Bacterial display systems for engineering of affinity proteins

Filippa Fleetwood

Kungliga Tekniska Högskolan, KTH Royal Institute of Technology School of Biotechnology Stockholm 2014

© Filippa Fleetwood Stockholm 2014

Royal Institute of Technology School of Biotechnology AlbaNova University Center SE-106 91 Stockholm Sweden

Printed by Universitetsservice US-AB Drottning Kristinas väg 53B SE-100 44 Stockholm Sweden

ISBN 978-91-7595-374-8 TRITA BIO-Report 2014:18 ISSN 1654-2312

ABSTRACT

Directed evolution is a powerful method for engineering of specific affinity proteins such as antibodies and alternative scaffold proteins. For selections from combinatorial protein libraries, robust and high-throughput selection platforms are needed. An attractive technology for this purpose is cell surface display, offering many advantages, such as the quantitative isolation of high-affinity library members using flow-cytometric cell sorting. This thesis describes the development, evaluation and use of bacterial display technologies for the engineering of affinity proteins. Affinity proteins used in therapeutic and diagnostic applications commonly aim to specifically bind to disease-related drug targets. Angiogenesis, the formation of new blood vessels from pre-existing vasculature, is a critical process in various types of cancer and vascular eye disorders. Vascular Growth Factor Receptor 2 (VEGFR2) is one of the main regulators of angiogenesis. The first two studies presented in this thesis describe the engineering of a biparatopic Affibody molecule targeting VEGFR2, intended for therapeutic and in vivo imaging applications. Monomeric VEGFR2-specific Affibody molecules were generated by combining phage and staphylococcal display technologies, and the engineering of two Affibody molecules, targeting distinct epitopes on VEGFR2 into a biparatopic construct, resulted in a dramatic increase in affinity. The biparatopic construct was able to block the ligand VEGF-A from binding to VEGFR2-expressing cells, resulting in an efficient inhibition of VEGFR2 phosphorylation and angiogenesis-like tube formation in vitro. In the third study, the staphylococcal display system was evaluated for the selection from a single-domain antibody library. This was the first demonstration of successful selection from an antibody-based library on Gram-positive bacteria. A direct comparison to the selection from the same library displayed on phage resulted in different sets of binders, and higher affinities among the clones selected by staphylococcal display. These results highlight the importance of choosing a display system that is suitable for the intended application. The last study describes the development and evaluation of an autotransporter-based display system intended for display of Affibody libraries on E. coli. A dual-purpose expression vector was designed, allowing efficient display of Affibody molecules, as well as small-scale protein production and purification of selected candidates without the need for sub-cloning. The use of E. coli would allow the display of large Affibody libraries due to a high transformation frequency. In combination with the facilitated means for protein production, this system has potential to improve the throughput of the engineering process of Affibody molecules.

In summary, this thesis describes the development, evaluation and use of bacterial display systems for engineering of affinity proteins. The results demonstrate great potential of these display systems and the generated affinity proteins for future biotechnological and therapeutic use.

Keywords: Combinatorial protein engineering, staphylococcal display, Affibody, biparatopic, VEGFR2, nanobody, E. coli display, autotransporter

List of publications

This thesis is based on the work described in the following publications or manuscripts, referred to as papers numbered in roman numerals (I-IV). The papers are presented in the appendix.

I Fleetwood, F., Klint, S., Hanze, M., Gunneriusson, E., Frejd, F.Y., Ståhl, S., Löfblom, J. Simultaneous targeting of two ligand-binding sites on VEGFR2 using biparatopic Affibody molecules results in dramatically improved affinity, in press, Sci. Rep.

II Fleetwood, F., Frejd, F.Y., Ståhl, S., Löfblom, J. Efficient blocking of VEGFR2-mediated signaling using biparatopic Affibody constructs,

manuscript

III Fleetwood, F., Devoogdt, N., Pellis, M., Wernery, U., Muyldermans, S., Ståhl, S., Löfblom, J. Surface display of a single-domain antibody library on Gram-positive bacteria, Cell Mol. Life Sci. (2013) 70:1081-1093

IV Fleetwood, F.*, Andersson, K. G.*, Ståhl, S., Löfblom, J. An engineered autotransporter-based surface expression vector enables efficient display of Affibody molecules on OmpT-negative E. coli as well as protease-mediated secretion in OmpT-positive strains, conditionally accepted, revised manuscript submitted, Microbial Cell Factories

*These authors contributed equally

All papers are reproduced with permission from the copyright holders.

ABBREVIATIONS

ABD Albumin-binding domain ABP Albumin-binding protein AIDA-I Adhesin involved in diffuse adherence CDR Complementary determining region DNA Deoxyribonucleic acid ELISA Enzyme-linked immunosorbent assay Fab Fragment, antigen-binding Fc Fragment, crystallizable FACS Fluorescence-activated cell sorting FDA Food and drug administration HSA Human serum albumin Fv Fragment, variable GFP Green fluorescent protein IgG Immunoglobulin G IMAC Immobilized metal ion affinity chromatography

KD Equilibrium dissociation constant kd Dissociation rate constant Nb Nanobody OmpT Outer membrane protease T PCR Polymerase chain reaction scFv Single chain fragment variable SDS-PAGE Sodium dodecyl sulfate polyacrylamide gel electrophoresis VEGF Vascular endothelial growth factor VEGFR2 Vascular endothelial growth factor receptor 2

CONTENTS

1. PROTEIN ENGINEERING ...... 1 1.1 PROTEINS IN THERAPEUTICS AND DIAGNOSTICS ...... 1 1.2 PROTEIN ENGINEERING BY RATIONAL DESIGN ...... 3 1.3 COMBINATORIAL PROTEIN ENGINEERING AND PROTEIN LIBRARIES ...... 4 2. AFFINITY PROTEINS ...... 8 2.1 ANTIBODIES AND ANTIBODY FRAGMENTS ...... 8 2.2 NANOBODIES ...... 12 2.3 ALTERNATIVE SCAFFOLD PROTEINS ...... 13 3. SELECTION SYSTEMS ...... 20 3.1 PHAGE DISPLAY ...... 20 3.2. CELL SURFACE DISPLAY ...... 22 3.3. CELL-FREE DISPLAY AND NON-DISPLAY SELECTION PLATFORMS ...... 33 4. TARGETING OF ANGIOGENESIS ...... 35 4.1 ANGIOGENESIS ...... 35 4.2 THE VEGF/VEGFR FAMILY ...... 36 4.3 TAREGTING OF THE VEGF LIGANDS AND RECEPTORS IN THERAPEUTICS AND DIAGNOSTICS ...... 39 5. PRESENT INVESTIGATION ...... 43 5.1 PAPER I ...... 45 SIMULTANEOUS TARGETING OF TWO LIGAND-BINDING SITES ON VEGFR2 USING BIPARATOPIC AFFIBODY MOLECULES RESULTS IN DRAMATICALLY IMPROVED AFFINITY .. 45 5.2 PAPER II ...... 53 EFFICIENT BLOCKING OF VEGFR2-MEDIATED SIGNALING USING BIPARATOPIC AFFIBODY MOLECULES ...... 53 5.3 PAPER III ...... 59 SURFACE DISPLAY OF A SINGLE-DOMAIN ANTIBODY LIBRARY ON GRAM-POSITIVE BACTERIA ...... 59 5.4 PAPER IV ...... 64 AN ENGINEERED AUTOTRANSPORTER-BASED SURFACE EXPRESSION VECTOR ENABLES EFFICIENT DISPLAY OF AFFIBODY MOLECULES ON OMPT-NEGATIVE E. COLI AS WELL AS PROTEASE-MEDIATED SECRETION IN OMPT-POSITIVE STRAINS ...... 64 POPULAR SCIENCE SUMMARY ...... 76 ACKNOWLEDGEMENTS ...... 79 REFERENCES ...... 81 APPENDIX ...... 89

“Play is the highest form of research”

- Albert Einstein Filippa Fleetwood

1. PROTEIN ENGINEERING

Proteins are polymeric biomolecules made of long chains of amino acids, which are usually folded into three-dimensional structures. The amino acid sequence is encoded by a DNA sequence, called a gene. A large part of our bodies is made of proteins, and proteins are responsible for many important functions, including for example enzymatic reactions, cell signaling and transportation of molecules. Many diseases are caused by mutated proteins or an up- or down-regulation of a certain protein, commonly leading to disruption or over-activation of various signaling pathways. The purpose of many therapeutics is to reset such unbalances, or in other ways interact with proteins in the body.

Over the last decades, advancements in molecular biology have made it possible to produce, modify or design proteins in the laboratory – a field known as ‘protein engineering’. Today, technologies are available that enables the production of proteins using the information in an isolated DNA sequence as starting material and cell cultures of microbial, fungal, plant, insect or mammalian cells as production machinery [1]. Such proteins are called recombinant proteins.

Protein engineering and recombinant proteins are used in a variety of applications, including agriculture, food, materials science, biotechnology and medicine. Protein engineering has enabled the development of protein-based therapeutics, which is a rapidly growing class of therapeutics, with many advantages compared to traditional small-molecule therapeutics. This chapter provides an overview of the main protein engineering strategies, with focus on therapeutic applications.

1.1 PROTEINS IN THERAPEUTICS AND DIAGNOSTICS

Most therapeutics that are available today are based on small organic molecules. However, many disease conditions involve the complex and specific actions of proteins, which are often difficult to efficiently regulate using small chemical compounds. Protein therapeutics have a much more advanced structure, which enables more specific interaction with disease-related proteins in the body, and thereby generally leads to higher potency and fewer side effects [2].

1 Bacterial display systems for engineering of affinity proteins

Many protein therapeutics are simply recombinant copies of natural human proteins and are used as a sort of replacement therapy to restore the concentration of a missing, deficient or down-regulated protein. An example of this is the first recombinant protein therapeutic, recombinant human insulin [3]. Other protein therapeutics are so- called targeted therapeutics, aiming to interact with specific molecules in the body. Since the first production of recombinant human insulin in 1979, the field of protein therapeutics has increased dramatically. With the advancements in molecular cloning technology and site-directed mutagenesis [4], it has become possible to ‘engineer’ proteins to give them novel or improved designs and functions [5]. Protein engineering can be used for the development of therapeutic agents that are capable of binding with a high affinity and specificity to proteins involved in a disease. Proteins that are designed to bind to a ‘target’ protein are sometimes called affinity proteins [6].

A natural source of affinity proteins are antibodies, also known as immunoglobulins (Ig’s), which are mediators of the immune system (see chapter 2). Antibodies are constantly being produced in our bodies, and some are matured to specifically bind to different pathogens or foreign molecules [7]. Presently, antibodies are the fastest growing class of protein therapeutics [5]. In the last few decades, the potential of engineering the specificity of antibodies to target in principal any protein of interest has been realized [8].

Protein-based targeted therapeutics, such as antibodies, can be engineered to have a very high affinity and specificity for a drug target (the antigen), and thereby blocking its function, targeting it for destruction by the immune system or stimulating a signaling pathway [2]. In addition, they can be used for the specific delivery of a drug compound, toxin, or therapeutic radionuclide to a defined site in the body, such as a tumor [2, 9]. Affinity proteins can also be labeled with a reporter molecule for various diagnostic purposes, such as in vivo molecular imaging. For example, radionuclide- conjugated affinity proteins can be used to identify sites in the body expressing the target protein (for example a tumor-associated antigen) by detecting the local concentration of radioactivity [10].

A challenge of using engineered proteins as therapeutics is the risk of eliciting an immune response, since the protein might be recognized as ‘foreign’ by the immune system [11]. This can cause neutralization of the therapeutic, diminishing the therapeutic effect, or even result in potentially harmful side effects. Therefore, much effort is put into de-immunization strategies, which can also be achieved by protein engineering [11]. Other drawbacks compared to small molecule therapeutics are the

2 Filippa Fleetwood typically poorer chemical, proteolytic and thermal stability and the relatively complicated and costly production, particularly for larger and more complex proteins like antibodies [2, 12].

1.2 PROTEIN ENGINEERING BY RATIONAL DESIGN

One way of generating an affinity protein for a given target is by so-called rational design. Thorough analysis of the relationship between structure and function of the target and the affinity protein can be used to identify amino acid positions that can be mutated in order to increase (or introduce) functions such as affinity or catalytic activity (Fig. 1.1a). When such detailed understanding is available, this technology is relatively straightforward and inexpensive. Simple site-directed mutagenesis and recombinant DNA methods can be used to acquire the necessary mutations in order to alter the function of a protein [13, 14] or even design new proteins [15]. Rational design can also involve fusions or deletions of larger protein sequences, in order to add or remove functions [16-18].

In recent years, advancement in computational methods has enabled de novo design of protein-protein interaction complexes [19] as well as enzyme design [20]. Computational algorithms are used for finding an amino acid sequence that forms energetically favorable interactions with its target [21]. Commonly, the target structure is specified, and calculations attempt to identify an amino acid sequence that fits into the pre-defined fold at a global free-energy minimum [22]. Although these methods have generated promising results they are still facing many challenges, including the dependence on accurate prediction of protein structure and interaction with other molecules, as well as issues with protein stability [23].

Generally, re-design of natural binding sites (such as optimizing an existing affinity or grafting of a known binding sequence to achieve new specificities) is more easily achieved than de novo design, since the starting point is a binding site that has been evolved by nature [24]. The main challenge of protein engineering by rational design is that the complex nature of the relationship between protein secondary structure and amino acid sequence often makes it difficult to predict how point mutations will affect the overall function of the protein. Furthermore, knowledge about the effect of single residues or secondary and three-dimensional structure on the function of the protein is often not available.

3 Bacterial display systems for engineering of affinity proteins

Directed evolution Gene

Gene library

Instroduce random mutations

Protein library

Apply selection Isolate pressure genes Selected variants with improved function

Rational design Protein with known structure and function

Analyze structure/function Produce recombinant relationship to protein with mutations identify bene!cial ?? & verify experimentally mutations

Figure 1.1. Schematic overview of the engineering process by directed evolution and rational design. a) In directed evolution, a selection pressure is applied in order to isolate variants with a desired function from a protein library. The genes encoding the selected variants can be amplified and subjected to additional rounds of selection. b) In rational design, knowledge about the structure/function relationship is used to identify mutations that can lead to a desired function.

1.3 COMBINATORIAL PROTEIN ENGINEERING AND PROTEIN LIBRARIES

An alternative approach for the generation of proteins with novel functions is combinatorial protein engineering, also known as directed evolution. As the name implies, this strategy attempts to mimic the natural evolution of a protein function by introducing random mutations in combination with a selection pressure. A protein library is a large repertoire of mutated variants of a protein. In order to isolate a protein variant with a desired function, a selection pressure is applied. A commonly

4 Filippa Fleetwood desired function is a high affinity towards a target protein. The selection pressure can be introduced by giving all library members a chance to interact with the target protein, and then ‘fishing out’ the variants that bind, while discarding all other library members (Fig. 1.1b).

Generally, the chance of finding a suitable candidate with the desired properties increases using a larger library, and the affinity of selected antibodies have been found to be proportional to the library size [25]. In order to be able to cover large library sizes, high-throughput selection- or screening methods are necessary.

For amplification and downstream characterization of selected variants, each protein variant in a library (the phenotype) should be coupled to its corresponding genotype, i. e. the gene encoding that particular protein sequence [26]. Various selection technologies are available today, and some of the most established technologies are described in more detail in chapter 3. For example, in so-called display technologies the protein is linked to its corresponding oligonucleotide sequence by expression for example in fusion to a surface protein of a microorganism. The displayed protein library is allowed to interact with the target protein, and following selection of binders with the desired function, the genes encoding the selected variants can be recovered and amplified. The recovered genes can then be used either for sub-cloning and downstream characterization, or as starting material for additional selection rounds.

The genes encoding a protein library can be synthesized in the laboratory, or be derived from a natural source, such as an antibody repertoire from an animal [25, 27]. The antibody-encoding genes from a B cell repertoire can be isolated and sub-cloned into a selection platform. If the repertoire is derived from infected patients, or if the animal from which the B cells are recovered was recently immunized with the target protein, the repertoire is pre-enriched for antibodies specific for that protein. Such libraries are called ‘immune’ libraries. If the library is not pre-enriched for any specific binding (or function), it is called a ‘naïve’ library [25, 27].

Synthetic libraries can be generated either by random diversification of a template gene, for example by error-prone PCR, or site-specifically, by incorporation of randomized codons in certain positions in the sequence (site-directed mutagenesis). In error-prone PCR, a polymerase with a high mutation rate is used for amplification of the template gene [28]. Mutations are consequently randomly distributed throughout the sequence, which can potentially be a drawback. This method also suffers from the introduction of unwanted biases in the library (which might hamper the enrichment of rare high-affinity clones), since some amino acids are more likely to be incorporated

5 Bacterial display systems for engineering of affinity proteins than others. For example, some amino acids are encoded by several codons, whereas other amino acids only are encoded by one codon [29]. Therefore, a mutation from one amino acid to another sometimes requires mutations in more than one base pair. In addition, the introduction of stop codons and unwanted amino acids can not be avoided. However, important advantages of error-prone PCR are that it is rapid and cost-efficient.

Various approaches based on site-directed mutagenesis can be used for the randomization of specified positions or regions within a protein sequence. For example, randomized oligonucleotides can be introduced during gene synthesis. However, like error-prone PCR, many of these strategies suffer from unwanted biases [29]. Certain types of more controlled oligonucleotide-directed mutagenesis can be used in order to avoid the introduction of codon biases [30]. Examples of this include the split-and-mix strategy [31], enzymatic digestion and ligation of double-stranded DNA codons [32], or the use of trinucleotide phoshoramidities for synthesis of the library oligonucleotides [33, 34]. Advantages of these methods include the possibility of introducing higher relative amounts of certain amino acids at a given position. For example, unwanted codons such as stop codons, cysteines (forming disulphide bonds) or prolines (disrupting helical secondary structure) can be avoided, and the level of hydrophilic, hydrophobic or charged amino acids can be adjusted. Such approaches for randomization are useful for example for the generation of affinity maturation libraries, where certain amino acids have been identified as important for the binding to the target protein, and therefore can be randomized to a lower degree compared to other library positions (as exemplified in paper I of this thesis). Consensus sequences from multiple selected clones can also be used to identify which amino acids to introduce at higher levels in certain positions.

A drawback of many approaches for randomization is that the complexity of the library can be very large, which makes it difficult to cover the whole library using most selection technologies [30]. By using fewer randomization options for each mutation during site-directed mutagenesis, the resulting library size can be decreased. Either, a limited number of residues can be selected for randomization, or a larger number of residues can be randomized using a limited number of amino acids. The optimal balance between the number of randomized positions, the number of amino acids used for incorporation, and the degree of library coverage is difficult to determine, but could be important to consider during library design. Extreme examples of reduction of library complexity include the selections from so-called minimalistic synthetic libraries, investigated by Sidhu and colleagues, using only a few amino acids (e.g. tyrosine/serine or tyrosine/serine/alanine/aspartate) for

6 Filippa Fleetwood randomization in the CDRs of antibody fragments [35-37]. However, it is not clear whether this type of library is suitable for any type of target molecule or epitope, such as for example flexible peptides, small molecules or sugars [37]. It is also not evident whether this strategy would work for other scaffolds than antibodies.

Diversification of protein libraries can also be achieved using gene shuffling [38]. In this technology, DNA sequences of homologues proteins are randomly fragmented and reassembled. This can be used for example for further diversification of variants selected from a library, or using genes from different species as starting material for library generation [39, 40].

Rational design and directed evolution both have advantages and drawbacks, and a common approach is to use a combination of both; so-called semi-rational approaches. For example, structure-based knowledge can be used to identify a number of positions that can be mutated to create smaller “smart libraries” with a high likeliness of generating improved variants [41, 42] or computational design can be used for design of novel protein binders, followed by affinity maturation by experimental directed evolution [43].

7 Bacterial display systems for engineering of affinity proteins

2. AFFINITY PROTEINS

As mentioned in chapter 1, one typical aim of combinatorial protein engineering is to generate novel affinity proteins, i.e. protein variants that bind with a high affinity and specificity to another target molecule (typically a protein). Affinity proteins are used in wide range of applications within biotechnology, therapeutics and diagnostics. The most common types of affinity proteins are antibodies. Antibody therapeutics is a rapidly growing field, with presently over 40 antibody-based therapeutics approved or in review (and many more in clinical trials) http://www.antibodysociety.org/news/approved_mabs.php (141017). Although antibodies have proven useful as therapeutics for various conditions, there are some limitations to the antibody format. This had led to the development of binding proteins based on alternative formats, derived from antibodies as well as non- immunoglobulin proteins, all with their respective advantages and limitations, which make them suitable for different applications.

2.1 ANTIBODIES AND ANTIBODY FRAGMENTS

2.1.1. Antibody structure and properties

An antibody is a large, Y-shaped protein, consisting of one constant part, and two variable parts at the ‘arms’ of the molecules [7]. There are five different classes of antibodies, based on the sequence of their constant part: IgM, IgD, IgG, IgE and IgA. Of these classes, IgG is the most commonly used for engineering of therapeutic antibodies [44].

An IgG is a 150 kDa protein, that consists of two heavy chains and two light chains

(Fig. 2.1) [7]. The heavy chain contains three constant domains (CH1, CH2 and CH3) and one variable domain (VH). The light chain is shorter, consisting of a shorter constant (CL) and one variable domain (VL). The light chain associates with the CH1 and VH domains, and is connected by disulphide bonds between the CH1 and CL domains. The heavy chain forms homodimers, connected by disulphide bonds in the

‘hinge region’, between CH1 and CH2. The VH, VL, CH1 and CL constitute the so- called Fab region, and the CH2 and CH3 domains are called the Fc region. The VH and the VL domains form the variable region, called the Fv. Three loops in each of the VL and VH domains form the antigen binding sites of the antibody. These loops, also known as complementary determining regions (CDR), are highly variable and can adopt many possible conformations. The site on the antigen to which the antibody

8 Filippa Fleetwood binds is called an epitope, and the antigen-binding site on the antibody is called a paratope [7].

a b Antigen binding site

V H H V V L L V C H 1 1 H Fab C C L S L S S S S S C S S C 2 H H 2 C Fc C 3 H H 3 C

Fig. 2.1 a) Schematic representation of the structure of an IgG b) Structure of a mouse IgG2a (PDB entry 1IGT).

The Fc part of the antibody is responsible for various so-called effector functions. Upon binding of the Fc to Fc receptors (FcRs) on immune cells or other components of the immune system, processes leading to stimulation of the immune system to recognize and destroy the antigen can be initiated. These processes include antibody- dependent cell-mediated cytotoxicity (ADCC), complement-determined cytotoxicity (CDC) and phagocytosis [7].

In addition to mediating effector functions, the Fc portion is involved in prolonging the half-life of IgG in the serum. The Fc of IgG binds to the neonatal Fc receptor (FcRn) on a number of cell types in the body (mainly endothelial cells), and the antibody is thereby prevented from degradation in the endosomes and recycled back into the blood stream [45].

2.1.2. Antibody engineering

Antibodies are generated daily in the B cells of our immune system, and in response to exposure to pathogens and foreign molecules of different types, they are matured to develop different specificities. In order to obtain antibodies targeting a protein of interest, the protein can be injected into a mammal such as mouse, rabbit or goat. In

9 Bacterial display systems for engineering of affinity proteins response to the immunization, the immune system of the animal produces a number of different antibodies that are specific for the injected protein. These can be purified from the serum of the animal using for example the target protein for affinity chromatography. The recovered antibodies are a mixture of different antibodies, all binding to the same target protein, but not necessarily to the same epitope. This antibody mixture is called a ‘polyclonal’ antibody [7].

Since the target-specific antibodies obtained from an immunization of a laboratory animal are not possible to amplify or produce recombinantly, this strategy is laborious and time-consuming. Also, the risk of batch-to-batch variations could lead to problems for various applications. In 1975, the hybridoma technology was invented [46], which enabled the in vitro production of ‘monoclonal’ antibodies, i.e. homogenous antibody pools containing only one antibody variant recognizing a single epitope. The hybridoma cell line is a renewable source for antibody productions, and the genes encoding the monoclonal antibody can be recovered from the hybridoma cell.

Another beneficial strategy for generation of monoclonal antibodies is to use combinatorial protein engineering platforms such as phage display for selection from antibody libraries of either synthetic or natural origin (see sections 1.3 and 3.1). Advantages of in vitro selection for the generation of monoclonal antibodies compared to the hybridoma technology include the possibility to apply various types of engineering strategies and selection pressures, such as selection for high stability, multiple specificities or the elimination of cross-reactivity.

The full-length antibody has many valuable properties for therapeutic purposes. Nevertheless, in many biotechnological, therapeutic and diagnostic applications, a smaller molecule with a simpler structure can be advantageous. Therefore, several alternative antibody formats have been explored (Fig. 2.2) [47]. The most common types of antibody fragments used in various antibody engineering and therapeutic applications are the Fab fragment (The VH, VL, CH1 and CL domains) and the single- chain Fv (scFv; only the VL an the VH domains directly connected by a linker). If the

VL and the VH are fused by a linker that is too short for the correct assembly into an Fv, the scFvs are forced to dimerize. This format is called a diabody [48]. Like the full-length antibody or the Fab, the diabody format has the advantage of bivalency, which can give rise to an avidity effect and thereby a slower dissociation from multivalent targets. ScFvs, on the other hand, are smaller, which simplifies protein production and is advantageous for several applications, such as design of multi- domain constructs or display on the surface of a microorganism. Even smaller

10 Filippa Fleetwood antibody fragments, consisting of only the VH or the VL, have been used. These antibody fragments, also known as domain antibodies (dAbs) were initially associated with aggregation problems due to the hydrophobic surface which is normally facing the VL or the VH, respectively [49-51]. However, further research has generated VH as well as VL domains with highly improved stability [52-54]. Single-domain antibody fragments with a typically very high stability can also be derived from a type of antibodies lacking a light chain, found in camelids and sharks [55]. The variable domains of the heavy-chain antibodies (VHH) from camelids and sharks are also known as nanobodies or VNARs, respectively [56]. Nanobodies are an attractive antibody format for antibody engineering, and are described in the following section (section 2.2).

