GNE-235: A Lead Compound Selective for the Second Bromodomain of PBRM1

Bromodomains are acetyl-lysine binding modules that are found in different classes of chromatin-interacting proteins. Among these are large chromatin remodeling complexes such as BAF and PBAF (variants of human SWI/SNF). Previous work has identified chemical probes targeting a subset of the bromodomains present in the BAF and PBAF complexes. Selective inhibitors of the individual bromodomains have proven challenging to discover, as the domains are highly similar. Here, elaboration of an aminopyridazine scaffold used previously to develop probes for the bromodomains of SMARCA2, SMARCA4, and the fifth bromodomain of PBRM1 yielded compounds with both potency and unusual selectivity for the second bromodomain of PBRM1. One of these, GNE-235, and its enantiomer control GNE-234 are suggested for initial cellular investigations of the function of the second bromodomain of PBRM1.

GNE-064: A Potent, Selective, and Orally Bioavailable Chemical Probe for the Bromodomains of SMARCA2 and SMARCA4 and the Fifth Bromodomain of PBRM1

Bromodomains are acetyllysine recognition domains present in a variety of human proteins. Bromodomains also bind small molecules that compete with acetyllysine, and therefore bromodomains have been targets for drug discovery efforts. Highly potent and selective ligands with good cellular permeability have been proposed as chemical probes for use in exploring the functions of many of the bromodomain proteins. We report here the discovery of a class of such inhibitors targeting the family VIII bromodomains of SMARCA2 (BRM) and SMARCA4 (BRG1), and PBRM1 (polybromo-1) bromodomain 5. We propose one example from this series, GNE-064, as a chemical probe for the bromodomains SMARCA2, SMARCA4, and PBRM1(5) with the potential for in vivo use.

Trifluoroacetyl Lysine as a Bromodomain Binding Mimic of Lysine Acetylation

Genetic code expansion has proven invaluable to the elucidation of functions of defined protein modifications through the site-specific incorporation of noncanonical amino acids. The use of nonhydrolyzable derivatives of post-translational modifications can greatly increase site stoichiometry and half-life. Investigating acetyllysine reader domain (bromodomain) interactions with acetylated nonhistone proteins is challenging due to the limited tools available and dynamic nature of this post-translational modification. Here, we demonstrate that bromodomains bind acetyllysine peptides and those substituted with an acetyllysine derivative, trifluoroacetyllysine, with similar affinity and selectivity. Importantly, both trifluoroacetyllysine and acetyllysine can be site-specifically incorporated into proteins expressed in bacterial and mammalian cells, and the strong electron-withdrawing trifluoro substituent makes the latter resistant to deacetylation by sirtuins (SIRTs). The controlled expression of SIRT-resistant, site-specifically acetylated transcription factors expands the set of available tools for determining the function of acetylation, and it serves as a template for investigating bromodomain interactions with acetylated transcription factors.

Expansion of the ISWI chromatin remodeler family with new active complexes

ISWI chromatin remodelers mobilize nucleosomes to control DNA accessibility. Complexes isolated to date pair one of six regulatory subunits with one of two highly similar ATPases. However, we find that each endogenously expressed ATPase co-purifies with every regulatory subunit, substantially increasing the diversity of ISWI complexes, and we additionally identify BAZ2B as a novel, seventh regulatory subunit. Through reconstitution of catalytically active human ISWI complexes, we demonstrate that the new interactions described here are stable and direct. Finally, we profile the nucleosome remodeling functions of the now expanded family of ISWI chromatin remodelers. By revealing the combinatorial nature of ISWI complexes, we provide a basis for better understanding ISWI function in normal settings and disease.

GNE-886: A Potent and Selective Inhibitor of the Cat Eye Syndrome Chromosome Region Candidate 2 Bromodomain (CECR2)

The biological function of bromodomains, epigenetic readers of acetylated lysine residues, remains largely unknown. Herein we report our efforts to discover a potent and selective inhibitor of the bromodomain of cat eye syndrome chromosome region candidate 2 (CECR2). Screening of our internal medicinal chemistry collection led to the identification of a pyrrolopyridone chemical lead, and subsequent structure-based drug design led to a potent and selective CECR2 bromodomain inhibitor (GNE-886) suitable for use as an in vitro tool compound.

Inhibition of bromodomain-containing protein 9 for the prevention of epigenetically-defined drug resistance

Bromodomain-containing protein 9 (BRD9), an epigenetic "reader" of acetylated lysines on post-translationally modified histone proteins, is upregulated in multiple cancer cell lines. To assess the functional role of BRD9 in cancer cell lines, we identified a small-molecule inhibitor of the BRD9 bromodomain. Starting from a pyrrolopyridone lead, we used structure-based drug design to identify a potent and highly selective in vitro tool compound 11, (GNE-375). While this compound showed minimal effects in cell viability or gene expression assays, it showed remarkable potency in preventing the emergence of a drug tolerant population in EGFR mutant PC9 cells treated with EGFR inhibitors. Such tolerance has been linked to an altered epigenetic state, and 11 decreased BRD9 binding to chromatin, and this was associated with decreased expression of ALDH1A1, a gene previously shown to be important in drug tolerance. BRD9 inhibitors may therefore show utility in preventing epigenetically-defined drug resistance.

