Stabilization of antibody structure upon association to a human carbonic anhydrase IX epitope studied by X-ray crystallography, microcalorimetry, and molecular dynamics simulations

Specific antibodies interfere with the function of human tumor-associated carbonic anhydrase IX (CA IX), and show potential as tools for anticancer interventions. In this work, a correlation between structural elements and thermodynamic parameters of the association of antibody fragment Fab M75 to a peptide corresponding to its epitope in the proteoglycan-like domain of CA IX, is presented. Comparisons of the crystal structures of free Fab M75 and its complex with the epitope peptide reveal major readjustments of CDR-H1 and CDR-H3. In contrast, the overall conformations and positions of CDR-H2 and CDR-L2 remain unaltered, and their positively charged residues may thus present a fixed frame for epitope recognition. Adoption of the altered CDR-H3 conformation in the structure of the complex is accompanied by an apparent local stabilization. Analysis of domain mobility with translation-libration-screw (TLS) method shows that librations of the entire heavy chain variable domain (V(H)) decrease and reorient in the complex, which correlates well with participation of the heavy chain in ligand binding. Isothermal titration microcalorimetry (ITC) experiments revealed a highly unfavorable entropy term, which can be attributed mainly to the decrease in the degrees of freedom of the system, the loss of conformational freedom of peptide and partially to a local stabilization of CDR-H3. Moreover, it was observed that one proton is transferred from the environment to the protein-ligand complex upon binding. Molecular dynamics simulations followed by molecular mechanics/generalized Born surface area (MM-GBSA) calculations of the ligand (epitope peptide) binding energy yielded energy values that were in agreement with the ITC measurements and indicated that the charged residues play crucial role in the epitope binding. Theoretical arguments presented in this work indicate that two adjacent arginine residues (ArgH50 and ArgH52) are responsible for the observed proton transfer.     

Biophysical properties of human antibody variable domains

There are great demands on the stability, expression yield and resistance to aggregation of antibody fragments. To untangle intrinsic domain effects from domain interactions, we present first a systematic evaluation of the isolated human immunoglobulin variable heavy (V(H)) and light (V(L)) germline family consensus domains and then a systematic series of V(H)-V(L) combinations in the scFv format. The constructs were evaluated in terms of their expression behavior, oligomeric state in solution and denaturant-induced unfolding equilibria under non-reducing conditions. The seven V(H) and seven V(L) domains represent the consensus sequences of the major human germline subclasses, derived from the Human Combinatorial Antibody Library (HuCAL). The isolated V(H) and V(L) domains with the highest thermodynamic stability and yield of soluble protein were V(H)3 and V(kappa)3, respectively. Similar measurements on all domain combinations in scFv fragments allowed the scFv fragments to be classified according to thermodynamic stability and in vivo folding yield. The scFv fragments containing the variable domain combinations H3kappa3, H1bkappa3, H5kappa3 and H3kappa1 show superior properties concerning yield and stability. Domain interactions diminish the intrinsic differences of the domains. ScFv fragments containing V(lambda) domains show high levels of stability, even though V(lambda) domains are surprisingly unstable by themselves. This is due to a strong interaction with the V(H) domain and depends on the amino acid sequence of the CDR-L3. On the basis of these analyses and model structures, we suggest possibilities for further improvement of the biophysical properties of individual frameworks and give recommendations for library design.     

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.     

Antigen binding and stability properties of non-covalently linked anti-CD22 single-chain Fv dimers

By varying linker length and domain orientation three multivalent derivatives of a monovalent anti-CD22 single-chain fragment variable (scFv) antibody were generated. Shortening the linker of the V(H)-V(L) oriented scFv to 5 or 0 residues resulted in the formation of diabodies or a mixture of tetramers and trimers, respectively. Unexpectedly, a V(L)-0-V(H) scFv assembled to homogenous dimers, remained substantially more stable than the V(H)-5-V(L) diabody when incubated in human serum at 37 degrees C, and retained its dimeric state when concentrated up to 4 mg/ml. These properties suggest the V(L)-0-V(H) scFv could become an attractive vehicle for the selective delivery of multiple effector molecules to CD22(+) tumor cells.     