Fab scFv Diabody dAb (V H ) V H H H H H H L L L H V V V V V V V V H 1 L H L H C C V S S V

Fig. 2.2 Schematic representation of different antibody formats (a Fab fragment, an scFv, a diabody, a dAb, and a VHH).

In order to maintain the effector functions, an scFv or a diabody can be fused to an Fc fragment [57]. Promising classes of antibody therapeutics are the antibody-drug conjugates (ADCs) and immunotoxins, which are used to selectively deliver for example a drug, radionuclide or toxin to a tumor [58-60]. Three ADCs have been FDA approved (Mylotarg® (Pfizer), Adcetris® (Seattle Genetics) and Kadcyla® (Roche-Genentech) (although Mylotarg® has been redrawn from the market due to lack of improvement in overall survival [59]). A related group of antibody therapeutics are immunocytokines, engineered antibody-cytokine fusions that are used for selective delivery of a cytokine to the site of disease [61]. Bi-specific or multi- specific antibody constructs can be used for example for enhanced selectivity, or to direct the cytotoxic effects of T-cells to the tumor cells [62]. In addition to engineering of the antigen binding sites, the Fc part of the antibody can also be engineered for example for enhancement of effector functions or half-life extension [63-66].

11 Bacterial display systems for engineering of affinity proteins

One of the main problems associated with antibody-based therapeutics is immunogenicity [11]. Antibodies that are not of a human origin are likely to elicit an immune response. This has lead to the development of various humanization strategies, for example the grafting of CDRs from a mouse antibody to a scaffold of a human antibody, creating a so-called chimeric antibody [44]. Alternatively, fully human antibodies can be selected from synthetic libraries or transgenic animals. Although this decreases the risk of immunogenicity, it does not eliminate the problem completely [11, 67].

Another problem is the complicated recombinant production due to the complexity of the antibody structure. Due to post-translational modifications, there is usually a need for mammalian protein production systems, which are costly and time-consuming [68]. Other common problems include protein aggregation, as well as thermal and proteolytic instability [69].

2.2 NANOBODIES

As described in the previous section, for many biotechnological applications using affinity proteins, properties such as a small size and monomeric behavior are attractive. For such applications, alternative antibody formats such as the scFv are attractive. However, attempts to further reduce the size by using only the VH or the VL have proven difficult, due to aggregation and stability problems [49-51].

A solution to this problem can be found in the antibody repertoire of camelids and sharks. In addition to their normal antibodies, camelids and sharks have antibodies lacking the light chain [70, 71]. The variable domains (VHH) of the camelid heavy- chain antibodies are attractive for protein engineering purposes, as they are small (around 15 kDa), monomeric and generally have a very high solubility and stability [70]. The VHH, also known as nanobody, has been used in various therapeutic, diagnostic and biotechnological applications [72, 73]. Nanobodies can be efficiently expressed in microbial systems such as bacteria and yeast [74-76].

The lack of the light chain has been compensated for by mutations and structural adaptations that result in full functional diversity maintained in the VHH [55]. For example, the CDR3 of a nanobody is longer and more exposed than the CDR3 of a human VH [77, 78]. This unique structure is attractive for protein engineering purposes since it has the potential to interact with epitopes that are inaccessible to conventional antibodies, such as the active site of enzymes [79, 80].

12 Filippa Fleetwood

Cloning of B cell repertoires from an immunized camel, llama or dromedary is a straightforward way of creating a functional antibody library, since the full diversity of the whole antibody repertoire can be isolated using a single primer pair, and no pairing of the heavy and light chain is required. Since an affinity for the antigen has already been developed in the immunization step, selections from isolated immune nanobody libraries of around 106-107 variants usually generate binders of high affinity, without the need for affinity maturation [55]. However, although the immunization step is beneficial for generating binders with a high target-affinity, it is also possible to select target-specific nanobodies from naïve, synthetic or semisynthetic libraries [81-83].

Nanobodies have been generated for numerous therapeutic purposes, including for example toxin neutralization, oncology, immunology, inflammation, enzyme inhibition and antiviral therapy [84-86]. A number of therapeutic nanobodies are currently in clinical trials (www.ablynx.com). The small size and monomeric behavior of the nanobody make them ideal for design of multi-specific and multivalent constructs, as well as fusion proteins for extended half-life, added effector functions or the carrying of payloads [84, 86]. Because of the non-human origin, humanization is an important issue, and strategies have been developed for this purpose [87, 88].

In addition to their therapeutic potential, nanobodies have also been found to be well suited for in vivo imaging, including both nuclear and optical imaging [85]. Nanobodies have also been used for biotech applications such as immunoaffinity purification [89], biosensors [90], and in anti-dandruff shampoo [91].

2.3 ALTERNATIVE SCAFFOLD PROTEINS

2.3.1 General properties of alternative scaffold proteins

As mentioned in section 2.1, some drawbacks of the use of antibodies for certain applications are related to the large size, relatively low stability and complicated production of these macromolecules. Affinity proteins derived from non- immunoglobulin proteins, commonly known as ‘alternative scaffold proteins’ can overcome some of these problems [92]. In principle, any protein with a suitable structure and biophysical and biochemical properties could be subjected to randomized mutations to create a library intended for selection of novel target binding

13 Bacterial display systems for engineering of affinity proteins proteins. Commonly, a protein scaffold chosen for diversification is a small, monomeric protein with a low degree of post-translational modifications and stable fold, which is tolerant to diversification [93].

An advantage of many alternative scaffold proteins is the possibility to cost- efficiently produce the recombinant protein in bacteria or by chemical peptide synthesis due to a small, monomeric structure and lack of post-translational modifications [92, 94]. Several alternative scaffold sequences are cysteine-free, which facilitates thiol-based site-specific modifications by the introduction of a single free cysteine [95, 96]. The generally high chemical, thermal and proteolytic stability of alternative scaffold proteins also facilitates modifications like chemical labeling, or other processes involving harsh conditions [94, 95, 97, 98].

A small size is beneficial for tissue penetration, while unbound protein is also cleared faster from the body [99]. A fast clearance is desired in imaging applications, in order to obtain a high signal-to-background ratio, resulting in a high image contrast [100, 101].

The small size and single-domain structure of alternative scaffold proteins also facilitates further engineering. For therapeutic purposes, where a short half-life is generally unfavorable, half-life prolongation can be achieved through various engineering strategies [102]. Hence, alternative scaffolds can usually easily be modified, for example by genetic fusion to a protein or peptide that binds to a protein with a long in vivo half-life, such as albumin [97, 102, 103]. Similarly, the affinity protein can be fused to an Fc molecule in order to gain effector functions as well as increased serum half-life [5, 104]. The size, stability and monomeric structure of alternative scaffold proteins also facilitate engineering of multivalent or multi-specific constructs. For example, affinity proteins targeting two different antigens can be fused together in order to increase specificity, or simply target two antigens with one therapeutic agent [18, 105]. The fusion of two identical binding domains into a bivalent construct can be used in order to obtain a higher apparent affinity due to avidity effects [106]. Alternatively, targeting of multiple epitopes on the same antigen can result in avidity effects or even enforce a structural conformation change [107, 108]. An example of the generation and characterization of such ‘biparatopic’ binders is described in papers (I) and (II) of this thesis.

An additional potential advantage of the small size and high chemical and proteolytic stability of many alternative scaffold proteins is the possibility of using alternative delivery routes, such as subcutaneous injection or inhalation [97, 104].

14 Filippa Fleetwood

Unlike monoclonal antibodies, alternative scaffold proteins cannot be generated by immunization or hybridoma technology. Instead, selections from synthetic libraries are typically performed, using various selection systems (see chapter 3).

Today, a large number of scaffolds have been investigated for engineering of novel binding proteins [97, 109, 110]. It should be noted that all scaffolds have different advantages as well as limitations, and the optimal choice of scaffold might vary depending on the application. Some of the most investigated alternative scaffolds are described in the following sections, and an overview of their structures is presented in Fig. 2.3.

DARPin Adnectin Anticalin

Cystine knot miniprotein A!body

Fig. 2.3 The structures of a few alternative scaffold proteins. a) a DARPin (PDB entry 4DUI), b) an adnectin (PBD entry 3QWQ), c) an anticalin (PBD entry 1LNM), d) a cystine-knot miniprotein (PBD entry 1MR0), e) an Affibody molecule (PDB entry 2B88).

2.3.2 DARPins

Ankyrin repeat proteins are involved in protein-protein interactions in a variety of species [111]. The Designed Ankyrin Repeat Protein (DARPin) scaffold was designed by structure and sequence consensus analysis of a large number of ankyrin repeat proteins [112]. The resulting scaffold is a 33 amino acid domain, which is used as a building block to produce a multimeric repeat protein. Typically, a DARPin consists of 2-3 repeats containing randomized positions, sealed at both ends with stabilizing “capping repeats”, which have a hydrophobic side pointing towards the internal repeats, and are hydrophilic on the outside [113]. DARPins have been selected against

15 Bacterial display systems for engineering of affinity proteins a broad range of different targets, often generating highly stable and well-expressed binders [114].

2.3.3 Adnectins

The adnectin scaffold (also known as monobody) is a single domain protein consisting of 94 residues, derived from the human tenth fibronectin domain (10Fn3) [115]. It has the structure of a β-sandwich, much resembling the Ig-fold, though it consists of a single domain and contains no cysteines [115]. Diversification for library construction is usually performed in three of the loops connecting the β-sheets [116]. The highest-affinity binder obtained from an adnectin library has an affinity of approximately 1 pM [117]. The human origin makes the scaffold attractive for therapeutic applications, since this decreases the risk for problems related to immunogenicity. The most successful adnectin so far, CT-322, targeting VEGFR2, has shown promising results in phase I clinical trials and is currently in phase II clinical trials [118].

2.3.4 Anticalins

Lipocalins are another class of single-domain proteins with similarities to the Ig fold, from which a group of alternative scaffold proteins called anticalins have been derived [119, 120]. The anticalin structure is a β-barrel, composed of eight antiparallel β-strands connected by loops, which resembles the antigen-binding site of antibodies. These loops are randomized for library purposes [119]. Advantageous features of the anticalins include the unusually high thermal stability, with melting temperatures sometimes above 70°C [121]. Anticalins are <20 kDa in size [119]. Binders of down to picomolar affinities have been generated for various different targets [122]. The deep ligand-binding pocket structure of the anticalins make them particularly suitable for binding to haptens and other small molecules, in addition to larger proteins [122]. Anticalins have been derived from lipocalins from a number of different species including human lipocalins, which have been used for generation of several adnectins targeting therapeutically relevant proteins [123]. A first-in-human Phase I study was recently performed using the VEGF-binding anticalin PRS-050 [124].

2.3.5 Cystine-knot miniproteins

The cystine-knot miniproteins (also known as knottins) are a group of proteins of 30- 50 amino acids, consisting of a core of antiparallel β-strands stabilized by at least three disulphide bridges [125]. Due to this constrained structure, knottins generally

16 Filippa Fleetwood have an extraordinary high thermal, chemical and proteolytic stability, which make them very attractive for in vivo applications and alternative administration routes including oral application [125, 126]. Knottins exist in a variety of organisms, and carry out many different functions [127]. Engineering of novel binding specificity into a knottin can be done by grafting of known target-binding peptide sequences into the loops [128], or by screening of knottin libraries [129]. Several knottins have been generated for therapeutic purposes, and many have also shown promising results in in vivo imaging studies [125].

A number of other cysteine-rich alternative scaffold proteins have been used for engineering of novel affinity proteins, including avimers [130], kringle domains [131] and kunitz domains [132].

2.3.6 Affibody molecules

Much of the work presented in this thesis involves a type of alternative scaffold proteins called Affibody molecules. The Affibody molecule is derived from staphylococcal protein A [133]. The B domain, like the other four homologous domains of staphylococcal protein A, is involved in the binding to IgG [134]. This domain has been engineered in order to increase the chemical stability, and called the Z domain [135].

The Z domain is a three-helix bundle protein consisting of 58 amino acids, only around 7 kDa in size [135]. Two of the three helices (helix 1 and 2) are involved in the binding to IgG [135]. In order to create an Affibody library, 13 positions in these two helices are typically randomized in order to create an Affibody library [133] (Fig. 2.4). When the composition of amino acids used for randomization is not skewed for binding to a particular target (as in an affinity maturation library), the library is often called a “naïve” Affibody library. The third helix is not randomized, but its presence has been found to be important for the stability of the Affibody molecule [136, 137].

17 Bacterial display systems for engineering of affinity proteins

24 25 18 28 27 17

14 32 13 11 9 10 35

Figure 2.4 Affibody structure and randomized positions (PDB entry 2B88).

Affibody libraries are typically created by chemical oligonucleotide synthesis [94]. Error-prone PCR can also be used [138], although this leads to randomization in other positions than only the thirteen intended library positions. Various selection systems have been investigated for selections of Affibody molecules, including phage display [139], staphylococcal display [140], ribosome display [141], microbead display [142], protein complementation assay (PCA) [143] and E. coli-display (see paper IV of this thesis). Some of these platforms and their respective advantages and limitations are described in chapter 3.

The B-domain of staphylococcal protein A has a high stability and solubility as well as extremely fast folding kinetics [144, 145]. This high stability and the three-helical structure is usually maintained in Affibody molecules selected for binding to different targets [94]. Affibody molecules binding to a number of targets have been selected with affinities down to the picomolar range [94, 146, 147].

Affibody molecules are well suited for in vivo imaging purposes, due to their high stability and target affinity along with the small size and the absence of cysteines [100]. Promising results have been achieved using for example HER2- and EGFR- specific Affibody molecules labeled with different radionuclides for in vivo imaging [148-151]. Recently, a HER2-binding Affibody molecule has also shown promising results in imaging of HER2-expressing breast cancer tumors and metastases in humans [152].

Affibody molecules are also promising agents for therapeutics purposes. Affibody molecules for blocking of protein-protein interactions have been developed [147], and promising results have been achieved using Affibody molecules equipped with a payload [151, 153, 154]. An Affibody molecule fused to a truncated version of

18 Filippa Fleetwood

Pseudomonas Endotoxin A demonstrated cytotoxic effects in vitro [153]. A 186Re- maSGS-Affibody conjugate for potential systemic therapy [151], and a 177Lu-labeled Affibody molecule fused to an albumin binding domain (ABD) [154] have also shown promising results. Other studies involving Affibody molecules for therapeutic applications include redirection of adenoviral particles [155] and nanoparticle- or liposome-based drug delivery [156, 157].

In addition to therapeutic and diagnostic applications, Affibody molecules have been investigated as biotechnological tools, for applications such as affinity chromatography [158, 159], biosensors [160] and other fluorescence-based assays [161, 162], among several other applications [94].

19 Bacterial display systems for engineering of affinity proteins

3. SELECTION SYSTEMS

In order to select candidates with a desired function from combinatorial protein libraries it is important to use a selection technology that is suitable for the application. For example, for large libraries, the selection platform should ideally allow for sampling of as many library members as possible to avoid the risk of losing valuable library members. If the purpose is affinity maturation, a selection technology that can discriminate between high- and intermediate affinity binders should be used. For improvement of thermal or proteolytic stability, the platform should preferably not be affected by the conditions. Some of the most commonly used technologies for selection of affinity proteins, with an emphasis on display technologies, are described in this chapter.

3.1 PHAGE DISPLAY

Bacteriophages are viruses that infect bacteria [163]. The types of bacteriophages that have gained the most interest for biotechnological purposes are the M13 and fd phage, which both belong to the Ff class of filamentous phages [164]. A filamentous phage is tube-shaped, around 6 nm in diameter and 1 µm long [164]. All the genetic information encoding a phage particle is present inside the phage. In 1985, George Smith showed that a peptide of interest could be genetically fused to a phage coat protein, and thereby be ‘displayed’ on the phage surface [165]. In the following years, the technology was further developed for the display of peptide and protein libraries [164, 166-168].

Phage-displayed antibody libraries can be ‘panned’ for example against immobilized or biotinylated antigens or against antigens on cell surfaces [164, 169]. Typically, after incubation of the phage library with the immobilized target protein, the phage particles expressing a library member with an affinity for the target are captured, while non-binding phage are washed away. Subsequent elution of bound phage can be performed in several ways. Commonly, disruption of the protein-protein interaction is achieved by for example a change in pH. Eluted phages are used to infect E. coli cells for amplification, and can be recovered and used for subsequent panning rounds. A typical phage panning principle is illustrated in Fig. 3.1a.

Several formats for display on Ff filamentous phage have been investigated, for example by fusing the recombinant peptide or protein to different coat proteins. The most commonly used coat proteins are pIII and pVIII [164]. The minor coat protein

20 Filippa Fleetwood pIII is present in only five copies at the tip of the phage, whereas the major coat protein pVIII accounts for the majority of the phage mass, with around 2700 copies expressed [163, 164]. Fusion to pIII or pVIII results in multivalent display, which might give rise to avidity effects when screening against an immobilized target, which can be advantageous for distinguishing binders from non-binders, but can also impede the isolation of high-affinity binders [25]. Monovalent display on pIII has been enabled through the development of so-called phagemid systems [170-173], which have become the most commonly used type of phage display system [174].

Today, phage display is widely used for selection from protein and peptide libraries, and several monoclonal antibodies generated by this technology are now in clinical trials or FDA approved [175](www.clinicaltrials.gov, http://www.antibodysociety.org/news/approved_mabs.php (141017)). In addition to the display of peptides and scFvs, it has been used for example for display of Fab’s [170] and full-length IgG’s [176] and various other affinity proteins [75, 94, 116, 177]. By adding different types of selection pressure, phage display can be used for combinatorial engineering of various other functions in addition to specific affinity. Examples include selections for high thermal and proteolytic stability [178, 179], introduction of new specificity [180], evolution of enzymatic activity [181] or selection of tissue-specific peptides by in vivo panning [182].

Phage display is a robust and versatile technology, and no expensive or complex laboratory equipment is required. Large libraries can be displayed (library sizes of >1012 are possible to reach using multiple electroporations [183]) and automation is possible. A drawback is that the selection process is relatively time-consuming, as several panning rounds are usually needed. Another challenge is that the display efficiency can be influenced by properties such as protein size [184]. A limitation of the typical ‘capture-and-elution’ principle of phage display is the risk of using elution conditions that are not strong enough to elute library members with a very high target affinity. An increased stringency usually decreases the yield, and the elution conditions must therefore be carefully balanced in order to recover high-affinity variants [163]. An interesting strategy that allows for very stringent wash conditions in order to remove non-specific binding and thereby decrease the number of required selection rounds, was presented earlier this year [185]. This strategy uses the selective biotinylation of phage particles that are in close proximity (i.e. bound) to the antigen, through the so-called triple catalytic reporter deposition approach, and rescuing of the biotinylated phage particles using streptavidin coated beads in order to withstand the stringent wash conditions [185].

21 Bacterial display systems for engineering of affinity proteins

Although phage display is a very well-established and widely used display technology today, limitations for certain purposes have lead to the development of other display systems, which are described in the following sections.

Panning a

Washing Library displayed on phage

Immobilized target protein

Phage display

Elution Recovery of phage for additional panning rounds

E. coli infection and ampli!cation

b Incubation with "uorescently labeled target protein

Library displayed on cells FACS for isolation of highly "ourescent cells

Cell display

Ampli!cation

Additional sorting rounds

Figure 3.1 Overview of the selection procedure using phage display (a) and cell display (b).

3.2. CELL SURFACE DISPLAY

In addition to phage display, various microbial display technologies based on the display on the surface of cells have been developed [186]. A major advantage of using whole cells for display of combinatorial protein libraries is the possibility to isolate

22 Filippa Fleetwood target-binding variants using fluorescence-activated cell sorting (FACS) [187]. This is possible due to the cell size, which is large enough for detection in a flow cytometer, and the multivalent display of recombinant protein on the surface. The multivalent display can also be used to improve the ability to discriminate between high and low affinity binders, since a cell displaying a high-affinity binder will have a higher proportion of target protein bound to it. This quantitative affinity discrimination can be further improved by normalizing the target-binding signal against the total surface expression level, which can be monitored using a reporter tag [188, 189]. FACS- based screening provides a means for real-time visualization of the selection process. The possibility to use different stringency settings and normalized gating in the flow cytometer is beneficial for the isolation of high-affinity variants. In addition, selection of clones with multiple specificities can be performed using multi-parameter sorting [190]. An additional advantage is the possibility to perform on-cell assays such as affinity determination using flow-cytometry [191].

In display systems such as phage display and ribosomal display, a ‘capture-and- elution’ principle is used for selection of target binding library members (see section 3.1 and 3.3). In order to avoid loosing library members with a very high target affinity, finding the right elution conditions is important [163]. Contrastingly, in cell surface display, cells displaying a protein library are typically incubated with soluble target protein, which can either be directly fluorescently labeled or detected using fluorescently labeled secondary reagent. Cells expressing high-affinity clones should in principle give rise to a higher fluorescent signal, and can be sorted out using FACS, and amplified prior to the subsequent sorting round (Fig. 3.1 b).

The first example of cell surface display in combination with flow cytometry was the display of scFvs on E. coli in 1993 [192]. Several bacterial display systems exploiting primarily Gram-negative but also Gram-positive bacteria have since then been used for successful display of single recombinant proteins as well as combinatorial protein libraries [187]. Bacterial display is described in sections 3.2.3-3.2.4. An overview of some of the display systems discussed in this chapter is provided in Fig. 3.2.

3.2.1 Yeast display

Although bacterial surface display has been used in many successful studies, the most established cell display system today is yeast display. In yeast display, the recombinant protein is typically displayed in fusion to the aga2p subunit of the agglutinin receptor on the surface of Saccharomyces cerevisiae, first described Boder and Wittrup in 1997 [193]. A typical display construct can also include an N-terminal

23 Bacterial display systems for engineering of affinity proteins hemaglutinin tag for normalization of expression level in the flow cytometer, and a C- terminal c-myc tag which can be used for either detection of correctly displayed full- length proteins or normalization purposes [194].

A typical yeast library contains 108 – 109 variants, although library sizes of 1010 have been reported [195]. As for all display systems exploiting the use of cells, the limitation in library size depends on the efficiency of transformation of recombinant DNA to the cells. Systems based on yeast and Gram-positive bacteria are generally more limited in this respect compared to technologies such as phage display, E. coli display and non-display platforms.

A major advantage of yeast display for antibody engineering is the possibility to display proteins containing post-translational modifications such as N-linked glycosylations, due to the eukaryotic host [196]. Yeast display has for example proved very successful for affinity maturations. The highest affinity reported for an scFv was achieved using yeast display in combination with error-prone PCR for affinity maturation of a fluorescein-binding scFv, which resulted in an equilibrium constant of around 50 fM [197]. An additional advantage of the use of a eukaryotic host is the quality control machinery in the endoplasmic reticulum, which leads to a lower percentage of unstable or incorrectly folded protein variants displayed on the yeast surface, and thereby promotes the selection of well-behaved, stable variants [198].

In an interesting study by Burton and colleagues, the selection from an scFv library displayed on yeast was directly compared to the selection from the same library displayed on phage [199]. A broader repertoire of binders were selected using yeast display. However, the affinities for the target protein were similar in both output repertoires. Presumably, the differences in output repertoire could be due to the use of one eukaryotic and one prokaryotic host. By contrast, paper III of this thesis describes how the parallel selections from the same nanobody library on phage and on bacteria, which are both microbial hosts, generated higher target affinities on average among the clones selected by bacterial display.

3.2.2 Mammalian display

In order to even further facilitate the display of complex mammalian proteins such as antibodies, display systems based on mammalian cells have been developed and used for display of scFvs as well as full-length IgGs [200, 201]. Several mammalian display systems have been described, including systems based on transiently or stably

24 Filippa Fleetwood transfected cell lines as well as B cells retrieved from human donors [200, 202-204]. Selections from mammalian display systems using a hypermutating B lymphoma cell line [205] or using activation-induced cytidine deaminase (AID) for introduction of somatic hypermutations between sorting rounds [206] are elegant ways of mimicking the natural antibody maturation process.

c-Myc a b c d HA ABD Aga2p ABD Aga1p XM pIII

S. cerevisiae B cell S. carnosus

Phage display Yeast display Mammalian Staphylococcal display display

e f

E. coli display E. coli E. coli

g h

P Ribosome

Ribosome mRNA display display

Fig. 3.2 Schematic overview of the display mechanism in different display technologies: a) phage display, b) yeast display, c) mammalian display, d) staphylococcal display, e) E. coli display on the inner membrane, f) E. coli display on the outer membrane, g) ribosome display, h) mRNA display.

3.2.3 Staphylococcal display

Although the vast majority of bacterial display systems utilize Gram-negative hosts, a few examples of display of recombinant proteins on different Gram-positive bacteria have also been described [207, 208].

