ACTAUNIVERSITATIS

UPSALIENSISUPPSALA

2017

Digital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Medicine 1354

Development of ADAPT-basedtracers for radionuclide molecularimaging of cancer

JAVAD GAROUSI

ISSN 1651-6206ISBN 978-91-513-0030-6urn:nbn:se:uu:diva-327419

Dissertation presented at Uppsala University to be publicly examined in Fåhraeus Hall,Rudbecklaboratoriet, Dag Hammarskjölds väg 20, Uppsala, Friday, 29 September 2017 at13:00 for the degree of Doctor of Philosophy (Faculty of Medicine). The examination willbe conducted in English. Faculty examiner: Professor Tony Lahoutte (In vivo Cellular andMolecular Imaging laboratory, Vrije Universiteit Brussel (VUB), Brussels, Belgium ).

AbstractGarousi, J. 2017. Development of ADAPT-based tracers for radionuclide molecular imagingof cancer. Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty ofMedicine 1354. 99 pp. Uppsala: Acta Universitatis Upsaliensis. ISBN 978-91-513-0030-6.

ABD-Derived Affinity Proteins (ADAPTs) is a novel class of small engineered scaffold proteinsbased on albumin-binding domain (ABD) of streptococcal protein G. High affinity ADAPT binders against various therapeutic targets can be selected. In this thesis, we report adevelopment of ADAPT-based radionuclide imaging agents providing high sensitivity andspecificity of molecular imaging of HER2 expression in disseminated cancers.

We investigated the feasibility of the use of ADAPTs as imaging agents and influence ofmolecular design and radiolabeling chemistry on in vivo targeting and biodistribution propertiesof the tracers.

In Paper I we demonstrated the feasibility of the use of anti-HER2 ADAPT6 molecule as ahigh contrast imaging agent;

In Paper II we evaluated the influence of composition of histidine-containing tag on invivo biodistribution of ADAPT-based tracers labeled with 99mTc using 99mTc(CO)3 binding tohistidine-containing tags and 111In using DOTA chelator at N-terminus;

In Paper III we evaluated the influence of different aspects of N-terminus leading sequenceon targeting including effect of sequence size on clearance rate and effect of the compositionof the sequence on biodistribution profile;

In Paper IV, we evaluated the influence of residualizing properties and positioning of thelabel on biodistribution and targeting; and

In Paper V, we compared tumor-targeting properties of the ADAPT6 labeled at C-terminuswith 99mTc using N3S chelator and 111In using DOTA chelator.

In conclusion, ADAPTs constitute a very promising class of targeting probes for molecularimaging providing high contrast. Molecular design of the ADAPT proteins and chelators/linkersfor labeling has an appreciable effect on their imaging properties.

Keywords: ADAPT, scaffold protein, albumin-binding domain (ABD), Affinity protein,HER2, radionuclide molecular imaging

Javad Garousi, Department of Immunology, Genetics and Pathology, Rudbecklaboratoriet,Uppsala University, SE-751 85 Uppsala, Sweden.

© Javad Garousi 2017

ISSN 1651-6206ISBN 978-91-513-0030-6urn:nbn:se:uu:diva-327419 (http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-327419)

To my beloved family, for all their love, kindness and support.

List of Papers

This thesis is based on the following papers, which are referred to in the text by their Roman numerals.

I Garousi, J.,* Lindbo, S.,* Nilvebrant, J., Åstrand, M., Buijs, J., Sandström, M., Honarvar, H., Orlova, A., Tolmachev, V.,Hober, S. (2015). ADAPT, a Novel Scaffold Protein-Based Probe for Radionuclide Imaging of Molecular Targets That Are Expressed in Disseminated Cancers. Cancer res. 75(20):4364-4371.

II Lindbo, S.,* Garousi, J.,* Åstrand, M., Honarvar, H., Orlova, A., Hober, S.,Tolmachev, V. (2015). Influence of Histidine-Containing Tags on the Biodistribution of ADAPT Scaffold Proteins. Bioconjugate Chem. 27(3):716-26.

III Garousi, J.,* Lindbo, S.,* Honarvar, H., Orlova, A., Velletta, J., Mitran, B., Altai, M., Tolmachev, V., Hober, S. (2016). In-fluence of the N-terminal composition on targeting properties of radiometal-labeled anti-HER2 scaffold protein ADAPT6. Bio-conjugate Chem. 27(11):2678-2688.

IV Lindbo, S.,* Garousi, J.,* Mitran, B., Buijs, J., Altai, M., Orlo-va, A., Hober, S., Tolmachev, V. Radionuclide tumor targeting using ADAPT scaffold proteins: aspects of label positioning and residualizing properties of the label. Accepted to Journal of Nuclear Medicine.

V Garousi, J.,* Lindbo, S.,* Mitran, B., Buijs, J., Vorobyeva, A., Orlova, A., Tolmachev, V., Hober, S. Comparative evaluation of tumor targeting using anti-HER2 ADAPT scaffold protein labeled at C-terminus with indium-111 and technetium-99m. Manuscript.

* Equal contribution.

Reprints were made with permission from the respective publishers.

Related papers not included in this thesis:

I. Sarabadani, P., Zeynali, B., Garousi, J., Nourkojouri, H., Amini, M. P., Satari, A. (2007). Chemical recovery and purifi-cation of 203Tl stable isotope enriched by using an electromag-netic isotope separator. J Label Compd Radiopharma. 50(5-6):638-639.

II. Jalilian, A. R., Garousi, J., Gholami, E., Akhlaghi, M., Bolourinovin, F., Rajabifar, S. (2007). Preparation and biodis-tribution of [67Ga]-insulin for SPECT purposes. Nukleonika 52(4):145-151.

III. Jalilian, A. R., Hakimi, A., Garousi, J., Bolourinovin, F., Ka-mali-Dehghan, M., Aslani, G. (2008). Development of [201Tl](III) oxinate complex for in vitro cell labeling. Iran J Radt Res. 6(3):145-150.

IV. Jalilian, A. R., Sardari, D., Kia, L., Rowshanfarzad, P., Garou-si, J., Akhlaghi, M., Shanehsazzadeh, S., Mirzaii, M. (2008). Preparation, quality control and biodistribution studies of two [ 111In]-rituximab immunoconjugates. Sci Pharm. 76(2):151-170.

V. Jalilian, A. R., Shanehsazzadeh, S., Akhlaghi, M., Garousi, J., Rajabifar, S., Tavakoli, M. B. (2008). Preparation and biodistri-bution of [67Ga]-DTPA-gonadorelin in normal rats. Journal of Radioanal Nucl Chem., 278(1), 123-129.

VI. Jalilian, A. R., Sadeghi, H., Zandi, H., Rowshanfarzad, P., Shafaii, K., Kamali-Dehghan, M Garousi, J., Majdabadi, A., Tavakoli, M. B. (2008). Synthesis and preclinical studies of [61Cu]-N-(2- hydroxyacetophenone)glycinate as a possible PET radiopharmaceutical. Sci Pharm. 76(4):637-651.

VII. Jalilian, A. R., Mirazizi, F., Nazem, H., Garousi, J., Bolouri-novin, F., Sadeghpour, H. (2009). Preparation and quality con-trol of radiolabeled streptokinase for thrombosis imaging. Iran J Radt Res. 6(4):195-200.

VIII. Jalilian, A. R., Tajik, M., Zandi, H., Garousi, J., Bolourinovin, F. (2009). Preparation and biodistribution of [67Ga]-labeled-oxytocin for SPECT purposes. Iran J Radt Res. 7(2):105-111.

IX. Jalilian, A. R., Garousi, J., Akhlaghi, M., & Rowshanfarzad, P. (2009). Development of 111In labeled insulin for receptor im-aging/therapy. J Radioanal Nucl Chem. 279(3):791-796.

X. Akhlaghi, M., Ahi, L. P., Jalilian, A. R., Garousi, J., Pour-Heravi, M. R. A. (2009). Radiosynthesis and biodistribution of [18F]-tetracosactide using a semi-automated [18F]SFB produc-tion module. Nucl Sci Tec. 20(3):163-169.

XI. Jalilian, A., Khoshdel, M., Garousi, J., Yousefnia, H., Hos-seini, M., Rajabifar, S., Sardari, D. (2009). Development of a radiolabeled β-human chorionic gonadotropin. Acta Pharm. 59(4):421-429.

XII. Jalilian, A. R., Zolghadri, S., Faghihi, R., Yousefnia, H., Ga-rousi, J., Shafaii, K., Bolourinovin, F. (2009). Preparation, quality control and biodistribution studies of [61Cu]-oxinate for PET tumor imaging. Nukleonika 54(3), 175-179.

XIII. Bahrami-Samani, A., Ghannadi-Maragheh, M., Jalilian, A. R., Yousefnia, H., Garousi, J., Moradkhani, S. (2009). Develop-ment of 153Sm-DTPA-rituximab for radioimmunotherapy. Nukleonika 54(4):271-277.

XIV. Jalilian, A. R., Yousefnia, H., Garousi, J., Novinrouz, A., Ra-jamand, A. A., Shafaee, K. (2009). The development of radio-gallium-acetylacetonate bis(thiosemicarbazone) complex for tumour imaging. Nucl Med Rev. 12(2):65-71.

XV. Jalilian, A. R., Jouiaei, M., Doroudi, A. R., Bolourinovin, F., Garousi, J. (2010). Development of a radiolabeled glucagon compound for imaging. Nukleonika 55(2):219-224.

XVI. Jalilian, A. R., Jouiaei, M., Doroudi, A., Garousi, J., Mo-radkhani, S. (2010). Preparation and biological evaluation of radiogallium labeled glucagon for SPECT imaging. Journal of Radioanal Nucl Chem. 285(3):555-561.

XVII. Oommen, O. P., Garousi, J., Sloff, M., Varghese, O. P. (2014). Tailored doxorubicin-hyaluronan conjugate as a potent anti-cancer glyco-drug: An alternative to prodrug approach. Macro-mol Biosci. 14(3):327-33.

XVIII. Honarvar, H., Garousi, J., Gunneriusson, E., Höidén-Guthenberg, I., Altai, M., Widström, C., Tolmachev, V., Frejd, F. Y. (2015). Imaging of CAIX-expressing xenografts in vivo using 99mTc-HEHEHE-ZCAIX:1 affibody molecule. Int J On-col. 46(2):513-20.

XIX. Garousi, J., Andersson, K. G., Mitran, B., Pichl, M., Stahl, S., Orlova, A., Buijs, J., Ståhl, S., Löfblom, J., Thisgaard., H., Tolmachev, V. (2016). PET imaging of epidermal growth factor receptor expression in tumours using 89Zr-labelled ZE-GFR:2377 affibody molecules. Int J Oncol. 48(4):1325-32.

XX. Garousi, J., Honarvar, H., Andersson, K. G., Mitran, B., Or-lova, A., Buijs, J., Löfblom, J., Frejd, F.Y., Tolmachev, V. (2016). Comparative evaluation of affibody molecules for radi-onuclide imaging of in vivo expression of carbonic anhydrase IX. Mol Pharm. 13(11):3676-87.

XXI. Andersson, K. G., Oroujeni, M., Garousi, J., Mitran, B., Ståhl, S., Orlova, A., Löfblom, J., Tolmachev, V. (2016). Feasibility of imaging of epidermal growth factor receptor expression with ZEGFR:2377 affibody molecule labeled with 99mTc using a peptide-based cysteine-containing chelator. Int J Oncol. 49(6):2285-93.

XXII. Garousi, J., Andersson, K. G., Dam, J. H., Olsen, B. B., Mi-tran, B., Orlova, A., Buijs, J., Ståhl, S., Löfblom, J., Thisgaard., H., Tolmachev, V. (2017). The use of radiocobalt as a label im-proves imaging of EGFR using DOTA-conjugated affibody molecule. Sci Rep. 7(1):5961.

XXIII. Tolmachev, V., Grönroos, T.J., Yim, C.B., Garousi, J., Yue, Y., Rajander, J., Perols, A., Haaparanta-Solin M., Vehmanen, M., Solin, O., Orlova, A., Anderson, C., Erikson Karlström, A. Optimal molecular design of radiocopper-labelled affibody molecules Submitted to Eur J Nucl Med Molecular Imaging

XXIV. Mitran, B.,* Andersson KG*, Rosestedt M, Garousi J, Tolma-chev V, Ståhl S, Orlova A, Löfblom J. Imaging of EGFR in prostate cancer murine model DU-145 using radiocobalt labeled anti-EGFR affibody molecule DOTA-ZEGFR:2377. Manu-script

XXV. Andersson, K.,* Garousi, J.,* Altai, M., Orlova, A., Ståhl, S., Tolmachev, V. Effect of radionuclide-chelator combination on targeting properties of anti-EGFR Affibody molecules. Manu-script

XXVI. Lindbo, S.,* Garousi, J.,* Mitran, B., Buijs, J., Vorobyeva, A., Oroujeni, M., Orlova, A., Hober, S., Tolmachev, V. On the op-timal molecular design of anti-HER2 ADAPT-based imaging probe labeled with 111In and 68Ga. Manuscript

XXVII. Summer, D.,* Garousi, J.,* Oroujeni, M.,* Mitran, B., Anders-son, K.G., Vorobyeva, A., Löfblom, J., Orlova, A., Tolmachev, V., Decristoforo, C. Cyclic versus noncyclic chelating scaffold for 89Zr-labeled ZEGFR:2377 affibody bioconjugates targeting epidermal growth factor receptor overexpression. Manuscript

* Equal contribution.

Content

Introduction ................................................................................................... 13 Tumor-targeted therapy ............................................................................ 13 Diagnosis of cancer .................................................................................. 14 Molecular imaging ................................................................................... 16 

Optical imaging ................................................................................... 16 Radionuclide molecular imaging using PET and SPECT .................... 17 

Imaging probes for visualization of molecular targets ............................. 20 Labeling of polypeptide-based imaging probes ....................................... 21 Radionuclides for labeling of polypeptide-based probes ......................... 22 Labeling techniques .................................................................................. 24 

Radioiodination ................................................................................... 24 Labeling with indium and gallium. ...................................................... 25 Labeling with technetium-99m ............................................................ 27 

Polypeptide-based imaging probes ........................................................... 28 Intact IgG monoclonal antibodies ........................................................ 28 Proteolytic antibody fragments ............................................................ 29 Engineered antibody fragments ........................................................... 29 Natural ligands and their analogs ........................................................ 30 

New developments in the field of imaging probes: application of non-immunoglobulin engineered scaffold proteins ......................................... 31 Lessons from development of imaging probes based on Affibody molecules .................................................................................................. 32 

Residualizing vs non-residualizing labels ............................................ 32 Different combinations of chelators and nuclides ............................... 33 Histidine-containing tags ..................................................................... 35 Effects of affinity, expression and dosage ........................................... 36 

ABD and ADAPT .................................................................................... 37 HER2 (neu/ErbB2) as a target for therapy ............................................... 39 

Aims of this thesis ......................................................................................... 41 

Paper I ........................................................................................................... 43 

Paper II .......................................................................................................... 51 

Paper III ........................................................................................................ 57 

Paper IV ........................................................................................................ 63 

Paper V ......................................................................................................... 70 

Concluding remarks ...................................................................................... 77 

Ongoing and future studies ........................................................................... 79 

Acknowledgement ........................................................................................ 81 

References ..................................................................................................... 84 

Abbreviations

ABD ADAPT CT DOTA EGFR EPR FDG HER1 HER2 HPEM HSA IC50 IGF-1R IgG KD keV LS174T MRI NMRI NODAGA NOTA PET PDGFR p.i. Ramos SKOV-3 SPECT SPG T½ TNF-α VEGFR ZHER2 GSSC %ID/g

Albumin binding domain ABD Derived Affinity Proteins. Computed tomography 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid Epidermal growth factor receptor Enhanced permeability and retention 18F-2-Fluoro-2-deoxy-D-glucose Human epidermal growth factor receptor 1 Human epidermal growth factor receptor 2 ((4-hydroxyphenyl)ethyl) maleimide Human serum albumin Half maximal inhibitory concentration Insulin-like growth factor-1 receptor Immunoglobulin G Dissociation constant Kilo electron-Volt Dukes' type B colorectal adenocarcinoma cell line Magnetic resonance imaging Naval Medical Research Institute 1,4,7-triazacyclononane-1-glutaric acid-4,7-diacetic acid 1,4,7-triazacyclononane-N,N',N''-triacetic acid Positron emission tomography Platelet-derived growth factor receptor Post injection Human Caucasian Burkitt′s lymphoma cell line Human ovarian carcinoma cell line Single photon emission computed tomography Streptococcal protein G Half-life Tumor necrosis factor-alpha Vascular endothelial growth factor receptor Anti-HER2 Affibody molecule Glycyl-seryl-seryl-cysteinyl Percentage of injected dose per gram

13

Introduction

Tumor-targeted therapy Over the past two decades, the development of therapeutics for the treatment of disseminated cancer has focused on targeting, i.e., molecular recognition of gene products that are aberrantly expressed (i.e., mutated or overex-pressed) in malignant tumors. The most suitable targets for cancer therapy are the receptors or cell adhesion molecules that are overexpressed in tumor cells but that are not or are only weakly expressed in normal tissues. The EGFR/HER family, VEGF receptors, CEA, G protein-coupled receptors (GPCRs), CD20 and CD44v6 are examples of targets that have been used in the development of anticancer therapeutics (Lappano and Maggiolini, 2017; Stigbrand et al., 2008). Monoclonal antibodies, which are specific to such molecular targets, can destroy cancer cells by complement-dependent cyto-toxicity (CDC), antibody-dependent cellular cytotoxicity (ADCC) or the delivery of cytotoxic radionuclides or drugs (Sharkey and Goldenberg, 2006).

An important mechanism of action in targeted therapy is disruption of the signaling of growth factor receptors. Cell regulation in multicellular organ-isms is achievable in different manners, such as signaling via secreted cyto-kines or growth factors, cell-cell or cell-extracellular matrix interactions and the stimulation of immune cells by antigens. Growth factors play essential roles in the homeostasis of normal tissue and in embryonic development. Polypeptide growth factors interact with cell surface-exposed receptors to regulate cellular processes, i.e., metabolism, apoptosis/survival, differentia-tion and proliferation. One group of growth factor receptors is the receptor tyrosine kinases (RTKs). Malignant transformation is often accompanied by mutation or overexpression of growth factor receptors. Excessive signaling of aberrantly expressed growth factor receptors result in the suppression of apoptosis, high proliferation rates and increased motility, conferring malig-nant properties to the transformed cells. Consequently, preventing growth factor signaling by the blocking of ligands or ligand-binding sites on recep-tors using monoclonal antibodies or intracellular active sites with tyrosine kinase inhibitors (TKIs) can inhibit the proliferation and migration of malig-nant cells or the recruitment of neovasculature (Tannock et al., 2013).

14

The RTK superfamily is subdivided into 20 main groups, and all of them have a similar structure. The extracellular region of receptors binds to lig-ands and activates downstream signaling. The intracellular region is in-volved in protein activation, using tyrosine kinase residues and communica-tion with the nucleus. Phosphorylation of protein receptor tyrosine kinases (RTKs) or cytoplasmic proteins leads to activation of a series of signaling cascades in cells. The lipophilic transmembrane part is a connector between the extracellular and intracellular regions. The EGFR/HER family (class I), insulin receptor family (class II) and VEGFR family (class V) RTK classes are often used as molecular targets for therapy (Lemmon and Schlessinger, 2010).

Diagnosis of cancer

Selection of a cancer treatment strategy depends on the stage of disease, particularly in the presence of metastasis. Investigation of the metastasis of primary tumors to other tissues/organs is possible by sampling of lymph nodes and imaging techniques, such as computed tomography (CT), magnet-ic resonance imaging (MRI) or metabolic imaging using 18F-2-fluoro-2-deoxu-D-glucose (18F-FDG). To facilitate the exchange of information be-tween specialists, an international TNM staging system is used. Thisv sys-tem includes the size and extension of the primary tumor (T), the number of involved nearby lymph nodes (N) and metastases of the tumor to distant tissues (M) (Telloni, 2017). Medical imaging is also helpful in the monitor-ing of disseminated cancer response to systemic therapy. The current ap-proach to investigating cancer progression and therapy response is based on the standard guidelines of the World Health Organization (WHO) (Organiza-tion and others, 1979) criteria and Response Evaluation Criteria In Solid Tumors (RECIST) (Eisenhauer et al., 2009). According to these criteria, size measurements of tumors using CT and MRI are the standard clinical meth-ods. It should be noted that the use of the TNM system and RECIST criteria is insufficient for the optimal use of targeted therapy. For example, these mod-els do not provide information about molecular target expression in particu-lar tumors. Due to the genetic heterogeneity of cancer (Bedard et al., 2013), tumors of the same origin might overexpress different molecular targets and require different therapeutics. Furthermore, many targeted therapeutics are cytostatic than rather cytotoxic, and even efficient treatment does not cause appreciable tumor shrinkage (Contractor and Aboagye, 2009; van der Meel et al., 2010). The limitations of anatomic (CT, MRI) and metabolic (18F-

15

FDG) imaging necessitate using biopsy (tissue sampling) methods for the molecular characterization of cancer. The analysis of biopsy samples can contribute to an accurate cancer diagno-sis in following manners. 1. Morphologic characterization using histochemistry staining techniques helps with the grading of tumor malignancy. Evaluation of the invasiveness of tumors, infiltration into neighboring tissues, tumor growth and necrosis is possible using this technique. 2. Immunohistochemistry (IHC) is the current routine method for the detec-tion of proteins, including cancer-associated antigens, in clinical pathology. IHC is a semi-quantitative method based on antigen-antibody interaction. IHC can be used for the staging/grading of tumors for and identifying the origin of metastasis in connection with primary tumors (Wolff et al., 2013). The newly developed quantitative immunohistochemistry (qIHC) method might help to determine the scoring of therapeutic targets (e.g., HER2 score) for therapy decisions (Jensen et al., 2017). Scoring is important for the pre-diction of response to anti-cancer drugs. 3. Molecular pathology provides information about molecular changes in different types of cancer using gene analysis techniques (Cree et al., 2014). Fluorescent in situ hybridization (FISH), silver in situ hybridization (SISH) and chromogenic in situ hybridization (CISH) are examples of molecular pathology techniques. It should be noted that, while biopsy analysis provides important infor-mation, biopsy is invasive, often painful and sometimes accompanied by the risk of tissue damaging. Due to limitations in the number of biopsies and the phenotypic heterogeneity of tumors, biopsy samples might not even be rep-resentative of the whole tumor (Bedard et al., 2013). Moreover, biopsy sam-ples will only represent a fraction of the overall disease when there are mul-tiple metastatic lesions. Non-invasive detection and measurement of chemi-cal or biochemical alterations in malignant tumors would facilitate both the stratification of patients for targeted therapy and the monitoring of response to such therapy. Molecular imaging is a methodology permitting such detec-tion and measurements.

16

Molecular imaging

Molecular imaging is based on the detection of radiation emitted by a mo-lecular probe in a living body by a detector or detector array placed outside the body. Optical imaging is based on non-ionizing radiation, while radionu-clide molecular imaging utilizes ionizing radiation of nuclides conjugated to a molecular probe.

