Engineering high affinity superantigens by phage display

Protein L (PpL) is a B-cell superantigen from Peptostreptococcus magnus known to bind to mammalian Vkappa light chains. PpL from P.magnus strain 312 comprises five homologous immunoglobulin (Ig) binding domains. We first analysed the binding of the individual domains (B1-B5) of PpL(312) to human Vkappa light chains (huVkappa) subtypes 1 (huVkappaI) and 3 (huVkappaIII). Using a combination of rational design and phage selection we isolated mutants of the N-terminal B1 domain with a 14-fold increased affinity for huVkappa1 (B1kappa1) and >tenfold increased affinity for huVkappaIII (B1kappa3). We investigated the potential of the selected domains, in particular the B1kappa1 domain, as reagents in immunochemistry and immunotherapy. B1kappa1 proved a superior reagent than the wild-type domain, allowing up to tenfold more sensitive detection of human Vkappa antibody fragments in ELISA. A fusion protein of B1kappa1 with a human Vlambda antibody scFv fragment promoted the efficient recruitment of antibody encoded effector functions including complement, mononuclear phagocyte respiratory burst and phagocytosis through retargeting of IgGkappa and IgMkappa. Our results suggest that superantigens with improved affinity and/or specificity are easily accessible through protein engineering. Such engineered superantigens should prove useful as reagents in immunochemistry and may have potential as agents in immunotherapy.     

A method to confer Protein L binding ability to any antibody fragment

Recombinant antibody single-chain variable fragments (scFv) are difficult to purify homogeneously from a protein complex mixture. The most effective, specific and fastest method of purification is an affinity chromatography on Protein L (PpL) matrix. This protein is a multi-domain bacterial surface protein that is able to interact with conformational patterns on kappa light chains. It mainly recognizes amino acid residues located at the VL FR1 and some residues in the variable and constant (C_L) domain. Not all kappa chains are recognized, however, and the lack of C_L can reduce the interaction. From a scFv composed of IGKV10-94 according to IMGT®, it is possible, with several mutations, to transfer the motif from the IGKV12-46 naturally recognized by the PpL, and, with the single mutation T8P, to confer PpL recognition with a higher affinity. A second mutation S24R greatly improves the affinity, in particular by modifying the dissociation rate (k_d). The equilibrium dissociation constant (K_D) was measured at 7.2 10^-11 M by surface plasmon resonance. It was possible to confer PpL recognition to all kappa chains. This protein interaction can be modulated according to the characteristics of scFv (e.g., stability) and their use with conjugated PpL. This work could be extrapolated to recombinant monoclonal antibodies, and offers an alternative for protein A purification and detection.     

Protein L from Peptostreptococcus magnus binds to the kappa light chain variable domain.

Protein L is an immunoglobulin light chain-binding protein expressed by some strains of the anaerobic bacterial species Peptostreptococcus magnus. The major variable region subgroups of human kappa and lambda light chains were tested for protein L binding; V kappa I, V kappa III, and V kappa IV bound protein L, whereas no binding occurred with proteins of the V kappa II subgroup or with any lambda light chain subgroups. Studies of the protein L binding capacity of naturally occurring VL fragments, and VL- and CL-related trypsin- and pepsin-derived peptides prepared from a kappa I light chain, localized the site of interaction to the VL domain. The affinity constant for the binding to an isolated V kappa I fragment was comparable to that for the native protein (Ka 0.9 x 10(9) M-1 and Ka 1.5 x 10(9) M-1, respectively). No binding occurred with CL-related fragments. Extensive reduction and alkylation of the V kappa fragment or the native kappa chain resulted in complete loss of protein L binding. Although it is possible, from comparative amino acid sequence data, to identify certain VL-framework region residues that account for the selective binding of protein L by kappa I, kappa III, and kappa IV proteins, our studies indicate that this interaction is essentially dependent upon the tertiary structural integrity of the kappa chain VL domain.     

