Rapid discovery of protein interactions by cell-free protein technologies.

Cell-free transcription and translation provides an open, controllable environment for production of correctly folded, soluble proteins and allows the rapid generation of proteins from DNA without the need for cloning. Thus it is becoming an increasingly attractive alternative to conventional in vivo expression systems, especially when parallel expression of multiple proteins is required. Through novel design and exploitation, powerful cell-free technologies of ribosome display and protein in situ arrays have been developed for in vitro production and isolation of protein-binding molecules from large libraries. These technologies can be combined for rapid detection of protein interactions.

Autoantibodies to alpha-fodrin in primary Sjögren's syndrome and SLE detected by an in vitro transcription and translation assay.

OBJECTIVE: To investigate the prevalence of alpha-fodrin autoantibodies in primary Sjögren's syndrome (SS) and systemic lupus erythematosus (SLE) with and without secondary SS, using an in vitro transcription and translation assay (ITT). METHODS: cDNA encoding JS-1, the amino-terminal portion of alpha-fodrin, was used for ITT. Immunoprecipitation was performed with sera from 56 primary SS patients and 67 SLE patients, 14 with and 53 without secondary SS. Correlations to RF, ANA, anti-dsDNA, anti-SS-A and anti-SS-B antibodies, hypergammaglobulinemia, labial salivary gland biopsy grade, extraglandular manifestations and a modified SLE disease activity index (mSLEDAI) were made. RESULTS: Autoantibodies against alpha-fodrin were detected in 16/56 (29%) of primary SS patients and in 25/53 (47%) of sera from SLE patients without secondary SS. In SLE patients with secondary SS the prevalence was 3/14 (21%). None of the blood donors showed alpha-fodrin reactivity. Correlations were found to RF, ANA, anti-dsDNA antibodies and a positive mSLEDAI score. CONCLUSION: The frequency of alpha-fodrin autoantibodies detected by this method is similar in sera from primary SS patients and SLE patients with or without secondary SS. The presence of alpha-fodrin autoantibodies seems to reflect non-organ-specific autoimmunity in primary SS and SLE and to be of limited discriminating value.

Complex between Peptostreptococcus magnus protein L and a human antibody reveals structural convergence in the interaction modes of Fab binding proteins.

BACKGROUND: Peptostreptococcus magnus protein L (PpL) is a multidomain, bacterial surface protein whose presence correlates with virulence. It consists of up to five homologous immunoglobulin binding domains that interact with the variable (VL) regions of kappa light chains found on two thirds of mammalian antibodies. RESULTS: We refined the crystal structure of the complex between a human antibody Fab fragment (2A2) and a single PpL domain (61 residues) to 2.7 A. The asymmetric unit contains two Fab molecules sandwiching a single PpL domain, which contacts similar VL framework regions of two light chains via independent interfaces. The residues contacted on VL are remote from the hypervariable loops. One PpL-Vkappa interface agrees with previous biochemical data, while the second is novel. Site-directed mutagenesis and analytical-centrifugation studies suggest that the two PpL binding sites have markedly different affinities for VL. The PpL residues in both interactions are well conserved among different Peptostreptococcus magnus strains. The Fab contact positions identified in the complex explain the high specificity of PpL for antibodies with kappa rather than lambda chains. CONCLUSIONS: The PpL-Fab complex shows the first interaction of a bacterial virulence factor with a Fab light chain outside the conventional combining site. Structural comparison with two other bacterial proteins interacting with the Fab heavy chain shows that PpL, structurally homologous to streptococcal SpG domains, shares with the latter a similar binding mode. These two bacterial surface proteins interact with their respective immunoglobulin regions through a similar beta zipper interaction.

Single step generation of protein arrays from DNA by cell-free expression and in situ immobilisation (PISA method).