Optical imaging

Optical imaging can be based on two phenomena: bioluminescence and fluo-rescence. Bioluminescence is the emission of light following an enzymatic reaction between luciferin and enzyme luciferase in the presence of oxygen and ATP. Emitted light from excited oxy-luciferin during the relaxation period is in the yellow-green/orange part of the spectrum and can be used for imaging. Luciferin + O2 + ATP Oxy-luciferin + CO2 + AMP + PP + visible light

Today, this reaction is widely used in preclinical investigations as a part of bioluminescence molecular imaging techniques. This approach facilitates the evaluation of novel cancer therapies by correlating tumor size with emitted light. This technique is applicable in research in two ways: 1. Transfected cells that express luciferase in vitro are injected into animals’ bodies; and 2. The cancer cell line owning luciferase reporter gene can be used to pro-duce bioluminescent tumors in transgenic animal models. Bioluminescence molecular imaging has some advantages and disadvantages compared with other imaging techniques, such as single photon emission computed tomography (SPECT) and positron emission tomography (PET). It is a non-invasive technique, which allows researchers to use the same animal during the course of the study. The cost of equipment is less than for SPECT and PET. Furthermore, it does not use radioactive compounds. Conversely, it is not as quantitative or accurate as nuclear techniques. Due to the very strong dependence of light signals on depth of tissue, this imaging technique is not applicable in humans (Zinn et al., 2008).

luciferase

Mg2+/Ca2+

17

Fluorescence imaging utilizes light emitted from fluorophores in the visible or near-IR range. Proteins conjugated with dyes, e.g., Alexa 750 or indocya-nine green (ICG), molecular beacons (MBs) and quantum dots (QDs) are some examples of near IR probes that have been used for the evaluation of tumor progression and the detection of molecular marker expression in tu-mors (van der Meel et al., 2010). Recently, a 3D map of mouse brain vascu-lature was created using indium-arsenide-based quantum dots, which clearly showed the detail of the vasculature (Bruns et al., 2017). Reporter genes, such as the fluorescent proteins β-galactosidase and lactamase, also facilitate the study of gene expression and regulation (Rao et al., 2007). Limitations in the application of these probes, such as autofluorescence, absorption and scattering by blood and surrounding tissue, have favored the use of short-wavelength IR (1000-2000 nm). Overall, due to limits of depth penetration in optical imaging compared to nuclear techniques, the former are less appli-cable in humans. One area of clinical application for fluorescent imaging is the use of intraoperative probes for exact tumor delineation and to search for metastases during surgery (Zhang et al., 2017).

Radionuclide molecular imaging using PET and SPECT

Positron emission tomography (PET) and single-photon emission tomogra-phy (SPECT) are two key in vivo imaging modalities, which are widely used in clinical routine, as well as in preclinical and clinical research. These mo-dalities are based on tracing of a radionuclide attached to or incorporated into a bioactive molecule (e.g., sugar, amino acid, lipid, peptide receptor ligand) that accumulates in tumors due to cancer-associated biochemical alterations. Depending on body-penetrating radiation, which is emitted by a radionuclide, different physical principles are used for the acquisition of images. PET imaging is based on coincidental detection of paired photons (511 keV) formed during the annihilation of a positron, emitted by a radionuclide, and an electron. The paired annihilation photons move in nearly opposite direc-tions. Due to their high energy, they can penetrate the human body and be registered by detector rings surrounding the patient (Figure 1). Information concerning multiple coincidences is processed by computers to reconstruct the distribution of positron-emitting radionuclides within a living body at each time point. The physical features of PET offer some advantages over SPECT, such as the former’s greater sensitivity, higher spatial and temporal resolution and greater accuracy of quantification of radioactivity concentra-tions in vivo.

18

Figure 1. A schematic overview of positron-electron annihilation and PET camera acquisition of images.

In clinical oncology, several different categories of PET tracers have been developed to characterize:

1. Tumor metabolism, for example, 18F-fluoro-2-deoxy-D-glucose (18F-FDG) and L-[methyl-11C] methionine;

2. Tumor receptor expression, for example, 18F-fluoroestradiol and 68Ga-somatostatin analogs; and

3. Tumor hypoxia, for example, 64Cu-diacetyl-bis(N4-methylthiosemicarbazone) (64Cu-ATSM) and 18F-fluoromisonidazole (18FMISO).

Initially, PET was developed for the labeling of small biologically active molecules. The biogenic, short-lived positron-emitting radionuclides 11C (T1/2 = 20.4 min), 15O (T1/2 = 2.04 min), 13N (T1/2 = 9.97 min), and 18F (T1/2 = 109.8 min) were used as labels. The short half-lives of these radionuclides complicate or completely exclude their transportation to distant imaging facilities. Therefore, so-called PET centers were established for research or for the clinical application of PET. A typical PET center includes a cyclotron for radionuclide production, a radiochemical laboratory for production and quality control of radiopharmaceuticals, and one or several PET cameras. These requirements make the use of PET appreciably more expensive than SPECT. A more cost-effective alternative to cyclotron production of short-lived posi-tron-emitting radionuclides is the use of radionuclide generators. In a typical generator, a long-lived mother nuclide is immobilized on an ion-exchange column. Decay of the mother nuclide generates a short-lived daughter radio-nuclide, which could be periodically eluted from the column (Knapp and Mirzadeh, 1994). This process creates an easily available source of a short-lived radionuclide, which is independent from a cyclotron. In particular, the

19

68Ge (T1/2 = 270.8 d)/68Ga (T1/2 = 67.6 min) generator is gaining broad ac-ceptance in nuclear pharmacy (Maecke and André, 2007). The advantages of PET in terms of sensitivity and resolution make its appli-cation attractive for imaging not only using small molecules but also using larger molecules, such as antibodies and their fragments. However, the blood clearance of these probes is slow, and the half-lives of biogenic radionu-clides are too short for such applications. Therefore, technologies for the production of long-lived positron emitters, such as 89Zr (T1/2 = 78.4 h), 124I (T1/2 =4.17 d), 64Cu (T1/2 = 12.7 h), 86Y (T1/2 = 14.7 h), and 55Co (T1/2 = 17.5 h), have been developed, and the chemistry of their conjugation to biomole-cules has been established (Tolmachev and Stone-Elander, 2010). SPECT imaging relies on the detection of gamma quanta irradiated from radionuclides decaying by electron capture or isomeric transition, such as 111In (T1/2 = 2.8 d) or 99mTc (T1/2 = 6.0 h). Rotation of the collimated detector of a gamma camera around the subject provides for acquisition of multiple 2D projections of radioactivity distribution in the human body. Using a tomographic reconstruction program, 2D images are converted into a 3D image (Figure 2). Currently, SPECT imaging is inferior to PET in sensitivity, resolution and accuracy of quantitative analysis. However, SPECT facilities are less expen-sive than those for PET, and the former are more available for clinical imag-ing. Moreover, recent advances in the development of cadmium zinc tellu-ride (CZT) detectors have resulted in an appreciable increase in the resolu-tion and sensitivity of SPECT imaging. Currently, such cameras are applied mainly for cardiac imaging, but we might expect rapid translation of this technology to other areas of radionuclide imaging. Some important properties of the different imaging instruments available in clinics are compared in Table 1.

Figure 2. A schematic overview of SPECT camera acquisition of images.

20

Table 1. Properties of different imaging instrumentations in clinical use.

Modality Spatial resolution (mm Sensitivity Quantification capability Availability Cost

PET 4-8 pmol Straightforward + High

SPECT 4-15* nmol Medium ++ Medium

CT 0.03-1 - N/A ++ Medium MRI 0.02 mmol Low ++ Medium * Spatial resolution of 3.9 mm was obtained for novel CZT cameras for cardiac imaging. The combination of molecular imaging (PET and SPECT) with anatomical computed tomography (CT) imaging modalities is currently widely used in clinics. Combined imaging of structures and biochemistry using PET/CT and SPECT/CT has provided superior diagnostic accuracy in oncology. CT in-formation-based attenuation correction can expedite imaging processes and reduce imaging costs. In addition, diagnostics based on two different modali-ties can facilitate accurate diagnosis by removing imaging and breathing artifacts. PET/CT has demonstrated better diagnostics in head and neck tu-mors, bronchial carcinoma and melanomas. Use of SPECT/CT in some cas-es, including bone imaging, could be a gold standard for malignancy staging. Hybrid imaging using tumor-specific tracers, e.g., octreotide labeled with 68Ga or 111In, could be advantageous for visualizing neoplastic foci (Bock-isch et al., 2009). Anatomical and functional imaging using PET/CT and SPECT/CT multimodalities is sequential. Judenhofer and co-workers devel-oped PET/MRI to synergize anatomical and functional imaging in preclinical research (Judenhofer et al., 2008). In the future, it might be possible to per-form simultaneous PET and MRI imaging on humans.

Imaging probes for visualization of molecular targets

Both radiolabeled small molecules and polypeptide-based imaging probes can be used for radionuclide molecular imaging. Small molecules are indis-pensable for the imaging of targets located in the cytoplasm since polypep-tides cannot usually penetrate the cellular membranes of living cancer cells. In contrast, many molecular targets are expressed on the cell surface. Rapid progress in biotechnology has enabled efficient selection of specific high-affinity polypeptide binders to such targets. In addition, natural ligands to receptors that are overexpressed in malignant tumors or their analogs might be used as imaging probes. This thesis is focused on the development of im-aging probes based on polypeptide-based binders.

21

Several factors should be considered during the selection of an imaging probe (James and Gambhir, 2012). The probe design should provide the highest possible sensitivity and specificity for imaging. Both characteristics require the difference between the sites of specific accumulation of a probe and the sites of its unspecific accumulation to be as large as possible. High accumulation of a probe in normal tissues decreases imaging contrast and therefore reduces imaging sensitivity. Factors favoring high sensitivity and specificity of an imaging probe include:

- High affinity to a molecular target; - High stability of a label because the release of a label in circulation

might cause its accumulation in normal tissues; - Absence of unspecific cross-reactivity with molecules expressed in

normal tissues; - Stability of a probe to peptidases and proteases encountered in circu-

lation; - Appropriate pharmacokinetics -- rapid binding to specific targets,

rapid excretion of unbound tracer from the blood and non-target or-gans and low re-absorption by excretion organs are important factors improving the sensitivity of imaging;

- Safety profile -- low toxicity of the agent is essential to avoid any harm to normal tissues;

- Potential for clinical translation -- large-scale production must be possible and cost effective for clinical use; and

- Selection of the most appropriated labeling strategy.

Labeling of polypeptide-based imaging probes

Labeling chemistry might influence the cellular-processing and tumor-targeting properties of polypeptide-base imaging probes. Consequently, the sensitivity of imaging might be improved. Previous experiences with Af-fibody molecules, bombesin analogs and somatostatin analogs have exerted a profound influence of labeling chemistry on the specificity and sensitivity of imaging (Tolmachev and Orlova, 2010). For Affibody molecules, the strong influence of different labeling approaches on binding affinity, blood clearance rate, uptake, tumor-to-organ ratio and retention in the kidney has been observed. The influence of different chelators and different radionu-clides with the same chelator on the biodistribution of labeled Affibody mol-ecules was demonstrated (Tolmachev et al., 2010c). Interaction of radionu-clides with different chelators can form complexes with different geometry,

22

net charge and lipophilicity, which in turn can alter the interaction of labeled polypeptides with a target, as well as plasma proteins. Interaction of an im-aging probe with plasma proteins is an important factor determining the rate of clearance from the blood. Moreover, a chemical form of radiocatabolites from intracellularly degraded tracers has a strong effect on their redistribu-tion to different organs. Site-specific labeling of proteins has several advantages over random label-ing. Site-specifically labeled tracers are chemically uniform and highly re-producible in different batches. In the case of random labeling, control over the reactions is weak. Heterogeneous tracers with radionuclides in different positions can show different biodistribution profiles. Optimization of label-ing chemistry for improvement of the biodistribution of an imaging probe is complicated in cases of random labeling. Site-specific labeling might be facilitated by the use of engineered proteins, based on a cysteine-free scaf-fold. For example, a unique cysteine has been introduced into the C-terminal of recombinantly produced Affibody molecules to direct thiol-specific label-ing (Ahlgren et al., 2008; Tolmachev et al., 2010b). This outcome enabled site-specific labeling of recombinantly produced Affibody molecules using different radiometals and radiohalogens. Some examples of such radionu-clides used for labeling of Affibody molecules include 111In (Ahlgren et al., 2008; Tolmachev et al., 2008, 2010b, 2011), 64Cu (Cheng et al., 2010; Miao et al., 2010), 68Ga (Kramer-Marek et al., 2011), 57Co (Wållberg et al., 2010), 18F (Cheng et al., 2008; Kramer-Marek et al., 2008), 76Br (Mume et al., 2005), and 131I(Tran et al., 2007b).

Radionuclides for labeling of polypeptide-based probes

A radionuclide must meet several requirements to be suitable for use as a label in molecular imaging.

- For use in PET, a radionuclide should have a high yield of positrons, no beta particles and preferably no co-emitted gamma rays.

- For use in SPECT, a radionuclide should emit gamma rays with en-ergy in the range of 100-300 keV and no beta particles.

- The half-life of a radionuclide should match the expected timing of imaging. Overly short-lived radionuclides might require injection of high levels of radioactivity. An overly long-lived one would remain in the body for a long time after imaging is completed. Both cases are associated with elevated absorbed dose burdens in patients.

- To meet the economic requirements for routine clinical diagnostics,

a radionuclide should be produced using a generator or low-to-

23

medium energy cyclotron, limiting the possible nuclear reactions for radionuclide production to (p, n), (p, ), (d, n) and (p, 2n).

- Finally, the radionuclide should possess chemical properties ena-

bling efficient radiolabeling. Table 2. Available radionuclides for radionuclide molecular imaging (National Nuclear Data Centre, 2015).

Nuclide Half-life Production reac-tion

Decay (%)

γ energy (keV) Application

13N 10 min 13N (p, n) 13C β+ (100) 511 PET 11C 20 min 14N (p, α) 11C β+ (99.8)

EC (0.2) 511 PET

68Ga 67.6 min

Generator β+ (89) EC (17)

511, 1077 PET

18F 109.8 min

18O (p, n) 18F β+ (97) EC (3)

511 PET

44Sc 3.97 h Generator β+ (94) EC (6)

511, 1157 PET

64Cu 12.7 h 64Ni (p, n) 64Cu β+ (18) β- (37) EC (24)

511, 1346 PET

86Y 14.7 h 86Sr (p, n)86Y β+ (33) EC (67)

511, 443, 628, 646, 703, 778, 1077

PET

76Br 16.2 h 76Se (p, n)76Br β+ (54) EC (46)

511, 559, 657, 1216, 1854, 2391, 2792, 2950

PET

55Co 17.5 h 54Fe (d, n) 55Co β+ (76) EC (24)

511, 477, 931, 1317, 1408 PET

89Zr 78.4 h 89Y (p, n) 89Zr β+ (23) EC (77)

511, 909 PET

124I 4.17 d 124Te (p, n) 124I 124Te (d, 2n) 124I

β+ (23) EC (77)

511, 603, 723, 1691 PET

99mTc 6 h Generator IT (100) 140.5 SPECT 123I 13.2 h 124Te (p, 2n) 123I EC (100) 159 SPECT 111In 2.8 d 111Cd (p, n) 111In

112Cd (p, 2n) 111In EC (100) 171.3, 245 SPECT

The properties of radionuclides used or considered for use in molecular im-aging are presented in Table 2.

24

Labeling techniques Radioiodination Until recently, radioimmunotherapy was mainly focused on the use of 131I (Tolmachev, 2008). The chemistry of iodination has been well investigated, and some excellent reviews are available (Adam and Wilbur, 2005; Wilbur, 1992).

Iodination is possible by two methods: direct and indirect (Figure 3). With the direct method, radioiodide is oxidized to the electrophilic form (I+) by an oxidant, such as chloramine-T, and it attacks an activated amino acid residue (mainly tyrosine) in a protein. This method is fast and provides a high yield and specific radioactivity. A disadvantage of direct iodination is high accu-mulation of catabolized radioiodine in the thyroid and stomach. This issue could be resolved to some extent by blocking the radioactivity uptake by these organs with non-radioactive iodide. Behr and co-workers optimized a protocol for direct iodination (Behr et al., 2002). The position of tyrosine residues in protein is important. If a tyrosine locates to a binding site, direct iodination might alter the binding to the target. For example, the binding of an anti-HER2 Affibody molecule to HER2-expressing cells was appreciably reduced after direct labeling with 125I (Steffen et al., 2005). With the indirect radioiodination method, a radioiodine is bound to a protein via a linker (Figure 3). The linker is bifunctional. In the first step, a radioio-dine is coupled to a linker by an electrophilic reaction. In the second step, the radioiodinated linker binds to an amino group of an N-terminal residue or lysine or to the thiol group of a cysteine residue in the structure of the protein. This method is especially valuable when tyrosine is in the binding site of the protein, and direct labeling at this site might alter the binding properties of the protein. Moreover, by choosing a suitable linker, the biodis-tribution properties of the protein can be improved. The drawbacks of the indirect method are its lower yield and low specific radioactivity, compared with direct radioiodination. Detailed information regarding the linker and indirect iodination using N-succinimidyl 3-[*I]iodobenzoate were provided by Vaidyanathan and Zalutsky (Vaidyanathan and Zalutsky, 2006).

25

Figure 3. Direct iodination (A). Indirect radioiodination using N-succinimidyl tri-methylstannyl-benzoate (B). The linker molecule is radioiodinated first under acidic conditions and is then coupled to a free amine (N-terminal of ω-amino group of lysine) under alkaline conditions. Both meta- and para-iododerivatives of benzoate have been described in the literature

Labeling with indium and gallium.

Direct stable labeling of proteins with many radiometal nuclides, labeling with 111In and 68Ga is not possible. An appropriate chelator is needed as a linker between a nuclide and a protein or peptide. The choice of chelator depends on the labeling condition, type of nuclide, properties of the imaging agent and association/dissociation rate of the complexing reaction. Some-times in vivo, proteins with chelating properties, such as transferrin, and metal-binding enzymes can compete with radioligands to release the nuclide. The released nuclide can be then taken up by non-target organs, reducing imaging contrast. Therefore, choosing a suitable macrocyclic/acyclic chela-tor might alter the biodistribution profile dramatically.

The choice of an appropriate chelator for the labeling of proteins and pep-tides with different radionuclides for in vivo study is very important. Ther-modynamic stability and kinetic inertness are two important factors in the

26

formation of a radionuclide-chelator complex. Polyaminopolycarboxylates are common chelators use to label proteins and peptides with 111In, 68Ga, 90Y and lanthanides (e.g., 153Sm, 177Lu), (Liu and Edwards, 2001). Macrocyclic and acyclic chelators are two different classes of polyaminopolycarboxylate chelators. Derivatives of DOTA (1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid), NOTA (1,4,7-triazacyclononane-N,N′,N′′triacetic acid) and NODAGA (1-(1,3- carboxypropyl)-4,7 carboxymethyl-1,4,7 triazacy-clononane) are macrocyclic chelators, which are often used for labeling of peptides (Figure 4). These chelators have high kinetic inertness, indicating that the association and dissociation rates are low. Therefore, a high tem-perature for labeling is necessary for the formation of highly stable tracers. These chelators are suitable for short peptides and robust scaffold proteins. DOTA was widely used for labeling Affibody molecules with different radi-onuclides, e.g., 111In (Ahlgren and Tolmachev, 2010), 68Ga (Kramer-Marek et al., 2011), 57Co (Garousi et al., 2017; Wållberg et al., 2010) and 64Cu (Cheng et al., 2010). However, DOTA is not the best chelator for all radio-nuclides. NOTA has a smaller cavity than DOTA and provides greater sta-bility of labeling of peptides with 68Ga and 64Cu (Clarke and Martell, 1991a, 1991b; Garrison et al., 2007). NODAGA and NOTA have similar structures. However, the extra hydroxyl group of NODAGA enables its conjugation to peptide while carrying a three-valent radionuclide with a neutral charge of the complex.

Figure 4. Metal complexes of the chelators 1,4,7,10-tetraazacylododecane-1,4,7,10-tetraacetic acid (DOTA), 1,4,7-triazacyclononane-1,4,7-triacetic acid (NOTA), and 1-(1,3-carboxypropyl)-1,4,7-triazacyclononane-4.7-diacetic acid (NODAGA) conju-gated to the N-terminal amino group via amide bonds.

Derivatives of DTPA (diethylenetriaminepentaacetic acid) are the most common acyclic chelators used for the radiolabeling of proteins and pep-

27

tides. Complexes of transitional metals with acyclic chelators are less kinet-ically inert than complexes with macrocyclic chelators. Therefore, some release of a radionuclide is possible in vivo. However, labeling using these chelators is very fast even at room temperature. These properties match mainly with antibodies or other molecules, which are sensitive to heating. There have been some publications concerning labeling using acyclic poly-aminopolycarboxylate chelators with detailed protocols (Cooper et al., 2006; Sosabowski and Mather, 2006).

Labeling with technetium-99m

99mTc is the most common and available radionuclide for nuclear medicine application. There are several different strategies to conjugate 99mTc to pep-tides or proteins. Previous studies have shown that thiol-containing moieties, together with adjacent amide nitrogen atoms, provide a suitable combination for the che-lating of 99mTc (Schwarz and others, n.d.). Below, we discuss two different thiol-containing moieties. Mercaptoacetyl-containing peptide-based chelators. Extension of the N-terminus of a peptide with at least two amino acids with further coupling of a mercaptoacetyl forms an SN3 chelator suitable for the labeling of proteins with 99mTc. (Figure 5). Importantly, the use of amino acids with different side chains can modify the local charge and lipophilicity of a tracer, which might, in turn, influence the biodistribution of the labeled peptide.

Figure 5. Cysteine-containing peptide-based chelators on the C-terminus (A). Mer-captoacetyl- containing peptide-based chelators on the N-terminus (B). x1, x2 and x3 represents different side groups of different amino acids. The clue color indicates the 13 residues responsible for molecular recognition in Affibody molecules.

Cysteine-containing peptide-based chelators. Introduction of cysteine at the C-terminal of recombinantly produced proteins can form an N3S chelator for

28

labeling with 99mTc. Such chelators resemble mercaptoacetyl-containing chelators at the N-terminus of peptides (Figure 5). Application of this ap-proach to the Affibody molecule ZHER2:2395-Cys provided a tracer with very good biodistribution, except for high kidney uptake (Ahlgren et al., 2009a).

Histidine tags as chelators for 99mTc. Engineering of a polyhistidine tag (His-tag) in a recombinantly produced protein enables facile purification using immobilized metal ion affinity chromatography (IMAC). This tag can also be used for labeling of the protein with 99mTc. Waibel and co-workers (Wai-bel et al., 1999) developed an excellent method to label peptides and proteins with a 99mTc(1)-carbonyl compound via His-tag. However, several studies demonstrated that the use of a hexahistidine tag in Affibody molecules re-sulted in elevated hepatic uptake (Ahlgren et al., 2008, 2009a; Orlova et al., 2006b). Replacing every second histidine with a more hydrophilic glutamic acid moiety in His-tag ((HE)3-tag) reduced the hepatic uptake dramatically (Hofstrom et al., 2011a; Tolmachev et al., 2010a).

Polypeptide-based imaging probes

Currently, several types of molecular imaging agents with different charac-teristics and pharmacokinetics are available. Some of their important fea-tures, which are essential in the nuclear medicine field, are briefly described below.

