Ab1 level, was found to increase with the stage of immunization. Concurrently, hydrolysis of PADAC was assayed with 25-fold diluted antiserum previously incubated on plates coated with goat anti-mouse Ig (H+L) antibodies. The hydrolysis rate was found to increase with the amount of anti-7AF9 antibodies ( Table 1). When studied in buffered solution, PADAC hydrolysis catalyzed by polyclonal anti-7AF9 antibodies followed Michaelis-Menten kinetics, with a K_M value of 3 µM. Polyclonal antiidiotypic catalytic antibodies may therefore be regarded as highly selective catalysts. This efficiency of polyclonal preparation was described previously for catalytic antibodies from different sources (16, 22, 23). Furthermore, 20% of the cephalosporin hydrolysis activity linked to polyclonal antibodies was inhibited by increasing amounts of Ab1 7AF9, whereas other antibodies specific for ß-lactamase had no effect on catalytic activity. These results, obtained at a polyclonal level, unambiguously indicate that an enzyme-like activity, usually absent in mouse serum, appears after immunization with Ab1 7AF9. However, the polyclonal Ab2 preparations are heterogeneous mixtures of antibodies, with only some of them exhibiting catalytic activity. To further characterize the catalytic activity measured, monoclonal Ab2 were prepared.

Of 620 clones, 16 IgM and 17 IgG specific for 7AF9 F(ab')2 fragments were obtained. Monoclonal Ab2 display different immunological activities. Isotype subclasses were determined using the mouse monoclonal antibody isotyping Kit RPN 29 (Amersham Arlington Heights, Ill.). Screening for catalytic activity was performed with antibodies 99% purified on protein A according to BioRad instructions. The spectrum of hydrolysis of ß-lactam substrates by monoclonal Ab2 was investigated with both a penicillin (ampicillin) and a cephalosporin (PADAC). One IgG2b, 9G4H9, turned out to hydrolyze both substrates, following Michaelis-Menten features. Kinetic parameters for substrate hydrolysis were determined ( Table 2). As indicated by comparing K_M values related to ampicillin hydrolysis, the monoclonal IgG appears to have lost some substrate affinity properties. The high affinity for PADAC may be assigned to hydrophobic interactions involving the imine group of the C3 side chain. k_cat values exhibit moderate amplitudes, 10⁵-fold less than that measured with the natural enzyme. However, ampicillin hydrolysis efficiency by 9G4H9 is one of the highest when compared with antibodies previously generated to catalyze amide bond hydrolysis (24). The significance of experimental kinetic data is supported by rate enhancement factor calculations (for ampicillin, k_cat/fk_uncat=16x10³). Since IgG catalytic efficiency (k_cat/K_M value) is apparently lowered by a factor of 10⁶ in both cases when compared with the enzyme, the appearance of the acyl–abzyme complex is assumed to be very slow. Other investigations of IgG 9G4H9 catalytic properties show that catalytic activity slowly decreases when catalyst concentration increases beyond a critical point, experimentally assessed at 0.8 mg/ml ( Fig. 2). Although some reports mentioned the necessity of limited abzyme denaturation in order to obtain optimal activity (25), this phenomenon may also be explained by an inactivating oligomerization. In any case, this result corroborates the major contribution of 9G4H9 in catalysis. Environmental factors were shown to alter catalytic potencies: abzymatic activity appears to be 70% inhibited by 1M NaCl. Nevertheless, kinetic parameters were determined to vary with pH; for both substrates, the k_cat/K_M plots display an optimum at pH 7.0. According to curve inflection point analysis, a key group involved in catalysis ionizes at about pH 5.4, with a pKa similar to that of ß-lactamase (26).

View larger version (11K): [in this window] [in a new window]   Figure 2. Catalytic activity of Mab2 9G4H9 related to purified antibody concentration. Hydrolysis of PADAC was followed at 560 nm in 0.1 M phosphate buffer (pH 7.4).

