A novel type of receptor protein, based on the lipocalin scaffold, with specificity for digoxigenin

Elarekibep (PRS-060/AZD1402), a new class of inhaled Anticalin medicine targeting IL-4Ra for type 2 endotype asthma

BACKGROUND

Type 2 endotype asthma is driven by IL-4 and IL-13 signaling via IL-4Ra, which is highly expressed on airway epithelium, airway smooth muscle, and immunocytes in the respiratory mucosa, suggesting potential advantages of an inhalable antagonist. Lipocalin 1 (Lcn1), a 16 kDa protein abundant in human periciliary fluid, has a robust drug-like structure well suited to protein engineering, but it has never been used to make an inhaled Anticalin protein therapeutic.

OBJECTIVES

We sought to reengineer Lcn1 into an inhalable IL-4Ra antagonist and assess its pharmacodynamic/kinetic profile.

METHODS

Lcn1 was systematically modified by directed protein mutagenesis yielding a high-affinity, slowly dissociating, long-acting full antagonist of IL-4Ra designated PRS-060 with properties analogous to dupilumab, competitively antagonizing IL-4Ra-dependent cell proliferation, mucus induction, and eotaxin expression in vitro. Because PRS-060 displayed exquisite specificity for human IL-4Ra, with no cross-reactivity to rodents or higher primates, we created a new triple-humanized mouse model substituting human IL-4Ra, IL-4, and IL-13 at their correct syntenic murine loci to model clinical dosing.

RESULTS

Inhaled PRS-060 strongly suppressed acute allergic inflammation indexes in triple-humanized mice with a duration of action longer than its bulk clearance, suggesting that it may act locally in the lung.

CONCLUSION

Lcn1 can be reengineered into the Anticalin antagonist PRS-060 (elarekibep), exemplifying a new class of inhaled topical, long-acting therapeutic drugs with the potential to treat type 2 endotype asthma.

Pharmacokinetic Engineering of OX40-Blocking Anticalin Proteins Using Monomeric Plasma Half-Life Extension Domains

Anticalin^® proteins have been proven as versatile clinical stage biotherapeutics. Due to their small size (∼20 kDa), they harbor a short intrinsic plasma half-life which can be extended, e.g., by fusion with IgG or Fc. However, for antagonism of co-immunostimulatory Tumor Necrosis Factor Receptor Superfamily (TNFRSF) members in therapy of autoimmune and inflammatory diseases, a monovalent, pharmacokinetically optimized Anticalin protein format that avoids receptor clustering and therefore potential activation is favored. We investigated the suitability of an affinity-improved streptococcal Albumin-Binding Domain (ABD) and the engineered Fab-selective Immunoglobulin-Binding Domain (IgBD) SpGC3Fab for plasma Half-Life Extension (HLE) of an OX40-specific Anticalin and bispecific Duocalin proteins, neutralizing OX40 and a second co-immunostimulatory TNFRSF member. The higher affinity of ABD fusion proteins to human serum albumin (HSA) and Mouse Serum Albumin (MSA), with a 4 to 5-order of magnitude lower K_D compared with the binding affinity of IgBD fusions to human/mouse IgG, translated into longer terminal plasma half-lives (t_1/2). Hence, the anti-OX40 Anticalin-ABD protein reached t_1/2 values of ∼40 h in wild-type mice and 110 h in hSA/hFcRn double humanized mice, in contrast to ∼7 h observed for anti-OX40 Anticalin-IgBD in wild-type mice. The pharmacokinetics of an anti-OX40 Anticalin-Fc fusion protein was the longest in both models (t_1/2 of 130 h and 146 h, respectively). Protein formats composed of two ABDs or IgBDs instead of one single HLE domain clearly showed longer presence in the circulation. Importantly, Anticalin-ABD and -IgBD fusions showed OX40 receptor binding and functional competition with OX40L-induced cellular reactivity in the presence of albumin or IgG, respectively. Our results suggest that fusion to ABD or IgBD can be a versatile platform to tune the plasma half-life of Anticalin proteins in response to therapeutic needs.

A sweet protein monellin as a non-antibody scaffold for synthetic binding proteins

Synthetic binding proteins that have the ability to bind with molecules can be generated using various protein domains as non-antibody scaffolds. These designer proteins have been used widely in research studies, as their properties overcome the disadvantages of using antibodies. Here, we describe the first application of a phage display to generate synthetic binding proteins using a sweet protein, monellin, as a non-antibody scaffold. Single-chain monellin (scMonellin), in which two polypeptide chains of natural monellin are connected by a short linker, has two loops on one side of the molecule. We constructed phage display libraries of scMonellin, in which the amino acid sequence of the two loops is diversified. To validate the performance of these libraries, we sorted them against the folding mutant of the green fluorescent protein variant (GFPuv) and yeast small ubiquitin-related modifier. We successfully obtained scMonellin variants exhibiting moderate but significant affinities for these target proteins. Crystal structures of one of the GFPuv-binding variants in complex with GFPuv revealed that the two diversified loops were involved in target recognition. scMonellin, therefore, represents a promising non-antibody scaffold in the design and generation of synthetic binding proteins. We termed the scMonellin-derived synthetic binding proteins 'SWEEPins'.