Intact IgG monoclonal antibodies

For years, monoclonal antibodies (mAbs) that bind to RTK have been used for targeted therapy in different types of cancers. The same mAbs labeled with radionuclides have been used successfully to visualize therapeutic tar-gets (Dijkers et al., 2010; Even et al., 2016; Behr et al., 2001) Currently, more than 8 mAbs labeled with radionuclides have been approved by the Food and Drug Administration (FDA) for imaging (Olafsen and Wu, 2010). Due to the long residence time of mAbs in the blood, relatively long-lived gamma emitting radionuclides, such as 111In, have mainly been used for la-beling on intact IgG. In the past few years, there has been an understanding that better sensitivity and resolution of PET compared to SPECT might im-prove imaging using mAbs. Therefore, the application of mAbs labeled with long-lived positron emitters, such as 89Zr, 64Cu, 124I and 86Y, for imaging is attracting increasing attention (Dongen et al., 2007).

29

Long residence time in blood is the main drawback of radiolabeled mAbs, which cause a low signal-to-background ratio in imaging and reduces sensi-tivity. Clinical studies indicated that the best imaging contrast could be ob-tained only 4-7 days after injection (Dijkers et al., 2010; Even et al., 2016). The large size of the mAbs is another drawback, which influences negatively their penetration in tumors (Mankoff et al., 2016). Moreover, mAbs are bulky proteins with a molecular weight of 150 kDa, which exceeds a thresh-old value of ca. 45 kDa for enhanced retention and permeability effect (EPR-effect) (Wester and Kessler, 2005) (Figure 4). This is associated with an unspecific accumulation in tumors of antibodies and risk of false-positive diagnoses.

Proteolytic antibody fragments To overcome the aforementioned limitation of mAbs, antibody fragments produced by the digestion of antibodies with enzymes the pepsin or papain, i.e., Fab (ca. 55 kDa) and (Fab)2 (ca. 110 kDa), are used for imaging (Figure 6). Clearance of unbound fragments from the blood is faster than clearance of full-length mAbs. Greater contrast of images compared to the contrast of images acquired using intact mAbs was obtained, and imaging on the day after injection was possible (Freise and Wu, 2015), enabling the use of more short-lived radionuclides, such as 99mTc, or medium the half-life positron emitters 64Cu, 76Br and 86Y, for imaging, which might reduce the dose bur-den to patients. The evaluation of Fab and (Fab)2 fragments demonstrated that the size reduction was an efficient way to improve radionuclide imaging contrast. However, this approach is associated with some disadvantages. For example, preparation of fragments is often associated with a decrease in affinity to a molecular target. Both Fab and (Fab)2 are too large to have effi-cient extravasation. In addition, the issue of nonspecific EPR-effect-mediated uptake in tumors has not been solved for such fragments.

Engineered antibody fragments

Protein engineering was used to improve the pharmacokinetics of antibody fragments while preserving their affinity and specificity. Radiolabeled single chain variable fragments (scFv) with molecular weight of ~25 kDa and one binding site (monovalent) were evaluated as imaging agents. It was found, however, that the relationship between the clearance and extravasation rates of radiolabeled scFv was suboptimal for efficient tumor accumulation (Thurber et al., 2007). The dimeric form of scFv (diabody, ~50 kDa) and scFv fused with proteins/Fc region (minibody, ~80 kDa and scFV-Fc, ~105 kDa) (Figure 6) showed higher accumulation in the target (Olafsen and Wu,

30

2010). A very promising method for developing imaging agents is to use single domain fragments (VHH, nanobodies) derived from camelid antibod-ies that contain only one heavy chain (Vaneycken et al., 2011), allowing for the obtaining of an imaging probe with a molecular weight of only 15 kDa. A VHH labeled with 68Ga provided very good imaging quality in humans (Keyaerts et al., 2016). 68Ga, 18F and 44Sc (T1/2 = 3.9 h) are some radionu-clides suitable for labeling of the engineered fragments.

Figure 6. Schematic overview and properties of some targeting agents.

Natural ligands and their analogs

Regulatory peptide ligands binding to receptors, which are overexpressed in key cancers, have attracted increasing attention over the past three decades. Regulatory peptides are small, highly specific, and non-immunogenic. They extravasate rapidly and clear rapidly from the blood. Their production and modification are simple. Therefore, their importance in diagnostics and ther-apy is comparable to that of immuno-targeting agents. Short peptide ligands to G-protein coupled receptors (GPCRs) and their analogs have been active-ly investigated over the past 30 years (Reubi, 2003; Reubi et al., 2005). These peptides can be conjugated to suitable chelators for labeling with radi-onuclides for PET (e.g., 68Ga, 18F and 44Sc) and SPECT (e.g., 111In and 99mTc) imaging and for radionuclide therapy (177Lu and 90Y). However, there are limitations to the use of them, such as:

- Often strong physiologic actions and the need either to use very highly specific activities or to develop antagonists; and

- GPCR receptors not being overexpressed in all cancers.

31

Somatostatin analogs are currently widely used in clinics for diagnosis and therapy of neuroendocrine tumors (Bodei et al., 2015; Johnbeck et al., 2014). Some other radiolabeled peptide ligands to GRPRs have been investigated (Reubi and Maecke, 2008): • cholecystokinin (CCK) and gastrin for targeting of medullary thyroid

cancers (Behr and Béhé, 2002); • bombesin for targeting of prostate and breast cancers (Zhang et al., 2004,

2006); • neuropeptide-Y (NPY) for targeting of breast tumors and sarcomas

(Körner and Reubi, 2007); and • gastrin-releasing peptide-1 (GLP-1) for targeting of insulinomas (Wild et

al., 2008). Various natural ligand binding RTKs have been evaluated for imaging, e.g., EGFR (Orlova et al., 2000), IGF-1R (Cai et al., 2006) and VEGFR (Shan, 2004; Wang et al., 2008). The use of peptide ligands as imaging probes has some limitations. Often, they have potent physiological action, causing appreciable discomfort or even toxic effects (Frejd and Kim, 2017; Löfblom et al., 2010; Ståhl et al., 2017) after injection. Not all molecular targets are receptors, which shortens a list of cancers to be treated. In some cases, there is no known ligand to a receptor, as in the case of HER2. Moreover, peptides are often rapidly de-graded in vivo by proteolytic enzymes (peptidases). Several techniques, for example, introducing D-amino acids or the use of unnatural amino acids, can help to increase the stability of peptide ligands (James and Gambhir, 2012).

New developments in the field of imaging probes: application of non-immunoglobulin engineered scaffold proteins

Further engineering to produce immunoglobulin-based proteins smaller than VHH (~15 kDa) seems to be impossible. Using molecular engineering and display techniques, it is possible to create libraries of non-immunoglobulin scaffold-based small affinity proteins (4-20 kDa) for the selection of appro-priate binders. These so-called scaffold proteins are usually stable in a wide range of pH values and temperatures and are capable of correctly refolding after thermal or chemical denaturing. Randomization of binding sites and molecular display selection techniques helps us to select high affinity scaf-fold proteins to bind to various specific targets (Nygren and Skerra, 2004).

32

Currently, there are a number of promising scaffold proteins under develop-ment to use as imaging agents, e.g., Affibody molecules, DARPins, cysteine knots, anticalins and adnectins (Vazquez-Lombardi et al., 2015). Affibody molecules are well-studied scaffold proteins. We have experience in our group concerning the development of Affibody molecules for imaging and therapy. Therefore, focusing on the properties of Affibody molecules pro-vide us with lessons for the development of new scaffold proteins.

Lessons from development of imaging probes based on Affibody molecules Affibody molecules are small affinity proteins, 6-7 kDa (58 amino acids), derived from the B domain of staphylococcal protein A. Randomization of 13 surface-exposed amino acids in 2 of the helices creates a library from which to select a high affinity protein binder against the target of interest. Affibody molecules are robust and refold rapidly into the initial 3D structure after exposure to harsh conditions, such as a non-physiologic pH or high temperature (Löfblom et al., 2010) (Frejd and Kim, 2017; Löfblom et al., 2010; Ståhl et al., 2017). Several Affibody molecules with subnanomolar affinity, for example, anti-HER2 (Orlova et al., 2006a), anti-HER3 (Orlova et al., 2014), anti-IGF-1R (Orlova et al., 2013), anti-EGFR (Tolmachev et al., 2009a), anti-CAIX (Honarvar et al., 2015), and anti-PDGFRβ (Tolma-chev et al., 2014), have been evaluated as radionuclide imaging probes.

Below we discuss some important aspects of our findings with Affibody molecules, including residualizing properties, types of radionuclides, types of chelators, the influence of histidine-containing tags, affinities, target ex-pression levels and dosages.

Residualizing vs non-residualizing labels

Binding of an imaging probe to a receptor, an antigen or other cell-surface molecule can trigger internalization. An internalized complex of a target and an imaging probe is trafficked through endosomes into lysosomes, where proteins are degraded by proteolytic enzymes. The chemical properties of radiocatabolites determine their further fate. Hydrophilicity and bulkiness cause the entrapping of radiometal catabolites inside the cells and conse-quently long intracellular retention (DeNardo et al., 2004). Such labels are called residualizing labels. The lipophilic radiocatabolites from radiohalogen labels can diffuse through lysosomal and cellular membranes and leak from

33

cells to be further redistributed through the blood and excreted in the urine (Tolmachev et al., 2003). This type of label is called non-residualizing. When a target-receptor complex is rapidly internalized, as in the case of GPCR targeting peptides, the use of residualizing labels is important to pro-vide good retention of radioactivity in tumors (Reubi, 2003). The kidney is the main excretory organ for protein-based molecules smaller than 60 kDa (Vegt et al., 2010). Several molecular mechanisms are involved in the reabsorption of proteins and peptides from primary urine in proximal tubuli to prevent loss of nutrients. The peptides and proteins are internalized by proximal tubuli cells and degraded into lysosomes and amino acids to return to the bloodstream. The use of residualizing radiometal labels results in long retention of radionuclides in the kidneys (Vegt et al., 2010). The size of scaffold proteins is smaller than the cut-off for glomerular filtration. Pre-vious studies have suggested that a high degree of tubular reabsorption is typical for small scaffold proteins, such as knottins, fibronectin domains and affibody molecules (Ackerman et al., 2014; Hackel et al., 2012; Tolmachev et al., 2010b), resulting in high radioactivity accumulation when these scaf-fold proteins are labeled using residualizing radiometal labels. Previous investigations have shown that radiolabeled affibody molecules are internalized slowly by cancer cells. The internalization of HER2-binding affibody molecules after 24 h was <30% (Wållberg et al., 2010). Similar internalization rates are also typical for EGFR-binding (Tolmachev et al., 2010b), CAIX-binding (Honarvar et al., 2015) and PDGFR-binding (Tol-machev et al., 2014) affibody molecules. Although the high affinity, residu-alizing properties of a label do not play an important role in the tumor reten-tion of affibody molecules, the internalization of affibody molecules in the proximal tubuli cells of kidneys is rapid. Thus, the use of non-residualizing labels creates an opportunity to reduce the renal retention of radioactivity without a reduction of the retention in tumors. This outcome has been con-firmed in a number of studies using radioiodine (Orlova et al., 2006a; Tol-machev et al., 2009b) and radiobromine labels (Mume et al., 2005). Moreo-ver, it was shown that 99mTc and 188Re, in combination with some peptide-based chelators, might also act as non-residualizing labels, resulting in de-creased renal retention of these radionuclides (Altai et al., 2014a, 2014b; Wållberg et al., 2011).

Different combinations of chelators and nuclides As already mentioned, coupling of a radiometal nuclide to a targeting protein requires conjugation of a chelator. Apparently, conjugation of a new moiety modifies the physicochemical properties of the protein surface. Such param-

34

eters as distribution of charge and “lipophilic patches” might be affected, which could, in turn, have effects on off-target interactions, e.g., reversible binding of a radiolabeled moiety to blood proteins or interactions with mo-lecular moieties on the surfaces of cells in the vasculature or healthy tissues. An example of this influence is a switch of predominantly hepatobiliary to predominantly renal excretion of octreotide after conjugation of a hydro-philic DTPA chelator (Bakker et al., 1991). Furthermore, different chelators might provide equally adequate thermodynamic stability and kinetic inert-ness of a complex with a radionuclide. At the same time, these chelators might have different numbers of functional groups and therefore modify the surface differently, which could translate into different influences on biodis-tribution profiles and, therefore, different imaging contrasts. Such phenome-na have been observed for radiolabeled short regulatory peptides (Eisenwie-ner et al., 2002; Ginj et al., 2008; Varasteh et al., 2015). Affibody molecules are larger than small regulatory peptides, and one night expect that this effect would be less pronounced in this case. However, it has been shown that there is a noticeable influence of chelators on the biodistribution of radiometal-labeled affibody molecules. For example, an increase in hydrophilicity of mercaptoacetyl-containing peptide based chelators at the N-terminus reduced hepatobiliary excretion of 99mTc-labeled affibody molecules (Engfeldt et al., 2007a, 2007b; Tran et al., 2007a). It was found that the conjugation of dif-ferent macrocyclic chelators (DOTA, NOTA and NODAGA) at the N-terminus for the labeling of synthetic anti-HER2 affibody molecule with 111In had a strong influence on biodistribution (Malmberg et al., 2012). For example, DOTA- and NODAGA-conjugated variants had nearly two-fold lower hepatic uptake than a NOTA-conjugated variant. In addition, a NODAGA conjugate had more rapid clearance from the blood. Further stud-ies have demonstrated that a further increase in the negative charge of af-fibody molecules using a DOTAGA chelator reduced hepatic uptake of 68Ga- and 111In-labeled affibody molecules with a second-generation scaffold (Westerlund et al., 2016). In the case of coupling of chelators at the C-terminus of recombinantly produced affibody molecules, the use of NODAGA provided lower hepatic uptake than DOTA for 68Ga- and 111In-labeled anti-HER2 affibody molecules (Altai et al., 2013). It should be noted that the same chelator might have a different complex geometry with different radiometals. For example, gallium in a complex of monoamide DOTA has a coordination number of six, and it forms a chelate with cis-pseudo-octahedral geometry, having a folded macrocyclic unit and one free carboxy group (Heppeler et al., 1999). Indium and yttrium are octa-coordinated in the complex with the same chelator, and the complexes have the geometry of a square antiprism (Heppeler et al., 1999). Despite this sub-tle difference, the uptake of 67Ga-labeled peptide DOTATOC in the kidneys and in tumors differs appreciably, compared with uptake of 111In- and 90Y-

35

labeled DOTATOC variants (Heppeler et al., 1999). In another study, the use of 68Ga instead of 111In for labeling DOTA-RGD peptide resulted in appre-ciably higher adhesion to blood proteins and noticeably slower clearance from the blood (Decristoforo et al., 2008). Studies with affibody molecules have also demonstrated that the use of different radionuclides for labeling of the same peptide-chelator conjugate had apparent influences on blood clear-ance and uptake in normal tissues, thus influencing imaging contrast (Altai et al., 2013; Garousi et al., 2017; Strand et al., 2014; Tolmachev et al., 2017, 2010b). Furthermore, the placement of a radionuclide-chelator complex at different sites of an affibody molecule (C-terminus, N-terminus or middle of the helix 3) also had significant influences on the biodistribution and tumor-targeting properties of anti-HER2 affibody molecules labeled with 111In or 68Ga (Honarvar et al., 2014; Perols et al., 2012). Taken together, this information suggests that the selection of an optimal labeling approach (a radionuclide, a chelator and the position of the chelator on a scaffold protein) influences the biodistribution of scaffold protein-based imaging probes. Detailed investigation of this influence could facilitate the selection of a conjugate, providing the best sensitivity and specificity for imaging.

Histidine-containing tags The introduction of histidine-containing tags (His-tag) at the N- or C-terminal enables efficient immobilized metal affinity chromatography (IMAC) purification of recombinantly produced proteins (Block et al., 2009; Bornhorst and Falke, 2000), thus simplifying appreciably the whole produc-tion process. Waibel et al. showed that the His-tags could be utilized as a chelator to bound 99mTc(CO)3 to proteins (Waibel et al., 1999). A previous investigation concerning labeling of anti-HER2 Affibody molecules with 99mTc(CO)3 using hexa-histidine tags at N-termini demonstrated high hepatic uptake (Orlova et al., 2006b). Later, Ahlgren et al. showed high hepatic up-take of anti-HER2 affibody molecules with N-terminal hexahistidine tags when an 111In or 99mTc label was placed at the C-terminus (Ahlgren et al., 2009b, 2008). When the same affibody molecules without hexahistidine tags were investigated, the hepatic uptake was significantly lower (Ahlgren et al., 2009b, 2008). This finding suggested that the N-terminal positioning of a hexahistidine tag was the reason for elevated uptake in the liver. By replac-ing every second histidine in a tag with glutamate, i.e., using a (HE)3-tag, the hepatic uptake of 99mTc(CO)3-labeled anti-HER2 affibody molecules was reduced dramatically by approximately 10 fold (Tolmachev et al., 2010a). At the same time, (HE)3-tags can also be utilized for IMAC purification. Further studies have demonstrated that the use of (HE)3-tags improved the biodistri-

36

bution of affibody molecules labeled at the C-terminus using 111In, 99mTc and 125I (Hofstrom et al., 2011). It was shown that the use of (HE)3-tag has simi-lar influences on other variants of Affibody molecules (HER3, CAIX and IGF-1R binders) (Honarvar et al., 2015; Orlova et al., 2014, 2013). A more detailed study showed that reduction of hepatic uptake of 99mTc(CO)3-labeled affibody molecules was correlated with the hydrophilicity of amino acids substituting histidine (Hofström et al., 2013). However, the use of positively charged lysine, instead of negatively charged glutamate, elevated the hepatic uptake, although lysine and glutamate are equally hydrophilic (Hofström et al., 2013). Interestingly, the effect of a composition of histidine-containing tags on biodistribution was much more pronounced for N-terminal position-ing of the tag, compared to the C-terminal (Hofström et al., 2013). Other groups working with different types of polypeptide-based tracer, including DARPins and two short peptides, have reported reductions in hepatic uptake using (HE)3-tags (Eder et al., 2013; Goldstein et al., 2015). We can conclude that histidine-containing tags might be used not only for IMAC purification and 99mTc(CO)3 labeling of scaffold-based proteins but also for the modification of their biodistribution.

Effects of affinity, expression and dosage Rates of penetration of targeting proteins in tumors and of the clearance of imaging agents from the blood are size dependent, and retention of radioac-tivity in tumors is affinity dependent. Schmidt and Wittrup demonstrated the correlations among size, uptake and affinity based on a mechanistic model. Small molecules (<25 kDa) penetrate more rapidly inti tumors, but for better retention, high affinity in a low nanomolar or sub-nanomolar range is essen-tial (Schmidt and Wittrup, 2009). The results of this modeling were in con-cordance with in vivo investigations (Orlova et al., 2006a). Monoclonal anti-bodies, which are not rapidly exerted, have similar accumulation as small molecules, but lower affinity (~100-fold) would be sufficient for the accu-mulation of antibodies in tumors. It should be noted that the activity in blood and normal tissues would also be high due to the large size. Therefore, good accumulation of monoclonal antibodies in tumors would not translate into high imaging contrast (Schmidt and Wittrup, 2009). Middle size molecule (25-60 kDa), such as scFv and Fab fragments, would have lower tumor up-take than intact IgG. Although such fragments have better penetration into tumors, it would be insufficient to compensate for reduced bioavailability due to the fast renal excretion (Schmidt and Wittrup, 2009). However, the imaging contrast might be better for fragments due to lower background radioactivity.

37

Interestingly, dimerization does not improve the in vivo tumor uptake of affibody molecules despite increased affinity (Ahlgren and Tolmachev, 2010). Apparently, the clearance rates of dimeric and monomeric forms are similar (Orlova et al., 2006a), but the smaller size of monomers should be favorable for rates of extravasation and diffusion in tissues. Ahlgren and Tolmachev (2010) concluded that the best way to improve affinity was affin-ity maturation of Affibody molecules, not the use of di- or multimeric con-structs. The optimal level of affinity depends on the level of expression of the target in tumors (Ahlgren and Tolmachev, 2010). Tolmachev and co-workers demonstrated using in vivo models that affinity of a single digit nanomole level was sufficient for imaging targets with high (> 106 targets per cell) expression (Tolmachev et al., 2012). However, affinity in the sub-nanomolar range would be preferable when the target expression is on the level of 4×104targets per cell.

ABD and ADAPT As discussed previously, the use of engineered small scaffold proteins for molecular imaging has advantages over larger antibodies/antibody frag-ments. Among the most important, favorable features of these molecules are rapid tissue penetration and tumor targeting, combined with fast clearance from non-target organs/tissues. In fact, a size smaller than 10 kDa is desira-ble because this is a cut-off size for the extravasation rate through the endo-thelium. The smaller size of the probe provides more rapid tissue penetration (Thurber et al., 2007; Wittrup et al., 2012). Although Affibody molecules are small (7 kDa), an albumin-binding domain is even smaller (Johansson et al., 2002; Linhult et al., 2002). ABDs were previously used for half-life extension of Affibody molecules to maintain the desired concentration of the protein in the blood for therapeutic applications (Tolmachev et al., 2007). ABDs are derived from streptococcal protein G (SPG), which is a multi-domain surface protein. SPG has two independent binding sites for albumin and immunoglobulins (Akerström et al., 1987; Fahnestock et al., 1986; Guss et al., 1986). Various streptococci can bind to human serum albumin (HSA) via protein G (Björck et al., 1987; Myhre and Kronvall, 1980). SPG contains three homologous albumin-binding domains, and the third albumin-binding domain, G148-ABD3, which has been well studied, is called ABD in this text. ABD is only 5 kDa in size, and it would be attractive to convert ABD into a general class of targeting agents, one of the smallest alternative scaf-folds developed so far. ABD consists of 46 amino acids, and its three-dimensional structure is not dependent on disulfide bridges. It is highly solu-ble in water (Figure 7). Experiments have confirmed that ABD is stable at

38

elevated temperatures, and after denaturation during chemical reactions, it refolds to preserve its initial 3D structure (Nilvebrant and Hober, 2013). Initial studies indicated that a second molecular recognition function could be engineered to ABD. A selection library was created by randomization of 11 amino acids on a surface, which was not involved in albumin binding (Nilvebrant et al., 2014, 2013, 2011). ABD variants binding to TNFα, HER3 and HER2 were selected using this library (Nilvebrant et al., 2014, 2013, 2011). The novel binding variants were termed ABD-derived affinity pro-teins (ADAPTs) (Nilvebrant et al., 2014). To use ADAPTs as imaging agents, fast blood clearance is favorable. Therefore, binding to albumin was completely eradicated from anti-HER2 variants. A HER2 variant of ADAPT after affinity maturation was called ADAPT6 (Figure 8). Its affinity to HER2 was approximately 1 nM with no binding to albumin. It binds to the same epitope as trastuzumab binds.

Figure 7. Schematic domain structure of streptococcal protein G and the albumin binding domain. The albumin-binding regions, BB and ABP, and the third albumin binding domain are displayed. The figure was modified from Linhult et al. (Linhult et al., 2002).

39

Figure 8. Selection of ADAPTs.

HER2 (neu/ErbB2) as a target for therapy Human growth factor 2 (HER2) and neu are homologous oncogenic growth factor receptors for human and rodent species, respectively HER2/ErbB2/neu, along with HER1/ErbB1, HER3/ErbB3 and HER4/ErbB4, belong to the HER family (type I transmembrane growth fac-tors) of the RTK superfamily (Moasser, 2007). No soluble ligand has been reported to bind to HER2. HER2 signaling is induced mainly by heterodi-merization of HER2 with other ligand-bound HER family members (Hynes and Lane, 2005). HER2 is a therapeutic target for several therapies using trastuzumab, pertuzumab and trastuzumab emtansine in breast cancer (Giordano et al., 2014). Twenty percent to 30% of breast cancer is HER2 positive (Yarden, 2001). Only tumors with high overexpression of HER2

40

respond to anti-HER2 therapies, Therefore, detection of HER2-positive tu-mors is crucial to select the optimal HER2 therapy modality (Wolff et al., 2013). HER2 is a very convenient model for evaluating the targeting proper-ties of new tracers. Very low HER2 expression levels in normal adult tissue (Natali et al., 1990) provides the preconditions to study the biodistribution of new targeting agents. The targeting properties of the agent in this condition are more dependent on biological, chemical and physical properties than on interaction with the target in normal tissues.