Discovery of a Potent and Selective in Vivo Probe (GNE-272) for the Bromodomains of CBP/EP300

The single bromodomain of the closely related transcriptional regulators CBP/EP300 is a target of much recent interest in cancer and immune system regulation. A co-crystal structure of a ligand-efficient screening hit and the CBP bromodomain guided initial design targeting the LPF shelf, ZA loop, and acetylated lysine binding regions. Structure-activity relationship studies allowed us to identify a more potent analogue. Optimization of permeability and microsomal stability and subsequent improvement of mouse hepatocyte stability afforded 59 (GNE-272, TR-FRET IC₅₀ = 0.02 μM, BRET IC₅₀ = 0.41 μM, BRD4(1) IC₅₀ = 13 μM) that retained the best balance of cell potency, selectivity, and in vivo PK. Compound 59 showed a marked antiproliferative effect in hematologic cancer cell lines and modulates MYC expression in vivo that corresponds with antitumor activity in an AML tumor model.

Diving into the Water: Inducible Binding Conformations for BRD4, TAF1(2), BRD9, and CECR2 Bromodomains

The biological role played by non-BET bromodomains remains poorly understood, and it is therefore imperative to identify potent and highly selective inhibitors to effectively explore the biology of individual bromodomain proteins. A ligand-efficient nonselective bromodomain inhibitor was identified from a 6-methyl pyrrolopyridone fragment. Small hydrophobic substituents replacing the N-methyl group were designed directing toward the conserved bromodomain water pocket, and two distinct binding conformations were then observed. The substituents either directly displaced and rearranged the conserved solvent network, as in BRD4(1) and TAF1(2), or induced a narrow hydrophobic channel adjacent to the lipophilic shelf, as in BRD9 and CECR2. The preference of distinct substituents for individual bromodomains provided selectivity handles useful for future lead optimization efforts for selective BRD9, CECR2, and TAF1(2) inhibitors.

Fragment-Based Discovery of a Selective and Cell-Active Benzodiazepinone CBP/EP300 Bromodomain Inhibitor (CPI-637)

CBP and EP300 are highly homologous, bromodomain-containing transcription coactivators involved in numerous cellular pathways relevant to oncology. As part of our effort to explore the potential therapeutic implications of selectively targeting bromodomains, we set out to identify a CBP/EP300 bromodomain inhibitor that was potent both in vitro and in cellular target engagement assays and was selective over the other members of the bromodomain family. Reported here is a series of cell-potent and selective probes of the CBP/EP300 bromodomains, derived from the fragment screening hit 4-methyl-1,3,4,5-tetrahydro-2H-benzo[b][1,4]diazepin-2-one.

Regulatory T Cell Modulation by CBP/EP300 Bromodomain Inhibition

Covalent modification of histones is a fundamental mechanism of regulated gene expression in eukaryotes, and interpretation of histone modifications is an essential feature of epigenetic control. Bromodomains are specialized binding modules that interact with acetylated histones, linking chromatin recognition to gene transcription. Because of their ability to function in a domain-specific fashion, selective disruption of bromodomain:acetylated histone interactions with chemical probes serves as a powerful means for understanding biological processes regulated by these chromatin adaptors. Here we describe the discovery and characterization of potent and selective small molecule inhibitors for the bromodomains of CREBBP/EP300 that engage their target in cellular assays. We use these tools to demonstrate a critical role for CREBBP/EP300 bromodomains in regulatory T cell biology. Because regulatory T cell recruitment to tumors is a major mechanism of immune evasion by cancer cells, our data highlight the importance of CREBBP/EP300 bromodomain inhibition as a novel, small molecule-based approach for cancer immunotherapy.

A Subset of Human Bromodomains Recognizes Butyryllysine and Crotonyllysine Histone Peptide Modifications

Bromodomains are epigenetic readers that are recruited to acetyllysine residues in histone tails. Recent studies have identified non-acetyl acyllysine modifications, raising the possibility that these might be read by bromodomains. Profiling the nearly complete human bromodomain family revealed that while most human bromodomains bind only the shorter acetyl and propionyl marks, the bromodomains of BRD9, CECR2, and the second bromodomain of TAF1 also recognize the longer butyryl mark. In addition, the TAF1 second bromodomain is capable of binding crotonyl marks. None of the human bromodomains tested binds succinyl marks. We characterized structurally and biochemically the binding to different acyl groups, identifying bromodomain residues and structural attributes that contribute to specificity. These studies demonstrate a surprising degree of plasticity in some human bromodomains but no single factor controlling specificity across the family. The identification of candidate butyryl- and crotonyllysine readers supports the idea that these marks could have specific physiological functions.

Deubiquitinase DUBA is a post-translational brake on interleukin-17 production in T cells

T-helper type 17 (TH17) cells that produce the cytokines interleukin-17A (IL-17A) and IL-17F are implicated in the pathogenesis of several autoimmune diseases. The differentiation of TH17 cells is regulated by transcription factors such as RORγt, but post-translational mechanisms preventing the rampant production of pro-inflammatory IL-17A have received less attention. Here we show that the deubiquitylating enzyme DUBA is a negative regulator of IL-17A production in T cells. Mice with DUBA-deficient T cells developed exacerbated inflammation in the small intestine after challenge with anti-CD3 antibodies. DUBA interacted with the ubiquitin ligase UBR5, which suppressed DUBA abundance in naive T cells. DUBA accumulated in activated T cells and stabilized UBR5, which then ubiquitylated RORγt in response to TGF-β signalling. Our data identify DUBA as a cell-intrinsic suppressor of IL-17 production.