Engineered Protein Scaffolds as Next-Generation Therapeutics

The concept of engineering robust protein scaffolds for novel binding functions emerged 20 years ago, one decade after the advent of recombinant antibody technology. Early examples were the Affibody, Monobody (Adnectin), and Anticalin proteins, which were derived from fragments of streptococcal protein A, from the tenth type III domain of human fibronectin, and from natural lipocalin proteins, respectively. Since then, this concept has expanded considerably, including many other protein templates. In fact, engineered protein scaffolds with useful binding specificities, mostly directed against targets of biomedical relevance, constitute an area of active research today, which has yielded versatile reagents as laboratory tools. However, despite strong interest from basic science, only a handful of those protein scaffolds have undergone biopharmaceutical development up to the clinical stage. This includes the abovementioned pioneering examples as well as designed ankyrin repeat proteins (DARPins). Here we review the current state and clinical validation of these next-generation therapeutics.

Anticalin® proteins: from bench to bedside

INTRODUCTION

Anticalin proteins are engineered versions of lipocalins that constitute a novel class of clinical-stage biopharmaceuticals. The lipocalins exhibit a central β-barrel with eight antiparallel β-strands and an α-helix attached to its side. Four structurally variable loops at the open end of the β-barrel form a pronounced binding pocket, which can be reshaped to generate specificities toward diverse disease-relevant molecular targets.

AREAS COVERED

This article reviews the current status of Anticalin engineering, from the basic principles to the development of Anticalins with high target affinity and specificity via combinatorial protein design and directed evolution, including examples of Anticalin-based drug candidates under preclinical and clinical development.

EXPERT OPINION

Combinatorial gene libraries together with powerful molecular selection techniques have enabled the expansion of the natural ligand specificities of lipocalins from small molecules to peptides and proteins. This biomolecular concept has been validated by structural analyses of a series of Anticalin•target complexes. Promising Anticalin lead candidates have reached different preclinical and clinical development stages in the areas of (immuno)oncology, metabolic, and respiratory diseases, as antidotes to treat intoxications and as novel antibiotics. Thus, Anticalins offer an alternative to antibodies with promising and potentially superior features as next-generation biologics.

Combinatorial mutagenesis with alternative CDR-L1 and -H2 loop lengths contributes to affinity maturation of antibodies

Loop length variation in the complementary determining regions (CDRs) 1 and 2 encoded in germline variable antibody genes provides structural diversity in naïve antibody libraries. In synthetic single framework libraries the parental CDR-1 and CDR-2 length is typically unchanged and alternative lengths are provided only at CDR-3 sites. Based on an analysis of the germline repertoire and structure-solved anti-hapten and anti-peptide antibodies, we introduced combinatorial diversity with alternative loop lengths into the CDR-L1, CDR-L3 and CDR-H2 loops of anti-digoxigenin and anti-microcystin-LR single chain Fv fragments (scFvs) sharing human IGKV3-20/IGHV3-23 frameworks. The libraries were phage display selected for folding and affinity, and analysed by single clone screening and deep sequencing. Among microcystin-LR binders the most frequently encountered alternative loop lengths were one amino acid shorter (6 aa) and four amino acids longer (11 aa) CDR-L1 loops leading up to 17- and 28-fold improved affinity, respectively. Among digoxigenin binders, 2 amino acids longer (10 aa) CDR-H2 loops were strongly enriched, but affinity improved anti-digoxigenin scFvs were also encountered with 7 aa CDR-H2 and 11 aa CDR-L1 loops. Despite the fact that CDR-L3 loop length variants were not specifically enriched in selections, one clone with 22-fold improved digoxigenin binding affinity was identified containing a 2 residues longer (10 aa) CDR-L3 loop. Based on our results the IGKV3-20/IGHV3-23 scaffold tolerates loop length variation, particularly in CDR-L1 and CDR-H2 loops, without compromising antibody stability, laying the foundation for developing novel synthetic antibody libraries with loop length combinations not existing in the natural human Ig gene repertoire.

Clonal array profiling of scFv-displaying phages for high-throughput discovery of affinity-matured antibody mutants

"Antibody-breeding" approach potentially generates therapeutic/diagnostic antibody mutants with greater performance than native antibodies. Therein, antibody fragments (e.g., single-chain Fv fragments; scFvs) with a variety of mutations are displayed on bacteriophage to generate diverse phage-antibody libraries. Rare clones with improved functions are then selected via panning against immobilized or tagged target antigens. However, this selection process often ended unsuccessful, mainly due to the biased propagation of phage-antibody clones and the competition with a large excess of undesirable clones with weaker affinities. To break radically from such panning-inherent problems, we developed a novel method, clonal array profiling of scFv-displaying phages (CAP), in which colonies of the initial bacterial libraries are examined one-by-one in microwells. Progenies of scFv-displaying phages generated are, if show sufficient affinity to target antigen, captured in the microwell via pre-coated antigen and detected using a luciferase-fused anti-phage scFv. The advantage of CAP was evidenced by its application with a small error-prone-PCR-based library (~ 10⁵ colonies) of anti-cortisol scFvs. Only two operations, each surveying only ~ 3% of the library (9,400 colonies), provided five mutants showing 32–63-fold improved K_a values (> 10¹⁰ M⁻¹), compared with the wild-type scFv (K_a = 3.8 × 10⁸ M⁻¹), none of which could be recovered via conventional panning procedures operated for the entire library.