Evidence for plasticity and structural mimicry at the immunoglobulin light chain-protein L interface

The multidomain bacterial surface protein L (PpL) is a virulence factor expressed by only 10% of Peptostreptococcus magnus strains, and its expression is correlated with bacterial vaginosis. The molecular basis for its ability to recognize 60% of mammalian immunoglobulin light chain variable regions (V(L)) has been described recently by x-ray crystallography, which suggested the presence of two V(L) binding sites on each protein L domain (Graille, M., Stura, E. A., Housden, N. G., Beckingham, J. A., Bottomley, S. P., Beale, D., Taussig, M. J., Sutton, B. J., Gore, M. G., and Charbonnier, J. (2001) Structure 9, 679-687). Here, we report the crystal structure at 2.1 A resolution of a protein L mutant complexed to an Fab' fragment with only 50% of the V(L) residues interacting with PpL site 1 conserved. Comparison of the site 1 interface from both structures shows how protein L is able to accommodate these sequence differences and therefore bind to a large repertoire of Ig. The x-ray structure and NMR results confirm the existence of two V(L) binding sites on a single protein L domain. These sites exhibit a remarkable structural mimicry of growth factors binding to their receptors. This could explain the protein L superantigenic activity on human B lymphocytes.     

Synthesis and screening of a rationally designed combinatorial library of affinity ligands mimicking protein L from Peptostreptococcus magnus

Rational design and combinatorial chemistry were utilized to search for lead protein L (PpL) mimetics for application as affinity ligands for the purification of antibodies and small fragments, such as Fab and scFv, and as potential diagnostic or therapeutic agents. Inspection of the key structural features of the complex between PpL and human Fab prompted the de novo design and combinatorial synthesis of a 169-membered solid-phase ligand library, which was assessed for binding to human IgG and subsequent selectivity for the Fab fragment. Eight ligands were selected, chemically characterized and compared with a commercial PpL-adsorbent for binding pure immunoglobulin fractions. The most promising lead, ligand 8/7, when immobilized on an agarose support, behaved in a similar fashion to PpL in isolating Fab fragments from papain digests of human IgG to a final purity of 97%.     

Recombinant production of a VL single domain antibody in Escherichia coli and analysis of its interaction with peptostreptococcal protein L

A kappa-light chain from a Fab expression system was truncated by the insertion of a stop codon in the gene sequence to produce a variable light (VL) single domain antibody (dAb). Here, we describe the expression of dAb in the periplasm of Escherichia coli through fermentation in a defined media. Immunoglobulin binding domains from peptostreptococcal protein L (PpL) have been shown to bind specifically to kappa-light chains. We have produced recombinant PpL, at high yield, and this was used to custom-produce PpL-Sepharose affinity columns. Here, we show that the affinity purification of VL dAb by this method is simple and efficient with no apparent loss in protein at any stage. The truncated dAb protein product was confirmed by electrospray mass spectrometry and N-terminal sequencing. When analyzed by SDS-PAGE it was shown to be over 95% pure and produced at yields of 35-65 mg/L of culture medium. The dAb protein produced was shown by NMR and CD to be a folded beta-sheet domain. This domain is bound by PpL with a Kd of approximately 50 nM as determined by stopped-flow fluorimetry.     

Isolation and characterization of an active variable domain from a homogeneous rabbit antibody light chain.

The variable domain (VL) of allotype b4 light chains of rabbit IgG was isolated from both nonimmune heterogeneous IgG and a homogeneous antibody directed against type III pneumococcal polysaccharide. Light chains were first isolated and then cleaved under mild acidic conditions between residues 109 and 110. Reduction with dithiothreitol in guanidine hydrochloride cleaved both intradomain disulfide bridges as well as the interdomain disulfide bridge joining the variable and constant domain. The sulfhydryl groups were protected after reduction by p-chloromercuribenzoate. VL was isolated from this mixture of variable and constant domains by affinity chromatography, utilizing sheep antibodies directed against a peptide including residues 110--211 from nonimmune IgG light chain. The isolated VL domain was identified by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and automated Edman degradation. VL from a homogeneous antibody was treated with dithiothreitol to remove p-chloromercuribenzoate, reoxidized, and recombined with homologous heavy chain. The binding of this recombinant to type III pneumococcal polysaccharide was identical with that of the light-chain--heavy-chain recombinant.     