41

Aims of this thesis

The main goal of this thesis was to develop ADAPT-based SPECT/PET imaging agents providing high sensitivity and specificity for molecular im-aging in disseminated cancers. To achieve this goal, we investigated the feasibility of using ADAPTs as imaging agents and the influence of molecu-lar design and radiolabeling chemistry on the in vivo targeting and biodistri-bution properties of the tracers. The HER2-binding variant of ADAPT was used as a model in this study (Figure 9). In particular, we seek to achieve the following goals. In Paper I, we evaluated the feasibility of using the anti-HER2 ADAPT6 molecule as a high contrast imaging agent. In Paper II, we evaluated the influence of the composition of histidine-containing tags on the in vivo biodistribution of ADAPT-based tracers la-beled with 99mTc via 99mTc(CO)3 binding to histidine-containing tags and 111In via DOTA chelators. Both labels were placed at the N-terminus. In Paper III, we evaluated the influence of different aspects of the N-terminus leading sequence on targeting, including:

The effect of sequence size on the clearance rate from blood and non-target tissues;

The effect of flanking amino acids in the presence of different histi-dine-containing tags; and

The effect of the composition of the sequence on off-target interac-tion of the tracers and their biodistribution profiles.

In Paper IV, we evaluated the influence of residualizing properties and site-specific positioning of the label on biodistribution. In Paper V, we compared the tumor-targeting properties of ADAPT6 (hav-ing an N-terminal DEAVDANS sequence and C-terminal Cys) when labeled with 99mTc through an N3S chelator and 111In through a DOTA chelator.

42

Figure 9. Schematic representation of papers 1, 2, 3, 4 and 5.

43

Paper I

ADAPT, a Novel Scaffold Protein-Based Probe for Ra-dionuclide Imaging of Molecular Targets That Are Ex-pressed in Disseminated Cancers

Background and aim

Imaging of molecular therapeutic targets in disseminated cancers requires high imaging contrast. One way to achieve this goal is the use of engineered affinity scaffold proteins as imaging agents. Affibody molecules (Ahlgren and Tolmachev, 2010), DARPins (Goldstein et al., 2015), knottins (Acker-man et al., 2014), and anticalins (Terwisscha van Scheltinga et al., 2014) are some examples of probes that were promising in preclinical studies. The potential of Affibody molecules as imaging probes has been proved in clini-cal investigations (Sörensen et al., 2016, 2014). The possibility of finding high affinity binders to a molecular target depends on matching of the geom-etry of scaffold proteins with the target. Therefore, different scaffolds might have different preferences of binding to the therapeutic target. The availabil-ity of various types of scaffold proteins provides an opportunity to select a highly specific high-affinity binder for the target of interest. The small size and high affinity of ADAPTs create preconditions for their successful use as imaging probes. However, binding to albumin might slow clearance and decrease imaging contrast. The aim of this study was to investigate the feasibility of molecular imaging using ADAPTs.

Methods

We used the following molecular design. - Binding of anti-HER2 ADAPT to albumin was completely prevented by engineering site-directed mutagenesis. The new variant was designated ADAPT6.

44

- A (HE)3-tag was introduced at the N-terminus to facilitate IMAC purifica-tion. This tag was selected because earlier studies with Affibody molecules showed that this tag provided lower hepatic uptake. - N-terminal cysteine was introduced for site-specific conjugation of a chela-tor. - A versatile chelator, DOTA, was selected for labeling. A maleimide derivative of the macrocyclic DOTA chelator was conjugated site-specifically to the N-terminal cysteine in (HE)3-ADAPT6. The conju-gate was labeled with 111In for SPECT imaging and with 68Ga for PET imag-ing. The in vitro and in vivo properties of 111In-DOTA-C-(HE)3-ADAPT6 and 68Ga-DOTA-C-(HE)3-ADAPT6 were studied.

Results and discussion

The purity of DOTA-C-(HE)3-ADAPT6, measured using reversed phase HPLC, was more than 95%. Measurements of binding kinetics using a BiacoreTM SPR system demonstrated a dissociation constant at equilibrium (KD) of 2.5 nmol/L for DOTA-C-(HE)3-ADAPT6 binding to HER2 and no binding of DOTA-C-(HE)3-ADAPT6 to human serum albumin (HSA) (Fig-ure 10 A and B).

Labeling with 111In was performed in 0.2 M ammonium acetate, pH 5.5, at 95 °C for 35 min. Labeling with 68Ga was performed in 1.25 M sodium ace-tate, pH 3.6, at 95 °C for 30 min. The labeling yield was more than 98% for conjugates labeled with 111In and 68Ga. There was no release of the labels after incubation with 500-fold EDTA for 2 h. LigandTracer measurements and InteractionMap calculations demonstrated that the affinity (KD) of 111In-DOTA-C-(HE)3-ADAPT6 binding to HER2-expressing cells was 1.13 ± 0.05 nmol/L. Significantly lower (P < 0.05) cell binding of 111In-DOTA-C-(HE)3-ADAPT6 to HER2-expressing SKOV-3 cells was observed after receptor saturation by adding a large amount of non-labeled (HE)3-ADAPT6 (Figure 10C). This outcome confirmed the in vitro specificity of ADAPT6 to HER2 targets. Moreover, the addition of a large excess of trastuzumab also reduced significantly the binding of 111In-DOTA-C-(HE)3-ADAPT6 to the cells. However, addition of the anti-HER2 affibody molecule ZHER2:342 did not change the uptake. This finding indicated that ADAPT6 and trastuzumab compete to bind to the same epitope, while ZHER2:342 binds to a different epitope. The experiment was repeated with the LS174T cell line. The results

45

showed the same binding pattern but with lower uptake compared to the uptake by SKOV-3 cells, due to lower HER2 expression in LS174T.

Figure 10. In vitro characterization of DOTA-C-(HE)3-ADAPT6. SPR sensor-grams of DOTA-C-(HE)3-ADAPT injected over immobilized HER2 (A) and human serum albumin (B). C. Specificity of the binding of 111In-DOTA-C-(HE)3-ADAPT6 to HER2-expressing SKOV-3 cells in vitro. For the pre-saturation of HER2, a 100-fold molar excess of non-labeled H6-ADAPT6, anti-HER2 ZHER2:342 affibody or trastuzumab was added. Data are presented as the mean ± SD (n= 3). D. Cell-associated radioactivity as a function of time during continuous incubation of HER2-expressing SKOV-3 cells with the 111In labeled ADAPT6 molecule. Data are presented as the mean ± SD (n= 3).

Data concerning the binding and internalization of 111In-DOTA-C-(HE)3-ADAPT6 by SKOV3 cells are shown in Figure 10D. Five percent to 26% of the total bounded radioactivity was internalized over 1-24 h of incubation, which is a slow internalization rate. The biodistribution results showed rapid clearance of 111In-DOTA-C-(HE)3-ADAPT6 from the blood in normal NMRI mice. At 1 h p.i., the blood radio-activity was less than 1%IA/g (Figure 11), indicating eradicated albumin binding affinity of ADAPT. Moreover, binding to other organs and tissues except for the kidney was lower than 1%IA/g, demonstrating that there was

46

no cross-reactivity with non-target organs/tissues. High kidney uptake is typical for small scaffold proteins.

Figure 11. Biodistribution of 111In-DOTA-C-(HE)3-ADAPT6 in female normal (NMRI) mice. Data, mean %IA/g ± SD (n = 4).

Figs. 12 and 13 demonstrate the biodistribution profile of 111In-DOTA-C-(HE)3-ADAPT6 in BALB/C nu/nu mice bearing SKOV-3 xenografts. In vivo specificity of the tracer to the HER2 target was demonstrated by pre-saturation of HER2 expressing SKOV-3 xenografts and the use of Ramos xenografts as negative controls. Three hundred micrograms of H6-ADAPT6 were used for the pre-saturation experiment (Figure 10). Uptake of 111In-DOTA-C-(HE)3-ADAPT6 in pre-saturated SKOV3 xenografts, 3.3±0.4%IA/g, and in Ramos xenografts, 0.112 ± 0.002%IA/g was signifi-cantly (P < 0.05) lower than in SKOV-3 xenografts, 19 ± 6%IA/g. There was a significant reduction (P < 0.05) in kidney uptake and a significant (P < 0.05) increase in uptake by the spleen in the case of pre-saturation.

47

Figure 12. In vivo targeting specificity of 111In-DOTA-C-(HE)3-ADAPT6 (3 µg) in mice bearing SKOV-3 xenografts at 1 h p.i. The blocked group was co-injected with an excess amount (300 µg) of non-labeled H6-ADAPT6 molecule. Ramos xenografts were used as HER2-negative controls. Data are presented as the mean % IA/g ± SD (n=4).

Biodistribution results in SKOV-3 bearing mice confirmed rapid clearance of 111In-DOTA-C-(HE)3-ADAPT6 from all of the normal organs and tissues except for the kidneys (Figure 11), in good agreement with biodistribution results in NMRI mice. Biodistribution was investigated at 1, 4 and 24 h p.i. The tumor uptake at 1 h p.i. was 19 ± 6% IA/g, and there was no significant difference between these values over time. The tumor to blood ratio at 1 h p.i. was 43.

48

Figure 13. Biodistribution (A) and tumor-to-organ ratios (B) of 111In-DOTA-C-(HE)3-ADAPT6 in mice bearing SKOV-3 xenografts. Data are the mean ± SD (n ¼ 4). Please note the logarithmic scale in A.

Table 3. Biodistribution of 68Ga-DOTA-C-(HE)3-ADAPT6 in BALB/C nu/nu mice bearing SKOV-3 xenografts (high HER2 expression) and LS174T (low HER2 ex-pression) at 1 h p.i. SKOV3 xenografts LS174T xenografts

1 µg 15 µg 1 µg 15 µg

Blood 0.59 ± 0.07 0.50 ± 0.06 0.60 ± 0.09 0.6 ± 0.2

Lung 0.70 ± 0.02 0.53 ± 0.06 0.67 ± 0.09 0.7 ± 0.2

Liver 1.7 ± 0.1 1.21 ± 0.05 1.8 ± 0.1 1.6 ± 0.1

Spleen 0.62 ± 0.05 0.50 ± 0.07 0.74 ± 0.07 0.56 ± 0.34

Stomach 0.63 ± 0.11 0.35 ± 0.07 0.54 ± 0.06 0.36 ± 0.09

Kidney 308 ± 24 277 ± 5 323 ± 27 311 ± 31

Tumor 10.3 ± 1.0a, c 8.4 ± 0.6d 2.9 ± 0.5b 0.9 ± 0.6

Muscle 0.18 ± 0.01 0.12 ± 0.02 0.14 ± 0.02 0.14 ± 0.03

Bone 0.5 ± 0.1 0.34 ± 0.07 0.5 ± 0.1 0.4 ± 0.1

NOTE: Data are presented as the average %IA/g and SD for four mice. Difference was significant (P < 0.05) between: aInjected doses of 1 and 15 mg in SKOV3 xenografts bInjected doses of 1 and 15 mg in LS174T xenografts cSKOV3 and LS174T xenografts at an injected dose of 1 mg dSKOV3 and LS174T xenografts an at injected dose of 15 mg

Table 3 shows the biodistribution results of 68Ga-DOTA-C-(HE)3-ADAPT6 in SKOV3 and LS174T xenograft-bearing BALB/C nu/nu mice at 1 h p.i. Table 4 shows the tumor-to-organ ratios. SKOV-3 had higher HER2 expres-sion than LS174T cells. The results of biodistribution measurements after 2 different injected protein doses (1 and 15 mg) indicated much higher tumor uptake in SKOV-3 xenografts than in LS174T. These results demonstrated that tumor uptake depends on the level of receptor expression. At higher

49

injected protein doses, discrimination between two types of tumors was bet-ter (Table 4).

Table 4. Tumor-to-organ ratios of 68Ga-DOTA-C-(HE)3-ADAPT6 in BALB/C nu/nu mice bearing SKOV-3 xenografts (high HER2 expression) and LS174T (low HER2 expression) at 1 h p.i.

SKOV-3 xenografts LS174T xenografts

1 μg 15 μg 1 μg 15 μg

Blood 17.5 ± 0.3 17 ± 2 4.8 ± 0.2 1.5 ± 0.8

Lung 15 ± 1 16 ± 1 4.3 ± 0.2 1.4 ± 0.7

Liver 6.0 ± 0.1 7.0 ± 0.4 1.6 ± 0.2 0.6 ± 0.4

Spleen 17 ± 2 17 ± 1 3.9 ± 0.4 2.1 ± 1.0

Stomach 16 ± 1 24 ± 4 5 ± 1 2 ± 1

Kidney 0.034 ± 0.005 0.030 ± 0.003 0.009 ± 0.002 0.003 ± 0.002

Muscle 56 ± 9 70 ± 6 21 ± 4 7 ± 4

Bone 20 ± 4 25 ± 4 7 ± 3 2 ± 1

NOTE: Data are presented as an average value and SD for four mice.

Figure 14. Imaging of HER2 expression in xenografts using radiolabeled DOTA-C-(HE)3-ADAPT6. A. Gamma camera imaging of SKOV-3 xenografts using 111In-DOTA-C-(HE)3-ADAPT6 (3 µg). Contours were derived from a digital photograph and were superimposed over images to facilitate interpretation. A control animal (right) was injected with a saturating amount of nonlabeled ADAPT6. B. PET/CT imaging of SKOV-3 (high expression) and LS174T (low expression) xenografts using 68Ga-DOTA-C-(HE)3-ADAPT6. The injected protein dose was 15 µg. Arrows indi-cate tumors (T) and kidneys (K). The biodistribution data and specificity of radiolabeled DOTA-C-(HE)3-ADAPT6 to HER2 expressing xenografts were confirmed by imaging exper-iments (Figure 14). Reduction of tumor uptake was observed after saturation of HER2 with DOTA-C-(HE)3-ADAPT6 (Fig 14A). PET imaging using 15 µg 68Ga-DOTA-C-(HE)3-ADAPT6 could provide clear discrimination be-

50

tween high (SKOV-3) and low HER2 expression (LS174T) in xenografts (Figure 14B).

The main findings of this study were the following. - Radiolabeled ADAPT6 binds with high specificity and affinity to

HER2-expressing cells. - The binding epitope of ADAPT6 overlaps with the epitope of

trastuzumab, but it is different from the epitope of anti-HER2 Af-fibody molecules. This finding suggests that different scaffolds have different preferential sites to bind to the same receptors. Thus, when the selection of a high affinity binder is not possible using one scaf-fold, it might be possible using another.

- After eradication of binding to albumin, ADAPT clears rapidly from the blood, enabling its use as an imaging agent.

- At concentrations relevant for imaging, the uptake of radiolabeled ADAPT is low in all tissues except the kidney, creating a precondi-tion for high imaging contrast.

- Both 111In-DOTA-C-(HE)3-ADAPT6 and 68Ga-DOTA-C-(HE)3-ADAPT6 bind to HER2-expressing xenografts with high specificity.

- Both 111In-DOTA-C-(HE)3-ADAPT6 and 68Ga-DOTA-C-(HE)3-ADAPT6 provide high-contrast imaging of xenografts with high HER2 expression.

- PET imaging using 68Ga-DOTA-C-(HE)3-ADAPT6 enables discrimi-nation between xenografts with high and low HER2 expression.

In conclusion, an ABD with eradicated binding to albumin is a very promis-ing scaffold for the development of targeting probes for radionuclide molec-ular imaging.

51

Paper II

Influence of Histidine-Containing Tags on the Biodistri-bution of ADAPT Scaffold Proteins

Background and aim

Addition of a histidine-containing tag to a terminus of recombinantly pro-duced scaffold proteins has two potential advantageous:

It makes possible facile IMAC purification; and It allows for the labeling of scaffold protein with 99mTc(CO)3.

Earlier studies with Affibody molecules in our group demonstrated reduction of hepatic uptake when every second histidine in the hexahistidine H6-tag was replaced with a glutamate bearing a negative charge (Tolmachev et al., 2010a). This effect has been confirmed by other groups as well, e.g., with DARPins and some short peptides (Goldstein et al., 2015). The aim of this study was to investigate the effect of the composition of his-tidine-containing tags on the biodistribution of ADAPT-based imaging probes labeled with 99mTc via 99mTc(CO)3 binding to histidine-containing tags and with111In via DOTA chelator. Both labels were placed at the N-terminus.

Methods

We used the following molecular design of imaging probes (Figure 15): Two variants of HER2-binding ADAPT6 having an H6-tag or (HE)3-

tag and a DEAVDANS leading sequence at the N-terminal were created;

N-terminal cysteine for site-specific conjugation of the chelator for 111In-labeling was introduced;

Maleimide-DOTA chelator for 111In-labeling was coupled; and There was no N-terminal cysteine in the variants for labeling using

99mTc(CO)3.

52

A maleimide derivative of macrocyclic DOTA chelator was site-specifically conjugated to the N-terminal cysteine in H6-ADAPT6 and (HE)3-ADAPT6. The DOTA-C-H6-ADAPT6 and DOTA-C-(HE)3-ADAPT6 conjugate were labeled with 111In. An IsoLink kit was used for the labeling of (H)6-ADAPT6 and (HE)3-ADAPT6 with 99mTc(CO)3 (Figure 15). The in vitro and in vivo properties of 99mTc(CO)3-H6-ADAPT6, 99mTc(CO)3-(HE)3-ADAPT6, 111In-DOTA-C-(H)6-ADAPT6 and 111In -DOTA-C-(HE)3-ADAPT6 were studied.

Figure 15. ADAPT6 protein domain. (A) Primary sequence of the protein domain. (B) Visualization of the N-terminus of the different constructs evaluated in this study. (C) Three-dimensional structure of the ABD domain, which was used as a scaffold for the development of ADAPTs (Kraulis et al., 1996).

Results

DOTA-C-(H)6-ADAPT6, DOTA-C-(HE)3-ADAPT6, (H)6-ADAPT6 and (HE)3-ADAPT6 were successfully produced and purified by our collabora-tors at the KTH-Royal Institute of Technology (Stockholm). The identity of the proteins was confirmed using mass spectrometry. The purity of proteins, as measured using a reversed phase HPLC, was greater than 95%. Measure-ments of binding kinetics using a BiacoreTM SPR system demonstrated a dissociation constant at equilibrium (KD) of 1-8 nmol/L for different ADAPT6 variants binding to HER2 (Table 1 in Paper II). The labeling with 111In was performed in 0.2 M ammonium acetate, pH 5.5, at 95 °C for 35 min. The labeling yield was more than 96%. There was no measurable release of the111In labels after incubation with 500-fold EDTA for 2 h. Labeling with 99mTc(CO)3 was performed using an IsoLink kit at 50 °C for 2 h. The labeling yield of H6-ADAPT6 with 99mTc(CO)3 was greater than 95%. The label was stable under a challenge with 5000-fold excess of L-histidine for 4 h.

53

Labeling of (HE)3-ADAPT6 with 99mTc(CO)3 was not stable under the same labeling conditions as for H6-ADAPT6. After challenge with 5000-fold ex-cess of L-histidine, the radiochemical purity was reduced from 96 to 82%. It is possible that some cluster of amino acids on the ADAPT6 surface can also bind to 99mTc(CO)3, although not as strongly as the histidine-containing sequence. Very strong binding to hexahistidine tags competed with this “chelating cluster” successfully and bound to 99mTc(CO)3 completely. The binding of 99mTc(CO)3 to (HE)3-tags might be not as strong, and a part of 99mTc(CO)3 is bound to a weaker “chelating cluster”. Pre-separation chal-lenge of the conjugate with 5000‐foldexcessofL‐histidineat50°Cfor25minstripped99mTcfromaweakchelatingsitebutnotfromthe(HE)3‐tag.Inthismanner,theoverallradiochemicalyieldwasreduced(ca.45%),buttheremaining99mTc‐activitywasstronglyboundtoADAPT6andcouldnotberemovedbyanadditionalchallengewithL‐histidine.

All conjugates labeled with 111In and 99mTc(CO)3 indicated specific binding to SKOV-3 cells with single digit affinities of 1.1-2.8 nM (Table 2 and Fig-ure 2 in Paper II).

The internalization of all labeled variants by SKOV-3 cells was slow (Figure 4 in Paper II), in agreement with data published in the Paper I. Figure 16 shows the biodistribution of 111In-DOTA-C-(HE)3-ADAPT6 and 111In-DOTA-C-H6-ADAPT6 in NMRI mice, demonstrating rapid clearance of both conjugates from the blood in normal NMRI mice. Both variants had equally high renal re-absorption. Interestingly, the influence of histidine-containing tags on the distribution of ADAPT6 was different, compared to the influence on Affibody molecules and other targeting proteins. The blood concertation of 111In-DOTA-C-H6-ADAPT6 was significantly higher than the concentration of 111In-DOTA-C-(HE)3-ADAPT6 at only 1 h after injection, and the difference was rather small. The liver uptake of 111In-DOTA-C-(HE)3-ADAPT6 was not much lower than the uptake of 111In-DOTA-C-H6-ADAPT6.

To explain such abnormal behavior, we proposed the following explana-tion in Paper II: “A possible explanation for the reduced hepatic uptake of hexahistidine-containing constructs might be the influence of the diserine- and DEAVDANS sequences flanking the histidine tags. Incorporation of hydrophilic serines into peptide-based chelators at the N-terminus of af-fibody molecules has been shown to reduce hepatic uptake and hepatobiliary excretion of 99mTc-labeled imaging probes. The presence of a glutamate and two aspartates, as well as a serine and an asparagine, renders the DEAV-

54

DANS-sequence quite hydrophilic.” In this case, placement of addition glu-tamate does not increase the hydrophilicity of the tag.

Figure 16. Comparative biodistribution of 111In-DOTA-C-(HE)3-ADAPT6 and 111In-DOTA-C-H6-ADAPT6 in NMRI mice. The data are presented as mean values with standard deviations from four mice. An asterisk indicates a significant difference (p < 0.05) between values.

We attempted to reduce the renal uptake of 99mTc(CO)3-H6-ADAPT6 in NMRI mice using co-injection of Gelofusine and L-lysine, based on a proto-col that was successfully used for reduction of the renal uptake of HER2-binding nanobody (D’huyvetter et al., 2014). The results of the experiment are displayed in Figure 17. There was no reduction in the kidney uptake of 99mTc(CO)3-H6-ADAPT6. Surprisingly, there was a slight but significant (P < 0.05) reduction in the liver uptake of radioactivity for both Gelofusine and L-lysine.

Figure 17. Comparative biodistribution of 99mTc(CO)3-(HE)3-ADAPT6 and 99mTc(CO)3-H6-ADAPT6 in NMRI mice. (A) Biodistribution after injection of 1 μg of protein in NMRI mice. (B) Effects of coinjection of lysine and Gelofusine on the biodistribution of 99mTc(CO)3-H6-ADAPT6 in NMRI mice. The data presented are the mean values with standard deviations from four mice, 1 h p.i. An asterisk indi-cates a significant difference (p <0.05) between values.