An engineered lipocalin that tightly complexes the plant poison colchicine for use as antidote and in bioanalytical applications

Colchicine is a toxic alkaloid prevalent in autumn crocus (Colchicum autumnale) that binds to tubulin and inhibits polymerization of microtubules. Using combinatorial and rational protein design, we have developed an artificial binding protein based on the human lipocalin 2 that binds colchicine with a dissociation constant of 120 pm, i.e. 10000-fold stronger than tubulin. Crystallographic analysis of the engineered lipocalin, dubbed Colchicalin, revealed major structural changes in the flexible loop region that forms the ligand pocket at the open end of the eight-stranded β-barrel, resulting in a lid-like structure over the deeply buried colchicine. A cis-peptide bond between residues Phe71 and Pro72 in loop #2 constitutes a peculiar feature and allows intimate contact with the tricyclic ligand. Using directed evolution, we achieved an extraordinary dissociation half-life of more than 9 h for the Colchicalin-colchicine complex. Together with the chemical robustness of colchicine and availability of activated derivatives, this also opens applications as a general-purpose affinity reagent, including facile quantification of colchicine in biological samples. Given that engineered lipocalins, also known as Anticalin® proteins, represent a class of clinically validated biopharmaceuticals, Colchicalin may offer a therapeutic antidote to scavenge colchicine and reverse its poisoning effect in situations of acute intoxication.

Sustained plasma hepcidin suppression and iron elevation by Anticalin-derived hepcidin antagonist in cynomolgus monkey

BACKGROUND AND PURPOSE

Anaemia of chronic disease (ACD) has been linked to iron-restricted erythropoiesis imposed by high circulating levels of hepcidin, a 25 amino acid hepatocyte-derived peptide that controls systemic iron homeostasis. Here, we report the engineering of the human lipocalin-derived, small protein-based anticalin PRS-080 hepcidin antagonist with high affinity and selectivity.

EXPERIMENTAL APPROACH

Anticalin- and hepcidin-specific pharmacokinetic (PK)/pharmacodynamic modelling (PD) was used to design and select the suitable drug candidate based on t_1/2 extension and duration of hepcidin suppression. The development of a novel free hepcidin assay enabled accurate analysis of bioactive hepcidin suppression and elucidation of the observed plasma iron levels after PRS-080-PEG30 administration in vivo.

KEY RESULTS

PRS-080 had a hepcidin-binding affinity of 0.07 nM and, after coupling to 30 kD PEG (PRS-080-PEG30), a t_1/2 of 43 h in cynomolgus monkeys. Dose-dependent iron mobilization and hepcidin suppression were observed after a single i.v. dose of PRS-080-PEG30 in cynomolgus monkeys. Importantly, in these animals, suppression of free hepcidin and subsequent plasma iron elevation were sustained during repeated s.c. dosing. After repeated dosing and followed by a treatment-free interval, all iron parameters returned to pre-dose values.

CONCLUSIONS AND IMPLICATIONS

In conclusion, we developed a dose-dependent and safe approach for the direct suppression of hepcidin, resulting in prolonged iron mobilization to alleviate iron-restricted erythropoiesis that can address the root cause of ACD. PRS-080-PEG30 is currently in early clinical development.

Protein labeling for live cell fluorescence microscopy with a highly photostable renewable signal

A novel method of protein labeling uses the highly dynamic reversible association of a cell-permeable fluorogenic dye and lipocalin Blc mutants.

We present protein-PAINT – the implementation of the general principles of PAINT (Point Accumulation for Imaging in Nanoscale Topography) for live-cell protein labeling. Our method employs the specific binding of cell-permeable fluorogenic dyes to genetically encoded protein tags. We engineered three mutants of the bacterial lipocalin Blc that possess different affinities to a fluorogenic dye and exhibit a strong increase in fluorescence intensity upon binding. This allows for rapid labeling and washout of intracellular targets on a time scale from seconds to a few minutes. We demonstrate an order of magnitude higher photostability of the fluorescence signal in comparison with spectrally similar fluorescent proteins. Protein-PAINT ensures prolonged super-resolution fluorescence microscopy of living cells in both single molecule detection and stimulated emission depletion regimes.

Affimer proteins are versatile and renewable affinity reagents

Molecular recognition reagents are key tools for understanding biological processes and are used universally by scientists to study protein expression, localisation and interactions. Antibodies remain the most widely used of such reagents and many show excellent performance, although some are poorly characterised or have stability or batch variability issues, supporting the use of alternative binding proteins as complementary reagents for many applications. Here we report on the use of Affimer proteins as research reagents. We selected 12 diverse molecular targets for Affimer selection to exemplify their use in common molecular and cellular applications including the (a) selection against various target molecules; (b) modulation of protein function in vitro and in vivo; (c) labelling of tumour antigens in mouse models; and (d) use in affinity fluorescence and super-resolution microscopy. This work shows that Affimer proteins, as is the case for other alternative binding scaffolds, represent complementary affinity reagents to antibodies for various molecular and cell biology applications.

Monobodies and other synthetic binding proteins for expanding protein science

Synthetic binding proteins are constructed using nonantibody molecular scaffolds. Over the last two decades, in‐depth structural and functional analyses of synthetic binding proteins have improved combinatorial library designs and selection strategies, which have resulted in potent platforms that consistently generate binding proteins to diverse targets with affinity and specificity that rival those of antibodies. Favorable attributes of synthetic binding proteins, such as small size, freedom from disulfide bond formation and ease of making fusion proteins, have enabled their unique applications in protein science, cell biology and beyond. Here, we review recent studies that illustrate how synthetic binding proteins are powerful probes that can directly link structure and function, often leading to new mechanistic insights. We propose that synthetic proteins will become powerful standard tools in diverse areas of protein science, biotechnology and medicine.

High-affinity Anticalins with aggregation-blocking activity directed against the Alzheimer β-amyloid peptide

Anticalins engineered for high affinity and specificity towards the central VFFAED epitope in Aβ peptides potently inhibit their aggregation, thus providing novel reagents to study the molecular pathology of Alzheimer's disease (AD) and alternative drug candidates compared with current biopharmaceutical treatments.

