FEATURE ARTICLE

Leveraging Lipocalins for Drug Development1 Executive summary Lipocalins are a naturally occurring family of small proteins involved in the transport or storage of chemically sensitive or waterinsoluble biological molecules. Although lipocalins have therapeutic potential in their own right, it is the use of engineered derivatives as alternatives to antibodies that is exciting most interest. The architecture of lipocalins – specifically, the presence of a hypervariable ligand-binding region at the end of a rigid structural element – creates the possibility of reshaping the binding site to produce versatile binding proteins against a variety of molecular targets for antagonistic or agonistic purposes. Anticalins® are a novel class of synthetic ligand-binding proteins engineered from lipocalins via targeted random mutagenesis. Anticalins can be selected for binding either small ligands or macromolecules, and have potential applications as bioanalytics, diagnostics, and therapeutics. Consisting of a single, stable polypeptide chain, lipocalins and their Anticalin derivatives are much smaller than antibodies or antibody fragments, are generally not glycosylated, and have lower immunogenic potential. Lipocalin-based drugs have several obvious therapeutic applications, such as receptor antagonists, antidotes against toxic compounds and vehicles for drug delivery. The potential advantages of lipocalin derivatives over antibody-based drugs – including improved drug delivery, cheaper manufacture and simpler manipulation – may seem them emerge as therapeutic drugs of tomorrow.

Lipocalins – versatile binding proteins Lipocalins are a family of ligand-binding, functionally diverse proteins found in a variety of organisms from humans to bacteria. In humans, lipocalins are abundant in plasma, tissues, and secretory fluids. The physiological function of most lipocalins is to store or transport hydrophobic or volatile compounds such as steroids, vitamins, and various metabolites. Less commonly, lipocalins can act as carriers of pheromones and odorants, and a few have been shown to possess enzymatic function. Approximately ten different lipocalins have been identified in humans, the best characterized of which is the plasma retinol-binding protein (RBP), which transports the poorly water-soluble and oxygen-sensitive vitamin A from the liver to several target tissues.

Like antibodies but smaller and more stable Although lipocalins exhibit a conserved threedimensional structure, heterogeneity at the ligand-binding site gives rise to a diverse range of ligand-binding properties. Interestingly, there is a striking resemblance between the three-dimensional structure of lipocalins and antibodies – including an element in lipocalins that seems to mimic the hypervariable antigen-binding region of immunoglobulins. Nevertheless, whereas antibodies possess a large size and complex architecture consisting of two pairs of polypeptide chains, lipocalins possess a simpler structure composed of a single short polypeptide chain of 160–180 residues that is generally not glycosylated. Due to their stably folded monomeric structure, lipocalins are generally robust proteins with significant thermostability.

Ligand transport throughout the body As would be expected for proteins involved in the binding and release of molecules, the ligand affinity of lipocalins is usually only moderate, with dissociation constants generally in the low micromolar range. Although some lipocalins are highly specific in terms of ligandbinding, others bind a wide range of compounds. For example, the human neutrophil gelatinase-associated lipocalin (NGAL) only binds carboxymycobactin molecules from mycobacteria whereas human tear lipocalin (Tlc) binds a wide range of lipophilic compounds. Lipocalins that bind ligands with intracellular destinations could potentially interact with cellular receptors. However, only two such cell-surface receptors have been identified – the NGAL-binding megalin receptor and the Tlc-binding LIMR receptor – and many lipocalins appear to have no receptor interaction. Thus, it appears that most lipocalins function as soluble ligandtransport proteins that are freely distributed throughout the bloodstream and tissue fluids.

Drug development of lipocalins As the discovery and characterization of lipocalins continues, there is growing interest in their potential application as bioanalytics, diagnostics and therapeutics. Several different approaches can be envisaged for therapeutic use of lipocalins: • The physiological effects of certain human lipocalins have therapeutic potential. For example, α-1-acid glycoprotein (AGP) enhances capillary barrier function during shock, thus helping maintain organ perfusion. Therefore, AGP may be of therapeutic use in patients experiencing trauma.