We describe a format for production of protein arrays termed 'protein in situ array' (PISA). A PISA is rapidly generated in one step directly from PCR-generated DNA fragments by cell-free protein expression and in situ immobilisation at a surface. The template for expression is DNA encoding individual proteins or domains, which is produced by PCR using primers designed from information in DNA databases. Coupled transcription and translation is carried out on a surface to which the tagged protein adheres as soon as it is synthesised. Because proteins generated by cell-free synthesis are usually soluble and functional, this method can overcome problems of insolubility or degradation associated with bacterial expression of recombinant proteins. Moreover, the use of PCR-generated DNA enables rapid production of proteins or domains based on genome information alone and will be particularly useful where cloned material is not available. Here we show that human single-chain antibody fragments (three domain, V(H)/K form) and an enzyme (luciferase) can be functionally arrayed by the PISA method.

Crystal structure of alkaline phosphatase from human placenta at 1.8 A resolution. Implication for a substrate specificity.

Human placental alkaline phosphatase (PLAP) is one of three tissue-specific human APs extensively studied because of its ectopic expression in tumors. The crystal structure, determined at 1.8-A resolution, reveals that during evolution, only the overall features of the enzyme have been conserved with respect to Escherichia coli. The surface is deeply mutated with 8% residues in common, and in the active site, only residues strictly necessary to perform the catalysis have been preserved. Additional structural elements aid an understanding of the allosteric property that is specific for the mammalian enzyme (Hoylaerts, M. F., Manes, T., and Millán, J. L. (1997) J. Biol. Chem. 272, 22781-22787). Allostery is probably favored by the quality of the dimer interface, by a long N-terminal alpha-helix from one monomer that embraces the other one, and similarly by the exchange of a residue from one monomer in the active site of the other. In the neighborhood of the catalytic serine, the orientation of Glu-429, a residue unique to PLAP, and the presence of a hydrophobic pocket close to the phosphate product, account for the specific uncompetitive inhibition of PLAP by l-amino acids, consistent with the acquisition of substrate specificity. The location of the active site at the bottom of a large valley flanked by an interfacial crown-shaped domain and a domain containing an extra metal ion on the other side suggest that the substrate of PLAP could be a specific phosphorylated protein.

Protein arrays: issues to be addressed.

Protein arrays are fast becoming established as a means to monitor protein expression levels and investigate protein interactions and function. They present particular technical demands that will need to be solved in order to achieve the maximum capability of efficient and sensitive protein analysis in the high throughput setting of functional genomics. The following resumé of some major issues around this new technology was made as the chairperson's introduction to the workshop session on peptide and protein chips.

Anti-anti-idiotypic (Ab3) antibodies that bind progesterone-11alpha-bovine serum albumin differ in their combining sites from antibodies raised directly against the antigen.

Polyclonal rabbit anti-idiotypic (Ab2) antibodies raised against the antiprogesterone mAb DB3 (Ab1) were used to induce an Ab3 antiprogesterone response in BALB/c mice. While the affinity of Ab3 sera for progesterone was 10-50-times lower than that of DB3, their steroid-binding specificity showed considerable similarity to DB3. Two immunoglobulin M (IgM) Ab3 monoclonal antibodies (mAbs), 1A4 and 3B11, were obtained, both of which bound progesterone conjugated to bovine serum albumin (progesterone-BSA). 1A4 also bound free progesterone, although with low affinity and very broad cross-reactivity. Like DB3, 1A4 is encoded by a heavy-chain variable region (VH) gene segment from the small VGAM3.8 family, a restriction that is characteristic of antibodies raised against progesterone-11alpha-BSA. In contrast, 3B11 binds progesterone-11alpha-BSA but not free progesterone and is encoded by an unrelated VH gene from the J558 family. The light chain variable region (VL) of 1A4 lacks the intradomain disulphide bridge owing to replacement of CysL23 by Tyr. Both the 1A4 and 3B11 heavy chains have extremely short complementarity determining region (CDR) H3 loops, comprising three and four amino acids, respectively. Modelling of the combining site of 1A4 from the X-ray crystallographic structure of DB3 indicates that the short H3 loop is a major factor in the loss of affinity and specificity for steroid.