Abstract A160: Recognition of UbcH5c and the nucleosome by the Bmi1/Ring1b ubiquitin ligase complex

Background and Objectives: The Polycomb repressive complex 1 (PRC1) mediates gene silencing, in part by monoubiquitination of histone H2A on lysine 119 (uH2A). Bmi1 and Ring1b are critical components of PRC1 that heterodimerize via their N-terminal RING domains to form an active E3 ubiquitin ligase. A longstanding question in this field is how a single lysine residue in the nucleosome is singled out for ubiquitin modification and why only one ubiquitin is added rather than a chain of ubiquitins. To better understand how Bmi1/Ring1b controls monoubiquitination of H2A, we structurally characterized the E2 and nucleosome interaction sites on Bmi1/Ring1b.

Methods: X-ray crystallography was used to determine the structure of a complex between Bmi1/Ring1b and UbcH5c, its cognate E2 enzyme. Fluorescence-polarization assays were used to measure the binding of Bmi1/Ring1b to a synthetic DNA duplex. Site-directed mutagenesis was used to assess the importance of residues at the potential Bmi1/Ring1b-DNA interface, and in vitro ubiquitin ligase assays were used to measure the catalytic activity of mutant complexes. Molecular modeling studies were performed using HADDOCK v2.0.

Results: We have determined the crystal structure of a complex between the Bmi1/Ring1b RING-RING heterodimer and the E2 enzyme UbcH5c and find that UbcH5c interacts with Ring1b only, in a manner fairly typical of E2-E3 interactions. However, we further show that the Bmi1/Ring1b RING domains bind directly to duplex DNA through a basic surface patch unique to the Bmi1/Ring1b RING-RING dimer. Mutation of residues on this interaction surface leads to a loss of H2A ubiquitination activity. Through site-directed mutagenesis we have been able to uncouple the E2 binding and the DNA binding activities of Bmi1/Ring1b, indicating that Bmi1/Ring1b uses distinct binding surfaces to recognize E2 and the nucleosomal substrate. Computational modeling of the interface between Bmi1/Ring1b-UbcH5c and the nucleosome suggests that Bmi1/Ring1b interacts with both nucleosomal DNA and an acidic patch on histone H4 to achieve specific mono-ubiquitination of H2A.

Conclusions: Our data show that the direct interaction of the RING domains with nucleosomal DNA is crucial for the ubiquitin ligase activity of Bmi1/Ring1b. To our knowledge, this is the first example of a RING-domain E3 ligase binding directly to its substrate via the RING domain. Our results point to a novel mechanism of substrate recognition, and control of product formation, by Bmi1/Ring1b.

Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2011 Nov 12-16; San Francisco, CA. Philadelphia (PA): AACR; Mol Cancer Ther 2011;10(11 Suppl):Abstract nr A160.

Wnt antagonists bind through a short peptide to the first β-propeller domain of LRP5/6

The Wnt pathway inhibitors DKK1 and sclerostin (SOST) are important therapeutic targets in diseases involving bone loss or damage. It has been appreciated that Wnt coreceptors LRP5/6 are also important, as human missense mutations that result in bone overgrowth (bone mineral density, or BMD, mutations) cluster to the E1 propeller domain of LRP5. Here, we report a crystal structure of LRP6 E1 bound to an antibody, revealing that the E1 domain is a peptide recognition module. Remarkably, the consensus E1 binding sequence is a close match to a conserved tripeptide motif present in all Wnt inhibitors that bind LRP5/6. We show that this motif is important for DKK1 and SOST binding to LRP6 and for inhibitory function, providing a detailed structural explanation for the effect of the BMD mutations.

Reconstitution of a frizzled8.Wnt3a.LRP6 signaling complex reveals multiple Wnt and Dkk1 binding sites on LRP6

Wnt/beta-catenin signaling is initiated at the cell surface by association of secreted Wnt with its receptors Frizzled (Fz) and low density lipoprotein receptor-related protein 5/6 (LRP5/6). The study of these molecular interactions has been a significant technical challenge because the proteins have been inaccessible in sufficient purity and quantity. In this report we describe insect cell expression and purification of soluble mouse Fz8 cysteine-rich domain and human LRP6 extracellular domain and show that they inhibit Wnt/beta-catenin signaling in cellular assays. We determine the binding affinities of Wnts and Dickkopf 1 (Dkk1) to the relevant co-receptors and reconstitute in vitro the Fz8 CRD.Wnt3a.LRP6 signaling complex. Using purified fragments of LRP6, we further show that Wnt3a binds to a region including only the third and fourth beta-propeller domains of LRP6 (E3E4). Surprisingly, we find that Wnt9b binds to a different part of the LRP6 extracellular domain, E1E2, and we demonstrate that Wnt3a and Wnt9b can bind to LRP6 simultaneously. Dkk1 binds to both E1E2 and E3E4 fragments and competes with both Wnt3a and Wnt9b for binding to LRP6. The existence of multiple, independent Wnt binding sites on the LRP6 co-receptor suggests new possibilities for the architecture of Wnt signaling complexes and a model for broad-spectrum inhibition of Wnt/beta-catenin signaling by Dkk1.

