Antibody recognition imaging by force microscopy.

We have developed a method that combines dynamic force microscopy with the simultaneous molecular recognition of an antigen by an antibody, during imaging. A magnetically oscillated atomic force microscopy tip carrying a tethered antibody was scanned over a surface to which lysozyme was bound. By oscillating the probe at an amplitude of only a few nanometers, the antibody was kept in close proximity to the surface, allowing fast and efficient antigen recognition and gentle interaction between tip and sample. Antigenic sites were evident from reduction of the oscillation amplitude, as a result of antibody-antigen recognition during the lateral scan. Lysozyme molecules bound to the surface were recognized by the antibody on the scanning tip with a few nanometers lateral resolution. In principle, any ligand can be tethered to the tip; thus, this technique could potentially be used for nanometer-scale epitope mapping of biomolecules and localizing receptor sites during biological processes.

Association of the anti-hen egg lysozyme antibody HyHEL-5 with avian species variant and mutant lysozymes.

The energetics of association of the murine anti-hen egg lysozyme antibody HyHEL-5 with bobwhite quail lysozyme, California quail lysozyme, and the Arg45-->Lys mutant of hen egg lysozyme was characterized by isothermal titration calorimetry. The association of each lysozyme with HyHEL-5 is enthalpically driven in the temperature range 10 degrees C to 37 degrees C. The calorimetric results indicate that the salt-links between Arg45 and Arg68 of hen egg lysozyme and GluH50 on the HyHEL-5 paratope are energetically important in HyHEL-5/HEL association. In contrast to previous studies, the results suggest that the three characteristic 'quail' mutations affect the energetics of antibody/antigen association, even though they are buried and not in direct contact with the antibody.

Cloning, expression, and characterization of the Fab fragment of the anti-lysozyme antibody HyHEL-5.

Hybridoma cDNAs encoding the individual chains of the Fab fragment of the well characterized murine monoclonal antibody HyHEL-5 were cloned and sequenced. The recombinant Fab fragment was produced by expressing each chain in a separate Escherichia coli pET vector, denaturing inclusion bodies and co-refolding. Characterization of the purified Fab by MALDI-TOF mass spectrometry and N-terminal amino acid sequencing demonstrated proper processing of the individual chains. The association of the recombinant Fab fragment with hen egg lysozyme and the avian epitope variant bobwhite quail lysozyme was found by isothermal titration calorimetry to have energetics very similar to that of the HyHEL-5 IgG. Heterologous expression of the HyHEL-5 Fab fragment opens the way to structure/function studies in this well-known system.

A three-dimensional model of an anti-lysozyme antibody.

The primary amino acid structure of the lysozyme-binding antibody, HyHEL-10, as determined by amino acid and nucleotide sequencing was utilized to construct a scale model of the Fv (variable region domain of immunoglobulin) using energy-minimized torsional angles of the McPC603 Fv as a prototype template. This model was in turn used as a template for generating a computer-built set of co-ordinates, which were subjected to a total of 600 steps of Adopted Basis Newton-Raphson energy minimizations using the program CHARMM. Only minimal shifts of the backbone (root-mean-square 0.76 A) were required to give an energetically stable structure with a favorable van der Waals' energy. Several notable features were evident from both the scale model and the energy-minimized computer model: (1) the shape of the antibody combining region is that of a very shallow concavity approximately 20 A X 25 A; (2) the concavity is acidic and non-hydrophobic and is bordered by hydrophobic segments; (3) the lower portion of the combining site is dominated by a cluster of tyrosine residues over the L3 and H2 areas; (4) a somatic mutation encoded by the J region of the heavy chain (JH) may contribute significantly to the complementarity of heavy chain H3 to the epitope on hen egg white lysozyme. In addition, the space-filling energy-minimized model revealed that residue 49L, a framework residue, was prominently exposed and accessible in the center of the combining-site concavity. The model suggests that variation in length of complementarity-determining regions may function not only to change directly the shape of the antibody combining site, but may also influence indirectly the nature of the antibody surface by changing the accessibility of residues not usually involved in antigen binding.