Alternative mechanism of protein A-immunoglobulin interaction the VH-associated reactivity of a monoclonal human IgM

The immunoglobulin site(s) that mediates the alternative mechanism of interaction between immunoglobulins and staphylococcal protein A (SpA) was studied by using a monoclonal human IgM. Several IgM fragments were tested for their inhibitory effect in a competitive binding assay of 125I-IgM to SpA. Only those fragments containing Fab mu pieces showed some inhibitory activity. The reactivity of the Fab mu region was retained in some of its subfragments, such as Fv or the VH domain, unlike isolated light chains or VL domains. Furthermore, antibodies specific for the VH domain completely inhibited the SpA-IgM interaction. These results indicate that the alternative SpA-binding site of IgM is located in VH regions.     

Domain association in immunoglobulin molecules. The packing of variable domains

We have analyzed the structure of the interface between VL and VH domains in three immunoglobulin fragments: Fab KOL, Fab NEW and Fab MCPC 603. About 1800 A2 of protein surface is buried between the domains. Approximately three quarters of this interface is formed by the packing of the VL and VH beta-sheets in the conserved "framework" and one quarter from contacts between the hypervariable regions. The beta-sheets that form the interface have edge strands that are strongly twisted (coiled) by beta-bulges. As a result, the edge strands fold back over their own beta-sheet at two diagonally opposite corners. When the VL and VH domains pack together, residues from these edge strands form the central part of the interface and give what we call a three-layer packing; i.e. there is a third layer composed of side-chains inserted between the two backbone side-chain layers that are usually in contact. This three-layer packing is different from previously described beta-sheet packings. The 12 residues that form the central part of the three observed VL-VH packings are absolutely or very strongly conserved in all immunoglobulin sequences. This strongly suggests that the structure described here is a general model for the association of VL and VH domains and that the three-layer packing plays a central role in forming the antibody combining site.     

Protein LG: a hybrid molecule with unique immunoglobulin binding properties.

Immunoglobulin (Ig)-binding bacterial proteins have attracted theoretical interest for their role in molecular host-parasite interactions, and they are widely used as tools in immunology, biochemistry, medicine, and biotechnology. Protein L of the anaerobic bacterial species Peptostreptococcus magnus binds Ig light chains, whereas streptococcal protein G has affinity for the constant (Fc) region of IgG. In this report, Ig binding parts of protein L and protein G were combined to form a hybrid molecule, protein LG, which was found to bind a large majority of intact human Igs as well as Fc and Fab fragments, and Ig light chains. Binding to Ig was specific, and the affinity constants of the reactions between protein LG and human IgG, IgGFc fragments, and kappa light chains, determined by Scatchard plots, were 5.9 x 10(9), 2.2 x 10(9), and 2.0 x 10(9) M-1, respectively. The binding properties of protein LG were more complete as compared with previously described Ig-binding proteins when also tested against mouse and rat Igs. This hybrid protein thus represents a powerful tool for the binding, detection, and purification of antibodies and antibody fragments.     

The specificity of chain interactions among immunoglobulins. Combinations of γ chains with κ chains of the same subgroup as in the parent immunoglobulin G

1. The specificity of combination of heavy and light chains from selected human immunoglobulins was examined in the light of greater structural information than in previous studies. Heavy (gamma) chains from immunoglobulin G (kappa) myeloma proteins were allowed to combine with their homologous light (kappa) chains or with other kappa chains of the same variable-region subgroup. The affinity of each such pairing was assessed by having the test kappa chain compete with a standard population of normal light chains. 2. There was a spread of affinities among the heavy-light pairings with the homologous pairings having an average affinity significantly higher than the heterologous pairings. 3. It follows that (a) the preference shown for homologous heavy-light pairings is not explicable simply in terms of the known subdivisions of the variable and constant regions of the chains, and (b) it is unlikely that those residues specifying the subgroups of kappa-chain variable regions have a predominant role in the formation of interchain bonds with the gamma-chain variable regions.     