A class of 2,4-bisanilinopyrimidine Aurora A inhibitors with unusually high selectivity against Aurora B

The two major Aurora kinases carry out critical functions at distinct mitotic stages. Selective inhibitors of these kinases, as well as pan-Aurora inhibitors, show antitumor efficacy and are now under clinical investigation. However, the ATP-binding sites of Aurora A and Aurora B are virtually identical, and the structural basis for selective inhibition has therefore not been clear. We report here a class of bisanilinopyrimidine Aurora A inhibitors with excellent selectivity for Aurora A over Aurora B, both in enzymatic assays and in cellular phenotypic assays. Crystal structures of two of the inhibitors in complex with Aurora A implicate a single amino acid difference in Aurora B as responsible for poor inhibitory activity against this enzyme. Mutation of this residue in Aurora B (E161T) or Aurora A (T217E) is sufficient to swap the inhibition profile, suggesting that this difference might be exploited more generally to achieve high selectivity for Aurora A.

Aurora A: target invalidated?

Aurora kinases are important mitotic regulators, and Aurora kinase inhibitors are under investigation as treatments for cancer. An ongoing debate in the field is which Aurora kinase is the better drug target. A new study (Girdler et al., 2008, in this issue of Chemistry & Biology) pushes the case forward for Aurora B.

A pentacyclic aurora kinase inhibitor (AKI-001) with high in vivo potency and oral bioavailability

Aurora kinase inhibitors have attracted a great deal of interest as a new class of antimitotic agents. We report a novel class of Aurora inhibitors based on a pentacyclic scaffold. A prototype pentacyclic inhibitor 32 (AKI-001) derived from two early lead structures improves upon the best properties of each parent and compares favorably to a previously reported Aurora inhibitor, 39 (VX-680). The inhibitor exhibits low nanomolar potency against both Aurora A and Aurora B enzymes, excellent cellular potency (IC50 < 100 nM), and good oral bioavailability. Phenotypic cellular assays show that both Aurora A and Aurora B are inhibited at inhibitor concentrations sufficient to block proliferation. Importantly, the cellular activity translates to potent inhibition of tumor growth in vivo. An oral dose of 5 mg/kg QD is well tolerated and results in near stasis (92% TGI) in an HCT116 mouse xenograft model.

The mitotic regulator Survivin binds as a monomer to its functional interactor Borealin

Survivin is a member of the IAP (inhibitor of apoptosis) protein family, defined in part by the presence of a zinc-binding baculoviral inhibitory repeat (BIR) domain. Most BIR domains bind short sequences beginning with alanine, and in this manner, they recognize and block the action of key targets in apoptotic pathways. However, Survivin binds only very weakly to typical IAP ligands. Unique features of Survivin are the long C-terminal helix following the BIR domain and a short segment (linking the helix and BIR domains) that mediates Survivin homodimerization. Despite this detailed knowledge of the structure of Survivin itself, there is a current lack of understanding about how Survivin recognizes cellular binding partners, and consequently, many questions about Survivin function remain unanswered. We determined two co-crystal structures of Survivin and a minimal binding fragment from the chromosomal passenger protein Borealin, a well validated functional interactor. The interaction between Survivin and Borealin involves extensive packing between the long C-terminal helix of Survivin and a long Borealin helix. Surprisingly, an additional important interaction occurs between the Survivin homodimerization interface and a short segment of Borealin. This segment both structurally mimics and displaces one Survivin monomer. The relevance of this unexpected interaction was tested by mutagenesis of two key Borealin residues. Mutant Borealin introduced into HeLa cells failed to localize properly during mitosis and also caused mislocalization of other chromosomal passenger proteins. This suggests that the mutant is dominant-negative and confirms the functional importance of the interaction surface identified in the crystal structures.

Activation of the proapoptotic death receptor DR5 by oligomeric peptide and antibody agonists

The cell-extrinsic apoptotic pathway triggers programmed cell death in response to certain ligands that bind to cell-surface death receptors. Apoptosis is essential for normal development and homeostasis in metazoans, and furthermore, selective activation of the cell-extrinsic pathway in tumor cells holds considerable promise for cancer therapy. We used phage display to identify peptides and synthetic antibodies that specifically bind to the human proapoptotic death receptor DR5. Despite great differences in overall size and structure, the DR5-binding peptides and antibodies shared a tripeptide motif, which was conserved within a disulfide-constrained loop of the peptides and the third complementarity determining region of the antibody heavy chains. The X-ray crystal structure of an antibody in complex with DR5 revealed that the tripeptide motif is buried at the core of the interface, confirming its central role in antigen recognition. We found that certain peptides and antibodies exhibited potent proapoptotic activity against DR5-expressing SK-MES-1 lung carcinoma cells. These phage-derived ligands may be useful for elucidating DR5 activation at the molecular level and for creating synthetic agonists of proapoptotic death receptors.

Substrate capacity considerations in developing kinase assays

In developing a screening assay for a serine/threonine kinase, we evaluated various formats of an in-plate enzyme-linked immunosorbent assay (ELISA), as well as solution-phase kinase assays using either ELISA or AlphaScreen detection. Substrate was available both as a biotinylated 15-residue peptide and as a 25-residue peptide containing the same sequence expressed as a glutathione S-transferase fusion protein. When increasing concentrations of either of these substrates were coated directly onto ELISA plates, the rates of the kinase reactions progressively increased. In contrast, when the biotin-peptide was captured onto NeutrAvidin-coated plates, the finite peptide binding capacity of the plates limited the amount of substrate that could be incorporated into the assay system and thereby limited the rate of the reaction at a given kinase concentration. Solution-phase kinase reactions can tolerate high substrate concentrations; however, analysis of kinase reaction samples containing biotin-peptide concentrations higher than the binding capacity of NeutrAvidin-coated plates resulted in an inability to detect differences between reactions run at different substrate concentrations. For AlphaScreen detection following solution-phase kinase reactions, limitations in the binding capacity of the donor and acceptor beads caused loss of signal for substrate concentrations above the maximum binding capacity. Overall, the solution-phase assays required significantly more kinase than the in-plate assays (1-4 microg/ml versus <100 ng/ml, respectively). These studies demonstrate that the amount of substrate that can be incorporated into an assay system substantially affects the rate of the kinase reaction and therefore the amount of kinase required for the assay.