25 Bacterial display systems for engineering of affinity proteins

Gram-positive bacteria are attractive for display purposes for several reasons. In contrast to Gram-negative bacteria, Gram-positive bacteria only have one cell membrane, making translocation of a recombinant protein to the cell surface less complicated [187]. Also, the relatively thick layer of peptidoglycan present in the Gram-positive bacterial cell wall make the cells robust and highly tolerant to harsh treatment such as flow-cytometric cell sorting [209].

Gram-positive bacterial display for combinatorial protein engineering applications has been much investigated using Staphylococcus carnosus [147, 209-212]. It is a non- pathogenic bacterium that has been used extensively for example in the food industry [213, 214]. In addition to the general advantageous properties related to the Gram- positive cell wall structure, its low proteolytic activity make it suitable for display purposes [213]. Since it grows mainly as single cells of 0.5-1.5 µm, it is also very suitable for analysis and sorting using flow cytometry [209].

S. carnosus was first used for display of recombinant proteins in 1995 [215]. The display mechanism is based on a lipase gene from the related species Staphylococcus hyicus [215]. A promoter, signal peptide and a propeptide from this lipase gene are included in the display vector. Staphylococcal as well as E. coli origins of replication, and two different antibiotics resistance genes enables cloning work using E. coli and transfer to S. carnosus for surface display. The display vector also includes an albumin-binding protein (ABP) [216] gene for normalization of variations in surface expression level in flow-cytometric applications [188]. Recently, the ABP gene was replaced by two engineered albumin-binding domains engineered to a femtomolar affinity for human serum albumin [217] (Jonsson et al., personal communication), which facilitates long sortings and off-rate based selections. For the possibility to release soluble affinity protein into the supernatant, intended for a straightforward small scale production and purification, a 3C protease site is included between the ABD normalization tag and the affinity protein (Z) [218]. A TEV protease site is included upstream of the affinity protein, for the possibility to cleave off the propeptide (Jonsson et al., personal communication). A schematic overview of the staphylococcal display vector is shown in Fig. 3.3.

26 Filippa Fleetwood

a b Target protein Z

S PP ABD ABD Z XM HSA ABD

HSA r ABD Display vector Bla Cml pSCZ1 XM

OriS OriE S. carnosus

Fig. 3.3 A bacterial display system based on S. carnosus. a) The staphylococcal display vector. b) Schematic representation of the display construct on the staphylococcal surface.

Staphylococcal display has been investigated for a number of various applications, including delivery of vaccine [219-223], display of metal-binding peptides for applications such as water purification [224, 225], and display of enzymes [226]. Today however, the system is mainly used for combinatorial protein engineering and epitope mapping purposes [209]. The first selection from a combinatorial protein library was the selection of Affibody molecules targeting TNF-α [140].

A challenge for the staphylococcal display system is the limited transformation frequency of Gram-positive bacteria, owing to the thick peptidoglycan wall [209, 227]. In order to increase the transformation frequency, optimized transformation protocols have been developed, including for example a heat treatment step [227]. Following these protocols, staphylococcal-displayed library sizes of 107-108 transformants can typically be obtained in a straight-forward manner. In order to enable the use of larger libraries as starting material, a pre-enrichment step using for example phage display can be performed, before the library is transformed into staphylococci [140, 228]. Staphylococcal display has proven to be suitable for display of affinity maturation libraries, as exemplified in paper I of this thesis [147, 210]. Such libraries are based on a protein variant which has already demonstrated binding to the target protein, and the libraries have therefore already been “skewed” for target binding. Since the aim of an affinity maturation effort is to generate binders with a very high affinity, such selections particularly benefit from the quantitative affinity discrimination using FACS. Recently, the staphylococcal display system was also

27 Bacterial display systems for engineering of affinity proteins used for display of a naïve Affibody library of >109 variants (Jonsson et al., personal communication). Since the maximum sorting speed of the flow cytometer limits the library size that can be used, a pre-enrichment step using magnetic-activated cell sorting (MACS) was used in order to downsize the library before FACS (Jonsson et al., personal communication).

Previously, the combinatorial protein libraries displayed on staphylococci have been libraries of small alternative binding scaffolds such as Affibody molecules and an albumin-binding bi-specific domain [209, 210]. In paper III of this thesis, the first example of the display of an immunoglobulin-based library on staphylococci is presented.

3.2.4 E. coli display

As discussed earlier, the library size is important for successful selection from combinatorial libraries (see chapter 1). A major challenge of cell-based display systems is to reach a high number of transformants. For yeast and staphylococcal display, creating cell-displayed libraries that are bigger than 108-109 variants is challenging. Phage libraries are generally larger than yeast libraries, since the limiting factor is the propagation in E. coli, which has a higher transformation frequency. Libraries of >1012 can be created by multiple electroporations [183]. Consequently, the use of E. coli surface display for similar library sizes should be possible.

The combination of large surface-displayed libraries and the possibility to use flow cytometry for screening make E. coli an attractive host for combinatorial protein engineering purposes. The rapid growth rate [229] and the many available cloning and protein expression technologies (strains, promoters etc.) are additional advantages.

Various platforms for display of recombinant proteins on the surface of E. coli have been described [229-231]. Some of the most commonly used display systems, with focus on combinatorial protein engineering applications, are described in this section and the next, and a schematic overview of some different display mechanisms used for combinatorial protein engineering of affinity proteins is presented in Fig. 3.4.

28 Filippa Fleetwood

a b c

Lpp 1-9 OmpA’ 46-159 OmpA Circularly permutated OmpX

Lpp-OmpA OmpA eCPX

d e f

Autotransporter β-domain NlpA 1-6 NlpA 1-6 ZZ

APEx E-clonal Autotransporter

Fig. 3.4 Schematic overview of the display mechanism of a few different E. coli display systems used for combinatorial protein engineering of affinity proteins. a) Fusion to the C-terminal of Lpp- OmpA [192], b) Insertion in to a surface-exposed loop of OmpA [232], c) eCPX, C- or N-terminal fusion to an engineered version of circularly permutated OmpX [233], d) APEx, N-terminal fusion to a six amino acid sequence derived from NlpA [234], e) E-clonal, display of a dimeric Z-domain as an N- terminal fusion to a six amino acid sequence derived from NlpA, and capture of the full-length IgG by the dimeric Z-domain [235], f) display in fusion to an autotransporter [236].

In 1993, Francisco et al provided the first example of E. coli cell surface display in combination with flow-cytometric sorting [192]. ScFvs were fused to the C-terminal of Lpp-OmpA a fusion protein derived from E. coli lipoprotein A and Outer membrane protease A and a target-specific scFv was selectively enriched from a background of non-target binding cells using FACS. A few years later, the same system was used for selections from an affinity maturation library of scFvs [237]. Display of proteins and peptides using Lpp-OmpA has been used for several additional purposes [128, 238, 239]. However, drawbacks include leakiness of the outer membrane and cellular toxicity [238, 240].

The insertion of libraries into a surface-exposed loop of OmpA and other OMPs has also been demonstrated [232, 241-243]. However, a limitation of most OMP-based display systems is that the insertion of large protein sequences in a loop can be problematic [231, 244]. Also, the selection of target-binding peptides displayed in a loop conformation can result in a decrease in target affinity when expressing soluble peptides in the absence of the loop conformational constraint [244].

In order to overcome the problems associated with selections from libraries of peptides displayed in a loop conformation, Rice et al used circularly permutated

29 Bacterial display systems for engineering of affinity proteins versions of Outer membrane protein X (OmpX) for display of peptide libraries as C- or N-terminal fusions [244]. This display system, denoted CPX, and the improved engineered version denoted enhanced CPX (eCPX [233]), have proved to be a successful tool for selection of high-affinity binders from peptide libraries [245-250], probably due to a closer resemblance of a terminally fused peptide to the free soluble form.

An alternative strategy is the display on the periplasmic side of the inner membrane of E. coli. The first example of this approach was the development of the Anchored periplasmic expression (APEx) system in 2004 [234], which was used for the affinity maturation of scFvs. Using the APEx system, the recombinant protein can be displayed either as an N-terminal fusion to the six first amino acids of the E. coli lipoprotein NlpA, or as a C-terminal fusion to the M13 phage pIII coat protein [234]. Incubation with fluorescently labeled target protein for flow cytometry requires disruption of the outer membrane, which prevents further amplification by cell growth. The genes encoding the isolated clones must be recovered by PCR, sub- cloned and re-transformed for further sorting rounds, which is time-consuming and laborious. On the other hand, the APEx system has advantages such as the possibility to express soluble antigen in fusion with a fluorescent protein in the periplasm. After disruption of the outer membrane and subsequent washing steps, only cells expressing a target-binding protein will be fluorescent, and can be sorted by flow cytometry. This application can be useful for example for studies of protein interaction pairs.

An advantage of display on the inner membrane is the possibility to display full- length antibodies, as demonstrated by Mazor et al in 2007 [235]. The soluble full length IgG is produced in the periplasm, and is captured and tethered to the inner membrane by a dimeric version of the Fc-binding Z domain fused to the lipoprotein NlpA, displayed on the inner membrane. The very slow dissociation of Fc from the dimeric Z domain ensures that the IgG remains captured on the inner membrane even after disruption of the outer membrane [235].

For successful display on the outer surface of E. coli, the recombinant protein must cross two membranes, avoid folding of the peptide during the transportation to the surface and finally integrating into the outer membrane. Autotransporters are a family of surface-expressed proteins in Gram-negative bacteria, that contain all the information needed for translocation through both membranes in a single gene [251]. Surface display by fusing the protein of interest to various autotransporters has been much investigated for different purposes, and recently also for combinatorial protein

30 Filippa Fleetwood engineering applications [236, 252]. The use of autotransporters for library purposes is further discussed in section 3.2.6.

Another group of proteins that has shown great potential for surface display on bacteria are the intimins and invasins, which show many similarities to autotransporters [253-255]. The intimin EaeA has been successfully used for example for display of passenger proteins of varying size and structure [256], screening of an epitope mapping peptide library [257] and FACS enrichment from a mock library of a 1:200 000 ratio of target-binding cells spiked in a background of negative cells [258] as well as MACS selection from an immune single-domain antibody library [252].

3.2.4.1 Autotransporters

The autotransporter system is a naturally occurring way of secreting proteins to the outer cell surface of Gram-negative bacteria [259]. The classical autotransporters are encoded by a single gene, which makes them very attractive for biotechnological applications [230, 231].

Another attractive property of autotransporters for recombinant surface display is the efficient mechanism for transport of a passenger protein through both membranes. Yet another advantage is the high expression level [230], which facilitates discrimination between high-affinity and intermediate- or low-affinity binders due to a higher signal-to-background ratio in the flow cytometer. Since the passenger protein is C-terminally anchored to the outer membrane, display of large proteins is more efficient than for passenger proteins inserted in a loop [231].

Two of the most investigated autotransporters are AIDA-I (adhesion involved in diffuse adherence) and IgA protease from Neisseria gonorrhea [230]. The IgA protease was the first autotransporter to be discovered and characterized [260, 261], and AIDA-I was discovered a few years later [262]. Today, the autotransporter family comprises more than 1000 proteins with a variety of different functions [251, 263], and it is believed that autotransporters exist in all Gram-negative bacteria [263].

The classical autotransporter proteins consist of three parts: a leader sequence at the amino-terminus, a β-domain (also known as translocator domain) at the carboxy- terminus, and the passenger protein in the middle [263]. The leader peptide transports the protein in an unfolded state through the inner membrane to the periplasm. This transportation is mediated by the so-called sec system, which translocates unfolded proteins across the cytoplasmic membrane by the help of a protein complex know as

31 Bacterial display systems for engineering of affinity proteins the sec translocase [264]. After the passage of the secretory protein, the leader peptide is cleaved off. The protein probably remains at least partially unfolded in the periplasm, in a complex with periplasmic chaperones [265]. Translocation through the outer membrane occurs by insertion of the β-domain into the outer membrane where it forms a β-barrel structure, a process mediated by the β-barrel assembly machinery (BAM) [265-267]. The β-barrel acts as a pore and is amphipathic: hydrophobic side chains on the outside, in contact with the lipid bilayer, and hydrophilic on the inside, creating an aqueous environment where the passenger protein can pass through [265]. After secretion, the passenger protein is typically cleaved off, and is either released from the surface or stays connected to the β-domain via non-covalent interactions [265].

Various autotransporters have been used for the display of recombinant protein passengers of a wide range of sizes and functions [230, 268]. These include for example scFvs [269], metal-binding proteins [270], toxins, [271, 272], enzyme inhibitors [273] and enzymes [274, 275]. One of the largest proteins displayed using an autotransporter is the 613 amino acid Serratia marcescence lipase, which was displayed on the surface of E. coli using the autotransporter EstA [268, 276].

Although expression of an autotransporter in a bacterium other than its original host is possible, incompatibility problems have been observed [277]. In that respect, the AIDA-I autotransporter has potential biotechnological advantages, as it originates from E. coli [262]. AIDA-I has previously been used for display of for example toxins [271] enzymes, [263, 275, 278, 279], and small libraries for engineering of enzyme inhibitors [280, 281]. To date, AIDA-I has not been used for combinatorial protein engineering of affinity proteins using large libraries. Binder et al. investigated AIDA- I along with a few other autotransporters for the display of anticalins with the purpose of using one of the investigated autotransporters for display of an anticalin library [236]. However, AIDA-I was not the most suitable candidate under the conditions tested in that study.

Binder et al. used the EspP autotransporter for the display of an anticalin library [236]. This was the first example of display of a combinatorial protein library for engineering of affinity using an autotransporter. A few years later, selections from a nanobody library were succesfully preformed using the autotransporter EhaA [252].

The main challenges using autotransporters for recombinant surface display are related to the transportation of the passenger protein through the β-barrel. Since a folded structure can be too large to pass through the narrow pore, display of

32 Filippa Fleetwood disulphide-containing protein can be problematic [272, 282, 283]. However, this problem can sometimes be overcome by using a host strain lacking the enzyme DsbA, preventing disulphide bonds to be formed in the periplasm [283]. In contrast to this hypothesis, successful display of disulphide containing proteins has also been achieved in bacterial strains capable of forming disulphide bonds in the periplasm [269, 284].

3.3. CELL-FREE DISPLAY AND NON-DISPLAY SELECTION PLATFORMS

As described in chapter 1, the chances of finding a protein variant with a desired property within a library increases proportionally to the library size [25]. Display systems depending on a host organism are limited by the efficiency of the transformation (or transfection) of DNA. For applications where a large library is necessary, alternatives based on cell-free methods are attractive.

The first cell-free display system was ribosome display, which was described in 1994. A peptide library was ‘displayed’ on ribosomes from an E. coli lysate, forming a complex with the corresponding mRNA sequences due to the lack of a stop codon [285]. After panning against an immobilized target, the mRNA could be released by the addition of EDTA, copied into cDNA and amplified using PCR. A few years later, the technology was further developed to enable the display and selection from scFv libraries [286]. Ribosome display has also been successfully used for the display of alternative scaffold proteins such as DARPins and Affibody molecules [141, 287].

A similar cell-free display technology is mRNA display. Puromycin coupled to a DNA primer leads to the formation of a covalent link between the displayed protein and the mRNA, and before selection, the mRNA can be stabilized by the synthesis of a complementary DNA strand [288].

The circumvention of a transformation step represents an important advantage of cell- free display technologies such as mRNA and ribosome display, allowing for theoretical library sizes of >1014 members to be displayed [287, 289]. The limiting factor that determines what library size can be used is the amount of genetic starting material and volumes that can be practically handled in the laboratory [287, 289]. The PCR amplification step between each selection round accounts for an opportunity to further expand the diversity of the library by on-going evolution through error-prone PCR or DNA shuffling [287, 290]. Since neither amplification steps in living

33 Bacterial display systems for engineering of affinity proteins organisms nor secretory pathways for display are necessary, the introduction of bias is reduced [289].

However, a drawback of cell-free technologies such as ribosome display and mRNA display, in which the phenotype-genotype link relies on mRNA, is the sensitivity to RNAse degradation [289]. CIS display and CAD display are two alternative approaches, where the peptide is cross-linked to double-stranded DNA, either by the bacterial replication initiator protein, RepA (CIS display) [291], or cis-acting DNA binding protein (CAD display) [292].

In addition to display systems, other non-display methods have been used for the selection from combinatorial libraries. An example of such a technology is the protein fragment complementation assay (PCA) [293]. In this assay, the antigen is fused to a ‘bait’ reporter protein, and members of the protein library are fused to a ‘prey’ protein. The bait and the prey can be for example two parts of a split protein, which when assembled confer antibiotic resistance. Binding of a library member to the target protein leads to interaction between the bait and the prey, which will give rise to a detectable activity such as antibiotic resistance [143, 293, 294].

A type of protein-fragment complementation assay that has been used for selection of intracellular antibodies is the two-hybrid approach in yeast or mammalian cells, in which the bait and the prey are based on a transcriptional transactivation domain and a DNA-binding domain, which upon interaction will activate a downstream reporter gene [295]. Similar systems based on bacteria have also been described [296].

34 Filippa Fleetwood

4. TARGETING OF ANGIOGENESIS

A common goal of antibody-based therapeutics and other affinity proteins is to target proteins that play an important role in a disease. Angiogenesis – the formation of new blood vessels from pre-existing vasculature – is an essential factor in several pathological conditions, including tumor growth and metastasis and several ocular disorders [297, 298]. The idea that inhibition of angiogenesis would lead to inhibition of tumor growth has been around for several decades [299]. This chapter describes the targeting of angiogenesis as a therapeutic and diagnostic strategy.

4.1 ANGIOGENESIS

Angiogenesis is required in natural processes such as wound healing and embryogenesis, but is also associated with several disease conditions including cancer and vascular eye diseases [300]. Tumor vascularization is required for tumor growth beyond sizes of 1-2 mm3. The angiogenic phenotype is attained when a flow of oxygen and metabolites is required, and can be triggered by for example oncogenic mutations or hypoxia [301, 302]. Angiogenesis is regulated by activators and inhibitors, and in normal tissue the activity of these pro-angiogenic and anti- angiogenic factors is tightly balanced, maintaining the endothelial cells of the vasculature in a dormant state. The point where the angiogenic inhibitors are outweighed by the activators is known as the ‘angiogenic switch’ [303].

Tumor vascularization can be achieved in several ways [304, 305]. Normal angiogenesis, also referred to as ‘sprouting’ angiogenesis, is the formation of new tubular structures induced by the proliferation and migration of endothelial cells in existing vasculature [305]. This process is regulated by different angiogenic factors, predominantly by the Vascular Endothelial Growth Factor (VEGF) family, which consequently have been recognized as attractive targets for anti-angiogenic therapy [306]. However, tumor vasculature development can also be achieved by other mechanisms [304, 305]. These include putative cancer stem-like cell or endothelial progenitor cell differentiation into endothelial cells [305, 307], vessel splitting by intussusception [308, 309] vessel co-opting [310], and vascular mimicry [311, 312]. Such sprouting-independent mechanisms of vessel growth can be less sensitive to blockade of VEGF or the VEGF receptors [305, 313].

Tumor vessel morphology differs from normal vasculature [298]. The structure of tumor vasculature is more disorganized than normal vasculature, vessels are more

35 Bacterial display systems for engineering of affinity proteins prone to leakage, and the blood flow is chaotic [314]. Since the vessels are often malfunctioned, hypoxia and nutrient deprivation is common in tumors despite the presence of vasculature [315].

Angiogenesis is also an important factor in various eye disorders, such as macular degeneration, retinopathy of prematurity, proliferative diabetic retinopathy and ischemic retinal vein occlusions [316, 317]. In these diseases, damage in the vascular beds supplying the retina (as a result of for example age, environment, systemic disease or genetic predisposition) can induce neovascularization. This leads to complications such as hemorrhage and retinal detachment, which can ultimately cause blindness [297, 317, 318]. In the 90’s, members of the VEGF family were found to be up-regulated in several retinal disorders and to promote retinal angiogenesis under hypoxic conditions [319, 320]. Since then, much evidence for the significant role of the VEGFs in these diseases has been found, and antiangiogenic treatment has been demonstrated to be an effective therapeutic strategy [316, 317].

4.2 THE VEGF/VEGFR FAMILY

Among the major regulators of both physiological and pathological angiogenesis are the members of the Vascular Endothelial Growth Factor (VEGF) family and the VEGF receptors (VEGFRs) [321]. The members of the VEGF family in mammals are VEGF-A, VEGF-B, VEGF-C, VEGF-D and Placental Growth Factor (PlGF) [321]. Additional VEGF family members have been identified in parapoxvirus and snake venom (VEGF-E and VEGF-F) [322-324].

The VEGF proteins are secreted glycoproteins that form antiparallel homodimers consisting of two monomers that are covalently linked and stabilized by disulphide bonds [325]. The homodimer contains one VEGFR binding site at each end [326- 328].

VEGF-A is the most abundant VEGF family member in humans, and is also the most potent regulator of angiogenesis [300]. VEGF-A has for example been shown to be critical for the vasculature development in mice embryos [329]. VEGF-A expression can be stimulated by for example hypoxia, hypoglycemia or oncogenic mutations [301, 302, 330]. Several isoforms of VEGF-A exist due to alternative splicing, of which VEGF-165 is the most abundant variant [331], a homodimeric protein of 46 kDa [332].

36 Filippa Fleetwood

The functions of the VEGF ligands are mediated by the VEGF receptors (VEGFRs): VEGFR1, VEGFR2 and VEGFR3, which are receptor protein-tyrosine kinases. In addition, it is activated by the two non-enzymatic co-receptors neuropilin-1 and -2 [300], though these will not be discussed here.

All three VEGF receptors are involved in the regulation of angiogenesis, though VEGFR2 (also known as kinase-insert domain receptor, KDR) has been found to play the most important role in this regulation, mainly through the action of VEGF-A [332]. VEGFR2 is, however, also activated by different isoforms of VEGF-C and VEGF-D (as well as VEGF-E, in parapoxvirus infections) [300]. In adults, VEGFR2 is expressed mainly on vascular endothelial cells [332], where it is involved in proliferation, migration and survival of endothelial cells, as well as in vascular permeability [332, 333].

VEGFR2 consists of an extracellular domain with a VEGF binding site, a transmembrane domain and an intracellular kinase domain. The extracellular domain consists of seven immunoglobulin (Ig)-like domains (I-VII), of which Ig-like domain II and III are involved in VEGF-A binding [334]. The Ig-like domain seven has been shown to stabilize the interaction between two receptors following ligand-induced dimerization, and to position the kinase domain in a way that facilitates autophosphorylation [335]. The intracellular kinase domain comprises two kinase domains (Kinase domain 1 and 2), separated by a Kinase insert domain [332]. The structure of VEGFR2 is represented by a schematic figure in Fig. 4.1. The murine form of VEGFR2 (foetal liver kinase (Flk-1)) has 85% sequence homology to human VEGFR2 [332]. An alternatively spliced variant of VEGFR2 is expressed as a soluble receptor, which acts as an inhibitor of lymphangiogenesis [336].

37 Bacterial display systems for engineering of affinity proteins

I I

II VEGF II III III Receptor dimerization IV IV

V V Extracellular domain VI VI

VII VII

P P Autophosphorylation P P Intracellular Intracellular kinase domain

Downstream signaling

Angiogenesis

Fig. 4.1 A structural overview of VEGFR2. The extracellular domain consists of seven Ig-like domains (I-VII). Binding of the homodimeric ligand (primarily VEGF-A) induces receptor dimerization, which leads to autophosphorylation of the intracellular kinase domain. This activates various signaling pathways, ultimately leading to angiogenesis.

Signaling of the VEGF receptors is achieved by receptor dimerization upon binding of the dimeric VEGF ligand, which enables autophosphorylation of a number of tyrosines in the kinase domain. These phosphorylations create binding sites for downstream signaling proteins [337-340]. In particular, Tyr1175 has been identified as one of the major phosphorylation sites, which is crucial for VEGF-dependent cell proliferation [339]. VEGFR2 can either form a homodimer or a heterodimer with VEGFR1.

VEGFR1 plays a complex and not fully understood role in angiogenesis [341], and it has been described as a positive as well as a negative regulator of VEGFR2 [321]. VEGFR1 signaling is important for the recruitment of haematopoetic stem cells and the migration of monocytes and macrophages, and the receptor is expressed mainly on these cells. It is also thought to serve as a negative regulator of VEGFR2 signaling,

38 Filippa Fleetwood partly by the actions of an alternatively spliced variant of VEGFR1, which is expressed as a soluble receptor [321, 342]. VEGFR1 binds to VEGF-A with a higher affinity than VEGFR2, though the signaling activity of VEGFR2 is more potent [343]. In addition VEGF-A, VEGFR1 also binds VEGF-B and PlGF [332]. VEGFR3 is mainly involved in regulation of lymphatic endothelial cell function [300]. In adults, VEGFR3 is expressed only in lymphatic vessels [300]. VEGFR3 is activated by VEGF-C and VEGF-D [332].

4.3 TAREGTING OF THE VEGF LIGANDS AND RECEPTORS IN THERAPEUTICS AND DIAGNOSTICS

The expression of VEGF and VEGFR2 is higher in tumor vasculature than in normal vasculature [300], making these proteins suitable targets for both therapeutic applications as well as diagnostics involving for example in vivo imaging. Targeting of the tumor vasculature instead of the tumor cells has some potential advantages. In many tumors, there is a high interstitial pressure, preventing therapeutics from efficiently penetrating into the tumor tissue [344]. It has also been suggested that targeting of the host endothelial cells, which are “normal”, genetically stable cells, is preferable to targeting of genetically unstable cancer cells, in order to lower the risks of development of drug resistance due to genetic mutations [345].