Immunoglobulin superantigen protein L induces IL-4 and IL-13 secretion from human Fc epsilon RI+ cells through interaction with the kappa light chains of IgE

Peptostreptococcus magnus protein L is a multidomain bacterial surface protein that correlates with virulence. It consists of up to five homologous Ig-binding domains (B1-B5) that interact with the variable domain of Ig kappa L chains. Intact protein L stimulates the synthesis and the release of IL-4 and IL-13 from human basophils in vitro. A protein L fragment covering the Ig-binding domains B1-B4 also induced IL-4 and IL-13 release from basophils. There was an excellent correlation (r(s) = 0.82; p < 0.001) between the maximal percent IL-4 release induced by protein L and that induced by anti-IgE and between intact protein L and the B1-B4 fragment (r(s) = 0.90; p < 0.01). Removal of IgE bound to basophils markedly reduced the IL-4 release induced by anti-IgE, protein L, and B1-B4. Preincubation of basophils with protein L or anti-IgE caused complete cross-desensitization to subsequent challenge with the heterologous stimulus. IgE purified from myeloma patients PS and PP (lambda chains) blocked anti-IgE-induced IL-4 release, but not the releasing activity of protein L. In contrast, IgE purified from myeloma patient ADZ (kappa chains) blocked both anti-IgE- and protein L-induced secretion. Cyclosporin A, but not cyclosporin H, inhibited protein L-induced release of IL-4 and IL-13 from basophils. Thus, protein L acts as a bacterial Ig superantigen to induce the synthesis and release of IL-4 and IL-13 from basophils by interacting with kappa L chains of the IgE isotype.     

Murine V kappa 25 and V kappa 27 amino-acid sequences of C57B1/6 origin: monoclonal antibodies 17S29.1 and 22S25.1 specific for the group A-streptococcal polysaccharide

Antibodies 17S29.1 and 22S25.1 are monoclonal, hybridoma-derived gamma 3 kappa murine immunoglobulins with specificity for N-acetyl-glucosamine beta 1----3-linked to the L-rhamnose backbone structure, the immunodeterminant of the streptococcal Group A polysaccharide. The VL 17S29.1 amino-acid sequence is the third complete one reported from an antibody with this specificity, the second fully determined V kappa 25 structure and the first complete V kappa sequence of C57B1/6 origin derived from a carbohydrate-specific antibody. VL22S25.1 is a member of the V kappa 27 isotype of murine immunoglobulin VL regions. V kappa 17S29.1 and the determined part of the V kappa 22S25.1 sequence are compared to the previously described V kappa regions of streptococcal Group A polysaccharide-specific antibodies and to 12 selected partial and complete V kappa regions of antibodies with other specificities, predominantly to carbohydrate antigens. Both V kappa 17S29.1 and V kappa 22S25.1 increase the variability of known murine V kappa regions. They are the most homologous to the other V kappa regions derived from antibodies with streptococcal Group A polysaccharide specificity and share with them the amino-acid residue Arg74, so far characteristic for V kappa regions from antibodies with this specificity. The analysis of groups of independently expressed, highly homologous V kappa regions, namely V kappa 17S29.1 and V kappa 2S1.3 as one and V kappa 7S34.1 and V kappa 22S25.1 as a second group, offers the possibility of estimating the minimal number of V kappa germline genes involved in the immune response to the structurally defined streptococcal Group A polysaccharide antigen.(ABSTRACT TRUNCATED AT 250 WORDS)     

Light chain determines the binding property of human anti-dsDNA IgG autoantibodies

We have previously prepared human anti-double-stranded (ds) DNA IgG Fab clones using phage-display technology. Nucleotide sequence analysis of genes of immunoglobulin (Ig) heavy and light chain variable regions in these Fab clones suggested that the DNA-binding activity of the clones depended on light chain usage. To confirm the role of the light chain in antibody binding to DNA, we constructed in the present study's new recombined Fab clones by heavy and light chain shuffling between the original anti-dsDNA Fab clones. Clones constructed by pairing Fdgamma fragments with the light chain from a high DNA-binding clone showed high DNA-binding activities, whereas other constructed clones using light chains from low DNA-binding clones showed low DNA-binding activities. Our results indicate that light chains in anti-dsDNA antibodies can determine the DNA-binding activity of the antibodies. Ig chain shuffling of phage-display antibodies may be useful for investigating the molecular mechanisms for antigen-antibody binding of human autoantibodies.     