Systematic classification of CDR-L3 in antibodies: implications of the light chain subtypes and the VL-VH interface

Antibody modeling is widely used for the analysis of antibody-antigen interactions and for the design of potent antibody drugs. The antibody combining site is composed of six complementarity determining regions (CDRs). The CDRs, except for CDR-H3, which is the most diverse CDR, form limited numbers of canonical structures, which can be identified from the amino acid sequences. A method to classify the CDR-H3 structure from its amino acid sequence was previously proposed. However, since those CDR structures were classified, many more antibody crystal structures have been determined. We performed systematic analyses of the CDR-L3 structures and found novel canonical structures, and we also classified a previously identified canonical structure into two subtypes. In addition, two differently defined canonical structures in the kappa and lambda subtypes were classified into the same canonical structure. We also identified a key residue in CDR-L3, which determines the conformation of CDR-H3. Several analyses of CDR-L3 loops longer than nine residues were performed. These new findings should be useful for structural modeling and are eventually expected to accelerate the design of antibody drugs.     

Single-chain Fv fragments derived from an anti-11-deoxycortisol antibody. Affinity,specificity, and idiotype analysis

Single-chain Fv fragments (scFvs) against a corticosteroid, 11-deoxycortisol (11-DC), have been generated as a template antibody fragment from which a comprehensive mutated antibody library containing various anti-steroid antibodies could be constructed. The cDNAs encoding variable heavy (V(H)) and light (V(L)) domains of a mouse anti-11-DC antibody (CET-M8), were amplified by RT-PCR, combined via a common linker to construct the sequence of 5'-V(H)-(Gly(4)Ser)(3)-V(L)-3', and cloned into a phagemid vector, pEXmide 5. The phage clones exhibiting binding activity to 11-DC were isolated after single panning against a hapten-immobilizing immunotube. The scFv gene in one of these clones was reamplified to introduce the ochre codons, and then expressed in the bacterial periplasm as the soluble antibody fragment. Two different scFvs (#6 and #12) were cloned, whose binding characteristics were examined by a radioimmunoassay using a tritium-labeled 11-DC. Both of them showed high affinity (K(a)=1.3x10(10)M(-1)) and practical specificity (cross-reactivity: cortisol, <0.2%; cortisone, <0.3%) to 11-DC, and furthermore, strong reactivity with an anti-idiotype antibody which recognizes the paratope of CET-M8. These results suggest that the present scFvs retain the three-dimensional structure of the paratope of the original monoclonal antibody.     

Using a heavy chain-loss hybridoma 26.4.1LL for studying the structural basis of immunoglobulin chain association

One of the mechanisms contributing to antibody diversity is created by the association of different heavy and light chains. The combinability of heavy and light chains has been studied previously in two systems: in vitro chain recombination and hybrid hybridoma. Here, a novel in vivo chain combination assay system involving a heavy chain-loss variant, 26.4.1LL, producing two kappa light chains (L(DEX) and L(MPC)) different in size is described. In conjunction with DNA transfection, immunoprecipitation and SDS-PAGE, the structural basis of noncovalent interaction between heavy and light chains can be elucidated systematically by examining the relative association tendency of a heavy chain with two light chains. To demonstrate the usefulness of this system, three stably transfected 26.4.1LL cell lines expressing gamma2b heavy chains, designated as H(DEX), H(CHI) and H(ARS), respectively, with structural interrelated variable regions were generated: H(DEX) differs from H(CHI) only in framework regions whereas H(CHI) differs from H(ARS) in complementarity-determining regions. The relative amounts (R values) of L(DEX) and L(MPC) associated with the heavy chains H(DEX), H(CHI) and H(ARS) in the assembled immunoglobulin molecules were found to be 1.02, 0.64 and 0.05, respectively, suggesting that the complementarity-determining regions and framework regions contribute equally to the V(L)-V(H) interaction. This conclusion is consistent with previous observations based on calculation of the buried area in the V(L)-V(H) interface, thus demonstrating the usefulness of this system.     