Protein-protein interactions: structural motifs and molecular recognition.

The past year has brought new insights into common structural motifs used for protein-protein interactions by DNA-binding proteins. In addition, there have been significant advances in our understanding of antibody-protein complexes.

A substantial proportion of the adult BALB/c available B cell repertoire consists of multireactive B cells.

A variety of studies have documented multireactive antibodies in both the preimmune and naturally activated repertoire, but the relationship of these primarily IgM multireactive antibodies to antigen-specific primary and secondary response antibodies is currently not defined. In order to characterize the BALB/c preimmunization specificity repertoire and the baseline of naturally activated antibodies from which the immune response to a specific antigen (hen egg-white lysozyme, HEL) develops, panels of polyclonally activated blast-derived hybridomas (BlAbs) and natural antibody hybridomas (NAbs) from the spleens of unimmunized mice were screened for binding to a panel of nine complex antigens. Over half of the IgM-secreting BlAbs produced antibodies that were antigen-reactive; of these, over half were multireactive, i.e. capable of binding more than one complex antigen. There was no bias towards self vs foreign or thymus-dependent vs thymus-independent antigens. The frequency of antigen-reactive NAbs was about half the frequency of antigen-reactive antibodies found among the BlAbs. However, over half of the antigen-reactive NAbs were also multireactive, and the reactivity profile within the antigen-reactive subset of NAbs was similar to that within the antigen-reactive subset of BlAbs. These results suggest that the available repertoire of adult spleen cells contains a high proportion of multireactive antibodies, and that a subset of the available repertoire is randomly activated, yielding a small proportion of natural antibodies which closely reflect a random sampling of the available repertoire. Although monospecific precursor cells are rare, monospecific IgM BlAbs were found for all antigens in the panel except staphylococcal nuclease and mouse IgG. Monospecific as well as multireactive HEL-binding BlAbs were found at frequencies comparable to other protein antigens in the panel, and HEL-reactive NAbs were also present. On the other hand, it has previously been shown that HEL-reactive IgM antibodies (including multireactive antibodies whose specificities include HEL) are rare or absent in both the primary and secondary response to HEL. This cannot be attributed to an absence of available precursor B cells, and most likely reflects an early recruitment of HEL-reactive clones into the peripheral B cell pool. The possibility that polyreactive B cells may serve as precursors for some HEL-specific IgG antibodies is discussed.

Plasmacytoma-refractory BALB/cAnPt mice have naive T cell and highly specific B cell responses to antigen.

Recently, a reduction in the incidence of pristane-induced plasmacytomas in BALB/cAnPt (BALB/c) mice that were kept in viral specific pathogen-free (SPF) conditions has been reported. Environmentally, these SPF-BALB/c mice differed from conventionally-housed (CON) mice only in viral exposure and diet (i.e. sterilization of mouse chow), since microbial colonization of the intestinal tract was seen to be equivalent. This report assessed the ability of SPF- and CON-BALB/c mice to respond to immunologic challenge with soluble antigen, i.e. hen egg white lyzosyme (HEL), as a means of evaluating differences in T and B cell function and, indirectly, evaluating the possible effects these differences might have on plasmacytoma development. When cultured in vitro for 5 days with HEL, HEL-primed lymph node cells (LNC) from SPF-BALB/c mice proliferated to a significantly lesser extent than HEL-primed CON-BALB/c LNC. Moreover, HEL-induced production of IFN-gamma and IL-5 was significantly lower in SPF LNC. Serum IgG1 levels were 10-fold lower in SPF-BALB/c mice with, or without prior immunization with HEL and were not reconstituted by repeated injections of HEL in adjuvant. Serum IgM levels of SPF- and CON-BALB/c mice were equivalent. This reduction in immune responses could not be attributed to a lack of colonization of secondary lymphoid organs, since flow cytometric analysis of LNC revealed no difference in the number of recoverable cells and the proportion of lymphocyte subsets (CD4+, CD8+ and CD45+ cells) obtained from SPF- and CON-BALB/c mice. However, only CON LNC were induced to increase surface expression of CD44 after antigenic or mitogenic stimulation in vitro. Antibody responsiveness to HEL, as evidenced by serum anti-HEL binding or splenic hybridoma studies, demonstrated higher levels of IgG1 antibodies in CON BALB/c mice than in SPF mice. However, a greater proportion of the SPF IgG1 antibodies present were specifically directed against HEL, so that specific activity was greater in SPF-BALB/c mice. Therefore, while SPF BALB/c mice have a more restricted response to HEL than CON-BALB/c mice, those antibodies that are produced are more specifically directed against HEL with very little apparent bystander/polyclonal activation of multireactive cells. Resistance to plasmacytomas in SPF-BALB/c mice, therefore, may stem from a reduced number of circulating memory T and B cells, which are capable of reacting and/or crossreacting with a chronic inflammatory stimulus.