Crystal structure of a Staphylococcus aureus protein A domain complexed with the Fab fragment of a human IgM antibody: structural basis for recognition of B-cell receptors and superantigen activity.

Staphylococcus aureus produces a virulence factor, protein A (SpA), that contains five homologous Ig-binding domains. The interactions of SpA with the Fab region of membrane-anchored Igs can stimulate a large fraction of B cells, contributing to lymphocyte clonal selection. To understand the molecular basis for this activity, we have solved the crystal structure of the complex between domain D of SpA and the Fab fragment of a human IgM antibody to 2.7-A resolution. In the complex, helices II and III of domain D interact with the variable region of the Fab heavy chain (V(H)) through framework residues, without the involvement of the hypervariable regions implicated in antigen recognition. The contact residues are highly conserved in human V(H)3 antibodies but not in other families. The contact residues from domain D also are conserved among all SpA Ig-binding domains, suggesting that each could bind in a similar manner. Features of this interaction parallel those reported for staphylococcal enterotoxins that are superantigens for many T cells. The structural homology between Ig V(H) regions and the T-cell receptor V(beta) regions facilitates their comparison, and both types of interactions involve lymphocyte receptor surface remote from the antigen binding site. However, T-cell superantigens reportedly interact through hydrogen bonds with T-cell receptor V(beta) backbone atoms in a primary sequence-independent manner, whereas SpA relies on a sequence-restricted conformational binding with residue side chains, suggesting that this common bacterial pathogen has adopted distinct molecular recognition strategies for affecting large sets of B and T lymphocytes.

Antibody repertoires of four- and five-feature translocus mice carrying human immunoglobulin heavy chain and kappa and lambda light chain yeast artificial chromosomes.

We have produced mice that carry the human Ig heavy (IgH) and both kappa and lambda light chain transloci in a background in which the endogenous IgH and kappa loci have been inactivated. The B lymphocyte population in these translocus mice is restored to about one-third of normal levels, with preferential (3:1) expression of human lambda over human kappa. Human IgM is found in the serum at levels between 50 and 400 microg/ml and is elevated following immunization. This primary human Ab repertoire is sufficient to yield diverse Ag-specific responses as judged by analysis of mAbs. The use of DH and J segments is similar to that seen in human B cells, with an analogous pattern of N nucleotide insertion. Maturation of the response is accompanied by somatic hypermutation, which is particularly effective in the light chain transloci. These mice therefore allow the production of Ag-specific repertoires of both IgM,kappa and IgM,lambda Abs and should prove useful for the production of human mAbs for clinical use.

Rapid appearance of M cells after microbial challenge is restricted at the periphery of the follicle-associated epithelium of Peyer's patch.

M cells within the follicle-associated epithelium (FAE) of the gut play a central role in the initiation of mucosal immune responses by transporting antigens to the intestinal lymphoid tissue. We have previously demonstrated that the instillation into the gut of a nonenteric microorganism, Streptococcus pneumoniae R36a, is an excellent experimental model to investigate the highly dynamic nature of the FAE in response to microbial challenge. In the present study, S. pneumoniae was introduced into rabbit ileal loops, each one containing a Peyer's patch (PP), and the number of M cells was assessed by morphological and functional characteristics in different areas of the FAE after a short time (1-3 hours). We report that a marked increase in the number of M cells was detected in the periphery, but not in the apical area, of the FAE as early as 1 hour after exposure to S. pneumoniae. Furthermore, a variant of this experiment enabled us to establish that the increased numbers of M cells led to an improved capability of the FAE to transport latex fluorescent microspheres (0.5 microm), highly specific to rabbit M cells, from the gut lumen to the intestinal lymphatic system. In these animals the cisterna chyli was cannulated, and the microparticles were introduced into the intestinal loops after stimulation with pneumococci. The microparticles reaching the lymph were then counted by flow cytometer. We interpreted these results as showing that only enterocytes located within the periphery of the FAE are converted to fully operational M cells by certain microbial interaction and the ability of enterocytes to undergo this conversion may depend on their stage of differentiation.