One strategy to inhibit VEGFR2-mediated angiogenesis is to block the tyrosine kinase activity of the receptor, and thereby inhibit downstream signaling pathways. Sorafenib (Nexavar®) [346], Sunitinib (Sutent®) [347] and Pazopanib [348] are examples of small molecule inhibitors that have been FDA approved for treatment of different cancers (www.cancer.gov). Such tyrosine kinase inhibitors generally target other kinases in addition to VEGFR2 [321], which could have advantageous as well as disadvantageous effects. Several small molecule drugs acting as antagonists of the VEGF receptors have also been developed [349, 350].

Several protein-based therapeutics targeting the VEGF ligands and receptors have also been developed. For example, the mouse anti-human VEGF mAb Bevacizumab (Avastin®, Genentech, Inc.) [351] has been approved by the FDA for treatment of a number of cancers in combination with chemotherapy (www.cancer.gov)[341]. An affinity-matured Fab variant of Bevacizumab, named Ranibizumab (Lucentis, Genentech Inc.), has been developed specifically for the treatment of vascular eye diseases [352]. The smaller Fab format allows for penetration across retinal layers

39 Bacterial display systems for engineering of affinity proteins upon intraocular administration, which is preferable to systemic administration for the treatment of eye diseases, due to potential toxic side effects [352].

Another interesting approach for blocking the VEGFR binding site on VEGF is the VEGF-trap (Aflibercept®, Regeneron Pharmaceuticals, Inc ), which consists of the Ig-like domain 2 of human VEGFR1 and the Ig-like domain 3 of human VEGFR2 fused to the Fc part of human IgG1 [310]. Precursors of this agent were based on a VEGFR1-Fc construct (which has a higher VEGF-A affinity) and a VEGFR2-Fc construct. VEGFR1 has a higher affinity for VEGF-A, but due to poor pharmacokinetic properties and toxic side effects (which were not observed for the VEGFR2-Fc construct), the new construct comprising domains from each of the receptors was engineered.

In addition to Bevacizumab and VEGF-Trap, other therapeutics targeting VEGF include the antisense oligonucleotide VEGF-AS (Veglin®; Vasgene Therapeutics, Inc., Sharon Hill, PA), the RNA aptamer Pegaptanib (Pegaptanib sodium; Macugen®, Eyetech Pharmaceuticals, Inc., New York, NY), which was the first antiangiogenic agent approved for treatment of neovascular macular degeneration [353], and the peptide Aplidin (Aplidin® Dehydrodidemnin B, Pharma Mar, Madrid, Spain) [354].

Monoclonal antibodies and other agents targeting the VEGF receptors are also under development. Potential advantages of targeting the receptor instead of the ligand include the blocking of other species of VEGF (VEGF-C and VEGF-D) from binding and activating the receptor [355]. A drawback of targeting only VEGFR2 could potentially arise from the ability of VEGF-A from activating VEGFR1 signaling. However, since the potency of VEGFR1 is significantly lower compared to that of VEGFR2, and it has also been described as a negative regulator of VEGFR2 signaling, (see section 4.2), it is possible that this risk is negligible. Another possible drawback is that soluble VEGFR2, which is an inhibitor of lymphangiogenesis, could also be targeted by anti-VEGFR2 agents. Since lymphangiogenesis is important for metastasis [356], this could potentially lead to indirect promotion of metastasis and tumor progression. The first anti-VEGFR2 antibody to be approved by the FDA was the fully human monoclonal antibody IMC-1121B (Ramucirumab, ImClone Systems) [357, 358] (http://www.fda.gov/Drugs/InformationOnDrugs/ApprovedDrugs/ucm394260.htm, 141029). Another example of a VEGFR2-targeting antibody-based therapeutic candidate is a nanobody-immunotoxin, comprising a VEGFR2-specific nanobody fused to Pseudomonas exotoxin A [359]. An example of a monoclonal antibody targeting VEGFR1 is the fully human monoclonal antibody IMC-18F1 [360]. The

40 Filippa Fleetwood adnectin CT-322 is an example of an alternative scaffold protein targeting VEGFR2 [361, 362]. Alternative scaffolds or antibody fragments might be advantageous for this application. The presence of a large and complex molecule such as the Fc is not necessarily required for blocking of protein-protein interactions. The improved circulation time that is generally provided by the Fc can be achieved by other means [102].

Anti-angiogenesis agents have also shown successful results in the treatment of ocular diseases [316]. However, the delivery of large proteins such as antibodies to the eye is complicated [363]. Small therapeutic agents that could be delivered by eye drops could therefore be a potential advantageous alternative, as has been demonstrated with other therapeutics [364].

Although targeting of the VEGF-VEGFR2 signaling pathways has proved effective, and several promising candidates are in clinical development, there are still challenges to be addressed. First of all, a number of side effects have been observed although these are usually manageable [341, 365]. A more serious challenge is that the disease is sometimes refractory or resistance to drugs is acquired [366-369]. Resistance can be developed for example by up-regulation of other pro-angiogenic factors [366, 369]. Alternatively, as a result of the hypoxia caused by the anti-VEGF treatment, cells that are less sensitive to hypoxia, and therefore do not require angiogenesis, gain a growth advantage [366]. In addition, it has been reported that anti-angiogenic compounds may accelerate metastasis [367, 368]. Furthermore, although sprouting angiogenesis is inhibited by anti-VEGF therapy, tumor vasculature development can also be achieved in other ways [304], which is one reason why these therapeutics are not always effective.

A therapeutic strategy that is not dependent on the expression of typical angiogenic mediators is vascular targeting [370]. In this approach, therapeutic agents specific to any target expressed exclusively in the tumor vasculature destroy the established tumor vasculature independent of the function of the target molecule. The therapeutic agent can consist of for example an antibody fused to a cytokine [61].

Another issue is the relatively moderate efficiency of many therapeutics. Bevacizumab was the first anti-angiogenic drug to be approved by the FDA, and has been used extensively in the clinic. However, it is only protective in combination with chemotherapy [341]. Many tyrosine kinase inhibitors are effective as monotherapy only in some cancers, and are sometimes toxic in combination with chemotherapy [354, 371].

41 Bacterial display systems for engineering of affinity proteins

Better prediction of the outcome and evaluation of the response of an anti-angiogenic therapy might be helpful for understanding the problem of refractoriness to disease and the lack of response in some patients. This might be achieved by the identification of better biomarkers and developing more sensitive assays [341, 355, 366, 371]. Imaging of angiogenesis is a potential way of achieving this [372, 373]. Imaging of angiogenesis could also be useful in studying the progression of a non-angiogenic cancer therapy. Previous studies describe imaging of angiogenesis using for example radiolabeled anti-VEGF antibodies [374, 375], fragments of VEGF [376, 377] and small molecules targeting VEGFR2 [378]. However, antibodies generally have a circulation time that is longer than optimal for imaging purposes [101]. The small size of small organic molecules leads to limitations in terms of what reporter molecule that can be coupled to them [379]. The size of peptides and small proteins, such as the Affibody molecule, are attractive alternatives for imaging applications [100, 379].

An alternative idea is to ‘normalize’ the abnormal morphology of the tumor vasculature in order to make the vessel walls more stable, rather than to destroy them [380]. Stabilization of the tumor vasculature could potentially facilitate the delivery of therapeutics, the improved supply of oxygen could lower the risk of for example hypoxia-resistant cells, and the decreased leakiness of the vessel walls could impede the escape of metastasizing tumor cells [313, 315, 380, 381].

42 Filippa Fleetwood

5. PRESENT INVESTIGATION

Combinatorial protein engineering is a powerful method for the generation of specific affinity proteins. As discussed in chapter 3, cell surface display technologies are attractive for this purpose, mainly due to the possibility to use flow cytometry for soring of the library. A bacterial display system based on the Gram-positive S. carnosus has previously been developed at our department, and used for successful generation of high-affinity Affibody molecules [147, 209, 228].

A common aim of the use of affinity proteins as therapeutics is to specifically target a protein that plays an important role in a disease. As described in chapter 4, angiogenesis is critical in many disease conditions, including various types of cancer and vascular eye disorders [300]. The receptor VEGFR2 is one of the main regulators of angiogenesis [306]. The goal of papers I and II of this thesis was to generate an Affibody molecule which could bind to VEGFR2, and thereby inhibit its signaling ability. This Affibody molecule is intended for therapeutic as well as in vivo imaging applications. Paper I describes the selection of two VEGFR2-binding Affibody molecules using a combination of phage and staphylococcal display, and the engineering of these two Affibody molecules into a biparatopic construct with a dramatically increased affinity for VEGFR2 due to simultaneous binding of the two Affibody domains. Paper II describes the further evaluation of the effects of this construct on VEGFR2-expressing cells.

In paper I of this thesis, and in several other studies [209], the staphylococcal display system has been demonstrated to be excellent for the display of Affibody libraries, but it has previously never been evaluated for display of an antibody-based library. In paper III, the suitability of the staphylococcal display system for selections from a single-domain antibody (nanobody) library was investigated. Furthermore, the outcome of the selection was directly compared to the selection from the same library using phage display.

It is generally believed that a larger size of a combinatorial library increases the likeliness of selecting a candidate with the desired properties [25]. The limiting factor determining the size of the library using microbial display systems is commonly the number transformants that can be achieved when transforming the DNA encoding the library into the host cells. In paper IV, a display system based on E. coli was evaluated as a means to engineer Affibody molecules. The use of a Gram-negative bacterium enables the display of larger libraries due to a higher transformation

43 Bacterial display systems for engineering of affinity proteins frequency. Furthermore, the expression system was designed to facilitate small-scale soluble protein production for characterization of selected candidates through OmpT- mediated release of the Affibody molecule into the medium.

In summary, the papers of this thesis describe the use, development and evaluation of bacterial display technologies for the engineering of specific affinity proteins.

44 Filippa Fleetwood

5.1 PAPER I

Simultaneous targeting of two ligand-binding sites on VEGFR2 using biparatopic Affibody molecules results in dramatically improved affinity

As described in chapter 4, the formation of new vasculature is a key step in tumor development, and also plays an important role in several other diseases such as age- related macular degeneration of the eye and diabetic retinopathy [297, 300]. The Vascular Endothelial Growth Factor Receptor 2 (VEGFR2) is an important mediator of angiogenesis, and is therefore an attractive target for anti-cancer therapies [300, 306]. As mentioned in section 4.3, targeting of angiogenesis as a strategy to treat cancer is however associated with many challenges, such as the development of resistance and the lack of response in some patients. Imaging of VEGFR2 expression could potentially be used for identification of responders and development of resistance. Since VEGFR2 is overexpressed during angiogenesis, it is also a potentially attractive target for in vivo imaging, for example to study the effect of an on-going therapy [382, 383].

VEGFR2 signaling is mediated by the binding of the ligand, Vascular Endothelial Growth Factor A (VEGF-A), to the extracellular domain of the receptor, inducing receptor dimerization [321]. Dimerization, in turn, induces phosphorylation of the intracellular kinase domain, activating downstream signaling pathways, ultimately leading to proliferation and migration of endothelial cells and vascular permeability [337].

Affibody molecules are a family of alternative scaffold proteins that have shown great potential for in vivo imaging as well as therapeutic applications [94]. The properties of Affibody molecules are discussed in section 2.3.6. General advantages of most alternative scaffold proteins include a small size and single-domain structure, which for example facilitates engineering of multi-domain constructs. Multiple specificities can be introduced for simultaneous targeting of different receptors, or bivalent constructs can be engineered in order to decrease dissociation from the target through the gain of an avidity effect or to target multiple sites on the same target [18, 106, 108]. One application of multi-specific proteins can be for example half-life prolongation for therapeutic purposes, which can be achieved through fusion to domains with a long in vivo half-life [102]. The small format and high stability is also potentially advantageous for applications where an alternative delivery route is beneficial. For example, topical administration of an antiangiogenic

45 Bacterial display systems for engineering of affinity proteins agent for treatment of intraocular disorders, might be preferable to systemic administration, due to potential toxic side effects [352].

The aim of paper I was to generate an Affibody molecule that could inhibit human VEGFR2 signaling by blocking the VEGF binding site, and which had adjustable pharmacokinetics in order to enable long circulation time for therapeutic purposes as well as a fast clearance for in vivo imaging. In order to potentially obtain a very high target affinity, two Affibody molecules binding to different sites on VEGFR2 were engineered into a biparatopic construct, which also included an albumin-binding domain for potential future half-life extension.

Selection of Affibody molecules binding to human VEGFR2 from a “naïve” Affibody library [384] was performed using phage display. ELISA screening identified two variants, ZVEGFR2_1 and ZVEGFR2_2, which demonstrated cross-reactivity to murine VEGFR2, a property required for pre-clinical studies of angiogenesis in mice. These two clones had a low sequence homology, while characterization using surface plasmon resonance (SPR) and circular dichroism (CD) showed that the two Affibody molecules were equal in terms of affinity for human and murine VEGFR2 (equilibrium constants ranging from 78-258 nM) (Fig. 1a, Appendix 1, and Supplementary Fig. S3, Appendix 1) and stability (melting temperatures of around

45°C for ZVEGFR2_1 and around 50°C for ZVEGFR2_2). ELISA and SPR experiments showed that none of the clones were cross-reactive to VEGFR1 or VEGFR3, and that both clones competed with VEGF-A for binding to VEGFR2 (Fig. 5.1a). Furthermore, an SPR-based competition assay demonstrated that the two Affibody molecules could bind simultaneously to VEGFR2 (Fig. 5.1b). Taken together, these results suggested that the two candidates bound close, but not overlapping binding- sites.

46 Filippa Fleetwood

a b

hVEGFR-2/Fc + PBS I. 1 M ZVEGFR2_1 hVEGFR-2/Fc + II. Combination 25 x ZVEGFR2_1 I. 1 M ZVEGFR2_1 hVEGFR-2/Fc + II. 2 M ZVEGFR2_1 50 25 x Z 250 VEGFR2_2 II

40 200 ) ) U U 30 150 R R ( (

I e e s s n n o o p p 20 s

s 100 e e R R

10 50

0 0 0 500 1000 1500 0 100 200 300 400 Time (s) Time (s)

Fig. 5.1 SPR competition assays for the first generation VEGFR2-binding Affibody molecules ZVEGFR2_1 and ZVEGFR2_2. a) Competition with VEGF-A. Human VEGFR2 was pre-incubated with 25 × molar excess of ZVEGFR2_1, ZVEGFR2_2 or PBS (control), and injected over immobilized VEGF-A. b) Competition of the two Affibody molecules for simultaneous binding to VEGFR2. A double injection was performed, where a first injection of 1 µM of ZVEGFR2_1 (I) was immediately followed by a second injection (II) of either a combination of 1 µM of ZVEGFR2_1 and 1 µM of ZVEGFR2_2, or 2 µM of ZVEGFR2_1, over immobilized human VEGFR2.

High target affinities (low nanomolar to picomolar) are desirable for in vivo imaging and therapeutic applications. Therefore, an affinity maturation of each binder was performed using staphylococcal display. Staphylococcal display has previously been successfully used for the affinity maturation of Affibody molecules [147]. Since

ZVEGFR2_1 and ZVEGFR2_2 were able to bind to VEGFR2 simultaneously, a strategy for potentially further increasing the VEGFR2 affinity was the engineering of a biparatopic construct based on these two Affibody molecules.

Two affinity maturation libraries were designed (ZVEGFR2_1matlib and ZVEGFR2_2matlib). First, an alanine scan, where each of the thirteen randomized library positions was mutated to alanine, was used for the evaluation of the contribution of each position to the VEGFR2 binding (Fig. 2, Appendix 1). The results from the alanine scan were used for the determination of the degree of randomization in each library position. Four rounds of staphylococcal display were performed, using decreasing concentrations of recombinant VEGFR2 in round 1-3, and an off-rate selection strategy was applied in round 4. In this sorting round, an incubation with an excess of unlabeled VEGFR2 following the incubation with labeled VEGFR2 allowed for the potential isolation of binders with slow dissociation rates. Dot plots from the sorting of ZVEGFR2_1matlib is shown in Fig. 5.2. Dot plots from the sorting of ZVEGFR2_2matlib can be found in Fig. 3, Appendix 1. The. Sequencing of selected clones revealed a number

47 Bacterial display systems for engineering of affinity proteins of unique variants from each library. Unexpectedly, almost all selected clones from

ZVEGFR2_2matlib contained an unintended mutation in the scaffold; K33N.

Unsorted library Sort 1 Sort 2 Sort 3

Sort 4, Sort 4, Sort 4, ZVEGFR2_1 no off-rate selection 0.5 h off-rate selection 4.5 h off-rate selection VEGFR2-binding (Log fluorescence intensity)

Surface expression level (Log fluorescence intensity)

Fig. 5.2 Sorting of affinity maturation libraries ZVEGFR2_1matlib. (Dot plots from the sorting of ZVEGFR2_2matlib is shown in Fig. 3, Appendix 1). The VEGFR2 binding signal (monitored by the binding of fluorescently labeled anti-Fc antibody to human VEGFR2/Fc) is represented on the Y axis and the surface expression level (monitored by fluorescently labeled HSA binding) is represented on the X axis. The dot plots show staphylococcal cell populations from the original unsorted library as well as cell populations isolated in the 1st, 2nd, 3rd and 4th selection round, respectively. For the 4th selection round, dot plots are shown for selections including an off-rate experiment of 0 min, 30 min or 4.5 h.

On cell-ranking was performed in order to select four top candidates from each library for further characterization. These variants (ZVEGFR2_11, ZVEGFR2_16, ZVEGFR2_19 and ZVEGFR2_22 from ZVEGFR2_1matlib, and ZVEGFR2_33, ZVEGFR2_38, ZVEGFR2_41 and ZVEGFR2_40 from ZVEGFR2_2matlib) were sub-cloned for soluble protein production in E. coli, and proteins were purified using IMAC, followed by characterization using CD and SPR. CD analysis showed a decreased thermal stability of the variant from ZVEGFR2_2-matlib containing the K33N mutation (Table 5.1). Therefore, ZVEGFR2_40, which was the best-ranked clone from

ZVEGFR2_2-matlib that did not contain the scaffold mutation, was chosen for further studies. From ZVEGFR2_1-matlib, a variant denoted ZVEGFR2_22 was selected for further studies, based on the affinity determination using SPR. The equilibrium constants of the two affinity matured Affibody molecules ZVEGFR2_22 and ZVEGFR2_40 were in the low nanomolar range for human as well as murine VEGFR2 (Table 5.1).

48 Filippa Fleetwood

Table 5.1. Equilibrium dissociation constants (KD), dissociation rate constants (kd) and estimates melting temperatures (Tm) for affinity-matured Affibody molecules

-1 -1 Denotation Average KD (nM) Average KD (nM) Average kd (s ) for Average kd (s ) for Estimated for murine for human human VEGFR2 (± murine VEGFR2 (± Tm (°C) VEGFR2 (± SD) VEGFR2 (± SD) SD) SD)

ZVEGFR2_22 9.8 ± 3.5 5.0 ± 0.2 5.8×10-3 ± 4.2×10-4 8.6×10-3 ± 4.0×10-4 49

ZVEGFR2_19 10.0 ± 2.4 5.4 ± 0.8 4.2×10-3 ± 6.3×10-4 1.1×10-2 ± 1.5×10-4 47

ZVEGFR2_11 ND ND ND ND 47

ZVEGFR2_16 11.9 ± 3.9 6.7 ± 1.1 5.6×10-3 ± 3.1×10-4 1.2×10-2 ± 5.5×10-4 46

ZVEGFR2_41 ND ND ND ND 34

ZVEGFR2_38 ND ND ND ND 34

ZVEGFR2_33 ND ND ND ND 39

ZVEGFR2_40 7.8 ± 3.4 10.9 ± 3.1 1.5×10-2 ± 1.5×10-3 1.1×10-2 ± 1.9×10-3 45

SPR experiments demonstrated that the two affinity matured VEGFR2-binding Affibody molecules had retained their ability to block VEGF-A from binding to the receptor, and that they could still bind simultaneously to VEGFR2 (Fig. 4, Appendix 1). Encouraged by these results, we designed dimeric constructs comprising

ZVEGFR2_22 and ZVEGFR2_40 separated by two flexible linkers and an albumin-binding domain (ABD035). ABD035 has a femtomolar affinity for human serum albumin (HSA) [217], and was intended to function as a linker as well as a tag for affinity purification, directed immobilization in SPR experiments and detection in flow- cytometry experiments. It also has the ability to increase the half-life of the construct

in future animal studies. Since the dimeric construct comprising ZVEGFR2_22 and

ZVEGFR2_40 contained two Affibody molecules that could bind to two non-overlapping binding sites on VEGFR2, they were termed ‘biparatopic’ Affibody molecules. The following four dimeric Affibody constructs were designed (Fig. 5.3):

ZVEGFR2_22-(S4G)4-ABD035-(S4G)4-ZVEGFR2_40,

ZVEGFR2_40-(S4G)4-ABD035-(S4G)4-ZVEGFR2_22,

ZVEGFR2_22-(S4G)4-ABD035-(S4G)4-ZVEGFR2_22 and

ZVEGFR2_40-(S4G)4-ABD035-(S4G)4-ZVEGFR2_40.

49 Bacterial display systems for engineering of affinity proteins

Heterodimers (biparatopic):

Z VEGFR2_22 Z VEGFR2_40 Z VEGFR2_40 Z VEGFR2_22 ABD ABD

Homodimers:

Z VEGFR2_22 Z VEGFR2_22 Z VEGFR2_40 Z VEGFR2_40 ABD ABD

Fig. 5.3 Schematic representation of the four dimeric Affibody constructs (heterodimers and homodimers) comprising ZVEGFR2_22, ZVEGFR2_40 and ABD035.

The latter two constructs were homodimeric, and were intended to serve as controls. These constructs were produced in E. coli and purified by affinity chromatography using immobilized HSA. Dissociation constants were analyzed using an SPR setting where HSA was immobilized on the chip surface, and dimeric constructs were injected directly followed by a second injection of monomeric human or murine VEGFR2 (Fig. 5.4a). This allowed for a directed immobilization of the dimeric constructs, resulting in functional available VEGFR2 binding sites. The results revealed a substantially slower dissociation of the heterodimeric constructs compared to the homodimeric controls (Fig. 5.4b). For ZVEGFR2_22-(S4G)4-ABD035-(S4G)4-

ZVEGFR2_40, which was the best-performing of the heterodimeric constructs, the dissociation rate constant was estimated to around 8.4 × 10-5 s-1 for human VEGFR2.

For the monomeric ZVEGFR2_22 and ZVEGFR2_40, the respective dissociation rate constants were around 5.8 × 10-3 s-1 and 1.5 × 10-2 s-1 (Table 5.2). The formatting the Affibody molecules as biparatopic binders resulted in around 70-fold and 180-fold slower off-rate, respectively. In comparison to the respective homodimeric constructs, the dissociation rate constants were around a 14-fold and a 20-fold improved. Dissociation rate constants for the dimeric Affibody constructs are summarized in Table 5.2. Since the SPR setup was such that avidity effects should be avoided (using immobilized Affibody construct and monomeric VEGFR2), these results indicate that the fusion of ZVEGFR2_22 and ZVEGFR2_40 in a dimeric construct resulted in an increased affinity due to simultaneous binding of the two Affibody molecules. These experiments also demonstrate that the biparatopic constructs were able to bind to HSA and VEGFR2 simultaneously, which is promising for future half-life extension in vivo.

50 Filippa Fleetwood

a b ZVEGFR2_22-ABD-ZVEGFR2_40 ZVEGFR2_40-ABD-ZVEGFR2_22 Z -ABD-Z I. Immobilization of II. Directed III. Injection of VEGFR2_22 VEGFR2_22 Z -ABD-Z HSA on chip non-covalent monomeric 400 VEGFR2_40 VEGFR2_40 immobilization of VEGFR2 350 biparatopic Affibody 300 ) U

R 250 (

e

VEGFR2 s 200 n o p

Z2 Z2 s 150 e

R 100 Z1 ABD ABD Z1 50 HSA HSA HSA HSA HSA 0 0 200 400 600 800

Fig. 5.4. SPR analysis of dimeric Affibody constructs. a) A schematic overview of the experimental setup in the SPR analysis of the dimeric Affibody constructs. Dimeric constructs were injected over a surface of immobilized HSA, directly followed by an injection of monomeric VEGFR2. b) SPR sensorgram showing the simultaneous injection of all four dimeric Affibody constructs, demonstrating a substantially slower dissociation rate for the heterodimers.

Table 5.2 Dissociation rate constants (kd) for dimeric Affibody constructs Construct Average k (s-1) for human Average k (s-1) for murine d d VEGFR2 (± SD) VEGFR2 (± SD) -5 -5 -4 -5 ZVEGFR2_22-ABD-ZVEGFR2_40 8.4×10 ± 3.5×10 1.4×10 ± 1.8×10 -4 -5 -4 -5 ZVEGFR2_40-ABD-ZVEGFR2_22 1.5×10 ± 4.2×10 1.8×10 ± 2.2×10 -3 -5 -4 -4 ZVEGFR2_22-ABD-ZVEGFR2_22 1.6×10 ± 8.7×10 9.9×10 ± 2.0×10 -3 -4 -4 -4 ZVEGFR2_40-ABD-ZVEGFR2_40 1.2×10 ± 1.5×10 7.9×10 ± 4.9×10

The findings from the SPR analysis were supported by the results from flow- cytometric analysis of the binding of the Affibody constructs to VEGFR2-expressing mammalian cells. A stronger cell-binding signal of the biparatopic constructs compared to the homodimeric constructs also indicate that fusion into a heterodimeric construct results in an affinity increase due to simultaneous VEGFR2 binding in addition to avidity effects (Fig. 5.5).