Amyloid beta (Aβ) peptides, in particular Aβ42 and Aβ40, exert neurotoxic effects and their overproduction leads to amyloid deposits in the brain, thus constituting an important biomolecular target for treatments of Alzheimer's disease (AD). We describe the engineering of cognate Anticalins as a novel type of neutralizing protein reagent based on the human lipocalin scaffold. Phage display selection from a genetic random library comprising variants of the human lipocalin 2 (Lcn2) with mutations targeted at 20 exposed amino acid positions in the four loops that form the natural binding site was performed using both recombinant and synthetic target peptides and resulted in three different Anticalins. Biochemical characterization of the purified proteins produced by periplasmic secretion in Escherichia coli revealed high folding stability in a monomeric state, with T_m values ranging from 53.4°C to 74.5°C, as well as high affinities for Aβ40, between 95 pM and 563 pM, as measured by real-time surface plasmon resonance analysis. The central linear VFFAED epitope within the Aβ sequence was mapped using a synthetic peptide array on membranes and was shared by all three Anticalins, despite up to 13 mutual amino acid differences in their binding sites. All Anticalins had the ability–with varying extent–to inhibit Aβ aggregation in vitro according to the thioflavin-T fluorescence assay and, furthermore, they abolished Aβ42-mediated toxicity in neuronal cell culture. Thus, these Anticalins provide not only useful protein reagents to study the molecular pathology of AD but they also show potential as alternative drug candidates compared with antibodies.

Improving Binding Affinity and Selectivity of Computationally Designed Ligand-Binding Proteins Using Experiments

The ability to de novo design proteins that can bind small molecules has wide implications for synthetic biology and medicine. Combining computational protein design with the high-throughput screening of mutagenic libraries of computationally designed proteins is emerging as a general approach for creating binding proteins with programmable binding modes, affinities, and selectivities. The computational step enables the creation of a binding site in a protein that otherwise does not (measurably) bind the intended ligand, and targeted mutagenic screening allows for validation and refinement of the computational model as well as provides orders-of-magnitude increases in the binding affinity. Deep sequencing of mutagenic libraries can provide insights into the mutagenic binding landscape and enable further affinity improvements. Moreover, in such a combined computational-experimental approach where the binding mode is preprogrammed and iteratively refined, selectivity can be achieved (and modulated) by the placement of specified amino acid side chain groups around the ligand in defined orientations. Here, we describe the experimental aspects of a combined computational-experimental approach for designing-using the software suite Rosetta-proteins that bind a small molecule of choice and engineering, using fluorescence-activated cell sorting and high-throughput yeast surface display, high affinity and ligand selectivity. We illustrated the utility of this approach by performing the design of a selective digoxigenin (DIG)-binding protein that, after affinity maturation, binds DIG with picomolar affinity and high selectivity over structurally related steroids.

A high-performance thioredoxin-based scaffold for peptide immunogen construction: proof-of-concept testing with a human papillomavirus epitope

Escherichia coli thioredoxin has been previously exploited as a scaffold for the presentation/stabilization of peptide aptamers as well as to confer immunogenicity to peptide epitopes. Here we focused on other key features of thioredoxin that are of general interest for the production of safer and more effective peptide immunogens, such as a high thermal stability, lack of cross-reactivity and a low-cost of production. We identified thioredoxin from the archaebacterium Pyrococcus furiosus (PfTrx) as a novel scaffold meeting all the above criteria. PfTrx is a highly thermostable and protease-resistant scaffold with a strong (poly)peptide solubilisation capacity. Anti-PfTrx antibodies did not cross-react with mouse, nor human thioredoxin. Untagged PfTrx bearing a previously identified HPV16-L2 peptide epitope was obtained in a >90% pure form with a one-step thermal purification procedure and effectively elicited the production of neutralizing anti-HPV antibodies. We thus propose PfTrx as a superior, general-purpose scaffold for the construction of safe, stable, and low-cost peptide immunogens.

AXM mutagenesis: an efficient means for the production of libraries for directed evolution of proteins

Affinity maturation is an important part of the recombinant antibody development process. There are several well-established approaches for generating libraries of mutated antibody genes for affinity maturation, but these approaches are generally too laborious or expensive to allow high-throughput, parallel processing of multiple antibodies. Here, we describe a scalable approach that enables the generation of libraries with greater than 10(8) clones from a single Escherichia coli transformation. In our method, a mutated DNA fragment is produced using PCR conditions that promote nucleotide misincorporation into newly synthesized DNA. In the PCR reaction, one of the primers contains at least three phosphorothioate linkages at its 5' end, and treatment of the PCR product with a 5' to 3' exonuclease is used to preferentially remove the strand synthesized with the non-modified primer, resulting in a single-stranded DNA fragment. This fragment then serves as a megaprimer to prime DNA synthesis on a uracilated, circular, single-stranded template in a Kunkel-like mutagenesis reaction that biases nucleotide base-changes between the megaprimer and uracilated DNA sequence in favor of the in vitro synthesized megaprimer. This method eliminates the inefficient subcloning steps that are normally required for the construction of affinity maturation libraries from randomly mutagenized antibody genes.