1 The full text of this article was published in BioDrugs 2005; 19 (5): 279-288. The original article was written by Steffen Schlehuber from PIERIS Proteolab and Arne Skerra from the Technical University of Munich, Germany. 1176-3469/05/0010-003/$34.95 © 2005 Adis Data Information BV. All rights reserved.

4

•

Several lipocalins produced by blood-sucking insects tightly bind inflammatory compounds and therefore may have direct application as drugs. For example, the tick Rhipicephalus appendiculatus produces a lipocalin that tightly binds histamine. This histamine-binding protein (HBP) has therapeutic potential in allergic disorders in which histamine plays a key proinflammatory role. Based on its efficacy in preclinical models of allergy, HBP has entered clinical trials with Evolutec Group (under the code name rEV131) for the treatment of allergic disorders such as rhinitis and conjunctivitis. • The ligand-binding properties of natural lipocalins render them particularly attractive not only as scavengers for physiologically active compounds but also as carrier vehicles for pharmaceutical drugs. Although the therapeutic use of natural lipocalins holds promise, the engineering of synthetic derivatives with novel binding properties may represent the most useful application of lipocalins – potentially providing an entire new class of protein-based targeted therapeutics.

Anticalins: novel binding proteins derived from the lipocalin scaffold The architecture of lipocalins – specifically, the presence of a hypervariable ligand-binding region at the end of a rigid structural element – allows the possibility of reshaping the binding site to produce versatile binding proteins against a variety of molecular targets for antagonistic or agonistic purposes. Such an approach is analogous to the use of engineered antibodies or antibody fragments. The structure of naturally occurring antibodies provides a type of biomolecular scaffold whose hypervariable antigen-binding site can be engineered to recognize novel molecular targets. With approximately 20 antibody-based products now approved as biopharmaceuticals – mainly for the treatment of cancer and autoimmune diseases – the application of antibody technology for drug development is well established.

polypeptides – the light and heavy chains – which causes unstable domain association when dealing with small variable (Fv) fragments and necessitates complicated cloning steps for the pair of genes. Third, the complex architecture of the binding site, which is formed by six hypervariable loops, is difficult to manipulate simultaneously if synthetic libraries are to be generated. Fourth, the constant (Fc) region mediates immunological effector reactions, which are normally not necessary for biopharmaceutical use and are often undesired. Finally, antibody production usually requires expensive mammalian expression systems, hence restricting chronic medical applications where large amounts of active substance are needed over a long period.

Lipocalin-derived antibody alternatives Consequently, there is a need for an alternative protein scaffold to create artificial binding proteins with improved pharmacological properties. Lipocalins appear particularly promising in this respect because of their small size, simple structure, ease of production and genetic manipulation, and high folding stability. The German company PIERIS ProteoLab is developing engineered lipocalin derivatives using combinatorial protein design to modify the four loops forming the entrance to the ligand-binding site in lipocalins. Using this strategy, PIERIS has generated molecular libraries containing proteins with novel binding specificities that can be selected against desired target molecules. This novel class of engineered proteins, that possess antibody-like ligand-binding activity but which are based on the lipocalin scaffold, has been termed Anticalins®.

Applications of Anticalins As is the case with antibodies, the diverse binding specificity of Anticalins offers the potential for use in many diagnostic and therapeutic biomedical applications.

Antidotes Engineering synthetic derivatives with novel binding properties may represent the most useful application of lipocalins – potentially providing an entire new class of protein-based targeted therapeutics. Despite their usefulness, antibodies suffer from some fundamental limitations. First, they are large macromolecules with limited ability to penetrate tissue, a feature that hinders the efficient treatment of solid tumors. Second, immunoglobulins consist of two different Pharmaceutical & Diagnostic Innovation 2005; Vol. 3, No. 10

