Pharmaceutical Editorial

3 Editorial 2015/09/26 ‘Engineered protein scaffolds’: have they lived up to expectations?

“... the growing number of clinical programs clearly indicates that non-Ig protein scaffolds have emerged beyond the stage of academic research.”

Pharm. Bioprocess. Keywords: bioprocessing • developability • pharmacology • safety • scaffold

Monoclonal antibodies are currently the • Affibodies based on the Z-domain of most successful class of therapeutic agents. Staphylococcal protein A [3]; However, the conventional immunoglobulin • Adnectins (previously called monobod- (Ig) format is not always optimally suited to ies) derived from the 10th extracellu- meet clinical demands. First, immunologi- lar domain of human fibronectin III cal effector functions mediated by the Fc 10 ( Fn3) [4]; region can evoke undesired side effects. Sec- ond, poor tissue penetration due to the large • Anticalins, obtained by engineering the molecular size hampers successful treatment natural binding site of the cup-shaped Arne Skerra Author for correspondence: of solid tumors. Also, the long circulation human lipocalins [5]; 385 Lehrstuhl für Biologische Chemie, in blood resulting from both the large size Technische Universität München, and FcRn-mediated endosomal recycling is • Avimers, based on a multimerized engi- Emil-Erlenmeyer-Forum 5, 85350 unfavorable both for therapies that require neered LDLR-A module [6]; Freising-Weihenstephan, Germany flexible adjustment of dosing and for in vivo Tel.: +49 8161 71 4351 • DARPins, designed ankyrin repeat pro- imaging applications. Finally, due to their Fax: +49 8161 71 4352 teins based on a consensus scaffold [7]; [email protected] complex biomolecular architecture, includ- ing four polypeptide chains with around • Fynomers, engineered from the SH3- 1500 amino acids and at least two glycosyl- domain of the human Fyn tyrosine ation sites, the production of full size anti- kinase and applied as fusion protein with bodies is costly and requires mammalian a monoclonal antibody [8]. expression systems. As a consequence, during the last two Neglecting business aspects such as access to 6 decades more than 50 alternative types of capital, competitive landscape or intellectual binding proteins have been proposed with property, the chance of a biopharmaceutical the intention to overcome some of the inher- to reach the clinic or, eventually, the market is ent limitations of antibodies. However, only primarily dependent on the so-called develop- 2015 a minority of these ‘alternative scaffolds’ ability of the molecule [9]. To this end, three Stefan R Schmidt have reached the clinic so far, which can major factors need to be considered: manufac- Rentschler Biotechnologie GmbH, be seen as the ultimate success in pharma- turing, safety, and pharmacology (including Erwin-Rentschler-Strasse 21, 88471 Laupheim, Germany ceutical biotechnology. According to recent biological activity). All three together ulti- reviews, ten drug candidates based on mately determine the overall chance of success seven different protein scaffolds have been to develop a biological drug. In principle, for tested in clinical trials while one biological most of these parameters the non-Ig scaffolds received market approval (the engineered should offer advantages if compared with Kunitz domain/protease inhibitor ecallan- classical antibodies and, therefore, present a tide) [1,2]. At present, the biopharmaceutical worthwhile alternative. development is dominated by the following Engineered protein scaffolds are generally protein scaffolds: easier and more cost-efficient to manufacture part of

10.4155/pbp.15.20 © 2015 Future Science Ltd Pharm. Bioprocess. (2015) 3(6), 383–385 ISSN 2048-9145 383 Editorial Skerra & Schmidt