	THE ROLE OF ANTI-IDIOTYPIC FEATURES IN CATALYSIS

TOP ABSTRACT INTRODUCTION ANTI-IDIOTYPES AND CATALYTIC... THE ROLE OF ANTI-IDIOTYPIC... CONCLUSIONS REFERENCES

 
The possibility that the presence of some nonspecific serine esterases or proteases could interfere with activity measurements by increasing the background substrate hydrolysis can be checked by using diisopropylfluorophosphate (DFP). This broad-spectrum, irreversible inhibitor of serine esterases failed to inhibit either ß-lactamase or antibody-related catalytic activity. DFP inhibition experiments were carried out in phosphate buffer 0.1 M (pH 7.4). 9G4H9 antibodies (10 nM of binding sites) and ß-lactamase (14 nM) were incubated for 2 h with 27 µM DFP. Residual activity was measured using 11 µM PADAC, and was higher than 95% of initial activity in both cases. Two arguments illustrate that the detected activity is independent of contingent interference: 1) ß-lactamase is strictly a bacterial activity; and 2) all preparations of other monoclonal Ab2 failed to display any catalytic activity. Because of all these data, IgG2b obviously appears to exhibit catalytic features derived from B. cereus ß-lactamase.

The main question concerns the effective part played by idiotypic interactions in the catalyst generation process. Since antibody–antibody interaction stoichiometry is complex, an extensive study of the inhibition of Ab2 catalytic activity by Ab1 7AF9 would not be the simplest way to demonstrate the role of the idiotypic cascade in the emergence of catalytic Ab2. The internal image assumption implies that the Ab2 are part of a complex network, with structural similarities between the antigen and Ab2, as well as Ab1 and Ab3, and so on. Consequently, the generation of anti-Ab2 antibodies (i.e. Ab3) properly connected to immunological characterization of expected anti-ß-lactamase properties turns out to provide the most valuable answer to this question (27). Ab2 9G4H9 (200 µg) injected into a Biozzi mouse and the resulting polyclonal Ab3 were tested for their complementary properties for both Ab2 9G4H9 and B. cereus ß-lactamase. Immunological evaluation of the recognition of Ab2 F(ab')₂ fragments by polyclonal Ab3 clearly showed that anti-Ab2 were generated during the successive steps of immunization ( Fig. 3A, B ). Inhibition of catalytic activity related to Ab2 9G4H9 might be erratic in this case, as generated Ab3 are polyclonal antibodies. An ELISA test performed using adsorbed ß-lactamase indicates that the appearance of anti-ß-lactamase antibodies is tightly linked to that of anti-Ab2 antibodies. This argument provides evidence that catalytic Ab2 are involved in a network, since the rise of anti ß-lactamase Ab3 can only be explained by anti-idiotypic interactions. These results support the hypothesis that 9G4H9 is an anti-idiotypic antibody because it displays abilities to elicit anti-ß-lactamase antibodies even though the immunized mouse has never been in contact with the enzyme. The significant loss of catalytic efficiency of Ab2 9G4H9 compared with the initial enzyme implies an expected decay in chemical information transmitted throughout the idiotypic cascade.

View larger version (47K): [in this window] [in a new window]   Figure 3. Production of Ab3. Binding of Ab3 antiserum on 2 µg ß-lactamase from Bacillus cereus () and 0.25 µg F(ab')2 fragments of Ab2 9G4H9 () evaluated by ELISA during the course of immunization. (A) IgG generation: dilution of antiserum is 1/320 () and 1/160 (). (B) IgM generation: antiserum is 320-fold diluted. The nonimmunized serum corresponds to the blood collected from the Biozzi mouse used for experiments before any immunization. I1, I2, and I3 correspond to each of the three immunizations. Error bars are represented.

	CONCLUSIONS

TOP ABSTRACT INTRODUCTION ANTI-IDIOTYPES AND CATALYTIC... THE ROLE OF ANTI-IDIOTYPIC... CONCLUSIONS REFERENCES

 
The demonstration that anti-idiotypic antibodies mimic a corresponding antigen is largely documented. These internal image properties of Ab2 not only were shown to resemble the original antigen at the level of primary structure, but they are also able to exhibit a functional activity that accurately mimics the activity of the antigen. Furthermore, the relevance of mimicry by anti-idiotypic antibodies has been backed up by many examples where anti-idiotypic antibodies elicit functional effects normally associated with the binding of physiological ligands (28, 29).