PIERIS has produced a novel Anticalin DigA16 that binds digoxin – a compound used to treat ventricular tachyarrhythmias. There is only a narrow margin between therapeutic dose and onset of toxic effects with digoxin, such that precise adjustment of serum levels is mandatory to prevent fatal outcome. Nevertheless, fatal digitalis poisoning resulting from digoxin overdose remains a risk. Several digoxin-specific antibody fragments (Fab) have been developed for use as antidotes to digitalis overdose – notably DigiFab™ from Protherics and Digibind® from GlaxoSmithKline. Although these compounds have been

modified for safety purposes, they are still associated with a risk of anaphylactic reactions due to their animal origin. Consequently, the availability of safer digoxin-binding agents could provide a major advantage in the therapy of digitalis overdose. The engineered lipocalin DigA16 offers a promising lead candidate for this therapeutic application. To increase its usefulness, derivatives were produced with dramatically improved affinities for digoxin. One of the derivatives, Digical®, was highly effective as an antidote to digoxin-induced toxicity in both guinea pig and porcine models of overdose. Importantly, the toxicity profile was favorable, with no signs of acute toxicity observed after administration of DigA16 25 mg/kg to mice over 2 weeks. Taken together, these results demonstrate for the first time that the lipocalin scaffold can serve as a basis for the development of a therapeutically active compound via protein design. Furthermore, the data from the Digical studies provide an in vivo proof of concept for lipocalin derivatives as antidotes to poisoning with small molecules.

Antagonists of soluble proteins and receptors Most biopharmaceuticals target cell-surface receptors or soluble plasma proteins and peptides such as hormones, interleukins and cytokines. The binding function of a biopharmaceutical may be directed against either the cellsurface receptor or its physiological ligand. By interrupting their mutual interaction, the therapeutic protein exerts an antagonistic effect and thereby inhibits cellular responses such as proliferation, apoptosis, or differentiation. To provide an alternative to traditional biopharmaceuticals, PIERIS has developed specific Anticalin libraries for the recognition of proteins. During the design of suitable combinatorial libraries, the bulky nature of proteins – which cannot penetrate into the lipocalin pocket as deeply as small molecules – was taken into consideration. Therefore, accessible amino acid positions at tips of the four loops were preferentially subjected to site-directed random mutagenesis. Furthermore, to reduce the potential for anaphylaxis, the libraries were derived from human lipocalins. For example, a combinatorial library based on the NGAL lipocalin scaffold was constructed and employed for the selection of Anticalins with antagonistic properties toward the cytotoxic T lymphocyte-associated antigen (CTLA-4; CD152) – a T-cell co-receptor with central immunomodulatory activity that is a target for cancer immunotherapy. Antibodies that bind to human CTLA-4 and block its interaction with the CD80 receptor on antigen-presenting cells have been shown to be effective for cancer treatment in preclinical and clinical studies.

Screening of the NGAL-derived library against the target human CTLA4-Ig – a fusion of the extracellular region of the receptor and an immunoglobulin Fc fragment – yielded a number of derivatives binding different epitopes with affinities in the low nanomolar range. The antagonistic activity of one particularly promising CTLA-4-specific Anticalin was demonstrated in several in vitro cell culture tests, where T-cell proliferation was enhanced to a comparable extent as with commercially available antibodies against CTLA-4. The affinity and epitope specificity of this CTLA-4-binding Anticalin were very similar to those of a fully human antibody, MDX-010, currently in clinical development with Medarex. MDX-010 has been employed successfully for tumor immunotherapy in cynomolgus macaques. However, in contrast to MDX-010, which does not bind to murine CTLA-4, the Anticalin analog exhibited full crossreactivity with the target receptor from mice, thus facilitating its ongoing preclinical evaluation. These data indicate that lipocalin derivatives should generally be applicable for blockade of disease-associated cell-surface receptors.

Delivery of cytotoxic agents Monoclonal antibodies have demonstrated remarkable efficacy in the clinical treatment of cancers such as nonHodgkin lymphoma (Rituxan®; rituximab) and metastatic breast cancer (Herceptin®; trastuzumab). This has prompted interest in further increasing their efficacy by ‘arming’ them with cytotoxic drugs. In this approach, antibodies directed at internalizing receptors on the surface of cancer cells are used to deliver cytotoxic compounds specifically to a tumor, thus avoiding the toxic adverse effects of chemotherapy on healthy tissue. Generally, intact monoclonal antibodies have been used for drug targeting but there is a current trend toward employing antibody fragments.