Heterodimers Homodimers Z -ABD-Z ZVEGFR2_22-ABD-ZVEGFR2_40 VEGFR2_22 VEGFR2_22 Z -ABD-Z ZVEGFR2_40-ABD-ZVEGFR2_22 VEGFR2_40 VEGFR2_40 Z -ABD-Z ZTaq-ABD-ZTaq Taq Taq Secondary reagent Secondary reagent Cell count Cell count

Affibody binding Affibody binding (Log fluorescence intensity) (Log fluorescence intensity)

Fig. 5.5 Flow-cytometric analysis of the binding of the dimeric Affibody constructs to VEGFR2- expressing 293/KDR cells. A larger shift in fluorescence intensity is observed for the heterodimers.

51 Bacterial display systems for engineering of affinity proteins

The ability of the VEGFR2-specific Affibody molecules to inhibit VEGFR2 signaling was investigated in an ELISA-based phosphorylation assay. 293/KDR cells were treated with the biparatopic binder ZVEGFR2_22-(S4G)4-ABD035-(S4G)4-ZVEGFR2_40, a combination of each of ZVEGFR2_22 and ZVEGFR2_40, the negative control construct or PBS only. VEGFR2 phosphorylation was thereafter stimulated with human VEGF-A. Cells were lysed and the phosphorylation level on VEGFR2 Tyr1175 was analyzed by ELISA using an anti-Tyr1175 mAb for detection. A decrease in phosphorylation level was observed for cells incubated with VEGFR2- specific Affibody molecules (Fig. 5.6). A more potent inhibition was observed for the heterodimer than the combination of the monomeric binders. These results suggest a great potential of the biparatopic Affibody molecules for inhibition of VEGFR2- mediated signaling, which is promising for potential therapeutic applications.

10 VEGFR2 phosphorylation ) s n l l o i e t c a

l 8 d y r e t o a h e p r t s n o 6 u h

t p

s d n i e t a a g l

a 4 u

d m i e t z s i l - a F 2 m G r E o V n ( 0

M M n n nM 30 30 3 F F G ol F ol F F F F F tr F tr F G F G G G G VE G on on No Affibody+ No Affibody- VEG VE VE VE VE VE . c VE . c + - VEG + + - VEG + + - VEG + Monomers Heterodimer 3 nM Neg Neg. controlNeg 3 nM Monomers 30 nM Monomers 30 nM Heterodimer 30 nM Heterodimer 30 nM

Fig. 5.6. Inhibition of VEGFR2 phosphorylation. Pre-incubation with the biparatopic ZVEGFR2_22- (S4G)4-ABD035-(S4G)4-ZVEGFR2_40 resulted in a decrease in phosphorylation level compared to the controls, and a more potent inhibition than the combination of the two monomers. VEGFR2 phosphorylation was determined by ELISA. The data is presented as the OD450 for each sample normalized against the OD450 of untreated cells.

In summary, we have generated two Affibody molecules that both block VEGF-A from binding to VEGFR2, but are able to bind simultaneously. The fusion of these two binders into a biparatopic Affibody construct resulted in a dramatically improved affinity for recombinant VEGFR2 as well as VEGFR2 expressed on the surface of mammalian cells. Furthermore, the biparatopic molecule demonstrated a potent ability to inhibit VEGF-A induced VEGFR2 phosphorylation, which is promising for future therapeutic and imaging applications. This is the first report of the engineering of a biparatopic Affibody molecule, and the positive results obtained here might encourage others to use this strategy for an improvement in target-binding affinity in the future.

52 Filippa Fleetwood

5.2 PAPER II

Efficient blocking of VEGFR2-mediated signaling using biparatopic Affibody molecules

Paper I describes the engineering of a biparatopic Affibody construct (ZVEGFR2_22-

(S4G)4-ABD035-(S4G)4-ZVEGFR2_40) comprising the two VEGFR2-binding Affibody molecules ZVEGFR2_22 and ZVEGFR2_40, which resulted in an increased VEGFR2 affinity due to simultaneous binding of the two Affibody domains to different epitopes on VEGFR2. This Affibody construct is intended for future in vivo imaging studies and therapeutic applications. For these purposes, a high affinity is desirable. The desired pharmacokinetics profile is different for imaging than therapeutic applications (fast clearance for imaging and long circulation time for therapeutics). In order to enable adjustment of the pharmacokinetics depending on the application, the biparatopic construct designed in paper I also contained an albumin-binding domain (ABD035, [217]).

Affibody molecules have previously shown very promising results for in vivo imaging in mice as well as humans [94, 152]. This is possibly due to their small size, which facilitates tumor penetration and clearance of unbound protein, yielding high-contrast images. For in vivo imaging of angiogenesis targeting VEGFR2, however, a small size may not be as crucial since the target is expressed on the vasculature instead of on tumor cells, and tumor penetration is therefore not needed. Therefore, the gain of a high affinity could potentially outweigh the common problems associated with a larger size. However, in order to minimize the size, the construct designed in paper I could be further engineered. Firstly, the ADB035 domain between the two Affibody molecules, which is intended as a tag for in vitro purposes such as detection and purification, as well as a means for half-life extension, is not necessary for imaging purposes. Secondly, the total protein size could be minimized by decreasing the length of the linker length. A shorter linker length could also result in higher serum stability, since linear peptides are generally more susceptible to protease degradation than folded proteins [385]. The aim of paper II was (i) to investigate whether a decrease or an increase in linker length between the two Affibody molecules had any effect on the VEGFR2 affinity, and (ii) to further evaluate the therapeutic potential of biparatopic Affibody constructs by investigating whether they could block VEGFR2 signaling.

53 Bacterial display systems for engineering of affinity proteins

In order to investigate the effect of the linker length between the two Affibody molecules, four new constructs were designed (Fig. 5.7):

ZVEGFR2_Bp1: ZVEGFR2_22 -(S4G)-ABD035-(S4G)-ZVEGFR2_40,

ZVEGFR2_Bp2: ZVEGFR2_22 -(S4G)-ZVEGFR2_40-(S4G)3-ABD035,

ZVEGFR2_Bp3: ZVEGFR2_22 -(S4G)3-ZVEGFR2_40-(S4G)3-ABD035, and

ZVEGFR2_Bp4: ZVEGFR2_22 -(S4G)8-ZVEGFR2_40-(S4G)3-ABD035.

ZVEGFR2_Bp1 ZVEGFR2_Bp2 ABD

Z S4 Z VEGFR2_22 G 4G VEGFR2_40 S S4G ABD Z Z VEGFR2_22 ZVEGFR2_40

ZVEGFR2_Bp3 ZVEGFR2_Bp4 ABD ABD

(S4G)3 (S4G)8 Z VEGFR2_22 ZVEGFR2_40 Z VEGFR2_22 ZVEGFR2_40

Fig. 5.7 Schematic representation of the four dimeric Affibody constructs ZVEGFR2_Bp1, ZVEGFR2_Bp2, ZVEGFR2_Bp3 and ZVEGFR2_Bp4.

The results obtained in paper I demonstrated that the dissociation rate of the construct

ZVEGFR2_22-(S4G)4-ABD035-(S4G)4-ZVEGFR2_40 was somewhat slower than for

ZVEGFR2_40-(S4G)4-ABD035-(S4G)4-ZVEGFR2_22. Therefore, the orientation of the Affibody molecule domains in former construct was used for design of the novel biparatopic constructs. In three of the constructs, the ABD035 was positioned at the C-terminal for potential removal in future in vivo imaging studies. To investigate whether the positioning of the ABD035 affected the VEGFR2 binding, one construct was designed to contain an ABD035 in the middle. Flexible (S4G)n linkers of 5, 15 or

40 amino acids separated the two Affibody domains in ZVEGFR2_Bp2 - ZVEGFR2_Bp4, and in ZVEGFR2_Bp1, the total number of amino acids (including ABD035) between the two Affibody domains was 56.

The four new constructs were tested for VEGFR2 binding using flow cytometry. The results demonstrated that all four constructs were able to bind to VEGFR2 expressed on the surface of 293/KDR cells (Sibtech, Inc.) as well as Human Umbilical Vein Endothelial cells (HUVEC, Lonza) (Fig. 5.8). 293/KDR is a human embryonic kidney cell line, which has been transfected with VEGFR2 [386], and has a relatively high VEGFR2 surface expression. HUVECs are cells derived from the endothelium of

54 Filippa Fleetwood veins from the umbilical cord [387], with a lower VEGFR2 expression compared to 293/KDR. The results demonstrated that shortening of the linker length did not lead to any substantial decrease in VEGFR2 binding. Interestingly, the construct with shortest linker, ZVEGFR2_Bp2, and the construct with ABD035 in the middle,

ZVEGFR2_Bp1, showed a somewhat higher binding signal than the two longer constructs

ZVEGFR2_Bp3 and ZVEGFR2_Bp4. The biparatopic Affibody molecule ZVEGFR2_22-(S4G)4-

ABD035-(S4G)4-ZVEGFR2_40, which had been shown to bind to VEGFR2 with both Affibody molecules simultaneously in paper I, was included for comparison. The binding signal of all four biparatopic constructs ZVEGFR2_Bp1 – ZVEGFR2_Bp4 was similar or higher than the binding signal of ZVEGFR2_22-(S4G)4-ABD035-(S4G)4-ZVEGFR2_40, indicating that the ability of both Affibody molecules to bind simultaneously to VEGFR2 was retained in the new constructs.

a b ) ) l l 15 293/KDR 2.5 HUVEC o o r r t t n n o o c c l l

2.0 a a e e n n v v i i t t g g i i a a s s 10

g g g g e e 1.5 n n n i i n

t t d d s s n n i i n n i i b b

a a e e 1.0 g g v v i i a a t t

5 a a d d l l e e e e z z i i R R l l 0.5 a a m m r r o o n n ( 0 ( 0.0

1 2 3 4 0 ) ) 1 2 3 4 0 ) ) p p p p p p p p B B B B B B B B R2_ R2_ R2_ R2_ dimer R2_ R2_ R2_ R2_ dimer q q a a Z VEGFR2_4 Z t Z VEGFR2_4 Z t Z VEGF Z VEGF Z VEGF Z VEGF - Z VEGF Z VEGF Z VEGF Z VEGF - ) 4 ) 4 G G S 4 S 4 Neg ctrl ( Neg ctrl (

-ABD035-( Pos ctrl (anti-VEGFR2 mAb -ABD035-( Pos ctrl (anti-VEGFR2 mAb ) 4 ) 4 G G S 4 S 4 -( -( 2 2

Z VEGFR2_2 Z VEGFR2_2

Fig. 5.8 Flow-cytometric analysis of dimeric Affibody constructs ZVEGFR2_Bp1, ZVEGFR2_Bp2, ZVEGFR2_Bp3 and ZVEGFR2_Bp4 binding to 293/KDR cells (a) or HUVECs (b). All constructs show a higher VEGFR2-binding signal compared to the negative control Affibody construct.

For investigation of the therapeutic potential of the different biparatopic Affibody constructs, the ability to block VEGF-A from binding to VEGFR2-expressing cells was analyzed in a flow-cytometry assay. 293/KDR cells were pre-incubated with varying concentrations of each Affibody construct before incubation with biotinylated VEGF-A. VEGF-A binding was thereafter detected using streptavidin-conjugated phycoerythrin (SAPE). All constructs demonstrated a concentration-dependent ability to block VEGF-A binding (Fig. 5.9). The IC50 values were in the same range for all four constructs (35-68 nM), though slightly lower values were obtained for the two

55 Bacterial display systems for engineering of affinity proteins shorter constructs ZVEGFR2_Bp1 and ZVEGFR2_Bp2, which was in accordance with the observations from the flow cytometry binding assays.

ZVEGFR2_Bp1 (IC50=35 nM) VEGF inhibition ZVEGFR2_Bp2 1.0 (IC50=36 nM)

ZVEGFR2_Bp3

g (IC50=60 nM) n i

d 0.8 n i ZVEGFR2_Bp4 b

d (IC50=58 nM) e z i l a

m 0.6 r o N

0.4 0 1 2 3 log[Affibody construct], nM

Fig. 5.9 Flow-cytometric analysis of the ability of the biparatopic VEGFR2-binding Affibody constructs to block VEGF from binding to 293/KDR cells. Cells were incubated with varying concentrations of each of the biparatopic Affibody molecules, followed by incubation with biotinylated VEGF-A. Binding of VEGF-A was monitored by the binding of fluorescently labeled streptavidin. A concentration-dependent negative shift in fluorescent signal was observed for all four Affibody constructs.

In order to investigate whether the observed VEGF-A blocking of ZVEGFR2_Bp1 -

ZVEGFR2_Bp4 could lead to an inhibition of VEGFR2 signaling, a phosphorylation assay was performed. As described in chapter 4, VEGFR2 signaling is initiated by the binding of VEGF-A to the extracellular domain of VEGFR2, which leads to phosphorylation of the intracellular kinase domain, induced by receptor dimerization. Using an ELISA setup with an antibody specific for the phosphorylation on VEGFR2 Tyr1175, the degree of phosphorylation could be determined. 293/KDR cells were pre-incubated with each Affibody construct followed by VEGF-A stimulation and immediate cell lysis. The cell extract was used in the subsequent ELISA for detection of Tyr1175 phosphorylation. The results from the ELISA showed that all four constructs could inhibit VEGF-induced phosphorylation (Fig. 5.10), and the higher Affibody concentration resulted in a phosphorylation level that was comparable to that of the cells that were not stimulated with VEGF-A. These results are encouraging for future therapeutic applications.

56 Filippa Fleetwood

150 200 nM Affibody Inhibition of phosphorylation construct

) + 1 nM VEGF % ( 20 nM Affibody n o i

t construct a l + 1 nM VEGF y 100 r

o 200 nM Affibody h

p construct + PBS s o h p

f o

n 50 o i t i b i h n I

0

1 2 3 4 r y

dime q Z ta Z VEGFR2_Bp Z VEGFR2_Bp Z VEGFR2_Bp Z VEGFR2_Bp No Affibod

Fig. 5.10 Analysis of the ability of the VEGFR2-binding Affibody constructs to inhibit VEGFR2 phosphorylation on 293/KDR cells. VEGFR2 phosphorylation level was detected using a Tyr1175 phosphorylation-specific antibody. Incubation of 293/KDR cells with 20 nM or 200 nM of each of the four biparatopic Affibody molecules prior to VEGF-A stimulation resulted in a decreased phosphorylation level compared to the incubation with the negative control construct ZTaq-ABD-ZTaq or incubation with no Affibody. A substantial decrease in phosphorylation level was observed for all four constructs.

In order to further investigate the therapeutic potential in terms of the ability to affect angiogenesis, an in vitro angiogenesis assay was performed. The biparatopic Affibody construct with the shortest linker, ZVEGFR2_Bp2, was tested for the ability to inhibit tube formation on a Matrigel matrix. This product is a basement membrane extracellular matrix, which has been extensively used for the studying of formation of capillary- like structures of endothelial cells [388]. Incubation of VEGF-stimulated HUVEC cells on the Matrigel matrix resulted in the formation in a tubular network. The tube formation was substantially decreased for cells incubated with ZVEGFR2_Bp2 compared to cells incubated with the negative control dimeric Ztaq construct (Fig. 5.11).

57 Bacterial display systems for engineering of affinity proteins

a b 500 nM Z -ABD-Z 500 nM ZVEGFR2_Bp2 taq taq

Fig. 5.11 Matrigel assay for evaluation of in vitro capillary-like tube formation by HUVECs. a) HUVECs treated with 500 nM ZVEGFR2_Bp2. b) HUVECs treated with the negative control Ztaq-ABD- Ztaq. A substantial decrease in tube formation was observed for ZVEGFR2_Bp2 compared to Ztaq-ABD-Ztaq.

In summary, the findings of this study demonstrated that a linker length as short as 5 amino acids can be used for connecting the two VEGFR2-binding Affibody molecules ZVEGFR2_22 and ZVEGFR2_40 in a biparatopic construct that is able to bind to VEGFR2 expressed on the surface of human cells, and thereby block VEGF-A binding, which leads to inhibition of VEGFR2 phosphorylation and inhibition of angiogenic effects in vitro. These results suggest that the binding sites for ZVEGFR2_22 and ZVEGFR2_40 are closely situated in space, perhaps on the same domain of VEGFR2. The orientation of the two Affibody domains do not appear to be sterically constrained by a short linker. This construct has potential for in vivo imaging applications as well as for therapeutic applications.

58 Filippa Fleetwood

5.3 PAPER III

Surface display of a single-domain antibody library on Gram-positive bacteria

Staphylococcal display has previously been successfully used for the selection from Affibody libraries (see section 3.2.3). The use of a cell-based display platform has many advantages, mainly due to the possibility to screen the library using FACS. The aims of paper III were (i) to investigate the suitability of the staphylococcal display system for the display of antibody fragments, and (ii) to directly compare the outcome of the selections from the same library using staphylococcal display and phage display. Since the selection from libraries displayed on phage and on a cell surface are based on fairly different principles, the outcome of a selection could potentially differ. In a study by Bowley et al., selections from a library displayed on yeast and on phage were performed in parallel, yielding a broader diversity among the binders selected by yeast display [199]. In paper III, we investigated whether the staphylococcal display system could be used as a complement to existing technologies for the selection of single-domain antibodies.

Nanobodies are derived from the variable domain (VHH) of a type of heavy-chain antibodies (HCAb’s), found in camelids (see section 2.2). HCAb’s are lacking a light chain, but structural adaptations, such as an extended CDR3, have been evolved to compensate for the reduced number of CDRs involved in binding the antigen [55]. Hence, a nanobody library comprises full diversity in a single-domain 15 kDa protein [55]. A small size and a single-domain structure are attractive properties for cell surface display. Nanobodies have also been demonstrated to be efficiently produced in bacteria [55]. In addition, due to the pre-enrichment of target-binding clones during immunization, selections from immune nanobody libraries of around 106 variants usually generate high-affinity binders [55]. This diversity is relatively easy to cover using the staphylococcal display system, for which a typical transformation frequency is around 105-106 transformants per transformation event [227], which means that libraries of 107-108 can be generated in a straightforward manner.

In this study, an immune nanobody library targeting GFP was used for proof-of- principle. The use of a fluorescent target protein provides a direct readout for flow- cytometry, without the need for labeling.

An immune nanobody library was recovered by PCR from the B cell repertoire of a dromedary immunized with Green Fluorescent Protein (GFP). The library was cloned

59 Bacterial display systems for engineering of affinity proteins into the phage display vector pHEN4, and transformed into E. coli for the production of phage, yielding a library of approximately 4×107 transformants. Three rounds of phage against GFP coated in microtiter plates was performed. The selected nanobodies could be categorized into five groups based on their amino acid sequence in the CDR3. One nanobody from each group was chosen for further characterization (P-Nb1, P-Nb2, P-Nb3, P-Nb5 and P-Nb6).

Nanobodies were sub-cloned and expressed in the periplasm of E. coli. The recombinant proteins were extracted by osmotic shock and purified IMAC and gel filtration. GFP binding was analyzed by SPR, and equilibrium constants (KD) were determined to 0.5 nM - 42.5 nM (Table 5.3).

Table 5.3. Equilibrium constants (KD) and GFP fluorescence enhancement effect of GFP-binding nanobodies

Nanobody KD (nM) GFP enhancing effect (%) S-Nb1 0.5 ± 0.2 20 ± 8 S-Nb2 0.25 ± 0.15 47 ± 11 S-Nb3 0.14 ± 0.04 43 ± 7 S-Nb4 10.1 ± 1.2 20 ± 4 S-Nb5 0.21 ± 0.05 45 ± 8 S-Nb6 13.8 ± 2.2 45 ± 21 S-Nb7 3.1 ± 0.3 86 ± 16

Nanobody KD (nM) GFP enhancing effect (%) P-Nb1 3.2 ± 0.05 40 P-Nb2 42.5 ± 1.3 8 P-Nb3 20.5 ±1.0 1 P-Nb5 0.53 ± 0.05 -49 P-Nb6 42.2 ± 5.4 5

As demonstrated by Kirchhofer et al, nanobodies can shift the fluorescence spectrum of GFP upon binding [389]. Among the nanobodies selected by phage display, P-Nb1 was demonstrated to significantly enhance GFP fluorescence (by 40%), and P-Nb5 reduced GFP fluorescence (by 49%), (Table 5.3). This is in concordance with the results of the study by Kirchhofer et al (where the nanobody here denoted P-Nb5 is denoted “minimizer”) [389].

In order to investigate whether nanobodies could be displayed and correctly folded on the surface of S. carnosus, three of the nanobodies selected by phage display (P-Nb1,

60 Filippa Fleetwood

P-Nb3 and P-Nb6) were subcloned into the staphylococcal display vector and transformed to S. carnosus. Flow-cytometric analysis showed that all three nanobodies could be functionally displayed with retained binding to the target protein (Fig. 2b, Appendix 3)

Encouraged by these results, we next investigated the possibility to use staphylococcal display in combination with FACS for the enrichment of GFP-binding nanobodies from an immune library. The library was subcloned from the phage display vector into the staphylococcal display vector, and a library of approximately 107 transformants was created in S. carnosus. The cell-displayed library was subjected to four rounds of FACS, demonstrating an enrichment in GFP-binding cells in each round (Fig. 5.12a). The output population was analyzed by sequencing and on-cell ranking of individual clones. Sequence analysis revealed 27 unique variants (out of 178 sequenced clones), which could be divided into six different groups of nanobodies, corresponding to different clusters in a phylogenetic tree (Fig. 5.12b). Among these six clusters, two contained variants that were homologous to nanobodies selected by phage display (P-Nb1 and P-Nb2). One of these clusters (containing P- Nb2) was dominating the output repertoire, containing 19 unique variants. Interestingly, four clusters were unique for the staphylococcal selection and did not contain any homologous variants from the phage selection. Similarly, three of the nanobodies selected by phage display were unique for the phage selection (P-Nb3, P- Nb5 and P-Nb6).

61 Bacterial display systems for engineering of affinity proteins

a b

4

4 10 1 10 2

1 2

2 2

10 10

3

0

0 10 0 4 10 4

10 102 10 100 102 10 4

4 4 10

3 10 4 2

2 5

10 10

eGFP-binding (FL-1)

0

0 10 0 2 4 10 4 10 10 10 100 102 10

Surface expression level (FL-4)

6

Fig. 5.12 Selection of GFP-specific nanobodies by staphylococcal display a) Sorting of the nanobody library displayed on S. carnosus. Four rounds of flow-cytometric sorting were performed using normalized gating, with an enrichment of GFP-binding clones after each round. In the first three rounds 200 nM GFP was used, and in the last round 10 nM (shown in the figure), 50 nM (data not shown) or 200 nM GFP (data not shown) was used. b) Phylogenetic tree showing clustering and genetic distances between sequences of different selected Nanobodies. S-Nb# denotes a Nanobody selected by staphylococcal display and P-Nb# denotes a Nanobody selected by phage display (indicated in blue). The Nanobodies that were selected for further characterization are indicated in orange. The number of identical clones is indicated after each name. The clusters are denoted cluster 1- 6.

One nanobody from each of the clusters 1-5 and two nanobodies from the dominant cluster 6 were selected for further characterization (denoted S-Nb1, S-Nb2, S-Nb3, S- Nb4, S-Nb5, S-Nb6 and S-Nb7). The GFP binding of these individual clones was verified by flow-cytometry (data not shown). Next, the nanobodies were sub-cloned and expressed in the periplasm of E. coli, followed by purification by IMAC and gel filtration. Equilibrium constants determined by SPR ranged from 140 pM to 13.8 nM (Table 5.3). All nanobodies selected by staphylococcal display were found to enhance GFP fluorescence, with enhancing effects of up to 80% (Table 5.3).

62 Filippa Fleetwood

The results of this study demonstrate that the staphylococcal display platform is suitable as a complement to phage display for selection from single-domain antibody libraries. This is the first report on the display of an antibody-based library on Gram- positive bacteria. The direct comparison of the outcome from the selection from the same library displayed on bacteria and on phage revealed some interesting differences. Both repertoires contained unique variants that were not found using the other selection system, as well as homologous variants. One of the nanobodies that were unique for the phage selection (P-Nb5) reduced GFP fluorescence intensity by almost 50%. Since the selection principle of staphylococcal display system in this case was based on the isolation of cells with a high GFP fluorescence intensity, it is not surprising that this nanobody was not selected by the staphylococcal system. The fact that all nanobodies selected by staphylococcal display were able to enhance GFP fluorescence can also be explained by this bias during the FACS selection. However, even though the selection outcome was influenced by the GFP enhancing effect, the average affinities among the nanobodies selected by staphylococcal display were higher than those of the nanobodies selected by phage display. A plausible reason for this outcome is the fine affinity discrimination enabled by the use of FACS for isolation of target-binding clones. Nevertheless, these results demonstrate that different hosts may generate different repertoires of selected binders. Therefore, we conclude that in order to cover a high degree of diversity, it is important to choose a display system that is suitable for the purpose of the selection, and parallel selections using different display system could potentially broaden the diversity of the total set of isolated binders.