in prokaryotic hosts for several reasons: usually, they data so far have proven, the level of immunogenicity contain no or only few disulfide bridges and they lack even for the non-endogenous protein scaffolds is usu- complex posttranslational modifications (PTMs) such ally low, most likely due to their small size, compact as glycosylation. In addition, they are significantly structure as well as low aggregation propensity. Addi- smaller than antibodies and monomeric, which does tionally, critical T-cell epitopes can be removed by pro- not only considerably facilitate downstream processing tein engineering. Off-target interactions have not been and quality control but, in principle, might even allow reported so far, but corresponding side effects could full chemical synthesis. Furthermore, when choosing limit the therapeutic window with regard to applied a microbial expression system such as Escherichia coli dose and frequency of administration. or yeast, the threat of viral contamination can largely The third crucial aspect of developability concerns be ignored during processing as no animal cells are pharmacology, in particular bioavailability and phar- involved; on the other hand, endotoxin removal must macokinetics. While the small size of most alternative be carefully monitored in case of a bacterial host cell. binding proteins is generally considered a benefit with One potential caveat of engineered scaffolds is the regard to high specific activity and good tissue penetra- lack of platform downstream processes since an estab- tion, the resulting fast excretion via kidney filtration lished affinity matrix like, for example, immobilized usually also constitutes a disadvantage. Nevertheless, protein A, which simplifies the first capture step dur- today there are several options available to increase ing antibody purification, is often not available. Initial plasma half-life either by chemically or genetically con- capture might also be complicated when intracellular jugating a larger molecule (polymer) or by implement- expression is performed (in E. coli) and the protein of ing FcRn-mediated endosomal recycling via fusion interest must be separated from abundant host pro- with Fc or albumin. teins and other cell constituents. In some rare cases, For the site-specific chemical conjugation of PEG, such as for Affibody molecules, their stability may radioactive labels or chemical toxins, the presence of a be utilized for enrichment under denaturing condi- free Cys residue is helpful and can easily be engineered tions (e.g., ­elevated temperature, extreme pH) that into most of the non-Ig scaffolds. While PEGylation precipitate or degrade most of the host cell proteins. leads to retarded renal filtration via a molecular size Alternatively, bacterial [10] and yeast expression sys- effect and has become the gold standard for plasma tems that allow highly efficient secretion are available half-life extension during the last decades, alternative today, offering a preferred route for some of the non-Ig technologies were recently proposed; PASylation, for p­rotein scaffolds. example, utilizes a genetically encoded polypeptide Physical stability of the biological can be improved made of the natural amino acids Pro, Ala and/or Ser by protein engineering, which is particularly advanced and provides the same biophysical properties as PEG for some types of protein scaffolds, whereas in the but, in contrast, allows cheaper manufacturing and is case of antibodies the current rules for ‘humaniza- biodegradable [11] . tion’ allow only moderate alterations to their frame- In principle, non-Ig scaffolds also offer additional work and constant regions. PTM, which might give modes of action compared with antibodies. This is rise to a wide range of isoforms that have to be sepa- partly related to their differently shaped binding inter- rated in a tedious downstream process, can either be face, which can address target epitopes not easily acces- eliminated by site-directed mutagenesis of the non-Ig sible by conventional Igs. In the context of the most scaffold, may simply be omitted through choice of a advanced scaffolds mentioned further above, one can prokaryotic expression system or is absent from the distinguish between binding via flat surfaces, cavi- beginning. Most of the alternative scaffolds do not ties or protruding loops, for example, to reach steri- require PTM for activity; therefore, product homo- cally hidden clefts or to tightly embrace small hapten geneity can be generally achieved easier. The absence compounds [12] . Furthermore, due to their small size, of unwanted glycosylation can also reduce the risk of scaffolds are ideal components for the design of fusion hypersensitivity in patients or suppress the tendency proteins, where two unrelated proteins are genetically to form aggregates. joined to combine the functionalities of both par- The second parameter for developability, safety, is ent proteins in a single polypeptide chain. Therefore, primarily related to immunogenicity and target speci- bispecific fusion proteins enable identical biodistribu- ficity. With just few exceptions, for example, Anti- tion for both protein components and simplify manu- calins [5], most of the engineered scaffolds are not of facturing [13] . Finally, the small size and robust nature human origin and, thus, unknown to the body. This of alternative binding proteins opens novel routes to can quickly lead to neutralizing antibodies or, in the administration, for example, via transdermal delivery worst case, cause allergic reactions. However, as c­linical or inhalation.