Contribution of the VK4 light chain to antibody specificity for lysozyme and beta (1,6)D-galactan.

The VL amino acid sequence of an anti-lysozyme hybridoma protein, HyHEL-5, was determined. HyHEL-5 expresses a V region of the VK4 family and JK1. The VK4 family also includes light chains from galactan binding antibodies, although sequence comparisons suggest that a different member of this family is used to encode HyHEL-5. The HyHEL-5 light chain has a deletion of residue 96, such that L3 is one residue shorter than the majority of murine L3. Chain recombination experiments, employing H and L chains from different anti-galactan and anti-lysozyme binding antibodies, were performed to examine the contribution of the H and L chain in dictating specificity for either galactan or the lysozyme epitope recognized by HyHEL-5. The results indicate that, although the ability to bind galactan vs lysozyme is absolutely heavy-chain dependent, having the appropriate heavy chain is not sufficient for specific high affinity binding. Both the L chains from HyHEL-5 and J539 (a galactan-binding myeloma protein) were capable of supporting binding to galactan in combination with the J539 H chain, but affinity for galactan is less with the HyHEL-5 L chain. Only VK4 L chains supported binding of the HyHEL-5 heavy chain to the HyHEL-5 epitope, although binding with the J539 L chain was low affinity and relatively nonspecific.

Organ reactive autoantibodies from non-immunized adult BALB/c mice are polyreactive and express non-biased VH gene usage.

To examine the naturally activated autoreactive B cell repertoire, we analyzed a panel of hybridomas from unmanipulated adult BALB/c spleen cells for reactivity patterns and VH gene usage. We found a pattern of VH usage that was diverse and appeared to reflect the germline repertoire. Although all but one natural antibody hybridoma (NAb) were initially selected for organ rather than antigen binding, the majority of organ reactive IgM NAbs were polyreactive, expressing a broad range of reactivity patterns for both self and foreign antigens, that were unique for each NAb and were not indiscriminate. Our results are consistent with the hypothesis that many naturally activated adult B cells are highly polyreactive and that autoreactivity is a consequence of polyreactivity. We suggest that the population of NAbs exhibiting organ reactivity overlaps the populations of other IgM autoantibodies that have been described previously, and that these all derive from a pool of polyreactive IgM antibodies which are polyclonally activated in the early immune response. These polyreactive natural antibodies may represent a first line of defense and offer protection for the host against a variety of foreign agents.

Structural differences among monoclonal antibodies with distinct fine specificities and kinetic properties.