Structural basis of HIV-1 neutralization by affinity matured Fabs directed against the internal trimeric coiled-coil of gp41

The conserved internal trimeric coiled-coil of the N-heptad repeat (N-HR) of HIV-1 gp41 is transiently exposed during the fusion process by forming a pre-hairpin intermediate, thus representing an attractive target for the design of fusion inhibitors and neutralizing antibodies. In previous studies we reported a series of broadly neutralizing mini-antibodies derived from a synthetic naïve human combinatorial antibody library by panning against a mimetic of the trimeric N-HR coiled coil, followed by affinity maturation using targeted diversification of the CDR-H2 loop. Here we report crystal structures of the N-HR mimetic 5-Helix with two Fabs that represent the extremes of this series: Fab 8066 is broadly neutralizing across a wide panel of B and C type HIV-1 viruses, whereas Fab 8062 is non-neutralizing. The crystal structures reveal important differences in the conformations of the CDR-H2 loops in the complexes that propagate into other regions of the antigen-antibody interface, and suggest that both neutralization properties and affinity for the target can be attributed, at least in part, to the differences in the interactions of the CDR-H2 loops with the antigen. Furthermore, modeling of the complex of an N-HR trimer with three Fabs suggests that the CDR-H2 loop may be involved in close intermolecular contacts between neighboring antibody molecules, and that such contacts may hinder the formation of complexes between the N-HR trimer and more than one antibody molecule depending on the conformation of the bound CDR-H2 loop which is defined by its interactions with antigen. Comparison with the crystal structure of the complex of 5-Helix with another neutralizing monoclonal antibody known as D5, derived using an entirely different antibody library and panning procedure, reveals remarkable convergence in the optimal sequence and conformation of the CDR-H2 loop.     

ScFv multimers of the anti-neuraminidase antibody NC10: shortening of the linker in single-chain Fv fragment assembled in V(L) to V(H) orientation drives the formation of dimers, trimers, tetramers and higher molecular mass multimers

Synthetic genes encoding single-chain variable fragments (scFvs) of NC10 anti-neuraminidase antibody were constructed by joining the V(L) and V(H) domains with linkers of fifteen, five, four, three, two, one and zero residues. These V(L)-V(H) constructs were expressed in Escherichia coli and the resulting proteins were characterized and compared with the previously characterized NC10 scFv proteins assembled in V(H)-V(L) orientation. Size-exclusion chromatography and electron microscope images of complexes formed between various NC10 scFvs and anti-idiotype Fab' were used to analyse the oligomeric status of these scFvs. The result showed that as the linker length between V(L) and V(H) was reduced, different patterns of oligomerization were observed compared with those with V(H)-V(L) isomers. As was the case for V(H)-V(L) orientation, the scFv-15 V(L)-V(H) protein existed mainly as a monomer whereas dimer (diabody) was a predominant conformation for the scFv-5, scFv-4 and scFv-3 V(L)-V(H) proteins. In contrast to the V(H)-V(L) isomer, direct ligation of V(L) to V(H) led to the formation of predominantly a tetramer (tetrabody) rather than to an expected trimer (triabody). Furthermore, the transition between dimers and higher order oligomers was not as distinct as for V(H)-V(L). Thus reducing the linker length in V(L)-V(H) from three to two residues did not precisely dictate a transition between dimers and tetramers. Instead, two-residue as well as one-residue linked scFvs formed a mixture of dimers, trimers and tetramers.     