55

The biodistribution of 99mTc(CO)3-H6-ADAPT6 in BALB/C nu/nu mice is displayed in Figure 18. Apparently, this tracer provides specific uptake in HER2-expressing xenografts and provides very good tumor-to-organ ratios, enabling its use as an imaging probe.

Figure 18. Biodistribution of 99mTc(CO)3-H6-ADAPT6 in BALB/c nu/nu mice bear-ing SKOV-3 expressing xenografts. (A) Presented as the injected dose per gram of tissue. (B) Presented as the tumor-to-organ ratios. The data displayed are average ratios with standard deviations from four mice. An asterisk indicates a significant difference (p < 0.05) between values.

Figure 19. Micro SPECT/CT tumor imaging using 99mTc(CO)3-H6- ADAPT6 at 4 h p.i. Imaging was performed (A) without or (B) with presaturation of HER2 by pre-injection of trastuzumab. Both animals were injected with 1 MBq (1 μg) 99mTc(CO)3-H6-ADAPT6. To saturate the HER2 receptors, one animal (panel B) was subcutane-ously injected with 10 mg of trastuzumab at 48 and 24 h before injection of 99mTc(CO)3-H6-ADAPT6. The scale was adjusted to provide clear visualization of the nonsaturated tumor.

56

The main findings of this study were the following. - DOTA-C-(H)6-ADAPT6 and DOTA-C-(HE)3-ADAPT6 labeled with

111In, and (H)6-ADAPT6 and (HE)3-ADAPT6 labeled with 99mTc(CO)3 bind with high specificity and affinity to HER2-expressing cells.

- Labeling efficiency of the tested ADAPTs with 111In is high regard-less of the composition of the purification tag. In the case of 99mTc(CO)3-labeled ADAPTs, the labeling efficiency is dependent on the composition of the tag. The overall yield was more than 90% for H6-tag variant, compared to approximately 45% for (HE)3-tag.

- The biodistribution of 111In-labeled and 99mTc(CO)3-labeled ADAPTs was different. One hour after injection, hepatic uptake of 111In-DOTA-C-(HE)3-ADAPT6 was slightly but significantly (P < 0.05) higher than the uptake of 111In-DOTA-C-H6-ADAPT6. At a later time point, 4 h, there was no significant difference between the uptakes of 111In-labeled ADAPTs. Interestingly, liver, blood and bone uptake of 99mTc(CO)3-H6-ADAPT6 was significantly higher than uptake of 99mTc(CO)3-H6-ADAPT6 at 1 h after injection. At the 4 h time point, liver uptake was similar, but the blood and bone uptake of 99mTc(CO)3-H6-ADAPT6 was lower than the uptake of 99mTc(CO)3-(HE)3-ADAPT6.

- 99mTc(CO)3-H6-ADAPT6 provides high-contrast imaging of xeno-grafts with high HER2 expression.

- The influence of histidine-containing tags on the biodistribution of scaffold proteins depends on the composition of the scaffold protein and might differ appreciably. This finding should be considered in the molecular design of imaging probes based on engineered scaffold proteins.

In conclusion, the influence of histidine-containing tags on the bio-distribution of scaffold proteins depends on the composition of the scaffold protein and might differ appreciably. This finding should be considered in the molecular design of imaging probes based on engi-neered scaffold proteins. The use of hexahistidine tags at the N-terminus for labeling of ADAPTs with 99mTc(CO)3 provides imaging agents with favorable biodistribution and high tumor to-organ ratios.

57

Paper III

Influence of the N-terminal composition on the targeting properties of the radiometal-labeled anti-HER2 scaffold protein ADAPT6

Background and aim

The results of the previous study demonstrated the influence of histidine-containing tags on the biodistribution profile of ADAPTs. Moreover, it indi-cated that the composition of the N-terminal leading sequence, DEAV-DANCE, could affect the biodistribution of ADAPT6 (Paper II). The aim of this study was to evaluate the influences of different aspects of N-terminus leading sequences on targeting including:

The effect of sequence size on the clearance rate from blood and non-target tissues;

The effect of flanking amino acids in the presence of different histi-dine-containing tags; and

The effect of the composition of the sequence on off-target interac-tion of the tracers and their biodistribution profiles.

Methods

We used the following molecular design (Figure 20): A series of ADAPT6 having different leading sequences --

GCH6DANS, GC(HE)3DANS, GCDEAVDANS and GSVDANS at the N-terminal -- was created;

GCSS(HE)3DANS as a parental variant was used for comparison; N-terminal cysteine for site-specific conjugation of the chelator for

111In-labeling was introduced; and A maleimide-DOTA chelator was site-specifically conjugated for

111In-labeling. Stably, in vitro target binding, biodistribution and in vivo targeting properties of radiolabeled conjugates were characterized as de-scribed in the Material and methods section of Paper III.

58

Figure 20. A. Sequences of ADAPT6 variants evaluated in this study. N-terminal sequences are highlighted in bold font. B. Conjugation of a monoamide-maleimide derivative of DOTA to a cysteine.

Results

DOTA-GCSS(HE)3DEAVDANS-ADAPT6 (1), DOTA-GCH6DANS-ADAPT6 (2), DOTA-GC(HE)3DANS-ADAPT6 (3), DOTA-GCDEAVDANS-ADAPT6 (4) and DOTA-GCVDANS-ADAPT6 (5) were successfully produced and purified by our collaborators at the KTH-Royal Institute of Technology (Stockholm). The identity of the proteins was con-firmed using mass spectrometry. The purity of the proteins, as measured using reversed phase HPLC, was greater than 95%. Measurements of bind-ing kinetics using a BiacoreTM SPR system demonstrated a dissociation con-stant at equilibrium (KD) of 1-3 nmol/L for different ADAPT6 variants bind-ing to HER2. The calculated lipophilicity of the leading sequences was in the range from -1.3 to -5.9 (Table 1 in Paper III).

The labeling with 111In was performed in 0.2 M ammonium acetate, pH 5.5, at 95 °C for 35 min. The labeling yield for all the variants was more than 98%. There was no measurable release of the111In labels after incubation with 500-fold EDTA for 3 h (Table 2 and in Paper III). All conjugates labeled with 111In demonstrated specific binding to SKOV-3 cells with two different interactions. The affinity for the first interaction was 0.16-0.22 nM, and for the second, it was between 5 and 9 nM (Table 3 and in Paper III).

The internalization of all labeled variants by SKOV-3 and BT474 cells was slow (Figure 3 in paper III), in agreement with the data published in Papers I and II.

59

The biodistribution results of 111In-DOTA-1, 111In-DOTA-2 and 111In-DOTA-3 in NMRI mice 1 h and 4 h p.i. (Figure 21) demonstrated that pres-ence of an H6-tag was associated with elevated hepatic uptake. Already at 1 h p.i., the blood concentration of 111In-DOTA-3 was significantly lower than for 111In-DOTA-1 and 111In-DOTA-2. Uptake in the blood, lung, liver, colon and bone of the 111In-DOTA-3 variant was significantly lower than for 111In-DOTA-1 and 111In-DOTA-2 at 4 h p.i.

Figure 21. Biodistribution of 111In-DOTA-1 (N-terminal sequence GCSS(HE)3DEAVDANS), 111In-DOTA-2 (N-terminal sequence GCH6DANS) and 111In-DOTA-3 (N-terminal sequence GC(HE)3DANS) in NMRI mice after i.v. injec-tion. Data are presented as the mean values with standard deviations from four mice. Data for the gastrointestinal tract with content are presented as %ID per whole sample. An asterisk indicates a significant difference between values (deter-mined by one-way ANOVA with Bonferroni’s multiple comparison test).

The results of biodistribution in NMRI mice suggested using shorter variants in future studies because of favorable biodistribution. 111In-DOTA-2 (H6-tag containing variant) showed unfavorable biodistribution and was excluded from further investigations. Biodistribution of 111In-DOTA-3, 111In-DOTA-4 and 111In-DOTA-5 was studied in BALB/C nu/nu mice bearing SKOV-3 xenografts (4 h p.i.) (Table 5). Uptake of all variants in the blood, lung, spleen and muscle was less than that of 111In-DOTA-1 (parental variant). Hepatic uptake of 111In-DOTA-3 was also significantly lower. Surprisingly, clearance of 111In-DOTA-3 and 111In-DOTA-5 from the colon and muscle was faster than that of 111In-DOTA-4. Despite faster blood clearance of shorter variants, tumor uptake was not lower than the parental variant. Tu-mor uptake of 111In-DOTA-3 was significantly higher than that of 111In-DOTA-1. There were no significant differences between the tumor uptake of the shorter variants.

60

Table 5. Comparative biodistribution values of 111In-labeled ADAPT6 variants in BALB/C nu/nu mice bearing SKOV-3 xenografts (4 h p.i.). Data are presented as mean values with standard deviations from four mice. Data for the gastrointestinal tract with content are presented as %ID per whole sample. One-way ANOVA with Bonferroni’s multiple comparison test was performed to find significant differences.

111In-DOTA-1

GCSS(HE)3DEAVDANS

111In-DOTA-3

GC(HE)3DANS

111In-DOTA-4

GCDEAVDANS

111In-DOTA-5

GCVDANS

Blood 0.43±0.02 a 0.037±0.003 c 0.27±0.03 d, e 0.031±0.007 b

Lung 0.32±0.02 0.11±0.01 c 0.24±0.05 d, e 0.15±0.05 b

Liver 0.50±0.04 a 0.40±0.06 c 0.40±0.05 0.43±0.09

Spleen 0.22±0.01 0.15±0.03 c 0.18±0.03 0.14±0.04 b

Colon 0.2±0.1 0.11±0.02 c 0.16±0.01 d, e 0.09±0.02 b

Kidney 319±14 334±15 272±29 e 304±34

Tumor 9.1±2.0 14.6±2.4 c 10.1±2.4 12.5±1.3

Muscle 0.09±0.03 a 0.028±0.004 c 0.050±0.007 d, e 0.022±0.004 b

Bone 0.17±0.07 0.13±0.04 0.09±0.01 0.08±0.01 b

GI

tract*

0.46±0.11 0.25±0.06 c 0.59±0.34 0.28±0.04 b

*Data for the gastrointestinal tract with content are presented as %ID/whole sample a significant difference between 111In-DOTA-1 and 111In-DOTA-4 b significant difference between 111In-DOTA-1 and 111In-DOTA-5 c significant difference between 111In-DOTA-1 and 111In-DOTA-3 d significant difference between 111In-DOTA-4 and 111In-DOTA-5 e significant difference between 111In-DOTA-4 and 111In-DOTA-3 f significant difference between 111In-DOTA-3 and 111In-DOTA-5 The tumor-to-organ ratios of 111In-DOTA-3 and 111In-DOTA-5 for the major-ity of the organs were significantly higher than the tumor-to-organ ratios of 111In-DOTA-1 and 111In-DOTA-4 (Table 6). Tumor-to-blood ratios of 419±91 and 395±75, were obtained for 111In-DOTA-5 and 111In-DOTA-3, respectively. These ratios were much higher than the best-known ratios of 186 ± 29 and 175 ± 26, respectively, for Affibody molecules and DARPins in the same murine model (Goldstein et al., 2015; Wållberg et al., 2011). MicroSPECT/CT imaging at 4 h p.i. of 111In-DOTA-3 and 111In-DOTA-5 in BALB/C nu/nu mice bearing SKOV-3 xenografts showed the tumors clearly. Except for high kidney uptake, there was no uptake in other normal organs Figure 22).

61

Table 6. Tumor-to-organ ratios for 11IIn-labeled ADAPT6 variants in BALB/C nu/nu mice bearing SKOV-3 xenografts (4 h p.i.). Data are presented as mean values with standard deviations from four mice. One-way ANOVA with Bonferroni’s multiple comparison test was performed to find significant difference.

111In-DOTA-1

GCSS(HE)3DEAVDANS

111In-DOTA-3

GC(HE)3DANS

111In-DOTA-4

GCDEAVDANS

111In-DOTA-5

GCVDANS

Blood 21±4 a, b 395±75 c 37±6 d 419±91

Lung 29±7 a, b 129±15 c 41±4 d 90±28 e

Liver 19±5 a, b 37±7 c 25±5 30±4

Spleen 41±10 a, b 100±15 c 54±8 d 94±24

Colon 40±21 a, b 127±7 c 63±11 d 144±35

Kidney 0.029±0.006 a 0.044±0.008 0.037±0.005 0.041±0.004

Muscle 110±50 a, b 530±53 c 201±40 d 568±88

Bone 61±31 b 123±45 113±20 162±19 a significant difference between 111In-DOTA-1 and 111In-DOTA-3 b significant difference between 111In-DOTA-1 and 111In-DOTA-5 c significant difference between 111In-DOTA-3 and 111In-DOTA-4 d significant difference between 111In-DOTA-4 and 111In-DOTA-5 e significant difference between 111In-DOTA-3 and 111In-DOTA-5

Figure 22. Imaging of HER2 expression in SKOV-3 ovarian cancer xenografts in BALB/C nu/nu mice using 111In-DOTA-3 (left) and 111In-DOTA-5 (right). Micro-SPECT/CT images were acquired 4 h after injection. A rainbow color scale was adjusted to set a red color to the volume of maximum uptake in the tumor.

62

The main findings of this study were: - The size reduction of ADAPT6 at the N-terminus demonstrated the

potential to increase the blood clearance rate. The impact of His-tags on biodistribution is influenced by the size and charge/lipophilicity of the leading sequence. Replacing the H6-tag with an (HE)3-tag im-proved biodistribution of the ADAPT6 molecule with a short leading sequence. This finding was in agreement with earlier data concern-ing other scaffold proteins (Eder et al., 2013; Goldstein et al., 2015; Hofstrom et al., 2011; Tolmachev et al., 2010).

- 111In-DOTA-GC(HE)3DANS-ADAPT6 and 111In-DOTA-- GCVDANS-ADAPT6 were the best variants and provided tumor uptakes of 14.6±2.4 and 12.5±1.3%ID/g at 4 h p.i., respectively. Up-take in tumors for these variants was significantly (p < 0.05) higher than that for parental 111In-DOTA-GCSS(HE)3DEAVDANS-ADAPT6 (9.1±2.0%ID/g).

- Tumor-to-blood ratios for 111In-DOTA-GC(HE)3DANS-ADAPT6 and 111In-DOTA-GCVDANS-ADAPT6, were 395±75 and 419±91 at 4 h p.i., respectively. A high tumor-to-background ratio is crucial to overcome partial volume effects. By increasing the signal-to-noise ratio, the sensitivity of small metastasis imaging will be increased.

In conclusion, we showed that the N-terminal composition affects the biodistribution and targeting properties of ADAPT proteins. Moreover, it can modulate the influence of histidine-containing tags on the biodistribution. Thus, optimization of the composition of the N-terminus could improve the tumor-to-organ ratios of other radio-labeled scaffold proteins as well. This finding should be considered during the development of new scaffold-protein-based imaging probes.

63

Paper IV

Radionuclide tumor targeting using ADAPT scaffold proteins: aspects of label positioning and residualizing properties of the label

Background and aim

As discussed in the introduction to this thesis, the residualizing properties of a label and label position might alter the biodistribution and targeting proper-ties of the tracer.

The aim of this study was to evaluate the influence of the residualizing prop-erties and site-specific positioning of the label on the biodistribution of ADAPT6.

Methods

We used the following molecular design (Figure 23): Variants of ADAPT6 having (HE)3-tags and DANS leading se-

quences at the N-terminal were selected as the variants with the best biodistribution found in study 3;

N-terminal cysteine (Cys2-ADAPT6) or C-terminal cysteine (Cys59-ADAPT6) for site-specific conjugation of a chelator for 111In and a linker for 125I labeling was introduced;

A maleimide-DOTA chelator was used for 111In-labeling; and A ((4-hydroxyphenyl)ethyl) maleimide (HPEM) linker was used for

125I-labeling (Figure 23D).

64

Figure 23. Structures of Cys2-ADAPT6(A), Cys59-ADAPT6(B), maleimide-DOTA(C) and HPEM (D) conjugated to a cysteine.

To provide DOTA-Cys2-ADAPT6 and DOTA-Cys59-ADAPT6 for labeling with 111In, the maleimide-DOTA chelator was site-specifically conjugated to the cysteine in ADAPT6 variants. The conjugates were labeled with 111In according to the method in paper I. Indirect site-specific labeling of HPEM-Cys2-ADAPT6 and HPEM-Cys59-ADAPT6 with 125I was based on the meth-od used to label Affibody molecules (Tolmachev et al., 2009b). In vitro and in vivo properties of 111In-DOTA-Cys2-ADAPT6, 111In-DOTA-Cys59-ADAPT6, 125I-HPEM-Cys2-ADAPT6 and 125I-HPEM-Cys59-ADAPT6 were studied. Results

DOTA-Cys2-ADAPT6 and DOTA-Cys59-ADAPT6 for labeling with 111In and Cys2-ADAPT6 and Cys59-ADAPT6 for labeling with 125I were success-fully produced and purified. The identity of the proteins was confirmed us-ing mass spectrometry. The purity of the proteins, as measured using re-versed phase HPLC, was greater than 98%. Measurements of binding kinet-ics using a BiacoreTM SPR system demonstrated a dissociation constant at equilibrium (KD) values, 2.2 and 2.6 nM for Cys2-ADAPT6 and Cys59-ADAPT6, respectively. The labeling with 111In was performed in 0.2 M ammonium acetate, pH 5.5, at 95 °C for 35 min. The labeling yield for all of the variants was greater than 98%. There was no measurable release of the111In labels after incuba-tion with 500-fold EDTA for 4 h.

All radiolabeled conjugates labeled with 111In demonstrated specific binding to SKOV-3 cells with two different interactions. The affinity for the first

65

interaction was 0.1-0.2 nM, and for the second, it was between 3 and 11 nM (Figure 1 and Table 1 in Paper IV).

The internalization of all labeled variants by SKOV-3 and BT474 cells was slow and less than 30% over 24 h (Figure 3 in paper IV), in agreement with the data published in Papers I, II and III.

The in vivo specificity test indicated that 111In-DOTA-Cys59-ADAPT6 and 125I-HPEM-Cys59-ADAPT6 molecules accumulated specifically in HER2-expressing SKOV3 xenografts (Figure 3 in paper IV).

The biodistribution data for 111In-DOTA-Cys2-ADAPT6, 111In-DOTA-Cys59-ADAPT6, 125I-HPEM-Cys2-ADAPT6 and 125I-HPEM-Cys59-ADAPT6 at 1 and 4 h p.i. in mice bearing HER2-expressing SKOV-3 xenografts are displayed in Table 7. The tumor-to-organ ratios are presented in Table 8. MicroSPECT/CT imaging of HER2-expressing xenografts in mice using radiolabeled ADPT6 variants is presented in Figure 24.

66

Table 7. Biodistribution of radiolabeled ADAPT6 in BALB/C nu/nu mice bearing SKOV-3 xenografts.

111In-DOTA-Cys59-

ADAPT6

125I-HPEM-Cys59-

ADAPT6

111In-DOTA-Cys2-

ADAPT6

125I-HPEM-Cys2-

ADAPT6

1 h

Blood 0.35±0.06a,c 1.0±0.1b,h 0.48±0.10d,f 0.54±0.14e,h

Lung 0.46±0.06a,c 1.4±0.2b,h 0.5±0.1d,f 0.6±0.1e,h

Liver 0.21±0.03c 2.7±0.8 0.26±0.05f 2.4±0.7e

Spleen 0.14±0.02a,c,g 0.7±0.2b,h 0.23±0.06d,f 0.45±0.04e,h

Stomach 0.23±0.03a,c 1.6±0.3 0.25±0.07d,f 1.2±0.3e

Kidney 272±21c 10±2b 258±27f 10±3

Tumor 13±2c 19±3 14±3f 16±3

Muscle 0.10±0.02a,c 0.34±0.03b,h 0.10±0.05d,f 0.2±0.1e,h

Bone 0.21±0.05a,c 0.9±0.2b,h 0.22±0.04d 0.31±0.08e,h

4 h

Blood 0.053±0.004c 0.41±0.06h 0.04±0.01f 0.20±0.05 h

Lung 0.13±0.03c,g 0.39±0.05h 0.19±0.02f 0.17±0.04h

Liver 0.17±0.02g 1.6±0.8 0.22±0.01 0.6±0.3

Spleen 0.10±0.00 0.3±0.2 0.12±0.01 0.15±0.03

Stomach 0.10±0.01c 1.3±0.3 0.09±0.02f 0.8±0.4f

Kidney 284±21c 1.7±0.3 241±32f 1.7±0.3

Tumor 15±2c,g 22±3h 10.6±0.5f 11.8±0.5h

Muscle 0.04±0.01c 0.18±0.08 0.046±0.005 0.09±0.04

Bone 0.07±0.02c 0.31±0.06b,h 0.07±0.02 0.12±0.06h

*Data for the gastrointestinal tract with content are presented as %ID/whole sample. a,b significant difference between uptake of Cys59-ADAPT6 at 1 and 4 h p.i. c significant difference between uptake of Cys59-ADAPT6 labeled with 111In and 125I at the same time point d,e significant difference between uptake of Cys2-ADAPT6 at 1 and 4 h p.i. f significant difference between uptake of Cys2-ADAPT6 labeled with 111In and 125I at the same time point g,h significant difference between uptake of Cys59-ADAPT6 and Cys2-ADAPT6 at the same time point

67

Table 8. Tumor-to-organ ratios of 111In-DOTA-Cys2-ADAPT6, 111In-DOTA-Cys59-ADAPT6, 125I-HPEM-Cys2-ADAPT6 and 125I-HPEM-Cys59-ADAPT6 molecules in BALB/C nu/nu mice bearing SKOV3 xenografts at 4 h p.i. Results are presented as the average and standard deviation of 4 animals.

111In-DOTA-Cys59-

ADAPT6

125I-HPEM-Cys59-

ADAPT6

111In-DOTA-Cys2-

ADAPT6

125I-HPEM-Cys2-

ADAPT6

1 h

Blood 38±3 a,c,g 18.8±0.8 b,h 28±2 d,f 31±3e

Lung 29.4±4.0a,c 13.1±0.7 b,h 26±4f, d 28±2e

Liver 64±10c 7±1 53±5f 7±2e

Spleen 94±12a,c,g 27±4b 59±11d,f 37±9e

Stomach 59±6a,c 12±1b 56±9 d,f 14±2

Kidney 0.049±0.003c 1.9±0.2b 0.05±0.01f 1.6±0.2e

Muscle 144±37a,c 54±7b 111±9 d,f 79±29

Bone 66±11c 22±5b,h 62±4 d,f 54±4e

4 h

Blood 277±35 c 53±10 b 254±37 f 61±14 e

Lung 119±35a,c,g 56±3 55±6d,f 73±14e

Liver 85±17c,g 17±9 48±3f 24±12e

Spleen 132±4a,c,g 59±12b 93±7d,f 82±15e

Stomach 147±24a,c 17±2b 116±19d,f 17±8

Kidney 0.05±0.01c 13±3b,h 0.04±0.01f 7±1e

Muscle 343±85c,g 134±43b 231±18d,f, 158±63

Bone 246±163 72±21b 152±42d,f 119±45e

*Data for the gastrointestinal tract with content are presented as %ID/whole sample. a,b significant difference between values for Cys59-ADAPT6 at 1 and 4 h p.i. c significant difference (paired t-test) between values for Cys59-ADAPT6 labeled with 111In and 125I at the same time point d,e significant difference between values for Cys2-ADAPT6 at 1 and 4 h p.i. f significant difference (paired t-test) between values for Cys2-ADAPT6 labeled with 111In and 125I at the same time point g,h significant difference between values for Cys59-ADAPT6 and Cys2-ADAPT6 at the same time point

68

Figure 24. MicroSPECT/CT imaging of HER2-expressing xenografts using ADAPT6 molecules at 4 h p.i. (maximum intensity projections). (A and C) Imaging using 111In-DOTA-Cys2-ADAPT6 and 111In-DOTA-Cys59-ADAPT6. The linear color scale was adjusted to provide clear visualization of the tumors. (B and D) Imaging using 125I-HPEM-Cys2-ADAPT6 and 125I-HPEM-Cys59-ADAPT6. Full linear color scale was applied.