Steered molecular dynamics simulations of ligand-receptor interaction in lipocalins

Retinol binding protein (RBP) and an engineered lipocalin, DigA16, have been studied using molecular dynamics simulations. Special emphasis has been placed on explaining the ligand-receptor interaction in RBP-retinol and DigA16-digoxigenin complexes, and steered molecular dynamics simulations of 10-20 ns have been carried out for the ligand expulsion process. Digoxigenin is bound deep inside the cavity of DigA16 and forms several stable hydrogen bonds in addition to the hydrophobic van der Waals interaction with the aromatic side-chains. Four crystalline water molecules inside the ligand-binding cavity remain trapped during the simulations. The strongly hydrophobic receptor site of RBP differs considerably from DigA16, and the main source of ligand attraction comes from the phenyl side-chains. The hydrogen bonds between digoxigenin and DigA16 cause the rupture forces on ligand removal in DigA16 and RBP to differ. The mutated DigA16 residues contribute approximately one-half of the digoxigenin interaction energy with DigA16 and, of these, the energetically most important are residues His35, Arg58, Ser87, Tyr88, and Phe114. Potential "sensor loops" were found for both receptors. These are the outlier loops between residues 114-121 and 63-67 for DigA16 and RBP, respectively, and they are located near the entrance of the ligand-binding cavity. Especially, the residues Glu119 (DigA16) and Leu64 (RBP) are critical for sensing. The ligand binding energies have been estimated based on the linear response approximation of binding affinity by using a previous parametrization for retinoids and RBP.

Anticalins: the lipocalin family as a novel protein scaffold for the development of next-generation immunotherapies

Anticalins are engineered ligand-binding proteins based on the human lipocalin scaffold. Their architecture is characterized by a rigid beta-barrel that supports four structurally hypervariable loops. Similar to antibodies, these loops form the natural ligand-binding site, usually for vitamins, hormones or secondary metabolites. Anticalins with novel specificities can be engineered by reshaping this loop region, using targeted random mutagenesis in combination with functional display and guided selection. Several drug candidates with specificities for exogenous low-molecular-weight substances, peptides and even protein targets (e.g., several disease-related cell surface receptors) have been obtained in this way. Owing to their exquisite specificity and high affinity, Anticalins are particularly attractive as antagonists for the manipulation of immune mechanisms, leading to either inhibitory or stimulatory effects. Compared with antibodies, Anticalins offer several practical advantages as they are much smaller, consist of a single polypeptide chain and can be produced easily in microbial expression systems.

NMR structure and dynamics of the engineered fluorescein-binding lipocalin FluA reveal rigidification of beta-barrel and variable loops upon enthalpy-driven ligand binding

The NMR structure of the 21 kDa lipocalin FluA, which was previously obtained by combinatorial design, elucidates a reshaped binding site specific for the dye fluorescein resulting from 21 side chain replacements with respect to the parental lipocalin, the naturally occurring bilin-binding protein (BBP). As expected, FluA exhibits the lipocalin fold of BBP, comprising eight antiparallel beta-strands forming a beta-barrel with an alpha-helix attached to its side. Comparison of the NMR structure of free FluA with the X-ray structures of BBP.biliverdin IX(gamma) and FluA.fluorescein complexes revealed significant conformational changes in the binding pocket, which is formed by four loops at the open end of the beta-barrel as well as adjoining beta-strand segments. An "induced fit" became apparent for the side chain conformations of Arg 88 and Phe 99, which contact the bound fluorescein in the complex and undergo concerted rearrangement upon ligand binding. Moreover, slower internal motional modes of the polypeptide backbone were identified by measuring transverse (15)N backbone spin relaxation times in the rotating frame for free FluA and also for the FluA.fluorescein complex. A reduction in the level of such motions was detected upon complex formation, indicating rigidification of the protein structure and loss of conformational entropy. This hypothesis was confirmed by isothermal titration calorimetry, showing that ligand binding is enthalpy-driven, thus overcompensating for the negative entropy associated with both ligand binding per se and rigidification of the protein. Our investigation of the solution structure and dynamics as well as thermodynamics of lipocalin-ligand interaction not only provides insight into the general mechanism of small molecule accommodation in the deep and narrow cavity of this abundant class of proteins but also supports the future design of corresponding binding proteins with novel specificities, so-called "anticalins".

Development of GFP-based biosensors possessing the binding properties of antibodies

Proteins that can bind specifically to targets that also have an intrinsic property allowing for easy detection could facilitate a multitude of applications. While the widely used green fluorescent protein (GFP) allows for easy detection, attempts to insert multiple binding loops into GFP to impart affinity for a specific target have been met with limited success because of the structural sensitivity of the GFP chromophore. In this study, directed evolution using a surrogate loop approach and yeast surface display yielded a family of GFP scaffolds capable of accommodating 2 proximal, randomized binding loops. The library of potential GFP-based binders or ''GFAbs'' was subsequently mined for GFAbs capable of binding to protein targets. Identified GFAbs bound with nanomolar affinity and required binding contributions from both loops indicating the advantage of a dual loop GFAb platform. Finally, GFAbs were solubly produced and used as fluorescence detection reagents to demonstrate their utility.