The present results provide evidence of the potential to use the remarkable properties of anti-idiotypic antibodies for generating antibodies that exhibit the activity of a model enzyme. The use of ß-lactamase in the present work lowers the probability of artifactual results due to contamination by serum enzymes, since this activity is absent in the serum of mammals. It therefore allows one to easily measure the emergence of a catalytic activity at the mere polyclonal level. The catalytic properties of the monoclonal Ab2 9G4H9 differ from ß-lactamase activity for both substrate affinity and catalytic efficiency. A structural study of its active site and the comparison with the structural features of the enzyme site will be the next step in understanding the links between catalytic activity and internal image properties. In any case, the generation of Ab3-recognizing ß-lactamase is indicative of a good conservation of structural features from Ab1 to Ab3 (30). The catalytic constant obtained for the amidase reaction on ampicillin hydrolysis is lowered when compared to enzyme, but remains high when compared to other amide–hydrolysis reactions catalyzed by antibodies. In addition to the indisputable contribution of this demonstration in explaining the amazing appearance of catalytic antibodies in some autoimmune diseases (31), this approach offers prospects of main therapeutic interest to vaccinology, either to replace a lacking enzyme or to induce an activity targeting a foreign substance (32).

	ACKNOWLEDGMENTS

 
We wish to thank Dr. Ladan Izadyar for fruitful scientific discussions and Dominique Mistro for technical assistance. Financial support was obtained from the French League against Cancer.

	FOOTNOTES

 
¹ Correspondence: alain.friboulet{at}utc.fr

² Abbreviations: DFP, diisopropylfluorophosphate; ELISA, enzyme-linked immunosorbent assay; Ig, immunoglobulin.

Received for publication January 5, 1998. Accepted for publication February 27, 1998.

	REFERENCES

TOP ABSTRACT INTRODUCTION ANTI-IDIOTYPES AND CATALYTIC... THE ROLE OF ANTI-IDIOTYPIC... CONCLUSIONS REFERENCES

 