The main findings of this study were the following. - 111In-DOTA-Cys59-ADAPT6, 125I-HPEM-Cys59-ADAPT6, 111In-

DOTA-Cys2-ADAPT6, and 125I-HPEM-Cys2-ADAPT6 bind to HER2-expressing cells with high specificity and affinity;

- Efficiency of 111In-labeling of ADAPT6 variants is high. The overall yield is greater than 98%.

- The radioiodination yield of HPEM is 49±8%. The overall yield of radioiodination of Cys2-ADAPT6 and Cys59-ADAPT6 using HPEM is 48% and 35% respectively.

- The uptake of radioiodinated conjugates in SKOV3 xenografts was significantly higher than the uptake of their radiometal-labeled coun-terparts. At 4 h p.i., the tumor uptake was 15±2, 22±3, 10.6±0.5, and 11.8±0.5% ID/g for 111In-DOTA-Cys59-ADAPT6, 125I-HPEM-Cys59-ADAPT6, 111In-DOTA-Cys2-ADAPT6, and 125I-HPEM-Cys2-ADAPT6, respectively;

- Tumor-to-blood ratios were 277±35, 53±10, 254±37 and 61±14, re-spectively, at the 4 h time point for 111In-DOTA-Cys59-ADAPT6, 125I-HPEM-Cys59-ADAPT6, 111In-DOTA-Cys2-ADAPT6, and 125I-HPEM-Cys2-ADAPT6.

- We found that renal retention of the radioiodinated variants (renal up-take of 1.7±0.3 ID/g at 4 h p.i., regardless of position) was ca. 150 lower compared to 111In-labeled variants. However, the uptake in oth-er organs was higher for radioiodine, most likely due to release of ra-diometabolites from the kidneys into the blood. Therefore, the tumor-to-organ ratios (excluding kidneys) were higher for 111In-DOTA-labeled tracers.

69

- Overall, 111In-DOTA-Cys59-ADAPT6 provided significantly higher tumor-to-lung, tumor-to-liver, tumor-to-spleen and tumor-to-muscle ratios than 111In-DOTA-Cys2-ADAPT6.

- The tumor-to-organ ratios were similar for 125I-HPEM-Cys59-ADAPT6 and 125I-HPEM-Cys2-ADAPT6, but 125I-HPEM-Cys59-ADAPT6 had twice as high a tumor uptake and provided twice as high a tumor-to-kidney ratio.

In conclusion, both the character and position of labels had substan-tial influences on the imaging properties of ADAPT6. Non-residualizing radiohalogen labels decreased renal uptake dramatical-ly. An optimal position for both 125I-HPEM and 111In-DOTA labels is at the C-terminus.

70

Paper V

Comparative evaluation of tumor targeting using an an-ti-HER2 ADAPT scaffold protein labeled at the C-terminus with indium-111 or technetium-99m.

Background and aims

Previous studies with Affibody molecules have demonstrated a profound influence of the chemical properties of the radionuclide (Strand et al., 2014; Westerlund et al., 2016), the chelator (Altai et al., 2013) and the label posi-tion (Honarvar et al., 2014) on the imaging properties of scaffold proteins. Introduction of cysteine at the C-terminal of recombinantly produced Af-fibody molecules can form an N3S chelator for labeling with 99mTc in a sin-gle process (Ahlgren et al., 2009a). The choice of an optimal composition of the N3S chelator might help us to improve biodistribution with reduced renal uptake (Wållberg et al., 2011). The aim of this study was to the compare tumor-targeting properties of ADAPT6 (having an N-terminal DEAVDANS sequence and C-terminal Cys) when labeled with 99mTc through an N3S chelator and 111In through a DOTA chelator.

Methods

We used the following molecular design (Figure 23). Variants of HER2-binding ADAPT6 having a DEAVDANS leading

sequence at the N-terminal were selected for this study. Although a variant with the (HE)3-tag had a better biodistribution profile, it was shown that the presence of this tag creates an alternative chelator for 99mTc=O and should be avoided due to loss of labeling site specifici-ty (Lindberg et al., 2012).

C-terminal cysteine was introduced for site-specific conjugation of a chelator for 111In-labeling.

A maleimide-DOTA chelator was conjugated for 111In-labeling. The C-terminal cysteine-containing peptide-based chelator GSSC

was engineered in ADAPT6 for labeling using 99mTc.

71

A maleimide-derivative of macrocyclic DOTA chelator was site-specifically conjugated to the C-terminal cysteine in DEAVDANS-ADAPT6-GSSC. The DEAVDANS-ADAPT6-GSSC-DOTA conjugate was labeled with 111In ac-cording to the methods in paper I. GSSC-(N3S chelator) in DEAVDANS-ADAPT6-GSSC was used for labeling with 99mTc, based on the method used for labeling of Affibody molecules (Figure 25) (Ahlgren et al., 2009a; Wåll-berg et al., 2011). The in vitro and in vivo properties of 111In- DEAVDANS-ADAPT6-GSSC-DOTA and 99mTc-DEAVDANS-ADAPT6-GSSC were studied as described in Paper V.

GDEAVDANSLAAAKETALYHLDRLGVADAYKDLIDKAKTYEGV-KARYFEILHALPGSSC

Figure 25. Amino acid sequence of DEAVDANS-ADAPT6-GSSC and structure of the labels 111In-DOTA-MMA-Cys59 (left) and 99mTc-GSSC (right).

Results

DEAVDANS-ADAPT6-GSSC-DOTA for labeling with 111In and DEAV-DANS-ADAPT6-GSSC for labeling with 99mTc were successfully produced and purified. The authenticity of the proteins was confirmed using mass spectrometry. The purity of the proteins, measured using reversed phase HPLC, was greater than 98%. Measurements of binding kinetics using a BiacoreTM SPR system demonstrated a dissociation constant at equilibrium (KD) value of 4 nM. Labeling with 111In was performed in 0.2 M ammonium acetate, pH 5.5, at 95 °C for 35 min. The labeling yield for all the variants was greater than 98%. There was no measurable release of the111In labels after incubation with 500-fold EDTA for 4 h.

DEAVDANS-ADAPT6-GSSC was labeled with 99mTc with the isolated yield of 31±17%. The radiochemical purity after size-exclusion purification

N

N

N

N

NH

O

N

O

O

O

O-

O

O--O

O

S

In3+

N OH

O

N

O

N

O

O

S

Tc

O

HO

HO

Pr

Pr

72

was 98±2%. Post-purification incubation with 300-fold cysteine for 4 h showed less than 2% radionuclide release from the conjugate.

All conjugates labeled with 111In and 99mTc indicated specific binding to SKOV-3 cells with two different interactions. The affinity for the first inter-action was 30-220 pM, and for the second, it was between 1 and 2 nM (Fig-ure 3 and Table 1 in Paper V).

The internalization of all labeled variants by SKOV-3 and BT474 cells was slow and was less than 30% over 24 h (Figure 4 in paper V), in agreement with the data published in Papers I, II and III and IV.

The in vivo specificity test indicated that the 111In-DEAVDANS-ADAPT6-GSSC-DOTA and 99mTc-DEAVDANS-ADAPT6-GSSC molecules specifi-cally accumulated in HER2-expressing SKOV-3 xenografts (Figure 5 in paper V).

The biodistribution results of 111In- DEAVDANS-ADAPT6-GSSC-DOTA and 99mTc-DEAVDANS-ADAPT6-GSSC at 1 and 4 h p.i. in mice bearing HER2-expressing SKOV-3 xenografts are displayed in Tables 9 and 10, respectively.

73

Table 9. Biodistribution of 111In-DEAVDANS-ADAPT6-GSSC-DOTA and 99mTc - DEAVDANS-ADAPT6-GSSC molecules in BALB/C nu/nu mice bearing SKOV-3 xenografts at 1 and 4 h p.i. Results are presented as the average %ID/g and stand-ard deviation of 4 animals.

Data for the gastrointestinal tract with content and the carcass are presented as %ID/whole sample. a significant difference between uptake of tracer at 1 and 4 h p.i. b significant difference between uptake of conjugate labeled with 111In and 99mTc at the same time point

1 h 4 h

111In-

DEAVDANS-

ADAPT6-GSSC-

DOTA

99mTc -

DEAVDANS-

ADAPT6-

GSSC

111In-

DEAVDANS-

ADAPT6-

GSSC-DOTA

99mTc -

DEAVDANS-

ADAPT6-

GSSC

Blood 0.4±0.1a 1.2±0.1a,b 0.10±0.02 b 0.34±0.08b

Salivary gland 0.18±0.06 0.6±0.1a,b 0.10±0.02 b 0.27±0.05 b

Lung 0.5±0.1a 1.2±0.1a,b 0.19±0.04 b 0.41±0.15 b

Liver 0.26±0.07 1.9±0.2b 0.31±0.05 b 1.7±0.2b

Spleen 0.21±0.05 0.82±0.04b 0.14±0.03 b 0.8±0.1b

Stomach 0.2±0.1 1.0±0.5b 0.10±0.02 b 0.5±0.3b

Kidney 415±67 312±12a,b 410±52 b 118±10a,b

tumor 18±4 19±3 21±6 19±3

Muscle 0.10±0.03a 0.33±0.09a,b 0.05±0.01 b 0.13±0.03 b

Bone 0.20±0.07 0.75±0.07a,b 0.11±0.04 b 0.47±0.17 ab

GI tract 2.3±0.7 1.6±0.4 1.5±0.5 1.9±0.99

Carcass 0.6±0.1 10±4b 0.09±0.03 b 6±2b

Blood 0.4±0.1a 1.2±0.1a,b 0.10±0.02 b 0.34±0.08b

74

Table 10. Tumor-to-organ ratio of 111In-DEAVDANS-ADAPT6-GSSC-DOTA and 99mTc-DEAVDANS-ADAPT6-GSSC in BALB/C nu/nu mice bearing SKOV-3 xeno-grafts at 1 and 4 h p.i. Results are presented as the average % ID/g and standard deviation of 4 animals. 1 h 4 h

111In-

DEAVDANS-

ADAPT6-

GSSC-DOTA

99mTc-

DEAVDANS-

ADAPT6-GSSC

111In-

DEAVDANS-

ADAPT6-

GSSC-DOTA

99mTc-

DEAVDANS-

ADAPT6-GSSC

Blood 49±6a,b 15±2a,b 206±15a,b 54±4a,b

Salivary gland 104±21a,b 28±7a,b 225±54a,b 64±8a,b

Lung 38±5a,b 15±2a,b 114±20a,b 47±7a,b

Liver 69±4b 10±1b 68±11b 11±1b

Spleen 87±26a,b 23±4b 148±17a,b 25±2b

Stomach 100±29a,b 22±9a,b 210±34a,b 46±16a,b

Kidney 0.04±0.01b 0,06±0.01a,b 0.05±0.01b 0.16±0.03a,b

Muscle 181±20a,b 57±12b 472±93a,b 134±21b

Bone 99±30a,b 25±2a,b 210±31a,b 42±10a,b

Data for the gastrointestinal tract with content are presented as %ID/whole sample. a significant difference between uptake of tracer at 1 and 4 h p.i. b significant difference between uptake of conjugate labeled with 111In and 99mTc at the same time point

HER2-expressing SKOV-3 xenografts in mice were clearly visualized using microSPECT/CT with radiolabeled ADAPT6 variants, as presented in Figure 26.

Figure 26. MicroSPECT/CT imaging of HER2-expressing SKOV-3 xenografts (left hind legs) at 4 h p.i. of 99mTc-DEAVDANS-ADAPT6-GSSC (left panel) and 111In-DEAVDANS-ADAPT6-GSSC-DOTA (right panel). Color scales were adjusted to provide clear visualization of tumors.

75

A comparison between the biodistribution of 111In-DEAVDANS-ADAPT6-GSSC-DOTA (C-terminal label) and 111In- DOTA-DEAVDANS-ADAPT6 (N-terminal label) is presented in Figure 27.

Up

take

, % ID

/g

blood

lung

liver

sple

en

kidney

tum

our

musc

lebone

0.0

0.2

0.4

0.6

0.8

1.0

100

200

300

400

5004 h In C4 h In N

* *

*A

Tu

mo

ur-

to-o

rgan

rat

io

blood

lung

liver

sple

en

musc

lebone

0

200

400

600

C-terminalN-terminal

*

** *

*

*

*

B

Figure 27. Comparison of 111In-DEAVDANS-ADAPT6-GSSC-DOTA (C-terminal label) and 111In- DOTA-DEAVDANS-ADAPT6 (N-terminal label). Biodistribution (A) and tumor-to-organ ratios (B) at 4 h p.i. in BALB/C nu/nu mice bearing SKOV-3 xenografts. Data for 111In-DOTA-DEAVDANS-ADAPT6 were obtained from (Ga-rousi 2016).

The main findings of this study were the following. - 111In-DEAVDANS-ADAPT6-GSSC-DOTA and 99mTc-DEAVDANS-

ADAPT6-GSSC bind to HER2-expressing cells with high specificity and affinity.

- The labeling efficiency of 111In-labeled ADAPT6 variants is high. The overall yield was greater than 98%.

- The overall labeling yield of DEAVDANS-ADAPT6-GSSC with 99mTc was 31±17%.

- The biodistribution profile demonstrated dramatic differences be-tween the renal retention of 111In-DEAVDANS-ADAPT6-GSSC-DOTA and 99mTc - DEAVDANS-ADAPT6-GSSC (Table 9), perhaps because of the weaker residualizing properties of 99mTc-label. Redis-tribution of the leaked radio-metabolites of 99mTc-DEAVDANS-ADAPT6-GSSC from the kidney can increase the uptake of radionu-clide in other organs. Higher uptake of 99mTc-DEAVDANS-ADAPT6-GSSC compared to 111In-DEAVDANS-ADAPT6-GSSC-DOTA in normal organs and in the blood confirms this hypothesis.

76

- There was no significant difference between the tumor uptakes of 99mTc- and 111In-labeled ADAPT6 at any time point (18-21%ID/g for both conjugates).

- The tumor-to-organ ratios in the case of 111In-label were significantly (P < 0.05) higher at both 1 and 4 h p.i. for all of the organs except for the kidneys (Table 10). At 4 h, the ratios for the liver and spleen were approximately six-fold higher, for stomach, the ratio was approxi-mately five-fold higher, and for the blood and salivary gland, it was approximately two-fold higher.

- Even with the suboptimal 99mTc-GSSC label, 99mTc-DEAVDANS-ADAPT6-GSSC provided much higher tumor-to-organ ratios com-pared with monoclonal antibodies. For example, the tumor-to-blood, tumor-to-liver and tumor-to-lung ratios in murine models have been shown to be less than 10 at 72 h for any HER2-targeting antibody (Tolmachev 2008).

- A comparison (Figure 27) demonstrated that the C-terminal 111In-label had an appreciably lower blood radioactivity concentration and hepatic uptake but higher tumor uptake (21±6 vs 10.1±2.4%ID/g) than the N-terminal 111In-label.

In conclusion, placement of an 111In-DOTA label at the C-terminus of DEAVDANS-ADAPT6 provides appreciably higher tumor-to-organ ratios than the same placement of a 99mTc-GSSC label. This placement should provide better sensitivity for the imaging of small metastases. Therefore, a C-terminal 111In-DOTA would be the pre-ferred label for ADAPT6.

77

Concluding remarks

We proved the feasibility of molecular imaging using ADAPTs in paper I. Based on experiences with Affibody molecules, we planned several studies to evaluate the effects of different factors that might influence the targeting properties and biodistribution of ADAPTs. In paper II, we evaluated the influence of the composition of histidine-containing tags at the N-terminus of ADAPT6. In paper III, the influences of the size and composition of leading sequences and their proximity to histidine-containing tags were in-vestigated. In paper IV, aspects of label positioning and the residualizing properties of the label were evaluated. We evaluated the tumor-targeting properties of ADAPT6 (having an N-terminal DEAVDANS sequence and C-terminal Cys), which was labeled with 99mTc through an N3S chelator and 111In through a DOTA chelator in paper V

The main findings of this thesis were:

Paper I

An ABD with eradicated binding to albumin is a very promising scaf-fold for the development of targeting probes for radionuclide molecu-lar imaging.

Paper II

The influence of a histidine-containing tag on the biodistribution of scaffold proteins depends on the composition of the scaffold protein and might differ appreciably. This finding should be considered in the molecular design of imaging probes based on engineered scaffold proteins. The use of hexahistidine tags for labeling of ADAPTs with 99mTc(CO)3 provides imaging agents with favorable biodistribution and high tumor to-organ ratios.

78

Paper III

We showed that the N-terminal composition affects the biodistribu-tion and targeting properties of ADAPT proteins. Moreover, it can modulate the influence of histidine-containing tags on the biodistribu-tion. Thus, optimization of the composition of the N-terminus could improve the tumor-to-organ ratios of other radiolabeled scaffold pro-teins as well. This finding should be considered during the develop-ment of new scaffold-protein-based imaging probes.

Paper IV

Both the character and position of labels had substantial influences on the imaging properties of ADAPT6. Non-residualizing radiohal-ogen labels decreased renal uptake dramatically. An optimal position for both thee 125I-HPEM and 111In-DOTA labels is at the C-terminus.

Paper V

Placement of an 111In-DOTA label at the C-terminus of DEAVDANS-ADAPT6 provides appreciably higher tumor-to-organ ratios than the same placement of a 99mTc-GSSC label. This placement should pro-vide better sensitivity in the imaging of small metastases. Therefore, a C-terminal 111In-DOTA would be the preferred label for ADAPT6.

Overall, ADAPTs constitute a very promising class of targeting probes

for molecular imaging, providing high contrast. The composition of the ADAPT proteins and the chelators/linkers for labeling have appreciable effects on their biodistribution. Further studies concerning the structure-properties relationship might additionally improve the imaging properties of ADAPTs and reveal the full potential of these imaging probes.

79

Ongoing and future studies

Based on the results of this thesis and an analysis of the field of molecular imaging, I propose the following studies to continue the research concerning the use of ADAPT for molecular imaging.

Comparative in vitro and in vivo evaluations of G-(HE)3DANS-ADAPT6-GSSC-DOTA and G-DEAVDANS-ADAPT6-GSSC-DOTA labeled with 68Ga at the C-terminus

Evaluation of the influences of different chelators (NOTA, NODAGA, DOTA and DOTAGA) positioned at the C-terminus on the biodistribution of 68Ga-labeled ADAPT6, which should be performed for the best variant selected in the previous study

The first-in-humans study should be performed to confirm the translational potential of ADAPTs. A protocol used by Sörensen and co-workers (2016) for the evaluation of 68Ga-ABY-25 Af-fibody molecules might be applied for such a study.

Development of labeling of ADAPT6 with 18F; for this purpose, I consider two different strategies: - The use of chelation of 18F-aluminum fluoride (18F-ALF), which usually provides relatively high yield but might cause high kidney uptake (Heskamp et al., 2012); and - The use of an organic linker, such as 18F-fluorobenzaldehyde (Rosik et al., 2013) or 18F-fluorophenyloxadiazole methylsulfone (Chiotellis et al., 2016). Considering that the use of a hydropho-bic linker increases the risk of hepatobiliary excretion, I consider the use of a highly hydrophilic triglutamyl spacer between the la-bel and ADAPT protein.

Evaluation of 131I-labeled ADAPT6 for therapy. For this purpose, biodistribution of 131I-ADAPT6 should be measured at several time points ranging from 20 min to 2 weeks after injection. Based on the biodistribution data, the residence of nuclides in tumors and healthy tissues will be calculated, and the dosimetry will be evaluated. In the case of favorable dosimetry (a tumor dose of more than 60 Gy, a renal dose of less than 23 Gy, and a bone marrow dose of less than 2 Gy), experimental therapy will be per-formed.

80

Evaluation of synthetically produced ADAPT6-based imaging probes. The relatively small number of amino acids in ADAPTs should allow for their production using peptide synthesis. Peptide synthesis might offer less expensive GMP production, compared to recombinant production.

Selection and evaluation of ADAPTs targeting other molecular targets, such as PSMA (prostate cancer), EpCAM, CEA, TAG72, and MUC1 (pancarcinoma antigens), to use them for therapy and for companion diagnostics.

81

Acknowledgement

I am thankful to everyone who has supported, helped and encouraged me during my thesis work. This thesis work would not have been pos-sible without help from these people.

I would like to express my special appreciation to the following: Vladimir, thank you so much for offering me the opportunity to work on very exciting projects! I learned many things from you about the logical way of thinking, planning, being precise and serious in doing a good job. I enjoyed the scientific discussions we had together and during the group meetings. Your suggestions to solve the problems during studies were really smart. Anna, thank you for your help and useful advices during in vitro and in vivo experiments! You helped me a lot with biodistribution studies. Your plans for celebrations on different occasions and your hospitality were fantastic. Jos, thank you for sharing your knowledge with me about affinity measurements! It was nice to work with you. Fredrik, thank you for your help and advices especially regarding the anti-CAIX project! I enjoyed working with you. Hadis, you taught me almost all the common techniques which we use in BMS group. It was great to work with you in a friendly atmosphere. Thanks a lot! Mohammad, you were a user-friendly encyclopedia in our room. It was easy to talk to you about everything even when you were so busy. Thank you so much! Joanna, I really enjoyed to talk with you about religion, culture and women rights in Iran and Sweden. Bogdan, your glorious PET/SPECT images are in many of our publi-cations. Thank you for the nice collaboration. You are an easy-going friend and helped me a lot. Thank you very much! Maria, it was a pleasure to work with you. I think you are a strong person that plans everything very well. I wish you good luck for your disputation!

82

Zohreh, it’s great to meet you sometimes in conferences with new presentations. Good luck in your research! Maryam, it was nice to discuss experimental results with you in Per-sian. I wish you good luck in your research! Anzhelika, thank you for helping me with Ga-generator milking! Your scientific questions about our projects encouraged me to study more. Sara R., thank you for sharing your experimental results at the group meetings! It was helpful for me to learn something new from you. Thank you for the chocolate and cookies you shared with me! Kicki (Christina), you helped us a lot for ordering chemicals and taking care of the cell labs. Thank you for arranging fika and nice ac-tivities for all BMS!

All BMS, Bo, Marika, Hanna, Jonas, John, Karl, Sina, Anja, Diana, Sara L., Andris, Amelie, Jorgen, and Lars, thank you for your help, the nice discussions and activities we had together.

All PPP, Ram, Ole, Sergio, Olof, and Veronika, thank you for helping me to use the facilities in your group!

Akademiska Sjukhuset, Aseel, Madina, Matias, Ulrike Garske, Anders Sundin, thank you all for your help and support. Dawoud, thank you for helping me to find accommodation in Uppala!

Our collaborators in KTH, Sophia, thank you for the amazing collaboration, meetings and nice discussions that we had together about the future plans of our projects! Sarah, thank you so much for providing me all the ADAPT variants for the projects! You did a great job and I wish you the best for your future! Stefan, it was a great pleasure to work in your group and produce an-ti-EGFR Affibody molecules. Ken, your supervision in producing anti-EGFR protein was great and I learned many things from you and your colleagues. I wish you good luck in your disputation! I would like to thank all the people who supported me during my pro-ject work in KTH.