63 Bacterial display systems for engineering of affinity proteins

5.4 PAPER IV

An engineered autotransporter-based surface expression vector enables efficient display of Affibody molecules on OmpT-negative E. coli as well as protease-mediated secretion in OmpT-positive strains

As mentioned in section 2.3.6., selections of high-affinity Affibody molecules have been performed using various display systems [140, 141, 228]. The highest target- affinity of an Affibody generated so far was reported for a HER3-binding Affibody, with an equilibrium constant of 21 pM, which was generated by affinity maturation using staphylococcal display [147]. As mentioned in previous chapters, an important advantage of using cell display for directed evolution is the possibility to use FACS for enrichment of high-affinity binders. This enables fine affinity discrimination and real-time visualization of the selection process. In paper III, higher affinities on average were obtained among the binders selected by staphylococcal display compared to the selections from the same library displayed on phage. These results suggest that cell display can be used as an advantageous complement to phage display, which today is the most established display technology.

However, a limitation of yeast as well as staphylococcal display is the number of library transformants that can be achieved using electroporation (see section 3.2). In phage display, the library size is restricted to the number of transformants when the library is prepared in E. coli. Transformation of DNA to Gram-negative bacteria, such as E. coli, generally results in higher transformation frequencies than yeast and Gram- positive bacteria and yeast [183, 195, 227]. As mentioned in chapter 1, the chance of finding high-affinity binders increase with library size. Therefore, the combination of high transformation frequency and the possibility to use FACS make E. coli an attractive host for directed evolution purposes. In addition, E. coli is a well- characterized host for recombinant protein expression, and has a rapid growth rate [229]. Affibody molecules are usually efficiently expressed in E. coli [94].

Affibody selections using display technologies are usually followed by sub-cloning of the top ranked selected clones for soluble protein expression and downstream characterization. This step is relatively time-consuming and laborious. An efficient and straightforward means of producing and purifying soluble protein candidates would potentially improve the speed and throughput of the engineering process for Affibody molecules.

64 Filippa Fleetwood

The aim of paper IV was to engineer and investigate a dual-purpose expression system for the engineering of Affibody molecules. A set of expression vectors were designed and evaluated. These vectors were intended for (i) surface display of Affibody libraries on E. coli for directed evolution, and (ii) post-selection small-scale soluble protein production by secretion of Affibody candidates into the medium.

Autotransporters are a family of proteins in Gram-negative bacteria, whose natural function is secretion and anchoring of passenger proteins to the outer surface. Since translocation of a recombinant protein through the two membranes and the periplasm of Gram-negative bacteria can be challenging , and autotransporters have shown great potential for display of recombinant proteins [231], we chose to use an autotransporter for our system. The autotransporter Adhesin Involved in Diffuse Adherence (AIDA-I) origins from E. coli, and has previously been used for successful display of several different recombinant passengers, including small libraries for enzyme engineering [263], but to our knowledge it has not been used for selection from affinity protein libraries before.

65 Bacterial display systems for engineering of affinity proteins

a Expression cassette (EC)

Promoter

Signal peptide Affibody His6 Normalization tag AIDA-C

Lead signal OmpT site specific cleavage site Recombinant protein ȕGRPDLQ

b Display Secretion

Affibody molecule

+LVWDJ OmpTVLWH Normalization tag

$,'$& OmpT

E. coli E. coli BL21(DE3) '+Į ǻ2PS7

Fig. 5.13 Schematic illustration of the expression cassette, including the recombinant protein to be displayed on E. coli as well as the optional process for selection from displayed libraries or secreted production. a) The expression cassette (EC) contains a promoter region, a signal peptide for translocation over the inner membrane, an Affibody molecule as binding protein, a His-tag for purification, an OmpT cleavage site for secretion, a surface expression reporter-tag for normalization and the AIDA-C for insertion into the outer membrane. b) The recombinant fusion protein expressed in two different strains of E. coli. In an OmpT negative strain the recombinant protein remains tethered in the outer membrane and thus displayed on the bacterial surface, allowing for phenotype-genotype linkage and FACS. In an OmpT positive strain, OmpT will cleave the recombinant protein and release the Affibody molecule fused to a His-tag into the medium.

We designed an expression vector for surface display of Affibody molecules by fusion to the AIDA-I autotransporter (Fig. 5.13a). An IgG-binding Affibody molecule

(ZIgG) [390] was used as a model Affibody for initial experiments. For flow- cytometric experiments, an Albumin Binding Protein (ABP), derived from Streptococcal Protein G [391] was included C-terminally to the Affibody molecule as a tag for normalization of the target-binding signal against expression level. In order to enable small-scale protein production without sub-cloning, an Outer Membrane Protease T (OmpT) cleavage site was inserted between the Affibody molecule and the

ABP tag. A His6-tag was included at the C-terminal end of the Affibody molecule in order to facilitate purification directly from the supernatant. Transformation of

66 Filippa Fleetwood selected candidates from an Affibody library to an OmpT positive strain of E. coli, would potentially enable rapid and straightforward small scale protein production (Fig. 5.13b).

Initial experiments were performed using the pMK90 vector [392] containing the aidA promoter, which is a constitutive promoter. The expression vector was transformed to OmpT negative E. coli BL21(DE3) cells, and the functional display of

ZIgG as well as ABP was verified by flow cytometry using fluorescently labeled IgG and human serum albumin (HSA) (Fig. 5.14a). a b 10

paidA-ZIgG-EC Expression cassette (EC) 10 paidA-Z-EC pRhaBAD-Z-EC

10 pAraBAD-Z-EC pT7-Z-EC

Target binding IgG(log FI) Target 10

10 10 10 10 Surface Expression Level HSA (log FI) c d 10 10 pRhaBAD-ZIgG-EC pAraBAD-ZIgG-EC

10 10

10 10

10 10 Target binding IgG (log FI) Target binding IgG(log FI) Target

10 10 10 10 10 10 10 10 Surface Expression Level Surface Expression Level HSA (log FI) HSA (log FI) Fig. 5.14 Different inducible promoters compared to the constitutive aidA promoter for surface expression of recombinant protein. a) Representative dot plot from flow-cytometric analysis of E. coli BL21 (paidA-ZIgG-EC). Fluorescence intensity corresponding to IgG-binding on the y-axis and fluorescence intensity corresponding to surface expression level (HSA-binding) on the x-axis. b) The different inducible promoters AraBAD, RhaBAD and T7 before the expression cassette (EC). c) Representative dot plot from flow-cytometric analysis of E. coli BL21 (pRhaBAD-ZIgG-EC). d) Representative dot plot from flow-cytometric analysis of E. coli BL21 (pAraBAD-ZIgG-EC).

In order to avoid toxicity caused by overexpression of recombinant protein on the surface, which could complicate future library work by introducing growth biases, three inducible promoters were investigated: T7, araBAD and rhaBAD (Fig. 5.14b). The two latter promoters are titratable, which enables fine tuning of protein

67 Bacterial display systems for engineering of affinity proteins expression level [393, 394], whereas the T7 promoter is very strong and practically non-titratable [395]. Cultivation of cells containing the T7 promoter resulted in decreased cell viability and growth rate, and occasional decrease in surface expression level (data not shown). The T7 promoter was therefore not included in further studies. In contrast, the araBAD and rhaBAD promoters both resulted in stable, functional display, without any observed toxicity issues (Fig. 5.14c-d).

Cultivation conditions were investigated by flow cytometric analysis after induction at different temperatures (25 °C, 30 °C and 37 °C) for different amounts of time (1 h, 3 h, 6 h and 16 h), and using different concentrations of L-arabinose and L-rhamnose for induction (0.01%, 0.1%, 0.2%, 0.4%, 0.6%, 0.8% and 1.0%). The results demonstrated that induction could be performed at either of the three temperatures tested, for at least 3 hours and using at least 0.1% of inductor (Additional file 1: Figure S1, Appendix 4). In general, higher concentrations of inductor yielded higher expression levels. Induction at higher temperature resulted in only a slightly elevated expression level, and a larger population of non-expressing cells was observed for arabinose-induced cultures. This observation could be an indication of toxicity caused by overexpression of displayed protein. Since addition of glucose to the medium is required for tight regulation of the rhaBAD promoter [394], time-consuming assays like sorting of large libraries could be complicated by a potential risk of glucose depletion. In order to minimize such risks, we used the araBAD promoter in the following evaluations of the technology for library purposes.

Before evaluating the system for library purposes, three Affibody molecules with different target specificities (targeting TNF-α, HER3 and HER2, respectively) were cloned into the expression vector for verification of functional display. All three Affibody molecules were efficiently displayed on the bacterial cell surface, and demonstrated retained target binding. These results are encouraging for future selections from Affibody libraries using different targets (Fig. 5.15).

68 Filippa Fleetwood

10 pAraBAD-ZHER2-EC10 pAraBAD-ZHER3-EC 10 pAraBAD-ZTNFα-EC

10 10 10

10 10 10

10 10 10 Target binding IgG(log FI) Target (log FI) (log TNF α binding Target Target binding HER3 (log FI) 10 10 10 10 10 10 10 10 10 10 10 10 Surface Expression Level Surface Expression Level Surface Expression Level HSA (log FI) HSA (log FI) HSA (log FI)

Fig. 5.15 Surface expression of different Affibody molecules on E. coli. Representative dot plots from flow-cytometric analysis of E. coli BL21 (pAraBAD-ZHER2-EC), (pAraBAD-ZHER3-EC) and (pAraBAD-ZTNFα-EC), respectively. Fluorescence intensity corresponding to target-binding on the y- axis and fluorescence intensity corresponding to surface expression level (HSA-binding) on the x-axis.

Next, we investigated whether the display system could be used for the enrichment of target-binding cells from an excess of non-binding cells by sorting a mock library of cells displaying a Her2-specific Affibody molecule spiked with ZIgG-displaying cells at a ratio of 1:100,000. One round of sorting resulted in an enrichment factor of around 20,000-fold, as determined by flow-cytometry (Fig. 5.16). In addition, we investigated the cell viability after FACS. The high shear stress induced by high- speed sorting could potentially lead to a decrease in cell viability, which would result in a need for re-transformation steps or a greater need for oversampling of the library, leading to longer sortings. All of the cells that were sorted directly to semi-solid medium yielded colonies after overnight incubation (Fig. 3c. Appendix 4), demonstrating a higher degree of viability than reported in similar studies [236].

69 Bacterial display systems for engineering of affinity proteins

Sorting 1:100 000 10 10 UR < 1 % Gate UR > 21±11 %

10 10 1 round FACS

10 10 get binding IgG (log FI) 10 r 10 Target binding IgG (log FI) Ta

10 10 10 10 10 10 10 10 Surface Expression Level Surface Expression Level HSA (log FI) HSA (log FI) Fig. 5.16 Representative dot plots from flow-cytometric sorting of a 1:100,000 mix of E. coli BL21 (pAraBAD-ZIgG-EC) in a background of E. coli BL21 (pAraBAD-ZHER2-EC). Leftmost dot plot shows the mixed population before FACS with the sorting gate indicated in the dot plot. Rightmost dot plot shows the enriched bacteria after sorting and overnight growth. The percentage (mean ± standard deviation) of events in the upper right (UR) quadrant is indicated in the dot plots. The entire experiment was performed in duplicates on different days, using freshly prepared mixtures of bacteria expressing ZHER2 and ZIgG, and freshly prepared reagents.

Finally, the expression system was evaluated for small-scale production of soluble Affibody molecules using an OmpT positive E. coli host strain. Flow-cytometric analysis showed a reduction in target binding signal in from OmpT-positive cells, and IMAC purification of Affibody molecules from the medium resulted in approximately 3 mg purified protein per liter cell culture, and a pure product of the correct size (Fig. 5.17a-c), with retained functionality (Fig. 5.17d).

70 Filippa Fleetwood ab pAraBAD-Z -EC pAraBAD-Z -EC BL21 DH5α IgG IgG Induced BL21(DE3) DH5α +- +- kDa

45.0

30.0

20.1

14.4 Target binding IgG (log FI) Target binding IgG (log FI)

Surface Expression Level Surface Expression Level ZIgG HSA (log FI) HSA (log FI) (9.7 kDa) c d

Cleaved ZIgG ZIgG over immobilized IgG 6 150 x10 9711.6 12 5 nM ZIgG 100 2.5 nM ZIgG 8 1 nM ZIgG 50 Counts 4 Response units (RU) 0 0 6000 7000 8000 9000 10000 11000 0200 400 Molecular weight (Da) Time (s) Fig. 5.17 Evaluation of OmpT-mediated release and small-scale purification of displayed Affibody molecule. a) Flow-cytometric analysis for comparison of surface expression of recombinant proteins between pAraBAD-ZIgG-EC in OmpT-negative E. coli BL21 and pAraBAD-ZIgG-EC in OmpT- positive E. coli DH5α. b) SDS-PAGE showing IMAC-purified supernatants from pAraBAD-ZIgG-EC in OmpT-negative E. coli BL21 and pAraBAD-ZIgG-EC in OmpT-positive E. coli DH5α. Supernatants from non-induced samples, indicated with (-), were included as controls. c) Mass spectrum from ESI- MS analysis of IMAC-purified supernatants from pAraBAD-ZIgG-EC in OmpT-positive E. coli DH5α. Theoretical molecular weight of the OmpT-cleaved ZIgG-His6 is 9712 Da. d) SPR-based biosensor analysis on the IMAC-purified ZIgG-His6. Response units (RU) on the y-axis and time on the x-axis. Representative sensorgrams from injection of ZIgG-His6 at three different concentrations over human IgG immobilized on the chip surface. Injections were performed in duplicates.

In conclusion, our results suggest that the dual-purpose AIDA-I based expression system is promising for selection of target-specific Affibody molecules as well as small-scale protein production by OmpT-mediated secretion. The combination of the high transformation frequency of E. coli, the high enrichment factor in the mock selection, the high cell viability and the use of a single expression vector without the need for subcloning for downstream characterization gives this expression system a potential to increase the throughput of the entire engineering process for Affibody molecules.

71 Bacterial display systems for engineering of affinity proteins

5.5 CONCLUDING REMARKS AND FUTURE OUTLOOK

This thesis describes the development, evaluation and use of bacterial display technologies for the engineering of specific affinity proteins.

In paper I, a staphylococcal display system was used for the affinity maturation of two VEGFR2-specific Affibody molecules. A dramatic increase in affinity was obtained by the engineering of these two Affibody molecules into a biparatopic Affibody construct. If an even further increase in affinity would be desired, an additional affinity maturation effort could be performed for each of the two Affibody molecules individually, or fused in the biparatopic construct. Nevertheless, the constructs based on the two Affibody molecules selected in paper I demonstrated dissociation rate constants in the same range as a HER2-specific Affibody molecule with a KD of 22 pM. The biparatopic binder could also bind specifically to VEGFR2 on the surface of mammalian cells and inhibit VEGFR2 phosphorylation. In paper II, new variants of the biparatopic construct were designed in order to investigate the effect of the linker length. These constructs were also demonstrated to block VEGF-A from binding to the receptor, and thereby inhibit VEGFR2 phosphorylation an tubular network formation of endothelial cells in vitro.

The results presented in paper II demonstrate that the VEGFR2-binding and VEGF-A blocking activity is retained using linker lengths as short as five amino acids. The somewhat surprising observation that no sterical constraint was introduced indicates that the epitopes on VEGFR2 are situated very close to each other, and that the orientation of the Affibody molecules in the construct is suitable for binding to both epitopes simultaneously. It would be interesting to determine the structure of the biparatopic binder in complex with VEGFR2 to verify this. For imaging purposes, the cellular effects are generally irrelevant, whereas the target affinity and the size of the tracer are important [100]. However, as discussed earlier, a small size of the tracer might not be as important for imaging of angiogenesis, since no extravasation from the blood and tumor penetration is required. Therefore, it would be interesting to compare the constructs tested in paper II, expressed without the ABD035, for imaging. For therapeutic applications, however, the biparatopic construct should be expressed in fusion to the ABD035 tag (or similar) for an increased in vivo half-life. The shortest linker might be advantageous due to potentially higher serum stability.

The next step towards imaging studies will be the production of the biparatopic binder

(initially ZVEGFR2_Bp2) without the ABD035 tag, coupled to a C-terminal cysteine for site-specific radionuclide labeling. This tracer will be used for initial preclinical in

72 Filippa Fleetwood vivo imaging experiments. The benefits of a tracer for imaging of angiogenesis include the possibility to identify non-responders to antiangiogenic therapy at an early stage (see section 4.3). This could potentially improve the therapeutic outcome and save resources.

For in vitro investigation of therapeutic potential, the biparatopic Affibody construct was tested for the ability to inhibit tube formation on a Matrigel matrix (in paper II). For further investigation, the same matrix can also be used for studies of vasculature formation in vivo, by implanting it into mice [396].

The results presented in paper I and II indicate that Affibody molecules have potential for blocking of protein-protein interactions and inhibition of signaling pathways. The possibility to engineer a small, biparatopic construct represents an advantage compared to monoclonal antibodies. This is the first example of the engineering of a biparatopic Affibody molecule. This is a relatively fast and straightforward strategy for improvement of target affinity compared to the construction of an affinity maturation library, and could be applied to Affibody molecules targeting other proteins in the future. In order to select for binders to new epitopes, blocking with a previously selected binder during selections could be performed.

In paper III, the staphylococcal display system was evaluated for the display of a nanobody library. This was the first example of display of an antibody-based library on Gram-positive bacteria. The positive results obtained here demonstrate that antibody fragments can indeed be displayed on a Gram-positive bacterial surface, in spite of the presence of disulphide bonds and the larger size compared to Affibody molecules. As mentioned in section 3.2.3, Gram-positive bacteria have some attractive properties for surface display purposes. The results of this study indicate that Gram-positive bacteria can be used for display of a wider range of protein scaffolds than those previously tested. These results have encouraged others to use the staphylococcal display systems for the display of scFv libraries, generating positive results (Hu et al, personal communication). Evaluation of the system for display of even larger affinity proteins would be interesting. The display of Fabs or full-length IgGs could be problematic due to the need for association of the heavy and the light chain. However, other systems based for example on a “secretion and capture” mechanism have been developed in order to solve this problem for yeast display [397], and such solutions could be investigated for staphylococcal display as well. Alternatively, association of the heavy and the light chain could potentially be achieved by separate display of each chain on the same cell.

73 Bacterial display systems for engineering of affinity proteins

Another aim of paper III was to perform a direct comparison of the selections from the same library displayed on different hosts. The selections generated overlapping but slightly different sets of binders, with higher affinities on average among the binders selected by staphylococcal display. The differences in affinity might potentially be a result of the use of flow cytometry in the staphylococcal selection. Whatever the reason, it can be concluded that the choice of selection platform can influence the outcome of the selection, and parallel selections using multiple platforms can be used in order to cover a larger diversity. The question arises whether repeated selections from the same library using the same platform would generate the same set of binders. In theory, this should be the case if the selection system is able to cover the whole library and the selection platform is robust. Such a study is on-going using the staphylococcal display system (Åstrand et al., personal communication), and the outcome of these selections will be interesting.

A larger diversity in the outcome of a selection should in theory be achieved with a larger library. A larger diversity is important for increasing the chances of generating high-affinity binders, but also for selecting proteins with a specified function, such as binding to a certain epitope. A larger library is more likely to generate binders targeting several different epitopes, which could be desirable for example for the possibility to engineer biparatopic constructs, as discussed above. For this purpose, the use of flow-cytometry can potentially be very beneficial, since it enables the isolation of different parts of the binding population using different sorting gates. For example, the binders targeting a certain pre-defined epitope might not be the strongest binders. A range of different sorting gates could be used for dividing the binding population into several sub-populations, which could facilitate the selection of binders targeting different epitopes.

For microbial display platforms, the main factor limiting the library size that can be used is commonly the transformation frequency of DNA to the microorganism. In order to facilitate bacterial display of large Affibody libraries, we investigated the use of an autotransporter for display on E. coli, which has high transformation frequency and is a well-characterized organism for recombinant protein expression. In order to further improve the process of generation of new Affibody molecules, we investigated a strategy for facilitated small-scale soluble protein production for characterization through OmpT-mediated secretion of selected candidates. The dual-purpose expression system was successfully used for selections from a mock library and soluble protein production using an IgG-binding Affibody as a model Affibody. The next step will be to evaluate the system for generation of new Affibody molecules, by selections from a ‘naïve’ Affibody library.

74 Filippa Fleetwood

The new E. coli display system has potential to be used as an attractive complement to staphylococcal display for the display of large Affibody libraries. Using this system, the limiting factor for the size of the library that can be analyzed should no longer be the transformation frequency, but the sorting speed in the flow cytometer. A strategy to get around this limitation could be to perform a round of selection against target immobilized on magnetic beads (magnetic-activated cell sorting, MACS) in order to enrich for binders followed by FACS of the downsized library. Since the advantages of FACS are the most important in later sorting rounds, this should probably not affect the final outcome of the selection. This workflow could potentially be optimized by the implementation of automatization.

Affibody molecules are well suited for the surface display using an autotransporter, since they are small and cysteine-free. The described display system could also be investigated for the display of other scaffolds and antibody fragments. However, problems related to the presence of disulphide bonds might need to be circumvented, for example by the investigation of the use of alternative E. coli strains that lack the possibility to form disulphide bonds in the periplasm, such as DsbA negative strains (as discussed in section 3.2.4.1). Another limitation for the display of antibodies containing certain post-translational modifications could potentially be the lack of a eukaryotic post-translational modification machinery.

75 Bacterial display systems for engineering of affinity proteins

POPULAR SCIENCE SUMMARY

Protein-based drugs is a rapidly growing class of therapeutics, which are promising for a numerous applications. In particular, antibody-based drugs have demonstrated great potential for the treatment of cancer and other diseases. Antibodies are proteins that are naturally produced by the immune system, and are developed to bind with a high affinity (binding strength) to for example foreign molecules, viruses and bacteria. By the use of so-called protein engineering methods, antibodies can be produced in the laboratory and be designed to bind to any molecule of interest. In cancer treatment, a common goal is to generate antibodies that bind specifically to a receptor (a signaling protein molecule) that is only present on the surface of cancer cells, and not on other cells in the body.

A protein that binds with a high affinity and specificity to a target molecule is sometimes called an affinity protein. Antibodies are the most common examples of affinity proteins. Although antibody therapeutics have shown very promising results, there are still many challenges to be addressed. Many of the common problems of antibodies are related to their relatively large size and complex composition. For example, the production of antibodies is relatively complicated and expensive. Also, for some applications, smaller sized molecules with a higher stability than antibodies generate better results. Therefore, a class of affinity proteins termed ‘alternative scaffold proteins’ has emerged. These proteins are commonly designed to overcome some of the limitations of antibodies, and typically have a very small size, a high stability and a cost-efficient production. An example of an alternative scaffold is the Affibody molecule.

A way of generating high-affinity binders to a given target molecule is by so-called combinatorial protein engineering. This method is based on the construction of a large pool of mutant versions of an affinity protein, called a protein library. From this library of billions of different mutated variants of the affinity protein, variants with a high affinity for the target molecule can be ‘fished out’, using the target protein as a kind of ‘bait’.

In order to do this, a robust system which enables the isolation of variants that bind specifically to the target protein is necessary. Also, it should allow for the identification of the DNA sequence encoding that protein variant. The DNA sequence is necessary in order to produce more of the protein candidate, to analyze it further. So-called display systems are typically used for this purpose. A display system is a

76 Filippa Fleetwood way of coupling each protein library member to its corresponding DNA sequence. When a target-binding variant has been ‘fished out’, the DNA sequence encoding that particular variant can be identified.

The PhD projects presented in this thesis have been focused on the development, evaluation and use of a certain type of display system called bacterial display. In these systems, the DNA sequence encoding each affinity protein in a library is attached to bacterial DNA sequence. The bacterial DNA sequence encodes for a protein that is attached to the outer surface of the bacterial cell. By connecting the DNA sequence of the bacterial surface protein to the DNA sequence of the affinity protein, the affinity protein will also be attached to the outer cell surface. Each cell is then said to ‘display’ a single variant of the affinity protein library. Each cell has tens of thousands of copies of one unique affinity protein variant on the surface. If the cells are allowed to interact with the target protein, cells that display a high-affinity variant will have more target protein molecules bound to it. Using fluorescently labeled target protein, the cells that display high-affinity variants can be sorted out from the cells displaying non-binding variants using a methodology called fluorescence-activated cell sorting (FACS). The DNA encoding these high-affinity variants can thereafter be recovered from each isolated cell, and used for further analysis.

In paper I of this thesis, a bacterial display system based on staphylococci, previously developed by our group, was used in order to generate Affibody molecules binding to a cellular receptor called the Vascular Endothelial Growth Factor Receptor 2 (VEGFR2). The signal cascades that are initiated by VEGFR2 are important for regulating the formation of new blood vessels, which is a critical step in the development of tumors and other diseases such as eye diseases. The long-term aim is to use these Affibody molecules as i) diagnostics, in order to identify new blood vessel formation in for example cancer, and ii) as therapeutics, in order to stop blood vessel formation in tumors (and other diseases) and thereby inhibit the progression of the disease. The results showed that the Affibody molecules could bind with a very high affinity to VEGFR2. In paper I and II, these Affibody molecules were also demonstrated to bind to mammalian cells that had VEGFR2 on their surface, and to inhibit VEGFR2 phosphorylation (which is the first step in VEGFR2 signaling), and also to inhibit angiogenesis-like tube formation of the cells. These results suggest that these Affibody molecules have a great potential as therapeutics and diagnostics.

