Eric S. Day

Small-Molecule Inhibition of TNF-α

We have identified a small-molecule inhibitor of tumor necrosis factor α (TNF-α) that promotes subunit disassembly of this trimeric cytokine family member. The compound inhibits TNF-α activity in biochemical and cell-based assays with median inhibitory concentrations of 22 and 4.6 micromolar, respectively. Formation of an intermediate complex between the compound and the intact trimer results in a 600-fold accelerated subunit dissociation rate that leads to trimer dissociation. A structure solved by x-ray crystallography reveals that a single compound molecule displaces a subunit of the trimer to form a complex with a dimer of TNF-α subunits.

Comparison of Soluble Decoy IgG Fusion Proteins of BAFF-R and BCMA as Antagonists for BAFF

BAFF is considered a therapeutic target because dysregulated production of BAFF can induce systemic lupus erythematosus-like phenotype in mice, and elevated levels of BAFF are associated with disease severity in systemic lupus erythematosus and rheumatoid arthritis patients. Fc fusion decoy receptors, BCMA-Fc and BAFF-R-Fc, are therapeutic candidates for blocking BAFF. While studying their interactions with BAFF, we found that BAFF-R-Fc is more effective than BCMA-Fc for blocking BAFF binding to its receptors. We also found that a trimeric BAFF can bind more than one BAFF-R-Fc but only one BCMA-Fc. Moreover, we show that, in contrast to monovalent BAFF-R-Fc, monovalent BCMA does not form stable complexes with BAFF. Differences in their interaction with BAFF predict BAFF-R-Fc would be a better inhibitor. Indeed, we show BAFF-R-Fc is 10-fold more efficacious than BCMA-Fc for blocking BAFF-induced B cell proliferation in vitro and for blocking BAFF-mediated survival of mouse splenic B lymphocytes in vivo.

Development of an Fn14 agonistic antibody as an anti-tumor agent

TWEAK, a TNF family ligand with pleiotropic cellular functions, was originally described as capable of inducing tumor cell death in vitro. TWEAK functions by binding its receptor, Fn14, which is up-regulated on many human solid tumors. Herein, we show that intratumoral administration of TWEAK, delivered either by an adenoviral vector or in an immunoglobulin Fc-fusion form, results in significant inhibition of tumor growth in a breast xenograft model. To exploit the TWEAK-Fn14 pathway as a therapeutic target in oncology, we developed an anti-Fn14 agonistic antibody, BIIB036. Studies described herein show that BIIB036 binds specifically to Fn14 but not other members of the TNF receptor family, induces Fn14 signaling, and promotes tumor cell apoptosis in vitro. In vivo, BIIB036 effectively inhibits growth of tumors in multiple xenograft models, including colon (WiDr), breast (MDA-MB-231), and gastric (NCI-N87) tumors, regardless of tumor cell growth inhibition response observed to BIIB036 in vitro. The anti-tumor activity in these cell lines is not TNF-dependent. Increasing the antigen-binding valency of BIB036 significantly enhances its anti-tumor effect, suggesting the contribution of higher order cross-linking of the Fn14 receptor. Full Fc effector function is required for maximal activity of BIIB036 in vivo, likely due to the cross-linking effect and/or ADCC mediated tumor killing activity. Taken together, the anti-tumor properties of BIIB036 validate Fn14 as a promising target in oncology and demonstrate its potential therapeutic utility in multiple solid tumor indications.

Small Molecule Inhibition of the TNF Family Cytokine CD40 Ligand Through a Subunit Fracture Mechanism