Up-regulation of microsphere transport across the follicle-associated epithelium of Peyer's patch by exposure to Streptococcus pneumoniae R36a.

Transport of antigens through the follicle-associated epithelium (FAE) of Peyer's patch (PP) is the critical first step in the induction of mucosal immune responses. We have previously described that short-term exposure to Streptococcus pneumoniae R36a induced dramatic morphological alterations of the FAE in rabbit PP. These results prompted us to investigate whether the pneumococci-induced modifications were accompanied by enhanced ability of the FAE to transport antigens. We addressed this problem by evaluating the ability of the FAE to bind, internalize, and transport fluorescent polystyrene microparticles, highly specific to rabbit M cells, after exposure to S. pneumoniae. Quantitative study revealed a marked increase in the number of microspheres in PP tissues exposed to S. pneumoniae compared to tissues exposed to either phosphate-buffered saline or Escherichia coli DH5alpha as controls. No sign of bacterially induced damage to the epithelial barrier was observed. Further confocal microscopy analysis of the FAE surface showed that a significant increase in the number of cells that showed both morphological and functional features of M cells took place within pneumococci-treated PP tissues. These data provide the first direct evidence that the FAE-specific antigen sampling function may be manipulated to improve antigen and drug delivery to the intestinal immune system.

Selection of a human anti-progesterone antibody fragment from a transgenic mouse library by ARM ribosome display.

In antibody-ribosome-mRNA complex (ARM) ribosome display, stable complexes of nascent protein, mRNA and ribosomes are produced in a eukaryotic in vitro expression system, through coupled transcription and translation of DNA lacking a 3' stop codon. Selection of the protein simultaneously captures the relevant mRNA, which is recovered as DNA by coupled reverse transcription-polymerase chain reaction (RT-PCR) performed on the intact complexes. Here, we describe the use of ARM display to select a specific human antibody fragment from a transgenic mouse library. The mice carry unrearranged gene segments of the human heavy (H) and kappa light (L) chain loci, while the endogenous murine H and kappa loci are functionally silenced; they respond to immunisation by production of fully human IgM antibodies. A library encoding human single-chain (sc) antibody (V(H)/K) fragments, in which V(H) domains and kappa light chains were combined at random by PCR, was prepared from spleen cells of transgenic mice immunised with progesterone-bovine serum albumin (BSA). Library diversity was demonstrated by sequencing. Progesterone-binding fragments were selected over five cycles of ARM display and the selected DNA cloned and expressed in Escherichia coli. Soluble V(H)/K fragments obtained in periplasmic extracts had the same specificity as ribosome-bound V(H)/K, supporting the view that folding and specificity of the displayed and soluble proteins are equivalent. The affinity of the expressed V(H)/K was approximately 10(-8) M. Sequencing showed that ARM display selected a single V(H)/V(L) combination (V(H)1-2, Vkappa4-1) and rearrangement, with a few mutational differences between clones. Monoclonal antibodies against progesterone-BSA obtained from hybridomas were encoded by the same V(H) and V(L) segments and had similar properties to the fragments obtained in vitro. The combination of ribosome display and transgenic mouse technologies is a rapid means of generating fully human antibody fragments in vitro for expression and further manipulation.

The structure of a human rheumatoid factor bound to IgG Fc.