Phage-display as a tool for quantifying protein stability determinants

To address questions of protein stability, researchers have increasingly turned to combinatorial approaches that permit the rapid analysis of libraries of protein variants. Phage-display has proved to be a powerful tool for analyzing protein stability due to the large library size and the robustness of the phage particle to a variety of denaturing conditions. With the B1 domain of protein G (GB1) and a camelid heavy chain antibody as model systems, we are using phage-display libraries to experimentally address questions that have generally been addressed in silico, either through computational studies or statistical analysis of known protein structures. One effort has focused on identifying novel solutions to repacking the hydrophobic core of GB1, while maintaining stability comparable to the wild type protein. In a second study, a small set of substitutions in complimentarity-determining region 3 was found to stabilize the framework of the camelid antibody. Another major focus has been to obtain quantitative data on beta-sheet stability determinants. We have successfully adapted a phage-display method for quantitating affinities of protein variants (shotgun alanine scanning) to analysis of GB1 stability. Using this method, we have analyzed the energetic contributions of cross-strand side chain-side chain interactions. Finally, we discuss parameters to consider in using phage-display to discriminate subtle stability differences among fully folded variants. Overall, this method provides a fast approach for quantitatively addressing biophysical questions.

BAFF/BLyS receptor 3 comprises a minimal TNF receptor-like module that encodes a highly focused ligand-binding site

BAFF/BLyS, a member of the tumor necrosis family (TNF) superfamily of ligands, is a crucial survival factor for B cells. BAFF binds three receptors, TACI, BCMA, and BR3, with signaling through BR3 being essential for promoting B cell function. Typical TNF receptor (TNFR) family members bind their cognate ligands through interactions with two cysteine-rich domains (CRDs). However, the extracellular domain (ECD) of BR3 consists of only a partial CRD, with cysteine spacing distinct from other modules described previously. Herein, we report the solution structure of the BR3 ECD. A core region of only 19 residues adopts a stable structure in solution. The BR3 fold is analogous to the first half of a canonical TNFR CRD but is stabilized by an additional noncanonical disulfide bond. BAFF-binding determinants were identified by shotgun alanine-scanning mutagenesis of the BR3 ECD expressed on phage. Several of the key BAFF-binding residues are presented from a beta-turn that we have shown previously to be sufficient for ligand binding when transferred to a structured beta-hairpin scaffold [Kayagaki, N., Yan, M., Seshasayee, D., Wang, H., Lee, W., French, D. M., Grewal, I. S., Cochran, A. G., Gordon, N. C., Yin, J., Starovasnik, M. A, and Dixit, V. M. (2002) Immunity 10, 515-524]. Outside of the turn, mutagenesis identifies additional hydrophobic contacts that enhance the BAFF-BR3 interaction. The crystal structure of the minimal hairpin peptide, bhpBR3, in complex with BAFF reveals intimate packing of the six-residue BR3 turn into a cavity on the ligand surface. Thus, BR3 binds BAFF through a highly focused interaction site, unprecedented in the TNFR family.

Turn stability in beta-hairpin peptides: Investigation of peptides containing 3:5 type I G1 bulge turns

The turn-forming ability of a series of three-residue sequences was investigated by substituting them into a well-characterized beta-hairpin peptide. The starting scaffold, bhpW, is a disulfide-cyclized 10-residue peptide that folds into a stable beta-hairpin with two antiparallel strands connected by a two-residue reverse turn. Substitution of the central two residues with the three-residue test sequences leads to less stable hairpins, as judged by thiol-disulfide equilibrium measurements. However, analysis of NMR parameters indicated that each molecule retains a significant folded population, and that the type of turn adopted by the three-residue sequence is the same in all cases. The solution structure of a selected peptide with a PDG turn contained an antiparallel beta-hairpin with a 3:5 type I + G1 bulge turn. Analysis of the energetic contributions of individual turn residues in the series of peptides indicates that substitution effects have significant context dependence, limiting the predictive power of individual amino acid propensities for turn formation. The most stable and least stable sequences were also substituted into a more stable disulfide-cyclized scaffold and a linear beta-hairpin scaffold. The relative stabilities remained the same, suggesting that experimental measurements in the bhpW context are a useful way to evaluate turn stability for use in protein design projects. Moreover, these scaffolds are capable of displaying a diverse set of turns, which can be exploited for the mimicry of protein loops or for generating libraries of reverse turns.