Alternate arrangement of PpL B3 domain and SpA D domain creates synergistic double-site binding to VH3 and Vkappa regions of fab

In our previous study, a kind of novel hybrid immunoglobulin (Ig)-binding proteins (IBPs) was obtained with the characteristic structure of alternately arranged Finegoldia magna (formerly Peptostreptococcus magnus) protein L (P. magnus protein L, PpL) B3 domain (B3) and Staphylococcal protein A (SpA) D domain (D). In this study, two representative molecules of these novel proteins, LD3 (B3-D-B3) and LD5 (B3-D-B3-D-B3) (LD3/5), showed substantially higher affinity for IgG-F(ab')2, IgM, and IgA than 4L (B3-B3-B3-B3) or SpA, which were also demonstrated by surface plasmon resonance detection. Further, LD5 showed much stronger binding to single-chain Fv (scFv) KM38 (V(H)3-V(kappa)I) than to KM41 (V(H)1-V(kappa)III) or KM36 (V(H)3-V(kappa)III). Competitive inhibition studies showed that 4L alone or in combination with SpA (4L + SpA) was a weaker inhibitor than LD3/5 in inhibiting LD3/5's binding to IgG-F(ab')2, IgM, or IgA. The computer modeling suggested that the B3-D arrangement in LD3/5 could simultaneously bind to V(H)3 and V(kappa). Thus, our results indicated for the first time that alternate arrangement of B3 and D domains creates synergistic double-site binding to V(H)3 and V(kappa) regions of fragment of antigen binding. The different competitive inhibition pattern of binding of LD5 to scFv KM38 by 4L + SpA suggested strict use of antibody conformation for this simultaneous double-site binding. The demonstration of this novel binding property would promote to achieve the designed hybrid IBPs for useful immunological applications.     

The critical role of arginine residues in the binding of human monoclonal antibodies to cardiolipin

Previously we reported that the variable heavy chain region (VH) of a human beta2 glycoprotein I-dependent monoclonal antiphospholipid antibody (IS4) was dominant in conferring the ability to bind cardiolipin (CL). In contrast, the identity of the paired variable light chain region (VL) determined the strength of CL binding. In the present study, we examine the importance of specific arginine residues in IS4VH and paired VL in CL binding. The distribution of arginine residues in complementarity determining regions (CDRs) of VH and VL sequences was altered by site-directed mutagenesis or by CDR exchange. Ten different 2a2 germline gene-derived VL sequences were expressed with IS4VH and the VH of an anti-dsDNA antibody, B3. Six variants of IS4VH, containing different patterns of arginine residues in CDR3, were paired with B3VL and IS4VL. The ability of the 32 expressed heavy chain/light chain combinations to bind CL was determined by ELISA. Of four arginine residues in IS4VH CDR3 substituted to serines, two residues at positions 100 and 100 g had a major influence on the strength of CL binding while the two residues at positions 96 and 97 had no effect. In CDR exchange studies, VL containing B3VL CDR1 were associated with elevated CL binding, which was reduced significantly by substitution of a CDR1 arginine residue at position 27a with serine. In contrast, arginine residues in VL CDR2 or VL CDR3 did not enhance CL binding, and in one case may have contributed to inhibition of this binding. Subsets of arginine residues at specific locations in the CDRs of heavy chains and light chains of pathogenic antiphospholipid antibodies are important in determining their ability to bind CL.     