BIO8898 is one of several synthetic organic molecules that have recently been reported to inhibit receptor binding and function of the constitutively trimeric tumor necrosis factor (TNF) family cytokine CD40 ligand (CD40L, aka CD154). Small molecule inhibitors of protein-protein interfaces are relatively rare, and their discovery is often very challenging. Therefore, to understand how BIO8898 achieves this feat, we characterized its mechanism of action using biochemical assays and X-ray crystallography. BIO8898 inhibited soluble CD40L binding to CD40-Ig with a potency of IC(50) = 25 μM and inhibited CD40L-dependent apoptosis in a cellular assay. A co-crystal structure of BIO8898 with CD40L revealed that one inhibitor molecule binds per protein trimer. Surprisingly, the compound binds not at the surface of the protein but by intercalating deeply between two subunits of the homotrimeric cytokine, disrupting a constitutive protein-protein interface and breaking the proteins 3-fold symmetry. The compound forms several hydrogen bonds with the protein, within an otherwise hydrophobic binding pocket. In addition to the translational splitting of the trimer, binding of BIO8898 was accompanied by additional local and longer-range conformational perturbations of the protein, both in the core and in a surface loop. Binding of BIO8898 is reversible, and the resulting complex is stable and does not lead to detectable dissociation of the protein trimer. Our results suggest that a set of core aromatic residues that are conserved across a subset of TNF family cytokines might represent a generic hot-spot for the induced-fit binding of trimer-disrupting small molecules.

Binding Efficiency of Protein-Protein Complexes

We examine the relationship between binding affinity and interface size for reversible protein-protein interactions (PPIs), using cytokines from the tumor necrosis factor (TNF) superfamily and their receptors as a test case. Using surface plasmon resonance, we measured single-site binding affinities for binding of the large receptor TNFR1 to its ligands TNFα (K(D) = 1.4 ± 0.4 nM) and lymphotoxin-α (K(D) = 50 ± 10 nM), and also for binding of the small receptor Fn14 to TWEAK (K(D) = 70 ± 10 nM). We additionally assembled data for all other TNF-TNFR family complexes for which reliable single-site binding affinities have been reported. We used these values to calculate the binding efficiencies, defined as binding energy per square angstrom of surface area buried at the contact interface, for nine of these complexes for which cocrystal structures are available, and compared the results to those for a set of 144 protein-protein complexes with published affinities. The results show that the most efficient PPI complexes generate ~20 cal mol(-1) Å(-2) of binding energy. A minimal contact area of ~500 Å(2) is required for a stable complex, required to generate sufficient interaction energy to pay the entropic cost of colocalizing two proteins from 1 M solution. The most compact and efficient TNF-TNFR complex was the BAFF-BR3 complex, which achieved ~80% of the maximal achievable binding efficiency. Other small receptors also gave high binding efficiencies, while the larger receptors generated only 44-49% of this limit despite interacting primarily through just a single small domain. The results provide new insight into how much binding energy can be generated by a PPI interface of a given size, and establish a quantitative method for predicting how large a natural or engineered contact interface must be to achieve a given level of binding affinity.

Effect of Divalent Cations on the Affinity and Selectivity of α4 Integrins Towards the Integrin Ligands Vascular Cell Adhesion Molecule-1 and Mucosal Addressin Cell Adhesion Molecule-1: Ca 2+ Activation of Integrin α4β1 Confers a Distinct Ligand Specificity

A microtiter plate assay measuring the binding of cells expressing integrins f 4 g 1 or f 4 g 7 to VCAM-1 and MAdCAM-1, expressed as Ig fusion proteins, was used to explore the interplay between the variables of integrin g -chain, identity and density of ligand, and identity and concentration of activating cations. Both Mn 2+ and Mg 2+ supported binding of either integrin to either ligand. Ca 2+ supported only the binding of f 4 g 1 to VCAM-Ig. Cation concentrations required for half-maximal binding (EC 50 ) ranged from 0.8-280 w M for Mn 2+ and 0.8-30 mM for Mg 2+, being thus 2-3 logs lower for Mn 2+ compared to Mg 2+ independent of ligand. EC 50 values for binding of f 4 g 1 to VCAM-Ig were 30-45-fold lower compared to MAdCAM-Ig, while f 4 g 7 showed an opposite 3-15-fold selectivity for MAdCAM-Ig over VCAM-Ig. The density of ligand required for adhesion via f 4 g 1 was markedly lower with Mn 2+ versus Mg 2+, and with VCAM-Ig versus MAdCAM-Ig. These results were interpreted in terms of a coupled equilibrium model, in which binding of activating metal ions and of integrin ligands each stabilizes activated integrin. We conclude that Mn 2+ and Mg 2+ bind to common regulatory sites with different affinities, producing similar activated states of the integrin. The resulting activated f 4 g 1 binds more strongly to VCAM-Ig versus MAdCAM-Ig by 30-45-fold, while similarly activated f 4 g 7 binds more strongly to MAdCAM-Ig versus VCAM-Ig by 3-15-fold. Inhibition studies showed that Ca 2+ also binds to regulatory sites on both integrins. However, the Ca 2+ -activated state of f 4 g 1 is distinct from that achieved by Mn 2+ and Mg 2+, possessing increased selectivity for binding to VCAM-1 versus MAdCAM-1.