The stable monomeric nature of lipocalins and their Anticalin derivatives should offer practical benefits for conjugation or fusion with cytotoxic agents. Although intact antibodies usually bind the target receptor more tightly than antibody fragments and may be more effective for eliciting antitumor immune responses, they have greater potential to cause anaphylaxis and their long serum half-lives prolong patient exposure to cytotoxic agents, thus increasing the risk of adverse effects. Although the use of antibody fragments avoids such problems, such derivatives have limited stability and 1176-3469/05/0010-005/$34.95 © 2005 Adis Data Information BV. All rights reserved.

6

a pronounced tendency to form oligomers or aggregates, which is a significant disadvantage for the production of well-defined pharmaceutical preparations. In contrast, the stable monomeric nature of lipocalins and their Anticalin derivatives should offer practical benefits for conjugation or fusion with cytotoxic agents. In addition, Anticalins display both their amino- and carboxy-terminus in an accessible manner remote from the binding region, in contrast to antibody fragments in which the amino-terminus is close to the antigen binding site. Therefore, Anticalins are well suited for fusion with toxic peptides or other functional domains without compromising their binding activity. A fusion of an Anticalin® targeting a specific tumor cell receptor with an enzyme that generates a cytotoxic compound from an inactive prodrug is of special interest as the lipocalin version of antibody-directed enzyme prodrug therapy (ADEPT).

Advantages of lipocalin-based drugs By using their natural ligand-binding functions, lipocalins can be exploited as unmodified recombinant proteins – such as HBP from blood-feeding ticks for the treatment of allergies. Artificially engineered lipocalin derivatives – Anticalins – with tailor-made specificities for low-molecular-weight compounds or macromolecular targets, represent a novel class of promising new drug candidates. Anticalins that block the interaction between a receptor and its physiological ligand offer the advantage of a well defined pharmacological function. Unlike antibody-based antagonists, which are generally manufactured in complex and expensive mammalian-based protein expression systems, large amounts of lipocalin derivatives can be produced in cheaper bacterial expression systems. Importantly, the smaller size of Anticalins compared with antibodies should result in greater tissue penetration. The application of Anticalins as antidotes, an approach validated in several in vivo models, could be useful to reverse toxicity caused by overdosing of established smallmolecule therapeutics. Data available so far indicate that natural lipocalins as well as Anticalins possess a serum half-life of less than 1 hour when administered as unconjugated proteins and are cleared from the body mainly via the renal route. This rapid excretion is an obvious advantage for an antidote, such as the digoxin-

Pharmaceutical & Diagnostic Innovation 2005; Vol. 3, No. 10

specific Digical, because the bound toxin is rapidly removed from circulation. In chronic therapy, for which sustained exposure to the therapeutic protein is necessary, serum half-life can be prolonged by well-established techniques such as conjugation to polyethylene glycol (PEGylation). This method could also help to avoid potential problems with immunogenicity, although immunogenic effects due to side chain replacements in the loop region should not be more pronounced than with antibody fragments – particularly with Anticalins derived from human lipocalins. Anticalins targeting internalizing cell-surface receptors overexpressed on tumor cells are of interest for the targeted delivery of cytotoxic compounds. Their robust monomeric structure, resulting in prolonged shelf life, may confer advantages over antibody fragments, which have recently attracted attention as part of recombinant immunotoxins for drug-targeting strategies.

Looking ahead Antibody-based drugs have proven to be a remarkable success story in modern medicine, with approximately 20 such biopharmaceuticals approved by the US FDA and a global market for antibody therapeutics projected to be worth $US16.7 billion by 2008. Nevertheless, antibodies do have significant limitations as biopharmaceuticals and, consequently, alternative protein scaffolds for drug design are under investigation to complement the toolbox of therapeutic proteins. Synthetic lipocalin derivatives – Anticalins – appear to be particularly promising in this respect. As is the case with antibodies, the remarkable plasticity of the ligandbinding site permits the selection of Anticalins against a diverse range of target molecules. Compared with antibodies, the structural features of Anticalins promise improved drug delivery and simpler, cheaper manufacture for biopharmaceutical purposes. Lipocalin-derived drugs have the potential to be used in a variety of therapeutic applications– including as blockers of specific proteins and receptors, as antidotes to poisons, and as vehicles for cytotoxic drug delivery. The unique properties of Anticalins should allow their versatile application as therapeutic drugs of tomorrow.