High-affinity recognition of lanthanide(III) chelate complexes by a reprogrammed human lipocalin 2

Human lipocalin 2 (Lcn2), also known as neutrophil gelatinase-associated lipocalin (NGAL), which naturally scavenges bacterial ferric siderophores, has been engineered to specifically bind rare-earth and related metal ions as chelate complexes with [(R)-2-amino-3-(4-aminophenyl)propyl]-trans-(S,S)-cyclohexane-1,2-diaminepentaacetic acid (p-NH(2)-Bn-CHX-A''-DTPA). To this end, 12 amino acid residues in the ligand pocket of Lcn2, which is formed by four loops at the open end of an eight-stranded beta-barrel, were subjected to targeted random mutagenesis, and from the resulting library, variants with binding activity for the Me x DTPA group were selected using the method of bacterial phage display. One promising candidate was further developed in several cycles of in vitro affinity maturation using partial random mutagenesis and selection (via phage display and/or Escherichia coli colony screening) under conditions of increasing stringency. As result, an Lcn2 variant was obtained that binds Y x DTPA with a dissociation constant as low as 400 pM. The Lcn2 variant specifically recognizes the artificial ligand, as exemplified in (competitive) ELISA and real-time surface plasmon resonance analyses. DTPA-complexed Y(3+), Tb(3+), Gd(3+), and Lu(3+) are most tightly bound, comprising metal ions whose isotopes are in common use for radiotherapy and imaging. All of the Lcn2 variants are stably folded and can be functionally produced in high yield in E. coli. X-ray crystallographic analyses show that the new ligand is well-accommodated in the central cavity of the engineered lipocalin, whose fold is largely preserved, but that the mode of binding differs from the one seen with the natural ligand Fe x enterobactin. This structural study reveals analogies but also differences with respect to previously described antibody-metal chelate complexes. Notably, the functionalized side chain of DTPA protrudes from the ligand pocket of the lipocalin in such a way that its conjugates (with other haptens, for example) are recognized too. With their small sizes and robust fold based on a single polypeptide chain, the engineered Lcn2 variants provide novel modules and/or fusion partners for radionuclide-chelate capturing strategies that bear promise for medical diagnostics and therapy.

An engineered lipocalin specific for CTLA-4 reveals a combining site with structural and conformational features similar to antibodies

Biomolecular reagents that enable the specific molecular recognition of proteins play a crucial role in basic research as well as medicine. Up to now, antibodies (immunoglobulins) have been widely used for this purpose. Their predominant feature is the vast repertoire of antigen-binding sites that arise from a set of 6 hypervariable loops. However, antibodies suffer from practical disadvantages because of their complicated architecture, large size, and multiple functions. The lipocalins, on the other hand, have evolved as a protein family that primarily serves for the binding of small molecules. Here, we show that an engineered lipocalin, derived from human Lcn2, can specifically bind the T cell coreceptor CTLA-4 as a prescribed protein target with subnanomolar affinity. Crystallographic analysis reveals that its reshaped cup-like binding site, which is formed by 4 variable loops, provides perfect structural complementarity with this "antigen." Furthermore, comparison with the crystal structure of the uncomplexed engineered lipocalin indicates a pronounced induced-fit mechanism, a phenomenon so far considered typical for antibodies. By recognizing the same epitope on CTLA-4 that interacts with the counterreceptors B7.1/B7.2 on antigen-presenting cells the engineered Lcn2 exhibits strong, cross-species antagonistic activity, as evidenced by biological effects comparable with a CTLA-4-specific antibody. With its proven stimulatory activity on T cells in vivo, the CTLA-4 blocking lipocalin offers potential for immunotherapy of cancer and infectious disease. Beyond that, lipocalins with engineered antigen-binding sites, so-called Anticalins, provide a class of small ( approximately 180 residues), structurally simple, and robust binding proteins with applications in the life sciences in general.

Functional humanization of an anti-CD16 Fab fragment: obstacles of switching from murine {lambda} to human {lambda} or {kappa} light chains

An alphaCD30xalphaCD16 bispecific monoclonal antibody (MAb) was previously shown to induce remission of Hodgkin's disease refractory to chemo- and radiotherapy through specific activation of natural killer (NK) cells, but the appearance of a human anti-mouse antibody (HAMA) response prevented its use for prolonged therapy. Here, we describe an effort to humanize the Fab arm directed against FcgammaRIII (CD16), which-in context with the previously humanized CD30 Fab fragment-provides the necessary component for the design of a clinically useful bispecific antibody. Thus, the CDRs of the anti-CD16 mouse IgG1/lambda MAb A9 were grafted onto human Ig sequences. In a first attempt, the murine V(lambda) domain was converted to a humanized lambda chain, which led, however, to complete loss of antigen-binding activity and extremely poor folding efficiency upon periplasmic expression in Escherichia coli. Hence, its CDRs were transplanted onto a human kappa light chain in a second attempt, which resulted in a functional recombinant Fab fragment, yet with 100-fold decreased antigen affinity. In the next step, an in vitro affinity maturation was performed, wherein random mutations were introduced into the humanized V(H) and V(kappa) domains through error-prone PCR, followed by a filter sandwich colony screening assay for increased binding activity towards the bacterially produced extracellular CD16 fragment. Finally, an optimized Fab fragment was obtained, which carries nine additional amino acid exchanges and exhibits an affinity that is within a factor of 2 identical to that of the original murine A9 Fab fragment. The resulting humanized Fab fragment was fully functional with respect to binding of the recombinant CD16 antigen in enzyme-linked immunosorbent assay and in cytofluorimetry with CD16-positive granulocytes, thus providing a promising starting point for the preparation of a fully human bispecific antibody that permits the therapeutic recruitment of NK cells.