Single-chain Fv multimers of the anti-neuraminidase antibody NC10: the residue at position 15 in the V(L) domain of the scFv-0 (V(L)-V(H)) molecule is primarily responsible for formation of a tetramer-trimer equilibrium

Single-chain variable fragment of the murine monoclonal antibody NC10 specific to influenza virus N9 neuraminidase, joined directly in the V(L) to V(H) orientation (scFv-0), forms an equilibrium mixture of tetramer and trimer with the tetramer as the preferred multimeric species. In contrast, the V(H)-V(L) isomer was previously shown to exist exclusively as a trimer. Computer-generated trimeric and tetrameric scFv models, based on the refined crystal structure for NC10 Fv domain, were constructed and used to evaluate factors influencing the transition between V(L)-V(H) trimer and tetramer. These model structures indicated that steric restrictions between loops spanning amino acid residues L55-L59 and L13-L17 from the two adjacent V(L) domains within the V(L)-V(H) trimer were responsible for four scFv-0 molecules assembling to form a tetramer. In particular, leucine at position L15 and glutamate at position L57 appeared to interfere significantly with each other. To minimize this steric interference, the site-directed mutagenesis technique was used to construct several NC10 scFv-0 clones with mutations at these positions. Size-exclusion chromatographic analyses revealed that several of these mutations resulted in the production of NC10 scFv-0 proteins with significantly altered tetramer-trimer equilibrium ratios. In particular, introduction of a polar residue, such as asparagine or threonine, at position L15 generated a highly stable NC10 scFv-0 trimer.     

Crystal structure of HEL4, a soluble, refoldable human V(H) single domain with a germ-line scaffold

The antigen binding site of antibodies usually comprises associated heavy (V(H)) and light (V(L)) chain variable domains, but in camels and llamas, the binding site frequently comprises the heavy chain variable domain only (referred to as V(HH)). In contrast to reported human V(H) domains, V(HH) domains are well expressed from bacteria and yeast, are readily purified in soluble form and refold reversibly after heat-denaturation. These desirable properties have been attributed to highly conserved substitutions of the hydrophobic residues of V(H) domains, which normally interact with complementary V(L) domains. Here, we describe the discovery and characterisation of an isolated human V(H) domain (HEL4) with properties similar to those of V(HH) domains. HEL4 is highly soluble at concentrations of > or =3 mM, essentially monomeric and resistant to aggregation upon thermodenaturation at concentrations as high as 56 microM. However, in contrast to V(HH) domains, the hydrophobic framework residues of the V(H):V(L) interface are maintained and the only sequence changes from the corresponding human germ-line segment (V3-23/DP-47) are located in the loops comprising the complementarity determining regions (CDRs). The crystallographic structure of HEL4 reveals an unusual feature; the side-chain of a framework residue (Trp47) is flipped into a cavity formed by Gly35 of CDR1, thereby increasing the hydrophilicity of the V(H):V(L) interface. To evaluate the specific contribution of Gly35 to domain properties, Gly35 was introduced into a V(H) domain with poor solution properties. This greatly enhanced the recovery of the mutant from a gel filtration matrix, but had little effect on its ability to refold reversibly after heat denaturation. Our results confirm the importance of a hydrophilic V(H):V(L) interface for purification of isolated V(H) domains, and constitute a step towards the design of isolated human V(H) domains with practical properties for immunotherapy.     

The structural repertoire of the human V kappa domain.

In humans, the gene for the V kappa domain is produced by the recombination of one of 40 functional V kappa segments and one of five functional J kappa segments. We have analysed the sequences of these germline segments and of 736 rearranged V kappa genes to determine the repertoire of main chain conformations, or canonical structures, they encode. Over 96% of the sequences correspond to one of four canonical structures for the first antigen binding loop (L1) and one canonical structure for the second antigen binding loop (L2). Junctional diversity produces some variation in the length of the third antigen binding loop (L3) and in the identity of residues at the V kappa-J kappa join. However, this is limited and 70% of the rearranged sequences correspond to one of three known canonical structures for the L3 region. Furthermore, we show that the canonical structures selected during the primary response are conserved during affinity maturation: the key residues that determine the conformations of the antigen binding loops are unmutated or undergo conservative mutation. The implications of these results for immune recognition are discussed.     