83

Former and current staff at BMS and IGP administration, Christle Richter Frohm, Heléne Lyngå, Christina Magnusson, Ulla Steimer, Camilla Sandgren, Helene Norlin and all others, thank you very much for all your support during my PhD!

Ångstrom, Oommen, OP, and Jons, thank you so much for your help and sup-port during my master thesis work in Ångstrom! You are great friends!

I’m thankful to all people who helped me in this journey here in Swe-den and my friends and former colleagues in Iran. My beloved parents, You are not in this world, but I will never forget you. Thank you for everything you have done for me!

My beloved siblings and their families, Without your support, I could not reach this point. I am thankful to all of you!

My beloved daughters, Saba and Setayesh, you are my angels. You are full of energy and happiness. Thank you for bringing me happiness!

Last but not least to my beloved Wife, Fariba, I don’t say just thank you for love, support…I’m proud of you! You are the most important person in my life. Thank you for your wonderful lovely companionship!

84

References

Ackerman, S.E., Currier, N.V., Bergen, J.M., Cochran, J.R., 2014. Cystine-knot peptides: emerging tools for cancer imaging and therapy. Ex-pert Rev. Proteomics 11, 561–572.

Adam, M.J., Wilbur, D.S., 2005. Radiohalogens for imaging and therapy. Chem. Soc. Rev. 34, 153–163.

Ahlgren, S., Orlova, A., Rosik, D., Sandström, M., Sjöberg, A., Baastrup, B., Widmark, O., Fant, G., Feldwisch, J., Tolmachev, V., 2008. Evalua-tion of maleimide derivative of DOTA for site-specific labeling of recombinant affibody molecules. Bioconjug. Chem. 19, 235–243.

Ahlgren, S., Tolmachev, V., 2010. Radionuclide molecular imaging using Affibody molecules. Curr. Pharm. Biotechnol. 11, 581–589.

Ahlgren, S., Wållberg, H., Tran, T.A., Widström, C., Hjertman, M., Abrahmsén, L., Berndorff, D., Dinkelborg, L.M., Cyr, J.E., Feld-wisch, J., others, 2009a. Targeting of HER2-expressing tumors with a site-specifically 99mTc-labeled recombinant affibody molecule, ZHER2: 2395, with C-terminally engineered cysteine. J. Nucl. Med. 50, 781–789.

Ahlgren, S., Wållberg, H., Tran, T.A., Widström, C., Hjertman, M., Abrahmsén, L., Berndorff, D., Dinkelborg, L.M., Cyr, J.E., Feld-wisch, J., Orlova, A., Tolmachev, V., 2009b. Targeting of HER2-expressing tumors with a site-specifically 99mTc-labeled recombi-nant affibody molecule, ZHER2:2395, with C-terminally engineered cysteine. J. Nucl. Med. Off. Publ. Soc. Nucl. Med. 50, 781–789.

Altai, M., Honarvar, H., Wållberg, H., Strand, J., Varasteh, Z., Rosestedt, M., Orlova, A., Dunås, F., Sandström, M., Löfblom, J., Tolmachev, V., Ståhl, S., 2014a. Selection of an optimal cysteine-containing peptide-based chelator for labeling of affibody molecules with (188)Re. Eur. J. Med. Chem. 87, 519–528.

Altai, M., Strand, J., Rosik, D., Selvaraju, R.K., Eriksson Karlström, A., Orlova, A., Tolmachev, V., 2013. Influence of nuclides and chela-tors on imaging using affibody molecules: comparative evaluation of recombinant affibody molecules site-specifically labeled with 68Ga and 111In via maleimido derivatives of DOTA and NODAGA. Bi-oconjug. Chem. 24, 1102–1109.

Altai, M., Wållberg, H., Honarvar, H., Strand, J., Orlova, A., Varasteh, Z., Sandström, M., Löfblom, J., Larsson, E., Strand, S.-E., Lubberink,

85

M., Ståhl, S., Tolmachev, V., 2014b. 188Re-ZHER2:V2, a promis-ing affibody-based targeting agent against HER2-expressing tumors: preclinical assessment. J. Nucl. Med. Off. Publ. Soc. Nucl. Med. 55, 1842–1848.

Akerström, B., Nielsen, E., Björck, L., 1987. Definition of IgG-and albumin-binding regions of streptococcal protein G. J. Biol. Chem. 262, 13388–13391.

Bakker, W.H., Krenning, E.P., Reubi, J.-C., Breeman, W.A.P., Setyono-Han, B., De Jong, M., Kooij, P.P.M., Bruns, C., Van Hagen, P.M., Mar-bach, P., others, 1991. In vivo application of [111In-DTPA-D-Phe1]-octreotide for detection of somatostatin receptor-positive tu-mors in rats. Life Sci. 49, 1593–1601.

Bedard, P.L., Hansen, A.R., Ratain, M.J., Siu, L.L., 2013. Tumour heteroge-neity in the clinic. Nature 501, 355–364.

Behr, T.M., Béhé M., Wörmann, B., 2001. Trastuzumab and breast cancer. N Engl J Med 345, 995–996.

Behr, T.M., Béhé, M.P., 2002. Cholecystokinin-B/gastrin receptor-targeting peptides for staging and therapy of medullary thyroid cancer and other cholecystokinin-B receptor-expressing malignancies, in: Sem-inars in Nuclear Medicine. Elsevier, pp. 97–109.

Behr, T.M., Gotthardt, M., Becker, W., Behe, M., others, 2002. Radioio-dination of monoclonal antibodies, proteins and peptides for diagno-sis and therapy. Nukl. Arch. 41, 71–79.

Björck, L., Kastern, W., Lindahl, G., Widebäck, K., 1987. Streptococcal protein G, expressed by streptococci or by Escherichia coli, has sep-arate binding sites for human albumin and IgG. Mol. Immunol. 24, 1113–1122.

Block, H., Maertens, B., Spriestersbach, A., Brinker, N., Kubicek, J., Fabis, R., Labahn, J., Schäfer, F., 2009. Immobilized-metal affinity chro-matography (IMAC): a review. Methods Enzymol. 463, 439–473.

Bockisch, A., Freudenberg, L.S., Schmidt, D., Kuwert, T., 2009. Hybrid imaging by SPECT/CT and PET/CT: proven outcomes in cancer im-aging, in: Seminars in Nuclear Medicine. Elsevier, pp. 276–289.

Bodei, L., Kidd, M., Prasad, V., Modlin, I.M., 2015. Peptide receptor radio-nuclide therapy of neuroendocrine tumors, in: Neuroendocrine Tu-mors: A Multidisciplinary Approach. Karger Publishers, pp. 198–215.

Bornhorst, J.A., Falke, J.J., 2000. [16] Purification of proteins using polyhis-tidine affinity tags. Methods Enzymol., Applications of Chimeric Genes and Hybrid Proteins Part A: Gene Expression and Protein Pu-rification 326, 245–254.

Bruns, O.T., Bischof, T.S., Harris, D.K., Franke, D., Shi, Y., Riedemann, L., Bartelt, A., Jaworski, F.B., Carr, J.A., Rowlands, C.J., others, 2017. Next-generation in vivo optical imaging with short-wave infrared quantum dots. Nat. Biomed. Eng. 1, 0056.

86

Cai, W., Chen, K., Mohamedali, K.A., Cao, Q., Gambhir, S.S., Rosenblum, M.G., Chen, X., 2006. PET of vascular endothelial growth factor re-ceptor expression. J. Nucl. Med. 47, 2048–2056.

Cheng, Z., De Jesus, O.P., Kramer, D.J., De, A., Webster, J.M., Gheysens, O., Levi, J., Namavari, M., Wang, S., Park, J.M., others, 2010. 64 Cu-labeled affibody molecules for imaging of HER2 expressing tu-mors. Mol. Imaging Biol. 12, 316–324.

Cheng, Z., De Jesus, O.P., Namavari, M., De, A., Levi, J., Webster, J.M., Zhang, R., Lee, B., Syud, F.A., Gambhir, S.S., 2008. Small-animal PET imaging of human epidermal growth factor receptor type 2 ex-pression with site-specific 18F-labeled protein scaffold molecules. J. Nucl. Med. 49, 804–813.

Chiotellis, A., Sladojevich, F., Mu, L., Herde, A.M., Valverde, I.E., Tolma-chev, V., Schibli, R., Ametamey, S.M., Mindt, T.L., 2016. Novel chemoselective 18 F-radiolabeling of thiol-containing biomolecules under mild aqueous conditions. Chem. Commun. 52, 6083–6086.

Clarke, E.T., Martell, A.E., 1991a. Stabilities of the Fe (III), Ga (III) and In (III) chelates of N, N′, N ″-triazacyclononanetriacetic acid. Inorgani-ca Chim. Acta 181, 273–280.

Clarke, E.T., Martell, A.E., 1991b. Stabilities of trivalent metal ion com-plexes of the tetraacetate derivatives of 12-, 13-and 14-membered tetraazamacrocycles. Inorganica Chim. Acta 190, 37–46.

Contractor, K.B., Aboagye, E.O., 2009. Monitoring predominantly cytostatic treatment response with 18F-FDG PET. J. Nucl. Med. Off. Publ. Soc. Nucl. Med. 50 Suppl 1, 97S–105S.

Cooper, M.S., Sabbah, E., Mather, S.J., 2006. Conjugation of chelating agents to proteins and radiolabeling with trivalent metallic isotopes. Nat. Protoc. 1, 314.

Cree, I.A., Deans, Z., Ligtenberg, M.J.L., Normanno, N., Edsjö, A., Rouleau, E., Solé, F., Thunnissen, E., Timens, W., Schuuring, E., Dequeker, E., Murray, S., Dietel, M., Groenen, P., Krieken, J.H.V., 2014. Guidance for laboratories performing molecular pathology for can-cer patients. J. Clin. Pathol. 67, 923–931. doi:10.1136/jclinpath-2014-202404

Decristoforo, C., Hernandez Gonzalez, I., Carlsen, J., Rupprich, M., Huis-man, M., Virgolini, I., Wester, H.-J., Haubner, R., 2008. 68Ga- and 111In-labelled DOTA-RGD peptides for imaging of alphavbeta3 in-tegrin expression. Eur. J. Nucl. Med. Mol. Imaging 35, 1507–1515.

DeNardo, G.L., Kennel, S.J., Siegel, J.A., DeNardo, S.J., 2004. Radiometals as Payloads for Radioimmunotherapy for Lymphoma Lymphoma. Clin. Lymphoma 5, S5–S10.

D’huyvetter, M., Vincke, C., Xavier, C., Aerts, A., Impens, N., Baatout, S., De Raeve, H., Muyldermans, S., Caveliers, V., Devoogdt, N., others, 2014. Targeted radionuclide therapy with A 177Lu-labeled anti-HER2 nanobody. Theranostics 4, 708.

87

Dijkers, E.C., Oude Munnink, T.H., Kosterink, J.G., Brouwers, A.H., Jager, P.L., de Jong, J.R., van Dongen, G.A., Schröder, C.P., Lub-de Hooge, M.N., de Vries, E.G., 2010. Biodistribution of 89Zr-trastuzumab and PET Imaging of HER2-Positive Lesions in Patients With Metastatic Breast Cancer. Clin. Pharmacol. Ther. 87, 586–592.

Dongen, G.A.M.S. van, Visser, G.W.M., Hooge, M.N.L., Vries, E.G. de, Perk, L.R., 2007. Immuno-PET: A Navigator in Monoclonal Anti-body Development and Applications. The Oncologist 12, 1379–1389.

Eder, M., Löhr, T., Bauder-Wüst, U., Reber, M., Mier, W., Schäfer, M., Haberkorn, U., Eisenhut, M., 2013. Pharmacokinetic properties of peptidic radiopharmaceuticals: reduced uptake of (EH)3-conjugates in important organs. J. Nucl. Med. Off. Publ. Soc. Nucl. Med. 54, 1327–1330.

Eisenhauer, E.A., Therasse, P., Bogaerts, J., Schwartz, L.H., Sargent, D., Ford, R., Dancey, J., Arbuck, S., Gwyther, S., Mooney, M., Rubin-stein, L., Shankar, L., Dodd, L., Kaplan, R., Lacombe, D., Verweij, J., 2009. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur. J. Cancer Oxf. Engl. 1990 45, 228–247.

Eisenwiener, K.-P., Prata, M.I.M., Buschmann, I., Zhang, H.-W., Santos, A.C., Wenger, S., Reubi, J.C., Mäcke, H.R., 2002. NODAGATOC, a new chelator-coupled somatostatin analogue labeled with [67/68Ga] and [111In] for SPECT, PET, and targeted therapeutic applications of somatostatin receptor (hsst2) expressing tumors. Bi-oconjug. Chem. 13, 530–541.

Engfeldt, T., Orlova, A., Tran, T., Bruskin, A., Widström, C., Karlström, A.E., Tolmachev, V., 2007a. Imaging of HER2-expressing tumours using a synthetic Affibody molecule containing the 99mTc-chelating mercaptoacetyl-glycyl-glycyl-glycyl (MAG3) sequence. Eur. J. Nucl. Med. Mol. Imaging 34, 722–733.

Engfeldt, T., Tran, T., Orlova, A., Widström, C., Feldwisch, J., Abrahmsen, L., Wennborg, A., Karlström, A.E., Tolmachev, V., 2007b. 99mTc-chelator engineering to improve tumour targeting properties of a HER2-specific Affibody molecule. Eur. J. Nucl. Med. Mol. Imaging 34, 1843–1853.

Even, A.J.G., Hamming-Vrieze, O., van Elmpt, W., Winnepenninckx, V.J.L., Heukelom, J., Tesselaar, M.E.T., Vogel, W.V., Hoeben, A., Zegers, C.M.L., Vugts, D.J., van Dongen, G.A.M.S., Bartelink, H., Mottaghy, F.M., Hoebers, F., Lambin, P., 2016. Quantitative as-sessment of Zirconium-89 labeled cetuximab using PET/CT imaging in patients with advanced head and neck cancer: a theragnostic ap-proach. Oncotarget 8, 3870–3880.

88

Fahnestock, S.R., Alexander, P., Nagle, J., Filpula, D., 1986. Gene for an immunoglobulin-binding protein from a group G streptococcus. J. Bacteriol. 167, 870–880.

Frejd, F.Y., Kim, K.-T., 2017. Affibody molecules as engineered protein drugs. Exp. Mol. Med. 49, e306.

Freise, A.C., Wu, A.M., 2015. In vivo imaging with antibodies and engi-neered fragments. Mol. Immunol. 67, 142–152.

Garousi, J., Anderson, K.G., Dam, J.H., Olsen, B.B., Orlova, A., Buijs, J., Ståhl, S., Thisgaard, H., Tolmachev, V., 2017. The use of radio-cobalt as a label improves imaging of EGFR using DOTA-conjugated Affibody molecule. Sci. Rep. 7, 5961.

Garrison, J.C., Rold, T.L., Sieckman, G.L., Figueroa, S.D., Volkert, W.A., Jurisson, S.S., Hoffman, T.J., 2007. In vivo evaluation and small-animal PET/CT of a prostate cancer mouse model using 64Cu bombesin analogs: side-by-side comparison of the CB-TE2A and DOTA chelation systems. J. Nucl. Med. 48, 1327–1337.

Ginj, M., Zhang, H., Eisenwiener, K.-P., Wild, D., Schulz, S., Rink, H., Ces-cato, R., Reubi, J.C., Maecke, H.R., 2008. New pansomatostatin lig-ands and their chelated versions: affinity profile, agonist activity, in-ternalization, and tumor targeting. Clin. Cancer Res. 14, 2019–2027.

Giordano, S.H., Temin, S., Kirshner, J.J., Chandarlapaty, S., Crews, J.R., Davidson, N.E., Esteva, F.J., Gonzalez-Angulo, A.M., Krop, I., Lev-inson, J., others, 2014. Systemic therapy for patients with advanced human epidermal growth factor receptor 2–positive breast cancer: American Society of Clinical Oncology clinical practice guideline. J. Clin. Oncol. 32, 2078–2099.

Goldstein, R., Sosabowski, J., Livanos, M., Leyton, J., Vigor, K., Bhavsar, G., Nagy-Davidescu, G., Rashid, M., Miranda, E., Yeung, J., Tolner, B., Plückthun, A., Mather, S., Meyer, T., Chester, K., 2015. Devel-opment of the designed ankyrin repeat protein (DARPin) G3 for HER2 molecular imaging. Eur. J. Nucl. Med. Mol. Imaging 42, 288–301.

Guss, B., Eliasson, M., Olsson, A., Uhlen, M., Frej, A.K., Jörnvall, H., Flock, J.I., Lindberg, M., 1986. Structure of the IgG-binding regions of streptococcal protein G. EMBO J. 5, 1567.

Hackel, B.J., Kimura, R.H., Gambhir, S.S., 2012. Use of 64Cu-labeled fi-bronectin domain with EGFR-overexpressing tumor xenograft: mo-lecular imaging. Radiology 263, 179–188.

Heppeler, A., Froidevaux, S., Mäcke, H.R., Jermann, E., Béhé, M., Powell, P., Hennig, M., 1999. Radiometal-Labelled Macrocyclic Chelator-Derivatised Somatostatin Analogue with Superb Tumour-Targeting Properties and Potential for Receptor-Mediated Internal Radiothera-py. Chem.- Eur. J. 5, 1974–1981.

Heskamp, S., Laverman, P., Rosik, D., Boschetti, F., van der Graaf, W.T., Oyen, W.J., van Laarhoven, H.W., Tolmachev, V., Boerman, O.C.,

89

2012. Imaging of human epidermal growth factor receptor type 2 expression with 18F-labeled affibody molecule ZHER2: 2395 in a mouse model for ovarian cancer. J. Nucl. Med. 53, 146–153.

Hofström, C., Altai, M., Honarvar, H., Strand, J., Malmberg, J., Hosseini-mehr, S.J., Orlova, A., Gräslund, T., Tolmachev, V., 2013. HAHA-HA, HEHEHE, HIHIHI, or HKHKHK: influence of position and composition of histidine containing tags on biodistribution of [(99m)Tc(CO)3](+)-labeled affibody molecules. J. Med. Chem. 56, 4966–4974.

Hofström, C., Orlova, A., Altai, M., Wangsell, F., Graslund, T., Tolmachev, V., 2011. Use of a HEHEHE purification tag instead of a hexahisti-dine tag improves biodistribution of affibody molecules site-specifically labeled with (99m)Tc, (111)In, and (125)I. J. Med. Chem. 54, 3817–3826.

Honarvar, H., Garousi, J., Gunneriusson, E., Höidén-Guthenberg, I., Altai, M., Widström, C., Tolmachev, V., Frejd, F.Y., 2015. Imaging of CAIX-expressing xenografts in vivo using 99mTc-HEHEHE-ZCAIX:1 affibody molecule. Int. J. Oncol. 46, 513–520.

Honarvar, H., Strand, J., Perols, A., Orlova, A., Selvaraju, R.K., Karlström, A.E., Tolmachev, V., 2014. Position for site-specific attachment of a DOTA chelator to synthetic Affibody molecules has a different in-fluence on the targeting properties of 68Ga-compared to 111In-labeled conjugates. Mol. Imaging 13, 7290–2014.

Hynes, N.E., Lane, H.A., 2005. ERBB receptors and cancer: the complexity of targeted inhibitors. Nat. Rev. Cancer 5, 341–354.

James, M.L., Gambhir, S.S., 2012. A Molecular Imaging Primer: Modalities, Imaging Agents, and Applications. Physiol. Rev. 92, 897–965.

Jensen, K., Krusenstjerna-Hafstrøm, R., Lohse, J., Petersen, K.H., Derand, H., 2017. A novel quantitative immunohistochemistry method for precise protein measurements directly in formalin-fixed, paraffin-embedded specimens: analytical performance measuring HER2. Mod. Pathol. 30, 180–193.

Johansson, M.U., Frick, I.-M., Nilsson, H., Kraulis, P.J., Hober, S., Jonas-son, P., Linhult, M., Nygren, P.-A., Uhlén, M., Björck, L., Dra-kenberg, T., Forsén, S., Wikström, M., 2002. Structure, specificity, and mode of interaction for bacterial albumin-binding modules. J. Biol. Chem. 277, 8114–8120.

Johnbeck, C.B., Knigge, U., Kjær, A., 2014. PET tracers for somatostatin receptor imaging of neuroendocrine tumors: current status and re-view of the literature. Future Oncol. Lond. Engl. 10, 2259–2277.

Judenhofer, M.S., Wehrl, H.F., Newport, D.F., Catana, C., Siegel, S.B., Becker, M., Thielscher, A., Kneilling, M., Lichy, M.P., Eichner, M., others, 2008. Simultaneous PET-MRI: a new approach for function-al and morphological imaging. Nat. Med. 14, 459–465.

90

Keyaerts, M., Xavier, C., Heemskerk, J., Devoogdt, N., Everaert, H., Acka-ert, C., Vanhoeij, M., Duhoux, F.P., Gevaert, T., Simon, P., Schalli-er, D., Fontaine, C., Vaneycken, I., Vanhove, C., Greve, J.D., Lamote, J., Caveliers, V., Lahoutte, T., 2016. Phase I Study of 68Ga-HER2-Nanobody for PET/CT Assessment of HER2 Expres-sion in Breast Carcinoma. J. Nucl. Med. 57, 27–33.

Knapp, F.F., Mirzadeh, S., 1994. The continuing important role of radionu-clide generator systems for nuclear medicine. Eur. J. Nucl. Med. 21, 1151–1165.

Körner, M., Reubi, J.C., 2007. NPY receptors in human cancer: a review of current knowledge. Peptides 28, 419–425.

Kramer-Marek, G., Kiesewetter, D.O., Martiniova, L., Jagoda, E., Lee, S.B., Capala, J., 2008. [18 F] FBEM-Z HER2: 342–Affibody molecule—a new molecular tracer for in vivo monitoring of HER2 expression by positron emission tomography. Eur. J. Nucl. Med. Mol. Imaging 35, 1008–1018.

Kramer-Marek, G., Shenoy, N., Seidel, J., Griffiths, G., Choyke, P., Capala, J., 2011. 68Ga-radiolabeled DOTA-affibody molecule for in vivo assessment of HER2 expression with PET. AACR.

Kraulis, P.J., Jonasson, P., Nygren, P.Å., Uhlén, M., Jendeberg, L., Nilsson, B., Kördel, J., 1996. The serum albumin-binding domain of strepto-coccal protein G is a three-helical bundle: a heteronuclear NMR study. FEBS Lett. 378, 190–194.

Lappano, R., Maggiolini, M., 2017. Pharmacotherapeutic Targeting of G Protein-Coupled Receptors in Oncology: Examples of Approved Therapies and Emerging Concepts. Drugs 77, 951–965.

Lemmon, M.A., Schlessinger, J., 2010. Cell signaling by receptor tyrosine kinases. Cell 141, 1117–1134.

Lindberg, H., Hofström, C., Altai, M., Honorvar, H., Wållberg, H., Orlova, A., Ståhl, S., Gräslund, T., Tolmachev, V., 2012. Evaluation of a HER2-targeting affibody molecule combining an N-terminal HEHEHE-tag with a GGGC chelator for 99mTc-labelling at the C terminus. Tumor Biol. 33, 641–651.