Using T7 phage display to select GFP-based binders

Filamentous phage do not display cytoplasmic proteins very effectively. As T7 is a cytoplasmic phage, released by cell lysis, it has been prospected as being more efficient for the display of such proteins. Here we investigate this proposition, using a family of GFP-based cytoplasmic proteins that are poorly expressed by traditional phage display. Using two single-molecule detection techniques, fluorescence correlation spectroscopy and anti-bunching, we show that the number of displayed fluorescent proteins ranges from one to three. The GFP derivatives displayed on T7 contain binding loops able to recognize specific targets. By mixing these in a large background of non-binders, these derivatives were used to optimize selection conditions. Using the optimal selection conditions determined in these experiments, we then demonstrated the selection of specific binders from a library of GFP clones containing heavy chain CDR3 antibody binding loops derived from normal donors inserted at a single site. The selected GFP-based binders were successfully used to detect binding without the use of secondary reagents in flow cytometry, fluorescence-linked immunosorbant assays and immunoblotting. These results demonstrate that specific GFP-based affinity reagents, selected from T7-based libraries, can be used in applications in which only the intrinsic fluorescence is used for detection.

Alternative non-antibody scaffolds for molecular recognition

Originally proposed one decade ago, the idea of engineering proteins outside the immunoglobulin family for novel binding functions has evolved as a powerful technology. Several classes of protein scaffolds proved to yield reagents with specificities and affinities in a range that was previously considered unique to antibodies. Such engineered protein scaffolds are usually obtained by designing a random library with mutagenesis focused at a loop region or at an otherwise permissible surface area and by selection of variants against a given target via phage display or related techniques. Whereas a plethora of protein scaffolds has meanwhile been proposed, only few of them were actually demonstrated to yield specificities towards different kinds of targets and to offer practical benefits such as robustness, smaller size, and ease of expression that justify their use as a true alternative to conventional antibodies or their recombinant fragments. Currently, the most promising scaffolds with broader applicability are protein A, the lipocalins, a fibronectin domain, an ankyrin consensus repeat domain, and thioredoxin. Corresponding binding proteins are not only of interest as research reagents or for separation in biotechnology but also as potential biopharmaceuticals, especially in the areas of cancer, autoimmune and infectious diseases as well as for in vivo diagnostics. The medical prospects have boosted high commercial expectations, and many of the promising scaffolds are under development by biotech start-up companies. Although some issues still have to be addressed, for example immunogenicity, effector functions, and plasma half-life in the context of therapeutic use or low-cost high-throughput selection for applications in proteomics research, it has become clear that scaffold-derived binding proteins will play an increasing role in biotechnology and medicine.

Affilin-novel binding molecules based on human gamma-B-crystallin, an all beta-sheet protein

The concept of novel binding proteins as an alternative to antibodies has undergone rapid development and is now ready for practical use in a wide range of applications. Alternative binding proteins, based on suitable scaffolds with desirable properties, are selected from combinatorial libraries in vitro. Here, we describe an approach using a beta-sheet of human gamma-B-crystallin to generate a universal binding site through randomization of eight solvent-exposed amino acid residues selected according to structural and sequence analyses. Specific variants, so-called Affilin, have been isolated from a phage display library against a variety of targets that differ considerably in size and structure. The isolated Affilin variants can be produced in Escherichia coli as soluble proteins and have a high level of thermodynamic stability. The crystal structures of the human wild-type gamma-B-crystallin and a selected Affilin variant have been determined to 1.7 A and 2.0 A resolution, respectively. Comparison of the two molecules indicates that the human gamma-B-crystallin tolerates amino acid exchanges with no major structural change. We conclude that the intrinsically stable and easily expressed gamma-B-crystallin provides a suitable framework for the generation of novel binding molecules.

Epitope mapping of the neuronal growth inhibitor Nogo-A for the Nogo receptor and the cognate monoclonal antibody IN-1 by means of the SPOT technique

Nogo-A is a potent inhibitor of axonal outgrowth in the central nervous system of adult mammals, where it is expressed as a membrane protein on oligodendrocytes and in myelin. Here we describe an attempt to identify linear peptide epitopes in its sequence that are responsible for the interaction either with the Nogo receptor (NgR) or with the neutralizing monoclonal antibody IN-1. Analysis of an array of immobilized overlapping 15 mer peptides covering the entire amino acid sequence of human Nogo-A (1192 residues) revealed a single epitope with prominent binding activity both towards the recombinant NgR and the IN-1 F(ab) fragment. Further truncation and substitution analysis yielded the minimal epitope sequence 'IKxLRRL' (x not equal to P), which occurs within the so-called Nogo66 region (residues 1054-1120) of Nogo-A. The bacterially produced Nogo66 fragment exhibited binding activity both for the recombinant NgR and for the IN-1 F(ab) fragment on the Western blot as well as in ELISA. Unexpectedly, the synthetic epitope peptide and the recombinant Nogo66 showed cross-reactivity with the 8-18C5 F(ab) fragment, which is directed against myelin oligodendrocyte glycoprotein (MOG) as a structurally unrelated target. On the other hand, the recombinant N-terminal domain of Nogo-A (residues 334-966) was shown to specifically interact on the Western blot and in an ELISA with the IN-1 F(ab) fragment but not with the recombinant NgR, which is in agreement with previous results. Hence, our data suggest that there is a distinct binding site for the Nogo receptor in the Nogo66 region of Nogo-A, whereas its interaction with NgR is less specific than anticipated before. Although there probably exists a non-linear epitope for the neutralizing antibody IN-1 in the N-terminal region of Nogo-A, which is likely to be accessible from outside the cell, a previously postulated second binding site for NgR in this region (called Nogo-A-24) remains elusive.