The mAbs HyHEL-8, HyHEL-26 (HH8, and HH26, respectively) recognize epitopes on hen egg-white lysozyme (HEL) highly overlapping with the structurally defined HH10 epitope, while the structurally related XRPC-25 is specific for DNP and does not bind HEL. All four Abs appear to use the same Vk23 germ line gene, and all but HH8 use the same VH36-60 germ line gene. Of the three anti-HEL Abs, the sequences of HH26 variable regions are closest to those encoded by the respective germ line sequences. HH8 utilizes a different member of the VH36-60 gene family. Thus, the same Vk and VH genes, combined with somatically derived sequence differences, are used to recognize the unrelated Ags HEL and DNP. In contrast, different VH36-60 germ line genes are used to bind the same antigen (e.g. HH8 vs HH10 and HH26). While the affinities of HH10, HH8, and HH26 for HEL vary by less than 10-fold, their affinities for mutated Ag vary over several orders of magnitude. Analyses of Fab binding kinetics with natural species variants and site-directed mutants of lysozyme indicate that these cross-reactivity differences reflect the relative sensitivities of both the association and dissociation rates to antigenic mutation: HH8 has relatively mutation-insensitive association and dissociation rates, HH10 has a relatively mutation-sensitive association rate but more variable dissociation rates, and HH26 has variable association and dissociation rates. Only a few amino acid differences among the antibodies produce the observed differences in the robustness of the association and dissociation rates. Our results suggest that association and dissociation rates and mutation sensitivity of these rates may be independently modulated during antibody repertoire development.

Immunotolerance against a foreign antigen transgenically expressed in the lens.

PURPOSE: To extend our knowledge concerning immunotolerance against autologous lens crystallins, transgenic (Tg) mice that express a foreign antigen in their lens were generated, and the immune response against the antigen in these mice was analyzed. METHODS: Conventional techniques were used to generate lines of Tg mice that express soluble (S-) or membrane-bound (M-) hen egg lysozyme (HEL) under the control of the alphaA-crystallin promoter. The presence of HEL in various organs was determined by the particle concentration fluorescence immunoassay (PCFIA), and reverse transcription-polymerase chain reaction technique was used to detect mRNA transcripts of the molecule. To examine the development of immunity (or tolerance), Tg mice and their wild-type controls were immunized with HEL (25 microg) in Freund's complete adjuvant and 14 days later were tested for immune response against the antigen. Cellular immunity was measured by the lymphocyte proliferation assay and cytokine production, and humoral immunity was determined by enzyme-linked immunosorbent assay. RESULTS: Eyes of the high copy number M-HEL Tg mice were dystrophic, with disrupted lens, whereas no morphologic changes were detected in the eyes of the other Tg mouse lines. All Tg mice exhibited tolerance to HEL by their cellular and humoral immune compartments. The state of immunotolerance to HEL was retained in the Tg mice for as long as 10 months after removal of the main depot of this protein, by enucleation. Measurable amounts of HEL were found in the eyes of all Tg mice, but the protein could not be detected in the serum or in other organs by the sensitive PCFIA (with a threshold of 1 ng/ml). Yet, HEL mRNA was found in the thymus of the Tg mice, suggesting that minute amounts of the protein are expressed in this organ. CONCLUSIONS: The unresponsiveness to HEL in the Tg mice seems to be due to a "central" mechanism of tolerance, mediated by a minuscule amount of HEL in the thymus. Conversely, the much larger amounts of HEL in the peripheral depot, the eyes, play a minor role if any in the tolerogenic process. It is further proposed that a similar mechanism of central tolerance is responsible for the immunotolerance against autologous lens crystallins.

Molecular recognition of lysozyme by monoclonal antibodies.

HEL was one of the first proteins to be mapped antigenically using mAb, and panels of mAb have been used as a measure of antigenicity in order to study regulation of the immune response and the apparent 'antigenic structure' of HEL. These studies have confirmed the multideterminant hypothesis derived from pAb. However, although the entire surface of HEL is potentially antigenic, the mature immune response appears to be dominated by three functionally nonoverlapping antigenic regions, defined operationally by antibody complementation assays. Recent structural studies have confirmed the existence of three distinct epitope clusters. Functional epitopes, defined by immunoassays, are generally only a subset of the structural epitope, the 14-17 amino acid residues which contact antibody in the X-ray structure of the complex. An even smaller subset of 'critical residues', the 'energetic' epitope, may predominate the interaction energetically. Antibody complex formation with HEL is enthalpically driven, and is accompanied by an unfavorable entropy change. Mutations of either antibody or antigen which lower affinity appear to do so primarily by increasing dissociation rates, and also appear to be accompanied by entropy/enthalpy compensation. The current availability of six structurally defined antibody-lysozyme complexes presents excellent opportunities for comparative studies in order to understand the structural bases of affinity, specificity, and thermodynamic properties, as well as the interrelationships of functional, structural, and energetic epitopes.