This is the first crystal structure analysis of a complex between an autoantibody and its autoantigen, and it reveals a mode of interaction never before seen in an antibody-antigen complex. Not only are there relatively few antibody contact residues, contributing perhaps to its very low affinity, but these residues are to be found on only one side of the potential combining site surface. Indeed, so many CDR residues are not involved in Fc binding, including those in the central region of the combining site, that it is easy to envisage that this RF may have another, entirely different, specificity. The antibody may therefore have originated in response to another, as yet unidentified, antigen, and the reactivity with IgG Fc may be an unfortunate cross-reactivity. Certainly some of the CDR residues which do interact with IgG Fc are germline encoded, but significantly one of only two residues in the light chain, Pro56, which makes many contacts with Fc, is a somatic mutation. Since this mutation would appear to make a significant contribution to the binding affinity, it is therefore evidence for an antigen driven response to the IgG Fc in the generation of this autoantibody. The Fc epitope recognised by RF-AN is strikingly similar to the binding sites for the bacterial binding proteins A and G, but the significance of this is not clear. What is clear however is that the epitope does not include any part of the Fc carbohydrate residues, although the structure of the complex does reveal that there is an alteration in the carbohydrate conformation when the galactose residues are absent. Loss of the interaction between the terminal galactose residue on the alpha (1-6) linked branch and the C gamma 2 domain appears to allow the carbohydrate chains to become mobile, at the same time exposing a predominantly hydrophobic patch on the C gamma 2 surface. Accessibility to either the agalactosyl carbohydrate chains or the newly exposed residues may account for the enhanced reactivity for G0-IgG that has been reported for certain RFs, and such an epitope need not be very different to that recognised by RF-AN. In order to understand more completely the effect of the presence or absence of the terminal galactose residue, the fully galactosylated glycoform of Fc must be studied for comparison; this work is underway. It is also important now to study a RF which is known to sense this difference in oligosaccharide composition, and also to study RFs of higher affinity, of the IgG class, and from the synovium. RF-AN was the first RF to be immortalised as a cell line, and in many ways it is a typical RF (in terms of specificity, relationship to germline sequence and affinity), but we must now establish whether the novel structural features revealed in this analysis are indeed typical of other RFs. Only when comparisons can be made between RFs of different origin and with contrasting functional properties will we begin to understand what constitutes a pathogenic RF, and the mechanism by which such auto-reactive antibodies are generated.

Production of human antibody repertoires in transgenic mice.

Transgenic mice have been created that carry human immunoglobulin heavy and light chain genes in germline configuration and that have the corresponding endogenous genes silenced. The transgenes are either minigene constructs or large, almost authentic, transloci on yeast artificial chromosomes and undergo B-cell-specific DNA rearrangement and hypermutation in the mouse lymphoid tissue. Monoclonal antibodies with good affinities for human antigens have been obtained after immunisation. These mice may be a future source of human antibodies for therapy.

Recognition of porcine growth hormone by a panel of monoclonal antibodies.

A panel of murine monoclonal antibodies (MAbs) against porcine growth hormone (pGH) has been raised from BALB/c mice. MAbs were characterized for binding to growth hormones (GH), prolactins (PRL), and placental lactogen (PL) from different species and to the N-terminal peptides of GH. From their patterns of cross-reactivity MAbs were assigned into nine specificity groups. The sharing of pGH epitopes among hormones of different species was related to the sequence similarity to pGH, i.e., overlap was greatest for equine, ruminant, and rodent GHs and least for human GH, ovine, and porcine PRLs, and human PL. Partial epitope mapping was carried out by relating hormone cross-reactivity patterns with amino acid sequences. Two epitopes were localized to interhelical loops, around valine-73 and glycine-130, respectively. Direct mapping with synthetic peptides localized other epitopes (Groups 7, 8, and 9) to the N-terminal region of the GH molecule. Selected MAbs were studied for the enhancement of the somatogenic activity of pGH in the dwarf mouse bioassay, measuring weight gain and sulphate incorporation into costal cartilage. Only those antibodies with specificities for GHs and not PRL or PL showed significant enhancement in this assay.

Structure of human IgM rheumatoid factor Fab bound to its autoantigen IgG Fc reveals a novel topology of antibody-antigen interaction.