384 Pharm. Bioprocess. (2015) 3(6) future science group ‘Engineered protein scaffolds’: have they lived up to expectations? Editorial

In conclusion, the growing number of clinical pro- the l­aboratory design of innovative multifunctional grams clearly indicates that non-Ig protein scaffolds f­ormats, will show soon whether engineered protein have emerged beyond the stage of academic research. scaffolds really live up to expectations. However, it also seems that this area of discovery approaches consolidation since the majority of ≥50 Financial & competing interests disclosure scaffolds that were proposed in the scientific literature A Skerra is shareholder of Pieris Pharmaceuticals, Inc. and co- have never delivered more than an initial example. For founder and managing director of XL-protein GmbH. The au- the more promising members of this class on the route thors have no other relevant affiliations or financial involvement toward novel biopharmaceutical drugs major tasks with any organization or entity with a financial interest in or have been mastered, including the challenges arising financial conflict with the subject matter or materials discussed during bioprocess development. Forthcoming data in the manuscript apart from those disclosed. No writing assis- both from advanced clinical studies, and also from tance was utilized in the production of this manuscript.

References 8 Silacci M, Baenziger-Tobler N, Lembke W et al. Linker length matters: Fynomer-Fc fusion with an optimized linker 1 Škrlec K, Štrukelj B, Berlec A. Non-immunoglobulin displaying picomolar IL–17A inhibition potency. J. Biol. scaffolds: a focus on their targets. Trends Biotechnol. 33(7), Chem. 289(20), 14392–14398 (2014). 408–418 (2015). 9 Zurdo J, Arnell A, Obrezanova O et al. Early Implementation 2 Gebauer M, Skerra A. Alternative protein scaffolds as of QbD in biopharmaceutical development: a practical novel biotherapeutics. In: Biobetters – Protein Engineering example. Biomed. Res. Int. 2015, 1–19 (2015). to Approach the Curative. Rosenberg A, Demeule B (Eds). 10 Di Cesare S, Binder U, Maier T, Skerra A. High-yield Springer, NY, USA, 221–268 (2015). production of PASylated human growth hormone using 3 Feldwisch J, Tolmachev V. Engineering of affibody molecules secretory E. coli technology. Bioprocess Int. 11(4), 30–38 for therapy and diagnostics. Methods Mol. Biol. 899, 103–126 (2013). (2012). 11 Schlapschy M, Binder U, Börger C et al. PASylation: a 4 Lipovšek D. Adnectins: engineered target-binding protein biological alternative to PEGylation for extending the plasma therapeutics. Protein Eng. Des. Sel. 24(1–2), 3–9 (2011). half-life of pharmaceutically active proteins. Protein Eng. Des. 5 Richter A, Eggenstein E, Skerra A. Anticalins: exploiting a Sel. 26(8), 489–501 (2013). non-Ig scaffold with hypervariable loops for the engineering 12 Gebauer M, Skerra A. Engineered protein scaffolds as next- of binding proteins. FEBS Lett. 588(2), 213–218 (2014). generation antibody therapeutics. Curr. Opin. Chem. Biol. 6 Silverman J, Liu Q, Bakker A et al. Multivalent avimer 13(3), 245–255 (2009). proteins evolved by exon shuffling of a family of human 13 Schmidt SR. Fusion proteins: applications and challenges. receptor domains. Nat. Biotechnol. 23(12), 1556 –1561 In: Fusion Protein Technologies for Biopharmaceuticals: (2005). Applications and Challenges. Schmidt SR (Ed.). Wiley, 7 Plückthun A. Designed Ankyrin Repeat Proteins (DARPins): Hoboken, NJ, 3–24 (2013). binding proteins for research, diagnostics, and therapy. Annu. Rev. Pharmacol. Toxicol. 55(1), 489–511 (2015).

future science group www.future-science.com 385