Recombinant shark natural antibodies to thyroglobulin

As cartilaginous fish are the vertebrates most distal from man to produce antibodies, fundamental information regarding conservation and variation of the antigen binding site should be gained by comparing the properties of antibodies directed against the same antigen from the two species. Since monoclonal cell lines cannot be generated using shark B cells, we isolated antigen binding recombinant single chain Fv antibodies (scFv) comprising of the complete variable regions from shark light and heavy chains. Thyroglobulin was used as the selecting antigen as both sharks and humans express natural antibodies to mammalian thyroglobulin in the absence of purposeful immunization. We report that recombinant sandbar shark (Carcharhinus plumbeus) scFvs that bind bovine thyroglobulin consist of heavy chain variable regions (VH) homologous to those of the human VHIII subset and light chain variable regions (VL) homologous to those of the human Vlambda6 subgroup. The homology within the frameworks is sufficient to enable the building of three-dimensional models of the shark VH/VL structure using established human structures as templates. In natural antibodies of both species, the major variability lies in the third complementarity determining region (CDR3) of both VH and VL.     

Primary structures and chain dominance of anti-DNA antibodies

Using several anti-DNA autoantibodies, we analyzed the relative involvement of heavy and light chains in their interactions with DNA. We previously obtained eight hybridomas producing monoclonal anti-DNA autoantibodies by fusing spleen cells from an MRL-lpr/lpr mouse with myeloma cells. The chain dominance was analyzed by UV cross-linking experiments, in which the antibodies were covalently cross-linked with radioisotope-labeled oligonucleotides by short-wavelength UV-light, and the cross-linked H and L chains were analyzed by SDS-PAGE and densitometric scanning. Among these, three were found to be heavy chain dominant antibodies in which heavy chains are dominantly involved in DNA binding. The other five were co-dominant antibodies in which both heavy and light chains are involved in DNA binding. To determine the factor(s) that can explain the chain dominance in DNA binding, we determined the amino acid sequences of the variable regions of both heavy (VH) and light (VL) chains of all eight monoclonal antibodies. By analyzing the data, we were able to draw the following conclusions: (1) The arginine residues are found in the CDR3 regions of both VH and VL of the co-dominant antibodies; whereas, the same residues are found only in the CDR3s of VH, but not in VL, of the heavy chain dominant antibodies. (2) The net charges of the V regions affect the chain dominance. From the results of this study it is suggested that the presence of arginine residue in CDR3 is a critical factor in determining chain-dominance, as well as DNA binding of anti-DNA antibodies in general.     

Three-dimensional structure of an Fv from a human IgM immunoglobulin.

An IgM(kappa) immunoglobulin from a patient (Pot) with Waldenstrom's macroglobulinemia was hydrolyzed with pepsin to release a fragment consisting of the 'variable' (V) domains of the light and heavy chains plus eight residue 'tails' from the 'constant' (C) domains. The crystal structure of this fragment was determined at 2.3 A resolution by molecular replacement and crystallographic refinement methods. When examined separately, the light chain component closely resembles another human kappa chain (Rei) in both the beta-pleated sheet regions and the 'hypervariable' loops. The conserved pleated sheets in the heavy chain are similar to those in the human Kol IgG1 protein, but the third hypervariable loop in particular is different from that in any immunoglobulin structure described to date. As in the Kol protein, this loop blocks the access to any internal active site along the light-heavy chain interface. Unlike the Kol protein, however, the loop does not protrude beyond the boundaries of a conventional antigen combining site. Instead, it forms a very compact structure, which fills almost all residual space between the domains. This is an example of one dominant complementarity-determining region (CDR) essentially negating the diversity possible with five other CDRs in the two chains. Ordered water molecules are associated with light chain constituents along the interface, but not with CDR3 of the heavy chain. In screening exercises the Pot IgM failed to bind a wide variety of peptides. Together, the results suggest that ligand binding can only occur on external surfaces of the protein. These surfaces carry a limited number of side chains usually assigned to CDRs in more typical antibodies.     

Antibody domain mutants demonstrate autonomy of the antigen binding site.

We have constructed derivatives of a lambda I light chain-bearing anti-(4-hydroxy,3-nitrophenyl)acetyl (NP) antibody which have the V regions exchanged between heavy chains and light chains of the kappa or lambda I type. These antibodies are assembled and secreted normally, and bind haptenic and macromolecular ligands like the wild-type; similar results are obtained for monovalent heterodimers of VHCL and lambda I light chains. The observed independence of the binding site from the constant region context argues against a role of longitudinal interactions between constant and variable domains in antigen recognition, and therefore against cooperativity between binding sites.