Rheumatoid factors are the characteristic autoantibodies of rheumatoid arthritis, which bind to the Fc regions of IgG molecules. Here we report the crystal structure of the Fab fragment of a patient-derived IgM rheumatoid factor (RF-AN) complexed with human IgG4 Fc, at 3.2 A resolution. This is the first structure of an autoantibody-autoantigen complex. The epitope recognised in IgG Fc includes the C gamma 2/C gamma 3 cleft region, and overlaps the binding sites of bacterial Fc-binding proteins. The antibody residues involved in autorecognition are all located at the edge of the conventional combining site surface, leaving much of the latter available, potentially, for recognition of a different antigen. Since an important contact residue is somatic mutation, the structure implicates antigen-driven selection, following somatic mutation of germline genes, in the production of pathogenic rheumatoid factors.

Sequence, specificity and crystallization of an oestrone-3-glucuronide antibody (3910).

We describe the specificity profile and V region sequences of a high-affinity monoclonal antibody (mAb), 3910, directed against oestrone-3-glucuronide (E3G). Inhibition studies show that the D-ring is critical for steroid specificity, while the glucuronic acid attached to the A ring is required for high binding affinity, suggesting that both 'ends' of the E3G ligand are recognized. The VH domain is encoded by a gene from the VH7183 family, while VL appears to be encoded by the Vk5.1 gene (kappa II subgroup) with a deletion of six residues from complementarity-determining region-1 (CDR1). The VH CDR3 is 10 amino acid residues in length, of which D/N contributes five residues. Comparison of VH CDR of 3910 with those of mAb against progesterone (DB3) and digoxin (26-10, 40-50), for which crystal structures have been determined, suggests that aromatic side chains are important for E3G binding and that tyrosine residues H50, H97 and H100 may interact with the ligand. The Fab fragment of 3910 has been crystallized in its native and steroid (E3G and oestriol-3-glucuronide) complexed forms. An X-ray diffraction data set to 3 A resolution has been collected for the native Fab.

Antibody-ribosome-mRNA (ARM) complexes as efficient selection particles for in vitro display and evolution of antibody combining sites.

We describe a rapid, eukaryotic, in vitro method for selection and evolution of antibody combining sites using antibody-ribosome-mRNA (ARM) complexes as selection particles. ARMs carrying single-chain (VH/K) binding fragments specific for progesterone were selected using antigen-coupled magnetic beads; selection simultaneously captured the genetic information as mRNA, making it possible to generate and amplify cDNA by single-step RT-PCR on the ribosome-bound mRNA for further manipulation. Using mutant libraries, antigen-binding ARMs were enriched by a factor of 10(4)-10(5)-fold in a single cycle, with further enrichment in repeated cycles. While demonstrated here for antibodies, the method has the potential to be applied equally for selection of receptors or peptides from libraries.

Crystallization of a complex between the Fab fragment of a human immunoglobulin M (IgM) rheumatoid factor (RF-AN) and the Fc fragment of human IgG4.

Rheumatoid factors (RF) are the characteristic autoantibodies found in patients with rheumatoid arthritis. They recognize epitopes in the Fc region of immunoglobulin G (IgG) and are often of the IgM isotype. In order to analyse the nature of RF-Fc interactions, we have crystallized a complex between the Fab fragment of a human monoclonal IgM rheumatoid factor (RF-AN) and the Fc fragment of human IgG4. The stoichiometry of the complex within the crystals was found to be 2:1 Fab:Fc. The crystals diffracted X-rays to 0.3 nm resolution, and the space group was C2, with cell dimensions a = 16.03 nm, b = 8.19 nm, c = 6.42 nm, beta = 98.3 degrees. We have also determined the sequence of the variable region of the RF-AN light chain, not hitherto reported. This belongs to the V lambda III-a subgroup and is closely related to the germline gene Humlv318, from which it differs in three amino acid residues. This is the first reported crystallized complex between a human autoantibody and its autoantigen.