Stability of cyclic beta-hairpins: asymmetric contributions from side chains of a hydrogen-bonded cross-strand residue pair

Amino acid structural propensities measured in "host-guest" model studies are often used in protein structure prediction or to choose appropriate residues in de novo protein design. While this concept has proven useful for helical structures, it is more difficult to apply successfully to beta-sheets. We have developed a cyclic beta-hairpin scaffold as a host for measurement of individual residue contributions to hairpin structural stability. Previously, we have characterized substitutions in non-backbone-hydrogen-bonded strand sites; relative stability differences measured in the cyclic host are highly predictive of changes in folding free energy for linear beta-hairpin peptides. Here, we examine the hydrogen-bonded strand positions of our host. Surprisingly, we find a large favorable contribution to stability from a valine (or isoleucine) substitution immediately preceding the C-terminal cysteine of the host peptide, but not at the cross-strand position of the host or in either strand of a folded linear beta-hairpin (trpzip peptide). Further substitutions in the peptides and NMR structural analysis indicate that the stabilizing effect of valine is general for CX(8)C cyclic hairpins and cannot be explained by particular side-chain-side-chain interactions. Instead, a localized decrease in twist of the peptide backbone on the N-terminal side of the cysteine allows the valine side chain to adopt a unique conformation that decreases the solvent accessibility of the peptide backbone. The conformation differs from the highly twisted (coiled) conformation of the trpzip hairpins and is more typical of conformations present in multistranded beta-sheets. This unexpected structural fine-tuning may explain why cyclic hairpins selected from phage-displayed libraries often have valine in the same position, preceding the C-terminal cysteine. It also emphasizes the diversity of structures accessible to beta-strands and the importance of considering not only "beta-propensity", but also hydrogen-bonding pattern and strand twist, when designing beta structures. Finally, we observe correlated, cooperative stabilization from side-chain substitutions on opposite faces of the hairpin. This suggests that cooperative folding in beta-hairpins and other small beta-structures is driven by cooperative strand-strand association.

Short constrained peptides that inhibit HIV-1 entry

Peptides corresponding to the C-terminal heptad repeat of HIV-1 gp41 (C-peptides) are potent inhibitors of HIV-1 entry into cells. Their mechanism of inhibition involves binding in a helical conformation to the central coiled coil of HIV-1 gp41 in a dominant-negative manner. Short C-peptides, however, have low binding affinity for gp41 and poor inhibitory activity, which creates an obstacle to the development of small drug-like C-peptides. To improve the inhibitory potency of short C-peptides that target the hydrophobic pocket region of gp41, we use two strategies to stabilize the C-peptide helix: chemical crosslinking and substitution with unnatural helix-favoring amino acids. In this study, the short linear peptide shows no significant inhibitory activity, but a constrained peptide (C14linkmid) inhibits cell-cell fusion at micromolar potency. Structural studies confirm that the constrained peptides bind to the gp41 hydrophobic pocket. Calorimetry reveals that, of the peptides analyzed, the most potent are those that best balance the changes in binding enthalpy and entropy, and surprisingly not those with the highest helical propensity as measured by circular dichroism spectroscopy. Our study reveals the thermodynamic basis of inhibition of an HIV C-peptide, demonstrates the utility of constraining methods for a short antiviral peptide inhibitor, and has implications for the future design of constrained peptides.

BAFF/BLyS receptor 3 binds the B cell survival factor BAFF ligand through a discrete surface loop and promotes processing of NF-kappaB2

The TNF-like ligand BAFF/BLyS is a potent survival factor for B cells. It binds three receptors: TACI, BCMA, and BR3. We show that BR3 signaling promotes processing of the transcription factor NF-kappaB2/p100 to p52. NF-kappaB2/p100 cleavage was abrogated in B cells from A/WySnJ mice possessing a mutant BR3 gene, but not in TACI or BCMA null B cells. Furthermore, wild-type mice injected with BAFF-neutralizing BR3-Fc protein showed reduced basal NF-kappaB2 activation. BR3-Fc treatment of NZB/WF1 mice, which develop a fatal lupus-like syndrome, inhibited NF-kappaB2 processing and attenuated the disease process. Since inhibiting the BR3-BAFF interaction has therapeutic ramifications, the ligand binding interface of BR3 was investigated and found to reside within a 26 residue core domain. When stabilized within a structured beta-hairpin peptide, six of these residues were sufficient to confer binding to BAFF.

Quantifying beta-sheet stability by phage display

The small immunoglobulin G (IgG)-binding protein GB1 is a favored model system for the study of individual residue contributions to the stability of beta-sheets. Nevertheless, only a few of the many possible combinations of mutations have been characterized, leaving many questions unanswered. In order to allow the simultaneous evaluation of libraries of mutants, we have adapted a phage-display method, called shotgun scanning. This method combines a binding (i.e. stability) selection with high-throughput sequence analysis. Relative folding free energies determined from GB1-phage sequence data agree well with published GB1 thermal stability studies, validating the use of phage display to conduct quantitative stability studies on GB1, and further suggesting that this method is generally applicable to mutational analysis of protein stability. Examination of residue pairing in our large collection of GB1 mutants indicates that specific side-chain-side-chain interactions are much less important to beta-sheet stability than individual residue contributions. The discrepancy between this observation and published studies can be traced to anomalous stability of the alanine-substituted GB1 variants typically used as reference states in double mutant-cycle analyses. Finally, the combination of large library sizes and a quantitative stability selection should allow phage-based "computation" to be applied to protein design problems.