In vitro display technologies reveal novel biopharmaceutics

Display technologies are fundamental to the isolation of specific high-affinity binding proteins for diagnostic and therapeutic applications in cancer, neurodegenerative, and infectious diseases as well as autoimmune and inflammatory disorders. Applications extend into the broad field of antibody (Ab) engineering, synthetic enzymes, proteomics, and cell-free protein synthesis. Recently, in vitro display technologies have come to prominence due to the isolation of high-affinity human antibodies by phage display, the development of novel scaffolds for ribosome display, and the discovery of novel protein-protein interactions. In vitro display represents an emerging and innovative technology for the rapid isolation and evolution of high-affinity peptides and proteins. So far, only one clinical drug candidate produced by in vitro display technology has been approved by the FDA for use in humans, but several are in clinical or preclinical testing. This review highlights recent advances in various engineered biopharmaceutical products isolated by in vitro display with a focus on the commercial developments.

Lipocalins - a family portrait

Lipocalins are a widely distributed group of proteins whose common feature is the presence of six-or eight-stranded beta-barrel in their tertiary structure and highly conservative motifs short conserved region, (SCR) in their amino acid sequences. The presence of three SCRs is typical for kernel lipocalins, while outlier lipocalins have only one or two such regions. Owing to their ability to bind and transport small, hydrophobic molecules, lipocalins participate in the distribution of such substances. However, the physiological significance of lipocalins is not limited to transfer processes. They play an important role in the regulation of immunological and developmental processes, and are also involved in the reactions of organisms to various stress factors and in the pathways of signal transduction. Of special interest is the enzymatic activity found in a few members of the lipocalin family, as well as the interaction with natural membranes, both directly with lipids and through membrane-localized protein receptors.

Directed evolution of ATP binding proteins from a zinc finger domain by using mRNA display

Antibodies have traditionally been used for isolating affinity reagents to new molecular targets, but alternative protein scaffolds are increasingly being used for the directed evolution of proteins with novel molecular recognition properties. We have designed a combinatorial library based on the DNA binding domain of the human retinoid-X-receptor (hRXRalpha). We chose this domain because of its small size, stable fold, and two closely juxtaposed recognition loops. We replaced the two loops with segments of random amino acids, and used mRNA display to isolate variants that specifically recognize adenosine triphosphate (ATP), demonstrating a significant alteration of the function of this protein domain from DNA binding to ATP recognition. Many novel independent sequences were recovered with moderate affinity and high specificity for ATP, validating this scaffold for the generation of functional molecules.

A new generation of protein display scaffolds for molecular recognition

Engineered antibodies and their fragments are invaluable tools for a vast range of biotechnological and pharmaceutical applications. However, they are facing increasing competition from a new generation of protein display scaffolds, specifically selected for binding virtually any target. Some of them have already entered clinical trials. Most of these nonimmunoglobulin proteins are involved in natural binding events and have amazingly diverse origins, frameworks, and functions, including even intrinsic enzyme activity. In many respects, they are superior over antibody-derived affinity molecules and offer an ever-extending arsenal of tools for, e.g., affinity purification, protein microarray technology, bioimaging, enzyme inhibition, and potential drug delivery. As excellent supporting frameworks for the presentation of polypeptide libraries, they can be subjected to powerful in vitro or in vivo selection and evolution strategies, enabling the isolation of high-affinity binding reagents. This article reviews the generation of these novel binding reagents, describing validated and advanced alternative scaffolds as well as the most recent nonimmunoglobulin libraries. Characteristics of these protein scaffolds in terms of structural stability, tolerance to multiple substitutions, ease of expression, and subsequent applications as specific targeting molecules are discussed. Furthermore, this review shows the close linkage between these novel protein tools and the constantly developing display, selection, and evolution strategies using phage display, ribosome display, mRNA display, cell surface display, or IVC (in vitro compartmentalization). Here, we predict the important role of these novel binding reagents as a toolkit for biotechnological and biomedical applications.

Rational engineering of a fluorescein-binding anticalin for improved ligand affinity

The anticalin FluA is an artificial lipocalin with novelspecificity for the fluorescein group, which was engineered from an insect bilin-binding protein by targeted random mutagenesis and selection. Based on the crystal structure of FluA, an attempt was made to improve the complementarity of its ligand pocket to fluorescein by rational protein design. Several side chains participating in sub-optimal interactions with the ligand were identified and replaced by residues that promised a better steric fit. As a result, the substitution of Ala45 by Ile and of Ser114 by Thr or Arg led to a tight affinity of ca. 1 nM, which is approximately 30-fold better than that of the parental anticalin. Similar to the original FluA, the improved version shows almost complete quenching of the bound ligand fluorescence. Interestingly, the quenching effect was significantly reduced when Trp129 was replaced by Tyr, thus supporting the previously postulated role of this residue, which closely packs against the bound ligand, for efficient electron transfer to the excited fluorescein. Circular dichroism spectra revealed that all variants investigated had retained the lipocalin fold. Corresponding thermal unfolding experiments confirmed similar folding stabilities, with melting temperatures ranging from 52.9 to 60.5 degrees C (i.e., for the high-affinity variant).

Artificial, non-antibody binding proteins for pharmaceutical and industrial applications

Using combinatorial chemistry to generate novel binding molecules based on protein frameworks ('scaffolds') is a concept that has been strongly promoted during the past five years in both academia and industry. Non-antibody recognition proteins derive from different structural families and mimic the binding principle of immunoglobulins to varying degrees. In addition to the specific binding of a pre-defined target, these proteins provide favourable characteristics such as robustness, ease of modification and cost-efficient production. The broad spectrum of potential applications, including research tools, separomics, diagnostics and therapy, has led to the commercial exploitation of this technology by various small- and medium-sized companies. It is predicted that scaffold-based affinity reagents will broaden and complement applications that are presently covered by natural or recombinant antibodies. Here, we provide an overview on current approaches in the biotech industry, considering both scientific and commercial aspects.