Determining the affinity and stoichiometry of interactions between unmodified proteins in solution using Biacore

We describe a general Biacore method for measuring equilibrium binding affinities and stoichiometries for interactions between unmodified proteins and their unmodified ligands free in solution. Mixtures of protein and ligand are preequilibrated at different ratios in solution and then analyzed by Biacore using a sensor chip surface that detects only unbound analyte. Performing the Biacore analysis under mass transport limited conditions allows the concentration of unbound analyte to be determined from the initial velocity of binding. Plots of initial velocity versus the concentration of the varied binding partner are fitted to a quadratic binding equation to give the affinity and stoichiometry of binding. We demonstrate the method using soluble Her2 extracellular domain binding to monovalent, bivalent, and trivalent forms of an anti-Her2 antibody. The affinity we measured agrees with that obtained from conventional Biacore kinetic analysis, and the stoichiometries for the resulting 1:1, 1:2, and 1:3 complexes were confirmed by gel filtration with in-line light scattering. The method is applicable over an affinity range of approximately 100 pM to 1 μM and is particularly useful when there is concern that covalently modifying one or the other binding partner might affect its binding properties or where multivalency might otherwise complicate a quantitative analysis of binding.

Structure of the extracellular domains of human and Xenopus Fn14: implications in the evolution of TWEAK and Fn14 interactions

TWEAK (TNF homologue with weak apoptosis‐inducing activity) and Fn14 (fibroblast growth factor‐inducible protein 14) are members of the tumor necrosis factor (TNF) ligand and receptor super‐families. Having observed that Xenopus Fn14 cross‐reacts with human TWEAK, despite its relatively low sequence homology to human Fn14, we examined the conservation in tertiary fold and binding interfaces between the two species. Our results, combining NMR solution structure determination, binding assays, extensive site‐directed mutagenesis and molecular modeling, reveal that, in addition to the known and previously characterized β−hairpin motif, the helix‐loop‐helix motif makes an essential contribution to the receptor/ligand binding interface. We further discuss the insight provided by the structural analyses regarding how the cysteine‐rich domains of the TNF receptor super‐family may have evolved over time.

An antibody loop replacement design feasibility study and a loop-swapped dimer structure

A design approach was taken to investigate the feasibility of replacing single complementarity determining region (CDR) antibody loops. This approach may complement simpler mutation-based strategies for rational antibody design by expanding conformation space. Enormous crystal structure diversity is available, making CDR loops logical targets for structure-based design. A detailed analysis for the L1 loop shows that each loop length takes a distinct conformation, thereby allowing control on a length scale beyond that accessible to simple mutations. The L1 loop in the anti-VLA1 antibody was replaced with the L2 loop residues longer in an attempt to add an additional hydrogen bond and fill space on the antibody-antigen interface. The designs expressed well, but failed to improve affinity. In an effort to learn more, one design was crystallized and data were collected at 1.9 A resolution. The designed L1 loop takes the qualitatively desired conformation; confirming that loop replacement by design is feasible. The crystal structure also shows that the outermost loop (residues Leu51-Ser68) is domain swapped with another monomer. Tryptophan fluorescence measurements were used to monitor unfolding as a function of temperature and indicate that the loop involved in domain swapping does not unfold below 60 degrees C. The domain-swapping is not directly responsible for the affinity loss, but is likely a side-effect of the structural instability which may contribute to affinity loss. A second round of design was successful in eliminating the dimerization through mutation of a residue (Leu51Ser) at the joint of the domain-swapped loop.