Amino acid determinants of beta-hairpin conformation in erythropoeitin receptor agonist peptides derived from a phage display library

Display of peptide libraries on filamentous phage has led to the identification of peptides of the form X(2-5)CX(2)GPXTWXCX(2-5) (where X is a variable residue) that bind to the extra-cellular portion of the erythropoietin receptor (EPO-R). These peptides adopt beta-hairpin conformations when co-crystallized with EPO-R. Solution NMR studies reveal that the peptide is conformationally heterogeneous in the absence of receptor due to cis-trans isomerization about the Gly-Pro peptide bond. Replacement of the conserved threonine residue with glycine at the turn i+3 position produces a stable beta-hairpin conformation in solution, although this peptide no longer has activity in an EPO-R-dependent cell proliferation assay. A truncated form of the EPO-R-binding peptide (containing the i+3 glycine residue) also forms a highly populated, monomeric beta-hairpin. In contrast, phage-derived peptide antagonists of insulin-like growth factor binding protein 1 (IGFBP-1) have a high level of sequence identity with the truncated EPO-R peptide (eight of 12 residues) yet adopt a turn-alpha-helix conformation in solution. Peptides containing all possible pairwise amino acid substitutions between the EPO-R and IGFBP-1 peptides have been analyzed to assess the degree to which the non-conserved residues stabilize the hairpin or helix conformation. All four residues present in the original sequence are required for maximum population of either the beta-hairpin or alpha-helix conformation, although some substitutions have a more dominant effect. The results demonstrate that, within a given sequence, the observed conformation can be dictated by a small subset of the residues (in this case four out of 12).

Solution structure of a phage-derived peptide antagonist in complex with vascular endothelial growth factor

Vascular endothelial growth factor (VEGF) is a potent endothelial cell-specific mediator of angiogenesis and vasculogenesis. VEGF is involved pathologically in cancer, proliferative retinopathy and rheumatoid arthritis, and as such represents an important therapeutic target. Three classes of disulfide-constrained peptides that antagonize binding of the VEGF dimer to its receptors, KDR and Flt-1, were identified previously using phage display methods. NMR studies of a representative peptide from the most potent class of these peptide antagonists, v107 (GGNECDAIRMWEWECFERL), were undertaken to characterize its interactions with VEGF. v107 has no defined structure free in solution, but binding to VEGF induces folding of the peptide. The solution structure of the VEGF receptor-binding domain-v107 complex was determined using 3940 (1970 per VEGF monomer) internuclear distance and 476 (238 per VEGF monomer) dihedral angle restraints derived from NMR data obtained using samples containing either (13)C/(15)N-labeled protein plus excess unlabeled peptide or (13)C/(15)N-labeled peptide plus excess unlabeled protein. Residual dipolar coupling restraints supplemented the structure determination of the complex and were found to increase significantly both the global precision of VEGF in the complex and the agreement with available crystal structures of VEGF. The calculated ensemble of structures is of high precision and is in excellent agreement with the experimental restraints. v107 has a turn-helix conformation with hydrophobic residues partitioned to one face of the peptide and polar or charged residues at the other face. Contacts between two v107 peptides and the VEGF dimer are mediated by primarily hydrophobic side-chain interactions. The v107-binding site on VEGF overlaps partially with the binding site of KDR and is similar to that for domain 2 of Flt-1. The structure of the VEGF-v107 complex provides new insight into how binding to VEGF can be achieved that may be useful for the design of small molecule antagonists.

Protein-protein interfaces: mimics and inhibitors

Many biological processes are mediated by specific molecular recognition between proteins. However, the thermodynamic and structural 'rules' for such recognition are incompletely understood, as is the potential for inhibition by small molecules. Recent progress has included the discovery of small-molecule inhibitors for several targets important in cancer.

Tryptophan zippers: stable, monomeric beta -hairpins

A structural motif, the tryptophan zipper (trpzip), greatly stabilizes the beta-hairpin conformation in short peptides. Peptides (12 or 16 aa in length) with four different turn sequences are monomeric and fold cooperatively in water, as has been observed previously for some hairpin peptides. However, the folding free energies of the trpzips exceed substantially those of all previously reported beta-hairpins and even those of some larger designed proteins. NMR structures of three of the trpzip peptides reveal exceptionally well-defined beta-hairpin conformations stabilized by cross-strand pairs of indole rings. The trpzips are the smallest peptides to adopt an unique tertiary fold without requiring metal binding, unusual amino acids, or disulfide crosslinks.

A minimal peptide scaffold for beta-turn display: optimizing a strand position in disulfide-cyclized beta-hairpins

Phage display of peptide libraries has become a powerful tool for the evolution of novel ligands that bind virtually any protein target. However, the rules governing conformational preferences in natural peptides are poorly understood, and consequently, structure-activity relationships in these molecules can be difficult to define. In an effort to simplify this process, we have investigated the structural stability of 10-residue, disulfide-constrained beta-hairpins and assessed their suitability as scaffolds for beta-turn display. Using disulfide formation as a probe, relative free energies of folding were measured for 19 peptides that differ at a one strand position. A tryptophan substitution promotes folding to a remarkable degree. NMR analysis confirms that the measured energies correlate well with the degree of beta-hairpin structure in the disulfide-cyclized peptides. Reexamination of a subset of the strand substitutions in peptides with different turn sequences reveals linear free energy relationships, indicating that turns and strand-strand interactions make independent, additive contributions to hairpin stability. Significantly, the tryptophan strand substitution is highly stabilizing with all turns tested, and peptides that display model turns or the less stable C'-C' ' turn of CD4 on this tryptophan "stem" are highly structured beta-hairpins in water. Thus, we have developed a small, structured beta-turn scaffold, containing only natural L-amino acids, that may be used to display peptide libraries of limited conformational diversity on phage.