In paper III, the same bacterial display system that was used in paper I was evaluated for the display of an antibody-based library. Previously, the display system had only been used for display of small alternative scaffolds like the Affibody molecule, but

77 Bacterial display systems for engineering of affinity proteins display of an antibody-based library is more challenging due to the more complex structure of antibodies. This was the first time that the display system had been used for antibodies. A type of antibodies called nanobodies were used in this project. The results were successful and high-affinity nanobodies were isolated, demonstrating that the staphylococcal display system is promising for the selection from antibody-based libraries. In this study, the staphylococcal display system was also compared to another display system, called phage display. Phage display is the most established display system today, and is similar to bacterial display, but uses a type of virus called phages instead of bacteria. Our results demonstrated that the staphylococcal display system generated binders with higher affinity towards the target protein than phage display, indicating that staphylococcal display can be an attractive alternative to phage display. Our hypothesis is that the higher affinity was due to the use of the FACS technology, which is a very powerful tool for isolation of binders. Phages are too small to be sorted using FACS, which can potentially make it more challenging to discriminate between the very high-affinity binders and the moderate-affinity binders.

In paper IV, a new display system for the display of Affibody libraries was developed and evaluated. This system was based on E. coli. An important advantage of using this bacterium (compared to many other cell-based display systems) is that it enables the use of much larger libraries. This is because the insertion of the DNA encoding the library members into the bacterial cells is more efficient in E. coli. A larger library increases the chances of finding many high-affinity binders, and is therefore an important advantage for combinatorial protein engineering. An additional feature of the display system presented in paper IV was the possibility to very efficiently produce protein for further characterization when a cell displaying a high-affinity variant has been isolated. This was done by cleaving the affinity protein directly from the bacterial surface using a specific enzyme. The common way of producing protein for further characterization is to isolate the DNA from inside the cells, and use it produce the protein by other means. Since this is relatively time-consuming and laborious, our approach offers another important advantage, in addition to the possibility to use larger libraries.

In summary, the work presented in this thesis describes the development and evaluation of bacterial display systems for combinatorial protein engineering purposes, and also provides successful examples of the use of bacterial display for engineering of affinity proteins.

78 Filippa Fleetwood

ACKNOWLEDGEMENTS

Först och främst vill jag tacka min huvudhandledare John Löfblom. Tack för ditt stora tålamod, ditt otroliga engagemang och att du alltid tar dig tid att diskutera små och stora frågor, oberoende av hur mycket du själv har att göra. Aldrig har jag sett dig stressad eller irriterad. Du kan väldigt mycket om allt från labbande till att ”knorra till” pek och ända sedan mitt exjobb har du alltid varit mycket pedagogisk! Du har lärt mig (nästan) allt jag kan!

Tack också till min bihandledare Stefan Ståhl för att du tog in mig i din härliga grupp, och för all coachning genom åren! Tack för din stora entusiasm och din förmåga att få en att tro på sig själv. Efter ett möte med dig är man alltid glad och taggad, oavsett hur motigt det kändes innan. Tack också för din generositet och positiva inställning till konferenser, kurser och internship utomlands.

Fredrik Frejd, min andra bihandledare, tack för bra input i projekten, och bra coachning! Tack för trevliga handledarluncher, karriärsnack och att du hjälpt mig att få en bättre inblick i hur det är att jobba i industrin!

Thank you all other co-authors and collaborators: Serge Muyldermans, Nick Devoogdt, Mireille Pellis, Ulrich Wernery, Ken Andersson, Susanne Klint, Elin Gunneriusson, and Martin Hanze.

Tack till Vetenskapsrådet (VR), ProNova excellence center for protein technology (VINNOVA) och Stiftelsen för Strategisk Forskning (SSF) för finansiering av projekten jag har arbetat med. Tack Gålöstiftelsen, Sixten Gemzéus fond, Apotekarsocieteten, KVA, ProNova och KTHs allmänna resestipendier för resestipendier. Tack till ProNova även för intressanta studiebesök och kul interaktion med andra forskargrupper och företag!

Thank you Dario Neri for letting me visit your lab those three months. It was an extremely valuable time for me. I learned a lot and had so much fun! Thank you also to the Neri lab for making me feel so welcome and helping me with everything!

Tack till mina duktiga exjobbare Ken Andersson och Martin Hanze. Tack också Ken för ett utmärkt trevligt samarbete under resten av doktorandtiden!

Tack till alla er andra som på ett eller annat sätt indirekt bidragit till de resultat som presenteras i den här avhandlingen, särskilt Tarek Bass för hjälp med fosforyleringsassayn och PDB-figurer, och Tetyana Tegnebratt för hjälp med Matrigel-assayn. Tack Gen Larsson och Martin Gustavsson för plasmiden med AIDA-I, olika E. coli-stammar, bidrag till felsökning vid utvecklingen av E. coli-displaysystemet, m.m.

Tack alla ni på Affibody AB som har hjälpt mig på olika sätt: Fredrik, Susanne, Elin, Caroline, Per och Lindvi.

Tack Torbjörn och Per-Åke för input i mina projekt och granskande vid 2-års-seminarium och Torbjörn även för granskande av avhandlingen.

Tack till medlemmarna av Sel-tag imaging-projektet (särskilt Sharon Stone-Elander och Erik Samén) för input och diskussioner i VEGFR2-projektet. Tack också till Vladimir Tolmachev och Anna Orlova för diskussioner och hjälp kring detta. Tack Kristian Pietras för cell-linjer.

79 Bacterial display systems for engineering of affinity proteins

Tack Christina Burtsoff-Asp från GE Healthcare för att du ville ställa upp som min mentor i IVAs mentorprogram Mentor4Research. Jag har lärt mig mycket om affärsplaner och kommersialisering av forskning tack vare dig!

Tack Bahram, Marika, Lili, Holger, Malin, Jenny, Mehri och alla ni andra i HPA MolBio för hjälp med sekvensering och annat! Tack till er i Proteinfabriken som bjudit på buffertar och tips och råd vid proteinrening. Tack Emma och Lan Lan för hjälp med beställningar.

Tack alla trevliga och glada kollegor på plan 3 för att ni gör det så roligt att gå till jobbet varje dag! Tack för trevliga tillställningar i och utanför labbet – AWs, fester, grillningar, spelkvällar, Oktoberfest…

Tack till alla er som har hjälpt mig och svarat på alla möjliga frågor genom åren, delat med sig av tips och råd och diskuterat problem. Bara för att nämna några: Andreas J, Nina, Magdalena, Johan N, Johan R, Danne, Micke, Sebastian G, Anna P, Hanna L, Lisa, Ken, Tarek och många fler… Listan kan göras mycket lång!

Alla kontorsgrannar genom åren! Magdalena, Johan N, Anna K, Burcu, Chi, Pavan, Andreas, Ali, Amrita, Naresh, Ronnie, Kristina, Veronika, Sebastian M… Tack för trevligt sällskap, och särskilt tack till Andreas och ni andra som har hjälp till att vattna och ge kärlek till chilin! (Just det - Tack Per för chilin!)

Härliga VBG staff-gruppen: Stefan, John, Nina, Helena, Magda, Ronny, Andreas, Hanna, Lisa, Tarek, Ken, Jonas och Sebastian! Tack för mysiga staff-dagar, middagar, lekar, retreats och resor! Är så glad att ha fått vara i en grupp med så många supertrevliga personer!

Tack John, Helena, Sebastian G, och Magda för trevligt sällskap i Keystone! Tack John för att du körde ner oss sakta och säkert från bergspasset. Tack Andreas, John, Johan, Camilla, Lisa och Feifan för häng i San Diego. Tack Anja och Christina för pärl-shopping och goda middagar i Kina. Tack Lisa, Paul och Josefine för kul sightseeing i San Francisco. Tack Hanna, Lisa, John, och Andreas för puderåkning i Whistler, och tack Johan S för charmiga bergspromenader i Italien!

Tack alla Bio-05or för nio trevliga år tillsammans! Johan S, Tove, Hanna, Lisa, Magnus, Anna H, Eva A, Cissi, Erik B, Anders J, Anna L, Staffan… Jag hoppas vi kommer kunna fortsätta ses även när vi inte jobbar tillsammans! Det gäller såklart även er som inte doktorerade! Tack förresten Emily för hjälp med korr-läsning av avhandlingen!

Tack Tove och Andreas för peppande träningssällskap! High five!! Kommer sakna KTH- hallen.

Tack alla tjejerna, särskilt Lisa, Hanna, Tove, Sara, Eva, Emilie, Sahar, Anna och Cissi för trevliga tjejkvällar och AWs! Skönt ibland att umgås med någon som förstår att man kanske kommer ifrån labbet klockan sju även om man sa klockan fem från början.

Tack min kära familj: Mamma, Pappa, Mathilde och Oliver för att ni alltid finns där och allt roligt vi har gjort tillsammans. Jag är så glad att jag har er. Thanks Joaquín for showing a special interest in my research, och tack favoriten Lill-Jocke för att du alltid gör mig så glad!

Och slutligen, tack min älskade fästman Anders! Hur ska jag kunna tacka dig nog? Du har varit helt fantastisk med all matlagning och hushållsarbete under avhandlingsskrivandet, allt stöd och uppmuntran när det har känts motigt, och hjälp med granskning av texter och figurer av olika slag. Tack för att du har stått ut med mig under avhandlingsskrivandet. Tack för allt roligt vi har gjort tillsammans. Men framför allt: tack för att du lyser upp varje dag med att bara finnas där när jag kommer hem på kvällen. Jag älskar dig! Längtar till augusti!

80 Filippa Fleetwood

REFERENCES

1. Demain, A.L. and P. Vaishnav, Biotechnol Adv, 2009. 27(3): 297-306. 2. Leader, B.B., Q. J. & Golan, D. E, Nature Rev. Drug Discov, 2008. 7: 21-39. 3. Goeddel, G., et al., P Natl Acad Sci USA, 1979. 76(1): 106-110. 4. Hutchison, C.A., S. Phillips, and M. Edgell, J Biol Chem, 1978. 253(18): 6551-6560. 5. Carter, P.J., Exp Cell Res, 2011. 317(9): 1261-9. 6. Ståhl, S., et al., Journal of Chemical Technology & Biotechnology, 2013. 88(1): 25-38. 7. Kindt, T.J., R.A. Goldsby, and B.A. Osborne, Immunology. 6 ed2007, New York: W. H. Freeman and Company. 8. Carter, P.J., Nature Reviews Immunology, 2006. 6(5): 343-357. 9. Sievers, E.L. and P.D. Senter, Annu Rev Med, 2013. 64: 15-29. 10. Britz-Cunningham, S.H. and S.J. Adelstein, The Journal of Nuclear Medicine, 2003. 2003(44): 1945-1961. 11. De Groot, A.S. and D.W. Scott, Trends in Immunology, 2007. 28(11): 482-90. 12. Goswami, S., et al., Antibodies, 2013. 2(3): 452-500. 13. Cedrone, F., A. Ménez, and E. Quéméneur, Current Opinion in Structural Biology, 2000. 10: 405-410. 14. Hytonen, V.P., et al., J Biol Chem, 2005. 280(11): 10228-33. 15. Koder, R.L., et al., Nature, 2009. 458(7236): 305-9. 16. Terskikh, A.V., et al., P Natl Acad Sci USA, 1997. 94: 1663-1668. 17. Luo, J., J. Choulet, and J.C. Samuelson, Protein Sci, 2009. 18(8): 1735-44. 18. Malm, M., et al., Biotechnology journal, 2014. 9(9): 1215-22. 19. Karanicolas, J., et al., Mol Cell, 2011. 42(2): 250-60. 20. Kiss, G., et al., Angew Chem Int Ed Engl, 2013. 52(22): 5700-25. 21. Khare, S.D. and S.J. Fleishman, FEBS Letters, 2013. 587(8): 1147-54. 22. Kortemme, T. and D. Baker, Current Opinion in Chemical Biology, 2004. 8(1): 91-7. 23. Pantazes, R.J., M.J. Grisewood, and C.D. Maranas, Current Opinion in Structural Biology, 2011. 21(4): 467-72. 24. Schreiber, G. and S.J. Fleishman, Current Opinion in Structural Biology, 2013. 23(6): 903-10. 25. Bradbury, A.R. and J.D. Marks, Journal of Immunological Methods, 2004. 290(1-2): 29-49. 26. Sergeeva, A., et al., Advanced Drug Delivery Reviews, 2006. 58(15): 1622-54. 27. Hoogenboom, H.R., Trends Biotechnol, 1997. 15: 62-70. 28. Cirino, P.C., K.M. Mayer, and D. Umeno, Generating Mutant Libraries Using Error-Prone PCR, in Directed Evolution Library Creation, F.H. Arnold and G. Georgiou, Editors. 2003, Humana Press Inc. p. 3-11. 29. Arunachalam, T.S., et al., Org Biomol Chem, 2012. 10(24): 4641-50. 30. Neylon, C., Nucleic Acids Res, 2004. 32(4): 1448-1459. 31. Lahr, S.J., A. Broadwater, and J. Carter, C. W., P Natl Acad Sci USA, 1999. 96(26): 14860- 18865. 32. Ashraf, M., et al., Biochem Soc Trans, 2013. 41(5): 1189-94. 33. Sondek, J. and D. Shortle, P Natl Acad Sci USA, 1992. 89: 3581-3585. 34. Virkenäs, B., et al., Nucleic Acids Res, 1994. 22(25): 5600-5607. 35. Fellouse, F.A., et al., Journal of Molecular Biology, 2005. 348(5): 1153-62. 36. Fellouse, F.A., C. Wiesmann, and S.S. Sidhu, P Natl Acad Sci USA, 2004. 101(34): 12467- 72. 37. Koide, S. and S.S. Sidhu, Acs Chem Biol, 2009. 4(5): 325-34. 38. Stemmer, W.P., P Natl Acad Sci USA, 1994. 91: 10747-10751. 39. Crameri, A., Nature, 1998. 391: 288-291. 40. Fermér, I., et al., Tumour Biology, 2004. 25(1-2): 7-13. 41. Chica, R.A., N. Doucet, and J.N. Pelletier, Current Opinion in Biotechnology, 2005. 16(4): 378-384. 42. Barderas, R., et al., Affinity Maturation by Semi-rational Approaches., in Antibody Engineering, P. Chames, Editor 2012, Humana Press Inc. 43. Whitehead, T.A., D. Baker, and S.J. Fleishman, Methods Enzymol, 2013. 523: 1-19. 44. Weiner, L.M., R. Surana, and S. Wang, Nat Rev Immunol, 2010. 10(5): 317-27. 45. Roopenian, D.C. and S. Akilesh, Nat Rev Immunol, 2007. 7(9): 715-25. 46. Kohler, G. and C. Milstein, Nature, 1975. 256(5517): 495-497. 47. Holliger, P. and P.J. Hudson, Nature Biotechnology, 2005. 23(9): 1126-36.

81 Bacterial display systems for engineering of affinity proteins

48. Holliger, P., T. Prospero, and G. Winter, P Natl Acad Sci USA, 1993. 90: 6444-6448. 49. Ward, E.S., et al., Nature, 1989. 341: 544-546. 50. Colby, D.W., et al., Journal of Molecular Biology, 2004. 342(3): 901-12. 51. Chen, W., et al., Construction of a Human Antibody Domain (VH) Library, in Therapeutic Antibodies, A.S. Dimitrov, Editor 2009, Humana Press. p. 81-100. 52. Chen, W., et al., Journal of Molecular Biology, 2008. 382(3): 779-89. 53. Chen, W., et al., P Natl Acad Sci USA, 2008. 105(44): 17121-6. 54. Kim, D.Y., et al., mAbs, 2014. 6(1): 219-35. 55. Muyldermans, S., Journal of Biotechnology, 2001. 74(4): 277-302. 56. Nguyen, V.K., et al., Immunogenetics, 2002. 54(1): 39-47. 57. Bujak, E., et al., Reformatting of scFv Antibodies into the scFv-Fc Format and Their Downstream Purification., in Monoclonal Antibodies, V. Ossipow and N. Fischer, Editors. 2014, Humana Press. p. 315-336. 58. Steiner, M. and D. Neri, Clin Cancer Res, 2011. 17(20): 6406-16. 59. Perez, H.L., et al., Drug Discov Today, 2014. 19(7): 869-81. 60. Madhumathi, J. and R.S. Verma, Curr Opin Microbiol, 2012. 15(3): 300-9. 61. Pasche, N. and D. Neri, Drug Discov Today, 2012. 17(11-12): 583-90. 62. Müller, D. and R.E. Kontermann, Biodrugs, 2010. 24(2): 89-98. 63. Idusogie, E.E., et al., The Journal of Immunology, 2001. 166(4): 2571-2575. 64. Acqua, W.F.D., et al., The Journal of Immunology, 2002. 169(9): 5171-5180. 65. Moore, G.L., et al., mAbs, 2010. 2(2): 181-189. 66. Lazar, G.A., et al., P Natl Acad Sci USA, 2006. 103(11): 4005-10. 67. Harding, F.A., et al., mAbs, 2010. 2(3): 256-65. 68. Beck, A., et al., Nat Rev Immunol, 2010. 10(5): 345-52. 69. Frokjaer, S. and D.E. Otzen, Nature reviews. Drug discovery, 2005. 4(4): 298-306. 70. Conrath, K.E., et al., Developmental and comparative immunology, 2003. 27(2): 87-103. 71. Hamers-Casterman, C., et al., Nature, 1993. 363(6428): 446-448. 72. Muyldermans, S., et al., Veterinary Immunology and Immunopathology, 2009. 128(1-3): 178- 83. 73. Hassanzadeh-Ghassabeh, G., et al., Nanomedicine, 2013. 8(6): 1013-1026. 74. van der Linden, R.H., et al., Journal of Biotechnology, 2000. 80(3): 261-270. 75. Arbabi Ghahroudi, M., et al., FEBS Letters, 1997. 414(3): 521-526. 76. Rahbarizadeh, F., et al., Protein Expression and Purification, 2005. 44(1): 32-8. 77. Vu, K.B., et al., Mol Immunol, 1997. 34(16-17): 1121-1131. 78. Harmsen, M.M. and H.J. De Haard, Applied Microbiology and Biotechnology, 2007. 77(1): 13-22. 79. Lauwereys, M., et al., The EMBO journal, 1998. 17(13): 3512-3520. 80. Desmyter, A., et al., Nat Struct Biol, 1996. 3(9): 803-811. 81. Yau, K.Y.F., et al., Journal of Immunological Methods, 2003. 281(1-2): 161-175. 82. Verheesen, P., et al., Biochim Biophys Acta, 2006. 1764(8): 1307-19. 83. Goldman, E.R., et al., Analytical Chemistry, 2006. 78(24): 8245-8255. 84. Saerens, D., G.H. Ghassabeh, and S. Muyldermans, Current Opinion in Pharmacology, 2008. 8(5): 600-8. 85. Oliveira, S., et al., J Control Release, 2013. 172(3): 607-17. 86. Kolkman, J.A. and D.A. Law, Drug Discov Today Technol, 2010. 7(2): e95-e146. 87. Conrath, K., et al., Journal of Molecular Biology, 2005. 350(1): 112-25. 88. Vincke, C., et al., Journal of Biological Chemistry, 2009. 284(5): 3273-3284. 89. Verheesen, P., et al., Biochimica et Biophysica Acta (BBA) - General Subjects, 2003. 1624(1- 3): 21-28. 90. Pleschberger, M., et al., Bioconjug Chem, 2004. 15: 664-671. 91. Dolk, E., et al., Applied and Environmental Microbiology, 2005. 71(1): 442-50. 92. Nygren, P.A. and A. Skerra, Journal of Immunological Methods, 2004. 290(1-2): 3-28. 93. Skerra, A., Current Opinion in Chemical Biology, 2003. 7(6): 683-693. 94. Lofblom, J., et al., FEBS Letters, 2010. 584(12): 2670-80. 95. Feldwisch, J. and V. Tolmachev, Engineering of Affibody Molecules for Therapy and Diagnostics, in Therapeutic Proteins, V. Voynov and J.A. Caravella, Editors. 2012, Humana Press. p. 103-126. 96. Tran, T., et al., Bioconjug Chem, 2007. 18: 549-558. 97. Gebauer, M. and A. Skerra, Current Opinion in Chemical Biology, 2009. 13(3): 245-55. 98. Kolmar, H. and A. Skerra, Febs J, 2008. 275(11): 2667.

82 Filippa Fleetwood

99. Beckman, R.A., L.M. Weiner, and H.M. Davis, Cancer, 2007. 109(2): 170-9. 100. Tolmachev, V., et al., Expert Opin Biol Ther, 2007. 7(4): 555-68. 101. Olafsen, T. and A.M. Wu, Semin Nucl Med, 2010. 40(3): 167-81. 102. Kontermann, R.E., Current Opinion in Biotechnology, 2011. 22(6): 868-76. 103. Andersen, J.T., et al., J Biol Chem, 2011. 286(7): 5234-41. 104. Li, J. and Z. Zhu, Acta Pharmacol Sin, 2010. 31(9): 1198-207. 105. Emanuel, S.L., et al., mAbs, 2011. 3(1): 38-48. 106. Steffen, A.-C., et al., Cancer Biother Radiopharm, 2005. 20(3): 239-248. 107. Roovers, R.C., et al., Int J Cancer, 2011. 129(8): 2013-24. 108. Jost, C., et al., Structure, 2013. 21(11): 1979-91. 109. Lofblom, J., F.Y. Frejd, and S. Stahl, Current Opinion in Biotechnology, 2011. 22(6): 843- 848. 110. Wurch, T., A. Pierre, and S. Depil, Trends Biotechnol, 2012. 30(11): 575-82. 111. Li, J., A. Mahajan, and M.D. Tsai, Biochemistry, 2006. 45(51): 15168-78. 112. Binz, H.K., et al., Journal of Molecular Biology, 2003. 332(2): 489-503. 113. Forrer, P., et al., FEBS Letters, 2003. 539(1-3): 2-6. 114. Tamaskovic, R., et al., Methods Enzymol, 2012. 503: 101-34. 115. Koide, A., et al., Journal of Molecular Biology, 1998. 284: 1141-1151. 116. Lipovsek, D., Protein Engineering Design and Selection, 2011. 24(1-2): 3-9. 117. Hackel, B.J., A. Kapila, and K.D. Wittrup, Journal of Molecular Biology, 2008. 381(5): 1238- 52. 118. Tolcher, A.W., et al., Clin Cancer Res, 2011. 17(2): 363-71. 119. Skerra, A., Febs J, 2008. 275(11): 2677-83. 120. Beste, G., et al., P Natl Acad Sci USA, 1999. 96: 1898-1903. 121. Schlehuber, S. and A. Skerra, Biophysical Chemistry, 2002. 96: 213-228. 122. Richter, A., E. Eggenstein, and A. Skerra, FEBS Letters, 2014. 588(2): 213-8. 123. Schlehuber, S. and A. Skerra, Drug Discov Today, 2005. 10(1): 23-33. 124. Mross, K., et al., PLoS ONE, 2013. 8(12): 1-11. 125. Ackerman, S.E., et al., Expert Rev Proteomics, 2014. 11(5): 561-72. 126. Kolmar, H., Febs J, 2008. 275(11): 2684-90. 127. Zhu, S., et al., The FASEB Journal, 2003. 17(12): 1756-1757. 128. Christmann, A., et al., Protein Eng, 1999. 12(9): 797-806. 129. Moore, S.J. and J.R. Cochran, Methods Enzymol, 2012. 503: 223-51. 130. Silverman, J., et al., Nature Biotechnology, 2005. 23(12): 1556-61. 131. Lee, C.-H., et al., P Natl Acad Sci USA, 2010. 107(21): 9567-9571. 132. Hosse, R.J., A. Rothe, and B.E. Power, Protein Sci, 2006. 15(1): 14-27. 133. Nord, K., et al., Nature Biotechnology, 1997. 15(8): 772-7. 134. Uhlén, M., et al., J Biol Chem, 1984. 259(3): 1695-1702. 135. Nilsson, B., et al., Protein Eng, 1987. 1(2): 107-13. 136. Ren, G., et al., J Nucl Med, 2009. 50(9): 1492-1499. 137. Webster, J.M., et al., Chembiochem, 2009. 10(8): 1293-1296. 138. Grimm, S., S. Salahshour, and P.A. Nygren, New biotechnology, 2012. 29(5): 534-42. 139. Nord, K., et al., Protein Eng, 1995. 8: 601-608. 140. Kronqvist, N., et al., Protein Eng Des Sel, 2008. 21(4): 247-55. 141. Grimm, S., F. Yu, and P.A. Nygren, Mol Biotechnol, 2011. 48(3): 263-76. 142. Nord, O., M. Uhlén, and P.-Å. Nygren, Journal of Biotechnology, 2003. 106(1): 1-13. 143. Lofdahl, P.A., et al., New biotechnology, 2009. 26(5): 251-9. 144. Hogbom, M., et al., P Natl Acad Sci USA, 2003. 100(6): 3191-6. 145. Myers, J.K. and T.G. Oas, Nat Struct Biol, 2001. 8(6): 552-8. 146. Orlova, A., et al., Cancer Res, 2006. 66(8): 4339-48. 147. Malm, M., et al., PLoS ONE, 2013. 8(5). 148. Wallberg, H. and A. Orlova, Cancer Biother Radiopharm, 2008. 23(4): 435-42. 149. Wållberg, H., et al., Molecular imaging and biology, 2010. 12(1): 54-62. 150. Tolmachev, V., et al., European journal of nuclear medicine and molecular imaging, 2009. 37(3): 613-622. 151. Orlova, A., et al., European journal of nuclear medicine and molecular imaging, 2010. 37(2): 260-9. 152. Sorensen, J., et al., J Nucl Med, 2014. 55(5): 730-5. 153. Zielinski, R., et al., J Immunother, 2009. 32(8): 817-25. 154. Tolmachev, V., et al., Cancer Res, 2007. 67(6): 2773-82.