Structural classification of CDR-H3 revisited: a lesson in antibody modeling

Among the six complementarity-determining regions (CDRs) in the variable domains of an antibody, the third CDR of the heavy chain (CDR-H3), which lies in the center of the antigen-binding site, plays a particularly important role in antigen recognition. CDR-H3 shows significant variability in its length, sequence, and structure. Although difficult, model building of this segment is the most critical step in antibody modeling. Since our first proposal of the "H3-rules," which classify CDR-H3 structure based on amino acid sequence, the number of experimentally determined antibody structures has increased. Here, we revise these H3-rules and propose an improved classification scheme for CDR-H3 structure modeling. In addition, we determine the common features of CDR-H3 in antibody drugs as well as discuss the concept of "antibody druggability," which can be applied as an indicator of antibody evaluation during drug discovery.     

CDR-H3 loop ensemble in solution – conformational selection upon antibody binding

ABSTRACT

We analyzed pairs of protein-binding, peptide-binding and hapten-binding antibodies crystallized as complex and in the absence of the antigen with and without conformational differences upon binding in the complementarity-determining region (CDR)-H3 loop. Here, we introduce a molecular dynamics-based approach to capture a diverse conformational ensemble of the CDR-H3 loop in solution. The results clearly indicate that the inherently flexible CDR-H3 loop indeed needs to be characterized as a conformational ensemble. The conformational changes of the CDR-H3 loop in all antibodies investigated follow the paradigm of conformation selection, because we observe the experimentally determined binding competent conformation without the presence of the antigen within the ensemble of pre-existing conformational states in solution before binding. We also demonstrate for several examples that the conformation observed in the antibody crystal structure without antigen present is actually selected to bind the carboxyterminal tail region of the antigen-binding fragment (Fab). Thus, special care must be taken when characterizing antibody CDR-H3 loops by Fab X-ray structures, and the possibility that pre-existing conformations are present should always be considered.     

Antiferritin VL homodimer binds human spleen ferritin with high specificity

The antiferritin variable light domain (VL) dimer binds human spleen ferritin ( approximately 85% L subunits) but with approximately 50-fold lower affinity, K(a)=4 x 10(7) x M(-1), than the parent F11 antibody (K(a)=2.1 x 10(9) x M(-1)). The VL dimer does not recognize either rL (100% L subunits) or rH (100% H subunits) human ferritin, whereas the parent antibody recognizes rL-ferritin. To help explain the differences in ferritin binding affinities and specificities, the crystal structure of the VL domain (2.8A resolution) was determined by molecular replacement and models of the antiferritin VL-VH dimer were made on the basis of antilysozyme antibody D1.3. The domain interface is smaller in the VL dimer but a larger number of interdomain hydrogen bonds may prevent rearrangement on antigen binding. The antigen binding surface of the VL dimer is flatter, lacking a negatively charged pocket found in the VL-VH models, contributed by the CDR3 loop of the VH domain. Loop CDR2 (VL dimer) is located away from the antigen binding site, while the corresponding loop of the VH domain would be located within the antigen binding site. Together these differences lead to 50-fold lower binding affinity in the VL dimer and to more restricted specificity than is seen for the parent antibody.     

Structure and dynamics of the anti-AMCV scFv(F8): effects of selected mutations on the antigen combining site

The recombinant antibody fragment scFv(F8), which recognizes the coat protein of the plant virus AMCV, is characterized by peculiar high in vitro stability and functional folding even in reducing environments, making it fit for designing stable antibodies with desired properties. Mutagenesis and functional analysis evidenced two residues, at positions 47 and 58 of the V(H) chain, playing a crucial role in the antigen binding recognition. Here, we used a computational procedure to assess the effects of these mutations on the stability, structure and dynamics of the antigen-binding site. Structural models of the wild type scFv(F8) and of its H47 and H58 mutants were built by homology modelling and assessed by multiple 15.5ns of molecular dynamics simulations. Computational results indicate that the 47H substitution strongly affects the CDR-H(2) conformation, destabilizes the V(H)/V(L) interface and confers high conformational flexibility to the antigen-binding site, leading the mutant to functional loss. The mutation at position H58 strenghtens the binding site, bestowing a high antigen specificity on the mutant. The essential dynamics and the analysis of the protein-solvent interface further corroborate the correspondence between the extent of the structurally-determined flexibility of the binding site with the different functional behaviours proved by the wild-type and its mutants. These results may have useful implications for structure-based design of antibody combining site.     