1. Jencks, W. (1969) Catalysis in Chemistry and Enzymology, McGraw-Hill, New York
2. Pollack, S. J., Jacobs, J. W., and Schultz, P. G. (1986) Selective chemical catalysis by an antibody. Science 234, 1570–1573
3. Tramontano, A., Janda, K. D., and Lerner, R. A. (1986) Catalytic antibodies. Science 234, 1566–1570
4. Janjic, N., and Tramontano, A. (1989) Antibody-catalyzed redox-reaction. J. Am. Chem. Soc. 111, 9109–9110
5. Wagner, J., Lerner, R. A., and Barbas, C. F., III (1995) Efficient aldolase catalytic antibodies that use the enamine mechanism of natural enzymes. Science 270, 1797–1800
6. Janda, K. D., Shevlin, C. G., and Lerner, R. A. (1993) Antibody catalysis of a disfavored chemical transformation. Science 259, 490–493
7. Danishefsky, S. (1993) Catalytic antibodies and disfavored reactions. Science 259, 469–470
8. Jacobsen, J. R., and Shultz, P. G. (1995) The scope of antibody catalysis. Curr. Opin. Struct. Biol. 5, 818–824
9. Benkovic, S. J. (1996) Antibody catalysis: the key is in the pocket. Nature (London) 383, 23–24
10. Schultz, P. G., and Lerner, R. A. (1995) From molecular diversity to catalysis: lessons from the immune system. Science 269, 1835–1842
11. Jerne, N. (1974) Toward a network theory of the immune system. Ann. Immunol. 125c, 373–389
12. Pan, Y., Yuhasz, S. C., and Amzel, L. M. (1995) Anti-idiotypic antibodies: biological function and structural studies. FASEB J. 9, 43–49
13. Fields, B. A., Goldbaum, F. A., Ysern, X., Poljak, R. J., and Mariuzza, R. A. (1995) Molecular basis of antigen mimicry by an anti-idiotope. Nature (London) 374, 739–742
14. Bronstein, I. B., Shuster, A. M., Gololobov, G. V., Gromova, I. I., Kvashuk, O. A., Belostotskaya, K. M., Alekberova, Z. S., Prokaeva, T. B., and Gabibov, A. G. (1992) DNA-specific antiidiotypic antibodies in the sera of patients with autoimmune diseases. FEBS Lett. 314, 259–263
15. Bronstein, I. B., Vorobyev, S., Timofeev, A., Jolles, C. J., Alder, S. L., and Holden, J. A. (1996) Elevations of DNA topoisomerase I catalytic activity and immunoprotein in human malignancies. Oncol. Res. 8, 17–25
16. Paul, S., Volle, D. J., Beach, C. M., Johnson, D. R., Powell, M. J., and Massey, R. J. (1989) Catalytic hydrolysis of vasoactive intestinal peptide by human autoantibody. Science 244, 1158–1162
17. Tyutyulkova, S., Gao, Q. S., Thompson, A., Rennard, S., and Paul, S. (1996) Efficient vasoactive intestinal polypeptide hydrolyzing autoantibody light chains selected by phage display. Biochim. Biophys. Acta 1316, 217–223
18. Izadyar, L., Friboulet, A., Remy, M. H., Roseto, A., and Thomas, D. (1993) Monoclonal anti-idiotypic antibodies as functional internal images of enzyme active sites: production of a catalytic antibody with a cholinesterase activity. Proc. Natl. Acad. Sci. USA 90, 8876–8880
19. Sorensen, K., Brodbeck, U., Rasmussen, A. G., and Norgaard-Pedersen, B. (1987) An inhibitory monoclonal antibody to human acetylcholinesterase. Biochim. Biophys. Acta 912, 56–62
20. Matagne, A., and Frère, J. M. (1995) Contribution of mutant analysis to the understanding of enzyme catalysis: the case of class A ß-lactamases. Biochim. Biophys. Acta 1246, 109–127
21. Kohler, G., and Milstein, C. (1975) Continuous cultures of fused cells secreting antibody of predefined specificity. Nature (London) 256, 495–497
22. Gallacher, G., Searcey, M., Jackson, C. S., and Brocklehurst, K. (1992) Polyclonal antibody-catalyzed amide hydrolysis. Biochem. J. 284, 675–680
23. Resmini, M., Vigna, R., Simms, C., Barber, N. J., Hagi-Pavli, E. P., Watts, A. B., Verma, C., Gallacher, G., and Brocklehurst, K. (1997) Characterization of the hydrolytic activity of a polyclonal catalytic antibody preparation by pH-dependence and chemical modification studies: evidence for the involvement of Tyr and Arg side chains as hydrogen-bond donors. Biochem. J. 326, 279–287
24. Janda, K. D., Schloeder, D., Benkovic, S. J., and Lerner, R. A. (1988) Induction of an antibody that catalyzes the hydrolysis of an amide bond. Science 241, 1188–1191
25. Paul, S., Li, L., Kalaga, R., Wilkins-Stevens, P., Stevens, F. J., and Solomon, A. (1995) Natural catalytic antibodies: peptide hydrolyzing activities of Bence-Jones proteins and V_L fragments. J. Biol. Chem. 270, 15257–15261
26. Brannigan, J., Matagne, A., Jacob, F., Damblon, C., Joris, B., Klein, D., Spratt, B. G., and Frère, J. M. (1991) The mutation Lys234His yields a class A ß-lactamase with a novel pH-dependence. Biochem. J. 278, 673–678
27. Goldbaum, F. A., Velikovsky, C. A., Dall'Acqua W., Fossati, C. A., Fields, B. A., Braden, B. C., Poljak, R. J., and Mariuzza, R. A. (1997) Characterization of anti-anti-idiotypic antibodies that bind antigen and an anti-idiotype. Proc. Natl. Acad. Sci. USA 94, 8697–8701
28. Sege, K., and Peterson, P. A. (1978) Use of the anti-idiotypic antibodies as cell-surface receptor probes. Proc. Natl. Acad. Sci. USA 75, 2443–2447
29. Couraud, P. O. (1986) Structural analysis of the epitopes recognized by monoclonal antibodies to angiotensin II. J. Immunol. 136, 3365–3370
30. Ban, N., Day, J., Wang, X. H., Ferrone, S., and Mcpherson, A. (1996) Crystal structure of an anti-anti-idiotype shows it to be self-complementary. J. Mol. Biol. 255, 617–627
31. Shuster, A. M., Gololobov, G. V., Kvashuk, O. A., Bogomolova, A. E., Smirnov, I. V., and Gabibov, A. G. (1992) DNA hydrolyzing autoantibodies. Science 256, 665–667
32. Magliani, W., Conti, S., De Bernardis, F., Gerloni, M., Bertolotti, D., Mozzoni, P., Cassone, A., and Polonelli, L. (1997) Therapeutic potential of antiidiotypic single chain antibodies with yeast killer toxin activity. Nat. Biotechnol. 15, 155–158