83 Bacterial display systems for engineering of affinity proteins

155. Myhre, S., et al., Gene Ther, 2009. 16(2): 252-61. 156. Alexis, F., et al., ChemMedChem, 2008. 3(12): 1839-43. 157. Beuttler, J., et al., Bioconjug Chem, 2009. 20: 1201-1208. 158. Nord, K., et al., Journal of Biotechnology, 2000. 80: 45-54. 159. Andersson, C., et al., Biotechnol. Appl. Biochem., 2001. 34: 25-32. 160. Engfeldt, T., et al., Chembiochem, 2005. 6(6): 1043-50. 161. Lyakhov, I., et al., Chembiochem, 2010. 11(3): 345-50. 162. Lundberg, E., et al., Journal of Immunological Methods, 2007. 319(1-2): 53-63. 163. Smith, G.P. and V.A. Petrenko, Chem Rev, 1997. 97(2): 391-410. 164. Barbas, C.F., 3rd, et al., Phage Display: A laboratory manual2001: Cold Spring Harbor Laboratory Press. 165. Smith, G.P., Science, 1985. 228(4705): 1315-1317. 166. McCafferty, J., et al., Nature, 1990. 348: 552-554. 167. Scott, J.K. and G.P. Smith, Science, 1990. 249(4967): 386-390. 168. Clackson, T., et al., Nature, 1991. 352: 624-628. 169. Hoogenboom, H.R., et al., Immunotechnology, 1998. 4: 1-20. 170. Hoogenboom, H.R., et al., Nucleic Acids Res, 1991. 19(15): 4133-4137. 171. Breitling, F., et al., Gene, 1991. 104: 147-153. 172. Bass, S., R. Greene, and J.A. Wells, Proteins, 1990. 8: 309-314. 173. Marks, J.D., et al., J Biol Chem, 1992. 267(23): 16007-16010. 174. Soltes, G., et al., Journal of Biotechnology, 2007. 127(4): 626-637. 175. Dantas-Barbosa, C., M. de Macedo Brigido, and A.Q. Maranhao, Int J Mol Sci, 2012. 13(5): 5420-40. 176. Mazor, Y., et al., Febs J, 2010. 277(10): 2291-303. 177. Schlehuber, S., G. Beste, and A. Skerra, Journal of Molecular Biology, 2000. 297(1105-1120). 178. Pershad, K. and B.K. Kay, Methods, 2013. 60(1): 38-45. 179. Jung, S., A. Honegger, and A. Plückthun, Journal of Molecular Biology, 1999. 294: 163-180. 180. Bostrom, J., et al., Science, 2009. 323: 1610-1614. 181. Pedersen, H., et al., P Natl Acad Sci USA, 1998. 95: 10523-10528. 182. Arap, W., et al., Nat Med, 2002. 8(2): 121-127. 183. Sidhu, S.S., et al., Method Enzymol, 2000. 328: 333-363. 184. Imai, S., et al., Pharmazie, 2008. 63(10): 760-764. 185. Maass, A., et al., Protein Eng Des Sel, 2014. 27(2): 41-7. 186. Doerner, A., et al., FEBS Letters, 2014. 588(2): 278-87. 187. Lofblom, J., Biotechnology journal, 2011. 6(9): 1115-1129. 188. Lofblom, J., H. Wernerus, and S. Stahl, FEMS Microbiology Letters, 2005. 248(2): 189-98. 189. VanAntwerp, J.J. and K.D. Wittrup, Biotechnology progress, 2000. 16(1): 31-37. 190. Garcia-Rodriguez, C., et al., Nature Biotechnology, 2007. 25(1): 107-16. 191. Lofblom, J., et al., Applied and Environmental Microbiology, 2007. 73(21): 6714-21. 192. Francisco, J.A., et al., P Natl Acad Sci USA, 1993. 90(22): 10444-10448. 193. Boder, E.T. and K.D. Wittrup, Nature Biotechnology, 1997. 15(6): 553-557. 194. Feldhaus, M.J. and R.W. Siegel, Journal of Immunological Methods, 2004. 290(1-2): 69-80. 195. Benatuil, L., et al., Protein Engineering Design & Selection, 2010. 23(4): 155-159. 196. Gai, S.A. and K.D. Wittrup, Current Opinion in Structural Biology, 2007. 17(4): 467-473. 197. Boder, E.T., K.S. Midelfort, and K.D. Wittrup, P Natl Acad Sci USA, 2000. 97(20): 10701-5. 198. Shusta, E.V., et al., Journal of Molecular Biology, 1999. 292: 949-956. 199. Bowley, D.R., et al., Protein Eng Des Sel, 2007. 20(2): 81-90. 200. Ho, M., S. Nagata, and I. Pastan, P Natl Acad Sci USA, 2006. 103(25): 9637-42. 201. Akamatsu, Y., et al., Journal of Immunological Methods, 2007. 327(1-2): 40-52. 202. Bowers, P.M., et al., Methods, 2014. 65(1): 44-56. 203. Beerli, R.R., et al., P Natl Acad Sci USA, 2008. 105(38): 14336-41. 204. Zhou, C., et al., mAbs, 2010. 2(5): 508-18. 205. Cumbers, S.J., et al., Nature Biotechnology, 2002. 20(11): 1129-34. 206. Bowers, P.M., et al., P Natl Acad Sci USA, 2011. 108(51): 20455-60. 207. Hansson, M., et al., J Bacteriol, 1992. 174(13): 4239-4245. 208. Pozzi, G., et al., Infection and Immunity, 1992. 60(5): 1902-1907. 209. Kronqvist, N., et al., Recent Pat Biotechnol, 2010. 4(3): 171-82. 210. Nilvebrant, J., et al., PLoS ONE, 2011. 6(10). 211. Fleetwood, F., et al., Cell Mol Life Sci, 2013. 70(6): 1081-93. 212. Lindberg, H., et al., Biotechnology journal, 2013. 8(1): 139-145.

84 Filippa Fleetwood

213. Gotz, F., Soc Appl Bacteriol Symp Ser, 1990. 19: 49S-53S. 214. Fadda, S., et al., FEMS Microbiology Letters, 2002. 210: 209-214. 215. Samuelson, P., et al., J Bacteriol. 177(6): 1470-1476. 216. Sjölander, A., et al., Journal of Immunological Methods, 1997. 201: 115-123. 217. Jonsson, A., et al., Protein Eng Des Sel, 2008. 21(8): 515-27. 218. Kronqvist, N., et al., FEMS Microbiology Letters, 2008. 278(1): 128-36. 219. Nguyen, T.N., et al., Gene, 1993. 128: 89-94. 220. Cano, F., et al., Vaccine, 2000. 18: 2743-2752. 221. Cano, F., et al., FEMS Microbiology Letters, 1999. 25: 289-298. 222. Liljeqvist, S., et al., Applied and Environmental Microbiology, 1997. 63(7): 2481-2488. 223. Liljeqvist, S. and S. Ståhl, Journal of Biotechnology, 1999. 73: 1-33. 224. Samuelson, P., et al., Applied and Environmental Microbiology, 2000. 66(3): 1243-1248. 225. Wernerus, H., et al., Applied and Environmental Microbiology, 2001. 67(10): 4678-4684. 226. Strauss, A. and F. Götz, Mol Microbiol, 1996. 21: 491-500. 227. Lofblom, J., et al., Journal of Applied Microbiology, 2007. 102(3): 736-47. 228. Kronqvist, N., et al., Protein Eng Des Sel, 2011. 24(4): 385-96. 229. Daugherty, P.S., Current Opinion in Structural Biology, 2007. 17(4): 474-80. 230. Jose, J. and T.F. Meyer, Microbiol Mol Biol Rev, 2007. 71(4): 600-19. 231. van Bloois, E., et al., Trends Biotechnol, 2011. 29(2): 79-86. 232. Bessette, P.H., J.J. Rice, and P.S. Daugherty, Protein Engineering Design & Selection, 2004. 17(10): 731-739. 233. Rice, J.J. and P.S. Daugherty, Protein Engineering Design & Selection, 2008. 21(7): 435-442. 234. Harvey, B.R., et al., P Natl Acad Sci USA, 2004. 101(25): 9193-9198. 235. Mazor, Y., et al., Nature Biotechnology, 2007. 25(5): 563-5. 236. Binder, U., et al., Journal of Molecular Biology, 2010. 400(4): 783-802. 237. Daugherty, P.S., et al., Protein Eng, 1998. 11(9): 825-832. 238. Stathopoulos, C., G. Georgiou, and C.F. Earhart, Applied Microbiology and Biotechnology, 1996. 45: 112-119. 239. Yang, C., et al., Applied and Environmental Microbiology, 2008. 74(24): 7733-9. 240. Georgiou, G., et al., Protein Eng, 1996. 9(2): 239-247. 241. Hall, S.S., S. Mitragotri, and P.S. Daugherty, Biotechnology progress, 2007. 23: 749-754. 242. Verhoeven, G.S., et al., PLoS ONE, 2009. 4(8): e6739. 243. Baek, J.H., et al., Applied and Environmental Microbiology, 2010. 76(3): 971-3. 244. Rice, J.J., et al., Protein Sci, 2006. 15(4): 825-836. 245. Dane, K.Y., et al., Journal of Immunological Methods, 2006. 309(1-2): 120-129. 246. Dane, K.Y., C. Gottstein, and P.S. Daugherty, Mol Cancer Ther, 2009. 8(5): 1312-1318. 247. Thomas, J.M. and P.S. Daugherty, Protein Sci, 2009. 18(10): 2053-9. 248. Kenrick, S.A. and P.S. Daugherty, Protein Eng Des Sel, 2010. 23(1): 9-17. 249. Little, L.E., et al., Biomaterials, 2011. 32(6): 1484-94. 250. Getz, J.A., J.J. Rice, and P.S. Daugherty, Acs Chem Biol, 2011. 6(8): 837-844. 251. Henderson, I.R., et al., Microbiol Mol Biol Rev, 2004. 68(4): 692-744. 252. Salema, V., et al., PLoS ONE, 2013. 8(9). 253. Bodelon, G., E. Marin, and L.A. Fernandez, J Bacteriol, 2009. 191(16): 5169-79. 254. Newman, C.L. and C. Stathopoulos, Crit Rev Microbiol, 2004. 30(4): 275-86. 255. Luo, Y., et al., Nature, 2000. 405: 1073-1077. 256. Adams, T.M., A. Wentzel, and H. Kolmar, J Bacteriol, 2005. 187(2): 522-33. 257. Christmann, A., et al., Journal of Immunological Methods, 2001. 257(1-2): 163-173. 258. Wentzel, A., et al., J Bacteriol, 2001. 183(24): 7273-84. 259. Henderson, I., Navarro-Garcia, F. and Nataro, J. P., Trends in Immunology, 1998. 6(9): 370- 378. 260. Halter, R., J. Pohlner, and T.F. Meyer, The EMBO journal, 1984. 3(7): 1595-1601. 261. Pohlner, J., et al., Nature, 1987. 325: 458-462. 262. Benz, I.S., M. A., Infection and Immunity, 1989. 57(5): 1506-1511. 263. Jose, J., R.M. Maas, and M.G. Teese, Journal of Biotechnology, 2012. 161(2): 92-103. 264. Natale, P., T. Bruser, and A. Driessen, Biochimica et Biophysica Acta (BBA) - Biomembranes, 2008. 1778(9): 1735-1756. 265. Jong, W.S., A. Sauri, and J. Luirink, Current Opinion in Biotechnology, 2010. 21(5): 646-52. 266. Sauri, A., et al., Microbiology, 2009. 155(Pt 12): 3982-91. 267. Ieva, R. and H.D. Bernstein, P Natl Acad Sci USA, 2009. 106(45): 19120-5. 268. Rutherford, N. and M. Mourez, Microb Cell Fact, 2006. 5: 22.

85 Bacterial display systems for engineering of affinity proteins

269. Veiga, E., V. de Lorenzo, and L.A. Fernandez, Mol Microbiol, 1999. 33(6): 1232-43. 270. Valls, M., et al., Nature Biotechnology, 2000. 18: 661-665. 271. Maurer, J., J. Jose, and T.F. Meyer, J Bacteriol, 1997. 179(3): 794-804. 272. Klauser, T., J. Pohlner, and T.F. Meyer, The EMBO journal, 1990. 9(6): 1991-1999. 273. Jose, J. and D. Zangen, Biochem Biophys Res Commun, 2005. 333(4): 1218-26. 274. Yang, T.H., et al., Applied and Environmental Microbiology, 2004. 70(12): 6968-76. 275. Lattemann, C.T., et al., J Bacteriol, 2000. 182(13): 3726-3733. 276. Becker, S., et al., FEBS Letters, 2005. 579(5): 1177-82. 277. Robert, V., et al., PloS Biology, 2006. 4(11): 1984-1995. 278. Jose, J., R. Bernhardt, and F. Hannemann, Journal of Biotechnology, 2002. 95(3): 257-268. 279. Jose, J. and S. von Schwichow, Chembiochem, 2004. 5(4): 491-499. 280. Jose, J., D. Betscheider, and D. Zangen, Analytical Biochemistry, 2005. 346(2): 258-267. 281. Jose, J. and T.F. Meyer, Microbiology and Molecular Biology Reviews, 2007. 71(4): 600-619. 282. Klauser, T., J. Pohlner, and T.F. Meyer, The EMBO journal, 1992. 11(6): 2327-2335. 283. Jose, J., et al., Gene, 1996. 178(1-2): 107-110. 284. Skillman, K.M., et al., Mol Microbiol, 2005. 58(4): 945-58. 285. Mattheakis, L.C., R.R. Bhatt, and W.J. Dower, P Natl Acad Sci USA, 1994. 91: 9022-9026. 286. Hanes, J. and A. Pluckthun, P Natl Acad Sci USA, 1997. 94(10): 4937-4942. 287. Plückthun, A., Ribosome Display: A Perspective, in Ribosome Display and Related Technologies, J.A. Douthwaite and R.H. Jackson, Editors. 2012, Humana Press. p. 3-30. 288. Roberts, R.W. and J.W. Szostak, P Natl Acad Sci USA, 1997. 94(23): 12297-12302. 289. Wang, H. and R. Liu, Expert Rev Proteomics, 2011. 8(3): 335-346. 290. Chodorge, M., et al., Protein Eng Des Sel, 2008. 21(5): 343-51. 291. Odegrip, R., et al., P Natl Acad Sci USA, 2004. 101(9): 2806-10. 292. Reiersen, H., et al., Nucleic Acids Res, 2005. 33(1): e10. 293. Koch, H., et al., Journal of Molecular Biology, 2006. 357(2): 427-41. 294. Secco, P., et al., Protein Eng Des Sel, 2009. 22(3): 149-58. 295. Visintin, M., et al., P Natl Acad Sci USA, 1999. 96(21): 11723-11728. 296. Pellis, M., et al., Arch Biochem Biophys, 2012: -. 297. Rajappa, M., P. Saxena, and J. Kaur, 2010. 50: 103-121. 298. Claesson-Welsh, L. and M. Welsh, J Intern Med, 2013. 273(2): 114-27. 299. Folkman, J., The New England Journal of Medicine, 1971. 285(21): 1182-1186. 300. Roskoski, R., Critical reviews in oncology/hematology, 2007. 62(3): 179-213. 301. Ferrara, N., Endocrine reviews, 2004. 25(4): 581-611. 302. Hoeben, A., et al., Pharmacological reviews, 2004. 56(4): 549-580. 303. Hanahan, D. and J. Folkman, Cell, 1996. 86(3): 353-64. 304. Auguste, P., et al., Critical reviews in oncology/hematology, 2005. 54(1): 53-61. 305. Carmeliet, P. and R.K. Jain, Nature, 2011. 473(7347): 298-307. 306. Tugues, S., et al., Mol Aspects Med, 2011. 32(2): 88-111. 307. Wang, R., et al., Nature, 2010. 468(7325): 829-33. 308. Patan, S., L.L. Munn, and R.K. Jain, Microvascular Research, 1996. 51: 260-272. 309. Ribatti, D. and V. Djonov, Cancer Lett, 2012. 316(2): 126-31. 310. Holash, J., et al., P Natl Acad Sci USA, 2002. 99(17): 11393-8. 311. Maniotis, A.J., et al., The American Journal of Pathology, 1999. 155(3): 739-752. 312. Folberg, R. and A.J. Maniotis, APMIS, 2004. 112: 508-525. 313. Potente, M., H. Gerhardt, and P. Carmeliet, Cell, 2011. 146(6): 873-87. 314. Jain, R.K., Nat Med, 2003. 9(6): 685-693. 315. Carmeliet, P. and R.K. Jain, Nature reviews. Drug discovery, 2011. 10(6): 417-27. 316. Saeed, M.U., E. Gkaragkani, and K. Ali, Clin Ophthalmol, 2013. 6: 533-43. 317. Penn, J.S., et al., Prog Retin Eye Res, 2008. 27(4): 331-71. 318. Miller, J.W., et al., Ophthalmology, 2013. 120(1): 106-14. 319. Aiello, L.P., et al., The New England Journal of Medicine, 1994. 331(22): 1480-1487. 320. Hata, Y., et al., Virchows Arch, 1995. 426: 479-486. 321. Olsson, A.K., et al., Nat Rev Mol Cell Biol, 2006. 7(5): 359-71. 322. Lyttle, D.J., et al., J Virol, 1994. 68(1): 84-92. 323. Yamazaki, Y., et al., J Biol Chem, 2003. 278(52): 51985-8. 324. Ogawa, S., et al., Journal of Biological Chemistry, 1998. 273(47): 31273-31282. 325. Pötgens, A.J.G., et al., J Biol Chem, 1994. 269(52): 32879-32885. 326. Muller, Y.A., et al., Structure, 1997. 5(10): 1325-38. 327. Muller, Y.A., et al., P Natl Acad Sci USA, 1997. 94(14): 7192-7.

86 Filippa Fleetwood

328. Wiesmann, C., et al., Cell, 1997. 91: 695-704. 329. Carmeliet, P., et al., Nature, 1996. 380(6573): 435-439. 330. Mazure, N.M., et al., Cancer Res, 1996. 56(15): 3436-40. 331. Robinson, C.J. and S.A. Stringer, Journal of Cell Science, 2001. 114: 853-865. 332. Holmes, K., et al., Cell Signal, 2007. 19(10): 2003-12. 333. Zachary, I. and G. Gliki, Cardiovascular Research, 2001. 49: 568-581. 334. Fuh, G., et al., Journal of Biological Chemistry, 1998. 273(18): 11197-11204. 335. Ruch, C., et al., Nat Struct Mol Biol, 2007. 14(3): 249-50. 336. Albuquerque, R.J., et al., Nat Med, 2009. 15(9): 1023-30. 337. Koch, S. and L. Claesson-Welsh, Cold Spring Harb Perspect Med, 2012. 2(7): a006502. 338. Holmqvist, K., et al., J Biol Chem, 2004. 279(21): 22267-75. 339. Takahashi, T., et al., Embo J, 2001. 20(11): 2768-2778. 340. Matsumoto, T., et al., The EMBO journal, 2005. 24: 2342-2353. 341. Waldner, M.J. and M.F. Neurath, Expert Opin. Ther. Targets, 2012. 16(1): 5-13. 342. Kendall, R.L. and K.A. Thomas, P Natl Acad Sci USA, 1993. 90(22): 10705-9. 343. Waltenberger, J., et al., J Biol Chem, 1994. 269(43): 26988-95. 344. Ferrara, N., Oncologist, 2004. 9 Suppl 1: 2-10. 345. Kerbel, R. and J. Folkman, Nat Rev Cancer, 2002. 2(10): 727-39. 346. Wilhelm, S.M., et al., Cancer Res, 2004. 64(19): 7099-109. 347. Mendel, D.B., et al., Clin Cancer Res, 2003. 9(1): 327-37. 348. Kumar, R., et al., Mol Cancer Ther, 2007. 6(7): 2012-21. 349. Polverino, A., et al., Cancer Res, 2006. 66(17): 8715-21. 350. Matsui, J., et al., Clin Cancer Res, 2008. 14(17): 5459-65. 351. Presta, L.G., et al., Cancer Res, 1997. 57(20): 4593-9. 352. Ferrara, N., et al., Retina, 2006. 26(8): 859-870. 353. Doggrell, S.A., Expert Opin. Pharmacother., 2005. 6(8): 1421-1423. 354. Jain, R.K., et al., Nat Clin Pract Oncol, 2006. 3(1): 24-40. 355. Sullivan, L.A. and R.A. Brekken, mAbs, 2010. 2(2): 165-175. 356. Saharinen, P., et al., Trends in Immunology, 2004. 25(7): 387-95. 357. Lu, D., et al., Int J Cancer, 2002. 97(3): 393-9. 358. Lu, D., et al., J Biol Chem, 2003. 278(44): 43496-507. 359. Behdani, M., et al., New biotechnology, 2013. 30(2): 205-9. 360. Wu, Y., et al., Clin Cancer Res, 2006. 12(21): 6573-84. 361. Dineen, S.P., et al., BMC Cancer, 2008. 8: 352. 362. Mamluk, R., et al., mAbs, 2010. 2(2): 199-208. 363. Davis, B.M., et al., Small, 2014. 10(8): 1575-84. 364. Thiel, M.A., et al., Ophthalmology, 2013. 120(7): 1403-8. 365. Hsu, J.Y. and A. Wakelee, Biodrugs, 2009. 23(5): 289-304. 366. Giuliano, S. and G. Pages, Biochimie, 2013. 95(6): 1110-9. 367. Shojaei, F., Cancer Lett, 2012. 320(2): 130-7. 368. Bellou, S., et al., Cancer Lett, 2013. 338(2): 219-28. 369. Bergers, G. and D. Hanahan, Nat Rev Cancer, 2008. 8(8): 592-603. 370. Neri, D. and R. Bicknell, Nat Rev Cancer, 2005. 5(6): 436-46. 371. Jain, R.K., et al., Nature reviews. Clinical oncology, 2009. 6(6): 327-38. 372. Cai, W. and X. Chen, J Nucl Med, 2008. 49 Suppl 2: 113S-28S. 373. Nagengast, W.B., et al., Cancer Res, 2011. 71(1): 143-53. 374. Collingridge, D.R., et al., Cancer Res, 2002. 62(5912-5919). 375. Nagengast, W.B., et al., J Nucl Med, 2007. 48(8): 1313-9. 376. Cai, W., et al., The Journal of Nuclear Medicine, 2006. 47(12): 2048-2056. 377. Backer, M.V., et al., Nat Med, 2007. 13(4): 504-9. 378. Samen, E., et al., European journal of nuclear medicine and molecular imaging, 2009. 36(8): 1283-95. 379. James, M.L. and S.S. Gambhir, Physiol Rev, 2012. 92(2): 897-965. 380. Goel, S., et al., Physiol Rev, 2011. 91(3): 1071-121. 381. Goel, S., A.H. Wong, and R.K. Jain, Cold Spring Harb Perspect Med, 2012. 2(3): a006486. 382. Sessa, C., et al., Nat Clin Pract Oncol, 2008. 5(7): 378-91. 383. Choe, Y.S. and K.H. Lee, Curr Pharm Des, 2007. 13(1): 17-31. 384. Gronwall, C., et al., Journal of Biotechnology, 2007. 128(1): 162-83. 385. Sato, A.K., et al., Current Opinion in Biotechnology, 2006. 17(6): 638-42. 386. Backer, M.V. and J.M. Backer, Protein Expression and Purification, 2001. 23(1): 1-7.

87 Bacterial display systems for engineering of affinity proteins

387. Siow, R.M., Culture of Human Endothelial Cells from Umbilical Veins, in Human Cell Culture Protocols, R.R. Mitry and R.D. Hughes, Editors. 2012, Humana Press. p. 265-274. 388. Arnaoutova, I., et al., Angiogenesis, 2009. 12(3): 267-74. 389. Kirchhofer, A., et al., Nat Struct Mol Biol, 2010. 17(1): 133-8. 390. Feldwisch, J., et al., Journal of Molecular Biology, 2010. 398(2): 232-47. 391. Nygren, P.-Å., et al., J Mol Recognit, 1988. 1(69-74). 392. Casali, N., et al., Infection and Immunity, 2002. 70(12): 6846-6852. 393. Guzman, L.M., et al., J Bacteriol, 1995. 177(14): 4121-4130. 394. Giacalone, M.J., et al., Biotechniques, 2006. 40(3): 355-364. 395. Schlegel, S., et al., Journal of Molecular Biology, 2012. 423(4): 648-59. 396. Malinda, K.M., In Vivo Matrigel Migration and Angiogenesis Assay, in Angiogenesis Protocols, S. Martin and C. Murray, Editors. 2009, Humana Press. 397. Rakestraw, J.A., et al., Protein Eng Des Sel, 2011. 24(6): 525-30.

88 Filippa Fleetwood

APPENDIX

Paper I-IV

Contribution to the papers

Paper I Performed characterization of first-generation VEGFR2 binders, library design, selection, and characterization of second-generation binders. Designed and characterized the biparatopic binders together with coauthor. Wrote the manuscript together with coauthors.

Paper II Performed all experimental work and wrote the manuscript together with coauthors.

Paper III Performed all experimental work except for the construction of the immune nanobody library, the phage selection and the GFP enhancement assay for the nanobodies selected by phage display. Wrote the manuscript together with coauthors.

Paper IV Performed all experimental work together with coauthor. Wrote the manuscript together with coauthors.

89