A variable light domain fluorogen activating protein homodimerizes to activate dimethylindole red

Novel fluorescent tools such as green fluorescent protein analogues and fluorogen activating proteins (FAPs) are useful in biological imaging for tracking protein dynamics in real time with a low fluorescence background. FAPs are single-chain variable fragments (scFvs) selected from a yeast surface display library that produce fluorescence upon binding a specific dye or fluorogen that is normally not fluorescent when present in solution. FAPs generally consist of human immunoglobulin variable heavy (V(H)) and variable light (V(L)) domains covalently attached via a glycine- and serine-rich linker. Previously, we determined that the yeast surface clone, V(H)-V(L) M8, could bind and activate the fluorogen dimethylindole red (DIR) but that the fluorogen activation properties were localized to the M8V(L) domain. We report here that both nuclear magnetic resonance and X-ray diffraction methods indicate the M8V(L) forms noncovalent, antiparallel homodimers that are the fluorogen activating species. The M8V(L) homodimers activate DIR by restriction of internal rotation of the bound dye. These structural results, together with directed evolution experiments with both V(H)-V(L) M8 and M8V(L), led us to rationally design tandem, covalent homodimers of M8V(L) domains joined by a flexible linker that have a high affinity for DIR and good quantum yields.     

Human immunodeficiency virus type 1 V1-V2 envelope loop sequences expand and add glycosylation sites over the course of infection, and these modifications affect antibody neutralization sensitivity

Over the course of infection, human immunodeficiency virus type 1 (HIV-1) continuously adapts to evade the evolving host neutralizing antibody responses. Changes in the envelope variable loop sequences, particularly the extent of glycosylation, have been implicated in antibody escape. To document modifications that potentially influence antibody susceptibility, we compared envelope variable loops 1 and 2 (V1-V2) from multiple sequences isolated at the primary phase of infection to those isolated around 2 to 3 years into the chronic phase of infection in nine women with HIV-1 subtype A. HIV-1 sequences isolated during chronic infection had significantly longer V1-V2 loops, with a significantly higher number of potential N-linked glycosylation sites, than the sequences isolated early in infection. To assess the effects of these V1-V2 changes on antibody neutralization and infectivity, we created chimeric envelope sequences, which incorporated a subject's V1-V2 sequences into a common subtype A envelope backbone and then used them to generate pseudotyped viruses. Compared to the parent virus, the introduction of a subject's early-infection V1-V2 envelope variable loops rendered the chimeric envelope more sensitive to that subject's plasma samples but only to plasma samples collected >6 months after the sequences were isolated. Neutralization was not detected with the same plasma when the early-infection V1-V2 sequences were replaced with chronic-infection V1-V2 sequences, suggesting that changes in V1-V2 contribute to antibody escape. Pseudotyped viruses with V1-V2 segments from different times in infection, however, showed no significant difference in neutralization sensitivity to heterologous pooled plasma, suggesting that viruses with V1-V2 loops from early in infection were not inherently more neutralization sensitive. Pseudotyped viruses bearing chimeric envelopes with early-infection V1-V2 sequences showed a trend in infecting cells with low CD4 concentrations more efficiently, while engineered viruses with V1-V2 sequences isolated during chronic infection were moderately better at infecting cells with low CCR5 concentrations. These studies suggest that changes within the V1-V2 envelope domains over the course of an infection influence sensitivity to autologous neutralizing antibodies and may also impact host receptor/coreceptor interactions.