Cytophysiological basis of disruptive pigmentary patterns in the leopard frog Rana pipiens. II. Wild type and mutant cell-specific patterns.

A new pattern index, Ip, is introduced and used to compare patterns of wild type, burnsi, and kandiyohi chromatophores in the leopard frog, Rana pipiens. Wild type chromatophores are hyperdispersed over distances within cellular contact, and it is concluded that this hyperdispersion results from contact-mediated negative interactions. The hyperdispersion is less strong in spot cells than interspot, and extends over larger areas in burnsi than in wild type epidermis. Over areas greater than chromatophore size, patterns are either random or clumped. Patterning of kandiyohi melanophores is clumped into aggregates small enough to be within the range of cellular contact, suggesting a lack of contact inhibition among these cells. The possible roles of cellular properties and the extracellular environment in pattern determination are discussed.

Production and characterization of a monoclonal antibody to bovine beta-crystallin.

A monoclonal antibody to a bovine lens 27K beta-crystallin polypeptide has been produced from a rat x mouse hybridoma. The antibody reacts with the 27K polypeptide of bovine, mouse, rat, and human lenses, but does not react to the 27K polypeptide of monkey lens nor does it react with any component in the Philly mouse lens which is missing the 27K polypeptide. The antibody does not recognize any of the other major bovine beta-crystallin polypeptides but does recognize a large number of native beta-crystallin proteins. The antigenic specificity and species cross-reactivity of this antibody provides an excellent opportunity to study many aspects of lens development and cataractogenesis.

Immunocytochemical localization of the 27K beta-crystallin polypeptide in the mouse lens during development using a specific monoclonal antibody: implications for cataract formation in the Philly mouse.

The Philly mouse develops a hereditary cataract about 5 weeks after birth. Although the causative agent is not known, data suggest that there is a correlation between cataract formation and the selective absence of a 27 kilodalton (27K) beta-crystallin lens polypeptide. The ontogeny of the 27K beta-crystallin polypeptide was examined in normal mice in order to evaluate its role in normal development and determine what impact its absence may have on the Philly mouse lens. A monoclonal antibody was used with the PAP method to immunocytochemically localize the 27K polypeptide in lenses of normal mice during development. beta-Crystallins detected with polyclonal antisera were found in differentiated fiber cells throughout the lens. In contrast, the 27K beta-crystallin polypeptide detected with a specific monoclonal antibody was not found in the fiber cells of the inner part of the lens (nucleus), but was specifically localized in the fiber cells of the outer part of the lens called the cortex. The polypeptide was found only in elongating and differentiated fiber cells and not in mitotically active epithelial cells. Although a minor component of the 2-day-old lens, the 27K polypeptide comprised a large portion of the 16-day-old lens including the anterior and posterior poles. These data show that the 27K polypeptide is a minor component of the embryonic lens, but becomes a major contributor to the postnatal lens. The 27K beta-crystallin lens polypeptide is abundant in the fiber cells of the normal postnatal mouse lens. The absence of the 27K polypeptide in the Philly mouse may contribute to the observed failure of fiber cells to differentiate in the Philly mouse after birth or may be deleterious in some other manner to normal lens development. The selective absence of the 27K beta-crystallin polypeptide, a defect which precedes cataract formation in the Philly mouse, is intriguing since it suggests a relationship between this major lens polypeptide and lens clarity.