Antagonists of protein-protein interactions

Protein-protein interactions are often attractive, but not straightforward, targets for disease therapy. Two strategies for identifying inhibitors of these interactions, peptide phage display and high-throughput screening, have recently shown new promise.

Novel peptides selected to bind vascular endothelial growth factor target the receptor-binding site

Peptides that inhibit binding of vascular endothelial growth factor (VEGF) to its receptors, KDR and Flt-1, have been produced using phage display. Libraries of short disulfide-constrained peptides yielded three distinct classes of peptides that bind to the receptor-binding domain of VEGF with micromolar affinities. The highest affinity peptide was also shown to antagonize VEGF-induced proliferation of primary human umbilical vascular endothelial cells. The peptides bind to a region of VEGF known to contain the contact surface for Flt-1 and the functional determinants for KDR binding. This suggests that the receptor-binding region of VEGF is a binding "hot spot" that is readily targeted by selected peptides and supports earlier assertions that phage-derived peptides frequently target protein-protein interaction sites. Such peptides may lead to the development of pharmacologically useful VEGF antagonists.

Crystal structure of the complex between VEGF and a receptor-blocking peptide

Vascular endothelial growth factor (VEGF) is a specific and potent angiogenic factor and, therefore, a prime therapeutic target for the development of antagonists for the treatment of cancer. As a first step toward this goal, phage display was used to generate peptides that bind to the receptor-binding domain (residues 8-109) of VEGF and compete with receptor [Fairbrother, W. J., Christinger, H. W., Cochran, A. G., Fuh, G., Keenan, C. J., Quan, C., Shriver, S. K., Tom, J. Y. K., Wells, J. A., and Cunningham, B. C. (1999) Biochemistry 38, 17754-17764]. The crystal structure of VEGF in complex with one of these peptides was solved and refined to a resolution of 1.9 A. The 20-mer peptide is unstructured in solution and adopts a largely extended conformation when bound to VEGF. Residues 3-8 form a beta-strand which pairs with strand beta6 of VEGF via six hydrogen bonds. The C-terminal four residues of the peptide point away from the growth factor, consistent with NMR data indicating that these residues are flexible in the complex in solution. In contrast, shortening the N-terminus of the peptide leads to decreased binding affinities. Truncation studies show that the peptide can be reduced to 14 residues with only moderate effect on binding affinity. However, because of the extended conformation and the scarcity of specific side-chain interactions with VEGF, the peptide is not a promising lead for small-molecule development. The interface between the peptide and VEGF contains a subset of the residues recognized by a neutralizing Fab fragment and overlaps partially with the binding site for the Flt-1 receptor. The location of the peptide-binding site and the hydrophilic character of the interactions with VEGF resemble more the binding mode of the Fab fragment than that of the receptor.

Computational learning reveals coiled coil-like motifs in histidine kinase linker domains

The recent rapid growth of protein sequence databases is outpacing the capacity of researchers to biochemically and structurally characterize new proteins. Accordingly, new methods for recognition of motifs and homologies in protein primary sequences may be useful in determining how these proteins might function. We have applied such a method, an iterative learning algorithm, to analyze possible coiled coil domains in histidine kinase receptors. The potential coiled coils have not yet been structurally characterized in any histidine kinase, and they appear outside previously noted kinase homology regions. The learning algorithm uses a combination of established sequence patterns in known coiled coil proteins and histidine kinase sequence data to learn to recognize efficiently this coiled coil-like motif in the histidine kinases. The common appearance of the structural motif in a functionally important part of the receptors suggests hypotheses for kinase regulation and signal transduction.

Imitation of Escherichia coli aspartate receptor signaling in engineered dimers of the cytoplasmic domain

Transmembrane signaling by bacterial chemotaxis receptors appears to require a conformational change within a receptor dimer. Dimers were engineered of the cytoplasmic domain of the Escherichia coli aspartate receptor that stimulated the kinase CheA in vitro. The folding free energy of the leucine-zipper dimerization domain was harnessed to twist the dimer interface of the receptor, which markedly affected the extent of CheA activation. Response to this twist was attenuated by modification of receptor regulatory sites, in the same manner as adaptation resets sensitivity to ligand in vivo. These results suggest that the normal allosteric activation of the chemotaxis receptor has been mimicked in a system that lacks both ligand-binding and transmembrane domains. The most stimulatory receptor dimer formed a species of tetrameric size.

Antibody-catalyzed porphyrin metallation

An antibody elicited to a distorted N-methyl porphyrin catalyzed metal ion chelation by the planar porphyrin. At fixed Zn2+ and Cu2+ concentrations, the antibody-catalyzed reaction showed saturation kinetics with respect to the substrate mesoporphyrin IX (2) and was inhibited by the hapten, N-methylmesoporphyrin IX (1). The turnover number of 80 hour-1 for antibody-catalyzed metallation of 2 with Zn2+ compares with an estimated value of 800 hour-1 for ferrochelatase. The antibody also catalyzed the insertion of Co2+ and Mn2+ into 2, but it did not catalyze the metallation of protoporphyrin IX (3) or deuteroporphyrin IX (4). The antibody has high affinity for several metalloporphyrins, suggesting an approach to developing antibody-heme catalysts for redox or electron transfer reactions.