Linhult, M., Binz, H.K., Uhlén, M., Hober, S., 2002. Mutational analysis of the interaction between albumin-binding domain from streptococcal protein G and human serum albumin. Protein Sci. 11, 206–213.

Liu, S., Edwards, D.S., 2001. Bifunctional chelators for therapeutic lantha-nide radiopharmaceuticals. Bioconjug. Chem. 12, 7–34.

Löfblom, J., Feldwisch, J., Tolmachev, V., Carlsson, J., Ståhl, S., Frejd, F.Y., 2010. Affibody molecules: engineered proteins for therapeutic, diagnostic and biotechnological applications. FEBS Lett. 584, 2670–2680.

Maecke, H.R., André, J.P., 2007. 68Ga-PET radiopharmacy: A generator-based alternative to 18F-radiopharmacy. Ernst Scher. Res. Found. Workshop 215–242.

91

Malmberg, J., Perols, A., Varasteh, Z., Altai, M., Braun, A., Sandström, M., Garske, U., Tolmachev, V., Orlova, A., Karlström, A.E., 2012. Comparative evaluation of synthetic anti-HER2 Affibody molecules site-specifically labelled with 111In using N-terminal DOTA, NO-TA and NODAGA chelators in mice bearing prostate cancer xeno-grafts. Eur. J. Nucl. Med. Mol. Imaging 39, 481–492.

Mankoff, D.A., Edmonds, C.E., Farwell, M.D., Pryma, D.A., 2016. Devel-opment of Companion Diagnostics. Semin. Nucl. Med. 46, 47–56. doi:10.1053/j.semnuclmed.2015.09.002

Miao, Z., Ren, G., Liu, H., Jiang, L., Cheng, Z., 2010. Small-animal PET imaging of human epidermal growth factor receptor positive tumor with a 64Cu labeled affibody protein. Bioconjug. Chem. 21, 947–954.

Moasser, M.M., 2007. The oncogene HER2; Its signaling and transforming functions and its role in human cancer pathogenesis. Oncogene 26, 6469–6487.

Mume, E., Orlova, A., Larsson, B., Nilsson, A.-S., Nilsson, F.Y., Sjöberg, S., Tolmachev, V., 2005. Evaluation of ((4-hydroxyphenyl)ethyl)maleimide for site-specific radiobromination of anti-HER2 affibody. Bioconjug. Chem. 16, 1547–1555.

Myhre, E.B., Kronvall, G., 1980. Demonstration of a new type of immuno-globulin G receptor in Streptococcus zooepidemicus strains. Infect. Immun. 27, 808–816.

Natali, P.G., Nicotra, M.R., Bigotti, A., Venturo, I., Slamon, D.J., Fendly, B.M., Ullrich, A., 1990. Expression of the p185 encoded by HER2 oncogene in normal and transformed human tissues. Int. J. Cancer 45, 457–461.

Nilvebrant, J., \AAstrand, M., Georgieva-Kotseva, M., Björnmalm, M., Löfblom, J., Hober, S., 2014. Engineering of bispecific affinity pro-teins with high affinity for ERBB2 and adaptable binding to albu-min. PloS One 9, e103094.

Nilvebrant, J., Åstrand, M., Löfblom, J., Hober, S., 2013. Development and characterization of small bispecific albumin-binding domains with high affinity for ErbB3. Cell. Mol. Life Sci. 70, 3973–3985.

Nilvebrant, J., Alm, T., Hober, S., Löfblom, J., 2011. Engineering bispecific-ity into a single albumin-binding domain. PLoS One 6, e25791.

Nilvebrant, J., Hober, S., 2013. The albumin-binding domain as a scaffold for protein engineering. Comput. Struct. Biotechnol. J. 6, 1–8.

Nygren, P.Å., Skerra, A., 2004. Binding proteins from alternative scaffolds. J. Immunol. Methods 290, 3–28.

Olafsen, T., Wu, A.M., 2010. Antibody Vectors for Imaging. Semin. Nucl. Med., Antibodies and Peptides in Nuclear Medicine Imaging and Therapy 40, 167–181. doi:10.1053/j.semnuclmed.2009.12.005

Organization, W.H., others, 1979. WHO handbook for reporting results of cancer treatment.

92

Orlova, A., Bruskin, A., Sjöström, A., Lundqvist, H., Gedda, L., Tolmachev, V., 2000. Cellular processing of 125 I-and 111 In-labeled epidermal growth factor (EGF) bound to cultured A431 tumor cells. Nucl. Med. Biol. 27, 827–835.

Orlova, A., Hofström, C., Strand, J., Varasteh, Z., Sandstrom, M., Anders-son, K., Tolmachev, V., Gräslund, T., 2013. [99mTc (CO) 3]+-(HE) 3-ZIGF1R: 4551, a new Affibody conjugate for visualization of in-sulin-like growth factor-1 receptor expression in malignant tumours. Eur. J. Nucl. Med. Mol. Imaging 40, 439–449.

Orlova, A., Magnusson, M., Eriksson, T.L., Nilsson, M., Larsson, B., Höidén-Guthenberg, I., Widström, C., Carlsson, J., Tolmachev, V., Ståhl, S., others, 2006a. Tumor imaging using a picomolar affinity HER2 binding affibody molecule. Cancer Res. 66, 4339–4348.

Orlova, A., Malm, M., Rosestedt, M., Varasteh, Z., Andersson, K., Selva-raju, R.K., Altai, M., Honarvar, H., Strand, J., Ståhl, S., others, 2014. Imaging of HER3-expressing xenografts in mice using a 99mTc (CO) 3-HEHEHE-ZHER3: 08699 affibody molecule. Eur. J. Nucl. Med. Mol. Imaging 41, 1450–1459.

Orlova, A., Nilsson, F.Y., Wikman, M., Widström, C., Ståhl, S., Carlsson, J., Tolmachev, V., 2006b. Comparative in vivo evaluation of techneti-um and iodine labels on an anti-HER2 affibody for single-photon imaging of HER2 expression in tumors. J. Nucl. Med. Off. Publ. Soc. Nucl. Med. 47, 512–519.

Perols, A., Honarvar, H., Strand, J., Selvaraju, R., Orlova, A., Eriksson Karl-ström, A., Tolmachev, V., 2012. Influence of DOTA chelator posi-tion on biodistribution and targeting properties of 111In-labeled syn-thetic anti-HER2 affibody molecules. Bioconjug. Chem. 23, 1661–1670.

Rao, J., Dragulescu-Andrasi, A., Yao, H., 2007. Fluorescence imaging in vivo: recent advances. Curr. Opin. Biotechnol., Analytical biotech-nology 18, 17–25.

Reubi, J.C., 2003. Peptide receptors as molecular targets for cancer diagnosis and therapy. Endocr. Rev. 24, 389–427.

Reubi, J.C., Mäcke, H.R., Krenning, E.P., 2005. Candidates for peptide re-ceptor radiotherapy today and in the future. J. Nucl. Med. 46, 67S–75S.

Reubi, J.C., Maecke, H.R., 2008. Peptide-based probes for cancer imaging. J. Nucl. Med. 49, 1735–1738.

Rosik, D., Thibblin, A., Antoni, G., Honarvar, H., Strand, J., Selvaraju, R.K., Altai, M., Orlova, A., Eriksson Karlström, A., Tolmachev, V., 2013. Incorporation of a triglutamyl spacer improves the biodistribution of synthetic affibody molecules radiofluorinated at the N-terminus via oxime formation with 18F-4-fluorobenzaldehyde. Bioconjug. Chem. 25, 82–92.

93

Schmidt, M.M., Wittrup, K.D., 2009. A modeling analysis of the effects of molecular size and binding affinity on tumor targeting. Mol. Cancer Ther. 8, 2861–2871.

Schwarz, A., others, n.d. A Novel Approach to Tc 99m Labeled Monoclonal Antibodies, Poster Sessions. Proc. 34th Ann Mtg J Nucl Med 28, 721.

Shan, L., 2004. (64)Cu-Labeled 1,4,7,10-Tetraazacyclododedane-N,N’,N’’,N’’’-tetraacetic acid–conjugated vascular endothelial growth factor A isoform 121-gelonin fusion protein, in: Molecular Imaging and Contrast Agent Database (MICAD). National Center for Biotechnology Information (US), Bethesda (MD).

Sharkey, R.M., Goldenberg, D.M., 2006. Targeted therapy of cancer: new prospects for antibodies and immunoconjugates. CA. Cancer J. Clin. 56, 226–243.

Sörensen, J., Sandberg, D., Sandström, M., Wennborg, A., Feldwisch, J., Tolmachev, V., Åström, G., Lubberink, M., Garske-Román, U., Carlsson, J., others, 2014. First-in-human molecular imaging of HER2 expression in breast cancer metastases using the 111In-ABY-025 affibody molecule. J. Nucl. Med. 55, 730–735.

Sörensen, J., Velikyan, I., Sandberg, D., Wennborg, A., Feldwisch, J., Tolmachev, V., Orlova, A., Sandström, M., Lubberink, M., Olofs-son, H., others, 2016. Measuring HER2-receptor expression in meta-static breast cancer using [68Ga] ABY-025 Affibody PET/CT. Theranostics 6, 262.

Sosabowski, J.K., Mather, S.J., 2006. Conjugation of DOTA-like chelating agents to peptides and radiolabeling with trivalent metallic isotopes. Nat. Protoc. 1, 972–976.

Ståhl, S., Gräslund, T., Eriksson Karlström, A., Frejd, F.Y., Nygren, P.-Å., Löfblom, J., 2017. Affibody Molecules in Biotechnological and Medical Applications. Trends Biotechnol. 35, 691–712.

Steffen, A.-C., Wikman, M., Tolmachev, V., Adams, G.P., Nilsson, F.Y., Ståhl, S., Carlsson, J., 2005. In vitro characterization of a bivalent anti-HER-2 affibody with potential for radionuclide-based diagnos-tics. Cancer Biother. Radiopharm. 20, 239–248.

Stigbrand, T., Carlsson, J., Adams, G.P., 2008. Targeted radionuclide tumor therapy: biological aspects. Springer.

Strand, J., Varasteh, Z., Eriksson, O., Abrahmsen, L., Orlova, A., Tolmachev, V., 2014. Gallium-68-labeled affibody molecule for PET imaging of PDGFRβ expression in vivo. Mol. Pharm. 11, 3957–3964.

Tannock, I., Hill, R., Bristow, R., Harrington, L., 2013. Basic Science of Oncology, Fifth Edition, 5 edition. ed. McGraw-Hill Education / Medical, New York.

Telloni, S.M., 2017. Tumor Staging and Grading: A Primer. Methods Mol. Biol. Clifton NJ 1606, 1–17. doi:10.1007/978-1-4939-6990-6_1

94

Terwisscha van Scheltinga, A.G.T., Lub-de Hooge, M.N., Hinner, M.J., Verheijen, R.B., Allersdorfer, A., Hülsmeyer, M., Nagengast, W.B., Schröder, C.P., Kosterink, J.G.W., de Vries, E.G.E., Audoly, L., Olwill, S.A., 2014. In vivo visualization of MET tumor expression and anticalin biodistribution with the MET-specific anticalin 89Zr-PRS-110 PET tracer. J. Nucl. Med. Off. Publ. Soc. Nucl. Med. 55, 665–671. doi:10.2967/jnumed.113.124941

Thurber, G.M., Zajic, S.C., Wittrup, K.D., 2007. Theoretic criteria for anti-body penetration into solid tumors and micrometastases. J. Nucl. Med. Off. Publ. Soc. Nucl. Med. 48, 995–999. doi:10.2967/jnumed.106.037069

Tolmachev, V., 2008. Choice of radionuclides and radiolabelling techniques, in: Targeted Radionuclide Tumor Therapy. Springer, pp. 145–174.

Tolmachev, V., Altai, M., Sandström, M., Perols, A., Karlström, A.E., Boschetti, F., Orlova, A., 2011. Evaluation of a maleimido deriva-tive of NOTA for site-specific labeling of affibody molecules. Bio-conjug. Chem. 22, 894–902.

Tolmachev, V., Friedman, M., Sandström, M., Eriksson, T.L., Rosik, D., Hodik, M., Ståhl, S., Frejd, F.Y., Orlova, A., 2009a. Affibody mole-cules for epidermal growth factor receptor targeting in vivo: aspects of dimerization and labeling chemistry. J. Nucl. Med. 50, 274–283.

Tolmachev, V., Hofström, C., Malmberg, J., Ahlgren, S., Hosseinimehr, S.J., Sandström, M., Abrahmsén, L., Orlova, A., Gräslund, T., 2010a. HEHEHE-tagged affibody molecule may be purified by IMAC, is conveniently labeled with [99(m)Tc(CO)₃](+), and shows improved biodistribution with reduced hepatic radioactivity accumulation. Bi-oconjug. Chem. 21, 2013–2022.

Tolmachev, V., Mume, E., Sjöberg, S., Frejd, F.Y., Orlova, A., 2009b. Influ-ence of valency and labelling chemistry on in vivo targeting using radioiodinated HER2-binding Affibody molecules. Eur. J. Nucl. Med. Mol. Imaging 36, 692–701.

Tolmachev, V., Orlova, A., 2010. Influence of labelling methods on biodis-tribution and imaging properties of radiolabelled peptides for visual-isation of molecular therapeutic targets. Curr. Med. Chem. 17, 2636–2655.

Tolmachev, V., Orlova, A., Lundqvist, H., 2003. Approaches to improve cellular retention of radiohalogen labels delivered by internalising tumour-targeting proteins and peptides. Curr. Med. Chem. 10, 2447–2460.

Tolmachev, V., Orlova, A., Pehrson, R., Galli, J., Baastrup, B., Andersson, K., Sandström, M., Rosik, D., Carlsson, J., Lundqvist, H., others, 2007. Radionuclide therapy of HER2-positive microxenografts using a 177Lu-labeled HER2-specific Affibody molecule. Cancer Res. 67, 2773–2782.

95

Tolmachev, V., Rosik, D., W\a allberg, H., Sjöberg, A., Sandström, M., Hansson, M., Wennborg, A., Orlova, A., 2010b. Imaging of EGFR expression in murine xenografts using site-specifically labelled anti-EGFR 111In-DOTA-ZEGFR: 2377 Affibody molecule: aspect of the injected tracer amount. Eur. J. Nucl. Med. Mol. Imaging 37, 613–622.

Tolmachev, V., Stone-Elander, S., 2010. Radiolabelled proteins for positron emission tomography: Pros and cons of labelling methods. Biochim. Biophys. Acta 1800, 487–510.

Tolmachev, V., Tran, T.A., Rosik, D., Sjöberg, A., Abrahmsén, L., Orlova, A., 2012. Tumor targeting using affibody molecules: interplay of af-finity, target expression level, and binding site composition. J. Nucl. Med. 53, 953–960.

Tolmachev, V., Varasteh, Z., Honarvar, H., Hosseinimehr, S.J., Eriksson, O., Jonasson, P., Frejd, F.Y., Abrahmsen, L., Orlova, A., 2014. Imaging of platelet-derived growth factor receptor β expression in glioblas-toma xenografts using affibody molecule 111In-DOTA-Z09591. J. Nucl. Med. 55, 294–300.

Tolmachev, V., Velikyan, I., Sandström, M., Orlova, A., 2010c. A HER2-binding Affibody molecule labelled with 68Ga for PET imaging: di-rect in vivo comparison with the 111In-labelled analogue. Eur. J. Nucl. Med. Mol. Imaging 37, 1356–1367.

Tolmachev, V., Xu, H., Wållberg, H., Ahlgren, S., Hjertman, M., Sjöberg, A., Sandström, M., Abrahmsén, L., Brechbiel, M.W., Orlova, A., 2008. Evaluation of a maleimido derivative of CHX-A′′ DTPA for site-specific labeling of affibody molecules. Bioconjug. Chem. 19, 1579–1587.

Tolmachev, V., Yim, C.-B., Rajander, J., Perols, A., Karlström, A.E., Haapa-ranta-Solin, M., Grönroos, T.J., Solin, O., Orlova, A., 2017. Com-parative Evaluation of Anti-HER2 Affibody Molecules Labeled with 64Cu Using NOTA and NODAGA. Contrast Media Mol. Imaging 2017.

Tran, T., Engfeldt, T., Orlova, A., Sandström, M., Feldwisch, J., Abrahmsén, L., Wennborg, A., Tolmachev, V., Karlström, A.E., 2007a. 99mTc-maEEE-ZHER2: 342, an Affibody molecule-based tracer for the de-tection of HER2 expression in malignant tumors. Bioconjug. Chem. 18, 1956–1964.

Tran, T., Orlova, A., Sivaev, I., Sandström, M., Tolmachev, V., 2007b. Comparison of benzoate-and dodecaborate-based linkers for attach-ment of radioiodine to HER2-targeting Affibody ligand. Int. J. Mol. Med. 19, 485–493.

Vaidyanathan, G., Zalutsky, M.R., 2006. Preparation of N-succinimidyl 3-[* I] iodobenzoate: an agent for the indirect radioiodination of proteins. Nat. Protoc. 1, 707–713.

96

van der Meel, R., Gallagher, W.M., Oliveira, S., O’Connor, A.E., Schiffelers, R.M., Byrne, A.T., 2010. Recent advances in molecular imaging biomarkers in cancer: application of bench to bedside tech-nologies. Drug Discov. Today 15, 102–114.

Vaneycken, I., D’huyvetter, M., Hernot, S., De Vos, J., Xavier, C., Devoogdt, N., Caveliers, V., Lahoutte, T., 2011. Immuno-imaging using nanobodies. Curr. Opin. Biotechnol. 22, 877–881.

Varasteh, Z., Mitran, B., Rosenström, U., Velikyan, I., Rosestedt, M., Lindeberg, G., Sörensen, J., Larhed, M., Tolmachev, V., Orlova, A., 2015. The effect of macrocyclic chelators on the targeting properties of the 68 Ga-labeled gastrin releasing peptide receptor antagonist PEG 2-RM26. Nucl. Med. Biol. 42, 446–454.

Vazquez-Lombardi, R., Phan, T.G., Zimmermann, C., Lowe, D., Jermutus, L., Christ, D., 2015. Challenges and opportunities for non-antibody scaffold drugs. Drug Discov. Today 20, 1271–1283.

Vegt, E., de Jong, M., Wetzels, J.F.M., Masereeuw, R., Melis, M., Oyen, W.J.G., Gotthardt, M., Boerman, O.C., 2010. Renal toxicity of ra-diolabeled peptides and antibody fragments: mechanisms, impact on radionuclide therapy, and strategies for prevention. J. Nucl. Med. Off. Publ. Soc. Nucl. Med. 51, 1049–1058.

Wållberg, H., Ahlgren, S., Widström, C., Orlova, A., 2010. Evaluation of the radiocobalt-labeled [MMA-DOTA-Cys61]-ZHER2: 2395-Cys af-fibody molecule for targeting of HER2-expressing tumors. Mol. Im-aging Biol. 12, 54–62.

Waibel, R., Alberto, R., Willuda, J., Finnern, R., Schibli, R., Stichelberger, A., Egli, A., Abram, U., Mach, J.P., Plückthun, A., Schubiger, P.A., 1999. Stable one-step technetium-99m labeling of His-tagged re-combinant proteins with a novel Tc(I)-carbonyl complex. Nat. Bio-technol. 17, 897–901.

Wållberg, H., Orlova, A., Altai, M., Hosseinimehr, S.J., Widström, C., Malmberg, J., Ståhl, S., Tolmachev, V., 2011. Molecular design and optimization of 99mTc-labeled recombinant affibody molecules im-proves their biodistribution and imaging properties. J. Nucl. Med. Off. Publ. Soc. Nucl. Med. 52, 461–469.

Wang, H., Chen, K., Niu, G., Chen, X., 2008. Site-specifically biotinylated VEGF121 for near-infrared fluorescence imaging of tumor angio-genesis. Mol. Pharm. 6, 285–294.

Wester, H.-J., Kessler, H., 2005. Molecular targeting with peptides or pep-tide-polymer conjugates: just a question of size? J. Nucl. Med. 46, 1940–1945.

Westerlund, K., Honarvar, H., Norrström, E., Strand, J., Mitran, B., Orlova, A., Eriksson Karlström, A., Tolmachev, V., 2016. Increasing the Net Negative Charge by Replacement of DOTA Chelator with DOT-AGA Improves the Biodistribution of Radiolabeled Second-

97

Generation Synthetic Affibody Molecules. Mol. Pharm. 13, 1668–1678.

Wilbur, D.S., 1992. Radiohalogenation of proteins: an overview of radionu-clides, labeling methods and reagents for conjugate labeling. Bio-conjug. Chem. 3, 433–470.

Wild, D., Mäcke, H., Christ, E., Gloor, B., Reubi, J.C., 2008. Glucagon-like peptide 1–receptor scans to localize occult insulinomas. N. Engl. J. Med. 359, 766–768.

Wittrup, K.D., Thurber, G.M., Schmidt, M.M., Rhoden, J.J., 2012. Practical theoretic guidance for the design of tumor-targeting agents. Methods Enzymol. 503, 255.

Wolff, A.C., Hammond, M.E.H., Hicks, D.G., Dowsett, M., McShane, L.M., Allison, K.H., Allred, D.C., Bartlett, J.M., Bilous, M., Fitzgibbons, P., others, 2013. Recommendations for human epidermal growth factor receptor 2 testing in breast cancer: American Society of Clini-cal Oncology/College of American Pathologists clinical practice guideline update. J. Clin. Oncol. 31, 3997–4013.

Yarden, Y., 2001. Biology of HER2 and its importance in breast cancer. Oncology 61, 1–13.

Zhang, H., Chen, J., Waldherr, C., Hinni, K., Waser, B., Reubi, J.C., Maecke, H.R., 2004. Synthesis and evaluation of bombesin deriva-tives on the basis of pan-bombesin peptides labeled with indium-111, lutetium-177, and yttrium-90 for targeting bombesin receptor-expressing tumors. Cancer Res. 64, 6707–6715.

Zhang, R.R., Schroeder, A.B., Grudzinski, J.J., Rosenthal, E.L., Warram, J.M., Pinchuk, A.N., Eliceiri, K.W., Kuo, J.S., Weichert, J.P., 2017. Beyond the margins: real-time detection of cancer using targeted fluorophores. Nat. Rev. Clin. Oncol. 14, 347–364.

Zhang, X., Cai, W., Cao, F., Schreibmann, E., Wu, Y., Wu, J.C., Xing, L., Chen, X., 2006. 18F-labeled bombesin analogs for targeting GRP receptor-expressing prostate cancer. J. Nucl. Med. 47, 492–501.

Zinn, K.R., Chaudhuri, T.R., Szafran, A.A., O’Quinn, D., Weaver, C., Dug-ger, K., Lamar, D., Kesterson, R.A., Wang, X., Frank, S.J., 2008. Noninvasive bioluminescence imaging in small animals. ILAR J. 49, 103–115.

Acta Universitatis UpsaliensisDigital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Medicine 1354

Editor: The Dean of the Faculty of Medicine

A doctoral dissertation from the Faculty of Medicine, UppsalaUniversity, is usually a summary of a number of papers. A fewcopies of the complete dissertation are kept at major Swedishresearch libraries, while the summary alone is distributedinternationally through the series Digital ComprehensiveSummaries of Uppsala Dissertations from the Faculty ofMedicine. (Prior to January, 2005, the series was publishedunder the title “Comprehensive Summaries of UppsalaDissertations from the Faculty of Medicine”.)

Distribution: publications.uu.seurn:nbn:se:uu:diva-327419

ACTAUNIVERSITATIS

UPSALIENSISUPPSALA

2017